diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,100718 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 14384,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.514305409155901,
+      "learning_rate": 2.3148148148148148e-08,
+      "loss": 0.4529,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.186428649571248,
+      "learning_rate": 4.6296296296296295e-08,
+      "loss": 0.4755,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.599727134663028,
+      "learning_rate": 6.944444444444444e-08,
+      "loss": 0.4499,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1632219608161094,
+      "learning_rate": 9.259259259259259e-08,
+      "loss": 0.4345,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.075053948187206,
+      "learning_rate": 1.1574074074074074e-07,
+      "loss": 0.4422,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.327217811309906,
+      "learning_rate": 1.3888888888888888e-07,
+      "loss": 0.4549,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.812472576558059,
+      "learning_rate": 1.6203703703703703e-07,
+      "loss": 0.4849,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.3812694250719675,
+      "learning_rate": 1.8518518518518518e-07,
+      "loss": 0.4814,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.414189972844575,
+      "learning_rate": 2.0833333333333333e-07,
+      "loss": 0.4719,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.018682633623476,
+      "learning_rate": 2.3148148148148148e-07,
+      "loss": 0.4976,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.556132594006582,
+      "learning_rate": 2.5462962962962963e-07,
+      "loss": 0.4778,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.85742079987303,
+      "learning_rate": 2.7777777777777776e-07,
+      "loss": 0.4513,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.522536638902696,
+      "learning_rate": 3.0092592592592594e-07,
+      "loss": 0.4484,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.38931730291115324,
+      "learning_rate": 3.2407407407407406e-07,
+      "loss": 0.1537,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.435737422113521,
+      "learning_rate": 3.472222222222223e-07,
+      "loss": 0.3885,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.957469265176187,
+      "learning_rate": 3.7037037037037036e-07,
+      "loss": 0.4956,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5683658632477204,
+      "learning_rate": 3.9351851851851854e-07,
+      "loss": 0.4674,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.1531191860528,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 0.4536,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.082546892276986,
+      "learning_rate": 4.3981481481481484e-07,
+      "loss": 0.4583,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.9302538146975605,
+      "learning_rate": 4.6296296296296297e-07,
+      "loss": 0.4886,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.242222335712642,
+      "learning_rate": 4.861111111111112e-07,
+      "loss": 0.4504,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.549333182510484,
+      "learning_rate": 5.092592592592593e-07,
+      "loss": 0.4494,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.730230869777737,
+      "learning_rate": 5.324074074074074e-07,
+      "loss": 0.4234,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.737008726548776,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 0.4698,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3802882906233807,
+      "learning_rate": 5.787037037037038e-07,
+      "loss": 0.4327,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3014795813797186,
+      "learning_rate": 6.018518518518519e-07,
+      "loss": 0.4667,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.091569429579109,
+      "learning_rate": 6.25e-07,
+      "loss": 0.4297,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.0743862732039755,
+      "learning_rate": 6.481481481481481e-07,
+      "loss": 0.422,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.491423120172929,
+      "learning_rate": 6.712962962962964e-07,
+      "loss": 0.4393,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.815451275632443,
+      "learning_rate": 6.944444444444446e-07,
+      "loss": 0.4638,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8156187528083185,
+      "learning_rate": 7.175925925925927e-07,
+      "loss": 0.4429,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.634878993456416,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.4449,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2500519209340575,
+      "learning_rate": 7.63888888888889e-07,
+      "loss": 0.4629,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9729631088282535,
+      "learning_rate": 7.870370370370371e-07,
+      "loss": 0.4096,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.3809582006425653,
+      "learning_rate": 8.101851851851853e-07,
+      "loss": 0.1574,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6845273110335195,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.428,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.908807881173384,
+      "learning_rate": 8.564814814814816e-07,
+      "loss": 0.4355,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.099124260332298,
+      "learning_rate": 8.796296296296297e-07,
+      "loss": 0.4744,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.804907964172545,
+      "learning_rate": 9.027777777777779e-07,
+      "loss": 0.4367,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.764129241549869,
+      "learning_rate": 9.259259259259259e-07,
+      "loss": 0.4391,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6778347974911214,
+      "learning_rate": 9.490740740740742e-07,
+      "loss": 0.4006,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2473065093666382,
+      "learning_rate": 9.722222222222224e-07,
+      "loss": 0.4265,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 18.600724944683304,
+      "learning_rate": 9.953703703703704e-07,
+      "loss": 0.4695,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3743515033110096,
+      "learning_rate": 1.0185185185185185e-06,
+      "loss": 0.4661,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8155510504817043,
+      "learning_rate": 1.0416666666666667e-06,
+      "loss": 0.437,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3811819350832386,
+      "learning_rate": 1.0648148148148149e-06,
+      "loss": 0.489,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.946467214484382,
+      "learning_rate": 1.087962962962963e-06,
+      "loss": 0.4415,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.445540978007164,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.4621,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.362203737808535,
+      "learning_rate": 1.1342592592592594e-06,
+      "loss": 0.4334,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.690667122154501,
+      "learning_rate": 1.1574074074074076e-06,
+      "loss": 0.4722,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8332809173439215,
+      "learning_rate": 1.1805555555555556e-06,
+      "loss": 0.4267,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.948145344460917,
+      "learning_rate": 1.2037037037037037e-06,
+      "loss": 0.4395,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.575342995276088,
+      "learning_rate": 1.226851851851852e-06,
+      "loss": 0.4751,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0567409303787114,
+      "learning_rate": 1.25e-06,
+      "loss": 0.4605,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.4524857490408865,
+      "learning_rate": 1.2731481481481483e-06,
+      "loss": 0.4654,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.115494320983145,
+      "learning_rate": 1.2962962962962962e-06,
+      "loss": 0.4253,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.2180498468354966,
+      "learning_rate": 1.3194444444444446e-06,
+      "loss": 0.439,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.25376794551028,
+      "learning_rate": 1.3425925925925928e-06,
+      "loss": 0.4765,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.339844570776696,
+      "learning_rate": 1.3657407407407408e-06,
+      "loss": 0.4675,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.455329750246264,
+      "learning_rate": 1.3888888888888892e-06,
+      "loss": 0.4377,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.087880756447201,
+      "learning_rate": 1.4120370370370371e-06,
+      "loss": 0.4175,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.6170705957053,
+      "learning_rate": 1.4351851851851853e-06,
+      "loss": 0.452,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.253387395693687,
+      "learning_rate": 1.4583333333333335e-06,
+      "loss": 0.4109,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.512485748096887,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.4942,
+      "step": 64
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.502930790545099,
+      "learning_rate": 1.5046296296296298e-06,
+      "loss": 0.415,
+      "step": 65
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.402421549531558,
+      "learning_rate": 1.527777777777778e-06,
+      "loss": 0.4425,
+      "step": 66
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.5965610814870876,
+      "learning_rate": 1.550925925925926e-06,
+      "loss": 0.4154,
+      "step": 67
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.603621041567123,
+      "learning_rate": 1.5740740740740742e-06,
+      "loss": 0.4288,
+      "step": 68
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9344560589041366,
+      "learning_rate": 1.5972222222222221e-06,
+      "loss": 0.4476,
+      "step": 69
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.4923666252251553,
+      "learning_rate": 1.6203703703703705e-06,
+      "loss": 0.4369,
+      "step": 70
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.230254972635425,
+      "learning_rate": 1.6435185185185187e-06,
+      "loss": 0.4787,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 23.25230349800371,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.4075,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.319343606401032,
+      "learning_rate": 1.689814814814815e-06,
+      "loss": 0.4013,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.902763495866052,
+      "learning_rate": 1.7129629629629632e-06,
+      "loss": 0.4568,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.941850704381902,
+      "learning_rate": 1.7361111111111112e-06,
+      "loss": 0.4362,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.508785478529396,
+      "learning_rate": 1.7592592592592594e-06,
+      "loss": 0.4692,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.361170477717208,
+      "learning_rate": 1.7824074074074073e-06,
+      "loss": 0.4322,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.875147974816881,
+      "learning_rate": 1.8055555555555557e-06,
+      "loss": 0.4245,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.9931579225032765,
+      "learning_rate": 1.828703703703704e-06,
+      "loss": 0.4021,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3465020759904041,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.1467,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.695669938677484,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.4497,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.69273699098346,
+      "learning_rate": 1.8981481481481484e-06,
+      "loss": 0.4679,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.09127528663891,
+      "learning_rate": 1.921296296296296e-06,
+      "loss": 0.4188,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.328880960866712,
+      "learning_rate": 1.944444444444445e-06,
+      "loss": 0.4414,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.732387946550619,
+      "learning_rate": 1.967592592592593e-06,
+      "loss": 0.4319,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.014515378716921,
+      "learning_rate": 1.9907407407407407e-06,
+      "loss": 0.4386,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.896661503191542,
+      "learning_rate": 2.0138888888888893e-06,
+      "loss": 0.4575,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.929195484104582,
+      "learning_rate": 2.037037037037037e-06,
+      "loss": 0.4319,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.639610554876977,
+      "learning_rate": 2.0601851851851853e-06,
+      "loss": 0.4508,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.541334565341399,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.4245,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.036355003479364,
+      "learning_rate": 2.1064814814814816e-06,
+      "loss": 0.4051,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2360901576212675,
+      "learning_rate": 2.1296296296296298e-06,
+      "loss": 0.4061,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.279200649589632,
+      "learning_rate": 2.152777777777778e-06,
+      "loss": 0.4397,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.1613271862073145,
+      "learning_rate": 2.175925925925926e-06,
+      "loss": 0.3934,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.840127181935883,
+      "learning_rate": 2.1990740740740743e-06,
+      "loss": 0.4119,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.291223965966735,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.4465,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.938125861829971,
+      "learning_rate": 2.2453703703703707e-06,
+      "loss": 0.4674,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.138946229334704,
+      "learning_rate": 2.268518518518519e-06,
+      "loss": 0.4346,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.290809510333111,
+      "learning_rate": 2.2916666666666666e-06,
+      "loss": 0.4086,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.012758715403518,
+      "learning_rate": 2.314814814814815e-06,
+      "loss": 0.3833,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.6583895420680745,
+      "learning_rate": 2.3379629629629634e-06,
+      "loss": 0.4438,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7502875744071793,
+      "learning_rate": 2.361111111111111e-06,
+      "loss": 0.4322,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5483185758460394,
+      "learning_rate": 2.3842592592592593e-06,
+      "loss": 0.4465,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.916244965307016,
+      "learning_rate": 2.4074074074074075e-06,
+      "loss": 0.4108,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4434338951593024,
+      "learning_rate": 2.4305555555555557e-06,
+      "loss": 0.4352,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.186633547496308,
+      "learning_rate": 2.453703703703704e-06,
+      "loss": 0.4316,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8831567496295896,
+      "learning_rate": 2.476851851851852e-06,
+      "loss": 0.4363,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8878897391046396,
+      "learning_rate": 2.5e-06,
+      "loss": 0.4632,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9048612244629046,
+      "learning_rate": 2.5231481481481484e-06,
+      "loss": 0.4362,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.576083701616785,
+      "learning_rate": 2.5462962962962966e-06,
+      "loss": 0.4444,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.134843658676051,
+      "learning_rate": 2.5694444444444443e-06,
+      "loss": 0.4358,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3446391684348655,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.4329,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.792284698988999,
+      "learning_rate": 2.615740740740741e-06,
+      "loss": 0.4531,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.240729726107102,
+      "learning_rate": 2.6388888888888893e-06,
+      "loss": 0.4331,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.463018534198631,
+      "learning_rate": 2.6620370370370374e-06,
+      "loss": 0.4533,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.97401210516448,
+      "learning_rate": 2.6851851851851856e-06,
+      "loss": 0.4124,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.459597349522393,
+      "learning_rate": 2.7083333333333334e-06,
+      "loss": 0.4582,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.70929250624902,
+      "learning_rate": 2.7314814814814816e-06,
+      "loss": 0.4421,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.99480573065225,
+      "learning_rate": 2.7546296296296297e-06,
+      "loss": 0.432,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.272022369943344,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 0.4855,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.810024802840501,
+      "learning_rate": 2.8009259259259265e-06,
+      "loss": 0.4064,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1447130116759814,
+      "learning_rate": 2.8240740740740743e-06,
+      "loss": 0.4347,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.89942966165132,
+      "learning_rate": 2.8472222222222224e-06,
+      "loss": 0.4369,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6629208747146893,
+      "learning_rate": 2.8703703703703706e-06,
+      "loss": 0.3839,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.918475011032535,
+      "learning_rate": 2.893518518518519e-06,
+      "loss": 0.3997,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7810890318572214,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 0.4252,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.683692618147881,
+      "learning_rate": 2.9398148148148147e-06,
+      "loss": 0.4004,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.755946572232162,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.4299,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.09948253931909,
+      "learning_rate": 2.986111111111111e-06,
+      "loss": 0.4367,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.721816116315901,
+      "learning_rate": 3.0092592592592597e-06,
+      "loss": 0.4519,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.225132290110673,
+      "learning_rate": 3.032407407407408e-06,
+      "loss": 0.4003,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.354432184876654,
+      "learning_rate": 3.055555555555556e-06,
+      "loss": 0.4267,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.513304489674287,
+      "learning_rate": 3.078703703703704e-06,
+      "loss": 0.4236,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.1274515109993475,
+      "learning_rate": 3.101851851851852e-06,
+      "loss": 0.4422,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.31220114956726,
+      "learning_rate": 3.125e-06,
+      "loss": 0.4862,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.033760872508533,
+      "learning_rate": 3.1481481481481483e-06,
+      "loss": 0.4584,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.677163998440382,
+      "learning_rate": 3.171296296296297e-06,
+      "loss": 0.4173,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6703463670563803,
+      "learning_rate": 3.1944444444444443e-06,
+      "loss": 0.3953,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.960071076950346,
+      "learning_rate": 3.217592592592593e-06,
+      "loss": 0.4105,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.829536350189737,
+      "learning_rate": 3.240740740740741e-06,
+      "loss": 0.4362,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.909245628246572,
+      "learning_rate": 3.2638888888888892e-06,
+      "loss": 0.386,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.738948356387575,
+      "learning_rate": 3.2870370370370374e-06,
+      "loss": 0.4364,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.520561792208101,
+      "learning_rate": 3.3101851851851856e-06,
+      "loss": 0.3945,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.7527144448970047,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.377,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.7706941629268504,
+      "learning_rate": 3.3564814814814815e-06,
+      "loss": 0.401,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.393483935968832,
+      "learning_rate": 3.37962962962963e-06,
+      "loss": 0.4315,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.219850724068039,
+      "learning_rate": 3.4027777777777783e-06,
+      "loss": 0.4413,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.9623707554846694,
+      "learning_rate": 3.4259259259259265e-06,
+      "loss": 0.4672,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.200726928407887,
+      "learning_rate": 3.449074074074074e-06,
+      "loss": 0.4433,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.222689730732137,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 0.396,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1246501752903546,
+      "learning_rate": 3.4953703703703706e-06,
+      "loss": 0.4403,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.559246973149497,
+      "learning_rate": 3.5185185185185187e-06,
+      "loss": 0.454,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.121530759283131,
+      "learning_rate": 3.5416666666666673e-06,
+      "loss": 0.4023,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 11.264952507387648,
+      "learning_rate": 3.5648148148148147e-06,
+      "loss": 0.3992,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.847301563229513,
+      "learning_rate": 3.5879629629629633e-06,
+      "loss": 0.4017,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.6229148197827294,
+      "learning_rate": 3.6111111111111115e-06,
+      "loss": 0.3949,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.1765667579530765,
+      "learning_rate": 3.6342592592592596e-06,
+      "loss": 0.4463,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.064252503022133,
+      "learning_rate": 3.657407407407408e-06,
+      "loss": 0.4488,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.150728833966288,
+      "learning_rate": 3.680555555555556e-06,
+      "loss": 0.4651,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.6316342692543957,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 0.425,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.023683629708625,
+      "learning_rate": 3.726851851851852e-06,
+      "loss": 0.4425,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.05590411224082,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.4127,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3539255273171037,
+      "learning_rate": 3.7731481481481487e-06,
+      "loss": 0.1415,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.096578666175323,
+      "learning_rate": 3.796296296296297e-06,
+      "loss": 0.4194,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.306721377017303,
+      "learning_rate": 3.819444444444444e-06,
+      "loss": 0.4308,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.8988022789372865,
+      "learning_rate": 3.842592592592592e-06,
+      "loss": 0.4319,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.5523156091918855,
+      "learning_rate": 3.865740740740741e-06,
+      "loss": 0.402,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3057884217756085,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 0.1446,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8020267610329905,
+      "learning_rate": 3.912037037037038e-06,
+      "loss": 0.4554,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 9.212518340883127,
+      "learning_rate": 3.935185185185186e-06,
+      "loss": 0.4536,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.112811801112081,
+      "learning_rate": 3.958333333333333e-06,
+      "loss": 0.4336,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.803824132545977,
+      "learning_rate": 3.9814814814814814e-06,
+      "loss": 0.3988,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.168966583653092,
+      "learning_rate": 4.00462962962963e-06,
+      "loss": 0.4229,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.281744335879201,
+      "learning_rate": 4.027777777777779e-06,
+      "loss": 0.4195,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.000675285117169,
+      "learning_rate": 4.050925925925927e-06,
+      "loss": 0.4406,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.848386901658452,
+      "learning_rate": 4.074074074074074e-06,
+      "loss": 0.3806,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.531747457562333,
+      "learning_rate": 4.097222222222222e-06,
+      "loss": 0.4338,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.630156643198267,
+      "learning_rate": 4.1203703703703705e-06,
+      "loss": 0.4243,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.754516331594139,
+      "learning_rate": 4.143518518518519e-06,
+      "loss": 0.4036,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.440587745868907,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.4203,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.527720154121761,
+      "learning_rate": 4.189814814814815e-06,
+      "loss": 0.4382,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 74.1185697219274,
+      "learning_rate": 4.212962962962963e-06,
+      "loss": 0.4019,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.4807658307255775,
+      "learning_rate": 4.236111111111111e-06,
+      "loss": 0.4096,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 8.81229826912245,
+      "learning_rate": 4.2592592592592596e-06,
+      "loss": 0.4506,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.158663475664476,
+      "learning_rate": 4.282407407407408e-06,
+      "loss": 0.4119,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.278691053988669,
+      "learning_rate": 4.305555555555556e-06,
+      "loss": 0.4108,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.762645906435576,
+      "learning_rate": 4.328703703703704e-06,
+      "loss": 0.4023,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.3120010725331186,
+      "learning_rate": 4.351851851851852e-06,
+      "loss": 0.4356,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.877923950733879,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.4561,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.075363029771738,
+      "learning_rate": 4.398148148148149e-06,
+      "loss": 0.4144,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.783775219349133,
+      "learning_rate": 4.421296296296297e-06,
+      "loss": 0.392,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7549103888739914,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.4159,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.013715256279632,
+      "learning_rate": 4.467592592592593e-06,
+      "loss": 0.3806,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.988747595018511,
+      "learning_rate": 4.490740740740741e-06,
+      "loss": 0.4247,
+      "step": 194
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.642404558999347,
+      "learning_rate": 4.5138888888888895e-06,
+      "loss": 0.435,
+      "step": 195
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.530297932788812,
+      "learning_rate": 4.537037037037038e-06,
+      "loss": 0.4357,
+      "step": 196
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.875846102196212,
+      "learning_rate": 4.560185185185186e-06,
+      "loss": 0.38,
+      "step": 197
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.829648635565988,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 0.4546,
+      "step": 198
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.34942790926101686,
+      "learning_rate": 4.606481481481481e-06,
+      "loss": 0.143,
+      "step": 199
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 6.0141949253030536,
+      "learning_rate": 4.62962962962963e-06,
+      "loss": 0.449,
+      "step": 200
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.113424136487239,
+      "learning_rate": 4.652777777777779e-06,
+      "loss": 0.4245,
+      "step": 201
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.373220319808973,
+      "learning_rate": 4.675925925925927e-06,
+      "loss": 0.433,
+      "step": 202
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.3270005478730719,
+      "learning_rate": 4.699074074074074e-06,
+      "loss": 0.1516,
+      "step": 203
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.990572550655989,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 0.4058,
+      "step": 204
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.707868299079014,
+      "learning_rate": 4.7453703703703705e-06,
+      "loss": 0.395,
+      "step": 205
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.077624671179822,
+      "learning_rate": 4.768518518518519e-06,
+      "loss": 0.3864,
+      "step": 206
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 7.840887885095195,
+      "learning_rate": 4.791666666666668e-06,
+      "loss": 0.4325,
+      "step": 207
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.027747895394712,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.3919,
+      "step": 208
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.400950970014762,
+      "learning_rate": 4.837962962962963e-06,
+      "loss": 0.4303,
+      "step": 209
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.1384830980694645,
+      "learning_rate": 4.861111111111111e-06,
+      "loss": 0.4472,
+      "step": 210
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.550974811246842,
+      "learning_rate": 4.8842592592592595e-06,
+      "loss": 0.4323,
+      "step": 211
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.982094203671558,
+      "learning_rate": 4.907407407407408e-06,
+      "loss": 0.4458,
+      "step": 212
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.369900086458573,
+      "learning_rate": 4.930555555555556e-06,
+      "loss": 0.3956,
+      "step": 213
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.8458678780166236,
+      "learning_rate": 4.953703703703704e-06,
+      "loss": 0.4022,
+      "step": 214
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.448501843754455,
+      "learning_rate": 4.976851851851852e-06,
+      "loss": 0.44,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 12.026042580051215,
+      "learning_rate": 5e-06,
+      "loss": 0.4413,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.247371920929824,
+      "learning_rate": 5.023148148148148e-06,
+      "loss": 0.45,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 28.132091958510625,
+      "learning_rate": 5.046296296296297e-06,
+      "loss": 0.4072,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.657539802846649,
+      "learning_rate": 5.069444444444445e-06,
+      "loss": 0.4168,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.133009719501354,
+      "learning_rate": 5.092592592592593e-06,
+      "loss": 0.4293,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.525665707155952,
+      "learning_rate": 5.115740740740741e-06,
+      "loss": 0.4675,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.250697544965712,
+      "learning_rate": 5.138888888888889e-06,
+      "loss": 0.4027,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 16.36979317449679,
+      "learning_rate": 5.162037037037038e-06,
+      "loss": 0.4183,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.707821986157068,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.4587,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8097979568494766,
+      "learning_rate": 5.208333333333334e-06,
+      "loss": 0.4708,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4625199898166055,
+      "learning_rate": 5.231481481481482e-06,
+      "loss": 0.451,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.40859486622742,
+      "learning_rate": 5.2546296296296295e-06,
+      "loss": 0.439,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.652576462110194,
+      "learning_rate": 5.2777777777777785e-06,
+      "loss": 0.4183,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.590122323810632,
+      "learning_rate": 5.300925925925926e-06,
+      "loss": 0.4624,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.693506693970371,
+      "learning_rate": 5.324074074074075e-06,
+      "loss": 0.4573,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.667141145463999,
+      "learning_rate": 5.347222222222222e-06,
+      "loss": 0.4227,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5950774838371506,
+      "learning_rate": 5.370370370370371e-06,
+      "loss": 0.4083,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4581729188154227,
+      "learning_rate": 5.3935185185185194e-06,
+      "loss": 0.3973,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.454809161643927,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 0.4048,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.166397758445659,
+      "learning_rate": 5.439814814814816e-06,
+      "loss": 0.4002,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.969509451149642,
+      "learning_rate": 5.462962962962963e-06,
+      "loss": 0.453,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.602148387009082,
+      "learning_rate": 5.486111111111112e-06,
+      "loss": 0.421,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.192831441200742,
+      "learning_rate": 5.5092592592592595e-06,
+      "loss": 0.407,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.343292998664184,
+      "learning_rate": 5.532407407407408e-06,
+      "loss": 0.3952,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7983890846874173,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.4088,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.75389781821996,
+      "learning_rate": 5.578703703703704e-06,
+      "loss": 0.4326,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.071352747242176,
+      "learning_rate": 5.601851851851853e-06,
+      "loss": 0.4441,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7860487973028016,
+      "learning_rate": 5.625e-06,
+      "loss": 0.4186,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.246371270094856,
+      "learning_rate": 5.6481481481481485e-06,
+      "loss": 0.4167,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.562623005260343,
+      "learning_rate": 5.671296296296297e-06,
+      "loss": 0.4139,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.339463917747065,
+      "learning_rate": 5.694444444444445e-06,
+      "loss": 0.4072,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9750254857541085,
+      "learning_rate": 5.717592592592593e-06,
+      "loss": 0.4016,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.828396133059335,
+      "learning_rate": 5.740740740740741e-06,
+      "loss": 0.4334,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.858246108047801,
+      "learning_rate": 5.7638888888888886e-06,
+      "loss": 0.3992,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.826572996279525,
+      "learning_rate": 5.787037037037038e-06,
+      "loss": 0.4425,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.857236374240313,
+      "learning_rate": 5.810185185185186e-06,
+      "loss": 0.4123,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.9221905536329675,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.4413,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.381680690682695,
+      "learning_rate": 5.856481481481482e-06,
+      "loss": 0.4027,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.393263307790279,
+      "learning_rate": 5.8796296296296295e-06,
+      "loss": 0.4071,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.4523043768394779,
+      "learning_rate": 5.9027777777777785e-06,
+      "loss": 0.1603,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.435549780168575,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.4094,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.809424871989361,
+      "learning_rate": 5.949074074074075e-06,
+      "loss": 0.3917,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0788836831196096,
+      "learning_rate": 5.972222222222222e-06,
+      "loss": 0.4019,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.967472820868868,
+      "learning_rate": 5.995370370370371e-06,
+      "loss": 0.4285,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.397425161596995,
+      "learning_rate": 6.018518518518519e-06,
+      "loss": 0.4124,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.246014141112839,
+      "learning_rate": 6.041666666666667e-06,
+      "loss": 0.4029,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3818127905487234,
+      "learning_rate": 6.064814814814816e-06,
+      "loss": 0.4186,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6190456138814913,
+      "learning_rate": 6.087962962962963e-06,
+      "loss": 0.4027,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.286087845404559,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 0.4296,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.6728188040617407,
+      "learning_rate": 6.134259259259259e-06,
+      "loss": 0.4299,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.370956554535659,
+      "learning_rate": 6.157407407407408e-06,
+      "loss": 0.4202,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.784251656823484,
+      "learning_rate": 6.180555555555557e-06,
+      "loss": 0.4342,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.578037377219002,
+      "learning_rate": 6.203703703703704e-06,
+      "loss": 0.4457,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.258249473479173,
+      "learning_rate": 6.226851851851853e-06,
+      "loss": 0.401,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.291740800243557,
+      "learning_rate": 6.25e-06,
+      "loss": 0.4124,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6706353058421115,
+      "learning_rate": 6.2731481481481485e-06,
+      "loss": 0.4398,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.926435600604493,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 0.4387,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7873878792369475,
+      "learning_rate": 6.319444444444445e-06,
+      "loss": 0.3814,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.55604687864791,
+      "learning_rate": 6.342592592592594e-06,
+      "loss": 0.3807,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.880645728741349,
+      "learning_rate": 6.365740740740741e-06,
+      "loss": 0.4245,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.737698222678566,
+      "learning_rate": 6.3888888888888885e-06,
+      "loss": 0.406,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.374360633506666,
+      "learning_rate": 6.4120370370370375e-06,
+      "loss": 0.4361,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.123037405860801,
+      "learning_rate": 6.435185185185186e-06,
+      "loss": 0.418,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.164633201038273,
+      "learning_rate": 6.458333333333334e-06,
+      "loss": 0.4442,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.509955063685664,
+      "learning_rate": 6.481481481481482e-06,
+      "loss": 0.1744,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.947935328914975,
+      "learning_rate": 6.504629629629629e-06,
+      "loss": 0.405,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.241680310110642,
+      "learning_rate": 6.5277777777777784e-06,
+      "loss": 0.4277,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.603193475021532,
+      "learning_rate": 6.550925925925926e-06,
+      "loss": 0.4483,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.744351438429756,
+      "learning_rate": 6.574074074074075e-06,
+      "loss": 0.4519,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.811434480324238,
+      "learning_rate": 6.597222222222223e-06,
+      "loss": 0.4096,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.985680466861682,
+      "learning_rate": 6.620370370370371e-06,
+      "loss": 0.4057,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.457954785870698,
+      "learning_rate": 6.643518518518519e-06,
+      "loss": 0.4072,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7225382834117053,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.4689,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.643794799406263,
+      "learning_rate": 6.689814814814816e-06,
+      "loss": 0.4131,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6096422128223619,
+      "learning_rate": 6.712962962962963e-06,
+      "loss": 0.1737,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.8399414459914345,
+      "learning_rate": 6.736111111111112e-06,
+      "loss": 0.4353,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.29205371952207,
+      "learning_rate": 6.75925925925926e-06,
+      "loss": 0.4187,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.3458845460828797,
+      "learning_rate": 6.7824074074074075e-06,
+      "loss": 0.4529,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.986403738637075,
+      "learning_rate": 6.8055555555555566e-06,
+      "loss": 0.3895,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.631970764059017,
+      "learning_rate": 6.828703703703704e-06,
+      "loss": 0.4014,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.229390167786846,
+      "learning_rate": 6.851851851851853e-06,
+      "loss": 0.4554,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8855455384311304,
+      "learning_rate": 6.875e-06,
+      "loss": 0.392,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.113347473239308,
+      "learning_rate": 6.898148148148148e-06,
+      "loss": 0.4489,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.711017881486881,
+      "learning_rate": 6.9212962962962974e-06,
+      "loss": 0.4416,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.543326933185458,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.4319,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.793221154232639,
+      "learning_rate": 6.967592592592594e-06,
+      "loss": 0.3801,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.9727946872298996,
+      "learning_rate": 6.990740740740741e-06,
+      "loss": 0.4213,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.1186123106544725,
+      "learning_rate": 7.013888888888889e-06,
+      "loss": 0.3916,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6755061441288377,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.4388,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.6707660273935936,
+      "learning_rate": 7.060185185185186e-06,
+      "loss": 0.4174,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.992270886724995,
+      "learning_rate": 7.083333333333335e-06,
+      "loss": 0.4198,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.321404245050796,
+      "learning_rate": 7.106481481481482e-06,
+      "loss": 0.3935,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.154859536347844,
+      "learning_rate": 7.129629629629629e-06,
+      "loss": 0.4421,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7599089968783749,
+      "learning_rate": 7.152777777777778e-06,
+      "loss": 0.1637,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6266394065613256,
+      "learning_rate": 7.1759259259259266e-06,
+      "loss": 0.1837,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.812903853107458,
+      "learning_rate": 7.199074074074075e-06,
+      "loss": 0.4238,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.961709885207106,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.4198,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8615650492264475,
+      "learning_rate": 7.245370370370371e-06,
+      "loss": 0.4291,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.03222144778231,
+      "learning_rate": 7.268518518518519e-06,
+      "loss": 0.4306,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.2695013921790568,
+      "learning_rate": 7.291666666666667e-06,
+      "loss": 0.4194,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 11.969888160011184,
+      "learning_rate": 7.314814814814816e-06,
+      "loss": 0.3827,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2536623781787937,
+      "learning_rate": 7.337962962962964e-06,
+      "loss": 0.2276,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.270260065312885,
+      "learning_rate": 7.361111111111112e-06,
+      "loss": 0.4451,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.328970138850622,
+      "learning_rate": 7.38425925925926e-06,
+      "loss": 0.4003,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.523144812248686,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.4169,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.100936978230398,
+      "learning_rate": 7.4305555555555565e-06,
+      "loss": 0.4612,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.2063513556927274,
+      "learning_rate": 7.453703703703704e-06,
+      "loss": 0.428,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.9276741748597646,
+      "learning_rate": 7.476851851851853e-06,
+      "loss": 0.4065,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5005682705242562,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3729,
+      "step": 324
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.938966590612719,
+      "learning_rate": 7.523148148148148e-06,
+      "loss": 0.4296,
+      "step": 325
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.02202074202606,
+      "learning_rate": 7.546296296296297e-06,
+      "loss": 0.4182,
+      "step": 326
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.332983901915014,
+      "learning_rate": 7.569444444444445e-06,
+      "loss": 0.3918,
+      "step": 327
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 9.50034593935045,
+      "learning_rate": 7.592592592592594e-06,
+      "loss": 0.448,
+      "step": 328
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5635058946416076,
+      "learning_rate": 7.615740740740741e-06,
+      "loss": 0.4082,
+      "step": 329
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.354910839401507,
+      "learning_rate": 7.638888888888888e-06,
+      "loss": 0.3899,
+      "step": 330
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.47491322384198,
+      "learning_rate": 7.662037037037037e-06,
+      "loss": 0.4369,
+      "step": 331
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.0933061351976745,
+      "learning_rate": 7.685185185185185e-06,
+      "loss": 0.4544,
+      "step": 332
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.288295182297491,
+      "learning_rate": 7.708333333333334e-06,
+      "loss": 0.3998,
+      "step": 333
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.576795089542426,
+      "learning_rate": 7.731481481481483e-06,
+      "loss": 0.4651,
+      "step": 334
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4209782063141483,
+      "learning_rate": 7.75462962962963e-06,
+      "loss": 0.4124,
+      "step": 335
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.989128713852441,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.4463,
+      "step": 336
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5770683513171702,
+      "learning_rate": 7.800925925925926e-06,
+      "loss": 0.1874,
+      "step": 337
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.214773840147203,
+      "learning_rate": 7.824074074074076e-06,
+      "loss": 0.4211,
+      "step": 338
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 22.634834059248394,
+      "learning_rate": 7.847222222222223e-06,
+      "loss": 0.4362,
+      "step": 339
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.021072939934938,
+      "learning_rate": 7.870370370370372e-06,
+      "loss": 0.4485,
+      "step": 340
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.503756770896912,
+      "learning_rate": 7.89351851851852e-06,
+      "loss": 0.4165,
+      "step": 341
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3659348312728342,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 0.1717,
+      "step": 342
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.861198015751531,
+      "learning_rate": 7.939814814814816e-06,
+      "loss": 0.3908,
+      "step": 343
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.044169360394368,
+      "learning_rate": 7.962962962962963e-06,
+      "loss": 0.4325,
+      "step": 344
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.564694490066405,
+      "learning_rate": 7.986111111111112e-06,
+      "loss": 0.4441,
+      "step": 345
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 6.1536493464199875,
+      "learning_rate": 8.00925925925926e-06,
+      "loss": 0.4208,
+      "step": 346
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.689479103128829,
+      "learning_rate": 8.032407407407408e-06,
+      "loss": 0.4489,
+      "step": 347
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.41662260281352,
+      "learning_rate": 8.055555555555557e-06,
+      "loss": 0.4139,
+      "step": 348
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 7.09412077103489,
+      "learning_rate": 8.078703703703705e-06,
+      "loss": 0.4202,
+      "step": 349
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.359371679864202,
+      "learning_rate": 8.101851851851854e-06,
+      "loss": 0.4049,
+      "step": 350
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.5328907262598066,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 0.4215,
+      "step": 351
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.191835756799257,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 0.3845,
+      "step": 352
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.356755472377848,
+      "learning_rate": 8.171296296296297e-06,
+      "loss": 0.4203,
+      "step": 353
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.5763177324854725,
+      "learning_rate": 8.194444444444445e-06,
+      "loss": 0.1805,
+      "step": 354
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 5.311981526642287,
+      "learning_rate": 8.217592592592594e-06,
+      "loss": 0.4498,
+      "step": 355
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 14.770680335419337,
+      "learning_rate": 8.240740740740741e-06,
+      "loss": 0.4262,
+      "step": 356
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.568932045200297,
+      "learning_rate": 8.263888888888888e-06,
+      "loss": 0.4408,
+      "step": 357
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.909806050485566,
+      "learning_rate": 8.287037037037037e-06,
+      "loss": 0.4136,
+      "step": 358
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.3795237336163993,
+      "learning_rate": 8.310185185185186e-06,
+      "loss": 0.1662,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.6014056557686445,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.3951,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.865163688675442,
+      "learning_rate": 8.356481481481483e-06,
+      "loss": 0.4352,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.35292165283931,
+      "learning_rate": 8.37962962962963e-06,
+      "loss": 0.4617,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.238227391053612,
+      "learning_rate": 8.402777777777779e-06,
+      "loss": 0.4201,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.543075364709646,
+      "learning_rate": 8.425925925925926e-06,
+      "loss": 0.4304,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.354617467670506,
+      "learning_rate": 8.449074074074075e-06,
+      "loss": 0.3651,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.966124983947846,
+      "learning_rate": 8.472222222222223e-06,
+      "loss": 0.4434,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7984483895517465,
+      "learning_rate": 8.495370370370372e-06,
+      "loss": 0.4312,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.3764360429761595,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 0.4618,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9670840574187864,
+      "learning_rate": 8.541666666666666e-06,
+      "loss": 0.4226,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4931629089361644,
+      "learning_rate": 8.564814814814816e-06,
+      "loss": 0.4306,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.4753012433885257,
+      "learning_rate": 8.587962962962963e-06,
+      "loss": 0.1749,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.258865664587859,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 0.412,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.963416925233422,
+      "learning_rate": 8.63425925925926e-06,
+      "loss": 0.4295,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.649339114883378,
+      "learning_rate": 8.657407407407408e-06,
+      "loss": 0.4372,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.695369443865818,
+      "learning_rate": 8.680555555555557e-06,
+      "loss": 0.41,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8269605120195296,
+      "learning_rate": 8.703703703703705e-06,
+      "loss": 0.4354,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.038261574538483,
+      "learning_rate": 8.726851851851854e-06,
+      "loss": 0.4066,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.283411701362511,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.4274,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.490318728098695,
+      "learning_rate": 8.773148148148148e-06,
+      "loss": 0.3949,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.672795909859017,
+      "learning_rate": 8.796296296296297e-06,
+      "loss": 0.363,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.443011978646804,
+      "learning_rate": 8.819444444444445e-06,
+      "loss": 0.4296,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6515210638474467,
+      "learning_rate": 8.842592592592594e-06,
+      "loss": 0.444,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 9.486008626846813,
+      "learning_rate": 8.865740740740741e-06,
+      "loss": 0.3953,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0890622604915574,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.442,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.147860947532613,
+      "learning_rate": 8.912037037037037e-06,
+      "loss": 0.4049,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.6935392538917595,
+      "learning_rate": 8.935185185185186e-06,
+      "loss": 0.42,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.915516808393219,
+      "learning_rate": 8.958333333333334e-06,
+      "loss": 0.4454,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.027913281660746,
+      "learning_rate": 8.981481481481483e-06,
+      "loss": 0.4075,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3066000950550047,
+      "learning_rate": 9.00462962962963e-06,
+      "loss": 0.4344,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.155656158091638,
+      "learning_rate": 9.027777777777779e-06,
+      "loss": 0.3672,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.273013947365108,
+      "learning_rate": 9.050925925925926e-06,
+      "loss": 0.4249,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.073354181647686,
+      "learning_rate": 9.074074074074075e-06,
+      "loss": 0.4246,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.280799616166756,
+      "learning_rate": 9.097222222222223e-06,
+      "loss": 0.4542,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.4384153073453825,
+      "learning_rate": 9.120370370370372e-06,
+      "loss": 0.4247,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.123798088447263,
+      "learning_rate": 9.143518518518519e-06,
+      "loss": 0.4518,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.707449867026008,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.3966,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.985106570845797,
+      "learning_rate": 9.189814814814815e-06,
+      "loss": 0.3992,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.9396480461581245,
+      "learning_rate": 9.212962962962963e-06,
+      "loss": 0.4024,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3923038606678495,
+      "learning_rate": 9.236111111111112e-06,
+      "loss": 0.4093,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.988374078026385,
+      "learning_rate": 9.25925925925926e-06,
+      "loss": 0.4365,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.372617123591882,
+      "learning_rate": 9.282407407407408e-06,
+      "loss": 0.3955,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.188622953221999,
+      "learning_rate": 9.305555555555557e-06,
+      "loss": 0.4177,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.652272640749697,
+      "learning_rate": 9.328703703703705e-06,
+      "loss": 0.3949,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.344940279523254,
+      "learning_rate": 9.351851851851854e-06,
+      "loss": 0.4019,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.032972525241581,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 0.419,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6367524736596977,
+      "learning_rate": 9.398148148148148e-06,
+      "loss": 0.403,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.5047013015330055,
+      "learning_rate": 9.421296296296297e-06,
+      "loss": 0.4726,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6568626948642207,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 0.4253,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.679919950424501,
+      "learning_rate": 9.467592592592594e-06,
+      "loss": 0.4534,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.464777140571788,
+      "learning_rate": 9.490740740740741e-06,
+      "loss": 0.4118,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.721894066716717,
+      "learning_rate": 9.51388888888889e-06,
+      "loss": 0.3928,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.085059380503294,
+      "learning_rate": 9.537037037037037e-06,
+      "loss": 0.4466,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.954399764203338,
+      "learning_rate": 9.560185185185186e-06,
+      "loss": 0.4675,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7959383702538427,
+      "learning_rate": 9.583333333333335e-06,
+      "loss": 0.4002,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7677335671998806,
+      "learning_rate": 9.606481481481483e-06,
+      "loss": 0.4095,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.936358583384487,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 0.4077,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.779672782902478,
+      "learning_rate": 9.652777777777779e-06,
+      "loss": 0.4175,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.305894400568077,
+      "learning_rate": 9.675925925925926e-06,
+      "loss": 0.4642,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.694301987072033,
+      "learning_rate": 9.699074074074075e-06,
+      "loss": 0.4697,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.795137005450328,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 0.4302,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.067715970529424,
+      "learning_rate": 9.745370370370372e-06,
+      "loss": 0.4163,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.153724776224593,
+      "learning_rate": 9.768518518518519e-06,
+      "loss": 0.4127,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.187897038928874,
+      "learning_rate": 9.791666666666666e-06,
+      "loss": 0.4672,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.4247643314079665,
+      "learning_rate": 9.814814814814815e-06,
+      "loss": 0.4346,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.592825552325313,
+      "learning_rate": 9.837962962962964e-06,
+      "loss": 0.4114,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.846616180051988,
+      "learning_rate": 9.861111111111112e-06,
+      "loss": 0.4196,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.423563870727898,
+      "learning_rate": 9.88425925925926e-06,
+      "loss": 0.4115,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.303613656017128,
+      "learning_rate": 9.907407407407408e-06,
+      "loss": 0.4125,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.695609035539208,
+      "learning_rate": 9.930555555555557e-06,
+      "loss": 0.4147,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.667459650442308,
+      "learning_rate": 9.953703703703704e-06,
+      "loss": 0.4726,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5199840606060415,
+      "learning_rate": 9.976851851851853e-06,
+      "loss": 0.4041,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.673122350761594,
+      "learning_rate": 1e-05,
+      "loss": 0.4395,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.438503016448449,
+      "learning_rate": 9.99999987320815e-06,
+      "loss": 0.3994,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.562585156785935,
+      "learning_rate": 9.999999492832608e-06,
+      "loss": 0.3709,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.069977253406105,
+      "learning_rate": 9.999998858873394e-06,
+      "loss": 0.4369,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.80667140655336,
+      "learning_rate": 9.999997971330537e-06,
+      "loss": 0.4526,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.45411349157404,
+      "learning_rate": 9.999996830204085e-06,
+      "loss": 0.4177,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.290543497167991,
+      "learning_rate": 9.999995435494093e-06,
+      "loss": 0.476,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6772228663076767,
+      "learning_rate": 9.999993787200635e-06,
+      "loss": 0.398,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.64478368993465,
+      "learning_rate": 9.999991885323793e-06,
+      "loss": 0.4162,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.241914314214585,
+      "learning_rate": 9.999989729863661e-06,
+      "loss": 0.422,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.1773675678511046,
+      "learning_rate": 9.999987320820353e-06,
+      "loss": 0.4213,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.589152863463461,
+      "learning_rate": 9.99998465819399e-06,
+      "loss": 0.4368,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.269170348146434,
+      "learning_rate": 9.999981741984703e-06,
+      "loss": 0.4243,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3348841033805288,
+      "learning_rate": 9.999978572192644e-06,
+      "loss": 0.4344,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 28.90882861950758,
+      "learning_rate": 9.999975148817974e-06,
+      "loss": 0.4442,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6534914724919942,
+      "learning_rate": 9.999971471860864e-06,
+      "loss": 0.4441,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.2980673073846827,
+      "learning_rate": 9.999967541321503e-06,
+      "loss": 0.431,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.517670830589429,
+      "learning_rate": 9.999963357200089e-06,
+      "loss": 0.4205,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 6.095093485694381,
+      "learning_rate": 9.999958919496832e-06,
+      "loss": 0.4549,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.221797546587668,
+      "learning_rate": 9.999954228211962e-06,
+      "loss": 0.4405,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.170999908900734,
+      "learning_rate": 9.999949283345714e-06,
+      "loss": 0.4148,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 10.232964795330545,
+      "learning_rate": 9.999944084898339e-06,
+      "loss": 0.4353,
+      "step": 453
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.49373144014666237,
+      "learning_rate": 9.999938632870099e-06,
+      "loss": 0.1745,
+      "step": 454
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.364899884392843,
+      "learning_rate": 9.999932927261275e-06,
+      "loss": 0.4171,
+      "step": 455
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.274156658288236,
+      "learning_rate": 9.999926968072151e-06,
+      "loss": 0.4428,
+      "step": 456
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 25.912711629991538,
+      "learning_rate": 9.999920755303033e-06,
+      "loss": 0.4212,
+      "step": 457
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.22535060009852,
+      "learning_rate": 9.999914288954235e-06,
+      "loss": 0.4246,
+      "step": 458
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6096611965585508,
+      "learning_rate": 9.999907569026086e-06,
+      "loss": 0.4211,
+      "step": 459
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7666126734814998,
+      "learning_rate": 9.999900595518925e-06,
+      "loss": 0.4524,
+      "step": 460
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.12923476007949,
+      "learning_rate": 9.999893368433106e-06,
+      "loss": 0.4244,
+      "step": 461
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.705110554277285,
+      "learning_rate": 9.999885887768996e-06,
+      "loss": 0.404,
+      "step": 462
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.7536288515178344,
+      "learning_rate": 9.999878153526974e-06,
+      "loss": 0.4228,
+      "step": 463
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.805531209801563,
+      "learning_rate": 9.999870165707434e-06,
+      "loss": 0.4274,
+      "step": 464
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.857101802640722,
+      "learning_rate": 9.999861924310779e-06,
+      "loss": 0.4398,
+      "step": 465
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7242651935150906,
+      "learning_rate": 9.999853429337427e-06,
+      "loss": 0.1951,
+      "step": 466
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.269919226588655,
+      "learning_rate": 9.99984468078781e-06,
+      "loss": 0.4409,
+      "step": 467
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.991834512770562,
+      "learning_rate": 9.999835678662372e-06,
+      "loss": 0.4419,
+      "step": 468
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.163112895864879,
+      "learning_rate": 9.999826422961567e-06,
+      "loss": 0.4461,
+      "step": 469
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 7.354046390209344,
+      "learning_rate": 9.999816913685867e-06,
+      "loss": 0.4156,
+      "step": 470
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.107889077293269,
+      "learning_rate": 9.999807150835754e-06,
+      "loss": 0.4529,
+      "step": 471
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5975380296513355,
+      "learning_rate": 9.999797134411724e-06,
+      "loss": 0.1887,
+      "step": 472
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0703896203137,
+      "learning_rate": 9.99978686441428e-06,
+      "loss": 0.4395,
+      "step": 473
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.266606011479328,
+      "learning_rate": 9.99977634084395e-06,
+      "loss": 0.4283,
+      "step": 474
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.817100355473826,
+      "learning_rate": 9.999765563701263e-06,
+      "loss": 0.4143,
+      "step": 475
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.341675417538916,
+      "learning_rate": 9.999754532986766e-06,
+      "loss": 0.4355,
+      "step": 476
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.838952954663778,
+      "learning_rate": 9.99974324870102e-06,
+      "loss": 0.4368,
+      "step": 477
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.424006027777981,
+      "learning_rate": 9.999731710844596e-06,
+      "loss": 0.4389,
+      "step": 478
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5002582718596282,
+      "learning_rate": 9.999719919418078e-06,
+      "loss": 0.1959,
+      "step": 479
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.16173828731522,
+      "learning_rate": 9.999707874422069e-06,
+      "loss": 0.4184,
+      "step": 480
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9320141342893034,
+      "learning_rate": 9.999695575857175e-06,
+      "loss": 0.4545,
+      "step": 481
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.65782676364887,
+      "learning_rate": 9.999683023724021e-06,
+      "loss": 0.4558,
+      "step": 482
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.135450930651972,
+      "learning_rate": 9.999670218023245e-06,
+      "loss": 0.463,
+      "step": 483
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.0872297505907307,
+      "learning_rate": 9.999657158755495e-06,
+      "loss": 0.4225,
+      "step": 484
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.705367314320185,
+      "learning_rate": 9.999643845921433e-06,
+      "loss": 0.4272,
+      "step": 485
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.523125695558921,
+      "learning_rate": 9.999630279521735e-06,
+      "loss": 0.4262,
+      "step": 486
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.363754160518768,
+      "learning_rate": 9.99961645955709e-06,
+      "loss": 0.4541,
+      "step": 487
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.308182491515727,
+      "learning_rate": 9.999602386028198e-06,
+      "loss": 0.4109,
+      "step": 488
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.225044374692758,
+      "learning_rate": 9.999588058935772e-06,
+      "loss": 0.456,
+      "step": 489
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.777479012036398,
+      "learning_rate": 9.999573478280538e-06,
+      "loss": 0.4516,
+      "step": 490
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.936982907388557,
+      "learning_rate": 9.999558644063238e-06,
+      "loss": 0.4314,
+      "step": 491
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 15.490812951880088,
+      "learning_rate": 9.999543556284623e-06,
+      "loss": 0.415,
+      "step": 492
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.063930167817286,
+      "learning_rate": 9.999528214945457e-06,
+      "loss": 0.4417,
+      "step": 493
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.941144696082619,
+      "learning_rate": 9.999512620046523e-06,
+      "loss": 0.386,
+      "step": 494
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.369084683985732,
+      "learning_rate": 9.999496771588606e-06,
+      "loss": 0.4332,
+      "step": 495
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.440386845128256,
+      "learning_rate": 9.99948066957251e-06,
+      "loss": 0.4208,
+      "step": 496
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.214974198202042,
+      "learning_rate": 9.999464313999057e-06,
+      "loss": 0.4252,
+      "step": 497
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.067365501365649,
+      "learning_rate": 9.99944770486907e-06,
+      "loss": 0.4463,
+      "step": 498
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.097055874905206,
+      "learning_rate": 9.999430842183397e-06,
+      "loss": 0.4308,
+      "step": 499
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.5304950321016975,
+      "learning_rate": 9.999413725942891e-06,
+      "loss": 0.4651,
+      "step": 500
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 18.233488599535587,
+      "learning_rate": 9.999396356148418e-06,
+      "loss": 0.4285,
+      "step": 501
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.083029644599197,
+      "learning_rate": 9.999378732800862e-06,
+      "loss": 0.421,
+      "step": 502
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6466124174106275,
+      "learning_rate": 9.999360855901115e-06,
+      "loss": 0.1857,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.408309994694042,
+      "learning_rate": 9.999342725450085e-06,
+      "loss": 0.4582,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.38584231055951973,
+      "learning_rate": 9.99932434144869e-06,
+      "loss": 0.1731,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3905744656842725,
+      "learning_rate": 9.999305703897864e-06,
+      "loss": 0.4299,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.780340948465748,
+      "learning_rate": 9.99928681279855e-06,
+      "loss": 0.4248,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9698591176214792,
+      "learning_rate": 9.999267668151709e-06,
+      "loss": 0.421,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7306522286089417,
+      "learning_rate": 9.999248269958308e-06,
+      "loss": 0.1929,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.784652315245995,
+      "learning_rate": 9.999228618219337e-06,
+      "loss": 0.4477,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.277439988304026,
+      "learning_rate": 9.999208712935786e-06,
+      "loss": 0.4189,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.496292324961903,
+      "learning_rate": 9.999188554108667e-06,
+      "loss": 0.4524,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.8631250949190825,
+      "learning_rate": 9.999168141739003e-06,
+      "loss": 0.4171,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.646935460856349,
+      "learning_rate": 9.999147475827832e-06,
+      "loss": 0.4179,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8211252939766327,
+      "learning_rate": 9.999126556376197e-06,
+      "loss": 0.3931,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.68685502019799,
+      "learning_rate": 9.999105383385161e-06,
+      "loss": 0.405,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6325839771664556,
+      "learning_rate": 9.9990839568558e-06,
+      "loss": 0.3935,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.529921280081829,
+      "learning_rate": 9.999062276789196e-06,
+      "loss": 0.459,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 15.410960704111279,
+      "learning_rate": 9.999040343186452e-06,
+      "loss": 0.3975,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4385910155822463,
+      "learning_rate": 9.999018156048681e-06,
+      "loss": 0.4248,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.677706601932677,
+      "learning_rate": 9.998995715377005e-06,
+      "loss": 0.4611,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.737969636502209,
+      "learning_rate": 9.998973021172564e-06,
+      "loss": 0.4492,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.462222267585546,
+      "learning_rate": 9.998950073436512e-06,
+      "loss": 0.4259,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.329341454329197,
+      "learning_rate": 9.998926872170006e-06,
+      "loss": 0.4161,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.685936619726238,
+      "learning_rate": 9.998903417374228e-06,
+      "loss": 0.4272,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.7204278924282885,
+      "learning_rate": 9.998879709050366e-06,
+      "loss": 0.443,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9402735976042274,
+      "learning_rate": 9.998855747199623e-06,
+      "loss": 0.4571,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.374326454954753,
+      "learning_rate": 9.998831531823212e-06,
+      "loss": 0.4413,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.634749589794055,
+      "learning_rate": 9.998807062922365e-06,
+      "loss": 0.4407,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.293018319588153,
+      "learning_rate": 9.99878234049832e-06,
+      "loss": 0.4179,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.405361065075835,
+      "learning_rate": 9.998757364552332e-06,
+      "loss": 0.4173,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.326983424624245,
+      "learning_rate": 9.998732135085665e-06,
+      "loss": 0.4112,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.839059472253815,
+      "learning_rate": 9.998706652099603e-06,
+      "loss": 0.4402,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.249840672069888,
+      "learning_rate": 9.998680915595436e-06,
+      "loss": 0.4307,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9484293731550095,
+      "learning_rate": 9.99865492557447e-06,
+      "loss": 0.3936,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.757118843248669,
+      "learning_rate": 9.998628682038023e-06,
+      "loss": 0.4458,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 20.34357830392036,
+      "learning_rate": 9.998602184987425e-06,
+      "loss": 0.4863,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.668886041534006,
+      "learning_rate": 9.998575434424021e-06,
+      "loss": 0.4221,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.010253343031133,
+      "learning_rate": 9.998548430349167e-06,
+      "loss": 0.426,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.527831520042611,
+      "learning_rate": 9.998521172764233e-06,
+      "loss": 0.3959,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.1336877523210935,
+      "learning_rate": 9.9984936616706e-06,
+      "loss": 0.4284,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.830505011647217,
+      "learning_rate": 9.998465897069665e-06,
+      "loss": 0.4305,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.34223496547748,
+      "learning_rate": 9.998437878962837e-06,
+      "loss": 0.4596,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8809226461238493,
+      "learning_rate": 9.998409607351534e-06,
+      "loss": 0.2225,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.037780771764948,
+      "learning_rate": 9.998381082237193e-06,
+      "loss": 0.4216,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.4485190194824,
+      "learning_rate": 9.998352303621258e-06,
+      "loss": 0.4159,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.6872447958806225,
+      "learning_rate": 9.99832327150519e-06,
+      "loss": 0.4029,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.3620920158269305,
+      "learning_rate": 9.998293985890461e-06,
+      "loss": 0.4516,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.506067366196322,
+      "learning_rate": 9.998264446778557e-06,
+      "loss": 0.4662,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.569735230540628,
+      "learning_rate": 9.998234654170975e-06,
+      "loss": 0.4053,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.196260554352589,
+      "learning_rate": 9.998204608069226e-06,
+      "loss": 0.4255,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9896031228537154,
+      "learning_rate": 9.998174308474836e-06,
+      "loss": 0.3936,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 19.178182191744757,
+      "learning_rate": 9.998143755389338e-06,
+      "loss": 0.4058,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.3496646863183965,
+      "learning_rate": 9.998112948814286e-06,
+      "loss": 0.4351,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0824274904315447,
+      "learning_rate": 9.998081888751239e-06,
+      "loss": 0.4353,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.41045298090762,
+      "learning_rate": 9.998050575201772e-06,
+      "loss": 0.4747,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.698669131804171,
+      "learning_rate": 9.998019008167476e-06,
+      "loss": 0.4326,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5999841780453314,
+      "learning_rate": 9.99798718764995e-06,
+      "loss": 0.4208,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.696124255631736,
+      "learning_rate": 9.997955113650808e-06,
+      "loss": 0.4719,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.0922334764113932,
+      "learning_rate": 9.997922786171677e-06,
+      "loss": 0.4277,
+      "step": 560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5242065263255165,
+      "learning_rate": 9.997890205214196e-06,
+      "loss": 0.384,
+      "step": 561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.1795019676127945,
+      "learning_rate": 9.997857370780018e-06,
+      "loss": 0.4286,
+      "step": 562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 10.439233658984238,
+      "learning_rate": 9.99782428287081e-06,
+      "loss": 0.4279,
+      "step": 563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.470396520665919,
+      "learning_rate": 9.997790941488246e-06,
+      "loss": 0.4428,
+      "step": 564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.290403600296553,
+      "learning_rate": 9.997757346634019e-06,
+      "loss": 0.4668,
+      "step": 565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 25.81799875123413,
+      "learning_rate": 9.997723498309834e-06,
+      "loss": 0.4448,
+      "step": 566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.1630312652202797,
+      "learning_rate": 9.997689396517408e-06,
+      "loss": 0.4153,
+      "step": 567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.935114131212486,
+      "learning_rate": 9.997655041258466e-06,
+      "loss": 0.3958,
+      "step": 568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.534275706451298,
+      "learning_rate": 9.997620432534756e-06,
+      "loss": 0.4381,
+      "step": 569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.116541785895124,
+      "learning_rate": 9.99758557034803e-06,
+      "loss": 0.4345,
+      "step": 570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4751517465335375,
+      "learning_rate": 9.997550454700056e-06,
+      "loss": 0.4492,
+      "step": 571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.260680190998716,
+      "learning_rate": 9.997515085592618e-06,
+      "loss": 0.4509,
+      "step": 572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.612093670157932,
+      "learning_rate": 9.997479463027506e-06,
+      "loss": 0.4331,
+      "step": 573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 34.19060943655095,
+      "learning_rate": 9.99744358700653e-06,
+      "loss": 0.4312,
+      "step": 574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.654469099710379,
+      "learning_rate": 9.997407457531507e-06,
+      "loss": 0.4285,
+      "step": 575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.576463128132242,
+      "learning_rate": 9.99737107460427e-06,
+      "loss": 0.4319,
+      "step": 576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.077357576605425,
+      "learning_rate": 9.997334438226664e-06,
+      "loss": 0.4594,
+      "step": 577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.230730235964226,
+      "learning_rate": 9.997297548400548e-06,
+      "loss": 0.4409,
+      "step": 578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6134990753068688,
+      "learning_rate": 9.997260405127792e-06,
+      "loss": 0.412,
+      "step": 579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.583804445674505,
+      "learning_rate": 9.997223008410281e-06,
+      "loss": 0.4554,
+      "step": 580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.4901268376312835,
+      "learning_rate": 9.99718535824991e-06,
+      "loss": 0.4132,
+      "step": 581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.927686790050747,
+      "learning_rate": 9.99714745464859e-06,
+      "loss": 0.4195,
+      "step": 582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7096960126301917,
+      "learning_rate": 9.997109297608244e-06,
+      "loss": 0.405,
+      "step": 583
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 12.231614305006246,
+      "learning_rate": 9.997070887130803e-06,
+      "loss": 0.4614,
+      "step": 584
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.6280535457515715,
+      "learning_rate": 9.997032223218222e-06,
+      "loss": 0.4344,
+      "step": 585
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.360393086483778,
+      "learning_rate": 9.996993305872456e-06,
+      "loss": 0.4288,
+      "step": 586
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5685682480145453,
+      "learning_rate": 9.99695413509548e-06,
+      "loss": 0.4572,
+      "step": 587
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.30426895830812,
+      "learning_rate": 9.99691471088928e-06,
+      "loss": 0.3905,
+      "step": 588
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.072135341937139,
+      "learning_rate": 9.99687503325586e-06,
+      "loss": 0.4393,
+      "step": 589
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.053237839610123,
+      "learning_rate": 9.996835102197227e-06,
+      "loss": 0.4473,
+      "step": 590
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.978008870488449,
+      "learning_rate": 9.996794917715408e-06,
+      "loss": 0.4432,
+      "step": 591
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.4137118861227265,
+      "learning_rate": 9.996754479812443e-06,
+      "loss": 0.4368,
+      "step": 592
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9833766138979456,
+      "learning_rate": 9.99671378849038e-06,
+      "loss": 0.4514,
+      "step": 593
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4024563154733873,
+      "learning_rate": 9.996672843751283e-06,
+      "loss": 0.4294,
+      "step": 594
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.9326024425452806,
+      "learning_rate": 9.99663164559723e-06,
+      "loss": 0.4106,
+      "step": 595
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.78687324244639,
+      "learning_rate": 9.996590194030311e-06,
+      "loss": 0.4412,
+      "step": 596
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.158213978306523,
+      "learning_rate": 9.996548489052627e-06,
+      "loss": 0.4132,
+      "step": 597
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.557517357906297,
+      "learning_rate": 9.996506530666292e-06,
+      "loss": 0.4103,
+      "step": 598
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.535610002140052,
+      "learning_rate": 9.996464318873435e-06,
+      "loss": 0.42,
+      "step": 599
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.077200734405768,
+      "learning_rate": 9.9964218536762e-06,
+      "loss": 0.4269,
+      "step": 600
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.184301209805311,
+      "learning_rate": 9.996379135076736e-06,
+      "loss": 0.4274,
+      "step": 601
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.562522838227557,
+      "learning_rate": 9.996336163077212e-06,
+      "loss": 0.3977,
+      "step": 602
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3761175658918636,
+      "learning_rate": 9.996292937679805e-06,
+      "loss": 0.2978,
+      "step": 603
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.012452085261214,
+      "learning_rate": 9.996249458886711e-06,
+      "loss": 0.4204,
+      "step": 604
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.7083492279618415,
+      "learning_rate": 9.996205726700133e-06,
+      "loss": 0.4411,
+      "step": 605
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.739417383706326,
+      "learning_rate": 9.996161741122288e-06,
+      "loss": 0.4492,
+      "step": 606
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8418974558893253,
+      "learning_rate": 9.99611750215541e-06,
+      "loss": 0.4334,
+      "step": 607
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4441466808445993,
+      "learning_rate": 9.996073009801738e-06,
+      "loss": 0.4304,
+      "step": 608
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.7221680617709656,
+      "learning_rate": 9.996028264063531e-06,
+      "loss": 0.3998,
+      "step": 609
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.322044654025656,
+      "learning_rate": 9.995983264943062e-06,
+      "loss": 0.4398,
+      "step": 610
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 13.240166428663265,
+      "learning_rate": 9.995938012442605e-06,
+      "loss": 0.4323,
+      "step": 611
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8717000851447314,
+      "learning_rate": 9.995892506564461e-06,
+      "loss": 0.4063,
+      "step": 612
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.886687297093467,
+      "learning_rate": 9.99584674731094e-06,
+      "loss": 0.4715,
+      "step": 613
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.270847202747545,
+      "learning_rate": 9.995800734684356e-06,
+      "loss": 0.4053,
+      "step": 614
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.657471032574639,
+      "learning_rate": 9.995754468687046e-06,
+      "loss": 0.4367,
+      "step": 615
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.7144463490182584,
+      "learning_rate": 9.995707949321358e-06,
+      "loss": 0.4511,
+      "step": 616
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8394598075428297,
+      "learning_rate": 9.995661176589649e-06,
+      "loss": 0.4289,
+      "step": 617
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9802515649926677,
+      "learning_rate": 9.995614150494293e-06,
+      "loss": 0.4575,
+      "step": 618
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.048848652188439,
+      "learning_rate": 9.995566871037671e-06,
+      "loss": 0.4188,
+      "step": 619
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.247314537173422,
+      "learning_rate": 9.995519338222187e-06,
+      "loss": 0.3954,
+      "step": 620
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.421027775618705,
+      "learning_rate": 9.995471552050247e-06,
+      "loss": 0.4644,
+      "step": 621
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.483394996208876,
+      "learning_rate": 9.995423512524277e-06,
+      "loss": 0.4376,
+      "step": 622
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.035059674544443,
+      "learning_rate": 9.995375219646711e-06,
+      "loss": 0.4306,
+      "step": 623
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.575283057812508,
+      "learning_rate": 9.99532667342e-06,
+      "loss": 0.4294,
+      "step": 624
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6247766323035497,
+      "learning_rate": 9.995277873846606e-06,
+      "loss": 0.4295,
+      "step": 625
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.724002865629507,
+      "learning_rate": 9.995228820929007e-06,
+      "loss": 0.4418,
+      "step": 626
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.354167867342193,
+      "learning_rate": 9.995179514669683e-06,
+      "loss": 0.4255,
+      "step": 627
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8665658701191628,
+      "learning_rate": 9.99512995507114e-06,
+      "loss": 0.2386,
+      "step": 628
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.24168865047806,
+      "learning_rate": 9.995080142135893e-06,
+      "loss": 0.4158,
+      "step": 629
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.279449671574902,
+      "learning_rate": 9.995030075866465e-06,
+      "loss": 0.426,
+      "step": 630
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.8337688494336306,
+      "learning_rate": 9.994979756265396e-06,
+      "loss": 0.4188,
+      "step": 631
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6989417502166497,
+      "learning_rate": 9.994929183335237e-06,
+      "loss": 0.4388,
+      "step": 632
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.9939551773928037,
+      "learning_rate": 9.994878357078556e-06,
+      "loss": 0.4499,
+      "step": 633
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 8.502823437287358,
+      "learning_rate": 9.99482727749793e-06,
+      "loss": 0.4715,
+      "step": 634
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.121220784221396,
+      "learning_rate": 9.994775944595947e-06,
+      "loss": 0.4458,
+      "step": 635
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.447197174697635,
+      "learning_rate": 9.994724358375212e-06,
+      "loss": 0.4129,
+      "step": 636
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.859633037470014,
+      "learning_rate": 9.99467251883834e-06,
+      "loss": 0.4229,
+      "step": 637
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8404291485198234,
+      "learning_rate": 9.994620425987964e-06,
+      "loss": 0.4239,
+      "step": 638
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 9.338708432492984,
+      "learning_rate": 9.994568079826721e-06,
+      "loss": 0.4713,
+      "step": 639
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.6680384352850903,
+      "learning_rate": 9.99451548035727e-06,
+      "loss": 0.4421,
+      "step": 640
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8186026589149895,
+      "learning_rate": 9.994462627582276e-06,
+      "loss": 0.4014,
+      "step": 641
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.287508457712858,
+      "learning_rate": 9.99440952150442e-06,
+      "loss": 0.4091,
+      "step": 642
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.6866456334056785,
+      "learning_rate": 9.994356162126397e-06,
+      "loss": 0.4695,
+      "step": 643
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 6.52074212722914,
+      "learning_rate": 9.994302549450912e-06,
+      "loss": 0.4187,
+      "step": 644
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.431487480815511,
+      "learning_rate": 9.994248683480683e-06,
+      "loss": 0.4345,
+      "step": 645
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.865754501325746,
+      "learning_rate": 9.994194564218445e-06,
+      "loss": 0.4255,
+      "step": 646
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 4.277927772522175,
+      "learning_rate": 9.99414019166694e-06,
+      "loss": 0.4467,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.000453554522621,
+      "learning_rate": 9.994085565828924e-06,
+      "loss": 0.4068,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.979615386643229,
+      "learning_rate": 9.994030686707171e-06,
+      "loss": 0.4803,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 26.983363553412246,
+      "learning_rate": 9.993975554304462e-06,
+      "loss": 0.4146,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.7183480802602418,
+      "learning_rate": 9.993920168623596e-06,
+      "loss": 0.4216,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7695479821725962,
+      "learning_rate": 9.99386452966738e-06,
+      "loss": 0.2116,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.159789585266918,
+      "learning_rate": 9.993808637438635e-06,
+      "loss": 0.4234,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.753700601617187,
+      "learning_rate": 9.993752491940198e-06,
+      "loss": 0.4594,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.208305637079963,
+      "learning_rate": 9.993696093174915e-06,
+      "loss": 0.4458,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.227650196030697,
+      "learning_rate": 9.993639441145645e-06,
+      "loss": 0.4304,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.13696418038436,
+      "learning_rate": 9.993582535855265e-06,
+      "loss": 0.4101,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.211248235229461,
+      "learning_rate": 9.993525377306657e-06,
+      "loss": 0.4396,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6337698394240996,
+      "learning_rate": 9.993467965502724e-06,
+      "loss": 0.4604,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.675801704764478,
+      "learning_rate": 9.993410300446372e-06,
+      "loss": 0.4474,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.83883973464124,
+      "learning_rate": 9.99335238214053e-06,
+      "loss": 0.4354,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.591426030745776,
+      "learning_rate": 9.993294210588135e-06,
+      "loss": 0.4213,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.9632097934186445,
+      "learning_rate": 9.993235785792138e-06,
+      "loss": 0.4165,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.320998413286234,
+      "learning_rate": 9.993177107755501e-06,
+      "loss": 0.4204,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.371199976227478,
+      "learning_rate": 9.993118176481199e-06,
+      "loss": 0.4512,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.242672358515438,
+      "learning_rate": 9.99305899197222e-06,
+      "loss": 0.3736,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.803886480019118,
+      "learning_rate": 9.992999554231568e-06,
+      "loss": 0.4203,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9921916270005107,
+      "learning_rate": 9.992939863262257e-06,
+      "loss": 0.4332,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.18441612224597,
+      "learning_rate": 9.992879919067314e-06,
+      "loss": 0.488,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.29320508758055,
+      "learning_rate": 9.99281972164978e-06,
+      "loss": 0.4343,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.92352043692322,
+      "learning_rate": 9.992759271012706e-06,
+      "loss": 0.4641,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8995353776003046,
+      "learning_rate": 9.99269856715916e-06,
+      "loss": 0.4558,
+      "step": 672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.922024144744522,
+      "learning_rate": 9.99263761009222e-06,
+      "loss": 0.4478,
+      "step": 673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.131207247022434,
+      "learning_rate": 9.992576399814976e-06,
+      "loss": 0.4007,
+      "step": 674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8674334456735835,
+      "learning_rate": 9.992514936330535e-06,
+      "loss": 0.4086,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7213528146586929,
+      "learning_rate": 9.992453219642012e-06,
+      "loss": 0.2049,
+      "step": 676
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.786819502767285,
+      "learning_rate": 9.992391249752538e-06,
+      "loss": 0.4339,
+      "step": 677
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3808769819267503,
+      "learning_rate": 9.992329026665258e-06,
+      "loss": 0.4042,
+      "step": 678
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.693112806314303,
+      "learning_rate": 9.992266550383323e-06,
+      "loss": 0.4375,
+      "step": 679
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5818543287825575,
+      "learning_rate": 9.992203820909906e-06,
+      "loss": 0.4447,
+      "step": 680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.253147188191536,
+      "learning_rate": 9.992140838248185e-06,
+      "loss": 0.4615,
+      "step": 681
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.994021657601774,
+      "learning_rate": 9.992077602401358e-06,
+      "loss": 0.4247,
+      "step": 682
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.41347828575026235,
+      "learning_rate": 9.99201411337263e-06,
+      "loss": 0.1807,
+      "step": 683
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.045456468794376,
+      "learning_rate": 9.99195037116522e-06,
+      "loss": 0.4247,
+      "step": 684
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 15.983282617616258,
+      "learning_rate": 9.991886375782362e-06,
+      "loss": 0.4311,
+      "step": 685
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0860603500024597,
+      "learning_rate": 9.991822127227301e-06,
+      "loss": 0.4535,
+      "step": 686
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.989522473438917,
+      "learning_rate": 9.991757625503298e-06,
+      "loss": 0.4419,
+      "step": 687
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.725001285913696,
+      "learning_rate": 9.99169287061362e-06,
+      "loss": 0.4014,
+      "step": 688
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.967278057363632,
+      "learning_rate": 9.991627862561555e-06,
+      "loss": 0.4608,
+      "step": 689
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5735422190487567,
+      "learning_rate": 9.9915626013504e-06,
+      "loss": 0.3894,
+      "step": 690
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.091407392829559,
+      "learning_rate": 9.99149708698346e-06,
+      "loss": 0.4349,
+      "step": 691
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0954980952291327,
+      "learning_rate": 9.991431319464062e-06,
+      "loss": 0.4198,
+      "step": 692
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.015432804152544,
+      "learning_rate": 9.99136529879554e-06,
+      "loss": 0.443,
+      "step": 693
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0369945634934563,
+      "learning_rate": 9.991299024981246e-06,
+      "loss": 0.4705,
+      "step": 694
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.124984783656927,
+      "learning_rate": 9.991232498024534e-06,
+      "loss": 0.4547,
+      "step": 695
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.573081281601767,
+      "learning_rate": 9.991165717928783e-06,
+      "loss": 0.4165,
+      "step": 696
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.2043178020593994,
+      "learning_rate": 9.99109868469738e-06,
+      "loss": 0.4336,
+      "step": 697
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 96.60821509465472,
+      "learning_rate": 9.991031398333725e-06,
+      "loss": 0.4351,
+      "step": 698
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.731951578128704,
+      "learning_rate": 9.990963858841225e-06,
+      "loss": 0.4694,
+      "step": 699
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.414378107286052,
+      "learning_rate": 9.990896066223313e-06,
+      "loss": 0.4155,
+      "step": 700
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.405995871812458,
+      "learning_rate": 9.990828020483423e-06,
+      "loss": 0.4178,
+      "step": 701
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.991826496743744,
+      "learning_rate": 9.990759721625005e-06,
+      "loss": 0.4714,
+      "step": 702
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.697256649360404,
+      "learning_rate": 9.990691169651527e-06,
+      "loss": 0.4066,
+      "step": 703
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.7589370860569127,
+      "learning_rate": 9.990622364566463e-06,
+      "loss": 0.4072,
+      "step": 704
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6197094178244231,
+      "learning_rate": 9.990553306373303e-06,
+      "loss": 0.1903,
+      "step": 705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3165095518090966,
+      "learning_rate": 9.990483995075549e-06,
+      "loss": 0.4422,
+      "step": 706
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.336880373852064,
+      "learning_rate": 9.990414430676716e-06,
+      "loss": 0.4659,
+      "step": 707
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9052298354220802,
+      "learning_rate": 9.990344613180334e-06,
+      "loss": 0.404,
+      "step": 708
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3501984936189286,
+      "learning_rate": 9.990274542589942e-06,
+      "loss": 0.181,
+      "step": 709
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.717497615295185,
+      "learning_rate": 9.990204218909095e-06,
+      "loss": 0.3777,
+      "step": 710
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3737384723791938,
+      "learning_rate": 9.990133642141359e-06,
+      "loss": 0.1731,
+      "step": 711
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3750849645823806,
+      "learning_rate": 9.990062812290313e-06,
+      "loss": 0.1749,
+      "step": 712
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8449994529344758,
+      "learning_rate": 9.989991729359549e-06,
+      "loss": 0.42,
+      "step": 713
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.43623347513008,
+      "learning_rate": 9.989920393352675e-06,
+      "loss": 0.4285,
+      "step": 714
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.060890182262364,
+      "learning_rate": 9.989848804273306e-06,
+      "loss": 0.472,
+      "step": 715
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 13.971489786230688,
+      "learning_rate": 9.989776962125072e-06,
+      "loss": 0.4474,
+      "step": 716
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.814868742052839,
+      "learning_rate": 9.989704866911617e-06,
+      "loss": 0.4211,
+      "step": 717
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.634389615049348,
+      "learning_rate": 9.989632518636602e-06,
+      "loss": 0.4254,
+      "step": 718
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8320099125088154,
+      "learning_rate": 9.989559917303691e-06,
+      "loss": 0.4303,
+      "step": 719
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3271585015379004,
+      "learning_rate": 9.989487062916568e-06,
+      "loss": 0.429,
+      "step": 720
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.234623022136248,
+      "learning_rate": 9.989413955478928e-06,
+      "loss": 0.4158,
+      "step": 721
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6539755141699196,
+      "learning_rate": 9.98934059499448e-06,
+      "loss": 0.2104,
+      "step": 722
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.335071716573553,
+      "learning_rate": 9.98926698146694e-06,
+      "loss": 0.431,
+      "step": 723
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4524938524436046,
+      "learning_rate": 9.989193114900048e-06,
+      "loss": 0.4181,
+      "step": 724
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.058618093036001,
+      "learning_rate": 9.989118995297545e-06,
+      "loss": 0.4075,
+      "step": 725
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.220882001121576,
+      "learning_rate": 9.989044622663194e-06,
+      "loss": 0.4537,
+      "step": 726
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0433186401727883,
+      "learning_rate": 9.988969997000761e-06,
+      "loss": 0.4967,
+      "step": 727
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9694835973133262,
+      "learning_rate": 9.988895118314038e-06,
+      "loss": 0.4466,
+      "step": 728
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47071770517974065,
+      "learning_rate": 9.98881998660682e-06,
+      "loss": 0.2053,
+      "step": 729
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9815632637717466,
+      "learning_rate": 9.988744601882913e-06,
+      "loss": 0.4414,
+      "step": 730
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.899296160273161,
+      "learning_rate": 9.988668964146147e-06,
+      "loss": 0.4287,
+      "step": 731
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.3503402835180476,
+      "learning_rate": 9.988593073400354e-06,
+      "loss": 0.1799,
+      "step": 732
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.561127491165165,
+      "learning_rate": 9.988516929649386e-06,
+      "loss": 0.4291,
+      "step": 733
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.076799594725559,
+      "learning_rate": 9.9884405328971e-06,
+      "loss": 0.4432,
+      "step": 734
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.206670080300507,
+      "learning_rate": 9.988363883147376e-06,
+      "loss": 0.4851,
+      "step": 735
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.38210906057920074,
+      "learning_rate": 9.988286980404096e-06,
+      "loss": 0.1968,
+      "step": 736
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.159839720585042,
+      "learning_rate": 9.988209824671164e-06,
+      "loss": 0.4468,
+      "step": 737
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.572744003968207,
+      "learning_rate": 9.988132415952494e-06,
+      "loss": 0.4548,
+      "step": 738
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.798431435586214,
+      "learning_rate": 9.988054754252009e-06,
+      "loss": 0.4426,
+      "step": 739
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 7.02316896527487,
+      "learning_rate": 9.987976839573649e-06,
+      "loss": 0.3986,
+      "step": 740
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.5359568211843624,
+      "learning_rate": 9.987898671921365e-06,
+      "loss": 0.4133,
+      "step": 741
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8090794345813634,
+      "learning_rate": 9.987820251299121e-06,
+      "loss": 0.4153,
+      "step": 742
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.2531753860783414,
+      "learning_rate": 9.987741577710897e-06,
+      "loss": 0.431,
+      "step": 743
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3105552527919273,
+      "learning_rate": 9.987662651160682e-06,
+      "loss": 0.4054,
+      "step": 744
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9898068390071644,
+      "learning_rate": 9.987583471652476e-06,
+      "loss": 0.4373,
+      "step": 745
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.261030975097089,
+      "learning_rate": 9.987504039190297e-06,
+      "loss": 0.4518,
+      "step": 746
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.653210450696857,
+      "learning_rate": 9.987424353778172e-06,
+      "loss": 0.4183,
+      "step": 747
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.4869778179062503,
+      "learning_rate": 9.987344415420145e-06,
+      "loss": 0.4609,
+      "step": 748
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.391218815621294,
+      "learning_rate": 9.98726422412027e-06,
+      "loss": 0.3976,
+      "step": 749
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.560476242494493,
+      "learning_rate": 9.987183779882611e-06,
+      "loss": 0.4139,
+      "step": 750
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.180268393439651,
+      "learning_rate": 9.98710308271125e-06,
+      "loss": 0.4045,
+      "step": 751
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.114403042819518,
+      "learning_rate": 9.987022132610282e-06,
+      "loss": 0.4045,
+      "step": 752
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.122146567597501,
+      "learning_rate": 9.986940929583808e-06,
+      "loss": 0.4168,
+      "step": 753
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.875453078710568,
+      "learning_rate": 9.98685947363595e-06,
+      "loss": 0.4487,
+      "step": 754
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.993036952452852,
+      "learning_rate": 9.986777764770836e-06,
+      "loss": 0.4024,
+      "step": 755
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.104206548055548,
+      "learning_rate": 9.986695802992613e-06,
+      "loss": 0.4621,
+      "step": 756
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6937357740330894,
+      "learning_rate": 9.986613588305435e-06,
+      "loss": 0.3857,
+      "step": 757
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1736964602207687,
+      "learning_rate": 9.986531120713474e-06,
+      "loss": 0.4286,
+      "step": 758
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1531571181508173,
+      "learning_rate": 9.986448400220912e-06,
+      "loss": 0.3731,
+      "step": 759
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.860378463193414,
+      "learning_rate": 9.986365426831943e-06,
+      "loss": 0.4125,
+      "step": 760
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3291620753888687,
+      "learning_rate": 9.986282200550775e-06,
+      "loss": 0.4148,
+      "step": 761
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.611240907772357,
+      "learning_rate": 9.98619872138163e-06,
+      "loss": 0.3806,
+      "step": 762
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9869485918742193,
+      "learning_rate": 9.986114989328745e-06,
+      "loss": 0.4391,
+      "step": 763
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.111707205305628,
+      "learning_rate": 9.986031004396362e-06,
+      "loss": 0.4157,
+      "step": 764
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.3345480356930426,
+      "learning_rate": 9.98594676658874e-06,
+      "loss": 0.4661,
+      "step": 765
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.964982622258637,
+      "learning_rate": 9.985862275910154e-06,
+      "loss": 0.4506,
+      "step": 766
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5973152057688806,
+      "learning_rate": 9.985777532364889e-06,
+      "loss": 0.3968,
+      "step": 767
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.278162974905271,
+      "learning_rate": 9.985692535957243e-06,
+      "loss": 0.4186,
+      "step": 768
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.69617552051413,
+      "learning_rate": 9.985607286691524e-06,
+      "loss": 0.4364,
+      "step": 769
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.116502889629469,
+      "learning_rate": 9.985521784572057e-06,
+      "loss": 0.4485,
+      "step": 770
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 12.028750394959877,
+      "learning_rate": 9.985436029603181e-06,
+      "loss": 0.4317,
+      "step": 771
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6090953638384504,
+      "learning_rate": 9.985350021789242e-06,
+      "loss": 0.4783,
+      "step": 772
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.237780340385099,
+      "learning_rate": 9.985263761134602e-06,
+      "loss": 0.4381,
+      "step": 773
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.09523968493673,
+      "learning_rate": 9.985177247643637e-06,
+      "loss": 0.4253,
+      "step": 774
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.361428700688823,
+      "learning_rate": 9.985090481320736e-06,
+      "loss": 0.4381,
+      "step": 775
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.533242058256771,
+      "learning_rate": 9.985003462170297e-06,
+      "loss": 0.4533,
+      "step": 776
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.296169870406036,
+      "learning_rate": 9.984916190196736e-06,
+      "loss": 0.4428,
+      "step": 777
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 9.879459925006454,
+      "learning_rate": 9.984828665404477e-06,
+      "loss": 0.4615,
+      "step": 778
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.528092504207168,
+      "learning_rate": 9.984740887797958e-06,
+      "loss": 0.4565,
+      "step": 779
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.9965667731735923,
+      "learning_rate": 9.984652857381633e-06,
+      "loss": 0.4176,
+      "step": 780
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9881324629636428,
+      "learning_rate": 9.984564574159968e-06,
+      "loss": 0.4394,
+      "step": 781
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4967343823746737,
+      "learning_rate": 9.984476038137437e-06,
+      "loss": 0.4421,
+      "step": 782
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.1684352821134047,
+      "learning_rate": 9.98438724931853e-06,
+      "loss": 0.4389,
+      "step": 783
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.642064040438759,
+      "learning_rate": 9.984298207707754e-06,
+      "loss": 0.4131,
+      "step": 784
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 4.616670542897823,
+      "learning_rate": 9.984208913309623e-06,
+      "loss": 0.4479,
+      "step": 785
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.462432446505075,
+      "learning_rate": 9.984119366128663e-06,
+      "loss": 0.1953,
+      "step": 786
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 6.059786883305137,
+      "learning_rate": 9.98402956616942e-06,
+      "loss": 0.4657,
+      "step": 787
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.6165464018313993,
+      "learning_rate": 9.983939513436447e-06,
+      "loss": 0.4231,
+      "step": 788
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 8.413549178267427,
+      "learning_rate": 9.983849207934308e-06,
+      "loss": 0.4123,
+      "step": 789
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 5.048882737792652,
+      "learning_rate": 9.983758649667587e-06,
+      "loss": 0.4961,
+      "step": 790
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 11.925092419239943,
+      "learning_rate": 9.983667838640874e-06,
+      "loss": 0.4351,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8682665366165905,
+      "learning_rate": 9.983576774858776e-06,
+      "loss": 0.464,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9484610846331303,
+      "learning_rate": 9.983485458325913e-06,
+      "loss": 0.4352,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.079503619358625,
+      "learning_rate": 9.983393889046914e-06,
+      "loss": 0.4378,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.925655663958739,
+      "learning_rate": 9.983302067026424e-06,
+      "loss": 0.4517,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.968821169414754,
+      "learning_rate": 9.983209992269098e-06,
+      "loss": 0.4148,
+      "step": 796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.580580642696761,
+      "learning_rate": 9.98311766477961e-06,
+      "loss": 0.4468,
+      "step": 797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.6451499435251,
+      "learning_rate": 9.983025084562637e-06,
+      "loss": 0.4441,
+      "step": 798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6166500698972333,
+      "learning_rate": 9.982932251622878e-06,
+      "loss": 0.4698,
+      "step": 799
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8183649278110146,
+      "learning_rate": 9.982839165965042e-06,
+      "loss": 0.4333,
+      "step": 800
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.49484191154582147,
+      "learning_rate": 9.982745827593848e-06,
+      "loss": 0.203,
+      "step": 801
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.391164823332556,
+      "learning_rate": 9.98265223651403e-06,
+      "loss": 0.4411,
+      "step": 802
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2584227531981114,
+      "learning_rate": 9.982558392730334e-06,
+      "loss": 0.447,
+      "step": 803
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 18.186929615161915,
+      "learning_rate": 9.982464296247523e-06,
+      "loss": 0.4819,
+      "step": 804
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.936850161959208,
+      "learning_rate": 9.982369947070365e-06,
+      "loss": 0.3993,
+      "step": 805
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.4970008728724453,
+      "learning_rate": 9.982275345203647e-06,
+      "loss": 0.4131,
+      "step": 806
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.327801061823603,
+      "learning_rate": 9.982180490652165e-06,
+      "loss": 0.4423,
+      "step": 807
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.734121425371093,
+      "learning_rate": 9.982085383420734e-06,
+      "loss": 0.4446,
+      "step": 808
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7910394436074357,
+      "learning_rate": 9.981990023514173e-06,
+      "loss": 0.3787,
+      "step": 809
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.518556006322825,
+      "learning_rate": 9.98189441093732e-06,
+      "loss": 0.406,
+      "step": 810
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4513347607676222,
+      "learning_rate": 9.981798545695026e-06,
+      "loss": 0.1788,
+      "step": 811
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.333896755319483,
+      "learning_rate": 9.98170242779215e-06,
+      "loss": 0.4515,
+      "step": 812
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 14.544978620141507,
+      "learning_rate": 9.98160605723357e-06,
+      "loss": 0.4397,
+      "step": 813
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5009996549968765,
+      "learning_rate": 9.98150943402417e-06,
+      "loss": 0.3969,
+      "step": 814
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7270447782262695,
+      "learning_rate": 9.981412558168853e-06,
+      "loss": 0.4221,
+      "step": 815
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.212473178206575,
+      "learning_rate": 9.981315429672531e-06,
+      "loss": 0.4084,
+      "step": 816
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7075338233436947,
+      "learning_rate": 9.98121804854013e-06,
+      "loss": 0.4554,
+      "step": 817
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3794986189141595,
+      "learning_rate": 9.981120414776591e-06,
+      "loss": 0.188,
+      "step": 818
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2283961836580537,
+      "learning_rate": 9.981022528386862e-06,
+      "loss": 0.3983,
+      "step": 819
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2527952870368395,
+      "learning_rate": 9.98092438937591e-06,
+      "loss": 0.4253,
+      "step": 820
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7457077624176693,
+      "learning_rate": 9.980825997748712e-06,
+      "loss": 0.444,
+      "step": 821
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.220948389916808,
+      "learning_rate": 9.980727353510257e-06,
+      "loss": 0.4192,
+      "step": 822
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.323081668024819,
+      "learning_rate": 9.98062845666555e-06,
+      "loss": 0.4331,
+      "step": 823
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.624621376573992,
+      "learning_rate": 9.980529307219605e-06,
+      "loss": 0.4482,
+      "step": 824
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.824875501379335,
+      "learning_rate": 9.98042990517745e-06,
+      "loss": 0.4801,
+      "step": 825
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.006855776021493,
+      "learning_rate": 9.98033025054413e-06,
+      "loss": 0.4404,
+      "step": 826
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.8833328365360513,
+      "learning_rate": 9.980230343324695e-06,
+      "loss": 0.445,
+      "step": 827
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.275466431049486,
+      "learning_rate": 9.980130183524213e-06,
+      "loss": 0.4253,
+      "step": 828
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.239312806132201,
+      "learning_rate": 9.980029771147764e-06,
+      "loss": 0.4642,
+      "step": 829
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.744655122668165,
+      "learning_rate": 9.97992910620044e-06,
+      "loss": 0.4613,
+      "step": 830
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.427778405691158,
+      "learning_rate": 9.97982818868735e-06,
+      "loss": 0.4507,
+      "step": 831
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.182987572966574,
+      "learning_rate": 9.979727018613607e-06,
+      "loss": 0.4653,
+      "step": 832
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.433539613628279,
+      "learning_rate": 9.979625595984345e-06,
+      "loss": 0.4054,
+      "step": 833
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.077450741571009,
+      "learning_rate": 9.979523920804709e-06,
+      "loss": 0.4237,
+      "step": 834
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.249589384192771,
+      "learning_rate": 9.979421993079853e-06,
+      "loss": 0.4345,
+      "step": 835
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.40522970595977315,
+      "learning_rate": 9.979319812814947e-06,
+      "loss": 0.172,
+      "step": 836
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.577962880238718,
+      "learning_rate": 9.979217380015173e-06,
+      "loss": 0.4456,
+      "step": 837
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1247921552713076,
+      "learning_rate": 9.979114694685727e-06,
+      "loss": 0.4265,
+      "step": 838
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.168060972125359,
+      "learning_rate": 9.979011756831818e-06,
+      "loss": 0.4401,
+      "step": 839
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9378855664809396,
+      "learning_rate": 9.978908566458663e-06,
+      "loss": 0.401,
+      "step": 840
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0769147947968714,
+      "learning_rate": 9.9788051235715e-06,
+      "loss": 0.4682,
+      "step": 841
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37545937739110763,
+      "learning_rate": 9.97870142817557e-06,
+      "loss": 0.1811,
+      "step": 842
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.82028874983575,
+      "learning_rate": 9.978597480276138e-06,
+      "loss": 0.4269,
+      "step": 843
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.881254131117299,
+      "learning_rate": 9.978493279878472e-06,
+      "loss": 0.4954,
+      "step": 844
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6963899048792426,
+      "learning_rate": 9.978388826987856e-06,
+      "loss": 0.4389,
+      "step": 845
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5451111119606424,
+      "learning_rate": 9.978284121609591e-06,
+      "loss": 0.4583,
+      "step": 846
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.835299896064639,
+      "learning_rate": 9.978179163748985e-06,
+      "loss": 0.4356,
+      "step": 847
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7425873958049465,
+      "learning_rate": 9.978073953411362e-06,
+      "loss": 0.4365,
+      "step": 848
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.796992814596756,
+      "learning_rate": 9.977968490602056e-06,
+      "loss": 0.4451,
+      "step": 849
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.676741528314537,
+      "learning_rate": 9.977862775326419e-06,
+      "loss": 0.4023,
+      "step": 850
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.33903465643817,
+      "learning_rate": 9.977756807589812e-06,
+      "loss": 0.443,
+      "step": 851
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1597840887352397,
+      "learning_rate": 9.977650587397606e-06,
+      "loss": 0.4378,
+      "step": 852
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0937251095700784,
+      "learning_rate": 9.97754411475519e-06,
+      "loss": 0.42,
+      "step": 853
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1585767109435934,
+      "learning_rate": 9.977437389667966e-06,
+      "loss": 0.4394,
+      "step": 854
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.628686734390035,
+      "learning_rate": 9.977330412141343e-06,
+      "loss": 0.4421,
+      "step": 855
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.775332733891346,
+      "learning_rate": 9.977223182180749e-06,
+      "loss": 0.4385,
+      "step": 856
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0855744951514854,
+      "learning_rate": 9.977115699791622e-06,
+      "loss": 0.4042,
+      "step": 857
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.390001860682499,
+      "learning_rate": 9.977007964979414e-06,
+      "loss": 0.4246,
+      "step": 858
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1983718818001137,
+      "learning_rate": 9.976899977749587e-06,
+      "loss": 0.4556,
+      "step": 859
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7091540936402483,
+      "learning_rate": 9.976791738107618e-06,
+      "loss": 0.4604,
+      "step": 860
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.566813430990463,
+      "learning_rate": 9.976683246059e-06,
+      "loss": 0.4441,
+      "step": 861
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.566464250643466,
+      "learning_rate": 9.976574501609231e-06,
+      "loss": 0.4087,
+      "step": 862
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.13908773964016,
+      "learning_rate": 9.976465504763828e-06,
+      "loss": 0.4535,
+      "step": 863
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.952815037143065,
+      "learning_rate": 9.976356255528318e-06,
+      "loss": 0.4024,
+      "step": 864
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5591291483203764,
+      "learning_rate": 9.976246753908243e-06,
+      "loss": 0.4504,
+      "step": 865
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.5084751162194094,
+      "learning_rate": 9.976136999909156e-06,
+      "loss": 0.4492,
+      "step": 866
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.6219542664417075,
+      "learning_rate": 9.976026993536625e-06,
+      "loss": 0.4602,
+      "step": 867
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 9.222294710479785,
+      "learning_rate": 9.975916734796226e-06,
+      "loss": 0.398,
+      "step": 868
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.294876809580078,
+      "learning_rate": 9.975806223693554e-06,
+      "loss": 0.4002,
+      "step": 869
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.988487784865998,
+      "learning_rate": 9.975695460234211e-06,
+      "loss": 0.4379,
+      "step": 870
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.395556043461477,
+      "learning_rate": 9.975584444423816e-06,
+      "loss": 0.4564,
+      "step": 871
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.37624466903822446,
+      "learning_rate": 9.975473176268001e-06,
+      "loss": 0.1745,
+      "step": 872
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.410302230374194,
+      "learning_rate": 9.975361655772406e-06,
+      "loss": 0.4131,
+      "step": 873
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.732833855210534,
+      "learning_rate": 9.975249882942688e-06,
+      "loss": 0.464,
+      "step": 874
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 7.5974859942824775,
+      "learning_rate": 9.975137857784518e-06,
+      "loss": 0.4601,
+      "step": 875
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.112082438283715,
+      "learning_rate": 9.975025580303575e-06,
+      "loss": 0.3994,
+      "step": 876
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.343997989408455,
+      "learning_rate": 9.974913050505553e-06,
+      "loss": 0.4439,
+      "step": 877
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.36167139935516057,
+      "learning_rate": 9.97480026839616e-06,
+      "loss": 0.1827,
+      "step": 878
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.7670218540498217,
+      "learning_rate": 9.97468723398112e-06,
+      "loss": 0.4439,
+      "step": 879
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1760270370789474,
+      "learning_rate": 9.974573947266158e-06,
+      "loss": 0.4605,
+      "step": 880
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.289994554046978,
+      "learning_rate": 9.974460408257024e-06,
+      "loss": 0.4369,
+      "step": 881
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.298675500685065,
+      "learning_rate": 9.974346616959476e-06,
+      "loss": 0.4204,
+      "step": 882
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.703408265531322,
+      "learning_rate": 9.974232573379285e-06,
+      "loss": 0.3888,
+      "step": 883
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.7178884781876564,
+      "learning_rate": 9.974118277522235e-06,
+      "loss": 0.4418,
+      "step": 884
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8834384756708538,
+      "learning_rate": 9.974003729394124e-06,
+      "loss": 0.4182,
+      "step": 885
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4278415114097722,
+      "learning_rate": 9.973888929000758e-06,
+      "loss": 0.4573,
+      "step": 886
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.357082861743962,
+      "learning_rate": 9.973773876347962e-06,
+      "loss": 0.4384,
+      "step": 887
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.424877259465171,
+      "learning_rate": 9.97365857144157e-06,
+      "loss": 0.4449,
+      "step": 888
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.37454431474936,
+      "learning_rate": 9.97354301428743e-06,
+      "loss": 0.4472,
+      "step": 889
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.388276341007746,
+      "learning_rate": 9.973427204891404e-06,
+      "loss": 0.1918,
+      "step": 890
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.550566799072139,
+      "learning_rate": 9.973311143259365e-06,
+      "loss": 0.3939,
+      "step": 891
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.3717447258948247,
+      "learning_rate": 9.973194829397198e-06,
+      "loss": 0.1823,
+      "step": 892
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.194709447351789,
+      "learning_rate": 9.973078263310803e-06,
+      "loss": 0.4069,
+      "step": 893
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.735003652355403,
+      "learning_rate": 9.97296144500609e-06,
+      "loss": 0.4391,
+      "step": 894
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.006563903093048,
+      "learning_rate": 9.972844374488986e-06,
+      "loss": 0.5087,
+      "step": 895
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.423442854505595,
+      "learning_rate": 9.97272705176543e-06,
+      "loss": 0.4218,
+      "step": 896
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.8035192019582,
+      "learning_rate": 9.972609476841368e-06,
+      "loss": 0.4276,
+      "step": 897
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9942823033503605,
+      "learning_rate": 9.972491649722763e-06,
+      "loss": 0.4504,
+      "step": 898
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0150890481917245,
+      "learning_rate": 9.972373570415596e-06,
+      "loss": 0.3879,
+      "step": 899
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5174823043176358,
+      "learning_rate": 9.972255238925853e-06,
+      "loss": 0.3931,
+      "step": 900
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.334015287408728,
+      "learning_rate": 9.972136655259532e-06,
+      "loss": 0.4375,
+      "step": 901
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.455519947588375,
+      "learning_rate": 9.972017819422652e-06,
+      "loss": 0.4782,
+      "step": 902
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 6.7235566860136275,
+      "learning_rate": 9.971898731421237e-06,
+      "loss": 0.4198,
+      "step": 903
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6685654480765835,
+      "learning_rate": 9.971779391261328e-06,
+      "loss": 0.4373,
+      "step": 904
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.778502871930368,
+      "learning_rate": 9.971659798948977e-06,
+      "loss": 0.4767,
+      "step": 905
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.3837594305548677,
+      "learning_rate": 9.971539954490248e-06,
+      "loss": 0.3711,
+      "step": 906
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.806610924148066,
+      "learning_rate": 9.971419857891223e-06,
+      "loss": 0.4367,
+      "step": 907
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.567113569630713,
+      "learning_rate": 9.97129950915799e-06,
+      "loss": 0.4456,
+      "step": 908
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.580009558362219,
+      "learning_rate": 9.971178908296652e-06,
+      "loss": 0.3956,
+      "step": 909
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.928280860576343,
+      "learning_rate": 9.971058055313327e-06,
+      "loss": 0.4325,
+      "step": 910
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.91054217902358,
+      "learning_rate": 9.970936950214145e-06,
+      "loss": 0.4137,
+      "step": 911
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9750045196354744,
+      "learning_rate": 9.970815593005248e-06,
+      "loss": 0.4502,
+      "step": 912
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.1799565156455825,
+      "learning_rate": 9.970693983692787e-06,
+      "loss": 0.4354,
+      "step": 913
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.620172339091545,
+      "learning_rate": 9.970572122282935e-06,
+      "loss": 0.4473,
+      "step": 914
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 5.6825153805696855,
+      "learning_rate": 9.970450008781869e-06,
+      "loss": 0.4237,
+      "step": 915
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.9656166024133217,
+      "learning_rate": 9.970327643195784e-06,
+      "loss": 0.4399,
+      "step": 916
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.028258415004872,
+      "learning_rate": 9.970205025530884e-06,
+      "loss": 0.4387,
+      "step": 917
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.214998968248516,
+      "learning_rate": 9.97008215579339e-06,
+      "loss": 0.4441,
+      "step": 918
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.9128106996782845,
+      "learning_rate": 9.969959033989532e-06,
+      "loss": 0.4711,
+      "step": 919
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0493745926186713,
+      "learning_rate": 9.969835660125554e-06,
+      "loss": 0.407,
+      "step": 920
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.2129731818904617,
+      "learning_rate": 9.969712034207715e-06,
+      "loss": 0.4607,
+      "step": 921
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.307804782565934,
+      "learning_rate": 9.969588156242282e-06,
+      "loss": 0.4552,
+      "step": 922
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0435267860467468,
+      "learning_rate": 9.969464026235543e-06,
+      "loss": 0.4418,
+      "step": 923
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 4.185032958091327,
+      "learning_rate": 9.969339644193787e-06,
+      "loss": 0.4448,
+      "step": 924
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.025917074022424,
+      "learning_rate": 9.969215010123325e-06,
+      "loss": 0.4483,
+      "step": 925
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 13.769929631514204,
+      "learning_rate": 9.96909012403048e-06,
+      "loss": 0.4177,
+      "step": 926
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.695544746610935,
+      "learning_rate": 9.968964985921584e-06,
+      "loss": 0.405,
+      "step": 927
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.0254448797314404,
+      "learning_rate": 9.968839595802982e-06,
+      "loss": 0.4587,
+      "step": 928
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.080358976286633,
+      "learning_rate": 9.968713953681037e-06,
+      "loss": 0.4483,
+      "step": 929
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.034181004501432,
+      "learning_rate": 9.968588059562119e-06,
+      "loss": 0.4468,
+      "step": 930
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.95598406673221,
+      "learning_rate": 9.968461913452612e-06,
+      "loss": 0.4091,
+      "step": 931
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.6802367522981196,
+      "learning_rate": 9.968335515358916e-06,
+      "loss": 0.4569,
+      "step": 932
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.05893103859969,
+      "learning_rate": 9.968208865287438e-06,
+      "loss": 0.4268,
+      "step": 933
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 8.494249475695591,
+      "learning_rate": 9.968081963244606e-06,
+      "loss": 0.4177,
+      "step": 934
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.760342136220472,
+      "learning_rate": 9.967954809236852e-06,
+      "loss": 0.4756,
+      "step": 935
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.612496219875474,
+      "learning_rate": 9.96782740327063e-06,
+      "loss": 0.4372,
+      "step": 936
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9269675889443536,
+      "learning_rate": 9.967699745352395e-06,
+      "loss": 0.4328,
+      "step": 937
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.3703901002014165,
+      "learning_rate": 9.967571835488625e-06,
+      "loss": 0.4239,
+      "step": 938
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.835515826211019,
+      "learning_rate": 9.967443673685806e-06,
+      "loss": 0.4671,
+      "step": 939
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6080022373048894,
+      "learning_rate": 9.96731525995044e-06,
+      "loss": 0.4498,
+      "step": 940
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.991200508317681,
+      "learning_rate": 9.967186594289038e-06,
+      "loss": 0.4374,
+      "step": 941
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.619159377879737,
+      "learning_rate": 9.967057676708126e-06,
+      "loss": 0.4365,
+      "step": 942
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.567396814718496,
+      "learning_rate": 9.966928507214241e-06,
+      "loss": 0.1848,
+      "step": 943
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1690165165774227,
+      "learning_rate": 9.96679908581394e-06,
+      "loss": 0.4155,
+      "step": 944
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.392849991951101,
+      "learning_rate": 9.966669412513776e-06,
+      "loss": 0.4107,
+      "step": 945
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.882748327343743,
+      "learning_rate": 9.966539487320335e-06,
+      "loss": 0.3969,
+      "step": 946
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3296570859639965,
+      "learning_rate": 9.9664093102402e-06,
+      "loss": 0.4661,
+      "step": 947
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.199756487133542,
+      "learning_rate": 9.96627888127998e-06,
+      "loss": 0.4189,
+      "step": 948
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.027963668973956,
+      "learning_rate": 9.966148200446284e-06,
+      "loss": 0.4234,
+      "step": 949
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.676099394458873,
+      "learning_rate": 9.966017267745743e-06,
+      "loss": 0.426,
+      "step": 950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.248100104996111,
+      "learning_rate": 9.965886083184996e-06,
+      "loss": 0.4053,
+      "step": 951
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6846759399697686,
+      "learning_rate": 9.965754646770695e-06,
+      "loss": 0.4115,
+      "step": 952
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7528109436064465,
+      "learning_rate": 9.96562295850951e-06,
+      "loss": 0.4288,
+      "step": 953
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.613963371220857,
+      "learning_rate": 9.965491018408115e-06,
+      "loss": 0.425,
+      "step": 954
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.2057955738299615,
+      "learning_rate": 9.965358826473206e-06,
+      "loss": 0.4426,
+      "step": 955
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.326978726558206,
+      "learning_rate": 9.965226382711485e-06,
+      "loss": 0.4794,
+      "step": 956
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.0103449711463135,
+      "learning_rate": 9.965093687129669e-06,
+      "loss": 0.4278,
+      "step": 957
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.880588148073968,
+      "learning_rate": 9.964960739734489e-06,
+      "loss": 0.4319,
+      "step": 958
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6894743091627528,
+      "learning_rate": 9.964827540532685e-06,
+      "loss": 0.215,
+      "step": 959
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3309041155262973,
+      "learning_rate": 9.964694089531017e-06,
+      "loss": 0.4354,
+      "step": 960
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.093724379920781,
+      "learning_rate": 9.964560386736248e-06,
+      "loss": 0.4299,
+      "step": 961
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 10.904833698336109,
+      "learning_rate": 9.964426432155163e-06,
+      "loss": 0.4531,
+      "step": 962
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.957562308019668,
+      "learning_rate": 9.964292225794556e-06,
+      "loss": 0.4141,
+      "step": 963
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.3825475676010295,
+      "learning_rate": 9.96415776766123e-06,
+      "loss": 0.171,
+      "step": 964
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.763204777909579,
+      "learning_rate": 9.964023057762005e-06,
+      "loss": 0.4558,
+      "step": 965
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.955874486062647,
+      "learning_rate": 9.963888096103716e-06,
+      "loss": 0.4943,
+      "step": 966
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4236258889156219,
+      "learning_rate": 9.963752882693205e-06,
+      "loss": 0.194,
+      "step": 967
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6500125975184554,
+      "learning_rate": 9.96361741753733e-06,
+      "loss": 0.4301,
+      "step": 968
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2822565739741343,
+      "learning_rate": 9.963481700642964e-06,
+      "loss": 0.4324,
+      "step": 969
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.286274086225351,
+      "learning_rate": 9.963345732016988e-06,
+      "loss": 0.4206,
+      "step": 970
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3881138665601704,
+      "learning_rate": 9.963209511666296e-06,
+      "loss": 0.4426,
+      "step": 971
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.401256073275571,
+      "learning_rate": 9.963073039597798e-06,
+      "loss": 0.424,
+      "step": 972
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.548996892830632,
+      "learning_rate": 9.962936315818417e-06,
+      "loss": 0.3884,
+      "step": 973
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2777486597124366,
+      "learning_rate": 9.962799340335088e-06,
+      "loss": 0.4281,
+      "step": 974
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5549533802198434,
+      "learning_rate": 9.962662113154753e-06,
+      "loss": 0.4579,
+      "step": 975
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.653361905846184,
+      "learning_rate": 9.962524634284377e-06,
+      "loss": 0.4197,
+      "step": 976
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.273846433301272,
+      "learning_rate": 9.96238690373093e-06,
+      "loss": 0.425,
+      "step": 977
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5080291193537128,
+      "learning_rate": 9.962248921501396e-06,
+      "loss": 0.4694,
+      "step": 978
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.686681132539784,
+      "learning_rate": 9.962110687602778e-06,
+      "loss": 0.4345,
+      "step": 979
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.981321248801702,
+      "learning_rate": 9.96197220204208e-06,
+      "loss": 0.442,
+      "step": 980
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.081450497579051,
+      "learning_rate": 9.96183346482633e-06,
+      "loss": 0.4543,
+      "step": 981
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.5630396031419176,
+      "learning_rate": 9.961694475962562e-06,
+      "loss": 0.4273,
+      "step": 982
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.307416813328998,
+      "learning_rate": 9.961555235457829e-06,
+      "loss": 0.4358,
+      "step": 983
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.868744890909495,
+      "learning_rate": 9.961415743319187e-06,
+      "loss": 0.4449,
+      "step": 984
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7302354014268078,
+      "learning_rate": 9.961275999553717e-06,
+      "loss": 0.4709,
+      "step": 985
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9773383586329865,
+      "learning_rate": 9.9611360041685e-06,
+      "loss": 0.4105,
+      "step": 986
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9737507659216402,
+      "learning_rate": 9.960995757170639e-06,
+      "loss": 0.4086,
+      "step": 987
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.304495726463494,
+      "learning_rate": 9.96085525856725e-06,
+      "loss": 0.4571,
+      "step": 988
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7061684586343624,
+      "learning_rate": 9.960714508365454e-06,
+      "loss": 0.4145,
+      "step": 989
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3893993443197656,
+      "learning_rate": 9.960573506572391e-06,
+      "loss": 0.4362,
+      "step": 990
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.982193940254108,
+      "learning_rate": 9.960432253195212e-06,
+      "loss": 0.4288,
+      "step": 991
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.295093041759435,
+      "learning_rate": 9.96029074824108e-06,
+      "loss": 0.4488,
+      "step": 992
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.110496338514903,
+      "learning_rate": 9.960148991717174e-06,
+      "loss": 0.458,
+      "step": 993
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.461435590370466,
+      "learning_rate": 9.96000698363068e-06,
+      "loss": 0.4356,
+      "step": 994
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3588919809230675,
+      "learning_rate": 9.959864723988805e-06,
+      "loss": 0.4236,
+      "step": 995
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6385015326697312,
+      "learning_rate": 9.959722212798762e-06,
+      "loss": 0.206,
+      "step": 996
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.751317769360924,
+      "learning_rate": 9.959579450067776e-06,
+      "loss": 0.4642,
+      "step": 997
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0277548241647003,
+      "learning_rate": 9.95943643580309e-06,
+      "loss": 0.4031,
+      "step": 998
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.926422987564575,
+      "learning_rate": 9.959293170011956e-06,
+      "loss": 0.4371,
+      "step": 999
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.053250105449407,
+      "learning_rate": 9.95914965270164e-06,
+      "loss": 0.426,
+      "step": 1000
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.105187554697718,
+      "learning_rate": 9.959005883879425e-06,
+      "loss": 0.4631,
+      "step": 1001
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4530096916356534,
+      "learning_rate": 9.958861863552596e-06,
+      "loss": 0.442,
+      "step": 1002
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8959907934464297,
+      "learning_rate": 9.95871759172846e-06,
+      "loss": 0.4406,
+      "step": 1003
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5431736429498706,
+      "learning_rate": 9.958573068414334e-06,
+      "loss": 0.4012,
+      "step": 1004
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.377780598570852,
+      "learning_rate": 9.95842829361755e-06,
+      "loss": 0.45,
+      "step": 1005
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 8.756992903872982,
+      "learning_rate": 9.958283267345446e-06,
+      "loss": 0.3921,
+      "step": 1006
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.086511455648567,
+      "learning_rate": 9.95813798960538e-06,
+      "loss": 0.4173,
+      "step": 1007
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.62747936880494,
+      "learning_rate": 9.957992460404721e-06,
+      "loss": 0.4232,
+      "step": 1008
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.716850648514961,
+      "learning_rate": 9.957846679750846e-06,
+      "loss": 0.4459,
+      "step": 1009
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6976038424089195,
+      "learning_rate": 9.957700647651154e-06,
+      "loss": 0.2099,
+      "step": 1010
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.598563741320774,
+      "learning_rate": 9.957554364113047e-06,
+      "loss": 0.4938,
+      "step": 1011
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.2910913454509245,
+      "learning_rate": 9.957407829143943e-06,
+      "loss": 0.4526,
+      "step": 1012
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.336467088329202,
+      "learning_rate": 9.957261042751278e-06,
+      "loss": 0.445,
+      "step": 1013
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 15.573603460432617,
+      "learning_rate": 9.957114004942495e-06,
+      "loss": 0.4176,
+      "step": 1014
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.379417664994578,
+      "learning_rate": 9.95696671572505e-06,
+      "loss": 0.4379,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.5572878883944705,
+      "learning_rate": 9.956819175106414e-06,
+      "loss": 0.4055,
+      "step": 1016
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.3553663908527214,
+      "learning_rate": 9.95667138309407e-06,
+      "loss": 0.4494,
+      "step": 1017
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.534180638809289,
+      "learning_rate": 9.956523339695513e-06,
+      "loss": 0.3894,
+      "step": 1018
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.365951147209446,
+      "learning_rate": 9.956375044918249e-06,
+      "loss": 0.4226,
+      "step": 1019
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.573375279307627,
+      "learning_rate": 9.956226498769805e-06,
+      "loss": 0.4321,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.354938408372415,
+      "learning_rate": 9.95607770125771e-06,
+      "loss": 0.424,
+      "step": 1021
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7215870713492025,
+      "learning_rate": 9.95592865238951e-06,
+      "loss": 0.442,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.219564923076984,
+      "learning_rate": 9.955779352172767e-06,
+      "loss": 0.4141,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.717860222142295,
+      "learning_rate": 9.955629800615055e-06,
+      "loss": 0.3819,
+      "step": 1024
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6545366550664493,
+      "learning_rate": 9.955479997723951e-06,
+      "loss": 0.4102,
+      "step": 1025
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6601579541139335,
+      "learning_rate": 9.955329943507061e-06,
+      "loss": 0.2207,
+      "step": 1026
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 9.318324383118442,
+      "learning_rate": 9.95517963797199e-06,
+      "loss": 0.3927,
+      "step": 1027
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.429242938317406,
+      "learning_rate": 9.955029081126362e-06,
+      "loss": 0.417,
+      "step": 1028
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9942324464981427,
+      "learning_rate": 9.954878272977814e-06,
+      "loss": 0.4543,
+      "step": 1029
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.604465399987888,
+      "learning_rate": 9.954727213533994e-06,
+      "loss": 0.4316,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5471461345014337,
+      "learning_rate": 9.954575902802563e-06,
+      "loss": 0.4384,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.634927961881907,
+      "learning_rate": 9.954424340791195e-06,
+      "loss": 0.4314,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7795162392994426,
+      "learning_rate": 9.954272527507577e-06,
+      "loss": 0.3976,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.7653034767596902,
+      "learning_rate": 9.954120462959408e-06,
+      "loss": 0.4368,
+      "step": 1034
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9251314012104164,
+      "learning_rate": 9.953968147154401e-06,
+      "loss": 0.4197,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9074332632823827,
+      "learning_rate": 9.953815580100279e-06,
+      "loss": 0.4358,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 49.93258726114038,
+      "learning_rate": 9.953662761804784e-06,
+      "loss": 0.4507,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.444959713541837,
+      "learning_rate": 9.95350969227566e-06,
+      "loss": 0.4174,
+      "step": 1038
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.181957015299123,
+      "learning_rate": 9.953356371520675e-06,
+      "loss": 0.4727,
+      "step": 1039
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.873111025963367,
+      "learning_rate": 9.953202799547606e-06,
+      "loss": 0.4278,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.729202561087373,
+      "learning_rate": 9.953048976364237e-06,
+      "loss": 0.3957,
+      "step": 1041
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.479921433608631,
+      "learning_rate": 9.952894901978371e-06,
+      "loss": 0.4317,
+      "step": 1042
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1606146859864888,
+      "learning_rate": 9.952740576397826e-06,
+      "loss": 0.4073,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 7.246735613117482,
+      "learning_rate": 9.952585999630423e-06,
+      "loss": 0.4275,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9897776083067105,
+      "learning_rate": 9.952431171684005e-06,
+      "loss": 0.4581,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.427114067647274,
+      "learning_rate": 9.952276092566426e-06,
+      "loss": 0.4469,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.711986725951927,
+      "learning_rate": 9.952120762285546e-06,
+      "loss": 0.438,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5482344386467966,
+      "learning_rate": 9.951965180849247e-06,
+      "loss": 0.4133,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.0418367709613703,
+      "learning_rate": 9.95180934826542e-06,
+      "loss": 0.3937,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.27010730817319,
+      "learning_rate": 9.951653264541964e-06,
+      "loss": 0.4175,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.448823662032704,
+      "learning_rate": 9.951496929686799e-06,
+      "loss": 0.4193,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.979583418906353,
+      "learning_rate": 9.951340343707852e-06,
+      "loss": 0.4471,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.9362308530902523,
+      "learning_rate": 9.951183506613066e-06,
+      "loss": 0.4306,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.9065215181159543,
+      "learning_rate": 9.951026418410394e-06,
+      "loss": 0.4175,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.084568406687551,
+      "learning_rate": 9.950869079107804e-06,
+      "loss": 0.4621,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.1048326408943305,
+      "learning_rate": 9.950711488713273e-06,
+      "loss": 0.4595,
+      "step": 1056
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7166070699901694,
+      "learning_rate": 9.950553647234798e-06,
+      "loss": 0.477,
+      "step": 1057
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7117568730987686,
+      "learning_rate": 9.95039555468038e-06,
+      "loss": 0.2148,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5867906153209554,
+      "learning_rate": 9.950237211058038e-06,
+      "loss": 0.2092,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.774081852056289,
+      "learning_rate": 9.950078616375804e-06,
+      "loss": 0.4548,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.652149632196855,
+      "learning_rate": 9.949919770641722e-06,
+      "loss": 0.4486,
+      "step": 1061
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.575197742336424,
+      "learning_rate": 9.949760673863846e-06,
+      "loss": 0.4367,
+      "step": 1062
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7656452263646492,
+      "learning_rate": 9.949601326050247e-06,
+      "loss": 0.2141,
+      "step": 1063
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.407268063186484,
+      "learning_rate": 9.949441727209004e-06,
+      "loss": 0.4704,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.4071714763582808,
+      "learning_rate": 9.949281877348214e-06,
+      "loss": 0.4392,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.6583809466659014,
+      "learning_rate": 9.94912177647598e-06,
+      "loss": 0.4024,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.50303313881635,
+      "learning_rate": 9.948961424600428e-06,
+      "loss": 0.4216,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.842483119339154,
+      "learning_rate": 9.948800821729685e-06,
+      "loss": 0.4271,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.8443755429187103,
+      "learning_rate": 9.948639967871897e-06,
+      "loss": 0.3912,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5781680127825783,
+      "learning_rate": 9.948478863035227e-06,
+      "loss": 0.4603,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.743494519941873,
+      "learning_rate": 9.948317507227841e-06,
+      "loss": 0.4482,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7853165911878444,
+      "learning_rate": 9.948155900457922e-06,
+      "loss": 0.4175,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.584455976464318,
+      "learning_rate": 9.947994042733668e-06,
+      "loss": 0.4577,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.40777349528737,
+      "learning_rate": 9.947831934063288e-06,
+      "loss": 0.4399,
+      "step": 1074
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.7369927059890204,
+      "learning_rate": 9.947669574455002e-06,
+      "loss": 0.4172,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5992249036677036,
+      "learning_rate": 9.947506963917047e-06,
+      "loss": 0.2011,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 4.694552074315017,
+      "learning_rate": 9.947344102457669e-06,
+      "loss": 0.4532,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 3.5589258715531056,
+      "learning_rate": 9.947180990085125e-06,
+      "loss": 0.4665,
+      "step": 1078
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.286369863405638,
+      "learning_rate": 9.94701762680769e-06,
+      "loss": 0.428,
+      "step": 1079
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6368977305595624,
+      "learning_rate": 9.946854012633653e-06,
+      "loss": 0.4516,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.663220798678074,
+      "learning_rate": 9.946690147571305e-06,
+      "loss": 0.4229,
+      "step": 1081
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.111531132656279,
+      "learning_rate": 9.94652603162896e-06,
+      "loss": 0.435,
+      "step": 1082
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.3083954011625005,
+      "learning_rate": 9.946361664814942e-06,
+      "loss": 0.4471,
+      "step": 1083
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.355456136094462,
+      "learning_rate": 9.946197047137587e-06,
+      "loss": 0.431,
+      "step": 1084
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.532041968035137,
+      "learning_rate": 9.946032178605242e-06,
+      "loss": 0.437,
+      "step": 1085
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.97214275346169,
+      "learning_rate": 9.94586705922627e-06,
+      "loss": 0.4224,
+      "step": 1086
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.032771876564066,
+      "learning_rate": 9.945701689009045e-06,
+      "loss": 0.4588,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.250505186550476,
+      "learning_rate": 9.945536067961956e-06,
+      "loss": 0.4258,
+      "step": 1088
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.510047607598261,
+      "learning_rate": 9.9453701960934e-06,
+      "loss": 0.4577,
+      "step": 1089
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4408208600540333,
+      "learning_rate": 9.94520407341179e-06,
+      "loss": 0.4148,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6114133648286826,
+      "learning_rate": 9.945037699925552e-06,
+      "loss": 0.4493,
+      "step": 1091
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.328340638258748,
+      "learning_rate": 9.944871075643125e-06,
+      "loss": 0.4327,
+      "step": 1092
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0177008608355465,
+      "learning_rate": 9.944704200572956e-06,
+      "loss": 0.4403,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3015363059593374,
+      "learning_rate": 9.944537074723513e-06,
+      "loss": 0.425,
+      "step": 1094
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.332419965604982,
+      "learning_rate": 9.944369698103267e-06,
+      "loss": 0.4141,
+      "step": 1095
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.356307193059925,
+      "learning_rate": 9.944202070720711e-06,
+      "loss": 0.4084,
+      "step": 1096
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.26835400268666,
+      "learning_rate": 9.944034192584345e-06,
+      "loss": 0.4216,
+      "step": 1097
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.779939663688997,
+      "learning_rate": 9.943866063702683e-06,
+      "loss": 0.4521,
+      "step": 1098
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.236916391095376,
+      "learning_rate": 9.943697684084253e-06,
+      "loss": 0.3967,
+      "step": 1099
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.198740614312695,
+      "learning_rate": 9.943529053737593e-06,
+      "loss": 0.4339,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3844593096209667,
+      "learning_rate": 9.943360172671255e-06,
+      "loss": 0.4345,
+      "step": 1101
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.993297773487913,
+      "learning_rate": 9.943191040893807e-06,
+      "loss": 0.4757,
+      "step": 1102
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.6912353134541824,
+      "learning_rate": 9.943021658413825e-06,
+      "loss": 0.3988,
+      "step": 1103
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.208987199218649,
+      "learning_rate": 9.9428520252399e-06,
+      "loss": 0.4102,
+      "step": 1104
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5454072510901516,
+      "learning_rate": 9.942682141380634e-06,
+      "loss": 0.2002,
+      "step": 1105
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.1667834818075615,
+      "learning_rate": 9.942512006844644e-06,
+      "loss": 0.4103,
+      "step": 1106
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1319650821034606,
+      "learning_rate": 9.942341621640558e-06,
+      "loss": 0.4229,
+      "step": 1107
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.16725144776658,
+      "learning_rate": 9.942170985777018e-06,
+      "loss": 0.4429,
+      "step": 1108
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.472278139952999,
+      "learning_rate": 9.942000099262679e-06,
+      "loss": 0.3997,
+      "step": 1109
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.851181435513924,
+      "learning_rate": 9.941828962106205e-06,
+      "loss": 0.4227,
+      "step": 1110
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.681358461067616,
+      "learning_rate": 9.941657574316281e-06,
+      "loss": 0.4548,
+      "step": 1111
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.07464651228969,
+      "learning_rate": 9.941485935901593e-06,
+      "loss": 0.4012,
+      "step": 1112
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.337874258312825,
+      "learning_rate": 9.941314046870848e-06,
+      "loss": 0.4477,
+      "step": 1113
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.590107170299154,
+      "learning_rate": 9.941141907232766e-06,
+      "loss": 0.4321,
+      "step": 1114
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2174251618952763,
+      "learning_rate": 9.940969516996074e-06,
+      "loss": 0.4275,
+      "step": 1115
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.4935705467156297,
+      "learning_rate": 9.940796876169516e-06,
+      "loss": 0.4667,
+      "step": 1116
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.76872456007815,
+      "learning_rate": 9.94062398476185e-06,
+      "loss": 0.4456,
+      "step": 1117
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.480616137193377,
+      "learning_rate": 9.940450842781842e-06,
+      "loss": 0.4019,
+      "step": 1118
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4941100204773473,
+      "learning_rate": 9.940277450238273e-06,
+      "loss": 0.1817,
+      "step": 1119
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7980285897794146,
+      "learning_rate": 9.94010380713994e-06,
+      "loss": 0.416,
+      "step": 1120
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.330982349482997,
+      "learning_rate": 9.939929913495647e-06,
+      "loss": 0.4204,
+      "step": 1121
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 18.46693260787135,
+      "learning_rate": 9.939755769314215e-06,
+      "loss": 0.4475,
+      "step": 1122
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.206529283351921,
+      "learning_rate": 9.939581374604472e-06,
+      "loss": 0.4178,
+      "step": 1123
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6830424775922337,
+      "learning_rate": 9.939406729375268e-06,
+      "loss": 0.4031,
+      "step": 1124
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.870403439021354,
+      "learning_rate": 9.939231833635457e-06,
+      "loss": 0.4129,
+      "step": 1125
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.0426769502741475,
+      "learning_rate": 9.93905668739391e-06,
+      "loss": 0.4406,
+      "step": 1126
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8861235398908875,
+      "learning_rate": 9.93888129065951e-06,
+      "loss": 0.4247,
+      "step": 1127
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.728802233925717,
+      "learning_rate": 9.938705643441154e-06,
+      "loss": 0.4288,
+      "step": 1128
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.222358328928819,
+      "learning_rate": 9.938529745747748e-06,
+      "loss": 0.4077,
+      "step": 1129
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41055630584458347,
+      "learning_rate": 9.938353597588215e-06,
+      "loss": 0.1985,
+      "step": 1130
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.3957724594094618,
+      "learning_rate": 9.938177198971486e-06,
+      "loss": 0.1708,
+      "step": 1131
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8070448688814866,
+      "learning_rate": 9.938000549906509e-06,
+      "loss": 0.4565,
+      "step": 1132
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2052518312972587,
+      "learning_rate": 9.937823650402242e-06,
+      "loss": 0.4602,
+      "step": 1133
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.637978153885302,
+      "learning_rate": 9.93764650046766e-06,
+      "loss": 0.4354,
+      "step": 1134
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.346258351142239,
+      "learning_rate": 9.937469100111743e-06,
+      "loss": 0.4576,
+      "step": 1135
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.24906511860012,
+      "learning_rate": 9.937291449343492e-06,
+      "loss": 0.4597,
+      "step": 1136
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.696004202405495,
+      "learning_rate": 9.937113548171914e-06,
+      "loss": 0.4584,
+      "step": 1137
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1270208191118725,
+      "learning_rate": 9.936935396606034e-06,
+      "loss": 0.4159,
+      "step": 1138
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.533674325664934,
+      "learning_rate": 9.936756994654885e-06,
+      "loss": 0.4564,
+      "step": 1139
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1460806516841977,
+      "learning_rate": 9.936578342327516e-06,
+      "loss": 0.4294,
+      "step": 1140
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.4987995311397215,
+      "learning_rate": 9.936399439632989e-06,
+      "loss": 0.4471,
+      "step": 1141
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 10.820221969386628,
+      "learning_rate": 9.936220286580375e-06,
+      "loss": 0.4176,
+      "step": 1142
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.676369467053964,
+      "learning_rate": 9.93604088317876e-06,
+      "loss": 0.4481,
+      "step": 1143
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.2547965490393205,
+      "learning_rate": 9.935861229437245e-06,
+      "loss": 0.4013,
+      "step": 1144
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.2082889272789954,
+      "learning_rate": 9.93568132536494e-06,
+      "loss": 0.4464,
+      "step": 1145
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 16.016363357610103,
+      "learning_rate": 9.935501170970969e-06,
+      "loss": 0.4392,
+      "step": 1146
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 6.3284297313980415,
+      "learning_rate": 9.93532076626447e-06,
+      "loss": 0.413,
+      "step": 1147
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9977556484155463,
+      "learning_rate": 9.935140111254593e-06,
+      "loss": 0.4285,
+      "step": 1148
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.7522821400322757,
+      "learning_rate": 9.934959205950496e-06,
+      "loss": 0.4355,
+      "step": 1149
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.544065942363322,
+      "learning_rate": 9.93477805036136e-06,
+      "loss": 0.4903,
+      "step": 1150
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.035204136899857,
+      "learning_rate": 9.934596644496369e-06,
+      "loss": 0.459,
+      "step": 1151
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6661872807875917,
+      "learning_rate": 9.934414988364722e-06,
+      "loss": 0.2021,
+      "step": 1152
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.131251316993209,
+      "learning_rate": 9.934233081975634e-06,
+      "loss": 0.4358,
+      "step": 1153
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.277501768723896,
+      "learning_rate": 9.934050925338333e-06,
+      "loss": 0.462,
+      "step": 1154
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 11.85162956359785,
+      "learning_rate": 9.933868518462052e-06,
+      "loss": 0.427,
+      "step": 1155
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.984646790622588,
+      "learning_rate": 9.933685861356047e-06,
+      "loss": 0.4426,
+      "step": 1156
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3956177237768874,
+      "learning_rate": 9.93350295402958e-06,
+      "loss": 0.4362,
+      "step": 1157
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4863020865542325,
+      "learning_rate": 9.933319796491927e-06,
+      "loss": 0.4508,
+      "step": 1158
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.3020487920672057,
+      "learning_rate": 9.933136388752378e-06,
+      "loss": 0.4342,
+      "step": 1159
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.982170068129495,
+      "learning_rate": 9.932952730820234e-06,
+      "loss": 0.46,
+      "step": 1160
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.351854993863421,
+      "learning_rate": 9.932768822704807e-06,
+      "loss": 0.3792,
+      "step": 1161
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.426568911751664,
+      "learning_rate": 9.93258466441543e-06,
+      "loss": 0.4131,
+      "step": 1162
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.767536468815671,
+      "learning_rate": 9.93240025596144e-06,
+      "loss": 0.4184,
+      "step": 1163
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.031337442726975,
+      "learning_rate": 9.932215597352189e-06,
+      "loss": 0.4294,
+      "step": 1164
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.393954433305972,
+      "learning_rate": 9.932030688597043e-06,
+      "loss": 0.3941,
+      "step": 1165
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.521550002994359,
+      "learning_rate": 9.93184552970538e-06,
+      "loss": 0.4368,
+      "step": 1166
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8005162829912957,
+      "learning_rate": 9.93166012068659e-06,
+      "loss": 0.4473,
+      "step": 1167
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1002768067058355,
+      "learning_rate": 9.931474461550078e-06,
+      "loss": 0.4298,
+      "step": 1168
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1273636615678875,
+      "learning_rate": 9.931288552305256e-06,
+      "loss": 0.3977,
+      "step": 1169
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5501700533299467,
+      "learning_rate": 9.931102392961558e-06,
+      "loss": 0.3866,
+      "step": 1170
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.678064111580878,
+      "learning_rate": 9.930915983528423e-06,
+      "loss": 0.4445,
+      "step": 1171
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5750436085497768,
+      "learning_rate": 9.930729324015304e-06,
+      "loss": 0.4355,
+      "step": 1172
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.535683543122053,
+      "learning_rate": 9.93054241443167e-06,
+      "loss": 0.4059,
+      "step": 1173
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.112776096652475,
+      "learning_rate": 9.930355254786998e-06,
+      "loss": 0.4615,
+      "step": 1174
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.9605642939225154,
+      "learning_rate": 9.930167845090782e-06,
+      "loss": 0.4223,
+      "step": 1175
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.48718870356136,
+      "learning_rate": 9.929980185352525e-06,
+      "loss": 0.4293,
+      "step": 1176
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.349663972101028,
+      "learning_rate": 9.929792275581747e-06,
+      "loss": 0.4184,
+      "step": 1177
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4311990338339093,
+      "learning_rate": 9.929604115787976e-06,
+      "loss": 0.4585,
+      "step": 1178
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1632823509908334,
+      "learning_rate": 9.929415705980756e-06,
+      "loss": 0.4375,
+      "step": 1179
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7364020251341993,
+      "learning_rate": 9.929227046169644e-06,
+      "loss": 0.2196,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.412574251135926,
+      "learning_rate": 9.929038136364203e-06,
+      "loss": 0.4327,
+      "step": 1181
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4840731473119786,
+      "learning_rate": 9.92884897657402e-06,
+      "loss": 0.4346,
+      "step": 1182
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0855202073514745,
+      "learning_rate": 9.928659566808685e-06,
+      "loss": 0.4199,
+      "step": 1183
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.711117802808915,
+      "learning_rate": 9.928469907077805e-06,
+      "loss": 0.4499,
+      "step": 1184
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9991791981909928,
+      "learning_rate": 9.928279997391e-06,
+      "loss": 0.4657,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.651551307853599,
+      "learning_rate": 9.928089837757898e-06,
+      "loss": 0.4743,
+      "step": 1186
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 24.748604278838794,
+      "learning_rate": 9.927899428188148e-06,
+      "loss": 0.4129,
+      "step": 1187
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.078369410923003,
+      "learning_rate": 9.927708768691404e-06,
+      "loss": 0.4336,
+      "step": 1188
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6372150539573447,
+      "learning_rate": 9.927517859277338e-06,
+      "loss": 0.456,
+      "step": 1189
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.21357561378674,
+      "learning_rate": 9.927326699955629e-06,
+      "loss": 0.4169,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.653867907914646,
+      "learning_rate": 9.927135290735976e-06,
+      "loss": 0.3925,
+      "step": 1191
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9196778365294138,
+      "learning_rate": 9.926943631628084e-06,
+      "loss": 0.4272,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.440449875776898,
+      "learning_rate": 9.92675172264167e-06,
+      "loss": 0.3891,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.32532011897416,
+      "learning_rate": 9.926559563786475e-06,
+      "loss": 0.4395,
+      "step": 1194
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5708573342430707,
+      "learning_rate": 9.926367155072239e-06,
+      "loss": 0.4195,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.142548534286809,
+      "learning_rate": 9.926174496508722e-06,
+      "loss": 0.449,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.10539495669056,
+      "learning_rate": 9.925981588105695e-06,
+      "loss": 0.4399,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.9574955232778217,
+      "learning_rate": 9.92578842987294e-06,
+      "loss": 0.4374,
+      "step": 1198
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3120212939722986,
+      "learning_rate": 9.925595021820256e-06,
+      "loss": 0.3964,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.732786840229121,
+      "learning_rate": 9.925401363957451e-06,
+      "loss": 0.3841,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.528052609984117,
+      "learning_rate": 9.925207456294345e-06,
+      "loss": 0.4478,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3741524076267106,
+      "learning_rate": 9.925013298840777e-06,
+      "loss": 0.4059,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7048468570653954,
+      "learning_rate": 9.924818891606589e-06,
+      "loss": 0.4319,
+      "step": 1203
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.7844645252753706,
+      "learning_rate": 9.924624234601644e-06,
+      "loss": 0.4665,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.0709629172473045,
+      "learning_rate": 9.92442932783581e-06,
+      "loss": 0.4551,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.5026817091010987,
+      "learning_rate": 9.924234171318978e-06,
+      "loss": 0.435,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.701133909677434,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 0.4138,
+      "step": 1207
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.157322661148928,
+      "learning_rate": 9.923843109071912e-06,
+      "loss": 0.4419,
+      "step": 1208
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.782107886249264,
+      "learning_rate": 9.923647203361512e-06,
+      "loss": 0.5042,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6381231442721589,
+      "learning_rate": 9.92345104793978e-06,
+      "loss": 0.2122,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5785606909716376,
+      "learning_rate": 9.92325464281666e-06,
+      "loss": 0.4665,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3792103906381246,
+      "learning_rate": 9.923057988002117e-06,
+      "loss": 0.4684,
+      "step": 1212
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.0784189102588533,
+      "learning_rate": 9.922861083506122e-06,
+      "loss": 0.4485,
+      "step": 1213
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.39294450479478127,
+      "learning_rate": 9.922663929338662e-06,
+      "loss": 0.219,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.850122639417442,
+      "learning_rate": 9.922466525509738e-06,
+      "loss": 0.4358,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.24475959689342,
+      "learning_rate": 9.92226887202936e-06,
+      "loss": 0.4524,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.558078889848628,
+      "learning_rate": 9.922070968907548e-06,
+      "loss": 0.4364,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 5.520584857788954,
+      "learning_rate": 9.921872816154348e-06,
+      "loss": 0.4587,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8002460155417075,
+      "learning_rate": 9.921674413779804e-06,
+      "loss": 0.446,
+      "step": 1219
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.715159478025085,
+      "learning_rate": 9.921475761793978e-06,
+      "loss": 0.4394,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 7.513522702539402,
+      "learning_rate": 9.921276860206948e-06,
+      "loss": 0.4284,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.8215245196030474,
+      "learning_rate": 9.9210777090288e-06,
+      "loss": 0.4286,
+      "step": 1222
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6453986421215535,
+      "learning_rate": 9.920878308269632e-06,
+      "loss": 0.2061,
+      "step": 1223
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.150336532746913,
+      "learning_rate": 9.92067865793956e-06,
+      "loss": 0.4176,
+      "step": 1224
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.326373236708268,
+      "learning_rate": 9.92047875804871e-06,
+      "loss": 0.3968,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1819016661345327,
+      "learning_rate": 9.920278608607218e-06,
+      "loss": 0.4107,
+      "step": 1226
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5823969591094946,
+      "learning_rate": 9.920078209625235e-06,
+      "loss": 0.4885,
+      "step": 1227
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.398766574010292,
+      "learning_rate": 9.919877561112927e-06,
+      "loss": 0.4253,
+      "step": 1228
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.692137002339529,
+      "learning_rate": 9.91967666308047e-06,
+      "loss": 0.4615,
+      "step": 1229
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1220297958821526,
+      "learning_rate": 9.91947551553805e-06,
+      "loss": 0.4205,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.526910298384765,
+      "learning_rate": 9.919274118495871e-06,
+      "loss": 0.4447,
+      "step": 1231
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.420380652014693,
+      "learning_rate": 9.919072471964146e-06,
+      "loss": 0.4199,
+      "step": 1232
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.453333444769558,
+      "learning_rate": 9.918870575953102e-06,
+      "loss": 0.4399,
+      "step": 1233
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.500538153047233,
+      "learning_rate": 9.91866843047298e-06,
+      "loss": 0.4504,
+      "step": 1234
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.63153810285414,
+      "learning_rate": 9.91846603553403e-06,
+      "loss": 0.4275,
+      "step": 1235
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.061780561343687,
+      "learning_rate": 9.918263391146518e-06,
+      "loss": 0.4909,
+      "step": 1236
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5912449416796384,
+      "learning_rate": 9.918060497320722e-06,
+      "loss": 0.4512,
+      "step": 1237
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6542848604510274,
+      "learning_rate": 9.91785735406693e-06,
+      "loss": 0.4355,
+      "step": 1238
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.744369334194906,
+      "learning_rate": 9.917653961395448e-06,
+      "loss": 0.4633,
+      "step": 1239
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6730910108847223,
+      "learning_rate": 9.917450319316589e-06,
+      "loss": 0.451,
+      "step": 1240
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.419234981708987,
+      "learning_rate": 9.91724642784068e-06,
+      "loss": 0.4161,
+      "step": 1241
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.642932123563812,
+      "learning_rate": 9.917042286978064e-06,
+      "loss": 0.4574,
+      "step": 1242
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.813775114349482,
+      "learning_rate": 9.916837896739095e-06,
+      "loss": 0.4574,
+      "step": 1243
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4842120935396106,
+      "learning_rate": 9.916633257134137e-06,
+      "loss": 0.4021,
+      "step": 1244
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.493155173950058,
+      "learning_rate": 9.916428368173567e-06,
+      "loss": 0.4373,
+      "step": 1245
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.945310679817467,
+      "learning_rate": 9.916223229867782e-06,
+      "loss": 0.4599,
+      "step": 1246
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.49065452496578427,
+      "learning_rate": 9.916017842227182e-06,
+      "loss": 0.1995,
+      "step": 1247
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.415091658307623,
+      "learning_rate": 9.915812205262182e-06,
+      "loss": 0.4241,
+      "step": 1248
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 20.253531226863608,
+      "learning_rate": 9.915606318983214e-06,
+      "loss": 0.4451,
+      "step": 1249
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.398089078185956,
+      "learning_rate": 9.915400183400722e-06,
+      "loss": 0.4151,
+      "step": 1250
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.218118326731287,
+      "learning_rate": 9.915193798525156e-06,
+      "loss": 0.4395,
+      "step": 1251
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.6387480458646655,
+      "learning_rate": 9.914987164366984e-06,
+      "loss": 0.4438,
+      "step": 1252
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.378456453628441,
+      "learning_rate": 9.914780280936689e-06,
+      "loss": 0.4404,
+      "step": 1253
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4639545101048483,
+      "learning_rate": 9.91457314824476e-06,
+      "loss": 0.4391,
+      "step": 1254
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3965394591749213,
+      "learning_rate": 9.914365766301704e-06,
+      "loss": 0.1915,
+      "step": 1255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.577866149168461,
+      "learning_rate": 9.914158135118038e-06,
+      "loss": 0.4285,
+      "step": 1256
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4281518225578393,
+      "learning_rate": 9.913950254704291e-06,
+      "loss": 0.4501,
+      "step": 1257
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4995749899434667,
+      "learning_rate": 9.91374212507101e-06,
+      "loss": 0.413,
+      "step": 1258
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.36787871861181126,
+      "learning_rate": 9.913533746228745e-06,
+      "loss": 0.2046,
+      "step": 1259
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 8.024046175010731,
+      "learning_rate": 9.913325118188068e-06,
+      "loss": 0.4421,
+      "step": 1260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.891086428187552,
+      "learning_rate": 9.913116240959561e-06,
+      "loss": 0.4059,
+      "step": 1261
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.967021333814803,
+      "learning_rate": 9.912907114553815e-06,
+      "loss": 0.4433,
+      "step": 1262
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3985954952843727,
+      "learning_rate": 9.912697738981438e-06,
+      "loss": 0.1882,
+      "step": 1263
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.385145443179772,
+      "learning_rate": 9.912488114253048e-06,
+      "loss": 0.4611,
+      "step": 1264
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6111550284943945,
+      "learning_rate": 9.912278240379277e-06,
+      "loss": 0.4021,
+      "step": 1265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.197810362479608,
+      "learning_rate": 9.912068117370766e-06,
+      "loss": 0.4318,
+      "step": 1266
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.897659540714329,
+      "learning_rate": 9.911857745238176e-06,
+      "loss": 0.4122,
+      "step": 1267
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.993072779277048,
+      "learning_rate": 9.911647123992175e-06,
+      "loss": 0.4171,
+      "step": 1268
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2195390100586305,
+      "learning_rate": 9.911436253643445e-06,
+      "loss": 0.3956,
+      "step": 1269
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.253377871135009,
+      "learning_rate": 9.911225134202678e-06,
+      "loss": 0.4109,
+      "step": 1270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.765403483876604,
+      "learning_rate": 9.911013765680585e-06,
+      "loss": 0.4288,
+      "step": 1271
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3244104939160093,
+      "learning_rate": 9.910802148087887e-06,
+      "loss": 0.4202,
+      "step": 1272
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.6463544760084274,
+      "learning_rate": 9.910590281435313e-06,
+      "loss": 0.4102,
+      "step": 1273
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3983631775535119,
+      "learning_rate": 9.910378165733608e-06,
+      "loss": 0.1926,
+      "step": 1274
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.101240643252105,
+      "learning_rate": 9.910165800993532e-06,
+      "loss": 0.4178,
+      "step": 1275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8126184738309203,
+      "learning_rate": 9.909953187225853e-06,
+      "loss": 0.4491,
+      "step": 1276
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.771073582195782,
+      "learning_rate": 9.90974032444136e-06,
+      "loss": 0.415,
+      "step": 1277
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.002110498527838,
+      "learning_rate": 9.90952721265084e-06,
+      "loss": 0.4304,
+      "step": 1278
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 21.906951635406976,
+      "learning_rate": 9.909313851865108e-06,
+      "loss": 0.5027,
+      "step": 1279
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8291527859116354,
+      "learning_rate": 9.909100242094982e-06,
+      "loss": 0.4422,
+      "step": 1280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.607394676133592,
+      "learning_rate": 9.908886383351297e-06,
+      "loss": 0.3872,
+      "step": 1281
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.062496321929604,
+      "learning_rate": 9.908672275644898e-06,
+      "loss": 0.4249,
+      "step": 1282
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.840696509209897,
+      "learning_rate": 9.908457918986644e-06,
+      "loss": 0.4009,
+      "step": 1283
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2390821491725768,
+      "learning_rate": 9.908243313387407e-06,
+      "loss": 0.4222,
+      "step": 1284
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.881441035583195,
+      "learning_rate": 9.908028458858072e-06,
+      "loss": 0.4148,
+      "step": 1285
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.6353031998161915,
+      "learning_rate": 9.907813355409532e-06,
+      "loss": 0.4104,
+      "step": 1286
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.197812978684478,
+      "learning_rate": 9.907598003052701e-06,
+      "loss": 0.4239,
+      "step": 1287
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.811758219087973,
+      "learning_rate": 9.9073824017985e-06,
+      "loss": 0.4387,
+      "step": 1288
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.7963292273811176,
+      "learning_rate": 9.90716655165786e-06,
+      "loss": 0.4319,
+      "step": 1289
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.188113711758418,
+      "learning_rate": 9.906950452641732e-06,
+      "loss": 0.466,
+      "step": 1290
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.427051054552438,
+      "learning_rate": 9.906734104761075e-06,
+      "loss": 0.4437,
+      "step": 1291
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.70159752668156,
+      "learning_rate": 9.906517508026862e-06,
+      "loss": 0.4543,
+      "step": 1292
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.38102945247589187,
+      "learning_rate": 9.906300662450074e-06,
+      "loss": 0.1923,
+      "step": 1293
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.947810166560389,
+      "learning_rate": 9.906083568041716e-06,
+      "loss": 0.4238,
+      "step": 1294
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.15298924622279,
+      "learning_rate": 9.905866224812791e-06,
+      "loss": 0.4552,
+      "step": 1295
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3476246200862114,
+      "learning_rate": 9.905648632774327e-06,
+      "loss": 0.1914,
+      "step": 1296
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.3648993497820476,
+      "learning_rate": 9.905430791937356e-06,
+      "loss": 0.1866,
+      "step": 1297
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.154010162443714,
+      "learning_rate": 9.905212702312929e-06,
+      "loss": 0.4433,
+      "step": 1298
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8595326657285423,
+      "learning_rate": 9.904994363912106e-06,
+      "loss": 0.4377,
+      "step": 1299
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0302655840679535,
+      "learning_rate": 9.904775776745959e-06,
+      "loss": 0.4389,
+      "step": 1300
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.708333658107862,
+      "learning_rate": 9.904556940825576e-06,
+      "loss": 0.4339,
+      "step": 1301
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.825488333706638,
+      "learning_rate": 9.904337856162054e-06,
+      "loss": 0.4387,
+      "step": 1302
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3123234722617942,
+      "learning_rate": 9.904118522766505e-06,
+      "loss": 0.4664,
+      "step": 1303
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.654506720238037,
+      "learning_rate": 9.903898940650052e-06,
+      "loss": 0.4323,
+      "step": 1304
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.063381350777765,
+      "learning_rate": 9.903679109823834e-06,
+      "loss": 0.4281,
+      "step": 1305
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.166128292532498,
+      "learning_rate": 9.903459030299e-06,
+      "loss": 0.4675,
+      "step": 1306
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.678468945698042,
+      "learning_rate": 9.903238702086707e-06,
+      "loss": 0.4193,
+      "step": 1307
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.578265412320871,
+      "learning_rate": 9.903018125198133e-06,
+      "loss": 0.4399,
+      "step": 1308
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 7.798696290643071,
+      "learning_rate": 9.902797299644464e-06,
+      "loss": 0.3984,
+      "step": 1309
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6327166714679118,
+      "learning_rate": 9.902576225436902e-06,
+      "loss": 0.3947,
+      "step": 1310
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.9659144637568176,
+      "learning_rate": 9.902354902586656e-06,
+      "loss": 0.4038,
+      "step": 1311
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.521752353319252,
+      "learning_rate": 9.902133331104952e-06,
+      "loss": 0.413,
+      "step": 1312
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.047301985307986,
+      "learning_rate": 9.901911511003028e-06,
+      "loss": 0.412,
+      "step": 1313
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.541187403729893,
+      "learning_rate": 9.901689442292133e-06,
+      "loss": 0.4151,
+      "step": 1314
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5934682493270507,
+      "learning_rate": 9.90146712498353e-06,
+      "loss": 0.4062,
+      "step": 1315
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.4050351485235373,
+      "learning_rate": 9.901244559088493e-06,
+      "loss": 0.4412,
+      "step": 1316
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.999456588742732,
+      "learning_rate": 9.90102174461831e-06,
+      "loss": 0.4251,
+      "step": 1317
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 6.722554977351457,
+      "learning_rate": 9.900798681584287e-06,
+      "loss": 0.4777,
+      "step": 1318
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6491133277439083,
+      "learning_rate": 9.90057536999773e-06,
+      "loss": 0.5076,
+      "step": 1319
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.259821871177595,
+      "learning_rate": 9.900351809869965e-06,
+      "loss": 0.4535,
+      "step": 1320
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5439173295487407,
+      "learning_rate": 9.900128001212333e-06,
+      "loss": 0.4107,
+      "step": 1321
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.3635711799698256,
+      "learning_rate": 9.899903944036185e-06,
+      "loss": 0.441,
+      "step": 1322
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.391353505198506,
+      "learning_rate": 9.899679638352884e-06,
+      "loss": 0.4194,
+      "step": 1323
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.706148731480394,
+      "learning_rate": 9.899455084173805e-06,
+      "loss": 0.4266,
+      "step": 1324
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4022408240101885,
+      "learning_rate": 9.899230281510337e-06,
+      "loss": 0.4232,
+      "step": 1325
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1042207040924685,
+      "learning_rate": 9.899005230373883e-06,
+      "loss": 0.4085,
+      "step": 1326
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7708257036871626,
+      "learning_rate": 9.898779930775854e-06,
+      "loss": 0.4222,
+      "step": 1327
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9327551145206243,
+      "learning_rate": 9.898554382727678e-06,
+      "loss": 0.432,
+      "step": 1328
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.277371731023874,
+      "learning_rate": 9.898328586240794e-06,
+      "loss": 0.4218,
+      "step": 1329
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.179494896509849,
+      "learning_rate": 9.898102541326652e-06,
+      "loss": 0.4145,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.423862236456824,
+      "learning_rate": 9.89787624799672e-06,
+      "loss": 0.4237,
+      "step": 1331
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.881261143743926,
+      "learning_rate": 9.897649706262474e-06,
+      "loss": 0.484,
+      "step": 1332
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7045856869554368,
+      "learning_rate": 9.897422916135399e-06,
+      "loss": 0.4455,
+      "step": 1333
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.4969299142313623,
+      "learning_rate": 9.897195877627002e-06,
+      "loss": 0.4204,
+      "step": 1334
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.1577634046450855,
+      "learning_rate": 9.896968590748795e-06,
+      "loss": 0.4376,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.1639893510021087,
+      "learning_rate": 9.896741055512307e-06,
+      "loss": 0.4594,
+      "step": 1336
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.362881454833505,
+      "learning_rate": 9.896513271929075e-06,
+      "loss": 0.4215,
+      "step": 1337
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.722700550196609,
+      "learning_rate": 9.896285240010658e-06,
+      "loss": 0.413,
+      "step": 1338
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.847706165728031,
+      "learning_rate": 9.896056959768612e-06,
+      "loss": 0.4506,
+      "step": 1339
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7093360199674783,
+      "learning_rate": 9.89582843121452e-06,
+      "loss": 0.3817,
+      "step": 1340
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.9384354891306943,
+      "learning_rate": 9.895599654359972e-06,
+      "loss": 0.3968,
+      "step": 1341
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5214411974180595,
+      "learning_rate": 9.89537062921657e-06,
+      "loss": 0.4219,
+      "step": 1342
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.6325340493306615,
+      "learning_rate": 9.895141355795929e-06,
+      "loss": 0.437,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.515557444740171,
+      "learning_rate": 9.894911834109677e-06,
+      "loss": 0.4672,
+      "step": 1344
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5222382794626941,
+      "learning_rate": 9.894682064169457e-06,
+      "loss": 0.1843,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.543313722648741,
+      "learning_rate": 9.894452045986918e-06,
+      "loss": 0.4345,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.026490847010613,
+      "learning_rate": 9.894221779573729e-06,
+      "loss": 0.3922,
+      "step": 1347
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.360372710480551,
+      "learning_rate": 9.893991264941568e-06,
+      "loss": 0.4604,
+      "step": 1348
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7634091389614683,
+      "learning_rate": 9.893760502102126e-06,
+      "loss": 0.4294,
+      "step": 1349
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.305344259713021,
+      "learning_rate": 9.893529491067104e-06,
+      "loss": 0.4142,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 10.278225944514825,
+      "learning_rate": 9.89329823184822e-06,
+      "loss": 0.4509,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.5996550380780117,
+      "learning_rate": 9.893066724457206e-06,
+      "loss": 0.415,
+      "step": 1352
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.272017005265138,
+      "learning_rate": 9.892834968905797e-06,
+      "loss": 0.3942,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.8347514420560933,
+      "learning_rate": 9.89260296520575e-06,
+      "loss": 0.4308,
+      "step": 1354
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.8512987303965227,
+      "learning_rate": 9.892370713368832e-06,
+      "loss": 0.4233,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.608258574673595,
+      "learning_rate": 9.892138213406822e-06,
+      "loss": 0.4052,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.595552277691766,
+      "learning_rate": 9.89190546533151e-06,
+      "loss": 0.4351,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5595346351700794,
+      "learning_rate": 9.891672469154703e-06,
+      "loss": 0.4572,
+      "step": 1358
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5988707179466175,
+      "learning_rate": 9.891439224888215e-06,
+      "loss": 0.4605,
+      "step": 1359
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.629174730092055,
+      "learning_rate": 9.891205732543876e-06,
+      "loss": 0.4068,
+      "step": 1360
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4720779611628841,
+      "learning_rate": 9.89097199213353e-06,
+      "loss": 0.2119,
+      "step": 1361
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 5.1228754884183765,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.4348,
+      "step": 1362
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2731065977279643,
+      "learning_rate": 9.890503767162241e-06,
+      "loss": 0.4286,
+      "step": 1363
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 4.87239730378908,
+      "learning_rate": 9.890269282625046e-06,
+      "loss": 0.4027,
+      "step": 1364
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.69823914777617,
+      "learning_rate": 9.890034550069337e-06,
+      "loss": 0.4149,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.7228595725918816,
+      "learning_rate": 9.889799569507017e-06,
+      "loss": 0.4382,
+      "step": 1366
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4774306623991786,
+      "learning_rate": 9.889564340950005e-06,
+      "loss": 0.4779,
+      "step": 1367
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9420991055033188,
+      "learning_rate": 9.88932886441023e-06,
+      "loss": 0.4605,
+      "step": 1368
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7047514932470524,
+      "learning_rate": 9.889093139899637e-06,
+      "loss": 0.4166,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.107589852033877,
+      "learning_rate": 9.888857167430177e-06,
+      "loss": 0.4479,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 17.173566256945588,
+      "learning_rate": 9.888620947013821e-06,
+      "loss": 0.4364,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6216358852212953,
+      "learning_rate": 9.88838447866255e-06,
+      "loss": 0.4325,
+      "step": 1372
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4141353380257384,
+      "learning_rate": 9.888147762388353e-06,
+      "loss": 0.4184,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4619377783511947,
+      "learning_rate": 9.887910798203237e-06,
+      "loss": 0.1875,
+      "step": 1374
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5908482234590173,
+      "learning_rate": 9.887673586119225e-06,
+      "loss": 0.4348,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.572550738151608,
+      "learning_rate": 9.88743612614834e-06,
+      "loss": 0.4351,
+      "step": 1376
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.38447415326295686,
+      "learning_rate": 9.887198418302629e-06,
+      "loss": 0.1839,
+      "step": 1377
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6121068550807234,
+      "learning_rate": 9.886960462594146e-06,
+      "loss": 0.4444,
+      "step": 1378
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.2451872594560434,
+      "learning_rate": 9.886722259034964e-06,
+      "loss": 0.404,
+      "step": 1379
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37630951973226284,
+      "learning_rate": 9.88648380763716e-06,
+      "loss": 0.2044,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7143470418980256,
+      "learning_rate": 9.886245108412826e-06,
+      "loss": 0.4317,
+      "step": 1381
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.844463840435968,
+      "learning_rate": 9.88600616137407e-06,
+      "loss": 0.4058,
+      "step": 1382
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.100360126563574,
+      "learning_rate": 9.885766966533014e-06,
+      "loss": 0.4436,
+      "step": 1383
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5743990629611617,
+      "learning_rate": 9.885527523901783e-06,
+      "loss": 0.4408,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5854765616252844,
+      "learning_rate": 9.885287833492525e-06,
+      "loss": 0.4643,
+      "step": 1385
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4017264251304207,
+      "learning_rate": 9.885047895317396e-06,
+      "loss": 0.4138,
+      "step": 1386
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.087828587727349,
+      "learning_rate": 9.884807709388563e-06,
+      "loss": 0.4087,
+      "step": 1387
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7248167184098806,
+      "learning_rate": 9.88456727571821e-06,
+      "loss": 0.4498,
+      "step": 1388
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.3519375759136665,
+      "learning_rate": 9.884326594318525e-06,
+      "loss": 0.417,
+      "step": 1389
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.178480391399827,
+      "learning_rate": 9.884085665201723e-06,
+      "loss": 0.4737,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1316478806567836,
+      "learning_rate": 9.883844488380016e-06,
+      "loss": 0.4115,
+      "step": 1391
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6481743069293215,
+      "learning_rate": 9.883603063865642e-06,
+      "loss": 0.4224,
+      "step": 1392
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.377458177932767,
+      "learning_rate": 9.883361391670841e-06,
+      "loss": 0.4469,
+      "step": 1393
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.05232642145485,
+      "learning_rate": 9.883119471807869e-06,
+      "loss": 0.4137,
+      "step": 1394
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.538100799786533,
+      "learning_rate": 9.882877304289e-06,
+      "loss": 0.4087,
+      "step": 1395
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.096909815865425,
+      "learning_rate": 9.882634889126513e-06,
+      "loss": 0.4548,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8014023819863745,
+      "learning_rate": 9.8823922263327e-06,
+      "loss": 0.422,
+      "step": 1397
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.98489518248428,
+      "learning_rate": 9.882149315919873e-06,
+      "loss": 0.426,
+      "step": 1398
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.074324681773418,
+      "learning_rate": 9.881906157900349e-06,
+      "loss": 0.4619,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6855118378531335,
+      "learning_rate": 9.88166275228646e-06,
+      "loss": 0.4367,
+      "step": 1400
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.087480394419512,
+      "learning_rate": 9.881419099090553e-06,
+      "loss": 0.5158,
+      "step": 1401
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.293781495543259,
+      "learning_rate": 9.881175198324982e-06,
+      "loss": 0.451,
+      "step": 1402
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.571875982116657,
+      "learning_rate": 9.880931050002119e-06,
+      "loss": 0.4609,
+      "step": 1403
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2598479573039434,
+      "learning_rate": 9.880686654134345e-06,
+      "loss": 0.4091,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.742482930441708,
+      "learning_rate": 9.880442010734057e-06,
+      "loss": 0.4461,
+      "step": 1405
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.369320665302325,
+      "learning_rate": 9.880197119813662e-06,
+      "loss": 0.409,
+      "step": 1406
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.028565082168225,
+      "learning_rate": 9.879951981385577e-06,
+      "loss": 0.4375,
+      "step": 1407
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.244226721228595,
+      "learning_rate": 9.87970659546224e-06,
+      "loss": 0.4679,
+      "step": 1408
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.098076617356885,
+      "learning_rate": 9.879460962056091e-06,
+      "loss": 0.4231,
+      "step": 1409
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7685714448732517,
+      "learning_rate": 9.87921508117959e-06,
+      "loss": 0.417,
+      "step": 1410
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.579549858277636,
+      "learning_rate": 9.87896895284521e-06,
+      "loss": 0.4253,
+      "step": 1411
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.488877871154544,
+      "learning_rate": 9.878722577065427e-06,
+      "loss": 0.4048,
+      "step": 1412
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.054934292011723,
+      "learning_rate": 9.878475953852743e-06,
+      "loss": 0.4638,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.049527211209861,
+      "learning_rate": 9.878229083219663e-06,
+      "loss": 0.4238,
+      "step": 1414
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.826493240702327,
+      "learning_rate": 9.877981965178708e-06,
+      "loss": 0.4562,
+      "step": 1415
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4382431765096597,
+      "learning_rate": 9.87773459974241e-06,
+      "loss": 0.4539,
+      "step": 1416
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3706126371210314,
+      "learning_rate": 9.877486986923317e-06,
+      "loss": 0.4121,
+      "step": 1417
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4398745213713076,
+      "learning_rate": 9.877239126733984e-06,
+      "loss": 0.442,
+      "step": 1418
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.67941938081869,
+      "learning_rate": 9.876991019186983e-06,
+      "loss": 0.4286,
+      "step": 1419
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.245241456597094,
+      "learning_rate": 9.876742664294897e-06,
+      "loss": 0.4143,
+      "step": 1420
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.794478767568897,
+      "learning_rate": 9.876494062070325e-06,
+      "loss": 0.3955,
+      "step": 1421
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.641648531167604,
+      "learning_rate": 9.87624521252587e-06,
+      "loss": 0.449,
+      "step": 1422
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.829429588447433,
+      "learning_rate": 9.875996115674157e-06,
+      "loss": 0.4513,
+      "step": 1423
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.686669406926293,
+      "learning_rate": 9.875746771527817e-06,
+      "loss": 0.4348,
+      "step": 1424
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5634609278891123,
+      "learning_rate": 9.875497180099495e-06,
+      "loss": 0.4687,
+      "step": 1425
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9840078829316004,
+      "learning_rate": 9.875247341401854e-06,
+      "loss": 0.4321,
+      "step": 1426
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.249527072919854,
+      "learning_rate": 9.87499725544756e-06,
+      "loss": 0.4565,
+      "step": 1427
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 67.76183811606278,
+      "learning_rate": 9.874746922249298e-06,
+      "loss": 0.4156,
+      "step": 1428
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.783642243797997,
+      "learning_rate": 9.874496341819767e-06,
+      "loss": 0.4326,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.050877953699164,
+      "learning_rate": 9.874245514171672e-06,
+      "loss": 0.4501,
+      "step": 1430
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.806083937603903,
+      "learning_rate": 9.873994439317735e-06,
+      "loss": 0.4432,
+      "step": 1431
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.554274889015211,
+      "learning_rate": 9.873743117270691e-06,
+      "loss": 0.4193,
+      "step": 1432
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.661095995843297,
+      "learning_rate": 9.873491548043287e-06,
+      "loss": 0.4059,
+      "step": 1433
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.431696410187974,
+      "learning_rate": 9.873239731648277e-06,
+      "loss": 0.4119,
+      "step": 1434
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 15.342963702876913,
+      "learning_rate": 9.872987668098438e-06,
+      "loss": 0.4449,
+      "step": 1435
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.678512697794886,
+      "learning_rate": 9.872735357406548e-06,
+      "loss": 0.4578,
+      "step": 1436
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.770335551828142,
+      "learning_rate": 9.87248279958541e-06,
+      "loss": 0.4247,
+      "step": 1437
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7025096088660074,
+      "learning_rate": 9.872229994647828e-06,
+      "loss": 0.4386,
+      "step": 1438
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7375202993159387,
+      "learning_rate": 9.871976942606626e-06,
+      "loss": 0.4257,
+      "step": 1439
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 6.595134377671582,
+      "learning_rate": 9.871723643474637e-06,
+      "loss": 0.4411,
+      "step": 1440
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.390428763346444,
+      "learning_rate": 9.871470097264705e-06,
+      "loss": 0.4493,
+      "step": 1441
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.827326138171663,
+      "learning_rate": 9.871216303989694e-06,
+      "loss": 0.4058,
+      "step": 1442
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1963076323650452,
+      "learning_rate": 9.870962263662473e-06,
+      "loss": 0.4608,
+      "step": 1443
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.7662231431248516,
+      "learning_rate": 9.870707976295927e-06,
+      "loss": 0.423,
+      "step": 1444
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.9338748443434186,
+      "learning_rate": 9.870453441902951e-06,
+      "loss": 0.4516,
+      "step": 1445
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.459568560110359,
+      "learning_rate": 9.870198660496455e-06,
+      "loss": 0.4304,
+      "step": 1446
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.148913428796746,
+      "learning_rate": 9.869943632089361e-06,
+      "loss": 0.4358,
+      "step": 1447
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.5667494937013355,
+      "learning_rate": 9.869688356694602e-06,
+      "loss": 0.45,
+      "step": 1448
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.881103726139674,
+      "learning_rate": 9.869432834325125e-06,
+      "loss": 0.4389,
+      "step": 1449
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5323633947540873,
+      "learning_rate": 9.869177064993892e-06,
+      "loss": 0.4427,
+      "step": 1450
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.798193060596963,
+      "learning_rate": 9.868921048713872e-06,
+      "loss": 0.3984,
+      "step": 1451
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4134700445360007,
+      "learning_rate": 9.868664785498049e-06,
+      "loss": 0.1932,
+      "step": 1452
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6302894226995432,
+      "learning_rate": 9.868408275359422e-06,
+      "loss": 0.4364,
+      "step": 1453
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.318929624432891,
+      "learning_rate": 9.868151518311e-06,
+      "loss": 0.4486,
+      "step": 1454
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.443387484611594,
+      "learning_rate": 9.867894514365802e-06,
+      "loss": 0.4747,
+      "step": 1455
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.955783440152845,
+      "learning_rate": 9.867637263536866e-06,
+      "loss": 0.4316,
+      "step": 1456
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.4557830525208613,
+      "learning_rate": 9.867379765837237e-06,
+      "loss": 0.4248,
+      "step": 1457
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5133249578274666,
+      "learning_rate": 9.867122021279974e-06,
+      "loss": 0.3898,
+      "step": 1458
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.231639414920015,
+      "learning_rate": 9.86686402987815e-06,
+      "loss": 0.416,
+      "step": 1459
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.37417405350543465,
+      "learning_rate": 9.86660579164485e-06,
+      "loss": 0.1858,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5033047070093146,
+      "learning_rate": 9.86634730659317e-06,
+      "loss": 0.4226,
+      "step": 1461
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6054670728447684,
+      "learning_rate": 9.86608857473622e-06,
+      "loss": 0.4061,
+      "step": 1462
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.715477976127685,
+      "learning_rate": 9.865829596087121e-06,
+      "loss": 0.4019,
+      "step": 1463
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.008205862471925,
+      "learning_rate": 9.86557037065901e-06,
+      "loss": 0.4101,
+      "step": 1464
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5621119914753017,
+      "learning_rate": 9.865310898465031e-06,
+      "loss": 0.4073,
+      "step": 1465
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.155181032064002,
+      "learning_rate": 9.865051179518346e-06,
+      "loss": 0.4232,
+      "step": 1466
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.8143092975412607,
+      "learning_rate": 9.864791213832125e-06,
+      "loss": 0.4104,
+      "step": 1467
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 5.206599039078415,
+      "learning_rate": 9.864531001419554e-06,
+      "loss": 0.4423,
+      "step": 1468
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.760323331995137,
+      "learning_rate": 9.864270542293833e-06,
+      "loss": 0.4364,
+      "step": 1469
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.6661672141227406,
+      "learning_rate": 9.864009836468165e-06,
+      "loss": 0.3872,
+      "step": 1470
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6077184734675236,
+      "learning_rate": 9.863748883955776e-06,
+      "loss": 0.4211,
+      "step": 1471
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8627610824016703,
+      "learning_rate": 9.8634876847699e-06,
+      "loss": 0.4413,
+      "step": 1472
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.172032244162322,
+      "learning_rate": 9.863226238923786e-06,
+      "loss": 0.4592,
+      "step": 1473
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.970515443209019,
+      "learning_rate": 9.862964546430692e-06,
+      "loss": 0.4668,
+      "step": 1474
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1089593107397038,
+      "learning_rate": 9.86270260730389e-06,
+      "loss": 0.4603,
+      "step": 1475
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9409282058608497,
+      "learning_rate": 9.862440421556665e-06,
+      "loss": 0.4472,
+      "step": 1476
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 9.105071395695271,
+      "learning_rate": 9.862177989202315e-06,
+      "loss": 0.4275,
+      "step": 1477
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.373943624802557,
+      "learning_rate": 9.861915310254149e-06,
+      "loss": 0.4342,
+      "step": 1478
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7014415254291713,
+      "learning_rate": 9.861652384725488e-06,
+      "loss": 0.4082,
+      "step": 1479
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.3873193695921231,
+      "learning_rate": 9.861389212629668e-06,
+      "loss": 0.1988,
+      "step": 1480
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.945752583540824,
+      "learning_rate": 9.861125793980038e-06,
+      "loss": 0.4174,
+      "step": 1481
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.341133461336426,
+      "learning_rate": 9.860862128789954e-06,
+      "loss": 0.4746,
+      "step": 1482
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.4553490383594565,
+      "learning_rate": 9.86059821707279e-06,
+      "loss": 0.4449,
+      "step": 1483
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8174875011861498,
+      "learning_rate": 9.860334058841933e-06,
+      "loss": 0.4574,
+      "step": 1484
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.53537574941966,
+      "learning_rate": 9.860069654110776e-06,
+      "loss": 0.4657,
+      "step": 1485
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.959643460175519,
+      "learning_rate": 9.859805002892733e-06,
+      "loss": 0.4237,
+      "step": 1486
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.837493222058365,
+      "learning_rate": 9.859540105201221e-06,
+      "loss": 0.3984,
+      "step": 1487
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.6267164457541017,
+      "learning_rate": 9.85927496104968e-06,
+      "loss": 0.4434,
+      "step": 1488
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.5442644752885815,
+      "learning_rate": 9.859009570451555e-06,
+      "loss": 0.4718,
+      "step": 1489
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.678112666587531,
+      "learning_rate": 9.858743933420306e-06,
+      "loss": 0.4472,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.254753553242987,
+      "learning_rate": 9.858478049969405e-06,
+      "loss": 0.4225,
+      "step": 1491
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8539231467954824,
+      "learning_rate": 9.858211920112338e-06,
+      "loss": 0.4503,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.7287874483494052,
+      "learning_rate": 9.857945543862599e-06,
+      "loss": 0.4162,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5097514954940823,
+      "learning_rate": 9.8576789212337e-06,
+      "loss": 0.4217,
+      "step": 1494
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3813436603987985,
+      "learning_rate": 9.857412052239164e-06,
+      "loss": 0.4417,
+      "step": 1495
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.3208429189946935,
+      "learning_rate": 9.857144936892524e-06,
+      "loss": 0.4125,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.010858763314755,
+      "learning_rate": 9.85687757520733e-06,
+      "loss": 0.4361,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.3636910068223513,
+      "learning_rate": 9.856609967197137e-06,
+      "loss": 0.4142,
+      "step": 1498
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0000169077311822,
+      "learning_rate": 9.856342112875521e-06,
+      "loss": 0.4049,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.751750023734578,
+      "learning_rate": 9.856074012256065e-06,
+      "loss": 0.4466,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.66653638808876,
+      "learning_rate": 9.855805665352368e-06,
+      "loss": 0.4459,
+      "step": 1501
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 4.820455541189917,
+      "learning_rate": 9.85553707217804e-06,
+      "loss": 0.3978,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.772174806748058,
+      "learning_rate": 9.855268232746698e-06,
+      "loss": 0.4431,
+      "step": 1503
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.774315512903627,
+      "learning_rate": 9.854999147071983e-06,
+      "loss": 0.3894,
+      "step": 1504
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.42877365900714753,
+      "learning_rate": 9.854729815167537e-06,
+      "loss": 0.1849,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.539752480274535,
+      "learning_rate": 9.854460237047023e-06,
+      "loss": 0.4215,
+      "step": 1506
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.385630952645305,
+      "learning_rate": 9.854190412724114e-06,
+      "loss": 0.4238,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0214131954335257,
+      "learning_rate": 9.85392034221249e-06,
+      "loss": 0.3782,
+      "step": 1508
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.9525162407193237,
+      "learning_rate": 9.853650025525855e-06,
+      "loss": 0.441,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.713643332295143,
+      "learning_rate": 9.85337946267791e-06,
+      "loss": 0.4386,
+      "step": 1510
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4123816215327359,
+      "learning_rate": 9.853108653682384e-06,
+      "loss": 0.201,
+      "step": 1511
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.757281450803158,
+      "learning_rate": 9.85283759855301e-06,
+      "loss": 0.4332,
+      "step": 1512
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.058882383010419,
+      "learning_rate": 9.852566297303531e-06,
+      "loss": 0.4409,
+      "step": 1513
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3510848283516224,
+      "learning_rate": 9.852294749947712e-06,
+      "loss": 0.1843,
+      "step": 1514
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.389136185516389,
+      "learning_rate": 9.852022956499323e-06,
+      "loss": 0.4226,
+      "step": 1515
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.36720953796588734,
+      "learning_rate": 9.851750916972147e-06,
+      "loss": 0.1918,
+      "step": 1516
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.72109699848047,
+      "learning_rate": 9.851478631379982e-06,
+      "loss": 0.4482,
+      "step": 1517
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0010442757498694,
+      "learning_rate": 9.851206099736638e-06,
+      "loss": 0.4446,
+      "step": 1518
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.260090996216169,
+      "learning_rate": 9.850933322055938e-06,
+      "loss": 0.4588,
+      "step": 1519
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.698571647713963,
+      "learning_rate": 9.850660298351714e-06,
+      "loss": 0.4457,
+      "step": 1520
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.979396656059586,
+      "learning_rate": 9.850387028637812e-06,
+      "loss": 0.4311,
+      "step": 1521
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2990956988449653,
+      "learning_rate": 9.850113512928094e-06,
+      "loss": 0.4013,
+      "step": 1522
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.37135049067502,
+      "learning_rate": 9.849839751236431e-06,
+      "loss": 0.419,
+      "step": 1523
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.267435106905391,
+      "learning_rate": 9.849565743576708e-06,
+      "loss": 0.4195,
+      "step": 1524
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.311545674672463,
+      "learning_rate": 9.84929148996282e-06,
+      "loss": 0.425,
+      "step": 1525
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.344056375707935,
+      "learning_rate": 9.849016990408678e-06,
+      "loss": 0.4616,
+      "step": 1526
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0353920476547547,
+      "learning_rate": 9.848742244928202e-06,
+      "loss": 0.4656,
+      "step": 1527
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.211840524559711,
+      "learning_rate": 9.848467253535327e-06,
+      "loss": 0.4141,
+      "step": 1528
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2471169117922205,
+      "learning_rate": 9.848192016244001e-06,
+      "loss": 0.3958,
+      "step": 1529
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.570566883670183,
+      "learning_rate": 9.847916533068183e-06,
+      "loss": 0.4019,
+      "step": 1530
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7899014704397134,
+      "learning_rate": 9.84764080402184e-06,
+      "loss": 0.4165,
+      "step": 1531
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7208298725829017,
+      "learning_rate": 9.847364829118963e-06,
+      "loss": 0.3932,
+      "step": 1532
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4851314296655453,
+      "learning_rate": 9.847088608373544e-06,
+      "loss": 0.4675,
+      "step": 1533
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.458183853944245,
+      "learning_rate": 9.846812141799592e-06,
+      "loss": 0.4665,
+      "step": 1534
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0553084250901605,
+      "learning_rate": 9.84653542941113e-06,
+      "loss": 0.3998,
+      "step": 1535
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.9345882733822037,
+      "learning_rate": 9.846258471222194e-06,
+      "loss": 0.4524,
+      "step": 1536
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.285474893255596,
+      "learning_rate": 9.845981267246826e-06,
+      "loss": 0.4161,
+      "step": 1537
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.253272112336449,
+      "learning_rate": 9.845703817499086e-06,
+      "loss": 0.4154,
+      "step": 1538
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.54541378219394,
+      "learning_rate": 9.84542612199305e-06,
+      "loss": 0.4368,
+      "step": 1539
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.109765083784128,
+      "learning_rate": 9.845148180742794e-06,
+      "loss": 0.4264,
+      "step": 1540
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.4632681155251746,
+      "learning_rate": 9.84486999376242e-06,
+      "loss": 0.4825,
+      "step": 1541
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4391497804442636,
+      "learning_rate": 9.844591561066035e-06,
+      "loss": 0.1974,
+      "step": 1542
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.5768569461817235,
+      "learning_rate": 9.844312882667762e-06,
+      "loss": 0.4101,
+      "step": 1543
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 12.304105662817761,
+      "learning_rate": 9.84403395858173e-06,
+      "loss": 0.4146,
+      "step": 1544
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3435948850147565,
+      "learning_rate": 9.84375478882209e-06,
+      "loss": 0.4061,
+      "step": 1545
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5633053125233216,
+      "learning_rate": 9.843475373402997e-06,
+      "loss": 0.4272,
+      "step": 1546
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.0858824462114764,
+      "learning_rate": 9.843195712338624e-06,
+      "loss": 0.4251,
+      "step": 1547
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.819791063765534,
+      "learning_rate": 9.842915805643156e-06,
+      "loss": 0.4415,
+      "step": 1548
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.029087327231807,
+      "learning_rate": 9.842635653330787e-06,
+      "loss": 0.4624,
+      "step": 1549
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.111178652107101,
+      "learning_rate": 9.842355255415724e-06,
+      "loss": 0.4533,
+      "step": 1550
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8971734918676546,
+      "learning_rate": 9.84207461191219e-06,
+      "loss": 0.4226,
+      "step": 1551
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.0104727898619865,
+      "learning_rate": 9.84179372283442e-06,
+      "loss": 0.4296,
+      "step": 1552
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4054739389124779,
+      "learning_rate": 9.841512588196653e-06,
+      "loss": 0.1909,
+      "step": 1553
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.837839366173351,
+      "learning_rate": 9.841231208013155e-06,
+      "loss": 0.448,
+      "step": 1554
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.95670563914984,
+      "learning_rate": 9.840949582298193e-06,
+      "loss": 0.4476,
+      "step": 1555
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.627316893850428,
+      "learning_rate": 9.84066771106605e-06,
+      "loss": 0.4061,
+      "step": 1556
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.5892253121426947,
+      "learning_rate": 9.840385594331022e-06,
+      "loss": 0.4289,
+      "step": 1557
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.0255509396199125,
+      "learning_rate": 9.840103232107418e-06,
+      "loss": 0.4488,
+      "step": 1558
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.778465438292273,
+      "learning_rate": 9.839820624409558e-06,
+      "loss": 0.4185,
+      "step": 1559
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.488914703395598,
+      "learning_rate": 9.839537771251773e-06,
+      "loss": 0.4435,
+      "step": 1560
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.139305766060006,
+      "learning_rate": 9.839254672648411e-06,
+      "loss": 0.4157,
+      "step": 1561
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.39778715149488303,
+      "learning_rate": 9.838971328613829e-06,
+      "loss": 0.192,
+      "step": 1562
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.0685268957512544,
+      "learning_rate": 9.838687739162397e-06,
+      "loss": 0.4597,
+      "step": 1563
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1452379781785234,
+      "learning_rate": 9.838403904308497e-06,
+      "loss": 0.403,
+      "step": 1564
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.878654288981148,
+      "learning_rate": 9.838119824066528e-06,
+      "loss": 0.4139,
+      "step": 1565
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.39111794992111,
+      "learning_rate": 9.837835498450892e-06,
+      "loss": 0.4223,
+      "step": 1566
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9109611594258387,
+      "learning_rate": 9.837550927476012e-06,
+      "loss": 0.4241,
+      "step": 1567
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.765427802264075,
+      "learning_rate": 9.837266111156322e-06,
+      "loss": 0.4486,
+      "step": 1568
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.206194225122574,
+      "learning_rate": 9.836981049506264e-06,
+      "loss": 0.4582,
+      "step": 1569
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9185986926269205,
+      "learning_rate": 9.836695742540296e-06,
+      "loss": 0.4377,
+      "step": 1570
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.824846871172513,
+      "learning_rate": 9.836410190272889e-06,
+      "loss": 0.4145,
+      "step": 1571
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.543069422297324,
+      "learning_rate": 9.836124392718526e-06,
+      "loss": 0.4611,
+      "step": 1572
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1927337257093247,
+      "learning_rate": 9.8358383498917e-06,
+      "loss": 0.4388,
+      "step": 1573
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 7.865098181560225,
+      "learning_rate": 9.835552061806917e-06,
+      "loss": 0.4307,
+      "step": 1574
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.853621067015978,
+      "learning_rate": 9.835265528478699e-06,
+      "loss": 0.4566,
+      "step": 1575
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.590639849409993,
+      "learning_rate": 9.834978749921578e-06,
+      "loss": 0.4727,
+      "step": 1576
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.775287927239054,
+      "learning_rate": 9.834691726150097e-06,
+      "loss": 0.423,
+      "step": 1577
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.927752620060652,
+      "learning_rate": 9.834404457178814e-06,
+      "loss": 0.4427,
+      "step": 1578
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.876934925911743,
+      "learning_rate": 9.834116943022299e-06,
+      "loss": 0.4439,
+      "step": 1579
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.0806091193320375,
+      "learning_rate": 9.83382918369513e-06,
+      "loss": 0.3891,
+      "step": 1580
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6097829171420477,
+      "learning_rate": 9.833541179211906e-06,
+      "loss": 0.3994,
+      "step": 1581
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.68682413092953,
+      "learning_rate": 9.833252929587231e-06,
+      "loss": 0.4193,
+      "step": 1582
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.015341501103128,
+      "learning_rate": 9.832964434835726e-06,
+      "loss": 0.4266,
+      "step": 1583
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.858733037578786,
+      "learning_rate": 9.83267569497202e-06,
+      "loss": 0.4774,
+      "step": 1584
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.093568980436586,
+      "learning_rate": 9.832386710010758e-06,
+      "loss": 0.4202,
+      "step": 1585
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.729055140652726,
+      "learning_rate": 9.832097479966599e-06,
+      "loss": 0.4195,
+      "step": 1586
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.2706612446519867,
+      "learning_rate": 9.831808004854207e-06,
+      "loss": 0.4341,
+      "step": 1587
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7912192758837766,
+      "learning_rate": 9.831518284688265e-06,
+      "loss": 0.4351,
+      "step": 1588
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.68159499457458,
+      "learning_rate": 9.831228319483469e-06,
+      "loss": 0.4227,
+      "step": 1589
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.146593145726965,
+      "learning_rate": 9.830938109254523e-06,
+      "loss": 0.4256,
+      "step": 1590
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.648277159645017,
+      "learning_rate": 9.830647654016143e-06,
+      "loss": 0.4043,
+      "step": 1591
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.2778381491293915,
+      "learning_rate": 9.830356953783068e-06,
+      "loss": 0.4435,
+      "step": 1592
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.888094090610566,
+      "learning_rate": 9.830066008570032e-06,
+      "loss": 0.4347,
+      "step": 1593
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.511560295751105,
+      "learning_rate": 9.829774818391794e-06,
+      "loss": 0.4489,
+      "step": 1594
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5103533201954282,
+      "learning_rate": 9.829483383263128e-06,
+      "loss": 0.4244,
+      "step": 1595
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.094876451215981,
+      "learning_rate": 9.829191703198805e-06,
+      "loss": 0.4254,
+      "step": 1596
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.1436523404268195,
+      "learning_rate": 9.828899778213628e-06,
+      "loss": 0.445,
+      "step": 1597
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.154084401599859,
+      "learning_rate": 9.828607608322393e-06,
+      "loss": 0.4253,
+      "step": 1598
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4041579943129077,
+      "learning_rate": 9.828315193539925e-06,
+      "loss": 0.1944,
+      "step": 1599
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.600230626783794,
+      "learning_rate": 9.82802253388105e-06,
+      "loss": 0.4482,
+      "step": 1600
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.3209057870285528,
+      "learning_rate": 9.827729629360613e-06,
+      "loss": 0.4509,
+      "step": 1601
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.34942651062491864,
+      "learning_rate": 9.827436479993468e-06,
+      "loss": 0.1815,
+      "step": 1602
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 17.880964524428794,
+      "learning_rate": 9.827143085794484e-06,
+      "loss": 0.442,
+      "step": 1603
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.022084707986386,
+      "learning_rate": 9.826849446778539e-06,
+      "loss": 0.4303,
+      "step": 1604
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.16600826146723,
+      "learning_rate": 9.826555562960526e-06,
+      "loss": 0.4559,
+      "step": 1605
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.320449145152882,
+      "learning_rate": 9.826261434355352e-06,
+      "loss": 0.4305,
+      "step": 1606
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.766188979688829,
+      "learning_rate": 9.825967060977933e-06,
+      "loss": 0.4632,
+      "step": 1607
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.385905088554126,
+      "learning_rate": 9.825672442843197e-06,
+      "loss": 0.4398,
+      "step": 1608
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.7089648690655914,
+      "learning_rate": 9.825377579966086e-06,
+      "loss": 0.445,
+      "step": 1609
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5646684113012976,
+      "learning_rate": 9.825082472361558e-06,
+      "loss": 0.4315,
+      "step": 1610
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.9466185981909017,
+      "learning_rate": 9.824787120044576e-06,
+      "loss": 0.4009,
+      "step": 1611
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5682644633446796,
+      "learning_rate": 9.824491523030123e-06,
+      "loss": 0.4179,
+      "step": 1612
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.413544987084189,
+      "learning_rate": 9.824195681333187e-06,
+      "loss": 0.4104,
+      "step": 1613
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.287234645975878,
+      "learning_rate": 9.823899594968776e-06,
+      "loss": 0.4321,
+      "step": 1614
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5290259823353782,
+      "learning_rate": 9.823603263951902e-06,
+      "loss": 0.4456,
+      "step": 1615
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.2440084434826995,
+      "learning_rate": 9.823306688297595e-06,
+      "loss": 0.4168,
+      "step": 1616
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.0273086364390425,
+      "learning_rate": 9.823009868020901e-06,
+      "loss": 0.4149,
+      "step": 1617
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.168899590736704,
+      "learning_rate": 9.822712803136869e-06,
+      "loss": 0.415,
+      "step": 1618
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.343132380364082,
+      "learning_rate": 9.822415493660567e-06,
+      "loss": 0.4429,
+      "step": 1619
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.409106798597856,
+      "learning_rate": 9.822117939607072e-06,
+      "loss": 0.4519,
+      "step": 1620
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 21.294344136374924,
+      "learning_rate": 9.821820140991478e-06,
+      "loss": 0.4288,
+      "step": 1621
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.070708197780658,
+      "learning_rate": 9.821522097828884e-06,
+      "loss": 0.4545,
+      "step": 1622
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.644226185962365,
+      "learning_rate": 9.82122381013441e-06,
+      "loss": 0.4157,
+      "step": 1623
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.085501622992438,
+      "learning_rate": 9.820925277923183e-06,
+      "loss": 0.396,
+      "step": 1624
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 10.386360016151755,
+      "learning_rate": 9.820626501210342e-06,
+      "loss": 0.4501,
+      "step": 1625
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4548362191000783,
+      "learning_rate": 9.82032748001104e-06,
+      "loss": 0.415,
+      "step": 1626
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.43668020029292287,
+      "learning_rate": 9.820028214340446e-06,
+      "loss": 0.1955,
+      "step": 1627
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.8043339341165927,
+      "learning_rate": 9.819728704213733e-06,
+      "loss": 0.4611,
+      "step": 1628
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.565628845138884,
+      "learning_rate": 9.819428949646094e-06,
+      "loss": 0.4606,
+      "step": 1629
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.592332431063105,
+      "learning_rate": 9.819128950652733e-06,
+      "loss": 0.4349,
+      "step": 1630
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.3683428483997673,
+      "learning_rate": 9.81882870724886e-06,
+      "loss": 0.1924,
+      "step": 1631
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 11.093165024680529,
+      "learning_rate": 9.818528219449705e-06,
+      "loss": 0.4507,
+      "step": 1632
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6983791684119414,
+      "learning_rate": 9.818227487270508e-06,
+      "loss": 0.4101,
+      "step": 1633
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4155778265176515,
+      "learning_rate": 9.817926510726522e-06,
+      "loss": 0.4333,
+      "step": 1634
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.22194836124524,
+      "learning_rate": 9.817625289833013e-06,
+      "loss": 0.447,
+      "step": 1635
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.446611447453024,
+      "learning_rate": 9.817323824605254e-06,
+      "loss": 0.4125,
+      "step": 1636
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.063187411737793,
+      "learning_rate": 9.817022115058534e-06,
+      "loss": 0.4304,
+      "step": 1637
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 6.012874079056184,
+      "learning_rate": 9.81672016120816e-06,
+      "loss": 0.4499,
+      "step": 1638
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7431316532305705,
+      "learning_rate": 9.81641796306944e-06,
+      "loss": 0.4237,
+      "step": 1639
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.6754014852578476,
+      "learning_rate": 9.816115520657705e-06,
+      "loss": 0.4517,
+      "step": 1640
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.45734557053622005,
+      "learning_rate": 9.815812833988292e-06,
+      "loss": 0.1886,
+      "step": 1641
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.53023421913002,
+      "learning_rate": 9.815509903076553e-06,
+      "loss": 0.4478,
+      "step": 1642
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.431411779093751,
+      "learning_rate": 9.81520672793785e-06,
+      "loss": 0.4365,
+      "step": 1643
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.539184611267868,
+      "learning_rate": 9.81490330858756e-06,
+      "loss": 0.4385,
+      "step": 1644
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.6358336372375066,
+      "learning_rate": 9.814599645041071e-06,
+      "loss": 0.4025,
+      "step": 1645
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.522591650135087,
+      "learning_rate": 9.814295737313786e-06,
+      "loss": 0.4188,
+      "step": 1646
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.778115402195751,
+      "learning_rate": 9.813991585421118e-06,
+      "loss": 0.3843,
+      "step": 1647
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.848966575304247,
+      "learning_rate": 9.81368718937849e-06,
+      "loss": 0.4096,
+      "step": 1648
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.0388607065582365,
+      "learning_rate": 9.81338254920134e-06,
+      "loss": 0.3834,
+      "step": 1649
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.166320260949917,
+      "learning_rate": 9.813077664905122e-06,
+      "loss": 0.4238,
+      "step": 1650
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.7751507697045175,
+      "learning_rate": 9.812772536505296e-06,
+      "loss": 0.4215,
+      "step": 1651
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.4201002697521314,
+      "learning_rate": 9.812467164017337e-06,
+      "loss": 0.4089,
+      "step": 1652
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.453156132904249,
+      "learning_rate": 9.81216154745673e-06,
+      "loss": 0.4544,
+      "step": 1653
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.301599427243765,
+      "learning_rate": 9.811855686838981e-06,
+      "loss": 0.4388,
+      "step": 1654
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.721639361597199,
+      "learning_rate": 9.811549582179599e-06,
+      "loss": 0.4454,
+      "step": 1655
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.003272786109217,
+      "learning_rate": 9.811243233494108e-06,
+      "loss": 0.4109,
+      "step": 1656
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5611867028889876,
+      "learning_rate": 9.810936640798046e-06,
+      "loss": 0.436,
+      "step": 1657
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4583710650113515,
+      "learning_rate": 9.810629804106964e-06,
+      "loss": 0.4048,
+      "step": 1658
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.318566221259358,
+      "learning_rate": 9.810322723436419e-06,
+      "loss": 0.4596,
+      "step": 1659
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.309395536843011,
+      "learning_rate": 9.810015398801989e-06,
+      "loss": 0.4463,
+      "step": 1660
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.138349992228137,
+      "learning_rate": 9.80970783021926e-06,
+      "loss": 0.4694,
+      "step": 1661
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8850090609222843,
+      "learning_rate": 9.80940001770383e-06,
+      "loss": 0.4284,
+      "step": 1662
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.9427537183107395,
+      "learning_rate": 9.809091961271311e-06,
+      "loss": 0.4378,
+      "step": 1663
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.655392634482903,
+      "learning_rate": 9.808783660937325e-06,
+      "loss": 0.4513,
+      "step": 1664
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5232353459008283,
+      "learning_rate": 9.80847511671751e-06,
+      "loss": 0.1928,
+      "step": 1665
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1453189545407865,
+      "learning_rate": 9.808166328627515e-06,
+      "loss": 0.4419,
+      "step": 1666
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.434258139210199,
+      "learning_rate": 9.807857296682999e-06,
+      "loss": 0.4263,
+      "step": 1667
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.273050609515693,
+      "learning_rate": 9.807548020899634e-06,
+      "loss": 0.4344,
+      "step": 1668
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.375183825997632,
+      "learning_rate": 9.807238501293108e-06,
+      "loss": 0.3962,
+      "step": 1669
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.25554634280617,
+      "learning_rate": 9.806928737879116e-06,
+      "loss": 0.4638,
+      "step": 1670
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.043551826191814,
+      "learning_rate": 9.806618730673371e-06,
+      "loss": 0.3912,
+      "step": 1671
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.7406666766012147,
+      "learning_rate": 9.806308479691595e-06,
+      "loss": 0.4315,
+      "step": 1672
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.65254769694183,
+      "learning_rate": 9.805997984949522e-06,
+      "loss": 0.4108,
+      "step": 1673
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4725749048128147,
+      "learning_rate": 9.8056872464629e-06,
+      "loss": 0.4053,
+      "step": 1674
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5373983226901777,
+      "learning_rate": 9.805376264247488e-06,
+      "loss": 0.4385,
+      "step": 1675
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.769911597758844,
+      "learning_rate": 9.805065038319057e-06,
+      "loss": 0.4594,
+      "step": 1676
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.323000657552079,
+      "learning_rate": 9.804753568693395e-06,
+      "loss": 0.4515,
+      "step": 1677
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.538462691572563,
+      "learning_rate": 9.804441855386296e-06,
+      "loss": 0.414,
+      "step": 1678
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2208350893078377,
+      "learning_rate": 9.804129898413569e-06,
+      "loss": 0.4704,
+      "step": 1679
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.55156934526264,
+      "learning_rate": 9.803817697791033e-06,
+      "loss": 0.4045,
+      "step": 1680
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9692700624572703,
+      "learning_rate": 9.80350525353453e-06,
+      "loss": 0.3969,
+      "step": 1681
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.524809353987789,
+      "learning_rate": 9.803192565659898e-06,
+      "loss": 0.4185,
+      "step": 1682
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.753004098538356,
+      "learning_rate": 9.802879634182998e-06,
+      "loss": 0.4435,
+      "step": 1683
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.92611731501989,
+      "learning_rate": 9.802566459119703e-06,
+      "loss": 0.4733,
+      "step": 1684
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.34511112499839,
+      "learning_rate": 9.802253040485892e-06,
+      "loss": 0.4304,
+      "step": 1685
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.56860768078146,
+      "learning_rate": 9.801939378297464e-06,
+      "loss": 0.4426,
+      "step": 1686
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5408712817781502,
+      "learning_rate": 9.801625472570326e-06,
+      "loss": 0.462,
+      "step": 1687
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.058489379238367,
+      "learning_rate": 9.8013113233204e-06,
+      "loss": 0.4254,
+      "step": 1688
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7145447320432435,
+      "learning_rate": 9.800996930563615e-06,
+      "loss": 0.2112,
+      "step": 1689
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5300544452706017,
+      "learning_rate": 9.800682294315917e-06,
+      "loss": 0.1995,
+      "step": 1690
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.46802894952802,
+      "learning_rate": 9.800367414593266e-06,
+      "loss": 0.3801,
+      "step": 1691
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6570004901538313,
+      "learning_rate": 9.80005229141163e-06,
+      "loss": 0.4307,
+      "step": 1692
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2881219278375093,
+      "learning_rate": 9.79973692478699e-06,
+      "loss": 0.4165,
+      "step": 1693
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.615683491541322,
+      "learning_rate": 9.799421314735341e-06,
+      "loss": 0.4175,
+      "step": 1694
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5524891447511697,
+      "learning_rate": 9.79910546127269e-06,
+      "loss": 0.449,
+      "step": 1695
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.365402326551262,
+      "learning_rate": 9.798789364415056e-06,
+      "loss": 0.4666,
+      "step": 1696
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.220330429206987,
+      "learning_rate": 9.798473024178471e-06,
+      "loss": 0.3904,
+      "step": 1697
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.023472594441508,
+      "learning_rate": 9.798156440578977e-06,
+      "loss": 0.4257,
+      "step": 1698
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.71546721812304,
+      "learning_rate": 9.797839613632632e-06,
+      "loss": 0.4147,
+      "step": 1699
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.618981618388146,
+      "learning_rate": 9.797522543355504e-06,
+      "loss": 0.4092,
+      "step": 1700
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.686287347881566,
+      "learning_rate": 9.797205229763674e-06,
+      "loss": 0.3917,
+      "step": 1701
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2870737495235023,
+      "learning_rate": 9.796887672873233e-06,
+      "loss": 0.4223,
+      "step": 1702
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5615119172326586,
+      "learning_rate": 9.796569872700287e-06,
+      "loss": 0.4023,
+      "step": 1703
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.06700329167484,
+      "learning_rate": 9.796251829260958e-06,
+      "loss": 0.4228,
+      "step": 1704
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.893876516534122,
+      "learning_rate": 9.795933542571372e-06,
+      "loss": 0.4282,
+      "step": 1705
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.603988691741243,
+      "learning_rate": 9.79561501264767e-06,
+      "loss": 0.4402,
+      "step": 1706
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.469923686872543,
+      "learning_rate": 9.795296239506011e-06,
+      "loss": 0.4341,
+      "step": 1707
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 57.70000636705232,
+      "learning_rate": 9.79497722316256e-06,
+      "loss": 0.46,
+      "step": 1708
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.073548150752569,
+      "learning_rate": 9.794657963633497e-06,
+      "loss": 0.4271,
+      "step": 1709
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4044646614342566,
+      "learning_rate": 9.794338460935014e-06,
+      "loss": 0.3375,
+      "step": 1710
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.5929459502564223,
+      "learning_rate": 9.794018715083312e-06,
+      "loss": 0.3944,
+      "step": 1711
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.101299286642421,
+      "learning_rate": 9.793698726094613e-06,
+      "loss": 0.4391,
+      "step": 1712
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.392539729754144,
+      "learning_rate": 9.793378493985144e-06,
+      "loss": 0.4192,
+      "step": 1713
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.035316013701524,
+      "learning_rate": 9.793058018771142e-06,
+      "loss": 0.4716,
+      "step": 1714
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.013790625841869,
+      "learning_rate": 9.792737300468865e-06,
+      "loss": 0.3938,
+      "step": 1715
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.025162163945536,
+      "learning_rate": 9.792416339094577e-06,
+      "loss": 0.4197,
+      "step": 1716
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.668951174150466,
+      "learning_rate": 9.792095134664556e-06,
+      "loss": 0.4179,
+      "step": 1717
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.41213321504429,
+      "learning_rate": 9.791773687195094e-06,
+      "loss": 0.4355,
+      "step": 1718
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 10.05463352664172,
+      "learning_rate": 9.791451996702492e-06,
+      "loss": 0.4059,
+      "step": 1719
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.911595991522413,
+      "learning_rate": 9.791130063203066e-06,
+      "loss": 0.4674,
+      "step": 1720
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.5093218298334317,
+      "learning_rate": 9.790807886713145e-06,
+      "loss": 0.4076,
+      "step": 1721
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.459656641979224,
+      "learning_rate": 9.790485467249065e-06,
+      "loss": 0.3651,
+      "step": 1722
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.8737598151514474,
+      "learning_rate": 9.790162804827179e-06,
+      "loss": 0.3855,
+      "step": 1723
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.847999905211759,
+      "learning_rate": 9.789839899463854e-06,
+      "loss": 0.4106,
+      "step": 1724
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.495725341735586,
+      "learning_rate": 9.789516751175466e-06,
+      "loss": 0.45,
+      "step": 1725
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.747811237040192,
+      "learning_rate": 9.7891933599784e-06,
+      "loss": 0.4454,
+      "step": 1726
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.9041222650492537,
+      "learning_rate": 9.788869725889062e-06,
+      "loss": 0.4259,
+      "step": 1727
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 8.160708017054626,
+      "learning_rate": 9.788545848923865e-06,
+      "loss": 0.4612,
+      "step": 1728
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 19.003008822131818,
+      "learning_rate": 9.788221729099232e-06,
+      "loss": 0.4368,
+      "step": 1729
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1835675147632556,
+      "learning_rate": 9.787897366431605e-06,
+      "loss": 0.4113,
+      "step": 1730
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.662141961601872,
+      "learning_rate": 9.787572760937435e-06,
+      "loss": 0.4215,
+      "step": 1731
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4004112432417415,
+      "learning_rate": 9.78724791263318e-06,
+      "loss": 0.4308,
+      "step": 1732
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.843725536009387,
+      "learning_rate": 9.786922821535317e-06,
+      "loss": 0.4617,
+      "step": 1733
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.2875862244151435,
+      "learning_rate": 9.786597487660336e-06,
+      "loss": 0.4424,
+      "step": 1734
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.990159324948382,
+      "learning_rate": 9.786271911024737e-06,
+      "loss": 0.4165,
+      "step": 1735
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8157260041138295,
+      "learning_rate": 9.785946091645031e-06,
+      "loss": 0.417,
+      "step": 1736
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.30903837127934,
+      "learning_rate": 9.785620029537741e-06,
+      "loss": 0.4562,
+      "step": 1737
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.811531433760126,
+      "learning_rate": 9.785293724719404e-06,
+      "loss": 0.4454,
+      "step": 1738
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2309477393633985,
+      "learning_rate": 9.784967177206572e-06,
+      "loss": 0.4084,
+      "step": 1739
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.883505545824626,
+      "learning_rate": 9.784640387015805e-06,
+      "loss": 0.408,
+      "step": 1740
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1051991455378656,
+      "learning_rate": 9.784313354163677e-06,
+      "loss": 0.4489,
+      "step": 1741
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.693215977731212,
+      "learning_rate": 9.783986078666772e-06,
+      "loss": 0.4116,
+      "step": 1742
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.631186942170326,
+      "learning_rate": 9.783658560541689e-06,
+      "loss": 0.4628,
+      "step": 1743
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.871145820394486,
+      "learning_rate": 9.78333079980504e-06,
+      "loss": 0.408,
+      "step": 1744
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.081393638670145,
+      "learning_rate": 9.783002796473449e-06,
+      "loss": 0.4172,
+      "step": 1745
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0945910511839125,
+      "learning_rate": 9.782674550563548e-06,
+      "loss": 0.3795,
+      "step": 1746
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.873738665270679,
+      "learning_rate": 9.782346062091988e-06,
+      "loss": 0.4467,
+      "step": 1747
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.2977236680315674,
+      "learning_rate": 9.782017331075425e-06,
+      "loss": 0.4343,
+      "step": 1748
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.124236244140499,
+      "learning_rate": 9.781688357530534e-06,
+      "loss": 0.4131,
+      "step": 1749
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.404419863000247,
+      "learning_rate": 9.781359141473998e-06,
+      "loss": 0.424,
+      "step": 1750
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.297418939933673,
+      "learning_rate": 9.781029682922515e-06,
+      "loss": 0.425,
+      "step": 1751
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.6695574747832325,
+      "learning_rate": 9.780699981892793e-06,
+      "loss": 0.4032,
+      "step": 1752
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.603812285783971,
+      "learning_rate": 9.780370038401557e-06,
+      "loss": 0.4485,
+      "step": 1753
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.170568131909116,
+      "learning_rate": 9.780039852465533e-06,
+      "loss": 0.439,
+      "step": 1754
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.046956028467354,
+      "learning_rate": 9.779709424101474e-06,
+      "loss": 0.4135,
+      "step": 1755
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.4063236060665814,
+      "learning_rate": 9.779378753326136e-06,
+      "loss": 0.4288,
+      "step": 1756
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.244205492921262,
+      "learning_rate": 9.779047840156288e-06,
+      "loss": 0.3977,
+      "step": 1757
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.886539885672464,
+      "learning_rate": 9.778716684608715e-06,
+      "loss": 0.3802,
+      "step": 1758
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.854511090952523,
+      "learning_rate": 9.778385286700212e-06,
+      "loss": 0.4137,
+      "step": 1759
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.240993648540068,
+      "learning_rate": 9.778053646447585e-06,
+      "loss": 0.4281,
+      "step": 1760
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6159482955890196,
+      "learning_rate": 9.777721763867653e-06,
+      "loss": 0.2396,
+      "step": 1761
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5344005443983365,
+      "learning_rate": 9.777389638977251e-06,
+      "loss": 0.2165,
+      "step": 1762
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.687966334307013,
+      "learning_rate": 9.777057271793222e-06,
+      "loss": 0.3938,
+      "step": 1763
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.8834144915715743,
+      "learning_rate": 9.776724662332422e-06,
+      "loss": 0.4553,
+      "step": 1764
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.3581200404361795,
+      "learning_rate": 9.776391810611719e-06,
+      "loss": 0.1903,
+      "step": 1765
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.567532690358918,
+      "learning_rate": 9.776058716647996e-06,
+      "loss": 0.4254,
+      "step": 1766
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.705341978298066,
+      "learning_rate": 9.775725380458145e-06,
+      "loss": 0.4091,
+      "step": 1767
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.559717826472271,
+      "learning_rate": 9.775391802059074e-06,
+      "loss": 0.4121,
+      "step": 1768
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.977967649930481,
+      "learning_rate": 9.775057981467699e-06,
+      "loss": 0.4716,
+      "step": 1769
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 7.278007285879158,
+      "learning_rate": 9.77472391870095e-06,
+      "loss": 0.4625,
+      "step": 1770
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.009355837203735,
+      "learning_rate": 9.774389613775772e-06,
+      "loss": 0.477,
+      "step": 1771
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6153202118585279,
+      "learning_rate": 9.774055066709117e-06,
+      "loss": 0.2189,
+      "step": 1772
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.758805375296945,
+      "learning_rate": 9.773720277517954e-06,
+      "loss": 0.4178,
+      "step": 1773
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.078512762675106,
+      "learning_rate": 9.77338524621926e-06,
+      "loss": 0.4242,
+      "step": 1774
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3540910305484934,
+      "learning_rate": 9.773049972830029e-06,
+      "loss": 0.4213,
+      "step": 1775
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.466002756244284,
+      "learning_rate": 9.772714457367265e-06,
+      "loss": 0.4351,
+      "step": 1776
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.372050291135736,
+      "learning_rate": 9.772378699847984e-06,
+      "loss": 0.4328,
+      "step": 1777
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.863475367728371,
+      "learning_rate": 9.772042700289214e-06,
+      "loss": 0.4495,
+      "step": 1778
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.581391919390761,
+      "learning_rate": 9.771706458707995e-06,
+      "loss": 0.4415,
+      "step": 1779
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.917605744105319,
+      "learning_rate": 9.771369975121383e-06,
+      "loss": 0.4039,
+      "step": 1780
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.45151829444185,
+      "learning_rate": 9.77103324954644e-06,
+      "loss": 0.4309,
+      "step": 1781
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.840298294447079,
+      "learning_rate": 9.770696282000245e-06,
+      "loss": 0.4423,
+      "step": 1782
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.0036265602849483,
+      "learning_rate": 9.77035907249989e-06,
+      "loss": 0.4139,
+      "step": 1783
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.6413465466355714,
+      "learning_rate": 9.77002162106247e-06,
+      "loss": 0.4895,
+      "step": 1784
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.202058502801711,
+      "learning_rate": 9.76968392770511e-06,
+      "loss": 0.4184,
+      "step": 1785
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.414137056053964,
+      "learning_rate": 9.76934599244493e-06,
+      "loss": 0.4388,
+      "step": 1786
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.876829211520033,
+      "learning_rate": 9.769007815299068e-06,
+      "loss": 0.4133,
+      "step": 1787
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.717028602294696,
+      "learning_rate": 9.76866939628468e-06,
+      "loss": 0.4053,
+      "step": 1788
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.174071413359127,
+      "learning_rate": 9.768330735418924e-06,
+      "loss": 0.4105,
+      "step": 1789
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 9.247180501430384,
+      "learning_rate": 9.767991832718982e-06,
+      "loss": 0.4176,
+      "step": 1790
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.337654017219103,
+      "learning_rate": 9.767652688202037e-06,
+      "loss": 0.4546,
+      "step": 1791
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.599009388318706,
+      "learning_rate": 9.767313301885291e-06,
+      "loss": 0.437,
+      "step": 1792
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 17.693737363891945,
+      "learning_rate": 9.766973673785957e-06,
+      "loss": 0.4204,
+      "step": 1793
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 5.353307446850516,
+      "learning_rate": 9.76663380392126e-06,
+      "loss": 0.4137,
+      "step": 1794
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.14339014377872,
+      "learning_rate": 9.766293692308436e-06,
+      "loss": 0.4122,
+      "step": 1795
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1804072470183997,
+      "learning_rate": 9.765953338964736e-06,
+      "loss": 0.4235,
+      "step": 1796
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.3457292899676347,
+      "learning_rate": 9.76561274390742e-06,
+      "loss": 0.3987,
+      "step": 1797
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46845724971757635,
+      "learning_rate": 9.765271907153762e-06,
+      "loss": 0.2001,
+      "step": 1798
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9156942705693623,
+      "learning_rate": 9.76493082872105e-06,
+      "loss": 0.4395,
+      "step": 1799
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.8482674952615925,
+      "learning_rate": 9.76458950862658e-06,
+      "loss": 0.4109,
+      "step": 1800
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.4721491827244595,
+      "learning_rate": 9.764247946887661e-06,
+      "loss": 0.4483,
+      "step": 1801
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8265889083140694,
+      "learning_rate": 9.763906143521621e-06,
+      "loss": 0.4598,
+      "step": 1802
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.760682039449551,
+      "learning_rate": 9.763564098545792e-06,
+      "loss": 0.3962,
+      "step": 1803
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5407172236063653,
+      "learning_rate": 9.763221811977522e-06,
+      "loss": 0.4277,
+      "step": 1804
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.806567349678776,
+      "learning_rate": 9.762879283834173e-06,
+      "loss": 0.455,
+      "step": 1805
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5545589548989622,
+      "learning_rate": 9.762536514133112e-06,
+      "loss": 0.4226,
+      "step": 1806
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 55.22584189316968,
+      "learning_rate": 9.762193502891726e-06,
+      "loss": 0.4443,
+      "step": 1807
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.008126587127299,
+      "learning_rate": 9.761850250127412e-06,
+      "loss": 0.4142,
+      "step": 1808
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.805982443405227,
+      "learning_rate": 9.761506755857576e-06,
+      "loss": 0.478,
+      "step": 1809
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9052202913358043,
+      "learning_rate": 9.761163020099642e-06,
+      "loss": 0.409,
+      "step": 1810
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.945329461583586,
+      "learning_rate": 9.760819042871043e-06,
+      "loss": 0.4198,
+      "step": 1811
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.286862192621361,
+      "learning_rate": 9.760474824189222e-06,
+      "loss": 0.4078,
+      "step": 1812
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2956264277485423,
+      "learning_rate": 9.760130364071639e-06,
+      "loss": 0.4477,
+      "step": 1813
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.734774295865657,
+      "learning_rate": 9.759785662535761e-06,
+      "loss": 0.4075,
+      "step": 1814
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.193367896671881,
+      "learning_rate": 9.759440719599074e-06,
+      "loss": 0.4266,
+      "step": 1815
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.224036319221474,
+      "learning_rate": 9.75909553527907e-06,
+      "loss": 0.3882,
+      "step": 1816
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.161141473708103,
+      "learning_rate": 9.758750109593255e-06,
+      "loss": 0.4152,
+      "step": 1817
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.788522929805673,
+      "learning_rate": 9.758404442559149e-06,
+      "loss": 0.4789,
+      "step": 1818
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7973588580816493,
+      "learning_rate": 9.758058534194283e-06,
+      "loss": 0.4447,
+      "step": 1819
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.587626939657292,
+      "learning_rate": 9.757712384516201e-06,
+      "loss": 0.4184,
+      "step": 1820
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3290211857835708,
+      "learning_rate": 9.757365993542457e-06,
+      "loss": 0.4103,
+      "step": 1821
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.571323140014648,
+      "learning_rate": 9.757019361290621e-06,
+      "loss": 0.4561,
+      "step": 1822
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4665567977245244,
+      "learning_rate": 9.75667248777827e-06,
+      "loss": 0.4285,
+      "step": 1823
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8564379776655566,
+      "learning_rate": 9.756325373023002e-06,
+      "loss": 0.4163,
+      "step": 1824
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.088610846354025,
+      "learning_rate": 9.755978017042414e-06,
+      "loss": 0.4275,
+      "step": 1825
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3587301913809453,
+      "learning_rate": 9.755630419854126e-06,
+      "loss": 0.4545,
+      "step": 1826
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9847459504602964,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.4327,
+      "step": 1827
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.078494677553509,
+      "learning_rate": 9.754934501924982e-06,
+      "loss": 0.41,
+      "step": 1828
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5478730098902521,
+      "learning_rate": 9.754586181219419e-06,
+      "loss": 0.2237,
+      "step": 1829
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6448751619752375,
+      "learning_rate": 9.754237619376746e-06,
+      "loss": 0.4562,
+      "step": 1830
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3436062206955275,
+      "learning_rate": 9.753888816414641e-06,
+      "loss": 0.4507,
+      "step": 1831
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.4975377156818452,
+      "learning_rate": 9.753539772350792e-06,
+      "loss": 0.3987,
+      "step": 1832
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9350601321468277,
+      "learning_rate": 9.753190487202903e-06,
+      "loss": 0.459,
+      "step": 1833
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.622520654234611,
+      "learning_rate": 9.752840960988693e-06,
+      "loss": 0.464,
+      "step": 1834
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.445741437452843,
+      "learning_rate": 9.75249119372588e-06,
+      "loss": 0.4282,
+      "step": 1835
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.997569761713137,
+      "learning_rate": 9.75214118543221e-06,
+      "loss": 0.4475,
+      "step": 1836
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.619938818650272,
+      "learning_rate": 9.751790936125431e-06,
+      "loss": 0.4345,
+      "step": 1837
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.767177561786431,
+      "learning_rate": 9.751440445823309e-06,
+      "loss": 0.4112,
+      "step": 1838
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7122599600628177,
+      "learning_rate": 9.751089714543615e-06,
+      "loss": 0.4605,
+      "step": 1839
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.859368008247097,
+      "learning_rate": 9.750738742304144e-06,
+      "loss": 0.3933,
+      "step": 1840
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.489741112189442,
+      "learning_rate": 9.750387529122688e-06,
+      "loss": 0.4222,
+      "step": 1841
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.423641871472593,
+      "learning_rate": 9.750036075017068e-06,
+      "loss": 0.4395,
+      "step": 1842
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0855474986883737,
+      "learning_rate": 9.749684380005101e-06,
+      "loss": 0.4457,
+      "step": 1843
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.0740255725186145,
+      "learning_rate": 9.749332444104628e-06,
+      "loss": 0.4405,
+      "step": 1844
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1233228369185997,
+      "learning_rate": 9.7489802673335e-06,
+      "loss": 0.4139,
+      "step": 1845
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8476436738246886,
+      "learning_rate": 9.748627849709572e-06,
+      "loss": 0.4162,
+      "step": 1846
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.4108419418851454,
+      "learning_rate": 9.748275191250722e-06,
+      "loss": 0.4665,
+      "step": 1847
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7633363968282834,
+      "learning_rate": 9.747922291974834e-06,
+      "loss": 0.4265,
+      "step": 1848
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7086895866122638,
+      "learning_rate": 9.747569151899807e-06,
+      "loss": 0.4333,
+      "step": 1849
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.9658954429985505,
+      "learning_rate": 9.747215771043551e-06,
+      "loss": 0.4293,
+      "step": 1850
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.4072622796158205,
+      "learning_rate": 9.746862149423989e-06,
+      "loss": 0.4533,
+      "step": 1851
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.127448895585339,
+      "learning_rate": 9.746508287059052e-06,
+      "loss": 0.4413,
+      "step": 1852
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.433188943423877,
+      "learning_rate": 9.74615418396669e-06,
+      "loss": 0.4473,
+      "step": 1853
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.7065250984705775,
+      "learning_rate": 9.74579984016486e-06,
+      "loss": 0.4201,
+      "step": 1854
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.2788539893720845,
+      "learning_rate": 9.745445255671538e-06,
+      "loss": 0.3998,
+      "step": 1855
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6360838313412474,
+      "learning_rate": 9.7450904305047e-06,
+      "loss": 0.2318,
+      "step": 1856
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.054080798396606,
+      "learning_rate": 9.744735364682347e-06,
+      "loss": 0.4328,
+      "step": 1857
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.721873235957679,
+      "learning_rate": 9.744380058222483e-06,
+      "loss": 0.454,
+      "step": 1858
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.803609769580776,
+      "learning_rate": 9.744024511143132e-06,
+      "loss": 0.4156,
+      "step": 1859
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.138096648755742,
+      "learning_rate": 9.743668723462325e-06,
+      "loss": 0.4333,
+      "step": 1860
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.964177944436524,
+      "learning_rate": 9.743312695198102e-06,
+      "loss": 0.4348,
+      "step": 1861
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1041027550295475,
+      "learning_rate": 9.742956426368526e-06,
+      "loss": 0.3983,
+      "step": 1862
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.418369495624715,
+      "learning_rate": 9.742599916991663e-06,
+      "loss": 0.4542,
+      "step": 1863
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.8350872640605056,
+      "learning_rate": 9.742243167085593e-06,
+      "loss": 0.4513,
+      "step": 1864
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.527098288212603,
+      "learning_rate": 9.741886176668412e-06,
+      "loss": 0.4305,
+      "step": 1865
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7080174091665747,
+      "learning_rate": 9.74152894575822e-06,
+      "loss": 0.4497,
+      "step": 1866
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.350167402469041,
+      "learning_rate": 9.741171474373141e-06,
+      "loss": 0.4687,
+      "step": 1867
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.1637081452674645,
+      "learning_rate": 9.740813762531301e-06,
+      "loss": 0.4148,
+      "step": 1868
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8421865746639545,
+      "learning_rate": 9.740455810250844e-06,
+      "loss": 0.4248,
+      "step": 1869
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.260097891064686,
+      "learning_rate": 9.740097617549921e-06,
+      "loss": 0.4308,
+      "step": 1870
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.728398145168935,
+      "learning_rate": 9.739739184446702e-06,
+      "loss": 0.4562,
+      "step": 1871
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.682817741780763,
+      "learning_rate": 9.739380510959365e-06,
+      "loss": 0.4134,
+      "step": 1872
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6527275059292108,
+      "learning_rate": 9.739021597106098e-06,
+      "loss": 0.2195,
+      "step": 1873
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.627290778547735,
+      "learning_rate": 9.738662442905105e-06,
+      "loss": 0.3667,
+      "step": 1874
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2555153888063653,
+      "learning_rate": 9.738303048374605e-06,
+      "loss": 0.471,
+      "step": 1875
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9673893711053987,
+      "learning_rate": 9.73794341353282e-06,
+      "loss": 0.4315,
+      "step": 1876
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8761648108880022,
+      "learning_rate": 9.73758353839799e-06,
+      "loss": 0.4367,
+      "step": 1877
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5034588247723626,
+      "learning_rate": 9.73722342298837e-06,
+      "loss": 0.3876,
+      "step": 1878
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.9372370928711184,
+      "learning_rate": 9.736863067322223e-06,
+      "loss": 0.4521,
+      "step": 1879
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.731312885679788,
+      "learning_rate": 9.736502471417823e-06,
+      "loss": 0.4251,
+      "step": 1880
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.782468675791109,
+      "learning_rate": 9.73614163529346e-06,
+      "loss": 0.4504,
+      "step": 1881
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.643554441805601,
+      "learning_rate": 9.735780558967434e-06,
+      "loss": 0.4294,
+      "step": 1882
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7567657102193848,
+      "learning_rate": 9.735419242458059e-06,
+      "loss": 0.4051,
+      "step": 1883
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3040164824672202,
+      "learning_rate": 9.735057685783656e-06,
+      "loss": 0.4047,
+      "step": 1884
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3490341078793158,
+      "learning_rate": 9.734695888962567e-06,
+      "loss": 0.4151,
+      "step": 1885
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.368101946141865,
+      "learning_rate": 9.734333852013136e-06,
+      "loss": 0.4076,
+      "step": 1886
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.0823378990831367,
+      "learning_rate": 9.733971574953726e-06,
+      "loss": 0.4415,
+      "step": 1887
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.661188546151563,
+      "learning_rate": 9.733609057802715e-06,
+      "loss": 0.4348,
+      "step": 1888
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.885304936847872,
+      "learning_rate": 9.733246300578482e-06,
+      "loss": 0.4414,
+      "step": 1889
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5274055503280521,
+      "learning_rate": 9.73288330329943e-06,
+      "loss": 0.2011,
+      "step": 1890
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.254781395577418,
+      "learning_rate": 9.732520065983967e-06,
+      "loss": 0.4197,
+      "step": 1891
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.025686060785087,
+      "learning_rate": 9.732156588650511e-06,
+      "loss": 0.3885,
+      "step": 1892
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.705963892453821,
+      "learning_rate": 9.731792871317504e-06,
+      "loss": 0.4096,
+      "step": 1893
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.016874595282583,
+      "learning_rate": 9.73142891400339e-06,
+      "loss": 0.4658,
+      "step": 1894
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3393139834174501,
+      "learning_rate": 9.731064716726626e-06,
+      "loss": 0.1908,
+      "step": 1895
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.524100781002577,
+      "learning_rate": 9.730700279505683e-06,
+      "loss": 0.4496,
+      "step": 1896
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5551132454271146,
+      "learning_rate": 9.730335602359044e-06,
+      "loss": 0.4544,
+      "step": 1897
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.476123370121122,
+      "learning_rate": 9.729970685305207e-06,
+      "loss": 0.3975,
+      "step": 1898
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.9040028322408387,
+      "learning_rate": 9.729605528362677e-06,
+      "loss": 0.3921,
+      "step": 1899
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.319029478225494,
+      "learning_rate": 9.729240131549973e-06,
+      "loss": 0.4241,
+      "step": 1900
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.6700713576214365,
+      "learning_rate": 9.72887449488563e-06,
+      "loss": 0.4054,
+      "step": 1901
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.914731168625475,
+      "learning_rate": 9.728508618388186e-06,
+      "loss": 0.4462,
+      "step": 1902
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.1374565999522765,
+      "learning_rate": 9.728142502076204e-06,
+      "loss": 0.4344,
+      "step": 1903
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.2772680184196865,
+      "learning_rate": 9.727776145968246e-06,
+      "loss": 0.4202,
+      "step": 1904
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.401245693044898,
+      "learning_rate": 9.727409550082899e-06,
+      "loss": 0.3955,
+      "step": 1905
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.369470934313828,
+      "learning_rate": 9.727042714438749e-06,
+      "loss": 0.4753,
+      "step": 1906
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.34122782648653,
+      "learning_rate": 9.726675639054403e-06,
+      "loss": 0.4484,
+      "step": 1907
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5066781943848885,
+      "learning_rate": 9.726308323948481e-06,
+      "loss": 0.195,
+      "step": 1908
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.0001947606079105,
+      "learning_rate": 9.725940769139608e-06,
+      "loss": 0.4179,
+      "step": 1909
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7235320331140898,
+      "learning_rate": 9.725572974646427e-06,
+      "loss": 0.432,
+      "step": 1910
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 9.087652018172758,
+      "learning_rate": 9.725204940487592e-06,
+      "loss": 0.38,
+      "step": 1911
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 6.734992963133724,
+      "learning_rate": 9.724836666681764e-06,
+      "loss": 0.4046,
+      "step": 1912
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3747937501911074,
+      "learning_rate": 9.724468153247627e-06,
+      "loss": 0.4262,
+      "step": 1913
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.2242490397646253,
+      "learning_rate": 9.724099400203868e-06,
+      "loss": 0.4172,
+      "step": 1914
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.6600974012341356,
+      "learning_rate": 9.723730407569188e-06,
+      "loss": 0.4014,
+      "step": 1915
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.461042570525083,
+      "learning_rate": 9.723361175362302e-06,
+      "loss": 0.4158,
+      "step": 1916
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3994686557280668,
+      "learning_rate": 9.722991703601936e-06,
+      "loss": 0.1938,
+      "step": 1917
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.862245272924053,
+      "learning_rate": 9.72262199230683e-06,
+      "loss": 0.4637,
+      "step": 1918
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.8588234171679305,
+      "learning_rate": 9.722252041495733e-06,
+      "loss": 0.4404,
+      "step": 1919
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.1632517236589206,
+      "learning_rate": 9.721881851187406e-06,
+      "loss": 0.4165,
+      "step": 1920
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.3675226013046123,
+      "learning_rate": 9.721511421400626e-06,
+      "loss": 0.1912,
+      "step": 1921
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.604213045970537,
+      "learning_rate": 9.721140752154182e-06,
+      "loss": 0.4198,
+      "step": 1922
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 7.36216141524642,
+      "learning_rate": 9.720769843466868e-06,
+      "loss": 0.416,
+      "step": 1923
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.935159536319073,
+      "learning_rate": 9.7203986953575e-06,
+      "loss": 0.4227,
+      "step": 1924
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.840978797555273,
+      "learning_rate": 9.7200273078449e-06,
+      "loss": 0.47,
+      "step": 1925
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.0658238807224345,
+      "learning_rate": 9.7196556809479e-06,
+      "loss": 0.441,
+      "step": 1926
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.513253793729869,
+      "learning_rate": 9.719283814685354e-06,
+      "loss": 0.4662,
+      "step": 1927
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.073580733641338,
+      "learning_rate": 9.718911709076118e-06,
+      "loss": 0.47,
+      "step": 1928
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.7548014965949763,
+      "learning_rate": 9.718539364139066e-06,
+      "loss": 0.4252,
+      "step": 1929
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5197292252742636,
+      "learning_rate": 9.718166779893077e-06,
+      "loss": 0.4334,
+      "step": 1930
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.00682749583975,
+      "learning_rate": 9.717793956357055e-06,
+      "loss": 0.4327,
+      "step": 1931
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.23406678963495,
+      "learning_rate": 9.717420893549902e-06,
+      "loss": 0.414,
+      "step": 1932
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.444954806762024,
+      "learning_rate": 9.717047591490541e-06,
+      "loss": 0.4131,
+      "step": 1933
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6479447138336645,
+      "learning_rate": 9.716674050197905e-06,
+      "loss": 0.4114,
+      "step": 1934
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5890781218066685,
+      "learning_rate": 9.716300269690939e-06,
+      "loss": 0.4122,
+      "step": 1935
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.534900526114226,
+      "learning_rate": 9.715926249988599e-06,
+      "loss": 0.4209,
+      "step": 1936
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.958924227591897,
+      "learning_rate": 9.715551991109855e-06,
+      "loss": 0.4465,
+      "step": 1937
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.239848493029939,
+      "learning_rate": 9.715177493073686e-06,
+      "loss": 0.4661,
+      "step": 1938
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.5517526907921875,
+      "learning_rate": 9.71480275589909e-06,
+      "loss": 0.4362,
+      "step": 1939
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 4.331416409591004,
+      "learning_rate": 9.714427779605066e-06,
+      "loss": 0.4539,
+      "step": 1940
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.8624309517114637,
+      "learning_rate": 9.714052564210636e-06,
+      "loss": 0.4439,
+      "step": 1941
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.281704835805374,
+      "learning_rate": 9.71367710973483e-06,
+      "loss": 0.4184,
+      "step": 1942
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.524588811148899,
+      "learning_rate": 9.713301416196686e-06,
+      "loss": 0.419,
+      "step": 1943
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2771450047152717,
+      "learning_rate": 9.71292548361526e-06,
+      "loss": 0.4381,
+      "step": 1944
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.923295718780752,
+      "learning_rate": 9.712549312009622e-06,
+      "loss": 0.4713,
+      "step": 1945
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.023521991431972,
+      "learning_rate": 9.712172901398844e-06,
+      "loss": 0.3908,
+      "step": 1946
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7157691807999145,
+      "learning_rate": 9.71179625180202e-06,
+      "loss": 0.486,
+      "step": 1947
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.361887678150912,
+      "learning_rate": 9.711419363238252e-06,
+      "loss": 0.3952,
+      "step": 1948
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.599707564013279,
+      "learning_rate": 9.711042235726652e-06,
+      "loss": 0.4112,
+      "step": 1949
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.519805425312636,
+      "learning_rate": 9.710664869286351e-06,
+      "loss": 0.4199,
+      "step": 1950
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3273849443249577,
+      "learning_rate": 9.710287263936485e-06,
+      "loss": 0.4377,
+      "step": 1951
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.114682716743529,
+      "learning_rate": 9.709909419696204e-06,
+      "loss": 0.3843,
+      "step": 1952
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.0628958925959617,
+      "learning_rate": 9.709531336584675e-06,
+      "loss": 0.4212,
+      "step": 1953
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.333006953265754,
+      "learning_rate": 9.709153014621066e-06,
+      "loss": 0.4237,
+      "step": 1954
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.208532635533771,
+      "learning_rate": 9.708774453824573e-06,
+      "loss": 0.4252,
+      "step": 1955
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4416069940806575,
+      "learning_rate": 9.70839565421439e-06,
+      "loss": 0.4488,
+      "step": 1956
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 43.51594643741771,
+      "learning_rate": 9.70801661580973e-06,
+      "loss": 0.4262,
+      "step": 1957
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.413202147644786,
+      "learning_rate": 9.707637338629816e-06,
+      "loss": 0.4367,
+      "step": 1958
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.889877836146757,
+      "learning_rate": 9.707257822693883e-06,
+      "loss": 0.4372,
+      "step": 1959
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4436141361399632,
+      "learning_rate": 9.70687806802118e-06,
+      "loss": 0.2071,
+      "step": 1960
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.224191122660243,
+      "learning_rate": 9.706498074630967e-06,
+      "loss": 0.4073,
+      "step": 1961
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.0464126740517194,
+      "learning_rate": 9.706117842542517e-06,
+      "loss": 0.4297,
+      "step": 1962
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6597415092755563,
+      "learning_rate": 9.705737371775112e-06,
+      "loss": 0.4334,
+      "step": 1963
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.295677720401901,
+      "learning_rate": 9.705356662348048e-06,
+      "loss": 0.3926,
+      "step": 1964
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.4427889781432155,
+      "learning_rate": 9.704975714280636e-06,
+      "loss": 0.4313,
+      "step": 1965
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.975453342452855,
+      "learning_rate": 9.704594527592195e-06,
+      "loss": 0.3971,
+      "step": 1966
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.3443209464959285,
+      "learning_rate": 9.704213102302056e-06,
+      "loss": 0.4316,
+      "step": 1967
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9128211623903018,
+      "learning_rate": 9.703831438429564e-06,
+      "loss": 0.4078,
+      "step": 1968
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9135009619391523,
+      "learning_rate": 9.70344953599408e-06,
+      "loss": 0.4334,
+      "step": 1969
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5910627628188276,
+      "learning_rate": 9.703067395014969e-06,
+      "loss": 0.3971,
+      "step": 1970
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.5149237495411474,
+      "learning_rate": 9.702685015511611e-06,
+      "loss": 0.4137,
+      "step": 1971
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.892602927996236,
+      "learning_rate": 9.702302397503401e-06,
+      "loss": 0.4124,
+      "step": 1972
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.45665042788628774,
+      "learning_rate": 9.701919541009745e-06,
+      "loss": 0.1997,
+      "step": 1973
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.292565901606779,
+      "learning_rate": 9.701536446050058e-06,
+      "loss": 0.4444,
+      "step": 1974
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.202605394121794,
+      "learning_rate": 9.70115311264377e-06,
+      "loss": 0.4362,
+      "step": 1975
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.560815252684841,
+      "learning_rate": 9.700769540810324e-06,
+      "loss": 0.4202,
+      "step": 1976
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7157569143844893,
+      "learning_rate": 9.700385730569171e-06,
+      "loss": 0.42,
+      "step": 1977
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.39721154294964417,
+      "learning_rate": 9.70000168193978e-06,
+      "loss": 0.1705,
+      "step": 1978
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.141486950764716,
+      "learning_rate": 9.699617394941623e-06,
+      "loss": 0.383,
+      "step": 1979
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.267165401555559,
+      "learning_rate": 9.699232869594196e-06,
+      "loss": 0.4202,
+      "step": 1980
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.824224558081431,
+      "learning_rate": 9.698848105916998e-06,
+      "loss": 0.4189,
+      "step": 1981
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1743578539784725,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.4044,
+      "step": 1982
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.8527996172447496,
+      "learning_rate": 9.698077863651356e-06,
+      "loss": 0.409,
+      "step": 1983
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.221614344996864,
+      "learning_rate": 9.697692385101978e-06,
+      "loss": 0.4561,
+      "step": 1984
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.210080191375841,
+      "learning_rate": 9.697306668300958e-06,
+      "loss": 0.3911,
+      "step": 1985
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.5246105702263035,
+      "learning_rate": 9.696920713267856e-06,
+      "loss": 0.4444,
+      "step": 1986
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.745310358524645,
+      "learning_rate": 9.69653452002225e-06,
+      "loss": 0.436,
+      "step": 1987
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.06563252324837,
+      "learning_rate": 9.696148088583724e-06,
+      "loss": 0.3992,
+      "step": 1988
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.66825870522148,
+      "learning_rate": 9.695761418971875e-06,
+      "loss": 0.4536,
+      "step": 1989
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9724089794607833,
+      "learning_rate": 9.69537451120632e-06,
+      "loss": 0.4141,
+      "step": 1990
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.646237176067811,
+      "learning_rate": 9.694987365306676e-06,
+      "loss": 0.4434,
+      "step": 1991
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.956756025464975,
+      "learning_rate": 9.694599981292578e-06,
+      "loss": 0.3995,
+      "step": 1992
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.949354936554114,
+      "learning_rate": 9.694212359183676e-06,
+      "loss": 0.4402,
+      "step": 1993
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.213156668518619,
+      "learning_rate": 9.693824498999626e-06,
+      "loss": 0.4122,
+      "step": 1994
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8332059923225854,
+      "learning_rate": 9.693436400760101e-06,
+      "loss": 0.4139,
+      "step": 1995
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.950629551728867,
+      "learning_rate": 9.693048064484782e-06,
+      "loss": 0.4283,
+      "step": 1996
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.881122089244038,
+      "learning_rate": 9.692659490193366e-06,
+      "loss": 0.3923,
+      "step": 1997
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.398090798480296,
+      "learning_rate": 9.69227067790556e-06,
+      "loss": 0.4513,
+      "step": 1998
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.42576699505152,
+      "learning_rate": 9.691881627641083e-06,
+      "loss": 0.4768,
+      "step": 1999
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6281597933013643,
+      "learning_rate": 9.691492339419666e-06,
+      "loss": 0.4036,
+      "step": 2000
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.304418912113632,
+      "learning_rate": 9.691102813261052e-06,
+      "loss": 0.4448,
+      "step": 2001
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.54167987734115,
+      "learning_rate": 9.690713049184998e-06,
+      "loss": 0.4256,
+      "step": 2002
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49447394468972117,
+      "learning_rate": 9.690323047211273e-06,
+      "loss": 0.1942,
+      "step": 2003
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.627529159261454,
+      "learning_rate": 9.68993280735965e-06,
+      "loss": 0.4181,
+      "step": 2004
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8543749566947816,
+      "learning_rate": 9.689542329649928e-06,
+      "loss": 0.4341,
+      "step": 2005
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.91702050778977,
+      "learning_rate": 9.689151614101905e-06,
+      "loss": 0.3864,
+      "step": 2006
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 9.775627562042342,
+      "learning_rate": 9.688760660735403e-06,
+      "loss": 0.4416,
+      "step": 2007
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.222128244177447,
+      "learning_rate": 9.688369469570245e-06,
+      "loss": 0.4118,
+      "step": 2008
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9284854120097314,
+      "learning_rate": 9.687978040626273e-06,
+      "loss": 0.452,
+      "step": 2009
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.295245629615492,
+      "learning_rate": 9.687586373923339e-06,
+      "loss": 0.4061,
+      "step": 2010
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1505689043604725,
+      "learning_rate": 9.687194469481304e-06,
+      "loss": 0.4113,
+      "step": 2011
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.695434607648616,
+      "learning_rate": 9.686802327320049e-06,
+      "loss": 0.3932,
+      "step": 2012
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8333609416265175,
+      "learning_rate": 9.68640994745946e-06,
+      "loss": 0.3925,
+      "step": 2013
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.826432609718342,
+      "learning_rate": 9.686017329919435e-06,
+      "loss": 0.4544,
+      "step": 2014
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.102213666254364,
+      "learning_rate": 9.685624474719888e-06,
+      "loss": 0.4182,
+      "step": 2015
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.611183961537692,
+      "learning_rate": 9.685231381880747e-06,
+      "loss": 0.4314,
+      "step": 2016
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6404131909841575,
+      "learning_rate": 9.68483805142194e-06,
+      "loss": 0.4553,
+      "step": 2017
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.758455012668206,
+      "learning_rate": 9.684444483363424e-06,
+      "loss": 0.4184,
+      "step": 2018
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5144842415980853,
+      "learning_rate": 9.684050677725157e-06,
+      "loss": 0.2197,
+      "step": 2019
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.7210678940933755,
+      "learning_rate": 9.683656634527109e-06,
+      "loss": 0.4153,
+      "step": 2020
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.948528518164944,
+      "learning_rate": 9.683262353789266e-06,
+      "loss": 0.4171,
+      "step": 2021
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.905210866478491,
+      "learning_rate": 9.682867835531624e-06,
+      "loss": 0.4602,
+      "step": 2022
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2197964106328096,
+      "learning_rate": 9.682473079774194e-06,
+      "loss": 0.4278,
+      "step": 2023
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.518343190951148,
+      "learning_rate": 9.682078086536994e-06,
+      "loss": 0.4111,
+      "step": 2024
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.764899776640584,
+      "learning_rate": 9.681682855840057e-06,
+      "loss": 0.4161,
+      "step": 2025
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.2257522187598617,
+      "learning_rate": 9.68128738770343e-06,
+      "loss": 0.3977,
+      "step": 2026
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4179929242597987,
+      "learning_rate": 9.68089168214717e-06,
+      "loss": 0.1841,
+      "step": 2027
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9553104621454556,
+      "learning_rate": 9.680495739191343e-06,
+      "loss": 0.4503,
+      "step": 2028
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.16756950790231,
+      "learning_rate": 9.680099558856031e-06,
+      "loss": 0.46,
+      "step": 2029
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.699661409150518,
+      "learning_rate": 9.679703141161327e-06,
+      "loss": 0.4311,
+      "step": 2030
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.888200998329191,
+      "learning_rate": 9.679306486127339e-06,
+      "loss": 0.4361,
+      "step": 2031
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.898030193549815,
+      "learning_rate": 9.67890959377418e-06,
+      "loss": 0.4186,
+      "step": 2032
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.170449615815803,
+      "learning_rate": 9.678512464121978e-06,
+      "loss": 0.4372,
+      "step": 2033
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.126352066530889,
+      "learning_rate": 9.67811509719088e-06,
+      "loss": 0.4416,
+      "step": 2034
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.93424104547389,
+      "learning_rate": 9.677717493001035e-06,
+      "loss": 0.4256,
+      "step": 2035
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.1745459755724035,
+      "learning_rate": 9.677319651572607e-06,
+      "loss": 0.4317,
+      "step": 2036
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8264340891621633,
+      "learning_rate": 9.676921572925777e-06,
+      "loss": 0.447,
+      "step": 2037
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.9069474221219123,
+      "learning_rate": 9.676523257080733e-06,
+      "loss": 0.4035,
+      "step": 2038
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4547842644316673,
+      "learning_rate": 9.676124704057675e-06,
+      "loss": 0.3988,
+      "step": 2039
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4109700268703053,
+      "learning_rate": 9.675725913876816e-06,
+      "loss": 0.1984,
+      "step": 2040
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.756161755742205,
+      "learning_rate": 9.675326886558384e-06,
+      "loss": 0.4467,
+      "step": 2041
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1319010394740916,
+      "learning_rate": 9.674927622122615e-06,
+      "loss": 0.4071,
+      "step": 2042
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3510455815864259,
+      "learning_rate": 9.674528120589757e-06,
+      "loss": 0.1892,
+      "step": 2043
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.079762681446131,
+      "learning_rate": 9.674128381980073e-06,
+      "loss": 0.4393,
+      "step": 2044
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.4234157949295465,
+      "learning_rate": 9.673728406313836e-06,
+      "loss": 0.444,
+      "step": 2045
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 18.44175316488595,
+      "learning_rate": 9.673328193611331e-06,
+      "loss": 0.4543,
+      "step": 2046
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.306175348594615,
+      "learning_rate": 9.672927743892856e-06,
+      "loss": 0.4368,
+      "step": 2047
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.665172181630232,
+      "learning_rate": 9.672527057178722e-06,
+      "loss": 0.4498,
+      "step": 2048
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.3754793674476926,
+      "learning_rate": 9.672126133489248e-06,
+      "loss": 0.454,
+      "step": 2049
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.642206303429718,
+      "learning_rate": 9.671724972844767e-06,
+      "loss": 0.4112,
+      "step": 2050
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.222700395745871,
+      "learning_rate": 9.671323575265628e-06,
+      "loss": 0.4221,
+      "step": 2051
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.429944947664622,
+      "learning_rate": 9.670921940772186e-06,
+      "loss": 0.3986,
+      "step": 2052
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.571411361121704,
+      "learning_rate": 9.67052006938481e-06,
+      "loss": 0.4237,
+      "step": 2053
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.009482010071301,
+      "learning_rate": 9.670117961123885e-06,
+      "loss": 0.4352,
+      "step": 2054
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.368757921535206,
+      "learning_rate": 9.6697156160098e-06,
+      "loss": 0.389,
+      "step": 2055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.3896813489624547,
+      "learning_rate": 9.669313034062965e-06,
+      "loss": 0.2008,
+      "step": 2056
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.311218425561208,
+      "learning_rate": 9.668910215303797e-06,
+      "loss": 0.4085,
+      "step": 2057
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.205309367640686,
+      "learning_rate": 9.668507159752723e-06,
+      "loss": 0.3794,
+      "step": 2058
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.878219466659039,
+      "learning_rate": 9.668103867430185e-06,
+      "loss": 0.4516,
+      "step": 2059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.965138994516143,
+      "learning_rate": 9.667700338356639e-06,
+      "loss": 0.4152,
+      "step": 2060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.014185425581324,
+      "learning_rate": 9.66729657255255e-06,
+      "loss": 0.4378,
+      "step": 2061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.1163786106970512,
+      "learning_rate": 9.666892570038393e-06,
+      "loss": 0.4059,
+      "step": 2062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.249815616004288,
+      "learning_rate": 9.66648833083466e-06,
+      "loss": 0.4195,
+      "step": 2063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.732241287589123,
+      "learning_rate": 9.666083854961854e-06,
+      "loss": 0.4433,
+      "step": 2064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.819641982394309,
+      "learning_rate": 9.665679142440488e-06,
+      "loss": 0.4428,
+      "step": 2065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.40303851933935614,
+      "learning_rate": 9.665274193291085e-06,
+      "loss": 0.2111,
+      "step": 2066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.362900943011466,
+      "learning_rate": 9.664869007534185e-06,
+      "loss": 0.4378,
+      "step": 2067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.407412493534397,
+      "learning_rate": 9.664463585190338e-06,
+      "loss": 0.4446,
+      "step": 2068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6524696050436862,
+      "learning_rate": 9.664057926280103e-06,
+      "loss": 0.4281,
+      "step": 2069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.702243620895761,
+      "learning_rate": 9.663652030824058e-06,
+      "loss": 0.4226,
+      "step": 2070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.889760231798988,
+      "learning_rate": 9.663245898842785e-06,
+      "loss": 0.4271,
+      "step": 2071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.224754104746096,
+      "learning_rate": 9.662839530356883e-06,
+      "loss": 0.4578,
+      "step": 2072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 7.973305156205782,
+      "learning_rate": 9.662432925386963e-06,
+      "loss": 0.4554,
+      "step": 2073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.385082546899468,
+      "learning_rate": 9.662026083953645e-06,
+      "loss": 0.395,
+      "step": 2074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.9781182730766935,
+      "learning_rate": 9.661619006077562e-06,
+      "loss": 0.4477,
+      "step": 2075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 13.729440124552402,
+      "learning_rate": 9.661211691779362e-06,
+      "loss": 0.4141,
+      "step": 2076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.7735823882178545,
+      "learning_rate": 9.6608041410797e-06,
+      "loss": 0.4202,
+      "step": 2077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.202039239129145,
+      "learning_rate": 9.660396353999249e-06,
+      "loss": 0.4173,
+      "step": 2078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 12.250496185571361,
+      "learning_rate": 9.659988330558686e-06,
+      "loss": 0.413,
+      "step": 2079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.035437089644244,
+      "learning_rate": 9.659580070778708e-06,
+      "loss": 0.4173,
+      "step": 2080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 5.71328088316059,
+      "learning_rate": 9.659171574680023e-06,
+      "loss": 0.4141,
+      "step": 2081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.8955949008089927,
+      "learning_rate": 9.658762842283343e-06,
+      "loss": 0.4014,
+      "step": 2082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.6659985238563557,
+      "learning_rate": 9.6583538736094e-06,
+      "loss": 0.4248,
+      "step": 2083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 8.230493137686702,
+      "learning_rate": 9.657944668678937e-06,
+      "loss": 0.4078,
+      "step": 2084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 6.753030571168581,
+      "learning_rate": 9.657535227512705e-06,
+      "loss": 0.4086,
+      "step": 2085
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3828353709018439,
+      "learning_rate": 9.657125550131471e-06,
+      "loss": 0.2023,
+      "step": 2086
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.2159149448395787,
+      "learning_rate": 9.656715636556012e-06,
+      "loss": 0.4512,
+      "step": 2087
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.513729330388177,
+      "learning_rate": 9.656305486807119e-06,
+      "loss": 0.4033,
+      "step": 2088
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.372494515276379,
+      "learning_rate": 9.65589510090559e-06,
+      "loss": 0.4203,
+      "step": 2089
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.25103944811376,
+      "learning_rate": 9.65548447887224e-06,
+      "loss": 0.4274,
+      "step": 2090
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.3061107537251075,
+      "learning_rate": 9.655073620727898e-06,
+      "loss": 0.4557,
+      "step": 2091
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.308719270306247,
+      "learning_rate": 9.654662526493397e-06,
+      "loss": 0.3996,
+      "step": 2092
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.632517393025309,
+      "learning_rate": 9.654251196189588e-06,
+      "loss": 0.4154,
+      "step": 2093
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.712445289891215,
+      "learning_rate": 9.653839629837332e-06,
+      "loss": 0.4123,
+      "step": 2094
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.626249462387227,
+      "learning_rate": 9.653427827457503e-06,
+      "loss": 0.4025,
+      "step": 2095
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.526026290505726,
+      "learning_rate": 9.653015789070982e-06,
+      "loss": 0.4476,
+      "step": 2096
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.271216560033576,
+      "learning_rate": 9.652603514698674e-06,
+      "loss": 0.4373,
+      "step": 2097
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.02247230567852,
+      "learning_rate": 9.652191004361482e-06,
+      "loss": 0.4323,
+      "step": 2098
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.455246007491682,
+      "learning_rate": 9.65177825808033e-06,
+      "loss": 0.397,
+      "step": 2099
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.5769463477288777,
+      "learning_rate": 9.65136527587615e-06,
+      "loss": 0.408,
+      "step": 2100
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4378166250737899,
+      "learning_rate": 9.650952057769888e-06,
+      "loss": 0.1696,
+      "step": 2101
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9710592967632614,
+      "learning_rate": 9.650538603782502e-06,
+      "loss": 0.4084,
+      "step": 2102
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.556031143663162,
+      "learning_rate": 9.650124913934957e-06,
+      "loss": 0.4536,
+      "step": 2103
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.13042980958627,
+      "learning_rate": 9.649710988248238e-06,
+      "loss": 0.4033,
+      "step": 2104
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.39099931563634355,
+      "learning_rate": 9.649296826743337e-06,
+      "loss": 0.1938,
+      "step": 2105
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.988677875751061,
+      "learning_rate": 9.648882429441258e-06,
+      "loss": 0.4584,
+      "step": 2106
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.456857227722525,
+      "learning_rate": 9.648467796363019e-06,
+      "loss": 0.4186,
+      "step": 2107
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6216948294872897,
+      "learning_rate": 9.648052927529648e-06,
+      "loss": 0.4011,
+      "step": 2108
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.36616498432963,
+      "learning_rate": 9.647637822962184e-06,
+      "loss": 0.4302,
+      "step": 2109
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.98952116804909,
+      "learning_rate": 9.647222482681684e-06,
+      "loss": 0.4083,
+      "step": 2110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.29441705615441,
+      "learning_rate": 9.646806906709209e-06,
+      "loss": 0.4212,
+      "step": 2111
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 19.28111886102844,
+      "learning_rate": 9.646391095065838e-06,
+      "loss": 0.4341,
+      "step": 2112
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.476253256732907,
+      "learning_rate": 9.64597504777266e-06,
+      "loss": 0.4427,
+      "step": 2113
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.0362981545285885,
+      "learning_rate": 9.645558764850773e-06,
+      "loss": 0.4207,
+      "step": 2114
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9767691677395574,
+      "learning_rate": 9.645142246321291e-06,
+      "loss": 0.3921,
+      "step": 2115
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.626983481876399,
+      "learning_rate": 9.644725492205338e-06,
+      "loss": 0.491,
+      "step": 2116
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.41405382318195,
+      "learning_rate": 9.644308502524052e-06,
+      "loss": 0.4052,
+      "step": 2117
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.884398607268568,
+      "learning_rate": 9.643891277298578e-06,
+      "loss": 0.4324,
+      "step": 2118
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.410409922511797,
+      "learning_rate": 9.64347381655008e-06,
+      "loss": 0.4237,
+      "step": 2119
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.048573967446438,
+      "learning_rate": 9.643056120299728e-06,
+      "loss": 0.4227,
+      "step": 2120
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.13097853663911,
+      "learning_rate": 9.642638188568706e-06,
+      "loss": 0.4152,
+      "step": 2121
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.7354470829458375,
+      "learning_rate": 9.642220021378212e-06,
+      "loss": 0.4211,
+      "step": 2122
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.4613579693988985,
+      "learning_rate": 9.641801618749454e-06,
+      "loss": 0.4313,
+      "step": 2123
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 16.88458410373081,
+      "learning_rate": 9.641382980703648e-06,
+      "loss": 0.4567,
+      "step": 2124
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.968670326042318,
+      "learning_rate": 9.640964107262031e-06,
+      "loss": 0.4714,
+      "step": 2125
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.467804524528303,
+      "learning_rate": 9.640544998445847e-06,
+      "loss": 0.397,
+      "step": 2126
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.3481566451753455,
+      "learning_rate": 9.640125654276347e-06,
+      "loss": 0.3852,
+      "step": 2127
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.456438191700505,
+      "learning_rate": 9.639706074774801e-06,
+      "loss": 0.4417,
+      "step": 2128
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.341457177808936,
+      "learning_rate": 9.63928625996249e-06,
+      "loss": 0.3927,
+      "step": 2129
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.35585290494804,
+      "learning_rate": 9.638866209860706e-06,
+      "loss": 0.4169,
+      "step": 2130
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.257197318470475,
+      "learning_rate": 9.63844592449075e-06,
+      "loss": 0.4053,
+      "step": 2131
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.85325807593556,
+      "learning_rate": 9.638025403873939e-06,
+      "loss": 0.4454,
+      "step": 2132
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.170878774756297,
+      "learning_rate": 9.637604648031603e-06,
+      "loss": 0.41,
+      "step": 2133
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.517560462438786,
+      "learning_rate": 9.637183656985075e-06,
+      "loss": 0.4342,
+      "step": 2134
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5800814211574763,
+      "learning_rate": 9.63676243075571e-06,
+      "loss": 0.4274,
+      "step": 2135
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.24241645368486,
+      "learning_rate": 9.636340969364875e-06,
+      "loss": 0.417,
+      "step": 2136
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.886343464205175,
+      "learning_rate": 9.635919272833938e-06,
+      "loss": 0.4212,
+      "step": 2137
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.876386049944942,
+      "learning_rate": 9.63549734118429e-06,
+      "loss": 0.4204,
+      "step": 2138
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.3783254329869195,
+      "learning_rate": 9.63507517443733e-06,
+      "loss": 0.4224,
+      "step": 2139
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.515110326175763,
+      "learning_rate": 9.634652772614469e-06,
+      "loss": 0.413,
+      "step": 2140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.684650846942196,
+      "learning_rate": 9.634230135737128e-06,
+      "loss": 0.4754,
+      "step": 2141
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.9482801847347377,
+      "learning_rate": 9.633807263826745e-06,
+      "loss": 0.4099,
+      "step": 2142
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.86130693957164,
+      "learning_rate": 9.633384156904762e-06,
+      "loss": 0.3706,
+      "step": 2143
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.357166083163546,
+      "learning_rate": 9.632960814992642e-06,
+      "loss": 0.4344,
+      "step": 2144
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.351906578705544,
+      "learning_rate": 9.632537238111852e-06,
+      "loss": 0.4083,
+      "step": 2145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.4157910435430314,
+      "learning_rate": 9.632113426283877e-06,
+      "loss": 0.4606,
+      "step": 2146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.499677353111395,
+      "learning_rate": 9.631689379530211e-06,
+      "loss": 0.406,
+      "step": 2147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.668992716493397,
+      "learning_rate": 9.63126509787236e-06,
+      "loss": 0.4315,
+      "step": 2148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.6198891022731843,
+      "learning_rate": 9.630840581331841e-06,
+      "loss": 0.4249,
+      "step": 2149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.265882320599197,
+      "learning_rate": 9.630415829930186e-06,
+      "loss": 0.4194,
+      "step": 2150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.715218067706965,
+      "learning_rate": 9.629990843688936e-06,
+      "loss": 0.4012,
+      "step": 2151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.080788585328057,
+      "learning_rate": 9.629565622629646e-06,
+      "loss": 0.3924,
+      "step": 2152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.5701524308128665,
+      "learning_rate": 9.62914016677388e-06,
+      "loss": 0.424,
+      "step": 2153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.647805248978985,
+      "learning_rate": 9.628714476143217e-06,
+      "loss": 0.4534,
+      "step": 2154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.148793911334401,
+      "learning_rate": 9.628288550759246e-06,
+      "loss": 0.441,
+      "step": 2155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.183937596476974,
+      "learning_rate": 9.627862390643568e-06,
+      "loss": 0.4454,
+      "step": 2156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.9268533040451405,
+      "learning_rate": 9.627435995817799e-06,
+      "loss": 0.4691,
+      "step": 2157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.690018321531943,
+      "learning_rate": 9.627009366303561e-06,
+      "loss": 0.4157,
+      "step": 2158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.436045549865946,
+      "learning_rate": 9.626582502122494e-06,
+      "loss": 0.3975,
+      "step": 2159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.556630429829577,
+      "learning_rate": 9.626155403296245e-06,
+      "loss": 0.4,
+      "step": 2160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.246251217705703,
+      "learning_rate": 9.625728069846476e-06,
+      "loss": 0.4201,
+      "step": 2161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.061502839700778,
+      "learning_rate": 9.625300501794862e-06,
+      "loss": 0.4066,
+      "step": 2162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.071170016165437,
+      "learning_rate": 9.624872699163084e-06,
+      "loss": 0.4045,
+      "step": 2163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.014472991300183,
+      "learning_rate": 9.624444661972841e-06,
+      "loss": 0.4266,
+      "step": 2164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.459648910736254,
+      "learning_rate": 9.624016390245841e-06,
+      "loss": 0.4194,
+      "step": 2165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.03617086553855,
+      "learning_rate": 9.623587884003804e-06,
+      "loss": 0.4148,
+      "step": 2166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.614564962708371,
+      "learning_rate": 9.623159143268466e-06,
+      "loss": 0.3819,
+      "step": 2167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.468313604660949,
+      "learning_rate": 9.622730168061568e-06,
+      "loss": 0.441,
+      "step": 2168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.050823243244909,
+      "learning_rate": 9.622300958404864e-06,
+      "loss": 0.4368,
+      "step": 2169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.744116903365635,
+      "learning_rate": 9.621871514320128e-06,
+      "loss": 0.4301,
+      "step": 2170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.302134487194424,
+      "learning_rate": 9.621441835829135e-06,
+      "loss": 0.4068,
+      "step": 2171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.3203325106123325,
+      "learning_rate": 9.621011922953681e-06,
+      "loss": 0.458,
+      "step": 2172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.431063570689035,
+      "learning_rate": 9.620581775715565e-06,
+      "loss": 0.4339,
+      "step": 2173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.712292093846029,
+      "learning_rate": 9.620151394136608e-06,
+      "loss": 0.4379,
+      "step": 2174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.565478446147961,
+      "learning_rate": 9.619720778238632e-06,
+      "loss": 0.4078,
+      "step": 2175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.970828690617024,
+      "learning_rate": 9.619289928043481e-06,
+      "loss": 0.3949,
+      "step": 2176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.000281160508373,
+      "learning_rate": 9.618858843573004e-06,
+      "loss": 0.4123,
+      "step": 2177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.0398951508365295,
+      "learning_rate": 9.618427524849066e-06,
+      "loss": 0.4104,
+      "step": 2178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.058685096588917,
+      "learning_rate": 9.61799597189354e-06,
+      "loss": 0.449,
+      "step": 2179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.5272628309661638,
+      "learning_rate": 9.617564184728313e-06,
+      "loss": 0.4038,
+      "step": 2180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.033332862373067,
+      "learning_rate": 9.617132163375287e-06,
+      "loss": 0.4638,
+      "step": 2181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.268343983951499,
+      "learning_rate": 9.616699907856368e-06,
+      "loss": 0.4468,
+      "step": 2182
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.569097653944869,
+      "learning_rate": 9.616267418193483e-06,
+      "loss": 0.3738,
+      "step": 2183
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.928140195215341,
+      "learning_rate": 9.615834694408564e-06,
+      "loss": 0.4179,
+      "step": 2184
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.705174987789969,
+      "learning_rate": 9.615401736523556e-06,
+      "loss": 0.3901,
+      "step": 2185
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.199853319935032,
+      "learning_rate": 9.61496854456042e-06,
+      "loss": 0.3946,
+      "step": 2186
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.036739948870444,
+      "learning_rate": 9.614535118541126e-06,
+      "loss": 0.4242,
+      "step": 2187
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.900563296814383,
+      "learning_rate": 9.614101458487653e-06,
+      "loss": 0.4133,
+      "step": 2188
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.8339670589359605,
+      "learning_rate": 9.613667564421999e-06,
+      "loss": 0.4443,
+      "step": 2189
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.758043832018357,
+      "learning_rate": 9.613233436366167e-06,
+      "loss": 0.399,
+      "step": 2190
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.192458508622665,
+      "learning_rate": 9.612799074342174e-06,
+      "loss": 0.4144,
+      "step": 2191
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.804829240584094,
+      "learning_rate": 9.612364478372051e-06,
+      "loss": 0.4248,
+      "step": 2192
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.350758328805776,
+      "learning_rate": 9.61192964847784e-06,
+      "loss": 0.4177,
+      "step": 2193
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.349428413734345,
+      "learning_rate": 9.611494584681591e-06,
+      "loss": 0.4196,
+      "step": 2194
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8251583763132597,
+      "learning_rate": 9.611059287005374e-06,
+      "loss": 0.4126,
+      "step": 2195
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.317884058674744,
+      "learning_rate": 9.61062375547126e-06,
+      "loss": 0.4377,
+      "step": 2196
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 10.047024600353602,
+      "learning_rate": 9.610187990101341e-06,
+      "loss": 0.4471,
+      "step": 2197
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.154691378493771,
+      "learning_rate": 9.609751990917719e-06,
+      "loss": 0.3889,
+      "step": 2198
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.977822885507551,
+      "learning_rate": 9.609315757942504e-06,
+      "loss": 0.4699,
+      "step": 2199
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.01134848976511,
+      "learning_rate": 9.60887929119782e-06,
+      "loss": 0.4084,
+      "step": 2200
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.940561145291456,
+      "learning_rate": 9.608442590705804e-06,
+      "loss": 0.4092,
+      "step": 2201
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.406499795454971,
+      "learning_rate": 9.608005656488605e-06,
+      "loss": 0.4587,
+      "step": 2202
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.7902540845089305,
+      "learning_rate": 9.607568488568382e-06,
+      "loss": 0.4394,
+      "step": 2203
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.1373061138415155,
+      "learning_rate": 9.607131086967307e-06,
+      "loss": 0.3886,
+      "step": 2204
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.552152323137182,
+      "learning_rate": 9.606693451707565e-06,
+      "loss": 0.4448,
+      "step": 2205
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.860076820377096,
+      "learning_rate": 9.606255582811346e-06,
+      "loss": 0.4219,
+      "step": 2206
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.625733373748926,
+      "learning_rate": 9.605817480300863e-06,
+      "loss": 0.4115,
+      "step": 2207
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 12.876158238013767,
+      "learning_rate": 9.605379144198335e-06,
+      "loss": 0.3915,
+      "step": 2208
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.910333086986523,
+      "learning_rate": 9.604940574525988e-06,
+      "loss": 0.4108,
+      "step": 2209
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.067974149965869,
+      "learning_rate": 9.60450177130607e-06,
+      "loss": 0.4448,
+      "step": 2210
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.194652283409591,
+      "learning_rate": 9.604062734560832e-06,
+      "loss": 0.3931,
+      "step": 2211
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.359051613525942,
+      "learning_rate": 9.603623464312545e-06,
+      "loss": 0.4207,
+      "step": 2212
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 17.052775107502747,
+      "learning_rate": 9.603183960583482e-06,
+      "loss": 0.3918,
+      "step": 2213
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.473867988338985,
+      "learning_rate": 9.602744223395937e-06,
+      "loss": 0.3998,
+      "step": 2214
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.71105590981024,
+      "learning_rate": 9.602304252772211e-06,
+      "loss": 0.4231,
+      "step": 2215
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 13.432287616027082,
+      "learning_rate": 9.601864048734617e-06,
+      "loss": 0.4041,
+      "step": 2216
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.098370681000494,
+      "learning_rate": 9.601423611305481e-06,
+      "loss": 0.4228,
+      "step": 2217
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.8803886669288747,
+      "learning_rate": 9.600982940507142e-06,
+      "loss": 0.4244,
+      "step": 2218
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.175366220995254,
+      "learning_rate": 9.600542036361948e-06,
+      "loss": 0.4418,
+      "step": 2219
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.616371901141268,
+      "learning_rate": 9.600100898892261e-06,
+      "loss": 0.4475,
+      "step": 2220
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.456068032698394,
+      "learning_rate": 9.599659528120453e-06,
+      "loss": 0.4026,
+      "step": 2221
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 6.712172658325925,
+      "learning_rate": 9.59921792406891e-06,
+      "loss": 0.3978,
+      "step": 2222
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 8.462780984872666,
+      "learning_rate": 9.598776086760026e-06,
+      "loss": 0.3943,
+      "step": 2223
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 15.775342624365384,
+      "learning_rate": 9.598334016216214e-06,
+      "loss": 0.4357,
+      "step": 2224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.500035252254414,
+      "learning_rate": 9.597891712459892e-06,
+      "loss": 0.4155,
+      "step": 2225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.841896071402818,
+      "learning_rate": 9.597449175513492e-06,
+      "loss": 0.3994,
+      "step": 2226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.722884294893017,
+      "learning_rate": 9.597006405399458e-06,
+      "loss": 0.4173,
+      "step": 2227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.621090389670379,
+      "learning_rate": 9.596563402140246e-06,
+      "loss": 0.4339,
+      "step": 2228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.100196509805762,
+      "learning_rate": 9.596120165758325e-06,
+      "loss": 0.4304,
+      "step": 2229
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.071164716970954,
+      "learning_rate": 9.595676696276173e-06,
+      "loss": 0.4023,
+      "step": 2230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.138056037236367,
+      "learning_rate": 9.595232993716282e-06,
+      "loss": 0.4148,
+      "step": 2231
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.415533664874702,
+      "learning_rate": 9.594789058101154e-06,
+      "loss": 0.4507,
+      "step": 2232
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.096232177242232,
+      "learning_rate": 9.594344889453305e-06,
+      "loss": 0.4618,
+      "step": 2233
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 16.390449103142405,
+      "learning_rate": 9.593900487795263e-06,
+      "loss": 0.4195,
+      "step": 2234
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.7691130009743725,
+      "learning_rate": 9.593455853149563e-06,
+      "loss": 0.3921,
+      "step": 2235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.360189095312188,
+      "learning_rate": 9.59301098553876e-06,
+      "loss": 0.3803,
+      "step": 2236
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.322474301825613,
+      "learning_rate": 9.592565884985412e-06,
+      "loss": 0.4104,
+      "step": 2237
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.121903724240498,
+      "learning_rate": 9.592120551512096e-06,
+      "loss": 0.3995,
+      "step": 2238
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.934588555388183,
+      "learning_rate": 9.591674985141395e-06,
+      "loss": 0.4108,
+      "step": 2239
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.593940510674432,
+      "learning_rate": 9.59122918589591e-06,
+      "loss": 0.4376,
+      "step": 2240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.65250490622457,
+      "learning_rate": 9.59078315379825e-06,
+      "loss": 0.4574,
+      "step": 2241
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.75148311009767,
+      "learning_rate": 9.590336888871034e-06,
+      "loss": 0.4081,
+      "step": 2242
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8620094126400257,
+      "learning_rate": 9.589890391136896e-06,
+      "loss": 0.3981,
+      "step": 2243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 14.872231266043983,
+      "learning_rate": 9.589443660618483e-06,
+      "loss": 0.4114,
+      "step": 2244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.44801344181408365,
+      "learning_rate": 9.58899669733845e-06,
+      "loss": 0.1937,
+      "step": 2245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.135318953211758,
+      "learning_rate": 9.588549501319462e-06,
+      "loss": 0.4198,
+      "step": 2246
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.524002472197025,
+      "learning_rate": 9.588102072584204e-06,
+      "loss": 0.4742,
+      "step": 2247
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.9331814268875003,
+      "learning_rate": 9.58765441115537e-06,
+      "loss": 0.4152,
+      "step": 2248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.503068558261996,
+      "learning_rate": 9.587206517055659e-06,
+      "loss": 0.3978,
+      "step": 2249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.369267157167057,
+      "learning_rate": 9.586758390307788e-06,
+      "loss": 0.4582,
+      "step": 2250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.1808194973109805,
+      "learning_rate": 9.586310030934484e-06,
+      "loss": 0.4478,
+      "step": 2251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.006627443719469,
+      "learning_rate": 9.58586143895849e-06,
+      "loss": 0.4329,
+      "step": 2252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.814010871646705,
+      "learning_rate": 9.585412614402554e-06,
+      "loss": 0.4371,
+      "step": 2253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.863479784993586,
+      "learning_rate": 9.58496355728944e-06,
+      "loss": 0.437,
+      "step": 2254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.505297623977466,
+      "learning_rate": 9.58451426764192e-06,
+      "loss": 0.4141,
+      "step": 2255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.134911374446605,
+      "learning_rate": 9.584064745482784e-06,
+      "loss": 0.3968,
+      "step": 2256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.824221685567036,
+      "learning_rate": 9.58361499083483e-06,
+      "loss": 0.4353,
+      "step": 2257
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.774417543701418,
+      "learning_rate": 9.583165003720866e-06,
+      "loss": 0.414,
+      "step": 2258
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.251339519225088,
+      "learning_rate": 9.582714784163715e-06,
+      "loss": 0.4105,
+      "step": 2259
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.468946144107196,
+      "learning_rate": 9.582264332186211e-06,
+      "loss": 0.3938,
+      "step": 2260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.445873664290071,
+      "learning_rate": 9.581813647811199e-06,
+      "loss": 0.4547,
+      "step": 2261
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.194242284142627,
+      "learning_rate": 9.581362731061537e-06,
+      "loss": 0.4359,
+      "step": 2262
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.527000723068008,
+      "learning_rate": 9.580911581960092e-06,
+      "loss": 0.4317,
+      "step": 2263
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.1907281646509,
+      "learning_rate": 9.580460200529746e-06,
+      "loss": 0.4061,
+      "step": 2264
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8109859342402883,
+      "learning_rate": 9.580008586793394e-06,
+      "loss": 0.4126,
+      "step": 2265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.795718143444829,
+      "learning_rate": 9.579556740773936e-06,
+      "loss": 0.4159,
+      "step": 2266
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 10.242049459673463,
+      "learning_rate": 9.579104662494293e-06,
+      "loss": 0.4196,
+      "step": 2267
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.824687834735094,
+      "learning_rate": 9.578652351977386e-06,
+      "loss": 0.4235,
+      "step": 2268
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.589094869990564,
+      "learning_rate": 9.578199809246162e-06,
+      "loss": 0.4105,
+      "step": 2269
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.720537332605927,
+      "learning_rate": 9.577747034323569e-06,
+      "loss": 0.4444,
+      "step": 2270
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.394144574282694,
+      "learning_rate": 9.577294027232571e-06,
+      "loss": 0.4242,
+      "step": 2271
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5257088251140799,
+      "learning_rate": 9.576840787996142e-06,
+      "loss": 0.213,
+      "step": 2272
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4428346085576393,
+      "learning_rate": 9.57638731663727e-06,
+      "loss": 0.4218,
+      "step": 2273
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.456883392066961,
+      "learning_rate": 9.575933613178954e-06,
+      "loss": 0.4236,
+      "step": 2274
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3968692899933295,
+      "learning_rate": 9.575479677644202e-06,
+      "loss": 0.435,
+      "step": 2275
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8854954333658327,
+      "learning_rate": 9.575025510056038e-06,
+      "loss": 0.4383,
+      "step": 2276
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.333044220455882,
+      "learning_rate": 9.574571110437496e-06,
+      "loss": 0.4287,
+      "step": 2277
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 12.097335635791383,
+      "learning_rate": 9.574116478811622e-06,
+      "loss": 0.4057,
+      "step": 2278
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.8820002020672995,
+      "learning_rate": 9.573661615201472e-06,
+      "loss": 0.4637,
+      "step": 2279
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4818808596908215,
+      "learning_rate": 9.573206519630117e-06,
+      "loss": 0.3998,
+      "step": 2280
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.850947154408965,
+      "learning_rate": 9.572751192120634e-06,
+      "loss": 0.4152,
+      "step": 2281
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.631280779952449,
+      "learning_rate": 9.57229563269612e-06,
+      "loss": 0.4582,
+      "step": 2282
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.628352955978802,
+      "learning_rate": 9.571839841379677e-06,
+      "loss": 0.4133,
+      "step": 2283
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.328220953747133,
+      "learning_rate": 9.571383818194424e-06,
+      "loss": 0.4615,
+      "step": 2284
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.243423101781134,
+      "learning_rate": 9.570927563163488e-06,
+      "loss": 0.4334,
+      "step": 2285
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.673319907769496,
+      "learning_rate": 9.570471076310006e-06,
+      "loss": 0.4365,
+      "step": 2286
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.28248063621346,
+      "learning_rate": 9.570014357657133e-06,
+      "loss": 0.394,
+      "step": 2287
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 35.90548926458689,
+      "learning_rate": 9.56955740722803e-06,
+      "loss": 0.4725,
+      "step": 2288
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.0362098975546505,
+      "learning_rate": 9.569100225045873e-06,
+      "loss": 0.4329,
+      "step": 2289
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.588388935558969,
+      "learning_rate": 9.56864281113385e-06,
+      "loss": 0.4481,
+      "step": 2290
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.727609038336485,
+      "learning_rate": 9.568185165515156e-06,
+      "loss": 0.3949,
+      "step": 2291
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.009817545658764,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.4348,
+      "step": 2292
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4788408025681177,
+      "learning_rate": 9.567269179250618e-06,
+      "loss": 0.4518,
+      "step": 2293
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5613268262607293,
+      "learning_rate": 9.566810838651228e-06,
+      "loss": 0.1997,
+      "step": 2294
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4947746138494727,
+      "learning_rate": 9.56635226643808e-06,
+      "loss": 0.1921,
+      "step": 2295
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.771788269639474,
+      "learning_rate": 9.565893462634432e-06,
+      "loss": 0.4166,
+      "step": 2296
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.082842065854804,
+      "learning_rate": 9.565434427263556e-06,
+      "loss": 0.402,
+      "step": 2297
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.406476639216313,
+      "learning_rate": 9.564975160348728e-06,
+      "loss": 0.4,
+      "step": 2298
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6095801047178955,
+      "learning_rate": 9.564515661913242e-06,
+      "loss": 0.4235,
+      "step": 2299
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.924015766701668,
+      "learning_rate": 9.564055931980403e-06,
+      "loss": 0.4429,
+      "step": 2300
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.4400594498860175,
+      "learning_rate": 9.563595970573527e-06,
+      "loss": 0.4228,
+      "step": 2301
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.807999172745121,
+      "learning_rate": 9.563135777715942e-06,
+      "loss": 0.464,
+      "step": 2302
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.7515798593436305,
+      "learning_rate": 9.562675353430988e-06,
+      "loss": 0.4056,
+      "step": 2303
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.62647295050162,
+      "learning_rate": 9.562214697742014e-06,
+      "loss": 0.4167,
+      "step": 2304
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.778841952704133,
+      "learning_rate": 9.561753810672385e-06,
+      "loss": 0.4034,
+      "step": 2305
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.0295234046821005,
+      "learning_rate": 9.561292692245473e-06,
+      "loss": 0.4034,
+      "step": 2306
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.22398102664201,
+      "learning_rate": 9.560831342484668e-06,
+      "loss": 0.4374,
+      "step": 2307
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.173530307236204,
+      "learning_rate": 9.560369761413366e-06,
+      "loss": 0.4019,
+      "step": 2308
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.80215250348137,
+      "learning_rate": 9.559907949054978e-06,
+      "loss": 0.4369,
+      "step": 2309
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2647234904085782,
+      "learning_rate": 9.559445905432923e-06,
+      "loss": 0.4796,
+      "step": 2310
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.364722322932441,
+      "learning_rate": 9.558983630570638e-06,
+      "loss": 0.4145,
+      "step": 2311
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1644130881818668,
+      "learning_rate": 9.558521124491567e-06,
+      "loss": 0.2642,
+      "step": 2312
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.7809098266705456,
+      "learning_rate": 9.558058387219165e-06,
+      "loss": 0.4962,
+      "step": 2313
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.773857427659792,
+      "learning_rate": 9.557595418776902e-06,
+      "loss": 0.4082,
+      "step": 2314
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.925602035253528,
+      "learning_rate": 9.557132219188257e-06,
+      "loss": 0.4177,
+      "step": 2315
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.715386594047724,
+      "learning_rate": 9.556668788476725e-06,
+      "loss": 0.4553,
+      "step": 2316
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8184217825519347,
+      "learning_rate": 9.556205126665805e-06,
+      "loss": 0.4169,
+      "step": 2317
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.438927848400107,
+      "learning_rate": 9.555741233779018e-06,
+      "loss": 0.4073,
+      "step": 2318
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.8265730276967815,
+      "learning_rate": 9.555277109839887e-06,
+      "loss": 0.4286,
+      "step": 2319
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.545714342307071,
+      "learning_rate": 9.55481275487195e-06,
+      "loss": 0.4243,
+      "step": 2320
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 11.090498590723703,
+      "learning_rate": 9.554348168898763e-06,
+      "loss": 0.4082,
+      "step": 2321
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5684781935427463,
+      "learning_rate": 9.553883351943882e-06,
+      "loss": 0.2134,
+      "step": 2322
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 8.689595900044818,
+      "learning_rate": 9.553418304030886e-06,
+      "loss": 0.3966,
+      "step": 2323
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.468701549124765,
+      "learning_rate": 9.552953025183358e-06,
+      "loss": 0.4532,
+      "step": 2324
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.8956051070026305,
+      "learning_rate": 9.552487515424896e-06,
+      "loss": 0.4124,
+      "step": 2325
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8138834994044006,
+      "learning_rate": 9.552021774779109e-06,
+      "loss": 0.4206,
+      "step": 2326
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.675251909025207,
+      "learning_rate": 9.551555803269618e-06,
+      "loss": 0.4222,
+      "step": 2327
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.492989594007183,
+      "learning_rate": 9.551089600920056e-06,
+      "loss": 0.4176,
+      "step": 2328
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.7955079532512315,
+      "learning_rate": 9.550623167754068e-06,
+      "loss": 0.5151,
+      "step": 2329
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.471606248211269,
+      "learning_rate": 9.550156503795307e-06,
+      "loss": 0.4455,
+      "step": 2330
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42460469086421354,
+      "learning_rate": 9.549689609067443e-06,
+      "loss": 0.1958,
+      "step": 2331
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 6.107276524682691,
+      "learning_rate": 9.549222483594154e-06,
+      "loss": 0.4465,
+      "step": 2332
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.122891092886755,
+      "learning_rate": 9.548755127399132e-06,
+      "loss": 0.4271,
+      "step": 2333
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3683584922983978,
+      "learning_rate": 9.54828754050608e-06,
+      "loss": 0.1749,
+      "step": 2334
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.793172644284434,
+      "learning_rate": 9.547819722938713e-06,
+      "loss": 0.4266,
+      "step": 2335
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.183643579429848,
+      "learning_rate": 9.547351674720757e-06,
+      "loss": 0.396,
+      "step": 2336
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.718204169373937,
+      "learning_rate": 9.546883395875947e-06,
+      "loss": 0.434,
+      "step": 2337
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.408815207650009,
+      "learning_rate": 9.546414886428037e-06,
+      "loss": 0.4823,
+      "step": 2338
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.284453547318475,
+      "learning_rate": 9.545946146400787e-06,
+      "loss": 0.4327,
+      "step": 2339
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.329322313399505,
+      "learning_rate": 9.545477175817966e-06,
+      "loss": 0.4482,
+      "step": 2340
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 7.474960785444928,
+      "learning_rate": 9.545007974703364e-06,
+      "loss": 0.4056,
+      "step": 2341
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.541270754496034,
+      "learning_rate": 9.544538543080773e-06,
+      "loss": 0.4183,
+      "step": 2342
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.03480369254328,
+      "learning_rate": 9.544068880974004e-06,
+      "loss": 0.4495,
+      "step": 2343
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.440145811954663,
+      "learning_rate": 9.543598988406877e-06,
+      "loss": 0.4207,
+      "step": 2344
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6706979742554577,
+      "learning_rate": 9.543128865403222e-06,
+      "loss": 0.4243,
+      "step": 2345
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.048096153311716,
+      "learning_rate": 9.542658511986882e-06,
+      "loss": 0.4335,
+      "step": 2346
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.543312532504392,
+      "learning_rate": 9.542187928181712e-06,
+      "loss": 0.4272,
+      "step": 2347
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.098496622082065,
+      "learning_rate": 9.541717114011579e-06,
+      "loss": 0.3833,
+      "step": 2348
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 9.33659932547455,
+      "learning_rate": 9.541246069500361e-06,
+      "loss": 0.4308,
+      "step": 2349
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.49886857117573313,
+      "learning_rate": 9.540774794671946e-06,
+      "loss": 0.2201,
+      "step": 2350
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 20.916944908596474,
+      "learning_rate": 9.54030328955024e-06,
+      "loss": 0.4119,
+      "step": 2351
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.580543838308559,
+      "learning_rate": 9.539831554159152e-06,
+      "loss": 0.4047,
+      "step": 2352
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.159728360637781,
+      "learning_rate": 9.539359588522608e-06,
+      "loss": 0.4435,
+      "step": 2353
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.3748573499934524,
+      "learning_rate": 9.538887392664544e-06,
+      "loss": 0.1935,
+      "step": 2354
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.7382853458072365,
+      "learning_rate": 9.53841496660891e-06,
+      "loss": 0.4305,
+      "step": 2355
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.66840677933595,
+      "learning_rate": 9.537942310379666e-06,
+      "loss": 0.4177,
+      "step": 2356
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.923639033020512,
+      "learning_rate": 9.53746942400078e-06,
+      "loss": 0.4242,
+      "step": 2357
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.44998627302475,
+      "learning_rate": 9.53699630749624e-06,
+      "loss": 0.427,
+      "step": 2358
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.73047764133375,
+      "learning_rate": 9.536522960890038e-06,
+      "loss": 0.4585,
+      "step": 2359
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.794338939242935,
+      "learning_rate": 9.536049384206182e-06,
+      "loss": 0.4084,
+      "step": 2360
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.382917501857542,
+      "learning_rate": 9.53557557746869e-06,
+      "loss": 0.4368,
+      "step": 2361
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.88917795466991,
+      "learning_rate": 9.53510154070159e-06,
+      "loss": 0.4248,
+      "step": 2362
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.362192855667103,
+      "learning_rate": 9.534627273928924e-06,
+      "loss": 0.3823,
+      "step": 2363
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.6514879789044046,
+      "learning_rate": 9.53415277717475e-06,
+      "loss": 0.4211,
+      "step": 2364
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.562747840049474,
+      "learning_rate": 9.533678050463126e-06,
+      "loss": 0.4412,
+      "step": 2365
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.3552098177267866,
+      "learning_rate": 9.533203093818135e-06,
+      "loss": 0.4202,
+      "step": 2366
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8260851941731846,
+      "learning_rate": 9.532727907263861e-06,
+      "loss": 0.4292,
+      "step": 2367
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 39.85421117366959,
+      "learning_rate": 9.532252490824404e-06,
+      "loss": 0.4194,
+      "step": 2368
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.691960697047073,
+      "learning_rate": 9.531776844523877e-06,
+      "loss": 0.3804,
+      "step": 2369
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.11340840498216,
+      "learning_rate": 9.531300968386404e-06,
+      "loss": 0.4046,
+      "step": 2370
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.5458189178131656,
+      "learning_rate": 9.530824862436117e-06,
+      "loss": 0.4331,
+      "step": 2371
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.300648871330333,
+      "learning_rate": 9.530348526697166e-06,
+      "loss": 0.443,
+      "step": 2372
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.499034426366422,
+      "learning_rate": 9.529871961193706e-06,
+      "loss": 0.1997,
+      "step": 2373
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.9585571146607887,
+      "learning_rate": 9.529395165949911e-06,
+      "loss": 0.429,
+      "step": 2374
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.698164047720833,
+      "learning_rate": 9.528918140989956e-06,
+      "loss": 0.4354,
+      "step": 2375
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.931498892861189,
+      "learning_rate": 9.528440886338042e-06,
+      "loss": 0.4191,
+      "step": 2376
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.312850409771848,
+      "learning_rate": 9.527963402018367e-06,
+      "loss": 0.4301,
+      "step": 2377
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.844446943500464,
+      "learning_rate": 9.527485688055152e-06,
+      "loss": 0.432,
+      "step": 2378
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.054328305820139,
+      "learning_rate": 9.527007744472622e-06,
+      "loss": 0.3553,
+      "step": 2379
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.595683448926774,
+      "learning_rate": 9.526529571295017e-06,
+      "loss": 0.4463,
+      "step": 2380
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.913289784381265,
+      "learning_rate": 9.526051168546593e-06,
+      "loss": 0.4143,
+      "step": 2381
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.241231628690263,
+      "learning_rate": 9.525572536251608e-06,
+      "loss": 0.3993,
+      "step": 2382
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.609206325002738,
+      "learning_rate": 9.525093674434337e-06,
+      "loss": 0.437,
+      "step": 2383
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4154327548259174,
+      "learning_rate": 9.524614583119067e-06,
+      "loss": 0.1951,
+      "step": 2384
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.125508769627743,
+      "learning_rate": 9.524135262330098e-06,
+      "loss": 0.4294,
+      "step": 2385
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.405759364895134,
+      "learning_rate": 9.523655712091739e-06,
+      "loss": 0.428,
+      "step": 2386
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.153630818030064,
+      "learning_rate": 9.523175932428307e-06,
+      "loss": 0.4225,
+      "step": 2387
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.79112778301043,
+      "learning_rate": 9.522695923364142e-06,
+      "loss": 0.3942,
+      "step": 2388
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.5231816020768565,
+      "learning_rate": 9.522215684923582e-06,
+      "loss": 0.4312,
+      "step": 2389
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.338567288993318,
+      "learning_rate": 9.521735217130987e-06,
+      "loss": 0.4332,
+      "step": 2390
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.7852596301695036,
+      "learning_rate": 9.521254520010722e-06,
+      "loss": 0.4059,
+      "step": 2391
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.760357444160125,
+      "learning_rate": 9.520773593587169e-06,
+      "loss": 0.4366,
+      "step": 2392
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.422894709858651,
+      "learning_rate": 9.520292437884718e-06,
+      "loss": 0.3993,
+      "step": 2393
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.082463099907226,
+      "learning_rate": 9.51981105292777e-06,
+      "loss": 0.4205,
+      "step": 2394
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.141868141042294,
+      "learning_rate": 9.519329438740744e-06,
+      "loss": 0.4503,
+      "step": 2395
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.197200828944037,
+      "learning_rate": 9.51884759534806e-06,
+      "loss": 0.4178,
+      "step": 2396
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6857069528862056,
+      "learning_rate": 9.518365522774157e-06,
+      "loss": 0.4063,
+      "step": 2397
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.721869218290839,
+      "learning_rate": 9.51788322104349e-06,
+      "loss": 0.4042,
+      "step": 2398
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.003077359930783,
+      "learning_rate": 9.517400690180512e-06,
+      "loss": 0.4098,
+      "step": 2399
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.387060659925024,
+      "learning_rate": 9.516917930209699e-06,
+      "loss": 0.403,
+      "step": 2400
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.42086749174433,
+      "learning_rate": 9.516434941155535e-06,
+      "loss": 0.4049,
+      "step": 2401
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.732656235429874,
+      "learning_rate": 9.515951723042514e-06,
+      "loss": 0.4723,
+      "step": 2402
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.656058253583432,
+      "learning_rate": 9.515468275895146e-06,
+      "loss": 0.4024,
+      "step": 2403
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.640487378146455,
+      "learning_rate": 9.514984599737947e-06,
+      "loss": 0.4173,
+      "step": 2404
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 15.231745883580396,
+      "learning_rate": 9.51450069459545e-06,
+      "loss": 0.4149,
+      "step": 2405
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.0183153803290494,
+      "learning_rate": 9.514016560492197e-06,
+      "loss": 0.4188,
+      "step": 2406
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.353070431809074,
+      "learning_rate": 9.513532197452737e-06,
+      "loss": 0.4403,
+      "step": 2407
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4609213367771535,
+      "learning_rate": 9.513047605501642e-06,
+      "loss": 0.2049,
+      "step": 2408
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.670182915320048,
+      "learning_rate": 9.512562784663484e-06,
+      "loss": 0.4436,
+      "step": 2409
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.238865523309299,
+      "learning_rate": 9.512077734962855e-06,
+      "loss": 0.463,
+      "step": 2410
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37436394607749235,
+      "learning_rate": 9.511592456424353e-06,
+      "loss": 0.1888,
+      "step": 2411
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.867177957080289,
+      "learning_rate": 9.511106949072588e-06,
+      "loss": 0.4384,
+      "step": 2412
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.755475906832055,
+      "learning_rate": 9.510621212932188e-06,
+      "loss": 0.4375,
+      "step": 2413
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.688794813869095,
+      "learning_rate": 9.510135248027786e-06,
+      "loss": 0.4208,
+      "step": 2414
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.126723836944493,
+      "learning_rate": 9.509649054384027e-06,
+      "loss": 0.4025,
+      "step": 2415
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.39545734123903126,
+      "learning_rate": 9.50916263202557e-06,
+      "loss": 0.2036,
+      "step": 2416
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.245273974673801,
+      "learning_rate": 9.508675980977085e-06,
+      "loss": 0.4493,
+      "step": 2417
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.880150898850358,
+      "learning_rate": 9.508189101263254e-06,
+      "loss": 0.3766,
+      "step": 2418
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.692286757687747,
+      "learning_rate": 9.507701992908768e-06,
+      "loss": 0.4094,
+      "step": 2419
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.352958442995817,
+      "learning_rate": 9.507214655938334e-06,
+      "loss": 0.3856,
+      "step": 2420
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.418085761387968,
+      "learning_rate": 9.506727090376668e-06,
+      "loss": 0.385,
+      "step": 2421
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.554800866625921,
+      "learning_rate": 9.506239296248497e-06,
+      "loss": 0.3843,
+      "step": 2422
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.591747944505454,
+      "learning_rate": 9.505751273578558e-06,
+      "loss": 0.4176,
+      "step": 2423
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.186041491798267,
+      "learning_rate": 9.505263022391605e-06,
+      "loss": 0.4138,
+      "step": 2424
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.377484286842116,
+      "learning_rate": 9.504774542712398e-06,
+      "loss": 0.4315,
+      "step": 2425
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.265143637691417,
+      "learning_rate": 9.504285834565714e-06,
+      "loss": 0.3892,
+      "step": 2426
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3747233913554616,
+      "learning_rate": 9.503796897976339e-06,
+      "loss": 0.1879,
+      "step": 2427
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.981969520186641,
+      "learning_rate": 9.503307732969065e-06,
+      "loss": 0.4195,
+      "step": 2428
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.092052728515408,
+      "learning_rate": 9.502818339568706e-06,
+      "loss": 0.4076,
+      "step": 2429
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.841953928944282,
+      "learning_rate": 9.502328717800082e-06,
+      "loss": 0.4496,
+      "step": 2430
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.2958964110378375,
+      "learning_rate": 9.501838867688022e-06,
+      "loss": 0.4647,
+      "step": 2431
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.32234532753915,
+      "learning_rate": 9.501348789257373e-06,
+      "loss": 0.4078,
+      "step": 2432
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.542607156347881,
+      "learning_rate": 9.500858482532989e-06,
+      "loss": 0.3971,
+      "step": 2433
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.522652847924362,
+      "learning_rate": 9.500367947539735e-06,
+      "loss": 0.4157,
+      "step": 2434
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 21.40545495386612,
+      "learning_rate": 9.499877184302491e-06,
+      "loss": 0.4513,
+      "step": 2435
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.75059088204097,
+      "learning_rate": 9.499386192846147e-06,
+      "loss": 0.4383,
+      "step": 2436
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.562063852933852,
+      "learning_rate": 9.498894973195603e-06,
+      "loss": 0.4357,
+      "step": 2437
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.225746551442026,
+      "learning_rate": 9.498403525375774e-06,
+      "loss": 0.4291,
+      "step": 2438
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.82743602313967,
+      "learning_rate": 9.497911849411585e-06,
+      "loss": 0.4756,
+      "step": 2439
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.172094114834434,
+      "learning_rate": 9.49741994532797e-06,
+      "loss": 0.4252,
+      "step": 2440
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.4880748841938045,
+      "learning_rate": 9.496927813149877e-06,
+      "loss": 0.4073,
+      "step": 2441
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.034797431614967,
+      "learning_rate": 9.496435452902268e-06,
+      "loss": 0.422,
+      "step": 2442
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4252063155945942,
+      "learning_rate": 9.49594286461011e-06,
+      "loss": 0.2039,
+      "step": 2443
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 17.12111287406401,
+      "learning_rate": 9.495450048298389e-06,
+      "loss": 0.421,
+      "step": 2444
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.247875774616606,
+      "learning_rate": 9.494957003992097e-06,
+      "loss": 0.4312,
+      "step": 2445
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.363590237112956,
+      "learning_rate": 9.494463731716241e-06,
+      "loss": 0.1973,
+      "step": 2446
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.813907814928439,
+      "learning_rate": 9.493970231495836e-06,
+      "loss": 0.411,
+      "step": 2447
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.214788353803758,
+      "learning_rate": 9.493476503355911e-06,
+      "loss": 0.4302,
+      "step": 2448
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.58923830369363,
+      "learning_rate": 9.492982547321509e-06,
+      "loss": 0.4774,
+      "step": 2449
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.638981913992788,
+      "learning_rate": 9.492488363417682e-06,
+      "loss": 0.4686,
+      "step": 2450
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37398678012560427,
+      "learning_rate": 9.49199395166949e-06,
+      "loss": 0.1973,
+      "step": 2451
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.918687774546759,
+      "learning_rate": 9.491499312102008e-06,
+      "loss": 0.3956,
+      "step": 2452
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.637321304038445,
+      "learning_rate": 9.491004444740326e-06,
+      "loss": 0.4066,
+      "step": 2453
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.524451349230535,
+      "learning_rate": 9.490509349609538e-06,
+      "loss": 0.5116,
+      "step": 2454
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.2031891647661075,
+      "learning_rate": 9.490014026734758e-06,
+      "loss": 0.4417,
+      "step": 2455
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.591430631209681,
+      "learning_rate": 9.489518476141105e-06,
+      "loss": 0.3894,
+      "step": 2456
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.313643097939643,
+      "learning_rate": 9.48902269785371e-06,
+      "loss": 0.408,
+      "step": 2457
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.233418231250482,
+      "learning_rate": 9.488526691897718e-06,
+      "loss": 0.4462,
+      "step": 2458
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 27.97608462668789,
+      "learning_rate": 9.488030458298288e-06,
+      "loss": 0.4052,
+      "step": 2459
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 10.675230855395904,
+      "learning_rate": 9.487533997080582e-06,
+      "loss": 0.4529,
+      "step": 2460
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.243894721719374,
+      "learning_rate": 9.487037308269784e-06,
+      "loss": 0.3859,
+      "step": 2461
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.037684493599818,
+      "learning_rate": 9.486540391891083e-06,
+      "loss": 0.4122,
+      "step": 2462
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.629910654555387,
+      "learning_rate": 9.48604324796968e-06,
+      "loss": 0.4253,
+      "step": 2463
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 35.61126524322489,
+      "learning_rate": 9.485545876530787e-06,
+      "loss": 0.4329,
+      "step": 2464
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.342964034707329,
+      "learning_rate": 9.485048277599632e-06,
+      "loss": 0.408,
+      "step": 2465
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 12.924133378036615,
+      "learning_rate": 9.484550451201451e-06,
+      "loss": 0.4078,
+      "step": 2466
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.3094897720317284,
+      "learning_rate": 9.48405239736149e-06,
+      "loss": 0.4512,
+      "step": 2467
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3817311999235716,
+      "learning_rate": 9.483554116105011e-06,
+      "loss": 0.1874,
+      "step": 2468
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.848705188930934,
+      "learning_rate": 9.483055607457285e-06,
+      "loss": 0.4096,
+      "step": 2469
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.415227823112266,
+      "learning_rate": 9.482556871443594e-06,
+      "loss": 0.4021,
+      "step": 2470
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.268508563800063,
+      "learning_rate": 9.482057908089233e-06,
+      "loss": 0.4371,
+      "step": 2471
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.919617972747443,
+      "learning_rate": 9.481558717419506e-06,
+      "loss": 0.4032,
+      "step": 2472
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.507279880045494,
+      "learning_rate": 9.481059299459731e-06,
+      "loss": 0.4281,
+      "step": 2473
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.18007404213255,
+      "learning_rate": 9.48055965423524e-06,
+      "loss": 0.4029,
+      "step": 2474
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.421807728343503,
+      "learning_rate": 9.48005978177137e-06,
+      "loss": 0.405,
+      "step": 2475
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.615449388493351,
+      "learning_rate": 9.479559682093473e-06,
+      "loss": 0.399,
+      "step": 2476
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.387085533966558,
+      "learning_rate": 9.479059355226912e-06,
+      "loss": 0.444,
+      "step": 2477
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.927265259341517,
+      "learning_rate": 9.478558801197065e-06,
+      "loss": 0.3909,
+      "step": 2478
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.3828764572900786,
+      "learning_rate": 9.478058020029316e-06,
+      "loss": 0.1935,
+      "step": 2479
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.184564821711417,
+      "learning_rate": 9.477557011749063e-06,
+      "loss": 0.4436,
+      "step": 2480
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.579941756713831,
+      "learning_rate": 9.477055776381715e-06,
+      "loss": 0.4273,
+      "step": 2481
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.579522345292181,
+      "learning_rate": 9.476554313952697e-06,
+      "loss": 0.4183,
+      "step": 2482
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.972075487423074,
+      "learning_rate": 9.476052624487437e-06,
+      "loss": 0.423,
+      "step": 2483
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.107864742154308,
+      "learning_rate": 9.47555070801138e-06,
+      "loss": 0.4051,
+      "step": 2484
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.500989886580422,
+      "learning_rate": 9.475048564549985e-06,
+      "loss": 0.4621,
+      "step": 2485
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.778639406044659,
+      "learning_rate": 9.474546194128713e-06,
+      "loss": 0.4674,
+      "step": 2486
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.0996892049230205,
+      "learning_rate": 9.474043596773048e-06,
+      "loss": 0.43,
+      "step": 2487
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.282764380679042,
+      "learning_rate": 9.473540772508476e-06,
+      "loss": 0.3968,
+      "step": 2488
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.920596190690415,
+      "learning_rate": 9.473037721360504e-06,
+      "loss": 0.4334,
+      "step": 2489
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 7.979223138575195,
+      "learning_rate": 9.472534443354637e-06,
+      "loss": 0.4719,
+      "step": 2490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.825460805768271,
+      "learning_rate": 9.472030938516408e-06,
+      "loss": 0.4329,
+      "step": 2491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 14.088546049939485,
+      "learning_rate": 9.471527206871348e-06,
+      "loss": 0.4036,
+      "step": 2492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.486711626448749,
+      "learning_rate": 9.471023248445007e-06,
+      "loss": 0.4054,
+      "step": 2493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.076160227229004,
+      "learning_rate": 9.470519063262941e-06,
+      "loss": 0.4437,
+      "step": 2494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.268286572907969,
+      "learning_rate": 9.470014651350726e-06,
+      "loss": 0.4113,
+      "step": 2495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.559008657553844,
+      "learning_rate": 9.46951001273394e-06,
+      "loss": 0.4386,
+      "step": 2496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.050118488135966,
+      "learning_rate": 9.469005147438176e-06,
+      "loss": 0.393,
+      "step": 2497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.5490628853743,
+      "learning_rate": 9.468500055489041e-06,
+      "loss": 0.3941,
+      "step": 2498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 8.015786395624941,
+      "learning_rate": 9.467994736912152e-06,
+      "loss": 0.3988,
+      "step": 2499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.30309848741402,
+      "learning_rate": 9.467489191733137e-06,
+      "loss": 0.4512,
+      "step": 2500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.5384292181022206,
+      "learning_rate": 9.466983419977634e-06,
+      "loss": 0.448,
+      "step": 2501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 9.41701585327888,
+      "learning_rate": 9.466477421671296e-06,
+      "loss": 0.413,
+      "step": 2502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.196864565868407,
+      "learning_rate": 9.465971196839786e-06,
+      "loss": 0.4429,
+      "step": 2503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 11.361316887058791,
+      "learning_rate": 9.465464745508775e-06,
+      "loss": 0.4056,
+      "step": 2504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.161235197414018,
+      "learning_rate": 9.464958067703948e-06,
+      "loss": 0.4438,
+      "step": 2505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 23.923329787349765,
+      "learning_rate": 9.464451163451008e-06,
+      "loss": 0.4376,
+      "step": 2506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.175333773737717,
+      "learning_rate": 9.46394403277566e-06,
+      "loss": 0.3887,
+      "step": 2507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.466054359345761,
+      "learning_rate": 9.463436675703622e-06,
+      "loss": 0.4221,
+      "step": 2508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40377093596073516,
+      "learning_rate": 9.46292909226063e-06,
+      "loss": 0.1924,
+      "step": 2509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.826462105350635,
+      "learning_rate": 9.46242128247242e-06,
+      "loss": 0.4105,
+      "step": 2510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.37170858099543613,
+      "learning_rate": 9.461913246364754e-06,
+      "loss": 0.2085,
+      "step": 2511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.114259335792079,
+      "learning_rate": 9.461404983963396e-06,
+      "loss": 0.3993,
+      "step": 2512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.693218010923052,
+      "learning_rate": 9.46089649529412e-06,
+      "loss": 0.4685,
+      "step": 2513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.094989431371425,
+      "learning_rate": 9.460387780382718e-06,
+      "loss": 0.4069,
+      "step": 2514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 6.049704763649189,
+      "learning_rate": 9.459878839254988e-06,
+      "loss": 0.4114,
+      "step": 2515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.40066490908750585,
+      "learning_rate": 9.459369671936744e-06,
+      "loss": 0.1987,
+      "step": 2516
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.302769130022875,
+      "learning_rate": 9.45886027845381e-06,
+      "loss": 0.4405,
+      "step": 2517
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.8989869230593825,
+      "learning_rate": 9.458350658832019e-06,
+      "loss": 0.3986,
+      "step": 2518
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.641160313610774,
+      "learning_rate": 9.457840813097217e-06,
+      "loss": 0.3939,
+      "step": 2519
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.46013831885637,
+      "learning_rate": 9.457330741275262e-06,
+      "loss": 0.4148,
+      "step": 2520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.818241947182049,
+      "learning_rate": 9.456820443392026e-06,
+      "loss": 0.3865,
+      "step": 2521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.988210278388063,
+      "learning_rate": 9.456309919473384e-06,
+      "loss": 0.438,
+      "step": 2522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.794423135532556,
+      "learning_rate": 9.455799169545233e-06,
+      "loss": 0.3951,
+      "step": 2523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.7858309111765855,
+      "learning_rate": 9.455288193633475e-06,
+      "loss": 0.4047,
+      "step": 2524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.484081810708125,
+      "learning_rate": 9.454776991764023e-06,
+      "loss": 0.4455,
+      "step": 2525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.110960144101636,
+      "learning_rate": 9.454265563962808e-06,
+      "loss": 0.4307,
+      "step": 2526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.320328371814233,
+      "learning_rate": 9.453753910255763e-06,
+      "loss": 0.4212,
+      "step": 2527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.781597345818058,
+      "learning_rate": 9.453242030668842e-06,
+      "loss": 0.4411,
+      "step": 2528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.932883416216587,
+      "learning_rate": 9.452729925228001e-06,
+      "loss": 0.4278,
+      "step": 2529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.401658527619097,
+      "learning_rate": 9.452217593959217e-06,
+      "loss": 0.3946,
+      "step": 2530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.966340934091045,
+      "learning_rate": 9.45170503688847e-06,
+      "loss": 0.3824,
+      "step": 2531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.9918018025644715,
+      "learning_rate": 9.451192254041759e-06,
+      "loss": 0.4045,
+      "step": 2532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 12.898319402376009,
+      "learning_rate": 9.450679245445086e-06,
+      "loss": 0.4126,
+      "step": 2533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.410010590115536,
+      "learning_rate": 9.450166011124475e-06,
+      "loss": 0.4489,
+      "step": 2534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.847671386158854,
+      "learning_rate": 9.44965255110595e-06,
+      "loss": 0.3901,
+      "step": 2535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 11.989755316896629,
+      "learning_rate": 9.449138865415554e-06,
+      "loss": 0.4103,
+      "step": 2536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.189219435028959,
+      "learning_rate": 9.448624954079342e-06,
+      "loss": 0.4545,
+      "step": 2537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.224655932011616,
+      "learning_rate": 9.448110817123375e-06,
+      "loss": 0.4385,
+      "step": 2538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.768627517056874,
+      "learning_rate": 9.447596454573728e-06,
+      "loss": 0.397,
+      "step": 2539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.95732947974361,
+      "learning_rate": 9.44708186645649e-06,
+      "loss": 0.4288,
+      "step": 2540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.227917280973172,
+      "learning_rate": 9.446567052797754e-06,
+      "loss": 0.452,
+      "step": 2541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.230769679485382,
+      "learning_rate": 9.446052013623639e-06,
+      "loss": 0.3926,
+      "step": 2542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.0010280786097745,
+      "learning_rate": 9.445536748960257e-06,
+      "loss": 0.4453,
+      "step": 2543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4722826802729252,
+      "learning_rate": 9.445021258833747e-06,
+      "loss": 0.2087,
+      "step": 2544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.885573969443931,
+      "learning_rate": 9.44450554327025e-06,
+      "loss": 0.4334,
+      "step": 2545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.25826786025685,
+      "learning_rate": 9.443989602295923e-06,
+      "loss": 0.4496,
+      "step": 2546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.028543831354116,
+      "learning_rate": 9.44347343593693e-06,
+      "loss": 0.3966,
+      "step": 2547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.004211839436081,
+      "learning_rate": 9.44295704421945e-06,
+      "loss": 0.4078,
+      "step": 2548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 23.873903079094365,
+      "learning_rate": 9.442440427169673e-06,
+      "loss": 0.4283,
+      "step": 2549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.565712901083609,
+      "learning_rate": 9.441923584813803e-06,
+      "loss": 0.376,
+      "step": 2550
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.114471231106574,
+      "learning_rate": 9.441406517178049e-06,
+      "loss": 0.4169,
+      "step": 2551
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.947368315386829,
+      "learning_rate": 9.440889224288637e-06,
+      "loss": 0.4185,
+      "step": 2552
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.172977087711491,
+      "learning_rate": 9.440371706171801e-06,
+      "loss": 0.4373,
+      "step": 2553
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.527069741103883,
+      "learning_rate": 9.439853962853789e-06,
+      "loss": 0.4019,
+      "step": 2554
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.295190709816151,
+      "learning_rate": 9.43933599436086e-06,
+      "loss": 0.3864,
+      "step": 2555
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.946341632847267,
+      "learning_rate": 9.43881780071928e-06,
+      "loss": 0.4141,
+      "step": 2556
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.658918956986284,
+      "learning_rate": 9.438299381955333e-06,
+      "loss": 0.4339,
+      "step": 2557
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.852498042363686,
+      "learning_rate": 9.437780738095311e-06,
+      "loss": 0.4488,
+      "step": 2558
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.874125217943871,
+      "learning_rate": 9.437261869165519e-06,
+      "loss": 0.4428,
+      "step": 2559
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 18.15213663142237,
+      "learning_rate": 9.436742775192272e-06,
+      "loss": 0.4352,
+      "step": 2560
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.299024417820569,
+      "learning_rate": 9.436223456201894e-06,
+      "loss": 0.4513,
+      "step": 2561
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.857537342454432,
+      "learning_rate": 9.435703912220727e-06,
+      "loss": 0.4385,
+      "step": 2562
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.65473716439787,
+      "learning_rate": 9.435184143275117e-06,
+      "loss": 0.4237,
+      "step": 2563
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.32642281754923,
+      "learning_rate": 9.434664149391427e-06,
+      "loss": 0.4208,
+      "step": 2564
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.928982759813626,
+      "learning_rate": 9.43414393059603e-06,
+      "loss": 0.4104,
+      "step": 2565
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.689825114808394,
+      "learning_rate": 9.43362348691531e-06,
+      "loss": 0.4422,
+      "step": 2566
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.164495535460582,
+      "learning_rate": 9.43310281837566e-06,
+      "loss": 0.4544,
+      "step": 2567
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8104066044566864,
+      "learning_rate": 9.432581925003489e-06,
+      "loss": 0.4265,
+      "step": 2568
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8734963409785013,
+      "learning_rate": 9.432060806825211e-06,
+      "loss": 0.4285,
+      "step": 2569
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.726189978715601,
+      "learning_rate": 9.43153946386726e-06,
+      "loss": 0.406,
+      "step": 2570
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.476780408939133,
+      "learning_rate": 9.431017896156074e-06,
+      "loss": 0.4183,
+      "step": 2571
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.8931699732203775,
+      "learning_rate": 9.430496103718108e-06,
+      "loss": 0.4383,
+      "step": 2572
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.609887047222249,
+      "learning_rate": 9.429974086579823e-06,
+      "loss": 0.3932,
+      "step": 2573
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.509567460588471,
+      "learning_rate": 9.429451844767694e-06,
+      "loss": 0.4036,
+      "step": 2574
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.242477978192536,
+      "learning_rate": 9.428929378308209e-06,
+      "loss": 0.4286,
+      "step": 2575
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.478394188556644,
+      "learning_rate": 9.428406687227865e-06,
+      "loss": 0.4373,
+      "step": 2576
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.143500311259018,
+      "learning_rate": 9.427883771553172e-06,
+      "loss": 0.4659,
+      "step": 2577
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.304765871623583,
+      "learning_rate": 9.42736063131065e-06,
+      "loss": 0.4506,
+      "step": 2578
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0517255728926864,
+      "learning_rate": 9.426837266526829e-06,
+      "loss": 0.4177,
+      "step": 2579
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.81807501614713,
+      "learning_rate": 9.426313677228256e-06,
+      "loss": 0.4232,
+      "step": 2580
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.325988305122752,
+      "learning_rate": 9.425789863441484e-06,
+      "loss": 0.4235,
+      "step": 2581
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.619274248247116,
+      "learning_rate": 9.425265825193077e-06,
+      "loss": 0.4342,
+      "step": 2582
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.265979278801934,
+      "learning_rate": 9.424741562509617e-06,
+      "loss": 0.4155,
+      "step": 2583
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.884222688250183,
+      "learning_rate": 9.42421707541769e-06,
+      "loss": 0.4488,
+      "step": 2584
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.954938478378123,
+      "learning_rate": 9.423692363943895e-06,
+      "loss": 0.4382,
+      "step": 2585
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.264207975821252,
+      "learning_rate": 9.423167428114846e-06,
+      "loss": 0.4548,
+      "step": 2586
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.9904668015939,
+      "learning_rate": 9.422642267957165e-06,
+      "loss": 0.4099,
+      "step": 2587
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.1446959278369855,
+      "learning_rate": 9.422116883497488e-06,
+      "loss": 0.4105,
+      "step": 2588
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.693605356697632,
+      "learning_rate": 9.42159127476246e-06,
+      "loss": 0.3912,
+      "step": 2589
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.958676263464777,
+      "learning_rate": 9.421065441778736e-06,
+      "loss": 0.4127,
+      "step": 2590
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.121031090590371,
+      "learning_rate": 9.420539384572988e-06,
+      "loss": 0.4477,
+      "step": 2591
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.728167192798398,
+      "learning_rate": 9.420013103171893e-06,
+      "loss": 0.44,
+      "step": 2592
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.45777227300335804,
+      "learning_rate": 9.419486597602143e-06,
+      "loss": 0.2082,
+      "step": 2593
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.99484165190263,
+      "learning_rate": 9.418959867890443e-06,
+      "loss": 0.4485,
+      "step": 2594
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7026125495460214,
+      "learning_rate": 9.418432914063505e-06,
+      "loss": 0.4056,
+      "step": 2595
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.085670585450937,
+      "learning_rate": 9.417905736148053e-06,
+      "loss": 0.4335,
+      "step": 2596
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.134485744588443,
+      "learning_rate": 9.417378334170826e-06,
+      "loss": 0.4172,
+      "step": 2597
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.1484319263469676,
+      "learning_rate": 9.416850708158571e-06,
+      "loss": 0.392,
+      "step": 2598
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.634822228453592,
+      "learning_rate": 9.416322858138047e-06,
+      "loss": 0.4327,
+      "step": 2599
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.326819986908709,
+      "learning_rate": 9.415794784136027e-06,
+      "loss": 0.4254,
+      "step": 2600
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.491933823795761,
+      "learning_rate": 9.415266486179293e-06,
+      "loss": 0.422,
+      "step": 2601
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.604716385907938,
+      "learning_rate": 9.414737964294636e-06,
+      "loss": 0.4145,
+      "step": 2602
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 9.67698103121612,
+      "learning_rate": 9.414209218508862e-06,
+      "loss": 0.4402,
+      "step": 2603
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.991798199965399,
+      "learning_rate": 9.413680248848788e-06,
+      "loss": 0.4608,
+      "step": 2604
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.452787566208195,
+      "learning_rate": 9.41315105534124e-06,
+      "loss": 0.4233,
+      "step": 2605
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.59705581793515,
+      "learning_rate": 9.412621638013059e-06,
+      "loss": 0.4596,
+      "step": 2606
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.825978578475387,
+      "learning_rate": 9.412091996891097e-06,
+      "loss": 0.4404,
+      "step": 2607
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.0453902766237286,
+      "learning_rate": 9.41156213200221e-06,
+      "loss": 0.4191,
+      "step": 2608
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.2917046828878562,
+      "learning_rate": 9.411032043373275e-06,
+      "loss": 0.4447,
+      "step": 2609
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.949861280804999,
+      "learning_rate": 9.410501731031177e-06,
+      "loss": 0.3935,
+      "step": 2610
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.5611628666709,
+      "learning_rate": 9.409971195002809e-06,
+      "loss": 0.411,
+      "step": 2611
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.098839990977794,
+      "learning_rate": 9.409440435315078e-06,
+      "loss": 0.4414,
+      "step": 2612
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.202731428543846,
+      "learning_rate": 9.408909451994906e-06,
+      "loss": 0.4257,
+      "step": 2613
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.3584186501678905,
+      "learning_rate": 9.40837824506922e-06,
+      "loss": 0.4271,
+      "step": 2614
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.72330246027387,
+      "learning_rate": 9.40784681456496e-06,
+      "loss": 0.4382,
+      "step": 2615
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.392970575117358,
+      "learning_rate": 9.407315160509083e-06,
+      "loss": 0.4408,
+      "step": 2616
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.722922720694175,
+      "learning_rate": 9.406783282928547e-06,
+      "loss": 0.413,
+      "step": 2617
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 10.20609042013056,
+      "learning_rate": 9.40625118185033e-06,
+      "loss": 0.4215,
+      "step": 2618
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.140287940902852,
+      "learning_rate": 9.405718857301419e-06,
+      "loss": 0.4354,
+      "step": 2619
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8328721584630174,
+      "learning_rate": 9.405186309308812e-06,
+      "loss": 0.4042,
+      "step": 2620
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.353240225566997,
+      "learning_rate": 9.404653537899515e-06,
+      "loss": 0.4486,
+      "step": 2621
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.820799161915554,
+      "learning_rate": 9.404120543100553e-06,
+      "loss": 0.4225,
+      "step": 2622
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.407188600940772,
+      "learning_rate": 9.403587324938952e-06,
+      "loss": 0.4371,
+      "step": 2623
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.388869746257234,
+      "learning_rate": 9.40305388344176e-06,
+      "loss": 0.4225,
+      "step": 2624
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.592791078394982,
+      "learning_rate": 9.40252021863603e-06,
+      "loss": 0.3781,
+      "step": 2625
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.1969877310869075,
+      "learning_rate": 9.401986330548828e-06,
+      "loss": 0.4357,
+      "step": 2626
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.335656561259848,
+      "learning_rate": 9.40145221920723e-06,
+      "loss": 0.378,
+      "step": 2627
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.5652106965524455,
+      "learning_rate": 9.400917884638325e-06,
+      "loss": 0.441,
+      "step": 2628
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.706656383861546,
+      "learning_rate": 9.400383326869214e-06,
+      "loss": 0.4208,
+      "step": 2629
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5834214835395386,
+      "learning_rate": 9.399848545927006e-06,
+      "loss": 0.4109,
+      "step": 2630
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.760355481466171,
+      "learning_rate": 9.399313541838823e-06,
+      "loss": 0.4498,
+      "step": 2631
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.566316269574721,
+      "learning_rate": 9.398778314631801e-06,
+      "loss": 0.2047,
+      "step": 2632
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 7.014588985855932,
+      "learning_rate": 9.398242864333084e-06,
+      "loss": 0.3972,
+      "step": 2633
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.019055623596467,
+      "learning_rate": 9.397707190969826e-06,
+      "loss": 0.4069,
+      "step": 2634
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.5279820567866125,
+      "learning_rate": 9.397171294569198e-06,
+      "loss": 0.3942,
+      "step": 2635
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.860391460426057,
+      "learning_rate": 9.396635175158377e-06,
+      "loss": 0.4176,
+      "step": 2636
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.344369037887725,
+      "learning_rate": 9.396098832764555e-06,
+      "loss": 0.4024,
+      "step": 2637
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4547385888964791,
+      "learning_rate": 9.395562267414932e-06,
+      "loss": 0.1904,
+      "step": 2638
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 56.311271289283496,
+      "learning_rate": 9.395025479136722e-06,
+      "loss": 0.3992,
+      "step": 2639
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.253045189116717,
+      "learning_rate": 9.394488467957147e-06,
+      "loss": 0.406,
+      "step": 2640
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.2532557125888335,
+      "learning_rate": 9.393951233903442e-06,
+      "loss": 0.4474,
+      "step": 2641
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.541884650172126,
+      "learning_rate": 9.393413777002858e-06,
+      "loss": 0.4334,
+      "step": 2642
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.864222683738525,
+      "learning_rate": 9.39287609728265e-06,
+      "loss": 0.4176,
+      "step": 2643
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 24.162309281173698,
+      "learning_rate": 9.392338194770087e-06,
+      "loss": 0.4404,
+      "step": 2644
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.560904240044689,
+      "learning_rate": 9.391800069492452e-06,
+      "loss": 0.4621,
+      "step": 2645
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.049106192544743,
+      "learning_rate": 9.391261721477034e-06,
+      "loss": 0.4309,
+      "step": 2646
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 8.960115344154042,
+      "learning_rate": 9.39072315075114e-06,
+      "loss": 0.4199,
+      "step": 2647
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.3647220540016614,
+      "learning_rate": 9.39018435734208e-06,
+      "loss": 0.424,
+      "step": 2648
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 29.197978848494643,
+      "learning_rate": 9.389645341277183e-06,
+      "loss": 0.4388,
+      "step": 2649
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 4.681192367812698,
+      "learning_rate": 9.389106102583787e-06,
+      "loss": 0.4592,
+      "step": 2650
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.339645345242785,
+      "learning_rate": 9.388566641289238e-06,
+      "loss": 0.4556,
+      "step": 2651
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.207050367262977,
+      "learning_rate": 9.388026957420895e-06,
+      "loss": 0.4246,
+      "step": 2652
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.8876913907240676,
+      "learning_rate": 9.387487051006133e-06,
+      "loss": 0.4223,
+      "step": 2653
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.391826047867097,
+      "learning_rate": 9.386946922072329e-06,
+      "loss": 0.3869,
+      "step": 2654
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7622546682385547,
+      "learning_rate": 9.386406570646881e-06,
+      "loss": 0.4434,
+      "step": 2655
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.993764066806853,
+      "learning_rate": 9.385865996757192e-06,
+      "loss": 0.4302,
+      "step": 2656
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4143225296764841,
+      "learning_rate": 9.385325200430679e-06,
+      "loss": 0.2285,
+      "step": 2657
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.1875435816469615,
+      "learning_rate": 9.384784181694767e-06,
+      "loss": 0.4208,
+      "step": 2658
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.050341030711866,
+      "learning_rate": 9.384242940576898e-06,
+      "loss": 0.4233,
+      "step": 2659
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 6.454528192475213,
+      "learning_rate": 9.38370147710452e-06,
+      "loss": 0.4222,
+      "step": 2660
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.26692024727363,
+      "learning_rate": 9.383159791305096e-06,
+      "loss": 0.3742,
+      "step": 2661
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.464449481202213,
+      "learning_rate": 9.382617883206096e-06,
+      "loss": 0.4242,
+      "step": 2662
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5395058609035606,
+      "learning_rate": 9.382075752835006e-06,
+      "loss": 0.4309,
+      "step": 2663
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.361855400776557,
+      "learning_rate": 9.381533400219319e-06,
+      "loss": 0.4366,
+      "step": 2664
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.047127861676178,
+      "learning_rate": 9.380990825386542e-06,
+      "loss": 0.4573,
+      "step": 2665
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.997619738035955,
+      "learning_rate": 9.380448028364195e-06,
+      "loss": 0.4178,
+      "step": 2666
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.1204698357899723,
+      "learning_rate": 9.379905009179804e-06,
+      "loss": 0.4167,
+      "step": 2667
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.8880427954294445,
+      "learning_rate": 9.379361767860912e-06,
+      "loss": 0.4003,
+      "step": 2668
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.484849075175314,
+      "learning_rate": 9.378818304435066e-06,
+      "loss": 0.4163,
+      "step": 2669
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.437646763564684,
+      "learning_rate": 9.378274618929833e-06,
+      "loss": 0.4173,
+      "step": 2670
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.380664736000623,
+      "learning_rate": 9.377730711372786e-06,
+      "loss": 0.4269,
+      "step": 2671
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.39395930223513,
+      "learning_rate": 9.37718658179151e-06,
+      "loss": 0.4593,
+      "step": 2672
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.5045792368605335,
+      "learning_rate": 9.376642230213597e-06,
+      "loss": 0.4213,
+      "step": 2673
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.050607381833903,
+      "learning_rate": 9.376097656666663e-06,
+      "loss": 0.4092,
+      "step": 2674
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.544744967888022,
+      "learning_rate": 9.37555286117832e-06,
+      "loss": 0.4647,
+      "step": 2675
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7076416080960333,
+      "learning_rate": 9.375007843776202e-06,
+      "loss": 0.4054,
+      "step": 2676
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.560434270081909,
+      "learning_rate": 9.37446260448795e-06,
+      "loss": 0.4148,
+      "step": 2677
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.0169425210823215,
+      "learning_rate": 9.373917143341216e-06,
+      "loss": 0.4135,
+      "step": 2678
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.024513935546247,
+      "learning_rate": 9.373371460363665e-06,
+      "loss": 0.4066,
+      "step": 2679
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9344205526265825,
+      "learning_rate": 9.372825555582969e-06,
+      "loss": 0.4411,
+      "step": 2680
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.162507540361819,
+      "learning_rate": 9.37227942902682e-06,
+      "loss": 0.4505,
+      "step": 2681
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.2963024642941035,
+      "learning_rate": 9.371733080722911e-06,
+      "loss": 0.4316,
+      "step": 2682
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.247574662544041,
+      "learning_rate": 9.371186510698954e-06,
+      "loss": 0.4344,
+      "step": 2683
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.9580088283256587,
+      "learning_rate": 9.370639718982668e-06,
+      "loss": 0.417,
+      "step": 2684
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.418235587928655,
+      "learning_rate": 9.370092705601783e-06,
+      "loss": 0.4464,
+      "step": 2685
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.437991603937274,
+      "learning_rate": 9.369545470584045e-06,
+      "loss": 0.3919,
+      "step": 2686
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.701349127893263,
+      "learning_rate": 9.368998013957205e-06,
+      "loss": 0.4608,
+      "step": 2687
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8550753018611017,
+      "learning_rate": 9.36845033574903e-06,
+      "loss": 0.4385,
+      "step": 2688
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.718717007416196,
+      "learning_rate": 9.367902435987297e-06,
+      "loss": 0.4274,
+      "step": 2689
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48204956577213637,
+      "learning_rate": 9.367354314699791e-06,
+      "loss": 0.2047,
+      "step": 2690
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8229559055772646,
+      "learning_rate": 9.366805971914313e-06,
+      "loss": 0.4536,
+      "step": 2691
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.092851156839028,
+      "learning_rate": 9.366257407658673e-06,
+      "loss": 0.375,
+      "step": 2692
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 25.05429942420422,
+      "learning_rate": 9.365708621960692e-06,
+      "loss": 0.4469,
+      "step": 2693
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.220900588485158,
+      "learning_rate": 9.365159614848204e-06,
+      "loss": 0.4244,
+      "step": 2694
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.454666233115918,
+      "learning_rate": 9.364610386349048e-06,
+      "loss": 0.433,
+      "step": 2695
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.053752572327515,
+      "learning_rate": 9.364060936491086e-06,
+      "loss": 0.4147,
+      "step": 2696
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.420929073529593,
+      "learning_rate": 9.36351126530218e-06,
+      "loss": 0.4333,
+      "step": 2697
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.086235994016695,
+      "learning_rate": 9.362961372810208e-06,
+      "loss": 0.3829,
+      "step": 2698
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.447905552497702,
+      "learning_rate": 9.36241125904306e-06,
+      "loss": 0.4256,
+      "step": 2699
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.223465373266618,
+      "learning_rate": 9.361860924028635e-06,
+      "loss": 0.403,
+      "step": 2700
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.219525706507221,
+      "learning_rate": 9.361310367794845e-06,
+      "loss": 0.4044,
+      "step": 2701
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.031764408031619,
+      "learning_rate": 9.36075959036961e-06,
+      "loss": 0.4,
+      "step": 2702
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.48476297463059836,
+      "learning_rate": 9.360208591780867e-06,
+      "loss": 0.1999,
+      "step": 2703
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.41071784600345523,
+      "learning_rate": 9.35965737205656e-06,
+      "loss": 0.1995,
+      "step": 2704
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.432666988551365,
+      "learning_rate": 9.359105931224643e-06,
+      "loss": 0.4215,
+      "step": 2705
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.118392438519834,
+      "learning_rate": 9.358554269313085e-06,
+      "loss": 0.4261,
+      "step": 2706
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.12769960409258,
+      "learning_rate": 9.358002386349862e-06,
+      "loss": 0.4052,
+      "step": 2707
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.5047871612463135,
+      "learning_rate": 9.35745028236297e-06,
+      "loss": 0.4159,
+      "step": 2708
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.468761802114591,
+      "learning_rate": 9.356897957380404e-06,
+      "loss": 0.4164,
+      "step": 2709
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.347802746370226,
+      "learning_rate": 9.356345411430176e-06,
+      "loss": 0.4385,
+      "step": 2710
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.499868523841558,
+      "learning_rate": 9.355792644540313e-06,
+      "loss": 0.4115,
+      "step": 2711
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.908322176337467,
+      "learning_rate": 9.355239656738849e-06,
+      "loss": 0.4068,
+      "step": 2712
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.554092702801891,
+      "learning_rate": 9.354686448053828e-06,
+      "loss": 0.4592,
+      "step": 2713
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.913009930483106,
+      "learning_rate": 9.354133018513305e-06,
+      "loss": 0.4585,
+      "step": 2714
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.811400997389348,
+      "learning_rate": 9.353579368145353e-06,
+      "loss": 0.4329,
+      "step": 2715
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6038521278009426,
+      "learning_rate": 9.353025496978047e-06,
+      "loss": 0.4209,
+      "step": 2716
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.879092302261387,
+      "learning_rate": 9.352471405039481e-06,
+      "loss": 0.4241,
+      "step": 2717
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.177712921166728,
+      "learning_rate": 9.351917092357754e-06,
+      "loss": 0.4052,
+      "step": 2718
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.590780543106286,
+      "learning_rate": 9.351362558960982e-06,
+      "loss": 0.4073,
+      "step": 2719
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.028790940150286,
+      "learning_rate": 9.350807804877286e-06,
+      "loss": 0.4395,
+      "step": 2720
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.5373461366626024,
+      "learning_rate": 9.350252830134802e-06,
+      "loss": 0.4383,
+      "step": 2721
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.49193326758304,
+      "learning_rate": 9.34969763476168e-06,
+      "loss": 0.4451,
+      "step": 2722
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.756076045080907,
+      "learning_rate": 9.34914221878607e-06,
+      "loss": 0.4185,
+      "step": 2723
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.5372112939794995,
+      "learning_rate": 9.34858658223615e-06,
+      "loss": 0.4485,
+      "step": 2724
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.606538103824933,
+      "learning_rate": 9.348030725140095e-06,
+      "loss": 0.3984,
+      "step": 2725
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.85160072703201,
+      "learning_rate": 9.347474647526095e-06,
+      "loss": 0.4161,
+      "step": 2726
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.050388862326749,
+      "learning_rate": 9.346918349422356e-06,
+      "loss": 0.3929,
+      "step": 2727
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.7910398737292565,
+      "learning_rate": 9.34636183085709e-06,
+      "loss": 0.4372,
+      "step": 2728
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 17.202790787637863,
+      "learning_rate": 9.345805091858522e-06,
+      "loss": 0.4191,
+      "step": 2729
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.806845599478608,
+      "learning_rate": 9.345248132454887e-06,
+      "loss": 0.3797,
+      "step": 2730
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.663671419864553,
+      "learning_rate": 9.344690952674434e-06,
+      "loss": 0.3968,
+      "step": 2731
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.513718440280297,
+      "learning_rate": 9.34413355254542e-06,
+      "loss": 0.4049,
+      "step": 2732
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.244699896041918,
+      "learning_rate": 9.343575932096116e-06,
+      "loss": 0.409,
+      "step": 2733
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.806354738251809,
+      "learning_rate": 9.343018091354798e-06,
+      "loss": 0.4409,
+      "step": 2734
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.871442573946455,
+      "learning_rate": 9.342460030349764e-06,
+      "loss": 0.3364,
+      "step": 2735
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.805736723895427,
+      "learning_rate": 9.341901749109315e-06,
+      "loss": 0.3949,
+      "step": 2736
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.930720073703612,
+      "learning_rate": 9.341343247661765e-06,
+      "loss": 0.4503,
+      "step": 2737
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9478304585818568,
+      "learning_rate": 9.340784526035438e-06,
+      "loss": 0.2587,
+      "step": 2738
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.449688130792003,
+      "learning_rate": 9.34022558425867e-06,
+      "loss": 0.411,
+      "step": 2739
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.820618664755517,
+      "learning_rate": 9.339666422359815e-06,
+      "loss": 0.434,
+      "step": 2740
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.452932768443564,
+      "learning_rate": 9.339107040367223e-06,
+      "loss": 0.4381,
+      "step": 2741
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.206420582138921,
+      "learning_rate": 9.33854743830927e-06,
+      "loss": 0.406,
+      "step": 2742
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.946540055755152,
+      "learning_rate": 9.337987616214335e-06,
+      "loss": 0.4513,
+      "step": 2743
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.053333912835885,
+      "learning_rate": 9.33742757411081e-06,
+      "loss": 0.4409,
+      "step": 2744
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.3477725642746305,
+      "learning_rate": 9.3368673120271e-06,
+      "loss": 0.4618,
+      "step": 2745
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.74250975516451,
+      "learning_rate": 9.336306829991619e-06,
+      "loss": 0.4577,
+      "step": 2746
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.040681506530569,
+      "learning_rate": 9.335746128032793e-06,
+      "loss": 0.4317,
+      "step": 2747
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6115674950179839,
+      "learning_rate": 9.335185206179058e-06,
+      "loss": 0.2187,
+      "step": 2748
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.446451523180164,
+      "learning_rate": 9.334624064458864e-06,
+      "loss": 0.4214,
+      "step": 2749
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5345064436824104,
+      "learning_rate": 9.334062702900668e-06,
+      "loss": 0.2087,
+      "step": 2750
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.582565803465363,
+      "learning_rate": 9.333501121532942e-06,
+      "loss": 0.3852,
+      "step": 2751
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 15.494021314490185,
+      "learning_rate": 9.332939320384166e-06,
+      "loss": 0.4156,
+      "step": 2752
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 7.835471301241473,
+      "learning_rate": 9.332377299482835e-06,
+      "loss": 0.4293,
+      "step": 2753
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7345796644209073,
+      "learning_rate": 9.331815058857452e-06,
+      "loss": 0.4371,
+      "step": 2754
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.76834225139677,
+      "learning_rate": 9.33125259853653e-06,
+      "loss": 0.4191,
+      "step": 2755
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.687625072294203,
+      "learning_rate": 9.3306899185486e-06,
+      "loss": 0.4342,
+      "step": 2756
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 10.0133298000449,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.4347,
+      "step": 2757
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.772476284063403,
+      "learning_rate": 9.329563899685862e-06,
+      "loss": 0.4639,
+      "step": 2758
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.179282367903539,
+      "learning_rate": 9.329000560868167e-06,
+      "loss": 0.4577,
+      "step": 2759
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 16.61049736103823,
+      "learning_rate": 9.328437002497675e-06,
+      "loss": 0.459,
+      "step": 2760
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.083390498408454,
+      "learning_rate": 9.32787322460297e-06,
+      "loss": 0.4221,
+      "step": 2761
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.755813864375804,
+      "learning_rate": 9.327309227212644e-06,
+      "loss": 0.4127,
+      "step": 2762
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.753305005958116,
+      "learning_rate": 9.326745010355303e-06,
+      "loss": 0.417,
+      "step": 2763
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.963134952892425,
+      "learning_rate": 9.32618057405956e-06,
+      "loss": 0.4209,
+      "step": 2764
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.729009032794378,
+      "learning_rate": 9.325615918354043e-06,
+      "loss": 0.4387,
+      "step": 2765
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.181615251734948,
+      "learning_rate": 9.32505104326739e-06,
+      "loss": 0.4161,
+      "step": 2766
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.139661799356277,
+      "learning_rate": 9.324485948828248e-06,
+      "loss": 0.474,
+      "step": 2767
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.685519842093186,
+      "learning_rate": 9.323920635065275e-06,
+      "loss": 0.4196,
+      "step": 2768
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.057430487770868,
+      "learning_rate": 9.323355102007146e-06,
+      "loss": 0.4021,
+      "step": 2769
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.046429471614623,
+      "learning_rate": 9.322789349682541e-06,
+      "loss": 0.4211,
+      "step": 2770
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.259329179147788,
+      "learning_rate": 9.322223378120154e-06,
+      "loss": 0.45,
+      "step": 2771
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 9.498517374010838,
+      "learning_rate": 9.321657187348689e-06,
+      "loss": 0.4378,
+      "step": 2772
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.450745871227276,
+      "learning_rate": 9.32109077739686e-06,
+      "loss": 0.4419,
+      "step": 2773
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.670851624140156,
+      "learning_rate": 9.320524148293393e-06,
+      "loss": 0.4239,
+      "step": 2774
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.641420874244472,
+      "learning_rate": 9.319957300067029e-06,
+      "loss": 0.4673,
+      "step": 2775
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 23.40257614915093,
+      "learning_rate": 9.319390232746514e-06,
+      "loss": 0.4283,
+      "step": 2776
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.836853036410664,
+      "learning_rate": 9.318822946360608e-06,
+      "loss": 0.4246,
+      "step": 2777
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.502381558028888,
+      "learning_rate": 9.318255440938082e-06,
+      "loss": 0.4257,
+      "step": 2778
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8205991342613745,
+      "learning_rate": 9.31768771650772e-06,
+      "loss": 0.4133,
+      "step": 2779
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.264014001430785,
+      "learning_rate": 9.31711977309831e-06,
+      "loss": 0.4286,
+      "step": 2780
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9598308995895635,
+      "learning_rate": 9.316551610738664e-06,
+      "loss": 0.2529,
+      "step": 2781
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.0003001843137564,
+      "learning_rate": 9.31598322945759e-06,
+      "loss": 0.4303,
+      "step": 2782
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.467095371663017,
+      "learning_rate": 9.315414629283918e-06,
+      "loss": 0.4029,
+      "step": 2783
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.7119388336788175,
+      "learning_rate": 9.314845810246487e-06,
+      "loss": 0.4474,
+      "step": 2784
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.282202285447534,
+      "learning_rate": 9.314276772374143e-06,
+      "loss": 0.3867,
+      "step": 2785
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.37447520114226884,
+      "learning_rate": 9.313707515695745e-06,
+      "loss": 0.2096,
+      "step": 2786
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 11.998643642604707,
+      "learning_rate": 9.313138040240167e-06,
+      "loss": 0.4731,
+      "step": 2787
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.43737840365533126,
+      "learning_rate": 9.312568346036288e-06,
+      "loss": 0.2153,
+      "step": 2788
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.8504401005459066,
+      "learning_rate": 9.311998433113002e-06,
+      "loss": 0.4403,
+      "step": 2789
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.620642399312722,
+      "learning_rate": 9.311428301499215e-06,
+      "loss": 0.4333,
+      "step": 2790
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.028308993088884,
+      "learning_rate": 9.31085795122384e-06,
+      "loss": 0.4238,
+      "step": 2791
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.682690032906842,
+      "learning_rate": 9.310287382315804e-06,
+      "loss": 0.3827,
+      "step": 2792
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.636363095450747,
+      "learning_rate": 9.309716594804044e-06,
+      "loss": 0.4149,
+      "step": 2793
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.818995266997174,
+      "learning_rate": 9.309145588717507e-06,
+      "loss": 0.4272,
+      "step": 2794
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.6731111131539937,
+      "learning_rate": 9.308574364085157e-06,
+      "loss": 0.3966,
+      "step": 2795
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.8922563577915072,
+      "learning_rate": 9.30800292093596e-06,
+      "loss": 0.4204,
+      "step": 2796
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 8.342693160389675,
+      "learning_rate": 9.307431259298903e-06,
+      "loss": 0.3987,
+      "step": 2797
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.438977876362387,
+      "learning_rate": 9.306859379202973e-06,
+      "loss": 0.4491,
+      "step": 2798
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.284079653616667,
+      "learning_rate": 9.306287280677176e-06,
+      "loss": 0.4411,
+      "step": 2799
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.926871119227945,
+      "learning_rate": 9.305714963750528e-06,
+      "loss": 0.4214,
+      "step": 2800
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.643758861586152,
+      "learning_rate": 9.305142428452055e-06,
+      "loss": 0.4085,
+      "step": 2801
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.355991594586906,
+      "learning_rate": 9.304569674810794e-06,
+      "loss": 0.4496,
+      "step": 2802
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.987463740225345,
+      "learning_rate": 9.303996702855792e-06,
+      "loss": 0.4385,
+      "step": 2803
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 4.457400679872922,
+      "learning_rate": 9.30342351261611e-06,
+      "loss": 0.411,
+      "step": 2804
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 13.829350099141752,
+      "learning_rate": 9.302850104120815e-06,
+      "loss": 0.4212,
+      "step": 2805
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 37.468534553274296,
+      "learning_rate": 9.302276477398992e-06,
+      "loss": 0.4402,
+      "step": 2806
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.158955121113772,
+      "learning_rate": 9.301702632479734e-06,
+      "loss": 0.4004,
+      "step": 2807
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.0089725121040525,
+      "learning_rate": 9.301128569392142e-06,
+      "loss": 0.3994,
+      "step": 2808
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.6254906978715224,
+      "learning_rate": 9.300554288165331e-06,
+      "loss": 0.4181,
+      "step": 2809
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9626328609792045,
+      "learning_rate": 9.299979788828426e-06,
+      "loss": 0.4176,
+      "step": 2810
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.9387303934575857,
+      "learning_rate": 9.299405071410568e-06,
+      "loss": 0.3935,
+      "step": 2811
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.0942492960812,
+      "learning_rate": 9.298830135940899e-06,
+      "loss": 0.4451,
+      "step": 2812
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.86243377513068,
+      "learning_rate": 9.29825498244858e-06,
+      "loss": 0.4035,
+      "step": 2813
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.100571239184572,
+      "learning_rate": 9.297679610962783e-06,
+      "loss": 0.4369,
+      "step": 2814
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.9743293702713824,
+      "learning_rate": 9.297104021512687e-06,
+      "loss": 0.457,
+      "step": 2815
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.251235783160864,
+      "learning_rate": 9.296528214127484e-06,
+      "loss": 0.4214,
+      "step": 2816
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.375921686312159,
+      "learning_rate": 9.29595218883638e-06,
+      "loss": 0.442,
+      "step": 2817
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 11.730223166268477,
+      "learning_rate": 9.295375945668583e-06,
+      "loss": 0.4588,
+      "step": 2818
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.216119776968065,
+      "learning_rate": 9.294799484653323e-06,
+      "loss": 0.452,
+      "step": 2819
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.887993199431595,
+      "learning_rate": 9.294222805819834e-06,
+      "loss": 0.4187,
+      "step": 2820
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.4931771151797415,
+      "learning_rate": 9.293645909197366e-06,
+      "loss": 0.4342,
+      "step": 2821
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.954591807099251,
+      "learning_rate": 9.293068794815175e-06,
+      "loss": 0.4155,
+      "step": 2822
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.693215056772339,
+      "learning_rate": 9.29249146270253e-06,
+      "loss": 0.4311,
+      "step": 2823
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.386693474220225,
+      "learning_rate": 9.291913912888714e-06,
+      "loss": 0.4467,
+      "step": 2824
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.903747208411555,
+      "learning_rate": 9.291336145403016e-06,
+      "loss": 0.4123,
+      "step": 2825
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.567782948852773,
+      "learning_rate": 9.29075816027474e-06,
+      "loss": 0.4575,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.062922615790026,
+      "learning_rate": 9.290179957533198e-06,
+      "loss": 0.404,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.906356678341168,
+      "learning_rate": 9.289601537207715e-06,
+      "loss": 0.3986,
+      "step": 2828
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.5638317283830325,
+      "learning_rate": 9.289022899327627e-06,
+      "loss": 0.4344,
+      "step": 2829
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.54683945462736,
+      "learning_rate": 9.288444043922283e-06,
+      "loss": 0.3978,
+      "step": 2830
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.482732745260901,
+      "learning_rate": 9.287864971021035e-06,
+      "loss": 0.442,
+      "step": 2831
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.756632249083126,
+      "learning_rate": 9.287285680653254e-06,
+      "loss": 0.4103,
+      "step": 2832
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.30790057808906,
+      "learning_rate": 9.286706172848325e-06,
+      "loss": 0.4599,
+      "step": 2833
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.484578431566688,
+      "learning_rate": 9.286126447635632e-06,
+      "loss": 0.4057,
+      "step": 2834
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.944747870543114,
+      "learning_rate": 9.285546505044578e-06,
+      "loss": 0.4486,
+      "step": 2835
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.612429654456276,
+      "learning_rate": 9.28496634510458e-06,
+      "loss": 0.4566,
+      "step": 2836
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.976504939008477,
+      "learning_rate": 9.284385967845056e-06,
+      "loss": 0.4198,
+      "step": 2837
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.721289003075783,
+      "learning_rate": 9.283805373295444e-06,
+      "loss": 0.3838,
+      "step": 2838
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.20163396158375,
+      "learning_rate": 9.283224561485191e-06,
+      "loss": 0.4265,
+      "step": 2839
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.628898349123452,
+      "learning_rate": 9.282643532443751e-06,
+      "loss": 0.4392,
+      "step": 2840
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.539478771632611,
+      "learning_rate": 9.282062286200595e-06,
+      "loss": 0.4362,
+      "step": 2841
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.47936134819206533,
+      "learning_rate": 9.281480822785197e-06,
+      "loss": 0.1874,
+      "step": 2842
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.217707622975097,
+      "learning_rate": 9.280899142227052e-06,
+      "loss": 0.4141,
+      "step": 2843
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.436909390433916,
+      "learning_rate": 9.280317244555659e-06,
+      "loss": 0.4365,
+      "step": 2844
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5363257154155496,
+      "learning_rate": 9.279735129800531e-06,
+      "loss": 0.4268,
+      "step": 2845
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.066611659672954,
+      "learning_rate": 9.27915279799119e-06,
+      "loss": 0.3832,
+      "step": 2846
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.37757052517719897,
+      "learning_rate": 9.278570249157166e-06,
+      "loss": 0.2176,
+      "step": 2847
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 12.325640601348445,
+      "learning_rate": 9.277987483328012e-06,
+      "loss": 0.4223,
+      "step": 2848
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.763750024983852,
+      "learning_rate": 9.277404500533278e-06,
+      "loss": 0.4561,
+      "step": 2849
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.746179736982473,
+      "learning_rate": 9.276821300802535e-06,
+      "loss": 0.4395,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.089632920583343,
+      "learning_rate": 9.276237884165356e-06,
+      "loss": 0.3984,
+      "step": 2851
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.082645094912765,
+      "learning_rate": 9.275654250651334e-06,
+      "loss": 0.4177,
+      "step": 2852
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.497436841270476,
+      "learning_rate": 9.275070400290068e-06,
+      "loss": 0.4267,
+      "step": 2853
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.549891210493909,
+      "learning_rate": 9.27448633311117e-06,
+      "loss": 0.4393,
+      "step": 2854
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.470375751300592,
+      "learning_rate": 9.273902049144259e-06,
+      "loss": 0.433,
+      "step": 2855
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.310897974626245,
+      "learning_rate": 9.27331754841897e-06,
+      "loss": 0.4076,
+      "step": 2856
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.43314100571231906,
+      "learning_rate": 9.272732830964948e-06,
+      "loss": 0.1846,
+      "step": 2857
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.6267595390724425,
+      "learning_rate": 9.272147896811844e-06,
+      "loss": 0.4346,
+      "step": 2858
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.329495801336259,
+      "learning_rate": 9.27156274598933e-06,
+      "loss": 0.3929,
+      "step": 2859
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.005417271475056,
+      "learning_rate": 9.27097737852708e-06,
+      "loss": 0.3864,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.970069147478256,
+      "learning_rate": 9.270391794454779e-06,
+      "loss": 0.4854,
+      "step": 2861
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.226436983162832,
+      "learning_rate": 9.26980599380213e-06,
+      "loss": 0.4269,
+      "step": 2862
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 38.21001843424305,
+      "learning_rate": 9.26921997659884e-06,
+      "loss": 0.4328,
+      "step": 2863
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.2050516416762305,
+      "learning_rate": 9.26863374287463e-06,
+      "loss": 0.437,
+      "step": 2864
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.540002709221275,
+      "learning_rate": 9.268047292659236e-06,
+      "loss": 0.4248,
+      "step": 2865
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.313310012919356,
+      "learning_rate": 9.267460625982398e-06,
+      "loss": 0.4243,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.4948788555574235,
+      "learning_rate": 9.266873742873867e-06,
+      "loss": 0.4362,
+      "step": 2867
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.296348668680526,
+      "learning_rate": 9.266286643363414e-06,
+      "loss": 0.4316,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.44820282696579,
+      "learning_rate": 9.265699327480809e-06,
+      "loss": 0.4144,
+      "step": 2869
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.3457094289988705,
+      "learning_rate": 9.265111795255841e-06,
+      "loss": 0.3925,
+      "step": 2870
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.417235633422592,
+      "learning_rate": 9.264524046718309e-06,
+      "loss": 0.4229,
+      "step": 2871
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.838976473722317,
+      "learning_rate": 9.26393608189802e-06,
+      "loss": 0.3847,
+      "step": 2872
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.711260878065159,
+      "learning_rate": 9.263347900824795e-06,
+      "loss": 0.4196,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.809565340137739,
+      "learning_rate": 9.262759503528463e-06,
+      "loss": 0.4665,
+      "step": 2874
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.269577593080466,
+      "learning_rate": 9.262170890038867e-06,
+      "loss": 0.4704,
+      "step": 2875
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.793767650738679,
+      "learning_rate": 9.261582060385857e-06,
+      "loss": 0.4298,
+      "step": 2876
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.917279195748421,
+      "learning_rate": 9.2609930145993e-06,
+      "loss": 0.4261,
+      "step": 2877
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.467059262002994,
+      "learning_rate": 9.260403752709068e-06,
+      "loss": 0.4224,
+      "step": 2878
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.786467028353113,
+      "learning_rate": 9.259814274745047e-06,
+      "loss": 0.4212,
+      "step": 2879
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.499312781803817,
+      "learning_rate": 9.259224580737136e-06,
+      "loss": 0.4312,
+      "step": 2880
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.096460840000827,
+      "learning_rate": 9.25863467071524e-06,
+      "loss": 0.417,
+      "step": 2881
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.0403284550280905,
+      "learning_rate": 9.258044544709276e-06,
+      "loss": 0.3919,
+      "step": 2882
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.718408636671889,
+      "learning_rate": 9.257454202749174e-06,
+      "loss": 0.4475,
+      "step": 2883
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4365360234577663,
+      "learning_rate": 9.256863644864877e-06,
+      "loss": 0.2005,
+      "step": 2884
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 16.727161430365975,
+      "learning_rate": 9.256272871086333e-06,
+      "loss": 0.4553,
+      "step": 2885
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.918284635843323,
+      "learning_rate": 9.255681881443506e-06,
+      "loss": 0.4352,
+      "step": 2886
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.801502407848492,
+      "learning_rate": 9.255090675966368e-06,
+      "loss": 0.4,
+      "step": 2887
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.697253533136624,
+      "learning_rate": 9.2544992546849e-06,
+      "loss": 0.427,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.0645820545058475,
+      "learning_rate": 9.253907617629105e-06,
+      "loss": 0.4444,
+      "step": 2889
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.010234111887671,
+      "learning_rate": 9.253315764828984e-06,
+      "loss": 0.4538,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.3832204570726549,
+      "learning_rate": 9.25272369631455e-06,
+      "loss": 0.2041,
+      "step": 2891
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.015017972081327,
+      "learning_rate": 9.252131412115838e-06,
+      "loss": 0.4394,
+      "step": 2892
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.7673260351546745,
+      "learning_rate": 9.251538912262883e-06,
+      "loss": 0.3964,
+      "step": 2893
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.287426371644106,
+      "learning_rate": 9.250946196785736e-06,
+      "loss": 0.3856,
+      "step": 2894
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.327133691048507,
+      "learning_rate": 9.250353265714454e-06,
+      "loss": 0.4167,
+      "step": 2895
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.558878820263688,
+      "learning_rate": 9.249760119079114e-06,
+      "loss": 0.4168,
+      "step": 2896
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.146262235244052,
+      "learning_rate": 9.249166756909793e-06,
+      "loss": 0.4378,
+      "step": 2897
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.160957755471764,
+      "learning_rate": 9.24857317923659e-06,
+      "loss": 0.4275,
+      "step": 2898
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.633427086715611,
+      "learning_rate": 9.247979386089606e-06,
+      "loss": 0.4666,
+      "step": 2899
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.318337563623115,
+      "learning_rate": 9.247385377498957e-06,
+      "loss": 0.4033,
+      "step": 2900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.546535472002535,
+      "learning_rate": 9.246791153494767e-06,
+      "loss": 0.4221,
+      "step": 2901
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.054265952803588,
+      "learning_rate": 9.246196714107176e-06,
+      "loss": 0.4258,
+      "step": 2902
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.119367944958212,
+      "learning_rate": 9.24560205936633e-06,
+      "loss": 0.4525,
+      "step": 2903
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5789845232056265,
+      "learning_rate": 9.245007189302392e-06,
+      "loss": 0.4024,
+      "step": 2904
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.1496981578695,
+      "learning_rate": 9.244412103945525e-06,
+      "loss": 0.4491,
+      "step": 2905
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.658642715117026,
+      "learning_rate": 9.243816803325918e-06,
+      "loss": 0.4083,
+      "step": 2906
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.258968387031044,
+      "learning_rate": 9.243221287473755e-06,
+      "loss": 0.4486,
+      "step": 2907
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 8.851245419143986,
+      "learning_rate": 9.242625556419243e-06,
+      "loss": 0.4196,
+      "step": 2908
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.739091815967702,
+      "learning_rate": 9.242029610192596e-06,
+      "loss": 0.3842,
+      "step": 2909
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.7517849414529683,
+      "learning_rate": 9.241433448824036e-06,
+      "loss": 0.4622,
+      "step": 2910
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.43286819315746,
+      "learning_rate": 9.240837072343797e-06,
+      "loss": 0.3999,
+      "step": 2911
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.24663961369997,
+      "learning_rate": 9.24024048078213e-06,
+      "loss": 0.4206,
+      "step": 2912
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.584906394529119,
+      "learning_rate": 9.239643674169289e-06,
+      "loss": 0.4157,
+      "step": 2913
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.499284544861238,
+      "learning_rate": 9.239046652535544e-06,
+      "loss": 0.3892,
+      "step": 2914
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.4224246086029595,
+      "learning_rate": 9.238449415911173e-06,
+      "loss": 0.4278,
+      "step": 2915
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.7708504554899,
+      "learning_rate": 9.237851964326466e-06,
+      "loss": 0.4503,
+      "step": 2916
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.3260982703071855,
+      "learning_rate": 9.23725429781172e-06,
+      "loss": 0.4331,
+      "step": 2917
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.272637055704366,
+      "learning_rate": 9.236656416397254e-06,
+      "loss": 0.4588,
+      "step": 2918
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.991251067654061,
+      "learning_rate": 9.236058320113386e-06,
+      "loss": 0.3932,
+      "step": 2919
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.347692414257632,
+      "learning_rate": 9.235460008990449e-06,
+      "loss": 0.4295,
+      "step": 2920
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.861650619725927,
+      "learning_rate": 9.23486148305879e-06,
+      "loss": 0.4254,
+      "step": 2921
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.127753023556862,
+      "learning_rate": 9.234262742348764e-06,
+      "loss": 0.4088,
+      "step": 2922
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4275224817544927,
+      "learning_rate": 9.233663786890735e-06,
+      "loss": 0.2111,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.0168795403932815,
+      "learning_rate": 9.23306461671508e-06,
+      "loss": 0.4396,
+      "step": 2924
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.909096713192478,
+      "learning_rate": 9.232465231852189e-06,
+      "loss": 0.422,
+      "step": 2925
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.431252035250745,
+      "learning_rate": 9.23186563233246e-06,
+      "loss": 0.4083,
+      "step": 2926
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.25754113749316,
+      "learning_rate": 9.231265818186304e-06,
+      "loss": 0.4145,
+      "step": 2927
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.212246897788609,
+      "learning_rate": 9.23066578944414e-06,
+      "loss": 0.4338,
+      "step": 2928
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.9622408526298925,
+      "learning_rate": 9.2300655461364e-06,
+      "loss": 0.4127,
+      "step": 2929
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.273546931563286,
+      "learning_rate": 9.229465088293525e-06,
+      "loss": 0.4254,
+      "step": 2930
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.759725417014236,
+      "learning_rate": 9.22886441594597e-06,
+      "loss": 0.4496,
+      "step": 2931
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 18.179610233406837,
+      "learning_rate": 9.228263529124199e-06,
+      "loss": 0.4297,
+      "step": 2932
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.997607314943828,
+      "learning_rate": 9.227662427858687e-06,
+      "loss": 0.4149,
+      "step": 2933
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.588398289048089,
+      "learning_rate": 9.22706111217992e-06,
+      "loss": 0.4368,
+      "step": 2934
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 10.110133966927886,
+      "learning_rate": 9.226459582118394e-06,
+      "loss": 0.4012,
+      "step": 2935
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.599209423041298,
+      "learning_rate": 9.225857837704617e-06,
+      "loss": 0.4394,
+      "step": 2936
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.398663526542087,
+      "learning_rate": 9.225255878969108e-06,
+      "loss": 0.4231,
+      "step": 2937
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.013519969963583,
+      "learning_rate": 9.224653705942396e-06,
+      "loss": 0.4524,
+      "step": 2938
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.895361823082405,
+      "learning_rate": 9.224051318655021e-06,
+      "loss": 0.4325,
+      "step": 2939
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.796992935404245,
+      "learning_rate": 9.223448717137533e-06,
+      "loss": 0.3999,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.489356784752569,
+      "learning_rate": 9.222845901420496e-06,
+      "loss": 0.3843,
+      "step": 2941
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.650620977207599,
+      "learning_rate": 9.222242871534484e-06,
+      "loss": 0.4118,
+      "step": 2942
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.13626339910423,
+      "learning_rate": 9.221639627510076e-06,
+      "loss": 0.4244,
+      "step": 2943
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.417238051318458,
+      "learning_rate": 9.22103616937787e-06,
+      "loss": 0.4062,
+      "step": 2944
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.40321220412465586,
+      "learning_rate": 9.220432497168472e-06,
+      "loss": 0.1814,
+      "step": 2945
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 6.317007725630637,
+      "learning_rate": 9.219828610912497e-06,
+      "loss": 0.4091,
+      "step": 2946
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 5.452410511794313,
+      "learning_rate": 9.219224510640572e-06,
+      "loss": 0.4046,
+      "step": 2947
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.8944674818229466,
+      "learning_rate": 9.218620196383337e-06,
+      "loss": 0.4473,
+      "step": 2948
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.657974516784545,
+      "learning_rate": 9.218015668171435e-06,
+      "loss": 0.442,
+      "step": 2949
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.566334999531755,
+      "learning_rate": 9.217410926035534e-06,
+      "loss": 0.3878,
+      "step": 2950
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.532419167925844,
+      "learning_rate": 9.216805970006298e-06,
+      "loss": 0.4225,
+      "step": 2951
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.6895917931878435,
+      "learning_rate": 9.216200800114412e-06,
+      "loss": 0.4268,
+      "step": 2952
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.5414367200718555,
+      "learning_rate": 9.215595416390567e-06,
+      "loss": 0.4246,
+      "step": 2953
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7229226525492076,
+      "learning_rate": 9.214989818865465e-06,
+      "loss": 0.3985,
+      "step": 2954
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.880921696320508,
+      "learning_rate": 9.21438400756982e-06,
+      "loss": 0.4515,
+      "step": 2955
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.697875493786738,
+      "learning_rate": 9.213777982534361e-06,
+      "loss": 0.4646,
+      "step": 2956
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.32009655443897,
+      "learning_rate": 9.21317174378982e-06,
+      "loss": 0.4283,
+      "step": 2957
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.115243046063673,
+      "learning_rate": 9.212565291366942e-06,
+      "loss": 0.4701,
+      "step": 2958
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.689760289529784,
+      "learning_rate": 9.211958625296487e-06,
+      "loss": 0.391,
+      "step": 2959
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.163821098589746,
+      "learning_rate": 9.211351745609221e-06,
+      "loss": 0.3995,
+      "step": 2960
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.43505533838571,
+      "learning_rate": 9.210744652335926e-06,
+      "loss": 0.4434,
+      "step": 2961
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.926488420922358,
+      "learning_rate": 9.21013734550739e-06,
+      "loss": 0.3752,
+      "step": 2962
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.989618581026807,
+      "learning_rate": 9.209529825154413e-06,
+      "loss": 0.4016,
+      "step": 2963
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.495667805102563,
+      "learning_rate": 9.208922091307808e-06,
+      "loss": 0.4426,
+      "step": 2964
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.6015939054536945,
+      "learning_rate": 9.208314143998396e-06,
+      "loss": 0.4379,
+      "step": 2965
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.498624254683919,
+      "learning_rate": 9.20770598325701e-06,
+      "loss": 0.4771,
+      "step": 2966
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.2420159025534465,
+      "learning_rate": 9.207097609114495e-06,
+      "loss": 0.4107,
+      "step": 2967
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4105325443884024,
+      "learning_rate": 9.206489021601704e-06,
+      "loss": 0.1953,
+      "step": 2968
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.35926370585201,
+      "learning_rate": 9.205880220749505e-06,
+      "loss": 0.443,
+      "step": 2969
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.622045759116396,
+      "learning_rate": 9.205271206588773e-06,
+      "loss": 0.4114,
+      "step": 2970
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.625406917333633,
+      "learning_rate": 9.204661979150395e-06,
+      "loss": 0.4464,
+      "step": 2971
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.992881453016702,
+      "learning_rate": 9.20405253846527e-06,
+      "loss": 0.4115,
+      "step": 2972
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.362989233857905,
+      "learning_rate": 9.203442884564305e-06,
+      "loss": 0.4112,
+      "step": 2973
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.489005583416034,
+      "learning_rate": 9.202833017478421e-06,
+      "loss": 0.4085,
+      "step": 2974
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.202894163927917,
+      "learning_rate": 9.20222293723855e-06,
+      "loss": 0.4112,
+      "step": 2975
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.916188160232178,
+      "learning_rate": 9.20161264387563e-06,
+      "loss": 0.4356,
+      "step": 2976
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.155686703946275,
+      "learning_rate": 9.201002137420615e-06,
+      "loss": 0.4159,
+      "step": 2977
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.161800948583706,
+      "learning_rate": 9.20039141790447e-06,
+      "loss": 0.4091,
+      "step": 2978
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.286944569173264,
+      "learning_rate": 9.199780485358164e-06,
+      "loss": 0.4013,
+      "step": 2979
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.646996357731364,
+      "learning_rate": 9.199169339812685e-06,
+      "loss": 0.4114,
+      "step": 2980
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.897958147683878,
+      "learning_rate": 9.198557981299026e-06,
+      "loss": 0.4313,
+      "step": 2981
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.425782584488928,
+      "learning_rate": 9.197946409848196e-06,
+      "loss": 0.4488,
+      "step": 2982
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.605798794734962,
+      "learning_rate": 9.197334625491208e-06,
+      "loss": 0.4664,
+      "step": 2983
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.978979023951187,
+      "learning_rate": 9.196722628259094e-06,
+      "loss": 0.4057,
+      "step": 2984
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4340721343105711,
+      "learning_rate": 9.196110418182889e-06,
+      "loss": 0.1862,
+      "step": 2985
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.42392556612325616,
+      "learning_rate": 9.195497995293643e-06,
+      "loss": 0.1999,
+      "step": 2986
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.045664020936156,
+      "learning_rate": 9.194885359622418e-06,
+      "loss": 0.4361,
+      "step": 2987
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.337580321676738,
+      "learning_rate": 9.194272511200284e-06,
+      "loss": 0.4329,
+      "step": 2988
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.414436870950591,
+      "learning_rate": 9.193659450058321e-06,
+      "loss": 0.4347,
+      "step": 2989
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.617157220921327,
+      "learning_rate": 9.193046176227625e-06,
+      "loss": 0.4375,
+      "step": 2990
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.941238264219871,
+      "learning_rate": 9.192432689739294e-06,
+      "loss": 0.4257,
+      "step": 2991
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.542820832934386,
+      "learning_rate": 9.191818990624447e-06,
+      "loss": 0.4007,
+      "step": 2992
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.543551913569605,
+      "learning_rate": 9.191205078914207e-06,
+      "loss": 0.4376,
+      "step": 2993
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.363894686235422,
+      "learning_rate": 9.190590954639709e-06,
+      "loss": 0.4586,
+      "step": 2994
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 19.329194799799318,
+      "learning_rate": 9.189976617832101e-06,
+      "loss": 0.4076,
+      "step": 2995
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.075740181325786,
+      "learning_rate": 9.189362068522538e-06,
+      "loss": 0.3962,
+      "step": 2996
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.145990487942113,
+      "learning_rate": 9.18874730674219e-06,
+      "loss": 0.4692,
+      "step": 2997
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.421727325167598,
+      "learning_rate": 9.188132332522233e-06,
+      "loss": 0.4337,
+      "step": 2998
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.908463871913921,
+      "learning_rate": 9.187517145893859e-06,
+      "loss": 0.4175,
+      "step": 2999
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.837074569348908,
+      "learning_rate": 9.186901746888267e-06,
+      "loss": 0.443,
+      "step": 3000
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.407112973316162,
+      "learning_rate": 9.186286135536668e-06,
+      "loss": 0.4106,
+      "step": 3001
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.716992008158504,
+      "learning_rate": 9.185670311870286e-06,
+      "loss": 0.4195,
+      "step": 3002
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.397672637563424,
+      "learning_rate": 9.185054275920351e-06,
+      "loss": 0.432,
+      "step": 3003
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.451500517367528,
+      "learning_rate": 9.184438027718108e-06,
+      "loss": 0.4584,
+      "step": 3004
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9357411543919821,
+      "learning_rate": 9.18382156729481e-06,
+      "loss": 0.2395,
+      "step": 3005
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.597882324667976,
+      "learning_rate": 9.183204894681722e-06,
+      "loss": 0.4494,
+      "step": 3006
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.553365663878026,
+      "learning_rate": 9.182588009910119e-06,
+      "loss": 0.4187,
+      "step": 3007
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.063879501434369,
+      "learning_rate": 9.181970913011289e-06,
+      "loss": 0.4001,
+      "step": 3008
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.356221438625427,
+      "learning_rate": 9.181353604016528e-06,
+      "loss": 0.4296,
+      "step": 3009
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.4056962888463245,
+      "learning_rate": 9.180736082957142e-06,
+      "loss": 0.4503,
+      "step": 3010
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.455646762149157,
+      "learning_rate": 9.180118349864455e-06,
+      "loss": 0.4428,
+      "step": 3011
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.658198725970609,
+      "learning_rate": 9.179500404769792e-06,
+      "loss": 0.4024,
+      "step": 3012
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 43.420546858563945,
+      "learning_rate": 9.178882247704494e-06,
+      "loss": 0.4471,
+      "step": 3013
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.232669319376145,
+      "learning_rate": 9.178263878699913e-06,
+      "loss": 0.4061,
+      "step": 3014
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.665554580156446,
+      "learning_rate": 9.177645297787412e-06,
+      "loss": 0.4306,
+      "step": 3015
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 15.741372956043852,
+      "learning_rate": 9.177026504998358e-06,
+      "loss": 0.4324,
+      "step": 3016
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.899361564008753,
+      "learning_rate": 9.176407500364139e-06,
+      "loss": 0.4084,
+      "step": 3017
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.899924831078414,
+      "learning_rate": 9.175788283916149e-06,
+      "loss": 0.4457,
+      "step": 3018
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.949507374295805,
+      "learning_rate": 9.17516885568579e-06,
+      "loss": 0.4086,
+      "step": 3019
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 20.26231037988272,
+      "learning_rate": 9.174549215704476e-06,
+      "loss": 0.4473,
+      "step": 3020
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8873478601193737,
+      "learning_rate": 9.173929364003637e-06,
+      "loss": 0.2516,
+      "step": 3021
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.300670873835242,
+      "learning_rate": 9.17330930061471e-06,
+      "loss": 0.4075,
+      "step": 3022
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 26.40171868008595,
+      "learning_rate": 9.17268902556914e-06,
+      "loss": 0.4146,
+      "step": 3023
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.723225564119829,
+      "learning_rate": 9.172068538898386e-06,
+      "loss": 0.426,
+      "step": 3024
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.679755307811094,
+      "learning_rate": 9.171447840633918e-06,
+      "loss": 0.3886,
+      "step": 3025
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.707972087835451,
+      "learning_rate": 9.170826930807214e-06,
+      "loss": 0.4315,
+      "step": 3026
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.995483689890627,
+      "learning_rate": 9.170205809449768e-06,
+      "loss": 0.469,
+      "step": 3027
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.802801635898172,
+      "learning_rate": 9.169584476593078e-06,
+      "loss": 0.4179,
+      "step": 3028
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.09839711340566,
+      "learning_rate": 9.168962932268657e-06,
+      "loss": 0.4491,
+      "step": 3029
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 25.674136825911933,
+      "learning_rate": 9.168341176508028e-06,
+      "loss": 0.4101,
+      "step": 3030
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.850977161598976,
+      "learning_rate": 9.167719209342723e-06,
+      "loss": 0.4498,
+      "step": 3031
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 11.552746750713377,
+      "learning_rate": 9.167097030804289e-06,
+      "loss": 0.4271,
+      "step": 3032
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.213108593473931,
+      "learning_rate": 9.166474640924276e-06,
+      "loss": 0.4276,
+      "step": 3033
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.627652028597445,
+      "learning_rate": 9.165852039734255e-06,
+      "loss": 0.4703,
+      "step": 3034
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.754113641259598,
+      "learning_rate": 9.1652292272658e-06,
+      "loss": 0.4066,
+      "step": 3035
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.1927206937300925,
+      "learning_rate": 9.164606203550498e-06,
+      "loss": 0.3846,
+      "step": 3036
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.787895228033868,
+      "learning_rate": 9.163982968619946e-06,
+      "loss": 0.4213,
+      "step": 3037
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.785327772951596,
+      "learning_rate": 9.163359522505755e-06,
+      "loss": 0.4326,
+      "step": 3038
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.947546794545524,
+      "learning_rate": 9.16273586523954e-06,
+      "loss": 0.3941,
+      "step": 3039
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.367378177435612,
+      "learning_rate": 9.162111996852933e-06,
+      "loss": 0.4223,
+      "step": 3040
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5632110851985567,
+      "learning_rate": 9.161487917377578e-06,
+      "loss": 0.2127,
+      "step": 3041
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.29404216671367,
+      "learning_rate": 9.16086362684512e-06,
+      "loss": 0.4129,
+      "step": 3042
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4181082412120483,
+      "learning_rate": 9.160239125287224e-06,
+      "loss": 0.2002,
+      "step": 3043
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.866419848901377,
+      "learning_rate": 9.159614412735566e-06,
+      "loss": 0.3862,
+      "step": 3044
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.539358019799124,
+      "learning_rate": 9.158989489221824e-06,
+      "loss": 0.4244,
+      "step": 3045
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.589785067093592,
+      "learning_rate": 9.158364354777693e-06,
+      "loss": 0.4073,
+      "step": 3046
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.008620877564666,
+      "learning_rate": 9.15773900943488e-06,
+      "loss": 0.4712,
+      "step": 3047
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.935982670438646,
+      "learning_rate": 9.157113453225099e-06,
+      "loss": 0.4194,
+      "step": 3048
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.66199228199182,
+      "learning_rate": 9.156487686180078e-06,
+      "loss": 0.4146,
+      "step": 3049
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.559168128686713,
+      "learning_rate": 9.155861708331553e-06,
+      "loss": 0.4663,
+      "step": 3050
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.398312315638955,
+      "learning_rate": 9.155235519711269e-06,
+      "loss": 0.4251,
+      "step": 3051
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 12.877418716208984,
+      "learning_rate": 9.154609120350986e-06,
+      "loss": 0.4248,
+      "step": 3052
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 29.45370451757948,
+      "learning_rate": 9.153982510282475e-06,
+      "loss": 0.4207,
+      "step": 3053
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.715614576910248,
+      "learning_rate": 9.153355689537513e-06,
+      "loss": 0.4409,
+      "step": 3054
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.942294897783627,
+      "learning_rate": 9.152728658147891e-06,
+      "loss": 0.4109,
+      "step": 3055
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.9125248210881294,
+      "learning_rate": 9.152101416145411e-06,
+      "loss": 0.4068,
+      "step": 3056
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.485196627883217,
+      "learning_rate": 9.151473963561884e-06,
+      "loss": 0.443,
+      "step": 3057
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.016216452800905,
+      "learning_rate": 9.15084630042913e-06,
+      "loss": 0.4269,
+      "step": 3058
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.577056640769282,
+      "learning_rate": 9.150218426778985e-06,
+      "loss": 0.4077,
+      "step": 3059
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.806748461495675,
+      "learning_rate": 9.149590342643293e-06,
+      "loss": 0.4408,
+      "step": 3060
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.2295253186241775,
+      "learning_rate": 9.148962048053906e-06,
+      "loss": 0.3998,
+      "step": 3061
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.7328694189392952,
+      "learning_rate": 9.14833354304269e-06,
+      "loss": 0.4538,
+      "step": 3062
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.478015219843606,
+      "learning_rate": 9.147704827641524e-06,
+      "loss": 0.3997,
+      "step": 3063
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.166247854057637,
+      "learning_rate": 9.147075901882289e-06,
+      "loss": 0.3969,
+      "step": 3064
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.136988636122654,
+      "learning_rate": 9.146446765796884e-06,
+      "loss": 0.3575,
+      "step": 3065
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.828752260580864,
+      "learning_rate": 9.145817419417218e-06,
+      "loss": 0.4779,
+      "step": 3066
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9360396445066486,
+      "learning_rate": 9.145187862775208e-06,
+      "loss": 0.2592,
+      "step": 3067
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.466192708027079,
+      "learning_rate": 9.144558095902785e-06,
+      "loss": 0.4232,
+      "step": 3068
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.49916081167405374,
+      "learning_rate": 9.143928118831888e-06,
+      "loss": 0.2244,
+      "step": 3069
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.525259294993424,
+      "learning_rate": 9.143297931594465e-06,
+      "loss": 0.413,
+      "step": 3070
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 116.50983706244021,
+      "learning_rate": 9.14266753422248e-06,
+      "loss": 0.4335,
+      "step": 3071
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.15261662079567,
+      "learning_rate": 9.142036926747904e-06,
+      "loss": 0.3755,
+      "step": 3072
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.485981396166076,
+      "learning_rate": 9.141406109202718e-06,
+      "loss": 0.4719,
+      "step": 3073
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.538319677445016,
+      "learning_rate": 9.140775081618915e-06,
+      "loss": 0.4612,
+      "step": 3074
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.829820642860211,
+      "learning_rate": 9.1401438440285e-06,
+      "loss": 0.4047,
+      "step": 3075
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 56.17132413328362,
+      "learning_rate": 9.139512396463487e-06,
+      "loss": 0.4406,
+      "step": 3076
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.098571118344977,
+      "learning_rate": 9.138880738955902e-06,
+      "loss": 0.4433,
+      "step": 3077
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.1732082908042365,
+      "learning_rate": 9.138248871537778e-06,
+      "loss": 0.4556,
+      "step": 3078
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 13.167543730556622,
+      "learning_rate": 9.137616794241162e-06,
+      "loss": 0.4112,
+      "step": 3079
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.417606351465401,
+      "learning_rate": 9.136984507098114e-06,
+      "loss": 0.4316,
+      "step": 3080
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 9.668624561296612,
+      "learning_rate": 9.136352010140697e-06,
+      "loss": 0.4204,
+      "step": 3081
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 6.316347063614023,
+      "learning_rate": 9.135719303400995e-06,
+      "loss": 0.4082,
+      "step": 3082
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 8.10279975877751,
+      "learning_rate": 9.13508638691109e-06,
+      "loss": 0.3999,
+      "step": 3083
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.558743166505888,
+      "learning_rate": 9.134453260703084e-06,
+      "loss": 0.4095,
+      "step": 3084
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.183528459721309,
+      "learning_rate": 9.133819924809089e-06,
+      "loss": 0.4642,
+      "step": 3085
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.121841709742834,
+      "learning_rate": 9.133186379261224e-06,
+      "loss": 0.4181,
+      "step": 3086
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 10.64839190646957,
+      "learning_rate": 9.13255262409162e-06,
+      "loss": 0.3939,
+      "step": 3087
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.820552837526884,
+      "learning_rate": 9.131918659332422e-06,
+      "loss": 0.3969,
+      "step": 3088
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.1996341281403953,
+      "learning_rate": 9.131284485015778e-06,
+      "loss": 0.3084,
+      "step": 3089
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 4.615184163575239,
+      "learning_rate": 9.130650101173856e-06,
+      "loss": 0.4157,
+      "step": 3090
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 5.313838534675096,
+      "learning_rate": 9.130015507838827e-06,
+      "loss": 0.4447,
+      "step": 3091
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.904251374201751,
+      "learning_rate": 9.129380705042874e-06,
+      "loss": 0.4352,
+      "step": 3092
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.794199721612345,
+      "learning_rate": 9.128745692818198e-06,
+      "loss": 0.4623,
+      "step": 3093
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.25160484551708,
+      "learning_rate": 9.128110471196997e-06,
+      "loss": 0.3688,
+      "step": 3094
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.78052527047746,
+      "learning_rate": 9.127475040211493e-06,
+      "loss": 0.4135,
+      "step": 3095
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.86821662265676,
+      "learning_rate": 9.126839399893913e-06,
+      "loss": 0.4081,
+      "step": 3096
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.120945901607623,
+      "learning_rate": 9.12620355027649e-06,
+      "loss": 0.4046,
+      "step": 3097
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 39.071890316980294,
+      "learning_rate": 9.125567491391476e-06,
+      "loss": 0.4355,
+      "step": 3098
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.942752646322269,
+      "learning_rate": 9.12493122327113e-06,
+      "loss": 0.4033,
+      "step": 3099
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.887648260766308,
+      "learning_rate": 9.124294745947719e-06,
+      "loss": 0.4444,
+      "step": 3100
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.466802857257661,
+      "learning_rate": 9.123658059453526e-06,
+      "loss": 0.4369,
+      "step": 3101
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.420994226493532,
+      "learning_rate": 9.123021163820839e-06,
+      "loss": 0.4346,
+      "step": 3102
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.059391370270722,
+      "learning_rate": 9.122384059081961e-06,
+      "loss": 0.3974,
+      "step": 3103
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.571591456327003,
+      "learning_rate": 9.121746745269201e-06,
+      "loss": 0.4633,
+      "step": 3104
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.093224713535955,
+      "learning_rate": 9.121109222414888e-06,
+      "loss": 0.4256,
+      "step": 3105
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5246081881601121,
+      "learning_rate": 9.120471490551348e-06,
+      "loss": 0.2233,
+      "step": 3106
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.203486803286027,
+      "learning_rate": 9.119833549710927e-06,
+      "loss": 0.4648,
+      "step": 3107
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.467238557591629,
+      "learning_rate": 9.119195399925981e-06,
+      "loss": 0.4138,
+      "step": 3108
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.576383599972866,
+      "learning_rate": 9.118557041228875e-06,
+      "loss": 0.3949,
+      "step": 3109
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.032648905809606,
+      "learning_rate": 9.117918473651982e-06,
+      "loss": 0.4267,
+      "step": 3110
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.183358195195572,
+      "learning_rate": 9.117279697227687e-06,
+      "loss": 0.4653,
+      "step": 3111
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.896602349992456,
+      "learning_rate": 9.116640711988392e-06,
+      "loss": 0.4146,
+      "step": 3112
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.991336619061355,
+      "learning_rate": 9.1160015179665e-06,
+      "loss": 0.425,
+      "step": 3113
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.381147783125669,
+      "learning_rate": 9.11536211519443e-06,
+      "loss": 0.211,
+      "step": 3114
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.942795484388842,
+      "learning_rate": 9.11472250370461e-06,
+      "loss": 0.4257,
+      "step": 3115
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.801600573691575,
+      "learning_rate": 9.11408268352948e-06,
+      "loss": 0.4225,
+      "step": 3116
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.976686820131062,
+      "learning_rate": 9.113442654701487e-06,
+      "loss": 0.4311,
+      "step": 3117
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.7167242405242655,
+      "learning_rate": 9.112802417253094e-06,
+      "loss": 0.4122,
+      "step": 3118
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.39262170450212,
+      "learning_rate": 9.112161971216774e-06,
+      "loss": 0.469,
+      "step": 3119
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.904428672680078,
+      "learning_rate": 9.111521316625002e-06,
+      "loss": 0.4865,
+      "step": 3120
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.735743968874984,
+      "learning_rate": 9.110880453510274e-06,
+      "loss": 0.4294,
+      "step": 3121
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.346845316171707,
+      "learning_rate": 9.11023938190509e-06,
+      "loss": 0.4438,
+      "step": 3122
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.52412592115245,
+      "learning_rate": 9.109598101841967e-06,
+      "loss": 0.4168,
+      "step": 3123
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.391643487800187,
+      "learning_rate": 9.108956613353425e-06,
+      "loss": 0.4324,
+      "step": 3124
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.916499807874167,
+      "learning_rate": 9.108314916472e-06,
+      "loss": 0.4048,
+      "step": 3125
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.239865383583732,
+      "learning_rate": 9.107673011230238e-06,
+      "loss": 0.3831,
+      "step": 3126
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.097608942067433,
+      "learning_rate": 9.10703089766069e-06,
+      "loss": 0.4541,
+      "step": 3127
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.460675151440537,
+      "learning_rate": 9.106388575795925e-06,
+      "loss": 0.4107,
+      "step": 3128
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.463307261454544,
+      "learning_rate": 9.10574604566852e-06,
+      "loss": 0.4306,
+      "step": 3129
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.761619758424164,
+      "learning_rate": 9.105103307311062e-06,
+      "loss": 0.4688,
+      "step": 3130
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.332269136360888,
+      "learning_rate": 9.104460360756146e-06,
+      "loss": 0.4157,
+      "step": 3131
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40919683629029396,
+      "learning_rate": 9.103817206036383e-06,
+      "loss": 0.2097,
+      "step": 3132
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 26.391557661897824,
+      "learning_rate": 9.10317384318439e-06,
+      "loss": 0.415,
+      "step": 3133
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.468469665422968,
+      "learning_rate": 9.102530272232796e-06,
+      "loss": 0.4261,
+      "step": 3134
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.4637507240732175,
+      "learning_rate": 9.101886493214243e-06,
+      "loss": 0.4198,
+      "step": 3135
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.283559861470675,
+      "learning_rate": 9.101242506161378e-06,
+      "loss": 0.4204,
+      "step": 3136
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.312882332096308,
+      "learning_rate": 9.100598311106865e-06,
+      "loss": 0.4045,
+      "step": 3137
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.382990596475792,
+      "learning_rate": 9.099953908083375e-06,
+      "loss": 0.4774,
+      "step": 3138
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.98049591418215,
+      "learning_rate": 9.099309297123588e-06,
+      "loss": 0.4251,
+      "step": 3139
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.0552321793989,
+      "learning_rate": 9.098664478260199e-06,
+      "loss": 0.4132,
+      "step": 3140
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.826595622311663,
+      "learning_rate": 9.09801945152591e-06,
+      "loss": 0.4134,
+      "step": 3141
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.288269688263363,
+      "learning_rate": 9.097374216953435e-06,
+      "loss": 0.4115,
+      "step": 3142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.38051608832237643,
+      "learning_rate": 9.096728774575498e-06,
+      "loss": 0.213,
+      "step": 3143
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.803305379834374,
+      "learning_rate": 9.096083124424833e-06,
+      "loss": 0.4139,
+      "step": 3144
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.364077009267671,
+      "learning_rate": 9.095437266534186e-06,
+      "loss": 0.437,
+      "step": 3145
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.844529469692306,
+      "learning_rate": 9.094791200936312e-06,
+      "loss": 0.4298,
+      "step": 3146
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.165270827817134,
+      "learning_rate": 9.094144927663979e-06,
+      "loss": 0.4405,
+      "step": 3147
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.608213187872671,
+      "learning_rate": 9.093498446749964e-06,
+      "loss": 0.4115,
+      "step": 3148
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 18.137525782626305,
+      "learning_rate": 9.092851758227052e-06,
+      "loss": 0.4571,
+      "step": 3149
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.6131778565962565,
+      "learning_rate": 9.092204862128042e-06,
+      "loss": 0.4476,
+      "step": 3150
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.8247805839641815,
+      "learning_rate": 9.091557758485743e-06,
+      "loss": 0.402,
+      "step": 3151
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.9145433432358776,
+      "learning_rate": 9.090910447332975e-06,
+      "loss": 0.4108,
+      "step": 3152
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.361694424885167,
+      "learning_rate": 9.090262928702564e-06,
+      "loss": 0.4068,
+      "step": 3153
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.40690644349975,
+      "learning_rate": 9.089615202627353e-06,
+      "loss": 0.4247,
+      "step": 3154
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.4019069525512,
+      "learning_rate": 9.088967269140193e-06,
+      "loss": 0.4322,
+      "step": 3155
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.302612827841607,
+      "learning_rate": 9.088319128273944e-06,
+      "loss": 0.4286,
+      "step": 3156
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.366595714295528,
+      "learning_rate": 9.087670780061477e-06,
+      "loss": 0.4349,
+      "step": 3157
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.233451457509225,
+      "learning_rate": 9.087022224535673e-06,
+      "loss": 0.4503,
+      "step": 3158
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4061449044002199,
+      "learning_rate": 9.086373461729428e-06,
+      "loss": 0.2077,
+      "step": 3159
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.157308332664487,
+      "learning_rate": 9.085724491675642e-06,
+      "loss": 0.4327,
+      "step": 3160
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.752113146847724,
+      "learning_rate": 9.085075314407233e-06,
+      "loss": 0.419,
+      "step": 3161
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.818455725558156,
+      "learning_rate": 9.08442592995712e-06,
+      "loss": 0.4089,
+      "step": 3162
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.882177543350974,
+      "learning_rate": 9.08377633835824e-06,
+      "loss": 0.4465,
+      "step": 3163
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 15.595076343643342,
+      "learning_rate": 9.083126539643538e-06,
+      "loss": 0.4328,
+      "step": 3164
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.072258586753537,
+      "learning_rate": 9.082476533845968e-06,
+      "loss": 0.4372,
+      "step": 3165
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3702424122063232,
+      "learning_rate": 9.081826320998502e-06,
+      "loss": 0.2077,
+      "step": 3166
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.761456293828241,
+      "learning_rate": 9.081175901134107e-06,
+      "loss": 0.4147,
+      "step": 3167
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.495777371693405,
+      "learning_rate": 9.080525274285781e-06,
+      "loss": 0.3847,
+      "step": 3168
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.823630042633013,
+      "learning_rate": 9.079874440486514e-06,
+      "loss": 0.4004,
+      "step": 3169
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.759444495855737,
+      "learning_rate": 9.079223399769316e-06,
+      "loss": 0.4049,
+      "step": 3170
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.400385449238657,
+      "learning_rate": 9.078572152167207e-06,
+      "loss": 0.4401,
+      "step": 3171
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.633287007612118,
+      "learning_rate": 9.077920697713215e-06,
+      "loss": 0.4311,
+      "step": 3172
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.847244897782892,
+      "learning_rate": 9.07726903644038e-06,
+      "loss": 0.4279,
+      "step": 3173
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 16.899313637716098,
+      "learning_rate": 9.076617168381751e-06,
+      "loss": 0.4075,
+      "step": 3174
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.595449087447479,
+      "learning_rate": 9.075965093570391e-06,
+      "loss": 0.386,
+      "step": 3175
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 25.776262580455573,
+      "learning_rate": 9.07531281203937e-06,
+      "loss": 0.4276,
+      "step": 3176
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.540730848679918,
+      "learning_rate": 9.074660323821772e-06,
+      "loss": 0.3942,
+      "step": 3177
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.009979712370698,
+      "learning_rate": 9.074007628950684e-06,
+      "loss": 0.4281,
+      "step": 3178
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.55096844645871,
+      "learning_rate": 9.07335472745921e-06,
+      "loss": 0.4135,
+      "step": 3179
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.44816170635179,
+      "learning_rate": 9.072701619380464e-06,
+      "loss": 0.4366,
+      "step": 3180
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.691837079916965,
+      "learning_rate": 9.072048304747572e-06,
+      "loss": 0.3775,
+      "step": 3181
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 21.987774138644912,
+      "learning_rate": 9.071394783593664e-06,
+      "loss": 0.3978,
+      "step": 3182
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.382051570168105,
+      "learning_rate": 9.070741055951886e-06,
+      "loss": 0.3853,
+      "step": 3183
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.047414688259584,
+      "learning_rate": 9.070087121855394e-06,
+      "loss": 0.3886,
+      "step": 3184
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.6806953082262215,
+      "learning_rate": 9.069432981337352e-06,
+      "loss": 0.3862,
+      "step": 3185
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.7902044318099835,
+      "learning_rate": 9.068778634430935e-06,
+      "loss": 0.4288,
+      "step": 3186
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.147333788231215,
+      "learning_rate": 9.068124081169332e-06,
+      "loss": 0.4123,
+      "step": 3187
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.752902158450025,
+      "learning_rate": 9.067469321585738e-06,
+      "loss": 0.4445,
+      "step": 3188
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.204774104771996,
+      "learning_rate": 9.06681435571336e-06,
+      "loss": 0.4205,
+      "step": 3189
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.538022189452344,
+      "learning_rate": 9.066159183585419e-06,
+      "loss": 0.4261,
+      "step": 3190
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.780846675293603,
+      "learning_rate": 9.065503805235139e-06,
+      "loss": 0.4194,
+      "step": 3191
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.310796476955813,
+      "learning_rate": 9.06484822069576e-06,
+      "loss": 0.4335,
+      "step": 3192
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.709700977819107,
+      "learning_rate": 9.064192430000533e-06,
+      "loss": 0.4221,
+      "step": 3193
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.1484406729971175,
+      "learning_rate": 9.063536433182714e-06,
+      "loss": 0.4085,
+      "step": 3194
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.307054562290603,
+      "learning_rate": 9.062880230275576e-06,
+      "loss": 0.3985,
+      "step": 3195
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.587711206181401,
+      "learning_rate": 9.062223821312398e-06,
+      "loss": 0.3842,
+      "step": 3196
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.3631926365227929,
+      "learning_rate": 9.061567206326472e-06,
+      "loss": 0.1984,
+      "step": 3197
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.823933214921759,
+      "learning_rate": 9.060910385351098e-06,
+      "loss": 0.3826,
+      "step": 3198
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.90029124107031,
+      "learning_rate": 9.060253358419588e-06,
+      "loss": 0.4293,
+      "step": 3199
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 17.12238795857824,
+      "learning_rate": 9.059596125565266e-06,
+      "loss": 0.4318,
+      "step": 3200
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.151559987985154,
+      "learning_rate": 9.058938686821464e-06,
+      "loss": 0.4419,
+      "step": 3201
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.869462327904575,
+      "learning_rate": 9.058281042221523e-06,
+      "loss": 0.4042,
+      "step": 3202
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.406624905088302,
+      "learning_rate": 9.0576231917988e-06,
+      "loss": 0.3903,
+      "step": 3203
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 9.670269420534137,
+      "learning_rate": 9.056965135586657e-06,
+      "loss": 0.4215,
+      "step": 3204
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.008424807424479,
+      "learning_rate": 9.056306873618469e-06,
+      "loss": 0.4288,
+      "step": 3205
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.989670167110441,
+      "learning_rate": 9.05564840592762e-06,
+      "loss": 0.4612,
+      "step": 3206
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.0082572103751275,
+      "learning_rate": 9.054989732547507e-06,
+      "loss": 0.3777,
+      "step": 3207
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.539355853822536,
+      "learning_rate": 9.054330853511534e-06,
+      "loss": 0.416,
+      "step": 3208
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.093567404911165,
+      "learning_rate": 9.05367176885312e-06,
+      "loss": 0.3759,
+      "step": 3209
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.865114997743296,
+      "learning_rate": 9.053012478605685e-06,
+      "loss": 0.4444,
+      "step": 3210
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.796333250929409,
+      "learning_rate": 9.052352982802675e-06,
+      "loss": 0.3972,
+      "step": 3211
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.410314561120517,
+      "learning_rate": 9.05169328147753e-06,
+      "loss": 0.4467,
+      "step": 3212
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 11.592634092230314,
+      "learning_rate": 9.051033374663713e-06,
+      "loss": 0.4159,
+      "step": 3213
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.628556689782056,
+      "learning_rate": 9.05037326239469e-06,
+      "loss": 0.4121,
+      "step": 3214
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.804168027777236,
+      "learning_rate": 9.049712944703939e-06,
+      "loss": 0.3873,
+      "step": 3215
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.843738736207932,
+      "learning_rate": 9.04905242162495e-06,
+      "loss": 0.3906,
+      "step": 3216
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.507901063536563,
+      "learning_rate": 9.048391693191225e-06,
+      "loss": 0.3885,
+      "step": 3217
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.977210397561019,
+      "learning_rate": 9.047730759436271e-06,
+      "loss": 0.4239,
+      "step": 3218
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 8.300312155402837,
+      "learning_rate": 9.047069620393609e-06,
+      "loss": 0.4393,
+      "step": 3219
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.595911717235131,
+      "learning_rate": 9.046408276096771e-06,
+      "loss": 0.4323,
+      "step": 3220
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 35.348558428038835,
+      "learning_rate": 9.045746726579297e-06,
+      "loss": 0.4367,
+      "step": 3221
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.644116586092254,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.3714,
+      "step": 3222
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.562093483973793,
+      "learning_rate": 9.044423012016657e-06,
+      "loss": 0.4425,
+      "step": 3223
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.810017683395818,
+      "learning_rate": 9.043760847038628e-06,
+      "loss": 0.4522,
+      "step": 3224
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.532946485362394,
+      "learning_rate": 9.04309847697423e-06,
+      "loss": 0.4227,
+      "step": 3225
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.792847553111623,
+      "learning_rate": 9.042435901857059e-06,
+      "loss": 0.4104,
+      "step": 3226
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.616687163441182,
+      "learning_rate": 9.04177312172072e-06,
+      "loss": 0.3795,
+      "step": 3227
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 7.692082360516868,
+      "learning_rate": 9.041110136598823e-06,
+      "loss": 0.4355,
+      "step": 3228
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4242265047557697,
+      "learning_rate": 9.040446946524995e-06,
+      "loss": 0.2221,
+      "step": 3229
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 61.693863147848425,
+      "learning_rate": 9.039783551532872e-06,
+      "loss": 0.4231,
+      "step": 3230
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.915596042359083,
+      "learning_rate": 9.039119951656097e-06,
+      "loss": 0.4213,
+      "step": 3231
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 10.19179625735259,
+      "learning_rate": 9.038456146928325e-06,
+      "loss": 0.4145,
+      "step": 3232
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.43854939552787,
+      "learning_rate": 9.037792137383226e-06,
+      "loss": 0.3821,
+      "step": 3233
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.159803375622804,
+      "learning_rate": 9.037127923054472e-06,
+      "loss": 0.423,
+      "step": 3234
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.615885899862362,
+      "learning_rate": 9.036463503975752e-06,
+      "loss": 0.4073,
+      "step": 3235
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.637426138312616,
+      "learning_rate": 9.035798880180762e-06,
+      "loss": 0.4705,
+      "step": 3236
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.464403219115743,
+      "learning_rate": 9.03513405170321e-06,
+      "loss": 0.3839,
+      "step": 3237
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.363669099958051,
+      "learning_rate": 9.034469018576815e-06,
+      "loss": 0.4226,
+      "step": 3238
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 28.077556484191906,
+      "learning_rate": 9.033803780835304e-06,
+      "loss": 0.4356,
+      "step": 3239
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.651315449250973,
+      "learning_rate": 9.033138338512417e-06,
+      "loss": 0.4281,
+      "step": 3240
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.759540566862864,
+      "learning_rate": 9.032472691641902e-06,
+      "loss": 0.3987,
+      "step": 3241
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.9225641323640374,
+      "learning_rate": 9.031806840257519e-06,
+      "loss": 0.4228,
+      "step": 3242
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.130763923506015,
+      "learning_rate": 9.031140784393036e-06,
+      "loss": 0.4179,
+      "step": 3243
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.7189803328328175,
+      "learning_rate": 9.030474524082237e-06,
+      "loss": 0.4384,
+      "step": 3244
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.43205996997259505,
+      "learning_rate": 9.029808059358908e-06,
+      "loss": 0.2043,
+      "step": 3245
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.191472884562605,
+      "learning_rate": 9.029141390256854e-06,
+      "loss": 0.4271,
+      "step": 3246
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.494141899796256,
+      "learning_rate": 9.028474516809882e-06,
+      "loss": 0.4029,
+      "step": 3247
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.343962208855071,
+      "learning_rate": 9.02780743905182e-06,
+      "loss": 0.447,
+      "step": 3248
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.24520190422399,
+      "learning_rate": 9.027140157016491e-06,
+      "loss": 0.4163,
+      "step": 3249
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.976062529725095,
+      "learning_rate": 9.026472670737746e-06,
+      "loss": 0.4212,
+      "step": 3250
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.34041107813883453,
+      "learning_rate": 9.025804980249434e-06,
+      "loss": 0.1823,
+      "step": 3251
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.855786253925187,
+      "learning_rate": 9.025137085585417e-06,
+      "loss": 0.3853,
+      "step": 3252
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.232265806830231,
+      "learning_rate": 9.02446898677957e-06,
+      "loss": 0.4605,
+      "step": 3253
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.697526669380138,
+      "learning_rate": 9.023800683865777e-06,
+      "loss": 0.4395,
+      "step": 3254
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.078971021263662,
+      "learning_rate": 9.023132176877933e-06,
+      "loss": 0.4409,
+      "step": 3255
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.981924476589436,
+      "learning_rate": 9.02246346584994e-06,
+      "loss": 0.4283,
+      "step": 3256
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.699029313381935,
+      "learning_rate": 9.021794550815713e-06,
+      "loss": 0.427,
+      "step": 3257
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.425422470603515,
+      "learning_rate": 9.02112543180918e-06,
+      "loss": 0.4454,
+      "step": 3258
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 35.03725314380584,
+      "learning_rate": 9.020456108864273e-06,
+      "loss": 0.4133,
+      "step": 3259
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 20.396256478865677,
+      "learning_rate": 9.01978658201494e-06,
+      "loss": 0.4053,
+      "step": 3260
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.027424008471493,
+      "learning_rate": 9.019116851295138e-06,
+      "loss": 0.4537,
+      "step": 3261
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.612322969559645,
+      "learning_rate": 9.018446916738832e-06,
+      "loss": 0.4007,
+      "step": 3262
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 31.642994100155686,
+      "learning_rate": 9.017776778379998e-06,
+      "loss": 0.4159,
+      "step": 3263
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.088595621346779,
+      "learning_rate": 9.017106436252626e-06,
+      "loss": 0.434,
+      "step": 3264
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.005480048544744,
+      "learning_rate": 9.01643589039071e-06,
+      "loss": 0.4099,
+      "step": 3265
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.24164537943111,
+      "learning_rate": 9.01576514082826e-06,
+      "loss": 0.3897,
+      "step": 3266
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.932322916511554,
+      "learning_rate": 9.015094187599297e-06,
+      "loss": 0.4044,
+      "step": 3267
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.3165976861931385,
+      "learning_rate": 9.014423030737846e-06,
+      "loss": 0.4356,
+      "step": 3268
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.871352039875777,
+      "learning_rate": 9.013751670277946e-06,
+      "loss": 0.4098,
+      "step": 3269
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.910724622669947,
+      "learning_rate": 9.013080106253646e-06,
+      "loss": 0.4316,
+      "step": 3270
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.547819664465996,
+      "learning_rate": 9.012408338699006e-06,
+      "loss": 0.3907,
+      "step": 3271
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.182152410537883,
+      "learning_rate": 9.011736367648099e-06,
+      "loss": 0.3742,
+      "step": 3272
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.280378842574327,
+      "learning_rate": 9.011064193135e-06,
+      "loss": 0.4197,
+      "step": 3273
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.927859026065453,
+      "learning_rate": 9.010391815193803e-06,
+      "loss": 0.4203,
+      "step": 3274
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.448197593859326,
+      "learning_rate": 9.009719233858609e-06,
+      "loss": 0.4248,
+      "step": 3275
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.723047099258252,
+      "learning_rate": 9.009046449163526e-06,
+      "loss": 0.4048,
+      "step": 3276
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.800651593040428,
+      "learning_rate": 9.008373461142676e-06,
+      "loss": 0.4036,
+      "step": 3277
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.051948722768515,
+      "learning_rate": 9.007700269830196e-06,
+      "loss": 0.3789,
+      "step": 3278
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.57080450611769,
+      "learning_rate": 9.00702687526022e-06,
+      "loss": 0.4436,
+      "step": 3279
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 24.326167091400873,
+      "learning_rate": 9.006353277466908e-06,
+      "loss": 0.3642,
+      "step": 3280
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4346322913096551,
+      "learning_rate": 9.005679476484417e-06,
+      "loss": 0.1934,
+      "step": 3281
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.799089207246162,
+      "learning_rate": 9.005005472346923e-06,
+      "loss": 0.4084,
+      "step": 3282
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.720627521622296,
+      "learning_rate": 9.00433126508861e-06,
+      "loss": 0.4863,
+      "step": 3283
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.192769257674473,
+      "learning_rate": 9.003656854743667e-06,
+      "loss": 0.4136,
+      "step": 3284
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 13.087056591430041,
+      "learning_rate": 9.002982241346303e-06,
+      "loss": 0.3817,
+      "step": 3285
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.512142575516166,
+      "learning_rate": 9.00230742493073e-06,
+      "loss": 0.3711,
+      "step": 3286
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.33367892144978,
+      "learning_rate": 9.001632405531174e-06,
+      "loss": 0.4015,
+      "step": 3287
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.012459202704518,
+      "learning_rate": 9.000957183181866e-06,
+      "loss": 0.3876,
+      "step": 3288
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.429639214622577,
+      "learning_rate": 9.000281757917055e-06,
+      "loss": 0.4249,
+      "step": 3289
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.716845738173628,
+      "learning_rate": 8.999606129770995e-06,
+      "loss": 0.4284,
+      "step": 3290
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.915226440043912,
+      "learning_rate": 8.99893029877795e-06,
+      "loss": 0.4379,
+      "step": 3291
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.595464671847311,
+      "learning_rate": 8.9982542649722e-06,
+      "loss": 0.396,
+      "step": 3292
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.955922136522983,
+      "learning_rate": 8.997578028388026e-06,
+      "loss": 0.4203,
+      "step": 3293
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.402283827772164,
+      "learning_rate": 8.996901589059728e-06,
+      "loss": 0.4359,
+      "step": 3294
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.002345084608343,
+      "learning_rate": 8.996224947021613e-06,
+      "loss": 0.4002,
+      "step": 3295
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.347271286320872,
+      "learning_rate": 8.995548102307997e-06,
+      "loss": 0.3831,
+      "step": 3296
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.350716831860844,
+      "learning_rate": 8.994871054953207e-06,
+      "loss": 0.4641,
+      "step": 3297
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.28084405979713,
+      "learning_rate": 8.994193804991582e-06,
+      "loss": 0.4094,
+      "step": 3298
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.3681341161777,
+      "learning_rate": 8.993516352457467e-06,
+      "loss": 0.4507,
+      "step": 3299
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.882593704077331,
+      "learning_rate": 8.992838697385224e-06,
+      "loss": 0.4122,
+      "step": 3300
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.46988222272684665,
+      "learning_rate": 8.99216083980922e-06,
+      "loss": 0.2049,
+      "step": 3301
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 25.30833143948996,
+      "learning_rate": 8.991482779763833e-06,
+      "loss": 0.397,
+      "step": 3302
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.440590525731192,
+      "learning_rate": 8.99080451728345e-06,
+      "loss": 0.3882,
+      "step": 3303
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.448036018741277,
+      "learning_rate": 8.990126052402478e-06,
+      "loss": 0.4435,
+      "step": 3304
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.688692912285256,
+      "learning_rate": 8.989447385155316e-06,
+      "loss": 0.4377,
+      "step": 3305
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.523297025696438,
+      "learning_rate": 8.988768515576392e-06,
+      "loss": 0.4072,
+      "step": 3306
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 12.193300300627412,
+      "learning_rate": 8.988089443700131e-06,
+      "loss": 0.4154,
+      "step": 3307
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.40405735403329085,
+      "learning_rate": 8.987410169560977e-06,
+      "loss": 0.2085,
+      "step": 3308
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.697105727914362,
+      "learning_rate": 8.986730693193377e-06,
+      "loss": 0.4048,
+      "step": 3309
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.48478332457493,
+      "learning_rate": 8.986051014631796e-06,
+      "loss": 0.4209,
+      "step": 3310
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.216196106717879,
+      "learning_rate": 8.9853711339107e-06,
+      "loss": 0.445,
+      "step": 3311
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.37866990202785,
+      "learning_rate": 8.984691051064576e-06,
+      "loss": 0.4101,
+      "step": 3312
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 10.38757989384943,
+      "learning_rate": 8.98401076612791e-06,
+      "loss": 0.4148,
+      "step": 3313
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 23.57642385544971,
+      "learning_rate": 8.98333027913521e-06,
+      "loss": 0.4452,
+      "step": 3314
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.845025679975771,
+      "learning_rate": 8.982649590120982e-06,
+      "loss": 0.4493,
+      "step": 3315
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.568084188136874,
+      "learning_rate": 8.981968699119752e-06,
+      "loss": 0.3918,
+      "step": 3316
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.442680953464537,
+      "learning_rate": 8.981287606166052e-06,
+      "loss": 0.3937,
+      "step": 3317
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.344140038968924,
+      "learning_rate": 8.980606311294423e-06,
+      "loss": 0.449,
+      "step": 3318
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.599485261920849,
+      "learning_rate": 8.979924814539421e-06,
+      "loss": 0.4254,
+      "step": 3319
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.6252772292932525,
+      "learning_rate": 8.979243115935605e-06,
+      "loss": 0.4279,
+      "step": 3320
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 11.7986103752548,
+      "learning_rate": 8.978561215517554e-06,
+      "loss": 0.4176,
+      "step": 3321
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.748714316995197,
+      "learning_rate": 8.977879113319847e-06,
+      "loss": 0.442,
+      "step": 3322
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.241762702156642,
+      "learning_rate": 8.977196809377083e-06,
+      "loss": 0.3964,
+      "step": 3323
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.715637436582787,
+      "learning_rate": 8.97651430372386e-06,
+      "loss": 0.398,
+      "step": 3324
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.436176818197934,
+      "learning_rate": 8.975831596394796e-06,
+      "loss": 0.4114,
+      "step": 3325
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.820611595204566,
+      "learning_rate": 8.975148687424518e-06,
+      "loss": 0.4126,
+      "step": 3326
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 18.44751708041965,
+      "learning_rate": 8.974465576847655e-06,
+      "loss": 0.4253,
+      "step": 3327
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.041719089242733,
+      "learning_rate": 8.973782264698858e-06,
+      "loss": 0.4255,
+      "step": 3328
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.143922263606235,
+      "learning_rate": 8.973098751012779e-06,
+      "loss": 0.4271,
+      "step": 3329
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.944623667185053,
+      "learning_rate": 8.972415035824085e-06,
+      "loss": 0.4085,
+      "step": 3330
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.589661316795609,
+      "learning_rate": 8.97173111916745e-06,
+      "loss": 0.4262,
+      "step": 3331
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.605764006235195,
+      "learning_rate": 8.971047001077561e-06,
+      "loss": 0.4402,
+      "step": 3332
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.458211341578293,
+      "learning_rate": 8.970362681589116e-06,
+      "loss": 0.3839,
+      "step": 3333
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.792042352959507,
+      "learning_rate": 8.969678160736819e-06,
+      "loss": 0.4021,
+      "step": 3334
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.199041438942857,
+      "learning_rate": 8.968993438555387e-06,
+      "loss": 0.4096,
+      "step": 3335
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.940332993455602,
+      "learning_rate": 8.968308515079548e-06,
+      "loss": 0.3906,
+      "step": 3336
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.484123500594629,
+      "learning_rate": 8.967623390344038e-06,
+      "loss": 0.406,
+      "step": 3337
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.134087005970997,
+      "learning_rate": 8.966938064383606e-06,
+      "loss": 0.3788,
+      "step": 3338
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.30367351864369,
+      "learning_rate": 8.96625253723301e-06,
+      "loss": 0.3953,
+      "step": 3339
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3921707138052392,
+      "learning_rate": 8.965566808927012e-06,
+      "loss": 0.185,
+      "step": 3340
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.535520738864979,
+      "learning_rate": 8.964880879500394e-06,
+      "loss": 0.416,
+      "step": 3341
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.302944147499686,
+      "learning_rate": 8.964194748987948e-06,
+      "loss": 0.3624,
+      "step": 3342
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.22483918086355,
+      "learning_rate": 8.963508417424465e-06,
+      "loss": 0.361,
+      "step": 3343
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.85877767573349,
+      "learning_rate": 8.96282188484476e-06,
+      "loss": 0.4087,
+      "step": 3344
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.256164831288972,
+      "learning_rate": 8.962135151283644e-06,
+      "loss": 0.4027,
+      "step": 3345
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.9580668587633685,
+      "learning_rate": 8.961448216775955e-06,
+      "loss": 0.3656,
+      "step": 3346
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 17.947754326361768,
+      "learning_rate": 8.960761081356524e-06,
+      "loss": 0.3806,
+      "step": 3347
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4280223616007336,
+      "learning_rate": 8.960073745060206e-06,
+      "loss": 0.1967,
+      "step": 3348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.531993444307905,
+      "learning_rate": 8.959386207921858e-06,
+      "loss": 0.4065,
+      "step": 3349
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.4048255715577485,
+      "learning_rate": 8.958698469976348e-06,
+      "loss": 0.405,
+      "step": 3350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.411088855166403,
+      "learning_rate": 8.95801053125856e-06,
+      "loss": 0.3962,
+      "step": 3351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.6349381635493,
+      "learning_rate": 8.95732239180338e-06,
+      "loss": 0.4289,
+      "step": 3352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.4422257614082685,
+      "learning_rate": 8.956634051645713e-06,
+      "loss": 0.4021,
+      "step": 3353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 6.949103659098756,
+      "learning_rate": 8.955945510820463e-06,
+      "loss": 0.4304,
+      "step": 3354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 16.938704034295146,
+      "learning_rate": 8.955256769362555e-06,
+      "loss": 0.423,
+      "step": 3355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.409422804370306,
+      "learning_rate": 8.954567827306917e-06,
+      "loss": 0.3904,
+      "step": 3356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 9.018795969948043,
+      "learning_rate": 8.953878684688492e-06,
+      "loss": 0.4193,
+      "step": 3357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.304743443477408,
+      "learning_rate": 8.953189341542232e-06,
+      "loss": 0.4401,
+      "step": 3358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.526514951575284,
+      "learning_rate": 8.952499797903097e-06,
+      "loss": 0.4242,
+      "step": 3359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 29.784233033927556,
+      "learning_rate": 8.951810053806057e-06,
+      "loss": 0.379,
+      "step": 3360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.352661802379565,
+      "learning_rate": 8.951120109286096e-06,
+      "loss": 0.434,
+      "step": 3361
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.494746754357491,
+      "learning_rate": 8.950429964378202e-06,
+      "loss": 0.4047,
+      "step": 3362
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.932203995624338,
+      "learning_rate": 8.949739619117381e-06,
+      "loss": 0.399,
+      "step": 3363
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.1431396544412795,
+      "learning_rate": 8.949049073538642e-06,
+      "loss": 0.412,
+      "step": 3364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.902776449976627,
+      "learning_rate": 8.948358327677012e-06,
+      "loss": 0.4122,
+      "step": 3365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 14.93082040130215,
+      "learning_rate": 8.947667381567518e-06,
+      "loss": 0.446,
+      "step": 3366
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.32157463703174,
+      "learning_rate": 8.946976235245202e-06,
+      "loss": 0.4394,
+      "step": 3367
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3828149340093748,
+      "learning_rate": 8.946284888745123e-06,
+      "loss": 0.1979,
+      "step": 3368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.966725858489611,
+      "learning_rate": 8.945593342102337e-06,
+      "loss": 0.4551,
+      "step": 3369
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.425420969972168,
+      "learning_rate": 8.944901595351922e-06,
+      "loss": 0.4333,
+      "step": 3370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.148315075850866,
+      "learning_rate": 8.94420964852896e-06,
+      "loss": 0.4133,
+      "step": 3371
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.267030543047511,
+      "learning_rate": 8.943517501668541e-06,
+      "loss": 0.4113,
+      "step": 3372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 5.633774979399213,
+      "learning_rate": 8.942825154805773e-06,
+      "loss": 0.3678,
+      "step": 3373
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.467514523700281,
+      "learning_rate": 8.942132607975766e-06,
+      "loss": 0.4546,
+      "step": 3374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 7.833649219912236,
+      "learning_rate": 8.941439861213646e-06,
+      "loss": 0.4099,
+      "step": 3375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.992768909986775,
+      "learning_rate": 8.940746914554546e-06,
+      "loss": 0.462,
+      "step": 3376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3860267729127844,
+      "learning_rate": 8.94005376803361e-06,
+      "loss": 0.2073,
+      "step": 3377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.644520196121361,
+      "learning_rate": 8.939360421685993e-06,
+      "loss": 0.4519,
+      "step": 3378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.995467734467615,
+      "learning_rate": 8.938666875546858e-06,
+      "loss": 0.4034,
+      "step": 3379
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.670796782420588,
+      "learning_rate": 8.93797312965138e-06,
+      "loss": 0.4122,
+      "step": 3380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.912513955470739,
+      "learning_rate": 8.937279184034741e-06,
+      "loss": 0.4344,
+      "step": 3381
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.06156682079896,
+      "learning_rate": 8.936585038732143e-06,
+      "loss": 0.4098,
+      "step": 3382
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.919119166480303,
+      "learning_rate": 8.935890693778781e-06,
+      "loss": 0.4078,
+      "step": 3383
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.421749121356976,
+      "learning_rate": 8.935196149209878e-06,
+      "loss": 0.3978,
+      "step": 3384
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.441306377082623,
+      "learning_rate": 8.934501405060654e-06,
+      "loss": 0.4133,
+      "step": 3385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.239795418690411,
+      "learning_rate": 8.933806461366346e-06,
+      "loss": 0.4188,
+      "step": 3386
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.364182631638924,
+      "learning_rate": 8.9331113181622e-06,
+      "loss": 0.4278,
+      "step": 3387
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 16.13896013227105,
+      "learning_rate": 8.93241597548347e-06,
+      "loss": 0.395,
+      "step": 3388
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 38.32874543086283,
+      "learning_rate": 8.931720433365423e-06,
+      "loss": 0.3984,
+      "step": 3389
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.213754852178367,
+      "learning_rate": 8.931024691843334e-06,
+      "loss": 0.4468,
+      "step": 3390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.207746266059297,
+      "learning_rate": 8.930328750952485e-06,
+      "loss": 0.4159,
+      "step": 3391
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.088888239269108,
+      "learning_rate": 8.929632610728178e-06,
+      "loss": 0.3816,
+      "step": 3392
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 38.29007006607525,
+      "learning_rate": 8.928936271205716e-06,
+      "loss": 0.3979,
+      "step": 3393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 30.995883767889485,
+      "learning_rate": 8.928239732420416e-06,
+      "loss": 0.3855,
+      "step": 3394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.36341935804305747,
+      "learning_rate": 8.927542994407604e-06,
+      "loss": 0.1861,
+      "step": 3395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.092825818077386,
+      "learning_rate": 8.926846057202617e-06,
+      "loss": 0.4114,
+      "step": 3396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.269417139801784,
+      "learning_rate": 8.926148920840797e-06,
+      "loss": 0.3806,
+      "step": 3397
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.396946760938913,
+      "learning_rate": 8.925451585357506e-06,
+      "loss": 0.4407,
+      "step": 3398
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.605520188720552,
+      "learning_rate": 8.924754050788106e-06,
+      "loss": 0.4046,
+      "step": 3399
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.803300252414385,
+      "learning_rate": 8.924056317167976e-06,
+      "loss": 0.414,
+      "step": 3400
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.89785553983669,
+      "learning_rate": 8.923358384532506e-06,
+      "loss": 0.4336,
+      "step": 3401
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.847007480664125,
+      "learning_rate": 8.922660252917088e-06,
+      "loss": 0.4289,
+      "step": 3402
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.664785012104463,
+      "learning_rate": 8.92196192235713e-06,
+      "loss": 0.4329,
+      "step": 3403
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.221976532359281,
+      "learning_rate": 8.92126339288805e-06,
+      "loss": 0.4186,
+      "step": 3404
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 13.20899004276606,
+      "learning_rate": 8.920564664545274e-06,
+      "loss": 0.4533,
+      "step": 3405
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.37914057299680576,
+      "learning_rate": 8.919865737364242e-06,
+      "loss": 0.2001,
+      "step": 3406
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.694243239322333,
+      "learning_rate": 8.919166611380397e-06,
+      "loss": 0.4089,
+      "step": 3407
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.762211553008525,
+      "learning_rate": 8.9184672866292e-06,
+      "loss": 0.4242,
+      "step": 3408
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.399545521961725,
+      "learning_rate": 8.917767763146117e-06,
+      "loss": 0.4086,
+      "step": 3409
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.260212391895895,
+      "learning_rate": 8.917068040966625e-06,
+      "loss": 0.4249,
+      "step": 3410
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.38296920577373184,
+      "learning_rate": 8.916368120126213e-06,
+      "loss": 0.2048,
+      "step": 3411
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.537740587533522,
+      "learning_rate": 8.915668000660378e-06,
+      "loss": 0.4204,
+      "step": 3412
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.352674492313408,
+      "learning_rate": 8.914967682604628e-06,
+      "loss": 0.4471,
+      "step": 3413
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.364426227633205,
+      "learning_rate": 8.91426716599448e-06,
+      "loss": 0.3811,
+      "step": 3414
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.17698595367626,
+      "learning_rate": 8.913566450865462e-06,
+      "loss": 0.4087,
+      "step": 3415
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.9541669290975,
+      "learning_rate": 8.912865537253115e-06,
+      "loss": 0.4262,
+      "step": 3416
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.883956224953547,
+      "learning_rate": 8.912164425192983e-06,
+      "loss": 0.3914,
+      "step": 3417
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.7150710552018005,
+      "learning_rate": 8.911463114720626e-06,
+      "loss": 0.4077,
+      "step": 3418
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.006984687274587,
+      "learning_rate": 8.910761605871611e-06,
+      "loss": 0.4016,
+      "step": 3419
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.508070758276262,
+      "learning_rate": 8.910059898681519e-06,
+      "loss": 0.3819,
+      "step": 3420
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.087831410432681,
+      "learning_rate": 8.909357993185935e-06,
+      "loss": 0.4446,
+      "step": 3421
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.610732480070771,
+      "learning_rate": 8.90865588942046e-06,
+      "loss": 0.4323,
+      "step": 3422
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.928167807775627,
+      "learning_rate": 8.9079535874207e-06,
+      "loss": 0.3962,
+      "step": 3423
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.547171305912741,
+      "learning_rate": 8.907251087222275e-06,
+      "loss": 0.4059,
+      "step": 3424
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.317861511739668,
+      "learning_rate": 8.906548388860813e-06,
+      "loss": 0.4192,
+      "step": 3425
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.8744324368752805,
+      "learning_rate": 8.905845492371953e-06,
+      "loss": 0.4169,
+      "step": 3426
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.571238682258001,
+      "learning_rate": 8.905142397791345e-06,
+      "loss": 0.4409,
+      "step": 3427
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.529341149112951,
+      "learning_rate": 8.904439105154645e-06,
+      "loss": 0.4604,
+      "step": 3428
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 33.406765599629566,
+      "learning_rate": 8.903735614497522e-06,
+      "loss": 0.3874,
+      "step": 3429
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.805831817885569,
+      "learning_rate": 8.903031925855656e-06,
+      "loss": 0.3991,
+      "step": 3430
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.006905334917942,
+      "learning_rate": 8.902328039264737e-06,
+      "loss": 0.4222,
+      "step": 3431
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.287888175446652,
+      "learning_rate": 8.90162395476046e-06,
+      "loss": 0.4371,
+      "step": 3432
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.38747770557501,
+      "learning_rate": 8.900919672378536e-06,
+      "loss": 0.4489,
+      "step": 3433
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.621868570873233,
+      "learning_rate": 8.900215192154685e-06,
+      "loss": 0.42,
+      "step": 3434
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.900295705817765,
+      "learning_rate": 8.899510514124634e-06,
+      "loss": 0.4377,
+      "step": 3435
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.824513206079619,
+      "learning_rate": 8.898805638324125e-06,
+      "loss": 0.4313,
+      "step": 3436
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.146391923663932,
+      "learning_rate": 8.898100564788903e-06,
+      "loss": 0.4213,
+      "step": 3437
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.3105048999881,
+      "learning_rate": 8.897395293554728e-06,
+      "loss": 0.424,
+      "step": 3438
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.551502772425194,
+      "learning_rate": 8.896689824657371e-06,
+      "loss": 0.4142,
+      "step": 3439
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.248803819697771,
+      "learning_rate": 8.895984158132612e-06,
+      "loss": 0.4315,
+      "step": 3440
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.509044521664197,
+      "learning_rate": 8.895278294016235e-06,
+      "loss": 0.3956,
+      "step": 3441
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 25.26899947066078,
+      "learning_rate": 8.894572232344042e-06,
+      "loss": 0.4607,
+      "step": 3442
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.777570428240095,
+      "learning_rate": 8.893865973151846e-06,
+      "loss": 0.431,
+      "step": 3443
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.579721270261604,
+      "learning_rate": 8.89315951647546e-06,
+      "loss": 0.3898,
+      "step": 3444
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.4794702834072355,
+      "learning_rate": 8.892452862350715e-06,
+      "loss": 0.4206,
+      "step": 3445
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.695135099863244,
+      "learning_rate": 8.891746010813454e-06,
+      "loss": 0.4282,
+      "step": 3446
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.350764648147043,
+      "learning_rate": 8.891038961899521e-06,
+      "loss": 0.433,
+      "step": 3447
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4052998082967304,
+      "learning_rate": 8.890331715644778e-06,
+      "loss": 0.2008,
+      "step": 3448
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.532276655665202,
+      "learning_rate": 8.889624272085092e-06,
+      "loss": 0.4058,
+      "step": 3449
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.265977913898389,
+      "learning_rate": 8.888916631256346e-06,
+      "loss": 0.4054,
+      "step": 3450
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.477269068091746,
+      "learning_rate": 8.888208793194426e-06,
+      "loss": 0.4352,
+      "step": 3451
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.858078453441319,
+      "learning_rate": 8.887500757935233e-06,
+      "loss": 0.3872,
+      "step": 3452
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.6447921589876575,
+      "learning_rate": 8.886792525514674e-06,
+      "loss": 0.4452,
+      "step": 3453
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.93500606438134,
+      "learning_rate": 8.886084095968671e-06,
+      "loss": 0.412,
+      "step": 3454
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.382878437977981,
+      "learning_rate": 8.885375469333151e-06,
+      "loss": 0.355,
+      "step": 3455
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.011796677187269,
+      "learning_rate": 8.884666645644056e-06,
+      "loss": 0.4732,
+      "step": 3456
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.285820052366434,
+      "learning_rate": 8.883957624937333e-06,
+      "loss": 0.4118,
+      "step": 3457
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.61772611335125,
+      "learning_rate": 8.883248407248941e-06,
+      "loss": 0.4585,
+      "step": 3458
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.930499098747047,
+      "learning_rate": 8.88253899261485e-06,
+      "loss": 0.3574,
+      "step": 3459
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.116664271591459,
+      "learning_rate": 8.881829381071039e-06,
+      "loss": 0.4005,
+      "step": 3460
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.946021207439635,
+      "learning_rate": 8.881119572653498e-06,
+      "loss": 0.3811,
+      "step": 3461
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.624097280049816,
+      "learning_rate": 8.880409567398225e-06,
+      "loss": 0.4226,
+      "step": 3462
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.246309102380301,
+      "learning_rate": 8.87969936534123e-06,
+      "loss": 0.4421,
+      "step": 3463
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.738500994391188,
+      "learning_rate": 8.878988966518532e-06,
+      "loss": 0.408,
+      "step": 3464
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.593990090706731,
+      "learning_rate": 8.878278370966157e-06,
+      "loss": 0.4585,
+      "step": 3465
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.236979244934119,
+      "learning_rate": 8.87756757872015e-06,
+      "loss": 0.4,
+      "step": 3466
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.61915953908338,
+      "learning_rate": 8.876856589816555e-06,
+      "loss": 0.3757,
+      "step": 3467
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.5851781590111855,
+      "learning_rate": 8.876145404291436e-06,
+      "loss": 0.4298,
+      "step": 3468
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 11.078922849821483,
+      "learning_rate": 8.875434022180857e-06,
+      "loss": 0.434,
+      "step": 3469
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.46232124002319414,
+      "learning_rate": 8.874722443520898e-06,
+      "loss": 0.2131,
+      "step": 3470
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.0648625491496935,
+      "learning_rate": 8.874010668347651e-06,
+      "loss": 0.3819,
+      "step": 3471
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.853825603637389,
+      "learning_rate": 8.873298696697213e-06,
+      "loss": 0.4381,
+      "step": 3472
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.3438072181774709,
+      "learning_rate": 8.872586528605692e-06,
+      "loss": 0.1976,
+      "step": 3473
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.238838433241803,
+      "learning_rate": 8.871874164109206e-06,
+      "loss": 0.4112,
+      "step": 3474
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.858718186002319,
+      "learning_rate": 8.871161603243887e-06,
+      "loss": 0.4103,
+      "step": 3475
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 12.205786530811393,
+      "learning_rate": 8.870448846045873e-06,
+      "loss": 0.4337,
+      "step": 3476
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.748478405607919,
+      "learning_rate": 8.869735892551312e-06,
+      "loss": 0.4066,
+      "step": 3477
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.517784193553046,
+      "learning_rate": 8.86902274279636e-06,
+      "loss": 0.4334,
+      "step": 3478
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.205380608078882,
+      "learning_rate": 8.86830939681719e-06,
+      "loss": 0.38,
+      "step": 3479
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.0970491951798955,
+      "learning_rate": 8.86759585464998e-06,
+      "loss": 0.4084,
+      "step": 3480
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 14.23867593830122,
+      "learning_rate": 8.866882116330914e-06,
+      "loss": 0.4127,
+      "step": 3481
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.4431517803021,
+      "learning_rate": 8.866168181896198e-06,
+      "loss": 0.4223,
+      "step": 3482
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.64349646468008,
+      "learning_rate": 8.865454051382033e-06,
+      "loss": 0.4155,
+      "step": 3483
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.50761772221549,
+      "learning_rate": 8.864739724824643e-06,
+      "loss": 0.3981,
+      "step": 3484
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.6249296368799655,
+      "learning_rate": 8.864025202260252e-06,
+      "loss": 0.4151,
+      "step": 3485
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.321257193627888,
+      "learning_rate": 8.863310483725101e-06,
+      "loss": 0.4308,
+      "step": 3486
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.559776299180404,
+      "learning_rate": 8.862595569255436e-06,
+      "loss": 0.4377,
+      "step": 3487
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5330244504382744,
+      "learning_rate": 8.861880458887518e-06,
+      "loss": 0.2053,
+      "step": 3488
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.461715111430202,
+      "learning_rate": 8.861165152657614e-06,
+      "loss": 0.2117,
+      "step": 3489
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.082530330085976,
+      "learning_rate": 8.860449650602002e-06,
+      "loss": 0.4093,
+      "step": 3490
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8591173253331945,
+      "learning_rate": 8.859733952756967e-06,
+      "loss": 0.4086,
+      "step": 3491
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.663619682464771,
+      "learning_rate": 8.85901805915881e-06,
+      "loss": 0.4354,
+      "step": 3492
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.8003529240544776,
+      "learning_rate": 8.85830196984384e-06,
+      "loss": 0.3966,
+      "step": 3493
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.850593818019642,
+      "learning_rate": 8.857585684848373e-06,
+      "loss": 0.4163,
+      "step": 3494
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.548046170430667,
+      "learning_rate": 8.856869204208734e-06,
+      "loss": 0.4412,
+      "step": 3495
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 19.453416541914397,
+      "learning_rate": 8.856152527961264e-06,
+      "loss": 0.4335,
+      "step": 3496
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.72586278407872,
+      "learning_rate": 8.85543565614231e-06,
+      "loss": 0.4141,
+      "step": 3497
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.404278145853276,
+      "learning_rate": 8.854718588788229e-06,
+      "loss": 0.4212,
+      "step": 3498
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.915425259479051,
+      "learning_rate": 8.854001325935388e-06,
+      "loss": 0.3923,
+      "step": 3499
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.634055661868227,
+      "learning_rate": 8.853283867620167e-06,
+      "loss": 0.3925,
+      "step": 3500
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.758252930208663,
+      "learning_rate": 8.852566213878947e-06,
+      "loss": 0.4325,
+      "step": 3501
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.586601718851137,
+      "learning_rate": 8.85184836474813e-06,
+      "loss": 0.3707,
+      "step": 3502
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.7114461377754555,
+      "learning_rate": 8.851130320264122e-06,
+      "loss": 0.4341,
+      "step": 3503
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.353477158113464,
+      "learning_rate": 8.85041208046334e-06,
+      "loss": 0.3792,
+      "step": 3504
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.926169344829891,
+      "learning_rate": 8.849693645382209e-06,
+      "loss": 0.429,
+      "step": 3505
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 17.411012701495505,
+      "learning_rate": 8.848975015057169e-06,
+      "loss": 0.4542,
+      "step": 3506
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 15.773837784253056,
+      "learning_rate": 8.848256189524661e-06,
+      "loss": 0.4278,
+      "step": 3507
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 18.390408109925072,
+      "learning_rate": 8.847537168821148e-06,
+      "loss": 0.4452,
+      "step": 3508
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.372998098389802,
+      "learning_rate": 8.846817952983094e-06,
+      "loss": 0.4309,
+      "step": 3509
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9968841533236955,
+      "learning_rate": 8.846098542046972e-06,
+      "loss": 0.2653,
+      "step": 3510
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 10.189374279107382,
+      "learning_rate": 8.845378936049274e-06,
+      "loss": 0.3719,
+      "step": 3511
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.256351502332146,
+      "learning_rate": 8.844659135026493e-06,
+      "loss": 0.4062,
+      "step": 3512
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.8713307112033344,
+      "learning_rate": 8.843939139015132e-06,
+      "loss": 0.404,
+      "step": 3513
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 5.518231592871066,
+      "learning_rate": 8.843218948051712e-06,
+      "loss": 0.4259,
+      "step": 3514
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.976851016048127,
+      "learning_rate": 8.842498562172757e-06,
+      "loss": 0.4344,
+      "step": 3515
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.582102119745656,
+      "learning_rate": 8.841777981414802e-06,
+      "loss": 0.4124,
+      "step": 3516
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.460014527039142,
+      "learning_rate": 8.841057205814391e-06,
+      "loss": 0.4475,
+      "step": 3517
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.059335970656048,
+      "learning_rate": 8.840336235408082e-06,
+      "loss": 0.4308,
+      "step": 3518
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.741989341360535,
+      "learning_rate": 8.839615070232438e-06,
+      "loss": 0.3846,
+      "step": 3519
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 7.362418534532932,
+      "learning_rate": 8.838893710324037e-06,
+      "loss": 0.4449,
+      "step": 3520
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 6.8388087338794135,
+      "learning_rate": 8.838172155719463e-06,
+      "loss": 0.4373,
+      "step": 3521
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 9.530708570528821,
+      "learning_rate": 8.83745040645531e-06,
+      "loss": 0.4303,
+      "step": 3522
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 8.50888929351661,
+      "learning_rate": 8.836728462568182e-06,
+      "loss": 0.4586,
+      "step": 3523
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.840998250722504,
+      "learning_rate": 8.836006324094697e-06,
+      "loss": 0.4503,
+      "step": 3524
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 15.259374214297326,
+      "learning_rate": 8.835283991071478e-06,
+      "loss": 0.4608,
+      "step": 3525
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6723748410696522,
+      "learning_rate": 8.834561463535155e-06,
+      "loss": 0.234,
+      "step": 3526
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 38.72111877720752,
+      "learning_rate": 8.833838741522379e-06,
+      "loss": 0.46,
+      "step": 3527
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.83119447649275,
+      "learning_rate": 8.833115825069801e-06,
+      "loss": 0.4297,
+      "step": 3528
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 88.26892882943325,
+      "learning_rate": 8.832392714214082e-06,
+      "loss": 0.4174,
+      "step": 3529
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.544443563177653,
+      "learning_rate": 8.8316694089919e-06,
+      "loss": 0.4281,
+      "step": 3530
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.573128980933161,
+      "learning_rate": 8.83094590943994e-06,
+      "loss": 0.4249,
+      "step": 3531
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.801833347997774,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.4516,
+      "step": 3532
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.785415879882619,
+      "learning_rate": 8.829498327493458e-06,
+      "loss": 0.424,
+      "step": 3533
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.229262911171039,
+      "learning_rate": 8.828774245172357e-06,
+      "loss": 0.4179,
+      "step": 3534
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.630561486524478,
+      "learning_rate": 8.828049968668308e-06,
+      "loss": 0.4001,
+      "step": 3535
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.62565257304474,
+      "learning_rate": 8.827325498018042e-06,
+      "loss": 0.4602,
+      "step": 3536
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.3307428794113445,
+      "learning_rate": 8.826600833258307e-06,
+      "loss": 0.4385,
+      "step": 3537
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.4422800618049743,
+      "learning_rate": 8.825875974425853e-06,
+      "loss": 0.4128,
+      "step": 3538
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.776200913225567,
+      "learning_rate": 8.825150921557443e-06,
+      "loss": 0.4354,
+      "step": 3539
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.405150809151248,
+      "learning_rate": 8.824425674689849e-06,
+      "loss": 0.41,
+      "step": 3540
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.336930814360116,
+      "learning_rate": 8.823700233859854e-06,
+      "loss": 0.4027,
+      "step": 3541
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.41726024474670026,
+      "learning_rate": 8.822974599104246e-06,
+      "loss": 0.2151,
+      "step": 3542
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.6626001244009725,
+      "learning_rate": 8.822248770459832e-06,
+      "loss": 0.4449,
+      "step": 3543
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.39800700976199,
+      "learning_rate": 8.821522747963422e-06,
+      "loss": 0.3812,
+      "step": 3544
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.963857257770383,
+      "learning_rate": 8.820796531651836e-06,
+      "loss": 0.3896,
+      "step": 3545
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.376593060360611,
+      "learning_rate": 8.820070121561906e-06,
+      "loss": 0.4162,
+      "step": 3546
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.803908844489251,
+      "learning_rate": 8.819343517730476e-06,
+      "loss": 0.4077,
+      "step": 3547
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.374405473200704,
+      "learning_rate": 8.818616720194393e-06,
+      "loss": 0.3786,
+      "step": 3548
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 22.864042708533255,
+      "learning_rate": 8.81788972899052e-06,
+      "loss": 0.4205,
+      "step": 3549
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.193317246662718,
+      "learning_rate": 8.817162544155726e-06,
+      "loss": 0.434,
+      "step": 3550
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.64588608972558,
+      "learning_rate": 8.816435165726892e-06,
+      "loss": 0.3416,
+      "step": 3551
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.587730400816529,
+      "learning_rate": 8.815707593740909e-06,
+      "loss": 0.4451,
+      "step": 3552
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.599449755165275,
+      "learning_rate": 8.814979828234678e-06,
+      "loss": 0.3998,
+      "step": 3553
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.866755670253131,
+      "learning_rate": 8.814251869245106e-06,
+      "loss": 0.4292,
+      "step": 3554
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.747178522771021,
+      "learning_rate": 8.813523716809115e-06,
+      "loss": 0.4232,
+      "step": 3555
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.1579792196589045,
+      "learning_rate": 8.812795370963635e-06,
+      "loss": 0.3567,
+      "step": 3556
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.555981091757401,
+      "learning_rate": 8.812066831745602e-06,
+      "loss": 0.4381,
+      "step": 3557
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.328764160012112,
+      "learning_rate": 8.81133809919197e-06,
+      "loss": 0.4303,
+      "step": 3558
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 19.601574025143634,
+      "learning_rate": 8.810609173339693e-06,
+      "loss": 0.412,
+      "step": 3559
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.413785627924609,
+      "learning_rate": 8.809880054225744e-06,
+      "loss": 0.4138,
+      "step": 3560
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.665112138695362,
+      "learning_rate": 8.809150741887099e-06,
+      "loss": 0.3816,
+      "step": 3561
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.120627743263768,
+      "learning_rate": 8.808421236360745e-06,
+      "loss": 0.4146,
+      "step": 3562
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.0017893551955215,
+      "learning_rate": 8.807691537683685e-06,
+      "loss": 0.3854,
+      "step": 3563
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.192373395104961,
+      "learning_rate": 8.806961645892921e-06,
+      "loss": 0.4363,
+      "step": 3564
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.43452542405144895,
+      "learning_rate": 8.806231561025476e-06,
+      "loss": 0.2006,
+      "step": 3565
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.174053732112068,
+      "learning_rate": 8.805501283118374e-06,
+      "loss": 0.3966,
+      "step": 3566
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.354883476623773,
+      "learning_rate": 8.804770812208655e-06,
+      "loss": 0.428,
+      "step": 3567
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.147474485989454,
+      "learning_rate": 8.804040148333364e-06,
+      "loss": 0.3887,
+      "step": 3568
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.079177476270363,
+      "learning_rate": 8.803309291529557e-06,
+      "loss": 0.3727,
+      "step": 3569
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.519355813836595,
+      "learning_rate": 8.802578241834306e-06,
+      "loss": 0.4119,
+      "step": 3570
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.684476812880178,
+      "learning_rate": 8.80184699928468e-06,
+      "loss": 0.4174,
+      "step": 3571
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.721557404693078,
+      "learning_rate": 8.80111556391777e-06,
+      "loss": 0.437,
+      "step": 3572
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.507672768603951,
+      "learning_rate": 8.800383935770672e-06,
+      "loss": 0.422,
+      "step": 3573
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.01588023529658,
+      "learning_rate": 8.799652114880489e-06,
+      "loss": 0.4486,
+      "step": 3574
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.750668039364788,
+      "learning_rate": 8.79892010128434e-06,
+      "loss": 0.4728,
+      "step": 3575
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.741971360991096,
+      "learning_rate": 8.798187895019348e-06,
+      "loss": 0.42,
+      "step": 3576
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.627292086318342,
+      "learning_rate": 8.797455496122647e-06,
+      "loss": 0.3888,
+      "step": 3577
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.326938231970756,
+      "learning_rate": 8.796722904631386e-06,
+      "loss": 0.3958,
+      "step": 3578
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.28129754761106,
+      "learning_rate": 8.795990120582717e-06,
+      "loss": 0.44,
+      "step": 3579
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.863294244639911,
+      "learning_rate": 8.795257144013805e-06,
+      "loss": 0.3893,
+      "step": 3580
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.973528606699655,
+      "learning_rate": 8.794523974961822e-06,
+      "loss": 0.4044,
+      "step": 3581
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.6070562990337005,
+      "learning_rate": 8.793790613463956e-06,
+      "loss": 0.4696,
+      "step": 3582
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.529221769693967,
+      "learning_rate": 8.793057059557397e-06,
+      "loss": 0.4378,
+      "step": 3583
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.472785532101751,
+      "learning_rate": 8.79232331327935e-06,
+      "loss": 0.4286,
+      "step": 3584
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.8220608925293815,
+      "learning_rate": 8.791589374667029e-06,
+      "loss": 0.4017,
+      "step": 3585
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.75964295091102,
+      "learning_rate": 8.790855243757657e-06,
+      "loss": 0.4175,
+      "step": 3586
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.021320340785288,
+      "learning_rate": 8.790120920588466e-06,
+      "loss": 0.4187,
+      "step": 3587
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.43027129040124,
+      "learning_rate": 8.789386405196695e-06,
+      "loss": 0.4222,
+      "step": 3588
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.94320917364787,
+      "learning_rate": 8.788651697619604e-06,
+      "loss": 0.4101,
+      "step": 3589
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.567421986041884,
+      "learning_rate": 8.787916797894448e-06,
+      "loss": 0.3994,
+      "step": 3590
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.771838944462428,
+      "learning_rate": 8.787181706058502e-06,
+      "loss": 0.3991,
+      "step": 3591
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.572357425982563,
+      "learning_rate": 8.786446422149046e-06,
+      "loss": 0.4329,
+      "step": 3592
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.500810831638548,
+      "learning_rate": 8.785710946203374e-06,
+      "loss": 0.4353,
+      "step": 3593
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.117178587979517,
+      "learning_rate": 8.784975278258783e-06,
+      "loss": 0.4407,
+      "step": 3594
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.305652493252394,
+      "learning_rate": 8.784239418352588e-06,
+      "loss": 0.426,
+      "step": 3595
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.964834341484513,
+      "learning_rate": 8.783503366522103e-06,
+      "loss": 0.3777,
+      "step": 3596
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.240577245456373,
+      "learning_rate": 8.782767122804664e-06,
+      "loss": 0.4368,
+      "step": 3597
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 10.71743000362856,
+      "learning_rate": 8.78203068723761e-06,
+      "loss": 0.4017,
+      "step": 3598
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.970899283597831,
+      "learning_rate": 8.781294059858289e-06,
+      "loss": 0.4141,
+      "step": 3599
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.484059734425786,
+      "learning_rate": 8.780557240704062e-06,
+      "loss": 0.3943,
+      "step": 3600
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.095103941394403,
+      "learning_rate": 8.779820229812297e-06,
+      "loss": 0.434,
+      "step": 3601
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.342156817253132,
+      "learning_rate": 8.779083027220372e-06,
+      "loss": 0.4446,
+      "step": 3602
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.060496655059029,
+      "learning_rate": 8.778345632965677e-06,
+      "loss": 0.4622,
+      "step": 3603
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.682285449330078,
+      "learning_rate": 8.777608047085608e-06,
+      "loss": 0.3859,
+      "step": 3604
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.5723158326988744,
+      "learning_rate": 8.776870269617577e-06,
+      "loss": 0.4383,
+      "step": 3605
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.063637465863592,
+      "learning_rate": 8.776132300598998e-06,
+      "loss": 0.438,
+      "step": 3606
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.0826424965464305,
+      "learning_rate": 8.775394140067299e-06,
+      "loss": 0.4279,
+      "step": 3607
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.707453820283508,
+      "learning_rate": 8.774655788059918e-06,
+      "loss": 0.4533,
+      "step": 3608
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.663530281377186,
+      "learning_rate": 8.773917244614303e-06,
+      "loss": 0.4069,
+      "step": 3609
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 14.483439821482909,
+      "learning_rate": 8.773178509767907e-06,
+      "loss": 0.4015,
+      "step": 3610
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.6847482032693035,
+      "learning_rate": 8.772439583558198e-06,
+      "loss": 0.4342,
+      "step": 3611
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.833000448992967,
+      "learning_rate": 8.771700466022655e-06,
+      "loss": 0.4329,
+      "step": 3612
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.488274628596997,
+      "learning_rate": 8.770961157198759e-06,
+      "loss": 0.4438,
+      "step": 3613
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.7751949678592975,
+      "learning_rate": 8.770221657124007e-06,
+      "loss": 0.4426,
+      "step": 3614
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.859237574602504,
+      "learning_rate": 8.769481965835904e-06,
+      "loss": 0.4107,
+      "step": 3615
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.129422726351327,
+      "learning_rate": 8.768742083371967e-06,
+      "loss": 0.3932,
+      "step": 3616
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4490188707215232,
+      "learning_rate": 8.768002009769716e-06,
+      "loss": 0.2139,
+      "step": 3617
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.2144810266286,
+      "learning_rate": 8.767261745066689e-06,
+      "loss": 0.425,
+      "step": 3618
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.104841872373646,
+      "learning_rate": 8.766521289300428e-06,
+      "loss": 0.4461,
+      "step": 3619
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.937057950979441,
+      "learning_rate": 8.765780642508484e-06,
+      "loss": 0.4326,
+      "step": 3620
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.052813382900106,
+      "learning_rate": 8.765039804728425e-06,
+      "loss": 0.4185,
+      "step": 3621
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8191906416219057,
+      "learning_rate": 8.764298775997823e-06,
+      "loss": 0.396,
+      "step": 3622
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.555515743706787,
+      "learning_rate": 8.763557556354258e-06,
+      "loss": 0.3999,
+      "step": 3623
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.900231421238809,
+      "learning_rate": 8.762816145835325e-06,
+      "loss": 0.4169,
+      "step": 3624
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.954113646661395,
+      "learning_rate": 8.762074544478622e-06,
+      "loss": 0.443,
+      "step": 3625
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.167070939659235,
+      "learning_rate": 8.761332752321765e-06,
+      "loss": 0.4258,
+      "step": 3626
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.835097335828513,
+      "learning_rate": 8.760590769402372e-06,
+      "loss": 0.4362,
+      "step": 3627
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.7044733385440227,
+      "learning_rate": 8.759848595758077e-06,
+      "loss": 0.3852,
+      "step": 3628
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4979291011280819,
+      "learning_rate": 8.75910623142652e-06,
+      "loss": 0.2181,
+      "step": 3629
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.259943267284958,
+      "learning_rate": 8.75836367644535e-06,
+      "loss": 0.3957,
+      "step": 3630
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.165322732265774,
+      "learning_rate": 8.757620930852225e-06,
+      "loss": 0.3816,
+      "step": 3631
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.221015215884838,
+      "learning_rate": 8.756877994684818e-06,
+      "loss": 0.4218,
+      "step": 3632
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.477090269643131,
+      "learning_rate": 8.756134867980808e-06,
+      "loss": 0.4187,
+      "step": 3633
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.730356255989284,
+      "learning_rate": 8.755391550777883e-06,
+      "loss": 0.4233,
+      "step": 3634
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.019641142974036,
+      "learning_rate": 8.754648043113742e-06,
+      "loss": 0.4221,
+      "step": 3635
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8230255850721284,
+      "learning_rate": 8.753904345026091e-06,
+      "loss": 0.4046,
+      "step": 3636
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.598559680381784,
+      "learning_rate": 8.753160456552653e-06,
+      "loss": 0.4128,
+      "step": 3637
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.618946631136109,
+      "learning_rate": 8.752416377731151e-06,
+      "loss": 0.3939,
+      "step": 3638
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.149267179745354,
+      "learning_rate": 8.751672108599325e-06,
+      "loss": 0.3977,
+      "step": 3639
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.4239699153775565,
+      "learning_rate": 8.75092764919492e-06,
+      "loss": 0.4414,
+      "step": 3640
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 18.01808769564983,
+      "learning_rate": 8.750182999555692e-06,
+      "loss": 0.4114,
+      "step": 3641
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.186932871543475,
+      "learning_rate": 8.74943815971941e-06,
+      "loss": 0.4072,
+      "step": 3642
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.367906056696397,
+      "learning_rate": 8.748693129723849e-06,
+      "loss": 0.4442,
+      "step": 3643
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 33.07234108514747,
+      "learning_rate": 8.747947909606793e-06,
+      "loss": 0.4349,
+      "step": 3644
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.671782338461613,
+      "learning_rate": 8.747202499406039e-06,
+      "loss": 0.4245,
+      "step": 3645
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.9711227018701285,
+      "learning_rate": 8.74645689915939e-06,
+      "loss": 0.3949,
+      "step": 3646
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.400799837964464,
+      "learning_rate": 8.745711108904661e-06,
+      "loss": 0.3991,
+      "step": 3647
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.769934695123766,
+      "learning_rate": 8.744965128679676e-06,
+      "loss": 0.4503,
+      "step": 3648
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.4761828594571424,
+      "learning_rate": 8.74421895852227e-06,
+      "loss": 0.4413,
+      "step": 3649
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.053579870699075,
+      "learning_rate": 8.743472598470286e-06,
+      "loss": 0.4168,
+      "step": 3650
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 11.248084090261868,
+      "learning_rate": 8.742726048561574e-06,
+      "loss": 0.4392,
+      "step": 3651
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 9.07735898200353,
+      "learning_rate": 8.741979308834e-06,
+      "loss": 0.4272,
+      "step": 3652
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.580128734411428,
+      "learning_rate": 8.741232379325434e-06,
+      "loss": 0.4394,
+      "step": 3653
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.70194504359723,
+      "learning_rate": 8.740485260073761e-06,
+      "loss": 0.4113,
+      "step": 3654
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.100102477160323,
+      "learning_rate": 8.739737951116869e-06,
+      "loss": 0.4288,
+      "step": 3655
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.844681463655549,
+      "learning_rate": 8.73899045249266e-06,
+      "loss": 0.4405,
+      "step": 3656
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.391912952822545,
+      "learning_rate": 8.738242764239046e-06,
+      "loss": 0.4303,
+      "step": 3657
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 8.300383941548649,
+      "learning_rate": 8.737494886393946e-06,
+      "loss": 0.4443,
+      "step": 3658
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.24649663276522,
+      "learning_rate": 8.736746818995289e-06,
+      "loss": 0.4457,
+      "step": 3659
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.7863984233250365,
+      "learning_rate": 8.735998562081018e-06,
+      "loss": 0.4536,
+      "step": 3660
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.835423305963661,
+      "learning_rate": 8.735250115689079e-06,
+      "loss": 0.4159,
+      "step": 3661
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 23.535449868717127,
+      "learning_rate": 8.734501479857431e-06,
+      "loss": 0.4577,
+      "step": 3662
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 16.903572288787707,
+      "learning_rate": 8.733752654624044e-06,
+      "loss": 0.4248,
+      "step": 3663
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.860518118997537,
+      "learning_rate": 8.733003640026895e-06,
+      "loss": 0.391,
+      "step": 3664
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.6450420337407845,
+      "learning_rate": 8.732254436103971e-06,
+      "loss": 0.3794,
+      "step": 3665
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 4.711331249871591,
+      "learning_rate": 8.731505042893272e-06,
+      "loss": 0.4442,
+      "step": 3666
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.914490235724631,
+      "learning_rate": 8.7307554604328e-06,
+      "loss": 0.4309,
+      "step": 3667
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.6363906056941,
+      "learning_rate": 8.730005688760575e-06,
+      "loss": 0.4011,
+      "step": 3668
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.7835115520083935,
+      "learning_rate": 8.729255727914622e-06,
+      "loss": 0.4243,
+      "step": 3669
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.681601053854818,
+      "learning_rate": 8.728505577932977e-06,
+      "loss": 0.4428,
+      "step": 3670
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.006196629352439,
+      "learning_rate": 8.727755238853684e-06,
+      "loss": 0.4191,
+      "step": 3671
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.969705899134394,
+      "learning_rate": 8.7270047107148e-06,
+      "loss": 0.3823,
+      "step": 3672
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.742137681749228,
+      "learning_rate": 8.726253993554385e-06,
+      "loss": 0.4123,
+      "step": 3673
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.3808713232337,
+      "learning_rate": 8.725503087410518e-06,
+      "loss": 0.4327,
+      "step": 3674
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.723531284661849,
+      "learning_rate": 8.724751992321279e-06,
+      "loss": 0.4384,
+      "step": 3675
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.749963187152922,
+      "learning_rate": 8.724000708324763e-06,
+      "loss": 0.4297,
+      "step": 3676
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.231102629594711,
+      "learning_rate": 8.723249235459073e-06,
+      "loss": 0.3944,
+      "step": 3677
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.540938689611503,
+      "learning_rate": 8.722497573762319e-06,
+      "loss": 0.4253,
+      "step": 3678
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.865900393324832,
+      "learning_rate": 8.721745723272622e-06,
+      "loss": 0.4196,
+      "step": 3679
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.45534394439289844,
+      "learning_rate": 8.72099368402812e-06,
+      "loss": 0.2119,
+      "step": 3680
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.726861966855348,
+      "learning_rate": 8.720241456066947e-06,
+      "loss": 0.3804,
+      "step": 3681
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.438662721254923,
+      "learning_rate": 8.719489039427256e-06,
+      "loss": 0.4473,
+      "step": 3682
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.313141994040611,
+      "learning_rate": 8.718736434147208e-06,
+      "loss": 0.4207,
+      "step": 3683
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.717903360527084,
+      "learning_rate": 8.717983640264972e-06,
+      "loss": 0.4393,
+      "step": 3684
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.971657798667438,
+      "learning_rate": 8.717230657818725e-06,
+      "loss": 0.4402,
+      "step": 3685
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.228649417409564,
+      "learning_rate": 8.71647748684666e-06,
+      "loss": 0.3792,
+      "step": 3686
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 39.290699585473455,
+      "learning_rate": 8.715724127386971e-06,
+      "loss": 0.4218,
+      "step": 3687
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.330875035384628,
+      "learning_rate": 8.71497057947787e-06,
+      "loss": 0.408,
+      "step": 3688
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.047325156125641,
+      "learning_rate": 8.714216843157574e-06,
+      "loss": 0.4333,
+      "step": 3689
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.234745912577581,
+      "learning_rate": 8.713462918464306e-06,
+      "loss": 0.4907,
+      "step": 3690
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.523123290145355,
+      "learning_rate": 8.712708805436307e-06,
+      "loss": 0.4091,
+      "step": 3691
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.411585234786176,
+      "learning_rate": 8.711954504111821e-06,
+      "loss": 0.4262,
+      "step": 3692
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.063364683122515,
+      "learning_rate": 8.711200014529104e-06,
+      "loss": 0.4307,
+      "step": 3693
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.567193306740524,
+      "learning_rate": 8.710445336726424e-06,
+      "loss": 0.4513,
+      "step": 3694
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.448188020698635,
+      "learning_rate": 8.70969047074205e-06,
+      "loss": 0.3944,
+      "step": 3695
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.971872628689784,
+      "learning_rate": 8.708935416614271e-06,
+      "loss": 0.3834,
+      "step": 3696
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.833426255909726,
+      "learning_rate": 8.708180174381378e-06,
+      "loss": 0.3908,
+      "step": 3697
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.219527279932674,
+      "learning_rate": 8.707424744081678e-06,
+      "loss": 0.4035,
+      "step": 3698
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.155389897845797,
+      "learning_rate": 8.70666912575348e-06,
+      "loss": 0.4138,
+      "step": 3699
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5619337196559009,
+      "learning_rate": 8.705913319435111e-06,
+      "loss": 0.2211,
+      "step": 3700
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.23539479267877,
+      "learning_rate": 8.705157325164898e-06,
+      "loss": 0.4016,
+      "step": 3701
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.603823139717053,
+      "learning_rate": 8.704401142981184e-06,
+      "loss": 0.4594,
+      "step": 3702
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.65193962128843,
+      "learning_rate": 8.703644772922324e-06,
+      "loss": 0.3985,
+      "step": 3703
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.85363114567961,
+      "learning_rate": 8.702888215026674e-06,
+      "loss": 0.4242,
+      "step": 3704
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.107003728291734,
+      "learning_rate": 8.702131469332607e-06,
+      "loss": 0.4581,
+      "step": 3705
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.082622072202591,
+      "learning_rate": 8.701374535878501e-06,
+      "loss": 0.4451,
+      "step": 3706
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.305164361934774,
+      "learning_rate": 8.700617414702746e-06,
+      "loss": 0.4217,
+      "step": 3707
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.00805268471809,
+      "learning_rate": 8.69986010584374e-06,
+      "loss": 0.4411,
+      "step": 3708
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.587269992115975,
+      "learning_rate": 8.699102609339891e-06,
+      "loss": 0.4011,
+      "step": 3709
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.44057925237117,
+      "learning_rate": 8.698344925229618e-06,
+      "loss": 0.3919,
+      "step": 3710
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.26884568090387,
+      "learning_rate": 8.697587053551348e-06,
+      "loss": 0.4255,
+      "step": 3711
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.648906804673404,
+      "learning_rate": 8.696828994343518e-06,
+      "loss": 0.4486,
+      "step": 3712
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.957776416532607,
+      "learning_rate": 8.696070747644572e-06,
+      "loss": 0.4208,
+      "step": 3713
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.720021702669705,
+      "learning_rate": 8.69531231349297e-06,
+      "loss": 0.4497,
+      "step": 3714
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.560697697188086,
+      "learning_rate": 8.694553691927172e-06,
+      "loss": 0.4382,
+      "step": 3715
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 14.261071306669805,
+      "learning_rate": 8.693794882985657e-06,
+      "loss": 0.42,
+      "step": 3716
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.676837868312909,
+      "learning_rate": 8.693035886706909e-06,
+      "loss": 0.4,
+      "step": 3717
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.88429143147426,
+      "learning_rate": 8.692276703129421e-06,
+      "loss": 0.4396,
+      "step": 3718
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.765786697476335,
+      "learning_rate": 8.691517332291694e-06,
+      "loss": 0.4342,
+      "step": 3719
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.687634738951716,
+      "learning_rate": 8.690757774232243e-06,
+      "loss": 0.3797,
+      "step": 3720
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.786092695888376,
+      "learning_rate": 8.689998028989592e-06,
+      "loss": 0.4087,
+      "step": 3721
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.287567983941374,
+      "learning_rate": 8.68923809660227e-06,
+      "loss": 0.4001,
+      "step": 3722
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.848719170801143,
+      "learning_rate": 8.68847797710882e-06,
+      "loss": 0.436,
+      "step": 3723
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5898247899221543,
+      "learning_rate": 8.687717670547793e-06,
+      "loss": 0.2255,
+      "step": 3724
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.520142702180221,
+      "learning_rate": 8.686957176957747e-06,
+      "loss": 0.4019,
+      "step": 3725
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.535535314921744,
+      "learning_rate": 8.686196496377252e-06,
+      "loss": 0.4005,
+      "step": 3726
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.019487981691488,
+      "learning_rate": 8.68543562884489e-06,
+      "loss": 0.4346,
+      "step": 3727
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.8618918338051635,
+      "learning_rate": 8.684674574399248e-06,
+      "loss": 0.3909,
+      "step": 3728
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.555497377856864,
+      "learning_rate": 8.683913333078924e-06,
+      "loss": 0.442,
+      "step": 3729
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.19484446899886,
+      "learning_rate": 8.683151904922526e-06,
+      "loss": 0.4662,
+      "step": 3730
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4000598464838781,
+      "learning_rate": 8.682390289968671e-06,
+      "loss": 0.1888,
+      "step": 3731
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.979738098149098,
+      "learning_rate": 8.681628488255986e-06,
+      "loss": 0.3524,
+      "step": 3732
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.547202458785662,
+      "learning_rate": 8.68086649982311e-06,
+      "loss": 0.4704,
+      "step": 3733
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 16.616702852907736,
+      "learning_rate": 8.68010432470868e-06,
+      "loss": 0.4141,
+      "step": 3734
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 9.477385294409288,
+      "learning_rate": 8.67934196295136e-06,
+      "loss": 0.4264,
+      "step": 3735
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.828308450418506,
+      "learning_rate": 8.67857941458981e-06,
+      "loss": 0.4254,
+      "step": 3736
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.160729923629188,
+      "learning_rate": 8.677816679662705e-06,
+      "loss": 0.4239,
+      "step": 3737
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.712773547977068,
+      "learning_rate": 8.677053758208727e-06,
+      "loss": 0.4044,
+      "step": 3738
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.707791218041923,
+      "learning_rate": 8.676290650266571e-06,
+      "loss": 0.3886,
+      "step": 3739
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.32861835116594,
+      "learning_rate": 8.67552735587494e-06,
+      "loss": 0.4326,
+      "step": 3740
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.8219942822737885,
+      "learning_rate": 8.674763875072544e-06,
+      "loss": 0.4285,
+      "step": 3741
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 10.079969473787525,
+      "learning_rate": 8.674000207898103e-06,
+      "loss": 0.4212,
+      "step": 3742
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.353351692781997,
+      "learning_rate": 8.67323635439035e-06,
+      "loss": 0.414,
+      "step": 3743
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.094882434568403,
+      "learning_rate": 8.672472314588027e-06,
+      "loss": 0.4579,
+      "step": 3744
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8190706254257925,
+      "learning_rate": 8.67170808852988e-06,
+      "loss": 0.4396,
+      "step": 3745
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.841032938097722,
+      "learning_rate": 8.670943676254669e-06,
+      "loss": 0.3604,
+      "step": 3746
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.328883078698794,
+      "learning_rate": 8.67017907780116e-06,
+      "loss": 0.4096,
+      "step": 3747
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.3038467263931235,
+      "learning_rate": 8.669414293208136e-06,
+      "loss": 0.4087,
+      "step": 3748
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.792683916172212,
+      "learning_rate": 8.668649322514382e-06,
+      "loss": 0.4337,
+      "step": 3749
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.401417755239624,
+      "learning_rate": 8.667884165758695e-06,
+      "loss": 0.417,
+      "step": 3750
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.430761619413625,
+      "learning_rate": 8.667118822979881e-06,
+      "loss": 0.3963,
+      "step": 3751
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.127306584232546,
+      "learning_rate": 8.666353294216756e-06,
+      "loss": 0.4211,
+      "step": 3752
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.701740123472579,
+      "learning_rate": 8.665587579508145e-06,
+      "loss": 0.4068,
+      "step": 3753
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.980044771789518,
+      "learning_rate": 8.66482167889288e-06,
+      "loss": 0.4056,
+      "step": 3754
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.109582667098889,
+      "learning_rate": 8.66405559240981e-06,
+      "loss": 0.4026,
+      "step": 3755
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.18715340828828,
+      "learning_rate": 8.663289320097786e-06,
+      "loss": 0.4492,
+      "step": 3756
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.5865498026515255,
+      "learning_rate": 8.66252286199567e-06,
+      "loss": 0.419,
+      "step": 3757
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.374227645047119,
+      "learning_rate": 8.661756218142335e-06,
+      "loss": 0.4006,
+      "step": 3758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4548204930227124,
+      "learning_rate": 8.660989388576663e-06,
+      "loss": 0.2147,
+      "step": 3759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.283197397286836,
+      "learning_rate": 8.660222373337543e-06,
+      "loss": 0.4043,
+      "step": 3760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.626319186439722,
+      "learning_rate": 8.65945517246388e-06,
+      "loss": 0.4227,
+      "step": 3761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 13.602693814940988,
+      "learning_rate": 8.658687785994579e-06,
+      "loss": 0.3905,
+      "step": 3762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.826614088953525,
+      "learning_rate": 8.657920213968563e-06,
+      "loss": 0.441,
+      "step": 3763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.0121536949382985,
+      "learning_rate": 8.657152456424758e-06,
+      "loss": 0.4316,
+      "step": 3764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.149641450238995,
+      "learning_rate": 8.656384513402104e-06,
+      "loss": 0.3791,
+      "step": 3765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.40799535161194,
+      "learning_rate": 8.655616384939547e-06,
+      "loss": 0.4182,
+      "step": 3766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.233031238048907,
+      "learning_rate": 8.654848071076048e-06,
+      "loss": 0.4068,
+      "step": 3767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.485522811371648,
+      "learning_rate": 8.654079571850567e-06,
+      "loss": 0.3985,
+      "step": 3768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.399975965237251,
+      "learning_rate": 8.653310887302084e-06,
+      "loss": 0.4175,
+      "step": 3769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.614380508095142,
+      "learning_rate": 8.652542017469582e-06,
+      "loss": 0.4699,
+      "step": 3770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.112938331098477,
+      "learning_rate": 8.65177296239206e-06,
+      "loss": 0.4302,
+      "step": 3771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.010551672020339,
+      "learning_rate": 8.651003722108517e-06,
+      "loss": 0.4291,
+      "step": 3772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.8354445444681056,
+      "learning_rate": 8.650234296657969e-06,
+      "loss": 0.4205,
+      "step": 3773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.618292754255814,
+      "learning_rate": 8.649464686079436e-06,
+      "loss": 0.4307,
+      "step": 3774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.686392905743196,
+      "learning_rate": 8.648694890411953e-06,
+      "loss": 0.3869,
+      "step": 3775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.760076458103716,
+      "learning_rate": 8.647924909694562e-06,
+      "loss": 0.4454,
+      "step": 3776
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.043905293493525,
+      "learning_rate": 8.64715474396631e-06,
+      "loss": 0.4117,
+      "step": 3777
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.5735260252909247,
+      "learning_rate": 8.646384393266263e-06,
+      "loss": 0.382,
+      "step": 3778
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.6657072688869974,
+      "learning_rate": 8.645613857633486e-06,
+      "loss": 0.378,
+      "step": 3779
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.338904981148392,
+      "learning_rate": 8.644843137107058e-06,
+      "loss": 0.3767,
+      "step": 3780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.921513665590369,
+      "learning_rate": 8.64407223172607e-06,
+      "loss": 0.3998,
+      "step": 3781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.25349783094434,
+      "learning_rate": 8.643301141529619e-06,
+      "loss": 0.4373,
+      "step": 3782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.150245032994205,
+      "learning_rate": 8.642529866556813e-06,
+      "loss": 0.421,
+      "step": 3783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.194929717389148,
+      "learning_rate": 8.641758406846765e-06,
+      "loss": 0.4587,
+      "step": 3784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.863150372836321,
+      "learning_rate": 8.640986762438606e-06,
+      "loss": 0.4159,
+      "step": 3785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.387339035245621,
+      "learning_rate": 8.640214933371467e-06,
+      "loss": 0.4459,
+      "step": 3786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.9886037216872463,
+      "learning_rate": 8.639442919684495e-06,
+      "loss": 0.3772,
+      "step": 3787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.861040256613024,
+      "learning_rate": 8.638670721416844e-06,
+      "loss": 0.4486,
+      "step": 3788
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.053382014519526,
+      "learning_rate": 8.637898338607675e-06,
+      "loss": 0.3779,
+      "step": 3789
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.618115361420994,
+      "learning_rate": 8.637125771296164e-06,
+      "loss": 0.3923,
+      "step": 3790
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.086249778204319,
+      "learning_rate": 8.636353019521491e-06,
+      "loss": 0.4348,
+      "step": 3791
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.781105462869253,
+      "learning_rate": 8.635580083322847e-06,
+      "loss": 0.479,
+      "step": 3792
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 21.02064180747512,
+      "learning_rate": 8.634806962739436e-06,
+      "loss": 0.4411,
+      "step": 3793
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.136698389159077,
+      "learning_rate": 8.634033657810465e-06,
+      "loss": 0.4075,
+      "step": 3794
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4711647077187999,
+      "learning_rate": 8.633260168575155e-06,
+      "loss": 0.2186,
+      "step": 3795
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 12.444175760135666,
+      "learning_rate": 8.632486495072734e-06,
+      "loss": 0.3681,
+      "step": 3796
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 11.442015997000018,
+      "learning_rate": 8.63171263734244e-06,
+      "loss": 0.4512,
+      "step": 3797
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.151850225016953,
+      "learning_rate": 8.630938595423521e-06,
+      "loss": 0.433,
+      "step": 3798
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.85749882239632,
+      "learning_rate": 8.630164369355234e-06,
+      "loss": 0.3849,
+      "step": 3799
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.523331505951532,
+      "learning_rate": 8.629389959176846e-06,
+      "loss": 0.3677,
+      "step": 3800
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 8.368921545629892,
+      "learning_rate": 8.628615364927632e-06,
+      "loss": 0.424,
+      "step": 3801
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.529010383369329,
+      "learning_rate": 8.627840586646876e-06,
+      "loss": 0.4209,
+      "step": 3802
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.939698034697384,
+      "learning_rate": 8.627065624373872e-06,
+      "loss": 0.3552,
+      "step": 3803
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.7430702693916356,
+      "learning_rate": 8.626290478147925e-06,
+      "loss": 0.4106,
+      "step": 3804
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.881451867940645,
+      "learning_rate": 8.625515148008349e-06,
+      "loss": 0.4523,
+      "step": 3805
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.962458383707436,
+      "learning_rate": 8.624739633994463e-06,
+      "loss": 0.4101,
+      "step": 3806
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 5.510589392742145,
+      "learning_rate": 8.6239639361456e-06,
+      "loss": 0.4242,
+      "step": 3807
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.52747361611853,
+      "learning_rate": 8.623188054501101e-06,
+      "loss": 0.4122,
+      "step": 3808
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.447125371674564,
+      "learning_rate": 8.622411989100316e-06,
+      "loss": 0.408,
+      "step": 3809
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5234773627695464,
+      "learning_rate": 8.621635739982607e-06,
+      "loss": 0.2057,
+      "step": 3810
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 7.498215386357655,
+      "learning_rate": 8.620859307187339e-06,
+      "loss": 0.4127,
+      "step": 3811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 30.38026146315577,
+      "learning_rate": 8.620082690753891e-06,
+      "loss": 0.4275,
+      "step": 3812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.683065432986237,
+      "learning_rate": 8.61930589072165e-06,
+      "loss": 0.3964,
+      "step": 3813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.889148263652697,
+      "learning_rate": 8.618528907130017e-06,
+      "loss": 0.3741,
+      "step": 3814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.30454310591247,
+      "learning_rate": 8.617751740018391e-06,
+      "loss": 0.4244,
+      "step": 3815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.650325937597103,
+      "learning_rate": 8.616974389426194e-06,
+      "loss": 0.4285,
+      "step": 3816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.424197533183545,
+      "learning_rate": 8.616196855392849e-06,
+      "loss": 0.3947,
+      "step": 3817
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.132948665993321,
+      "learning_rate": 8.615419137957785e-06,
+      "loss": 0.4122,
+      "step": 3818
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.370879419051802,
+      "learning_rate": 8.614641237160453e-06,
+      "loss": 0.4116,
+      "step": 3819
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.101018958700343,
+      "learning_rate": 8.613863153040301e-06,
+      "loss": 0.3966,
+      "step": 3820
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.407821183748925,
+      "learning_rate": 8.61308488563679e-06,
+      "loss": 0.4004,
+      "step": 3821
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.433510384291052,
+      "learning_rate": 8.612306434989395e-06,
+      "loss": 0.4069,
+      "step": 3822
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.641849348103855,
+      "learning_rate": 8.611527801137593e-06,
+      "loss": 0.4169,
+      "step": 3823
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42867811231276115,
+      "learning_rate": 8.610748984120876e-06,
+      "loss": 0.2002,
+      "step": 3824
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.207904241460617,
+      "learning_rate": 8.609969983978741e-06,
+      "loss": 0.4465,
+      "step": 3825
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.821817008104182,
+      "learning_rate": 8.6091908007507e-06,
+      "loss": 0.456,
+      "step": 3826
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.0118853838717605,
+      "learning_rate": 8.608411434476267e-06,
+      "loss": 0.3878,
+      "step": 3827
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.686629476629486,
+      "learning_rate": 8.60763188519497e-06,
+      "loss": 0.4054,
+      "step": 3828
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38923892527340265,
+      "learning_rate": 8.606852152946345e-06,
+      "loss": 0.2094,
+      "step": 3829
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.4116431264347025,
+      "learning_rate": 8.606072237769938e-06,
+      "loss": 0.4384,
+      "step": 3830
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.35480714742963254,
+      "learning_rate": 8.605292139705303e-06,
+      "loss": 0.2019,
+      "step": 3831
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.982033088282748,
+      "learning_rate": 8.604511858792006e-06,
+      "loss": 0.4368,
+      "step": 3832
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.565475001081145,
+      "learning_rate": 8.603731395069618e-06,
+      "loss": 0.4214,
+      "step": 3833
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.383781257859957,
+      "learning_rate": 8.602950748577722e-06,
+      "loss": 0.4253,
+      "step": 3834
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.405047504561646,
+      "learning_rate": 8.60216991935591e-06,
+      "loss": 0.4343,
+      "step": 3835
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.028935952293998,
+      "learning_rate": 8.601388907443787e-06,
+      "loss": 0.4313,
+      "step": 3836
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6782040028035894,
+      "learning_rate": 8.600607712880956e-06,
+      "loss": 0.4303,
+      "step": 3837
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.11510924310188,
+      "learning_rate": 8.599826335707043e-06,
+      "loss": 0.4003,
+      "step": 3838
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.641377062594978,
+      "learning_rate": 8.599044775961671e-06,
+      "loss": 0.4425,
+      "step": 3839
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.905551191478285,
+      "learning_rate": 8.598263033684485e-06,
+      "loss": 0.4013,
+      "step": 3840
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.907513873698513,
+      "learning_rate": 8.597481108915127e-06,
+      "loss": 0.4087,
+      "step": 3841
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.8163415124322295,
+      "learning_rate": 8.596699001693257e-06,
+      "loss": 0.4341,
+      "step": 3842
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.529466485160263,
+      "learning_rate": 8.595916712058539e-06,
+      "loss": 0.4319,
+      "step": 3843
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43786198192251347,
+      "learning_rate": 8.595134240050648e-06,
+      "loss": 0.2034,
+      "step": 3844
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.736058269755547,
+      "learning_rate": 8.59435158570927e-06,
+      "loss": 0.4222,
+      "step": 3845
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 18.964190556170433,
+      "learning_rate": 8.593568749074098e-06,
+      "loss": 0.4132,
+      "step": 3846
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.793097533185397,
+      "learning_rate": 8.592785730184833e-06,
+      "loss": 0.4286,
+      "step": 3847
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.77173691945276,
+      "learning_rate": 8.59200252908119e-06,
+      "loss": 0.3912,
+      "step": 3848
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.418322705657429,
+      "learning_rate": 8.591219145802888e-06,
+      "loss": 0.4468,
+      "step": 3849
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.594515501935546,
+      "learning_rate": 8.590435580389663e-06,
+      "loss": 0.4346,
+      "step": 3850
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.18305698068446,
+      "learning_rate": 8.589651832881247e-06,
+      "loss": 0.4245,
+      "step": 3851
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.265898856497643,
+      "learning_rate": 8.588867903317395e-06,
+      "loss": 0.4075,
+      "step": 3852
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.340017655476044,
+      "learning_rate": 8.588083791737862e-06,
+      "loss": 0.4257,
+      "step": 3853
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.099700602818479,
+      "learning_rate": 8.587299498182418e-06,
+      "loss": 0.423,
+      "step": 3854
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.43459473034324875,
+      "learning_rate": 8.586515022690838e-06,
+      "loss": 0.2013,
+      "step": 3855
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.38872303692960347,
+      "learning_rate": 8.58573036530291e-06,
+      "loss": 0.2065,
+      "step": 3856
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7985424105545578,
+      "learning_rate": 8.584945526058426e-06,
+      "loss": 0.4247,
+      "step": 3857
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.063277523226818,
+      "learning_rate": 8.584160504997191e-06,
+      "loss": 0.4362,
+      "step": 3858
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.017033716585484,
+      "learning_rate": 8.583375302159026e-06,
+      "loss": 0.4168,
+      "step": 3859
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.568887619934926,
+      "learning_rate": 8.582589917583744e-06,
+      "loss": 0.4192,
+      "step": 3860
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.42147795846307134,
+      "learning_rate": 8.581804351311181e-06,
+      "loss": 0.2027,
+      "step": 3861
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.890366908385326,
+      "learning_rate": 8.581018603381181e-06,
+      "loss": 0.4082,
+      "step": 3862
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.726107659314696,
+      "learning_rate": 8.58023267383359e-06,
+      "loss": 0.409,
+      "step": 3863
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.691799133655408,
+      "learning_rate": 8.579446562708272e-06,
+      "loss": 0.4376,
+      "step": 3864
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 20.343543074607762,
+      "learning_rate": 8.578660270045092e-06,
+      "loss": 0.3834,
+      "step": 3865
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.3042107376873995,
+      "learning_rate": 8.57787379588393e-06,
+      "loss": 0.4008,
+      "step": 3866
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.529668776116852,
+      "learning_rate": 8.577087140264677e-06,
+      "loss": 0.3801,
+      "step": 3867
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.801712557329872,
+      "learning_rate": 8.576300303227224e-06,
+      "loss": 0.4366,
+      "step": 3868
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.200330883025147,
+      "learning_rate": 8.57551328481148e-06,
+      "loss": 0.4004,
+      "step": 3869
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.3045984787611262,
+      "learning_rate": 8.574726085057357e-06,
+      "loss": 0.4162,
+      "step": 3870
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.581257552793236,
+      "learning_rate": 8.573938704004783e-06,
+      "loss": 0.4332,
+      "step": 3871
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 13.730516162779534,
+      "learning_rate": 8.573151141693688e-06,
+      "loss": 0.4021,
+      "step": 3872
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.658322560584027,
+      "learning_rate": 8.572363398164017e-06,
+      "loss": 0.4126,
+      "step": 3873
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.561981165268786,
+      "learning_rate": 8.571575473455723e-06,
+      "loss": 0.4143,
+      "step": 3874
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.268920661741892,
+      "learning_rate": 8.570787367608762e-06,
+      "loss": 0.4093,
+      "step": 3875
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.662845227895199,
+      "learning_rate": 8.569999080663108e-06,
+      "loss": 0.4313,
+      "step": 3876
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.770267878383975,
+      "learning_rate": 8.569210612658741e-06,
+      "loss": 0.4465,
+      "step": 3877
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.143290929192842,
+      "learning_rate": 8.568421963635647e-06,
+      "loss": 0.4323,
+      "step": 3878
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.530809664815183,
+      "learning_rate": 8.567633133633824e-06,
+      "loss": 0.4238,
+      "step": 3879
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.92838457413379,
+      "learning_rate": 8.56684412269328e-06,
+      "loss": 0.4362,
+      "step": 3880
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.799765657348471,
+      "learning_rate": 8.56605493085403e-06,
+      "loss": 0.4478,
+      "step": 3881
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.974689903286326,
+      "learning_rate": 8.565265558156101e-06,
+      "loss": 0.4191,
+      "step": 3882
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.373769365610645,
+      "learning_rate": 8.564476004639528e-06,
+      "loss": 0.4248,
+      "step": 3883
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.872734374084031,
+      "learning_rate": 8.56368627034435e-06,
+      "loss": 0.4205,
+      "step": 3884
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.070470004847321,
+      "learning_rate": 8.562896355310624e-06,
+      "loss": 0.4136,
+      "step": 3885
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.131022749443039,
+      "learning_rate": 8.56210625957841e-06,
+      "loss": 0.4333,
+      "step": 3886
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.922579261504207,
+      "learning_rate": 8.561315983187781e-06,
+      "loss": 0.4227,
+      "step": 3887
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.245102776151354,
+      "learning_rate": 8.560525526178814e-06,
+      "loss": 0.4043,
+      "step": 3888
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.509308126494464,
+      "learning_rate": 8.559734888591602e-06,
+      "loss": 0.4211,
+      "step": 3889
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.473631317265875,
+      "learning_rate": 8.558944070466242e-06,
+      "loss": 0.4036,
+      "step": 3890
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.414401978499358,
+      "learning_rate": 8.55815307184284e-06,
+      "loss": 0.3986,
+      "step": 3891
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.0373243533261673,
+      "learning_rate": 8.557361892761514e-06,
+      "loss": 0.4175,
+      "step": 3892
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.398629141405738,
+      "learning_rate": 8.556570533262392e-06,
+      "loss": 0.4365,
+      "step": 3893
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.557923909358993,
+      "learning_rate": 8.555778993385607e-06,
+      "loss": 0.433,
+      "step": 3894
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.6577531948827122,
+      "learning_rate": 8.554987273171305e-06,
+      "loss": 0.4141,
+      "step": 3895
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 11.45838520841045,
+      "learning_rate": 8.554195372659638e-06,
+      "loss": 0.402,
+      "step": 3896
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.0665036297049175,
+      "learning_rate": 8.553403291890767e-06,
+      "loss": 0.438,
+      "step": 3897
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.321468148379923,
+      "learning_rate": 8.552611030904866e-06,
+      "loss": 0.3932,
+      "step": 3898
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 10.369148530058752,
+      "learning_rate": 8.551818589742118e-06,
+      "loss": 0.4719,
+      "step": 3899
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.8645440257702304,
+      "learning_rate": 8.55102596844271e-06,
+      "loss": 0.4184,
+      "step": 3900
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.300429930167817,
+      "learning_rate": 8.550233167046842e-06,
+      "loss": 0.4328,
+      "step": 3901
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.38120769110144,
+      "learning_rate": 8.549440185594722e-06,
+      "loss": 0.413,
+      "step": 3902
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.763814667870659,
+      "learning_rate": 8.548647024126568e-06,
+      "loss": 0.4574,
+      "step": 3903
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.221370002206253,
+      "learning_rate": 8.547853682682605e-06,
+      "loss": 0.4279,
+      "step": 3904
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.23856033221092,
+      "learning_rate": 8.547060161303071e-06,
+      "loss": 0.4396,
+      "step": 3905
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.017573647931714,
+      "learning_rate": 8.54626646002821e-06,
+      "loss": 0.4242,
+      "step": 3906
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.365713404862079,
+      "learning_rate": 8.545472578898276e-06,
+      "loss": 0.4102,
+      "step": 3907
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.799653900232872,
+      "learning_rate": 8.544678517953528e-06,
+      "loss": 0.4065,
+      "step": 3908
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.786628596055821,
+      "learning_rate": 8.543884277234245e-06,
+      "loss": 0.3746,
+      "step": 3909
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.629164193169609,
+      "learning_rate": 8.543089856780704e-06,
+      "loss": 0.4048,
+      "step": 3910
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.533025360340257,
+      "learning_rate": 8.542295256633198e-06,
+      "loss": 0.4438,
+      "step": 3911
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.453771592709116,
+      "learning_rate": 8.541500476832025e-06,
+      "loss": 0.4038,
+      "step": 3912
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.732528236958803,
+      "learning_rate": 8.540705517417494e-06,
+      "loss": 0.4168,
+      "step": 3913
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.281277018804368,
+      "learning_rate": 8.539910378429921e-06,
+      "loss": 0.4352,
+      "step": 3914
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.801971960832366,
+      "learning_rate": 8.539115059909634e-06,
+      "loss": 0.4324,
+      "step": 3915
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 12.715648231781966,
+      "learning_rate": 8.53831956189697e-06,
+      "loss": 0.4351,
+      "step": 3916
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.850750866664695,
+      "learning_rate": 8.537523884432275e-06,
+      "loss": 0.4097,
+      "step": 3917
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.535737865970934,
+      "learning_rate": 8.5367280275559e-06,
+      "loss": 0.4174,
+      "step": 3918
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.350135305319124,
+      "learning_rate": 8.53593199130821e-06,
+      "loss": 0.3925,
+      "step": 3919
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.315266650029613,
+      "learning_rate": 8.535135775729578e-06,
+      "loss": 0.4202,
+      "step": 3920
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.024697078470177,
+      "learning_rate": 8.534339380860383e-06,
+      "loss": 0.3948,
+      "step": 3921
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 8.466088646541689,
+      "learning_rate": 8.53354280674102e-06,
+      "loss": 0.4203,
+      "step": 3922
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.989579751771484,
+      "learning_rate": 8.532746053411883e-06,
+      "loss": 0.3833,
+      "step": 3923
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.290452632869084,
+      "learning_rate": 8.531949120913385e-06,
+      "loss": 0.4166,
+      "step": 3924
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.306856667211173,
+      "learning_rate": 8.531152009285942e-06,
+      "loss": 0.4299,
+      "step": 3925
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.695552531157985,
+      "learning_rate": 8.530354718569982e-06,
+      "loss": 0.386,
+      "step": 3926
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.935511234952024,
+      "learning_rate": 8.52955724880594e-06,
+      "loss": 0.4086,
+      "step": 3927
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.089520403803548,
+      "learning_rate": 8.52875960003426e-06,
+      "loss": 0.4502,
+      "step": 3928
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.173296305507457,
+      "learning_rate": 8.527961772295399e-06,
+      "loss": 0.4022,
+      "step": 3929
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.147490548332593,
+      "learning_rate": 8.527163765629819e-06,
+      "loss": 0.4069,
+      "step": 3930
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.104927253143151,
+      "learning_rate": 8.526365580077991e-06,
+      "loss": 0.4096,
+      "step": 3931
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.749287440443363,
+      "learning_rate": 8.525567215680397e-06,
+      "loss": 0.4633,
+      "step": 3932
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.386456470035492,
+      "learning_rate": 8.524768672477528e-06,
+      "loss": 0.3555,
+      "step": 3933
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.97999599131255,
+      "learning_rate": 8.523969950509886e-06,
+      "loss": 0.3892,
+      "step": 3934
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.7439811014612348,
+      "learning_rate": 8.523171049817974e-06,
+      "loss": 0.4264,
+      "step": 3935
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.446495990611158,
+      "learning_rate": 8.522371970442314e-06,
+      "loss": 0.4293,
+      "step": 3936
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 9.308912592460096,
+      "learning_rate": 8.52157271242343e-06,
+      "loss": 0.4043,
+      "step": 3937
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.278905284880584,
+      "learning_rate": 8.52077327580186e-06,
+      "loss": 0.4191,
+      "step": 3938
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.914621319422669,
+      "learning_rate": 8.519973660618149e-06,
+      "loss": 0.3859,
+      "step": 3939
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.080278444118906,
+      "learning_rate": 8.519173866912847e-06,
+      "loss": 0.4132,
+      "step": 3940
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.541644071129197,
+      "learning_rate": 8.51837389472652e-06,
+      "loss": 0.4162,
+      "step": 3941
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.876980434217718,
+      "learning_rate": 8.51757374409974e-06,
+      "loss": 0.4325,
+      "step": 3942
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.118298949995971,
+      "learning_rate": 8.516773415073089e-06,
+      "loss": 0.3833,
+      "step": 3943
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.722116604659064,
+      "learning_rate": 8.515972907687155e-06,
+      "loss": 0.4064,
+      "step": 3944
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.430200945694761,
+      "learning_rate": 8.515172221982537e-06,
+      "loss": 0.4143,
+      "step": 3945
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.332844872828631,
+      "learning_rate": 8.514371357999846e-06,
+      "loss": 0.4262,
+      "step": 3946
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.644327120743076,
+      "learning_rate": 8.513570315779696e-06,
+      "loss": 0.4282,
+      "step": 3947
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.315995912885699,
+      "learning_rate": 8.512769095362714e-06,
+      "loss": 0.4254,
+      "step": 3948
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.937436871475605,
+      "learning_rate": 8.511967696789535e-06,
+      "loss": 0.4457,
+      "step": 3949
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 6.289789869322682,
+      "learning_rate": 8.511166120100806e-06,
+      "loss": 0.4338,
+      "step": 3950
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 7.508693511565535,
+      "learning_rate": 8.510364365337178e-06,
+      "loss": 0.4302,
+      "step": 3951
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.783972458418909,
+      "learning_rate": 8.509562432539313e-06,
+      "loss": 0.4143,
+      "step": 3952
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 4.361300343573623,
+      "learning_rate": 8.508760321747884e-06,
+      "loss": 0.4335,
+      "step": 3953
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 5.260116959857058,
+      "learning_rate": 8.507958033003569e-06,
+      "loss": 0.4314,
+      "step": 3954
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.742911306925295,
+      "learning_rate": 8.507155566347063e-06,
+      "loss": 0.4205,
+      "step": 3955
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.629925600632549,
+      "learning_rate": 8.506352921819057e-06,
+      "loss": 0.3729,
+      "step": 3956
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.088759141008321,
+      "learning_rate": 8.505550099460264e-06,
+      "loss": 0.4127,
+      "step": 3957
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.402628514565033,
+      "learning_rate": 8.504747099311398e-06,
+      "loss": 0.4239,
+      "step": 3958
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4738674587554486,
+      "learning_rate": 8.503943921413186e-06,
+      "loss": 0.3969,
+      "step": 3959
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.095223932521769,
+      "learning_rate": 8.50314056580636e-06,
+      "loss": 0.4221,
+      "step": 3960
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.686351474989815,
+      "learning_rate": 8.502337032531667e-06,
+      "loss": 0.3973,
+      "step": 3961
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.9048843983303243,
+      "learning_rate": 8.501533321629859e-06,
+      "loss": 0.4289,
+      "step": 3962
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.809761406917024,
+      "learning_rate": 8.500729433141694e-06,
+      "loss": 0.4611,
+      "step": 3963
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.93937508740606,
+      "learning_rate": 8.499925367107947e-06,
+      "loss": 0.4157,
+      "step": 3964
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.87450948265819,
+      "learning_rate": 8.499121123569395e-06,
+      "loss": 0.4165,
+      "step": 3965
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.4948397107391234,
+      "learning_rate": 8.498316702566828e-06,
+      "loss": 0.4347,
+      "step": 3966
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.41806252134707783,
+      "learning_rate": 8.497512104141043e-06,
+      "loss": 0.2179,
+      "step": 3967
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3733402829165685,
+      "learning_rate": 8.496707328332846e-06,
+      "loss": 0.2072,
+      "step": 3968
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.317863686924785,
+      "learning_rate": 8.495902375183054e-06,
+      "loss": 0.4527,
+      "step": 3969
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.15837842217616,
+      "learning_rate": 8.495097244732489e-06,
+      "loss": 0.4471,
+      "step": 3970
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.454195385110939,
+      "learning_rate": 8.494291937021988e-06,
+      "loss": 0.4161,
+      "step": 3971
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.479977570746793,
+      "learning_rate": 8.493486452092391e-06,
+      "loss": 0.4196,
+      "step": 3972
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.0460701145109588,
+      "learning_rate": 8.492680789984553e-06,
+      "loss": 0.4114,
+      "step": 3973
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.970361043553756,
+      "learning_rate": 8.49187495073933e-06,
+      "loss": 0.4145,
+      "step": 3974
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.148737829648252,
+      "learning_rate": 8.491068934397596e-06,
+      "loss": 0.4256,
+      "step": 3975
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9773439208288104,
+      "learning_rate": 8.490262741000223e-06,
+      "loss": 0.4111,
+      "step": 3976
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.292620568973986,
+      "learning_rate": 8.489456370588106e-06,
+      "loss": 0.4066,
+      "step": 3977
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9157799017613164,
+      "learning_rate": 8.488649823202137e-06,
+      "loss": 0.4195,
+      "step": 3978
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4444745208464678,
+      "learning_rate": 8.487843098883222e-06,
+      "loss": 0.1952,
+      "step": 3979
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.181594315938987,
+      "learning_rate": 8.487036197672277e-06,
+      "loss": 0.4286,
+      "step": 3980
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3791953128815164,
+      "learning_rate": 8.486229119610225e-06,
+      "loss": 0.1932,
+      "step": 3981
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.433765415029255,
+      "learning_rate": 8.485421864737997e-06,
+      "loss": 0.4393,
+      "step": 3982
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.432687732274011,
+      "learning_rate": 8.484614433096535e-06,
+      "loss": 0.4328,
+      "step": 3983
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.766697005447135,
+      "learning_rate": 8.483806824726788e-06,
+      "loss": 0.4684,
+      "step": 3984
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.233942763147831,
+      "learning_rate": 8.48299903966972e-06,
+      "loss": 0.4465,
+      "step": 3985
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.4479635223007685,
+      "learning_rate": 8.482191077966293e-06,
+      "loss": 0.4049,
+      "step": 3986
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4383684392604411,
+      "learning_rate": 8.48138293965749e-06,
+      "loss": 0.2048,
+      "step": 3987
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.8975233389272885,
+      "learning_rate": 8.480574624784292e-06,
+      "loss": 0.4676,
+      "step": 3988
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3326735038651543,
+      "learning_rate": 8.479766133387698e-06,
+      "loss": 0.4032,
+      "step": 3989
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.720819473975927,
+      "learning_rate": 8.478957465508708e-06,
+      "loss": 0.4092,
+      "step": 3990
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4163189484751042,
+      "learning_rate": 8.47814862118834e-06,
+      "loss": 0.2082,
+      "step": 3991
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8195344253305823,
+      "learning_rate": 8.477339600467612e-06,
+      "loss": 0.4192,
+      "step": 3992
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.1698369274598335,
+      "learning_rate": 8.476530403387557e-06,
+      "loss": 0.3888,
+      "step": 3993
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.836042114932037,
+      "learning_rate": 8.475721029989212e-06,
+      "loss": 0.4217,
+      "step": 3994
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 28.61974736778765,
+      "learning_rate": 8.47491148031363e-06,
+      "loss": 0.4288,
+      "step": 3995
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1368496218799065,
+      "learning_rate": 8.474101754401866e-06,
+      "loss": 0.3591,
+      "step": 3996
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.9266806718263805,
+      "learning_rate": 8.473291852294986e-06,
+      "loss": 0.4227,
+      "step": 3997
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.589895583779776,
+      "learning_rate": 8.47248177403407e-06,
+      "loss": 0.4145,
+      "step": 3998
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6635154904008385,
+      "learning_rate": 8.471671519660196e-06,
+      "loss": 0.4067,
+      "step": 3999
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3707301553347317,
+      "learning_rate": 8.470861089214464e-06,
+      "loss": 0.4142,
+      "step": 4000
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.720525699305355,
+      "learning_rate": 8.47005048273797e-06,
+      "loss": 0.4248,
+      "step": 4001
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.3004429773018362,
+      "learning_rate": 8.46923970027183e-06,
+      "loss": 0.4027,
+      "step": 4002
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.000103579165123,
+      "learning_rate": 8.468428741857164e-06,
+      "loss": 0.4534,
+      "step": 4003
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.505025414088027,
+      "learning_rate": 8.467617607535096e-06,
+      "loss": 0.4194,
+      "step": 4004
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.94192937754645,
+      "learning_rate": 8.466806297346772e-06,
+      "loss": 0.4401,
+      "step": 4005
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.031811552552276,
+      "learning_rate": 8.465994811333334e-06,
+      "loss": 0.4443,
+      "step": 4006
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.816866950288662,
+      "learning_rate": 8.465183149535939e-06,
+      "loss": 0.3894,
+      "step": 4007
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.579022715008342,
+      "learning_rate": 8.464371311995752e-06,
+      "loss": 0.4391,
+      "step": 4008
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.375845390753202,
+      "learning_rate": 8.463559298753946e-06,
+      "loss": 0.4066,
+      "step": 4009
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.232641827366635,
+      "learning_rate": 8.462747109851703e-06,
+      "loss": 0.4255,
+      "step": 4010
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8488555628969987,
+      "learning_rate": 8.461934745330219e-06,
+      "loss": 0.4093,
+      "step": 4011
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.161125698116083,
+      "learning_rate": 8.461122205230689e-06,
+      "loss": 0.404,
+      "step": 4012
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.8945957761265038,
+      "learning_rate": 8.460309489594324e-06,
+      "loss": 0.4464,
+      "step": 4013
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.107711494910836,
+      "learning_rate": 8.459496598462345e-06,
+      "loss": 0.3849,
+      "step": 4014
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.287266916544053,
+      "learning_rate": 8.458683531875975e-06,
+      "loss": 0.4336,
+      "step": 4015
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.6185850236538295,
+      "learning_rate": 8.457870289876454e-06,
+      "loss": 0.4099,
+      "step": 4016
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5747997198497643,
+      "learning_rate": 8.457056872505024e-06,
+      "loss": 0.402,
+      "step": 4017
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.5406972892974045,
+      "learning_rate": 8.456243279802941e-06,
+      "loss": 0.4135,
+      "step": 4018
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.217116492583587,
+      "learning_rate": 8.455429511811466e-06,
+      "loss": 0.3757,
+      "step": 4019
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.329155800658797,
+      "learning_rate": 8.454615568571872e-06,
+      "loss": 0.4254,
+      "step": 4020
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9885169582735505,
+      "learning_rate": 8.453801450125439e-06,
+      "loss": 0.4033,
+      "step": 4021
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.544517455046838,
+      "learning_rate": 8.452987156513457e-06,
+      "loss": 0.4144,
+      "step": 4022
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.689681997780447,
+      "learning_rate": 8.452172687777223e-06,
+      "loss": 0.3535,
+      "step": 4023
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.086662214545773,
+      "learning_rate": 8.451358043958046e-06,
+      "loss": 0.4124,
+      "step": 4024
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.221859562894499,
+      "learning_rate": 8.45054322509724e-06,
+      "loss": 0.427,
+      "step": 4025
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.300631206703308,
+      "learning_rate": 8.449728231236132e-06,
+      "loss": 0.4264,
+      "step": 4026
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.283844661560317,
+      "learning_rate": 8.448913062416054e-06,
+      "loss": 0.4071,
+      "step": 4027
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.745772697588965,
+      "learning_rate": 8.44809771867835e-06,
+      "loss": 0.415,
+      "step": 4028
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.708567399118609,
+      "learning_rate": 8.447282200064371e-06,
+      "loss": 0.4153,
+      "step": 4029
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.695557055564441,
+      "learning_rate": 8.446466506615478e-06,
+      "loss": 0.4264,
+      "step": 4030
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.3073490486279296,
+      "learning_rate": 8.445650638373038e-06,
+      "loss": 0.4085,
+      "step": 4031
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.092896484238077,
+      "learning_rate": 8.444834595378434e-06,
+      "loss": 0.435,
+      "step": 4032
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.463465156694404,
+      "learning_rate": 8.444018377673048e-06,
+      "loss": 0.4439,
+      "step": 4033
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.129131842560376,
+      "learning_rate": 8.443201985298279e-06,
+      "loss": 0.3874,
+      "step": 4034
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.764378504948511,
+      "learning_rate": 8.442385418295532e-06,
+      "loss": 0.4084,
+      "step": 4035
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.530580243194052,
+      "learning_rate": 8.441568676706217e-06,
+      "loss": 0.42,
+      "step": 4036
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.797282412861643,
+      "learning_rate": 8.44075176057176e-06,
+      "loss": 0.3911,
+      "step": 4037
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.073213696700629,
+      "learning_rate": 8.439934669933591e-06,
+      "loss": 0.4143,
+      "step": 4038
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.456746757736086,
+      "learning_rate": 8.439117404833151e-06,
+      "loss": 0.4564,
+      "step": 4039
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.19852661286054,
+      "learning_rate": 8.438299965311887e-06,
+      "loss": 0.3997,
+      "step": 4040
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.276180546930875,
+      "learning_rate": 8.437482351411258e-06,
+      "loss": 0.4263,
+      "step": 4041
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.45696495322020253,
+      "learning_rate": 8.436664563172733e-06,
+      "loss": 0.2122,
+      "step": 4042
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.766144146583094,
+      "learning_rate": 8.435846600637786e-06,
+      "loss": 0.4056,
+      "step": 4043
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.014658454610659,
+      "learning_rate": 8.435028463847898e-06,
+      "loss": 0.3813,
+      "step": 4044
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.817795148668165,
+      "learning_rate": 8.434210152844566e-06,
+      "loss": 0.4254,
+      "step": 4045
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.095823788500936,
+      "learning_rate": 8.433391667669292e-06,
+      "loss": 0.4663,
+      "step": 4046
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9791175941462287,
+      "learning_rate": 8.432573008363587e-06,
+      "loss": 0.4099,
+      "step": 4047
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.40719716344499,
+      "learning_rate": 8.431754174968969e-06,
+      "loss": 0.388,
+      "step": 4048
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.417407400795511,
+      "learning_rate": 8.430935167526967e-06,
+      "loss": 0.4083,
+      "step": 4049
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.089441594213951,
+      "learning_rate": 8.43011598607912e-06,
+      "loss": 0.4385,
+      "step": 4050
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.155701183129664,
+      "learning_rate": 8.429296630666971e-06,
+      "loss": 0.4563,
+      "step": 4051
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.782353955397653,
+      "learning_rate": 8.428477101332079e-06,
+      "loss": 0.4395,
+      "step": 4052
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 12.270152607174534,
+      "learning_rate": 8.427657398116005e-06,
+      "loss": 0.4162,
+      "step": 4053
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.169787079768914,
+      "learning_rate": 8.426837521060322e-06,
+      "loss": 0.421,
+      "step": 4054
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.201655068244703,
+      "learning_rate": 8.426017470206612e-06,
+      "loss": 0.3863,
+      "step": 4055
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.107132765710986,
+      "learning_rate": 8.425197245596467e-06,
+      "loss": 0.4168,
+      "step": 4056
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 29.77802002950844,
+      "learning_rate": 8.424376847271483e-06,
+      "loss": 0.4211,
+      "step": 4057
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.956631877924124,
+      "learning_rate": 8.42355627527327e-06,
+      "loss": 0.4159,
+      "step": 4058
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.40023890469281,
+      "learning_rate": 8.422735529643445e-06,
+      "loss": 0.4114,
+      "step": 4059
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.7661425373508175,
+      "learning_rate": 8.421914610423631e-06,
+      "loss": 0.3945,
+      "step": 4060
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.26324072822136,
+      "learning_rate": 8.421093517655465e-06,
+      "loss": 0.4533,
+      "step": 4061
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.451676990552305,
+      "learning_rate": 8.42027225138059e-06,
+      "loss": 0.4198,
+      "step": 4062
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.617084733312273,
+      "learning_rate": 8.419450811640654e-06,
+      "loss": 0.443,
+      "step": 4063
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.661280232535554,
+      "learning_rate": 8.418629198477325e-06,
+      "loss": 0.4159,
+      "step": 4064
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.167126172217376,
+      "learning_rate": 8.417807411932265e-06,
+      "loss": 0.4079,
+      "step": 4065
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.2843545425769265,
+      "learning_rate": 8.416985452047157e-06,
+      "loss": 0.4462,
+      "step": 4066
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.30414914537432,
+      "learning_rate": 8.416163318863687e-06,
+      "loss": 0.4456,
+      "step": 4067
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.727451571303054,
+      "learning_rate": 8.415341012423552e-06,
+      "loss": 0.4087,
+      "step": 4068
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.179345192070316,
+      "learning_rate": 8.414518532768455e-06,
+      "loss": 0.4651,
+      "step": 4069
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.592641603412242,
+      "learning_rate": 8.41369587994011e-06,
+      "loss": 0.4254,
+      "step": 4070
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.515897117814267,
+      "learning_rate": 8.412873053980237e-06,
+      "loss": 0.4149,
+      "step": 4071
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.121065396659326,
+      "learning_rate": 8.412050054930571e-06,
+      "loss": 0.415,
+      "step": 4072
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.356544888757852,
+      "learning_rate": 8.41122688283285e-06,
+      "loss": 0.4313,
+      "step": 4073
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.866560291777348,
+      "learning_rate": 8.410403537728824e-06,
+      "loss": 0.4319,
+      "step": 4074
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.720915475774998,
+      "learning_rate": 8.409580019660246e-06,
+      "loss": 0.4353,
+      "step": 4075
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 13.684143139428022,
+      "learning_rate": 8.408756328668888e-06,
+      "loss": 0.4316,
+      "step": 4076
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.09642579925916,
+      "learning_rate": 8.407932464796521e-06,
+      "loss": 0.4329,
+      "step": 4077
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.396155365474958,
+      "learning_rate": 8.407108428084932e-06,
+      "loss": 0.4266,
+      "step": 4078
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.375630780682022,
+      "learning_rate": 8.40628421857591e-06,
+      "loss": 0.4425,
+      "step": 4079
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.632137922757318,
+      "learning_rate": 8.405459836311256e-06,
+      "loss": 0.4166,
+      "step": 4080
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.466719076389277,
+      "learning_rate": 8.404635281332783e-06,
+      "loss": 0.4182,
+      "step": 4081
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.053377173448509,
+      "learning_rate": 8.403810553682307e-06,
+      "loss": 0.4122,
+      "step": 4082
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.755473082968509,
+      "learning_rate": 8.402985653401659e-06,
+      "loss": 0.4399,
+      "step": 4083
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.6829856629257973,
+      "learning_rate": 8.40216058053267e-06,
+      "loss": 0.4438,
+      "step": 4084
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 7.031005843070603,
+      "learning_rate": 8.40133533511719e-06,
+      "loss": 0.4155,
+      "step": 4085
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.3438059571121626,
+      "learning_rate": 8.40050991719707e-06,
+      "loss": 0.3976,
+      "step": 4086
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.5726505574734344,
+      "learning_rate": 8.399684326814174e-06,
+      "loss": 0.4409,
+      "step": 4087
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1642261909042495,
+      "learning_rate": 8.398858564010373e-06,
+      "loss": 0.3955,
+      "step": 4088
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.982679911084919,
+      "learning_rate": 8.398032628827547e-06,
+      "loss": 0.4716,
+      "step": 4089
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.9095047383144697,
+      "learning_rate": 8.397206521307584e-06,
+      "loss": 0.4296,
+      "step": 4090
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.037905216935888,
+      "learning_rate": 8.396380241492382e-06,
+      "loss": 0.4099,
+      "step": 4091
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 10.050957191208445,
+      "learning_rate": 8.395553789423844e-06,
+      "loss": 0.4189,
+      "step": 4092
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.133121055036233,
+      "learning_rate": 8.39472716514389e-06,
+      "loss": 0.42,
+      "step": 4093
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.503006688416452,
+      "learning_rate": 8.393900368694442e-06,
+      "loss": 0.4234,
+      "step": 4094
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.54088185566446,
+      "learning_rate": 8.393073400117433e-06,
+      "loss": 0.4346,
+      "step": 4095
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.477404325559722,
+      "learning_rate": 8.392246259454801e-06,
+      "loss": 0.413,
+      "step": 4096
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.643738758132492,
+      "learning_rate": 8.3914189467485e-06,
+      "loss": 0.4053,
+      "step": 4097
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 5.104932287680142,
+      "learning_rate": 8.390591462040485e-06,
+      "loss": 0.4009,
+      "step": 4098
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.983387084696739,
+      "learning_rate": 8.389763805372725e-06,
+      "loss": 0.4182,
+      "step": 4099
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.013412923963579,
+      "learning_rate": 8.388935976787196e-06,
+      "loss": 0.4322,
+      "step": 4100
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.748380749654933,
+      "learning_rate": 8.388107976325885e-06,
+      "loss": 0.4553,
+      "step": 4101
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.0107896445942375,
+      "learning_rate": 8.387279804030781e-06,
+      "loss": 0.4305,
+      "step": 4102
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.91613339861678,
+      "learning_rate": 8.386451459943889e-06,
+      "loss": 0.422,
+      "step": 4103
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.75744359984365,
+      "learning_rate": 8.385622944107221e-06,
+      "loss": 0.4145,
+      "step": 4104
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.320916503247258,
+      "learning_rate": 8.384794256562793e-06,
+      "loss": 0.3947,
+      "step": 4105
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.934171651764611,
+      "learning_rate": 8.383965397352636e-06,
+      "loss": 0.4312,
+      "step": 4106
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.5358439667542436,
+      "learning_rate": 8.383136366518788e-06,
+      "loss": 0.3982,
+      "step": 4107
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.113544148719433,
+      "learning_rate": 8.382307164103291e-06,
+      "loss": 0.3988,
+      "step": 4108
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.369129019901582,
+      "learning_rate": 8.381477790148203e-06,
+      "loss": 0.4089,
+      "step": 4109
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 17.41823917514347,
+      "learning_rate": 8.380648244695584e-06,
+      "loss": 0.4203,
+      "step": 4110
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.65352961384922,
+      "learning_rate": 8.379818527787508e-06,
+      "loss": 0.4373,
+      "step": 4111
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.548502784826743,
+      "learning_rate": 8.378988639466058e-06,
+      "loss": 0.4463,
+      "step": 4112
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.807436142466519,
+      "learning_rate": 8.378158579773318e-06,
+      "loss": 0.4282,
+      "step": 4113
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.144553900030665,
+      "learning_rate": 8.377328348751388e-06,
+      "loss": 0.3761,
+      "step": 4114
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.58680390007866,
+      "learning_rate": 8.376497946442374e-06,
+      "loss": 0.4352,
+      "step": 4115
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.100736587950279,
+      "learning_rate": 8.375667372888394e-06,
+      "loss": 0.4505,
+      "step": 4116
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.521188504042746,
+      "learning_rate": 8.374836628131571e-06,
+      "loss": 0.4441,
+      "step": 4117
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.830438180978835,
+      "learning_rate": 8.374005712214035e-06,
+      "loss": 0.3776,
+      "step": 4118
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.939905207411626,
+      "learning_rate": 8.373174625177929e-06,
+      "loss": 0.4419,
+      "step": 4119
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.190703154448583,
+      "learning_rate": 8.372343367065404e-06,
+      "loss": 0.4282,
+      "step": 4120
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.2685263907975886,
+      "learning_rate": 8.371511937918616e-06,
+      "loss": 0.4516,
+      "step": 4121
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.049102755166142,
+      "learning_rate": 8.370680337779737e-06,
+      "loss": 0.4176,
+      "step": 4122
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.004993690736769,
+      "learning_rate": 8.369848566690941e-06,
+      "loss": 0.4339,
+      "step": 4123
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.408077662852948,
+      "learning_rate": 8.36901662469441e-06,
+      "loss": 0.3922,
+      "step": 4124
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.439741159050259,
+      "learning_rate": 8.36818451183234e-06,
+      "loss": 0.4479,
+      "step": 4125
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.090797904897043,
+      "learning_rate": 8.367352228146931e-06,
+      "loss": 0.4768,
+      "step": 4126
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.3543874017042645,
+      "learning_rate": 8.366519773680398e-06,
+      "loss": 0.4417,
+      "step": 4127
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.151642979549596,
+      "learning_rate": 8.365687148474955e-06,
+      "loss": 0.4518,
+      "step": 4128
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.443781595303874,
+      "learning_rate": 8.364854352572834e-06,
+      "loss": 0.3864,
+      "step": 4129
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.661523162827929,
+      "learning_rate": 8.364021386016269e-06,
+      "loss": 0.4094,
+      "step": 4130
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.49140696065018674,
+      "learning_rate": 8.363188248847508e-06,
+      "loss": 0.2052,
+      "step": 4131
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.680828022253013,
+      "learning_rate": 8.362354941108803e-06,
+      "loss": 0.3931,
+      "step": 4132
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.3949498754180745,
+      "learning_rate": 8.361521462842416e-06,
+      "loss": 0.4198,
+      "step": 4133
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.902825597392937,
+      "learning_rate": 8.360687814090622e-06,
+      "loss": 0.4123,
+      "step": 4134
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.383508442843816,
+      "learning_rate": 8.359853994895698e-06,
+      "loss": 0.4712,
+      "step": 4135
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.367761705104161,
+      "learning_rate": 8.359020005299931e-06,
+      "loss": 0.4163,
+      "step": 4136
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.9069618620508217,
+      "learning_rate": 8.358185845345623e-06,
+      "loss": 0.3999,
+      "step": 4137
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.133871467760438,
+      "learning_rate": 8.357351515075077e-06,
+      "loss": 0.4131,
+      "step": 4138
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.557870106425674,
+      "learning_rate": 8.356517014530608e-06,
+      "loss": 0.4249,
+      "step": 4139
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.116968668306993,
+      "learning_rate": 8.355682343754536e-06,
+      "loss": 0.435,
+      "step": 4140
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.6549944281399736,
+      "learning_rate": 8.354847502789198e-06,
+      "loss": 0.4195,
+      "step": 4141
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.332378048234875,
+      "learning_rate": 8.354012491676933e-06,
+      "loss": 0.4029,
+      "step": 4142
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4588562271582334,
+      "learning_rate": 8.353177310460088e-06,
+      "loss": 0.2134,
+      "step": 4143
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39461326556532567,
+      "learning_rate": 8.352341959181021e-06,
+      "loss": 0.1952,
+      "step": 4144
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.531499805741717,
+      "learning_rate": 8.3515064378821e-06,
+      "loss": 0.3874,
+      "step": 4145
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.919787742096565,
+      "learning_rate": 8.350670746605698e-06,
+      "loss": 0.4426,
+      "step": 4146
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.039132016000233,
+      "learning_rate": 8.3498348853942e-06,
+      "loss": 0.3612,
+      "step": 4147
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.1023731570611774,
+      "learning_rate": 8.348998854289997e-06,
+      "loss": 0.4223,
+      "step": 4148
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.653676189366551,
+      "learning_rate": 8.348162653335493e-06,
+      "loss": 0.3918,
+      "step": 4149
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.331182576579416,
+      "learning_rate": 8.347326282573093e-06,
+      "loss": 0.3991,
+      "step": 4150
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.2484663744311,
+      "learning_rate": 8.346489742045216e-06,
+      "loss": 0.4662,
+      "step": 4151
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.8435229571949705,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.4177,
+      "step": 4152
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.426773217193984,
+      "learning_rate": 8.34481615186275e-06,
+      "loss": 0.446,
+      "step": 4153
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.284525571072227,
+      "learning_rate": 8.343979102293043e-06,
+      "loss": 0.3999,
+      "step": 4154
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 14.976378848750826,
+      "learning_rate": 8.343141883127615e-06,
+      "loss": 0.3767,
+      "step": 4155
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.249592201852045,
+      "learning_rate": 8.342304494408929e-06,
+      "loss": 0.4286,
+      "step": 4156
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.943382963323283,
+      "learning_rate": 8.341466936179457e-06,
+      "loss": 0.4354,
+      "step": 4157
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.49073079706278,
+      "learning_rate": 8.340629208481675e-06,
+      "loss": 0.4458,
+      "step": 4158
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.035874253303919,
+      "learning_rate": 8.33979131135807e-06,
+      "loss": 0.4368,
+      "step": 4159
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.506797233305738,
+      "learning_rate": 8.338953244851137e-06,
+      "loss": 0.4636,
+      "step": 4160
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.719025045686887,
+      "learning_rate": 8.338115009003384e-06,
+      "loss": 0.3944,
+      "step": 4161
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.519704575731165,
+      "learning_rate": 8.33727660385732e-06,
+      "loss": 0.3807,
+      "step": 4162
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.27233778008192,
+      "learning_rate": 8.336438029455467e-06,
+      "loss": 0.4094,
+      "step": 4163
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2917974275286728,
+      "learning_rate": 8.335599285840353e-06,
+      "loss": 0.4457,
+      "step": 4164
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.8970522204511115,
+      "learning_rate": 8.334760373054517e-06,
+      "loss": 0.4184,
+      "step": 4165
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.898999863211553,
+      "learning_rate": 8.33392129114051e-06,
+      "loss": 0.3974,
+      "step": 4166
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.324462163799164,
+      "learning_rate": 8.333082040140884e-06,
+      "loss": 0.3921,
+      "step": 4167
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.767067748730468,
+      "learning_rate": 8.3322426200982e-06,
+      "loss": 0.424,
+      "step": 4168
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.203664292622449,
+      "learning_rate": 8.331403031055036e-06,
+      "loss": 0.4362,
+      "step": 4169
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.704006848888223,
+      "learning_rate": 8.330563273053973e-06,
+      "loss": 0.4096,
+      "step": 4170
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.2669947556472203,
+      "learning_rate": 8.329723346137597e-06,
+      "loss": 0.3882,
+      "step": 4171
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.599562803469652,
+      "learning_rate": 8.32888325034851e-06,
+      "loss": 0.4317,
+      "step": 4172
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.139991060583018,
+      "learning_rate": 8.328042985729316e-06,
+      "loss": 0.435,
+      "step": 4173
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.315005514313358,
+      "learning_rate": 8.327202552322634e-06,
+      "loss": 0.3954,
+      "step": 4174
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.455315959623293,
+      "learning_rate": 8.326361950171083e-06,
+      "loss": 0.4153,
+      "step": 4175
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.882501997342048,
+      "learning_rate": 8.325521179317298e-06,
+      "loss": 0.4544,
+      "step": 4176
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.7935011266627714,
+      "learning_rate": 8.324680239803922e-06,
+      "loss": 0.3621,
+      "step": 4177
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.701580607968931,
+      "learning_rate": 8.323839131673603e-06,
+      "loss": 0.4108,
+      "step": 4178
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.081955247684716,
+      "learning_rate": 8.322997854969e-06,
+      "loss": 0.42,
+      "step": 4179
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.627320368276557,
+      "learning_rate": 8.32215640973278e-06,
+      "loss": 0.3791,
+      "step": 4180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.212656926142657,
+      "learning_rate": 8.321314796007618e-06,
+      "loss": 0.4213,
+      "step": 4181
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.6237912907149052,
+      "learning_rate": 8.320473013836197e-06,
+      "loss": 0.4171,
+      "step": 4182
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.6040330267126475,
+      "learning_rate": 8.319631063261209e-06,
+      "loss": 0.4126,
+      "step": 4183
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.208808125589513,
+      "learning_rate": 8.318788944325357e-06,
+      "loss": 0.3897,
+      "step": 4184
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.667666033182372,
+      "learning_rate": 8.31794665707135e-06,
+      "loss": 0.4365,
+      "step": 4185
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.886073097856462,
+      "learning_rate": 8.317104201541904e-06,
+      "loss": 0.4626,
+      "step": 4186
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.690559890776747,
+      "learning_rate": 8.316261577779749e-06,
+      "loss": 0.3958,
+      "step": 4187
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.954530335360811,
+      "learning_rate": 8.315418785827618e-06,
+      "loss": 0.3857,
+      "step": 4188
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.06531305289575,
+      "learning_rate": 8.314575825728254e-06,
+      "loss": 0.3985,
+      "step": 4189
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.638734364799122,
+      "learning_rate": 8.313732697524413e-06,
+      "loss": 0.4291,
+      "step": 4190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.014028753389908,
+      "learning_rate": 8.31288940125885e-06,
+      "loss": 0.4146,
+      "step": 4191
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.2693945684002275,
+      "learning_rate": 8.312045936974339e-06,
+      "loss": 0.4131,
+      "step": 4192
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.286538716785039,
+      "learning_rate": 8.311202304713655e-06,
+      "loss": 0.4419,
+      "step": 4193
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.402866601480544,
+      "learning_rate": 8.310358504519587e-06,
+      "loss": 0.4387,
+      "step": 4194
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.071734699783866,
+      "learning_rate": 8.309514536434926e-06,
+      "loss": 0.3952,
+      "step": 4195
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.5395406002742265,
+      "learning_rate": 8.308670400502479e-06,
+      "loss": 0.4494,
+      "step": 4196
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.506085633870856,
+      "learning_rate": 8.307826096765054e-06,
+      "loss": 0.4124,
+      "step": 4197
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.4673084641870195,
+      "learning_rate": 8.306981625265476e-06,
+      "loss": 0.4629,
+      "step": 4198
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.596340691672042,
+      "learning_rate": 8.306136986046568e-06,
+      "loss": 0.4462,
+      "step": 4199
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.753987896015158,
+      "learning_rate": 8.305292179151175e-06,
+      "loss": 0.4272,
+      "step": 4200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.010962351709068,
+      "learning_rate": 8.304447204622135e-06,
+      "loss": 0.4138,
+      "step": 4201
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.459319142366127,
+      "learning_rate": 8.303602062502307e-06,
+      "loss": 0.4502,
+      "step": 4202
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.051308037071492,
+      "learning_rate": 8.302756752834553e-06,
+      "loss": 0.4042,
+      "step": 4203
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.914375422323402,
+      "learning_rate": 8.301911275661744e-06,
+      "loss": 0.4053,
+      "step": 4204
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.54634197407742,
+      "learning_rate": 8.301065631026759e-06,
+      "loss": 0.4131,
+      "step": 4205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.4585198591102735,
+      "learning_rate": 8.30021981897249e-06,
+      "loss": 0.3995,
+      "step": 4206
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.022418876545359,
+      "learning_rate": 8.299373839541829e-06,
+      "loss": 0.4113,
+      "step": 4207
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.093241696870406,
+      "learning_rate": 8.298527692777682e-06,
+      "loss": 0.4088,
+      "step": 4208
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.122738919576237,
+      "learning_rate": 8.297681378722967e-06,
+      "loss": 0.4315,
+      "step": 4209
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.781117210688171,
+      "learning_rate": 8.296834897420602e-06,
+      "loss": 0.4676,
+      "step": 4210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.458112069560963,
+      "learning_rate": 8.29598824891352e-06,
+      "loss": 0.4034,
+      "step": 4211
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 9.638228310401496,
+      "learning_rate": 8.29514143324466e-06,
+      "loss": 0.4086,
+      "step": 4212
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.191609314803523,
+      "learning_rate": 8.29429445045697e-06,
+      "loss": 0.461,
+      "step": 4213
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.553063489018637,
+      "learning_rate": 8.293447300593402e-06,
+      "loss": 0.4128,
+      "step": 4214
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.334501077389333,
+      "learning_rate": 8.292599983696927e-06,
+      "loss": 0.3698,
+      "step": 4215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.356816505330811,
+      "learning_rate": 8.291752499810515e-06,
+      "loss": 0.4897,
+      "step": 4216
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.1448950638886926,
+      "learning_rate": 8.290904848977149e-06,
+      "loss": 0.4501,
+      "step": 4217
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6916421901274756,
+      "learning_rate": 8.290057031239815e-06,
+      "loss": 0.2467,
+      "step": 4218
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.980506681043826,
+      "learning_rate": 8.289209046641518e-06,
+      "loss": 0.4367,
+      "step": 4219
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.038801326194554,
+      "learning_rate": 8.288360895225259e-06,
+      "loss": 0.4396,
+      "step": 4220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.566071011146666,
+      "learning_rate": 8.287512577034058e-06,
+      "loss": 0.388,
+      "step": 4221
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.7246887706610865,
+      "learning_rate": 8.286664092110935e-06,
+      "loss": 0.4198,
+      "step": 4222
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.990396688219263,
+      "learning_rate": 8.285815440498926e-06,
+      "loss": 0.4046,
+      "step": 4223
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.489801856488194,
+      "learning_rate": 8.284966622241069e-06,
+      "loss": 0.428,
+      "step": 4224
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4658962032051779,
+      "learning_rate": 8.284117637380415e-06,
+      "loss": 0.2123,
+      "step": 4225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.565065366143205,
+      "learning_rate": 8.28326848596002e-06,
+      "loss": 0.3941,
+      "step": 4226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.965162825295831,
+      "learning_rate": 8.282419168022953e-06,
+      "loss": 0.4289,
+      "step": 4227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.51066265743577,
+      "learning_rate": 8.281569683612288e-06,
+      "loss": 0.4508,
+      "step": 4228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.860244395607432,
+      "learning_rate": 8.280720032771105e-06,
+      "loss": 0.4239,
+      "step": 4229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.9187303525830535,
+      "learning_rate": 8.279870215542499e-06,
+      "loss": 0.4275,
+      "step": 4230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 12.101619988896651,
+      "learning_rate": 8.27902023196957e-06,
+      "loss": 0.3844,
+      "step": 4231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.73600749014312,
+      "learning_rate": 8.278170082095422e-06,
+      "loss": 0.4306,
+      "step": 4232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 5.240301335036803,
+      "learning_rate": 8.277319765963176e-06,
+      "loss": 0.4841,
+      "step": 4233
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.466262427336109,
+      "learning_rate": 8.276469283615958e-06,
+      "loss": 0.4066,
+      "step": 4234
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 10.4299021946191,
+      "learning_rate": 8.275618635096898e-06,
+      "loss": 0.4332,
+      "step": 4235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 8.005651062922066,
+      "learning_rate": 8.27476782044914e-06,
+      "loss": 0.4082,
+      "step": 4236
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.69721171108251,
+      "learning_rate": 8.273916839715835e-06,
+      "loss": 0.397,
+      "step": 4237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.461345181538372,
+      "learning_rate": 8.27306569294014e-06,
+      "loss": 0.4238,
+      "step": 4238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.821631259429967,
+      "learning_rate": 8.272214380165226e-06,
+      "loss": 0.404,
+      "step": 4239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 11.836357240360345,
+      "learning_rate": 8.271362901434266e-06,
+      "loss": 0.4175,
+      "step": 4240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 7.595457198143039,
+      "learning_rate": 8.270511256790445e-06,
+      "loss": 0.4151,
+      "step": 4241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 13.91624936491638,
+      "learning_rate": 8.269659446276955e-06,
+      "loss": 0.3577,
+      "step": 4242
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.44568378229249134,
+      "learning_rate": 8.268807469937e-06,
+      "loss": 0.2087,
+      "step": 4243
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.921453813755103,
+      "learning_rate": 8.267955327813783e-06,
+      "loss": 0.4151,
+      "step": 4244
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.299357345517884,
+      "learning_rate": 8.267103019950529e-06,
+      "loss": 0.4062,
+      "step": 4245
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.2935189481565095,
+      "learning_rate": 8.26625054639046e-06,
+      "loss": 0.3929,
+      "step": 4246
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.554750743570194,
+      "learning_rate": 8.26539790717681e-06,
+      "loss": 0.4019,
+      "step": 4247
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.117709467582226,
+      "learning_rate": 8.264545102352827e-06,
+      "loss": 0.3782,
+      "step": 4248
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.809827255418329,
+      "learning_rate": 8.263692131961758e-06,
+      "loss": 0.4063,
+      "step": 4249
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9842955467078847,
+      "learning_rate": 8.262838996046865e-06,
+      "loss": 0.4508,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.600901623567604,
+      "learning_rate": 8.261985694651416e-06,
+      "loss": 0.4089,
+      "step": 4251
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.663481815990723,
+      "learning_rate": 8.261132227818685e-06,
+      "loss": 0.3936,
+      "step": 4252
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.770149657272915,
+      "learning_rate": 8.260278595591962e-06,
+      "loss": 0.4068,
+      "step": 4253
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.434038271142298,
+      "learning_rate": 8.259424798014536e-06,
+      "loss": 0.4202,
+      "step": 4254
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.848336804661795,
+      "learning_rate": 8.25857083512971e-06,
+      "loss": 0.4224,
+      "step": 4255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.148332159290485,
+      "learning_rate": 8.257716706980796e-06,
+      "loss": 0.4117,
+      "step": 4256
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9386143524453794,
+      "learning_rate": 8.256862413611113e-06,
+      "loss": 0.3977,
+      "step": 4257
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.415796571609748,
+      "learning_rate": 8.256007955063983e-06,
+      "loss": 0.4248,
+      "step": 4258
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.479825475137746,
+      "learning_rate": 8.255153331382747e-06,
+      "loss": 0.4385,
+      "step": 4259
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.75984648679745,
+      "learning_rate": 8.254298542610745e-06,
+      "loss": 0.4061,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.819972039150718,
+      "learning_rate": 8.253443588791332e-06,
+      "loss": 0.4367,
+      "step": 4261
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4409814983844045,
+      "learning_rate": 8.252588469967866e-06,
+      "loss": 0.2116,
+      "step": 4262
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 13.91460892153716,
+      "learning_rate": 8.251733186183718e-06,
+      "loss": 0.4195,
+      "step": 4263
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.277990240670775,
+      "learning_rate": 8.250877737482264e-06,
+      "loss": 0.4203,
+      "step": 4264
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.330213206100834,
+      "learning_rate": 8.250022123906887e-06,
+      "loss": 0.3842,
+      "step": 4265
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.094801652845434,
+      "learning_rate": 8.249166345500984e-06,
+      "loss": 0.3979,
+      "step": 4266
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.454070072607039,
+      "learning_rate": 8.24831040230796e-06,
+      "loss": 0.4498,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.70969947769997,
+      "learning_rate": 8.247454294371221e-06,
+      "loss": 0.3952,
+      "step": 4268
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4929627724517824,
+      "learning_rate": 8.246598021734187e-06,
+      "loss": 0.4318,
+      "step": 4269
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3823983863597921,
+      "learning_rate": 8.245741584440286e-06,
+      "loss": 0.202,
+      "step": 4270
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 12.586899853320855,
+      "learning_rate": 8.244884982532954e-06,
+      "loss": 0.4547,
+      "step": 4271
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.756852115855495,
+      "learning_rate": 8.244028216055636e-06,
+      "loss": 0.4069,
+      "step": 4272
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.978206736178281,
+      "learning_rate": 8.24317128505178e-06,
+      "loss": 0.4156,
+      "step": 4273
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.10840115261451,
+      "learning_rate": 8.242314189564851e-06,
+      "loss": 0.416,
+      "step": 4274
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4239203046613467,
+      "learning_rate": 8.241456929638316e-06,
+      "loss": 0.4488,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.672529778504821,
+      "learning_rate": 8.240599505315656e-06,
+      "loss": 0.4222,
+      "step": 4276
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.413930792948165,
+      "learning_rate": 8.239741916640352e-06,
+      "loss": 0.4394,
+      "step": 4277
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.695890158035813,
+      "learning_rate": 8.238884163655901e-06,
+      "loss": 0.3874,
+      "step": 4278
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.503339790580438,
+      "learning_rate": 8.238026246405804e-06,
+      "loss": 0.4054,
+      "step": 4279
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.889229378802362,
+      "learning_rate": 8.237168164933572e-06,
+      "loss": 0.4451,
+      "step": 4280
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.793003088852384,
+      "learning_rate": 8.236309919282724e-06,
+      "loss": 0.4379,
+      "step": 4281
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.111399157484469,
+      "learning_rate": 8.23545150949679e-06,
+      "loss": 0.4478,
+      "step": 4282
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.3882187814494475,
+      "learning_rate": 8.2345929356193e-06,
+      "loss": 0.4072,
+      "step": 4283
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.274331521222838,
+      "learning_rate": 8.233734197693805e-06,
+      "loss": 0.3927,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.792473876086023,
+      "learning_rate": 8.232875295763853e-06,
+      "loss": 0.4224,
+      "step": 4285
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.565993939277687,
+      "learning_rate": 8.232016229873006e-06,
+      "loss": 0.4485,
+      "step": 4286
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7020009323570924,
+      "learning_rate": 8.231157000064833e-06,
+      "loss": 0.425,
+      "step": 4287
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.695800339795739,
+      "learning_rate": 8.230297606382911e-06,
+      "loss": 0.4211,
+      "step": 4288
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.325683224034369,
+      "learning_rate": 8.229438048870826e-06,
+      "loss": 0.4378,
+      "step": 4289
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.0487015796287595,
+      "learning_rate": 8.228578327572172e-06,
+      "loss": 0.4469,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.4716726128036886,
+      "learning_rate": 8.227718442530549e-06,
+      "loss": 0.4015,
+      "step": 4291
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.498043539206544,
+      "learning_rate": 8.226858393789572e-06,
+      "loss": 0.4248,
+      "step": 4292
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39224057163502135,
+      "learning_rate": 8.225998181392854e-06,
+      "loss": 0.2131,
+      "step": 4293
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 16.583412944841527,
+      "learning_rate": 8.225137805384029e-06,
+      "loss": 0.4249,
+      "step": 4294
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.014649158005886,
+      "learning_rate": 8.224277265806729e-06,
+      "loss": 0.403,
+      "step": 4295
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.945972532537996,
+      "learning_rate": 8.223416562704596e-06,
+      "loss": 0.4339,
+      "step": 4296
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.763418664258444,
+      "learning_rate": 8.222555696121283e-06,
+      "loss": 0.4104,
+      "step": 4297
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.6968970545756985,
+      "learning_rate": 8.221694666100455e-06,
+      "loss": 0.413,
+      "step": 4298
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.488226315402943,
+      "learning_rate": 8.220833472685774e-06,
+      "loss": 0.4325,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.188033437494355,
+      "learning_rate": 8.21997211592092e-06,
+      "loss": 0.4295,
+      "step": 4300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.3809614164204245,
+      "learning_rate": 8.21911059584958e-06,
+      "loss": 0.416,
+      "step": 4301
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.214068154930528,
+      "learning_rate": 8.218248912515443e-06,
+      "loss": 0.4393,
+      "step": 4302
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3789061311963566,
+      "learning_rate": 8.217387065962214e-06,
+      "loss": 0.1954,
+      "step": 4303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.861030560246592,
+      "learning_rate": 8.216525056233604e-06,
+      "loss": 0.4274,
+      "step": 4304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7094442259566147,
+      "learning_rate": 8.215662883373328e-06,
+      "loss": 0.433,
+      "step": 4305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.258467560971216,
+      "learning_rate": 8.214800547425115e-06,
+      "loss": 0.4046,
+      "step": 4306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.37270251235154356,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.2069,
+      "step": 4307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.804612608296434,
+      "learning_rate": 8.213075386439821e-06,
+      "loss": 0.4346,
+      "step": 4308
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.3602974317700065,
+      "learning_rate": 8.212212561490236e-06,
+      "loss": 0.3848,
+      "step": 4309
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3542585004975726,
+      "learning_rate": 8.211349573627703e-06,
+      "loss": 0.1967,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.35719856671217237,
+      "learning_rate": 8.210486422895989e-06,
+      "loss": 0.2114,
+      "step": 4311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.625903194991858,
+      "learning_rate": 8.209623109338872e-06,
+      "loss": 0.4359,
+      "step": 4312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.084221231461204,
+      "learning_rate": 8.208759633000132e-06,
+      "loss": 0.4013,
+      "step": 4313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.348298009479489,
+      "learning_rate": 8.207895993923566e-06,
+      "loss": 0.4223,
+      "step": 4314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.644709866014989,
+      "learning_rate": 8.207032192152974e-06,
+      "loss": 0.4171,
+      "step": 4315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.818760770116508,
+      "learning_rate": 8.206168227732164e-06,
+      "loss": 0.4188,
+      "step": 4316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.871391503099464,
+      "learning_rate": 8.205304100704953e-06,
+      "loss": 0.401,
+      "step": 4317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3695228348207821,
+      "learning_rate": 8.20443981111517e-06,
+      "loss": 0.1978,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.6402381929441505,
+      "learning_rate": 8.203575359006645e-06,
+      "loss": 0.3778,
+      "step": 4319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.296736376211266,
+      "learning_rate": 8.202710744423223e-06,
+      "loss": 0.4563,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.984556020897953,
+      "learning_rate": 8.201845967408753e-06,
+      "loss": 0.4203,
+      "step": 4321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.606341701157499,
+      "learning_rate": 8.200981028007095e-06,
+      "loss": 0.3937,
+      "step": 4322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.350646036081955,
+      "learning_rate": 8.200115926262116e-06,
+      "loss": 0.445,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.704526065594238,
+      "learning_rate": 8.199250662217688e-06,
+      "loss": 0.4099,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.266795408900784,
+      "learning_rate": 8.198385235917697e-06,
+      "loss": 0.3856,
+      "step": 4325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 18.59066458379753,
+      "learning_rate": 8.197519647406036e-06,
+      "loss": 0.4346,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.14112992942462,
+      "learning_rate": 8.1966538967266e-06,
+      "loss": 0.4775,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.654611826392639,
+      "learning_rate": 8.195787983923302e-06,
+      "loss": 0.4085,
+      "step": 4328
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.9770492238148596,
+      "learning_rate": 8.194921909040056e-06,
+      "loss": 0.4188,
+      "step": 4329
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 10.900253468028529,
+      "learning_rate": 8.194055672120788e-06,
+      "loss": 0.422,
+      "step": 4330
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.368830299942182,
+      "learning_rate": 8.193189273209428e-06,
+      "loss": 0.4108,
+      "step": 4331
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.306058395713658,
+      "learning_rate": 8.192322712349917e-06,
+      "loss": 0.4007,
+      "step": 4332
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.666832160930862,
+      "learning_rate": 8.191455989586208e-06,
+      "loss": 0.4158,
+      "step": 4333
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.512511764731039,
+      "learning_rate": 8.190589104962257e-06,
+      "loss": 0.41,
+      "step": 4334
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.97818371688031,
+      "learning_rate": 8.189722058522025e-06,
+      "loss": 0.4035,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.021182869251373,
+      "learning_rate": 8.18885485030949e-06,
+      "loss": 0.4405,
+      "step": 4336
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.657461075140102,
+      "learning_rate": 8.187987480368637e-06,
+      "loss": 0.4119,
+      "step": 4337
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.580098276280396,
+      "learning_rate": 8.18711994874345e-06,
+      "loss": 0.4186,
+      "step": 4338
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.002912498378594,
+      "learning_rate": 8.186252255477929e-06,
+      "loss": 0.4058,
+      "step": 4339
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.371698471346542,
+      "learning_rate": 8.185384400616082e-06,
+      "loss": 0.4434,
+      "step": 4340
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.703364251227013,
+      "learning_rate": 8.184516384201924e-06,
+      "loss": 0.4551,
+      "step": 4341
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.429690220456141,
+      "learning_rate": 8.183648206279476e-06,
+      "loss": 0.377,
+      "step": 4342
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.381758008026007,
+      "learning_rate": 8.18277986689277e-06,
+      "loss": 0.3941,
+      "step": 4343
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 15.687617541558998,
+      "learning_rate": 8.181911366085846e-06,
+      "loss": 0.4427,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.912843266938991,
+      "learning_rate": 8.181042703902751e-06,
+      "loss": 0.4141,
+      "step": 4345
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.81339832287152,
+      "learning_rate": 8.180173880387542e-06,
+      "loss": 0.404,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.939397241393293,
+      "learning_rate": 8.179304895584282e-06,
+      "loss": 0.4406,
+      "step": 4347
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.114389839857839,
+      "learning_rate": 8.17843574953704e-06,
+      "loss": 0.4225,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.7030941570870075,
+      "learning_rate": 8.177566442289902e-06,
+      "loss": 0.4019,
+      "step": 4349
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.068509172852253,
+      "learning_rate": 8.176696973886951e-06,
+      "loss": 0.4145,
+      "step": 4350
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.913657776569597,
+      "learning_rate": 8.17582734437229e-06,
+      "loss": 0.408,
+      "step": 4351
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.580699833126978,
+      "learning_rate": 8.174957553790015e-06,
+      "loss": 0.4442,
+      "step": 4352
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.255491654280537,
+      "learning_rate": 8.174087602184247e-06,
+      "loss": 0.3951,
+      "step": 4353
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.519577593009908,
+      "learning_rate": 8.173217489599102e-06,
+      "loss": 0.3925,
+      "step": 4354
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.048490531627075,
+      "learning_rate": 8.172347216078713e-06,
+      "loss": 0.4002,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.570515086205827,
+      "learning_rate": 8.171476781667215e-06,
+      "loss": 0.4278,
+      "step": 4356
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 8.42394868436176,
+      "learning_rate": 8.170606186408756e-06,
+      "loss": 0.4405,
+      "step": 4357
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.594092446306364,
+      "learning_rate": 8.169735430347486e-06,
+      "loss": 0.4458,
+      "step": 4358
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.047447095737779,
+      "learning_rate": 8.168864513527569e-06,
+      "loss": 0.4136,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.7888551972798643,
+      "learning_rate": 8.167993435993174e-06,
+      "loss": 0.4228,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 11.566189124815887,
+      "learning_rate": 8.167122197788482e-06,
+      "loss": 0.4371,
+      "step": 4361
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.2421731616127625,
+      "learning_rate": 8.166250798957676e-06,
+      "loss": 0.4211,
+      "step": 4362
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.39470180090256696,
+      "learning_rate": 8.165379239544953e-06,
+      "loss": 0.1909,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.028612719443949,
+      "learning_rate": 8.164507519594513e-06,
+      "loss": 0.4101,
+      "step": 4364
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.765685124752632,
+      "learning_rate": 8.16363563915057e-06,
+      "loss": 0.4667,
+      "step": 4365
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.461210087200872,
+      "learning_rate": 8.162763598257342e-06,
+      "loss": 0.3917,
+      "step": 4366
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 7.534899945074374,
+      "learning_rate": 8.161891396959056e-06,
+      "loss": 0.4263,
+      "step": 4367
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.758750086265754,
+      "learning_rate": 8.161019035299945e-06,
+      "loss": 0.4019,
+      "step": 4368
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.556482392505828,
+      "learning_rate": 8.160146513324256e-06,
+      "loss": 0.4312,
+      "step": 4369
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.372100316908464,
+      "learning_rate": 8.159273831076235e-06,
+      "loss": 0.4073,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.4828015042009755,
+      "learning_rate": 8.158400988600149e-06,
+      "loss": 0.3961,
+      "step": 4371
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.039306681692093,
+      "learning_rate": 8.157527985940259e-06,
+      "loss": 0.4111,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.6597465709826,
+      "learning_rate": 8.156654823140847e-06,
+      "loss": 0.3979,
+      "step": 4373
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.302008526465848,
+      "learning_rate": 8.15578150024619e-06,
+      "loss": 0.4597,
+      "step": 4374
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 14.965684513638582,
+      "learning_rate": 8.154908017300586e-06,
+      "loss": 0.4131,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.8049946341216465,
+      "learning_rate": 8.154034374348332e-06,
+      "loss": 0.4844,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.496500568910825,
+      "learning_rate": 8.153160571433738e-06,
+      "loss": 0.3878,
+      "step": 4377
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.278041509736323,
+      "learning_rate": 8.15228660860112e-06,
+      "loss": 0.3882,
+      "step": 4378
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 3.914611930452162,
+      "learning_rate": 8.151412485894801e-06,
+      "loss": 0.39,
+      "step": 4379
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.576626264740117,
+      "learning_rate": 8.150538203359114e-06,
+      "loss": 0.4258,
+      "step": 4380
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.0824834866659,
+      "learning_rate": 8.149663761038403e-06,
+      "loss": 0.3623,
+      "step": 4381
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.563889079186109,
+      "learning_rate": 8.148789158977012e-06,
+      "loss": 0.4081,
+      "step": 4382
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.02985513569214,
+      "learning_rate": 8.147914397219303e-06,
+      "loss": 0.3893,
+      "step": 4383
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 4.1999402963366235,
+      "learning_rate": 8.147039475809637e-06,
+      "loss": 0.404,
+      "step": 4384
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.119631967562611,
+      "learning_rate": 8.14616439479239e-06,
+      "loss": 0.4022,
+      "step": 4385
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 5.430082198007959,
+      "learning_rate": 8.145289154211941e-06,
+      "loss": 0.4393,
+      "step": 4386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.422073656116502,
+      "learning_rate": 8.14441375411268e-06,
+      "loss": 0.4647,
+      "step": 4387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.479857481470168,
+      "learning_rate": 8.143538194539007e-06,
+      "loss": 0.4171,
+      "step": 4388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.069092275428551,
+      "learning_rate": 8.142662475535323e-06,
+      "loss": 0.4157,
+      "step": 4389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.526033845981653,
+      "learning_rate": 8.141786597146045e-06,
+      "loss": 0.3864,
+      "step": 4390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.459992215799837,
+      "learning_rate": 8.140910559415594e-06,
+      "loss": 0.4297,
+      "step": 4391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.494555308466055,
+      "learning_rate": 8.140034362388398e-06,
+      "loss": 0.4274,
+      "step": 4392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.891029493519575,
+      "learning_rate": 8.139158006108897e-06,
+      "loss": 0.4376,
+      "step": 4393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.315216570029798,
+      "learning_rate": 8.138281490621535e-06,
+      "loss": 0.4006,
+      "step": 4394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.743984878482132,
+      "learning_rate": 8.137404815970769e-06,
+      "loss": 0.3961,
+      "step": 4395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.83605386918975,
+      "learning_rate": 8.136527982201059e-06,
+      "loss": 0.4579,
+      "step": 4396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.232950994937239,
+      "learning_rate": 8.135650989356875e-06,
+      "loss": 0.4008,
+      "step": 4397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.964159347451008,
+      "learning_rate": 8.134773837482695e-06,
+      "loss": 0.3975,
+      "step": 4398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.280784347219172,
+      "learning_rate": 8.133896526623007e-06,
+      "loss": 0.4244,
+      "step": 4399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.577233401581894,
+      "learning_rate": 8.133019056822303e-06,
+      "loss": 0.433,
+      "step": 4400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.330331881352542,
+      "learning_rate": 8.132141428125087e-06,
+      "loss": 0.4599,
+      "step": 4401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.818594739084031,
+      "learning_rate": 8.131263640575871e-06,
+      "loss": 0.4278,
+      "step": 4402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.5152907078983535,
+      "learning_rate": 8.130385694219169e-06,
+      "loss": 0.4148,
+      "step": 4403
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.87985403797708,
+      "learning_rate": 8.129507589099512e-06,
+      "loss": 0.4058,
+      "step": 4404
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4029658546312475,
+      "learning_rate": 8.128629325261432e-06,
+      "loss": 0.2031,
+      "step": 4405
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.656380580000618,
+      "learning_rate": 8.127750902749472e-06,
+      "loss": 0.4313,
+      "step": 4406
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.08050053842473,
+      "learning_rate": 8.126872321608185e-06,
+      "loss": 0.4229,
+      "step": 4407
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.825122193076309,
+      "learning_rate": 8.125993581882128e-06,
+      "loss": 0.417,
+      "step": 4408
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.889292678355407,
+      "learning_rate": 8.125114683615866e-06,
+      "loss": 0.3836,
+      "step": 4409
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.425517968849543,
+      "learning_rate": 8.124235626853976e-06,
+      "loss": 0.4109,
+      "step": 4410
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.370856240201135,
+      "learning_rate": 8.123356411641042e-06,
+      "loss": 0.3683,
+      "step": 4411
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.3109427063109615,
+      "learning_rate": 8.122477038021653e-06,
+      "loss": 0.4311,
+      "step": 4412
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.3215112751829015,
+      "learning_rate": 8.121597506040409e-06,
+      "loss": 0.4296,
+      "step": 4413
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.1659918493226185,
+      "learning_rate": 8.120717815741915e-06,
+      "loss": 0.4311,
+      "step": 4414
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4090471762646208,
+      "learning_rate": 8.119837967170787e-06,
+      "loss": 0.2039,
+      "step": 4415
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.966361243821813,
+      "learning_rate": 8.118957960371651e-06,
+      "loss": 0.4054,
+      "step": 4416
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.653904677455482,
+      "learning_rate": 8.118077795389135e-06,
+      "loss": 0.3601,
+      "step": 4417
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.024364964074834,
+      "learning_rate": 8.117197472267878e-06,
+      "loss": 0.4435,
+      "step": 4418
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.624963976955071,
+      "learning_rate": 8.116316991052527e-06,
+      "loss": 0.3998,
+      "step": 4419
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.478593252893143,
+      "learning_rate": 8.11543635178774e-06,
+      "loss": 0.3998,
+      "step": 4420
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.447703186082814,
+      "learning_rate": 8.114555554518178e-06,
+      "loss": 0.3968,
+      "step": 4421
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.229453042180427,
+      "learning_rate": 8.11367459928851e-06,
+      "loss": 0.4396,
+      "step": 4422
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.353014321934053,
+      "learning_rate": 8.11279348614342e-06,
+      "loss": 0.214,
+      "step": 4423
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.5030957081726175,
+      "learning_rate": 8.111912215127592e-06,
+      "loss": 0.4228,
+      "step": 4424
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.801380744012295,
+      "learning_rate": 8.11103078628572e-06,
+      "loss": 0.4412,
+      "step": 4425
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.360836623593141,
+      "learning_rate": 8.110149199662511e-06,
+      "loss": 0.3989,
+      "step": 4426
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.360104256508818,
+      "learning_rate": 8.109267455302674e-06,
+      "loss": 0.4178,
+      "step": 4427
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.565261134174427,
+      "learning_rate": 8.108385553250927e-06,
+      "loss": 0.3844,
+      "step": 4428
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.528997029469665,
+      "learning_rate": 8.107503493552002e-06,
+      "loss": 0.4326,
+      "step": 4429
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.660366602970748,
+      "learning_rate": 8.10662127625063e-06,
+      "loss": 0.4208,
+      "step": 4430
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.434365072190762,
+      "learning_rate": 8.105738901391553e-06,
+      "loss": 0.461,
+      "step": 4431
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.416349830111454,
+      "learning_rate": 8.104856369019525e-06,
+      "loss": 0.4726,
+      "step": 4432
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.677263110888786,
+      "learning_rate": 8.103973679179305e-06,
+      "loss": 0.4278,
+      "step": 4433
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.507700455810538,
+      "learning_rate": 8.103090831915659e-06,
+      "loss": 0.416,
+      "step": 4434
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.361306291388305,
+      "learning_rate": 8.102207827273362e-06,
+      "loss": 0.4273,
+      "step": 4435
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.958786920491476,
+      "learning_rate": 8.101324665297199e-06,
+      "loss": 0.4046,
+      "step": 4436
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.125919322524142,
+      "learning_rate": 8.100441346031958e-06,
+      "loss": 0.4016,
+      "step": 4437
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.532061541567203,
+      "learning_rate": 8.099557869522441e-06,
+      "loss": 0.4331,
+      "step": 4438
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.703181674398699,
+      "learning_rate": 8.098674235813456e-06,
+      "loss": 0.4034,
+      "step": 4439
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.445350106845257,
+      "learning_rate": 8.097790444949814e-06,
+      "loss": 0.4442,
+      "step": 4440
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4045072276705423,
+      "learning_rate": 8.09690649697634e-06,
+      "loss": 0.1883,
+      "step": 4441
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.578035389190888,
+      "learning_rate": 8.096022391937866e-06,
+      "loss": 0.4278,
+      "step": 4442
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.56949104542176,
+      "learning_rate": 8.095138129879228e-06,
+      "loss": 0.4176,
+      "step": 4443
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.46764824071613,
+      "learning_rate": 8.094253710845277e-06,
+      "loss": 0.4283,
+      "step": 4444
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.58153838407513,
+      "learning_rate": 8.093369134880864e-06,
+      "loss": 0.4367,
+      "step": 4445
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.1549738620037004,
+      "learning_rate": 8.092484402030853e-06,
+      "loss": 0.4056,
+      "step": 4446
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.973219573388222,
+      "learning_rate": 8.091599512340117e-06,
+      "loss": 0.3824,
+      "step": 4447
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.487383789437756,
+      "learning_rate": 8.090714465853534e-06,
+      "loss": 0.392,
+      "step": 4448
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.5972505102041366,
+      "learning_rate": 8.089829262615987e-06,
+      "loss": 0.4274,
+      "step": 4449
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.2853106503828045,
+      "learning_rate": 8.088943902672372e-06,
+      "loss": 0.3966,
+      "step": 4450
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.758533517795644,
+      "learning_rate": 8.088058386067597e-06,
+      "loss": 0.4175,
+      "step": 4451
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.347785082764759,
+      "learning_rate": 8.087172712846565e-06,
+      "loss": 0.4352,
+      "step": 4452
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 8.414888420725724,
+      "learning_rate": 8.086286883054198e-06,
+      "loss": 0.3878,
+      "step": 4453
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.739052657856908,
+      "learning_rate": 8.085400896735425e-06,
+      "loss": 0.4258,
+      "step": 4454
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4099737768252597,
+      "learning_rate": 8.084514753935173e-06,
+      "loss": 0.2061,
+      "step": 4455
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.38575035867055246,
+      "learning_rate": 8.083628454698394e-06,
+      "loss": 0.2013,
+      "step": 4456
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.472229577339026,
+      "learning_rate": 8.08274199907003e-06,
+      "loss": 0.4292,
+      "step": 4457
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.757940243169285,
+      "learning_rate": 8.081855387095041e-06,
+      "loss": 0.4225,
+      "step": 4458
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.576797035232156,
+      "learning_rate": 8.080968618818395e-06,
+      "loss": 0.4307,
+      "step": 4459
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.820715604686101,
+      "learning_rate": 8.080081694285067e-06,
+      "loss": 0.4283,
+      "step": 4460
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.861566933567739,
+      "learning_rate": 8.079194613540036e-06,
+      "loss": 0.4355,
+      "step": 4461
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.014225463146937,
+      "learning_rate": 8.078307376628292e-06,
+      "loss": 0.4185,
+      "step": 4462
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.0772617242493014,
+      "learning_rate": 8.077419983594835e-06,
+      "loss": 0.4566,
+      "step": 4463
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5138169897252161,
+      "learning_rate": 8.076532434484668e-06,
+      "loss": 0.2044,
+      "step": 4464
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.888791150782928,
+      "learning_rate": 8.075644729342808e-06,
+      "loss": 0.3758,
+      "step": 4465
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.865452325832907,
+      "learning_rate": 8.074756868214272e-06,
+      "loss": 0.4446,
+      "step": 4466
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.8212193262209486,
+      "learning_rate": 8.073868851144094e-06,
+      "loss": 0.3882,
+      "step": 4467
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 34.715393235650545,
+      "learning_rate": 8.07298067817731e-06,
+      "loss": 0.4387,
+      "step": 4468
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.710585252448927,
+      "learning_rate": 8.072092349358963e-06,
+      "loss": 0.3905,
+      "step": 4469
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 10.586349740337768,
+      "learning_rate": 8.071203864734108e-06,
+      "loss": 0.4116,
+      "step": 4470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.03036216262157,
+      "learning_rate": 8.070315224347805e-06,
+      "loss": 0.4449,
+      "step": 4471
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.648767566898222,
+      "learning_rate": 8.069426428245124e-06,
+      "loss": 0.4247,
+      "step": 4472
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.626198556999888,
+      "learning_rate": 8.06853747647114e-06,
+      "loss": 0.4153,
+      "step": 4473
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.677216344427246,
+      "learning_rate": 8.067648369070942e-06,
+      "loss": 0.3948,
+      "step": 4474
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.6328629775935,
+      "learning_rate": 8.06675910608962e-06,
+      "loss": 0.418,
+      "step": 4475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.34600373401435,
+      "learning_rate": 8.065869687572272e-06,
+      "loss": 0.428,
+      "step": 4476
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.10652007994312,
+      "learning_rate": 8.06498011356401e-06,
+      "loss": 0.3969,
+      "step": 4477
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.572331270871378,
+      "learning_rate": 8.064090384109949e-06,
+      "loss": 0.4214,
+      "step": 4478
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.886635684079407,
+      "learning_rate": 8.063200499255213e-06,
+      "loss": 0.4325,
+      "step": 4479
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.607006538189284,
+      "learning_rate": 8.062310459044933e-06,
+      "loss": 0.4088,
+      "step": 4480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 13.900339657500433,
+      "learning_rate": 8.06142026352425e-06,
+      "loss": 0.3968,
+      "step": 4481
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.702516950314712,
+      "learning_rate": 8.060529912738316e-06,
+      "loss": 0.4065,
+      "step": 4482
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.941386097321264,
+      "learning_rate": 8.059639406732279e-06,
+      "loss": 0.4072,
+      "step": 4483
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.318385998131724,
+      "learning_rate": 8.058748745551308e-06,
+      "loss": 0.413,
+      "step": 4484
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.0834301788051945,
+      "learning_rate": 8.057857929240571e-06,
+      "loss": 0.4185,
+      "step": 4485
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.238349229300431,
+      "learning_rate": 8.056966957845251e-06,
+      "loss": 0.41,
+      "step": 4486
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.127123600443296,
+      "learning_rate": 8.056075831410533e-06,
+      "loss": 0.4543,
+      "step": 4487
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.0739845813506115,
+      "learning_rate": 8.055184549981611e-06,
+      "loss": 0.4105,
+      "step": 4488
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.221447747861587,
+      "learning_rate": 8.05429311360369e-06,
+      "loss": 0.4241,
+      "step": 4489
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.621090945336679,
+      "learning_rate": 8.05340152232198e-06,
+      "loss": 0.4202,
+      "step": 4490
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.225633256992187,
+      "learning_rate": 8.052509776181698e-06,
+      "loss": 0.4264,
+      "step": 4491
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.531641556627536,
+      "learning_rate": 8.051617875228073e-06,
+      "loss": 0.4363,
+      "step": 4492
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.2610606016098833,
+      "learning_rate": 8.05072581950634e-06,
+      "loss": 0.4188,
+      "step": 4493
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.862492840794233,
+      "learning_rate": 8.049833609061737e-06,
+      "loss": 0.4027,
+      "step": 4494
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.095231297128892,
+      "learning_rate": 8.048941243939516e-06,
+      "loss": 0.3814,
+      "step": 4495
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.072306432157796,
+      "learning_rate": 8.048048724184936e-06,
+      "loss": 0.416,
+      "step": 4496
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.319086629534689,
+      "learning_rate": 8.047156049843264e-06,
+      "loss": 0.4269,
+      "step": 4497
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.2579201961442275,
+      "learning_rate": 8.046263220959767e-06,
+      "loss": 0.4131,
+      "step": 4498
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.187911982793212,
+      "learning_rate": 8.045370237579735e-06,
+      "loss": 0.4032,
+      "step": 4499
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.45539044154913316,
+      "learning_rate": 8.044477099748451e-06,
+      "loss": 0.2111,
+      "step": 4500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.943739687731137,
+      "learning_rate": 8.043583807511215e-06,
+      "loss": 0.418,
+      "step": 4501
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 15.27839303599988,
+      "learning_rate": 8.04269036091333e-06,
+      "loss": 0.407,
+      "step": 4502
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.649856824973525,
+      "learning_rate": 8.041796760000111e-06,
+      "loss": 0.4114,
+      "step": 4503
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 11.161955960517748,
+      "learning_rate": 8.040903004816874e-06,
+      "loss": 0.3733,
+      "step": 4504
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.644472890453655,
+      "learning_rate": 8.040009095408953e-06,
+      "loss": 0.4366,
+      "step": 4505
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.975392537275934,
+      "learning_rate": 8.039115031821683e-06,
+      "loss": 0.4402,
+      "step": 4506
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.530789164046277,
+      "learning_rate": 8.038220814100403e-06,
+      "loss": 0.4111,
+      "step": 4507
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 24.34061863935547,
+      "learning_rate": 8.037326442290472e-06,
+      "loss": 0.3976,
+      "step": 4508
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.937246514680089,
+      "learning_rate": 8.036431916437244e-06,
+      "loss": 0.4345,
+      "step": 4509
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 9.49852593571815,
+      "learning_rate": 8.035537236586088e-06,
+      "loss": 0.4541,
+      "step": 4510
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.279643596554263,
+      "learning_rate": 8.034642402782382e-06,
+      "loss": 0.4177,
+      "step": 4511
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.7223947831985535,
+      "learning_rate": 8.033747415071507e-06,
+      "loss": 0.4316,
+      "step": 4512
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.204449223833681,
+      "learning_rate": 8.032852273498852e-06,
+      "loss": 0.4369,
+      "step": 4513
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.7102438036877405,
+      "learning_rate": 8.031956978109817e-06,
+      "loss": 0.3803,
+      "step": 4514
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.244878943665134,
+      "learning_rate": 8.03106152894981e-06,
+      "loss": 0.3867,
+      "step": 4515
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.261554187040325,
+      "learning_rate": 8.030165926064244e-06,
+      "loss": 0.418,
+      "step": 4516
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.70558832309189,
+      "learning_rate": 8.029270169498539e-06,
+      "loss": 0.4296,
+      "step": 4517
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.164175654501331,
+      "learning_rate": 8.02837425929813e-06,
+      "loss": 0.4325,
+      "step": 4518
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 12.204427901111325,
+      "learning_rate": 8.02747819550845e-06,
+      "loss": 0.4208,
+      "step": 4519
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.613659722759715,
+      "learning_rate": 8.026581978174947e-06,
+      "loss": 0.4231,
+      "step": 4520
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.4852523505510185,
+      "learning_rate": 8.025685607343073e-06,
+      "loss": 0.3905,
+      "step": 4521
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.655653910494423,
+      "learning_rate": 8.024789083058289e-06,
+      "loss": 0.4228,
+      "step": 4522
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.400250896750641,
+      "learning_rate": 8.023892405366065e-06,
+      "loss": 0.4098,
+      "step": 4523
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.247424047158755,
+      "learning_rate": 8.022995574311876e-06,
+      "loss": 0.4214,
+      "step": 4524
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.63646266434907,
+      "learning_rate": 8.022098589941206e-06,
+      "loss": 0.4144,
+      "step": 4525
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 7.269079551708297,
+      "learning_rate": 8.021201452299549e-06,
+      "loss": 0.3991,
+      "step": 4526
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.650631467935833,
+      "learning_rate": 8.020304161432404e-06,
+      "loss": 0.4063,
+      "step": 4527
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.901765047382441,
+      "learning_rate": 8.01940671738528e-06,
+      "loss": 0.4275,
+      "step": 4528
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.315090496066733,
+      "learning_rate": 8.018509120203688e-06,
+      "loss": 0.4091,
+      "step": 4529
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.3914745744342705,
+      "learning_rate": 8.017611369933157e-06,
+      "loss": 0.411,
+      "step": 4530
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.854436281202684,
+      "learning_rate": 8.016713466619213e-06,
+      "loss": 0.4268,
+      "step": 4531
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.054235388345701,
+      "learning_rate": 8.0158154103074e-06,
+      "loss": 0.4492,
+      "step": 4532
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7483172538512117,
+      "learning_rate": 8.01491720104326e-06,
+      "loss": 0.4336,
+      "step": 4533
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.470579776129392,
+      "learning_rate": 8.014018838872348e-06,
+      "loss": 0.4158,
+      "step": 4534
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.444886250516879,
+      "learning_rate": 8.013120323840228e-06,
+      "loss": 0.3894,
+      "step": 4535
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.4496090009367135,
+      "learning_rate": 8.012221655992469e-06,
+      "loss": 0.4321,
+      "step": 4536
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.117266116088185,
+      "learning_rate": 8.011322835374646e-06,
+      "loss": 0.3808,
+      "step": 4537
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.739827320863962,
+      "learning_rate": 8.010423862032347e-06,
+      "loss": 0.3976,
+      "step": 4538
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.2241316606525,
+      "learning_rate": 8.009524736011166e-06,
+      "loss": 0.4701,
+      "step": 4539
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.997628528396867,
+      "learning_rate": 8.0086254573567e-06,
+      "loss": 0.4211,
+      "step": 4540
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.134864506684886,
+      "learning_rate": 8.00772602611456e-06,
+      "loss": 0.3959,
+      "step": 4541
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4744752126580921,
+      "learning_rate": 8.006826442330362e-06,
+      "loss": 0.2231,
+      "step": 4542
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.463207352321749,
+      "learning_rate": 8.005926706049728e-06,
+      "loss": 0.4366,
+      "step": 4543
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.5639172104100805,
+      "learning_rate": 8.005026817318293e-06,
+      "loss": 0.3644,
+      "step": 4544
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.730604897210335,
+      "learning_rate": 8.004126776181695e-06,
+      "loss": 0.4061,
+      "step": 4545
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.2843514690465065,
+      "learning_rate": 8.003226582685582e-06,
+      "loss": 0.4192,
+      "step": 4546
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3592563709415886,
+      "learning_rate": 8.002326236875604e-06,
+      "loss": 0.2014,
+      "step": 4547
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.006512468403673,
+      "learning_rate": 8.00142573879743e-06,
+      "loss": 0.4269,
+      "step": 4548
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 14.489260816932514,
+      "learning_rate": 8.000525088496727e-06,
+      "loss": 0.4161,
+      "step": 4549
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 15.467813785004248,
+      "learning_rate": 7.999624286019175e-06,
+      "loss": 0.423,
+      "step": 4550
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7594969837956445,
+      "learning_rate": 7.998723331410458e-06,
+      "loss": 0.4256,
+      "step": 4551
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.76012940012515,
+      "learning_rate": 7.997822224716269e-06,
+      "loss": 0.4537,
+      "step": 4552
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.7144405584603017,
+      "learning_rate": 7.99692096598231e-06,
+      "loss": 0.434,
+      "step": 4553
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.470444365887784,
+      "learning_rate": 7.996019555254293e-06,
+      "loss": 0.3923,
+      "step": 4554
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.201858843335293,
+      "learning_rate": 7.99511799257793e-06,
+      "loss": 0.4528,
+      "step": 4555
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.994350338527562,
+      "learning_rate": 7.994216277998947e-06,
+      "loss": 0.4519,
+      "step": 4556
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.650393643894921,
+      "learning_rate": 7.993314411563075e-06,
+      "loss": 0.4121,
+      "step": 4557
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.045187756056979,
+      "learning_rate": 7.992412393316059e-06,
+      "loss": 0.4482,
+      "step": 4558
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.892111462580771,
+      "learning_rate": 7.991510223303637e-06,
+      "loss": 0.3989,
+      "step": 4559
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.361211830344237,
+      "learning_rate": 7.990607901571573e-06,
+      "loss": 0.4346,
+      "step": 4560
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.761363436151266,
+      "learning_rate": 7.989705428165623e-06,
+      "loss": 0.4203,
+      "step": 4561
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4071301842542088,
+      "learning_rate": 7.988802803131563e-06,
+      "loss": 0.1946,
+      "step": 4562
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.08428585762336,
+      "learning_rate": 7.987900026515167e-06,
+      "loss": 0.4284,
+      "step": 4563
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.756414869959762,
+      "learning_rate": 7.986997098362225e-06,
+      "loss": 0.4144,
+      "step": 4564
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.749290536746905,
+      "learning_rate": 7.986094018718527e-06,
+      "loss": 0.3908,
+      "step": 4565
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.8289083551610315,
+      "learning_rate": 7.985190787629877e-06,
+      "loss": 0.411,
+      "step": 4566
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.893057863878409,
+      "learning_rate": 7.98428740514208e-06,
+      "loss": 0.3986,
+      "step": 4567
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.134299075860259,
+      "learning_rate": 7.983383871300957e-06,
+      "loss": 0.4206,
+      "step": 4568
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 9.851454900097018,
+      "learning_rate": 7.98248018615233e-06,
+      "loss": 0.3917,
+      "step": 4569
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.918025581175296,
+      "learning_rate": 7.981576349742031e-06,
+      "loss": 0.4341,
+      "step": 4570
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.609954360597754,
+      "learning_rate": 7.9806723621159e-06,
+      "loss": 0.466,
+      "step": 4571
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4365548056321501,
+      "learning_rate": 7.979768223319786e-06,
+      "loss": 0.2157,
+      "step": 4572
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.374311339239909,
+      "learning_rate": 7.978863933399542e-06,
+      "loss": 0.4188,
+      "step": 4573
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.771190248157933,
+      "learning_rate": 7.97795949240103e-06,
+      "loss": 0.4539,
+      "step": 4574
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.377981425749226,
+      "learning_rate": 7.97705490037012e-06,
+      "loss": 0.4317,
+      "step": 4575
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.542401078235475,
+      "learning_rate": 7.976150157352694e-06,
+      "loss": 0.4247,
+      "step": 4576
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.17691415014888,
+      "learning_rate": 7.975245263394632e-06,
+      "loss": 0.423,
+      "step": 4577
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.681308214908187,
+      "learning_rate": 7.974340218541831e-06,
+      "loss": 0.4117,
+      "step": 4578
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.0056032109525335,
+      "learning_rate": 7.973435022840192e-06,
+      "loss": 0.4155,
+      "step": 4579
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.809982479395813,
+      "learning_rate": 7.972529676335623e-06,
+      "loss": 0.4112,
+      "step": 4580
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.519735410086932,
+      "learning_rate": 7.971624179074038e-06,
+      "loss": 0.407,
+      "step": 4581
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.2800101017813175,
+      "learning_rate": 7.970718531101365e-06,
+      "loss": 0.3964,
+      "step": 4582
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.38339997410032,
+      "learning_rate": 7.96981273246353e-06,
+      "loss": 0.443,
+      "step": 4583
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.886723080235653,
+      "learning_rate": 7.968906783206478e-06,
+      "loss": 0.4117,
+      "step": 4584
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.055677500462456,
+      "learning_rate": 7.968000683376152e-06,
+      "loss": 0.4297,
+      "step": 4585
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.3257443986793405,
+      "learning_rate": 7.967094433018508e-06,
+      "loss": 0.3527,
+      "step": 4586
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45623295480528936,
+      "learning_rate": 7.966188032179507e-06,
+      "loss": 0.2148,
+      "step": 4587
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5940914859701887,
+      "learning_rate": 7.96528148090512e-06,
+      "loss": 0.446,
+      "step": 4588
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.280636389251532,
+      "learning_rate": 7.964374779241324e-06,
+      "loss": 0.4367,
+      "step": 4589
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.108555508748928,
+      "learning_rate": 7.963467927234103e-06,
+      "loss": 0.4068,
+      "step": 4590
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.573784689695833,
+      "learning_rate": 7.96256092492945e-06,
+      "loss": 0.4481,
+      "step": 4591
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 12.671395482234226,
+      "learning_rate": 7.961653772373366e-06,
+      "loss": 0.4103,
+      "step": 4592
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.227167594739421,
+      "learning_rate": 7.960746469611858e-06,
+      "loss": 0.3762,
+      "step": 4593
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.379239991828135,
+      "learning_rate": 7.959839016690941e-06,
+      "loss": 0.4404,
+      "step": 4594
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.102249205581771,
+      "learning_rate": 7.95893141365664e-06,
+      "loss": 0.4028,
+      "step": 4595
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.052972102613796,
+      "learning_rate": 7.958023660554983e-06,
+      "loss": 0.4216,
+      "step": 4596
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.804160836563882,
+      "learning_rate": 7.95711575743201e-06,
+      "loss": 0.4008,
+      "step": 4597
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.997499165052432,
+      "learning_rate": 7.956207704333767e-06,
+      "loss": 0.3798,
+      "step": 4598
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.710124903101995,
+      "learning_rate": 7.955299501306308e-06,
+      "loss": 0.4205,
+      "step": 4599
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9730511932431027,
+      "learning_rate": 7.954391148395692e-06,
+      "loss": 0.4253,
+      "step": 4600
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.631386008344835,
+      "learning_rate": 7.953482645647988e-06,
+      "loss": 0.3905,
+      "step": 4601
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.524548568143925,
+      "learning_rate": 7.952573993109273e-06,
+      "loss": 0.4006,
+      "step": 4602
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.181858643539375,
+      "learning_rate": 7.951665190825633e-06,
+      "loss": 0.3934,
+      "step": 4603
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 10.558450916503626,
+      "learning_rate": 7.950756238843155e-06,
+      "loss": 0.4662,
+      "step": 4604
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.4591061760053,
+      "learning_rate": 7.949847137207945e-06,
+      "loss": 0.4321,
+      "step": 4605
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.008963290935224,
+      "learning_rate": 7.9489378859661e-06,
+      "loss": 0.3727,
+      "step": 4606
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.845877041071682,
+      "learning_rate": 7.948028485163744e-06,
+      "loss": 0.432,
+      "step": 4607
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.788353160419769,
+      "learning_rate": 7.94711893484699e-06,
+      "loss": 0.406,
+      "step": 4608
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.107607554850754,
+      "learning_rate": 7.946209235061975e-06,
+      "loss": 0.4216,
+      "step": 4609
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.514639056776697,
+      "learning_rate": 7.945299385854831e-06,
+      "loss": 0.3991,
+      "step": 4610
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9424038209710135,
+      "learning_rate": 7.944389387271706e-06,
+      "loss": 0.4108,
+      "step": 4611
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.185742925162904,
+      "learning_rate": 7.94347923935875e-06,
+      "loss": 0.4288,
+      "step": 4612
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.983568672297312,
+      "learning_rate": 7.942568942162123e-06,
+      "loss": 0.3842,
+      "step": 4613
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.265623655507664,
+      "learning_rate": 7.941658495727991e-06,
+      "loss": 0.3976,
+      "step": 4614
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.131371063199334,
+      "learning_rate": 7.940747900102533e-06,
+      "loss": 0.4214,
+      "step": 4615
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.031816609717073,
+      "learning_rate": 7.939837155331927e-06,
+      "loss": 0.4389,
+      "step": 4616
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.039043553458549,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.3867,
+      "step": 4617
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 61.04051802527188,
+      "learning_rate": 7.938015218540047e-06,
+      "loss": 0.4078,
+      "step": 4618
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.14004551885247,
+      "learning_rate": 7.937104026611173e-06,
+      "loss": 0.4294,
+      "step": 4619
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4275622426409481,
+      "learning_rate": 7.93619268572196e-06,
+      "loss": 0.2088,
+      "step": 4620
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.503882701290631,
+      "learning_rate": 7.935281195918626e-06,
+      "loss": 0.402,
+      "step": 4621
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.893852140919412,
+      "learning_rate": 7.934369557247397e-06,
+      "loss": 0.4095,
+      "step": 4622
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.757283282529613,
+      "learning_rate": 7.933457769754513e-06,
+      "loss": 0.4091,
+      "step": 4623
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.492064715344707,
+      "learning_rate": 7.932545833486214e-06,
+      "loss": 0.3949,
+      "step": 4624
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3770417465576003,
+      "learning_rate": 7.931633748488752e-06,
+      "loss": 0.2,
+      "step": 4625
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.4600313019153255,
+      "learning_rate": 7.930721514808383e-06,
+      "loss": 0.4016,
+      "step": 4626
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.415812350395612,
+      "learning_rate": 7.929809132491374e-06,
+      "loss": 0.4248,
+      "step": 4627
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.403939385485686,
+      "learning_rate": 7.928896601583996e-06,
+      "loss": 0.4469,
+      "step": 4628
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.270418505174667,
+      "learning_rate": 7.927983922132533e-06,
+      "loss": 0.419,
+      "step": 4629
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.219184918705042,
+      "learning_rate": 7.92707109418327e-06,
+      "loss": 0.4057,
+      "step": 4630
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.311631483593866,
+      "learning_rate": 7.926158117782505e-06,
+      "loss": 0.4189,
+      "step": 4631
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 7.172643705481688,
+      "learning_rate": 7.925244992976538e-06,
+      "loss": 0.3722,
+      "step": 4632
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.522377186921123,
+      "learning_rate": 7.924331719811685e-06,
+      "loss": 0.4621,
+      "step": 4633
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.038419288574115,
+      "learning_rate": 7.923418298334257e-06,
+      "loss": 0.3825,
+      "step": 4634
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.031695586176827,
+      "learning_rate": 7.922504728590587e-06,
+      "loss": 0.3855,
+      "step": 4635
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.39086602841304,
+      "learning_rate": 7.921591010627005e-06,
+      "loss": 0.3904,
+      "step": 4636
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.321282865840615,
+      "learning_rate": 7.92067714448985e-06,
+      "loss": 0.3785,
+      "step": 4637
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.06480497524211,
+      "learning_rate": 7.919763130225473e-06,
+      "loss": 0.3847,
+      "step": 4638
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.906161447692416,
+      "learning_rate": 7.918848967880231e-06,
+      "loss": 0.4503,
+      "step": 4639
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 49.720844749116935,
+      "learning_rate": 7.917934657500483e-06,
+      "loss": 0.4088,
+      "step": 4640
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 13.170562267243803,
+      "learning_rate": 7.917020199132603e-06,
+      "loss": 0.4369,
+      "step": 4641
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.9565648812838456,
+      "learning_rate": 7.916105592822966e-06,
+      "loss": 0.4204,
+      "step": 4642
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.65325636087471,
+      "learning_rate": 7.915190838617964e-06,
+      "loss": 0.4495,
+      "step": 4643
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.43563421811186676,
+      "learning_rate": 7.914275936563986e-06,
+      "loss": 0.2021,
+      "step": 4644
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.439184413793307,
+      "learning_rate": 7.913360886707434e-06,
+      "loss": 0.4104,
+      "step": 4645
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.587951960679621,
+      "learning_rate": 7.912445689094715e-06,
+      "loss": 0.4077,
+      "step": 4646
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.208506296073501,
+      "learning_rate": 7.911530343772244e-06,
+      "loss": 0.4165,
+      "step": 4647
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.7860287168671825,
+      "learning_rate": 7.910614850786448e-06,
+      "loss": 0.4047,
+      "step": 4648
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.104920184653892,
+      "learning_rate": 7.909699210183757e-06,
+      "loss": 0.4459,
+      "step": 4649
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.847745589182942,
+      "learning_rate": 7.908783422010605e-06,
+      "loss": 0.4124,
+      "step": 4650
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 8.265631617014122,
+      "learning_rate": 7.907867486313444e-06,
+      "loss": 0.4352,
+      "step": 4651
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.577332613580643,
+      "learning_rate": 7.906951403138723e-06,
+      "loss": 0.3664,
+      "step": 4652
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.680772032397982,
+      "learning_rate": 7.906035172532904e-06,
+      "loss": 0.4223,
+      "step": 4653
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.5598809930839694,
+      "learning_rate": 7.905118794542455e-06,
+      "loss": 0.4139,
+      "step": 4654
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.660452056892742,
+      "learning_rate": 7.904202269213854e-06,
+      "loss": 0.4284,
+      "step": 4655
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.03600243656392,
+      "learning_rate": 7.903285596593579e-06,
+      "loss": 0.3984,
+      "step": 4656
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42728862773059045,
+      "learning_rate": 7.902368776728125e-06,
+      "loss": 0.2078,
+      "step": 4657
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.2079272478051415,
+      "learning_rate": 7.90145180966399e-06,
+      "loss": 0.4027,
+      "step": 4658
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.175217861234775,
+      "learning_rate": 7.900534695447677e-06,
+      "loss": 0.4182,
+      "step": 4659
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.8138201439218307,
+      "learning_rate": 7.899617434125703e-06,
+      "loss": 0.4129,
+      "step": 4660
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.191919648593301,
+      "learning_rate": 7.898700025744583e-06,
+      "loss": 0.4202,
+      "step": 4661
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.508012046560773,
+      "learning_rate": 7.89778247035085e-06,
+      "loss": 0.3937,
+      "step": 4662
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.818641377888404,
+      "learning_rate": 7.896864767991037e-06,
+      "loss": 0.4348,
+      "step": 4663
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.670631844178356,
+      "learning_rate": 7.895946918711688e-06,
+      "loss": 0.414,
+      "step": 4664
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.222748230770759,
+      "learning_rate": 7.895028922559351e-06,
+      "loss": 0.4309,
+      "step": 4665
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.89806091432922,
+      "learning_rate": 7.894110779580587e-06,
+      "loss": 0.409,
+      "step": 4666
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.1464120029745923,
+      "learning_rate": 7.893192489821958e-06,
+      "loss": 0.3996,
+      "step": 4667
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.7287056147285895,
+      "learning_rate": 7.89227405333004e-06,
+      "loss": 0.3771,
+      "step": 4668
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.668613476444213,
+      "learning_rate": 7.89135547015141e-06,
+      "loss": 0.4283,
+      "step": 4669
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.605360075237625,
+      "learning_rate": 7.890436740332656e-06,
+      "loss": 0.4284,
+      "step": 4670
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.451991153985221,
+      "learning_rate": 7.889517863920375e-06,
+      "loss": 0.3951,
+      "step": 4671
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 5.272056021363996,
+      "learning_rate": 7.888598840961168e-06,
+      "loss": 0.4019,
+      "step": 4672
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.019036667036754,
+      "learning_rate": 7.887679671501647e-06,
+      "loss": 0.3905,
+      "step": 4673
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 4.798495583646222,
+      "learning_rate": 7.886760355588425e-06,
+      "loss": 0.3832,
+      "step": 4674
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.2089544083685455,
+      "learning_rate": 7.88584089326813e-06,
+      "loss": 0.4221,
+      "step": 4675
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.593434031497945,
+      "learning_rate": 7.884921284587394e-06,
+      "loss": 0.4331,
+      "step": 4676
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.691391229026682,
+      "learning_rate": 7.884001529592855e-06,
+      "loss": 0.4397,
+      "step": 4677
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.149672738928196,
+      "learning_rate": 7.883081628331162e-06,
+      "loss": 0.4309,
+      "step": 4678
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.900351484149109,
+      "learning_rate": 7.882161580848966e-06,
+      "loss": 0.4061,
+      "step": 4679
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.4523826403892586,
+      "learning_rate": 7.881241387192933e-06,
+      "loss": 0.4123,
+      "step": 4680
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.569524012075366,
+      "learning_rate": 7.880321047409729e-06,
+      "loss": 0.433,
+      "step": 4681
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.6335183237065003,
+      "learning_rate": 7.879400561546033e-06,
+      "loss": 0.4658,
+      "step": 4682
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.398296225421954,
+      "learning_rate": 7.878479929648528e-06,
+      "loss": 0.39,
+      "step": 4683
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.745669256848296,
+      "learning_rate": 7.877559151763905e-06,
+      "loss": 0.4886,
+      "step": 4684
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.181424284364341,
+      "learning_rate": 7.876638227938862e-06,
+      "loss": 0.4062,
+      "step": 4685
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.837742446682897,
+      "learning_rate": 7.875717158220108e-06,
+      "loss": 0.4738,
+      "step": 4686
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.984365657110778,
+      "learning_rate": 7.874795942654354e-06,
+      "loss": 0.4046,
+      "step": 4687
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3871824402935495,
+      "learning_rate": 7.873874581288323e-06,
+      "loss": 0.1953,
+      "step": 4688
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 11.334861516743343,
+      "learning_rate": 7.872953074168741e-06,
+      "loss": 0.399,
+      "step": 4689
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.6872639970710575,
+      "learning_rate": 7.872031421342348e-06,
+      "loss": 0.4078,
+      "step": 4690
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.603872647486251,
+      "learning_rate": 7.871109622855882e-06,
+      "loss": 0.359,
+      "step": 4691
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.482740131323222,
+      "learning_rate": 7.870187678756099e-06,
+      "loss": 0.4363,
+      "step": 4692
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.267192436029936,
+      "learning_rate": 7.869265589089751e-06,
+      "loss": 0.3841,
+      "step": 4693
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.091334818262108,
+      "learning_rate": 7.868343353903611e-06,
+      "loss": 0.4277,
+      "step": 4694
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3699136616146399,
+      "learning_rate": 7.867420973244444e-06,
+      "loss": 0.209,
+      "step": 4695
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9129687693528052,
+      "learning_rate": 7.866498447159035e-06,
+      "loss": 0.3777,
+      "step": 4696
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.217250244450068,
+      "learning_rate": 7.86557577569417e-06,
+      "loss": 0.4112,
+      "step": 4697
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 10.544468789491514,
+      "learning_rate": 7.864652958896643e-06,
+      "loss": 0.3999,
+      "step": 4698
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.13247808486997,
+      "learning_rate": 7.863729996813258e-06,
+      "loss": 0.3844,
+      "step": 4699
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.45307144276865,
+      "learning_rate": 7.862806889490824e-06,
+      "loss": 0.4336,
+      "step": 4700
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.822787466639939,
+      "learning_rate": 7.861883636976158e-06,
+      "loss": 0.4047,
+      "step": 4701
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.887858271123595,
+      "learning_rate": 7.860960239316083e-06,
+      "loss": 0.3785,
+      "step": 4702
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.3924356743191355,
+      "learning_rate": 7.860036696557432e-06,
+      "loss": 0.4263,
+      "step": 4703
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.020329059953276,
+      "learning_rate": 7.859113008747046e-06,
+      "loss": 0.4032,
+      "step": 4704
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.623459316356217,
+      "learning_rate": 7.858189175931767e-06,
+      "loss": 0.3939,
+      "step": 4705
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.728464176528977,
+      "learning_rate": 7.857265198158453e-06,
+      "loss": 0.4342,
+      "step": 4706
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.113036821102433,
+      "learning_rate": 7.856341075473963e-06,
+      "loss": 0.395,
+      "step": 4707
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.001552782172715,
+      "learning_rate": 7.855416807925165e-06,
+      "loss": 0.4133,
+      "step": 4708
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.5934881339032145,
+      "learning_rate": 7.854492395558935e-06,
+      "loss": 0.4166,
+      "step": 4709
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.5439620985420195,
+      "learning_rate": 7.85356783842216e-06,
+      "loss": 0.4308,
+      "step": 4710
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.284024947125479,
+      "learning_rate": 7.852643136561724e-06,
+      "loss": 0.3895,
+      "step": 4711
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.8608831581812035,
+      "learning_rate": 7.85171829002453e-06,
+      "loss": 0.3851,
+      "step": 4712
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.7375410233160276,
+      "learning_rate": 7.850793298857481e-06,
+      "loss": 0.4258,
+      "step": 4713
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9305413971103964,
+      "learning_rate": 7.849868163107492e-06,
+      "loss": 0.4085,
+      "step": 4714
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.208886568743173,
+      "learning_rate": 7.848942882821479e-06,
+      "loss": 0.3987,
+      "step": 4715
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.1366570901558815,
+      "learning_rate": 7.848017458046371e-06,
+      "loss": 0.3469,
+      "step": 4716
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.034573646910651,
+      "learning_rate": 7.847091888829102e-06,
+      "loss": 0.4383,
+      "step": 4717
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.625594186391808,
+      "learning_rate": 7.846166175216617e-06,
+      "loss": 0.3988,
+      "step": 4718
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.1593918529104625,
+      "learning_rate": 7.84524031725586e-06,
+      "loss": 0.4166,
+      "step": 4719
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.31046756892374,
+      "learning_rate": 7.844314314993791e-06,
+      "loss": 0.4125,
+      "step": 4720
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38359659757897585,
+      "learning_rate": 7.843388168477373e-06,
+      "loss": 0.2082,
+      "step": 4721
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.627400597437153,
+      "learning_rate": 7.842461877753575e-06,
+      "loss": 0.4042,
+      "step": 4722
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.918913284271235,
+      "learning_rate": 7.841535442869381e-06,
+      "loss": 0.4254,
+      "step": 4723
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.624000925603266,
+      "learning_rate": 7.840608863871772e-06,
+      "loss": 0.3708,
+      "step": 4724
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.9594113427088695,
+      "learning_rate": 7.839682140807743e-06,
+      "loss": 0.3917,
+      "step": 4725
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.59160522841482,
+      "learning_rate": 7.838755273724293e-06,
+      "loss": 0.4091,
+      "step": 4726
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.4654955192201715,
+      "learning_rate": 7.83782826266843e-06,
+      "loss": 0.378,
+      "step": 4727
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.05417009517667,
+      "learning_rate": 7.83690110768717e-06,
+      "loss": 0.4075,
+      "step": 4728
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.902182655944951,
+      "learning_rate": 7.835973808827533e-06,
+      "loss": 0.4206,
+      "step": 4729
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.232156797725697,
+      "learning_rate": 7.835046366136552e-06,
+      "loss": 0.4395,
+      "step": 4730
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.436762894648998,
+      "learning_rate": 7.834118779661262e-06,
+      "loss": 0.43,
+      "step": 4731
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 35.50800271398428,
+      "learning_rate": 7.833191049448706e-06,
+      "loss": 0.3953,
+      "step": 4732
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.722052803038593,
+      "learning_rate": 7.832263175545938e-06,
+      "loss": 0.431,
+      "step": 4733
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.195596601497676,
+      "learning_rate": 7.831335158000015e-06,
+      "loss": 0.4136,
+      "step": 4734
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.0866428943213275,
+      "learning_rate": 7.830406996858004e-06,
+      "loss": 0.3931,
+      "step": 4735
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.792979933532434,
+      "learning_rate": 7.829478692166976e-06,
+      "loss": 0.435,
+      "step": 4736
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.685735875981046,
+      "learning_rate": 7.828550243974015e-06,
+      "loss": 0.4075,
+      "step": 4737
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.861294230243869,
+      "learning_rate": 7.827621652326207e-06,
+      "loss": 0.3941,
+      "step": 4738
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.802913017267737,
+      "learning_rate": 7.826692917270648e-06,
+      "loss": 0.4222,
+      "step": 4739
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.697457133118721,
+      "learning_rate": 7.825764038854438e-06,
+      "loss": 0.4182,
+      "step": 4740
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.4890250893784875,
+      "learning_rate": 7.82483501712469e-06,
+      "loss": 0.4318,
+      "step": 4741
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.394348164170169,
+      "learning_rate": 7.82390585212852e-06,
+      "loss": 0.4175,
+      "step": 4742
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.770985826494247,
+      "learning_rate": 7.82297654391305e-06,
+      "loss": 0.413,
+      "step": 4743
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.894896210046695,
+      "learning_rate": 7.822047092525413e-06,
+      "loss": 0.4065,
+      "step": 4744
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.936692489307726,
+      "learning_rate": 7.821117498012749e-06,
+      "loss": 0.463,
+      "step": 4745
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.457994572092795,
+      "learning_rate": 7.820187760422202e-06,
+      "loss": 0.4209,
+      "step": 4746
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.809222559750385,
+      "learning_rate": 7.819257879800927e-06,
+      "loss": 0.4345,
+      "step": 4747
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.405341931023856,
+      "learning_rate": 7.818327856196083e-06,
+      "loss": 0.414,
+      "step": 4748
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.495556688096118,
+      "learning_rate": 7.81739768965484e-06,
+      "loss": 0.4453,
+      "step": 4749
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.7625394502516425,
+      "learning_rate": 7.81646738022437e-06,
+      "loss": 0.3983,
+      "step": 4750
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.5323408266358145,
+      "learning_rate": 7.815536927951856e-06,
+      "loss": 0.4159,
+      "step": 4751
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.681286571614341,
+      "learning_rate": 7.81460633288449e-06,
+      "loss": 0.4296,
+      "step": 4752
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.789926216626468,
+      "learning_rate": 7.813675595069466e-06,
+      "loss": 0.431,
+      "step": 4753
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.462252058587235,
+      "learning_rate": 7.812744714553988e-06,
+      "loss": 0.4031,
+      "step": 4754
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.652962886869151,
+      "learning_rate": 7.811813691385269e-06,
+      "loss": 0.373,
+      "step": 4755
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.56267982566457,
+      "learning_rate": 7.810882525610526e-06,
+      "loss": 0.4225,
+      "step": 4756
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.856037179876042,
+      "learning_rate": 7.809951217276986e-06,
+      "loss": 0.4544,
+      "step": 4757
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.712292220815318,
+      "learning_rate": 7.809019766431882e-06,
+      "loss": 0.4016,
+      "step": 4758
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.653268111360772,
+      "learning_rate": 7.808088173122453e-06,
+      "loss": 0.4109,
+      "step": 4759
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.078834547984309,
+      "learning_rate": 7.807156437395944e-06,
+      "loss": 0.4157,
+      "step": 4760
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.424362176977293,
+      "learning_rate": 7.806224559299615e-06,
+      "loss": 0.3863,
+      "step": 4761
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.904189715231947,
+      "learning_rate": 7.805292538880724e-06,
+      "loss": 0.4095,
+      "step": 4762
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.072476662594974,
+      "learning_rate": 7.804360376186544e-06,
+      "loss": 0.4012,
+      "step": 4763
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 36.55634710400196,
+      "learning_rate": 7.803428071264345e-06,
+      "loss": 0.4243,
+      "step": 4764
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.268643645198272,
+      "learning_rate": 7.802495624161416e-06,
+      "loss": 0.3924,
+      "step": 4765
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.831314107200279,
+      "learning_rate": 7.801563034925044e-06,
+      "loss": 0.4401,
+      "step": 4766
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 17.05669511354116,
+      "learning_rate": 7.800630303602529e-06,
+      "loss": 0.3897,
+      "step": 4767
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.784632692328748,
+      "learning_rate": 7.799697430241178e-06,
+      "loss": 0.3935,
+      "step": 4768
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 22.161318104073988,
+      "learning_rate": 7.798764414888298e-06,
+      "loss": 0.4129,
+      "step": 4769
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 58.87938050663186,
+      "learning_rate": 7.797831257591213e-06,
+      "loss": 0.4253,
+      "step": 4770
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.341113368067874,
+      "learning_rate": 7.796897958397248e-06,
+      "loss": 0.4386,
+      "step": 4771
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.181882501803374,
+      "learning_rate": 7.795964517353734e-06,
+      "loss": 0.416,
+      "step": 4772
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.114367550965966,
+      "learning_rate": 7.79503093450802e-06,
+      "loss": 0.4411,
+      "step": 4773
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.03380951187569,
+      "learning_rate": 7.794097209907445e-06,
+      "loss": 0.411,
+      "step": 4774
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.384961400541184,
+      "learning_rate": 7.79316334359937e-06,
+      "loss": 0.4292,
+      "step": 4775
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.412680830049623,
+      "learning_rate": 7.792229335631158e-06,
+      "loss": 0.4297,
+      "step": 4776
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.73307638061394,
+      "learning_rate": 7.791295186050176e-06,
+      "loss": 0.3997,
+      "step": 4777
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.712730419950342,
+      "learning_rate": 7.790360894903802e-06,
+      "loss": 0.3859,
+      "step": 4778
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.323129421159086,
+      "learning_rate": 7.789426462239421e-06,
+      "loss": 0.4457,
+      "step": 4779
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.590483929677022,
+      "learning_rate": 7.788491888104423e-06,
+      "loss": 0.3778,
+      "step": 4780
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.937190508026869,
+      "learning_rate": 7.787557172546209e-06,
+      "loss": 0.415,
+      "step": 4781
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 36.85540686799523,
+      "learning_rate": 7.786622315612182e-06,
+      "loss": 0.4075,
+      "step": 4782
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.454695965045875,
+      "learning_rate": 7.785687317349757e-06,
+      "loss": 0.3825,
+      "step": 4783
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 8.582451922381948,
+      "learning_rate": 7.784752177806353e-06,
+      "loss": 0.3939,
+      "step": 4784
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.3587063949174016,
+      "learning_rate": 7.783816897029398e-06,
+      "loss": 0.3776,
+      "step": 4785
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4393466399739173,
+      "learning_rate": 7.782881475066324e-06,
+      "loss": 0.2025,
+      "step": 4786
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.444827700850803,
+      "learning_rate": 7.781945911964574e-06,
+      "loss": 0.4348,
+      "step": 4787
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.984617809085269,
+      "learning_rate": 7.7810102077716e-06,
+      "loss": 0.4178,
+      "step": 4788
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.193508817539221,
+      "learning_rate": 7.780074362534852e-06,
+      "loss": 0.4391,
+      "step": 4789
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.438149824703558,
+      "learning_rate": 7.779138376301797e-06,
+      "loss": 0.4188,
+      "step": 4790
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.801775149627759,
+      "learning_rate": 7.778202249119905e-06,
+      "loss": 0.4207,
+      "step": 4791
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.066962074811597,
+      "learning_rate": 7.77726598103665e-06,
+      "loss": 0.4285,
+      "step": 4792
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.164393790890697,
+      "learning_rate": 7.77632957209952e-06,
+      "loss": 0.4401,
+      "step": 4793
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.794181048495882,
+      "learning_rate": 7.775393022356006e-06,
+      "loss": 0.4312,
+      "step": 4794
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.03395629053253,
+      "learning_rate": 7.774456331853607e-06,
+      "loss": 0.3914,
+      "step": 4795
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.77992023314334,
+      "learning_rate": 7.773519500639826e-06,
+      "loss": 0.4188,
+      "step": 4796
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.290151507985233,
+      "learning_rate": 7.772582528762179e-06,
+      "loss": 0.4436,
+      "step": 4797
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.094385821548524,
+      "learning_rate": 7.771645416268186e-06,
+      "loss": 0.4107,
+      "step": 4798
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3902528704964731,
+      "learning_rate": 7.770708163205372e-06,
+      "loss": 0.2083,
+      "step": 4799
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 13.61469277156377,
+      "learning_rate": 7.769770769621274e-06,
+      "loss": 0.3907,
+      "step": 4800
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 9.417554665426158,
+      "learning_rate": 7.768833235563433e-06,
+      "loss": 0.4342,
+      "step": 4801
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.732883375882914,
+      "learning_rate": 7.767895561079397e-06,
+      "loss": 0.4033,
+      "step": 4802
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.252229380609423,
+      "learning_rate": 7.76695774621672e-06,
+      "loss": 0.3873,
+      "step": 4803
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 3.9379852699653575,
+      "learning_rate": 7.76601979102297e-06,
+      "loss": 0.3772,
+      "step": 4804
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.387244912505345,
+      "learning_rate": 7.765081695545712e-06,
+      "loss": 0.4127,
+      "step": 4805
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 5.447503843270306,
+      "learning_rate": 7.764143459832527e-06,
+      "loss": 0.38,
+      "step": 4806
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.7942246241594795,
+      "learning_rate": 7.763205083930995e-06,
+      "loss": 0.418,
+      "step": 4807
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.408091912507589,
+      "learning_rate": 7.762266567888712e-06,
+      "loss": 0.1969,
+      "step": 4808
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.803809374248124,
+      "learning_rate": 7.761327911753271e-06,
+      "loss": 0.4529,
+      "step": 4809
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.38145047874230165,
+      "learning_rate": 7.760389115572283e-06,
+      "loss": 0.1993,
+      "step": 4810
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.418671794960715,
+      "learning_rate": 7.759450179393358e-06,
+      "loss": 0.4258,
+      "step": 4811
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.650336958552111,
+      "learning_rate": 7.758511103264116e-06,
+      "loss": 0.4468,
+      "step": 4812
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 12.904515406616273,
+      "learning_rate": 7.757571887232185e-06,
+      "loss": 0.4458,
+      "step": 4813
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.608964016917629,
+      "learning_rate": 7.756632531345198e-06,
+      "loss": 0.3777,
+      "step": 4814
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.2490014047115805,
+      "learning_rate": 7.755693035650794e-06,
+      "loss": 0.4035,
+      "step": 4815
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.015538955546431,
+      "learning_rate": 7.754753400196627e-06,
+      "loss": 0.3852,
+      "step": 4816
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 4.812717626656842,
+      "learning_rate": 7.753813625030344e-06,
+      "loss": 0.3764,
+      "step": 4817
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.322039298606034,
+      "learning_rate": 7.752873710199616e-06,
+      "loss": 0.4276,
+      "step": 4818
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.8104250980393015,
+      "learning_rate": 7.751933655752106e-06,
+      "loss": 0.393,
+      "step": 4819
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7802037307524508,
+      "learning_rate": 7.750993461735495e-06,
+      "loss": 0.4031,
+      "step": 4820
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.391186071053815,
+      "learning_rate": 7.750053128197461e-06,
+      "loss": 0.4688,
+      "step": 4821
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.225321327451643,
+      "learning_rate": 7.7491126551857e-06,
+      "loss": 0.4048,
+      "step": 4822
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.974425364962373,
+      "learning_rate": 7.748172042747908e-06,
+      "loss": 0.407,
+      "step": 4823
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.148021832625603,
+      "learning_rate": 7.74723129093179e-06,
+      "loss": 0.3951,
+      "step": 4824
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.710969519240005,
+      "learning_rate": 7.746290399785055e-06,
+      "loss": 0.4135,
+      "step": 4825
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.760591945487501,
+      "learning_rate": 7.745349369355426e-06,
+      "loss": 0.3787,
+      "step": 4826
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7310787091180595,
+      "learning_rate": 7.744408199690628e-06,
+      "loss": 0.4087,
+      "step": 4827
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.4610227940589,
+      "learning_rate": 7.743466890838393e-06,
+      "loss": 0.3497,
+      "step": 4828
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.247705048724483,
+      "learning_rate": 7.742525442846461e-06,
+      "loss": 0.4067,
+      "step": 4829
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.087801679939327,
+      "learning_rate": 7.74158385576258e-06,
+      "loss": 0.4138,
+      "step": 4830
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.739851955732788,
+      "learning_rate": 7.740642129634503e-06,
+      "loss": 0.4166,
+      "step": 4831
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.546872195728064,
+      "learning_rate": 7.739700264509993e-06,
+      "loss": 0.414,
+      "step": 4832
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.19398227061394,
+      "learning_rate": 7.738758260436817e-06,
+      "loss": 0.443,
+      "step": 4833
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.827118337644696,
+      "learning_rate": 7.737816117462752e-06,
+      "loss": 0.435,
+      "step": 4834
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.126197242804143,
+      "learning_rate": 7.736873835635578e-06,
+      "loss": 0.4016,
+      "step": 4835
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.283762117280875,
+      "learning_rate": 7.73593141500309e-06,
+      "loss": 0.4266,
+      "step": 4836
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.368638934509039,
+      "learning_rate": 7.734988855613074e-06,
+      "loss": 0.4122,
+      "step": 4837
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.888399726818766,
+      "learning_rate": 7.734046157513342e-06,
+      "loss": 0.448,
+      "step": 4838
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.4133192467775455,
+      "learning_rate": 7.733103320751703e-06,
+      "loss": 0.4372,
+      "step": 4839
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.774550986496163,
+      "learning_rate": 7.732160345375974e-06,
+      "loss": 0.3953,
+      "step": 4840
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4710962854501215,
+      "learning_rate": 7.73121723143398e-06,
+      "loss": 0.398,
+      "step": 4841
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.190868034596277,
+      "learning_rate": 7.730273978973552e-06,
+      "loss": 0.4279,
+      "step": 4842
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.762206899018335,
+      "learning_rate": 7.72933058804253e-06,
+      "loss": 0.3801,
+      "step": 4843
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.049077743366745,
+      "learning_rate": 7.728387058688756e-06,
+      "loss": 0.4061,
+      "step": 4844
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.944201252791933,
+      "learning_rate": 7.727443390960087e-06,
+      "loss": 0.4238,
+      "step": 4845
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.029446112019142,
+      "learning_rate": 7.72649958490438e-06,
+      "loss": 0.4269,
+      "step": 4846
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.420684089568865,
+      "learning_rate": 7.725555640569506e-06,
+      "loss": 0.4696,
+      "step": 4847
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.745984346231855,
+      "learning_rate": 7.724611558003333e-06,
+      "loss": 0.456,
+      "step": 4848
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.747773995513522,
+      "learning_rate": 7.723667337253745e-06,
+      "loss": 0.4255,
+      "step": 4849
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.077277860335212,
+      "learning_rate": 7.722722978368631e-06,
+      "loss": 0.4027,
+      "step": 4850
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.984671513536532,
+      "learning_rate": 7.721778481395883e-06,
+      "loss": 0.4171,
+      "step": 4851
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3068338858709323,
+      "learning_rate": 7.720833846383403e-06,
+      "loss": 0.3866,
+      "step": 4852
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.166312610498274,
+      "learning_rate": 7.719889073379102e-06,
+      "loss": 0.4233,
+      "step": 4853
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2856175874162843,
+      "learning_rate": 7.718944162430894e-06,
+      "loss": 0.4045,
+      "step": 4854
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.301598493207437,
+      "learning_rate": 7.717999113586703e-06,
+      "loss": 0.4141,
+      "step": 4855
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.881711987651465,
+      "learning_rate": 7.717053926894458e-06,
+      "loss": 0.4292,
+      "step": 4856
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.255198339706457,
+      "learning_rate": 7.716108602402094e-06,
+      "loss": 0.402,
+      "step": 4857
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5067819694465272,
+      "learning_rate": 7.715163140157558e-06,
+      "loss": 0.2118,
+      "step": 4858
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.149534796691238,
+      "learning_rate": 7.714217540208802e-06,
+      "loss": 0.4152,
+      "step": 4859
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9212506114568852,
+      "learning_rate": 7.713271802603777e-06,
+      "loss": 0.3802,
+      "step": 4860
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.271822270080452,
+      "learning_rate": 7.712325927390454e-06,
+      "loss": 0.4632,
+      "step": 4861
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3623895869392957,
+      "learning_rate": 7.711379914616803e-06,
+      "loss": 0.443,
+      "step": 4862
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.34668631999048,
+      "learning_rate": 7.710433764330802e-06,
+      "loss": 0.4243,
+      "step": 4863
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.340153675521535,
+      "learning_rate": 7.709487476580436e-06,
+      "loss": 0.415,
+      "step": 4864
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.345673769287091,
+      "learning_rate": 7.7085410514137e-06,
+      "loss": 0.4241,
+      "step": 4865
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.052020580013375,
+      "learning_rate": 7.707594488878592e-06,
+      "loss": 0.42,
+      "step": 4866
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.203273827636585,
+      "learning_rate": 7.706647789023119e-06,
+      "loss": 0.4335,
+      "step": 4867
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.49213660101091583,
+      "learning_rate": 7.705700951895292e-06,
+      "loss": 0.2066,
+      "step": 4868
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.236753427375236,
+      "learning_rate": 7.704753977543138e-06,
+      "loss": 0.4453,
+      "step": 4869
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.3047102063869387,
+      "learning_rate": 7.703806866014677e-06,
+      "loss": 0.4097,
+      "step": 4870
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.658942399121796,
+      "learning_rate": 7.702859617357948e-06,
+      "loss": 0.4431,
+      "step": 4871
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.37858976350148127,
+      "learning_rate": 7.70191223162099e-06,
+      "loss": 0.2114,
+      "step": 4872
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.172672091820316,
+      "learning_rate": 7.700964708851852e-06,
+      "loss": 0.4051,
+      "step": 4873
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.38898238705340066,
+      "learning_rate": 7.70001704909859e-06,
+      "loss": 0.2015,
+      "step": 4874
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.051326453463482,
+      "learning_rate": 7.699069252409268e-06,
+      "loss": 0.4026,
+      "step": 4875
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.6319993637446952,
+      "learning_rate": 7.698121318831952e-06,
+      "loss": 0.3868,
+      "step": 4876
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.883808723951309,
+      "learning_rate": 7.697173248414718e-06,
+      "loss": 0.4025,
+      "step": 4877
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.057542275990353,
+      "learning_rate": 7.69622504120565e-06,
+      "loss": 0.4271,
+      "step": 4878
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.7837470201479046,
+      "learning_rate": 7.69527669725284e-06,
+      "loss": 0.3718,
+      "step": 4879
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.351994755211955,
+      "learning_rate": 7.69432821660438e-06,
+      "loss": 0.4672,
+      "step": 4880
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 8.08541334350025,
+      "learning_rate": 7.69337959930838e-06,
+      "loss": 0.4187,
+      "step": 4881
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 21.608607077033604,
+      "learning_rate": 7.692430845412946e-06,
+      "loss": 0.4332,
+      "step": 4882
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.572814920784045,
+      "learning_rate": 7.691481954966197e-06,
+      "loss": 0.4196,
+      "step": 4883
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.723025012815941,
+      "learning_rate": 7.690532928016259e-06,
+      "loss": 0.4392,
+      "step": 4884
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.052512009163433,
+      "learning_rate": 7.689583764611261e-06,
+      "loss": 0.4063,
+      "step": 4885
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.074690914803545,
+      "learning_rate": 7.688634464799344e-06,
+      "loss": 0.4305,
+      "step": 4886
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.971145958707373,
+      "learning_rate": 7.687685028628653e-06,
+      "loss": 0.4318,
+      "step": 4887
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.432396511841863,
+      "learning_rate": 7.68673545614734e-06,
+      "loss": 0.4324,
+      "step": 4888
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9465441484455495,
+      "learning_rate": 7.685785747403564e-06,
+      "loss": 0.4183,
+      "step": 4889
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.051669226835895,
+      "learning_rate": 7.684835902445489e-06,
+      "loss": 0.4512,
+      "step": 4890
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.072771570045427,
+      "learning_rate": 7.683885921321293e-06,
+      "loss": 0.4093,
+      "step": 4891
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 12.282148589001613,
+      "learning_rate": 7.68293580407915e-06,
+      "loss": 0.4182,
+      "step": 4892
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.2097657308589795,
+      "learning_rate": 7.68198555076725e-06,
+      "loss": 0.4317,
+      "step": 4893
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.677079081109461,
+      "learning_rate": 7.681035161433789e-06,
+      "loss": 0.4342,
+      "step": 4894
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.6504955024390355,
+      "learning_rate": 7.680084636126964e-06,
+      "loss": 0.3802,
+      "step": 4895
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.444976281441835,
+      "learning_rate": 7.679133974894984e-06,
+      "loss": 0.4164,
+      "step": 4896
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.072559034792754,
+      "learning_rate": 7.678183177786063e-06,
+      "loss": 0.409,
+      "step": 4897
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.615957282658494,
+      "learning_rate": 7.677232244848423e-06,
+      "loss": 0.4022,
+      "step": 4898
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 17.81227438094084,
+      "learning_rate": 7.67628117613029e-06,
+      "loss": 0.4329,
+      "step": 4899
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.027127600129088,
+      "learning_rate": 7.675329971679905e-06,
+      "loss": 0.4196,
+      "step": 4900
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.709644627700703,
+      "learning_rate": 7.674378631545504e-06,
+      "loss": 0.4019,
+      "step": 4901
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.289213077451036,
+      "learning_rate": 7.673427155775336e-06,
+      "loss": 0.3915,
+      "step": 4902
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.42459394239783943,
+      "learning_rate": 7.672475544417662e-06,
+      "loss": 0.211,
+      "step": 4903
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.748600363514438,
+      "learning_rate": 7.671523797520736e-06,
+      "loss": 0.4077,
+      "step": 4904
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.609579959999978,
+      "learning_rate": 7.670571915132837e-06,
+      "loss": 0.4465,
+      "step": 4905
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4826281661718426,
+      "learning_rate": 7.669619897302236e-06,
+      "loss": 0.4282,
+      "step": 4906
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.316514821991075,
+      "learning_rate": 7.668667744077215e-06,
+      "loss": 0.4142,
+      "step": 4907
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.563069775175914,
+      "learning_rate": 7.66771545550607e-06,
+      "loss": 0.4202,
+      "step": 4908
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 11.055726666061574,
+      "learning_rate": 7.666763031637092e-06,
+      "loss": 0.3869,
+      "step": 4909
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.8814373837033145,
+      "learning_rate": 7.66581047251859e-06,
+      "loss": 0.4391,
+      "step": 4910
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 14.856466894684543,
+      "learning_rate": 7.664857778198868e-06,
+      "loss": 0.4538,
+      "step": 4911
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4382114454060229,
+      "learning_rate": 7.66390494872625e-06,
+      "loss": 0.2158,
+      "step": 4912
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.432522098206094,
+      "learning_rate": 7.662951984149057e-06,
+      "loss": 0.3814,
+      "step": 4913
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.908420497901717,
+      "learning_rate": 7.66199888451562e-06,
+      "loss": 0.4328,
+      "step": 4914
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 9.478182827456736,
+      "learning_rate": 7.661045649874279e-06,
+      "loss": 0.4256,
+      "step": 4915
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.787293166179091,
+      "learning_rate": 7.66009228027338e-06,
+      "loss": 0.4052,
+      "step": 4916
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.845084882933924,
+      "learning_rate": 7.65913877576127e-06,
+      "loss": 0.426,
+      "step": 4917
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.096101860302526,
+      "learning_rate": 7.65818513638631e-06,
+      "loss": 0.375,
+      "step": 4918
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.8926361210457934,
+      "learning_rate": 7.65723136219687e-06,
+      "loss": 0.4241,
+      "step": 4919
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.261412512222484,
+      "learning_rate": 7.656277453241316e-06,
+      "loss": 0.3958,
+      "step": 4920
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 18.527003917272506,
+      "learning_rate": 7.65532340956803e-06,
+      "loss": 0.4201,
+      "step": 4921
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.914607105955681,
+      "learning_rate": 7.654369231225398e-06,
+      "loss": 0.4062,
+      "step": 4922
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.688311854338064,
+      "learning_rate": 7.653414918261812e-06,
+      "loss": 0.4249,
+      "step": 4923
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.1293757424018125,
+      "learning_rate": 7.652460470725671e-06,
+      "loss": 0.4545,
+      "step": 4924
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.634402586682809,
+      "learning_rate": 7.651505888665384e-06,
+      "loss": 0.3676,
+      "step": 4925
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.467027695423834,
+      "learning_rate": 7.650551172129362e-06,
+      "loss": 0.4134,
+      "step": 4926
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.529055662256247,
+      "learning_rate": 7.649596321166024e-06,
+      "loss": 0.4095,
+      "step": 4927
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.003828180147311,
+      "learning_rate": 7.648641335823802e-06,
+      "loss": 0.4002,
+      "step": 4928
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.4903171132283415,
+      "learning_rate": 7.647686216151124e-06,
+      "loss": 0.4132,
+      "step": 4929
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.9558991611565677,
+      "learning_rate": 7.646730962196436e-06,
+      "loss": 0.4121,
+      "step": 4930
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.921216215282887,
+      "learning_rate": 7.645775574008181e-06,
+      "loss": 0.3844,
+      "step": 4931
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.631323978883824,
+      "learning_rate": 7.644820051634813e-06,
+      "loss": 0.3958,
+      "step": 4932
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.027719947465937,
+      "learning_rate": 7.643864395124796e-06,
+      "loss": 0.4006,
+      "step": 4933
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 16.195371915968064,
+      "learning_rate": 7.642908604526596e-06,
+      "loss": 0.4557,
+      "step": 4934
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.864193043018837,
+      "learning_rate": 7.641952679888687e-06,
+      "loss": 0.392,
+      "step": 4935
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.448094538385497,
+      "learning_rate": 7.640996621259553e-06,
+      "loss": 0.428,
+      "step": 4936
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.219251828958124,
+      "learning_rate": 7.64004042868768e-06,
+      "loss": 0.3928,
+      "step": 4937
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.826209117659275,
+      "learning_rate": 7.639084102221563e-06,
+      "loss": 0.4418,
+      "step": 4938
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 23.59048310227658,
+      "learning_rate": 7.638127641909705e-06,
+      "loss": 0.3942,
+      "step": 4939
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.599611759723053,
+      "learning_rate": 7.637171047800613e-06,
+      "loss": 0.412,
+      "step": 4940
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.766793160050358,
+      "learning_rate": 7.636214319942804e-06,
+      "loss": 0.4428,
+      "step": 4941
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.81661342805067,
+      "learning_rate": 7.635257458384799e-06,
+      "loss": 0.4269,
+      "step": 4942
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.658163889258369,
+      "learning_rate": 7.634300463175128e-06,
+      "loss": 0.482,
+      "step": 4943
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 13.235530194149577,
+      "learning_rate": 7.633343334362324e-06,
+      "loss": 0.4,
+      "step": 4944
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 28.483918253155775,
+      "learning_rate": 7.632386071994932e-06,
+      "loss": 0.3911,
+      "step": 4945
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.041526899882395,
+      "learning_rate": 7.631428676121502e-06,
+      "loss": 0.4014,
+      "step": 4946
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.019743725642966,
+      "learning_rate": 7.630471146790586e-06,
+      "loss": 0.4147,
+      "step": 4947
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.714417533974368,
+      "learning_rate": 7.62951348405075e-06,
+      "loss": 0.4331,
+      "step": 4948
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.831533861938859,
+      "learning_rate": 7.628555687950563e-06,
+      "loss": 0.4242,
+      "step": 4949
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3936412707703119,
+      "learning_rate": 7.627597758538602e-06,
+      "loss": 0.2051,
+      "step": 4950
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 7.915361981602434,
+      "learning_rate": 7.626639695863447e-06,
+      "loss": 0.4235,
+      "step": 4951
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.492042359138862,
+      "learning_rate": 7.625681499973693e-06,
+      "loss": 0.4292,
+      "step": 4952
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.81993331471895,
+      "learning_rate": 7.624723170917931e-06,
+      "loss": 0.4384,
+      "step": 4953
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.48930039789246,
+      "learning_rate": 7.623764708744767e-06,
+      "loss": 0.3933,
+      "step": 4954
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.794589942914467,
+      "learning_rate": 7.622806113502813e-06,
+      "loss": 0.4271,
+      "step": 4955
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 10.585564752530418,
+      "learning_rate": 7.6218473852406826e-06,
+      "loss": 0.4031,
+      "step": 4956
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.285249619796292,
+      "learning_rate": 7.6208885240069995e-06,
+      "loss": 0.3894,
+      "step": 4957
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 6.701975125564328,
+      "learning_rate": 7.619929529850397e-06,
+      "loss": 0.4264,
+      "step": 4958
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.207319778707752,
+      "learning_rate": 7.618970402819509e-06,
+      "loss": 0.4324,
+      "step": 4959
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 5.88820040948218,
+      "learning_rate": 7.61801114296298e-06,
+      "loss": 0.4405,
+      "step": 4960
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.913182822852285,
+      "learning_rate": 7.617051750329462e-06,
+      "loss": 0.4337,
+      "step": 4961
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.472780416876222,
+      "learning_rate": 7.61609222496761e-06,
+      "loss": 0.4523,
+      "step": 4962
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.35433347530578,
+      "learning_rate": 7.6151325669260915e-06,
+      "loss": 0.3886,
+      "step": 4963
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.9545933177789614,
+      "learning_rate": 7.614172776253573e-06,
+      "loss": 0.4131,
+      "step": 4964
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 36.72723610483068,
+      "learning_rate": 7.613212852998735e-06,
+      "loss": 0.434,
+      "step": 4965
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.112139631067949,
+      "learning_rate": 7.61225279721026e-06,
+      "loss": 0.3951,
+      "step": 4966
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.416472234437258,
+      "learning_rate": 7.611292608936841e-06,
+      "loss": 0.4446,
+      "step": 4967
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.42437771887734477,
+      "learning_rate": 7.610332288227173e-06,
+      "loss": 0.2044,
+      "step": 4968
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.43448788949249,
+      "learning_rate": 7.609371835129962e-06,
+      "loss": 0.3777,
+      "step": 4969
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.953149490902296,
+      "learning_rate": 7.6084112496939185e-06,
+      "loss": 0.426,
+      "step": 4970
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.44310952196547,
+      "learning_rate": 7.6074505319677595e-06,
+      "loss": 0.4492,
+      "step": 4971
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.758806178349619,
+      "learning_rate": 7.606489682000211e-06,
+      "loss": 0.4017,
+      "step": 4972
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.894292999484922,
+      "learning_rate": 7.605528699840004e-06,
+      "loss": 0.388,
+      "step": 4973
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.11125430093527,
+      "learning_rate": 7.604567585535875e-06,
+      "loss": 0.4262,
+      "step": 4974
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.812399217590182,
+      "learning_rate": 7.6036063391365685e-06,
+      "loss": 0.3681,
+      "step": 4975
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.205932714010422,
+      "learning_rate": 7.6026449606908395e-06,
+      "loss": 0.3979,
+      "step": 4976
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.363373773894307,
+      "learning_rate": 7.6016834502474415e-06,
+      "loss": 0.3762,
+      "step": 4977
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.780623754120524,
+      "learning_rate": 7.600721807855141e-06,
+      "loss": 0.4062,
+      "step": 4978
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.692763688208174,
+      "learning_rate": 7.599760033562708e-06,
+      "loss": 0.4099,
+      "step": 4979
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.358477130101587,
+      "learning_rate": 7.598798127418923e-06,
+      "loss": 0.4534,
+      "step": 4980
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.703483790034664,
+      "learning_rate": 7.59783608947257e-06,
+      "loss": 0.4318,
+      "step": 4981
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.94019124261076,
+      "learning_rate": 7.596873919772438e-06,
+      "loss": 0.4542,
+      "step": 4982
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.555874651814693,
+      "learning_rate": 7.59591161836733e-06,
+      "loss": 0.3954,
+      "step": 4983
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.375578029800533,
+      "learning_rate": 7.5949491853060465e-06,
+      "loss": 0.3743,
+      "step": 4984
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.266893465152988,
+      "learning_rate": 7.5939866206374006e-06,
+      "loss": 0.3845,
+      "step": 4985
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.562013931600962,
+      "learning_rate": 7.593023924410209e-06,
+      "loss": 0.4338,
+      "step": 4986
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.311040909196945,
+      "learning_rate": 7.592061096673299e-06,
+      "loss": 0.4094,
+      "step": 4987
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 32.5123488145877,
+      "learning_rate": 7.591098137475501e-06,
+      "loss": 0.4426,
+      "step": 4988
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.775731369690057,
+      "learning_rate": 7.590135046865652e-06,
+      "loss": 0.3619,
+      "step": 4989
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.082641458373449,
+      "learning_rate": 7.589171824892599e-06,
+      "loss": 0.4048,
+      "step": 4990
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.115228411238699,
+      "learning_rate": 7.588208471605191e-06,
+      "loss": 0.4454,
+      "step": 4991
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.715040716386525,
+      "learning_rate": 7.587244987052287e-06,
+      "loss": 0.3875,
+      "step": 4992
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.697854116270032,
+      "learning_rate": 7.586281371282753e-06,
+      "loss": 0.4063,
+      "step": 4993
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.270139164628531,
+      "learning_rate": 7.585317624345459e-06,
+      "loss": 0.3919,
+      "step": 4994
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.475716440501883,
+      "learning_rate": 7.584353746289283e-06,
+      "loss": 0.4069,
+      "step": 4995
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.113986294105402,
+      "learning_rate": 7.583389737163112e-06,
+      "loss": 0.4344,
+      "step": 4996
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.727830461011553,
+      "learning_rate": 7.582425597015837e-06,
+      "loss": 0.3758,
+      "step": 4997
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.362646601156877,
+      "learning_rate": 7.5814613258963535e-06,
+      "loss": 0.411,
+      "step": 4998
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.644788779457551,
+      "learning_rate": 7.580496923853567e-06,
+      "loss": 0.3932,
+      "step": 4999
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.7898354871282764,
+      "learning_rate": 7.579532390936391e-06,
+      "loss": 0.3599,
+      "step": 5000
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.347948946574471,
+      "learning_rate": 7.578567727193739e-06,
+      "loss": 0.4288,
+      "step": 5001
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.680860940032272,
+      "learning_rate": 7.577602932674542e-06,
+      "loss": 0.3983,
+      "step": 5002
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.792506975556162,
+      "learning_rate": 7.576638007427725e-06,
+      "loss": 0.41,
+      "step": 5003
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.235728331761503,
+      "learning_rate": 7.575672951502229e-06,
+      "loss": 0.3938,
+      "step": 5004
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.113030910460528,
+      "learning_rate": 7.574707764946999e-06,
+      "loss": 0.4168,
+      "step": 5005
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.124778834152398,
+      "learning_rate": 7.5737424478109844e-06,
+      "loss": 0.4066,
+      "step": 5006
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.890886409176312,
+      "learning_rate": 7.572777000143145e-06,
+      "loss": 0.4223,
+      "step": 5007
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5009523138780653,
+      "learning_rate": 7.571811421992442e-06,
+      "loss": 0.2045,
+      "step": 5008
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.8300759162007605,
+      "learning_rate": 7.57084571340785e-06,
+      "loss": 0.4044,
+      "step": 5009
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.74313639068283,
+      "learning_rate": 7.5698798744383436e-06,
+      "loss": 0.4301,
+      "step": 5010
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.450104761123079,
+      "learning_rate": 7.568913905132911e-06,
+      "loss": 0.4088,
+      "step": 5011
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.3825114823358848,
+      "learning_rate": 7.567947805540536e-06,
+      "loss": 0.2057,
+      "step": 5012
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.433489427200484,
+      "learning_rate": 7.566981575710224e-06,
+      "loss": 0.4069,
+      "step": 5013
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.726992269791886,
+      "learning_rate": 7.566015215690973e-06,
+      "loss": 0.4266,
+      "step": 5014
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.514707987654532,
+      "learning_rate": 7.565048725531797e-06,
+      "loss": 0.381,
+      "step": 5015
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.4830376076911955,
+      "learning_rate": 7.564082105281711e-06,
+      "loss": 0.4201,
+      "step": 5016
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.212427736583935,
+      "learning_rate": 7.563115354989742e-06,
+      "loss": 0.4542,
+      "step": 5017
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 49.925677873840236,
+      "learning_rate": 7.562148474704916e-06,
+      "loss": 0.4021,
+      "step": 5018
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.407044559328034,
+      "learning_rate": 7.561181464476275e-06,
+      "loss": 0.4396,
+      "step": 5019
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 23.799408013846957,
+      "learning_rate": 7.560214324352858e-06,
+      "loss": 0.4178,
+      "step": 5020
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.5741732188942645,
+      "learning_rate": 7.559247054383718e-06,
+      "loss": 0.4525,
+      "step": 5021
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.334440931900414,
+      "learning_rate": 7.5582796546179125e-06,
+      "loss": 0.3944,
+      "step": 5022
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.526775132669506,
+      "learning_rate": 7.557312125104501e-06,
+      "loss": 0.4095,
+      "step": 5023
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.4779786474129875,
+      "learning_rate": 7.556344465892557e-06,
+      "loss": 0.4368,
+      "step": 5024
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.058308667141982,
+      "learning_rate": 7.555376677031156e-06,
+      "loss": 0.4006,
+      "step": 5025
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.373846245395796,
+      "learning_rate": 7.5544087585693805e-06,
+      "loss": 0.4067,
+      "step": 5026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 15.495239519716945,
+      "learning_rate": 7.553440710556321e-06,
+      "loss": 0.3847,
+      "step": 5027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.31099333628324,
+      "learning_rate": 7.552472533041073e-06,
+      "loss": 0.4114,
+      "step": 5028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.032121621747589,
+      "learning_rate": 7.551504226072739e-06,
+      "loss": 0.4297,
+      "step": 5029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 18.730816196919747,
+      "learning_rate": 7.55053578970043e-06,
+      "loss": 0.4359,
+      "step": 5030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.495916358428323,
+      "learning_rate": 7.549567223973261e-06,
+      "loss": 0.4162,
+      "step": 5031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 28.819882760074385,
+      "learning_rate": 7.548598528940354e-06,
+      "loss": 0.4084,
+      "step": 5032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5200112610391537,
+      "learning_rate": 7.547629704650839e-06,
+      "loss": 0.2065,
+      "step": 5033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.651256818797963,
+      "learning_rate": 7.546660751153851e-06,
+      "loss": 0.4428,
+      "step": 5034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.41741994257794596,
+      "learning_rate": 7.5456916684985325e-06,
+      "loss": 0.2017,
+      "step": 5035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.344689901110803,
+      "learning_rate": 7.54472245673403e-06,
+      "loss": 0.4372,
+      "step": 5036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.578537604777829,
+      "learning_rate": 7.543753115909504e-06,
+      "loss": 0.4212,
+      "step": 5037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.362524589409158,
+      "learning_rate": 7.542783646074111e-06,
+      "loss": 0.3877,
+      "step": 5038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.34174535786036,
+      "learning_rate": 7.541814047277023e-06,
+      "loss": 0.4089,
+      "step": 5039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.948728570215053,
+      "learning_rate": 7.540844319567411e-06,
+      "loss": 0.3858,
+      "step": 5040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.904312642656292,
+      "learning_rate": 7.53987446299446e-06,
+      "loss": 0.4166,
+      "step": 5041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.011801022622378,
+      "learning_rate": 7.5389044776073565e-06,
+      "loss": 0.4255,
+      "step": 5042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.030792202787653,
+      "learning_rate": 7.537934363455296e-06,
+      "loss": 0.3876,
+      "step": 5043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 11.998246859646317,
+      "learning_rate": 7.536964120587478e-06,
+      "loss": 0.3908,
+      "step": 5044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.372578034120173,
+      "learning_rate": 7.535993749053111e-06,
+      "loss": 0.4258,
+      "step": 5045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.977356384573518,
+      "learning_rate": 7.535023248901409e-06,
+      "loss": 0.4208,
+      "step": 5046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.978784915446848,
+      "learning_rate": 7.534052620181591e-06,
+      "loss": 0.3681,
+      "step": 5047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.214472463774554,
+      "learning_rate": 7.533081862942888e-06,
+      "loss": 0.4252,
+      "step": 5048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.614386164791594,
+      "learning_rate": 7.5321109772345305e-06,
+      "loss": 0.3895,
+      "step": 5049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.271585550331442,
+      "learning_rate": 7.531139963105758e-06,
+      "loss": 0.4054,
+      "step": 5050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.853263185072293,
+      "learning_rate": 7.530168820605819e-06,
+      "loss": 0.4106,
+      "step": 5051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.2184567092475795,
+      "learning_rate": 7.529197549783967e-06,
+      "loss": 0.4357,
+      "step": 5052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.298830406452331,
+      "learning_rate": 7.528226150689461e-06,
+      "loss": 0.4214,
+      "step": 5053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.141057722412565,
+      "learning_rate": 7.527254623371567e-06,
+      "loss": 0.4253,
+      "step": 5054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.229856835670705,
+      "learning_rate": 7.526282967879558e-06,
+      "loss": 0.4198,
+      "step": 5055
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.2539090787011675,
+      "learning_rate": 7.525311184262714e-06,
+      "loss": 0.4122,
+      "step": 5056
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.535071369083814,
+      "learning_rate": 7.524339272570317e-06,
+      "loss": 0.4406,
+      "step": 5057
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.544945023779245,
+      "learning_rate": 7.523367232851664e-06,
+      "loss": 0.3939,
+      "step": 5058
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.696954586638519,
+      "learning_rate": 7.522395065156049e-06,
+      "loss": 0.4231,
+      "step": 5059
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.916139063567723,
+      "learning_rate": 7.5214227695327825e-06,
+      "loss": 0.39,
+      "step": 5060
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.724232943151632,
+      "learning_rate": 7.520450346031172e-06,
+      "loss": 0.3985,
+      "step": 5061
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.610161780914758,
+      "learning_rate": 7.519477794700538e-06,
+      "loss": 0.4161,
+      "step": 5062
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.767574850549739,
+      "learning_rate": 7.518505115590204e-06,
+      "loss": 0.4455,
+      "step": 5063
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.607044301945914,
+      "learning_rate": 7.5175323087495025e-06,
+      "loss": 0.4148,
+      "step": 5064
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.438212922324761,
+      "learning_rate": 7.5165593742277676e-06,
+      "loss": 0.3795,
+      "step": 5065
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.628061510861352,
+      "learning_rate": 7.5155863120743475e-06,
+      "loss": 0.4064,
+      "step": 5066
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 19.78248878355449,
+      "learning_rate": 7.5146131223385895e-06,
+      "loss": 0.4362,
+      "step": 5067
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.5534747279773224,
+      "learning_rate": 7.513639805069855e-06,
+      "loss": 0.4606,
+      "step": 5068
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.947573033713241,
+      "learning_rate": 7.5126663603175e-06,
+      "loss": 0.4558,
+      "step": 5069
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.961724326225806,
+      "learning_rate": 7.511692788130902e-06,
+      "loss": 0.4149,
+      "step": 5070
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.1235247689175685,
+      "learning_rate": 7.510719088559433e-06,
+      "loss": 0.4356,
+      "step": 5071
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 16.378325490176596,
+      "learning_rate": 7.509745261652479e-06,
+      "loss": 0.43,
+      "step": 5072
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.646038322241565,
+      "learning_rate": 7.508771307459426e-06,
+      "loss": 0.3969,
+      "step": 5073
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.542580819848394,
+      "learning_rate": 7.507797226029672e-06,
+      "loss": 0.3925,
+      "step": 5074
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.80396983571785,
+      "learning_rate": 7.506823017412617e-06,
+      "loss": 0.378,
+      "step": 5075
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.649299680788743,
+      "learning_rate": 7.505848681657674e-06,
+      "loss": 0.3688,
+      "step": 5076
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.319339525292163,
+      "learning_rate": 7.504874218814253e-06,
+      "loss": 0.4013,
+      "step": 5077
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.603599031551753,
+      "learning_rate": 7.503899628931781e-06,
+      "loss": 0.4209,
+      "step": 5078
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.340378078172057,
+      "learning_rate": 7.502924912059679e-06,
+      "loss": 0.4023,
+      "step": 5079
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.821442786558005,
+      "learning_rate": 7.501950068247389e-06,
+      "loss": 0.4341,
+      "step": 5080
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.126329921345306,
+      "learning_rate": 7.500975097544346e-06,
+      "loss": 0.4345,
+      "step": 5081
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0234305173872178,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2439,
+      "step": 5082
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.606768342289106,
+      "learning_rate": 7.499024775663806e-06,
+      "loss": 0.405,
+      "step": 5083
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 12.332149055694824,
+      "learning_rate": 7.498049424585222e-06,
+      "loss": 0.4298,
+      "step": 5084
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.404460300655753,
+      "learning_rate": 7.497073946813714e-06,
+      "loss": 0.4234,
+      "step": 5085
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.8584913468012765,
+      "learning_rate": 7.496098342398758e-06,
+      "loss": 0.3811,
+      "step": 5086
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.460869612795348,
+      "learning_rate": 7.495122611389831e-06,
+      "loss": 0.4118,
+      "step": 5087
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.79616338752196,
+      "learning_rate": 7.49414675383642e-06,
+      "loss": 0.4236,
+      "step": 5088
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.52528149569849,
+      "learning_rate": 7.4931707697880175e-06,
+      "loss": 0.4567,
+      "step": 5089
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.436236216511352,
+      "learning_rate": 7.4921946592941206e-06,
+      "loss": 0.4496,
+      "step": 5090
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.641516932161435,
+      "learning_rate": 7.491218422404236e-06,
+      "loss": 0.4705,
+      "step": 5091
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 9.232664100695821,
+      "learning_rate": 7.490242059167876e-06,
+      "loss": 0.4244,
+      "step": 5092
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.705169210588128,
+      "learning_rate": 7.489265569634556e-06,
+      "loss": 0.409,
+      "step": 5093
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.668563315903459,
+      "learning_rate": 7.488288953853802e-06,
+      "loss": 0.4226,
+      "step": 5094
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.422031657653414,
+      "learning_rate": 7.487312211875144e-06,
+      "loss": 0.4244,
+      "step": 5095
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.201604952195401,
+      "learning_rate": 7.486335343748121e-06,
+      "loss": 0.3972,
+      "step": 5096
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.378601664737282,
+      "learning_rate": 7.4853583495222745e-06,
+      "loss": 0.3881,
+      "step": 5097
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 7.414488204990885,
+      "learning_rate": 7.484381229247156e-06,
+      "loss": 0.4324,
+      "step": 5098
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 10.283356914522756,
+      "learning_rate": 7.4834039829723214e-06,
+      "loss": 0.4194,
+      "step": 5099
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 27.847419955740435,
+      "learning_rate": 7.482426610747333e-06,
+      "loss": 0.3699,
+      "step": 5100
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.045541434531843,
+      "learning_rate": 7.481449112621758e-06,
+      "loss": 0.4059,
+      "step": 5101
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 8.308234450686331,
+      "learning_rate": 7.480471488645175e-06,
+      "loss": 0.4221,
+      "step": 5102
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 4.455894082211952,
+      "learning_rate": 7.479493738867166e-06,
+      "loss": 0.4277,
+      "step": 5103
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 6.241076429400722,
+      "learning_rate": 7.478515863337317e-06,
+      "loss": 0.4163,
+      "step": 5104
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 13.145020489913598,
+      "learning_rate": 7.477537862105224e-06,
+      "loss": 0.3933,
+      "step": 5105
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7514925806078635,
+      "learning_rate": 7.4765597352204885e-06,
+      "loss": 0.2498,
+      "step": 5106
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.469874624314357,
+      "learning_rate": 7.475581482732717e-06,
+      "loss": 0.3938,
+      "step": 5107
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.504708579477634,
+      "learning_rate": 7.4746031046915234e-06,
+      "loss": 0.4001,
+      "step": 5108
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.879062066249945,
+      "learning_rate": 7.473624601146529e-06,
+      "loss": 0.4134,
+      "step": 5109
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.80438656645539,
+      "learning_rate": 7.472645972147359e-06,
+      "loss": 0.4224,
+      "step": 5110
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 17.02996336988066,
+      "learning_rate": 7.471667217743646e-06,
+      "loss": 0.4188,
+      "step": 5111
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.246192594746278,
+      "learning_rate": 7.470688337985029e-06,
+      "loss": 0.3877,
+      "step": 5112
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.993094493113891,
+      "learning_rate": 7.469709332921155e-06,
+      "loss": 0.429,
+      "step": 5113
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.650660343230416,
+      "learning_rate": 7.4687302026016765e-06,
+      "loss": 0.4024,
+      "step": 5114
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.985232158570794,
+      "learning_rate": 7.46775094707625e-06,
+      "loss": 0.3997,
+      "step": 5115
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.07102608274262,
+      "learning_rate": 7.466771566394539e-06,
+      "loss": 0.4106,
+      "step": 5116
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.39487148124487,
+      "learning_rate": 7.465792060606218e-06,
+      "loss": 0.4456,
+      "step": 5117
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.357778617561368,
+      "learning_rate": 7.464812429760963e-06,
+      "loss": 0.4224,
+      "step": 5118
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.75598582878143,
+      "learning_rate": 7.463832673908458e-06,
+      "loss": 0.4164,
+      "step": 5119
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.088609229992308,
+      "learning_rate": 7.46285279309839e-06,
+      "loss": 0.3923,
+      "step": 5120
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.31102025124468,
+      "learning_rate": 7.4618727873804586e-06,
+      "loss": 0.4467,
+      "step": 5121
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.33841674065176,
+      "learning_rate": 7.460892656804366e-06,
+      "loss": 0.4286,
+      "step": 5122
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.431422087547383,
+      "learning_rate": 7.4599124014198225e-06,
+      "loss": 0.3992,
+      "step": 5123
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.077270008417891,
+      "learning_rate": 7.458932021276538e-06,
+      "loss": 0.4153,
+      "step": 5124
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 32.9288115970929,
+      "learning_rate": 7.457951516424241e-06,
+      "loss": 0.4184,
+      "step": 5125
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.223761316952459,
+      "learning_rate": 7.456970886912655e-06,
+      "loss": 0.4478,
+      "step": 5126
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4488678357873052,
+      "learning_rate": 7.455990132791516e-06,
+      "loss": 0.2143,
+      "step": 5127
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.49583238273943,
+      "learning_rate": 7.455009254110565e-06,
+      "loss": 0.4157,
+      "step": 5128
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.358252844398133,
+      "learning_rate": 7.454028250919547e-06,
+      "loss": 0.4753,
+      "step": 5129
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.37744063202451755,
+      "learning_rate": 7.453047123268218e-06,
+      "loss": 0.2212,
+      "step": 5130
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.739972566065215,
+      "learning_rate": 7.452065871206336e-06,
+      "loss": 0.4476,
+      "step": 5131
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.962237252233531,
+      "learning_rate": 7.451084494783668e-06,
+      "loss": 0.4023,
+      "step": 5132
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.474439490367816,
+      "learning_rate": 7.450102994049985e-06,
+      "loss": 0.4695,
+      "step": 5133
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.765372430197792,
+      "learning_rate": 7.449121369055066e-06,
+      "loss": 0.478,
+      "step": 5134
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.677383024135758,
+      "learning_rate": 7.448139619848695e-06,
+      "loss": 0.4333,
+      "step": 5135
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.5201022160790805,
+      "learning_rate": 7.4471577464806635e-06,
+      "loss": 0.4052,
+      "step": 5136
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.0748793879438505,
+      "learning_rate": 7.44617574900077e-06,
+      "loss": 0.4016,
+      "step": 5137
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.668554043836922,
+      "learning_rate": 7.445193627458818e-06,
+      "loss": 0.4275,
+      "step": 5138
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.611842177160016,
+      "learning_rate": 7.4442113819046154e-06,
+      "loss": 0.3984,
+      "step": 5139
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 20.830541162937944,
+      "learning_rate": 7.443229012387982e-06,
+      "loss": 0.4452,
+      "step": 5140
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.411912131926814,
+      "learning_rate": 7.442246518958738e-06,
+      "loss": 0.4206,
+      "step": 5141
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.5301051435593624,
+      "learning_rate": 7.441263901666711e-06,
+      "loss": 0.4445,
+      "step": 5142
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.652022326419658,
+      "learning_rate": 7.4402811605617395e-06,
+      "loss": 0.4036,
+      "step": 5143
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.447293608632584,
+      "learning_rate": 7.4392982956936644e-06,
+      "loss": 0.4115,
+      "step": 5144
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.115397927506166,
+      "learning_rate": 7.438315307112331e-06,
+      "loss": 0.4434,
+      "step": 5145
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.558507095310255,
+      "learning_rate": 7.437332194867595e-06,
+      "loss": 0.407,
+      "step": 5146
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.566968049876717,
+      "learning_rate": 7.436348959009316e-06,
+      "loss": 0.4532,
+      "step": 5147
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 14.306110382732054,
+      "learning_rate": 7.43536559958736e-06,
+      "loss": 0.4309,
+      "step": 5148
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.166927858337883,
+      "learning_rate": 7.4343821166516015e-06,
+      "loss": 0.4686,
+      "step": 5149
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.174566345465292,
+      "learning_rate": 7.433398510251918e-06,
+      "loss": 0.4158,
+      "step": 5150
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.784595429653403,
+      "learning_rate": 7.4324147804381956e-06,
+      "loss": 0.4177,
+      "step": 5151
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.90761546029646,
+      "learning_rate": 7.431430927260326e-06,
+      "loss": 0.3866,
+      "step": 5152
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.675126089562968,
+      "learning_rate": 7.430446950768207e-06,
+      "loss": 0.4566,
+      "step": 5153
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.5919454281657845,
+      "learning_rate": 7.4294628510117405e-06,
+      "loss": 0.4444,
+      "step": 5154
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.759170034243303,
+      "learning_rate": 7.428478628040841e-06,
+      "loss": 0.4559,
+      "step": 5155
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.9732592960330395,
+      "learning_rate": 7.427494281905421e-06,
+      "loss": 0.42,
+      "step": 5156
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.318967895916565,
+      "learning_rate": 7.4265098126554065e-06,
+      "loss": 0.4201,
+      "step": 5157
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.080054810797494,
+      "learning_rate": 7.425525220340723e-06,
+      "loss": 0.4221,
+      "step": 5158
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5671475888128275,
+      "learning_rate": 7.424540505011311e-06,
+      "loss": 0.2249,
+      "step": 5159
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.599681871550648,
+      "learning_rate": 7.423555666717107e-06,
+      "loss": 0.4053,
+      "step": 5160
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.248277143892327,
+      "learning_rate": 7.422570705508062e-06,
+      "loss": 0.4091,
+      "step": 5161
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.23152886471819,
+      "learning_rate": 7.421585621434128e-06,
+      "loss": 0.4177,
+      "step": 5162
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.155848387577581,
+      "learning_rate": 7.420600414545266e-06,
+      "loss": 0.4125,
+      "step": 5163
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.106288488173856,
+      "learning_rate": 7.419615084891443e-06,
+      "loss": 0.4307,
+      "step": 5164
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.578540079795062,
+      "learning_rate": 7.4186296325226315e-06,
+      "loss": 0.4098,
+      "step": 5165
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.371621034727422,
+      "learning_rate": 7.41764405748881e-06,
+      "loss": 0.4206,
+      "step": 5166
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.700722153886488,
+      "learning_rate": 7.416658359839964e-06,
+      "loss": 0.3955,
+      "step": 5167
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.037914959787077,
+      "learning_rate": 7.415672539626084e-06,
+      "loss": 0.3939,
+      "step": 5168
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.23179401668046,
+      "learning_rate": 7.414686596897169e-06,
+      "loss": 0.4344,
+      "step": 5169
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.261235856798153,
+      "learning_rate": 7.413700531703223e-06,
+      "loss": 0.4467,
+      "step": 5170
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.972961459597504,
+      "learning_rate": 7.4127143440942525e-06,
+      "loss": 0.4407,
+      "step": 5171
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.969967841071323,
+      "learning_rate": 7.411728034120279e-06,
+      "loss": 0.4009,
+      "step": 5172
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.011567418389424,
+      "learning_rate": 7.41074160183132e-06,
+      "loss": 0.4016,
+      "step": 5173
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.122307243066748,
+      "learning_rate": 7.409755047277409e-06,
+      "loss": 0.388,
+      "step": 5174
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.927825920083381,
+      "learning_rate": 7.408768370508577e-06,
+      "loss": 0.4179,
+      "step": 5175
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.4086224480153895,
+      "learning_rate": 7.407781571574866e-06,
+      "loss": 0.4056,
+      "step": 5176
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.3037132654479064,
+      "learning_rate": 7.406794650526324e-06,
+      "loss": 0.3863,
+      "step": 5177
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 33.28595589191784,
+      "learning_rate": 7.4058076074130044e-06,
+      "loss": 0.4594,
+      "step": 5178
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.262237217588735,
+      "learning_rate": 7.404820442284966e-06,
+      "loss": 0.3898,
+      "step": 5179
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.456940081999953,
+      "learning_rate": 7.403833155192275e-06,
+      "loss": 0.376,
+      "step": 5180
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.064406571512603,
+      "learning_rate": 7.4028457461850015e-06,
+      "loss": 0.394,
+      "step": 5181
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.454765571167476,
+      "learning_rate": 7.401858215313228e-06,
+      "loss": 0.4189,
+      "step": 5182
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.38660115807598,
+      "learning_rate": 7.400870562627034e-06,
+      "loss": 0.4221,
+      "step": 5183
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.004905194765643,
+      "learning_rate": 7.399882788176515e-06,
+      "loss": 0.4075,
+      "step": 5184
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.359174071633196,
+      "learning_rate": 7.398894892011762e-06,
+      "loss": 0.3905,
+      "step": 5185
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.547694898446448,
+      "learning_rate": 7.397906874182883e-06,
+      "loss": 0.404,
+      "step": 5186
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.944541435527931,
+      "learning_rate": 7.396918734739985e-06,
+      "loss": 0.4446,
+      "step": 5187
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.034959439649294,
+      "learning_rate": 7.395930473733183e-06,
+      "loss": 0.4168,
+      "step": 5188
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.41133514654461,
+      "learning_rate": 7.394942091212598e-06,
+      "loss": 0.381,
+      "step": 5189
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.887109660639239,
+      "learning_rate": 7.393953587228359e-06,
+      "loss": 0.4235,
+      "step": 5190
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.827965685904982,
+      "learning_rate": 7.392964961830598e-06,
+      "loss": 0.4045,
+      "step": 5191
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.192483192398559,
+      "learning_rate": 7.391976215069456e-06,
+      "loss": 0.3969,
+      "step": 5192
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.359163473317746,
+      "learning_rate": 7.390987346995078e-06,
+      "loss": 0.4255,
+      "step": 5193
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.92966531463988,
+      "learning_rate": 7.389998357657617e-06,
+      "loss": 0.4112,
+      "step": 5194
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.1464421778476845,
+      "learning_rate": 7.389009247107232e-06,
+      "loss": 0.4126,
+      "step": 5195
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.085484326901734,
+      "learning_rate": 7.388020015394086e-06,
+      "loss": 0.4508,
+      "step": 5196
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5628751697568759,
+      "learning_rate": 7.38703066256835e-06,
+      "loss": 0.2255,
+      "step": 5197
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 35.133842944081636,
+      "learning_rate": 7.386041188680202e-06,
+      "loss": 0.4,
+      "step": 5198
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.7429653665490825,
+      "learning_rate": 7.3850515937798225e-06,
+      "loss": 0.4136,
+      "step": 5199
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.693994496095793,
+      "learning_rate": 7.3840618779174034e-06,
+      "loss": 0.4484,
+      "step": 5200
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 18.66456921691234,
+      "learning_rate": 7.383072041143137e-06,
+      "loss": 0.4444,
+      "step": 5201
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.132392255296666,
+      "learning_rate": 7.382082083507226e-06,
+      "loss": 0.4215,
+      "step": 5202
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.798513612793256,
+      "learning_rate": 7.381092005059878e-06,
+      "loss": 0.3201,
+      "step": 5203
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.80657644665968,
+      "learning_rate": 7.380101805851307e-06,
+      "loss": 0.3764,
+      "step": 5204
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.093218837731931,
+      "learning_rate": 7.379111485931731e-06,
+      "loss": 0.3923,
+      "step": 5205
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.077140094829424,
+      "learning_rate": 7.378121045351378e-06,
+      "loss": 0.4167,
+      "step": 5206
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.5110221234933645,
+      "learning_rate": 7.3771304841604764e-06,
+      "loss": 0.418,
+      "step": 5207
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.25317344104094,
+      "learning_rate": 7.376139802409268e-06,
+      "loss": 0.4043,
+      "step": 5208
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4981179584455057,
+      "learning_rate": 7.375149000147995e-06,
+      "loss": 0.2087,
+      "step": 5209
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 12.23950221925532,
+      "learning_rate": 7.37415807742691e-06,
+      "loss": 0.3809,
+      "step": 5210
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.863910849448008,
+      "learning_rate": 7.3731670342962655e-06,
+      "loss": 0.4285,
+      "step": 5211
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.864929140930194,
+      "learning_rate": 7.372175870806327e-06,
+      "loss": 0.4127,
+      "step": 5212
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.909323738652096,
+      "learning_rate": 7.3711845870073615e-06,
+      "loss": 0.4187,
+      "step": 5213
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.311275919788103,
+      "learning_rate": 7.370193182949645e-06,
+      "loss": 0.4221,
+      "step": 5214
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.2955506618965,
+      "learning_rate": 7.369201658683456e-06,
+      "loss": 0.4516,
+      "step": 5215
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.16339518707435,
+      "learning_rate": 7.3682100142590855e-06,
+      "loss": 0.3991,
+      "step": 5216
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.586286586324904,
+      "learning_rate": 7.367218249726821e-06,
+      "loss": 0.4271,
+      "step": 5217
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.01431336507799,
+      "learning_rate": 7.366226365136966e-06,
+      "loss": 0.3866,
+      "step": 5218
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.766386983489,
+      "learning_rate": 7.365234360539824e-06,
+      "loss": 0.3832,
+      "step": 5219
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.34940456318696,
+      "learning_rate": 7.3642422359857055e-06,
+      "loss": 0.4276,
+      "step": 5220
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 13.119894422356165,
+      "learning_rate": 7.36324999152493e-06,
+      "loss": 0.4228,
+      "step": 5221
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 6.040986606844776,
+      "learning_rate": 7.362257627207818e-06,
+      "loss": 0.4077,
+      "step": 5222
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.252384001565092,
+      "learning_rate": 7.361265143084703e-06,
+      "loss": 0.4227,
+      "step": 5223
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 19.573671433294493,
+      "learning_rate": 7.360272539205917e-06,
+      "loss": 0.405,
+      "step": 5224
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.244004008406878,
+      "learning_rate": 7.359279815621801e-06,
+      "loss": 0.4036,
+      "step": 5225
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 10.980636628617425,
+      "learning_rate": 7.3582869723827065e-06,
+      "loss": 0.4176,
+      "step": 5226
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.647742184216636,
+      "learning_rate": 7.3572940095389845e-06,
+      "loss": 0.4095,
+      "step": 5227
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.456028185884802,
+      "learning_rate": 7.356300927140996e-06,
+      "loss": 0.3956,
+      "step": 5228
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.5426600330230364,
+      "learning_rate": 7.355307725239104e-06,
+      "loss": 0.4335,
+      "step": 5229
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.352463048906314,
+      "learning_rate": 7.3543144038836865e-06,
+      "loss": 0.4421,
+      "step": 5230
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.84523238226206,
+      "learning_rate": 7.3533209631251155e-06,
+      "loss": 0.436,
+      "step": 5231
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.209495285424516,
+      "learning_rate": 7.352327403013779e-06,
+      "loss": 0.3854,
+      "step": 5232
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 51.682981010682425,
+      "learning_rate": 7.351333723600065e-06,
+      "loss": 0.4189,
+      "step": 5233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.680126882858509,
+      "learning_rate": 7.35033992493437e-06,
+      "loss": 0.4674,
+      "step": 5234
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.601838849937643,
+      "learning_rate": 7.3493460070670975e-06,
+      "loss": 0.4007,
+      "step": 5235
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.8705398245768965,
+      "learning_rate": 7.348351970048654e-06,
+      "loss": 0.3652,
+      "step": 5236
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.279078038619245,
+      "learning_rate": 7.347357813929455e-06,
+      "loss": 0.3563,
+      "step": 5237
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.893110146975505,
+      "learning_rate": 7.34636353875992e-06,
+      "loss": 0.4443,
+      "step": 5238
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.522557903271032,
+      "learning_rate": 7.345369144590476e-06,
+      "loss": 0.4079,
+      "step": 5239
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.222066497437492,
+      "learning_rate": 7.3443746314715554e-06,
+      "loss": 0.3875,
+      "step": 5240
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 16.063721052888646,
+      "learning_rate": 7.343379999453597e-06,
+      "loss": 0.3963,
+      "step": 5241
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.46690152311351135,
+      "learning_rate": 7.342385248587044e-06,
+      "loss": 0.2056,
+      "step": 5242
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.629071475962925,
+      "learning_rate": 7.34139037892235e-06,
+      "loss": 0.3647,
+      "step": 5243
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 5.219877323712834,
+      "learning_rate": 7.340395390509968e-06,
+      "loss": 0.4306,
+      "step": 5244
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 7.984691732336158,
+      "learning_rate": 7.339400283400362e-06,
+      "loss": 0.4016,
+      "step": 5245
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 8.579978360721498,
+      "learning_rate": 7.338405057643999e-06,
+      "loss": 0.4366,
+      "step": 5246
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.738121439856403,
+      "learning_rate": 7.337409713291357e-06,
+      "loss": 0.4132,
+      "step": 5247
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 9.908283247519389,
+      "learning_rate": 7.336414250392914e-06,
+      "loss": 0.4436,
+      "step": 5248
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 11.73141040452063,
+      "learning_rate": 7.335418668999158e-06,
+      "loss": 0.3839,
+      "step": 5249
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.651571677695675,
+      "learning_rate": 7.33442296916058e-06,
+      "loss": 0.3767,
+      "step": 5250
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.879963366264922,
+      "learning_rate": 7.33342715092768e-06,
+      "loss": 0.4106,
+      "step": 5251
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.72109319341328,
+      "learning_rate": 7.332431214350962e-06,
+      "loss": 0.4108,
+      "step": 5252
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.518148760071401,
+      "learning_rate": 7.331435159480938e-06,
+      "loss": 0.426,
+      "step": 5253
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.396834056448487,
+      "learning_rate": 7.330438986368124e-06,
+      "loss": 0.4092,
+      "step": 5254
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.749311603463243,
+      "learning_rate": 7.329442695063041e-06,
+      "loss": 0.4209,
+      "step": 5255
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.785068081171343,
+      "learning_rate": 7.32844628561622e-06,
+      "loss": 0.4325,
+      "step": 5256
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.365319199719984,
+      "learning_rate": 7.327449758078194e-06,
+      "loss": 0.369,
+      "step": 5257
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.730224250210268,
+      "learning_rate": 7.326453112499503e-06,
+      "loss": 0.4261,
+      "step": 5258
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.893651956959236,
+      "learning_rate": 7.325456348930696e-06,
+      "loss": 0.4321,
+      "step": 5259
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.1116676654021935,
+      "learning_rate": 7.324459467422323e-06,
+      "loss": 0.4478,
+      "step": 5260
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.358012542225023,
+      "learning_rate": 7.323462468024947e-06,
+      "loss": 0.3999,
+      "step": 5261
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.539329986894286,
+      "learning_rate": 7.322465350789126e-06,
+      "loss": 0.4184,
+      "step": 5262
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.64117657976516,
+      "learning_rate": 7.321468115765438e-06,
+      "loss": 0.411,
+      "step": 5263
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.30149223916292,
+      "learning_rate": 7.320470763004452e-06,
+      "loss": 0.4262,
+      "step": 5264
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.847142390256384,
+      "learning_rate": 7.319473292556756e-06,
+      "loss": 0.4448,
+      "step": 5265
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.461832605740517,
+      "learning_rate": 7.318475704472936e-06,
+      "loss": 0.4206,
+      "step": 5266
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.598015509228789,
+      "learning_rate": 7.3174779988035885e-06,
+      "loss": 0.3828,
+      "step": 5267
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.032782865063568,
+      "learning_rate": 7.31648017559931e-06,
+      "loss": 0.452,
+      "step": 5268
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.17372102410605,
+      "learning_rate": 7.3154822349107106e-06,
+      "loss": 0.4468,
+      "step": 5269
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.79532614816795,
+      "learning_rate": 7.314484176788399e-06,
+      "loss": 0.4287,
+      "step": 5270
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.48952881897602335,
+      "learning_rate": 7.313486001282998e-06,
+      "loss": 0.2198,
+      "step": 5271
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.630029345777275,
+      "learning_rate": 7.312487708445127e-06,
+      "loss": 0.3466,
+      "step": 5272
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.112860177245382,
+      "learning_rate": 7.311489298325422e-06,
+      "loss": 0.3861,
+      "step": 5273
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.381269200226625,
+      "learning_rate": 7.310490770974512e-06,
+      "loss": 0.3995,
+      "step": 5274
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.977826655585646,
+      "learning_rate": 7.309492126443045e-06,
+      "loss": 0.4176,
+      "step": 5275
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.169152032647585,
+      "learning_rate": 7.308493364781666e-06,
+      "loss": 0.4252,
+      "step": 5276
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.992385458609154,
+      "learning_rate": 7.30749448604103e-06,
+      "loss": 0.4297,
+      "step": 5277
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.557204881638633,
+      "learning_rate": 7.3064954902717966e-06,
+      "loss": 0.379,
+      "step": 5278
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.82860788338529,
+      "learning_rate": 7.305496377524632e-06,
+      "loss": 0.3977,
+      "step": 5279
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.572903662442498,
+      "learning_rate": 7.304497147850207e-06,
+      "loss": 0.4082,
+      "step": 5280
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.44375334797874166,
+      "learning_rate": 7.3034978012991996e-06,
+      "loss": 0.1947,
+      "step": 5281
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.5365251918128005,
+      "learning_rate": 7.302498337922293e-06,
+      "loss": 0.4033,
+      "step": 5282
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.394452103401344,
+      "learning_rate": 7.301498757770178e-06,
+      "loss": 0.4109,
+      "step": 5283
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.196910803882833,
+      "learning_rate": 7.300499060893549e-06,
+      "loss": 0.4054,
+      "step": 5284
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.344210407741854,
+      "learning_rate": 7.299499247343108e-06,
+      "loss": 0.4011,
+      "step": 5285
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.267494404178477,
+      "learning_rate": 7.298499317169562e-06,
+      "loss": 0.3488,
+      "step": 5286
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.973533569408163,
+      "learning_rate": 7.297499270423623e-06,
+      "loss": 0.4276,
+      "step": 5287
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.1188008217871035,
+      "learning_rate": 7.296499107156013e-06,
+      "loss": 0.4003,
+      "step": 5288
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.464971571258209,
+      "learning_rate": 7.295498827417455e-06,
+      "loss": 0.4361,
+      "step": 5289
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.750104825846698,
+      "learning_rate": 7.29449843125868e-06,
+      "loss": 0.3944,
+      "step": 5290
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.399023641548359,
+      "learning_rate": 7.293497918730425e-06,
+      "loss": 0.4141,
+      "step": 5291
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.504163189732115,
+      "learning_rate": 7.292497289883432e-06,
+      "loss": 0.4365,
+      "step": 5292
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.096965444669477,
+      "learning_rate": 7.291496544768452e-06,
+      "loss": 0.3795,
+      "step": 5293
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.405023978267389,
+      "learning_rate": 7.290495683436236e-06,
+      "loss": 0.4505,
+      "step": 5294
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 14.097050040888883,
+      "learning_rate": 7.289494705937548e-06,
+      "loss": 0.3986,
+      "step": 5295
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.669974874329238,
+      "learning_rate": 7.2884936123231506e-06,
+      "loss": 0.4038,
+      "step": 5296
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.870394285669018,
+      "learning_rate": 7.28749240264382e-06,
+      "loss": 0.4179,
+      "step": 5297
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.91580861506884,
+      "learning_rate": 7.286491076950333e-06,
+      "loss": 0.4631,
+      "step": 5298
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.615425362362172,
+      "learning_rate": 7.285489635293472e-06,
+      "loss": 0.4252,
+      "step": 5299
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.573639812342404,
+      "learning_rate": 7.284488077724028e-06,
+      "loss": 0.3804,
+      "step": 5300
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.312835486552853,
+      "learning_rate": 7.283486404292796e-06,
+      "loss": 0.383,
+      "step": 5301
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.712424165905305,
+      "learning_rate": 7.282484615050578e-06,
+      "loss": 0.4092,
+      "step": 5302
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.142483334271349,
+      "learning_rate": 7.2814827100481825e-06,
+      "loss": 0.4364,
+      "step": 5303
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.746798780607321,
+      "learning_rate": 7.280480689336423e-06,
+      "loss": 0.3755,
+      "step": 5304
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.215766717601808,
+      "learning_rate": 7.279478552966115e-06,
+      "loss": 0.3895,
+      "step": 5305
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.7825004893686405,
+      "learning_rate": 7.2784763009880875e-06,
+      "loss": 0.378,
+      "step": 5306
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.96487920175618,
+      "learning_rate": 7.27747393345317e-06,
+      "loss": 0.4018,
+      "step": 5307
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.873378567504661,
+      "learning_rate": 7.276471450412201e-06,
+      "loss": 0.4208,
+      "step": 5308
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.0798322347829075,
+      "learning_rate": 7.27546885191602e-06,
+      "loss": 0.4063,
+      "step": 5309
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.43325409166124,
+      "learning_rate": 7.274466138015479e-06,
+      "loss": 0.1937,
+      "step": 5310
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.883111374208665,
+      "learning_rate": 7.27346330876143e-06,
+      "loss": 0.4321,
+      "step": 5311
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.507639475323459,
+      "learning_rate": 7.272460364204734e-06,
+      "loss": 0.3847,
+      "step": 5312
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.9303689458151325,
+      "learning_rate": 7.271457304396256e-06,
+      "loss": 0.4108,
+      "step": 5313
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.237491047139487,
+      "learning_rate": 7.2704541293868705e-06,
+      "loss": 0.4002,
+      "step": 5314
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.889741706196569,
+      "learning_rate": 7.269450839227452e-06,
+      "loss": 0.4252,
+      "step": 5315
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.37737777582555204,
+      "learning_rate": 7.268447433968887e-06,
+      "loss": 0.1977,
+      "step": 5316
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.82467251936771,
+      "learning_rate": 7.2674439136620626e-06,
+      "loss": 0.4053,
+      "step": 5317
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.516541082358229,
+      "learning_rate": 7.266440278357876e-06,
+      "loss": 0.4405,
+      "step": 5318
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 24.57334480358495,
+      "learning_rate": 7.265436528107227e-06,
+      "loss": 0.4258,
+      "step": 5319
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.129357192966867,
+      "learning_rate": 7.264432662961023e-06,
+      "loss": 0.4263,
+      "step": 5320
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.4940572423336,
+      "learning_rate": 7.263428682970178e-06,
+      "loss": 0.3841,
+      "step": 5321
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 15.82970163440369,
+      "learning_rate": 7.2624245881856094e-06,
+      "loss": 0.3955,
+      "step": 5322
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.759022822362435,
+      "learning_rate": 7.261420378658239e-06,
+      "loss": 0.3904,
+      "step": 5323
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.380321052087075,
+      "learning_rate": 7.260416054439002e-06,
+      "loss": 0.39,
+      "step": 5324
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.222792367993261,
+      "learning_rate": 7.259411615578832e-06,
+      "loss": 0.4021,
+      "step": 5325
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.838307421046968,
+      "learning_rate": 7.2584070621286705e-06,
+      "loss": 0.3664,
+      "step": 5326
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.724989062972647,
+      "learning_rate": 7.257402394139466e-06,
+      "loss": 0.39,
+      "step": 5327
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.705519328299962,
+      "learning_rate": 7.256397611662172e-06,
+      "loss": 0.4032,
+      "step": 5328
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.27474940753378,
+      "learning_rate": 7.255392714747744e-06,
+      "loss": 0.428,
+      "step": 5329
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.580232280720121,
+      "learning_rate": 7.254387703447154e-06,
+      "loss": 0.4178,
+      "step": 5330
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.46762349639097,
+      "learning_rate": 7.253382577811368e-06,
+      "loss": 0.4252,
+      "step": 5331
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.90326298317649,
+      "learning_rate": 7.2523773378913655e-06,
+      "loss": 0.4181,
+      "step": 5332
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.383938828333415,
+      "learning_rate": 7.251371983738127e-06,
+      "loss": 0.464,
+      "step": 5333
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.907689959608816,
+      "learning_rate": 7.250366515402643e-06,
+      "loss": 0.4021,
+      "step": 5334
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.255852824819446,
+      "learning_rate": 7.2493609329359025e-06,
+      "loss": 0.4117,
+      "step": 5335
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.763987922162709,
+      "learning_rate": 7.248355236388911e-06,
+      "loss": 0.3977,
+      "step": 5336
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.243337594016169,
+      "learning_rate": 7.247349425812671e-06,
+      "loss": 0.3951,
+      "step": 5337
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.276528875228324,
+      "learning_rate": 7.246343501258195e-06,
+      "loss": 0.4084,
+      "step": 5338
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.398448646394709,
+      "learning_rate": 7.2453374627765e-06,
+      "loss": 0.4061,
+      "step": 5339
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.445382513981863,
+      "learning_rate": 7.244331310418611e-06,
+      "loss": 0.3777,
+      "step": 5340
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.4693617438735105,
+      "learning_rate": 7.243325044235552e-06,
+      "loss": 0.3739,
+      "step": 5341
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.909609231185147,
+      "learning_rate": 7.242318664278363e-06,
+      "loss": 0.3789,
+      "step": 5342
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.401162927077003,
+      "learning_rate": 7.241312170598079e-06,
+      "loss": 0.4213,
+      "step": 5343
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.70756644558486,
+      "learning_rate": 7.240305563245751e-06,
+      "loss": 0.4277,
+      "step": 5344
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.705195961503122,
+      "learning_rate": 7.239298842272429e-06,
+      "loss": 0.4711,
+      "step": 5345
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.901702457697324,
+      "learning_rate": 7.238292007729169e-06,
+      "loss": 0.3916,
+      "step": 5346
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.741940828602917,
+      "learning_rate": 7.237285059667036e-06,
+      "loss": 0.4254,
+      "step": 5347
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.029990272574427,
+      "learning_rate": 7.236277998137099e-06,
+      "loss": 0.4273,
+      "step": 5348
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.037605727999152,
+      "learning_rate": 7.235270823190431e-06,
+      "loss": 0.4444,
+      "step": 5349
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.342137816063232,
+      "learning_rate": 7.234263534878116e-06,
+      "loss": 0.423,
+      "step": 5350
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.766784042182827,
+      "learning_rate": 7.2332561332512375e-06,
+      "loss": 0.3963,
+      "step": 5351
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.497276743428245,
+      "learning_rate": 7.232248618360889e-06,
+      "loss": 0.4284,
+      "step": 5352
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.567414459647972,
+      "learning_rate": 7.231240990258169e-06,
+      "loss": 0.4044,
+      "step": 5353
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 12.535976674214298,
+      "learning_rate": 7.2302332489941805e-06,
+      "loss": 0.4266,
+      "step": 5354
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.421395132576445,
+      "learning_rate": 7.229225394620032e-06,
+      "loss": 0.4,
+      "step": 5355
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.87833132998367,
+      "learning_rate": 7.228217427186839e-06,
+      "loss": 0.4229,
+      "step": 5356
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.273022021525362,
+      "learning_rate": 7.2272093467457226e-06,
+      "loss": 0.3787,
+      "step": 5357
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.040851732845939,
+      "learning_rate": 7.226201153347809e-06,
+      "loss": 0.4027,
+      "step": 5358
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.501170155033516,
+      "learning_rate": 7.225192847044231e-06,
+      "loss": 0.3906,
+      "step": 5359
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.512655099646057,
+      "learning_rate": 7.224184427886125e-06,
+      "loss": 0.4373,
+      "step": 5360
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.283273645429808,
+      "learning_rate": 7.223175895924638e-06,
+      "loss": 0.4086,
+      "step": 5361
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.816451651951046,
+      "learning_rate": 7.222167251210916e-06,
+      "loss": 0.4285,
+      "step": 5362
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.875867401301685,
+      "learning_rate": 7.2211584937961175e-06,
+      "loss": 0.4065,
+      "step": 5363
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.385291992689701,
+      "learning_rate": 7.220149623731401e-06,
+      "loss": 0.4082,
+      "step": 5364
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.1647910279785325,
+      "learning_rate": 7.2191406410679325e-06,
+      "loss": 0.4143,
+      "step": 5365
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.9325556996677316,
+      "learning_rate": 7.2181315458568864e-06,
+      "loss": 0.4521,
+      "step": 5366
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.030014315236126,
+      "learning_rate": 7.217122338149441e-06,
+      "loss": 0.4051,
+      "step": 5367
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 16.38966511786335,
+      "learning_rate": 7.216113017996776e-06,
+      "loss": 0.4428,
+      "step": 5368
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.78361263178797,
+      "learning_rate": 7.2151035854500866e-06,
+      "loss": 0.4444,
+      "step": 5369
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.806578874001166,
+      "learning_rate": 7.214094040560563e-06,
+      "loss": 0.4239,
+      "step": 5370
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.841791406174159,
+      "learning_rate": 7.213084383379409e-06,
+      "loss": 0.4399,
+      "step": 5371
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.908761792757708,
+      "learning_rate": 7.212074613957829e-06,
+      "loss": 0.3965,
+      "step": 5372
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.384051611455771,
+      "learning_rate": 7.211064732347038e-06,
+      "loss": 0.4184,
+      "step": 5373
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.85807716231179,
+      "learning_rate": 7.2100547385982515e-06,
+      "loss": 0.4106,
+      "step": 5374
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.014562526604299,
+      "learning_rate": 7.209044632762694e-06,
+      "loss": 0.3885,
+      "step": 5375
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.604894105980445,
+      "learning_rate": 7.208034414891594e-06,
+      "loss": 0.3683,
+      "step": 5376
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 13.327548404004444,
+      "learning_rate": 7.207024085036188e-06,
+      "loss": 0.4047,
+      "step": 5377
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.481108971683543,
+      "learning_rate": 7.2060136432477145e-06,
+      "loss": 0.4112,
+      "step": 5378
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 6.9635219600174585,
+      "learning_rate": 7.2050030895774225e-06,
+      "loss": 0.3768,
+      "step": 5379
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.750489731178858,
+      "learning_rate": 7.203992424076561e-06,
+      "loss": 0.3922,
+      "step": 5380
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.647662515727657,
+      "learning_rate": 7.2029816467963906e-06,
+      "loss": 0.4426,
+      "step": 5381
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.985292639559086,
+      "learning_rate": 7.201970757788172e-06,
+      "loss": 0.4268,
+      "step": 5382
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 7.500762133301598,
+      "learning_rate": 7.200959757103176e-06,
+      "loss": 0.4289,
+      "step": 5383
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 21.583543276750092,
+      "learning_rate": 7.199948644792676e-06,
+      "loss": 0.4227,
+      "step": 5384
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38617743324622017,
+      "learning_rate": 7.1989374209079545e-06,
+      "loss": 0.194,
+      "step": 5385
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.201230133231174,
+      "learning_rate": 7.197926085500294e-06,
+      "loss": 0.4107,
+      "step": 5386
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 10.364825148241529,
+      "learning_rate": 7.19691463862099e-06,
+      "loss": 0.4205,
+      "step": 5387
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 8.341067196013544,
+      "learning_rate": 7.195903080321338e-06,
+      "loss": 0.431,
+      "step": 5388
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.003786220849098,
+      "learning_rate": 7.19489141065264e-06,
+      "loss": 0.4354,
+      "step": 5389
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.3407747151804699,
+      "learning_rate": 7.193879629666205e-06,
+      "loss": 0.2159,
+      "step": 5390
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.535037284217324,
+      "learning_rate": 7.192867737413348e-06,
+      "loss": 0.4339,
+      "step": 5391
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 11.750273104722412,
+      "learning_rate": 7.191855733945388e-06,
+      "loss": 0.397,
+      "step": 5392
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.327562462930828,
+      "learning_rate": 7.190843619313652e-06,
+      "loss": 0.4242,
+      "step": 5393
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.656146394112388,
+      "learning_rate": 7.189831393569469e-06,
+      "loss": 0.4071,
+      "step": 5394
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.089784772015882,
+      "learning_rate": 7.188819056764178e-06,
+      "loss": 0.4121,
+      "step": 5395
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.987936751436136,
+      "learning_rate": 7.187806608949119e-06,
+      "loss": 0.4083,
+      "step": 5396
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.73321547804367,
+      "learning_rate": 7.186794050175643e-06,
+      "loss": 0.392,
+      "step": 5397
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.965498451110855,
+      "learning_rate": 7.1857813804951014e-06,
+      "loss": 0.435,
+      "step": 5398
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.503071516594955,
+      "learning_rate": 7.184768599958854e-06,
+      "loss": 0.409,
+      "step": 5399
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.122088169029156,
+      "learning_rate": 7.1837557086182676e-06,
+      "loss": 0.3966,
+      "step": 5400
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.165427263543344,
+      "learning_rate": 7.18274270652471e-06,
+      "loss": 0.391,
+      "step": 5401
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.353973822272696,
+      "learning_rate": 7.181729593729558e-06,
+      "loss": 0.3906,
+      "step": 5402
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.718373936225099,
+      "learning_rate": 7.180716370284195e-06,
+      "loss": 0.4075,
+      "step": 5403
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.37026484269062243,
+      "learning_rate": 7.1797030362400056e-06,
+      "loss": 0.2058,
+      "step": 5404
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.34528149670905406,
+      "learning_rate": 7.1786895916483855e-06,
+      "loss": 0.2042,
+      "step": 5405
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.806944052999283,
+      "learning_rate": 7.177676036560732e-06,
+      "loss": 0.4072,
+      "step": 5406
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.710218942327198,
+      "learning_rate": 7.17666237102845e-06,
+      "loss": 0.406,
+      "step": 5407
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.944159248868325,
+      "learning_rate": 7.175648595102949e-06,
+      "loss": 0.4265,
+      "step": 5408
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.596452976083425,
+      "learning_rate": 7.174634708835644e-06,
+      "loss": 0.411,
+      "step": 5409
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.897775622641636,
+      "learning_rate": 7.1736207122779555e-06,
+      "loss": 0.4286,
+      "step": 5410
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.805157535306486,
+      "learning_rate": 7.172606605481311e-06,
+      "loss": 0.3661,
+      "step": 5411
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.5591637536721645,
+      "learning_rate": 7.171592388497144e-06,
+      "loss": 0.3684,
+      "step": 5412
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.623656713276729,
+      "learning_rate": 7.170578061376889e-06,
+      "loss": 0.3719,
+      "step": 5413
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.218317865671977,
+      "learning_rate": 7.169563624171993e-06,
+      "loss": 0.4024,
+      "step": 5414
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.6175309157146,
+      "learning_rate": 7.168549076933902e-06,
+      "loss": 0.4095,
+      "step": 5415
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.395555397788842,
+      "learning_rate": 7.167534419714072e-06,
+      "loss": 0.4241,
+      "step": 5416
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.662660656272517,
+      "learning_rate": 7.1665196525639635e-06,
+      "loss": 0.417,
+      "step": 5417
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.052517250413951,
+      "learning_rate": 7.16550477553504e-06,
+      "loss": 0.4317,
+      "step": 5418
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.746129364789756,
+      "learning_rate": 7.1644897886787745e-06,
+      "loss": 0.4353,
+      "step": 5419
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.802311738053546,
+      "learning_rate": 7.163474692046644e-06,
+      "loss": 0.4022,
+      "step": 5420
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.904227620361935,
+      "learning_rate": 7.162459485690131e-06,
+      "loss": 0.438,
+      "step": 5421
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.861080387483056,
+      "learning_rate": 7.161444169660723e-06,
+      "loss": 0.3877,
+      "step": 5422
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.280689239621255,
+      "learning_rate": 7.160428744009913e-06,
+      "loss": 0.3935,
+      "step": 5423
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.3398438313408505,
+      "learning_rate": 7.1594132087892e-06,
+      "loss": 0.3872,
+      "step": 5424
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.34208837233173,
+      "learning_rate": 7.158397564050089e-06,
+      "loss": 0.404,
+      "step": 5425
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.963347644430881,
+      "learning_rate": 7.15738180984409e-06,
+      "loss": 0.4075,
+      "step": 5426
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.9485119458077,
+      "learning_rate": 7.156365946222721e-06,
+      "loss": 0.4184,
+      "step": 5427
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.864933284775071,
+      "learning_rate": 7.155349973237499e-06,
+      "loss": 0.4155,
+      "step": 5428
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.099880150172267,
+      "learning_rate": 7.154333890939955e-06,
+      "loss": 0.389,
+      "step": 5429
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.33354125676315,
+      "learning_rate": 7.1533176993816185e-06,
+      "loss": 0.424,
+      "step": 5430
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.644982071810338,
+      "learning_rate": 7.15230139861403e-06,
+      "loss": 0.3928,
+      "step": 5431
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.012044437255764,
+      "learning_rate": 7.151284988688731e-06,
+      "loss": 0.4544,
+      "step": 5432
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.069422884498098,
+      "learning_rate": 7.150268469657271e-06,
+      "loss": 0.399,
+      "step": 5433
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.895025158355615,
+      "learning_rate": 7.149251841571204e-06,
+      "loss": 0.4247,
+      "step": 5434
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.9553055042667244,
+      "learning_rate": 7.1482351044820895e-06,
+      "loss": 0.4111,
+      "step": 5435
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.404207265586661,
+      "learning_rate": 7.147218258441496e-06,
+      "loss": 0.397,
+      "step": 5436
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 13.296824663391266,
+      "learning_rate": 7.146201303500992e-06,
+      "loss": 0.4128,
+      "step": 5437
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.6258865304712735,
+      "learning_rate": 7.1451842397121556e-06,
+      "loss": 0.4007,
+      "step": 5438
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.97583112891564,
+      "learning_rate": 7.144167067126566e-06,
+      "loss": 0.3811,
+      "step": 5439
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.053058391379016,
+      "learning_rate": 7.143149785795815e-06,
+      "loss": 0.386,
+      "step": 5440
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5118150177676634,
+      "learning_rate": 7.1421323957714925e-06,
+      "loss": 0.2176,
+      "step": 5441
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.356638518524592,
+      "learning_rate": 7.141114897105202e-06,
+      "loss": 0.4237,
+      "step": 5442
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.244877888150813,
+      "learning_rate": 7.140097289848541e-06,
+      "loss": 0.3992,
+      "step": 5443
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.516744016397412,
+      "learning_rate": 7.139079574053125e-06,
+      "loss": 0.3836,
+      "step": 5444
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.764604280279145,
+      "learning_rate": 7.138061749770564e-06,
+      "loss": 0.4146,
+      "step": 5445
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.9297631898454775,
+      "learning_rate": 7.1370438170524835e-06,
+      "loss": 0.385,
+      "step": 5446
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.951520429883058,
+      "learning_rate": 7.136025775950507e-06,
+      "loss": 0.4089,
+      "step": 5447
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.408463991204865,
+      "learning_rate": 7.135007626516268e-06,
+      "loss": 0.4126,
+      "step": 5448
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.05682398011279,
+      "learning_rate": 7.133989368801402e-06,
+      "loss": 0.4485,
+      "step": 5449
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.862150725514324,
+      "learning_rate": 7.132971002857552e-06,
+      "loss": 0.4296,
+      "step": 5450
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.880863200811651,
+      "learning_rate": 7.131952528736367e-06,
+      "loss": 0.4171,
+      "step": 5451
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.836960866753584,
+      "learning_rate": 7.1309339464895e-06,
+      "loss": 0.3962,
+      "step": 5452
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.729858337585119,
+      "learning_rate": 7.12991525616861e-06,
+      "loss": 0.4276,
+      "step": 5453
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 28.911115556347802,
+      "learning_rate": 7.128896457825364e-06,
+      "loss": 0.3919,
+      "step": 5454
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.256066477987255,
+      "learning_rate": 7.1278775515114296e-06,
+      "loss": 0.4055,
+      "step": 5455
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.051589614423957,
+      "learning_rate": 7.1268585372784826e-06,
+      "loss": 0.4539,
+      "step": 5456
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.616933331736782,
+      "learning_rate": 7.125839415178204e-06,
+      "loss": 0.3824,
+      "step": 5457
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.6495234557519645,
+      "learning_rate": 7.124820185262282e-06,
+      "loss": 0.408,
+      "step": 5458
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 10.196857456717494,
+      "learning_rate": 7.123800847582407e-06,
+      "loss": 0.4014,
+      "step": 5459
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.606384938374651,
+      "learning_rate": 7.122781402190277e-06,
+      "loss": 0.4037,
+      "step": 5460
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.579263046130541,
+      "learning_rate": 7.121761849137594e-06,
+      "loss": 0.4137,
+      "step": 5461
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.414787004066804,
+      "learning_rate": 7.120742188476069e-06,
+      "loss": 0.4033,
+      "step": 5462
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.275418197711156,
+      "learning_rate": 7.1197224202574135e-06,
+      "loss": 0.3996,
+      "step": 5463
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.845919566981078,
+      "learning_rate": 7.118702544533348e-06,
+      "loss": 0.4673,
+      "step": 5464
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.995605652633808,
+      "learning_rate": 7.117682561355596e-06,
+      "loss": 0.3927,
+      "step": 5465
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.608526258229592,
+      "learning_rate": 7.116662470775889e-06,
+      "loss": 0.3962,
+      "step": 5466
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.156982424834403,
+      "learning_rate": 7.115642272845963e-06,
+      "loss": 0.4113,
+      "step": 5467
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5590296600463737,
+      "learning_rate": 7.1146219676175575e-06,
+      "loss": 0.2151,
+      "step": 5468
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 12.348617319253655,
+      "learning_rate": 7.11360155514242e-06,
+      "loss": 0.4289,
+      "step": 5469
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.117144052717797,
+      "learning_rate": 7.1125810354723016e-06,
+      "loss": 0.3927,
+      "step": 5470
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.394334944055897,
+      "learning_rate": 7.1115604086589605e-06,
+      "loss": 0.4107,
+      "step": 5471
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.50736115716845,
+      "learning_rate": 7.11053967475416e-06,
+      "loss": 0.4032,
+      "step": 5472
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.902438273142784,
+      "learning_rate": 7.109518833809668e-06,
+      "loss": 0.3975,
+      "step": 5473
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.49267245245203,
+      "learning_rate": 7.108497885877259e-06,
+      "loss": 0.4133,
+      "step": 5474
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.642264128307763,
+      "learning_rate": 7.10747683100871e-06,
+      "loss": 0.4325,
+      "step": 5475
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.123708852941979,
+      "learning_rate": 7.106455669255807e-06,
+      "loss": 0.4124,
+      "step": 5476
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 37.545428806305914,
+      "learning_rate": 7.105434400670341e-06,
+      "loss": 0.4187,
+      "step": 5477
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.907938453564955,
+      "learning_rate": 7.104413025304106e-06,
+      "loss": 0.3953,
+      "step": 5478
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 15.811403572198808,
+      "learning_rate": 7.103391543208903e-06,
+      "loss": 0.4264,
+      "step": 5479
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.898745252480878,
+      "learning_rate": 7.102369954436539e-06,
+      "loss": 0.4189,
+      "step": 5480
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.648546016160272,
+      "learning_rate": 7.101348259038823e-06,
+      "loss": 0.3844,
+      "step": 5481
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.24767727761796,
+      "learning_rate": 7.100326457067576e-06,
+      "loss": 0.4082,
+      "step": 5482
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.452941266014825,
+      "learning_rate": 7.099304548574617e-06,
+      "loss": 0.3812,
+      "step": 5483
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.211580641719308,
+      "learning_rate": 7.098282533611777e-06,
+      "loss": 0.3844,
+      "step": 5484
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.169517353955689,
+      "learning_rate": 7.0972604122308865e-06,
+      "loss": 0.4605,
+      "step": 5485
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 14.402799343394097,
+      "learning_rate": 7.096238184483786e-06,
+      "loss": 0.4444,
+      "step": 5486
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.2409143472031765,
+      "learning_rate": 7.095215850422318e-06,
+      "loss": 0.3667,
+      "step": 5487
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5203614667345012,
+      "learning_rate": 7.094193410098333e-06,
+      "loss": 0.2148,
+      "step": 5488
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.02239984135241,
+      "learning_rate": 7.093170863563688e-06,
+      "loss": 0.4075,
+      "step": 5489
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 11.944234822558894,
+      "learning_rate": 7.092148210870238e-06,
+      "loss": 0.4374,
+      "step": 5490
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.254338461499253,
+      "learning_rate": 7.091125452069853e-06,
+      "loss": 0.4039,
+      "step": 5491
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.838285438671714,
+      "learning_rate": 7.090102587214402e-06,
+      "loss": 0.3679,
+      "step": 5492
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.633235719940695,
+      "learning_rate": 7.0890796163557625e-06,
+      "loss": 0.4203,
+      "step": 5493
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.949664308300303,
+      "learning_rate": 7.088056539545815e-06,
+      "loss": 0.4145,
+      "step": 5494
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.325191801444929,
+      "learning_rate": 7.087033356836448e-06,
+      "loss": 0.4338,
+      "step": 5495
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.8066089864857435,
+      "learning_rate": 7.086010068279553e-06,
+      "loss": 0.4194,
+      "step": 5496
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.031295184491781,
+      "learning_rate": 7.084986673927028e-06,
+      "loss": 0.4046,
+      "step": 5497
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.484047528227244,
+      "learning_rate": 7.083963173830775e-06,
+      "loss": 0.406,
+      "step": 5498
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.564835877392661,
+      "learning_rate": 7.082939568042706e-06,
+      "loss": 0.3844,
+      "step": 5499
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.929882465041963,
+      "learning_rate": 7.081915856614733e-06,
+      "loss": 0.4118,
+      "step": 5500
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.49848546860599,
+      "learning_rate": 7.080892039598774e-06,
+      "loss": 0.3976,
+      "step": 5501
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 9.06152238364278,
+      "learning_rate": 7.079868117046755e-06,
+      "loss": 0.4358,
+      "step": 5502
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.3164870942122375,
+      "learning_rate": 7.078844089010607e-06,
+      "loss": 0.3541,
+      "step": 5503
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.252127971663864,
+      "learning_rate": 7.077819955542262e-06,
+      "loss": 0.406,
+      "step": 5504
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.17382267607268,
+      "learning_rate": 7.076795716693665e-06,
+      "loss": 0.3964,
+      "step": 5505
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.457431998107647,
+      "learning_rate": 7.075771372516759e-06,
+      "loss": 0.4308,
+      "step": 5506
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.614230889299753,
+      "learning_rate": 7.074746923063497e-06,
+      "loss": 0.3911,
+      "step": 5507
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.744919983724484,
+      "learning_rate": 7.073722368385833e-06,
+      "loss": 0.4243,
+      "step": 5508
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.600575277568147,
+      "learning_rate": 7.072697708535734e-06,
+      "loss": 0.4034,
+      "step": 5509
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.855349277402515,
+      "learning_rate": 7.071672943565164e-06,
+      "loss": 0.4152,
+      "step": 5510
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.307419035342346,
+      "learning_rate": 7.070648073526096e-06,
+      "loss": 0.4241,
+      "step": 5511
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.893239348359276,
+      "learning_rate": 7.069623098470508e-06,
+      "loss": 0.4226,
+      "step": 5512
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.556494902661889,
+      "learning_rate": 7.068598018450386e-06,
+      "loss": 0.4272,
+      "step": 5513
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.52836699724854,
+      "learning_rate": 7.067572833517713e-06,
+      "loss": 0.3857,
+      "step": 5514
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.954971668610774,
+      "learning_rate": 7.066547543724489e-06,
+      "loss": 0.4195,
+      "step": 5515
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.916372339883802,
+      "learning_rate": 7.06552214912271e-06,
+      "loss": 0.4195,
+      "step": 5516
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 8.72473457441375,
+      "learning_rate": 7.064496649764381e-06,
+      "loss": 0.3961,
+      "step": 5517
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.314891797949641,
+      "learning_rate": 7.063471045701513e-06,
+      "loss": 0.3852,
+      "step": 5518
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5266010507773082,
+      "learning_rate": 7.062445336986121e-06,
+      "loss": 0.2004,
+      "step": 5519
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.2166370811217835,
+      "learning_rate": 7.0614195236702255e-06,
+      "loss": 0.4503,
+      "step": 5520
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.877571377809386,
+      "learning_rate": 7.060393605805853e-06,
+      "loss": 0.455,
+      "step": 5521
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.309636361203786,
+      "learning_rate": 7.059367583445034e-06,
+      "loss": 0.4195,
+      "step": 5522
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.825689418688869,
+      "learning_rate": 7.058341456639806e-06,
+      "loss": 0.464,
+      "step": 5523
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.36974772388131,
+      "learning_rate": 7.057315225442207e-06,
+      "loss": 0.3821,
+      "step": 5524
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.324496763060697,
+      "learning_rate": 7.056288889904289e-06,
+      "loss": 0.4347,
+      "step": 5525
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.092470371405415,
+      "learning_rate": 7.055262450078102e-06,
+      "loss": 0.4226,
+      "step": 5526
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.426778271365022,
+      "learning_rate": 7.054235906015704e-06,
+      "loss": 0.4094,
+      "step": 5527
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.361530900903571,
+      "learning_rate": 7.053209257769157e-06,
+      "loss": 0.4197,
+      "step": 5528
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 6.12160556425555,
+      "learning_rate": 7.0521825053905326e-06,
+      "loss": 0.4619,
+      "step": 5529
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.279742755205019,
+      "learning_rate": 7.0511556489319e-06,
+      "loss": 0.4052,
+      "step": 5530
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 16.12102132823972,
+      "learning_rate": 7.0501286884453415e-06,
+      "loss": 0.4168,
+      "step": 5531
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.07343019510535,
+      "learning_rate": 7.049101623982938e-06,
+      "loss": 0.4168,
+      "step": 5532
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5066622656391284,
+      "learning_rate": 7.048074455596782e-06,
+      "loss": 0.2096,
+      "step": 5533
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.355635674179831,
+      "learning_rate": 7.047047183338967e-06,
+      "loss": 0.404,
+      "step": 5534
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.882551457196046,
+      "learning_rate": 7.046019807261593e-06,
+      "loss": 0.4212,
+      "step": 5535
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.541899505756805,
+      "learning_rate": 7.044992327416762e-06,
+      "loss": 0.3818,
+      "step": 5536
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.842550209730029,
+      "learning_rate": 7.0439647438565895e-06,
+      "loss": 0.4304,
+      "step": 5537
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.258688127537328,
+      "learning_rate": 7.042937056633188e-06,
+      "loss": 0.4113,
+      "step": 5538
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.287670393156383,
+      "learning_rate": 7.0419092657986796e-06,
+      "loss": 0.3744,
+      "step": 5539
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.0350146620977885,
+      "learning_rate": 7.04088137140519e-06,
+      "loss": 0.421,
+      "step": 5540
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.421040608143513,
+      "learning_rate": 7.039853373504851e-06,
+      "loss": 0.4034,
+      "step": 5541
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.3635267181606805,
+      "learning_rate": 7.0388252721498e-06,
+      "loss": 0.4058,
+      "step": 5542
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.50544365840271,
+      "learning_rate": 7.037797067392178e-06,
+      "loss": 0.4283,
+      "step": 5543
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.65336201894323,
+      "learning_rate": 7.036768759284131e-06,
+      "loss": 0.417,
+      "step": 5544
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.145918484210702,
+      "learning_rate": 7.035740347877815e-06,
+      "loss": 0.4354,
+      "step": 5545
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.526065678623729,
+      "learning_rate": 7.034711833225384e-06,
+      "loss": 0.4707,
+      "step": 5546
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 19.1618277186721,
+      "learning_rate": 7.033683215379002e-06,
+      "loss": 0.3976,
+      "step": 5547
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.079249918122431,
+      "learning_rate": 7.032654494390838e-06,
+      "loss": 0.4007,
+      "step": 5548
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.287329291575045,
+      "learning_rate": 7.031625670313065e-06,
+      "loss": 0.425,
+      "step": 5549
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 12.095929489472065,
+      "learning_rate": 7.030596743197862e-06,
+      "loss": 0.395,
+      "step": 5550
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.2043392899851915,
+      "learning_rate": 7.029567713097412e-06,
+      "loss": 0.4082,
+      "step": 5551
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.911095277479159,
+      "learning_rate": 7.028538580063903e-06,
+      "loss": 0.4312,
+      "step": 5552
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.066484812505748,
+      "learning_rate": 7.02750934414953e-06,
+      "loss": 0.4255,
+      "step": 5553
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.166617697692315,
+      "learning_rate": 7.026480005406496e-06,
+      "loss": 0.3967,
+      "step": 5554
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.42347578033247274,
+      "learning_rate": 7.025450563887002e-06,
+      "loss": 0.2138,
+      "step": 5555
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.166892785303816,
+      "learning_rate": 7.024421019643259e-06,
+      "loss": 0.3968,
+      "step": 5556
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.921854439297241,
+      "learning_rate": 7.02339137272748e-06,
+      "loss": 0.4204,
+      "step": 5557
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.341843009439446,
+      "learning_rate": 7.022361623191889e-06,
+      "loss": 0.4322,
+      "step": 5558
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7101259475870316,
+      "learning_rate": 7.021331771088708e-06,
+      "loss": 0.42,
+      "step": 5559
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3427972252970257,
+      "learning_rate": 7.020301816470172e-06,
+      "loss": 0.1898,
+      "step": 5560
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.131067506961215,
+      "learning_rate": 7.019271759388513e-06,
+      "loss": 0.4052,
+      "step": 5561
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.2422677137835,
+      "learning_rate": 7.018241599895974e-06,
+      "loss": 0.3948,
+      "step": 5562
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.2351758072718,
+      "learning_rate": 7.0172113380448e-06,
+      "loss": 0.4124,
+      "step": 5563
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.200006478874086,
+      "learning_rate": 7.016180973887244e-06,
+      "loss": 0.4131,
+      "step": 5564
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.197436537735924,
+      "learning_rate": 7.015150507475564e-06,
+      "loss": 0.4243,
+      "step": 5565
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.642687963489299,
+      "learning_rate": 7.0141199388620165e-06,
+      "loss": 0.4762,
+      "step": 5566
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.173415575085913,
+      "learning_rate": 7.013089268098874e-06,
+      "loss": 0.4007,
+      "step": 5567
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 14.046775324036922,
+      "learning_rate": 7.012058495238408e-06,
+      "loss": 0.3856,
+      "step": 5568
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.74227590537535,
+      "learning_rate": 7.011027620332893e-06,
+      "loss": 0.4263,
+      "step": 5569
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.628495944117145,
+      "learning_rate": 7.009996643434613e-06,
+      "loss": 0.4164,
+      "step": 5570
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44576867625431343,
+      "learning_rate": 7.008965564595857e-06,
+      "loss": 0.2084,
+      "step": 5571
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.521025677169133,
+      "learning_rate": 7.007934383868917e-06,
+      "loss": 0.4126,
+      "step": 5572
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.000522308824422,
+      "learning_rate": 7.006903101306091e-06,
+      "loss": 0.3767,
+      "step": 5573
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.398226274546043,
+      "learning_rate": 7.0058717169596845e-06,
+      "loss": 0.4152,
+      "step": 5574
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.8564416283069205,
+      "learning_rate": 7.004840230882002e-06,
+      "loss": 0.4171,
+      "step": 5575
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.537418140690025,
+      "learning_rate": 7.003808643125361e-06,
+      "loss": 0.4061,
+      "step": 5576
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.878342157963153,
+      "learning_rate": 7.002776953742078e-06,
+      "loss": 0.3939,
+      "step": 5577
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4960073814226686,
+      "learning_rate": 7.0017451627844765e-06,
+      "loss": 0.4149,
+      "step": 5578
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.24234820520591,
+      "learning_rate": 7.000713270304885e-06,
+      "loss": 0.4601,
+      "step": 5579
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.96409541769025,
+      "learning_rate": 6.999681276355641e-06,
+      "loss": 0.4062,
+      "step": 5580
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.057882215466905,
+      "learning_rate": 6.998649180989081e-06,
+      "loss": 0.4288,
+      "step": 5581
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.63591360791563,
+      "learning_rate": 6.9976169842575526e-06,
+      "loss": 0.4531,
+      "step": 5582
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.9532426770491735,
+      "learning_rate": 6.996584686213401e-06,
+      "loss": 0.3975,
+      "step": 5583
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.754615391262405,
+      "learning_rate": 6.9955522869089844e-06,
+      "loss": 0.4358,
+      "step": 5584
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.207926655121993,
+      "learning_rate": 6.9945197863966606e-06,
+      "loss": 0.4152,
+      "step": 5585
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.421023232649572,
+      "learning_rate": 6.9934871847287975e-06,
+      "loss": 0.4397,
+      "step": 5586
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.947924372323863,
+      "learning_rate": 6.992454481957762e-06,
+      "loss": 0.4261,
+      "step": 5587
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.036273739578165,
+      "learning_rate": 6.9914216781359324e-06,
+      "loss": 0.372,
+      "step": 5588
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.251085546239012,
+      "learning_rate": 6.990388773315687e-06,
+      "loss": 0.397,
+      "step": 5589
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8714129449078283,
+      "learning_rate": 6.989355767549413e-06,
+      "loss": 0.3758,
+      "step": 5590
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.020709947712186,
+      "learning_rate": 6.988322660889498e-06,
+      "loss": 0.4156,
+      "step": 5591
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.602159684193099,
+      "learning_rate": 6.9872894533883415e-06,
+      "loss": 0.4272,
+      "step": 5592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9164147533202294,
+      "learning_rate": 6.986256145098342e-06,
+      "loss": 0.4199,
+      "step": 5593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.439152651798667,
+      "learning_rate": 6.985222736071908e-06,
+      "loss": 0.3837,
+      "step": 5594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8997348128842857,
+      "learning_rate": 6.984189226361446e-06,
+      "loss": 0.4172,
+      "step": 5595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 13.483795670789345,
+      "learning_rate": 6.9831556160193795e-06,
+      "loss": 0.4111,
+      "step": 5596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.3655986059535214,
+      "learning_rate": 6.982121905098122e-06,
+      "loss": 0.3828,
+      "step": 5597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.0425776292589735,
+      "learning_rate": 6.981088093650104e-06,
+      "loss": 0.3931,
+      "step": 5598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.129233187872224,
+      "learning_rate": 6.980054181727759e-06,
+      "loss": 0.369,
+      "step": 5599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.3345815935867344,
+      "learning_rate": 6.979020169383519e-06,
+      "loss": 0.4236,
+      "step": 5600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.044711510232808,
+      "learning_rate": 6.977986056669829e-06,
+      "loss": 0.4227,
+      "step": 5601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.50163202192675,
+      "learning_rate": 6.976951843639135e-06,
+      "loss": 0.3831,
+      "step": 5602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.370285304535516,
+      "learning_rate": 6.975917530343887e-06,
+      "loss": 0.4044,
+      "step": 5603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.725625591668438,
+      "learning_rate": 6.974883116836546e-06,
+      "loss": 0.4088,
+      "step": 5604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.627823939340518,
+      "learning_rate": 6.97384860316957e-06,
+      "loss": 0.3812,
+      "step": 5605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.601969046457559,
+      "learning_rate": 6.972813989395429e-06,
+      "loss": 0.413,
+      "step": 5606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.135489520474719,
+      "learning_rate": 6.971779275566593e-06,
+      "loss": 0.4211,
+      "step": 5607
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.932164881819491,
+      "learning_rate": 6.9707444617355425e-06,
+      "loss": 0.3983,
+      "step": 5608
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 10.13734717866218,
+      "learning_rate": 6.9697095479547564e-06,
+      "loss": 0.3972,
+      "step": 5609
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.591471809819587,
+      "learning_rate": 6.968674534276724e-06,
+      "loss": 0.3995,
+      "step": 5610
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.162103302437942,
+      "learning_rate": 6.9676394207539375e-06,
+      "loss": 0.3954,
+      "step": 5611
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 17.939736320720066,
+      "learning_rate": 6.966604207438895e-06,
+      "loss": 0.4035,
+      "step": 5612
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.299571515891604,
+      "learning_rate": 6.965568894384098e-06,
+      "loss": 0.4475,
+      "step": 5613
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.679678819304252,
+      "learning_rate": 6.964533481642055e-06,
+      "loss": 0.4023,
+      "step": 5614
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.233671180129566,
+      "learning_rate": 6.963497969265278e-06,
+      "loss": 0.3903,
+      "step": 5615
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.287771908068208,
+      "learning_rate": 6.962462357306286e-06,
+      "loss": 0.4279,
+      "step": 5616
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.205888297180847,
+      "learning_rate": 6.9614266458176e-06,
+      "loss": 0.4202,
+      "step": 5617
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.418586652115281,
+      "learning_rate": 6.96039083485175e-06,
+      "loss": 0.4075,
+      "step": 5618
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.308199538764324,
+      "learning_rate": 6.959354924461269e-06,
+      "loss": 0.3962,
+      "step": 5619
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5281372449697406,
+      "learning_rate": 6.958318914698693e-06,
+      "loss": 0.3813,
+      "step": 5620
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.08107345458769,
+      "learning_rate": 6.957282805616567e-06,
+      "loss": 0.4252,
+      "step": 5621
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.40778666116620704,
+      "learning_rate": 6.956246597267438e-06,
+      "loss": 0.2146,
+      "step": 5622
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.296832102406445,
+      "learning_rate": 6.95521028970386e-06,
+      "loss": 0.3935,
+      "step": 5623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7671890364894978,
+      "learning_rate": 6.954173882978388e-06,
+      "loss": 0.4264,
+      "step": 5624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.201116547402052,
+      "learning_rate": 6.95313737714359e-06,
+      "loss": 0.4255,
+      "step": 5625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.4708016684652305,
+      "learning_rate": 6.952100772252031e-06,
+      "loss": 0.4342,
+      "step": 5626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.486527176879768,
+      "learning_rate": 6.951064068356285e-06,
+      "loss": 0.4128,
+      "step": 5627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.368529250746333,
+      "learning_rate": 6.950027265508929e-06,
+      "loss": 0.1993,
+      "step": 5628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.7196426876620885,
+      "learning_rate": 6.948990363762549e-06,
+      "loss": 0.3737,
+      "step": 5629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.183006378023702,
+      "learning_rate": 6.9479533631697316e-06,
+      "loss": 0.3864,
+      "step": 5630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.753622622426723,
+      "learning_rate": 6.94691626378307e-06,
+      "loss": 0.3428,
+      "step": 5631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.277213912259648,
+      "learning_rate": 6.945879065655164e-06,
+      "loss": 0.3929,
+      "step": 5632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.614698622968544,
+      "learning_rate": 6.944841768838615e-06,
+      "loss": 0.4037,
+      "step": 5633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.496069084293762,
+      "learning_rate": 6.943804373386032e-06,
+      "loss": 0.4192,
+      "step": 5634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 22.255329112130173,
+      "learning_rate": 6.942766879350029e-06,
+      "loss": 0.4178,
+      "step": 5635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.3272766433541556,
+      "learning_rate": 6.941729286783222e-06,
+      "loss": 0.4202,
+      "step": 5636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.50745326501857,
+      "learning_rate": 6.940691595738237e-06,
+      "loss": 0.4131,
+      "step": 5637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.9731206778124,
+      "learning_rate": 6.939653806267701e-06,
+      "loss": 0.4198,
+      "step": 5638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.603575713019232,
+      "learning_rate": 6.938615918424248e-06,
+      "loss": 0.3926,
+      "step": 5639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.41247343099771855,
+      "learning_rate": 6.9375779322605154e-06,
+      "loss": 0.2012,
+      "step": 5640
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.684277287608869,
+      "learning_rate": 6.936539847829147e-06,
+      "loss": 0.4451,
+      "step": 5641
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.8357822299024065,
+      "learning_rate": 6.935501665182791e-06,
+      "loss": 0.3826,
+      "step": 5642
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.571396273636788,
+      "learning_rate": 6.934463384374102e-06,
+      "loss": 0.4085,
+      "step": 5643
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.8513423556273776,
+      "learning_rate": 6.9334250054557354e-06,
+      "loss": 0.4081,
+      "step": 5644
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 11.74344639881525,
+      "learning_rate": 6.932386528480357e-06,
+      "loss": 0.3994,
+      "step": 5645
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.804459910480878,
+      "learning_rate": 6.931347953500632e-06,
+      "loss": 0.3919,
+      "step": 5646
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.853372413229044,
+      "learning_rate": 6.930309280569236e-06,
+      "loss": 0.4053,
+      "step": 5647
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.16376472177093,
+      "learning_rate": 6.929270509738847e-06,
+      "loss": 0.4202,
+      "step": 5648
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.996517906658066,
+      "learning_rate": 6.9282316410621476e-06,
+      "loss": 0.3994,
+      "step": 5649
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.139663961195154,
+      "learning_rate": 6.927192674591825e-06,
+      "loss": 0.3825,
+      "step": 5650
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.588623209750367,
+      "learning_rate": 6.9261536103805735e-06,
+      "loss": 0.41,
+      "step": 5651
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.660518169849267,
+      "learning_rate": 6.925114448481089e-06,
+      "loss": 0.4495,
+      "step": 5652
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3824592835554081,
+      "learning_rate": 6.924075188946079e-06,
+      "loss": 0.2011,
+      "step": 5653
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 8.243785469161427,
+      "learning_rate": 6.9230358318282454e-06,
+      "loss": 0.446,
+      "step": 5654
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 6.568106123465224,
+      "learning_rate": 6.921996377180305e-06,
+      "loss": 0.4466,
+      "step": 5655
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.940162301628218,
+      "learning_rate": 6.9209568250549765e-06,
+      "loss": 0.4241,
+      "step": 5656
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 9.96874751841989,
+      "learning_rate": 6.919917175504978e-06,
+      "loss": 0.4686,
+      "step": 5657
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.011579600379092,
+      "learning_rate": 6.91887742858304e-06,
+      "loss": 0.4047,
+      "step": 5658
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.151521556774672,
+      "learning_rate": 6.9178375843418956e-06,
+      "loss": 0.3932,
+      "step": 5659
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.25393199986534,
+      "learning_rate": 6.91679764283428e-06,
+      "loss": 0.4013,
+      "step": 5660
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.562329902174593,
+      "learning_rate": 6.915757604112938e-06,
+      "loss": 0.3648,
+      "step": 5661
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3729693996822226,
+      "learning_rate": 6.914717468230615e-06,
+      "loss": 0.1961,
+      "step": 5662
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.316288998689827,
+      "learning_rate": 6.913677235240066e-06,
+      "loss": 0.3957,
+      "step": 5663
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.36930555884126,
+      "learning_rate": 6.912636905194044e-06,
+      "loss": 0.3972,
+      "step": 5664
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.771546210648668,
+      "learning_rate": 6.911596478145317e-06,
+      "loss": 0.384,
+      "step": 5665
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.370117499926039,
+      "learning_rate": 6.910555954146647e-06,
+      "loss": 0.4243,
+      "step": 5666
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.4575398905330057,
+      "learning_rate": 6.909515333250809e-06,
+      "loss": 0.4092,
+      "step": 5667
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.7297339095910615,
+      "learning_rate": 6.908474615510576e-06,
+      "loss": 0.4101,
+      "step": 5668
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.718757792121937,
+      "learning_rate": 6.9074338009787355e-06,
+      "loss": 0.4023,
+      "step": 5669
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.436873888225965,
+      "learning_rate": 6.906392889708069e-06,
+      "loss": 0.4106,
+      "step": 5670
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.237196664447986,
+      "learning_rate": 6.905351881751372e-06,
+      "loss": 0.4381,
+      "step": 5671
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.995834543420085,
+      "learning_rate": 6.904310777161438e-06,
+      "loss": 0.4108,
+      "step": 5672
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.197217517496961,
+      "learning_rate": 6.903269575991071e-06,
+      "loss": 0.3995,
+      "step": 5673
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 7.366603467418724,
+      "learning_rate": 6.902228278293075e-06,
+      "loss": 0.4434,
+      "step": 5674
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 15.763021558135527,
+      "learning_rate": 6.901186884120264e-06,
+      "loss": 0.3838,
+      "step": 5675
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.646410610492553,
+      "learning_rate": 6.900145393525449e-06,
+      "loss": 0.4207,
+      "step": 5676
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.770948365291174,
+      "learning_rate": 6.899103806561458e-06,
+      "loss": 0.4538,
+      "step": 5677
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.214300044659968,
+      "learning_rate": 6.898062123281113e-06,
+      "loss": 0.4069,
+      "step": 5678
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 5.002702749006121,
+      "learning_rate": 6.897020343737244e-06,
+      "loss": 0.4038,
+      "step": 5679
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.5110845739080716,
+      "learning_rate": 6.895978467982687e-06,
+      "loss": 0.3532,
+      "step": 5680
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38880407018842295,
+      "learning_rate": 6.894936496070284e-06,
+      "loss": 0.2011,
+      "step": 5681
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.939134241559522,
+      "learning_rate": 6.893894428052881e-06,
+      "loss": 0.4316,
+      "step": 5682
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.998149487318448,
+      "learning_rate": 6.8928522639833255e-06,
+      "loss": 0.4386,
+      "step": 5683
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.931844331069037,
+      "learning_rate": 6.891810003914475e-06,
+      "loss": 0.4474,
+      "step": 5684
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.585246603114602,
+      "learning_rate": 6.890767647899189e-06,
+      "loss": 0.4147,
+      "step": 5685
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.301050589042923,
+      "learning_rate": 6.889725195990333e-06,
+      "loss": 0.4106,
+      "step": 5686
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.228463088744887,
+      "learning_rate": 6.888682648240773e-06,
+      "loss": 0.4305,
+      "step": 5687
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.2298694820404257,
+      "learning_rate": 6.887640004703389e-06,
+      "loss": 0.3987,
+      "step": 5688
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.533923372088039,
+      "learning_rate": 6.8865972654310584e-06,
+      "loss": 0.4068,
+      "step": 5689
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.9628679068949495,
+      "learning_rate": 6.885554430476664e-06,
+      "loss": 0.4017,
+      "step": 5690
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.47277707952345,
+      "learning_rate": 6.8845114998930965e-06,
+      "loss": 0.3795,
+      "step": 5691
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.441132667685079,
+      "learning_rate": 6.883468473733249e-06,
+      "loss": 0.4376,
+      "step": 5692
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.422343152431929,
+      "learning_rate": 6.882425352050021e-06,
+      "loss": 0.3956,
+      "step": 5693
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.564915354113813,
+      "learning_rate": 6.881382134896317e-06,
+      "loss": 0.4339,
+      "step": 5694
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.229938816634739,
+      "learning_rate": 6.880338822325043e-06,
+      "loss": 0.3975,
+      "step": 5695
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.441421927970226,
+      "learning_rate": 6.879295414389117e-06,
+      "loss": 0.4188,
+      "step": 5696
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.978757680137073,
+      "learning_rate": 6.8782519111414515e-06,
+      "loss": 0.4323,
+      "step": 5697
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.2296470227793215,
+      "learning_rate": 6.877208312634975e-06,
+      "loss": 0.4253,
+      "step": 5698
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.361018087761721,
+      "learning_rate": 6.876164618922612e-06,
+      "loss": 0.3927,
+      "step": 5699
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.270893863460063,
+      "learning_rate": 6.875120830057297e-06,
+      "loss": 0.4149,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.712940171131562,
+      "learning_rate": 6.874076946091966e-06,
+      "loss": 0.4046,
+      "step": 5701
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.8445350391894406,
+      "learning_rate": 6.873032967079562e-06,
+      "loss": 0.4475,
+      "step": 5702
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.5253358185611035,
+      "learning_rate": 6.871988893073032e-06,
+      "loss": 0.4351,
+      "step": 5703
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.279239597961945,
+      "learning_rate": 6.870944724125328e-06,
+      "loss": 0.4123,
+      "step": 5704
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.488725765588542,
+      "learning_rate": 6.869900460289407e-06,
+      "loss": 0.4395,
+      "step": 5705
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.562166295985152,
+      "learning_rate": 6.868856101618233e-06,
+      "loss": 0.3924,
+      "step": 5706
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.774674054874208,
+      "learning_rate": 6.867811648164769e-06,
+      "loss": 0.4147,
+      "step": 5707
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.105079481626103,
+      "learning_rate": 6.8667670999819876e-06,
+      "loss": 0.3931,
+      "step": 5708
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.451262765354113,
+      "learning_rate": 6.8657224571228645e-06,
+      "loss": 0.3936,
+      "step": 5709
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.730267393895296,
+      "learning_rate": 6.864677719640382e-06,
+      "loss": 0.4199,
+      "step": 5710
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.686617906412986,
+      "learning_rate": 6.863632887587524e-06,
+      "loss": 0.3813,
+      "step": 5711
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.580667409989504,
+      "learning_rate": 6.862587961017283e-06,
+      "loss": 0.4388,
+      "step": 5712
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.865723001944921,
+      "learning_rate": 6.861542939982652e-06,
+      "loss": 0.4162,
+      "step": 5713
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.38843359388646403,
+      "learning_rate": 6.860497824536631e-06,
+      "loss": 0.2007,
+      "step": 5714
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34588192913340854,
+      "learning_rate": 6.859452614732227e-06,
+      "loss": 0.2107,
+      "step": 5715
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.259618996777998,
+      "learning_rate": 6.8584073106224505e-06,
+      "loss": 0.3945,
+      "step": 5716
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.605354754688127,
+      "learning_rate": 6.857361912260311e-06,
+      "loss": 0.4065,
+      "step": 5717
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7835781446782386,
+      "learning_rate": 6.856316419698832e-06,
+      "loss": 0.4113,
+      "step": 5718
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.35981586338224547,
+      "learning_rate": 6.855270832991035e-06,
+      "loss": 0.1948,
+      "step": 5719
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.621401154756748,
+      "learning_rate": 6.854225152189951e-06,
+      "loss": 0.3979,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.709050487755505,
+      "learning_rate": 6.853179377348611e-06,
+      "loss": 0.423,
+      "step": 5721
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.1965660321437435,
+      "learning_rate": 6.852133508520057e-06,
+      "loss": 0.4256,
+      "step": 5722
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.140897786929135,
+      "learning_rate": 6.851087545757329e-06,
+      "loss": 0.4394,
+      "step": 5723
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.864040772056168,
+      "learning_rate": 6.850041489113475e-06,
+      "loss": 0.4181,
+      "step": 5724
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.501432858049021,
+      "learning_rate": 6.848995338641548e-06,
+      "loss": 0.4172,
+      "step": 5725
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.979968343532036,
+      "learning_rate": 6.847949094394606e-06,
+      "loss": 0.398,
+      "step": 5726
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.538096958647765,
+      "learning_rate": 6.846902756425709e-06,
+      "loss": 0.4513,
+      "step": 5727
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.3804409749697024,
+      "learning_rate": 6.845856324787928e-06,
+      "loss": 0.4175,
+      "step": 5728
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.37598761096224576,
+      "learning_rate": 6.844809799534331e-06,
+      "loss": 0.2013,
+      "step": 5729
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.832994529973318,
+      "learning_rate": 6.8437631807179945e-06,
+      "loss": 0.3766,
+      "step": 5730
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.538558938757041,
+      "learning_rate": 6.8427164683920015e-06,
+      "loss": 0.4497,
+      "step": 5731
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.907731018335102,
+      "learning_rate": 6.841669662609437e-06,
+      "loss": 0.4267,
+      "step": 5732
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.466014349931502,
+      "learning_rate": 6.840622763423391e-06,
+      "loss": 0.3916,
+      "step": 5733
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.1738629941856065,
+      "learning_rate": 6.839575770886959e-06,
+      "loss": 0.3699,
+      "step": 5734
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.7232224576602615,
+      "learning_rate": 6.8385286850532415e-06,
+      "loss": 0.41,
+      "step": 5735
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.795046211064529,
+      "learning_rate": 6.837481505975343e-06,
+      "loss": 0.4153,
+      "step": 5736
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.56407618466572,
+      "learning_rate": 6.836434233706372e-06,
+      "loss": 0.3915,
+      "step": 5737
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.662491996490291,
+      "learning_rate": 6.8353868682994465e-06,
+      "loss": 0.4745,
+      "step": 5738
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.498954141864035,
+      "learning_rate": 6.83433940980768e-06,
+      "loss": 0.409,
+      "step": 5739
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.284262669386792,
+      "learning_rate": 6.8332918582842e-06,
+      "loss": 0.4138,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.291324037277762,
+      "learning_rate": 6.832244213782133e-06,
+      "loss": 0.4314,
+      "step": 5741
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.7170886449726646,
+      "learning_rate": 6.831196476354615e-06,
+      "loss": 0.43,
+      "step": 5742
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.2920415048230645,
+      "learning_rate": 6.83014864605478e-06,
+      "loss": 0.4079,
+      "step": 5743
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.6860106400467005,
+      "learning_rate": 6.8291007229357735e-06,
+      "loss": 0.374,
+      "step": 5744
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.291240146364924,
+      "learning_rate": 6.828052707050741e-06,
+      "loss": 0.4456,
+      "step": 5745
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.106131784997431,
+      "learning_rate": 6.827004598452833e-06,
+      "loss": 0.44,
+      "step": 5746
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.63575563360998,
+      "learning_rate": 6.825956397195211e-06,
+      "loss": 0.4138,
+      "step": 5747
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3878342943441279,
+      "learning_rate": 6.824908103331031e-06,
+      "loss": 0.2205,
+      "step": 5748
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.462274129093137,
+      "learning_rate": 6.823859716913465e-06,
+      "loss": 0.4351,
+      "step": 5749
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.173375842055158,
+      "learning_rate": 6.822811237995677e-06,
+      "loss": 0.3825,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.711925765658599,
+      "learning_rate": 6.821762666630848e-06,
+      "loss": 0.4139,
+      "step": 5751
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.314986616823422,
+      "learning_rate": 6.820714002872154e-06,
+      "loss": 0.429,
+      "step": 5752
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.204301674019696,
+      "learning_rate": 6.819665246772783e-06,
+      "loss": 0.4191,
+      "step": 5753
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.235292264305598,
+      "learning_rate": 6.818616398385922e-06,
+      "loss": 0.4222,
+      "step": 5754
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.9252968168057505,
+      "learning_rate": 6.817567457764768e-06,
+      "loss": 0.3655,
+      "step": 5755
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.080508775299384,
+      "learning_rate": 6.816518424962517e-06,
+      "loss": 0.4059,
+      "step": 5756
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.916488221875882,
+      "learning_rate": 6.815469300032374e-06,
+      "loss": 0.4555,
+      "step": 5757
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.751890187879347,
+      "learning_rate": 6.814420083027547e-06,
+      "loss": 0.369,
+      "step": 5758
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.432893527083015,
+      "learning_rate": 6.813370774001248e-06,
+      "loss": 0.4234,
+      "step": 5759
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.56507226630878,
+      "learning_rate": 6.812321373006695e-06,
+      "loss": 0.4368,
+      "step": 5760
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.213862185883294,
+      "learning_rate": 6.8112718800971114e-06,
+      "loss": 0.4248,
+      "step": 5761
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.077910942920549,
+      "learning_rate": 6.810222295325723e-06,
+      "loss": 0.4637,
+      "step": 5762
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.44197932985088,
+      "learning_rate": 6.80917261874576e-06,
+      "loss": 0.4171,
+      "step": 5763
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.0604575505371585,
+      "learning_rate": 6.808122850410461e-06,
+      "loss": 0.4005,
+      "step": 5764
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.753608737570558,
+      "learning_rate": 6.807072990373067e-06,
+      "loss": 0.4234,
+      "step": 5765
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.774632430443921,
+      "learning_rate": 6.806023038686822e-06,
+      "loss": 0.3769,
+      "step": 5766
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 19.215417793432565,
+      "learning_rate": 6.804972995404977e-06,
+      "loss": 0.4075,
+      "step": 5767
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.855985550448826,
+      "learning_rate": 6.803922860580784e-06,
+      "loss": 0.3866,
+      "step": 5768
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.269957588967364,
+      "learning_rate": 6.802872634267507e-06,
+      "loss": 0.3836,
+      "step": 5769
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.797828076589096,
+      "learning_rate": 6.801822316518405e-06,
+      "loss": 0.4169,
+      "step": 5770
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.287615795879442,
+      "learning_rate": 6.800771907386752e-06,
+      "loss": 0.4326,
+      "step": 5771
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.6910951613422815,
+      "learning_rate": 6.7997214069258166e-06,
+      "loss": 0.3872,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.987222731525624,
+      "learning_rate": 6.79867081518888e-06,
+      "loss": 0.3973,
+      "step": 5773
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.13758814234029,
+      "learning_rate": 6.797620132229224e-06,
+      "loss": 0.3964,
+      "step": 5774
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 13.634305278094786,
+      "learning_rate": 6.796569358100135e-06,
+      "loss": 0.4253,
+      "step": 5775
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 12.45439774204591,
+      "learning_rate": 6.795518492854905e-06,
+      "loss": 0.4049,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 14.325851664676398,
+      "learning_rate": 6.794467536546831e-06,
+      "loss": 0.413,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.822873658445987,
+      "learning_rate": 6.793416489229214e-06,
+      "loss": 0.447,
+      "step": 5778
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3973043819588864,
+      "learning_rate": 6.79236535095536e-06,
+      "loss": 0.2079,
+      "step": 5779
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.489532124688638,
+      "learning_rate": 6.791314121778577e-06,
+      "loss": 0.4324,
+      "step": 5780
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.801659165823032,
+      "learning_rate": 6.790262801752184e-06,
+      "loss": 0.4023,
+      "step": 5781
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.041743754313979,
+      "learning_rate": 6.789211390929497e-06,
+      "loss": 0.4099,
+      "step": 5782
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.199691946190099,
+      "learning_rate": 6.788159889363842e-06,
+      "loss": 0.4069,
+      "step": 5783
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.749315993330506,
+      "learning_rate": 6.787108297108547e-06,
+      "loss": 0.42,
+      "step": 5784
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.509966053254247,
+      "learning_rate": 6.786056614216945e-06,
+      "loss": 0.3858,
+      "step": 5785
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.062091736856958,
+      "learning_rate": 6.7850048407423744e-06,
+      "loss": 0.4098,
+      "step": 5786
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.930253861957723,
+      "learning_rate": 6.7839529767381785e-06,
+      "loss": 0.3759,
+      "step": 5787
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.6138435598178855,
+      "learning_rate": 6.782901022257702e-06,
+      "loss": 0.3932,
+      "step": 5788
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.315321418175289,
+      "learning_rate": 6.7818489773543005e-06,
+      "loss": 0.4269,
+      "step": 5789
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.439969757739834,
+      "learning_rate": 6.780796842081328e-06,
+      "loss": 0.4194,
+      "step": 5790
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.881192429211216,
+      "learning_rate": 6.7797446164921445e-06,
+      "loss": 0.4453,
+      "step": 5791
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.641057948223435,
+      "learning_rate": 6.778692300640115e-06,
+      "loss": 0.406,
+      "step": 5792
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.437487302146505,
+      "learning_rate": 6.7776398945786135e-06,
+      "loss": 0.4022,
+      "step": 5793
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.191232079424982,
+      "learning_rate": 6.7765873983610095e-06,
+      "loss": 0.3707,
+      "step": 5794
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.272359609616597,
+      "learning_rate": 6.775534812040686e-06,
+      "loss": 0.4597,
+      "step": 5795
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.051488867276969,
+      "learning_rate": 6.774482135671026e-06,
+      "loss": 0.4161,
+      "step": 5796
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.817783397999703,
+      "learning_rate": 6.773429369305415e-06,
+      "loss": 0.365,
+      "step": 5797
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.3857439617814755,
+      "learning_rate": 6.772376512997249e-06,
+      "loss": 0.4239,
+      "step": 5798
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.481230330509551,
+      "learning_rate": 6.771323566799927e-06,
+      "loss": 0.3938,
+      "step": 5799
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.15230552812663,
+      "learning_rate": 6.770270530766847e-06,
+      "loss": 0.3939,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.869194603717327,
+      "learning_rate": 6.769217404951416e-06,
+      "loss": 0.4119,
+      "step": 5801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.239907341396193,
+      "learning_rate": 6.768164189407047e-06,
+      "loss": 0.4214,
+      "step": 5802
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.305476397466882,
+      "learning_rate": 6.7671108841871545e-06,
+      "loss": 0.4395,
+      "step": 5803
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.104748567840227,
+      "learning_rate": 6.766057489345159e-06,
+      "loss": 0.3677,
+      "step": 5804
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.640548787180301,
+      "learning_rate": 6.765004004934487e-06,
+      "loss": 0.4014,
+      "step": 5805
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.065835465239266,
+      "learning_rate": 6.763950431008564e-06,
+      "loss": 0.4153,
+      "step": 5806
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.974956815027482,
+      "learning_rate": 6.762896767620827e-06,
+      "loss": 0.4012,
+      "step": 5807
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.512352158137425,
+      "learning_rate": 6.761843014824712e-06,
+      "loss": 0.4048,
+      "step": 5808
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.4935973397799716,
+      "learning_rate": 6.760789172673664e-06,
+      "loss": 0.414,
+      "step": 5809
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.690376401045572,
+      "learning_rate": 6.759735241221131e-06,
+      "loss": 0.4099,
+      "step": 5810
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.482455432590134,
+      "learning_rate": 6.758681220520562e-06,
+      "loss": 0.4237,
+      "step": 5811
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.128211219374878,
+      "learning_rate": 6.757627110625417e-06,
+      "loss": 0.4158,
+      "step": 5812
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 11.398109112673058,
+      "learning_rate": 6.756572911589152e-06,
+      "loss": 0.4116,
+      "step": 5813
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.622700393706019,
+      "learning_rate": 6.755518623465238e-06,
+      "loss": 0.4087,
+      "step": 5814
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.495040134596108,
+      "learning_rate": 6.754464246307142e-06,
+      "loss": 0.4216,
+      "step": 5815
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.312450137398567,
+      "learning_rate": 6.753409780168341e-06,
+      "loss": 0.3685,
+      "step": 5816
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.550127974077462,
+      "learning_rate": 6.752355225102309e-06,
+      "loss": 0.3846,
+      "step": 5817
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 8.891901106123223,
+      "learning_rate": 6.751300581162537e-06,
+      "loss": 0.388,
+      "step": 5818
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.6519214917659,
+      "learning_rate": 6.7502458484025055e-06,
+      "loss": 0.4574,
+      "step": 5819
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.509141432515074,
+      "learning_rate": 6.749191026875712e-06,
+      "loss": 0.3743,
+      "step": 5820
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.004251523351565,
+      "learning_rate": 6.748136116635653e-06,
+      "loss": 0.4173,
+      "step": 5821
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.5279192286814,
+      "learning_rate": 6.747081117735829e-06,
+      "loss": 0.4007,
+      "step": 5822
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.893435314585663,
+      "learning_rate": 6.746026030229747e-06,
+      "loss": 0.4141,
+      "step": 5823
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 5.9393767858039075,
+      "learning_rate": 6.744970854170917e-06,
+      "loss": 0.4168,
+      "step": 5824
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4065728741890946,
+      "learning_rate": 6.7439155896128525e-06,
+      "loss": 0.2073,
+      "step": 5825
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.63919961395114,
+      "learning_rate": 6.7428602366090764e-06,
+      "loss": 0.3944,
+      "step": 5826
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.059025201672123,
+      "learning_rate": 6.741804795213112e-06,
+      "loss": 0.4192,
+      "step": 5827
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.385216854083309,
+      "learning_rate": 6.7407492654784855e-06,
+      "loss": 0.4268,
+      "step": 5828
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.40342890236172,
+      "learning_rate": 6.739693647458731e-06,
+      "loss": 0.4325,
+      "step": 5829
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.0940612878571105,
+      "learning_rate": 6.738637941207388e-06,
+      "loss": 0.4156,
+      "step": 5830
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.451897767819446,
+      "learning_rate": 6.737582146777995e-06,
+      "loss": 0.4077,
+      "step": 5831
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.350956000220896,
+      "learning_rate": 6.736526264224101e-06,
+      "loss": 0.373,
+      "step": 5832
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.440543335458844,
+      "learning_rate": 6.735470293599258e-06,
+      "loss": 0.4102,
+      "step": 5833
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.619879768259417,
+      "learning_rate": 6.7344142349570185e-06,
+      "loss": 0.3927,
+      "step": 5834
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.6685378949100285,
+      "learning_rate": 6.733358088350943e-06,
+      "loss": 0.3588,
+      "step": 5835
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.750120765269353,
+      "learning_rate": 6.732301853834597e-06,
+      "loss": 0.4256,
+      "step": 5836
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4325723345522329,
+      "learning_rate": 6.7312455314615475e-06,
+      "loss": 0.1968,
+      "step": 5837
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.291356966484295,
+      "learning_rate": 6.730189121285371e-06,
+      "loss": 0.4413,
+      "step": 5838
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.073385921082933,
+      "learning_rate": 6.729132623359641e-06,
+      "loss": 0.4214,
+      "step": 5839
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 12.917821995985447,
+      "learning_rate": 6.728076037737943e-06,
+      "loss": 0.4165,
+      "step": 5840
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.187616285965397,
+      "learning_rate": 6.727019364473861e-06,
+      "loss": 0.425,
+      "step": 5841
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.557315286265419,
+      "learning_rate": 6.725962603620988e-06,
+      "loss": 0.409,
+      "step": 5842
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.4721101412786113,
+      "learning_rate": 6.7249057552329175e-06,
+      "loss": 0.4112,
+      "step": 5843
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.838133520007875,
+      "learning_rate": 6.723848819363253e-06,
+      "loss": 0.3732,
+      "step": 5844
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.163309654322506,
+      "learning_rate": 6.722791796065596e-06,
+      "loss": 0.3854,
+      "step": 5845
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.428171814538863,
+      "learning_rate": 6.721734685393554e-06,
+      "loss": 0.4417,
+      "step": 5846
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.643137783189202,
+      "learning_rate": 6.7206774874007415e-06,
+      "loss": 0.407,
+      "step": 5847
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.326090009697792,
+      "learning_rate": 6.71962020214078e-06,
+      "loss": 0.4153,
+      "step": 5848
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.172625610798781,
+      "learning_rate": 6.718562829667284e-06,
+      "loss": 0.4529,
+      "step": 5849
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.024433188302006,
+      "learning_rate": 6.717505370033887e-06,
+      "loss": 0.4167,
+      "step": 5850
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.553014401754236,
+      "learning_rate": 6.716447823294215e-06,
+      "loss": 0.3806,
+      "step": 5851
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.357115212955779,
+      "learning_rate": 6.715390189501907e-06,
+      "loss": 0.4182,
+      "step": 5852
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.6972628852388985,
+      "learning_rate": 6.7143324687106e-06,
+      "loss": 0.4337,
+      "step": 5853
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.429777129923868,
+      "learning_rate": 6.713274660973941e-06,
+      "loss": 0.397,
+      "step": 5854
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.658271507546822,
+      "learning_rate": 6.712216766345575e-06,
+      "loss": 0.3698,
+      "step": 5855
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.055517736225985,
+      "learning_rate": 6.711158784879157e-06,
+      "loss": 0.4083,
+      "step": 5856
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.043160310372702,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.4041,
+      "step": 5857
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.676562149803669,
+      "learning_rate": 6.709042561646799e-06,
+      "loss": 0.3928,
+      "step": 5858
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.323688223745351,
+      "learning_rate": 6.7079843199881854e-06,
+      "loss": 0.4045,
+      "step": 5859
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.2143221216310165,
+      "learning_rate": 6.706925991706178e-06,
+      "loss": 0.4307,
+      "step": 5860
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.064294817605772,
+      "learning_rate": 6.705867576854447e-06,
+      "loss": 0.3816,
+      "step": 5861
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.864396170772591,
+      "learning_rate": 6.704809075486674e-06,
+      "loss": 0.3845,
+      "step": 5862
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.585005119677454,
+      "learning_rate": 6.703750487656544e-06,
+      "loss": 0.4194,
+      "step": 5863
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.039378823566235,
+      "learning_rate": 6.702691813417743e-06,
+      "loss": 0.4373,
+      "step": 5864
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9513576271454385,
+      "learning_rate": 6.7016330528239635e-06,
+      "loss": 0.4084,
+      "step": 5865
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.8524505776600226,
+      "learning_rate": 6.700574205928905e-06,
+      "loss": 0.4424,
+      "step": 5866
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.8224645532277455,
+      "learning_rate": 6.6995152727862654e-06,
+      "loss": 0.4189,
+      "step": 5867
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.146836105227825,
+      "learning_rate": 6.698456253449754e-06,
+      "loss": 0.4071,
+      "step": 5868
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.921339505623665,
+      "learning_rate": 6.697397147973077e-06,
+      "loss": 0.4448,
+      "step": 5869
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 25.432946203289188,
+      "learning_rate": 6.696337956409951e-06,
+      "loss": 0.396,
+      "step": 5870
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.950426810443767,
+      "learning_rate": 6.6952786788140936e-06,
+      "loss": 0.4222,
+      "step": 5871
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.02246829772119,
+      "learning_rate": 6.694219315239228e-06,
+      "loss": 0.4444,
+      "step": 5872
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.468438552506207,
+      "learning_rate": 6.693159865739083e-06,
+      "loss": 0.3981,
+      "step": 5873
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 13.744906495513693,
+      "learning_rate": 6.6921003303673895e-06,
+      "loss": 0.4131,
+      "step": 5874
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.569576369286281,
+      "learning_rate": 6.691040709177885e-06,
+      "loss": 0.4217,
+      "step": 5875
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.24248191804397,
+      "learning_rate": 6.689981002224306e-06,
+      "loss": 0.4224,
+      "step": 5876
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.424362935058442,
+      "learning_rate": 6.6889212095604036e-06,
+      "loss": 0.4083,
+      "step": 5877
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.859683367008419,
+      "learning_rate": 6.6878613312399225e-06,
+      "loss": 0.4047,
+      "step": 5878
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.814549036665552,
+      "learning_rate": 6.686801367316618e-06,
+      "loss": 0.4199,
+      "step": 5879
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.106087459964575,
+      "learning_rate": 6.685741317844245e-06,
+      "loss": 0.3963,
+      "step": 5880
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.19644929673357,
+      "learning_rate": 6.68468118287657e-06,
+      "loss": 0.4205,
+      "step": 5881
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.430339982658651,
+      "learning_rate": 6.6836209624673575e-06,
+      "loss": 0.4332,
+      "step": 5882
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.155835455007582,
+      "learning_rate": 6.68256065667038e-06,
+      "loss": 0.3868,
+      "step": 5883
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.34121855230557,
+      "learning_rate": 6.681500265539409e-06,
+      "loss": 0.4035,
+      "step": 5884
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.744374698463102,
+      "learning_rate": 6.680439789128228e-06,
+      "loss": 0.4059,
+      "step": 5885
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.092702610866848,
+      "learning_rate": 6.6793792274906185e-06,
+      "loss": 0.4004,
+      "step": 5886
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.581053394918015,
+      "learning_rate": 6.678318580680371e-06,
+      "loss": 0.3884,
+      "step": 5887
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3909550164185696,
+      "learning_rate": 6.677257848751276e-06,
+      "loss": 0.193,
+      "step": 5888
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.940112855933748,
+      "learning_rate": 6.676197031757131e-06,
+      "loss": 0.4438,
+      "step": 5889
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.066484956479594,
+      "learning_rate": 6.675136129751737e-06,
+      "loss": 0.4189,
+      "step": 5890
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.089808185191743,
+      "learning_rate": 6.6740751427889e-06,
+      "loss": 0.4101,
+      "step": 5891
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 14.764359497318813,
+      "learning_rate": 6.67301407092243e-06,
+      "loss": 0.4056,
+      "step": 5892
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.628660897482517,
+      "learning_rate": 6.6719529142061415e-06,
+      "loss": 0.3897,
+      "step": 5893
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.911681471901447,
+      "learning_rate": 6.67089167269385e-06,
+      "loss": 0.4049,
+      "step": 5894
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.0391924137562185,
+      "learning_rate": 6.669830346439381e-06,
+      "loss": 0.3891,
+      "step": 5895
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.784304661936996,
+      "learning_rate": 6.668768935496561e-06,
+      "loss": 0.4058,
+      "step": 5896
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.834461180800453,
+      "learning_rate": 6.667707439919221e-06,
+      "loss": 0.4239,
+      "step": 5897
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.30319298916032,
+      "learning_rate": 6.666645859761196e-06,
+      "loss": 0.4265,
+      "step": 5898
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.556720450895919,
+      "learning_rate": 6.665584195076329e-06,
+      "loss": 0.3829,
+      "step": 5899
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39814474409277756,
+      "learning_rate": 6.66452244591846e-06,
+      "loss": 0.2089,
+      "step": 5900
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.754488856490156,
+      "learning_rate": 6.6634606123414395e-06,
+      "loss": 0.4141,
+      "step": 5901
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.699905898896189,
+      "learning_rate": 6.662398694399119e-06,
+      "loss": 0.4306,
+      "step": 5902
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.606132371888938,
+      "learning_rate": 6.661336692145358e-06,
+      "loss": 0.4198,
+      "step": 5903
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.9283982598298905,
+      "learning_rate": 6.660274605634015e-06,
+      "loss": 0.377,
+      "step": 5904
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.6216946910619936,
+      "learning_rate": 6.659212434918958e-06,
+      "loss": 0.396,
+      "step": 5905
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.825754982265794,
+      "learning_rate": 6.658150180054054e-06,
+      "loss": 0.398,
+      "step": 5906
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.860429585597407,
+      "learning_rate": 6.657087841093179e-06,
+      "loss": 0.3832,
+      "step": 5907
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.023757946575023,
+      "learning_rate": 6.6560254180902115e-06,
+      "loss": 0.4401,
+      "step": 5908
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.191828655895043,
+      "learning_rate": 6.654962911099035e-06,
+      "loss": 0.4041,
+      "step": 5909
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.284377904686918,
+      "learning_rate": 6.653900320173533e-06,
+      "loss": 0.3992,
+      "step": 5910
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.3848958705326915,
+      "learning_rate": 6.652837645367601e-06,
+      "loss": 0.392,
+      "step": 5911
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.631906060772546,
+      "learning_rate": 6.651774886735132e-06,
+      "loss": 0.4146,
+      "step": 5912
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.558909847501529,
+      "learning_rate": 6.650712044330024e-06,
+      "loss": 0.4349,
+      "step": 5913
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.088267765863809,
+      "learning_rate": 6.649649118206184e-06,
+      "loss": 0.398,
+      "step": 5914
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.127055164659987,
+      "learning_rate": 6.648586108417519e-06,
+      "loss": 0.4079,
+      "step": 5915
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.060853097511792,
+      "learning_rate": 6.647523015017941e-06,
+      "loss": 0.4139,
+      "step": 5916
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.189946781485114,
+      "learning_rate": 6.646459838061367e-06,
+      "loss": 0.4089,
+      "step": 5917
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.516916126542758,
+      "learning_rate": 6.645396577601716e-06,
+      "loss": 0.4317,
+      "step": 5918
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.323812247276126,
+      "learning_rate": 6.644333233692917e-06,
+      "loss": 0.4408,
+      "step": 5919
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.596871010652083,
+      "learning_rate": 6.643269806388897e-06,
+      "loss": 0.4042,
+      "step": 5920
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.655196412908994,
+      "learning_rate": 6.642206295743589e-06,
+      "loss": 0.4336,
+      "step": 5921
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 59.7239000454647,
+      "learning_rate": 6.641142701810932e-06,
+      "loss": 0.4243,
+      "step": 5922
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.919167600368336,
+      "learning_rate": 6.6400790246448685e-06,
+      "loss": 0.4105,
+      "step": 5923
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.924602699826263,
+      "learning_rate": 6.639015264299342e-06,
+      "loss": 0.3981,
+      "step": 5924
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.055936304848405,
+      "learning_rate": 6.637951420828307e-06,
+      "loss": 0.4242,
+      "step": 5925
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.527423182807876,
+      "learning_rate": 6.636887494285713e-06,
+      "loss": 0.3978,
+      "step": 5926
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.6465236447870435,
+      "learning_rate": 6.635823484725525e-06,
+      "loss": 0.3926,
+      "step": 5927
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.9887404074789945,
+      "learning_rate": 6.6347593922017e-06,
+      "loss": 0.4082,
+      "step": 5928
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.505494283983386,
+      "learning_rate": 6.6336952167682116e-06,
+      "loss": 0.3878,
+      "step": 5929
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.607583621289899,
+      "learning_rate": 6.632630958479025e-06,
+      "loss": 0.4479,
+      "step": 5930
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.021431515300779,
+      "learning_rate": 6.63156661738812e-06,
+      "loss": 0.3959,
+      "step": 5931
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.040994247334928,
+      "learning_rate": 6.6305021935494755e-06,
+      "loss": 0.389,
+      "step": 5932
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.810607686255279,
+      "learning_rate": 6.629437687017075e-06,
+      "loss": 0.4194,
+      "step": 5933
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.490227221382474,
+      "learning_rate": 6.628373097844909e-06,
+      "loss": 0.3985,
+      "step": 5934
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.93775873630416,
+      "learning_rate": 6.6273084260869665e-06,
+      "loss": 0.3937,
+      "step": 5935
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.9721117964367703,
+      "learning_rate": 6.626243671797246e-06,
+      "loss": 0.4089,
+      "step": 5936
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.833813825826987,
+      "learning_rate": 6.625178835029749e-06,
+      "loss": 0.415,
+      "step": 5937
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.134902117247243,
+      "learning_rate": 6.624113915838479e-06,
+      "loss": 0.3847,
+      "step": 5938
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.009338491775443,
+      "learning_rate": 6.623048914277446e-06,
+      "loss": 0.3921,
+      "step": 5939
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.5915532578529,
+      "learning_rate": 6.621983830400665e-06,
+      "loss": 0.4091,
+      "step": 5940
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.684422149234575,
+      "learning_rate": 6.62091866426215e-06,
+      "loss": 0.399,
+      "step": 5941
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.301004367239392,
+      "learning_rate": 6.619853415915926e-06,
+      "loss": 0.41,
+      "step": 5942
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.92251150180856,
+      "learning_rate": 6.618788085416019e-06,
+      "loss": 0.4477,
+      "step": 5943
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4025452849087994,
+      "learning_rate": 6.617722672816457e-06,
+      "loss": 0.1839,
+      "step": 5944
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.455375137504579,
+      "learning_rate": 6.616657178171275e-06,
+      "loss": 0.4142,
+      "step": 5945
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.269260963363598,
+      "learning_rate": 6.615591601534513e-06,
+      "loss": 0.3657,
+      "step": 5946
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.005157525076663,
+      "learning_rate": 6.614525942960209e-06,
+      "loss": 0.4088,
+      "step": 5947
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.581519363918924,
+      "learning_rate": 6.613460202502414e-06,
+      "loss": 0.3832,
+      "step": 5948
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.377480243794984,
+      "learning_rate": 6.612394380215178e-06,
+      "loss": 0.4187,
+      "step": 5949
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.7396812533846076,
+      "learning_rate": 6.611328476152557e-06,
+      "loss": 0.3762,
+      "step": 5950
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 10.115470680332882,
+      "learning_rate": 6.610262490368606e-06,
+      "loss": 0.4341,
+      "step": 5951
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 8.578448628402768,
+      "learning_rate": 6.609196422917394e-06,
+      "loss": 0.4355,
+      "step": 5952
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.071295813398325,
+      "learning_rate": 6.608130273852984e-06,
+      "loss": 0.4338,
+      "step": 5953
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.2845985099756705,
+      "learning_rate": 6.60706404322945e-06,
+      "loss": 0.4174,
+      "step": 5954
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3805012225499294,
+      "learning_rate": 6.605997731100868e-06,
+      "loss": 0.206,
+      "step": 5955
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4047751037599087,
+      "learning_rate": 6.604931337521316e-06,
+      "loss": 0.1991,
+      "step": 5956
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.705582467975546,
+      "learning_rate": 6.603864862544879e-06,
+      "loss": 0.4378,
+      "step": 5957
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 7.436173667526142,
+      "learning_rate": 6.6027983062256455e-06,
+      "loss": 0.413,
+      "step": 5958
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.765339774450161,
+      "learning_rate": 6.6017316686177055e-06,
+      "loss": 0.4201,
+      "step": 5959
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.516999972089408,
+      "learning_rate": 6.60066494977516e-06,
+      "loss": 0.3739,
+      "step": 5960
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.369071457922579,
+      "learning_rate": 6.5995981497521045e-06,
+      "loss": 0.3848,
+      "step": 5961
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.229746942340359,
+      "learning_rate": 6.598531268602646e-06,
+      "loss": 0.3866,
+      "step": 5962
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.660203686720178,
+      "learning_rate": 6.5974643063808945e-06,
+      "loss": 0.4267,
+      "step": 5963
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 4.421315472685211,
+      "learning_rate": 6.59639726314096e-06,
+      "loss": 0.3736,
+      "step": 5964
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 5.413309299921669,
+      "learning_rate": 6.595330138936961e-06,
+      "loss": 0.4225,
+      "step": 5965
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 9.501736190408494,
+      "learning_rate": 6.594262933823019e-06,
+      "loss": 0.3883,
+      "step": 5966
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4181527158263478,
+      "learning_rate": 6.5931956478532585e-06,
+      "loss": 0.1929,
+      "step": 5967
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.854755575819862,
+      "learning_rate": 6.59212828108181e-06,
+      "loss": 0.4083,
+      "step": 5968
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.255117090346984,
+      "learning_rate": 6.591060833562803e-06,
+      "loss": 0.4038,
+      "step": 5969
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 15.315538647619308,
+      "learning_rate": 6.5899933053503804e-06,
+      "loss": 0.4261,
+      "step": 5970
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.673663283673004,
+      "learning_rate": 6.588925696498679e-06,
+      "loss": 0.3785,
+      "step": 5971
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.380455799238536,
+      "learning_rate": 6.587858007061848e-06,
+      "loss": 0.4219,
+      "step": 5972
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.033963607108938,
+      "learning_rate": 6.586790237094034e-06,
+      "loss": 0.401,
+      "step": 5973
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.16633749039613,
+      "learning_rate": 6.585722386649394e-06,
+      "loss": 0.41,
+      "step": 5974
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.8423221529299205,
+      "learning_rate": 6.584654455782083e-06,
+      "loss": 0.3914,
+      "step": 5975
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.085416783080153,
+      "learning_rate": 6.583586444546265e-06,
+      "loss": 0.4342,
+      "step": 5976
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.220409858267165,
+      "learning_rate": 6.5825183529961046e-06,
+      "loss": 0.4145,
+      "step": 5977
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.740120493902847,
+      "learning_rate": 6.581450181185773e-06,
+      "loss": 0.391,
+      "step": 5978
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.725543958335919,
+      "learning_rate": 6.580381929169446e-06,
+      "loss": 0.4332,
+      "step": 5979
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.381868996023768,
+      "learning_rate": 6.579313597001298e-06,
+      "loss": 0.3911,
+      "step": 5980
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.9358502420070005,
+      "learning_rate": 6.578245184735513e-06,
+      "loss": 0.3951,
+      "step": 5981
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.025134560234364,
+      "learning_rate": 6.5771766924262795e-06,
+      "loss": 0.4107,
+      "step": 5982
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8168970256500154,
+      "learning_rate": 6.576108120127783e-06,
+      "loss": 0.4745,
+      "step": 5983
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.972090310511548,
+      "learning_rate": 6.575039467894223e-06,
+      "loss": 0.4143,
+      "step": 5984
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.041460187679004,
+      "learning_rate": 6.573970735779796e-06,
+      "loss": 0.388,
+      "step": 5985
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.837721493006692,
+      "learning_rate": 6.572901923838704e-06,
+      "loss": 0.377,
+      "step": 5986
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4126982131728646,
+      "learning_rate": 6.571833032125154e-06,
+      "loss": 0.2008,
+      "step": 5987
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.0781299943847,
+      "learning_rate": 6.570764060693359e-06,
+      "loss": 0.457,
+      "step": 5988
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.109170087984225,
+      "learning_rate": 6.569695009597531e-06,
+      "loss": 0.3848,
+      "step": 5989
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.310275559828238,
+      "learning_rate": 6.56862587889189e-06,
+      "loss": 0.389,
+      "step": 5990
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.54636908545421,
+      "learning_rate": 6.567556668630657e-06,
+      "loss": 0.3988,
+      "step": 5991
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.92965390923459,
+      "learning_rate": 6.566487378868062e-06,
+      "loss": 0.4573,
+      "step": 5992
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.180905470872252,
+      "learning_rate": 6.565418009658332e-06,
+      "loss": 0.4255,
+      "step": 5993
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.417087229930676,
+      "learning_rate": 6.564348561055705e-06,
+      "loss": 0.3737,
+      "step": 5994
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.8790771030763835,
+      "learning_rate": 6.563279033114421e-06,
+      "loss": 0.3892,
+      "step": 5995
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.214078313217065,
+      "learning_rate": 6.562209425888718e-06,
+      "loss": 0.3798,
+      "step": 5996
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38546538506979994,
+      "learning_rate": 6.5611397394328465e-06,
+      "loss": 0.2057,
+      "step": 5997
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.7335666291182004,
+      "learning_rate": 6.56006997380106e-06,
+      "loss": 0.4142,
+      "step": 5998
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.32949098371959346,
+      "learning_rate": 6.559000129047607e-06,
+      "loss": 0.196,
+      "step": 5999
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.9269677647011925,
+      "learning_rate": 6.557930205226752e-06,
+      "loss": 0.4145,
+      "step": 6000
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.320020158848161,
+      "learning_rate": 6.556860202392757e-06,
+      "loss": 0.3883,
+      "step": 6001
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.009564518901034,
+      "learning_rate": 6.555790120599885e-06,
+      "loss": 0.4045,
+      "step": 6002
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.729073739983192,
+      "learning_rate": 6.554719959902414e-06,
+      "loss": 0.4451,
+      "step": 6003
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.478833211528522,
+      "learning_rate": 6.553649720354612e-06,
+      "loss": 0.4183,
+      "step": 6004
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.47519627438709,
+      "learning_rate": 6.5525794020107634e-06,
+      "loss": 0.4178,
+      "step": 6005
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.371020652147953,
+      "learning_rate": 6.551509004925149e-06,
+      "loss": 0.3844,
+      "step": 6006
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.812934540414634,
+      "learning_rate": 6.5504385291520554e-06,
+      "loss": 0.4454,
+      "step": 6007
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.060012383827558,
+      "learning_rate": 6.549367974745775e-06,
+      "loss": 0.459,
+      "step": 6008
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.365160848368284,
+      "learning_rate": 6.548297341760601e-06,
+      "loss": 0.3894,
+      "step": 6009
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.45802807348497987,
+      "learning_rate": 6.547226630250836e-06,
+      "loss": 0.2202,
+      "step": 6010
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.704333022870189,
+      "learning_rate": 6.546155840270778e-06,
+      "loss": 0.406,
+      "step": 6011
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.35932521413083696,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.2121,
+      "step": 6012
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.873812880785563,
+      "learning_rate": 6.544014025117025e-06,
+      "loss": 0.4151,
+      "step": 6013
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.47203856918351,
+      "learning_rate": 6.542943000051954e-06,
+      "loss": 0.3846,
+      "step": 6014
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.200822127264129,
+      "learning_rate": 6.541871896733845e-06,
+      "loss": 0.3926,
+      "step": 6015
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.74084073837881,
+      "learning_rate": 6.5408007152170194e-06,
+      "loss": 0.414,
+      "step": 6016
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.991828950514483,
+      "learning_rate": 6.539729455555805e-06,
+      "loss": 0.3786,
+      "step": 6017
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.827068056853985,
+      "learning_rate": 6.538658117804532e-06,
+      "loss": 0.4222,
+      "step": 6018
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.613078223068111,
+      "learning_rate": 6.537586702017538e-06,
+      "loss": 0.4322,
+      "step": 6019
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.873054048607244,
+      "learning_rate": 6.5365152082491566e-06,
+      "loss": 0.373,
+      "step": 6020
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.457703596408112,
+      "learning_rate": 6.5354436365537355e-06,
+      "loss": 0.1881,
+      "step": 6021
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.3687474628304335,
+      "learning_rate": 6.534371986985618e-06,
+      "loss": 0.4209,
+      "step": 6022
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.46724035398880903,
+      "learning_rate": 6.5333002595991555e-06,
+      "loss": 0.1973,
+      "step": 6023
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.841261931776982,
+      "learning_rate": 6.532228454448702e-06,
+      "loss": 0.4172,
+      "step": 6024
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3493318794288035,
+      "learning_rate": 6.531156571588618e-06,
+      "loss": 0.2063,
+      "step": 6025
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.414733309645092,
+      "learning_rate": 6.530084611073263e-06,
+      "loss": 0.4098,
+      "step": 6026
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.556857013720722,
+      "learning_rate": 6.5290125729570066e-06,
+      "loss": 0.4374,
+      "step": 6027
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.560614616635796,
+      "learning_rate": 6.527940457294215e-06,
+      "loss": 0.4343,
+      "step": 6028
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.430161273621094,
+      "learning_rate": 6.526868264139267e-06,
+      "loss": 0.4086,
+      "step": 6029
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.334919415654539,
+      "learning_rate": 6.5257959935465374e-06,
+      "loss": 0.3969,
+      "step": 6030
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.69838370827077,
+      "learning_rate": 6.52472364557041e-06,
+      "loss": 0.4192,
+      "step": 6031
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.2064855451498,
+      "learning_rate": 6.523651220265269e-06,
+      "loss": 0.3848,
+      "step": 6032
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.074732975013526,
+      "learning_rate": 6.522578717685507e-06,
+      "loss": 0.3539,
+      "step": 6033
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.4223689568574365,
+      "learning_rate": 6.521506137885516e-06,
+      "loss": 0.4298,
+      "step": 6034
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.819302136778513,
+      "learning_rate": 6.520433480919695e-06,
+      "loss": 0.3765,
+      "step": 6035
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.329647197703494,
+      "learning_rate": 6.519360746842442e-06,
+      "loss": 0.4062,
+      "step": 6036
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.717486321106547,
+      "learning_rate": 6.518287935708168e-06,
+      "loss": 0.4017,
+      "step": 6037
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.940411412784151,
+      "learning_rate": 6.517215047571277e-06,
+      "loss": 0.4229,
+      "step": 6038
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.882549081053072,
+      "learning_rate": 6.516142082486188e-06,
+      "loss": 0.3956,
+      "step": 6039
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.856783463299195,
+      "learning_rate": 6.5150690405073135e-06,
+      "loss": 0.3887,
+      "step": 6040
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.08306470478597,
+      "learning_rate": 6.513995921689078e-06,
+      "loss": 0.4392,
+      "step": 6041
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.470923585183673,
+      "learning_rate": 6.512922726085904e-06,
+      "loss": 0.384,
+      "step": 6042
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.7274344713574195,
+      "learning_rate": 6.5118494537522235e-06,
+      "loss": 0.3991,
+      "step": 6043
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.072983580960547,
+      "learning_rate": 6.510776104742467e-06,
+      "loss": 0.3672,
+      "step": 6044
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.886404697603455,
+      "learning_rate": 6.50970267911107e-06,
+      "loss": 0.4133,
+      "step": 6045
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6549086003919501,
+      "learning_rate": 6.508629176912478e-06,
+      "loss": 0.2098,
+      "step": 6046
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.185822490128354,
+      "learning_rate": 6.507555598201132e-06,
+      "loss": 0.4352,
+      "step": 6047
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.928878564663664,
+      "learning_rate": 6.506481943031479e-06,
+      "loss": 0.4375,
+      "step": 6048
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.11327236187709,
+      "learning_rate": 6.5054082114579744e-06,
+      "loss": 0.4352,
+      "step": 6049
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.906493289175532,
+      "learning_rate": 6.504334403535073e-06,
+      "loss": 0.3971,
+      "step": 6050
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.683930405392665,
+      "learning_rate": 6.503260519317236e-06,
+      "loss": 0.4568,
+      "step": 6051
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.389897306739791,
+      "learning_rate": 6.502186558858925e-06,
+      "loss": 0.4229,
+      "step": 6052
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.550848583013284,
+      "learning_rate": 6.50111252221461e-06,
+      "loss": 0.4095,
+      "step": 6053
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.056455500633531,
+      "learning_rate": 6.50003840943876e-06,
+      "loss": 0.4331,
+      "step": 6054
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.41451938785850645,
+      "learning_rate": 6.498964220585855e-06,
+      "loss": 0.2086,
+      "step": 6055
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.552728610486303,
+      "learning_rate": 6.49788995571037e-06,
+      "loss": 0.4048,
+      "step": 6056
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.39509090365344596,
+      "learning_rate": 6.496815614866792e-06,
+      "loss": 0.1978,
+      "step": 6057
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.022579348629582,
+      "learning_rate": 6.495741198109603e-06,
+      "loss": 0.4046,
+      "step": 6058
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 13.286719329713907,
+      "learning_rate": 6.494666705493299e-06,
+      "loss": 0.3839,
+      "step": 6059
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.62851326652457,
+      "learning_rate": 6.493592137072371e-06,
+      "loss": 0.401,
+      "step": 6060
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.635936261227098,
+      "learning_rate": 6.492517492901321e-06,
+      "loss": 0.3906,
+      "step": 6061
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.184899689568788,
+      "learning_rate": 6.4914427730346465e-06,
+      "loss": 0.3929,
+      "step": 6062
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.257916722452222,
+      "learning_rate": 6.4903679775268595e-06,
+      "loss": 0.4128,
+      "step": 6063
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.969967140172105,
+      "learning_rate": 6.489293106432466e-06,
+      "loss": 0.4221,
+      "step": 6064
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 42.57706969038808,
+      "learning_rate": 6.488218159805983e-06,
+      "loss": 0.3989,
+      "step": 6065
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.355165475636039,
+      "learning_rate": 6.487143137701925e-06,
+      "loss": 0.4105,
+      "step": 6066
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 43.42711152438496,
+      "learning_rate": 6.486068040174818e-06,
+      "loss": 0.3942,
+      "step": 6067
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.196153129308956,
+      "learning_rate": 6.484992867279184e-06,
+      "loss": 0.4241,
+      "step": 6068
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.002064864823556,
+      "learning_rate": 6.483917619069552e-06,
+      "loss": 0.4644,
+      "step": 6069
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.542396092674981,
+      "learning_rate": 6.482842295600458e-06,
+      "loss": 0.3638,
+      "step": 6070
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 22.425910903348413,
+      "learning_rate": 6.481766896926434e-06,
+      "loss": 0.379,
+      "step": 6071
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.49622427719428086,
+      "learning_rate": 6.480691423102028e-06,
+      "loss": 0.2147,
+      "step": 6072
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.944891781757379,
+      "learning_rate": 6.479615874181777e-06,
+      "loss": 0.4287,
+      "step": 6073
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.322614392775436,
+      "learning_rate": 6.4785402502202345e-06,
+      "loss": 0.4045,
+      "step": 6074
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.117932544545293,
+      "learning_rate": 6.477464551271949e-06,
+      "loss": 0.4057,
+      "step": 6075
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 12.263530427047217,
+      "learning_rate": 6.47638877739148e-06,
+      "loss": 0.4019,
+      "step": 6076
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 9.063959319489712,
+      "learning_rate": 6.475312928633386e-06,
+      "loss": 0.4055,
+      "step": 6077
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.762551276109271,
+      "learning_rate": 6.474237005052228e-06,
+      "loss": 0.4244,
+      "step": 6078
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.518979629785678,
+      "learning_rate": 6.473161006702578e-06,
+      "loss": 0.406,
+      "step": 6079
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 30.739033624959337,
+      "learning_rate": 6.472084933639002e-06,
+      "loss": 0.4099,
+      "step": 6080
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.168176764981702,
+      "learning_rate": 6.471008785916078e-06,
+      "loss": 0.3931,
+      "step": 6081
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.895354029840641,
+      "learning_rate": 6.469932563588386e-06,
+      "loss": 0.4545,
+      "step": 6082
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 11.363079560800637,
+      "learning_rate": 6.4688562667105035e-06,
+      "loss": 0.4126,
+      "step": 6083
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.306873733387228,
+      "learning_rate": 6.467779895337022e-06,
+      "loss": 0.396,
+      "step": 6084
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 16.237215651484863,
+      "learning_rate": 6.466703449522529e-06,
+      "loss": 0.4018,
+      "step": 6085
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 10.721116937611383,
+      "learning_rate": 6.4656269293216186e-06,
+      "loss": 0.4211,
+      "step": 6086
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 3.8013249348490983,
+      "learning_rate": 6.464550334788888e-06,
+      "loss": 0.4097,
+      "step": 6087
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.700288932909906,
+      "learning_rate": 6.463473665978941e-06,
+      "loss": 0.4434,
+      "step": 6088
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 17.508924419599506,
+      "learning_rate": 6.46239692294638e-06,
+      "loss": 0.4282,
+      "step": 6089
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 20.023638906966788,
+      "learning_rate": 6.461320105745815e-06,
+      "loss": 0.3854,
+      "step": 6090
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.62012144803128,
+      "learning_rate": 6.460243214431857e-06,
+      "loss": 0.3934,
+      "step": 6091
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.356598502279322,
+      "learning_rate": 6.459166249059125e-06,
+      "loss": 0.4113,
+      "step": 6092
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.156692239582553,
+      "learning_rate": 6.458089209682238e-06,
+      "loss": 0.398,
+      "step": 6093
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.635384508722433,
+      "learning_rate": 6.457012096355819e-06,
+      "loss": 0.3945,
+      "step": 6094
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.731867324573328,
+      "learning_rate": 6.4559349091344984e-06,
+      "loss": 0.4084,
+      "step": 6095
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 20.842682182049764,
+      "learning_rate": 6.4548576480729045e-06,
+      "loss": 0.3992,
+      "step": 6096
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.138865329590375,
+      "learning_rate": 6.4537803132256724e-06,
+      "loss": 0.4448,
+      "step": 6097
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.56643706308646,
+      "learning_rate": 6.452702904647445e-06,
+      "loss": 0.4011,
+      "step": 6098
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.56197788309211,
+      "learning_rate": 6.451625422392861e-06,
+      "loss": 0.4023,
+      "step": 6099
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.575710453189714,
+      "learning_rate": 6.450547866516569e-06,
+      "loss": 0.4318,
+      "step": 6100
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.870801783761907,
+      "learning_rate": 6.4494702370732186e-06,
+      "loss": 0.4291,
+      "step": 6101
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.972478314134777,
+      "learning_rate": 6.4483925341174625e-06,
+      "loss": 0.3778,
+      "step": 6102
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 8.853786681459782,
+      "learning_rate": 6.447314757703958e-06,
+      "loss": 0.3896,
+      "step": 6103
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.0592623794181115,
+      "learning_rate": 6.44623690788737e-06,
+      "loss": 0.4617,
+      "step": 6104
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.731509856858922,
+      "learning_rate": 6.445158984722358e-06,
+      "loss": 0.4156,
+      "step": 6105
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.537721937850361,
+      "learning_rate": 6.4440809882635966e-06,
+      "loss": 0.4388,
+      "step": 6106
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.794232188096371,
+      "learning_rate": 6.443002918565754e-06,
+      "loss": 0.433,
+      "step": 6107
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.9090938032700695,
+      "learning_rate": 6.441924775683508e-06,
+      "loss": 0.415,
+      "step": 6108
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 18.418054028256,
+      "learning_rate": 6.440846559671539e-06,
+      "loss": 0.3924,
+      "step": 6109
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.707262413233455,
+      "learning_rate": 6.439768270584529e-06,
+      "loss": 0.4448,
+      "step": 6110
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 6.236505484644505,
+      "learning_rate": 6.438689908477167e-06,
+      "loss": 0.4157,
+      "step": 6111
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.119586231776594,
+      "learning_rate": 6.437611473404143e-06,
+      "loss": 0.4182,
+      "step": 6112
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.44547021765452577,
+      "learning_rate": 6.436532965420151e-06,
+      "loss": 0.2256,
+      "step": 6113
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.998196846005078,
+      "learning_rate": 6.4354543845798925e-06,
+      "loss": 0.3836,
+      "step": 6114
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.952052915630052,
+      "learning_rate": 6.434375730938065e-06,
+      "loss": 0.4247,
+      "step": 6115
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.0288635848207095,
+      "learning_rate": 6.433297004549378e-06,
+      "loss": 0.458,
+      "step": 6116
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.797372498126834,
+      "learning_rate": 6.432218205468539e-06,
+      "loss": 0.3979,
+      "step": 6117
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.4926399371344745,
+      "learning_rate": 6.431139333750262e-06,
+      "loss": 0.4683,
+      "step": 6118
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.829727688735765,
+      "learning_rate": 6.430060389449265e-06,
+      "loss": 0.3955,
+      "step": 6119
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.099389548022148,
+      "learning_rate": 6.4289813726202665e-06,
+      "loss": 0.4215,
+      "step": 6120
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5561306263669774,
+      "learning_rate": 6.427902283317991e-06,
+      "loss": 0.3969,
+      "step": 6121
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.445707422447022,
+      "learning_rate": 6.426823121597169e-06,
+      "loss": 0.3912,
+      "step": 6122
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.026903864206094,
+      "learning_rate": 6.425743887512529e-06,
+      "loss": 0.4151,
+      "step": 6123
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.448068426841346,
+      "learning_rate": 6.424664581118807e-06,
+      "loss": 0.4225,
+      "step": 6124
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.142105275069848,
+      "learning_rate": 6.423585202470741e-06,
+      "loss": 0.4049,
+      "step": 6125
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.831630262064526,
+      "learning_rate": 6.422505751623076e-06,
+      "loss": 0.4012,
+      "step": 6126
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.877184798860002,
+      "learning_rate": 6.421426228630556e-06,
+      "loss": 0.3979,
+      "step": 6127
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.56943594569989,
+      "learning_rate": 6.4203466335479315e-06,
+      "loss": 0.405,
+      "step": 6128
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.604782994022029,
+      "learning_rate": 6.419266966429958e-06,
+      "loss": 0.398,
+      "step": 6129
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.580475024032916,
+      "learning_rate": 6.418187227331389e-06,
+      "loss": 0.4015,
+      "step": 6130
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.931539297493875,
+      "learning_rate": 6.417107416306989e-06,
+      "loss": 0.3821,
+      "step": 6131
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.329979284036523,
+      "learning_rate": 6.41602753341152e-06,
+      "loss": 0.4008,
+      "step": 6132
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.624533815335925,
+      "learning_rate": 6.4149475786997515e-06,
+      "loss": 0.412,
+      "step": 6133
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.020371361185728,
+      "learning_rate": 6.413867552226456e-06,
+      "loss": 0.4427,
+      "step": 6134
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.99955170575234,
+      "learning_rate": 6.412787454046407e-06,
+      "loss": 0.4009,
+      "step": 6135
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 18.18156076988601,
+      "learning_rate": 6.411707284214384e-06,
+      "loss": 0.3878,
+      "step": 6136
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.198897198179441,
+      "learning_rate": 6.410627042785169e-06,
+      "loss": 0.3842,
+      "step": 6137
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.086014247011054,
+      "learning_rate": 6.4095467298135495e-06,
+      "loss": 0.3988,
+      "step": 6138
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.741020519173649,
+      "learning_rate": 6.408466345354316e-06,
+      "loss": 0.4405,
+      "step": 6139
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.658925315710467,
+      "learning_rate": 6.407385889462261e-06,
+      "loss": 0.3848,
+      "step": 6140
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4365910952859019,
+      "learning_rate": 6.406305362192182e-06,
+      "loss": 0.2011,
+      "step": 6141
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.6113878211322845,
+      "learning_rate": 6.40522476359888e-06,
+      "loss": 0.3684,
+      "step": 6142
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.973031297270336,
+      "learning_rate": 6.404144093737159e-06,
+      "loss": 0.3862,
+      "step": 6143
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.182400181123027,
+      "learning_rate": 6.403063352661829e-06,
+      "loss": 0.3953,
+      "step": 6144
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.754569849998003,
+      "learning_rate": 6.401982540427698e-06,
+      "loss": 0.4127,
+      "step": 6145
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.100775538401168,
+      "learning_rate": 6.400901657089584e-06,
+      "loss": 0.4498,
+      "step": 6146
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.2308351857689726,
+      "learning_rate": 6.3998207027023056e-06,
+      "loss": 0.3926,
+      "step": 6147
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.616169033299027,
+      "learning_rate": 6.398739677320683e-06,
+      "loss": 0.4138,
+      "step": 6148
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.4909157352159745,
+      "learning_rate": 6.397658580999546e-06,
+      "loss": 0.4367,
+      "step": 6149
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.71811146798402,
+      "learning_rate": 6.396577413793721e-06,
+      "loss": 0.428,
+      "step": 6150
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.392453301819835,
+      "learning_rate": 6.395496175758045e-06,
+      "loss": 0.4092,
+      "step": 6151
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.045942800910539,
+      "learning_rate": 6.394414866947349e-06,
+      "loss": 0.4447,
+      "step": 6152
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.201441975969733,
+      "learning_rate": 6.39333348741648e-06,
+      "loss": 0.3885,
+      "step": 6153
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.534734938668984,
+      "learning_rate": 6.392252037220278e-06,
+      "loss": 0.4056,
+      "step": 6154
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.0640773770103875,
+      "learning_rate": 6.391170516413592e-06,
+      "loss": 0.4337,
+      "step": 6155
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.325186573864007,
+      "learning_rate": 6.390088925051273e-06,
+      "loss": 0.4075,
+      "step": 6156
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.49626408049261,
+      "learning_rate": 6.389007263188176e-06,
+      "loss": 0.3922,
+      "step": 6157
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.076484946466541,
+      "learning_rate": 6.387925530879158e-06,
+      "loss": 0.4261,
+      "step": 6158
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.16375985682998,
+      "learning_rate": 6.386843728179082e-06,
+      "loss": 0.3935,
+      "step": 6159
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.299109846351274,
+      "learning_rate": 6.385761855142813e-06,
+      "loss": 0.429,
+      "step": 6160
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.179958957917073,
+      "learning_rate": 6.384679911825222e-06,
+      "loss": 0.4111,
+      "step": 6161
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.1003464096285995,
+      "learning_rate": 6.383597898281179e-06,
+      "loss": 0.4678,
+      "step": 6162
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.43099186259721467,
+      "learning_rate": 6.382515814565563e-06,
+      "loss": 0.2081,
+      "step": 6163
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.8426027122318924,
+      "learning_rate": 6.38143366073325e-06,
+      "loss": 0.4096,
+      "step": 6164
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.2928842347621545,
+      "learning_rate": 6.380351436839127e-06,
+      "loss": 0.3658,
+      "step": 6165
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.72917552032938,
+      "learning_rate": 6.379269142938078e-06,
+      "loss": 0.4137,
+      "step": 6166
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.361224163330617,
+      "learning_rate": 6.378186779084996e-06,
+      "loss": 0.4124,
+      "step": 6167
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 15.352067802856155,
+      "learning_rate": 6.377104345334775e-06,
+      "loss": 0.4022,
+      "step": 6168
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.307200905433655,
+      "learning_rate": 6.3760218417423096e-06,
+      "loss": 0.4535,
+      "step": 6169
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.935373209472068,
+      "learning_rate": 6.374939268362501e-06,
+      "loss": 0.4489,
+      "step": 6170
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.860936951257205,
+      "learning_rate": 6.373856625250257e-06,
+      "loss": 0.4162,
+      "step": 6171
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.40654426396335,
+      "learning_rate": 6.372773912460484e-06,
+      "loss": 0.3628,
+      "step": 6172
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.018260754695194,
+      "learning_rate": 6.371691130048093e-06,
+      "loss": 0.3918,
+      "step": 6173
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.090087237240234,
+      "learning_rate": 6.370608278068e-06,
+      "loss": 0.4119,
+      "step": 6174
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.69567842017361,
+      "learning_rate": 6.369525356575123e-06,
+      "loss": 0.4206,
+      "step": 6175
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.155266889311313,
+      "learning_rate": 6.368442365624386e-06,
+      "loss": 0.415,
+      "step": 6176
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.807656371779544,
+      "learning_rate": 6.367359305270714e-06,
+      "loss": 0.3861,
+      "step": 6177
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.447654268859341,
+      "learning_rate": 6.366276175569035e-06,
+      "loss": 0.4554,
+      "step": 6178
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.9002297198689835,
+      "learning_rate": 6.365192976574284e-06,
+      "loss": 0.4497,
+      "step": 6179
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4000126354685055,
+      "learning_rate": 6.364109708341394e-06,
+      "loss": 0.1999,
+      "step": 6180
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.10145023179269,
+      "learning_rate": 6.363026370925309e-06,
+      "loss": 0.4101,
+      "step": 6181
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36830649848455255,
+      "learning_rate": 6.361942964380967e-06,
+      "loss": 0.2028,
+      "step": 6182
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.4593908946783505,
+      "learning_rate": 6.36085948876332e-06,
+      "loss": 0.2082,
+      "step": 6183
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.638758594095642,
+      "learning_rate": 6.3597759441273155e-06,
+      "loss": 0.402,
+      "step": 6184
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.766844521463274,
+      "learning_rate": 6.3586923305279104e-06,
+      "loss": 0.4041,
+      "step": 6185
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.855859355152573,
+      "learning_rate": 6.357608648020057e-06,
+      "loss": 0.4273,
+      "step": 6186
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.973385177041019,
+      "learning_rate": 6.356524896658721e-06,
+      "loss": 0.4488,
+      "step": 6187
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.447907293610308,
+      "learning_rate": 6.355441076498864e-06,
+      "loss": 0.4179,
+      "step": 6188
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.4557643718322595,
+      "learning_rate": 6.354357187595454e-06,
+      "loss": 0.4217,
+      "step": 6189
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.676752730217452,
+      "learning_rate": 6.353273230003465e-06,
+      "loss": 0.3658,
+      "step": 6190
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.157855038424498,
+      "learning_rate": 6.35218920377787e-06,
+      "loss": 0.4282,
+      "step": 6191
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.171802972443141,
+      "learning_rate": 6.351105108973644e-06,
+      "loss": 0.3999,
+      "step": 6192
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.1901478335271154,
+      "learning_rate": 6.350020945645773e-06,
+      "loss": 0.4109,
+      "step": 6193
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.309032981758959,
+      "learning_rate": 6.348936713849241e-06,
+      "loss": 0.4013,
+      "step": 6194
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.756686903558021,
+      "learning_rate": 6.347852413639036e-06,
+      "loss": 0.4151,
+      "step": 6195
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.744870907420943,
+      "learning_rate": 6.346768045070153e-06,
+      "loss": 0.4099,
+      "step": 6196
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.994890654572026,
+      "learning_rate": 6.345683608197582e-06,
+      "loss": 0.4032,
+      "step": 6197
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.085091935325514,
+      "learning_rate": 6.344599103076329e-06,
+      "loss": 0.4233,
+      "step": 6198
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.8816438539702025,
+      "learning_rate": 6.343514529761392e-06,
+      "loss": 0.4126,
+      "step": 6199
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.211180138255277,
+      "learning_rate": 6.3424298883077775e-06,
+      "loss": 0.4215,
+      "step": 6200
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.445757156793031,
+      "learning_rate": 6.3413451787704975e-06,
+      "loss": 0.4236,
+      "step": 6201
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.850943046244897,
+      "learning_rate": 6.340260401204561e-06,
+      "loss": 0.3913,
+      "step": 6202
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.150213387377782,
+      "learning_rate": 6.339175555664987e-06,
+      "loss": 0.4266,
+      "step": 6203
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.331907768887057,
+      "learning_rate": 6.3380906422067955e-06,
+      "loss": 0.3822,
+      "step": 6204
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 12.740889043089366,
+      "learning_rate": 6.337005660885007e-06,
+      "loss": 0.3967,
+      "step": 6205
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 11.857642500951954,
+      "learning_rate": 6.335920611754652e-06,
+      "loss": 0.4236,
+      "step": 6206
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.964547137681559,
+      "learning_rate": 6.334835494870759e-06,
+      "loss": 0.4136,
+      "step": 6207
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.287859886813505,
+      "learning_rate": 6.333750310288361e-06,
+      "loss": 0.4093,
+      "step": 6208
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.719428454746105,
+      "learning_rate": 6.3326650580624946e-06,
+      "loss": 0.4293,
+      "step": 6209
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.6377094035195565,
+      "learning_rate": 6.331579738248204e-06,
+      "loss": 0.4257,
+      "step": 6210
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.041420484777953,
+      "learning_rate": 6.330494350900529e-06,
+      "loss": 0.3869,
+      "step": 6211
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.274355950510989,
+      "learning_rate": 6.329408896074518e-06,
+      "loss": 0.3652,
+      "step": 6212
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 9.006906010215433,
+      "learning_rate": 6.32832337382522e-06,
+      "loss": 0.4448,
+      "step": 6213
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.593510358738081,
+      "learning_rate": 6.327237784207693e-06,
+      "loss": 0.4256,
+      "step": 6214
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.667090463405135,
+      "learning_rate": 6.326152127276992e-06,
+      "loss": 0.449,
+      "step": 6215
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.873254216535606,
+      "learning_rate": 6.32506640308818e-06,
+      "loss": 0.3976,
+      "step": 6216
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.867707104191886,
+      "learning_rate": 6.323980611696319e-06,
+      "loss": 0.3989,
+      "step": 6217
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.447378561086757,
+      "learning_rate": 6.3228947531564775e-06,
+      "loss": 0.3951,
+      "step": 6218
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.406388277924703,
+      "learning_rate": 6.3218088275237265e-06,
+      "loss": 0.3867,
+      "step": 6219
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.49457723462822994,
+      "learning_rate": 6.3207228348531425e-06,
+      "loss": 0.1984,
+      "step": 6220
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.48792169085395465,
+      "learning_rate": 6.3196367751998e-06,
+      "loss": 0.2074,
+      "step": 6221
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.118960478458926,
+      "learning_rate": 6.318550648618785e-06,
+      "loss": 0.416,
+      "step": 6222
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 14.797199845782192,
+      "learning_rate": 6.3174644551651795e-06,
+      "loss": 0.45,
+      "step": 6223
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3741628107595779,
+      "learning_rate": 6.316378194894072e-06,
+      "loss": 0.1904,
+      "step": 6224
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.564180362848293,
+      "learning_rate": 6.315291867860552e-06,
+      "loss": 0.4175,
+      "step": 6225
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.224509926200872,
+      "learning_rate": 6.3142054741197186e-06,
+      "loss": 0.4158,
+      "step": 6226
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.919862125091976,
+      "learning_rate": 6.313119013726667e-06,
+      "loss": 0.4251,
+      "step": 6227
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.019245612074601,
+      "learning_rate": 6.312032486736501e-06,
+      "loss": 0.3976,
+      "step": 6228
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.8639437690936544,
+      "learning_rate": 6.310945893204324e-06,
+      "loss": 0.4332,
+      "step": 6229
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.860481786518734,
+      "learning_rate": 6.309859233185246e-06,
+      "loss": 0.4168,
+      "step": 6230
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.7095095620729146,
+      "learning_rate": 6.308772506734377e-06,
+      "loss": 0.4048,
+      "step": 6231
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.819683479895555,
+      "learning_rate": 6.307685713906835e-06,
+      "loss": 0.4186,
+      "step": 6232
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 19.588156645434502,
+      "learning_rate": 6.306598854757735e-06,
+      "loss": 0.4281,
+      "step": 6233
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6927611565497742,
+      "learning_rate": 6.305511929342202e-06,
+      "loss": 0.2373,
+      "step": 6234
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.6523468831852255,
+      "learning_rate": 6.30442493771536e-06,
+      "loss": 0.3945,
+      "step": 6235
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.5947226957940863,
+      "learning_rate": 6.303337879932339e-06,
+      "loss": 0.4323,
+      "step": 6236
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 17.240491291934564,
+      "learning_rate": 6.302250756048267e-06,
+      "loss": 0.4371,
+      "step": 6237
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.0977282153087398,
+      "learning_rate": 6.301163566118285e-06,
+      "loss": 0.411,
+      "step": 6238
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 7.162376221197712,
+      "learning_rate": 6.300076310197526e-06,
+      "loss": 0.4439,
+      "step": 6239
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.9016520698828545,
+      "learning_rate": 6.298988988341137e-06,
+      "loss": 0.4315,
+      "step": 6240
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.36548562028373704,
+      "learning_rate": 6.297901600604261e-06,
+      "loss": 0.2046,
+      "step": 6241
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.258175525061917,
+      "learning_rate": 6.296814147042048e-06,
+      "loss": 0.4237,
+      "step": 6242
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.045787888726577,
+      "learning_rate": 6.2957266277096474e-06,
+      "loss": 0.3872,
+      "step": 6243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.676272703673346,
+      "learning_rate": 6.294639042662219e-06,
+      "loss": 0.4022,
+      "step": 6244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.1246076629713855,
+      "learning_rate": 6.293551391954918e-06,
+      "loss": 0.4507,
+      "step": 6245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 5.027711692995234,
+      "learning_rate": 6.292463675642908e-06,
+      "loss": 0.4143,
+      "step": 6246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 20.041085470058817,
+      "learning_rate": 6.291375893781353e-06,
+      "loss": 0.3761,
+      "step": 6247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.165604870715724,
+      "learning_rate": 6.290288046425424e-06,
+      "loss": 0.3913,
+      "step": 6248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.324782242616431,
+      "learning_rate": 6.289200133630291e-06,
+      "loss": 0.4191,
+      "step": 6249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.853213292214911,
+      "learning_rate": 6.288112155451132e-06,
+      "loss": 0.3811,
+      "step": 6250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.18477737682902,
+      "learning_rate": 6.28702411194312e-06,
+      "loss": 0.4395,
+      "step": 6251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 22.78283227391264,
+      "learning_rate": 6.285936003161445e-06,
+      "loss": 0.4069,
+      "step": 6252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.954876520871835,
+      "learning_rate": 6.284847829161285e-06,
+      "loss": 0.3953,
+      "step": 6253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.741899611821849,
+      "learning_rate": 6.283759589997835e-06,
+      "loss": 0.3925,
+      "step": 6254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.981395054389284,
+      "learning_rate": 6.282671285726283e-06,
+      "loss": 0.4226,
+      "step": 6255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 4.050145407486678,
+      "learning_rate": 6.281582916401825e-06,
+      "loss": 0.4395,
+      "step": 6256
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.436122462829789,
+      "learning_rate": 6.2804944820796596e-06,
+      "loss": 0.4063,
+      "step": 6257
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.562910629940574,
+      "learning_rate": 6.2794059828149874e-06,
+      "loss": 0.4217,
+      "step": 6258
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.6851882014879145,
+      "learning_rate": 6.278317418663017e-06,
+      "loss": 0.4112,
+      "step": 6259
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4519864237119642,
+      "learning_rate": 6.277228789678953e-06,
+      "loss": 0.1939,
+      "step": 6260
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.169987721569828,
+      "learning_rate": 6.276140095918011e-06,
+      "loss": 0.401,
+      "step": 6261
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.533029361079203,
+      "learning_rate": 6.275051337435401e-06,
+      "loss": 0.4043,
+      "step": 6262
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.840345961954252,
+      "learning_rate": 6.273962514286346e-06,
+      "loss": 0.3875,
+      "step": 6263
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.363834608025485,
+      "learning_rate": 6.272873626526064e-06,
+      "loss": 0.407,
+      "step": 6264
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.354262736351409,
+      "learning_rate": 6.271784674209783e-06,
+      "loss": 0.4253,
+      "step": 6265
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3130606115732446,
+      "learning_rate": 6.270695657392729e-06,
+      "loss": 0.3621,
+      "step": 6266
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3640046401155942,
+      "learning_rate": 6.269606576130135e-06,
+      "loss": 0.2037,
+      "step": 6267
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.002400065913683,
+      "learning_rate": 6.268517430477234e-06,
+      "loss": 0.4023,
+      "step": 6268
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.447154251179826,
+      "learning_rate": 6.267428220489263e-06,
+      "loss": 0.4192,
+      "step": 6269
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5825295002658475,
+      "learning_rate": 6.266338946221466e-06,
+      "loss": 0.4197,
+      "step": 6270
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.27074856069143,
+      "learning_rate": 6.265249607729087e-06,
+      "loss": 0.426,
+      "step": 6271
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1726126940465402,
+      "learning_rate": 6.264160205067371e-06,
+      "loss": 0.4095,
+      "step": 6272
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9235040893661277,
+      "learning_rate": 6.263070738291572e-06,
+      "loss": 0.4508,
+      "step": 6273
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.001740444126143,
+      "learning_rate": 6.261981207456942e-06,
+      "loss": 0.3901,
+      "step": 6274
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.2611676160007,
+      "learning_rate": 6.26089161261874e-06,
+      "loss": 0.4039,
+      "step": 6275
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.876449853444846,
+      "learning_rate": 6.259801953832226e-06,
+      "loss": 0.3923,
+      "step": 6276
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.521149214167065,
+      "learning_rate": 6.258712231152665e-06,
+      "loss": 0.423,
+      "step": 6277
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.679913021085149,
+      "learning_rate": 6.257622444635323e-06,
+      "loss": 0.421,
+      "step": 6278
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.32681794545359,
+      "learning_rate": 6.256532594335471e-06,
+      "loss": 0.4296,
+      "step": 6279
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.604945735477369,
+      "learning_rate": 6.255442680308381e-06,
+      "loss": 0.4093,
+      "step": 6280
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.20504448479475,
+      "learning_rate": 6.254352702609332e-06,
+      "loss": 0.3533,
+      "step": 6281
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 11.199093038902467,
+      "learning_rate": 6.2532626612936035e-06,
+      "loss": 0.4013,
+      "step": 6282
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4159775747233713,
+      "learning_rate": 6.2521725564164785e-06,
+      "loss": 0.205,
+      "step": 6283
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.325884614888303,
+      "learning_rate": 6.251082388033243e-06,
+      "loss": 0.4024,
+      "step": 6284
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 13.033052848418896,
+      "learning_rate": 6.249992156199189e-06,
+      "loss": 0.4268,
+      "step": 6285
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.20330315646929,
+      "learning_rate": 6.248901860969607e-06,
+      "loss": 0.3922,
+      "step": 6286
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.60410316948435,
+      "learning_rate": 6.247811502399795e-06,
+      "loss": 0.4422,
+      "step": 6287
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9544154600692,
+      "learning_rate": 6.2467210805450505e-06,
+      "loss": 0.414,
+      "step": 6288
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7378912066930483,
+      "learning_rate": 6.2456305954606784e-06,
+      "loss": 0.4205,
+      "step": 6289
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.513219822680442,
+      "learning_rate": 6.244540047201984e-06,
+      "loss": 0.411,
+      "step": 6290
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.405569867549418,
+      "learning_rate": 6.243449435824276e-06,
+      "loss": 0.4082,
+      "step": 6291
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.523953364107956,
+      "learning_rate": 6.242358761382864e-06,
+      "loss": 0.4123,
+      "step": 6292
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.24985539014654,
+      "learning_rate": 6.241268023933068e-06,
+      "loss": 0.4461,
+      "step": 6293
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0476955879788785,
+      "learning_rate": 6.240177223530204e-06,
+      "loss": 0.4468,
+      "step": 6294
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.545622171545131,
+      "learning_rate": 6.239086360229595e-06,
+      "loss": 0.4391,
+      "step": 6295
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.875170870840277,
+      "learning_rate": 6.237995434086565e-06,
+      "loss": 0.4193,
+      "step": 6296
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.334322118154398,
+      "learning_rate": 6.236904445156442e-06,
+      "loss": 0.4108,
+      "step": 6297
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.8111255741269,
+      "learning_rate": 6.235813393494558e-06,
+      "loss": 0.4637,
+      "step": 6298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.571653739481872,
+      "learning_rate": 6.234722279156249e-06,
+      "loss": 0.4627,
+      "step": 6299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.711836828350454,
+      "learning_rate": 6.233631102196851e-06,
+      "loss": 0.4272,
+      "step": 6300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 25.51091275390508,
+      "learning_rate": 6.232539862671706e-06,
+      "loss": 0.4527,
+      "step": 6301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.034682837844467,
+      "learning_rate": 6.231448560636157e-06,
+      "loss": 0.4002,
+      "step": 6302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.985275524995266,
+      "learning_rate": 6.2303571961455514e-06,
+      "loss": 0.4512,
+      "step": 6303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.426924642093491,
+      "learning_rate": 6.2292657692552394e-06,
+      "loss": 0.4242,
+      "step": 6304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.6785436721202185,
+      "learning_rate": 6.228174280020576e-06,
+      "loss": 0.4192,
+      "step": 6305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.15896255970827,
+      "learning_rate": 6.227082728496916e-06,
+      "loss": 0.4549,
+      "step": 6306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.5903140680605174,
+      "learning_rate": 6.225991114739622e-06,
+      "loss": 0.4152,
+      "step": 6307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0456762191519653,
+      "learning_rate": 6.2248994388040555e-06,
+      "loss": 0.3772,
+      "step": 6308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.497561209183315,
+      "learning_rate": 6.223807700745582e-06,
+      "loss": 0.4367,
+      "step": 6309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 10.483100356761444,
+      "learning_rate": 6.222715900619573e-06,
+      "loss": 0.3961,
+      "step": 6310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3218530440073533,
+      "learning_rate": 6.2216240384813985e-06,
+      "loss": 0.4296,
+      "step": 6311
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.96947617323781,
+      "learning_rate": 6.220532114386437e-06,
+      "loss": 0.4137,
+      "step": 6312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6140014663099485,
+      "learning_rate": 6.2194401283900654e-06,
+      "loss": 0.3965,
+      "step": 6313
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.077005023062614,
+      "learning_rate": 6.2183480805476646e-06,
+      "loss": 0.3856,
+      "step": 6314
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.2431475799735985,
+      "learning_rate": 6.217255970914623e-06,
+      "loss": 0.4064,
+      "step": 6315
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.144710536036482,
+      "learning_rate": 6.216163799546326e-06,
+      "loss": 0.3912,
+      "step": 6316
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 31.98190487704999,
+      "learning_rate": 6.215071566498167e-06,
+      "loss": 0.3934,
+      "step": 6317
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 15.412277316741344,
+      "learning_rate": 6.213979271825538e-06,
+      "loss": 0.3802,
+      "step": 6318
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0620484295831285,
+      "learning_rate": 6.212886915583839e-06,
+      "loss": 0.3776,
+      "step": 6319
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.974752913967924,
+      "learning_rate": 6.21179449782847e-06,
+      "loss": 0.4184,
+      "step": 6320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8294217974098514,
+      "learning_rate": 6.210702018614833e-06,
+      "loss": 0.4149,
+      "step": 6321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.542957608966548,
+      "learning_rate": 6.209609477998339e-06,
+      "loss": 0.4176,
+      "step": 6322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7559416389835683,
+      "learning_rate": 6.208516876034395e-06,
+      "loss": 0.4152,
+      "step": 6323
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.715412642782888,
+      "learning_rate": 6.207424212778416e-06,
+      "loss": 0.4314,
+      "step": 6324
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4392879991862775,
+      "learning_rate": 6.206331488285816e-06,
+      "loss": 0.408,
+      "step": 6325
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.0694764928388265,
+      "learning_rate": 6.205238702612016e-06,
+      "loss": 0.3791,
+      "step": 6326
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9160747501032533,
+      "learning_rate": 6.204145855812439e-06,
+      "loss": 0.4247,
+      "step": 6327
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9293290147430904,
+      "learning_rate": 6.203052947942509e-06,
+      "loss": 0.4357,
+      "step": 6328
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.453798095036561,
+      "learning_rate": 6.2019599790576565e-06,
+      "loss": 0.4301,
+      "step": 6329
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.299375340127552,
+      "learning_rate": 6.200866949213312e-06,
+      "loss": 0.4015,
+      "step": 6330
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.370824577867067,
+      "learning_rate": 6.19977385846491e-06,
+      "loss": 0.3459,
+      "step": 6331
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.215963996629145,
+      "learning_rate": 6.1986807068678926e-06,
+      "loss": 0.3902,
+      "step": 6332
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.334391849311387,
+      "learning_rate": 6.197587494477695e-06,
+      "loss": 0.3751,
+      "step": 6333
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.3849312400491085,
+      "learning_rate": 6.1964942213497645e-06,
+      "loss": 0.3713,
+      "step": 6334
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.991144979950912,
+      "learning_rate": 6.195400887539548e-06,
+      "loss": 0.3682,
+      "step": 6335
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.922235003390416,
+      "learning_rate": 6.194307493102496e-06,
+      "loss": 0.3877,
+      "step": 6336
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.886478086671188,
+      "learning_rate": 6.193214038094061e-06,
+      "loss": 0.3981,
+      "step": 6337
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3625947863349455,
+      "learning_rate": 6.1921205225697e-06,
+      "loss": 0.1879,
+      "step": 6338
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4768699197573008,
+      "learning_rate": 6.191026946584873e-06,
+      "loss": 0.3978,
+      "step": 6339
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.129131813301521,
+      "learning_rate": 6.189933310195042e-06,
+      "loss": 0.4215,
+      "step": 6340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3609611354523372,
+      "learning_rate": 6.188839613455673e-06,
+      "loss": 0.1898,
+      "step": 6341
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.9250849773446164,
+      "learning_rate": 6.187745856422236e-06,
+      "loss": 0.3971,
+      "step": 6342
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.8315091779433645,
+      "learning_rate": 6.186652039150199e-06,
+      "loss": 0.3992,
+      "step": 6343
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7559459131938278,
+      "learning_rate": 6.185558161695041e-06,
+      "loss": 0.3935,
+      "step": 6344
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7326168519095133,
+      "learning_rate": 6.1844642241122374e-06,
+      "loss": 0.3991,
+      "step": 6345
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.265030241232376,
+      "learning_rate": 6.183370226457272e-06,
+      "loss": 0.4342,
+      "step": 6346
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.47184371228287225,
+      "learning_rate": 6.182276168785624e-06,
+      "loss": 0.1942,
+      "step": 6347
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.57418251130217,
+      "learning_rate": 6.181182051152784e-06,
+      "loss": 0.4299,
+      "step": 6348
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7481558600544838,
+      "learning_rate": 6.180087873614241e-06,
+      "loss": 0.3956,
+      "step": 6349
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.6124760630867607,
+      "learning_rate": 6.1789936362254885e-06,
+      "loss": 0.397,
+      "step": 6350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.194520248368186,
+      "learning_rate": 6.177899339042023e-06,
+      "loss": 0.3888,
+      "step": 6351
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.498985985462224,
+      "learning_rate": 6.176804982119343e-06,
+      "loss": 0.4299,
+      "step": 6352
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.443548149397484,
+      "learning_rate": 6.17571056551295e-06,
+      "loss": 0.3777,
+      "step": 6353
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.8076382306426035,
+      "learning_rate": 6.1746160892783515e-06,
+      "loss": 0.4392,
+      "step": 6354
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.506727227246231,
+      "learning_rate": 6.173521553471054e-06,
+      "loss": 0.406,
+      "step": 6355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7900129540631045,
+      "learning_rate": 6.172426958146569e-06,
+      "loss": 0.4717,
+      "step": 6356
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.880377051998629,
+      "learning_rate": 6.171332303360411e-06,
+      "loss": 0.4227,
+      "step": 6357
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.625091313956386,
+      "learning_rate": 6.1702375891680985e-06,
+      "loss": 0.3885,
+      "step": 6358
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.740392620223943,
+      "learning_rate": 6.169142815625148e-06,
+      "loss": 0.4002,
+      "step": 6359
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.37878654338657686,
+      "learning_rate": 6.168047982787089e-06,
+      "loss": 0.1852,
+      "step": 6360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7254995685142718,
+      "learning_rate": 6.166953090709442e-06,
+      "loss": 0.3867,
+      "step": 6361
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.85510382684627,
+      "learning_rate": 6.165858139447739e-06,
+      "loss": 0.426,
+      "step": 6362
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.041879003732645,
+      "learning_rate": 6.164763129057511e-06,
+      "loss": 0.4164,
+      "step": 6363
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 9.794614441909793,
+      "learning_rate": 6.163668059594296e-06,
+      "loss": 0.4202,
+      "step": 6364
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.709273430866761,
+      "learning_rate": 6.1625729311136306e-06,
+      "loss": 0.4045,
+      "step": 6365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.611735781328435,
+      "learning_rate": 6.161477743671057e-06,
+      "loss": 0.3951,
+      "step": 6366
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.478001788841824,
+      "learning_rate": 6.16038249732212e-06,
+      "loss": 0.4267,
+      "step": 6367
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.716940450819446,
+      "learning_rate": 6.159287192122363e-06,
+      "loss": 0.437,
+      "step": 6368
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.661476806815387,
+      "learning_rate": 6.15819182812734e-06,
+      "loss": 0.3635,
+      "step": 6369
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 12.182077139383114,
+      "learning_rate": 6.157096405392604e-06,
+      "loss": 0.4339,
+      "step": 6370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.0254570137994765,
+      "learning_rate": 6.156000923973709e-06,
+      "loss": 0.413,
+      "step": 6371
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 8.355602610866756,
+      "learning_rate": 6.154905383926218e-06,
+      "loss": 0.4263,
+      "step": 6372
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.832872947212771,
+      "learning_rate": 6.153809785305689e-06,
+      "loss": 0.3735,
+      "step": 6373
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.37954609441308,
+      "learning_rate": 6.152714128167691e-06,
+      "loss": 0.4037,
+      "step": 6374
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.2384219432323262,
+      "learning_rate": 6.15161841256779e-06,
+      "loss": 0.4377,
+      "step": 6375
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.797302434171517,
+      "learning_rate": 6.150522638561558e-06,
+      "loss": 0.3905,
+      "step": 6376
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.645326381005364,
+      "learning_rate": 6.149426806204568e-06,
+      "loss": 0.4145,
+      "step": 6377
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8034586230749956,
+      "learning_rate": 6.148330915552398e-06,
+      "loss": 0.4308,
+      "step": 6378
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.804967254311515,
+      "learning_rate": 6.147234966660628e-06,
+      "loss": 0.3841,
+      "step": 6379
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.465115751952542,
+      "learning_rate": 6.1461389595848406e-06,
+      "loss": 0.3992,
+      "step": 6380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.539922631100997,
+      "learning_rate": 6.1450428943806214e-06,
+      "loss": 0.4205,
+      "step": 6381
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9729055593663833,
+      "learning_rate": 6.143946771103561e-06,
+      "loss": 0.4235,
+      "step": 6382
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.964558121469245,
+      "learning_rate": 6.142850589809248e-06,
+      "loss": 0.4037,
+      "step": 6383
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.574999389215214,
+      "learning_rate": 6.141754350553279e-06,
+      "loss": 0.3931,
+      "step": 6384
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.479092429632642,
+      "learning_rate": 6.140658053391253e-06,
+      "loss": 0.4037,
+      "step": 6385
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.539992592339131,
+      "learning_rate": 6.139561698378768e-06,
+      "loss": 0.3914,
+      "step": 6386
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.048190763426326,
+      "learning_rate": 6.1384652855714295e-06,
+      "loss": 0.4105,
+      "step": 6387
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.5487348433735955,
+      "learning_rate": 6.1373688150248434e-06,
+      "loss": 0.4305,
+      "step": 6388
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.730179399845977,
+      "learning_rate": 6.136272286794619e-06,
+      "loss": 0.3839,
+      "step": 6389
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.523316633728644,
+      "learning_rate": 6.135175700936369e-06,
+      "loss": 0.4065,
+      "step": 6390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3567978482157489,
+      "learning_rate": 6.134079057505708e-06,
+      "loss": 0.2173,
+      "step": 6391
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7731787342706857,
+      "learning_rate": 6.132982356558254e-06,
+      "loss": 0.399,
+      "step": 6392
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 6.996480590861989,
+      "learning_rate": 6.13188559814963e-06,
+      "loss": 0.4097,
+      "step": 6393
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.7052657407804226,
+      "learning_rate": 6.130788782335457e-06,
+      "loss": 0.3729,
+      "step": 6394
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.418794038096379,
+      "learning_rate": 6.129691909171366e-06,
+      "loss": 0.4123,
+      "step": 6395
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.70305558637709,
+      "learning_rate": 6.128594978712982e-06,
+      "loss": 0.418,
+      "step": 6396
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.454779445177805,
+      "learning_rate": 6.1274979910159405e-06,
+      "loss": 0.3973,
+      "step": 6397
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 18.356455214348355,
+      "learning_rate": 6.126400946135876e-06,
+      "loss": 0.435,
+      "step": 6398
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.542371242580093,
+      "learning_rate": 6.125303844128429e-06,
+      "loss": 0.4127,
+      "step": 6399
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.813820411127707,
+      "learning_rate": 6.124206685049241e-06,
+      "loss": 0.3914,
+      "step": 6400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.622121063009107,
+      "learning_rate": 6.123109468953954e-06,
+      "loss": 0.3873,
+      "step": 6401
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38252803412115555,
+      "learning_rate": 6.122012195898216e-06,
+      "loss": 0.1997,
+      "step": 6402
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8452773256965145,
+      "learning_rate": 6.1209148659376775e-06,
+      "loss": 0.43,
+      "step": 6403
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.571250157394901,
+      "learning_rate": 6.119817479127991e-06,
+      "loss": 0.4212,
+      "step": 6404
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.773346712549025,
+      "learning_rate": 6.1187200355248135e-06,
+      "loss": 0.4382,
+      "step": 6405
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.848584573614403,
+      "learning_rate": 6.117622535183802e-06,
+      "loss": 0.4142,
+      "step": 6406
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.286079160843698,
+      "learning_rate": 6.11652497816062e-06,
+      "loss": 0.4053,
+      "step": 6407
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.582216232005638,
+      "learning_rate": 6.11542736451093e-06,
+      "loss": 0.4097,
+      "step": 6408
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.827048522976344,
+      "learning_rate": 6.114329694290401e-06,
+      "loss": 0.3812,
+      "step": 6409
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.557694478786527,
+      "learning_rate": 6.113231967554704e-06,
+      "loss": 0.401,
+      "step": 6410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.669181084621289,
+      "learning_rate": 6.112134184359509e-06,
+      "loss": 0.3726,
+      "step": 6411
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.430658579659394,
+      "learning_rate": 6.111036344760497e-06,
+      "loss": 0.4131,
+      "step": 6412
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.37224327479302327,
+      "learning_rate": 6.109938448813342e-06,
+      "loss": 0.2046,
+      "step": 6413
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.857105880330205,
+      "learning_rate": 6.108840496573727e-06,
+      "loss": 0.4329,
+      "step": 6414
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.13526659018605,
+      "learning_rate": 6.107742488097338e-06,
+      "loss": 0.4253,
+      "step": 6415
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.40672925278250127,
+      "learning_rate": 6.10664442343986e-06,
+      "loss": 0.1822,
+      "step": 6416
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2204865387971195,
+      "learning_rate": 6.105546302656986e-06,
+      "loss": 0.3905,
+      "step": 6417
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.8560698513140164,
+      "learning_rate": 6.104448125804407e-06,
+      "loss": 0.4041,
+      "step": 6418
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.924583084901604,
+      "learning_rate": 6.103349892937821e-06,
+      "loss": 0.4634,
+      "step": 6419
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.183138649930166,
+      "learning_rate": 6.102251604112924e-06,
+      "loss": 0.4017,
+      "step": 6420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.1043225028032015,
+      "learning_rate": 6.10115325938542e-06,
+      "loss": 0.4755,
+      "step": 6421
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.007151070366321,
+      "learning_rate": 6.100054858811012e-06,
+      "loss": 0.416,
+      "step": 6422
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.1775917826535505,
+      "learning_rate": 6.098956402445409e-06,
+      "loss": 0.3996,
+      "step": 6423
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.4619205005739175,
+      "learning_rate": 6.097857890344321e-06,
+      "loss": 0.4346,
+      "step": 6424
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.185244166182526,
+      "learning_rate": 6.0967593225634584e-06,
+      "loss": 0.389,
+      "step": 6425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6472739397415594,
+      "learning_rate": 6.095660699158539e-06,
+      "loss": 0.4313,
+      "step": 6426
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1863310317830678,
+      "learning_rate": 6.094562020185281e-06,
+      "loss": 0.4105,
+      "step": 6427
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.555326349864971,
+      "learning_rate": 6.093463285699404e-06,
+      "loss": 0.4035,
+      "step": 6428
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.657319147741369,
+      "learning_rate": 6.0923644957566375e-06,
+      "loss": 0.4228,
+      "step": 6429
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.281388712049785,
+      "learning_rate": 6.091265650412702e-06,
+      "loss": 0.3769,
+      "step": 6430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.820382935068866,
+      "learning_rate": 6.0901667497233315e-06,
+      "loss": 0.4302,
+      "step": 6431
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.165054700582293,
+      "learning_rate": 6.089067793744258e-06,
+      "loss": 0.4236,
+      "step": 6432
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.870702761535284,
+      "learning_rate": 6.087968782531217e-06,
+      "loss": 0.4244,
+      "step": 6433
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.235370489471106,
+      "learning_rate": 6.086869716139946e-06,
+      "loss": 0.4451,
+      "step": 6434
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.013521770828623,
+      "learning_rate": 6.085770594626187e-06,
+      "loss": 0.3852,
+      "step": 6435
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.2740350970884062,
+      "learning_rate": 6.084671418045681e-06,
+      "loss": 0.3985,
+      "step": 6436
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.902270210076626,
+      "learning_rate": 6.083572186454178e-06,
+      "loss": 0.3774,
+      "step": 6437
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.921283883542401,
+      "learning_rate": 6.082472899907428e-06,
+      "loss": 0.3974,
+      "step": 6438
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.322462123803255,
+      "learning_rate": 6.08137355846118e-06,
+      "loss": 0.4031,
+      "step": 6439
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.778066268265571,
+      "learning_rate": 6.0802741621711906e-06,
+      "loss": 0.4093,
+      "step": 6440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.9859621235482185,
+      "learning_rate": 6.079174711093217e-06,
+      "loss": 0.4276,
+      "step": 6441
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.375428592130389,
+      "learning_rate": 6.078075205283022e-06,
+      "loss": 0.4236,
+      "step": 6442
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.684056409451163,
+      "learning_rate": 6.0769756447963666e-06,
+      "loss": 0.4391,
+      "step": 6443
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.147855099533549,
+      "learning_rate": 6.075876029689016e-06,
+      "loss": 0.4256,
+      "step": 6444
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.851616562391644,
+      "learning_rate": 6.074776360016743e-06,
+      "loss": 0.4343,
+      "step": 6445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.418946488832013,
+      "learning_rate": 6.073676635835317e-06,
+      "loss": 0.364,
+      "step": 6446
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0184168238642832,
+      "learning_rate": 6.072576857200512e-06,
+      "loss": 0.3939,
+      "step": 6447
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.972487134620782,
+      "learning_rate": 6.0714770241681054e-06,
+      "loss": 0.4293,
+      "step": 6448
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.247334714297527,
+      "learning_rate": 6.070377136793876e-06,
+      "loss": 0.4118,
+      "step": 6449
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.498738105968696,
+      "learning_rate": 6.069277195133609e-06,
+      "loss": 0.3862,
+      "step": 6450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.62604725072733,
+      "learning_rate": 6.06817719924309e-06,
+      "loss": 0.4027,
+      "step": 6451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.002374076474038,
+      "learning_rate": 6.067077149178105e-06,
+      "loss": 0.4467,
+      "step": 6452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.971491832626935,
+      "learning_rate": 6.065977044994445e-06,
+      "loss": 0.4589,
+      "step": 6453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.344083590694381,
+      "learning_rate": 6.064876886747905e-06,
+      "loss": 0.4085,
+      "step": 6454
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.770870141094499,
+      "learning_rate": 6.063776674494282e-06,
+      "loss": 0.4009,
+      "step": 6455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.943297135221426,
+      "learning_rate": 6.062676408289372e-06,
+      "loss": 0.4321,
+      "step": 6456
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.072404345120793,
+      "learning_rate": 6.061576088188981e-06,
+      "loss": 0.4155,
+      "step": 6457
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.903485990990774,
+      "learning_rate": 6.060475714248911e-06,
+      "loss": 0.43,
+      "step": 6458
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.635644695595873,
+      "learning_rate": 6.05937528652497e-06,
+      "loss": 0.3998,
+      "step": 6459
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.51888159630447,
+      "learning_rate": 6.0582748050729686e-06,
+      "loss": 0.392,
+      "step": 6460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.124318298918486,
+      "learning_rate": 6.057174269948719e-06,
+      "loss": 0.4149,
+      "step": 6461
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.936973445616139,
+      "learning_rate": 6.056073681208038e-06,
+      "loss": 0.3944,
+      "step": 6462
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.12694833911006,
+      "learning_rate": 6.054973038906741e-06,
+      "loss": 0.4083,
+      "step": 6463
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.087548306698317,
+      "learning_rate": 6.053872343100653e-06,
+      "loss": 0.4162,
+      "step": 6464
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.933812839300831,
+      "learning_rate": 6.0527715938455935e-06,
+      "loss": 0.3876,
+      "step": 6465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6312977211309523,
+      "learning_rate": 6.051670791197393e-06,
+      "loss": 0.4513,
+      "step": 6466
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5908961380851725,
+      "learning_rate": 6.050569935211878e-06,
+      "loss": 0.4382,
+      "step": 6467
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.532388869553097,
+      "learning_rate": 6.0494690259448805e-06,
+      "loss": 0.4086,
+      "step": 6468
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4209140721581079,
+      "learning_rate": 6.0483680634522346e-06,
+      "loss": 0.2097,
+      "step": 6469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.223402268928578,
+      "learning_rate": 6.0472670477897796e-06,
+      "loss": 0.3986,
+      "step": 6470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.288300543739512,
+      "learning_rate": 6.0461659790133545e-06,
+      "loss": 0.4276,
+      "step": 6471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.202448961371114,
+      "learning_rate": 6.045064857178801e-06,
+      "loss": 0.3822,
+      "step": 6472
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.694672534377448,
+      "learning_rate": 6.043963682341964e-06,
+      "loss": 0.3976,
+      "step": 6473
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9609436406649783,
+      "learning_rate": 6.042862454558694e-06,
+      "loss": 0.4159,
+      "step": 6474
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.412920077217396,
+      "learning_rate": 6.04176117388484e-06,
+      "loss": 0.4048,
+      "step": 6475
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.731637470429293,
+      "learning_rate": 6.040659840376256e-06,
+      "loss": 0.394,
+      "step": 6476
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.029757638449528,
+      "learning_rate": 6.039558454088796e-06,
+      "loss": 0.433,
+      "step": 6477
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.393616096777595,
+      "learning_rate": 6.038457015078323e-06,
+      "loss": 0.4103,
+      "step": 6478
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.57685264318926,
+      "learning_rate": 6.0373555234006946e-06,
+      "loss": 0.4193,
+      "step": 6479
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.204174213267637,
+      "learning_rate": 6.0362539791117755e-06,
+      "loss": 0.3982,
+      "step": 6480
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.745957272364804,
+      "learning_rate": 6.0351523822674325e-06,
+      "loss": 0.4107,
+      "step": 6481
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.9250829689530145,
+      "learning_rate": 6.034050732923538e-06,
+      "loss": 0.3881,
+      "step": 6482
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.459774009528093,
+      "learning_rate": 6.032949031135958e-06,
+      "loss": 0.3728,
+      "step": 6483
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.612501485058588,
+      "learning_rate": 6.031847276960573e-06,
+      "loss": 0.4035,
+      "step": 6484
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.737365060137538,
+      "learning_rate": 6.030745470453257e-06,
+      "loss": 0.3889,
+      "step": 6485
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.716091876717683,
+      "learning_rate": 6.029643611669892e-06,
+      "loss": 0.4016,
+      "step": 6486
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.593845108420798,
+      "learning_rate": 6.028541700666359e-06,
+      "loss": 0.4429,
+      "step": 6487
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.237201602157939,
+      "learning_rate": 6.027439737498546e-06,
+      "loss": 0.3992,
+      "step": 6488
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.38477799168293536,
+      "learning_rate": 6.026337722222337e-06,
+      "loss": 0.2025,
+      "step": 6489
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.361322685647168,
+      "learning_rate": 6.025235654893626e-06,
+      "loss": 0.379,
+      "step": 6490
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.768432339199781,
+      "learning_rate": 6.024133535568305e-06,
+      "loss": 0.4037,
+      "step": 6491
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.491161963739377,
+      "learning_rate": 6.02303136430227e-06,
+      "loss": 0.4233,
+      "step": 6492
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.698415829352545,
+      "learning_rate": 6.0219291411514195e-06,
+      "loss": 0.4137,
+      "step": 6493
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.813137850386624,
+      "learning_rate": 6.020826866171654e-06,
+      "loss": 0.4047,
+      "step": 6494
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.047334250274953,
+      "learning_rate": 6.019724539418878e-06,
+      "loss": 0.3839,
+      "step": 6495
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.54768185838219,
+      "learning_rate": 6.018622160948998e-06,
+      "loss": 0.398,
+      "step": 6496
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 9.106629073794132,
+      "learning_rate": 6.017519730817922e-06,
+      "loss": 0.4455,
+      "step": 6497
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 25.407372191317574,
+      "learning_rate": 6.016417249081564e-06,
+      "loss": 0.415,
+      "step": 6498
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.37143753584935,
+      "learning_rate": 6.0153147157958365e-06,
+      "loss": 0.4119,
+      "step": 6499
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.001923746813276,
+      "learning_rate": 6.014212131016657e-06,
+      "loss": 0.4071,
+      "step": 6500
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.444110741016343,
+      "learning_rate": 6.013109494799945e-06,
+      "loss": 0.4252,
+      "step": 6501
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.410718380540035,
+      "learning_rate": 6.012006807201623e-06,
+      "loss": 0.4122,
+      "step": 6502
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7073650091785177,
+      "learning_rate": 6.010904068277612e-06,
+      "loss": 0.4061,
+      "step": 6503
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.5659558874669575,
+      "learning_rate": 6.009801278083845e-06,
+      "loss": 0.3944,
+      "step": 6504
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.172160174500213,
+      "learning_rate": 6.008698436676248e-06,
+      "loss": 0.4157,
+      "step": 6505
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.325903647284693,
+      "learning_rate": 6.007595544110756e-06,
+      "loss": 0.4033,
+      "step": 6506
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.810387662418999,
+      "learning_rate": 6.006492600443301e-06,
+      "loss": 0.3871,
+      "step": 6507
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.298595406516803,
+      "learning_rate": 6.005389605729824e-06,
+      "loss": 0.361,
+      "step": 6508
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.9134342823695283,
+      "learning_rate": 6.004286560026264e-06,
+      "loss": 0.3842,
+      "step": 6509
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.5948661939648536,
+      "learning_rate": 6.003183463388565e-06,
+      "loss": 0.4008,
+      "step": 6510
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.001735124168289,
+      "learning_rate": 6.002080315872668e-06,
+      "loss": 0.4274,
+      "step": 6511
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.49001079478247,
+      "learning_rate": 6.000977117534527e-06,
+      "loss": 0.3675,
+      "step": 6512
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.297470768286402,
+      "learning_rate": 5.999873868430089e-06,
+      "loss": 0.4056,
+      "step": 6513
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.8565273063984415,
+      "learning_rate": 5.9987705686153085e-06,
+      "loss": 0.4254,
+      "step": 6514
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.540626283831394,
+      "learning_rate": 5.997667218146142e-06,
+      "loss": 0.4389,
+      "step": 6515
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 17.10767939682607,
+      "learning_rate": 5.996563817078544e-06,
+      "loss": 0.4212,
+      "step": 6516
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 12.090655737082727,
+      "learning_rate": 5.995460365468482e-06,
+      "loss": 0.4397,
+      "step": 6517
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.0450380710187686,
+      "learning_rate": 5.9943568633719115e-06,
+      "loss": 0.4293,
+      "step": 6518
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.900932727690322,
+      "learning_rate": 5.993253310844805e-06,
+      "loss": 0.4295,
+      "step": 6519
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.571580291862225,
+      "learning_rate": 5.992149707943129e-06,
+      "loss": 0.4055,
+      "step": 6520
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.766676573461422,
+      "learning_rate": 5.991046054722854e-06,
+      "loss": 0.3985,
+      "step": 6521
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.744357314335085,
+      "learning_rate": 5.989942351239954e-06,
+      "loss": 0.38,
+      "step": 6522
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.038289612801645,
+      "learning_rate": 5.988838597550406e-06,
+      "loss": 0.4226,
+      "step": 6523
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.499987776332777,
+      "learning_rate": 5.987734793710187e-06,
+      "loss": 0.3994,
+      "step": 6524
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.547799043121907,
+      "learning_rate": 5.986630939775281e-06,
+      "loss": 0.3913,
+      "step": 6525
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.14363314509529,
+      "learning_rate": 5.985527035801668e-06,
+      "loss": 0.4539,
+      "step": 6526
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.600938117233821,
+      "learning_rate": 5.984423081845339e-06,
+      "loss": 0.4127,
+      "step": 6527
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4579623447620285,
+      "learning_rate": 5.98331907796228e-06,
+      "loss": 0.4218,
+      "step": 6528
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.652499575093625,
+      "learning_rate": 5.982215024208482e-06,
+      "loss": 0.4097,
+      "step": 6529
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.1912059954519805,
+      "learning_rate": 5.981110920639941e-06,
+      "loss": 0.4022,
+      "step": 6530
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 8.739244996819933,
+      "learning_rate": 5.9800067673126526e-06,
+      "loss": 0.4197,
+      "step": 6531
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.6115433196354774,
+      "learning_rate": 5.978902564282616e-06,
+      "loss": 0.4287,
+      "step": 6532
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.738755615703403,
+      "learning_rate": 5.977798311605834e-06,
+      "loss": 0.3647,
+      "step": 6533
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.541922794855026,
+      "learning_rate": 5.976694009338309e-06,
+      "loss": 0.4032,
+      "step": 6534
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 6.309193678970691,
+      "learning_rate": 5.975589657536047e-06,
+      "loss": 0.3802,
+      "step": 6535
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.636103615712144,
+      "learning_rate": 5.974485256255058e-06,
+      "loss": 0.3525,
+      "step": 6536
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.599843196719478,
+      "learning_rate": 5.973380805551354e-06,
+      "loss": 0.4456,
+      "step": 6537
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.002939801563883,
+      "learning_rate": 5.97227630548095e-06,
+      "loss": 0.3896,
+      "step": 6538
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 10.61811338680911,
+      "learning_rate": 5.97117175609986e-06,
+      "loss": 0.3935,
+      "step": 6539
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.0524785113930815,
+      "learning_rate": 5.9700671574641054e-06,
+      "loss": 0.3778,
+      "step": 6540
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.586180041096793,
+      "learning_rate": 5.968962509629709e-06,
+      "loss": 0.4049,
+      "step": 6541
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4807305231297665,
+      "learning_rate": 5.967857812652691e-06,
+      "loss": 0.4006,
+      "step": 6542
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7714824580629567,
+      "learning_rate": 5.966753066589081e-06,
+      "loss": 0.4131,
+      "step": 6543
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.316079556973125,
+      "learning_rate": 5.965648271494906e-06,
+      "loss": 0.429,
+      "step": 6544
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7063475979351423,
+      "learning_rate": 5.9645434274262e-06,
+      "loss": 0.4333,
+      "step": 6545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.167266237954056,
+      "learning_rate": 5.963438534438997e-06,
+      "loss": 0.4293,
+      "step": 6546
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.774454853982152,
+      "learning_rate": 5.962333592589331e-06,
+      "loss": 0.4221,
+      "step": 6547
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.853571465105372,
+      "learning_rate": 5.961228601933242e-06,
+      "loss": 0.4209,
+      "step": 6548
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.944420834238513,
+      "learning_rate": 5.960123562526774e-06,
+      "loss": 0.4237,
+      "step": 6549
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.035875970254219,
+      "learning_rate": 5.959018474425967e-06,
+      "loss": 0.4189,
+      "step": 6550
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.586731558836648,
+      "learning_rate": 5.9579133376868715e-06,
+      "loss": 0.4053,
+      "step": 6551
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.782771179404623,
+      "learning_rate": 5.956808152365532e-06,
+      "loss": 0.3757,
+      "step": 6552
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.215977204262177,
+      "learning_rate": 5.955702918518005e-06,
+      "loss": 0.4047,
+      "step": 6553
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.674904889339668,
+      "learning_rate": 5.95459763620034e-06,
+      "loss": 0.3686,
+      "step": 6554
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.114328223090469,
+      "learning_rate": 5.953492305468597e-06,
+      "loss": 0.4387,
+      "step": 6555
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.361843504382842,
+      "learning_rate": 5.9523869263788314e-06,
+      "loss": 0.4248,
+      "step": 6556
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.158313818450819,
+      "learning_rate": 5.951281498987106e-06,
+      "loss": 0.4075,
+      "step": 6557
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.213176932427182,
+      "learning_rate": 5.950176023349484e-06,
+      "loss": 0.4643,
+      "step": 6558
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.098349201441112,
+      "learning_rate": 5.9490704995220325e-06,
+      "loss": 0.3903,
+      "step": 6559
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 13.496402842505665,
+      "learning_rate": 5.947964927560819e-06,
+      "loss": 0.3983,
+      "step": 6560
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.155322290243392,
+      "learning_rate": 5.946859307521915e-06,
+      "loss": 0.4162,
+      "step": 6561
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38074652889775207,
+      "learning_rate": 5.945753639461393e-06,
+      "loss": 0.1956,
+      "step": 6562
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.4513730943002034,
+      "learning_rate": 5.944647923435332e-06,
+      "loss": 0.454,
+      "step": 6563
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7009488748297605,
+      "learning_rate": 5.943542159499805e-06,
+      "loss": 0.4215,
+      "step": 6564
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.156491740548157,
+      "learning_rate": 5.942436347710899e-06,
+      "loss": 0.3736,
+      "step": 6565
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.771399432050034,
+      "learning_rate": 5.9413304881246915e-06,
+      "loss": 0.3926,
+      "step": 6566
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3695770137554398,
+      "learning_rate": 5.940224580797272e-06,
+      "loss": 0.195,
+      "step": 6567
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.451680350965636,
+      "learning_rate": 5.9391186257847275e-06,
+      "loss": 0.4102,
+      "step": 6568
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.102802198895869,
+      "learning_rate": 5.938012623143148e-06,
+      "loss": 0.4039,
+      "step": 6569
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.189906823727464,
+      "learning_rate": 5.936906572928625e-06,
+      "loss": 0.3907,
+      "step": 6570
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.1669506835784142,
+      "learning_rate": 5.9358004751972555e-06,
+      "loss": 0.4076,
+      "step": 6571
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.8693577452613885,
+      "learning_rate": 5.934694330005136e-06,
+      "loss": 0.3843,
+      "step": 6572
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.082568782697335,
+      "learning_rate": 5.933588137408368e-06,
+      "loss": 0.4227,
+      "step": 6573
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.0972458317424545,
+      "learning_rate": 5.932481897463054e-06,
+      "loss": 0.4195,
+      "step": 6574
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.81941439808215,
+      "learning_rate": 5.931375610225297e-06,
+      "loss": 0.4238,
+      "step": 6575
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.239160699981697,
+      "learning_rate": 5.930269275751206e-06,
+      "loss": 0.403,
+      "step": 6576
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.263209714962642,
+      "learning_rate": 5.929162894096891e-06,
+      "loss": 0.3773,
+      "step": 6577
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.386049044497889,
+      "learning_rate": 5.928056465318462e-06,
+      "loss": 0.4051,
+      "step": 6578
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.97193785710819,
+      "learning_rate": 5.926949989472036e-06,
+      "loss": 0.3969,
+      "step": 6579
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.1235033627207605,
+      "learning_rate": 5.925843466613729e-06,
+      "loss": 0.4004,
+      "step": 6580
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.302229118549784,
+      "learning_rate": 5.924736896799658e-06,
+      "loss": 0.4034,
+      "step": 6581
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 9.605902874072227,
+      "learning_rate": 5.923630280085948e-06,
+      "loss": 0.402,
+      "step": 6582
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6789863972246004,
+      "learning_rate": 5.922523616528719e-06,
+      "loss": 0.4052,
+      "step": 6583
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.863402774574456,
+      "learning_rate": 5.921416906184102e-06,
+      "loss": 0.4269,
+      "step": 6584
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8099545744333234,
+      "learning_rate": 5.920310149108223e-06,
+      "loss": 0.4166,
+      "step": 6585
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.033574724294757,
+      "learning_rate": 5.919203345357214e-06,
+      "loss": 0.4152,
+      "step": 6586
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8049489154235134,
+      "learning_rate": 5.918096494987206e-06,
+      "loss": 0.403,
+      "step": 6587
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.344180470446771,
+      "learning_rate": 5.91698959805434e-06,
+      "loss": 0.3803,
+      "step": 6588
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.2528205107374815,
+      "learning_rate": 5.91588265461475e-06,
+      "loss": 0.3895,
+      "step": 6589
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.533564943700897,
+      "learning_rate": 5.914775664724578e-06,
+      "loss": 0.3966,
+      "step": 6590
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.306521129759099,
+      "learning_rate": 5.913668628439965e-06,
+      "loss": 0.4358,
+      "step": 6591
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.177266090641907,
+      "learning_rate": 5.9125615458170595e-06,
+      "loss": 0.4164,
+      "step": 6592
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.138320659204481,
+      "learning_rate": 5.911454416912006e-06,
+      "loss": 0.3999,
+      "step": 6593
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0621148057934073,
+      "learning_rate": 5.910347241780958e-06,
+      "loss": 0.4066,
+      "step": 6594
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.795782259523838,
+      "learning_rate": 5.9092400204800636e-06,
+      "loss": 0.3909,
+      "step": 6595
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.204159878559541,
+      "learning_rate": 5.908132753065482e-06,
+      "loss": 0.4291,
+      "step": 6596
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.269822672747874,
+      "learning_rate": 5.907025439593366e-06,
+      "loss": 0.4211,
+      "step": 6597
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.227669645997175,
+      "learning_rate": 5.905918080119877e-06,
+      "loss": 0.4016,
+      "step": 6598
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.236864094781895,
+      "learning_rate": 5.904810674701177e-06,
+      "loss": 0.3899,
+      "step": 6599
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 37.601055546450304,
+      "learning_rate": 5.9037032233934284e-06,
+      "loss": 0.4176,
+      "step": 6600
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.339620324399052,
+      "learning_rate": 5.902595726252801e-06,
+      "loss": 0.411,
+      "step": 6601
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.199359831298512,
+      "learning_rate": 5.90148818333546e-06,
+      "loss": 0.4164,
+      "step": 6602
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.802172265371545,
+      "learning_rate": 5.900380594697576e-06,
+      "loss": 0.4152,
+      "step": 6603
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7793355932326462,
+      "learning_rate": 5.899272960395324e-06,
+      "loss": 0.4175,
+      "step": 6604
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4418729185987122,
+      "learning_rate": 5.89816528048488e-06,
+      "loss": 0.4178,
+      "step": 6605
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.179883682892611,
+      "learning_rate": 5.897057555022421e-06,
+      "loss": 0.3843,
+      "step": 6606
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0263210729247136,
+      "learning_rate": 5.895949784064126e-06,
+      "loss": 0.4341,
+      "step": 6607
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.664588395562136,
+      "learning_rate": 5.894841967666181e-06,
+      "loss": 0.4187,
+      "step": 6608
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.773814733814686,
+      "learning_rate": 5.893734105884768e-06,
+      "loss": 0.4178,
+      "step": 6609
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.6842368812977933,
+      "learning_rate": 5.892626198776075e-06,
+      "loss": 0.4112,
+      "step": 6610
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.307608958061455,
+      "learning_rate": 5.891518246396291e-06,
+      "loss": 0.4223,
+      "step": 6611
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8752344469577094,
+      "learning_rate": 5.890410248801608e-06,
+      "loss": 0.4453,
+      "step": 6612
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.233805726300036,
+      "learning_rate": 5.889302206048221e-06,
+      "loss": 0.4402,
+      "step": 6613
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.006083035748077,
+      "learning_rate": 5.8881941181923255e-06,
+      "loss": 0.4245,
+      "step": 6614
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8797567521365948,
+      "learning_rate": 5.8870859852901175e-06,
+      "loss": 0.4741,
+      "step": 6615
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.686774470943225,
+      "learning_rate": 5.885977807397803e-06,
+      "loss": 0.3847,
+      "step": 6616
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.193073427851895,
+      "learning_rate": 5.884869584571581e-06,
+      "loss": 0.3911,
+      "step": 6617
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.275741208126556,
+      "learning_rate": 5.88376131686766e-06,
+      "loss": 0.3813,
+      "step": 6618
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.58670089891873,
+      "learning_rate": 5.882653004342245e-06,
+      "loss": 0.3943,
+      "step": 6619
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0432828365150013,
+      "learning_rate": 5.881544647051547e-06,
+      "loss": 0.4266,
+      "step": 6620
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3120745511889305,
+      "learning_rate": 5.88043624505178e-06,
+      "loss": 0.3849,
+      "step": 6621
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.774723069001402,
+      "learning_rate": 5.879327798399155e-06,
+      "loss": 0.4136,
+      "step": 6622
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7318914542098973,
+      "learning_rate": 5.878219307149892e-06,
+      "loss": 0.3761,
+      "step": 6623
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.72653340031857,
+      "learning_rate": 5.87711077136021e-06,
+      "loss": 0.4004,
+      "step": 6624
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.513837026034654,
+      "learning_rate": 5.876002191086327e-06,
+      "loss": 0.3879,
+      "step": 6625
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.413095181673714,
+      "learning_rate": 5.874893566384471e-06,
+      "loss": 0.4193,
+      "step": 6626
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.952914258730389,
+      "learning_rate": 5.873784897310864e-06,
+      "loss": 0.3981,
+      "step": 6627
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.7377292154436246,
+      "learning_rate": 5.872676183921737e-06,
+      "loss": 0.3859,
+      "step": 6628
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0898268060923657,
+      "learning_rate": 5.871567426273317e-06,
+      "loss": 0.3893,
+      "step": 6629
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.568031987805853,
+      "learning_rate": 5.870458624421841e-06,
+      "loss": 0.4303,
+      "step": 6630
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.361950799027887,
+      "learning_rate": 5.86934977842354e-06,
+      "loss": 0.4366,
+      "step": 6631
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 12.068181225786875,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.4005,
+      "step": 6632
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.213339299388377,
+      "learning_rate": 5.867131954211418e-06,
+      "loss": 0.3899,
+      "step": 6633
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38277896754801277,
+      "learning_rate": 5.866022976110078e-06,
+      "loss": 0.1872,
+      "step": 6634
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8549448560292423,
+      "learning_rate": 5.864913954086876e-06,
+      "loss": 0.4125,
+      "step": 6635
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.837989742976454,
+      "learning_rate": 5.863804888198059e-06,
+      "loss": 0.4294,
+      "step": 6636
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.652816194242347,
+      "learning_rate": 5.862695778499872e-06,
+      "loss": 0.4054,
+      "step": 6637
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.146897274169035,
+      "learning_rate": 5.86158662504857e-06,
+      "loss": 0.4369,
+      "step": 6638
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.502875432585522,
+      "learning_rate": 5.860477427900401e-06,
+      "loss": 0.4094,
+      "step": 6639
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.906732845254486,
+      "learning_rate": 5.859368187111623e-06,
+      "loss": 0.3689,
+      "step": 6640
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.611819361275235,
+      "learning_rate": 5.858258902738493e-06,
+      "loss": 0.4336,
+      "step": 6641
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8022942256825454,
+      "learning_rate": 5.857149574837269e-06,
+      "loss": 0.4315,
+      "step": 6642
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0630388962747963,
+      "learning_rate": 5.856040203464213e-06,
+      "loss": 0.3876,
+      "step": 6643
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.91146141903056,
+      "learning_rate": 5.854930788675589e-06,
+      "loss": 0.4,
+      "step": 6644
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.34755236175972115,
+      "learning_rate": 5.853821330527662e-06,
+      "loss": 0.1987,
+      "step": 6645
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.671031866210766,
+      "learning_rate": 5.852711829076702e-06,
+      "loss": 0.4097,
+      "step": 6646
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.539352916309754,
+      "learning_rate": 5.851602284378978e-06,
+      "loss": 0.3897,
+      "step": 6647
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4432154151019723,
+      "learning_rate": 5.850492696490761e-06,
+      "loss": 0.3758,
+      "step": 6648
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.1713270169841445,
+      "learning_rate": 5.849383065468328e-06,
+      "loss": 0.425,
+      "step": 6649
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.414636634791064,
+      "learning_rate": 5.848273391367955e-06,
+      "loss": 0.4038,
+      "step": 6650
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4756893963751065,
+      "learning_rate": 5.8471636742459215e-06,
+      "loss": 0.411,
+      "step": 6651
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8029464932378425,
+      "learning_rate": 5.846053914158507e-06,
+      "loss": 0.4164,
+      "step": 6652
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4284769120136587,
+      "learning_rate": 5.844944111161999e-06,
+      "loss": 0.4106,
+      "step": 6653
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.347190866188768,
+      "learning_rate": 5.843834265312678e-06,
+      "loss": 0.427,
+      "step": 6654
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.7944542871480635,
+      "learning_rate": 5.842724376666834e-06,
+      "loss": 0.403,
+      "step": 6655
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.493483910015752,
+      "learning_rate": 5.841614445280759e-06,
+      "loss": 0.3734,
+      "step": 6656
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.013350297508667,
+      "learning_rate": 5.840504471210742e-06,
+      "loss": 0.4195,
+      "step": 6657
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.972859984480599,
+      "learning_rate": 5.8393944545130774e-06,
+      "loss": 0.4071,
+      "step": 6658
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.4320270972019005,
+      "learning_rate": 5.838284395244065e-06,
+      "loss": 0.3983,
+      "step": 6659
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.251500827930776,
+      "learning_rate": 5.837174293459999e-06,
+      "loss": 0.3984,
+      "step": 6660
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 18.675252306959752,
+      "learning_rate": 5.836064149217183e-06,
+      "loss": 0.4153,
+      "step": 6661
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3568212315953526,
+      "learning_rate": 5.834953962571919e-06,
+      "loss": 0.1907,
+      "step": 6662
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.731575001477312,
+      "learning_rate": 5.8338437335805124e-06,
+      "loss": 0.3881,
+      "step": 6663
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.566016332982923,
+      "learning_rate": 5.832733462299269e-06,
+      "loss": 0.4216,
+      "step": 6664
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.699112788416835,
+      "learning_rate": 5.831623148784501e-06,
+      "loss": 0.4095,
+      "step": 6665
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.1182246584198365,
+      "learning_rate": 5.830512793092517e-06,
+      "loss": 0.4384,
+      "step": 6666
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.858643372829432,
+      "learning_rate": 5.829402395279632e-06,
+      "loss": 0.4308,
+      "step": 6667
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.69102552490837,
+      "learning_rate": 5.828291955402161e-06,
+      "loss": 0.407,
+      "step": 6668
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.199858482501507,
+      "learning_rate": 5.827181473516423e-06,
+      "loss": 0.4679,
+      "step": 6669
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 8.589233599093049,
+      "learning_rate": 5.826070949678737e-06,
+      "loss": 0.3916,
+      "step": 6670
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.820338397887047,
+      "learning_rate": 5.824960383945427e-06,
+      "loss": 0.4334,
+      "step": 6671
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.719741776471575,
+      "learning_rate": 5.823849776372814e-06,
+      "loss": 0.4086,
+      "step": 6672
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.07788273902036,
+      "learning_rate": 5.822739127017227e-06,
+      "loss": 0.4153,
+      "step": 6673
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4330339759881796,
+      "learning_rate": 5.821628435934994e-06,
+      "loss": 0.4311,
+      "step": 6674
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.38843526864172273,
+      "learning_rate": 5.820517703182446e-06,
+      "loss": 0.1972,
+      "step": 6675
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.809245305804619,
+      "learning_rate": 5.819406928815912e-06,
+      "loss": 0.4019,
+      "step": 6676
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.967323721675069,
+      "learning_rate": 5.8182961128917336e-06,
+      "loss": 0.4072,
+      "step": 6677
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.3024532861046247,
+      "learning_rate": 5.817185255466244e-06,
+      "loss": 0.3905,
+      "step": 6678
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.402174412422531,
+      "learning_rate": 5.81607435659578e-06,
+      "loss": 0.423,
+      "step": 6679
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3600129184930638,
+      "learning_rate": 5.814963416336688e-06,
+      "loss": 0.1968,
+      "step": 6680
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4025909353778134,
+      "learning_rate": 5.813852434745306e-06,
+      "loss": 0.4293,
+      "step": 6681
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.74102409664118,
+      "learning_rate": 5.8127414118779825e-06,
+      "loss": 0.4432,
+      "step": 6682
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9926958008596163,
+      "learning_rate": 5.811630347791065e-06,
+      "loss": 0.3777,
+      "step": 6683
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.467780906129886,
+      "learning_rate": 5.8105192425409015e-06,
+      "loss": 0.4016,
+      "step": 6684
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.8956374612704034,
+      "learning_rate": 5.809408096183845e-06,
+      "loss": 0.4233,
+      "step": 6685
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.3521415006217563,
+      "learning_rate": 5.808296908776247e-06,
+      "loss": 0.1971,
+      "step": 6686
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 7.489507691150733,
+      "learning_rate": 5.807185680374467e-06,
+      "loss": 0.3573,
+      "step": 6687
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5108988771578256,
+      "learning_rate": 5.806074411034861e-06,
+      "loss": 0.4068,
+      "step": 6688
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 14.414199656283193,
+      "learning_rate": 5.8049631008137885e-06,
+      "loss": 0.4175,
+      "step": 6689
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.556541677480369,
+      "learning_rate": 5.8038517497676116e-06,
+      "loss": 0.4245,
+      "step": 6690
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5990741780064535,
+      "learning_rate": 5.802740357952695e-06,
+      "loss": 0.3937,
+      "step": 6691
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.34733014146845476,
+      "learning_rate": 5.801628925425403e-06,
+      "loss": 0.203,
+      "step": 6692
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.895896017553609,
+      "learning_rate": 5.800517452242107e-06,
+      "loss": 0.4254,
+      "step": 6693
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.960891032723665,
+      "learning_rate": 5.799405938459175e-06,
+      "loss": 0.4148,
+      "step": 6694
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.080552681083233,
+      "learning_rate": 5.798294384132981e-06,
+      "loss": 0.4183,
+      "step": 6695
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.703950911082579,
+      "learning_rate": 5.797182789319896e-06,
+      "loss": 0.394,
+      "step": 6696
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8565260355957705,
+      "learning_rate": 5.796071154076301e-06,
+      "loss": 0.3773,
+      "step": 6697
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.648407411424171,
+      "learning_rate": 5.794959478458572e-06,
+      "loss": 0.4139,
+      "step": 6698
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5310610213415354,
+      "learning_rate": 5.793847762523091e-06,
+      "loss": 0.4165,
+      "step": 6699
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.948553155977185,
+      "learning_rate": 5.792736006326237e-06,
+      "loss": 0.3938,
+      "step": 6700
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.232117147885311,
+      "learning_rate": 5.791624209924399e-06,
+      "loss": 0.3808,
+      "step": 6701
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.8774058314874345,
+      "learning_rate": 5.790512373373962e-06,
+      "loss": 0.4124,
+      "step": 6702
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 11.282049700899064,
+      "learning_rate": 5.789400496731315e-06,
+      "loss": 0.4097,
+      "step": 6703
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.532365663494982,
+      "learning_rate": 5.788288580052846e-06,
+      "loss": 0.4454,
+      "step": 6704
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.596582172248198,
+      "learning_rate": 5.787176623394952e-06,
+      "loss": 0.4445,
+      "step": 6705
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 17.993528676240842,
+      "learning_rate": 5.786064626814026e-06,
+      "loss": 0.388,
+      "step": 6706
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.503434849922304,
+      "learning_rate": 5.7849525903664636e-06,
+      "loss": 0.3671,
+      "step": 6707
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.09521001887811,
+      "learning_rate": 5.783840514108667e-06,
+      "loss": 0.3816,
+      "step": 6708
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.830517660551346,
+      "learning_rate": 5.782728398097033e-06,
+      "loss": 0.4017,
+      "step": 6709
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.830198877932207,
+      "learning_rate": 5.781616242387968e-06,
+      "loss": 0.3682,
+      "step": 6710
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 16.360867008435186,
+      "learning_rate": 5.780504047037876e-06,
+      "loss": 0.3895,
+      "step": 6711
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.50354270146342,
+      "learning_rate": 5.779391812103162e-06,
+      "loss": 0.4054,
+      "step": 6712
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.420710767636556,
+      "learning_rate": 5.778279537640236e-06,
+      "loss": 0.412,
+      "step": 6713
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.966126144810564,
+      "learning_rate": 5.7771672237055096e-06,
+      "loss": 0.4252,
+      "step": 6714
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.762435367371644,
+      "learning_rate": 5.776054870355395e-06,
+      "loss": 0.39,
+      "step": 6715
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.4408074813124,
+      "learning_rate": 5.774942477646308e-06,
+      "loss": 0.4238,
+      "step": 6716
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.7547528496872005,
+      "learning_rate": 5.773830045634664e-06,
+      "loss": 0.3739,
+      "step": 6717
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.138579164560818,
+      "learning_rate": 5.772717574376885e-06,
+      "loss": 0.4051,
+      "step": 6718
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.040048252554666,
+      "learning_rate": 5.771605063929388e-06,
+      "loss": 0.4232,
+      "step": 6719
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.768214349289105,
+      "learning_rate": 5.770492514348598e-06,
+      "loss": 0.3772,
+      "step": 6720
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5295067508822404,
+      "learning_rate": 5.76937992569094e-06,
+      "loss": 0.4146,
+      "step": 6721
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.590301589814552,
+      "learning_rate": 5.768267298012841e-06,
+      "loss": 0.4151,
+      "step": 6722
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.927851431813895,
+      "learning_rate": 5.767154631370728e-06,
+      "loss": 0.3842,
+      "step": 6723
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.551940819981742,
+      "learning_rate": 5.766041925821036e-06,
+      "loss": 0.375,
+      "step": 6724
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.97842550857553,
+      "learning_rate": 5.764929181420191e-06,
+      "loss": 0.3604,
+      "step": 6725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.997094052895467,
+      "learning_rate": 5.763816398224634e-06,
+      "loss": 0.4024,
+      "step": 6726
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.682869608624705,
+      "learning_rate": 5.762703576290798e-06,
+      "loss": 0.3937,
+      "step": 6727
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.219744467271052,
+      "learning_rate": 5.7615907156751246e-06,
+      "loss": 0.4065,
+      "step": 6728
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6878664034008777,
+      "learning_rate": 5.760477816434053e-06,
+      "loss": 0.3929,
+      "step": 6729
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.41708951789747284,
+      "learning_rate": 5.759364878624025e-06,
+      "loss": 0.2018,
+      "step": 6730
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.174608235350646,
+      "learning_rate": 5.758251902301486e-06,
+      "loss": 0.4341,
+      "step": 6731
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.861618493490422,
+      "learning_rate": 5.757138887522884e-06,
+      "loss": 0.4252,
+      "step": 6732
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.29413558436487,
+      "learning_rate": 5.756025834344664e-06,
+      "loss": 0.424,
+      "step": 6733
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.07345486355887,
+      "learning_rate": 5.754912742823281e-06,
+      "loss": 0.4024,
+      "step": 6734
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.356579073618122,
+      "learning_rate": 5.753799613015183e-06,
+      "loss": 0.4116,
+      "step": 6735
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.113988148867706,
+      "learning_rate": 5.7526864449768276e-06,
+      "loss": 0.3797,
+      "step": 6736
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.47901412814579,
+      "learning_rate": 5.751573238764667e-06,
+      "loss": 0.4101,
+      "step": 6737
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.75349620281534,
+      "learning_rate": 5.750459994435165e-06,
+      "loss": 0.4211,
+      "step": 6738
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.746509894406798,
+      "learning_rate": 5.749346712044777e-06,
+      "loss": 0.4427,
+      "step": 6739
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.817455042840113,
+      "learning_rate": 5.7482333916499665e-06,
+      "loss": 0.4218,
+      "step": 6740
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.066021095448877,
+      "learning_rate": 5.747120033307198e-06,
+      "loss": 0.4099,
+      "step": 6741
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.1321760581572144,
+      "learning_rate": 5.746006637072938e-06,
+      "loss": 0.3789,
+      "step": 6742
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.667773950396022,
+      "learning_rate": 5.744893203003653e-06,
+      "loss": 0.4225,
+      "step": 6743
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.78736248356012,
+      "learning_rate": 5.743779731155813e-06,
+      "loss": 0.3836,
+      "step": 6744
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.430287772102767,
+      "learning_rate": 5.74266622158589e-06,
+      "loss": 0.4107,
+      "step": 6745
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.447144854313463,
+      "learning_rate": 5.741552674350359e-06,
+      "loss": 0.4054,
+      "step": 6746
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.317916850649105,
+      "learning_rate": 5.740439089505691e-06,
+      "loss": 0.4147,
+      "step": 6747
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 28.447073590205125,
+      "learning_rate": 5.73932546710837e-06,
+      "loss": 0.4176,
+      "step": 6748
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.6051067835807005,
+      "learning_rate": 5.738211807214868e-06,
+      "loss": 0.3933,
+      "step": 6749
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.634287317303261,
+      "learning_rate": 5.737098109881673e-06,
+      "loss": 0.4214,
+      "step": 6750
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.444767635498102,
+      "learning_rate": 5.7359843751652625e-06,
+      "loss": 0.4344,
+      "step": 6751
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.935364196620398,
+      "learning_rate": 5.7348706031221245e-06,
+      "loss": 0.4319,
+      "step": 6752
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7386974390029706,
+      "learning_rate": 5.733756793808745e-06,
+      "loss": 0.3984,
+      "step": 6753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.9083232188307,
+      "learning_rate": 5.732642947281615e-06,
+      "loss": 0.4083,
+      "step": 6754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5682856465056045,
+      "learning_rate": 5.731529063597221e-06,
+      "loss": 0.4144,
+      "step": 6755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.245509845160347,
+      "learning_rate": 5.730415142812059e-06,
+      "loss": 0.4458,
+      "step": 6756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.690026617015436,
+      "learning_rate": 5.729301184982622e-06,
+      "loss": 0.4072,
+      "step": 6757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4884681763975105,
+      "learning_rate": 5.728187190165408e-06,
+      "loss": 0.4433,
+      "step": 6758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.259859373283074,
+      "learning_rate": 5.727073158416911e-06,
+      "loss": 0.356,
+      "step": 6759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.829048190130561,
+      "learning_rate": 5.725959089793636e-06,
+      "loss": 0.4074,
+      "step": 6760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.851532757966321,
+      "learning_rate": 5.724844984352081e-06,
+      "loss": 0.4078,
+      "step": 6761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1866998335266876,
+      "learning_rate": 5.723730842148752e-06,
+      "loss": 0.3977,
+      "step": 6762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.334795504369822,
+      "learning_rate": 5.722616663240155e-06,
+      "loss": 0.3859,
+      "step": 6763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.681167923213458,
+      "learning_rate": 5.721502447682796e-06,
+      "loss": 0.3819,
+      "step": 6764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4151770212568686,
+      "learning_rate": 5.720388195533185e-06,
+      "loss": 0.4144,
+      "step": 6765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3671848234076032,
+      "learning_rate": 5.719273906847834e-06,
+      "loss": 0.1874,
+      "step": 6766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.403882549555752,
+      "learning_rate": 5.7181595816832544e-06,
+      "loss": 0.4579,
+      "step": 6767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.364771882115749,
+      "learning_rate": 5.7170452200959635e-06,
+      "loss": 0.4023,
+      "step": 6768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.267684388068331,
+      "learning_rate": 5.715930822142477e-06,
+      "loss": 0.381,
+      "step": 6769
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8421283551184215,
+      "learning_rate": 5.714816387879312e-06,
+      "loss": 0.3909,
+      "step": 6770
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9308192601920937,
+      "learning_rate": 5.71370191736299e-06,
+      "loss": 0.3965,
+      "step": 6771
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.37851444789033944,
+      "learning_rate": 5.712587410650034e-06,
+      "loss": 0.1919,
+      "step": 6772
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.683259835660982,
+      "learning_rate": 5.7114728677969684e-06,
+      "loss": 0.3905,
+      "step": 6773
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.44722843614603,
+      "learning_rate": 5.710358288860317e-06,
+      "loss": 0.3989,
+      "step": 6774
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.16286810417561,
+      "learning_rate": 5.7092436738966105e-06,
+      "loss": 0.3991,
+      "step": 6775
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 13.785763956431003,
+      "learning_rate": 5.708129022962377e-06,
+      "loss": 0.396,
+      "step": 6776
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5791030594585598,
+      "learning_rate": 5.707014336114147e-06,
+      "loss": 0.384,
+      "step": 6777
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.5793272055211327,
+      "learning_rate": 5.705899613408456e-06,
+      "loss": 0.388,
+      "step": 6778
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.7524306023398175,
+      "learning_rate": 5.704784854901839e-06,
+      "loss": 0.3956,
+      "step": 6779
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.314864032630883,
+      "learning_rate": 5.70367006065083e-06,
+      "loss": 0.418,
+      "step": 6780
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.785056086309018,
+      "learning_rate": 5.70255523071197e-06,
+      "loss": 0.4269,
+      "step": 6781
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.1068467712833625,
+      "learning_rate": 5.701440365141799e-06,
+      "loss": 0.4246,
+      "step": 6782
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.6056437693425485,
+      "learning_rate": 5.7003254639968595e-06,
+      "loss": 0.3916,
+      "step": 6783
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.429044654172707,
+      "learning_rate": 5.6992105273336964e-06,
+      "loss": 0.4207,
+      "step": 6784
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.264672328237512,
+      "learning_rate": 5.698095555208854e-06,
+      "loss": 0.3963,
+      "step": 6785
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.395683784284995,
+      "learning_rate": 5.6969805476788805e-06,
+      "loss": 0.4353,
+      "step": 6786
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3193316351291124,
+      "learning_rate": 5.695865504800328e-06,
+      "loss": 0.4016,
+      "step": 6787
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 20.144547586924354,
+      "learning_rate": 5.694750426629744e-06,
+      "loss": 0.4309,
+      "step": 6788
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.192248233208088,
+      "learning_rate": 5.693635313223684e-06,
+      "loss": 0.388,
+      "step": 6789
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.910201712576569,
+      "learning_rate": 5.692520164638703e-06,
+      "loss": 0.4143,
+      "step": 6790
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.990126569515261,
+      "learning_rate": 5.691404980931357e-06,
+      "loss": 0.4236,
+      "step": 6791
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3571801974254644,
+      "learning_rate": 5.690289762158203e-06,
+      "loss": 0.1889,
+      "step": 6792
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.0828980112440405,
+      "learning_rate": 5.689174508375803e-06,
+      "loss": 0.4235,
+      "step": 6793
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.4299008044935966,
+      "learning_rate": 5.688059219640719e-06,
+      "loss": 0.4364,
+      "step": 6794
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.289556911375562,
+      "learning_rate": 5.686943896009516e-06,
+      "loss": 0.4126,
+      "step": 6795
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8137656366004706,
+      "learning_rate": 5.685828537538756e-06,
+      "loss": 0.4062,
+      "step": 6796
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.579610336063203,
+      "learning_rate": 5.68471314428501e-06,
+      "loss": 0.4448,
+      "step": 6797
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.73610628936591,
+      "learning_rate": 5.683597716304844e-06,
+      "loss": 0.4283,
+      "step": 6798
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.596732233967984,
+      "learning_rate": 5.682482253654832e-06,
+      "loss": 0.4058,
+      "step": 6799
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.03979311688344,
+      "learning_rate": 5.681366756391544e-06,
+      "loss": 0.4632,
+      "step": 6800
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.702203138367491,
+      "learning_rate": 5.680251224571557e-06,
+      "loss": 0.3998,
+      "step": 6801
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 9.92760299719288,
+      "learning_rate": 5.679135658251444e-06,
+      "loss": 0.37,
+      "step": 6802
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.304302306684123,
+      "learning_rate": 5.678020057487787e-06,
+      "loss": 0.4189,
+      "step": 6803
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.463596784698059,
+      "learning_rate": 5.676904422337161e-06,
+      "loss": 0.384,
+      "step": 6804
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.8673986969077334,
+      "learning_rate": 5.67578875285615e-06,
+      "loss": 0.3859,
+      "step": 6805
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.360246306709985,
+      "learning_rate": 5.674673049101336e-06,
+      "loss": 0.3984,
+      "step": 6806
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.138453518317409,
+      "learning_rate": 5.673557311129306e-06,
+      "loss": 0.3989,
+      "step": 6807
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.220142930251122,
+      "learning_rate": 5.672441538996643e-06,
+      "loss": 0.4142,
+      "step": 6808
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9728039489266807,
+      "learning_rate": 5.67132573275994e-06,
+      "loss": 0.4277,
+      "step": 6809
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0436218661325896,
+      "learning_rate": 5.670209892475782e-06,
+      "loss": 0.3673,
+      "step": 6810
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.171073093181882,
+      "learning_rate": 5.669094018200764e-06,
+      "loss": 0.4379,
+      "step": 6811
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.259338568014855,
+      "learning_rate": 5.667978109991481e-06,
+      "loss": 0.3964,
+      "step": 6812
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9727109809625767,
+      "learning_rate": 5.666862167904524e-06,
+      "loss": 0.4165,
+      "step": 6813
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.351993050493498,
+      "learning_rate": 5.665746191996492e-06,
+      "loss": 0.3974,
+      "step": 6814
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.383216709318127,
+      "learning_rate": 5.6646301823239845e-06,
+      "loss": 0.4445,
+      "step": 6815
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.642572628792951,
+      "learning_rate": 5.6635141389436e-06,
+      "loss": 0.3979,
+      "step": 6816
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.554867415300995,
+      "learning_rate": 5.662398061911943e-06,
+      "loss": 0.4346,
+      "step": 6817
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.759252550428663,
+      "learning_rate": 5.661281951285613e-06,
+      "loss": 0.4124,
+      "step": 6818
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.535993728157059,
+      "learning_rate": 5.66016580712122e-06,
+      "loss": 0.4225,
+      "step": 6819
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.249620442408545,
+      "learning_rate": 5.659049629475369e-06,
+      "loss": 0.4459,
+      "step": 6820
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9313135754579265,
+      "learning_rate": 5.657933418404671e-06,
+      "loss": 0.4061,
+      "step": 6821
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 8.317353683590145,
+      "learning_rate": 5.656817173965733e-06,
+      "loss": 0.4307,
+      "step": 6822
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.557171005589265,
+      "learning_rate": 5.655700896215171e-06,
+      "loss": 0.3946,
+      "step": 6823
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.984023813497867,
+      "learning_rate": 5.654584585209596e-06,
+      "loss": 0.4331,
+      "step": 6824
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.27057374130368,
+      "learning_rate": 5.6534682410056265e-06,
+      "loss": 0.3915,
+      "step": 6825
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.3875914151428725,
+      "learning_rate": 5.652351863659876e-06,
+      "loss": 0.3788,
+      "step": 6826
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.70110449408609,
+      "learning_rate": 5.651235453228967e-06,
+      "loss": 0.4221,
+      "step": 6827
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.393452957142705,
+      "learning_rate": 5.650119009769519e-06,
+      "loss": 0.3823,
+      "step": 6828
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.9294433417174632,
+      "learning_rate": 5.649002533338155e-06,
+      "loss": 0.3585,
+      "step": 6829
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.904339684022014,
+      "learning_rate": 5.647886023991496e-06,
+      "loss": 0.4118,
+      "step": 6830
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.806701575632174,
+      "learning_rate": 5.6467694817861715e-06,
+      "loss": 0.391,
+      "step": 6831
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.951871137257706,
+      "learning_rate": 5.645652906778808e-06,
+      "loss": 0.4164,
+      "step": 6832
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.39903502683620257,
+      "learning_rate": 5.644536299026033e-06,
+      "loss": 0.2086,
+      "step": 6833
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.622755962110787,
+      "learning_rate": 5.643419658584479e-06,
+      "loss": 0.435,
+      "step": 6834
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.35968414214305927,
+      "learning_rate": 5.642302985510778e-06,
+      "loss": 0.1908,
+      "step": 6835
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8029199017999926,
+      "learning_rate": 5.641186279861563e-06,
+      "loss": 0.377,
+      "step": 6836
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.8742115137675155,
+      "learning_rate": 5.64006954169347e-06,
+      "loss": 0.3681,
+      "step": 6837
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.875220033986675,
+      "learning_rate": 5.6389527710631374e-06,
+      "loss": 0.4067,
+      "step": 6838
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4544075304122415,
+      "learning_rate": 5.637835968027202e-06,
+      "loss": 0.441,
+      "step": 6839
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.287373768916172,
+      "learning_rate": 5.636719132642308e-06,
+      "loss": 0.4274,
+      "step": 6840
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38108057471855816,
+      "learning_rate": 5.635602264965093e-06,
+      "loss": 0.1906,
+      "step": 6841
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.586984977213552,
+      "learning_rate": 5.6344853650522045e-06,
+      "loss": 0.4011,
+      "step": 6842
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.328232095304275,
+      "learning_rate": 5.633368432960286e-06,
+      "loss": 0.4322,
+      "step": 6843
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3462134114546004,
+      "learning_rate": 5.632251468745987e-06,
+      "loss": 0.1873,
+      "step": 6844
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.666285699025433,
+      "learning_rate": 5.631134472465954e-06,
+      "loss": 0.3994,
+      "step": 6845
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2438220362890995,
+      "learning_rate": 5.630017444176838e-06,
+      "loss": 0.3792,
+      "step": 6846
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 8.263206901816478,
+      "learning_rate": 5.628900383935292e-06,
+      "loss": 0.4349,
+      "step": 6847
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.620089831375731,
+      "learning_rate": 5.627783291797968e-06,
+      "loss": 0.3906,
+      "step": 6848
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.568364995435934,
+      "learning_rate": 5.626666167821522e-06,
+      "loss": 0.4135,
+      "step": 6849
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.096387387875023,
+      "learning_rate": 5.625549012062611e-06,
+      "loss": 0.3912,
+      "step": 6850
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.24842093591508,
+      "learning_rate": 5.624431824577894e-06,
+      "loss": 0.415,
+      "step": 6851
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.305593114642639,
+      "learning_rate": 5.623314605424031e-06,
+      "loss": 0.4055,
+      "step": 6852
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.24648665373438,
+      "learning_rate": 5.622197354657682e-06,
+      "loss": 0.4024,
+      "step": 6853
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.389106565516562,
+      "learning_rate": 5.621080072335513e-06,
+      "loss": 0.3976,
+      "step": 6854
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.128153241098216,
+      "learning_rate": 5.619962758514186e-06,
+      "loss": 0.3883,
+      "step": 6855
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8398744845203847,
+      "learning_rate": 5.618845413250371e-06,
+      "loss": 0.3828,
+      "step": 6856
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4040906731784509,
+      "learning_rate": 5.617728036600734e-06,
+      "loss": 0.1945,
+      "step": 6857
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5745188238083587,
+      "learning_rate": 5.616610628621944e-06,
+      "loss": 0.4081,
+      "step": 6858
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.619151632941284,
+      "learning_rate": 5.615493189370673e-06,
+      "loss": 0.4068,
+      "step": 6859
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5679851171144215,
+      "learning_rate": 5.614375718903594e-06,
+      "loss": 0.3877,
+      "step": 6860
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.4198765847790655,
+      "learning_rate": 5.613258217277382e-06,
+      "loss": 0.4092,
+      "step": 6861
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.809984570124264,
+      "learning_rate": 5.612140684548713e-06,
+      "loss": 0.3993,
+      "step": 6862
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.764783324512602,
+      "learning_rate": 5.611023120774263e-06,
+      "loss": 0.4314,
+      "step": 6863
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.629454142791617,
+      "learning_rate": 5.609905526010713e-06,
+      "loss": 0.4087,
+      "step": 6864
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.682020448275841,
+      "learning_rate": 5.608787900314742e-06,
+      "loss": 0.4123,
+      "step": 6865
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7727918796498887,
+      "learning_rate": 5.607670243743036e-06,
+      "loss": 0.4252,
+      "step": 6866
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.134794077952061,
+      "learning_rate": 5.606552556352275e-06,
+      "loss": 0.4157,
+      "step": 6867
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.3300689684086775,
+      "learning_rate": 5.605434838199145e-06,
+      "loss": 0.4123,
+      "step": 6868
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7706795295314306,
+      "learning_rate": 5.604317089340336e-06,
+      "loss": 0.3908,
+      "step": 6869
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.393545235882686,
+      "learning_rate": 5.603199309832533e-06,
+      "loss": 0.4456,
+      "step": 6870
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.660215556120972,
+      "learning_rate": 5.602081499732427e-06,
+      "loss": 0.3879,
+      "step": 6871
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7811654397304966,
+      "learning_rate": 5.600963659096711e-06,
+      "loss": 0.3931,
+      "step": 6872
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7765846312316484,
+      "learning_rate": 5.599845787982077e-06,
+      "loss": 0.4218,
+      "step": 6873
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3951766997037773,
+      "learning_rate": 5.598727886445221e-06,
+      "loss": 0.4288,
+      "step": 6874
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.335928466459123,
+      "learning_rate": 5.597609954542838e-06,
+      "loss": 0.425,
+      "step": 6875
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.548679647533326,
+      "learning_rate": 5.596491992331626e-06,
+      "loss": 0.4114,
+      "step": 6876
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.196235286172363,
+      "learning_rate": 5.595373999868285e-06,
+      "loss": 0.4154,
+      "step": 6877
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.605144974852955,
+      "learning_rate": 5.594255977209516e-06,
+      "loss": 0.3883,
+      "step": 6878
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3526498142250092,
+      "learning_rate": 5.593137924412022e-06,
+      "loss": 0.1942,
+      "step": 6879
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3705674733077557,
+      "learning_rate": 5.592019841532507e-06,
+      "loss": 0.1927,
+      "step": 6880
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5069518869129386,
+      "learning_rate": 5.590901728627673e-06,
+      "loss": 0.4139,
+      "step": 6881
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.5919078120820958,
+      "learning_rate": 5.5897835857542315e-06,
+      "loss": 0.4265,
+      "step": 6882
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.159021882692171,
+      "learning_rate": 5.588665412968888e-06,
+      "loss": 0.3978,
+      "step": 6883
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.36943146744439026,
+      "learning_rate": 5.587547210328356e-06,
+      "loss": 0.2065,
+      "step": 6884
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 7.051313381357865,
+      "learning_rate": 5.586428977889342e-06,
+      "loss": 0.4113,
+      "step": 6885
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.648598350345457,
+      "learning_rate": 5.585310715708565e-06,
+      "loss": 0.378,
+      "step": 6886
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.9507769442160505,
+      "learning_rate": 5.584192423842735e-06,
+      "loss": 0.4022,
+      "step": 6887
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.182440127460199,
+      "learning_rate": 5.583074102348571e-06,
+      "loss": 0.4413,
+      "step": 6888
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4203979135012954,
+      "learning_rate": 5.581955751282788e-06,
+      "loss": 0.3735,
+      "step": 6889
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.238248219217864,
+      "learning_rate": 5.580837370702108e-06,
+      "loss": 0.4632,
+      "step": 6890
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 32.82634446515978,
+      "learning_rate": 5.579718960663251e-06,
+      "loss": 0.4398,
+      "step": 6891
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.40167878794969913,
+      "learning_rate": 5.5786005212229365e-06,
+      "loss": 0.1976,
+      "step": 6892
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3891019367216826,
+      "learning_rate": 5.577482052437891e-06,
+      "loss": 0.4178,
+      "step": 6893
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.033286758444969,
+      "learning_rate": 5.576363554364838e-06,
+      "loss": 0.4107,
+      "step": 6894
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9341704784594542,
+      "learning_rate": 5.575245027060503e-06,
+      "loss": 0.4456,
+      "step": 6895
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.92824100702026,
+      "learning_rate": 5.574126470581617e-06,
+      "loss": 0.3973,
+      "step": 6896
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9781688949335345,
+      "learning_rate": 5.573007884984909e-06,
+      "loss": 0.4134,
+      "step": 6897
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.115858451263442,
+      "learning_rate": 5.571889270327107e-06,
+      "loss": 0.4381,
+      "step": 6898
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.5405420472413205,
+      "learning_rate": 5.570770626664947e-06,
+      "loss": 0.4133,
+      "step": 6899
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.259387643983986,
+      "learning_rate": 5.569651954055163e-06,
+      "loss": 0.3881,
+      "step": 6900
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1298405045679236,
+      "learning_rate": 5.568533252554489e-06,
+      "loss": 0.3824,
+      "step": 6901
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.370005929517604,
+      "learning_rate": 5.5674145222196596e-06,
+      "loss": 0.434,
+      "step": 6902
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4900932948785943,
+      "learning_rate": 5.566295763107418e-06,
+      "loss": 0.3977,
+      "step": 6903
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.878806122592148,
+      "learning_rate": 5.5651769752744985e-06,
+      "loss": 0.4059,
+      "step": 6904
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.440645595113124,
+      "learning_rate": 5.564058158777647e-06,
+      "loss": 0.43,
+      "step": 6905
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.854365941693098,
+      "learning_rate": 5.562939313673605e-06,
+      "loss": 0.3843,
+      "step": 6906
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8268021257917355,
+      "learning_rate": 5.561820440019117e-06,
+      "loss": 0.4275,
+      "step": 6907
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.283399637854348,
+      "learning_rate": 5.560701537870926e-06,
+      "loss": 0.4419,
+      "step": 6908
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4797745431323803,
+      "learning_rate": 5.559582607285783e-06,
+      "loss": 0.4291,
+      "step": 6909
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.0120371273511175,
+      "learning_rate": 5.558463648320432e-06,
+      "loss": 0.4093,
+      "step": 6910
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3372160048771917,
+      "learning_rate": 5.557344661031628e-06,
+      "loss": 0.3754,
+      "step": 6911
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.030103414227308,
+      "learning_rate": 5.556225645476119e-06,
+      "loss": 0.3785,
+      "step": 6912
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8138415367885967,
+      "learning_rate": 5.555106601710659e-06,
+      "loss": 0.385,
+      "step": 6913
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 22.78250688881993,
+      "learning_rate": 5.553987529792001e-06,
+      "loss": 0.397,
+      "step": 6914
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 10.50745063852041,
+      "learning_rate": 5.552868429776902e-06,
+      "loss": 0.3727,
+      "step": 6915
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.7849683936433935,
+      "learning_rate": 5.551749301722118e-06,
+      "loss": 0.3708,
+      "step": 6916
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3769958966951805,
+      "learning_rate": 5.550630145684409e-06,
+      "loss": 0.2091,
+      "step": 6917
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.634185225796393,
+      "learning_rate": 5.549510961720533e-06,
+      "loss": 0.3869,
+      "step": 6918
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.979960578889654,
+      "learning_rate": 5.548391749887255e-06,
+      "loss": 0.4508,
+      "step": 6919
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.8221406559001023,
+      "learning_rate": 5.547272510241333e-06,
+      "loss": 0.4227,
+      "step": 6920
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3694846650784327,
+      "learning_rate": 5.546153242839535e-06,
+      "loss": 0.4337,
+      "step": 6921
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.121840447493471,
+      "learning_rate": 5.545033947738624e-06,
+      "loss": 0.391,
+      "step": 6922
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.721639966637442,
+      "learning_rate": 5.543914624995369e-06,
+      "loss": 0.4122,
+      "step": 6923
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.2366019813646556,
+      "learning_rate": 5.542795274666537e-06,
+      "loss": 0.4147,
+      "step": 6924
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.473133169864706,
+      "learning_rate": 5.541675896808899e-06,
+      "loss": 0.4251,
+      "step": 6925
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4117631782212414,
+      "learning_rate": 5.540556491479224e-06,
+      "loss": 0.4127,
+      "step": 6926
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0540193755402223,
+      "learning_rate": 5.539437058734287e-06,
+      "loss": 0.4379,
+      "step": 6927
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38156520964260604,
+      "learning_rate": 5.5383175986308605e-06,
+      "loss": 0.1909,
+      "step": 6928
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.868656178810441,
+      "learning_rate": 5.537198111225721e-06,
+      "loss": 0.3968,
+      "step": 6929
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.525293357814393,
+      "learning_rate": 5.5360785965756445e-06,
+      "loss": 0.4267,
+      "step": 6930
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 21.334780332466853,
+      "learning_rate": 5.53495905473741e-06,
+      "loss": 0.4082,
+      "step": 6931
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.949223218007834,
+      "learning_rate": 5.533839485767795e-06,
+      "loss": 0.3627,
+      "step": 6932
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.763431490835007,
+      "learning_rate": 5.532719889723582e-06,
+      "loss": 0.3883,
+      "step": 6933
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.161093701422762,
+      "learning_rate": 5.531600266661554e-06,
+      "loss": 0.4293,
+      "step": 6934
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.803317593978081,
+      "learning_rate": 5.530480616638492e-06,
+      "loss": 0.435,
+      "step": 6935
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.235125047682835,
+      "learning_rate": 5.529360939711184e-06,
+      "loss": 0.4253,
+      "step": 6936
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.7593446139256,
+      "learning_rate": 5.528241235936415e-06,
+      "loss": 0.4213,
+      "step": 6937
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3637305746617518,
+      "learning_rate": 5.527121505370971e-06,
+      "loss": 0.1927,
+      "step": 6938
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.78016078926331,
+      "learning_rate": 5.526001748071645e-06,
+      "loss": 0.3862,
+      "step": 6939
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.909434691962065,
+      "learning_rate": 5.524881964095222e-06,
+      "loss": 0.3904,
+      "step": 6940
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.495710359070669,
+      "learning_rate": 5.5237621534985e-06,
+      "loss": 0.3774,
+      "step": 6941
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.594384322590744,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.4079,
+      "step": 6942
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.108214909121619,
+      "learning_rate": 5.521522452671322e-06,
+      "loss": 0.383,
+      "step": 6943
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1501413076835174,
+      "learning_rate": 5.520402562554456e-06,
+      "loss": 0.388,
+      "step": 6944
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.0433432346316227,
+      "learning_rate": 5.519282646044471e-06,
+      "loss": 0.3773,
+      "step": 6945
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.154643024538289,
+      "learning_rate": 5.518162703198162e-06,
+      "loss": 0.4276,
+      "step": 6946
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.4430394227039245,
+      "learning_rate": 5.517042734072331e-06,
+      "loss": 0.462,
+      "step": 6947
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.573853808071206,
+      "learning_rate": 5.515922738723777e-06,
+      "loss": 0.4099,
+      "step": 6948
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.17366909259663,
+      "learning_rate": 5.514802717209304e-06,
+      "loss": 0.4081,
+      "step": 6949
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7871689401961444,
+      "learning_rate": 5.513682669585715e-06,
+      "loss": 0.4052,
+      "step": 6950
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.698185428462423,
+      "learning_rate": 5.512562595909816e-06,
+      "loss": 0.4145,
+      "step": 6951
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9477665589797155,
+      "learning_rate": 5.511442496238413e-06,
+      "loss": 0.3802,
+      "step": 6952
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.206979101056886,
+      "learning_rate": 5.510322370628315e-06,
+      "loss": 0.4303,
+      "step": 6953
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.914802238212199,
+      "learning_rate": 5.5092022191363285e-06,
+      "loss": 0.376,
+      "step": 6954
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.7639869456932322,
+      "learning_rate": 5.5080820418192674e-06,
+      "loss": 0.4399,
+      "step": 6955
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.602163147907719,
+      "learning_rate": 5.5069618387339395e-06,
+      "loss": 0.4086,
+      "step": 6956
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.123713588216426,
+      "learning_rate": 5.505841609937162e-06,
+      "loss": 0.3928,
+      "step": 6957
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.137739878307426,
+      "learning_rate": 5.504721355485746e-06,
+      "loss": 0.4165,
+      "step": 6958
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.052058425220952,
+      "learning_rate": 5.503601075436509e-06,
+      "loss": 0.4084,
+      "step": 6959
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.968273028750585,
+      "learning_rate": 5.502480769846267e-06,
+      "loss": 0.4012,
+      "step": 6960
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.9291767892133302,
+      "learning_rate": 5.501360438771838e-06,
+      "loss": 0.3933,
+      "step": 6961
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1342984790623785,
+      "learning_rate": 5.5002400822700415e-06,
+      "loss": 0.4105,
+      "step": 6962
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.338457239499782,
+      "learning_rate": 5.4991197003977e-06,
+      "loss": 0.4033,
+      "step": 6963
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.296470113510665,
+      "learning_rate": 5.497999293211636e-06,
+      "loss": 0.4226,
+      "step": 6964
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.188377727420085,
+      "learning_rate": 5.496878860768669e-06,
+      "loss": 0.3879,
+      "step": 6965
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9222505473633906,
+      "learning_rate": 5.495758403125626e-06,
+      "loss": 0.3913,
+      "step": 6966
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.358957460721612,
+      "learning_rate": 5.494637920339335e-06,
+      "loss": 0.4227,
+      "step": 6967
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.895254978322963,
+      "learning_rate": 5.49351741246662e-06,
+      "loss": 0.3854,
+      "step": 6968
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9571522561791976,
+      "learning_rate": 5.4923968795643115e-06,
+      "loss": 0.3812,
+      "step": 6969
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.708006920159842,
+      "learning_rate": 5.4912763216892386e-06,
+      "loss": 0.3774,
+      "step": 6970
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.735253838427071,
+      "learning_rate": 5.490155738898231e-06,
+      "loss": 0.4299,
+      "step": 6971
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.06351947505917,
+      "learning_rate": 5.489035131248124e-06,
+      "loss": 0.4271,
+      "step": 6972
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.6396445833810684,
+      "learning_rate": 5.487914498795748e-06,
+      "loss": 0.4118,
+      "step": 6973
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.519721586772924,
+      "learning_rate": 5.486793841597941e-06,
+      "loss": 0.4516,
+      "step": 6974
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.611317045825046,
+      "learning_rate": 5.485673159711535e-06,
+      "loss": 0.3964,
+      "step": 6975
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1442512624031242,
+      "learning_rate": 5.484552453193371e-06,
+      "loss": 0.4252,
+      "step": 6976
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.807460237284276,
+      "learning_rate": 5.4834317221002866e-06,
+      "loss": 0.4064,
+      "step": 6977
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.597710016495944,
+      "learning_rate": 5.482310966489122e-06,
+      "loss": 0.4025,
+      "step": 6978
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.156616646389269,
+      "learning_rate": 5.481190186416716e-06,
+      "loss": 0.4161,
+      "step": 6979
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5989452949487775,
+      "learning_rate": 5.480069381939914e-06,
+      "loss": 0.4157,
+      "step": 6980
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.899220995282168,
+      "learning_rate": 5.478948553115557e-06,
+      "loss": 0.3644,
+      "step": 6981
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8822622484901776,
+      "learning_rate": 5.477827700000492e-06,
+      "loss": 0.3974,
+      "step": 6982
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.50315348080426,
+      "learning_rate": 5.4767068226515614e-06,
+      "loss": 0.3495,
+      "step": 6983
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8424195843622386,
+      "learning_rate": 5.475585921125618e-06,
+      "loss": 0.424,
+      "step": 6984
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.026847967125983,
+      "learning_rate": 5.4744649954795045e-06,
+      "loss": 0.3705,
+      "step": 6985
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4881480070827395,
+      "learning_rate": 5.473344045770075e-06,
+      "loss": 0.43,
+      "step": 6986
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1041416349330433,
+      "learning_rate": 5.472223072054178e-06,
+      "loss": 0.4025,
+      "step": 6987
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.353789351163848,
+      "learning_rate": 5.471102074388667e-06,
+      "loss": 0.378,
+      "step": 6988
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.957224660674983,
+      "learning_rate": 5.4699810528303934e-06,
+      "loss": 0.4172,
+      "step": 6989
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.351391940502634,
+      "learning_rate": 5.468860007436214e-06,
+      "loss": 0.4043,
+      "step": 6990
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.194918119328018,
+      "learning_rate": 5.467738938262984e-06,
+      "loss": 0.4223,
+      "step": 6991
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.03472009799612,
+      "learning_rate": 5.46661784536756e-06,
+      "loss": 0.4389,
+      "step": 6992
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6137353882581156,
+      "learning_rate": 5.4654967288067985e-06,
+      "loss": 0.4249,
+      "step": 6993
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1926393710173446,
+      "learning_rate": 5.464375588637563e-06,
+      "loss": 0.4274,
+      "step": 6994
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.1459477834921294,
+      "learning_rate": 5.463254424916709e-06,
+      "loss": 0.4032,
+      "step": 6995
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.955650036944799,
+      "learning_rate": 5.462133237701103e-06,
+      "loss": 0.4137,
+      "step": 6996
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.023800070068386,
+      "learning_rate": 5.461012027047603e-06,
+      "loss": 0.3923,
+      "step": 6997
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.560274252160508,
+      "learning_rate": 5.459890793013078e-06,
+      "loss": 0.3959,
+      "step": 6998
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.093690077143668,
+      "learning_rate": 5.458769535654392e-06,
+      "loss": 0.4073,
+      "step": 6999
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.310048865100065,
+      "learning_rate": 5.457648255028409e-06,
+      "loss": 0.3984,
+      "step": 7000
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.35520191517918187,
+      "learning_rate": 5.4565269511920006e-06,
+      "loss": 0.1941,
+      "step": 7001
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.4563807033771115,
+      "learning_rate": 5.455405624202032e-06,
+      "loss": 0.3877,
+      "step": 7002
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8337882673528534,
+      "learning_rate": 5.454284274115376e-06,
+      "loss": 0.4072,
+      "step": 7003
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.779892403110818,
+      "learning_rate": 5.453162900988902e-06,
+      "loss": 0.4146,
+      "step": 7004
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1728434408839146,
+      "learning_rate": 5.452041504879483e-06,
+      "loss": 0.3897,
+      "step": 7005
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.251013065496736,
+      "learning_rate": 5.450920085843993e-06,
+      "loss": 0.408,
+      "step": 7006
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.43466587947268,
+      "learning_rate": 5.449798643939305e-06,
+      "loss": 0.4143,
+      "step": 7007
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.84439537584942,
+      "learning_rate": 5.448677179222298e-06,
+      "loss": 0.426,
+      "step": 7008
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.811924650413788,
+      "learning_rate": 5.447555691749845e-06,
+      "loss": 0.4257,
+      "step": 7009
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.552749735473949,
+      "learning_rate": 5.446434181578829e-06,
+      "loss": 0.4053,
+      "step": 7010
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8606348634781473,
+      "learning_rate": 5.445312648766125e-06,
+      "loss": 0.4188,
+      "step": 7011
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.638954334565788,
+      "learning_rate": 5.444191093368616e-06,
+      "loss": 0.4394,
+      "step": 7012
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5645109206126913,
+      "learning_rate": 5.443069515443184e-06,
+      "loss": 0.4309,
+      "step": 7013
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.290429792032968,
+      "learning_rate": 5.441947915046709e-06,
+      "loss": 0.4167,
+      "step": 7014
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.404791304475636,
+      "learning_rate": 5.440826292236076e-06,
+      "loss": 0.4243,
+      "step": 7015
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.348255104096188,
+      "learning_rate": 5.439704647068173e-06,
+      "loss": 0.1938,
+      "step": 7016
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.705161393120885,
+      "learning_rate": 5.4385829795998815e-06,
+      "loss": 0.424,
+      "step": 7017
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0707566533184267,
+      "learning_rate": 5.4374612898880926e-06,
+      "loss": 0.3796,
+      "step": 7018
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.211649835120833,
+      "learning_rate": 5.436339577989692e-06,
+      "loss": 0.4116,
+      "step": 7019
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.293491201765429,
+      "learning_rate": 5.435217843961572e-06,
+      "loss": 0.4194,
+      "step": 7020
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.010659379445601,
+      "learning_rate": 5.434096087860621e-06,
+      "loss": 0.4047,
+      "step": 7021
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7699183734465946,
+      "learning_rate": 5.4329743097437316e-06,
+      "loss": 0.4274,
+      "step": 7022
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.346819374599488,
+      "learning_rate": 5.4318525096677966e-06,
+      "loss": 0.4131,
+      "step": 7023
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.548511175850004,
+      "learning_rate": 5.430730687689711e-06,
+      "loss": 0.4116,
+      "step": 7024
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6680594044969808,
+      "learning_rate": 5.4296088438663695e-06,
+      "loss": 0.4108,
+      "step": 7025
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.93647469214339,
+      "learning_rate": 5.428486978254667e-06,
+      "loss": 0.4556,
+      "step": 7026
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.850345576980643,
+      "learning_rate": 5.4273650909115024e-06,
+      "loss": 0.3932,
+      "step": 7027
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2425010835710655,
+      "learning_rate": 5.4262431818937736e-06,
+      "loss": 0.4183,
+      "step": 7028
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.0921584581471198,
+      "learning_rate": 5.425121251258381e-06,
+      "loss": 0.3736,
+      "step": 7029
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.144885134649989,
+      "learning_rate": 5.423999299062222e-06,
+      "loss": 0.4126,
+      "step": 7030
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.956711666766488,
+      "learning_rate": 5.422877325362203e-06,
+      "loss": 0.3875,
+      "step": 7031
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9374407774854157,
+      "learning_rate": 5.421755330215223e-06,
+      "loss": 0.4072,
+      "step": 7032
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.999197302784572,
+      "learning_rate": 5.420633313678191e-06,
+      "loss": 0.4179,
+      "step": 7033
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.311937727457086,
+      "learning_rate": 5.4195112758080075e-06,
+      "loss": 0.4435,
+      "step": 7034
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.8004173204984486,
+      "learning_rate": 5.41838921666158e-06,
+      "loss": 0.4255,
+      "step": 7035
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.92519007582553,
+      "learning_rate": 5.417267136295813e-06,
+      "loss": 0.4242,
+      "step": 7036
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.244227732232376,
+      "learning_rate": 5.416145034767618e-06,
+      "loss": 0.4132,
+      "step": 7037
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3364015540788348,
+      "learning_rate": 5.415022912133904e-06,
+      "loss": 0.4327,
+      "step": 7038
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 11.70166053799732,
+      "learning_rate": 5.413900768451582e-06,
+      "loss": 0.406,
+      "step": 7039
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2969646091809697,
+      "learning_rate": 5.412778603777561e-06,
+      "loss": 0.3892,
+      "step": 7040
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1018200804636997,
+      "learning_rate": 5.4116564181687555e-06,
+      "loss": 0.3756,
+      "step": 7041
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.733389952612463,
+      "learning_rate": 5.410534211682078e-06,
+      "loss": 0.4113,
+      "step": 7042
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.450590871301966,
+      "learning_rate": 5.409411984374444e-06,
+      "loss": 0.363,
+      "step": 7043
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.297157886488537,
+      "learning_rate": 5.408289736302769e-06,
+      "loss": 0.3992,
+      "step": 7044
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3962305563856403,
+      "learning_rate": 5.407167467523971e-06,
+      "loss": 0.4284,
+      "step": 7045
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4151970673175662,
+      "learning_rate": 5.406045178094965e-06,
+      "loss": 0.3931,
+      "step": 7046
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.02767630132759,
+      "learning_rate": 5.404922868072673e-06,
+      "loss": 0.4226,
+      "step": 7047
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2176993955222217,
+      "learning_rate": 5.4038005375140105e-06,
+      "loss": 0.4021,
+      "step": 7048
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.997982195880383,
+      "learning_rate": 5.402678186475904e-06,
+      "loss": 0.3885,
+      "step": 7049
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.811292735626706,
+      "learning_rate": 5.40155581501527e-06,
+      "loss": 0.3988,
+      "step": 7050
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.507039350819468,
+      "learning_rate": 5.4004334231890355e-06,
+      "loss": 0.4216,
+      "step": 7051
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.4549550562716282,
+      "learning_rate": 5.399311011054122e-06,
+      "loss": 0.4251,
+      "step": 7052
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.448630107924563,
+      "learning_rate": 5.398188578667458e-06,
+      "loss": 0.4228,
+      "step": 7053
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6960218855488565,
+      "learning_rate": 5.397066126085967e-06,
+      "loss": 0.4218,
+      "step": 7054
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.1842410082249994,
+      "learning_rate": 5.3959436533665765e-06,
+      "loss": 0.4139,
+      "step": 7055
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.111418165072542,
+      "learning_rate": 5.394821160566214e-06,
+      "loss": 0.3891,
+      "step": 7056
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.216529697104171,
+      "learning_rate": 5.39369864774181e-06,
+      "loss": 0.4428,
+      "step": 7057
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.1004660222148415,
+      "learning_rate": 5.392576114950294e-06,
+      "loss": 0.4057,
+      "step": 7058
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 8.403145772585146,
+      "learning_rate": 5.391453562248597e-06,
+      "loss": 0.4362,
+      "step": 7059
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.48991129513098,
+      "learning_rate": 5.390330989693652e-06,
+      "loss": 0.435,
+      "step": 7060
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.419654800765527,
+      "learning_rate": 5.389208397342393e-06,
+      "loss": 0.4141,
+      "step": 7061
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.787527873477,
+      "learning_rate": 5.38808578525175e-06,
+      "loss": 0.4578,
+      "step": 7062
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.112047453592444,
+      "learning_rate": 5.386963153478663e-06,
+      "loss": 0.3992,
+      "step": 7063
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.55319981953245,
+      "learning_rate": 5.3858405020800666e-06,
+      "loss": 0.4023,
+      "step": 7064
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.471059974459809,
+      "learning_rate": 5.384717831112898e-06,
+      "loss": 0.3711,
+      "step": 7065
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.694881376240465,
+      "learning_rate": 5.383595140634093e-06,
+      "loss": 0.4073,
+      "step": 7066
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.150778322757169,
+      "learning_rate": 5.382472430700596e-06,
+      "loss": 0.3866,
+      "step": 7067
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3735124703557379,
+      "learning_rate": 5.381349701369343e-06,
+      "loss": 0.1716,
+      "step": 7068
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.0958749444882105,
+      "learning_rate": 5.380226952697278e-06,
+      "loss": 0.4166,
+      "step": 7069
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.985183537137761,
+      "learning_rate": 5.379104184741338e-06,
+      "loss": 0.383,
+      "step": 7070
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.843196487625502,
+      "learning_rate": 5.377981397558473e-06,
+      "loss": 0.4329,
+      "step": 7071
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3941165242552411,
+      "learning_rate": 5.376858591205622e-06,
+      "loss": 0.2022,
+      "step": 7072
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.2491985831529173,
+      "learning_rate": 5.375735765739733e-06,
+      "loss": 0.4279,
+      "step": 7073
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.037416572205365,
+      "learning_rate": 5.374612921217749e-06,
+      "loss": 0.3923,
+      "step": 7074
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.856288051039682,
+      "learning_rate": 5.373490057696621e-06,
+      "loss": 0.4037,
+      "step": 7075
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9474261897045198,
+      "learning_rate": 5.372367175233293e-06,
+      "loss": 0.3734,
+      "step": 7076
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.108419739120933,
+      "learning_rate": 5.371244273884718e-06,
+      "loss": 0.3915,
+      "step": 7077
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.648717532142566,
+      "learning_rate": 5.370121353707842e-06,
+      "loss": 0.3647,
+      "step": 7078
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.90911040870757,
+      "learning_rate": 5.3689984147596184e-06,
+      "loss": 0.3971,
+      "step": 7079
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9263177683060437,
+      "learning_rate": 5.367875457096999e-06,
+      "loss": 0.4089,
+      "step": 7080
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.162051304797106,
+      "learning_rate": 5.3667524807769355e-06,
+      "loss": 0.3881,
+      "step": 7081
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.231169337099022,
+      "learning_rate": 5.365629485856381e-06,
+      "loss": 0.4213,
+      "step": 7082
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.825834482130496,
+      "learning_rate": 5.364506472392291e-06,
+      "loss": 0.3904,
+      "step": 7083
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.5601315398970876,
+      "learning_rate": 5.36338344044162e-06,
+      "loss": 0.3791,
+      "step": 7084
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.465779682728074,
+      "learning_rate": 5.362260390061329e-06,
+      "loss": 0.3683,
+      "step": 7085
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.257040025069122,
+      "learning_rate": 5.3611373213083695e-06,
+      "loss": 0.3764,
+      "step": 7086
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.086769031672178,
+      "learning_rate": 5.360014234239704e-06,
+      "loss": 0.4249,
+      "step": 7087
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.017619881027401,
+      "learning_rate": 5.358891128912288e-06,
+      "loss": 0.4494,
+      "step": 7088
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 31.521598249649774,
+      "learning_rate": 5.357768005383085e-06,
+      "loss": 0.3945,
+      "step": 7089
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.039688254237937,
+      "learning_rate": 5.356644863709055e-06,
+      "loss": 0.3945,
+      "step": 7090
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.650328241285597,
+      "learning_rate": 5.355521703947161e-06,
+      "loss": 0.4473,
+      "step": 7091
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.388771152585652,
+      "learning_rate": 5.354398526154365e-06,
+      "loss": 0.384,
+      "step": 7092
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.047426163639351,
+      "learning_rate": 5.35327533038763e-06,
+      "loss": 0.3787,
+      "step": 7093
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.164558965957791,
+      "learning_rate": 5.352152116703925e-06,
+      "loss": 0.4146,
+      "step": 7094
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.8914146930528934,
+      "learning_rate": 5.35102888516021e-06,
+      "loss": 0.4243,
+      "step": 7095
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.855779704793778,
+      "learning_rate": 5.349905635813455e-06,
+      "loss": 0.4583,
+      "step": 7096
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9422536886081114,
+      "learning_rate": 5.348782368720627e-06,
+      "loss": 0.4538,
+      "step": 7097
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5046189008261726,
+      "learning_rate": 5.347659083938694e-06,
+      "loss": 0.3695,
+      "step": 7098
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 10.111849704515523,
+      "learning_rate": 5.346535781524626e-06,
+      "loss": 0.3994,
+      "step": 7099
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.223500768194598,
+      "learning_rate": 5.345412461535393e-06,
+      "loss": 0.4163,
+      "step": 7100
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9779898974008367,
+      "learning_rate": 5.344289124027967e-06,
+      "loss": 0.402,
+      "step": 7101
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.38364485929617836,
+      "learning_rate": 5.343165769059318e-06,
+      "loss": 0.1991,
+      "step": 7102
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.160269919154664,
+      "learning_rate": 5.34204239668642e-06,
+      "loss": 0.3813,
+      "step": 7103
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.5597356350344485,
+      "learning_rate": 5.3409190069662474e-06,
+      "loss": 0.4372,
+      "step": 7104
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 7.248140551076108,
+      "learning_rate": 5.339795599955772e-06,
+      "loss": 0.4106,
+      "step": 7105
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7860394460296645,
+      "learning_rate": 5.338672175711975e-06,
+      "loss": 0.3899,
+      "step": 7106
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.439062005107278,
+      "learning_rate": 5.337548734291827e-06,
+      "loss": 0.414,
+      "step": 7107
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.322386468779333,
+      "learning_rate": 5.336425275752308e-06,
+      "loss": 0.3874,
+      "step": 7108
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 6.003537997398053,
+      "learning_rate": 5.335301800150397e-06,
+      "loss": 0.426,
+      "step": 7109
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3120031659850784,
+      "learning_rate": 5.334178307543071e-06,
+      "loss": 0.4136,
+      "step": 7110
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.8262140064632497,
+      "learning_rate": 5.333054797987309e-06,
+      "loss": 0.4222,
+      "step": 7111
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 4.341591992390665,
+      "learning_rate": 5.331931271540096e-06,
+      "loss": 0.4111,
+      "step": 7112
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.057509944683106,
+      "learning_rate": 5.330807728258411e-06,
+      "loss": 0.4129,
+      "step": 7113
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.9417598439531787,
+      "learning_rate": 5.329684168199236e-06,
+      "loss": 0.4075,
+      "step": 7114
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.882962049801633,
+      "learning_rate": 5.328560591419554e-06,
+      "loss": 0.4021,
+      "step": 7115
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.425249953082573,
+      "learning_rate": 5.3274369979763514e-06,
+      "loss": 0.4343,
+      "step": 7116
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.905765819941617,
+      "learning_rate": 5.3263133879266105e-06,
+      "loss": 0.4239,
+      "step": 7117
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3758141619030323,
+      "learning_rate": 5.32518976132732e-06,
+      "loss": 0.2002,
+      "step": 7118
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.905954465117142,
+      "learning_rate": 5.324066118235462e-06,
+      "loss": 0.4073,
+      "step": 7119
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.399157366830174,
+      "learning_rate": 5.322942458708029e-06,
+      "loss": 0.3915,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8204600805950917,
+      "learning_rate": 5.321818782802007e-06,
+      "loss": 0.401,
+      "step": 7121
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.672908790915483,
+      "learning_rate": 5.320695090574386e-06,
+      "loss": 0.4298,
+      "step": 7122
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.226310520283435,
+      "learning_rate": 5.319571382082156e-06,
+      "loss": 0.4386,
+      "step": 7123
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.112014977345922,
+      "learning_rate": 5.318447657382306e-06,
+      "loss": 0.4123,
+      "step": 7124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.487399661921639,
+      "learning_rate": 5.317323916531831e-06,
+      "loss": 0.4216,
+      "step": 7125
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 16.311060175394534,
+      "learning_rate": 5.31620015958772e-06,
+      "loss": 0.403,
+      "step": 7126
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6812415691350893,
+      "learning_rate": 5.315076386606967e-06,
+      "loss": 0.3849,
+      "step": 7127
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.60808065532023,
+      "learning_rate": 5.3139525976465675e-06,
+      "loss": 0.3583,
+      "step": 7128
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.9256720623807344,
+      "learning_rate": 5.312828792763516e-06,
+      "loss": 0.3936,
+      "step": 7129
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9160700965105435,
+      "learning_rate": 5.311704972014808e-06,
+      "loss": 0.4475,
+      "step": 7130
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.995187234783653,
+      "learning_rate": 5.31058113545744e-06,
+      "loss": 0.3974,
+      "step": 7131
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6942698858783243,
+      "learning_rate": 5.30945728314841e-06,
+      "loss": 0.4217,
+      "step": 7132
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.050247533706259,
+      "learning_rate": 5.308333415144714e-06,
+      "loss": 0.4209,
+      "step": 7133
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.502460628568222,
+      "learning_rate": 5.307209531503354e-06,
+      "loss": 0.3836,
+      "step": 7134
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.558065484142996,
+      "learning_rate": 5.306085632281328e-06,
+      "loss": 0.4436,
+      "step": 7135
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.543890517331624,
+      "learning_rate": 5.304961717535636e-06,
+      "loss": 0.3902,
+      "step": 7136
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 29.344670091309148,
+      "learning_rate": 5.30383778732328e-06,
+      "loss": 0.4157,
+      "step": 7137
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.514175917093063,
+      "learning_rate": 5.302713841701263e-06,
+      "loss": 0.4055,
+      "step": 7138
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.470198783788359,
+      "learning_rate": 5.301589880726584e-06,
+      "loss": 0.4368,
+      "step": 7139
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.751462336022052,
+      "learning_rate": 5.300465904456252e-06,
+      "loss": 0.4296,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8464541971176627,
+      "learning_rate": 5.299341912947268e-06,
+      "loss": 0.4295,
+      "step": 7141
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.0809903575866,
+      "learning_rate": 5.298217906256638e-06,
+      "loss": 0.4022,
+      "step": 7142
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.730531308125624,
+      "learning_rate": 5.2970938844413684e-06,
+      "loss": 0.3747,
+      "step": 7143
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.3210312630076237,
+      "learning_rate": 5.295969847558467e-06,
+      "loss": 0.4165,
+      "step": 7144
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.823188577511101,
+      "learning_rate": 5.294845795664937e-06,
+      "loss": 0.3763,
+      "step": 7145
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1432769405027092,
+      "learning_rate": 5.2937217288177915e-06,
+      "loss": 0.4733,
+      "step": 7146
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6426556782716766,
+      "learning_rate": 5.292597647074037e-06,
+      "loss": 0.3981,
+      "step": 7147
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.881508231084036,
+      "learning_rate": 5.2914735504906845e-06,
+      "loss": 0.4325,
+      "step": 7148
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8273520128084364,
+      "learning_rate": 5.290349439124742e-06,
+      "loss": 0.3805,
+      "step": 7149
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.779928059770307,
+      "learning_rate": 5.289225313033224e-06,
+      "loss": 0.4079,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.254595211988487,
+      "learning_rate": 5.28810117227314e-06,
+      "loss": 0.4116,
+      "step": 7151
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2869717488551022,
+      "learning_rate": 5.286977016901503e-06,
+      "loss": 0.4628,
+      "step": 7152
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.794676866283558,
+      "learning_rate": 5.28585284697533e-06,
+      "loss": 0.4246,
+      "step": 7153
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.2490888578296766,
+      "learning_rate": 5.284728662551631e-06,
+      "loss": 0.425,
+      "step": 7154
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.702009226843884,
+      "learning_rate": 5.283604463687423e-06,
+      "loss": 0.4031,
+      "step": 7155
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.154430622164074,
+      "learning_rate": 5.282480250439721e-06,
+      "loss": 0.4486,
+      "step": 7156
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.829156812937423,
+      "learning_rate": 5.281356022865542e-06,
+      "loss": 0.4553,
+      "step": 7157
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.163437121900888,
+      "learning_rate": 5.2802317810219016e-06,
+      "loss": 0.4274,
+      "step": 7158
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.11153372997178,
+      "learning_rate": 5.27910752496582e-06,
+      "loss": 0.4053,
+      "step": 7159
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.189996614283321,
+      "learning_rate": 5.277983254754314e-06,
+      "loss": 0.4232,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.781428653653414,
+      "learning_rate": 5.276858970444404e-06,
+      "loss": 0.4134,
+      "step": 7161
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.145277208723174,
+      "learning_rate": 5.275734672093109e-06,
+      "loss": 0.3753,
+      "step": 7162
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.749009739939943,
+      "learning_rate": 5.274610359757452e-06,
+      "loss": 0.3923,
+      "step": 7163
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.726798549980412,
+      "learning_rate": 5.2734860334944504e-06,
+      "loss": 0.4278,
+      "step": 7164
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.192408239294352,
+      "learning_rate": 5.272361693361131e-06,
+      "loss": 0.421,
+      "step": 7165
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.503996088993601,
+      "learning_rate": 5.271237339414512e-06,
+      "loss": 0.3614,
+      "step": 7166
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3511764440921341,
+      "learning_rate": 5.2701129717116215e-06,
+      "loss": 0.1832,
+      "step": 7167
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8543368732419743,
+      "learning_rate": 5.2689885903094815e-06,
+      "loss": 0.4027,
+      "step": 7168
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.138086772877543,
+      "learning_rate": 5.267864195265118e-06,
+      "loss": 0.3931,
+      "step": 7169
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.224804449517897,
+      "learning_rate": 5.266739786635554e-06,
+      "loss": 0.4489,
+      "step": 7170
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.128056953359973,
+      "learning_rate": 5.265615364477818e-06,
+      "loss": 0.4398,
+      "step": 7171
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.488720467975089,
+      "learning_rate": 5.2644909288489365e-06,
+      "loss": 0.3338,
+      "step": 7172
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.434210629587866,
+      "learning_rate": 5.2633664798059395e-06,
+      "loss": 0.4155,
+      "step": 7173
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.990322092913884,
+      "learning_rate": 5.262242017405851e-06,
+      "loss": 0.4501,
+      "step": 7174
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.365856083687846,
+      "learning_rate": 5.2611175417057024e-06,
+      "loss": 0.3844,
+      "step": 7175
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.599310529582345,
+      "learning_rate": 5.259993052762523e-06,
+      "loss": 0.3966,
+      "step": 7176
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.649856521304783,
+      "learning_rate": 5.258868550633346e-06,
+      "loss": 0.3849,
+      "step": 7177
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8687410834316514,
+      "learning_rate": 5.257744035375198e-06,
+      "loss": 0.4021,
+      "step": 7178
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.479411446604593,
+      "learning_rate": 5.256619507045113e-06,
+      "loss": 0.3716,
+      "step": 7179
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.626042124487039,
+      "learning_rate": 5.2554949657001246e-06,
+      "loss": 0.4276,
+      "step": 7180
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.109786728257055,
+      "learning_rate": 5.254370411397264e-06,
+      "loss": 0.3632,
+      "step": 7181
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.423195779070191,
+      "learning_rate": 5.253245844193564e-06,
+      "loss": 0.3975,
+      "step": 7182
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.956758535024957,
+      "learning_rate": 5.252121264146062e-06,
+      "loss": 0.3778,
+      "step": 7183
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.728568256678654,
+      "learning_rate": 5.25099667131179e-06,
+      "loss": 0.3774,
+      "step": 7184
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.7689555528172574,
+      "learning_rate": 5.249872065747786e-06,
+      "loss": 0.3972,
+      "step": 7185
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.921617642630527,
+      "learning_rate": 5.248747447511085e-06,
+      "loss": 0.4061,
+      "step": 7186
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.453180583454455,
+      "learning_rate": 5.247622816658726e-06,
+      "loss": 0.4057,
+      "step": 7187
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.1163669097372795,
+      "learning_rate": 5.2464981732477424e-06,
+      "loss": 0.4346,
+      "step": 7188
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.164688569590988,
+      "learning_rate": 5.245373517335177e-06,
+      "loss": 0.3994,
+      "step": 7189
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.6993132761832195,
+      "learning_rate": 5.244248848978067e-06,
+      "loss": 0.3756,
+      "step": 7190
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.046595873588387,
+      "learning_rate": 5.243124168233452e-06,
+      "loss": 0.4252,
+      "step": 7191
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.72908631659617,
+      "learning_rate": 5.2419994751583704e-06,
+      "loss": 0.3963,
+      "step": 7192
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 12.146380782111512,
+      "learning_rate": 5.240874769809865e-06,
+      "loss": 0.3895,
+      "step": 7193
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.086863825696362,
+      "learning_rate": 5.239750052244975e-06,
+      "loss": 0.3891,
+      "step": 7194
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.8191944961945286,
+      "learning_rate": 5.238625322520746e-06,
+      "loss": 0.404,
+      "step": 7195
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.39543829286075,
+      "learning_rate": 5.237500580694218e-06,
+      "loss": 0.4055,
+      "step": 7196
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40382112169688017,
+      "learning_rate": 5.236375826822435e-06,
+      "loss": 0.1923,
+      "step": 7197
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.021043260550525,
+      "learning_rate": 5.23525106096244e-06,
+      "loss": 0.4484,
+      "step": 7198
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36782413032016376,
+      "learning_rate": 5.2341262831712795e-06,
+      "loss": 0.2011,
+      "step": 7199
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.424179879007857,
+      "learning_rate": 5.2330014935059945e-06,
+      "loss": 0.3941,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.1955555068448875,
+      "learning_rate": 5.231876692023636e-06,
+      "loss": 0.3995,
+      "step": 7201
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.579044187129049,
+      "learning_rate": 5.230751878781247e-06,
+      "loss": 0.3992,
+      "step": 7202
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.271481633633948,
+      "learning_rate": 5.229627053835876e-06,
+      "loss": 0.3871,
+      "step": 7203
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.881726700312318,
+      "learning_rate": 5.228502217244568e-06,
+      "loss": 0.4197,
+      "step": 7204
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.247432196486142,
+      "learning_rate": 5.2273773690643725e-06,
+      "loss": 0.4284,
+      "step": 7205
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.283933249407613,
+      "learning_rate": 5.2262525093523384e-06,
+      "loss": 0.4011,
+      "step": 7206
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.194731608160995,
+      "learning_rate": 5.225127638165514e-06,
+      "loss": 0.4148,
+      "step": 7207
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.907008536190267,
+      "learning_rate": 5.22400275556095e-06,
+      "loss": 0.4011,
+      "step": 7208
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.431332748554083,
+      "learning_rate": 5.222877861595698e-06,
+      "loss": 0.4011,
+      "step": 7209
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.722401225915565,
+      "learning_rate": 5.221752956326806e-06,
+      "loss": 0.4169,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 9.221995865382286,
+      "learning_rate": 5.220628039811328e-06,
+      "loss": 0.4036,
+      "step": 7211
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.870737334635783,
+      "learning_rate": 5.2195031121063145e-06,
+      "loss": 0.4364,
+      "step": 7212
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 13.730863763651284,
+      "learning_rate": 5.218378173268819e-06,
+      "loss": 0.4268,
+      "step": 7213
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.067676131979945,
+      "learning_rate": 5.217253223355895e-06,
+      "loss": 0.4017,
+      "step": 7214
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.54698108571447,
+      "learning_rate": 5.216128262424596e-06,
+      "loss": 0.3955,
+      "step": 7215
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 15.072741502270535,
+      "learning_rate": 5.215003290531974e-06,
+      "loss": 0.4223,
+      "step": 7216
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 10.302948920520503,
+      "learning_rate": 5.213878307735088e-06,
+      "loss": 0.4258,
+      "step": 7217
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.37517109652005404,
+      "learning_rate": 5.21275331409099e-06,
+      "loss": 0.197,
+      "step": 7218
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.388222305163791,
+      "learning_rate": 5.211628309656738e-06,
+      "loss": 0.4212,
+      "step": 7219
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.359296152699776,
+      "learning_rate": 5.210503294489388e-06,
+      "loss": 0.4016,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.508184442123071,
+      "learning_rate": 5.209378268645998e-06,
+      "loss": 0.4058,
+      "step": 7221
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6019889165597103,
+      "learning_rate": 5.208253232183625e-06,
+      "loss": 0.4017,
+      "step": 7222
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.321510854519537,
+      "learning_rate": 5.207128185159327e-06,
+      "loss": 0.4291,
+      "step": 7223
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.088895406098018,
+      "learning_rate": 5.206003127630163e-06,
+      "loss": 0.4082,
+      "step": 7224
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.715715006677683,
+      "learning_rate": 5.20487805965319e-06,
+      "loss": 0.4544,
+      "step": 7225
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6549677578854216,
+      "learning_rate": 5.203752981285472e-06,
+      "loss": 0.3821,
+      "step": 7226
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.8689142564346892,
+      "learning_rate": 5.2026278925840656e-06,
+      "loss": 0.3976,
+      "step": 7227
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.714702366463178,
+      "learning_rate": 5.201502793606034e-06,
+      "loss": 0.4056,
+      "step": 7228
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.688726877209641,
+      "learning_rate": 5.200377684408436e-06,
+      "loss": 0.4175,
+      "step": 7229
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.690974575439662,
+      "learning_rate": 5.199252565048337e-06,
+      "loss": 0.3743,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9393235997867584,
+      "learning_rate": 5.198127435582796e-06,
+      "loss": 0.3934,
+      "step": 7231
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.439349464142422,
+      "learning_rate": 5.197002296068878e-06,
+      "loss": 0.3915,
+      "step": 7232
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.667926639687773,
+      "learning_rate": 5.195877146563646e-06,
+      "loss": 0.4329,
+      "step": 7233
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.117569895303239,
+      "learning_rate": 5.194751987124165e-06,
+      "loss": 0.4121,
+      "step": 7234
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.562615524403924,
+      "learning_rate": 5.193626817807497e-06,
+      "loss": 0.3816,
+      "step": 7235
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.839690240567814,
+      "learning_rate": 5.192501638670709e-06,
+      "loss": 0.4013,
+      "step": 7236
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.637798414309839,
+      "learning_rate": 5.191376449770863e-06,
+      "loss": 0.4274,
+      "step": 7237
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.132251699957092,
+      "learning_rate": 5.1902512511650295e-06,
+      "loss": 0.4433,
+      "step": 7238
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.883196157430117,
+      "learning_rate": 5.189126042910271e-06,
+      "loss": 0.3984,
+      "step": 7239
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.112584048238779,
+      "learning_rate": 5.188000825063658e-06,
+      "loss": 0.3811,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.543727429651387,
+      "learning_rate": 5.186875597682255e-06,
+      "loss": 0.3959,
+      "step": 7241
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.36883389126359883,
+      "learning_rate": 5.18575036082313e-06,
+      "loss": 0.1904,
+      "step": 7242
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.40310187829899363,
+      "learning_rate": 5.184625114543354e-06,
+      "loss": 0.1854,
+      "step": 7243
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.530875163517801,
+      "learning_rate": 5.183499858899994e-06,
+      "loss": 0.4094,
+      "step": 7244
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.430646943367424,
+      "learning_rate": 5.182374593950119e-06,
+      "loss": 0.4329,
+      "step": 7245
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.846939423052928,
+      "learning_rate": 5.181249319750799e-06,
+      "loss": 0.3896,
+      "step": 7246
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.093239162523645,
+      "learning_rate": 5.180124036359106e-06,
+      "loss": 0.3983,
+      "step": 7247
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.795169139676427,
+      "learning_rate": 5.1789987438321075e-06,
+      "loss": 0.4154,
+      "step": 7248
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.9891159653262944,
+      "learning_rate": 5.177873442226876e-06,
+      "loss": 0.4228,
+      "step": 7249
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.9153223267643655,
+      "learning_rate": 5.176748131600484e-06,
+      "loss": 0.4164,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 19.010577288695426,
+      "learning_rate": 5.175622812010003e-06,
+      "loss": 0.3955,
+      "step": 7251
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.0645461436932155,
+      "learning_rate": 5.174497483512506e-06,
+      "loss": 0.3987,
+      "step": 7252
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.495700499628674,
+      "learning_rate": 5.1733721461650646e-06,
+      "loss": 0.382,
+      "step": 7253
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.271172645171455,
+      "learning_rate": 5.172246800024754e-06,
+      "loss": 0.3961,
+      "step": 7254
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.874066606311963,
+      "learning_rate": 5.171121445148646e-06,
+      "loss": 0.4028,
+      "step": 7255
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.366431289519165,
+      "learning_rate": 5.169996081593818e-06,
+      "loss": 0.4249,
+      "step": 7256
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.695726648069972,
+      "learning_rate": 5.168870709417342e-06,
+      "loss": 0.3849,
+      "step": 7257
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.063052653700488,
+      "learning_rate": 5.167745328676295e-06,
+      "loss": 0.3883,
+      "step": 7258
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.513233908394572,
+      "learning_rate": 5.166619939427749e-06,
+      "loss": 0.3576,
+      "step": 7259
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.6745735541340316,
+      "learning_rate": 5.165494541728786e-06,
+      "loss": 0.3414,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.1458102936036063,
+      "learning_rate": 5.164369135636478e-06,
+      "loss": 0.3677,
+      "step": 7261
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.060422826255353,
+      "learning_rate": 5.163243721207903e-06,
+      "loss": 0.4005,
+      "step": 7262
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.883017190005213,
+      "learning_rate": 5.162118298500139e-06,
+      "loss": 0.4166,
+      "step": 7263
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.8688947708631325,
+      "learning_rate": 5.160992867570265e-06,
+      "loss": 0.3713,
+      "step": 7264
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.085268518034481,
+      "learning_rate": 5.159867428475354e-06,
+      "loss": 0.391,
+      "step": 7265
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.582887873469438,
+      "learning_rate": 5.158741981272493e-06,
+      "loss": 0.4233,
+      "step": 7266
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.760002786603657,
+      "learning_rate": 5.157616526018753e-06,
+      "loss": 0.3998,
+      "step": 7267
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.0229169819977075,
+      "learning_rate": 5.156491062771217e-06,
+      "loss": 0.4101,
+      "step": 7268
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8755928505560906,
+      "learning_rate": 5.155365591586966e-06,
+      "loss": 0.408,
+      "step": 7269
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.3218972215853615,
+      "learning_rate": 5.154240112523078e-06,
+      "loss": 0.3971,
+      "step": 7270
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.962412507687046,
+      "learning_rate": 5.153114625636632e-06,
+      "loss": 0.4063,
+      "step": 7271
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.628467530791254,
+      "learning_rate": 5.151989130984715e-06,
+      "loss": 0.4046,
+      "step": 7272
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.42382298908292604,
+      "learning_rate": 5.150863628624402e-06,
+      "loss": 0.2011,
+      "step": 7273
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.995805087398833,
+      "learning_rate": 5.14973811861278e-06,
+      "loss": 0.416,
+      "step": 7274
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.497536274618485,
+      "learning_rate": 5.148612601006926e-06,
+      "loss": 0.43,
+      "step": 7275
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.465332771488364,
+      "learning_rate": 5.147487075863927e-06,
+      "loss": 0.4596,
+      "step": 7276
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8856278501459998,
+      "learning_rate": 5.146361543240864e-06,
+      "loss": 0.4222,
+      "step": 7277
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.560523705431943,
+      "learning_rate": 5.145236003194822e-06,
+      "loss": 0.3958,
+      "step": 7278
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.25487097275093,
+      "learning_rate": 5.144110455782881e-06,
+      "loss": 0.3979,
+      "step": 7279
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.114734302533178,
+      "learning_rate": 5.142984901062131e-06,
+      "loss": 0.3876,
+      "step": 7280
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.447633113424327,
+      "learning_rate": 5.14185933908965e-06,
+      "loss": 0.4147,
+      "step": 7281
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6510681997527246,
+      "learning_rate": 5.140733769922525e-06,
+      "loss": 0.4266,
+      "step": 7282
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6996713241679666,
+      "learning_rate": 5.139608193617846e-06,
+      "loss": 0.4468,
+      "step": 7283
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3594141407032057,
+      "learning_rate": 5.138482610232691e-06,
+      "loss": 0.3548,
+      "step": 7284
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.618440857441257,
+      "learning_rate": 5.1373570198241505e-06,
+      "loss": 0.3925,
+      "step": 7285
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.842421250470998,
+      "learning_rate": 5.13623142244931e-06,
+      "loss": 0.4201,
+      "step": 7286
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.128499472865874,
+      "learning_rate": 5.135105818165256e-06,
+      "loss": 0.4045,
+      "step": 7287
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4687916988679537,
+      "learning_rate": 5.133980207029075e-06,
+      "loss": 0.4042,
+      "step": 7288
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.011726105319139,
+      "learning_rate": 5.132854589097855e-06,
+      "loss": 0.3905,
+      "step": 7289
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 14.860291637060351,
+      "learning_rate": 5.131728964428684e-06,
+      "loss": 0.3658,
+      "step": 7290
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.612482519569758,
+      "learning_rate": 5.13060333307865e-06,
+      "loss": 0.4008,
+      "step": 7291
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.514050483238945,
+      "learning_rate": 5.129477695104839e-06,
+      "loss": 0.4358,
+      "step": 7292
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.851684835025362,
+      "learning_rate": 5.128352050564341e-06,
+      "loss": 0.4227,
+      "step": 7293
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5638177252725054,
+      "learning_rate": 5.127226399514246e-06,
+      "loss": 0.4035,
+      "step": 7294
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.803519109505555,
+      "learning_rate": 5.126100742011644e-06,
+      "loss": 0.4026,
+      "step": 7295
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.41539124936252675,
+      "learning_rate": 5.124975078113622e-06,
+      "loss": 0.1988,
+      "step": 7296
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.522100990473304,
+      "learning_rate": 5.123849407877273e-06,
+      "loss": 0.4152,
+      "step": 7297
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.820158044349714,
+      "learning_rate": 5.122723731359683e-06,
+      "loss": 0.3989,
+      "step": 7298
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.966247026700216,
+      "learning_rate": 5.1215980486179486e-06,
+      "loss": 0.3992,
+      "step": 7299
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.574117085852362,
+      "learning_rate": 5.120472359709155e-06,
+      "loss": 0.4044,
+      "step": 7300
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.643704161753959,
+      "learning_rate": 5.119346664690398e-06,
+      "loss": 0.3988,
+      "step": 7301
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2490095522584306,
+      "learning_rate": 5.118220963618767e-06,
+      "loss": 0.3701,
+      "step": 7302
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.5186561947822,
+      "learning_rate": 5.117095256551354e-06,
+      "loss": 0.3858,
+      "step": 7303
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.988369058286498,
+      "learning_rate": 5.115969543545252e-06,
+      "loss": 0.406,
+      "step": 7304
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.89246588088019,
+      "learning_rate": 5.114843824657552e-06,
+      "loss": 0.4119,
+      "step": 7305
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.758701471724572,
+      "learning_rate": 5.113718099945346e-06,
+      "loss": 0.3867,
+      "step": 7306
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.970531595155301,
+      "learning_rate": 5.112592369465731e-06,
+      "loss": 0.4586,
+      "step": 7307
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.641926581945703,
+      "learning_rate": 5.111466633275797e-06,
+      "loss": 0.4182,
+      "step": 7308
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.513425938502991,
+      "learning_rate": 5.11034089143264e-06,
+      "loss": 0.461,
+      "step": 7309
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.316502674391709,
+      "learning_rate": 5.109215143993352e-06,
+      "loss": 0.382,
+      "step": 7310
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.1647424943279425,
+      "learning_rate": 5.108089391015028e-06,
+      "loss": 0.4224,
+      "step": 7311
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.136237052890415,
+      "learning_rate": 5.106963632554762e-06,
+      "loss": 0.3527,
+      "step": 7312
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0704452252757086,
+      "learning_rate": 5.10583786866965e-06,
+      "loss": 0.393,
+      "step": 7313
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0106171002770648,
+      "learning_rate": 5.1047120994167855e-06,
+      "loss": 0.3708,
+      "step": 7314
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.578455151852747,
+      "learning_rate": 5.103586324853266e-06,
+      "loss": 0.4291,
+      "step": 7315
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1047272879424552,
+      "learning_rate": 5.102460545036184e-06,
+      "loss": 0.4356,
+      "step": 7316
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 10.01552023124748,
+      "learning_rate": 5.101334760022639e-06,
+      "loss": 0.3903,
+      "step": 7317
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.162537843409845,
+      "learning_rate": 5.100208969869723e-06,
+      "loss": 0.4548,
+      "step": 7318
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.50680962887044,
+      "learning_rate": 5.099083174634537e-06,
+      "loss": 0.3918,
+      "step": 7319
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.732178504304858,
+      "learning_rate": 5.097957374374174e-06,
+      "loss": 0.4128,
+      "step": 7320
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.983561424079755,
+      "learning_rate": 5.096831569145733e-06,
+      "loss": 0.4179,
+      "step": 7321
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 18.709698065370677,
+      "learning_rate": 5.095705759006311e-06,
+      "loss": 0.3905,
+      "step": 7322
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 9.627679235721232,
+      "learning_rate": 5.094579944013005e-06,
+      "loss": 0.4029,
+      "step": 7323
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.269198158788132,
+      "learning_rate": 5.093454124222912e-06,
+      "loss": 0.4003,
+      "step": 7324
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.260145748584352,
+      "learning_rate": 5.092328299693131e-06,
+      "loss": 0.3902,
+      "step": 7325
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.121176204979371,
+      "learning_rate": 5.0912024704807585e-06,
+      "loss": 0.4506,
+      "step": 7326
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.292169489065743,
+      "learning_rate": 5.090076636642896e-06,
+      "loss": 0.3921,
+      "step": 7327
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7444612401463693,
+      "learning_rate": 5.0889507982366385e-06,
+      "loss": 0.3772,
+      "step": 7328
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8959354767968364,
+      "learning_rate": 5.0878249553190864e-06,
+      "loss": 0.4024,
+      "step": 7329
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6506880409581997,
+      "learning_rate": 5.08669910794734e-06,
+      "loss": 0.3987,
+      "step": 7330
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.0807274898122765,
+      "learning_rate": 5.0855732561784966e-06,
+      "loss": 0.4154,
+      "step": 7331
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.801754866944131,
+      "learning_rate": 5.084447400069656e-06,
+      "loss": 0.4285,
+      "step": 7332
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.770258462210654,
+      "learning_rate": 5.08332153967792e-06,
+      "loss": 0.4174,
+      "step": 7333
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.347354585608972,
+      "learning_rate": 5.0821956750603866e-06,
+      "loss": 0.4201,
+      "step": 7334
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.611283419969454,
+      "learning_rate": 5.081069806274156e-06,
+      "loss": 0.3768,
+      "step": 7335
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.397681631308397,
+      "learning_rate": 5.07994393337633e-06,
+      "loss": 0.4389,
+      "step": 7336
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.214738761449275,
+      "learning_rate": 5.0788180564240084e-06,
+      "loss": 0.3916,
+      "step": 7337
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 12.39511174770064,
+      "learning_rate": 5.077692175474292e-06,
+      "loss": 0.4185,
+      "step": 7338
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.309829088576712,
+      "learning_rate": 5.076566290584282e-06,
+      "loss": 0.4169,
+      "step": 7339
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.971952847475386,
+      "learning_rate": 5.075440401811077e-06,
+      "loss": 0.387,
+      "step": 7340
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.1361124335259047,
+      "learning_rate": 5.074314509211783e-06,
+      "loss": 0.3779,
+      "step": 7341
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.737198267895121,
+      "learning_rate": 5.073188612843498e-06,
+      "loss": 0.3781,
+      "step": 7342
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.0985423129872647,
+      "learning_rate": 5.072062712763327e-06,
+      "loss": 0.3974,
+      "step": 7343
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.813760213806243,
+      "learning_rate": 5.070936809028368e-06,
+      "loss": 0.4235,
+      "step": 7344
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.84349165972975,
+      "learning_rate": 5.069810901695727e-06,
+      "loss": 0.4149,
+      "step": 7345
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.521633422233609,
+      "learning_rate": 5.068684990822505e-06,
+      "loss": 0.3617,
+      "step": 7346
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 28.1932748485103,
+      "learning_rate": 5.067559076465803e-06,
+      "loss": 0.4065,
+      "step": 7347
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.508388008643048,
+      "learning_rate": 5.066433158682725e-06,
+      "loss": 0.4168,
+      "step": 7348
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.40761688028910675,
+      "learning_rate": 5.065307237530373e-06,
+      "loss": 0.2027,
+      "step": 7349
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.910407495884714,
+      "learning_rate": 5.064181313065853e-06,
+      "loss": 0.4144,
+      "step": 7350
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.838654817157729,
+      "learning_rate": 5.063055385346264e-06,
+      "loss": 0.3688,
+      "step": 7351
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.686721228335012,
+      "learning_rate": 5.061929454428712e-06,
+      "loss": 0.3924,
+      "step": 7352
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.440523607055384,
+      "learning_rate": 5.0608035203703e-06,
+      "loss": 0.3891,
+      "step": 7353
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.636285263625897,
+      "learning_rate": 5.059677583228133e-06,
+      "loss": 0.4087,
+      "step": 7354
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3762324933252175,
+      "learning_rate": 5.058551643059312e-06,
+      "loss": 0.4068,
+      "step": 7355
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 7.877300275264105,
+      "learning_rate": 5.057425699920943e-06,
+      "loss": 0.3996,
+      "step": 7356
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.5626329420070597,
+      "learning_rate": 5.0562997538701295e-06,
+      "loss": 0.401,
+      "step": 7357
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.265634573352871,
+      "learning_rate": 5.0551738049639774e-06,
+      "loss": 0.3716,
+      "step": 7358
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.202343593804817,
+      "learning_rate": 5.0540478532595886e-06,
+      "loss": 0.4125,
+      "step": 7359
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.3370831594562915,
+      "learning_rate": 5.052921898814068e-06,
+      "loss": 0.4106,
+      "step": 7360
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.170838464680334,
+      "learning_rate": 5.051795941684522e-06,
+      "loss": 0.4085,
+      "step": 7361
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.79040192828889,
+      "learning_rate": 5.050669981928056e-06,
+      "loss": 0.4508,
+      "step": 7362
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3509040576227096,
+      "learning_rate": 5.049544019601772e-06,
+      "loss": 0.3989,
+      "step": 7363
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.121376302255741,
+      "learning_rate": 5.048418054762777e-06,
+      "loss": 0.4102,
+      "step": 7364
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8368938133715362,
+      "learning_rate": 5.047292087468176e-06,
+      "loss": 0.4315,
+      "step": 7365
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3858422063134753,
+      "learning_rate": 5.046166117775076e-06,
+      "loss": 0.1968,
+      "step": 7366
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.34396615971591,
+      "learning_rate": 5.04504014574058e-06,
+      "loss": 0.384,
+      "step": 7367
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.912865212914863,
+      "learning_rate": 5.043914171421794e-06,
+      "loss": 0.416,
+      "step": 7368
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.843493161925342,
+      "learning_rate": 5.0427881948758265e-06,
+      "loss": 0.3961,
+      "step": 7369
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.616491427907568,
+      "learning_rate": 5.04166221615978e-06,
+      "loss": 0.4093,
+      "step": 7370
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.528492965693116,
+      "learning_rate": 5.040536235330761e-06,
+      "loss": 0.4303,
+      "step": 7371
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.518865574088145,
+      "learning_rate": 5.039410252445878e-06,
+      "loss": 0.4226,
+      "step": 7372
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.6941736578034248,
+      "learning_rate": 5.038284267562234e-06,
+      "loss": 0.4203,
+      "step": 7373
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.078408660301972,
+      "learning_rate": 5.0371582807369365e-06,
+      "loss": 0.4001,
+      "step": 7374
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.3810272567893818,
+      "learning_rate": 5.036032292027092e-06,
+      "loss": 0.3636,
+      "step": 7375
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.910634904157183,
+      "learning_rate": 5.034906301489808e-06,
+      "loss": 0.404,
+      "step": 7376
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.873746028597553,
+      "learning_rate": 5.0337803091821905e-06,
+      "loss": 0.3672,
+      "step": 7377
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.697133961901802,
+      "learning_rate": 5.032654315161345e-06,
+      "loss": 0.3858,
+      "step": 7378
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.732073221845689,
+      "learning_rate": 5.03152831948438e-06,
+      "loss": 0.3498,
+      "step": 7379
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.113510752016988,
+      "learning_rate": 5.030402322208402e-06,
+      "loss": 0.3977,
+      "step": 7380
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.36496587257500335,
+      "learning_rate": 5.029276323390515e-06,
+      "loss": 0.1911,
+      "step": 7381
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.565384988893964,
+      "learning_rate": 5.0281503230878304e-06,
+      "loss": 0.4223,
+      "step": 7382
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3757665206252686,
+      "learning_rate": 5.027024321357452e-06,
+      "loss": 0.2035,
+      "step": 7383
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.421980923449741,
+      "learning_rate": 5.025898318256489e-06,
+      "loss": 0.406,
+      "step": 7384
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.4754569141965264,
+      "learning_rate": 5.024772313842047e-06,
+      "loss": 0.3825,
+      "step": 7385
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.495833121199979,
+      "learning_rate": 5.023646308171234e-06,
+      "loss": 0.4273,
+      "step": 7386
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.149536691295713,
+      "learning_rate": 5.022520301301157e-06,
+      "loss": 0.4274,
+      "step": 7387
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.9325352574623826,
+      "learning_rate": 5.021394293288925e-06,
+      "loss": 0.4305,
+      "step": 7388
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3537882917230986,
+      "learning_rate": 5.020268284191643e-06,
+      "loss": 0.1885,
+      "step": 7389
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.9616837497438,
+      "learning_rate": 5.019142274066421e-06,
+      "loss": 0.4085,
+      "step": 7390
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.596688536816625,
+      "learning_rate": 5.018016262970364e-06,
+      "loss": 0.4,
+      "step": 7391
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.087097299331338,
+      "learning_rate": 5.016890250960582e-06,
+      "loss": 0.386,
+      "step": 7392
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 16.2515936377932,
+      "learning_rate": 5.01576423809418e-06,
+      "loss": 0.3785,
+      "step": 7393
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.199597567782293,
+      "learning_rate": 5.014638224428269e-06,
+      "loss": 0.3832,
+      "step": 7394
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.241119440549563,
+      "learning_rate": 5.013512210019954e-06,
+      "loss": 0.4675,
+      "step": 7395
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.6961592067179,
+      "learning_rate": 5.012386194926344e-06,
+      "loss": 0.3852,
+      "step": 7396
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.5796397171109735,
+      "learning_rate": 5.011260179204546e-06,
+      "loss": 0.4641,
+      "step": 7397
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 5.970778887316613,
+      "learning_rate": 5.01013416291167e-06,
+      "loss": 0.3934,
+      "step": 7398
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.262429731071983,
+      "learning_rate": 5.00900814610482e-06,
+      "loss": 0.4384,
+      "step": 7399
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 11.26114388581254,
+      "learning_rate": 5.0078821288411074e-06,
+      "loss": 0.415,
+      "step": 7400
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.537740614339852,
+      "learning_rate": 5.006756111177639e-06,
+      "loss": 0.3922,
+      "step": 7401
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.8502234757360103,
+      "learning_rate": 5.005630093171523e-06,
+      "loss": 0.3802,
+      "step": 7402
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.393863507477546,
+      "learning_rate": 5.004504074879867e-06,
+      "loss": 0.3788,
+      "step": 7403
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 6.010938970231526,
+      "learning_rate": 5.003378056359779e-06,
+      "loss": 0.435,
+      "step": 7404
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.447928102891006,
+      "learning_rate": 5.002252037668366e-06,
+      "loss": 0.4085,
+      "step": 7405
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.136413678020397,
+      "learning_rate": 5.001126018862737e-06,
+      "loss": 0.4097,
+      "step": 7406
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.472648223717891,
+      "learning_rate": 5e-06,
+      "loss": 0.3914,
+      "step": 7407
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.018675134709104,
+      "learning_rate": 4.998873981137265e-06,
+      "loss": 0.4131,
+      "step": 7408
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.521409783599443,
+      "learning_rate": 4.997747962331634e-06,
+      "loss": 0.3779,
+      "step": 7409
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7308557596935894,
+      "learning_rate": 4.9966219436402225e-06,
+      "loss": 0.3593,
+      "step": 7410
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.630764629879076,
+      "learning_rate": 4.995495925120135e-06,
+      "loss": 0.4245,
+      "step": 7411
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 11.387372839119838,
+      "learning_rate": 4.994369906828479e-06,
+      "loss": 0.3729,
+      "step": 7412
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.501034052151704,
+      "learning_rate": 4.993243888822361e-06,
+      "loss": 0.4292,
+      "step": 7413
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9009675378098634,
+      "learning_rate": 4.992117871158893e-06,
+      "loss": 0.3951,
+      "step": 7414
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.375559974834698,
+      "learning_rate": 4.990991853895182e-06,
+      "loss": 0.3965,
+      "step": 7415
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34370430626850634,
+      "learning_rate": 4.989865837088334e-06,
+      "loss": 0.1647,
+      "step": 7416
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.163992452184963,
+      "learning_rate": 4.988739820795455e-06,
+      "loss": 0.4033,
+      "step": 7417
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 23.819078868529175,
+      "learning_rate": 4.9876138050736575e-06,
+      "loss": 0.3841,
+      "step": 7418
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.149886579801034,
+      "learning_rate": 4.986487789980048e-06,
+      "loss": 0.3894,
+      "step": 7419
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.453736118141464,
+      "learning_rate": 4.985361775571733e-06,
+      "loss": 0.3736,
+      "step": 7420
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3699582958676085,
+      "learning_rate": 4.98423576190582e-06,
+      "loss": 0.1789,
+      "step": 7421
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.2596643162959564,
+      "learning_rate": 4.9831097490394195e-06,
+      "loss": 0.3834,
+      "step": 7422
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.0052334586460425,
+      "learning_rate": 4.981983737029637e-06,
+      "loss": 0.4137,
+      "step": 7423
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.393138091505956,
+      "learning_rate": 4.980857725933581e-06,
+      "loss": 0.4256,
+      "step": 7424
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.375260206856598,
+      "learning_rate": 4.979731715808357e-06,
+      "loss": 0.3843,
+      "step": 7425
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.967561645546602,
+      "learning_rate": 4.978605706711077e-06,
+      "loss": 0.3694,
+      "step": 7426
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5204917124124706,
+      "learning_rate": 4.9774796986988446e-06,
+      "loss": 0.3912,
+      "step": 7427
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7249205226599855,
+      "learning_rate": 4.9763536918287685e-06,
+      "loss": 0.4294,
+      "step": 7428
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.319864765432864,
+      "learning_rate": 4.975227686157954e-06,
+      "loss": 0.4066,
+      "step": 7429
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5243086133089703,
+      "learning_rate": 4.974101681743513e-06,
+      "loss": 0.4214,
+      "step": 7430
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.298376310719366,
+      "learning_rate": 4.97297567864255e-06,
+      "loss": 0.4068,
+      "step": 7431
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39298812026114405,
+      "learning_rate": 4.971849676912172e-06,
+      "loss": 0.2059,
+      "step": 7432
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.752652882330548,
+      "learning_rate": 4.9707236766094855e-06,
+      "loss": 0.3593,
+      "step": 7433
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.545572175559413,
+      "learning_rate": 4.9695976777915995e-06,
+      "loss": 0.3893,
+      "step": 7434
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.571812292302837,
+      "learning_rate": 4.968471680515622e-06,
+      "loss": 0.3574,
+      "step": 7435
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.468944054880684,
+      "learning_rate": 4.967345684838656e-06,
+      "loss": 0.3823,
+      "step": 7436
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.108809279414297,
+      "learning_rate": 4.96621969081781e-06,
+      "loss": 0.4177,
+      "step": 7437
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.410598168313704,
+      "learning_rate": 4.965093698510192e-06,
+      "loss": 0.4086,
+      "step": 7438
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.04052515674128,
+      "learning_rate": 4.963967707972909e-06,
+      "loss": 0.3869,
+      "step": 7439
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.445684912435785,
+      "learning_rate": 4.962841719263066e-06,
+      "loss": 0.4178,
+      "step": 7440
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.6691761269100445,
+      "learning_rate": 4.961715732437767e-06,
+      "loss": 0.4289,
+      "step": 7441
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.369224617793952,
+      "learning_rate": 4.9605897475541245e-06,
+      "loss": 0.4341,
+      "step": 7442
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.684380998920587,
+      "learning_rate": 4.959463764669241e-06,
+      "loss": 0.4397,
+      "step": 7443
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.726430370186764,
+      "learning_rate": 4.958337783840223e-06,
+      "loss": 0.3901,
+      "step": 7444
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.580462935000123,
+      "learning_rate": 4.957211805124175e-06,
+      "loss": 0.428,
+      "step": 7445
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.730011056663571,
+      "learning_rate": 4.956085828578207e-06,
+      "loss": 0.3893,
+      "step": 7446
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.21288597038055,
+      "learning_rate": 4.954959854259422e-06,
+      "loss": 0.3886,
+      "step": 7447
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.501945594504654,
+      "learning_rate": 4.953833882224927e-06,
+      "loss": 0.4163,
+      "step": 7448
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.899077186853764,
+      "learning_rate": 4.952707912531824e-06,
+      "loss": 0.4275,
+      "step": 7449
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.9337473369104288,
+      "learning_rate": 4.951581945237224e-06,
+      "loss": 0.3864,
+      "step": 7450
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.226833573703141,
+      "learning_rate": 4.95045598039823e-06,
+      "loss": 0.433,
+      "step": 7451
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.785904317388335,
+      "learning_rate": 4.949330018071947e-06,
+      "loss": 0.452,
+      "step": 7452
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.587413078541499,
+      "learning_rate": 4.948204058315478e-06,
+      "loss": 0.4126,
+      "step": 7453
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.376166225718339,
+      "learning_rate": 4.947078101185933e-06,
+      "loss": 0.368,
+      "step": 7454
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.654762195637695,
+      "learning_rate": 4.945952146740414e-06,
+      "loss": 0.3985,
+      "step": 7455
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.83784573064741,
+      "learning_rate": 4.944826195036025e-06,
+      "loss": 0.4064,
+      "step": 7456
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.754780098878759,
+      "learning_rate": 4.943700246129871e-06,
+      "loss": 0.4391,
+      "step": 7457
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.4872831102150115,
+      "learning_rate": 4.942574300079058e-06,
+      "loss": 0.4201,
+      "step": 7458
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.494124156207137,
+      "learning_rate": 4.9414483569406905e-06,
+      "loss": 0.3979,
+      "step": 7459
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.214578703534903,
+      "learning_rate": 4.94032241677187e-06,
+      "loss": 0.3636,
+      "step": 7460
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4061683611808264,
+      "learning_rate": 4.9391964796297e-06,
+      "loss": 0.2014,
+      "step": 7461
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.6473879652518795,
+      "learning_rate": 4.938070545571289e-06,
+      "loss": 0.4,
+      "step": 7462
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9709417132723215,
+      "learning_rate": 4.936944614653738e-06,
+      "loss": 0.3686,
+      "step": 7463
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35491317433298414,
+      "learning_rate": 4.93581868693415e-06,
+      "loss": 0.1997,
+      "step": 7464
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5256296986657776,
+      "learning_rate": 4.9346927624696266e-06,
+      "loss": 0.4417,
+      "step": 7465
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.536504598227062,
+      "learning_rate": 4.9335668413172765e-06,
+      "loss": 0.4194,
+      "step": 7466
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.54060935529555,
+      "learning_rate": 4.932440923534199e-06,
+      "loss": 0.4085,
+      "step": 7467
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.262671314986386,
+      "learning_rate": 4.931315009177498e-06,
+      "loss": 0.4137,
+      "step": 7468
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.45524306910498,
+      "learning_rate": 4.9301890983042744e-06,
+      "loss": 0.4179,
+      "step": 7469
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9064277720515355,
+      "learning_rate": 4.929063190971633e-06,
+      "loss": 0.3999,
+      "step": 7470
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.245512616692113,
+      "learning_rate": 4.927937287236676e-06,
+      "loss": 0.4118,
+      "step": 7471
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.839042297868189,
+      "learning_rate": 4.926811387156502e-06,
+      "loss": 0.4187,
+      "step": 7472
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.147044696254284,
+      "learning_rate": 4.925685490788218e-06,
+      "loss": 0.3757,
+      "step": 7473
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.015090714864325,
+      "learning_rate": 4.924559598188924e-06,
+      "loss": 0.4159,
+      "step": 7474
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.267791028662053,
+      "learning_rate": 4.923433709415722e-06,
+      "loss": 0.4187,
+      "step": 7475
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.120586450436925,
+      "learning_rate": 4.922307824525709e-06,
+      "loss": 0.3878,
+      "step": 7476
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.608726007298672,
+      "learning_rate": 4.921181943575992e-06,
+      "loss": 0.4457,
+      "step": 7477
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4196442263950659,
+      "learning_rate": 4.920056066623671e-06,
+      "loss": 0.2055,
+      "step": 7478
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.586504405338923,
+      "learning_rate": 4.918930193725846e-06,
+      "loss": 0.4375,
+      "step": 7479
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.331024353033513,
+      "learning_rate": 4.917804324939614e-06,
+      "loss": 0.3906,
+      "step": 7480
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.584052676423924,
+      "learning_rate": 4.916678460322082e-06,
+      "loss": 0.3851,
+      "step": 7481
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.915253290869042,
+      "learning_rate": 4.915552599930345e-06,
+      "loss": 0.3882,
+      "step": 7482
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.277479007961027,
+      "learning_rate": 4.914426743821507e-06,
+      "loss": 0.414,
+      "step": 7483
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.271034688620959,
+      "learning_rate": 4.913300892052661e-06,
+      "loss": 0.4063,
+      "step": 7484
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.067412978739892,
+      "learning_rate": 4.912175044680914e-06,
+      "loss": 0.3919,
+      "step": 7485
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5425938583330074,
+      "learning_rate": 4.911049201763364e-06,
+      "loss": 0.3993,
+      "step": 7486
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.268028526936185,
+      "learning_rate": 4.909923363357107e-06,
+      "loss": 0.4265,
+      "step": 7487
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.925067810106691,
+      "learning_rate": 4.9087975295192415e-06,
+      "loss": 0.405,
+      "step": 7488
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.097800564039274,
+      "learning_rate": 4.907671700306871e-06,
+      "loss": 0.3759,
+      "step": 7489
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.579293023024187,
+      "learning_rate": 4.906545875777089e-06,
+      "loss": 0.4035,
+      "step": 7490
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.952526357147157,
+      "learning_rate": 4.905420055986997e-06,
+      "loss": 0.384,
+      "step": 7491
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.785960541412142,
+      "learning_rate": 4.904294240993689e-06,
+      "loss": 0.3878,
+      "step": 7492
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.573971798222901,
+      "learning_rate": 4.903168430854267e-06,
+      "loss": 0.3986,
+      "step": 7493
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.408258974835639,
+      "learning_rate": 4.902042625625827e-06,
+      "loss": 0.4202,
+      "step": 7494
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.365599561875153,
+      "learning_rate": 4.9009168253654645e-06,
+      "loss": 0.4001,
+      "step": 7495
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.1488344717162584,
+      "learning_rate": 4.899791030130276e-06,
+      "loss": 0.3809,
+      "step": 7496
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.817115598069229,
+      "learning_rate": 4.8986652399773625e-06,
+      "loss": 0.373,
+      "step": 7497
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.7052860301753987,
+      "learning_rate": 4.897539454963817e-06,
+      "loss": 0.4298,
+      "step": 7498
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5173896900404453,
+      "learning_rate": 4.896413675146737e-06,
+      "loss": 0.4153,
+      "step": 7499
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.248917644303711,
+      "learning_rate": 4.895287900583216e-06,
+      "loss": 0.41,
+      "step": 7500
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.130836806594309,
+      "learning_rate": 4.894162131330353e-06,
+      "loss": 0.3838,
+      "step": 7501
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.0579885799729376,
+      "learning_rate": 4.893036367445239e-06,
+      "loss": 0.3767,
+      "step": 7502
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 7.570504590941025,
+      "learning_rate": 4.891910608984974e-06,
+      "loss": 0.4218,
+      "step": 7503
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.320123825943609,
+      "learning_rate": 4.890784856006649e-06,
+      "loss": 0.385,
+      "step": 7504
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.541315665760254,
+      "learning_rate": 4.889659108567362e-06,
+      "loss": 0.398,
+      "step": 7505
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.33815211813945,
+      "learning_rate": 4.888533366724204e-06,
+      "loss": 0.3941,
+      "step": 7506
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.927607063509523,
+      "learning_rate": 4.887407630534271e-06,
+      "loss": 0.4074,
+      "step": 7507
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 10.482863333120724,
+      "learning_rate": 4.886281900054653e-06,
+      "loss": 0.414,
+      "step": 7508
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.292278165406564,
+      "learning_rate": 4.88515617534245e-06,
+      "loss": 0.3807,
+      "step": 7509
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 9.22092676600162,
+      "learning_rate": 4.8840304564547505e-06,
+      "loss": 0.4134,
+      "step": 7510
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.398233541770894,
+      "learning_rate": 4.882904743448647e-06,
+      "loss": 0.388,
+      "step": 7511
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.734616917623252,
+      "learning_rate": 4.881779036381234e-06,
+      "loss": 0.3572,
+      "step": 7512
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.486638480697668,
+      "learning_rate": 4.880653335309603e-06,
+      "loss": 0.4044,
+      "step": 7513
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.945172679013795,
+      "learning_rate": 4.879527640290846e-06,
+      "loss": 0.3995,
+      "step": 7514
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.726048253129453,
+      "learning_rate": 4.878401951382054e-06,
+      "loss": 0.4164,
+      "step": 7515
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.421807491978879,
+      "learning_rate": 4.877276268640317e-06,
+      "loss": 0.397,
+      "step": 7516
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.3528722556460195,
+      "learning_rate": 4.8761505921227295e-06,
+      "loss": 0.4016,
+      "step": 7517
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.555697226117061,
+      "learning_rate": 4.87502492188638e-06,
+      "loss": 0.4461,
+      "step": 7518
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.863174070822113,
+      "learning_rate": 4.873899257988359e-06,
+      "loss": 0.441,
+      "step": 7519
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 16.133677575833484,
+      "learning_rate": 4.872773600485754e-06,
+      "loss": 0.4032,
+      "step": 7520
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.8552727195067016,
+      "learning_rate": 4.87164794943566e-06,
+      "loss": 0.4548,
+      "step": 7521
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.10674923284175,
+      "learning_rate": 4.870522304895164e-06,
+      "loss": 0.4212,
+      "step": 7522
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.6616249425750027,
+      "learning_rate": 4.869396666921353e-06,
+      "loss": 0.3727,
+      "step": 7523
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.240463563704339,
+      "learning_rate": 4.8682710355713175e-06,
+      "loss": 0.4084,
+      "step": 7524
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.143397048755454,
+      "learning_rate": 4.867145410902146e-06,
+      "loss": 0.4011,
+      "step": 7525
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.125368643688767,
+      "learning_rate": 4.866019792970926e-06,
+      "loss": 0.388,
+      "step": 7526
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.458955526146529,
+      "learning_rate": 4.8648941818347465e-06,
+      "loss": 0.4022,
+      "step": 7527
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.581750772561724,
+      "learning_rate": 4.863768577550691e-06,
+      "loss": 0.4057,
+      "step": 7528
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.4411867255987705,
+      "learning_rate": 4.86264298017585e-06,
+      "loss": 0.4063,
+      "step": 7529
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.904271711369062,
+      "learning_rate": 4.8615173897673105e-06,
+      "loss": 0.4042,
+      "step": 7530
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.35074225408469195,
+      "learning_rate": 4.860391806382157e-06,
+      "loss": 0.1884,
+      "step": 7531
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.249482694235853,
+      "learning_rate": 4.859266230077474e-06,
+      "loss": 0.372,
+      "step": 7532
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.215023800469746,
+      "learning_rate": 4.858140660910352e-06,
+      "loss": 0.3826,
+      "step": 7533
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.663394116934857,
+      "learning_rate": 4.857015098937872e-06,
+      "loss": 0.4098,
+      "step": 7534
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.169895449815317,
+      "learning_rate": 4.855889544217119e-06,
+      "loss": 0.4023,
+      "step": 7535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 14.989174318168637,
+      "learning_rate": 4.854763996805179e-06,
+      "loss": 0.4171,
+      "step": 7536
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.882760985438017,
+      "learning_rate": 4.853638456759137e-06,
+      "loss": 0.4106,
+      "step": 7537
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.41266973947654517,
+      "learning_rate": 4.8525129241360746e-06,
+      "loss": 0.203,
+      "step": 7538
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.427499360321123,
+      "learning_rate": 4.851387398993073e-06,
+      "loss": 0.4103,
+      "step": 7539
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 4.9435118815984564,
+      "learning_rate": 4.850261881387222e-06,
+      "loss": 0.4188,
+      "step": 7540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.387774875653295,
+      "learning_rate": 4.849136371375599e-06,
+      "loss": 0.3831,
+      "step": 7541
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.563997719557452,
+      "learning_rate": 4.848010869015288e-06,
+      "loss": 0.3856,
+      "step": 7542
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5531760180233536,
+      "learning_rate": 4.846885374363367e-06,
+      "loss": 0.3913,
+      "step": 7543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.271593738305013,
+      "learning_rate": 4.845759887476924e-06,
+      "loss": 0.4279,
+      "step": 7544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 13.132599335948717,
+      "learning_rate": 4.844634408413036e-06,
+      "loss": 0.4528,
+      "step": 7545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.338051974100555,
+      "learning_rate": 4.8435089372287844e-06,
+      "loss": 0.3962,
+      "step": 7546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.978158382490954,
+      "learning_rate": 4.842383473981248e-06,
+      "loss": 0.3775,
+      "step": 7547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.34820033024356584,
+      "learning_rate": 4.8412580187275095e-06,
+      "loss": 0.2001,
+      "step": 7548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 6.714233409810886,
+      "learning_rate": 4.840132571524646e-06,
+      "loss": 0.4477,
+      "step": 7549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.867138975650887,
+      "learning_rate": 4.839007132429738e-06,
+      "loss": 0.3869,
+      "step": 7550
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.090255102584191,
+      "learning_rate": 4.837881701499861e-06,
+      "loss": 0.4463,
+      "step": 7551
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2023936240071396,
+      "learning_rate": 4.836756278792098e-06,
+      "loss": 0.4088,
+      "step": 7552
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.690944966893554,
+      "learning_rate": 4.835630864363524e-06,
+      "loss": 0.3714,
+      "step": 7553
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.984024571155635,
+      "learning_rate": 4.834505458271217e-06,
+      "loss": 0.4154,
+      "step": 7554
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.035895458091396,
+      "learning_rate": 4.833380060572251e-06,
+      "loss": 0.4191,
+      "step": 7555
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.760291601257955,
+      "learning_rate": 4.832254671323707e-06,
+      "loss": 0.4063,
+      "step": 7556
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.843220581929175,
+      "learning_rate": 4.83112929058266e-06,
+      "loss": 0.4112,
+      "step": 7557
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.1389030699592,
+      "learning_rate": 4.830003918406184e-06,
+      "loss": 0.407,
+      "step": 7558
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.255810652160049,
+      "learning_rate": 4.828878554851355e-06,
+      "loss": 0.4315,
+      "step": 7559
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6744560538802014,
+      "learning_rate": 4.827753199975248e-06,
+      "loss": 0.3895,
+      "step": 7560
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6311687518789166,
+      "learning_rate": 4.826627853834937e-06,
+      "loss": 0.4294,
+      "step": 7561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 31.629563896629524,
+      "learning_rate": 4.825502516487497e-06,
+      "loss": 0.4165,
+      "step": 7562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.883429241317097,
+      "learning_rate": 4.824377187989997e-06,
+      "loss": 0.4351,
+      "step": 7563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.882802243139063,
+      "learning_rate": 4.823251868399517e-06,
+      "loss": 0.4153,
+      "step": 7564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.3264203331794775,
+      "learning_rate": 4.8221265577731256e-06,
+      "loss": 0.377,
+      "step": 7565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.810582835554762,
+      "learning_rate": 4.821001256167895e-06,
+      "loss": 0.3804,
+      "step": 7566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4795745440569004,
+      "learning_rate": 4.819875963640896e-06,
+      "loss": 0.3873,
+      "step": 7567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.992531748927126,
+      "learning_rate": 4.818750680249202e-06,
+      "loss": 0.3799,
+      "step": 7568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.201568459771983,
+      "learning_rate": 4.817625406049883e-06,
+      "loss": 0.4425,
+      "step": 7569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 19.518534460136195,
+      "learning_rate": 4.8165001411000074e-06,
+      "loss": 0.4115,
+      "step": 7570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.292485242795883,
+      "learning_rate": 4.815374885456646e-06,
+      "loss": 0.3754,
+      "step": 7571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.677353270746864,
+      "learning_rate": 4.81424963917687e-06,
+      "loss": 0.4419,
+      "step": 7572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.9699526752119585,
+      "learning_rate": 4.813124402317747e-06,
+      "loss": 0.4039,
+      "step": 7573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.342863915664349,
+      "learning_rate": 4.8119991749363445e-06,
+      "loss": 0.3745,
+      "step": 7574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.139242775468365,
+      "learning_rate": 4.81087395708973e-06,
+      "loss": 0.3943,
+      "step": 7575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5228867222405444,
+      "learning_rate": 4.809748748834972e-06,
+      "loss": 0.415,
+      "step": 7576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.923389523132817,
+      "learning_rate": 4.808623550229139e-06,
+      "loss": 0.3992,
+      "step": 7577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6882541582318757,
+      "learning_rate": 4.807498361329295e-06,
+      "loss": 0.3898,
+      "step": 7578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.227127290968584,
+      "learning_rate": 4.806373182192505e-06,
+      "loss": 0.4104,
+      "step": 7579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.36139309909114564,
+      "learning_rate": 4.805248012875837e-06,
+      "loss": 0.1926,
+      "step": 7580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9764338778573696,
+      "learning_rate": 4.804122853436355e-06,
+      "loss": 0.4383,
+      "step": 7581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.487661557433378,
+      "learning_rate": 4.802997703931124e-06,
+      "loss": 0.3806,
+      "step": 7582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.013201791689612,
+      "learning_rate": 4.801872564417204e-06,
+      "loss": 0.3807,
+      "step": 7583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.243480172889367,
+      "learning_rate": 4.800747434951665e-06,
+      "loss": 0.3647,
+      "step": 7584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.2158894471186565,
+      "learning_rate": 4.799622315591566e-06,
+      "loss": 0.406,
+      "step": 7585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.9771813385423425,
+      "learning_rate": 4.798497206393969e-06,
+      "loss": 0.4067,
+      "step": 7586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35915953354126906,
+      "learning_rate": 4.797372107415935e-06,
+      "loss": 0.1969,
+      "step": 7587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.945053395708759,
+      "learning_rate": 4.79624701871453e-06,
+      "loss": 0.377,
+      "step": 7588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.188161161899293,
+      "learning_rate": 4.795121940346811e-06,
+      "loss": 0.3753,
+      "step": 7589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.548684228472511,
+      "learning_rate": 4.79399687236984e-06,
+      "loss": 0.37,
+      "step": 7590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.697240185042246,
+      "learning_rate": 4.792871814840674e-06,
+      "loss": 0.4013,
+      "step": 7591
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8689259154889704,
+      "learning_rate": 4.7917467678163764e-06,
+      "loss": 0.4338,
+      "step": 7592
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.108592045547587,
+      "learning_rate": 4.7906217313540035e-06,
+      "loss": 0.4109,
+      "step": 7593
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8415812543653325,
+      "learning_rate": 4.789496705510613e-06,
+      "loss": 0.4069,
+      "step": 7594
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.062712010189525,
+      "learning_rate": 4.788371690343262e-06,
+      "loss": 0.3729,
+      "step": 7595
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6398588956526083,
+      "learning_rate": 4.787246685909011e-06,
+      "loss": 0.387,
+      "step": 7596
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.764677833758133,
+      "learning_rate": 4.786121692264915e-06,
+      "loss": 0.3969,
+      "step": 7597
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.391880191897102,
+      "learning_rate": 4.784996709468029e-06,
+      "loss": 0.4338,
+      "step": 7598
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.600862995062789,
+      "learning_rate": 4.783871737575406e-06,
+      "loss": 0.3787,
+      "step": 7599
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.225566549304748,
+      "learning_rate": 4.782746776644107e-06,
+      "loss": 0.3854,
+      "step": 7600
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.707393171392065,
+      "learning_rate": 4.781621826731183e-06,
+      "loss": 0.3649,
+      "step": 7601
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.218817906643685,
+      "learning_rate": 4.780496887893686e-06,
+      "loss": 0.3923,
+      "step": 7602
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.121250523904275,
+      "learning_rate": 4.7793719601886735e-06,
+      "loss": 0.3695,
+      "step": 7603
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.750020654851918,
+      "learning_rate": 4.7782470436731954e-06,
+      "loss": 0.4032,
+      "step": 7604
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.4602758622531296,
+      "learning_rate": 4.777122138404304e-06,
+      "loss": 0.3925,
+      "step": 7605
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.280852586472201,
+      "learning_rate": 4.77599724443905e-06,
+      "loss": 0.3797,
+      "step": 7606
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.356554702684086,
+      "learning_rate": 4.7748723618344865e-06,
+      "loss": 0.377,
+      "step": 7607
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.418500557789932,
+      "learning_rate": 4.773747490647663e-06,
+      "loss": 0.3801,
+      "step": 7608
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.782577083580268,
+      "learning_rate": 4.772622630935629e-06,
+      "loss": 0.4206,
+      "step": 7609
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.052898107662577,
+      "learning_rate": 4.771497782755433e-06,
+      "loss": 0.3972,
+      "step": 7610
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.35370441714966494,
+      "learning_rate": 4.770372946164125e-06,
+      "loss": 0.1904,
+      "step": 7611
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.612204389551049,
+      "learning_rate": 4.7692481212187534e-06,
+      "loss": 0.383,
+      "step": 7612
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.465185717593721,
+      "learning_rate": 4.768123307976365e-06,
+      "loss": 0.3954,
+      "step": 7613
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.2134835258056125,
+      "learning_rate": 4.766998506494005e-06,
+      "loss": 0.3987,
+      "step": 7614
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.062200948152723,
+      "learning_rate": 4.765873716828723e-06,
+      "loss": 0.3595,
+      "step": 7615
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.164363661648857,
+      "learning_rate": 4.764748939037562e-06,
+      "loss": 0.423,
+      "step": 7616
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.952453868497788,
+      "learning_rate": 4.763624173177568e-06,
+      "loss": 0.3628,
+      "step": 7617
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.8480011771579905,
+      "learning_rate": 4.7624994193057825e-06,
+      "loss": 0.3944,
+      "step": 7618
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.565709991316697,
+      "learning_rate": 4.761374677479255e-06,
+      "loss": 0.4132,
+      "step": 7619
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8992267611688294,
+      "learning_rate": 4.7602499477550255e-06,
+      "loss": 0.4225,
+      "step": 7620
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6227922837945292,
+      "learning_rate": 4.759125230190138e-06,
+      "loss": 0.399,
+      "step": 7621
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2783246489764943,
+      "learning_rate": 4.758000524841632e-06,
+      "loss": 0.3907,
+      "step": 7622
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.708163816544075,
+      "learning_rate": 4.75687583176655e-06,
+      "loss": 0.4014,
+      "step": 7623
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.7623941895173685,
+      "learning_rate": 4.755751151021934e-06,
+      "loss": 0.4551,
+      "step": 7624
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.697068853424013,
+      "learning_rate": 4.754626482664824e-06,
+      "loss": 0.3947,
+      "step": 7625
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.272648339336723,
+      "learning_rate": 4.7535018267522575e-06,
+      "loss": 0.3704,
+      "step": 7626
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.08104332881365,
+      "learning_rate": 4.752377183341276e-06,
+      "loss": 0.4221,
+      "step": 7627
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.001855135228243,
+      "learning_rate": 4.7512525524889154e-06,
+      "loss": 0.4229,
+      "step": 7628
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5911505955115266,
+      "learning_rate": 4.750127934252216e-06,
+      "loss": 0.3771,
+      "step": 7629
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.2803493990434704,
+      "learning_rate": 4.74900332868821e-06,
+      "loss": 0.3808,
+      "step": 7630
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3415238899641413,
+      "learning_rate": 4.747878735853939e-06,
+      "loss": 0.1733,
+      "step": 7631
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.531391003333005,
+      "learning_rate": 4.746754155806437e-06,
+      "loss": 0.4414,
+      "step": 7632
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.757454721526316,
+      "learning_rate": 4.745629588602739e-06,
+      "loss": 0.4123,
+      "step": 7633
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.097629259882896,
+      "learning_rate": 4.744505034299876e-06,
+      "loss": 0.4182,
+      "step": 7634
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3623734541242755,
+      "learning_rate": 4.743380492954888e-06,
+      "loss": 0.2086,
+      "step": 7635
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.011547030077778,
+      "learning_rate": 4.742255964624804e-06,
+      "loss": 0.3703,
+      "step": 7636
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2965584011259983,
+      "learning_rate": 4.741131449366657e-06,
+      "loss": 0.4262,
+      "step": 7637
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.500540324103882,
+      "learning_rate": 4.740006947237477e-06,
+      "loss": 0.3766,
+      "step": 7638
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.152032054296632,
+      "learning_rate": 4.738882458294299e-06,
+      "loss": 0.4116,
+      "step": 7639
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.083120905589828,
+      "learning_rate": 4.737757982594151e-06,
+      "loss": 0.3966,
+      "step": 7640
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.157475475450779,
+      "learning_rate": 4.736633520194064e-06,
+      "loss": 0.4169,
+      "step": 7641
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.424988481067749,
+      "learning_rate": 4.7355090711510635e-06,
+      "loss": 0.3974,
+      "step": 7642
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.811110761154684,
+      "learning_rate": 4.7343846355221825e-06,
+      "loss": 0.4577,
+      "step": 7643
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3824802829298908,
+      "learning_rate": 4.733260213364447e-06,
+      "loss": 0.186,
+      "step": 7644
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.851995931291511,
+      "learning_rate": 4.732135804734885e-06,
+      "loss": 0.3534,
+      "step": 7645
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.05966774937447,
+      "learning_rate": 4.731011409690519e-06,
+      "loss": 0.4088,
+      "step": 7646
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.326080308200854,
+      "learning_rate": 4.729887028288379e-06,
+      "loss": 0.396,
+      "step": 7647
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.3674126368970794,
+      "learning_rate": 4.7287626605854895e-06,
+      "loss": 0.4051,
+      "step": 7648
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.088112803734968,
+      "learning_rate": 4.727638306638872e-06,
+      "loss": 0.4196,
+      "step": 7649
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.273461456804781,
+      "learning_rate": 4.7265139665055495e-06,
+      "loss": 0.4108,
+      "step": 7650
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.581929319437232,
+      "learning_rate": 4.725389640242551e-06,
+      "loss": 0.4042,
+      "step": 7651
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.881908496639747,
+      "learning_rate": 4.724265327906893e-06,
+      "loss": 0.4206,
+      "step": 7652
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5737379329635477,
+      "learning_rate": 4.723141029555599e-06,
+      "loss": 0.4434,
+      "step": 7653
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.276951365696513,
+      "learning_rate": 4.722016745245687e-06,
+      "loss": 0.3949,
+      "step": 7654
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 8.09285140057372,
+      "learning_rate": 4.720892475034181e-06,
+      "loss": 0.4157,
+      "step": 7655
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.546754095256626,
+      "learning_rate": 4.7197682189781e-06,
+      "loss": 0.4381,
+      "step": 7656
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.5153576560043875,
+      "learning_rate": 4.71864397713446e-06,
+      "loss": 0.404,
+      "step": 7657
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.4736250782691895,
+      "learning_rate": 4.717519749560281e-06,
+      "loss": 0.4119,
+      "step": 7658
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.56858063756365,
+      "learning_rate": 4.716395536312579e-06,
+      "loss": 0.4285,
+      "step": 7659
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.66622757742528,
+      "learning_rate": 4.715271337448371e-06,
+      "loss": 0.3679,
+      "step": 7660
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.93678880700122,
+      "learning_rate": 4.714147153024672e-06,
+      "loss": 0.4004,
+      "step": 7661
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.817679546467621,
+      "learning_rate": 4.713022983098496e-06,
+      "loss": 0.3969,
+      "step": 7662
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.8807709187527575,
+      "learning_rate": 4.711898827726862e-06,
+      "loss": 0.3943,
+      "step": 7663
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.260814483892727,
+      "learning_rate": 4.710774686966778e-06,
+      "loss": 0.3853,
+      "step": 7664
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.499405353448653,
+      "learning_rate": 4.709650560875258e-06,
+      "loss": 0.381,
+      "step": 7665
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.2732918004262976,
+      "learning_rate": 4.708526449509317e-06,
+      "loss": 0.3734,
+      "step": 7666
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.577703332879735,
+      "learning_rate": 4.707402352925964e-06,
+      "loss": 0.4078,
+      "step": 7667
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.833301843115611,
+      "learning_rate": 4.706278271182209e-06,
+      "loss": 0.3935,
+      "step": 7668
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.079898112352966,
+      "learning_rate": 4.705154204335063e-06,
+      "loss": 0.4056,
+      "step": 7669
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.147452755418713,
+      "learning_rate": 4.704030152441535e-06,
+      "loss": 0.4165,
+      "step": 7670
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.108437941578633,
+      "learning_rate": 4.702906115558632e-06,
+      "loss": 0.3737,
+      "step": 7671
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.923580869675884,
+      "learning_rate": 4.7017820937433635e-06,
+      "loss": 0.3984,
+      "step": 7672
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.615116622908354,
+      "learning_rate": 4.700658087052733e-06,
+      "loss": 0.4038,
+      "step": 7673
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.239787269862973,
+      "learning_rate": 4.699534095543749e-06,
+      "loss": 0.3837,
+      "step": 7674
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.3589550862275744,
+      "learning_rate": 4.698410119273417e-06,
+      "loss": 0.1786,
+      "step": 7675
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.094482554221229,
+      "learning_rate": 4.69728615829874e-06,
+      "loss": 0.4164,
+      "step": 7676
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.807043597921026,
+      "learning_rate": 4.696162212676721e-06,
+      "loss": 0.3836,
+      "step": 7677
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.650128550189748,
+      "learning_rate": 4.695038282464365e-06,
+      "loss": 0.4049,
+      "step": 7678
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 16.409566618136935,
+      "learning_rate": 4.693914367718673e-06,
+      "loss": 0.42,
+      "step": 7679
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.347391799414436,
+      "learning_rate": 4.6927904684966475e-06,
+      "loss": 0.4102,
+      "step": 7680
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 7.9529796958388195,
+      "learning_rate": 4.691666584855286e-06,
+      "loss": 0.3549,
+      "step": 7681
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 9.246211034525464,
+      "learning_rate": 4.6905427168515914e-06,
+      "loss": 0.382,
+      "step": 7682
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.546486807628696,
+      "learning_rate": 4.689418864542561e-06,
+      "loss": 0.4203,
+      "step": 7683
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.650525464129712,
+      "learning_rate": 4.688295027985194e-06,
+      "loss": 0.4226,
+      "step": 7684
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6337866066137567,
+      "learning_rate": 4.687171207236484e-06,
+      "loss": 0.4335,
+      "step": 7685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.039984758643192,
+      "learning_rate": 4.686047402353433e-06,
+      "loss": 0.3816,
+      "step": 7686
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.974175311608317,
+      "learning_rate": 4.684923613393035e-06,
+      "loss": 0.3986,
+      "step": 7687
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 13.81213995551743,
+      "learning_rate": 4.683799840412283e-06,
+      "loss": 0.4357,
+      "step": 7688
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.373492781994824,
+      "learning_rate": 4.682676083468172e-06,
+      "loss": 0.4342,
+      "step": 7689
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.229140501565141,
+      "learning_rate": 4.681552342617696e-06,
+      "loss": 0.4027,
+      "step": 7690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.8247116793822467,
+      "learning_rate": 4.680428617917846e-06,
+      "loss": 0.427,
+      "step": 7691
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.6962117631784053,
+      "learning_rate": 4.679304909425615e-06,
+      "loss": 0.3665,
+      "step": 7692
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.5492374065254944,
+      "learning_rate": 4.678181217197993e-06,
+      "loss": 0.3766,
+      "step": 7693
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.146817320750688,
+      "learning_rate": 4.677057541291972e-06,
+      "loss": 0.3995,
+      "step": 7694
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.1422541727937343,
+      "learning_rate": 4.6759338817645395e-06,
+      "loss": 0.399,
+      "step": 7695
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.077965061021107,
+      "learning_rate": 4.674810238672684e-06,
+      "loss": 0.436,
+      "step": 7696
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.457451050712827,
+      "learning_rate": 4.67368661207339e-06,
+      "loss": 0.4205,
+      "step": 7697
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 54.98911119533138,
+      "learning_rate": 4.67256300202365e-06,
+      "loss": 0.4002,
+      "step": 7698
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.8531534591596905,
+      "learning_rate": 4.671439408580447e-06,
+      "loss": 0.392,
+      "step": 7699
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.8150398682628404,
+      "learning_rate": 4.670315831800767e-06,
+      "loss": 0.3569,
+      "step": 7700
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.936826348059666,
+      "learning_rate": 4.669192271741591e-06,
+      "loss": 0.4078,
+      "step": 7701
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.5125780821154375,
+      "learning_rate": 4.668068728459905e-06,
+      "loss": 0.4147,
+      "step": 7702
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.038114656689894,
+      "learning_rate": 4.6669452020126915e-06,
+      "loss": 0.4095,
+      "step": 7703
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.210678135165998,
+      "learning_rate": 4.6658216924569325e-06,
+      "loss": 0.4473,
+      "step": 7704
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.0344776373668685,
+      "learning_rate": 4.664698199849604e-06,
+      "loss": 0.3899,
+      "step": 7705
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.192209586676284,
+      "learning_rate": 4.663574724247693e-06,
+      "loss": 0.4097,
+      "step": 7706
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.921706449354061,
+      "learning_rate": 4.662451265708174e-06,
+      "loss": 0.4018,
+      "step": 7707
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.908977472837299,
+      "learning_rate": 4.6613278242880275e-06,
+      "loss": 0.3996,
+      "step": 7708
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3673644989356645,
+      "learning_rate": 4.660204400044227e-06,
+      "loss": 0.1925,
+      "step": 7709
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.778731799891995,
+      "learning_rate": 4.659080993033754e-06,
+      "loss": 0.4044,
+      "step": 7710
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6056385027784774,
+      "learning_rate": 4.6579576033135815e-06,
+      "loss": 0.4154,
+      "step": 7711
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.151777269010656,
+      "learning_rate": 4.656834230940684e-06,
+      "loss": 0.3759,
+      "step": 7712
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.661094118381373,
+      "learning_rate": 4.655710875972035e-06,
+      "loss": 0.3665,
+      "step": 7713
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 12.593542107333436,
+      "learning_rate": 4.6545875384646075e-06,
+      "loss": 0.372,
+      "step": 7714
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.3679404987767745,
+      "learning_rate": 4.653464218475375e-06,
+      "loss": 0.4566,
+      "step": 7715
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.849149300537279,
+      "learning_rate": 4.652340916061308e-06,
+      "loss": 0.3997,
+      "step": 7716
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.46668745408231,
+      "learning_rate": 4.651217631279374e-06,
+      "loss": 0.4251,
+      "step": 7717
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.547397913047259,
+      "learning_rate": 4.6500943641865455e-06,
+      "loss": 0.3969,
+      "step": 7718
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.720245124503414,
+      "learning_rate": 4.648971114839792e-06,
+      "loss": 0.3859,
+      "step": 7719
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.466764824250384,
+      "learning_rate": 4.647847883296079e-06,
+      "loss": 0.4318,
+      "step": 7720
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.544545876676372,
+      "learning_rate": 4.646724669612369e-06,
+      "loss": 0.4196,
+      "step": 7721
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.298667160176665,
+      "learning_rate": 4.645601473845636e-06,
+      "loss": 0.3999,
+      "step": 7722
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.358399451500897,
+      "learning_rate": 4.64447829605284e-06,
+      "loss": 0.3989,
+      "step": 7723
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.916230246882302,
+      "learning_rate": 4.643355136290946e-06,
+      "loss": 0.3564,
+      "step": 7724
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7316414401037616,
+      "learning_rate": 4.642231994616916e-06,
+      "loss": 0.367,
+      "step": 7725
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.140007756386835,
+      "learning_rate": 4.641108871087714e-06,
+      "loss": 0.4405,
+      "step": 7726
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3116099513489554,
+      "learning_rate": 4.639985765760299e-06,
+      "loss": 0.3764,
+      "step": 7727
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4607699102972145,
+      "learning_rate": 4.638862678691631e-06,
+      "loss": 0.4189,
+      "step": 7728
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.447529712385899,
+      "learning_rate": 4.637739609938673e-06,
+      "loss": 0.3833,
+      "step": 7729
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.222575905292448,
+      "learning_rate": 4.6366165595583805e-06,
+      "loss": 0.395,
+      "step": 7730
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6288473908332524,
+      "learning_rate": 4.635493527607711e-06,
+      "loss": 0.4036,
+      "step": 7731
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 23.58604445608638,
+      "learning_rate": 4.63437051414362e-06,
+      "loss": 0.4237,
+      "step": 7732
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.71033327194461,
+      "learning_rate": 4.633247519223066e-06,
+      "loss": 0.4254,
+      "step": 7733
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.4367953288664985,
+      "learning_rate": 4.632124542903003e-06,
+      "loss": 0.4167,
+      "step": 7734
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.767005498457986,
+      "learning_rate": 4.631001585240382e-06,
+      "loss": 0.4022,
+      "step": 7735
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.900198863622219,
+      "learning_rate": 4.629878646292159e-06,
+      "loss": 0.3991,
+      "step": 7736
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.201889573575235,
+      "learning_rate": 4.628755726115284e-06,
+      "loss": 0.3918,
+      "step": 7737
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.867875086541487,
+      "learning_rate": 4.627632824766708e-06,
+      "loss": 0.385,
+      "step": 7738
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.140957909580603,
+      "learning_rate": 4.626509942303382e-06,
+      "loss": 0.4274,
+      "step": 7739
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.429136593102695,
+      "learning_rate": 4.625387078782252e-06,
+      "loss": 0.4059,
+      "step": 7740
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.285467231564058,
+      "learning_rate": 4.624264234260269e-06,
+      "loss": 0.4426,
+      "step": 7741
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.815750153403125,
+      "learning_rate": 4.62314140879438e-06,
+      "loss": 0.377,
+      "step": 7742
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3741337285387614,
+      "learning_rate": 4.62201860244153e-06,
+      "loss": 0.4158,
+      "step": 7743
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.6546166462449134,
+      "learning_rate": 4.620895815258662e-06,
+      "loss": 0.3895,
+      "step": 7744
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.165548707249975,
+      "learning_rate": 4.619773047302725e-06,
+      "loss": 0.3616,
+      "step": 7745
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.133323328786923,
+      "learning_rate": 4.618650298630658e-06,
+      "loss": 0.4456,
+      "step": 7746
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.05819062196841,
+      "learning_rate": 4.617527569299406e-06,
+      "loss": 0.426,
+      "step": 7747
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.973574922086011,
+      "learning_rate": 4.6164048593659076e-06,
+      "loss": 0.3794,
+      "step": 7748
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.0182983331977855,
+      "learning_rate": 4.615282168887104e-06,
+      "loss": 0.4084,
+      "step": 7749
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.0647584503797525,
+      "learning_rate": 4.614159497919936e-06,
+      "loss": 0.4047,
+      "step": 7750
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.6407427150735066,
+      "learning_rate": 4.613036846521339e-06,
+      "loss": 0.3919,
+      "step": 7751
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.7884906723211964,
+      "learning_rate": 4.61191421474825e-06,
+      "loss": 0.4012,
+      "step": 7752
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.226622356060291,
+      "learning_rate": 4.610791602657609e-06,
+      "loss": 0.4155,
+      "step": 7753
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.468108289681431,
+      "learning_rate": 4.609669010306349e-06,
+      "loss": 0.4186,
+      "step": 7754
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.317303566012148,
+      "learning_rate": 4.608546437751404e-06,
+      "loss": 0.3802,
+      "step": 7755
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.883446851141167,
+      "learning_rate": 4.6074238850497076e-06,
+      "loss": 0.4083,
+      "step": 7756
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.671465682564801,
+      "learning_rate": 4.606301352258192e-06,
+      "loss": 0.3735,
+      "step": 7757
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.071627379478876,
+      "learning_rate": 4.605178839433788e-06,
+      "loss": 0.3975,
+      "step": 7758
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.6805820447847744,
+      "learning_rate": 4.604056346633425e-06,
+      "loss": 0.4188,
+      "step": 7759
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.58328507123124,
+      "learning_rate": 4.602933873914034e-06,
+      "loss": 0.405,
+      "step": 7760
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3826704255759945,
+      "learning_rate": 4.601811421332543e-06,
+      "loss": 0.1757,
+      "step": 7761
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.942595725833063,
+      "learning_rate": 4.6006889889458784e-06,
+      "loss": 0.4264,
+      "step": 7762
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.478498369695445,
+      "learning_rate": 4.599566576810966e-06,
+      "loss": 0.4004,
+      "step": 7763
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.874860534768587,
+      "learning_rate": 4.598444184984731e-06,
+      "loss": 0.3907,
+      "step": 7764
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.150802234544424,
+      "learning_rate": 4.5973218135240985e-06,
+      "loss": 0.4142,
+      "step": 7765
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.311168296294149,
+      "learning_rate": 4.59619946248599e-06,
+      "loss": 0.357,
+      "step": 7766
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.231235165518236,
+      "learning_rate": 4.59507713192733e-06,
+      "loss": 0.4114,
+      "step": 7767
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.99014000638099,
+      "learning_rate": 4.593954821905036e-06,
+      "loss": 0.3968,
+      "step": 7768
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.860602412387214,
+      "learning_rate": 4.592832532476031e-06,
+      "loss": 0.406,
+      "step": 7769
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.88965267472235,
+      "learning_rate": 4.591710263697232e-06,
+      "loss": 0.3766,
+      "step": 7770
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.460398538260645,
+      "learning_rate": 4.5905880156255575e-06,
+      "loss": 0.4091,
+      "step": 7771
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.809364464294418,
+      "learning_rate": 4.589465788317923e-06,
+      "loss": 0.4154,
+      "step": 7772
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.369927870587357,
+      "learning_rate": 4.588343581831246e-06,
+      "loss": 0.3896,
+      "step": 7773
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.36372933562324633,
+      "learning_rate": 4.587221396222441e-06,
+      "loss": 0.2066,
+      "step": 7774
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9120566819773788,
+      "learning_rate": 4.586099231548421e-06,
+      "loss": 0.3981,
+      "step": 7775
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9949306309603227,
+      "learning_rate": 4.584977087866096e-06,
+      "loss": 0.4152,
+      "step": 7776
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.444599021225077,
+      "learning_rate": 4.583854965232383e-06,
+      "loss": 0.3999,
+      "step": 7777
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.163030363012195,
+      "learning_rate": 4.582732863704189e-06,
+      "loss": 0.4078,
+      "step": 7778
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.04096161138382,
+      "learning_rate": 4.581610783338424e-06,
+      "loss": 0.412,
+      "step": 7779
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.05421966753851,
+      "learning_rate": 4.580488724191995e-06,
+      "loss": 0.3694,
+      "step": 7780
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.253853587316453,
+      "learning_rate": 4.579366686321811e-06,
+      "loss": 0.3502,
+      "step": 7781
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.5029421225302535,
+      "learning_rate": 4.5782446697847775e-06,
+      "loss": 0.4084,
+      "step": 7782
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.473740239594705,
+      "learning_rate": 4.5771226746377985e-06,
+      "loss": 0.3875,
+      "step": 7783
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3644252590667338,
+      "learning_rate": 4.576000700937778e-06,
+      "loss": 0.1913,
+      "step": 7784
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.201678134988162,
+      "learning_rate": 4.574878748741621e-06,
+      "loss": 0.3883,
+      "step": 7785
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.584487376687964,
+      "learning_rate": 4.573756818106228e-06,
+      "loss": 0.3703,
+      "step": 7786
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3475953345357731,
+      "learning_rate": 4.5726349090885e-06,
+      "loss": 0.1958,
+      "step": 7787
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.349517123185602,
+      "learning_rate": 4.571513021745334e-06,
+      "loss": 0.4171,
+      "step": 7788
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.855828263555073,
+      "learning_rate": 4.570391156133631e-06,
+      "loss": 0.4104,
+      "step": 7789
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.45716216226089,
+      "learning_rate": 4.56926931231029e-06,
+      "loss": 0.4402,
+      "step": 7790
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.033624272284277,
+      "learning_rate": 4.568147490332203e-06,
+      "loss": 0.4341,
+      "step": 7791
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.006676313186082,
+      "learning_rate": 4.567025690256269e-06,
+      "loss": 0.3933,
+      "step": 7792
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.33553156737536804,
+      "learning_rate": 4.5659039121393805e-06,
+      "loss": 0.1852,
+      "step": 7793
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.639087059549869,
+      "learning_rate": 4.564782156038431e-06,
+      "loss": 0.4225,
+      "step": 7794
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.222551777158045,
+      "learning_rate": 4.563660422010308e-06,
+      "loss": 0.3587,
+      "step": 7795
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.3480384105262091,
+      "learning_rate": 4.562538710111909e-06,
+      "loss": 0.1773,
+      "step": 7796
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.943018956467897,
+      "learning_rate": 4.561417020400119e-06,
+      "loss": 0.3889,
+      "step": 7797
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.147130328627659,
+      "learning_rate": 4.56029535293183e-06,
+      "loss": 0.4008,
+      "step": 7798
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.827059303438836,
+      "learning_rate": 4.559173707763924e-06,
+      "loss": 0.4239,
+      "step": 7799
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.286663635871766,
+      "learning_rate": 4.558052084953292e-06,
+      "loss": 0.4217,
+      "step": 7800
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.477193237719916,
+      "learning_rate": 4.556930484556818e-06,
+      "loss": 0.4339,
+      "step": 7801
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.139765350938342,
+      "learning_rate": 4.5558089066313855e-06,
+      "loss": 0.3895,
+      "step": 7802
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.3982678532285155,
+      "learning_rate": 4.554687351233876e-06,
+      "loss": 0.3848,
+      "step": 7803
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.35730803415673,
+      "learning_rate": 4.553565818421173e-06,
+      "loss": 0.3989,
+      "step": 7804
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.551803479569781,
+      "learning_rate": 4.552444308250156e-06,
+      "loss": 0.4334,
+      "step": 7805
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.35013108221158445,
+      "learning_rate": 4.551322820777705e-06,
+      "loss": 0.1829,
+      "step": 7806
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.62317312022334,
+      "learning_rate": 4.550201356060695e-06,
+      "loss": 0.3552,
+      "step": 7807
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.422878082947484,
+      "learning_rate": 4.549079914156009e-06,
+      "loss": 0.4294,
+      "step": 7808
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.902306944736536,
+      "learning_rate": 4.5479584951205195e-06,
+      "loss": 0.4155,
+      "step": 7809
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 16.39917675831375,
+      "learning_rate": 4.546837099011101e-06,
+      "loss": 0.4014,
+      "step": 7810
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 13.368306063578968,
+      "learning_rate": 4.545715725884625e-06,
+      "loss": 0.4123,
+      "step": 7811
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9582562533079146,
+      "learning_rate": 4.544594375797969e-06,
+      "loss": 0.3804,
+      "step": 7812
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.018936310260227,
+      "learning_rate": 4.543473048808001e-06,
+      "loss": 0.4015,
+      "step": 7813
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.3478427747337935,
+      "learning_rate": 4.542351744971592e-06,
+      "loss": 0.3601,
+      "step": 7814
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.373465366665442,
+      "learning_rate": 4.541230464345609e-06,
+      "loss": 0.3876,
+      "step": 7815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.94148126212001,
+      "learning_rate": 4.540109206986923e-06,
+      "loss": 0.4036,
+      "step": 7816
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.715460676866276,
+      "learning_rate": 4.538987972952398e-06,
+      "loss": 0.3726,
+      "step": 7817
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 10.257892581277405,
+      "learning_rate": 4.537866762298901e-06,
+      "loss": 0.3964,
+      "step": 7818
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.915886429372999,
+      "learning_rate": 4.536745575083291e-06,
+      "loss": 0.4093,
+      "step": 7819
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.9231705943641435,
+      "learning_rate": 4.5356244113624394e-06,
+      "loss": 0.3868,
+      "step": 7820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 6.7767542130459075,
+      "learning_rate": 4.534503271193202e-06,
+      "loss": 0.3881,
+      "step": 7821
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.224251697542071,
+      "learning_rate": 4.533382154632442e-06,
+      "loss": 0.424,
+      "step": 7822
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.168608454557438,
+      "learning_rate": 4.532261061737017e-06,
+      "loss": 0.4235,
+      "step": 7823
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 7.323434819085391,
+      "learning_rate": 4.531139992563788e-06,
+      "loss": 0.4306,
+      "step": 7824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.027499069814009,
+      "learning_rate": 4.530018947169608e-06,
+      "loss": 0.4348,
+      "step": 7825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 17.130886385028187,
+      "learning_rate": 4.528897925611335e-06,
+      "loss": 0.374,
+      "step": 7826
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 9.204649428160392,
+      "learning_rate": 4.527776927945823e-06,
+      "loss": 0.4333,
+      "step": 7827
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.85559833992186,
+      "learning_rate": 4.526655954229926e-06,
+      "loss": 0.4528,
+      "step": 7828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.204777636645867,
+      "learning_rate": 4.525535004520496e-06,
+      "loss": 0.4143,
+      "step": 7829
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.36564592042116323,
+      "learning_rate": 4.524414078874385e-06,
+      "loss": 0.2005,
+      "step": 7830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.64637598172985,
+      "learning_rate": 4.5232931773484385e-06,
+      "loss": 0.396,
+      "step": 7831
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 18.890165702563515,
+      "learning_rate": 4.52217229999951e-06,
+      "loss": 0.3887,
+      "step": 7832
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 8.735445339012918,
+      "learning_rate": 4.521051446884445e-06,
+      "loss": 0.3858,
+      "step": 7833
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.008351175091351,
+      "learning_rate": 4.519930618060088e-06,
+      "loss": 0.3881,
+      "step": 7834
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.903155545996135,
+      "learning_rate": 4.518809813583284e-06,
+      "loss": 0.4053,
+      "step": 7835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 14.033223895463657,
+      "learning_rate": 4.517689033510881e-06,
+      "loss": 0.4044,
+      "step": 7836
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 11.94097927447596,
+      "learning_rate": 4.516568277899715e-06,
+      "loss": 0.3963,
+      "step": 7837
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.35874911906827683,
+      "learning_rate": 4.5154475468066315e-06,
+      "loss": 0.1852,
+      "step": 7838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 12.97145688086188,
+      "learning_rate": 4.514326840288466e-06,
+      "loss": 0.3884,
+      "step": 7839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3457525677734926,
+      "learning_rate": 4.513206158402061e-06,
+      "loss": 0.1787,
+      "step": 7840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.469368984022129,
+      "learning_rate": 4.512085501204254e-06,
+      "loss": 0.3656,
+      "step": 7841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.603221687600884,
+      "learning_rate": 4.510964868751879e-06,
+      "loss": 0.4504,
+      "step": 7842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.120315539253006,
+      "learning_rate": 4.509844261101769e-06,
+      "loss": 0.4274,
+      "step": 7843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.676443061585869,
+      "learning_rate": 4.508723678310763e-06,
+      "loss": 0.3824,
+      "step": 7844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.569066927733604,
+      "learning_rate": 4.50760312043569e-06,
+      "loss": 0.4494,
+      "step": 7845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.121514880176474,
+      "learning_rate": 4.506482587533382e-06,
+      "loss": 0.3603,
+      "step": 7846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.273599750973865,
+      "learning_rate": 4.5053620796606665e-06,
+      "loss": 0.3874,
+      "step": 7847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.826882347796291,
+      "learning_rate": 4.504241596874375e-06,
+      "loss": 0.4001,
+      "step": 7848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.570163818590559,
+      "learning_rate": 4.503121139231333e-06,
+      "loss": 0.3901,
+      "step": 7849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.559735695402663,
+      "learning_rate": 4.5020007067883675e-06,
+      "loss": 0.4048,
+      "step": 7850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.007011518662605,
+      "learning_rate": 4.500880299602301e-06,
+      "loss": 0.4074,
+      "step": 7851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.020239914533967,
+      "learning_rate": 4.499759917729959e-06,
+      "loss": 0.3876,
+      "step": 7852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.636504492747307,
+      "learning_rate": 4.498639561228164e-06,
+      "loss": 0.3777,
+      "step": 7853
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.399474289971665,
+      "learning_rate": 4.497519230153736e-06,
+      "loss": 0.4183,
+      "step": 7854
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.959880341799854,
+      "learning_rate": 4.496398924563492e-06,
+      "loss": 0.3808,
+      "step": 7855
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.629617326243105,
+      "learning_rate": 4.495278644514255e-06,
+      "loss": 0.4181,
+      "step": 7856
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5216339643508383,
+      "learning_rate": 4.49415839006284e-06,
+      "loss": 0.3575,
+      "step": 7857
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.286247770133685,
+      "learning_rate": 4.4930381612660605e-06,
+      "loss": 0.4105,
+      "step": 7858
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34095875804620424,
+      "learning_rate": 4.491917958180734e-06,
+      "loss": 0.1821,
+      "step": 7859
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4803866758709683,
+      "learning_rate": 4.490797780863672e-06,
+      "loss": 0.404,
+      "step": 7860
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.658547144146443,
+      "learning_rate": 4.489677629371688e-06,
+      "loss": 0.3535,
+      "step": 7861
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.904916375501904,
+      "learning_rate": 4.488557503761588e-06,
+      "loss": 0.4037,
+      "step": 7862
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.890427728684775,
+      "learning_rate": 4.4874374040901855e-06,
+      "loss": 0.402,
+      "step": 7863
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3567110618515549,
+      "learning_rate": 4.486317330414287e-06,
+      "loss": 0.1961,
+      "step": 7864
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 10.054698798563443,
+      "learning_rate": 4.485197282790698e-06,
+      "loss": 0.3716,
+      "step": 7865
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.817163605993874,
+      "learning_rate": 4.484077261276224e-06,
+      "loss": 0.3992,
+      "step": 7866
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 18.54104301441358,
+      "learning_rate": 4.48295726592767e-06,
+      "loss": 0.4257,
+      "step": 7867
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6390042778944225,
+      "learning_rate": 4.481837296801839e-06,
+      "loss": 0.4031,
+      "step": 7868
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.943102318797405,
+      "learning_rate": 4.48071735395553e-06,
+      "loss": 0.4067,
+      "step": 7869
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.761930619380596,
+      "learning_rate": 4.479597437445544e-06,
+      "loss": 0.3844,
+      "step": 7870
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.7838156751188645,
+      "learning_rate": 4.47847754732868e-06,
+      "loss": 0.4082,
+      "step": 7871
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.5593329122044235,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.3917,
+      "step": 7872
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.659136143652808,
+      "learning_rate": 4.476237846501502e-06,
+      "loss": 0.4107,
+      "step": 7873
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.792509391507792,
+      "learning_rate": 4.475118035904778e-06,
+      "loss": 0.3873,
+      "step": 7874
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.0892869798068885,
+      "learning_rate": 4.473998251928357e-06,
+      "loss": 0.3993,
+      "step": 7875
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 22.55021392378995,
+      "learning_rate": 4.4728784946290305e-06,
+      "loss": 0.3996,
+      "step": 7876
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3568391857290848,
+      "learning_rate": 4.471758764063588e-06,
+      "loss": 0.1808,
+      "step": 7877
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.853293427648464,
+      "learning_rate": 4.470639060288818e-06,
+      "loss": 0.4067,
+      "step": 7878
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.206377160365185,
+      "learning_rate": 4.469519383361508e-06,
+      "loss": 0.4211,
+      "step": 7879
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 14.93844658722292,
+      "learning_rate": 4.468399733338447e-06,
+      "loss": 0.4471,
+      "step": 7880
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.776927432323838,
+      "learning_rate": 4.4672801102764195e-06,
+      "loss": 0.4026,
+      "step": 7881
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.210043008592396,
+      "learning_rate": 4.466160514232206e-06,
+      "loss": 0.3703,
+      "step": 7882
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.013283499563583,
+      "learning_rate": 4.465040945262592e-06,
+      "loss": 0.3796,
+      "step": 7883
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.707252243738331,
+      "learning_rate": 4.463921403424357e-06,
+      "loss": 0.3937,
+      "step": 7884
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.576171696494141,
+      "learning_rate": 4.462801888774281e-06,
+      "loss": 0.4294,
+      "step": 7885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7532482378324796,
+      "learning_rate": 4.4616824013691395e-06,
+      "loss": 0.3816,
+      "step": 7886
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.652065176731428,
+      "learning_rate": 4.4605629412657145e-06,
+      "loss": 0.3806,
+      "step": 7887
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.881445084785621,
+      "learning_rate": 4.459443508520778e-06,
+      "loss": 0.4004,
+      "step": 7888
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.36920008171509117,
+      "learning_rate": 4.458324103191103e-06,
+      "loss": 0.1879,
+      "step": 7889
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.4458816470594575,
+      "learning_rate": 4.4572047253334645e-06,
+      "loss": 0.3844,
+      "step": 7890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4388846719410466,
+      "learning_rate": 4.456085375004633e-06,
+      "loss": 0.4131,
+      "step": 7891
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.067418517157954,
+      "learning_rate": 4.454966052261378e-06,
+      "loss": 0.3883,
+      "step": 7892
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6277011337057132,
+      "learning_rate": 4.4538467571604676e-06,
+      "loss": 0.382,
+      "step": 7893
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.218136212012108,
+      "learning_rate": 4.452727489758668e-06,
+      "loss": 0.4174,
+      "step": 7894
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8147641587136034,
+      "learning_rate": 4.451608250112747e-06,
+      "loss": 0.4179,
+      "step": 7895
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.294424381088363,
+      "learning_rate": 4.450489038279468e-06,
+      "loss": 0.4275,
+      "step": 7896
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.45865906992868,
+      "learning_rate": 4.449369854315594e-06,
+      "loss": 0.3914,
+      "step": 7897
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.654572618402616,
+      "learning_rate": 4.448250698277883e-06,
+      "loss": 0.4303,
+      "step": 7898
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.350842285406544,
+      "learning_rate": 4.447131570223099e-06,
+      "loss": 0.4526,
+      "step": 7899
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.9309632502745355,
+      "learning_rate": 4.446012470208001e-06,
+      "loss": 0.3964,
+      "step": 7900
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.086669511319041,
+      "learning_rate": 4.4448933982893435e-06,
+      "loss": 0.4088,
+      "step": 7901
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.080701010214851,
+      "learning_rate": 4.443774354523883e-06,
+      "loss": 0.3796,
+      "step": 7902
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8495879574189193,
+      "learning_rate": 4.442655338968373e-06,
+      "loss": 0.397,
+      "step": 7903
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.792624232298812,
+      "learning_rate": 4.441536351679569e-06,
+      "loss": 0.3816,
+      "step": 7904
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.22629702126163,
+      "learning_rate": 4.44041739271422e-06,
+      "loss": 0.4144,
+      "step": 7905
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8424358440911974,
+      "learning_rate": 4.439298462129074e-06,
+      "loss": 0.4062,
+      "step": 7906
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6575102155446517,
+      "learning_rate": 4.438179559980885e-06,
+      "loss": 0.3682,
+      "step": 7907
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.521832306942887,
+      "learning_rate": 4.437060686326397e-06,
+      "loss": 0.392,
+      "step": 7908
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.109640339780308,
+      "learning_rate": 4.435941841222354e-06,
+      "loss": 0.401,
+      "step": 7909
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.070341328172731,
+      "learning_rate": 4.4348230247255015e-06,
+      "loss": 0.3824,
+      "step": 7910
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.211695823704992,
+      "learning_rate": 4.433704236892584e-06,
+      "loss": 0.386,
+      "step": 7911
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1712530490077864,
+      "learning_rate": 4.432585477780342e-06,
+      "loss": 0.4035,
+      "step": 7912
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.8141437332113255,
+      "learning_rate": 4.431466747445514e-06,
+      "loss": 0.3938,
+      "step": 7913
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.918241561462104,
+      "learning_rate": 4.4303480459448384e-06,
+      "loss": 0.4079,
+      "step": 7914
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.649455633703835,
+      "learning_rate": 4.429229373335054e-06,
+      "loss": 0.3651,
+      "step": 7915
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.947931968136595,
+      "learning_rate": 4.428110729672894e-06,
+      "loss": 0.3728,
+      "step": 7916
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.261428494126607,
+      "learning_rate": 4.426992115015094e-06,
+      "loss": 0.3711,
+      "step": 7917
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.49294366027929,
+      "learning_rate": 4.425873529418384e-06,
+      "loss": 0.3993,
+      "step": 7918
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 11.318998442302581,
+      "learning_rate": 4.424754972939498e-06,
+      "loss": 0.3892,
+      "step": 7919
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.882118312438397,
+      "learning_rate": 4.423636445635165e-06,
+      "loss": 0.4569,
+      "step": 7920
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7776597844067257,
+      "learning_rate": 4.42251794756211e-06,
+      "loss": 0.3681,
+      "step": 7921
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.379592408999679,
+      "learning_rate": 4.421399478777064e-06,
+      "loss": 0.3675,
+      "step": 7922
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.107951700127315,
+      "learning_rate": 4.420281039336751e-06,
+      "loss": 0.3844,
+      "step": 7923
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.720570337682006,
+      "learning_rate": 4.419162629297894e-06,
+      "loss": 0.405,
+      "step": 7924
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.819722833578309,
+      "learning_rate": 4.418044248717212e-06,
+      "loss": 0.3645,
+      "step": 7925
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.649911801957676,
+      "learning_rate": 4.41692589765143e-06,
+      "loss": 0.3934,
+      "step": 7926
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.1415440568105946,
+      "learning_rate": 4.415807576157267e-06,
+      "loss": 0.3928,
+      "step": 7927
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.093146186555613,
+      "learning_rate": 4.414689284291438e-06,
+      "loss": 0.4008,
+      "step": 7928
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.351280040574338,
+      "learning_rate": 4.413571022110658e-06,
+      "loss": 0.3825,
+      "step": 7929
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.737845257898391,
+      "learning_rate": 4.4124527896716466e-06,
+      "loss": 0.4085,
+      "step": 7930
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.205274002795756,
+      "learning_rate": 4.411334587031113e-06,
+      "loss": 0.3866,
+      "step": 7931
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.705792638861901,
+      "learning_rate": 4.410216414245771e-06,
+      "loss": 0.4221,
+      "step": 7932
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.604853992027241,
+      "learning_rate": 4.409098271372328e-06,
+      "loss": 0.4415,
+      "step": 7933
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.782795989027395,
+      "learning_rate": 4.4079801584674955e-06,
+      "loss": 0.3991,
+      "step": 7934
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5712997775262343,
+      "learning_rate": 4.40686207558798e-06,
+      "loss": 0.3863,
+      "step": 7935
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.07475136365106,
+      "learning_rate": 4.405744022790485e-06,
+      "loss": 0.4298,
+      "step": 7936
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.254311211063186,
+      "learning_rate": 4.404626000131716e-06,
+      "loss": 0.399,
+      "step": 7937
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.430728646975824,
+      "learning_rate": 4.403508007668375e-06,
+      "loss": 0.4335,
+      "step": 7938
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.5488206285664274,
+      "learning_rate": 4.402390045457165e-06,
+      "loss": 0.4356,
+      "step": 7939
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.734324169892457,
+      "learning_rate": 4.401272113554782e-06,
+      "loss": 0.3678,
+      "step": 7940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7064351839612124,
+      "learning_rate": 4.400154212017923e-06,
+      "loss": 0.4315,
+      "step": 7941
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.191693446665543,
+      "learning_rate": 4.39903634090329e-06,
+      "loss": 0.4033,
+      "step": 7942
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.272193854175181,
+      "learning_rate": 4.397918500267575e-06,
+      "loss": 0.3558,
+      "step": 7943
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 205.27137857407513,
+      "learning_rate": 4.39680069016747e-06,
+      "loss": 0.4029,
+      "step": 7944
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.94796851022187,
+      "learning_rate": 4.395682910659666e-06,
+      "loss": 0.4154,
+      "step": 7945
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.282041113247016,
+      "learning_rate": 4.3945651618008565e-06,
+      "loss": 0.3985,
+      "step": 7946
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.297204152942186,
+      "learning_rate": 4.393447443647726e-06,
+      "loss": 0.4334,
+      "step": 7947
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.271332708283157,
+      "learning_rate": 4.3923297562569655e-06,
+      "loss": 0.4157,
+      "step": 7948
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.744336238472644,
+      "learning_rate": 4.391212099685258e-06,
+      "loss": 0.4133,
+      "step": 7949
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.8065835928244995,
+      "learning_rate": 4.390094473989288e-06,
+      "loss": 0.3585,
+      "step": 7950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35796475642623626,
+      "learning_rate": 4.388976879225739e-06,
+      "loss": 0.199,
+      "step": 7951
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.835146789338563,
+      "learning_rate": 4.38785931545129e-06,
+      "loss": 0.3936,
+      "step": 7952
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 13.260537353892584,
+      "learning_rate": 4.386741782722618e-06,
+      "loss": 0.4295,
+      "step": 7953
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3564226963785933,
+      "learning_rate": 4.385624281096407e-06,
+      "loss": 0.4158,
+      "step": 7954
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.4203639909553663,
+      "learning_rate": 4.384506810629329e-06,
+      "loss": 0.398,
+      "step": 7955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.987694857456762,
+      "learning_rate": 4.383389371378059e-06,
+      "loss": 0.3872,
+      "step": 7956
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 5.513013639537281,
+      "learning_rate": 4.382271963399268e-06,
+      "loss": 0.3734,
+      "step": 7957
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.03910480277697,
+      "learning_rate": 4.381154586749631e-06,
+      "loss": 0.4333,
+      "step": 7958
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.769650671786968,
+      "learning_rate": 4.380037241485815e-06,
+      "loss": 0.3902,
+      "step": 7959
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.248011821612291,
+      "learning_rate": 4.37891992766449e-06,
+      "loss": 0.3962,
+      "step": 7960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.894662064091779,
+      "learning_rate": 4.377802645342318e-06,
+      "loss": 0.3991,
+      "step": 7961
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.384113258275228,
+      "learning_rate": 4.376685394575971e-06,
+      "loss": 0.4383,
+      "step": 7962
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 8.268920296954223,
+      "learning_rate": 4.375568175422108e-06,
+      "loss": 0.3997,
+      "step": 7963
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.344553402268197,
+      "learning_rate": 4.374450987937391e-06,
+      "loss": 0.4018,
+      "step": 7964
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 9.536573707703667,
+      "learning_rate": 4.373333832178478e-06,
+      "loss": 0.4163,
+      "step": 7965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0740915658638657,
+      "learning_rate": 4.372216708202034e-06,
+      "loss": 0.3977,
+      "step": 7966
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.747848234292047,
+      "learning_rate": 4.37109961606471e-06,
+      "loss": 0.3533,
+      "step": 7967
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.6471477015490046,
+      "learning_rate": 4.3699825558231636e-06,
+      "loss": 0.3731,
+      "step": 7968
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 15.205180941674538,
+      "learning_rate": 4.368865527534047e-06,
+      "loss": 0.4266,
+      "step": 7969
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.5288863277955773,
+      "learning_rate": 4.367748531254015e-06,
+      "loss": 0.3767,
+      "step": 7970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.35764436758414997,
+      "learning_rate": 4.366631567039715e-06,
+      "loss": 0.1865,
+      "step": 7971
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.901761287676857,
+      "learning_rate": 4.365514634947798e-06,
+      "loss": 0.4132,
+      "step": 7972
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.830295111008195,
+      "learning_rate": 4.364397735034908e-06,
+      "loss": 0.3657,
+      "step": 7973
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.137387427788084,
+      "learning_rate": 4.363280867357694e-06,
+      "loss": 0.3918,
+      "step": 7974
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.3959630629076347,
+      "learning_rate": 4.362164031972799e-06,
+      "loss": 0.3935,
+      "step": 7975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.34056687142666214,
+      "learning_rate": 4.361047228936865e-06,
+      "loss": 0.1886,
+      "step": 7976
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.6973625298293715,
+      "learning_rate": 4.35993045830653e-06,
+      "loss": 0.3831,
+      "step": 7977
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.7699182056599265,
+      "learning_rate": 4.358813720138438e-06,
+      "loss": 0.432,
+      "step": 7978
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 23.29041690527599,
+      "learning_rate": 4.3576970144892236e-06,
+      "loss": 0.4265,
+      "step": 7979
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.9975212746620143,
+      "learning_rate": 4.356580341415522e-06,
+      "loss": 0.4396,
+      "step": 7980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 6.2123298616665075,
+      "learning_rate": 4.355463700973968e-06,
+      "loss": 0.4212,
+      "step": 7981
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.060335161572872,
+      "learning_rate": 4.354347093221194e-06,
+      "loss": 0.4318,
+      "step": 7982
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.600032937052077,
+      "learning_rate": 4.35323051821383e-06,
+      "loss": 0.4417,
+      "step": 7983
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3649067980263685,
+      "learning_rate": 4.352113976008504e-06,
+      "loss": 0.1822,
+      "step": 7984
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.366316359599057,
+      "learning_rate": 4.350997466661847e-06,
+      "loss": 0.3966,
+      "step": 7985
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.318347091364994,
+      "learning_rate": 4.349880990230483e-06,
+      "loss": 0.3805,
+      "step": 7986
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.638432499162179,
+      "learning_rate": 4.348764546771035e-06,
+      "loss": 0.4223,
+      "step": 7987
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.07047086562782,
+      "learning_rate": 4.347648136340124e-06,
+      "loss": 0.406,
+      "step": 7988
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.482997016386461,
+      "learning_rate": 4.346531758994375e-06,
+      "loss": 0.4081,
+      "step": 7989
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.424471720536896,
+      "learning_rate": 4.345415414790405e-06,
+      "loss": 0.4051,
+      "step": 7990
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.327507624492921,
+      "learning_rate": 4.344299103784831e-06,
+      "loss": 0.3644,
+      "step": 7991
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.736172216838679,
+      "learning_rate": 4.343182826034268e-06,
+      "loss": 0.4076,
+      "step": 7992
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3471765128832978,
+      "learning_rate": 4.342066581595331e-06,
+      "loss": 0.4094,
+      "step": 7993
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.879314054885762,
+      "learning_rate": 4.3409503705246325e-06,
+      "loss": 0.3961,
+      "step": 7994
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.391864530472838,
+      "learning_rate": 4.339834192878782e-06,
+      "loss": 0.3716,
+      "step": 7995
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.299583618964692,
+      "learning_rate": 4.3387180487143875e-06,
+      "loss": 0.4205,
+      "step": 7996
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.320527031755656,
+      "learning_rate": 4.33760193808806e-06,
+      "loss": 0.4521,
+      "step": 7997
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.326213672436449,
+      "learning_rate": 4.336485861056402e-06,
+      "loss": 0.3985,
+      "step": 7998
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.18725368800589,
+      "learning_rate": 4.335369817676018e-06,
+      "loss": 0.4109,
+      "step": 7999
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36890020952221664,
+      "learning_rate": 4.334253808003508e-06,
+      "loss": 0.1927,
+      "step": 8000
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.437647866767816,
+      "learning_rate": 4.333137832095477e-06,
+      "loss": 0.3882,
+      "step": 8001
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.668134871883089,
+      "learning_rate": 4.33202189000852e-06,
+      "loss": 0.3937,
+      "step": 8002
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4496856489344956,
+      "learning_rate": 4.3309059817992365e-06,
+      "loss": 0.4037,
+      "step": 8003
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.303551554339067,
+      "learning_rate": 4.329790107524218e-06,
+      "loss": 0.4027,
+      "step": 8004
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5034014712945707,
+      "learning_rate": 4.3286742672400615e-06,
+      "loss": 0.3787,
+      "step": 8005
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.774137416079403,
+      "learning_rate": 4.3275584610033575e-06,
+      "loss": 0.4009,
+      "step": 8006
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0549784954023624,
+      "learning_rate": 4.326442688870697e-06,
+      "loss": 0.3713,
+      "step": 8007
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.889881268335535,
+      "learning_rate": 4.325326950898664e-06,
+      "loss": 0.4145,
+      "step": 8008
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.987657222885691,
+      "learning_rate": 4.3242112471438515e-06,
+      "loss": 0.3882,
+      "step": 8009
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 19.642865464152283,
+      "learning_rate": 4.323095577662841e-06,
+      "loss": 0.3676,
+      "step": 8010
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.850758585801613,
+      "learning_rate": 4.321979942512216e-06,
+      "loss": 0.3927,
+      "step": 8011
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.932760621773938,
+      "learning_rate": 4.320864341748557e-06,
+      "loss": 0.3771,
+      "step": 8012
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 8.921656333231825,
+      "learning_rate": 4.3197487754284444e-06,
+      "loss": 0.3856,
+      "step": 8013
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.570974152174638,
+      "learning_rate": 4.318633243608457e-06,
+      "loss": 0.4108,
+      "step": 8014
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.869755672768588,
+      "learning_rate": 4.31751774634517e-06,
+      "loss": 0.4472,
+      "step": 8015
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.121248371335664,
+      "learning_rate": 4.316402283695156e-06,
+      "loss": 0.4359,
+      "step": 8016
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.406255294534557,
+      "learning_rate": 4.315286855714992e-06,
+      "loss": 0.3908,
+      "step": 8017
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.996653975622225,
+      "learning_rate": 4.314171462461246e-06,
+      "loss": 0.4073,
+      "step": 8018
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.50635272271622,
+      "learning_rate": 4.313056103990487e-06,
+      "loss": 0.4107,
+      "step": 8019
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.461818055651024,
+      "learning_rate": 4.311940780359281e-06,
+      "loss": 0.3932,
+      "step": 8020
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.985246384512811,
+      "learning_rate": 4.310825491624198e-06,
+      "loss": 0.3842,
+      "step": 8021
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.854238722197796,
+      "learning_rate": 4.3097102378417985e-06,
+      "loss": 0.4023,
+      "step": 8022
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.082367269353292,
+      "learning_rate": 4.308595019068646e-06,
+      "loss": 0.3735,
+      "step": 8023
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.245262168913023,
+      "learning_rate": 4.307479835361298e-06,
+      "loss": 0.4352,
+      "step": 8024
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.050080070590322,
+      "learning_rate": 4.3063646867763175e-06,
+      "loss": 0.4352,
+      "step": 8025
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.633071662824895,
+      "learning_rate": 4.305249573370258e-06,
+      "loss": 0.4055,
+      "step": 8026
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.760346096402295,
+      "learning_rate": 4.304134495199675e-06,
+      "loss": 0.3776,
+      "step": 8027
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8932120716843115,
+      "learning_rate": 4.3030194523211195e-06,
+      "loss": 0.4517,
+      "step": 8028
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.178002418704676,
+      "learning_rate": 4.301904444791147e-06,
+      "loss": 0.3756,
+      "step": 8029
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.138472629984705,
+      "learning_rate": 4.300789472666306e-06,
+      "loss": 0.3738,
+      "step": 8030
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3551879233227728,
+      "learning_rate": 4.299674536003142e-06,
+      "loss": 0.1788,
+      "step": 8031
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.926456761222759,
+      "learning_rate": 4.298559634858202e-06,
+      "loss": 0.3937,
+      "step": 8032
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.861580187820664,
+      "learning_rate": 4.297444769288031e-06,
+      "loss": 0.3856,
+      "step": 8033
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8773524685971004,
+      "learning_rate": 4.296329939349172e-06,
+      "loss": 0.3772,
+      "step": 8034
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.24487059304446,
+      "learning_rate": 4.295215145098164e-06,
+      "loss": 0.3985,
+      "step": 8035
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5199118883595997,
+      "learning_rate": 4.2941003865915445e-06,
+      "loss": 0.4281,
+      "step": 8036
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.34422999216080863,
+      "learning_rate": 4.292985663885854e-06,
+      "loss": 0.1858,
+      "step": 8037
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 9.698177894399528,
+      "learning_rate": 4.291870977037625e-06,
+      "loss": 0.4402,
+      "step": 8038
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.639931377659572,
+      "learning_rate": 4.290756326103392e-06,
+      "loss": 0.3765,
+      "step": 8039
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.283595397147811,
+      "learning_rate": 4.289641711139683e-06,
+      "loss": 0.4066,
+      "step": 8040
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.240282844008282,
+      "learning_rate": 4.288527132203033e-06,
+      "loss": 0.4367,
+      "step": 8041
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.9373891769962635,
+      "learning_rate": 4.287412589349967e-06,
+      "loss": 0.4091,
+      "step": 8042
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 20.289660779634417,
+      "learning_rate": 4.286298082637012e-06,
+      "loss": 0.3975,
+      "step": 8043
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.357126332210596,
+      "learning_rate": 4.285183612120689e-06,
+      "loss": 0.3905,
+      "step": 8044
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.842017417325781,
+      "learning_rate": 4.284069177857525e-06,
+      "loss": 0.3895,
+      "step": 8045
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.5236705086651146,
+      "learning_rate": 4.282954779904038e-06,
+      "loss": 0.3951,
+      "step": 8046
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6774439261402003,
+      "learning_rate": 4.2818404183167455e-06,
+      "loss": 0.4091,
+      "step": 8047
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.646274873729503,
+      "learning_rate": 4.280726093152167e-06,
+      "loss": 0.3685,
+      "step": 8048
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.528377719189188,
+      "learning_rate": 4.2796118044668165e-06,
+      "loss": 0.3928,
+      "step": 8049
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 48.222205055731685,
+      "learning_rate": 4.2784975523172064e-06,
+      "loss": 0.3577,
+      "step": 8050
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7247507645006195,
+      "learning_rate": 4.277383336759846e-06,
+      "loss": 0.3916,
+      "step": 8051
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.137557121933893,
+      "learning_rate": 4.2762691578512485e-06,
+      "loss": 0.4104,
+      "step": 8052
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.514200156607976,
+      "learning_rate": 4.27515501564792e-06,
+      "loss": 0.408,
+      "step": 8053
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 26.63195698464562,
+      "learning_rate": 4.2740409102063665e-06,
+      "loss": 0.3656,
+      "step": 8054
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.775048091124067,
+      "learning_rate": 4.2729268415830895e-06,
+      "loss": 0.4068,
+      "step": 8055
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8999056428629815,
+      "learning_rate": 4.271812809834594e-06,
+      "loss": 0.356,
+      "step": 8056
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.332254645972968,
+      "learning_rate": 4.270698815017379e-06,
+      "loss": 0.4315,
+      "step": 8057
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3643596625842111,
+      "learning_rate": 4.269584857187942e-06,
+      "loss": 0.1891,
+      "step": 8058
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.250489378656265,
+      "learning_rate": 4.26847093640278e-06,
+      "loss": 0.398,
+      "step": 8059
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.35181044122703076,
+      "learning_rate": 4.267357052718387e-06,
+      "loss": 0.1816,
+      "step": 8060
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.208263038749157,
+      "learning_rate": 4.266243206191256e-06,
+      "loss": 0.3691,
+      "step": 8061
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2824390337473166,
+      "learning_rate": 4.265129396877878e-06,
+      "loss": 0.343,
+      "step": 8062
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.441129705759765,
+      "learning_rate": 4.264015624834738e-06,
+      "loss": 0.3962,
+      "step": 8063
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.044918244919765,
+      "learning_rate": 4.26290189011833e-06,
+      "loss": 0.3804,
+      "step": 8064
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.319833997264082,
+      "learning_rate": 4.261788192785133e-06,
+      "loss": 0.3876,
+      "step": 8065
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.115736770567409,
+      "learning_rate": 4.2606745328916335e-06,
+      "loss": 0.4031,
+      "step": 8066
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.69299222317162,
+      "learning_rate": 4.2595609104943095e-06,
+      "loss": 0.4024,
+      "step": 8067
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0673070589475713,
+      "learning_rate": 4.258447325649643e-06,
+      "loss": 0.4026,
+      "step": 8068
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3224640114769475,
+      "learning_rate": 4.25733377841411e-06,
+      "loss": 0.3745,
+      "step": 8069
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 13.354748995326108,
+      "learning_rate": 4.256220268844188e-06,
+      "loss": 0.4219,
+      "step": 8070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.878391134620169,
+      "learning_rate": 4.255106796996348e-06,
+      "loss": 0.3533,
+      "step": 8071
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6727099720011354,
+      "learning_rate": 4.253993362927063e-06,
+      "loss": 0.4195,
+      "step": 8072
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.460733109475914,
+      "learning_rate": 4.2528799666928034e-06,
+      "loss": 0.3706,
+      "step": 8073
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.652617747952218,
+      "learning_rate": 4.251766608350035e-06,
+      "loss": 0.3668,
+      "step": 8074
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.050870563109905,
+      "learning_rate": 4.250653287955224e-06,
+      "loss": 0.3849,
+      "step": 8075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37582977885592694,
+      "learning_rate": 4.249540005564837e-06,
+      "loss": 0.1976,
+      "step": 8076
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.2045895633225925,
+      "learning_rate": 4.2484267612353335e-06,
+      "loss": 0.3697,
+      "step": 8077
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.3166831563437236,
+      "learning_rate": 4.247313555023175e-06,
+      "loss": 0.4125,
+      "step": 8078
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.45100639695611,
+      "learning_rate": 4.246200386984818e-06,
+      "loss": 0.3766,
+      "step": 8079
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.99844855419848,
+      "learning_rate": 4.245087257176721e-06,
+      "loss": 0.3588,
+      "step": 8080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.569965111001408,
+      "learning_rate": 4.243974165655336e-06,
+      "loss": 0.4022,
+      "step": 8081
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 16.387622505893475,
+      "learning_rate": 4.2428611124771184e-06,
+      "loss": 0.41,
+      "step": 8082
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.37490750022162656,
+      "learning_rate": 4.241748097698514e-06,
+      "loss": 0.1844,
+      "step": 8083
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.8257613860627235,
+      "learning_rate": 4.240635121375975e-06,
+      "loss": 0.3931,
+      "step": 8084
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.1723110122801588,
+      "learning_rate": 4.239522183565949e-06,
+      "loss": 0.4097,
+      "step": 8085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.152857159418727,
+      "learning_rate": 4.238409284324877e-06,
+      "loss": 0.4046,
+      "step": 8086
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2523250489964632,
+      "learning_rate": 4.2372964237092015e-06,
+      "loss": 0.4234,
+      "step": 8087
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 21.617431339620307,
+      "learning_rate": 4.2361836017753675e-06,
+      "loss": 0.3923,
+      "step": 8088
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.755647220507085,
+      "learning_rate": 4.23507081857981e-06,
+      "loss": 0.4532,
+      "step": 8089
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8472972624634574,
+      "learning_rate": 4.233958074178968e-06,
+      "loss": 0.375,
+      "step": 8090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.666122822597315,
+      "learning_rate": 4.232845368629273e-06,
+      "loss": 0.3861,
+      "step": 8091
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.204849854066963,
+      "learning_rate": 4.2317327019871614e-06,
+      "loss": 0.3884,
+      "step": 8092
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.091092403730136,
+      "learning_rate": 4.230620074309062e-06,
+      "loss": 0.4234,
+      "step": 8093
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.196709695501056,
+      "learning_rate": 4.229507485651404e-06,
+      "loss": 0.3659,
+      "step": 8094
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7844292735549687,
+      "learning_rate": 4.228394936070613e-06,
+      "loss": 0.4023,
+      "step": 8095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7316502834766396,
+      "learning_rate": 4.227282425623117e-06,
+      "loss": 0.4195,
+      "step": 8096
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.903280086034013,
+      "learning_rate": 4.226169954365337e-06,
+      "loss": 0.3997,
+      "step": 8097
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.29946513650802,
+      "learning_rate": 4.225057522353694e-06,
+      "loss": 0.3752,
+      "step": 8098
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 6.6559252906024815,
+      "learning_rate": 4.223945129644606e-06,
+      "loss": 0.3942,
+      "step": 8099
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.5484741345080395,
+      "learning_rate": 4.222832776294491e-06,
+      "loss": 0.412,
+      "step": 8100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 10.879035941272754,
+      "learning_rate": 4.221720462359765e-06,
+      "loss": 0.393,
+      "step": 8101
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.8137770200917167,
+      "learning_rate": 4.22060818789684e-06,
+      "loss": 0.3845,
+      "step": 8102
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 7.0387665943442785,
+      "learning_rate": 4.219495952962126e-06,
+      "loss": 0.4312,
+      "step": 8103
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7898567715167393,
+      "learning_rate": 4.218383757612033e-06,
+      "loss": 0.4087,
+      "step": 8104
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 56.81381655040793,
+      "learning_rate": 4.217271601902968e-06,
+      "loss": 0.4154,
+      "step": 8105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.113933027848267,
+      "learning_rate": 4.216159485891336e-06,
+      "loss": 0.4,
+      "step": 8106
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.804669151289418,
+      "learning_rate": 4.2150474096335356e-06,
+      "loss": 0.3845,
+      "step": 8107
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0478094247747456,
+      "learning_rate": 4.213935373185975e-06,
+      "loss": 0.4158,
+      "step": 8108
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.2722409120665708,
+      "learning_rate": 4.21282337660505e-06,
+      "loss": 0.4372,
+      "step": 8109
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.172181440754666,
+      "learning_rate": 4.211711419947156e-06,
+      "loss": 0.381,
+      "step": 8110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9274765701169807,
+      "learning_rate": 4.210599503268686e-06,
+      "loss": 0.3635,
+      "step": 8111
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.119167535997438,
+      "learning_rate": 4.209487626626039e-06,
+      "loss": 0.4084,
+      "step": 8112
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.791416332944095,
+      "learning_rate": 4.208375790075603e-06,
+      "loss": 0.3852,
+      "step": 8113
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.616158667050347,
+      "learning_rate": 4.2072639936737635e-06,
+      "loss": 0.3843,
+      "step": 8114
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.349879321592374,
+      "learning_rate": 4.206152237476912e-06,
+      "loss": 0.4088,
+      "step": 8115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.122287773913753,
+      "learning_rate": 4.20504052154143e-06,
+      "loss": 0.3827,
+      "step": 8116
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.043983345523658,
+      "learning_rate": 4.203928845923701e-06,
+      "loss": 0.3591,
+      "step": 8117
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.856180836109949,
+      "learning_rate": 4.202817210680103e-06,
+      "loss": 0.3698,
+      "step": 8118
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.442060860826741,
+      "learning_rate": 4.201705615867021e-06,
+      "loss": 0.3883,
+      "step": 8119
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.6666923385876693,
+      "learning_rate": 4.200594061540827e-06,
+      "loss": 0.3839,
+      "step": 8120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.816288504954941,
+      "learning_rate": 4.199482547757895e-06,
+      "loss": 0.3995,
+      "step": 8121
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.7640648104006704,
+      "learning_rate": 4.198371074574597e-06,
+      "loss": 0.3833,
+      "step": 8122
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.74588021066646,
+      "learning_rate": 4.197259642047306e-06,
+      "loss": 0.427,
+      "step": 8123
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.0955655138887046,
+      "learning_rate": 4.19614825023239e-06,
+      "loss": 0.3946,
+      "step": 8124
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 5.791594036072617,
+      "learning_rate": 4.195036899186213e-06,
+      "loss": 0.3868,
+      "step": 8125
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9847690314939896,
+      "learning_rate": 4.19392558896514e-06,
+      "loss": 0.3976,
+      "step": 8126
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4515011797441715,
+      "learning_rate": 4.192814319625534e-06,
+      "loss": 0.3615,
+      "step": 8127
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 17.349461420893693,
+      "learning_rate": 4.191703091223753e-06,
+      "loss": 0.3978,
+      "step": 8128
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7739551938539395,
+      "learning_rate": 4.190591903816157e-06,
+      "loss": 0.3902,
+      "step": 8129
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2486463489430513,
+      "learning_rate": 4.189480757459099e-06,
+      "loss": 0.4203,
+      "step": 8130
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.657863132262031,
+      "learning_rate": 4.188369652208936e-06,
+      "loss": 0.3691,
+      "step": 8131
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.384426914141851,
+      "learning_rate": 4.187258588122019e-06,
+      "loss": 0.3765,
+      "step": 8132
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.217082931429477,
+      "learning_rate": 4.186147565254696e-06,
+      "loss": 0.4006,
+      "step": 8133
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.942629331225813,
+      "learning_rate": 4.185036583663314e-06,
+      "loss": 0.4136,
+      "step": 8134
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.113906712788433,
+      "learning_rate": 4.183925643404222e-06,
+      "loss": 0.3915,
+      "step": 8135
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35010006980100244,
+      "learning_rate": 4.182814744533758e-06,
+      "loss": 0.1884,
+      "step": 8136
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.21710578086457,
+      "learning_rate": 4.181703887108268e-06,
+      "loss": 0.4409,
+      "step": 8137
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.613857524359582,
+      "learning_rate": 4.180593071184087e-06,
+      "loss": 0.3823,
+      "step": 8138
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.805494080795436,
+      "learning_rate": 4.179482296817557e-06,
+      "loss": 0.4135,
+      "step": 8139
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 16.429716573354934,
+      "learning_rate": 4.178371564065008e-06,
+      "loss": 0.4251,
+      "step": 8140
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.420115830191793,
+      "learning_rate": 4.177260872982774e-06,
+      "loss": 0.4067,
+      "step": 8141
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4521355767312496,
+      "learning_rate": 4.176150223627186e-06,
+      "loss": 0.3984,
+      "step": 8142
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.074031833461905,
+      "learning_rate": 4.175039616054575e-06,
+      "loss": 0.4281,
+      "step": 8143
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7657439666436936,
+      "learning_rate": 4.173929050321264e-06,
+      "loss": 0.3959,
+      "step": 8144
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1667295535225093,
+      "learning_rate": 4.1728185264835785e-06,
+      "loss": 0.4069,
+      "step": 8145
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.880162256962636,
+      "learning_rate": 4.1717080445978394e-06,
+      "loss": 0.3943,
+      "step": 8146
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.848522758669204,
+      "learning_rate": 4.17059760472037e-06,
+      "loss": 0.3904,
+      "step": 8147
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4035685213764513,
+      "learning_rate": 4.169487206907486e-06,
+      "loss": 0.3938,
+      "step": 8148
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 8.85694984628732,
+      "learning_rate": 4.168376851215502e-06,
+      "loss": 0.3919,
+      "step": 8149
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35123930385004853,
+      "learning_rate": 4.167266537700732e-06,
+      "loss": 0.1732,
+      "step": 8150
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5991287252722723,
+      "learning_rate": 4.166156266419489e-06,
+      "loss": 0.3698,
+      "step": 8151
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1722159259883744,
+      "learning_rate": 4.165046037428083e-06,
+      "loss": 0.4047,
+      "step": 8152
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.977449930048313,
+      "learning_rate": 4.163935850782819e-06,
+      "loss": 0.4088,
+      "step": 8153
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.36217169271818195,
+      "learning_rate": 4.162825706540001e-06,
+      "loss": 0.1712,
+      "step": 8154
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.2012718257916335,
+      "learning_rate": 4.161715604755937e-06,
+      "loss": 0.3939,
+      "step": 8155
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8317937655593797,
+      "learning_rate": 4.160605545486923e-06,
+      "loss": 0.4133,
+      "step": 8156
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.702413929206047,
+      "learning_rate": 4.15949552878926e-06,
+      "loss": 0.4253,
+      "step": 8157
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.745946966349028,
+      "learning_rate": 4.1583855547192425e-06,
+      "loss": 0.4053,
+      "step": 8158
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.522826675932665,
+      "learning_rate": 4.157275623333167e-06,
+      "loss": 0.3954,
+      "step": 8159
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.806281137022243,
+      "learning_rate": 4.156165734687324e-06,
+      "loss": 0.3711,
+      "step": 8160
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.251247460851789,
+      "learning_rate": 4.155055888838005e-06,
+      "loss": 0.4146,
+      "step": 8161
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.755982350302056,
+      "learning_rate": 4.153946085841493e-06,
+      "loss": 0.3983,
+      "step": 8162
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.434547215415008,
+      "learning_rate": 4.15283632575408e-06,
+      "loss": 0.3727,
+      "step": 8163
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8055075783590775,
+      "learning_rate": 4.151726608632047e-06,
+      "loss": 0.3735,
+      "step": 8164
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.408621446668482,
+      "learning_rate": 4.150616934531674e-06,
+      "loss": 0.382,
+      "step": 8165
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.743701754754888,
+      "learning_rate": 4.149507303509239e-06,
+      "loss": 0.3817,
+      "step": 8166
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.230313105452079,
+      "learning_rate": 4.148397715621024e-06,
+      "loss": 0.4165,
+      "step": 8167
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.749534998846493,
+      "learning_rate": 4.147288170923299e-06,
+      "loss": 0.3781,
+      "step": 8168
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3555074837544643,
+      "learning_rate": 4.146178669472339e-06,
+      "loss": 0.1922,
+      "step": 8169
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4859603794691125,
+      "learning_rate": 4.145069211324412e-06,
+      "loss": 0.3818,
+      "step": 8170
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.248878983907011,
+      "learning_rate": 4.143959796535789e-06,
+      "loss": 0.3929,
+      "step": 8171
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3148198669884077,
+      "learning_rate": 4.1428504251627335e-06,
+      "loss": 0.3959,
+      "step": 8172
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.9029999344559254,
+      "learning_rate": 4.1417410972615095e-06,
+      "loss": 0.4127,
+      "step": 8173
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.005729438658136,
+      "learning_rate": 4.140631812888378e-06,
+      "loss": 0.4535,
+      "step": 8174
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.251177537530013,
+      "learning_rate": 4.1395225720996e-06,
+      "loss": 0.3783,
+      "step": 8175
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.784171838691723,
+      "learning_rate": 4.138413374951434e-06,
+      "loss": 0.3667,
+      "step": 8176
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.34562780295601314,
+      "learning_rate": 4.137304221500128e-06,
+      "loss": 0.1807,
+      "step": 8177
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 259.59035350518667,
+      "learning_rate": 4.136195111801943e-06,
+      "loss": 0.4015,
+      "step": 8178
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.699154517192489,
+      "learning_rate": 4.135086045913125e-06,
+      "loss": 0.4087,
+      "step": 8179
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.563442324591459,
+      "learning_rate": 4.133977023889923e-06,
+      "loss": 0.3721,
+      "step": 8180
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.716249055099594,
+      "learning_rate": 4.1328680457885825e-06,
+      "loss": 0.4318,
+      "step": 8181
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.509238558804269,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.3789,
+      "step": 8182
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2077718907859984,
+      "learning_rate": 4.1306502215764625e-06,
+      "loss": 0.4283,
+      "step": 8183
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.42543956254113,
+      "learning_rate": 4.1295413755781625e-06,
+      "loss": 0.3952,
+      "step": 8184
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.507170993626518,
+      "learning_rate": 4.128432573726683e-06,
+      "loss": 0.4189,
+      "step": 8185
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.5505841421107345,
+      "learning_rate": 4.127323816078264e-06,
+      "loss": 0.3821,
+      "step": 8186
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.640836025434377,
+      "learning_rate": 4.126215102689137e-06,
+      "loss": 0.3801,
+      "step": 8187
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3440022260188096,
+      "learning_rate": 4.125106433615532e-06,
+      "loss": 0.1779,
+      "step": 8188
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.4859291778341115,
+      "learning_rate": 4.123997808913673e-06,
+      "loss": 0.4009,
+      "step": 8189
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.4563477471556237,
+      "learning_rate": 4.122889228639791e-06,
+      "loss": 0.401,
+      "step": 8190
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.5718045054936254,
+      "learning_rate": 4.1217806928501084e-06,
+      "loss": 0.3849,
+      "step": 8191
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.2882289409932066,
+      "learning_rate": 4.120672201600846e-06,
+      "loss": 0.3898,
+      "step": 8192
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.186358919653187,
+      "learning_rate": 4.119563754948221e-06,
+      "loss": 0.3798,
+      "step": 8193
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.284529745421748,
+      "learning_rate": 4.1184553529484536e-06,
+      "loss": 0.3769,
+      "step": 8194
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.855627959462953,
+      "learning_rate": 4.117346995657757e-06,
+      "loss": 0.3773,
+      "step": 8195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.361444945125891,
+      "learning_rate": 4.116238683132343e-06,
+      "loss": 0.466,
+      "step": 8196
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.59760830205458,
+      "learning_rate": 4.11513041542842e-06,
+      "loss": 0.3563,
+      "step": 8197
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.48923553945387,
+      "learning_rate": 4.114022192602198e-06,
+      "loss": 0.4156,
+      "step": 8198
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.586021631985201,
+      "learning_rate": 4.112914014709883e-06,
+      "loss": 0.4137,
+      "step": 8199
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.1252031875297432,
+      "learning_rate": 4.111805881807678e-06,
+      "loss": 0.3605,
+      "step": 8200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.115645038883087,
+      "learning_rate": 4.110697793951781e-06,
+      "loss": 0.403,
+      "step": 8201
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.752839858007589,
+      "learning_rate": 4.109589751198393e-06,
+      "loss": 0.4224,
+      "step": 8202
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.244198024989425,
+      "learning_rate": 4.108481753603712e-06,
+      "loss": 0.4044,
+      "step": 8203
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.112729108849075,
+      "learning_rate": 4.107373801223927e-06,
+      "loss": 0.4007,
+      "step": 8204
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.495216482911307,
+      "learning_rate": 4.106265894115233e-06,
+      "loss": 0.3702,
+      "step": 8205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.430252052919956,
+      "learning_rate": 4.10515803233382e-06,
+      "loss": 0.3537,
+      "step": 8206
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.027687599291121,
+      "learning_rate": 4.104050215935875e-06,
+      "loss": 0.3809,
+      "step": 8207
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 10.960181497806673,
+      "learning_rate": 4.102942444977581e-06,
+      "loss": 0.4059,
+      "step": 8208
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.58015588473123,
+      "learning_rate": 4.10183471951512e-06,
+      "loss": 0.3941,
+      "step": 8209
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.064608305976483,
+      "learning_rate": 4.100727039604676e-06,
+      "loss": 0.4157,
+      "step": 8210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8420519368634456,
+      "learning_rate": 4.099619405302426e-06,
+      "loss": 0.3833,
+      "step": 8211
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.603494375309393,
+      "learning_rate": 4.0985118166645426e-06,
+      "loss": 0.3618,
+      "step": 8212
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.118098165400446,
+      "learning_rate": 4.0974042737472005e-06,
+      "loss": 0.3783,
+      "step": 8213
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7975846176888806,
+      "learning_rate": 4.096296776606572e-06,
+      "loss": 0.4317,
+      "step": 8214
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.075743887853721,
+      "learning_rate": 4.095189325298825e-06,
+      "loss": 0.3817,
+      "step": 8215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.036194744405835,
+      "learning_rate": 4.094081919880125e-06,
+      "loss": 0.4339,
+      "step": 8216
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.645295327320767,
+      "learning_rate": 4.092974560406635e-06,
+      "loss": 0.4504,
+      "step": 8217
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.447958240749431,
+      "learning_rate": 4.09186724693452e-06,
+      "loss": 0.4183,
+      "step": 8218
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7141380869511997,
+      "learning_rate": 4.090759979519937e-06,
+      "loss": 0.416,
+      "step": 8219
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.630504685012677,
+      "learning_rate": 4.089652758219045e-06,
+      "loss": 0.3635,
+      "step": 8220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.346704414686733,
+      "learning_rate": 4.088545583087994e-06,
+      "loss": 0.4059,
+      "step": 8221
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.655436999098595,
+      "learning_rate": 4.087438454182942e-06,
+      "loss": 0.4119,
+      "step": 8222
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.654850493838649,
+      "learning_rate": 4.086331371560036e-06,
+      "loss": 0.3837,
+      "step": 8223
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.900597317171265,
+      "learning_rate": 4.085224335275425e-06,
+      "loss": 0.4067,
+      "step": 8224
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.601481860710583,
+      "learning_rate": 4.084117345385252e-06,
+      "loss": 0.4265,
+      "step": 8225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.125941488981163,
+      "learning_rate": 4.083010401945662e-06,
+      "loss": 0.4206,
+      "step": 8226
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.3781768842919515,
+      "learning_rate": 4.081903505012795e-06,
+      "loss": 0.4111,
+      "step": 8227
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.427412625800433,
+      "learning_rate": 4.0807966546427895e-06,
+      "loss": 0.4046,
+      "step": 8228
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.029986352149303,
+      "learning_rate": 4.079689850891778e-06,
+      "loss": 0.404,
+      "step": 8229
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3378064523611464,
+      "learning_rate": 4.078583093815899e-06,
+      "loss": 0.4268,
+      "step": 8230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.471904433379762,
+      "learning_rate": 4.077476383471281e-06,
+      "loss": 0.424,
+      "step": 8231
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 20.813857063619626,
+      "learning_rate": 4.076369719914055e-06,
+      "loss": 0.4085,
+      "step": 8232
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.158217436425787,
+      "learning_rate": 4.075263103200342e-06,
+      "loss": 0.4245,
+      "step": 8233
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.800769778638847,
+      "learning_rate": 4.074156533386273e-06,
+      "loss": 0.3923,
+      "step": 8234
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.231835820889319,
+      "learning_rate": 4.0730500105279655e-06,
+      "loss": 0.3691,
+      "step": 8235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.264123803672112,
+      "learning_rate": 4.071943534681539e-06,
+      "loss": 0.4399,
+      "step": 8236
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.296731336286872,
+      "learning_rate": 4.07083710590311e-06,
+      "loss": 0.3824,
+      "step": 8237
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.03676895530116,
+      "learning_rate": 4.069730724248795e-06,
+      "loss": 0.3663,
+      "step": 8238
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 29.809382406059548,
+      "learning_rate": 4.0686243897747045e-06,
+      "loss": 0.4254,
+      "step": 8239
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.992099018569553,
+      "learning_rate": 4.067518102536947e-06,
+      "loss": 0.4142,
+      "step": 8240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.809012654436309,
+      "learning_rate": 4.066411862591633e-06,
+      "loss": 0.4005,
+      "step": 8241
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.42917284334036,
+      "learning_rate": 4.065305669994866e-06,
+      "loss": 0.373,
+      "step": 8242
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 13.81119527014298,
+      "learning_rate": 4.064199524802748e-06,
+      "loss": 0.4371,
+      "step": 8243
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.164205978079646,
+      "learning_rate": 4.063093427071376e-06,
+      "loss": 0.3987,
+      "step": 8244
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.952974143808622,
+      "learning_rate": 4.0619873768568545e-06,
+      "loss": 0.3967,
+      "step": 8245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 12.173308448977146,
+      "learning_rate": 4.060881374215274e-06,
+      "loss": 0.4385,
+      "step": 8246
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.2297898137666605,
+      "learning_rate": 4.059775419202729e-06,
+      "loss": 0.4032,
+      "step": 8247
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.445847905196133,
+      "learning_rate": 4.0586695118753084e-06,
+      "loss": 0.3831,
+      "step": 8248
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.932256703588149,
+      "learning_rate": 4.057563652289102e-06,
+      "loss": 0.415,
+      "step": 8249
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.8011293706842,
+      "learning_rate": 4.056457840500195e-06,
+      "loss": 0.3958,
+      "step": 8250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.932700556792633,
+      "learning_rate": 4.055352076564671e-06,
+      "loss": 0.409,
+      "step": 8251
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.9648416288608868,
+      "learning_rate": 4.054246360538607e-06,
+      "loss": 0.3932,
+      "step": 8252
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.801718516464096,
+      "learning_rate": 4.053140692478086e-06,
+      "loss": 0.3967,
+      "step": 8253
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.583459318179778,
+      "learning_rate": 4.052035072439183e-06,
+      "loss": 0.3662,
+      "step": 8254
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 5.494904732604356,
+      "learning_rate": 4.05092950047797e-06,
+      "loss": 0.4212,
+      "step": 8255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.929943964755832,
+      "learning_rate": 4.049823976650517e-06,
+      "loss": 0.4236,
+      "step": 8256
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.295430547236237,
+      "learning_rate": 4.048718501012895e-06,
+      "loss": 0.3715,
+      "step": 8257
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 7.806595778834374,
+      "learning_rate": 4.04761307362117e-06,
+      "loss": 0.4163,
+      "step": 8258
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.414549059374785,
+      "learning_rate": 4.046507694531405e-06,
+      "loss": 0.4025,
+      "step": 8259
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.671609152390672,
+      "learning_rate": 4.0454023637996606e-06,
+      "loss": 0.3726,
+      "step": 8260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 6.839490499988788,
+      "learning_rate": 4.044297081481997e-06,
+      "loss": 0.3944,
+      "step": 8261
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 58.66471719583293,
+      "learning_rate": 4.043191847634469e-06,
+      "loss": 0.3741,
+      "step": 8262
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.156973118596295,
+      "learning_rate": 4.042086662313132e-06,
+      "loss": 0.421,
+      "step": 8263
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.35574829535201585,
+      "learning_rate": 4.040981525574034e-06,
+      "loss": 0.1992,
+      "step": 8264
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.684080230335014,
+      "learning_rate": 4.039876437473228e-06,
+      "loss": 0.3846,
+      "step": 8265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.37113286986738453,
+      "learning_rate": 4.038771398066759e-06,
+      "loss": 0.1906,
+      "step": 8266
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.3709240093527506,
+      "learning_rate": 4.037666407410671e-06,
+      "loss": 0.4032,
+      "step": 8267
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.508132003164672,
+      "learning_rate": 4.036561465561005e-06,
+      "loss": 0.36,
+      "step": 8268
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.195643169838496,
+      "learning_rate": 4.0354565725738015e-06,
+      "loss": 0.4147,
+      "step": 8269
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3553162017842957,
+      "learning_rate": 4.034351728505096e-06,
+      "loss": 0.179,
+      "step": 8270
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.867075608949488,
+      "learning_rate": 4.033246933410922e-06,
+      "loss": 0.384,
+      "step": 8271
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.225187121869004,
+      "learning_rate": 4.0321421873473095e-06,
+      "loss": 0.4115,
+      "step": 8272
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.548871552882097,
+      "learning_rate": 4.0310374903702934e-06,
+      "loss": 0.3956,
+      "step": 8273
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.16801359286386,
+      "learning_rate": 4.029932842535895e-06,
+      "loss": 0.3779,
+      "step": 8274
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.066300356806018,
+      "learning_rate": 4.028828243900141e-06,
+      "loss": 0.366,
+      "step": 8275
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.323832152306607,
+      "learning_rate": 4.027723694519051e-06,
+      "loss": 0.3971,
+      "step": 8276
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.410211224126975,
+      "learning_rate": 4.026619194448647e-06,
+      "loss": 0.4344,
+      "step": 8277
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.3875979419461135,
+      "learning_rate": 4.025514743744943e-06,
+      "loss": 0.4144,
+      "step": 8278
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.122492378371199,
+      "learning_rate": 4.024410342463956e-06,
+      "loss": 0.3807,
+      "step": 8279
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.872375466331468,
+      "learning_rate": 4.023305990661694e-06,
+      "loss": 0.3664,
+      "step": 8280
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.013035346314679,
+      "learning_rate": 4.022201688394168e-06,
+      "loss": 0.4156,
+      "step": 8281
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.855340128703485,
+      "learning_rate": 4.021097435717386e-06,
+      "loss": 0.3933,
+      "step": 8282
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.093535218302993,
+      "learning_rate": 4.019993232687349e-06,
+      "loss": 0.3941,
+      "step": 8283
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.245152840256067,
+      "learning_rate": 4.018889079360059e-06,
+      "loss": 0.3727,
+      "step": 8284
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.670540702961856,
+      "learning_rate": 4.0177849757915185e-06,
+      "loss": 0.3886,
+      "step": 8285
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.503909124786443,
+      "learning_rate": 4.016680922037723e-06,
+      "loss": 0.3829,
+      "step": 8286
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.515582997650358,
+      "learning_rate": 4.015576918154664e-06,
+      "loss": 0.4311,
+      "step": 8287
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.671296051326861,
+      "learning_rate": 4.014472964198332e-06,
+      "loss": 0.4084,
+      "step": 8288
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.548152019008989,
+      "learning_rate": 4.013369060224721e-06,
+      "loss": 0.3927,
+      "step": 8289
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.417002949399751,
+      "learning_rate": 4.012265206289814e-06,
+      "loss": 0.3609,
+      "step": 8290
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.831137777831959,
+      "learning_rate": 4.011161402449596e-06,
+      "loss": 0.4127,
+      "step": 8291
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3793390253828565,
+      "learning_rate": 4.0100576487600465e-06,
+      "loss": 0.191,
+      "step": 8292
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.901285779578863,
+      "learning_rate": 4.008953945277148e-06,
+      "loss": 0.4028,
+      "step": 8293
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.41505914171335,
+      "learning_rate": 4.007850292056873e-06,
+      "loss": 0.3792,
+      "step": 8294
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.948148771709953,
+      "learning_rate": 4.006746689155196e-06,
+      "loss": 0.4021,
+      "step": 8295
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0044614847983313,
+      "learning_rate": 4.005643136628088e-06,
+      "loss": 0.4175,
+      "step": 8296
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.5254030954104705,
+      "learning_rate": 4.00453963453152e-06,
+      "loss": 0.4128,
+      "step": 8297
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.0907656905402625,
+      "learning_rate": 4.0034361829214565e-06,
+      "loss": 0.4102,
+      "step": 8298
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.324706269460937,
+      "learning_rate": 4.002332781853861e-06,
+      "loss": 0.391,
+      "step": 8299
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4602466828741743,
+      "learning_rate": 4.0012294313846915e-06,
+      "loss": 0.4053,
+      "step": 8300
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.583908489524617,
+      "learning_rate": 4.000126131569912e-06,
+      "loss": 0.3672,
+      "step": 8301
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.624939790504457,
+      "learning_rate": 3.999022882465474e-06,
+      "loss": 0.4512,
+      "step": 8302
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.846668601467832,
+      "learning_rate": 3.997919684127332e-06,
+      "loss": 0.4013,
+      "step": 8303
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.456520779720257,
+      "learning_rate": 3.996816536611438e-06,
+      "loss": 0.4496,
+      "step": 8304
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.285317023176373,
+      "learning_rate": 3.995713439973738e-06,
+      "loss": 0.3955,
+      "step": 8305
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.591876559279323,
+      "learning_rate": 3.994610394270178e-06,
+      "loss": 0.4136,
+      "step": 8306
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.923701129854569,
+      "learning_rate": 3.993507399556699e-06,
+      "loss": 0.3927,
+      "step": 8307
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.0268432037854085,
+      "learning_rate": 3.992404455889246e-06,
+      "loss": 0.424,
+      "step": 8308
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.03633083755183,
+      "learning_rate": 3.991301563323754e-06,
+      "loss": 0.3648,
+      "step": 8309
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9457868900240984,
+      "learning_rate": 3.990198721916158e-06,
+      "loss": 0.3714,
+      "step": 8310
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.9324492678257745,
+      "learning_rate": 3.989095931722388e-06,
+      "loss": 0.4164,
+      "step": 8311
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.742032196683788,
+      "learning_rate": 3.98799319279838e-06,
+      "loss": 0.421,
+      "step": 8312
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 17.233497363714754,
+      "learning_rate": 3.986890505200057e-06,
+      "loss": 0.4218,
+      "step": 8313
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3774809369435017,
+      "learning_rate": 3.985787868983345e-06,
+      "loss": 0.4126,
+      "step": 8314
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.74597072969405,
+      "learning_rate": 3.984685284204164e-06,
+      "loss": 0.4214,
+      "step": 8315
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.1607320253037114,
+      "learning_rate": 3.983582750918437e-06,
+      "loss": 0.397,
+      "step": 8316
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.41881788894124,
+      "learning_rate": 3.982480269182079e-06,
+      "loss": 0.3888,
+      "step": 8317
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.357122468890216,
+      "learning_rate": 3.981377839051004e-06,
+      "loss": 0.3966,
+      "step": 8318
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.011168410607123,
+      "learning_rate": 3.980275460581122e-06,
+      "loss": 0.3896,
+      "step": 8319
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.363325973955091,
+      "learning_rate": 3.979173133828347e-06,
+      "loss": 0.4021,
+      "step": 8320
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.790440509186232,
+      "learning_rate": 3.978070858848583e-06,
+      "loss": 0.3968,
+      "step": 8321
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.090478494036732,
+      "learning_rate": 3.976968635697732e-06,
+      "loss": 0.3855,
+      "step": 8322
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.797486524278973,
+      "learning_rate": 3.975866464431696e-06,
+      "loss": 0.385,
+      "step": 8323
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5150584742972715,
+      "learning_rate": 3.974764345106376e-06,
+      "loss": 0.4085,
+      "step": 8324
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.492468372731286,
+      "learning_rate": 3.9736622777776635e-06,
+      "loss": 0.4025,
+      "step": 8325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.734409972327594,
+      "learning_rate": 3.972560262501456e-06,
+      "loss": 0.4059,
+      "step": 8326
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.450986298698997,
+      "learning_rate": 3.971458299333641e-06,
+      "loss": 0.4114,
+      "step": 8327
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.944310669137741,
+      "learning_rate": 3.970356388330109e-06,
+      "loss": 0.3823,
+      "step": 8328
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.333497371057772,
+      "learning_rate": 3.9692545295467445e-06,
+      "loss": 0.3954,
+      "step": 8329
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.159236696410423,
+      "learning_rate": 3.968152723039429e-06,
+      "loss": 0.4106,
+      "step": 8330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.741861024703573,
+      "learning_rate": 3.967050968864042e-06,
+      "loss": 0.4389,
+      "step": 8331
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.496705306014875,
+      "learning_rate": 3.965949267076465e-06,
+      "loss": 0.4372,
+      "step": 8332
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.402190637222661,
+      "learning_rate": 3.964847617732568e-06,
+      "loss": 0.3784,
+      "step": 8333
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3592891236166054,
+      "learning_rate": 3.963746020888226e-06,
+      "loss": 0.178,
+      "step": 8334
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.228033622008253,
+      "learning_rate": 3.962644476599306e-06,
+      "loss": 0.418,
+      "step": 8335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.1587651323142034,
+      "learning_rate": 3.961542984921679e-06,
+      "loss": 0.3803,
+      "step": 8336
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.753703410685494,
+      "learning_rate": 3.960441545911205e-06,
+      "loss": 0.4167,
+      "step": 8337
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.065958579290161,
+      "learning_rate": 3.959340159623747e-06,
+      "loss": 0.416,
+      "step": 8338
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 13.94622187372678,
+      "learning_rate": 3.9582388261151606e-06,
+      "loss": 0.361,
+      "step": 8339
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.978521371155678,
+      "learning_rate": 3.957137545441307e-06,
+      "loss": 0.4256,
+      "step": 8340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.806006248543477,
+      "learning_rate": 3.956036317658037e-06,
+      "loss": 0.3573,
+      "step": 8341
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.301393891569136,
+      "learning_rate": 3.954935142821202e-06,
+      "loss": 0.3713,
+      "step": 8342
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.408483051654355,
+      "learning_rate": 3.953834020986646e-06,
+      "loss": 0.3808,
+      "step": 8343
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.195657930988063,
+      "learning_rate": 3.952732952210221e-06,
+      "loss": 0.3968,
+      "step": 8344
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.944175605053839,
+      "learning_rate": 3.951631936547766e-06,
+      "loss": 0.3972,
+      "step": 8345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.581472726251818,
+      "learning_rate": 3.950530974055122e-06,
+      "loss": 0.3881,
+      "step": 8346
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.429611834000818,
+      "learning_rate": 3.949430064788124e-06,
+      "loss": 0.3864,
+      "step": 8347
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.773471218173916,
+      "learning_rate": 3.948329208802609e-06,
+      "loss": 0.4212,
+      "step": 8348
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.448359588906248,
+      "learning_rate": 3.947228406154408e-06,
+      "loss": 0.3846,
+      "step": 8349
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.854067227795289,
+      "learning_rate": 3.94612765689935e-06,
+      "loss": 0.4078,
+      "step": 8350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.532527786312949,
+      "learning_rate": 3.945026961093259e-06,
+      "loss": 0.364,
+      "step": 8351
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.615465149700801,
+      "learning_rate": 3.9439263187919635e-06,
+      "loss": 0.4081,
+      "step": 8352
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 10.64728122161649,
+      "learning_rate": 3.9428257300512825e-06,
+      "loss": 0.3894,
+      "step": 8353
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.554928726718259,
+      "learning_rate": 3.941725194927034e-06,
+      "loss": 0.3985,
+      "step": 8354
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 16.301579634931436,
+      "learning_rate": 3.940624713475031e-06,
+      "loss": 0.4257,
+      "step": 8355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.516654622185007,
+      "learning_rate": 3.93952428575109e-06,
+      "loss": 0.4258,
+      "step": 8356
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.054566338930098,
+      "learning_rate": 3.938423911811021e-06,
+      "loss": 0.3775,
+      "step": 8357
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.768880602599943,
+      "learning_rate": 3.937323591710629e-06,
+      "loss": 0.3773,
+      "step": 8358
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.781627894920806,
+      "learning_rate": 3.9362233255057205e-06,
+      "loss": 0.3824,
+      "step": 8359
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.705935712052339,
+      "learning_rate": 3.935123113252097e-06,
+      "loss": 0.4117,
+      "step": 8360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.581583903188265,
+      "learning_rate": 3.934022955005558e-06,
+      "loss": 0.3703,
+      "step": 8361
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.0418156574399475,
+      "learning_rate": 3.932922850821899e-06,
+      "loss": 0.4194,
+      "step": 8362
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.8302729477954984,
+      "learning_rate": 3.931822800756911e-06,
+      "loss": 0.3908,
+      "step": 8363
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.127180698518863,
+      "learning_rate": 3.9307228048663915e-06,
+      "loss": 0.3746,
+      "step": 8364
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.350642991020902,
+      "learning_rate": 3.929622863206125e-06,
+      "loss": 0.1803,
+      "step": 8365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8860929591096056,
+      "learning_rate": 3.928522975831898e-06,
+      "loss": 0.3897,
+      "step": 8366
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.4094767663987895,
+      "learning_rate": 3.927423142799489e-06,
+      "loss": 0.4319,
+      "step": 8367
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 9.591347713338596,
+      "learning_rate": 3.926323364164684e-06,
+      "loss": 0.3927,
+      "step": 8368
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.8120916725727505,
+      "learning_rate": 3.925223639983258e-06,
+      "loss": 0.4023,
+      "step": 8369
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.699757365400062,
+      "learning_rate": 3.924123970310984e-06,
+      "loss": 0.4031,
+      "step": 8370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.285477810118768,
+      "learning_rate": 3.923024355203635e-06,
+      "loss": 0.412,
+      "step": 8371
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.327906464205487,
+      "learning_rate": 3.92192479471698e-06,
+      "loss": 0.3795,
+      "step": 8372
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.739268849724161,
+      "learning_rate": 3.920825288906784e-06,
+      "loss": 0.4678,
+      "step": 8373
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3287194187872364,
+      "learning_rate": 3.91972583782881e-06,
+      "loss": 0.386,
+      "step": 8374
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 12.056573880282194,
+      "learning_rate": 3.918626441538821e-06,
+      "loss": 0.4119,
+      "step": 8375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.44401593555029,
+      "learning_rate": 3.917527100092574e-06,
+      "loss": 0.4849,
+      "step": 8376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.555196448776833,
+      "learning_rate": 3.916427813545823e-06,
+      "loss": 0.4545,
+      "step": 8377
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.672807605339252,
+      "learning_rate": 3.915328581954319e-06,
+      "loss": 0.3826,
+      "step": 8378
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.6520898273939446,
+      "learning_rate": 3.914229405373815e-06,
+      "loss": 0.3426,
+      "step": 8379
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.272133731253204,
+      "learning_rate": 3.913130283860056e-06,
+      "loss": 0.4153,
+      "step": 8380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.185781331585037,
+      "learning_rate": 3.9120312174687845e-06,
+      "loss": 0.3823,
+      "step": 8381
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.3371862191738737,
+      "learning_rate": 3.910932206255742e-06,
+      "loss": 0.4048,
+      "step": 8382
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.8315421340543585,
+      "learning_rate": 3.909833250276669e-06,
+      "loss": 0.4052,
+      "step": 8383
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 11.61984813682154,
+      "learning_rate": 3.908734349587299e-06,
+      "loss": 0.3924,
+      "step": 8384
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.300400409372574,
+      "learning_rate": 3.907635504243366e-06,
+      "loss": 0.3888,
+      "step": 8385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.010806802740918,
+      "learning_rate": 3.906536714300596e-06,
+      "loss": 0.3919,
+      "step": 8386
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.509960744114564,
+      "learning_rate": 3.905437979814721e-06,
+      "loss": 0.3901,
+      "step": 8387
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.6445484046521694,
+      "learning_rate": 3.904339300841464e-06,
+      "loss": 0.3602,
+      "step": 8388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.305584279407773,
+      "learning_rate": 3.903240677436544e-06,
+      "loss": 0.4011,
+      "step": 8389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.7866500175323115,
+      "learning_rate": 3.902142109655681e-06,
+      "loss": 0.4078,
+      "step": 8390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 70.01736238535075,
+      "learning_rate": 3.901043597554592e-06,
+      "loss": 0.3583,
+      "step": 8391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 7.45719626088778,
+      "learning_rate": 3.899945141188989e-06,
+      "loss": 0.3975,
+      "step": 8392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.905450982166024,
+      "learning_rate": 3.898846740614581e-06,
+      "loss": 0.3995,
+      "step": 8393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.5558528137014576,
+      "learning_rate": 3.897748395887077e-06,
+      "loss": 0.4077,
+      "step": 8394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2340364154703525,
+      "learning_rate": 3.8966501070621805e-06,
+      "loss": 0.4112,
+      "step": 8395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.765525312270878,
+      "learning_rate": 3.8955518741955945e-06,
+      "loss": 0.4121,
+      "step": 8396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.4545613006609237,
+      "learning_rate": 3.894453697343016e-06,
+      "loss": 0.3981,
+      "step": 8397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.081971828740132,
+      "learning_rate": 3.893355576560141e-06,
+      "loss": 0.4028,
+      "step": 8398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9426789567061262,
+      "learning_rate": 3.892257511902664e-06,
+      "loss": 0.4113,
+      "step": 8399
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.834636658090232,
+      "learning_rate": 3.891159503426275e-06,
+      "loss": 0.396,
+      "step": 8400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.198020717937872,
+      "learning_rate": 3.890061551186661e-06,
+      "loss": 0.4146,
+      "step": 8401
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.494327363839174,
+      "learning_rate": 3.888963655239505e-06,
+      "loss": 0.403,
+      "step": 8402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.4790705546616816,
+      "learning_rate": 3.8878658156404914e-06,
+      "loss": 0.3952,
+      "step": 8403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.090171172463893,
+      "learning_rate": 3.886768032445299e-06,
+      "loss": 0.4166,
+      "step": 8404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9076541858459857,
+      "learning_rate": 3.885670305709601e-06,
+      "loss": 0.3864,
+      "step": 8405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 8.27384929407164,
+      "learning_rate": 3.884572635489071e-06,
+      "loss": 0.3808,
+      "step": 8406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.380181853741292,
+      "learning_rate": 3.883475021839382e-06,
+      "loss": 0.4206,
+      "step": 8407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.831502940186633,
+      "learning_rate": 3.882377464816201e-06,
+      "loss": 0.4052,
+      "step": 8408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.2595539899193517,
+      "learning_rate": 3.88127996447519e-06,
+      "loss": 0.3793,
+      "step": 8409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.443549027283796,
+      "learning_rate": 3.88018252087201e-06,
+      "loss": 0.3928,
+      "step": 8410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3560818710827047,
+      "learning_rate": 3.879085134062324e-06,
+      "loss": 0.1748,
+      "step": 8411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.9924683266364966,
+      "learning_rate": 3.877987804101786e-06,
+      "loss": 0.3818,
+      "step": 8412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.4179212682171904,
+      "learning_rate": 3.876890531046048e-06,
+      "loss": 0.3644,
+      "step": 8413
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.490368890707445,
+      "learning_rate": 3.87579331495076e-06,
+      "loss": 0.3947,
+      "step": 8414
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.174301303622129,
+      "learning_rate": 3.8746961558715716e-06,
+      "loss": 0.3954,
+      "step": 8415
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 21.397694248136382,
+      "learning_rate": 3.873599053864125e-06,
+      "loss": 0.4056,
+      "step": 8416
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.442447210244744,
+      "learning_rate": 3.872502008984061e-06,
+      "loss": 0.4191,
+      "step": 8417
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.42802585474797,
+      "learning_rate": 3.871405021287019e-06,
+      "loss": 0.3695,
+      "step": 8418
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2728279778036873,
+      "learning_rate": 3.870308090828636e-06,
+      "loss": 0.3611,
+      "step": 8419
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.944581166902572,
+      "learning_rate": 3.869211217664544e-06,
+      "loss": 0.4124,
+      "step": 8420
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5350374053341134,
+      "learning_rate": 3.868114401850372e-06,
+      "loss": 0.4012,
+      "step": 8421
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.201816865277842,
+      "learning_rate": 3.867017643441746e-06,
+      "loss": 0.4179,
+      "step": 8422
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.725954339353397,
+      "learning_rate": 3.8659209424942925e-06,
+      "loss": 0.3931,
+      "step": 8423
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8125166265602766,
+      "learning_rate": 3.864824299063632e-06,
+      "loss": 0.373,
+      "step": 8424
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.19563128198087,
+      "learning_rate": 3.863727713205382e-06,
+      "loss": 0.3844,
+      "step": 8425
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.364457060584593,
+      "learning_rate": 3.862631184975157e-06,
+      "loss": 0.1969,
+      "step": 8426
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.452105077314102,
+      "learning_rate": 3.861534714428571e-06,
+      "loss": 0.4304,
+      "step": 8427
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.645053710353474,
+      "learning_rate": 3.860438301621233e-06,
+      "loss": 0.3837,
+      "step": 8428
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.4659804435347095,
+      "learning_rate": 3.85934194660875e-06,
+      "loss": 0.3931,
+      "step": 8429
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.037543196251115,
+      "learning_rate": 3.8582456494467214e-06,
+      "loss": 0.4253,
+      "step": 8430
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.711520494592668,
+      "learning_rate": 3.857149410190754e-06,
+      "loss": 0.4465,
+      "step": 8431
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.862251909097621,
+      "learning_rate": 3.856053228896442e-06,
+      "loss": 0.4198,
+      "step": 8432
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.447553173516939,
+      "learning_rate": 3.8549571056193785e-06,
+      "loss": 0.3872,
+      "step": 8433
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5469496550566944,
+      "learning_rate": 3.85386104041516e-06,
+      "loss": 0.3723,
+      "step": 8434
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.442186797816302,
+      "learning_rate": 3.8527650333393725e-06,
+      "loss": 0.4165,
+      "step": 8435
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.34413402466520876,
+      "learning_rate": 3.8516690844476035e-06,
+      "loss": 0.1938,
+      "step": 8436
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.39307489173561094,
+      "learning_rate": 3.850573193795433e-06,
+      "loss": 0.1942,
+      "step": 8437
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.897650038740742,
+      "learning_rate": 3.849477361438443e-06,
+      "loss": 0.409,
+      "step": 8438
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.088279722226474,
+      "learning_rate": 3.848381587432212e-06,
+      "loss": 0.4238,
+      "step": 8439
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 24.102012942456298,
+      "learning_rate": 3.847285871832311e-06,
+      "loss": 0.3715,
+      "step": 8440
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32906602762464776,
+      "learning_rate": 3.846190214694311e-06,
+      "loss": 0.189,
+      "step": 8441
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.146989002507709,
+      "learning_rate": 3.845094616073783e-06,
+      "loss": 0.3668,
+      "step": 8442
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 9.852296799740957,
+      "learning_rate": 3.843999076026292e-06,
+      "loss": 0.385,
+      "step": 8443
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.3453857681810195,
+      "learning_rate": 3.842903594607399e-06,
+      "loss": 0.3861,
+      "step": 8444
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3722257504918955,
+      "learning_rate": 3.841808171872661e-06,
+      "loss": 0.3836,
+      "step": 8445
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.267647574696349,
+      "learning_rate": 3.840712807877638e-06,
+      "loss": 0.4209,
+      "step": 8446
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8238306534708486,
+      "learning_rate": 3.839617502677883e-06,
+      "loss": 0.4228,
+      "step": 8447
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.080447961650498,
+      "learning_rate": 3.838522256328944e-06,
+      "loss": 0.3806,
+      "step": 8448
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.4644459192613475,
+      "learning_rate": 3.837427068886369e-06,
+      "loss": 0.4256,
+      "step": 8449
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1832410441500776,
+      "learning_rate": 3.8363319404057045e-06,
+      "loss": 0.39,
+      "step": 8450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.318005492470258,
+      "learning_rate": 3.83523687094249e-06,
+      "loss": 0.4478,
+      "step": 8451
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.466107676197985,
+      "learning_rate": 3.834141860552263e-06,
+      "loss": 0.4533,
+      "step": 8452
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.400021613507326,
+      "learning_rate": 3.833046909290559e-06,
+      "loss": 0.3869,
+      "step": 8453
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.463951804777352,
+      "learning_rate": 3.8319520172129125e-06,
+      "loss": 0.3826,
+      "step": 8454
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.34661253285368443,
+      "learning_rate": 3.8308571843748525e-06,
+      "loss": 0.2091,
+      "step": 8455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 25.178835384170405,
+      "learning_rate": 3.829762410831904e-06,
+      "loss": 0.3907,
+      "step": 8456
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.5466540083246105,
+      "learning_rate": 3.8286676966395895e-06,
+      "loss": 0.4218,
+      "step": 8457
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.929891429673055,
+      "learning_rate": 3.827573041853433e-06,
+      "loss": 0.3796,
+      "step": 8458
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.107054642694973,
+      "learning_rate": 3.826478446528948e-06,
+      "loss": 0.4031,
+      "step": 8459
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.042293120656734,
+      "learning_rate": 3.825383910721651e-06,
+      "loss": 0.4193,
+      "step": 8460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.220704139554048,
+      "learning_rate": 3.82428943448705e-06,
+      "loss": 0.452,
+      "step": 8461
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.236714266101366,
+      "learning_rate": 3.823195017880659e-06,
+      "loss": 0.3772,
+      "step": 8462
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.122577859155658,
+      "learning_rate": 3.822100660957979e-06,
+      "loss": 0.3811,
+      "step": 8463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.489706413917105,
+      "learning_rate": 3.821006363774513e-06,
+      "loss": 0.372,
+      "step": 8464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6284895794468275,
+      "learning_rate": 3.819912126385759e-06,
+      "loss": 0.3798,
+      "step": 8465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.048886488420461,
+      "learning_rate": 3.818817948847217e-06,
+      "loss": 0.4016,
+      "step": 8466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.457064795433167,
+      "learning_rate": 3.817723831214378e-06,
+      "loss": 0.3908,
+      "step": 8467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.754411670088749,
+      "learning_rate": 3.816629773542731e-06,
+      "loss": 0.3788,
+      "step": 8468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.5312304282466815,
+      "learning_rate": 3.815535775887763e-06,
+      "loss": 0.3838,
+      "step": 8469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.384373063799868,
+      "learning_rate": 3.8144418383049604e-06,
+      "loss": 0.4581,
+      "step": 8470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.139692282108119,
+      "learning_rate": 3.8133479608498026e-06,
+      "loss": 0.3653,
+      "step": 8471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.464160983708116,
+      "learning_rate": 3.812254143577767e-06,
+      "loss": 0.3891,
+      "step": 8472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.481604071131225,
+      "learning_rate": 3.8111603865443268e-06,
+      "loss": 0.4049,
+      "step": 8473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.348172719081364,
+      "learning_rate": 3.8100666898049586e-06,
+      "loss": 0.3794,
+      "step": 8474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.5613479068677125,
+      "learning_rate": 3.8089730534151286e-06,
+      "loss": 0.3991,
+      "step": 8475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.9252075103423945,
+      "learning_rate": 3.807879477430302e-06,
+      "loss": 0.352,
+      "step": 8476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.567882721657302,
+      "learning_rate": 3.8067859619059398e-06,
+      "loss": 0.4004,
+      "step": 8477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.668957948004222,
+      "learning_rate": 3.8056925068975055e-06,
+      "loss": 0.3978,
+      "step": 8478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.26012630287014,
+      "learning_rate": 3.8045991124604536e-06,
+      "loss": 0.4238,
+      "step": 8479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.7429277336420372,
+      "learning_rate": 3.803505778650237e-06,
+      "loss": 0.4053,
+      "step": 8480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.303989620339453,
+      "learning_rate": 3.802412505522306e-06,
+      "loss": 0.4172,
+      "step": 8481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.76577256653946,
+      "learning_rate": 3.8013192931321095e-06,
+      "loss": 0.4339,
+      "step": 8482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.780453096972776,
+      "learning_rate": 3.8002261415350905e-06,
+      "loss": 0.4021,
+      "step": 8483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8279760556028695,
+      "learning_rate": 3.7991330507866903e-06,
+      "loss": 0.3926,
+      "step": 8484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.6927829979794224,
+      "learning_rate": 3.798040020942344e-06,
+      "loss": 0.3756,
+      "step": 8485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.38492728976344553,
+      "learning_rate": 3.796947052057492e-06,
+      "loss": 0.1911,
+      "step": 8486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8755567510124718,
+      "learning_rate": 3.7958541441875628e-06,
+      "loss": 0.4239,
+      "step": 8487
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.911210466550823,
+      "learning_rate": 3.7947612973879853e-06,
+      "loss": 0.4091,
+      "step": 8488
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4701443757972763,
+      "learning_rate": 3.7936685117141847e-06,
+      "loss": 0.3835,
+      "step": 8489
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.6417297380309055,
+      "learning_rate": 3.792575787221585e-06,
+      "loss": 0.4322,
+      "step": 8490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 32.553908809804184,
+      "learning_rate": 3.7914831239656056e-06,
+      "loss": 0.3946,
+      "step": 8491
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.20895894045135,
+      "learning_rate": 3.790390522001662e-06,
+      "loss": 0.4129,
+      "step": 8492
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.925956381026897,
+      "learning_rate": 3.789297981385167e-06,
+      "loss": 0.3653,
+      "step": 8493
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7529781278442425,
+      "learning_rate": 3.7882055021715325e-06,
+      "loss": 0.3845,
+      "step": 8494
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.258137069022738,
+      "learning_rate": 3.7871130844161635e-06,
+      "loss": 0.38,
+      "step": 8495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.607320624738306,
+      "learning_rate": 3.7860207281744626e-06,
+      "loss": 0.3526,
+      "step": 8496
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 12.017745581838362,
+      "learning_rate": 3.784928433501835e-06,
+      "loss": 0.3899,
+      "step": 8497
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.1388530494483353,
+      "learning_rate": 3.7838362004536756e-06,
+      "loss": 0.4332,
+      "step": 8498
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3547439014874958,
+      "learning_rate": 3.7827440290853794e-06,
+      "loss": 0.1915,
+      "step": 8499
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.133836625163362,
+      "learning_rate": 3.7816519194523354e-06,
+      "loss": 0.4032,
+      "step": 8500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 8.77803717008846,
+      "learning_rate": 3.781823027158907e-06,
+      "loss": 0.3218,
+      "step": 8501
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.796127455133916,
+      "learning_rate": 3.7807311261720003e-06,
+      "loss": 0.3147,
+      "step": 8502
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8089960974202057,
+      "learning_rate": 3.77963928700471e-06,
+      "loss": 0.338,
+      "step": 8503
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.00802683193875,
+      "learning_rate": 3.7785475097123934e-06,
+      "loss": 0.349,
+      "step": 8504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30783039819046254,
+      "learning_rate": 3.777455794350403e-06,
+      "loss": 0.1111,
+      "step": 8505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.440080326556451,
+      "learning_rate": 3.776364140974097e-06,
+      "loss": 0.3313,
+      "step": 8506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.990369517119199,
+      "learning_rate": 3.7752725496388195e-06,
+      "loss": 0.3477,
+      "step": 8507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 22.309527828421373,
+      "learning_rate": 3.7741810203999202e-06,
+      "loss": 0.3192,
+      "step": 8508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.018978934801194,
+      "learning_rate": 3.773089553312741e-06,
+      "loss": 0.3375,
+      "step": 8509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.533800344950472,
+      "learning_rate": 3.7719981484326192e-06,
+      "loss": 0.3187,
+      "step": 8510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 20.609930104517808,
+      "learning_rate": 3.7709068058148954e-06,
+      "loss": 0.3604,
+      "step": 8511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.648232193785528,
+      "learning_rate": 3.7698155255149005e-06,
+      "loss": 0.33,
+      "step": 8512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.197853902390579,
+      "learning_rate": 3.768724307587965e-06,
+      "loss": 0.3366,
+      "step": 8513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3271761669973615,
+      "learning_rate": 3.7676331520894184e-06,
+      "loss": 0.1065,
+      "step": 8514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.2076060229576315,
+      "learning_rate": 3.7665420590745805e-06,
+      "loss": 0.3417,
+      "step": 8515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3405908071810675,
+      "learning_rate": 3.7654510285987765e-06,
+      "loss": 0.1147,
+      "step": 8516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.6215279021030184,
+      "learning_rate": 3.7643600607173216e-06,
+      "loss": 0.3302,
+      "step": 8517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 10.663297713424182,
+      "learning_rate": 3.7632691554855282e-06,
+      "loss": 0.3414,
+      "step": 8518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.40256319976418,
+      "learning_rate": 3.7621783129587132e-06,
+      "loss": 0.3525,
+      "step": 8519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 11.70971870824609,
+      "learning_rate": 3.7610875331921785e-06,
+      "loss": 0.344,
+      "step": 8520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.811118916276885,
+      "learning_rate": 3.759996816241235e-06,
+      "loss": 0.3549,
+      "step": 8521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.3296442241902735,
+      "learning_rate": 3.7589061621611805e-06,
+      "loss": 0.3199,
+      "step": 8522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.767165516185309,
+      "learning_rate": 3.7578155710073134e-06,
+      "loss": 0.2919,
+      "step": 8523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32622533141175414,
+      "learning_rate": 3.756725042834932e-06,
+      "loss": 0.115,
+      "step": 8524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.346030312426701,
+      "learning_rate": 3.7556345776993257e-06,
+      "loss": 0.3223,
+      "step": 8525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.022747192407015,
+      "learning_rate": 3.7545441756557844e-06,
+      "loss": 0.3618,
+      "step": 8526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.495209393835952,
+      "learning_rate": 3.753453836759595e-06,
+      "loss": 0.3517,
+      "step": 8527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.120832515943762,
+      "learning_rate": 3.752363561066039e-06,
+      "loss": 0.3397,
+      "step": 8528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.575261558353103,
+      "learning_rate": 3.7512733486303965e-06,
+      "loss": 0.3038,
+      "step": 8529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.31616243313549147,
+      "learning_rate": 3.75018319950794e-06,
+      "loss": 0.1184,
+      "step": 8530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.450663669471249,
+      "learning_rate": 3.749093113753949e-06,
+      "loss": 0.34,
+      "step": 8531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8812822658142254,
+      "learning_rate": 3.7480030914236887e-06,
+      "loss": 0.3213,
+      "step": 8532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9453511367264085,
+      "learning_rate": 3.746913132572425e-06,
+      "loss": 0.3229,
+      "step": 8533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.043311820932786,
+      "learning_rate": 3.745823237255425e-06,
+      "loss": 0.3359,
+      "step": 8534
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.7621151709954725,
+      "learning_rate": 3.744733405527945e-06,
+      "loss": 0.3228,
+      "step": 8535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 23.09011784069246,
+      "learning_rate": 3.7436436374452452e-06,
+      "loss": 0.3215,
+      "step": 8536
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.5989668784501685,
+      "learning_rate": 3.7425539330625776e-06,
+      "loss": 0.3115,
+      "step": 8537
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8466835469846536,
+      "learning_rate": 3.7414642924351917e-06,
+      "loss": 0.2985,
+      "step": 8538
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.092094845786697,
+      "learning_rate": 3.740374715618336e-06,
+      "loss": 0.3528,
+      "step": 8539
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4980722624292646,
+      "learning_rate": 3.7392852026672534e-06,
+      "loss": 0.3104,
+      "step": 8540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.564336198330846,
+      "learning_rate": 3.738195753637186e-06,
+      "loss": 0.3402,
+      "step": 8541
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 14.73555546522274,
+      "learning_rate": 3.7371063685833706e-06,
+      "loss": 0.3048,
+      "step": 8542
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.285729649821114,
+      "learning_rate": 3.7360170475610395e-06,
+      "loss": 0.3233,
+      "step": 8543
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.30022595837879085,
+      "learning_rate": 3.7349277906254278e-06,
+      "loss": 0.1156,
+      "step": 8544
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 5.2985746199298545,
+      "learning_rate": 3.7338385978317587e-06,
+      "loss": 0.334,
+      "step": 8545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0646496718984735,
+      "learning_rate": 3.7327494692352605e-06,
+      "loss": 0.2927,
+      "step": 8546
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.317093458147333,
+      "learning_rate": 3.7316604048911538e-06,
+      "loss": 0.2959,
+      "step": 8547
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.8726597728648624,
+      "learning_rate": 3.730571404854653e-06,
+      "loss": 0.3071,
+      "step": 8548
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2048758443948624,
+      "learning_rate": 3.7294824691809773e-06,
+      "loss": 0.3174,
+      "step": 8549
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.741325893895236,
+      "learning_rate": 3.728393597925335e-06,
+      "loss": 0.3297,
+      "step": 8550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.4868627803787913,
+      "learning_rate": 3.727304791142936e-06,
+      "loss": 0.322,
+      "step": 8551
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.400146894604596,
+      "learning_rate": 3.7262160488889854e-06,
+      "loss": 0.3396,
+      "step": 8552
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.539676437627521,
+      "learning_rate": 3.725127371218683e-06,
+      "loss": 0.319,
+      "step": 8553
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.508415403412691,
+      "learning_rate": 3.724038758187229e-06,
+      "loss": 0.3563,
+      "step": 8554
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 6.38912059085408,
+      "learning_rate": 3.7229502098498183e-06,
+      "loss": 0.3068,
+      "step": 8555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.307018445707124,
+      "learning_rate": 3.7218617262616398e-06,
+      "loss": 0.3119,
+      "step": 8556
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.757596940399695,
+      "learning_rate": 3.720773307477886e-06,
+      "loss": 0.3243,
+      "step": 8557
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.7722702436799755,
+      "learning_rate": 3.719684953553738e-06,
+      "loss": 0.3378,
+      "step": 8558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.859780814526762,
+      "learning_rate": 3.718596664544384e-06,
+      "loss": 0.3204,
+      "step": 8559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.360056763714683,
+      "learning_rate": 3.7175084405049978e-06,
+      "loss": 0.3306,
+      "step": 8560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.091704574475129,
+      "learning_rate": 3.7164202814907534e-06,
+      "loss": 0.3382,
+      "step": 8561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.484507361084051,
+      "learning_rate": 3.715332187556828e-06,
+      "loss": 0.3487,
+      "step": 8562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1644907693988578,
+      "learning_rate": 3.714244158758386e-06,
+      "loss": 0.3213,
+      "step": 8563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.137302355679623,
+      "learning_rate": 3.7131561951505947e-06,
+      "loss": 0.2933,
+      "step": 8564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7323651090490237,
+      "learning_rate": 3.7120682967886167e-06,
+      "loss": 0.352,
+      "step": 8565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31367522796944614,
+      "learning_rate": 3.71098046372761e-06,
+      "loss": 0.1158,
+      "step": 8566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.716763990425315,
+      "learning_rate": 3.709892696022731e-06,
+      "loss": 0.3046,
+      "step": 8567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.981631931611865,
+      "learning_rate": 3.7088049937291287e-06,
+      "loss": 0.3188,
+      "step": 8568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.42485464062791,
+      "learning_rate": 3.7077173569019566e-06,
+      "loss": 0.3197,
+      "step": 8569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.32768016891616103,
+      "learning_rate": 3.7066297855963584e-06,
+      "loss": 0.1169,
+      "step": 8570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.088885040235388,
+      "learning_rate": 3.7055422798674745e-06,
+      "loss": 0.3067,
+      "step": 8571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.000709729891122,
+      "learning_rate": 3.7044548397704472e-06,
+      "loss": 0.3558,
+      "step": 8572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.848009508317879,
+      "learning_rate": 3.703367465360408e-06,
+      "loss": 0.3098,
+      "step": 8573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.187586448836602,
+      "learning_rate": 3.7022801566924936e-06,
+      "loss": 0.3542,
+      "step": 8574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9988070832591003,
+      "learning_rate": 3.701192913821831e-06,
+      "loss": 0.3051,
+      "step": 8575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.246405837513622,
+      "learning_rate": 3.7001057368035446e-06,
+      "loss": 0.3189,
+      "step": 8576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.43555758215519,
+      "learning_rate": 3.6990186256927584e-06,
+      "loss": 0.336,
+      "step": 8577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 12.463699756341981,
+      "learning_rate": 3.6979315805445905e-06,
+      "loss": 0.3249,
+      "step": 8578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.547928253934768,
+      "learning_rate": 3.696844601414157e-06,
+      "loss": 0.3224,
+      "step": 8579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 17.009824115550003,
+      "learning_rate": 3.6957576883565705e-06,
+      "loss": 0.2981,
+      "step": 8580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.981797511598128,
+      "learning_rate": 3.6946708414269368e-06,
+      "loss": 0.3304,
+      "step": 8581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.348672499883253,
+      "learning_rate": 3.693584060680366e-06,
+      "loss": 0.3041,
+      "step": 8582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.090401831185819,
+      "learning_rate": 3.692497346171956e-06,
+      "loss": 0.3217,
+      "step": 8583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.271828293755197,
+      "learning_rate": 3.6914106979568098e-06,
+      "loss": 0.3286,
+      "step": 8584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.2994525365320029,
+      "learning_rate": 3.6903241160900204e-06,
+      "loss": 0.1196,
+      "step": 8585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.471107867558973,
+      "learning_rate": 3.689237600626679e-06,
+      "loss": 0.2967,
+      "step": 8586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.412760003589155,
+      "learning_rate": 3.6881511516218772e-06,
+      "loss": 0.3328,
+      "step": 8587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.696681599254114,
+      "learning_rate": 3.687064769130698e-06,
+      "loss": 0.3094,
+      "step": 8588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3755447505066543,
+      "learning_rate": 3.685978453208223e-06,
+      "loss": 0.3284,
+      "step": 8589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.637421354289064,
+      "learning_rate": 3.684892203909534e-06,
+      "loss": 0.3147,
+      "step": 8590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 16.787425443153534,
+      "learning_rate": 3.683806021289703e-06,
+      "loss": 0.3209,
+      "step": 8591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.665763422668422,
+      "learning_rate": 3.6827199054038043e-06,
+      "loss": 0.3383,
+      "step": 8592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.28432911516704,
+      "learning_rate": 3.681633856306904e-06,
+      "loss": 0.3162,
+      "step": 8593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.685756155646927,
+      "learning_rate": 3.6805478740540678e-06,
+      "loss": 0.3345,
+      "step": 8594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5801345360436034,
+      "learning_rate": 3.679461958700359e-06,
+      "loss": 0.298,
+      "step": 8595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.334445069871769,
+      "learning_rate": 3.6783761103008324e-06,
+      "loss": 0.3026,
+      "step": 8596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.236538700964453,
+      "learning_rate": 3.6772903289105477e-06,
+      "loss": 0.2961,
+      "step": 8597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.35437142133933974,
+      "learning_rate": 3.6762046145845544e-06,
+      "loss": 0.1212,
+      "step": 8598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.6297342798980825,
+      "learning_rate": 3.6751189673778975e-06,
+      "loss": 0.3286,
+      "step": 8599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.859105737277487,
+      "learning_rate": 3.6740333873456263e-06,
+      "loss": 0.3366,
+      "step": 8600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.456514179346155,
+      "learning_rate": 3.6729478745427794e-06,
+      "loss": 0.3189,
+      "step": 8601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.16594231030995,
+      "learning_rate": 3.6718624290243953e-06,
+      "loss": 0.3313,
+      "step": 8602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.121439285405922,
+      "learning_rate": 3.67077705084551e-06,
+      "loss": 0.3201,
+      "step": 8603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.319593996555127,
+      "learning_rate": 3.669691740061152e-06,
+      "loss": 0.115,
+      "step": 8604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.9174888021558045,
+      "learning_rate": 3.6686064967263513e-06,
+      "loss": 0.3194,
+      "step": 8605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.972164869141701,
+      "learning_rate": 3.667521320896128e-06,
+      "loss": 0.3071,
+      "step": 8606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5281097219712305,
+      "learning_rate": 3.6664362126255087e-06,
+      "loss": 0.3195,
+      "step": 8607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.100464090065067,
+      "learning_rate": 3.6653511719695077e-06,
+      "loss": 0.3189,
+      "step": 8608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.260770761632271,
+      "learning_rate": 3.6642661989831374e-06,
+      "loss": 0.3563,
+      "step": 8609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.6366520891174097,
+      "learning_rate": 3.6631812937214118e-06,
+      "loss": 0.3198,
+      "step": 8610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.3683209834222145,
+      "learning_rate": 3.6620964562393346e-06,
+      "loss": 0.2911,
+      "step": 8611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.495868334963013,
+      "learning_rate": 3.661011686591913e-06,
+      "loss": 0.3233,
+      "step": 8612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.368629296663916,
+      "learning_rate": 3.6599269848341457e-06,
+      "loss": 0.3091,
+      "step": 8613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.674808055863001,
+      "learning_rate": 3.6588423510210276e-06,
+      "loss": 0.2879,
+      "step": 8614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.995092046071646,
+      "learning_rate": 3.6577577852075536e-06,
+      "loss": 0.3117,
+      "step": 8615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.011185168037286,
+      "learning_rate": 3.656673287448713e-06,
+      "loss": 0.2924,
+      "step": 8616
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 14.065888563519707,
+      "learning_rate": 3.655588857799495e-06,
+      "loss": 0.326,
+      "step": 8617
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.209337824585103,
+      "learning_rate": 3.6545044963148795e-06,
+      "loss": 0.294,
+      "step": 8618
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.631845214713074,
+      "learning_rate": 3.6534202030498447e-06,
+      "loss": 0.3209,
+      "step": 8619
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.339495934535963,
+      "learning_rate": 3.6523359780593715e-06,
+      "loss": 0.3173,
+      "step": 8620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.2650344738413493,
+      "learning_rate": 3.651251821398427e-06,
+      "loss": 0.2906,
+      "step": 8621
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.985337928263551,
+      "learning_rate": 3.6501677331219854e-06,
+      "loss": 0.2728,
+      "step": 8622
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.886764319267541,
+      "learning_rate": 3.6490837132850097e-06,
+      "loss": 0.3488,
+      "step": 8623
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.388540078395927,
+      "learning_rate": 3.6479997619424605e-06,
+      "loss": 0.2926,
+      "step": 8624
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.36096380659067,
+      "learning_rate": 3.646915879149301e-06,
+      "loss": 0.2944,
+      "step": 8625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8969442379432406,
+      "learning_rate": 3.6458320649604834e-06,
+      "loss": 0.3199,
+      "step": 8626
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.100630984437303,
+      "learning_rate": 3.644748319430959e-06,
+      "loss": 0.338,
+      "step": 8627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.877128079482309,
+      "learning_rate": 3.6436646426156776e-06,
+      "loss": 0.3468,
+      "step": 8628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.31283884661669425,
+      "learning_rate": 3.6425810345695823e-06,
+      "loss": 0.126,
+      "step": 8629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.29310695287693544,
+      "learning_rate": 3.6414974953476174e-06,
+      "loss": 0.1178,
+      "step": 8630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.60830824351576,
+      "learning_rate": 3.6404140250047187e-06,
+      "loss": 0.3688,
+      "step": 8631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.550821877106817,
+      "learning_rate": 3.6393306235958185e-06,
+      "loss": 0.317,
+      "step": 8632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8037198035598467,
+      "learning_rate": 3.638247291175853e-06,
+      "loss": 0.3188,
+      "step": 8633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3251248882052185,
+      "learning_rate": 3.637164027799743e-06,
+      "loss": 0.3102,
+      "step": 8634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.140398427289543,
+      "learning_rate": 3.6360808335224178e-06,
+      "loss": 0.2848,
+      "step": 8635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.0068147837443755,
+      "learning_rate": 3.6349977083987963e-06,
+      "loss": 0.3344,
+      "step": 8636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 22.83337215886581,
+      "learning_rate": 3.633914652483792e-06,
+      "loss": 0.3395,
+      "step": 8637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3430519633605793,
+      "learning_rate": 3.6328316658323235e-06,
+      "loss": 0.3412,
+      "step": 8638
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 8.083328024126805,
+      "learning_rate": 3.6317487484992964e-06,
+      "loss": 0.2769,
+      "step": 8639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.867803201214721,
+      "learning_rate": 3.6306659005396195e-06,
+      "loss": 0.3527,
+      "step": 8640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.227064136143274,
+      "learning_rate": 3.6295831220081945e-06,
+      "loss": 0.3382,
+      "step": 8641
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.359537916767246,
+      "learning_rate": 3.628500412959921e-06,
+      "loss": 0.3397,
+      "step": 8642
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5163640276001304,
+      "learning_rate": 3.627417773449695e-06,
+      "loss": 0.3265,
+      "step": 8643
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.713658049532135,
+      "learning_rate": 3.626335203532406e-06,
+      "loss": 0.3383,
+      "step": 8644
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7741932675427643,
+      "learning_rate": 3.625252703262947e-06,
+      "loss": 0.3126,
+      "step": 8645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.016654259609859,
+      "learning_rate": 3.6241702726962015e-06,
+      "loss": 0.3153,
+      "step": 8646
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9160232376883255,
+      "learning_rate": 3.623087911887049e-06,
+      "loss": 0.3168,
+      "step": 8647
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.972602072418904,
+      "learning_rate": 3.6220056208903704e-06,
+      "loss": 0.3151,
+      "step": 8648
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.716008968850745,
+      "learning_rate": 3.6209233997610373e-06,
+      "loss": 0.3277,
+      "step": 8649
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.35913935227193,
+      "learning_rate": 3.6198412485539246e-06,
+      "loss": 0.3197,
+      "step": 8650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.894127398730685,
+      "learning_rate": 3.618759167323898e-06,
+      "loss": 0.3733,
+      "step": 8651
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.35516428071093,
+      "learning_rate": 3.6176771561258194e-06,
+      "loss": 0.3107,
+      "step": 8652
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.560015589761949,
+      "learning_rate": 3.616595215014551e-06,
+      "loss": 0.3332,
+      "step": 8653
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.249571570868275,
+      "learning_rate": 3.6155133440449496e-06,
+      "loss": 0.3087,
+      "step": 8654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 9.896096199652396,
+      "learning_rate": 3.6144315432718695e-06,
+      "loss": 0.2928,
+      "step": 8655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7777231075886086,
+      "learning_rate": 3.6133498127501587e-06,
+      "loss": 0.3152,
+      "step": 8656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.445158652221507,
+      "learning_rate": 3.612268152534662e-06,
+      "loss": 0.3343,
+      "step": 8657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.552105030587837,
+      "learning_rate": 3.6111865626802255e-06,
+      "loss": 0.2869,
+      "step": 8658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.29409604564096,
+      "learning_rate": 3.610105043241684e-06,
+      "loss": 0.3115,
+      "step": 8659
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.800596162747475,
+      "learning_rate": 3.609023594273878e-06,
+      "loss": 0.3107,
+      "step": 8660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.202564240872546,
+      "learning_rate": 3.6079422158316357e-06,
+      "loss": 0.3517,
+      "step": 8661
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.23073769332587,
+      "learning_rate": 3.6068609079697853e-06,
+      "loss": 0.3209,
+      "step": 8662
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.0307453919414975,
+      "learning_rate": 3.6057796707431537e-06,
+      "loss": 0.3304,
+      "step": 8663
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.3513040408426074,
+      "learning_rate": 3.6046985042065613e-06,
+      "loss": 0.3003,
+      "step": 8664
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.6420208291093354,
+      "learning_rate": 3.603617408414824e-06,
+      "loss": 0.3633,
+      "step": 8665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.8772739449044447,
+      "learning_rate": 3.602536383422757e-06,
+      "loss": 0.328,
+      "step": 8666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.464660039068109,
+      "learning_rate": 3.60145542928517e-06,
+      "loss": 0.319,
+      "step": 8667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.128261424071267,
+      "learning_rate": 3.600374546056871e-06,
+      "loss": 0.3132,
+      "step": 8668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.35545801051425,
+      "learning_rate": 3.599293733792663e-06,
+      "loss": 0.3191,
+      "step": 8669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.9547934451389364,
+      "learning_rate": 3.598212992547343e-06,
+      "loss": 0.3311,
+      "step": 8670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.173385728201676,
+      "learning_rate": 3.5971323223757103e-06,
+      "loss": 0.3221,
+      "step": 8671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.8093951978446725,
+      "learning_rate": 3.596051723332554e-06,
+      "loss": 0.3236,
+      "step": 8672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 7.170266567413728,
+      "learning_rate": 3.5949711954726674e-06,
+      "loss": 0.3139,
+      "step": 8673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.001510162099478,
+      "learning_rate": 3.593890738850833e-06,
+      "loss": 0.3382,
+      "step": 8674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.206371270812354,
+      "learning_rate": 3.5928103535218294e-06,
+      "loss": 0.362,
+      "step": 8675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.432000768406813,
+      "learning_rate": 3.59173003954044e-06,
+      "loss": 0.3238,
+      "step": 8676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 21.165918669949278,
+      "learning_rate": 3.5906497969614355e-06,
+      "loss": 0.3348,
+      "step": 8677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.112794002208354,
+      "learning_rate": 3.5895696258395873e-06,
+      "loss": 0.3209,
+      "step": 8678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.56790518027172,
+      "learning_rate": 3.5884895262296645e-06,
+      "loss": 0.3007,
+      "step": 8679
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4473384066382153,
+      "learning_rate": 3.587409498186427e-06,
+      "loss": 0.2726,
+      "step": 8680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.167224550362544,
+      "learning_rate": 3.586329541764637e-06,
+      "loss": 0.3296,
+      "step": 8681
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0202580435030226,
+      "learning_rate": 3.5852496570190485e-06,
+      "loss": 0.3349,
+      "step": 8682
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.433469876616673,
+      "learning_rate": 3.5841698440044183e-06,
+      "loss": 0.3172,
+      "step": 8683
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.394178579037431,
+      "learning_rate": 3.5830901027754916e-06,
+      "loss": 0.3325,
+      "step": 8684
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.392660699107114,
+      "learning_rate": 3.5820104333870125e-06,
+      "loss": 0.3492,
+      "step": 8685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.414696218175826,
+      "learning_rate": 3.5809308358937274e-06,
+      "loss": 0.3366,
+      "step": 8686
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.0711203082751215,
+      "learning_rate": 3.579851310350369e-06,
+      "loss": 0.3237,
+      "step": 8687
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.762815181588175,
+      "learning_rate": 3.5787718568116764e-06,
+      "loss": 0.321,
+      "step": 8688
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.5691905930217502,
+      "learning_rate": 3.5776924753323783e-06,
+      "loss": 0.3389,
+      "step": 8689
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3038856356777541,
+      "learning_rate": 3.5766131659672e-06,
+      "loss": 0.1119,
+      "step": 8690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.419317709524744,
+      "learning_rate": 3.5755339287708666e-06,
+      "loss": 0.2924,
+      "step": 8691
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.042157202103175,
+      "learning_rate": 3.5744547637980975e-06,
+      "loss": 0.312,
+      "step": 8692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.39813516186806,
+      "learning_rate": 3.5733756711036107e-06,
+      "loss": 0.3522,
+      "step": 8693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.500848086893026,
+      "learning_rate": 3.5722966507421152e-06,
+      "loss": 0.3183,
+      "step": 8694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.450632669929686,
+      "learning_rate": 3.5712177027683206e-06,
+      "loss": 0.339,
+      "step": 8695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.798740894880639,
+      "learning_rate": 3.570138827236933e-06,
+      "loss": 0.34,
+      "step": 8696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.615840662922747,
+      "learning_rate": 3.5690600242026523e-06,
+      "loss": 0.3515,
+      "step": 8697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.0204925283949597,
+      "learning_rate": 3.567981293720179e-06,
+      "loss": 0.3068,
+      "step": 8698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.328500929410092,
+      "learning_rate": 3.5669026358442047e-06,
+      "loss": 0.3341,
+      "step": 8699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.7775718541669763,
+      "learning_rate": 3.5658240506294196e-06,
+      "loss": 0.3169,
+      "step": 8700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.432162324931514,
+      "learning_rate": 3.564745538130512e-06,
+      "loss": 0.3348,
+      "step": 8701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.058019274494575,
+      "learning_rate": 3.563667098402164e-06,
+      "loss": 0.3071,
+      "step": 8702
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.625747267567016,
+      "learning_rate": 3.562588731499054e-06,
+      "loss": 0.3229,
+      "step": 8703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7551596195058923,
+      "learning_rate": 3.561510437475858e-06,
+      "loss": 0.344,
+      "step": 8704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 14.832292856101573,
+      "learning_rate": 3.5604322163872484e-06,
+      "loss": 0.3125,
+      "step": 8705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.450317592894977,
+      "learning_rate": 3.5593540682878946e-06,
+      "loss": 0.3206,
+      "step": 8706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.446290487396168,
+      "learning_rate": 3.5582759932324597e-06,
+      "loss": 0.3045,
+      "step": 8707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 12.391582884045535,
+      "learning_rate": 3.5571979912756017e-06,
+      "loss": 0.3188,
+      "step": 8708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.820709003144534,
+      "learning_rate": 3.5561200624719827e-06,
+      "loss": 0.3102,
+      "step": 8709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.802176647602665,
+      "learning_rate": 3.5550422068762525e-06,
+      "loss": 0.3071,
+      "step": 8710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.654144009065177,
+      "learning_rate": 3.5539644245430636e-06,
+      "loss": 0.3351,
+      "step": 8711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.111899792336968,
+      "learning_rate": 3.5528867155270604e-06,
+      "loss": 0.361,
+      "step": 8712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6833724201274642,
+      "learning_rate": 3.5518090798828832e-06,
+      "loss": 0.3489,
+      "step": 8713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.972970260271795,
+      "learning_rate": 3.5507315176651737e-06,
+      "loss": 0.359,
+      "step": 8714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.76371160201063,
+      "learning_rate": 3.549654028928565e-06,
+      "loss": 0.3241,
+      "step": 8715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8606980082695794,
+      "learning_rate": 3.5485766137276894e-06,
+      "loss": 0.3249,
+      "step": 8716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.4087601392752696,
+      "learning_rate": 3.547499272117172e-06,
+      "loss": 0.3365,
+      "step": 8717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.372039125958813,
+      "learning_rate": 3.546422004151638e-06,
+      "loss": 0.3554,
+      "step": 8718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.0662780393128175,
+      "learning_rate": 3.5453448098857085e-06,
+      "loss": 0.2968,
+      "step": 8719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9043562668433376,
+      "learning_rate": 3.5442676893739956e-06,
+      "loss": 0.3035,
+      "step": 8720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.260445773507498,
+      "learning_rate": 3.543190642671116e-06,
+      "loss": 0.3142,
+      "step": 8721
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.404233118812798,
+      "learning_rate": 3.542113669831677e-06,
+      "loss": 0.3036,
+      "step": 8722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.503798837764728,
+      "learning_rate": 3.541036770910281e-06,
+      "loss": 0.3201,
+      "step": 8723
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.061633096968052,
+      "learning_rate": 3.539959945961533e-06,
+      "loss": 0.3457,
+      "step": 8724
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.451018275343823,
+      "learning_rate": 3.5388831950400264e-06,
+      "loss": 0.3169,
+      "step": 8725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.746940692145462,
+      "learning_rate": 3.5378065182003597e-06,
+      "loss": 0.301,
+      "step": 8726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5882063358149785,
+      "learning_rate": 3.536729915497119e-06,
+      "loss": 0.3375,
+      "step": 8727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0126868004352376,
+      "learning_rate": 3.5356533869848918e-06,
+      "loss": 0.3228,
+      "step": 8728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9090866002245757,
+      "learning_rate": 3.5345769327182587e-06,
+      "loss": 0.2962,
+      "step": 8729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.461794434815245,
+      "learning_rate": 3.5335005527518006e-06,
+      "loss": 0.3323,
+      "step": 8730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.648133389865598,
+      "learning_rate": 3.5324242471400927e-06,
+      "loss": 0.313,
+      "step": 8731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 21.30611385049724,
+      "learning_rate": 3.531348015937705e-06,
+      "loss": 0.3314,
+      "step": 8732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.220861329825218,
+      "learning_rate": 3.5302718591992027e-06,
+      "loss": 0.3497,
+      "step": 8733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.190035587221634,
+      "learning_rate": 3.529195776979153e-06,
+      "loss": 0.328,
+      "step": 8734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 8.010564031158141,
+      "learning_rate": 3.5281197693321133e-06,
+      "loss": 0.3531,
+      "step": 8735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.179404974686961,
+      "learning_rate": 3.527043836312639e-06,
+      "loss": 0.3047,
+      "step": 8736
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.4137358396196245,
+      "learning_rate": 3.525967977975285e-06,
+      "loss": 0.3556,
+      "step": 8737
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.466090869145302,
+      "learning_rate": 3.5248921943745963e-06,
+      "loss": 0.3119,
+      "step": 8738
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4480302173561346,
+      "learning_rate": 3.5238164855651213e-06,
+      "loss": 0.2884,
+      "step": 8739
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.057147750992018,
+      "learning_rate": 3.5227408516013987e-06,
+      "loss": 0.2882,
+      "step": 8740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.342259151066055,
+      "learning_rate": 3.5216652925379646e-06,
+      "loss": 0.3212,
+      "step": 8741
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4607615837204553,
+      "learning_rate": 3.520589808429354e-06,
+      "loss": 0.3382,
+      "step": 8742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.99249311698555,
+      "learning_rate": 3.5195143993300953e-06,
+      "loss": 0.3107,
+      "step": 8743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0810275561304663,
+      "learning_rate": 3.5184390652947143e-06,
+      "loss": 0.301,
+      "step": 8744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.365410901624314,
+      "learning_rate": 3.5173638063777337e-06,
+      "loss": 0.3034,
+      "step": 8745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.227836338237756,
+      "learning_rate": 3.516288622633669e-06,
+      "loss": 0.3011,
+      "step": 8746
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5702033582174377,
+      "learning_rate": 3.5152135141170374e-06,
+      "loss": 0.3392,
+      "step": 8747
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.267404978836859,
+      "learning_rate": 3.5141384808823464e-06,
+      "loss": 0.3304,
+      "step": 8748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8040510675816748,
+      "learning_rate": 3.513063522984106e-06,
+      "loss": 0.3196,
+      "step": 8749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.32918733303941994,
+      "learning_rate": 3.5119886404768167e-06,
+      "loss": 0.1149,
+      "step": 8750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.107389427205292,
+      "learning_rate": 3.5109138334149756e-06,
+      "loss": 0.3294,
+      "step": 8751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 45.11878412050445,
+      "learning_rate": 3.509839101853082e-06,
+      "loss": 0.3013,
+      "step": 8752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.481978437316002,
+      "learning_rate": 3.508764445845624e-06,
+      "loss": 0.3275,
+      "step": 8753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.992954093846868,
+      "learning_rate": 3.5076898654470914e-06,
+      "loss": 0.3325,
+      "step": 8754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.239673875168192,
+      "learning_rate": 3.5066153607119658e-06,
+      "loss": 0.3207,
+      "step": 8755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3055474843740966,
+      "learning_rate": 3.5055409316947263e-06,
+      "loss": 0.3114,
+      "step": 8756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.496292529310704,
+      "learning_rate": 3.504466578449852e-06,
+      "loss": 0.3025,
+      "step": 8757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.860306767968694,
+      "learning_rate": 3.5033923010318117e-06,
+      "loss": 0.2878,
+      "step": 8758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.9109587345297565,
+      "learning_rate": 3.5023180994950764e-06,
+      "loss": 0.3195,
+      "step": 8759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.51677685930505,
+      "learning_rate": 3.501243973894109e-06,
+      "loss": 0.2991,
+      "step": 8760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.8975852307752294,
+      "learning_rate": 3.5001699242833682e-06,
+      "loss": 0.3247,
+      "step": 8761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.630477707637915,
+      "learning_rate": 3.4990959507173145e-06,
+      "loss": 0.2922,
+      "step": 8762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.08363969714308,
+      "learning_rate": 3.4980220532503974e-06,
+      "loss": 0.3273,
+      "step": 8763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3291639004503132,
+      "learning_rate": 3.4969482319370684e-06,
+      "loss": 0.114,
+      "step": 8764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7420652283141633,
+      "learning_rate": 3.495874486831772e-06,
+      "loss": 0.3236,
+      "step": 8765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.8465258736743033,
+      "learning_rate": 3.494800817988948e-06,
+      "loss": 0.3023,
+      "step": 8766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9983741508747848,
+      "learning_rate": 3.493727225463035e-06,
+      "loss": 0.3597,
+      "step": 8767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6812881635565544,
+      "learning_rate": 3.4926537093084654e-06,
+      "loss": 0.3034,
+      "step": 8768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.068037567706742,
+      "learning_rate": 3.4915802695796717e-06,
+      "loss": 0.3186,
+      "step": 8769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.833608677505199,
+      "learning_rate": 3.490506906331077e-06,
+      "loss": 0.3309,
+      "step": 8770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1675400453646225,
+      "learning_rate": 3.489433619617102e-06,
+      "loss": 0.3196,
+      "step": 8771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.7652255420813256,
+      "learning_rate": 3.4883604094921687e-06,
+      "loss": 0.334,
+      "step": 8772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.397079440694324,
+      "learning_rate": 3.487287276010688e-06,
+      "loss": 0.3014,
+      "step": 8773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.442047914707692,
+      "learning_rate": 3.48621421922707e-06,
+      "loss": 0.3352,
+      "step": 8774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.185644575905091,
+      "learning_rate": 3.485141239195724e-06,
+      "loss": 0.315,
+      "step": 8775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.3819844619223485,
+      "learning_rate": 3.4840683359710486e-06,
+      "loss": 0.3338,
+      "step": 8776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.304554404303062,
+      "learning_rate": 3.4829955096074464e-06,
+      "loss": 0.2935,
+      "step": 8777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.747388044762883,
+      "learning_rate": 3.4819227601593104e-06,
+      "loss": 0.3662,
+      "step": 8778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 9.943092040329434,
+      "learning_rate": 3.480850087681029e-06,
+      "loss": 0.3138,
+      "step": 8779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1070526243751853,
+      "learning_rate": 3.4797774922269923e-06,
+      "loss": 0.3016,
+      "step": 8780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.767218458447394,
+      "learning_rate": 3.478704973851581e-06,
+      "loss": 0.3025,
+      "step": 8781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5072180563488757,
+      "learning_rate": 3.477632532609177e-06,
+      "loss": 0.324,
+      "step": 8782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.882942332636521,
+      "learning_rate": 3.476560168554153e-06,
+      "loss": 0.3213,
+      "step": 8783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.5365513705333305,
+      "learning_rate": 3.475487881740879e-06,
+      "loss": 0.3313,
+      "step": 8784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.671524731252908,
+      "learning_rate": 3.474415672223726e-06,
+      "loss": 0.3128,
+      "step": 8785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.852052145217452,
+      "learning_rate": 3.473343540057054e-06,
+      "loss": 0.3037,
+      "step": 8786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.520230192794494,
+      "learning_rate": 3.472271485295225e-06,
+      "loss": 0.3207,
+      "step": 8787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.79447128177472,
+      "learning_rate": 3.471199507992594e-06,
+      "loss": 0.3013,
+      "step": 8788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.501139483317584,
+      "learning_rate": 3.4701276082035096e-06,
+      "loss": 0.3161,
+      "step": 8789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5804857133037253,
+      "learning_rate": 3.469055785982325e-06,
+      "loss": 0.3409,
+      "step": 8790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.719787152324805,
+      "learning_rate": 3.4679840413833787e-06,
+      "loss": 0.339,
+      "step": 8791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.354207031695661,
+      "learning_rate": 3.4669123744610146e-06,
+      "loss": 0.3159,
+      "step": 8792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.315885393512471,
+      "learning_rate": 3.4658407852695652e-06,
+      "loss": 0.3412,
+      "step": 8793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.7490059463182535,
+      "learning_rate": 3.4647692738633643e-06,
+      "loss": 0.3101,
+      "step": 8794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0201152476031194,
+      "learning_rate": 3.4636978402967404e-06,
+      "loss": 0.3227,
+      "step": 8795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.5793787527735343,
+      "learning_rate": 3.4626264846240143e-06,
+      "loss": 0.3193,
+      "step": 8796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.6997951154387785,
+      "learning_rate": 3.461555206899511e-06,
+      "loss": 0.311,
+      "step": 8797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.232189607384097,
+      "learning_rate": 3.460484007177543e-06,
+      "loss": 0.2963,
+      "step": 8798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 30.187057467147493,
+      "learning_rate": 3.4594128855124222e-06,
+      "loss": 0.2932,
+      "step": 8799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 10.902237650736884,
+      "learning_rate": 3.45834184195846e-06,
+      "loss": 0.3261,
+      "step": 8800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.028541324422556,
+      "learning_rate": 3.4572708765699568e-06,
+      "loss": 0.3296,
+      "step": 8801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.081427475613879,
+      "learning_rate": 3.456199989401218e-06,
+      "loss": 0.327,
+      "step": 8802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.975136741537669,
+      "learning_rate": 3.4551291805065356e-06,
+      "loss": 0.3465,
+      "step": 8803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 15.56789194731202,
+      "learning_rate": 3.454058449940203e-06,
+      "loss": 0.3303,
+      "step": 8804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.660223667560387,
+      "learning_rate": 3.4529877977565083e-06,
+      "loss": 0.3265,
+      "step": 8805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.949310950325331,
+      "learning_rate": 3.4519172240097364e-06,
+      "loss": 0.3282,
+      "step": 8806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.647063892802865,
+      "learning_rate": 3.450846728754169e-06,
+      "loss": 0.3282,
+      "step": 8807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.4812120472516135,
+      "learning_rate": 3.4497763120440817e-06,
+      "loss": 0.367,
+      "step": 8808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.604581673368928,
+      "learning_rate": 3.4487059739337446e-06,
+      "loss": 0.2983,
+      "step": 8809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.937986559254353,
+      "learning_rate": 3.4476357144774297e-06,
+      "loss": 0.341,
+      "step": 8810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1557298963319647,
+      "learning_rate": 3.446565533729401e-06,
+      "loss": 0.3065,
+      "step": 8811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.075376003548388,
+      "learning_rate": 3.445495431743915e-06,
+      "loss": 0.3238,
+      "step": 8812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.490475782697528,
+      "learning_rate": 3.444425408575234e-06,
+      "loss": 0.303,
+      "step": 8813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.28567747081402,
+      "learning_rate": 3.443355464277606e-06,
+      "loss": 0.3339,
+      "step": 8814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.072867872422119,
+      "learning_rate": 3.4422855989052827e-06,
+      "loss": 0.3191,
+      "step": 8815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.152280643525157,
+      "learning_rate": 3.441215812512508e-06,
+      "loss": 0.3269,
+      "step": 8816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.987957517895643,
+      "learning_rate": 3.4401461051535202e-06,
+      "loss": 0.3048,
+      "step": 8817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.037919496603898,
+      "learning_rate": 3.4390764768825575e-06,
+      "loss": 0.2616,
+      "step": 8818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3683083060418877,
+      "learning_rate": 3.438006927753852e-06,
+      "loss": 0.3222,
+      "step": 8819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.793982152524631,
+      "learning_rate": 3.436937457821634e-06,
+      "loss": 0.3197,
+      "step": 8820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3569040979302733,
+      "learning_rate": 3.435868067140126e-06,
+      "loss": 0.2926,
+      "step": 8821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.092192930338723,
+      "learning_rate": 3.4347987557635464e-06,
+      "loss": 0.3171,
+      "step": 8822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29230685930824774,
+      "learning_rate": 3.4337295237461173e-06,
+      "loss": 0.1158,
+      "step": 8823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.457391302610685,
+      "learning_rate": 3.4326603711420446e-06,
+      "loss": 0.3354,
+      "step": 8824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6388650773158986,
+      "learning_rate": 3.4315912980055433e-06,
+      "loss": 0.3035,
+      "step": 8825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.9160255747939554,
+      "learning_rate": 3.430522304390814e-06,
+      "loss": 0.3297,
+      "step": 8826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6387256671925727,
+      "learning_rate": 3.4294533903520553e-06,
+      "loss": 0.3116,
+      "step": 8827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.890143375268779,
+      "learning_rate": 3.4283845559434677e-06,
+      "loss": 0.3473,
+      "step": 8828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 11.377835634345725,
+      "learning_rate": 3.4273158012192407e-06,
+      "loss": 0.318,
+      "step": 8829
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.812069763706782,
+      "learning_rate": 3.426247126233564e-06,
+      "loss": 0.3204,
+      "step": 8830
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6287775174216184,
+      "learning_rate": 3.4251785310406203e-06,
+      "loss": 0.3056,
+      "step": 8831
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3150354142535959,
+      "learning_rate": 3.4241100156945907e-06,
+      "loss": 0.1165,
+      "step": 8832
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.371669326884425,
+      "learning_rate": 3.4230415802496513e-06,
+      "loss": 0.3594,
+      "step": 8833
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.6738366273508207,
+      "learning_rate": 3.4219732247599722e-06,
+      "loss": 0.329,
+      "step": 8834
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 6.020157054182436,
+      "learning_rate": 3.4209049492797252e-06,
+      "loss": 0.3222,
+      "step": 8835
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.3054824164077634,
+      "learning_rate": 3.419836753863072e-06,
+      "loss": 0.2937,
+      "step": 8836
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.734766093818795,
+      "learning_rate": 3.4187686385641707e-06,
+      "loss": 0.3179,
+      "step": 8837
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.337943255743587,
+      "learning_rate": 3.417700603437181e-06,
+      "loss": 0.3349,
+      "step": 8838
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.2964070795118188,
+      "learning_rate": 3.41663264853625e-06,
+      "loss": 0.3308,
+      "step": 8839
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.7447297888268722,
+      "learning_rate": 3.4155647739155308e-06,
+      "loss": 0.3377,
+      "step": 8840
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.29267385420557535,
+      "learning_rate": 3.4144969796291636e-06,
+      "loss": 0.1104,
+      "step": 8841
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 7.481606359693649,
+      "learning_rate": 3.413429265731287e-06,
+      "loss": 0.3198,
+      "step": 8842
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.2966797123706252,
+      "learning_rate": 3.41236163227604e-06,
+      "loss": 0.1101,
+      "step": 8843
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.6392240924398624,
+      "learning_rate": 3.4112940793175498e-06,
+      "loss": 0.2945,
+      "step": 8844
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.30493700516219424,
+      "learning_rate": 3.4102266069099465e-06,
+      "loss": 0.1158,
+      "step": 8845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.753731784003044,
+      "learning_rate": 3.4091592151073532e-06,
+      "loss": 0.3144,
+      "step": 8846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.005858657384829,
+      "learning_rate": 3.4080919039638862e-06,
+      "loss": 0.3138,
+      "step": 8847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.935576507179636,
+      "learning_rate": 3.407024673533665e-06,
+      "loss": 0.3147,
+      "step": 8848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5645652313158274,
+      "learning_rate": 3.4059575238707976e-06,
+      "loss": 0.2955,
+      "step": 8849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0341469884422403,
+      "learning_rate": 3.4048904550293903e-06,
+      "loss": 0.3028,
+      "step": 8850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 16.208629349372956,
+      "learning_rate": 3.4038234670635482e-06,
+      "loss": 0.3349,
+      "step": 8851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.843431064854775,
+      "learning_rate": 3.402756560027367e-06,
+      "loss": 0.3314,
+      "step": 8852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.54501596871545,
+      "learning_rate": 3.4016897339749454e-06,
+      "loss": 0.3101,
+      "step": 8853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.010317743962392,
+      "learning_rate": 3.4006229889603702e-06,
+      "loss": 0.2828,
+      "step": 8854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 9.464700827966675,
+      "learning_rate": 3.3995563250377293e-06,
+      "loss": 0.3131,
+      "step": 8855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.759707918069942,
+      "learning_rate": 3.3984897422611043e-06,
+      "loss": 0.3131,
+      "step": 8856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 21.90815889489206,
+      "learning_rate": 3.3974232406845723e-06,
+      "loss": 0.3243,
+      "step": 8857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.306660037008271,
+      "learning_rate": 3.3963568203622106e-06,
+      "loss": 0.32,
+      "step": 8858
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2007089722653026,
+      "learning_rate": 3.395290481348087e-06,
+      "loss": 0.3223,
+      "step": 8859
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6661802871794147,
+      "learning_rate": 3.394224223696264e-06,
+      "loss": 0.3239,
+      "step": 8860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7989169606219635,
+      "learning_rate": 3.3931580474608096e-06,
+      "loss": 0.311,
+      "step": 8861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8559875758918585,
+      "learning_rate": 3.392091952695775e-06,
+      "loss": 0.3207,
+      "step": 8862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3275585755697646,
+      "learning_rate": 3.391025939455219e-06,
+      "loss": 0.1221,
+      "step": 8863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.906612638277812,
+      "learning_rate": 3.389960007793189e-06,
+      "loss": 0.3743,
+      "step": 8864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2173116995111553,
+      "learning_rate": 3.388894157763727e-06,
+      "loss": 0.3134,
+      "step": 8865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9913042688739067,
+      "learning_rate": 3.3878283894208783e-06,
+      "loss": 0.3347,
+      "step": 8866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.15857833129414,
+      "learning_rate": 3.3867627028186776e-06,
+      "loss": 0.3499,
+      "step": 8867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.08222956872996,
+      "learning_rate": 3.3856970980111583e-06,
+      "loss": 0.3109,
+      "step": 8868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8005233458496384,
+      "learning_rate": 3.384631575052347e-06,
+      "loss": 0.2965,
+      "step": 8869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1051112845203206,
+      "learning_rate": 3.38356613399627e-06,
+      "loss": 0.2966,
+      "step": 8870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.675483878532522,
+      "learning_rate": 3.3825007748969485e-06,
+      "loss": 0.3283,
+      "step": 8871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8428162859644233,
+      "learning_rate": 3.381435497808394e-06,
+      "loss": 0.3008,
+      "step": 8872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4710022621358907,
+      "learning_rate": 3.380370302784624e-06,
+      "loss": 0.3164,
+      "step": 8873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2604211856609497,
+      "learning_rate": 3.3793051898796438e-06,
+      "loss": 0.285,
+      "step": 8874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.914479080086725,
+      "learning_rate": 3.3782401591474547e-06,
+      "loss": 0.332,
+      "step": 8875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.3874681650676814,
+      "learning_rate": 3.3771752106420596e-06,
+      "loss": 0.3339,
+      "step": 8876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.810258434159949,
+      "learning_rate": 3.3761103444174504e-06,
+      "loss": 0.3346,
+      "step": 8877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.277874638007083,
+      "learning_rate": 3.3750455605276227e-06,
+      "loss": 0.346,
+      "step": 8878
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.436798447658456,
+      "learning_rate": 3.373980859026561e-06,
+      "loss": 0.3181,
+      "step": 8879
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.745197374787657,
+      "learning_rate": 3.372916239968246e-06,
+      "loss": 0.3239,
+      "step": 8880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.488170215315223,
+      "learning_rate": 3.3718517034066593e-06,
+      "loss": 0.292,
+      "step": 8881
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9272416100159893,
+      "learning_rate": 3.3707872493957734e-06,
+      "loss": 0.332,
+      "step": 8882
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.551651904140145,
+      "learning_rate": 3.3697228779895585e-06,
+      "loss": 0.3248,
+      "step": 8883
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2107167410099584,
+      "learning_rate": 3.3686585892419825e-06,
+      "loss": 0.3201,
+      "step": 8884
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7050280083769866,
+      "learning_rate": 3.3675943832070036e-06,
+      "loss": 0.3214,
+      "step": 8885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.564644280973982,
+      "learning_rate": 3.366530259938584e-06,
+      "loss": 0.3367,
+      "step": 8886
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.219627930090962,
+      "learning_rate": 3.365466219490674e-06,
+      "loss": 0.3064,
+      "step": 8887
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.943514608492765,
+      "learning_rate": 3.3644022619172223e-06,
+      "loss": 0.3203,
+      "step": 8888
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9062326937239806,
+      "learning_rate": 3.3633383872721763e-06,
+      "loss": 0.3391,
+      "step": 8889
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30613337566449195,
+      "learning_rate": 3.362274595609474e-06,
+      "loss": 0.111,
+      "step": 8890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.29425622375869,
+      "learning_rate": 3.361210886983055e-06,
+      "loss": 0.3122,
+      "step": 8891
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30995548617564517,
+      "learning_rate": 3.360147261446851e-06,
+      "loss": 0.1166,
+      "step": 8892
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.5913162737449795,
+      "learning_rate": 3.3590837190547874e-06,
+      "loss": 0.3254,
+      "step": 8893
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.320461713980759,
+      "learning_rate": 3.3580202598607907e-06,
+      "loss": 0.1133,
+      "step": 8894
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0749547988278625,
+      "learning_rate": 3.3569568839187806e-06,
+      "loss": 0.2999,
+      "step": 8895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.332002371547879,
+      "learning_rate": 3.355893591282672e-06,
+      "loss": 0.3157,
+      "step": 8896
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.6792028197517315,
+      "learning_rate": 3.3548303820063764e-06,
+      "loss": 0.316,
+      "step": 8897
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.6748562557327964,
+      "learning_rate": 3.3537672561437984e-06,
+      "loss": 0.3161,
+      "step": 8898
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.0373303289876725,
+      "learning_rate": 3.3527042137488458e-06,
+      "loss": 0.2996,
+      "step": 8899
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.894234743051301,
+      "learning_rate": 3.3516412548754117e-06,
+      "loss": 0.3388,
+      "step": 8900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.671612251994345,
+      "learning_rate": 3.350578379577396e-06,
+      "loss": 0.3152,
+      "step": 8901
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.907258217754013,
+      "learning_rate": 3.3495155879086853e-06,
+      "loss": 0.2922,
+      "step": 8902
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.0827697132876715,
+      "learning_rate": 3.3484528799231643e-06,
+      "loss": 0.3357,
+      "step": 8903
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.095484965695594,
+      "learning_rate": 3.347390255674719e-06,
+      "loss": 0.3226,
+      "step": 8904
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.739045575747265,
+      "learning_rate": 3.346327715217223e-06,
+      "loss": 0.3066,
+      "step": 8905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.766712025915091,
+      "learning_rate": 3.345265258604552e-06,
+      "loss": 0.3297,
+      "step": 8906
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.014341611991793,
+      "learning_rate": 3.3442028858905716e-06,
+      "loss": 0.3798,
+      "step": 8907
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.4903121269875665,
+      "learning_rate": 3.3431405971291497e-06,
+      "loss": 0.3124,
+      "step": 8908
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2605280707976014,
+      "learning_rate": 3.342078392374146e-06,
+      "loss": 0.3501,
+      "step": 8909
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.583994728055591,
+      "learning_rate": 3.341016271679414e-06,
+      "loss": 0.3212,
+      "step": 8910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.692580627810526,
+      "learning_rate": 3.33995423509881e-06,
+      "loss": 0.3011,
+      "step": 8911
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.463516237224809,
+      "learning_rate": 3.3388922826861794e-06,
+      "loss": 0.3235,
+      "step": 8912
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.2877860026668615,
+      "learning_rate": 3.3378304144953623e-06,
+      "loss": 0.327,
+      "step": 8913
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.3843872176802026,
+      "learning_rate": 3.3367686305802035e-06,
+      "loss": 0.2896,
+      "step": 8914
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.702602354159114,
+      "learning_rate": 3.3357069309945328e-06,
+      "loss": 0.3259,
+      "step": 8915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.4986964500161095,
+      "learning_rate": 3.3346453157921854e-06,
+      "loss": 0.3195,
+      "step": 8916
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.88419374459863,
+      "learning_rate": 3.333583785026985e-06,
+      "loss": 0.3191,
+      "step": 8917
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.403365149324548,
+      "learning_rate": 3.332522338752752e-06,
+      "loss": 0.2685,
+      "step": 8918
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.967590780133046,
+      "learning_rate": 3.3314609770233074e-06,
+      "loss": 0.321,
+      "step": 8919
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.185046305612843,
+      "learning_rate": 3.3303996998924616e-06,
+      "loss": 0.3394,
+      "step": 8920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3362640064157164,
+      "learning_rate": 3.329338507414025e-06,
+      "loss": 0.1102,
+      "step": 8921
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.722825905075725,
+      "learning_rate": 3.328277399641804e-06,
+      "loss": 0.3237,
+      "step": 8922
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3284573431454269,
+      "learning_rate": 3.3272163766295947e-06,
+      "loss": 0.1223,
+      "step": 8923
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.30519797643608654,
+      "learning_rate": 3.326155438431198e-06,
+      "loss": 0.1004,
+      "step": 8924
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.671848124626758,
+      "learning_rate": 3.3250945851004046e-06,
+      "loss": 0.316,
+      "step": 8925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7596041060442067,
+      "learning_rate": 3.324033816690999e-06,
+      "loss": 0.3039,
+      "step": 8926
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0011484265173327,
+      "learning_rate": 3.3229731332567683e-06,
+      "loss": 0.2824,
+      "step": 8927
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 12.534812019993883,
+      "learning_rate": 3.321912534851489e-06,
+      "loss": 0.3128,
+      "step": 8928
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.650930803082423,
+      "learning_rate": 3.320852021528939e-06,
+      "loss": 0.3142,
+      "step": 8929
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.849572409757016,
+      "learning_rate": 3.3197915933428847e-06,
+      "loss": 0.355,
+      "step": 8930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.09086199819471,
+      "learning_rate": 3.3187312503470948e-06,
+      "loss": 0.3539,
+      "step": 8931
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9631396884854255,
+      "learning_rate": 3.3176709925953306e-06,
+      "loss": 0.346,
+      "step": 8932
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.729367831393781,
+      "learning_rate": 3.3166108201413465e-06,
+      "loss": 0.2922,
+      "step": 8933
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9132995003636144,
+      "learning_rate": 3.3155507330389004e-06,
+      "loss": 0.2997,
+      "step": 8934
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3283057531121843,
+      "learning_rate": 3.314490731341739e-06,
+      "loss": 0.1149,
+      "step": 8935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9887517155953467,
+      "learning_rate": 3.3134308151036044e-06,
+      "loss": 0.3182,
+      "step": 8936
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.686201720966138,
+      "learning_rate": 3.312370984378241e-06,
+      "loss": 0.3181,
+      "step": 8937
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.034678951261235,
+      "learning_rate": 3.3113112392193804e-06,
+      "loss": 0.3145,
+      "step": 8938
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.283739605337396,
+      "learning_rate": 3.310251579680757e-06,
+      "loss": 0.3294,
+      "step": 8939
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.5204739767745865,
+      "learning_rate": 3.3091920058160977e-06,
+      "loss": 0.295,
+      "step": 8940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5959827134695352,
+      "learning_rate": 3.3081325176791225e-06,
+      "loss": 0.3266,
+      "step": 8941
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8372886356438376,
+      "learning_rate": 3.3070731153235527e-06,
+      "loss": 0.3257,
+      "step": 8942
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.724289424296563,
+      "learning_rate": 3.3060137988031004e-06,
+      "loss": 0.2919,
+      "step": 8943
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.2936105499446751,
+      "learning_rate": 3.304954568171478e-06,
+      "loss": 0.1068,
+      "step": 8944
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.846225731647723,
+      "learning_rate": 3.3038954234823872e-06,
+      "loss": 0.2934,
+      "step": 8945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8318372470336914,
+      "learning_rate": 3.302836364789531e-06,
+      "loss": 0.3358,
+      "step": 8946
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3212367016737588,
+      "learning_rate": 3.3017773921466066e-06,
+      "loss": 0.1259,
+      "step": 8947
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.892333108461737,
+      "learning_rate": 3.3007185056073037e-06,
+      "loss": 0.3174,
+      "step": 8948
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8110892603405184,
+      "learning_rate": 3.299659705225313e-06,
+      "loss": 0.3326,
+      "step": 8949
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7404110638963806,
+      "learning_rate": 3.298600991054318e-06,
+      "loss": 0.3107,
+      "step": 8950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8330389620958267,
+      "learning_rate": 3.2975423631479943e-06,
+      "loss": 0.2906,
+      "step": 8951
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.8149721374007104,
+      "learning_rate": 3.2964838215600203e-06,
+      "loss": 0.3581,
+      "step": 8952
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.8169896165663375,
+      "learning_rate": 3.295425366344066e-06,
+      "loss": 0.3304,
+      "step": 8953
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.1118062083206546,
+      "learning_rate": 3.294366997553794e-06,
+      "loss": 0.3035,
+      "step": 8954
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.243301352483876,
+      "learning_rate": 3.2933087152428713e-06,
+      "loss": 0.3503,
+      "step": 8955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.484890383373716,
+      "learning_rate": 3.2922505194649508e-06,
+      "loss": 0.318,
+      "step": 8956
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.830460526768803,
+      "learning_rate": 3.291192410273688e-06,
+      "loss": 0.3168,
+      "step": 8957
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.32877013871339,
+      "learning_rate": 3.2901343877227287e-06,
+      "loss": 0.3182,
+      "step": 8958
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.5570087994824044,
+      "learning_rate": 3.289076451865719e-06,
+      "loss": 0.3278,
+      "step": 8959
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.95544464289463,
+      "learning_rate": 3.288018602756299e-06,
+      "loss": 0.3123,
+      "step": 8960
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.470381591229352,
+      "learning_rate": 3.2869608404481015e-06,
+      "loss": 0.3405,
+      "step": 8961
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9520331982737584,
+      "learning_rate": 3.2859031649947604e-06,
+      "loss": 0.3412,
+      "step": 8962
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.31117234144519895,
+      "learning_rate": 3.2848455764499003e-06,
+      "loss": 0.1232,
+      "step": 8963
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.675848537780098,
+      "learning_rate": 3.2837880748671423e-06,
+      "loss": 0.2934,
+      "step": 8964
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.180299951600039,
+      "learning_rate": 3.2827306603001073e-06,
+      "loss": 0.3272,
+      "step": 8965
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 6.499863726377749,
+      "learning_rate": 3.2816733328024046e-06,
+      "loss": 0.3188,
+      "step": 8966
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.543296550971289,
+      "learning_rate": 3.2806160924276474e-06,
+      "loss": 0.3328,
+      "step": 8967
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0387480279331527,
+      "learning_rate": 3.2795589392294376e-06,
+      "loss": 0.3188,
+      "step": 8968
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.025845716227599,
+      "learning_rate": 3.2785018732613744e-06,
+      "loss": 0.3244,
+      "step": 8969
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.696125665162951,
+      "learning_rate": 3.2774448945770553e-06,
+      "loss": 0.3203,
+      "step": 8970
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.560811730855951,
+      "learning_rate": 3.276388003230069e-06,
+      "loss": 0.3416,
+      "step": 8971
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.450896229196934,
+      "learning_rate": 3.2753311992740054e-06,
+      "loss": 0.3334,
+      "step": 8972
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.792604871494545,
+      "learning_rate": 3.274274482762445e-06,
+      "loss": 0.3205,
+      "step": 8973
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.7330802124016293,
+      "learning_rate": 3.273217853748964e-06,
+      "loss": 0.3093,
+      "step": 8974
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5422216676268747,
+      "learning_rate": 3.27216131228714e-06,
+      "loss": 0.3325,
+      "step": 8975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.924621345142096,
+      "learning_rate": 3.271104858430537e-06,
+      "loss": 0.3352,
+      "step": 8976
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.488767761226123,
+      "learning_rate": 3.270048492232724e-06,
+      "loss": 0.3473,
+      "step": 8977
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 10.68990168261805,
+      "learning_rate": 3.2689922137472596e-06,
+      "loss": 0.2986,
+      "step": 8978
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.848798977988201,
+      "learning_rate": 3.2679360230276967e-06,
+      "loss": 0.3598,
+      "step": 8979
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9181470992165734,
+      "learning_rate": 3.2668799201275916e-06,
+      "loss": 0.3167,
+      "step": 8980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.910327367169441,
+      "learning_rate": 3.2658239051004864e-06,
+      "loss": 0.3376,
+      "step": 8981
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.934147311294526,
+      "learning_rate": 3.2647679779999265e-06,
+      "loss": 0.329,
+      "step": 8982
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 8.374070259702037,
+      "learning_rate": 3.263712138879448e-06,
+      "loss": 0.3134,
+      "step": 8983
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.399981351566322,
+      "learning_rate": 3.2626563877925844e-06,
+      "loss": 0.3096,
+      "step": 8984
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.141278382664405,
+      "learning_rate": 3.261600724792866e-06,
+      "loss": 0.3322,
+      "step": 8985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.9841689002844234,
+      "learning_rate": 3.260545149933814e-06,
+      "loss": 0.3361,
+      "step": 8986
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.434464603967744,
+      "learning_rate": 3.259489663268953e-06,
+      "loss": 0.2979,
+      "step": 8987
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.526611033769907,
+      "learning_rate": 3.258434264851796e-06,
+      "loss": 0.3079,
+      "step": 8988
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.003402148796934,
+      "learning_rate": 3.2573789547358514e-06,
+      "loss": 0.3454,
+      "step": 8989
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4364435348539946,
+      "learning_rate": 3.256323732974631e-06,
+      "loss": 0.2774,
+      "step": 8990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2009236433682693,
+      "learning_rate": 3.255268599621635e-06,
+      "loss": 0.3249,
+      "step": 8991
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5937584370743445,
+      "learning_rate": 3.254213554730358e-06,
+      "loss": 0.2884,
+      "step": 8992
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1024157877391474,
+      "learning_rate": 3.253158598354296e-06,
+      "loss": 0.3105,
+      "step": 8993
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.476264868040448,
+      "learning_rate": 3.252103730546937e-06,
+      "loss": 0.3047,
+      "step": 8994
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30200197499922016,
+      "learning_rate": 3.2510489513617666e-06,
+      "loss": 0.1143,
+      "step": 8995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9603338805071377,
+      "learning_rate": 3.249994260852262e-06,
+      "loss": 0.3127,
+      "step": 8996
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8395211804661384,
+      "learning_rate": 3.248939659071898e-06,
+      "loss": 0.3112,
+      "step": 8997
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.070676395553002,
+      "learning_rate": 3.247885146074148e-06,
+      "loss": 0.3529,
+      "step": 8998
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4375829215670155,
+      "learning_rate": 3.246830721912475e-06,
+      "loss": 0.2995,
+      "step": 8999
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2571803658793277,
+      "learning_rate": 3.2457763866403442e-06,
+      "loss": 0.3159,
+      "step": 9000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.148204750197694,
+      "learning_rate": 3.2447221403112104e-06,
+      "loss": 0.3113,
+      "step": 9001
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.955233941656208,
+      "learning_rate": 3.243667982978525e-06,
+      "loss": 0.3593,
+      "step": 9002
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8982389787190805,
+      "learning_rate": 3.2426139146957397e-06,
+      "loss": 0.3011,
+      "step": 9003
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.906742535652686,
+      "learning_rate": 3.241559935516294e-06,
+      "loss": 0.3444,
+      "step": 9004
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.43247239672999,
+      "learning_rate": 3.240506045493631e-06,
+      "loss": 0.3168,
+      "step": 9005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.715814674584693,
+      "learning_rate": 3.2394522446811827e-06,
+      "loss": 0.3715,
+      "step": 9006
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.51635240847635,
+      "learning_rate": 3.2383985331323797e-06,
+      "loss": 0.3471,
+      "step": 9007
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.953686633880406,
+      "learning_rate": 3.2373449109006476e-06,
+      "loss": 0.313,
+      "step": 9008
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.3496144491529605,
+      "learning_rate": 3.2362913780394058e-06,
+      "loss": 0.2808,
+      "step": 9009
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.514540829793111,
+      "learning_rate": 3.235237934602074e-06,
+      "loss": 0.2842,
+      "step": 9010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.130557224807011,
+      "learning_rate": 3.2341845806420623e-06,
+      "loss": 0.3031,
+      "step": 9011
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.281252718046165,
+      "learning_rate": 3.2331313162127765e-06,
+      "loss": 0.293,
+      "step": 9012
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6415277419035927,
+      "learning_rate": 3.232078141367622e-06,
+      "loss": 0.3003,
+      "step": 9013
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.604794827729267,
+      "learning_rate": 3.2310250561599943e-06,
+      "loss": 0.3333,
+      "step": 9014
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.116747278819916,
+      "learning_rate": 3.2299720606432917e-06,
+      "loss": 0.3644,
+      "step": 9015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.534711954102813,
+      "learning_rate": 3.2289191548708997e-06,
+      "loss": 0.3146,
+      "step": 9016
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7501978562868175,
+      "learning_rate": 3.227866338896202e-06,
+      "loss": 0.3221,
+      "step": 9017
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.8020532570503383,
+      "learning_rate": 3.226813612772583e-06,
+      "loss": 0.3366,
+      "step": 9018
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.934720635348917,
+      "learning_rate": 3.225760976553414e-06,
+      "loss": 0.2815,
+      "step": 9019
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.083258709180319,
+      "learning_rate": 3.2247084302920693e-06,
+      "loss": 0.2981,
+      "step": 9020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.938501214314961,
+      "learning_rate": 3.2236559740419128e-06,
+      "loss": 0.3453,
+      "step": 9021
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.514293123752087,
+      "learning_rate": 3.222603607856307e-06,
+      "loss": 0.3125,
+      "step": 9022
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.4115228107212285,
+      "learning_rate": 3.2215513317886105e-06,
+      "loss": 0.3245,
+      "step": 9023
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.445729873348593,
+      "learning_rate": 3.220499145892173e-06,
+      "loss": 0.325,
+      "step": 9024
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.020459725209818,
+      "learning_rate": 3.219447050220346e-06,
+      "loss": 0.3517,
+      "step": 9025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.808842134896048,
+      "learning_rate": 3.2183950448264723e-06,
+      "loss": 0.308,
+      "step": 9026
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.094666719098748,
+      "learning_rate": 3.217343129763888e-06,
+      "loss": 0.2933,
+      "step": 9027
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.06629625294741,
+      "learning_rate": 3.2162913050859325e-06,
+      "loss": 0.3231,
+      "step": 9028
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.526858675992867,
+      "learning_rate": 3.215239570845932e-06,
+      "loss": 0.3108,
+      "step": 9029
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.164786124074065,
+      "learning_rate": 3.214187927097211e-06,
+      "loss": 0.3798,
+      "step": 9030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7859321012161287,
+      "learning_rate": 3.2131363738930937e-06,
+      "loss": 0.3121,
+      "step": 9031
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.865860620012564,
+      "learning_rate": 3.2120849112868936e-06,
+      "loss": 0.3403,
+      "step": 9032
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.174712272635474,
+      "learning_rate": 3.211033539331923e-06,
+      "loss": 0.3396,
+      "step": 9033
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5820579755020154,
+      "learning_rate": 3.2099822580814884e-06,
+      "loss": 0.3506,
+      "step": 9034
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.980014758477082,
+      "learning_rate": 3.2089310675888916e-06,
+      "loss": 0.3165,
+      "step": 9035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.383230338310903,
+      "learning_rate": 3.2078799679074324e-06,
+      "loss": 0.2978,
+      "step": 9036
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.152853922612817,
+      "learning_rate": 3.2068289590904e-06,
+      "loss": 0.3194,
+      "step": 9037
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6154934634690954,
+      "learning_rate": 3.205778041191088e-06,
+      "loss": 0.3267,
+      "step": 9038
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3207758378478053,
+      "learning_rate": 3.2047272142627774e-06,
+      "loss": 0.1097,
+      "step": 9039
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.4122493672799825,
+      "learning_rate": 3.2036764783587446e-06,
+      "loss": 0.3232,
+      "step": 9040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.231320872040723,
+      "learning_rate": 3.2026258335322696e-06,
+      "loss": 0.2948,
+      "step": 9041
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.803009488249549,
+      "learning_rate": 3.2015752798366184e-06,
+      "loss": 0.3166,
+      "step": 9042
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.42615465060344,
+      "learning_rate": 3.2005248173250593e-06,
+      "loss": 0.3316,
+      "step": 9043
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7860855013271344,
+      "learning_rate": 3.199474446050852e-06,
+      "loss": 0.2988,
+      "step": 9044
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.328958606394463,
+      "learning_rate": 3.1984241660672514e-06,
+      "loss": 0.3189,
+      "step": 9045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.86388259026596,
+      "learning_rate": 3.1973739774275115e-06,
+      "loss": 0.3531,
+      "step": 9046
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9118256236633457,
+      "learning_rate": 3.196323880184875e-06,
+      "loss": 0.3126,
+      "step": 9047
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.003994109638567,
+      "learning_rate": 3.1952738743925888e-06,
+      "loss": 0.3136,
+      "step": 9048
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.054187672677928,
+      "learning_rate": 3.194223960103888e-06,
+      "loss": 0.3421,
+      "step": 9049
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.130147843897508,
+      "learning_rate": 3.193174137372005e-06,
+      "loss": 0.3232,
+      "step": 9050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7388656338856574,
+      "learning_rate": 3.192124406250171e-06,
+      "loss": 0.2956,
+      "step": 9051
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1354862927656915,
+      "learning_rate": 3.191074766791605e-06,
+      "loss": 0.3384,
+      "step": 9052
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5355174979425645,
+      "learning_rate": 3.1900252190495316e-06,
+      "loss": 0.3238,
+      "step": 9053
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.29650020623560125,
+      "learning_rate": 3.1889757630771627e-06,
+      "loss": 0.1075,
+      "step": 9054
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.08836053063542,
+      "learning_rate": 3.1879263989277055e-06,
+      "loss": 0.3277,
+      "step": 9055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.170673868055886,
+      "learning_rate": 3.186877126654369e-06,
+      "loss": 0.3069,
+      "step": 9056
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8767458884885566,
+      "learning_rate": 3.1858279463103516e-06,
+      "loss": 0.3022,
+      "step": 9057
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30472503933180367,
+      "learning_rate": 3.18477885794885e-06,
+      "loss": 0.1122,
+      "step": 9058
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.737876277599949,
+      "learning_rate": 3.183729861623055e-06,
+      "loss": 0.3378,
+      "step": 9059
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.368371425516704,
+      "learning_rate": 3.18268095738615e-06,
+      "loss": 0.3392,
+      "step": 9060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5722538177831895,
+      "learning_rate": 3.1816321452913224e-06,
+      "loss": 0.3272,
+      "step": 9061
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.452375780997937,
+      "learning_rate": 3.1805834253917455e-06,
+      "loss": 0.3317,
+      "step": 9062
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.261834293137338,
+      "learning_rate": 3.179534797740591e-06,
+      "loss": 0.3367,
+      "step": 9063
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5234064919300447,
+      "learning_rate": 3.1784862623910307e-06,
+      "loss": 0.3069,
+      "step": 9064
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.11485191943227,
+      "learning_rate": 3.177437819396222e-06,
+      "loss": 0.302,
+      "step": 9065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5475944786798026,
+      "learning_rate": 3.1763894688093295e-06,
+      "loss": 0.3266,
+      "step": 9066
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.7534106734680646,
+      "learning_rate": 3.1753412106835036e-06,
+      "loss": 0.3179,
+      "step": 9067
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.122506553030889,
+      "learning_rate": 3.174293045071891e-06,
+      "loss": 0.3603,
+      "step": 9068
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.97138770918383,
+      "learning_rate": 3.1732449720276404e-06,
+      "loss": 0.3164,
+      "step": 9069
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8526404015169264,
+      "learning_rate": 3.1721969916038884e-06,
+      "loss": 0.3097,
+      "step": 9070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.7639585425070674,
+      "learning_rate": 3.1711491038537736e-06,
+      "loss": 0.327,
+      "step": 9071
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.43293682101874,
+      "learning_rate": 3.170101308830421e-06,
+      "loss": 0.3408,
+      "step": 9072
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9609067622618874,
+      "learning_rate": 3.1690536065869594e-06,
+      "loss": 0.338,
+      "step": 9073
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.47136343700558,
+      "learning_rate": 3.16800599717651e-06,
+      "loss": 0.3323,
+      "step": 9074
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9859087101117088,
+      "learning_rate": 3.1669584806521845e-06,
+      "loss": 0.3075,
+      "step": 9075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.30378279052366347,
+      "learning_rate": 3.1659110570671013e-06,
+      "loss": 0.1091,
+      "step": 9076
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.720289321908954,
+      "learning_rate": 3.1648637264743633e-06,
+      "loss": 0.2974,
+      "step": 9077
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 9.466202079599404,
+      "learning_rate": 3.1638164889270693e-06,
+      "loss": 0.322,
+      "step": 9078
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.073892275848359,
+      "learning_rate": 3.1627693444783235e-06,
+      "loss": 0.3256,
+      "step": 9079
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.129814892509375,
+      "learning_rate": 3.161722293181212e-06,
+      "loss": 0.3203,
+      "step": 9080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.988064148352971,
+      "learning_rate": 3.1606753350888276e-06,
+      "loss": 0.3367,
+      "step": 9081
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.160609846996354,
+      "learning_rate": 3.159628470254251e-06,
+      "loss": 0.3245,
+      "step": 9082
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2444266906194605,
+      "learning_rate": 3.1585816987305596e-06,
+      "loss": 0.3276,
+      "step": 9083
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.374746959045672,
+      "learning_rate": 3.157535020570829e-06,
+      "loss": 0.3482,
+      "step": 9084
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.668985289252401,
+      "learning_rate": 3.156488435828125e-06,
+      "loss": 0.3028,
+      "step": 9085
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.165952391800678,
+      "learning_rate": 3.1554419445555164e-06,
+      "loss": 0.3517,
+      "step": 9086
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.346292087916556,
+      "learning_rate": 3.1543955468060607e-06,
+      "loss": 0.3158,
+      "step": 9087
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.35743234924473,
+      "learning_rate": 3.1533492426328086e-06,
+      "loss": 0.3222,
+      "step": 9088
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3213154563906182,
+      "learning_rate": 3.152303032088817e-06,
+      "loss": 0.116,
+      "step": 9089
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 41.45778997294989,
+      "learning_rate": 3.1512569152271246e-06,
+      "loss": 0.3144,
+      "step": 9090
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.372879351040779,
+      "learning_rate": 3.1502108921007766e-06,
+      "loss": 0.335,
+      "step": 9091
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.114572723525513,
+      "learning_rate": 3.1491649627628074e-06,
+      "loss": 0.3358,
+      "step": 9092
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.0719888141286673,
+      "learning_rate": 3.1481191272662453e-06,
+      "loss": 0.3051,
+      "step": 9093
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.0633940848814,
+      "learning_rate": 3.1470733856641207e-06,
+      "loss": 0.3234,
+      "step": 9094
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.912736652825254,
+      "learning_rate": 3.1460277380094517e-06,
+      "loss": 0.2891,
+      "step": 9095
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 7.269390463252893,
+      "learning_rate": 3.1449821843552563e-06,
+      "loss": 0.3089,
+      "step": 9096
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5937260067211634,
+      "learning_rate": 3.143936724754547e-06,
+      "loss": 0.3251,
+      "step": 9097
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8550042154554496,
+      "learning_rate": 3.1428913592603285e-06,
+      "loss": 0.3385,
+      "step": 9098
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.22126375624085,
+      "learning_rate": 3.141846087925605e-06,
+      "loss": 0.3347,
+      "step": 9099
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.31963086880019,
+      "learning_rate": 3.140800910803375e-06,
+      "loss": 0.3296,
+      "step": 9100
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.580724098004392,
+      "learning_rate": 3.139755827946628e-06,
+      "loss": 0.3042,
+      "step": 9101
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1312617892103196,
+      "learning_rate": 3.1387108394083556e-06,
+      "loss": 0.3315,
+      "step": 9102
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.109690035503677,
+      "learning_rate": 3.137665945241537e-06,
+      "loss": 0.306,
+      "step": 9103
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.69733102452773,
+      "learning_rate": 3.1366211454991558e-06,
+      "loss": 0.3009,
+      "step": 9104
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3448360059417636,
+      "learning_rate": 3.1355764402341828e-06,
+      "loss": 0.1213,
+      "step": 9105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.887606324928629,
+      "learning_rate": 3.1345318294995847e-06,
+      "loss": 0.3235,
+      "step": 9106
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.428071536320367,
+      "learning_rate": 3.13348731334833e-06,
+      "loss": 0.3425,
+      "step": 9107
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9864585487461666,
+      "learning_rate": 3.1324428918333742e-06,
+      "loss": 0.3371,
+      "step": 9108
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 8.146573216775302,
+      "learning_rate": 3.131398565007675e-06,
+      "loss": 0.3141,
+      "step": 9109
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 10.305136366399019,
+      "learning_rate": 3.130354332924179e-06,
+      "loss": 0.3116,
+      "step": 9110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.315032497983523,
+      "learning_rate": 3.129310195635833e-06,
+      "loss": 0.3259,
+      "step": 9111
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3074471805616505,
+      "learning_rate": 3.128266153195577e-06,
+      "loss": 0.1153,
+      "step": 9112
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6845975808360993,
+      "learning_rate": 3.1272222056563445e-06,
+      "loss": 0.3176,
+      "step": 9113
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 14.268758489736227,
+      "learning_rate": 3.1261783530710686e-06,
+      "loss": 0.2992,
+      "step": 9114
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.9320369465208445,
+      "learning_rate": 3.125134595492674e-06,
+      "loss": 0.323,
+      "step": 9115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.8145344765189,
+      "learning_rate": 3.1240909329740786e-06,
+      "loss": 0.2863,
+      "step": 9116
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.461797064759532,
+      "learning_rate": 3.1230473655682034e-06,
+      "loss": 0.322,
+      "step": 9117
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.346813672383508,
+      "learning_rate": 3.1220038933279546e-06,
+      "loss": 0.3377,
+      "step": 9118
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.097819023843935,
+      "learning_rate": 3.120960516306243e-06,
+      "loss": 0.3446,
+      "step": 9119
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.331845828579167,
+      "learning_rate": 3.119917234555968e-06,
+      "loss": 0.3164,
+      "step": 9120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.6471600857542197,
+      "learning_rate": 3.1188740481300257e-06,
+      "loss": 0.3232,
+      "step": 9121
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.597311646734296,
+      "learning_rate": 3.1178309570813092e-06,
+      "loss": 0.3304,
+      "step": 9122
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.303133237935365,
+      "learning_rate": 3.116787961462703e-06,
+      "loss": 0.1113,
+      "step": 9123
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.536108356230827,
+      "learning_rate": 3.1157450613270936e-06,
+      "loss": 0.3226,
+      "step": 9124
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.032201848499094,
+      "learning_rate": 3.114702256727355e-06,
+      "loss": 0.339,
+      "step": 9125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.011223643848947,
+      "learning_rate": 3.113659547716359e-06,
+      "loss": 0.3251,
+      "step": 9126
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.59690625300531,
+      "learning_rate": 3.112616934346977e-06,
+      "loss": 0.3244,
+      "step": 9127
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.20316840109012,
+      "learning_rate": 3.1115744166720672e-06,
+      "loss": 0.3296,
+      "step": 9128
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.663807434523424,
+      "learning_rate": 3.1105319947444923e-06,
+      "loss": 0.2985,
+      "step": 9129
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 13.566988724726372,
+      "learning_rate": 3.109489668617103e-06,
+      "loss": 0.3313,
+      "step": 9130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.9774802549498864,
+      "learning_rate": 3.108447438342746e-06,
+      "loss": 0.3041,
+      "step": 9131
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.333474563573255,
+      "learning_rate": 3.107405303974268e-06,
+      "loss": 0.2984,
+      "step": 9132
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 5.015128074279973,
+      "learning_rate": 3.1063632655645048e-06,
+      "loss": 0.301,
+      "step": 9133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1197162140541965,
+      "learning_rate": 3.1053213231662917e-06,
+      "loss": 0.2698,
+      "step": 9134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5345379199705054,
+      "learning_rate": 3.104279476832457e-06,
+      "loss": 0.2794,
+      "step": 9135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.508987296748292,
+      "learning_rate": 3.103237726615822e-06,
+      "loss": 0.2766,
+      "step": 9136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.033170736007012,
+      "learning_rate": 3.1021960725692102e-06,
+      "loss": 0.293,
+      "step": 9137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.026766755348335,
+      "learning_rate": 3.1011545147454335e-06,
+      "loss": 0.3386,
+      "step": 9138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.288103002286617,
+      "learning_rate": 3.1001130531973002e-06,
+      "loss": 0.2728,
+      "step": 9139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2168926910621463,
+      "learning_rate": 3.099071687977617e-06,
+      "loss": 0.3202,
+      "step": 9140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5646001871753024,
+      "learning_rate": 3.098030419139181e-06,
+      "loss": 0.373,
+      "step": 9141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.4456596088486755,
+      "learning_rate": 3.096989246734789e-06,
+      "loss": 0.3408,
+      "step": 9142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.326694349141946,
+      "learning_rate": 3.09594817081723e-06,
+      "loss": 0.3104,
+      "step": 9143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.685852991362607,
+      "learning_rate": 3.094907191439287e-06,
+      "loss": 0.3343,
+      "step": 9144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.344589748469062,
+      "learning_rate": 3.0938663086537433e-06,
+      "loss": 0.344,
+      "step": 9145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9951283601470955,
+      "learning_rate": 3.0928255225133717e-06,
+      "loss": 0.3361,
+      "step": 9146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.628565503837688,
+      "learning_rate": 3.0917848330709434e-06,
+      "loss": 0.317,
+      "step": 9147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.811790920122874,
+      "learning_rate": 3.0907442403792233e-06,
+      "loss": 0.3304,
+      "step": 9148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.042974211329295,
+      "learning_rate": 3.0897037444909694e-06,
+      "loss": 0.3425,
+      "step": 9149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.671252719256132,
+      "learning_rate": 3.0886633454589416e-06,
+      "loss": 0.3338,
+      "step": 9150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8212555440094493,
+      "learning_rate": 3.0876230433358856e-06,
+      "loss": 0.3461,
+      "step": 9151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2763544839953704,
+      "learning_rate": 3.0865828381745515e-06,
+      "loss": 0.3179,
+      "step": 9152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.511661168344319,
+      "learning_rate": 3.0855427300276786e-06,
+      "loss": 0.322,
+      "step": 9153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.3780745430446455,
+      "learning_rate": 3.084502718947999e-06,
+      "loss": 0.3239,
+      "step": 9154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.513957139787239,
+      "learning_rate": 3.08346280498825e-06,
+      "loss": 0.3309,
+      "step": 9155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.612282214280781,
+      "learning_rate": 3.0824229882011516e-06,
+      "loss": 0.2901,
+      "step": 9156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.998635061718893,
+      "learning_rate": 3.0813832686394287e-06,
+      "loss": 0.348,
+      "step": 9157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3556517408696536,
+      "learning_rate": 3.0803436463557967e-06,
+      "loss": 0.3103,
+      "step": 9158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9871779276195087,
+      "learning_rate": 3.0793041214029644e-06,
+      "loss": 0.2859,
+      "step": 9159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.525439186886392,
+      "learning_rate": 3.0782646938336413e-06,
+      "loss": 0.3323,
+      "step": 9160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7220301354432213,
+      "learning_rate": 3.0772253637005244e-06,
+      "loss": 0.339,
+      "step": 9161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.059500964623454,
+      "learning_rate": 3.076186131056316e-06,
+      "loss": 0.2796,
+      "step": 9162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.226462033531091,
+      "learning_rate": 3.0751469959537027e-06,
+      "loss": 0.3348,
+      "step": 9163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.575601216047523,
+      "learning_rate": 3.0741079584453706e-06,
+      "loss": 0.2886,
+      "step": 9164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2270914446440826,
+      "learning_rate": 3.073069018584005e-06,
+      "loss": 0.3284,
+      "step": 9165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.7729561422807496,
+      "learning_rate": 3.0720301764222783e-06,
+      "loss": 0.3382,
+      "step": 9166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.644735797295677,
+      "learning_rate": 3.0709914320128666e-06,
+      "loss": 0.3336,
+      "step": 9167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5872227544296074,
+      "learning_rate": 3.0699527854084338e-06,
+      "loss": 0.2966,
+      "step": 9168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.900458683145582,
+      "learning_rate": 3.06891423666164e-06,
+      "loss": 0.3156,
+      "step": 9169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7879747172236105,
+      "learning_rate": 3.0678757858251452e-06,
+      "loss": 0.3126,
+      "step": 9170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.958410550187427,
+      "learning_rate": 3.0668374329516003e-06,
+      "loss": 0.3191,
+      "step": 9171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.008033527834155,
+      "learning_rate": 3.0657991780936502e-06,
+      "loss": 0.3159,
+      "step": 9172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8287116219427126,
+      "learning_rate": 3.064761021303939e-06,
+      "loss": 0.308,
+      "step": 9173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.317421798281584,
+      "learning_rate": 3.063722962635101e-06,
+      "loss": 0.3062,
+      "step": 9174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8131000734153604,
+      "learning_rate": 3.0626850021397714e-06,
+      "loss": 0.323,
+      "step": 9175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.346078832851333,
+      "learning_rate": 3.061647139870575e-06,
+      "loss": 0.2837,
+      "step": 9176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.328508508180722,
+      "learning_rate": 3.0606093758801327e-06,
+      "loss": 0.2885,
+      "step": 9177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.371268398959547,
+      "learning_rate": 3.0595717102210633e-06,
+      "loss": 0.3102,
+      "step": 9178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.359091449874637,
+      "learning_rate": 3.058534142945977e-06,
+      "loss": 0.2983,
+      "step": 9179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9465596200723727,
+      "learning_rate": 3.057496674107484e-06,
+      "loss": 0.3238,
+      "step": 9180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3095167729438048,
+      "learning_rate": 3.0564593037581836e-06,
+      "loss": 0.1009,
+      "step": 9181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3023237005474956,
+      "learning_rate": 3.0554220319506716e-06,
+      "loss": 0.1201,
+      "step": 9182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.163662797968025,
+      "learning_rate": 3.0543848587375436e-06,
+      "loss": 0.3383,
+      "step": 9183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8670193837389197,
+      "learning_rate": 3.0533477841713833e-06,
+      "loss": 0.3194,
+      "step": 9184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.4624176808858955,
+      "learning_rate": 3.0523108083047754e-06,
+      "loss": 0.3362,
+      "step": 9185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32795992113734856,
+      "learning_rate": 3.051273931190295e-06,
+      "loss": 0.1147,
+      "step": 9186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9026646265725184,
+      "learning_rate": 3.050237152880513e-06,
+      "loss": 0.3327,
+      "step": 9187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.29845068882007825,
+      "learning_rate": 3.049200473427999e-06,
+      "loss": 0.1053,
+      "step": 9188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.006193764203986,
+      "learning_rate": 3.0481638928853123e-06,
+      "loss": 0.3458,
+      "step": 9189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.341486283415114,
+      "learning_rate": 3.047127411305013e-06,
+      "loss": 0.2937,
+      "step": 9190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.200398179551201,
+      "learning_rate": 3.0460910287396516e-06,
+      "loss": 0.3204,
+      "step": 9191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 18.300655209295794,
+      "learning_rate": 3.0450547452417728e-06,
+      "loss": 0.3173,
+      "step": 9192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9550562010516304,
+      "learning_rate": 3.0440185608639215e-06,
+      "loss": 0.2981,
+      "step": 9193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.2088753188966526,
+      "learning_rate": 3.0429824756586333e-06,
+      "loss": 0.3347,
+      "step": 9194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.09111211886386,
+      "learning_rate": 3.0419464896784395e-06,
+      "loss": 0.3239,
+      "step": 9195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.765605617582746,
+      "learning_rate": 3.0409106029758685e-06,
+      "loss": 0.3263,
+      "step": 9196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.377413787828191,
+      "learning_rate": 3.0398748156034395e-06,
+      "loss": 0.3284,
+      "step": 9197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.696087432964233,
+      "learning_rate": 3.0388391276136717e-06,
+      "loss": 0.3321,
+      "step": 9198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.891365303361251,
+      "learning_rate": 3.0378035390590744e-06,
+      "loss": 0.2975,
+      "step": 9199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8806174798339375,
+      "learning_rate": 3.036768049992157e-06,
+      "loss": 0.3097,
+      "step": 9200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.037003087316495,
+      "learning_rate": 3.035732660465419e-06,
+      "loss": 0.3469,
+      "step": 9201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.924327991553876,
+      "learning_rate": 3.0346973705313564e-06,
+      "loss": 0.3268,
+      "step": 9202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.3160147687897754,
+      "learning_rate": 3.033662180242464e-06,
+      "loss": 0.3105,
+      "step": 9203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.492238986139472,
+      "learning_rate": 3.032627089651223e-06,
+      "loss": 0.3177,
+      "step": 9204
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8245651387473236,
+      "learning_rate": 3.03159209881012e-06,
+      "loss": 0.2739,
+      "step": 9205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6952865792403458,
+      "learning_rate": 3.0305572077716294e-06,
+      "loss": 0.3147,
+      "step": 9206
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.82128043793869,
+      "learning_rate": 3.02952241658822e-06,
+      "loss": 0.3237,
+      "step": 9207
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5892029820290063,
+      "learning_rate": 3.0284877253123618e-06,
+      "loss": 0.3202,
+      "step": 9208
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1637454092409025,
+      "learning_rate": 3.0274531339965147e-06,
+      "loss": 0.2985,
+      "step": 9209
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7070263099695935,
+      "learning_rate": 3.026418642693133e-06,
+      "loss": 0.3719,
+      "step": 9210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 10.967551446516696,
+      "learning_rate": 3.0253842514546704e-06,
+      "loss": 0.3021,
+      "step": 9211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.738856718675692,
+      "learning_rate": 3.024349960333569e-06,
+      "loss": 0.3263,
+      "step": 9212
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0049457365964245,
+      "learning_rate": 3.0233157693822747e-06,
+      "loss": 0.3399,
+      "step": 9213
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.322743549735181,
+      "learning_rate": 3.0222816786532196e-06,
+      "loss": 0.3124,
+      "step": 9214
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.768606220467343,
+      "learning_rate": 3.021247688198834e-06,
+      "loss": 0.3157,
+      "step": 9215
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.725503621002974,
+      "learning_rate": 3.020213798071546e-06,
+      "loss": 0.2745,
+      "step": 9216
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.303937046954423,
+      "learning_rate": 3.0191800083237743e-06,
+      "loss": 0.1185,
+      "step": 9217
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.5085590340150263,
+      "learning_rate": 3.0181463190079362e-06,
+      "loss": 0.2807,
+      "step": 9218
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.74550404988846,
+      "learning_rate": 3.0171127301764413e-06,
+      "loss": 0.3296,
+      "step": 9219
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7855235791775272,
+      "learning_rate": 3.016079241881692e-06,
+      "loss": 0.3351,
+      "step": 9220
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.5756436058614725,
+      "learning_rate": 3.015045854176093e-06,
+      "loss": 0.3371,
+      "step": 9221
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 14.58906108567871,
+      "learning_rate": 3.0140125671120356e-06,
+      "loss": 0.2871,
+      "step": 9222
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.2689425857453,
+      "learning_rate": 3.0129793807419127e-06,
+      "loss": 0.3129,
+      "step": 9223
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.276604754311221,
+      "learning_rate": 3.011946295118107e-06,
+      "loss": 0.3245,
+      "step": 9224
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.336498701751698,
+      "learning_rate": 3.0109133102929976e-06,
+      "loss": 0.3039,
+      "step": 9225
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3022228697378149,
+      "learning_rate": 3.009880426318962e-06,
+      "loss": 0.1032,
+      "step": 9226
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.6411342759287715,
+      "learning_rate": 3.008847643248366e-06,
+      "loss": 0.354,
+      "step": 9227
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.8918569279077735,
+      "learning_rate": 3.007814961133578e-06,
+      "loss": 0.3081,
+      "step": 9228
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.3721477416956933,
+      "learning_rate": 3.0067823800269558e-06,
+      "loss": 0.3399,
+      "step": 9229
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.636550550812921,
+      "learning_rate": 3.0057498999808503e-06,
+      "loss": 0.3643,
+      "step": 9230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.956445851623939,
+      "learning_rate": 3.0047175210476153e-06,
+      "loss": 0.3411,
+      "step": 9231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.142972262219213,
+      "learning_rate": 3.0036852432795925e-06,
+      "loss": 0.3094,
+      "step": 9232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.559103031214914,
+      "learning_rate": 3.00265306672912e-06,
+      "loss": 0.2779,
+      "step": 9233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.852771898106412,
+      "learning_rate": 3.0016209914485334e-06,
+      "loss": 0.3203,
+      "step": 9234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.282798270111778,
+      "learning_rate": 3.0005890174901585e-06,
+      "loss": 0.3405,
+      "step": 9235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.8476690948982983,
+      "learning_rate": 2.9995571449063223e-06,
+      "loss": 0.2833,
+      "step": 9236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6726350645979395,
+      "learning_rate": 2.998525373749338e-06,
+      "loss": 0.297,
+      "step": 9237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.4082200342006885,
+      "learning_rate": 2.9974937040715236e-06,
+      "loss": 0.3291,
+      "step": 9238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.812999687417906,
+      "learning_rate": 2.9964621359251844e-06,
+      "loss": 0.3391,
+      "step": 9239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.2007326446801745,
+      "learning_rate": 2.9954306693626223e-06,
+      "loss": 0.3156,
+      "step": 9240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.7030427401220067,
+      "learning_rate": 2.9943993044361386e-06,
+      "loss": 0.3358,
+      "step": 9241
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9015026613296784,
+      "learning_rate": 2.993368041198021e-06,
+      "loss": 0.3119,
+      "step": 9242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9391579189037085,
+      "learning_rate": 2.9923368797005612e-06,
+      "loss": 0.3308,
+      "step": 9243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.084902752564118,
+      "learning_rate": 2.9913058199960397e-06,
+      "loss": 0.3185,
+      "step": 9244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.918931569583835,
+      "learning_rate": 2.9902748621367315e-06,
+      "loss": 0.3288,
+      "step": 9245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.916045507039646,
+      "learning_rate": 2.9892440061749117e-06,
+      "loss": 0.2869,
+      "step": 9246
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.5692891502960125,
+      "learning_rate": 2.988213252162846e-06,
+      "loss": 0.307,
+      "step": 9247
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.787565557771287,
+      "learning_rate": 2.987182600152794e-06,
+      "loss": 0.3206,
+      "step": 9248
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8094267253027554,
+      "learning_rate": 2.986152050197015e-06,
+      "loss": 0.3181,
+      "step": 9249
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.100199968502183,
+      "learning_rate": 2.985121602347756e-06,
+      "loss": 0.3063,
+      "step": 9250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.59075300491386,
+      "learning_rate": 2.984091256657268e-06,
+      "loss": 0.2683,
+      "step": 9251
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8735819613320155,
+      "learning_rate": 2.983061013177788e-06,
+      "loss": 0.3067,
+      "step": 9252
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.6886369762286275,
+      "learning_rate": 2.982030871961552e-06,
+      "loss": 0.3081,
+      "step": 9253
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.6210573599044587,
+      "learning_rate": 2.981000833060793e-06,
+      "loss": 0.3268,
+      "step": 9254
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 16.022770790943436,
+      "learning_rate": 2.979970896527733e-06,
+      "loss": 0.3043,
+      "step": 9255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.32241594176774,
+      "learning_rate": 2.978941062414595e-06,
+      "loss": 0.2941,
+      "step": 9256
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.925572664273845,
+      "learning_rate": 2.9779113307735928e-06,
+      "loss": 0.2892,
+      "step": 9257
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.585709356111785,
+      "learning_rate": 2.976881701656934e-06,
+      "loss": 0.2832,
+      "step": 9258
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.000400877690716,
+      "learning_rate": 2.9758521751168266e-06,
+      "loss": 0.2986,
+      "step": 9259
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.533374652278073,
+      "learning_rate": 2.9748227512054672e-06,
+      "loss": 0.3089,
+      "step": 9260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 31.294092224927933,
+      "learning_rate": 2.9737934299750514e-06,
+      "loss": 0.3288,
+      "step": 9261
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4389957101432325,
+      "learning_rate": 2.9727642114777677e-06,
+      "loss": 0.3084,
+      "step": 9262
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.617503250943977,
+      "learning_rate": 2.9717350957657976e-06,
+      "loss": 0.3428,
+      "step": 9263
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1977278610879996,
+      "learning_rate": 2.9707060828913226e-06,
+      "loss": 0.3083,
+      "step": 9264
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.490816229436969,
+      "learning_rate": 2.969677172906513e-06,
+      "loss": 0.3313,
+      "step": 9265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.226422782931338,
+      "learning_rate": 2.968648365863541e-06,
+      "loss": 0.3165,
+      "step": 9266
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.0054458100069747,
+      "learning_rate": 2.9676196618145664e-06,
+      "loss": 0.3212,
+      "step": 9267
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.966815452914892,
+      "learning_rate": 2.9665910608117454e-06,
+      "loss": 0.3042,
+      "step": 9268
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 5.116142325370468,
+      "learning_rate": 2.9655625629072337e-06,
+      "loss": 0.2863,
+      "step": 9269
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.345510116804334,
+      "learning_rate": 2.9645341681531758e-06,
+      "loss": 0.3365,
+      "step": 9270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.852860097718681,
+      "learning_rate": 2.9635058766017155e-06,
+      "loss": 0.2996,
+      "step": 9271
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.9872866805083236,
+      "learning_rate": 2.962477688304989e-06,
+      "loss": 0.3223,
+      "step": 9272
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.1576963384449823,
+      "learning_rate": 2.9614496033151265e-06,
+      "loss": 0.2748,
+      "step": 9273
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.973134552402703,
+      "learning_rate": 2.9604216216842553e-06,
+      "loss": 0.3472,
+      "step": 9274
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.666648465277831,
+      "learning_rate": 2.959393743464495e-06,
+      "loss": 0.3627,
+      "step": 9275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.32280346200541726,
+      "learning_rate": 2.9583659687079634e-06,
+      "loss": 0.1173,
+      "step": 9276
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.06410120601963,
+      "learning_rate": 2.9573382974667697e-06,
+      "loss": 0.3047,
+      "step": 9277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4374332213915975,
+      "learning_rate": 2.9563107297930183e-06,
+      "loss": 0.3253,
+      "step": 9278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.9821937516835035,
+      "learning_rate": 2.955283265738811e-06,
+      "loss": 0.3318,
+      "step": 9279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.29615518803213325,
+      "learning_rate": 2.954255905356242e-06,
+      "loss": 0.1073,
+      "step": 9280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.162680078289285,
+      "learning_rate": 2.9532286486973983e-06,
+      "loss": 0.3602,
+      "step": 9281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 11.597954653349781,
+      "learning_rate": 2.9522014958143674e-06,
+      "loss": 0.3437,
+      "step": 9282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.111168839944875,
+      "learning_rate": 2.9511744467592266e-06,
+      "loss": 0.3204,
+      "step": 9283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.518886903654011,
+      "learning_rate": 2.9501475015840497e-06,
+      "loss": 0.3146,
+      "step": 9284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.035035779791845,
+      "learning_rate": 2.949120660340906e-06,
+      "loss": 0.3195,
+      "step": 9285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.650240321580107,
+      "learning_rate": 2.948093923081857e-06,
+      "loss": 0.3153,
+      "step": 9286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2356055382239335,
+      "learning_rate": 2.9470672898589625e-06,
+      "loss": 0.329,
+      "step": 9287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6332638975674825,
+      "learning_rate": 2.9460407607242715e-06,
+      "loss": 0.3299,
+      "step": 9288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31814916555959116,
+      "learning_rate": 2.9450143357298355e-06,
+      "loss": 0.116,
+      "step": 9289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.530442423473564,
+      "learning_rate": 2.943988014927695e-06,
+      "loss": 0.3213,
+      "step": 9290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.411479608196148,
+      "learning_rate": 2.942961798369884e-06,
+      "loss": 0.3218,
+      "step": 9291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3030917637540182,
+      "learning_rate": 2.941935686108439e-06,
+      "loss": 0.1123,
+      "step": 9292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.4445614520226915,
+      "learning_rate": 2.940909678195381e-06,
+      "loss": 0.3449,
+      "step": 9293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.30379562515669,
+      "learning_rate": 2.939883774682736e-06,
+      "loss": 0.2904,
+      "step": 9294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.124746999240081,
+      "learning_rate": 2.9388579756225168e-06,
+      "loss": 0.2914,
+      "step": 9295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8847000291761398,
+      "learning_rate": 2.937832281066731e-06,
+      "loss": 0.3332,
+      "step": 9296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.138152621062123,
+      "learning_rate": 2.936806691067389e-06,
+      "loss": 0.3242,
+      "step": 9297
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.866748848316517,
+      "learning_rate": 2.935781205676487e-06,
+      "loss": 0.3427,
+      "step": 9298
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.309524034325277,
+      "learning_rate": 2.93475582494602e-06,
+      "loss": 0.3114,
+      "step": 9299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.5205309247906555,
+      "learning_rate": 2.933730548927978e-06,
+      "loss": 0.3248,
+      "step": 9300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.494739565043533,
+      "learning_rate": 2.9327053776743426e-06,
+      "loss": 0.3362,
+      "step": 9301
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.069569218143307,
+      "learning_rate": 2.931680311237095e-06,
+      "loss": 0.3394,
+      "step": 9302
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.557481460177879,
+      "learning_rate": 2.930655349668205e-06,
+      "loss": 0.3191,
+      "step": 9303
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.744901129968667,
+      "learning_rate": 2.9296304930196447e-06,
+      "loss": 0.3364,
+      "step": 9304
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.243636500872045,
+      "learning_rate": 2.928605741343374e-06,
+      "loss": 0.3208,
+      "step": 9305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.574892238934352,
+      "learning_rate": 2.9275810946913487e-06,
+      "loss": 0.299,
+      "step": 9306
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2177417322274735,
+      "learning_rate": 2.9265565531155238e-06,
+      "loss": 0.3244,
+      "step": 9307
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.080356010437651,
+      "learning_rate": 2.925532116667843e-06,
+      "loss": 0.3023,
+      "step": 9308
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.31089546720510686,
+      "learning_rate": 2.924507785400249e-06,
+      "loss": 0.1074,
+      "step": 9309
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.359798607343355,
+      "learning_rate": 2.9234835593646793e-06,
+      "loss": 0.3189,
+      "step": 9310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7050100307573435,
+      "learning_rate": 2.9224594386130604e-06,
+      "loss": 0.2973,
+      "step": 9311
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.626269591853685,
+      "learning_rate": 2.921435423197321e-06,
+      "loss": 0.3391,
+      "step": 9312
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3867231096059376,
+      "learning_rate": 2.920411513169378e-06,
+      "loss": 0.3626,
+      "step": 9313
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.669261677188004,
+      "learning_rate": 2.9193877085811483e-06,
+      "loss": 0.3231,
+      "step": 9314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.342950121546021,
+      "learning_rate": 2.918364009484541e-06,
+      "loss": 0.3318,
+      "step": 9315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.050903046813849,
+      "learning_rate": 2.9173404159314577e-06,
+      "loss": 0.3315,
+      "step": 9316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.937121344715841,
+      "learning_rate": 2.916316927973799e-06,
+      "loss": 0.3048,
+      "step": 9317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6841098806216213,
+      "learning_rate": 2.915293545663457e-06,
+      "loss": 0.3261,
+      "step": 9318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.109628693629916,
+      "learning_rate": 2.914270269052319e-06,
+      "loss": 0.3342,
+      "step": 9319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4430418194975205,
+      "learning_rate": 2.91324709819227e-06,
+      "loss": 0.3447,
+      "step": 9320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.170428335992376,
+      "learning_rate": 2.9122240331351815e-06,
+      "loss": 0.3284,
+      "step": 9321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5066546161043193,
+      "learning_rate": 2.911201073932932e-06,
+      "loss": 0.3299,
+      "step": 9322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.4634794575339765,
+      "learning_rate": 2.9101782206373837e-06,
+      "loss": 0.3068,
+      "step": 9323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.877886954146468,
+      "learning_rate": 2.9091554733003975e-06,
+      "loss": 0.2944,
+      "step": 9324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0297242379254024,
+      "learning_rate": 2.90813283197383e-06,
+      "loss": 0.2927,
+      "step": 9325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.28857494770200587,
+      "learning_rate": 2.9071102967095304e-06,
+      "loss": 0.1142,
+      "step": 9326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.363901026262003,
+      "learning_rate": 2.906087867559345e-06,
+      "loss": 0.3671,
+      "step": 9327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 100.1058327083039,
+      "learning_rate": 2.905065544575114e-06,
+      "loss": 0.3096,
+      "step": 9328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.508506137227468,
+      "learning_rate": 2.9040433278086677e-06,
+      "loss": 0.3076,
+      "step": 9329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.330027715008663,
+      "learning_rate": 2.9030212173118366e-06,
+      "loss": 0.3292,
+      "step": 9330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.022293516857236,
+      "learning_rate": 2.901999213136445e-06,
+      "loss": 0.3257,
+      "step": 9331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.821860833887102,
+      "learning_rate": 2.900977315334309e-06,
+      "loss": 0.3118,
+      "step": 9332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.6447676574273515,
+      "learning_rate": 2.8999555239572454e-06,
+      "loss": 0.3112,
+      "step": 9333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.92689265029943,
+      "learning_rate": 2.8989338390570527e-06,
+      "loss": 0.2989,
+      "step": 9334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1178866175163593,
+      "learning_rate": 2.897912260685542e-06,
+      "loss": 0.3189,
+      "step": 9335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6537880858738547,
+      "learning_rate": 2.8968907888945013e-06,
+      "loss": 0.3247,
+      "step": 9336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0007470741244724,
+      "learning_rate": 2.8958694237357297e-06,
+      "loss": 0.292,
+      "step": 9337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.412011973274222,
+      "learning_rate": 2.8948481652610068e-06,
+      "loss": 0.3432,
+      "step": 9338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5950259660162347,
+      "learning_rate": 2.8938270135221134e-06,
+      "loss": 0.3266,
+      "step": 9339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2508043525141,
+      "learning_rate": 2.8928059685708258e-06,
+      "loss": 0.3132,
+      "step": 9340
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.315232298543406,
+      "learning_rate": 2.8917850304589124e-06,
+      "loss": 0.2945,
+      "step": 9341
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.23463239755481,
+      "learning_rate": 2.8907641992381384e-06,
+      "loss": 0.3611,
+      "step": 9342
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.806120543894992,
+      "learning_rate": 2.88974347496026e-06,
+      "loss": 0.3652,
+      "step": 9343
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.841818674109018,
+      "learning_rate": 2.8887228576770302e-06,
+      "loss": 0.3436,
+      "step": 9344
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.061775446980245,
+      "learning_rate": 2.8877023474401976e-06,
+      "loss": 0.313,
+      "step": 9345
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.049462441727743,
+      "learning_rate": 2.8866819443015047e-06,
+      "loss": 0.3239,
+      "step": 9346
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.913750122671095,
+      "learning_rate": 2.885661648312687e-06,
+      "loss": 0.3265,
+      "step": 9347
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.526962982200963,
+      "learning_rate": 2.8846414595254785e-06,
+      "loss": 0.285,
+      "step": 9348
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.663540949106458,
+      "learning_rate": 2.8836213779915988e-06,
+      "loss": 0.3355,
+      "step": 9349
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.391473002036314,
+      "learning_rate": 2.882601403762776e-06,
+      "loss": 0.3013,
+      "step": 9350
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6000805857701286,
+      "learning_rate": 2.881581536890718e-06,
+      "loss": 0.3382,
+      "step": 9351
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.7866740018949985,
+      "learning_rate": 2.880561777427141e-06,
+      "loss": 0.3064,
+      "step": 9352
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.282241286466443,
+      "learning_rate": 2.8795421254237443e-06,
+      "loss": 0.3365,
+      "step": 9353
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.70231742498146,
+      "learning_rate": 2.878522580932228e-06,
+      "loss": 0.3055,
+      "step": 9354
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.221569866619467,
+      "learning_rate": 2.8775031440042868e-06,
+      "loss": 0.3077,
+      "step": 9355
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0019973629069168,
+      "learning_rate": 2.8764838146916048e-06,
+      "loss": 0.2977,
+      "step": 9356
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 8.191399488780803,
+      "learning_rate": 2.8754645930458656e-06,
+      "loss": 0.307,
+      "step": 9357
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.026912443828602,
+      "learning_rate": 2.874445479118748e-06,
+      "loss": 0.354,
+      "step": 9358
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3559118936541648,
+      "learning_rate": 2.873426472961921e-06,
+      "loss": 0.2919,
+      "step": 9359
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.3287750208799647,
+      "learning_rate": 2.8724075746270513e-06,
+      "loss": 0.2916,
+      "step": 9360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.189753718549863,
+      "learning_rate": 2.8713887841658017e-06,
+      "loss": 0.3634,
+      "step": 9361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.288235659971159,
+      "learning_rate": 2.87037010162982e-06,
+      "loss": 0.3104,
+      "step": 9362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 26.180835378227755,
+      "learning_rate": 2.8693515270707654e-06,
+      "loss": 0.3453,
+      "step": 9363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.6081330865285106,
+      "learning_rate": 2.8683330605402725e-06,
+      "loss": 0.2907,
+      "step": 9364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.511375183020742,
+      "learning_rate": 2.867314702089988e-06,
+      "loss": 0.3359,
+      "step": 9365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.197469795175125,
+      "learning_rate": 2.8662964517715398e-06,
+      "loss": 0.3515,
+      "step": 9366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 18.114528245581035,
+      "learning_rate": 2.865278309636556e-06,
+      "loss": 0.2872,
+      "step": 9367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.161902690535186,
+      "learning_rate": 2.864260275736662e-06,
+      "loss": 0.3543,
+      "step": 9368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8493304637691668,
+      "learning_rate": 2.863242350123467e-06,
+      "loss": 0.2974,
+      "step": 9369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.972750274583479,
+      "learning_rate": 2.862224532848591e-06,
+      "loss": 0.3137,
+      "step": 9370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.968901051855156,
+      "learning_rate": 2.8612068239636335e-06,
+      "loss": 0.2942,
+      "step": 9371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.94235236063931,
+      "learning_rate": 2.860189223520196e-06,
+      "loss": 0.3346,
+      "step": 9372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4851110811583923,
+      "learning_rate": 2.859171731569874e-06,
+      "loss": 0.32,
+      "step": 9373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.323510182670194,
+      "learning_rate": 2.8581543481642558e-06,
+      "loss": 0.3063,
+      "step": 9374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.645035057244083,
+      "learning_rate": 2.8571370733549257e-06,
+      "loss": 0.3007,
+      "step": 9375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2943369630493518,
+      "learning_rate": 2.856119907193463e-06,
+      "loss": 0.2958,
+      "step": 9376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.631939761461812,
+      "learning_rate": 2.8551028497314346e-06,
+      "loss": 0.3372,
+      "step": 9377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.2986173448246112,
+      "learning_rate": 2.8540859010204164e-06,
+      "loss": 0.1018,
+      "step": 9378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.555468635848508,
+      "learning_rate": 2.8530690611119627e-06,
+      "loss": 0.2858,
+      "step": 9379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.85984436282857,
+      "learning_rate": 2.8520523300576315e-06,
+      "loss": 0.2863,
+      "step": 9380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.298780889349961,
+      "learning_rate": 2.851035707908974e-06,
+      "loss": 0.2808,
+      "step": 9381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.530655779798847,
+      "learning_rate": 2.850019194717535e-06,
+      "loss": 0.3415,
+      "step": 9382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.977062312481699,
+      "learning_rate": 2.8490027905348555e-06,
+      "loss": 0.3225,
+      "step": 9383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 12.614799816349423,
+      "learning_rate": 2.8479864954124643e-06,
+      "loss": 0.3085,
+      "step": 9384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.9708582142168565,
+      "learning_rate": 2.846970309401897e-06,
+      "loss": 0.3266,
+      "step": 9385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.508029083964004,
+      "learning_rate": 2.8459542325546706e-06,
+      "loss": 0.325,
+      "step": 9386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.951454228882249,
+      "learning_rate": 2.8449382649223045e-06,
+      "loss": 0.2925,
+      "step": 9387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.6099017158023825,
+      "learning_rate": 2.843922406556311e-06,
+      "loss": 0.3175,
+      "step": 9388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2432338416502153,
+      "learning_rate": 2.842906657508198e-06,
+      "loss": 0.3078,
+      "step": 9389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.710388108656313,
+      "learning_rate": 2.84189101782946e-06,
+      "loss": 0.3017,
+      "step": 9390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0518800589810438,
+      "learning_rate": 2.8408754875716006e-06,
+      "loss": 0.3444,
+      "step": 9391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2783193389690157,
+      "learning_rate": 2.8398600667861032e-06,
+      "loss": 0.2859,
+      "step": 9392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.8558688427487935,
+      "learning_rate": 2.8388447555244532e-06,
+      "loss": 0.303,
+      "step": 9393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.227253808941557,
+      "learning_rate": 2.8378295538381295e-06,
+      "loss": 0.3034,
+      "step": 9394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.032657800786509,
+      "learning_rate": 2.8368144617786066e-06,
+      "loss": 0.331,
+      "step": 9395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4423861829930176,
+      "learning_rate": 2.8357994793973516e-06,
+      "loss": 0.3006,
+      "step": 9396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.8856374863597924,
+      "learning_rate": 2.834784606745821e-06,
+      "loss": 0.3247,
+      "step": 9397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.158846452955258,
+      "learning_rate": 2.83376984387548e-06,
+      "loss": 0.3241,
+      "step": 9398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.806153286274767,
+      "learning_rate": 2.8327551908377724e-06,
+      "loss": 0.3284,
+      "step": 9399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.810700727253729,
+      "learning_rate": 2.8317406476841465e-06,
+      "loss": 0.2957,
+      "step": 9400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.1709915055576605,
+      "learning_rate": 2.8307262144660404e-06,
+      "loss": 0.3279,
+      "step": 9401
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.653896635749932,
+      "learning_rate": 2.829711891234889e-06,
+      "loss": 0.29,
+      "step": 9402
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.593978558020135,
+      "learning_rate": 2.82869767804212e-06,
+      "loss": 0.3407,
+      "step": 9403
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.438604777106031,
+      "learning_rate": 2.8276835749391583e-06,
+      "loss": 0.321,
+      "step": 9404
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.719622501366053,
+      "learning_rate": 2.826669581977419e-06,
+      "loss": 0.3321,
+      "step": 9405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 6.565964214617669,
+      "learning_rate": 2.8256556992083127e-06,
+      "loss": 0.3151,
+      "step": 9406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.2381663621129135,
+      "learning_rate": 2.824641926683248e-06,
+      "loss": 0.3216,
+      "step": 9407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1580607824743576,
+      "learning_rate": 2.823628264453625e-06,
+      "loss": 0.334,
+      "step": 9408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.668221663638304,
+      "learning_rate": 2.822614712570839e-06,
+      "loss": 0.2797,
+      "step": 9409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 7.934315942219548,
+      "learning_rate": 2.821601271086275e-06,
+      "loss": 0.3227,
+      "step": 9410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.705055479600213,
+      "learning_rate": 2.8205879400513234e-06,
+      "loss": 0.2842,
+      "step": 9411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.292556249560609,
+      "learning_rate": 2.8195747195173562e-06,
+      "loss": 0.3227,
+      "step": 9412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.217490072192803,
+      "learning_rate": 2.8185616095357525e-06,
+      "loss": 0.3399,
+      "step": 9413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 9.285463493069468,
+      "learning_rate": 2.817548610157873e-06,
+      "loss": 0.3025,
+      "step": 9414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.542814789598009,
+      "learning_rate": 2.816535721435082e-06,
+      "loss": 0.3183,
+      "step": 9415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3646891666478964,
+      "learning_rate": 2.8155229434187336e-06,
+      "loss": 0.1162,
+      "step": 9416
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.7350353503969242,
+      "learning_rate": 2.81451027616018e-06,
+      "loss": 0.3202,
+      "step": 9417
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.1465954046767206,
+      "learning_rate": 2.8134977197107662e-06,
+      "loss": 0.3066,
+      "step": 9418
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.212823988528186,
+      "learning_rate": 2.812485274121828e-06,
+      "loss": 0.2987,
+      "step": 9419
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.72870271034845,
+      "learning_rate": 2.811472939444699e-06,
+      "loss": 0.3004,
+      "step": 9420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.446589055657715,
+      "learning_rate": 2.8104607157307086e-06,
+      "loss": 0.3405,
+      "step": 9421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8975278939340683,
+      "learning_rate": 2.8094486030311776e-06,
+      "loss": 0.2878,
+      "step": 9422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5338608277487786,
+      "learning_rate": 2.808436601397423e-06,
+      "loss": 0.2822,
+      "step": 9423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.892785386290677,
+      "learning_rate": 2.8074247108807568e-06,
+      "loss": 0.3111,
+      "step": 9424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.094188180282545,
+      "learning_rate": 2.8064129315324783e-06,
+      "loss": 0.3083,
+      "step": 9425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.1081720021092325,
+      "learning_rate": 2.805401263403896e-06,
+      "loss": 0.2916,
+      "step": 9426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.614536175743133,
+      "learning_rate": 2.8043897065462968e-06,
+      "loss": 0.3162,
+      "step": 9427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.65497749400812,
+      "learning_rate": 2.8033782610109707e-06,
+      "loss": 0.2856,
+      "step": 9428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.538705321012816,
+      "learning_rate": 2.8023669268492002e-06,
+      "loss": 0.3158,
+      "step": 9429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7556737295198332,
+      "learning_rate": 2.801355704112263e-06,
+      "loss": 0.2816,
+      "step": 9430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1783161306459493,
+      "learning_rate": 2.8003445928514312e-06,
+      "loss": 0.2978,
+      "step": 9431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.218341160625351,
+      "learning_rate": 2.7993335931179674e-06,
+      "loss": 0.3005,
+      "step": 9432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.026071840016837,
+      "learning_rate": 2.798322704963132e-06,
+      "loss": 0.3721,
+      "step": 9433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.875299163760959,
+      "learning_rate": 2.797311928438181e-06,
+      "loss": 0.3146,
+      "step": 9434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.047671425070195,
+      "learning_rate": 2.7963012635943624e-06,
+      "loss": 0.3117,
+      "step": 9435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4410212672761915,
+      "learning_rate": 2.7952907104829186e-06,
+      "loss": 0.3539,
+      "step": 9436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.497162704133692,
+      "learning_rate": 2.7942802691550896e-06,
+      "loss": 0.3368,
+      "step": 9437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.255299178925132,
+      "learning_rate": 2.793269939662101e-06,
+      "loss": 0.3321,
+      "step": 9438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.508464054531784,
+      "learning_rate": 2.7922597220551856e-06,
+      "loss": 0.3357,
+      "step": 9439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8871488249020176,
+      "learning_rate": 2.7912496163855563e-06,
+      "loss": 0.3138,
+      "step": 9440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.255689507917378,
+      "learning_rate": 2.7902396227044366e-06,
+      "loss": 0.2837,
+      "step": 9441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.726324576656273,
+      "learning_rate": 2.7892297410630283e-06,
+      "loss": 0.3537,
+      "step": 9442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.180469500939944,
+      "learning_rate": 2.788219971512537e-06,
+      "loss": 0.3215,
+      "step": 9443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4962577943126814,
+      "learning_rate": 2.787210314104162e-06,
+      "loss": 0.3022,
+      "step": 9444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.804548191217695,
+      "learning_rate": 2.786200768889089e-06,
+      "loss": 0.3072,
+      "step": 9445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.61270329594324,
+      "learning_rate": 2.7851913359185133e-06,
+      "loss": 0.3572,
+      "step": 9446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.231981509625081,
+      "learning_rate": 2.7841820152436084e-06,
+      "loss": 0.3601,
+      "step": 9447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.177125028336843,
+      "learning_rate": 2.7831728069155504e-06,
+      "loss": 0.3243,
+      "step": 9448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.461750758769203,
+      "learning_rate": 2.7821637109855102e-06,
+      "loss": 0.3309,
+      "step": 9449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8996819852280353,
+      "learning_rate": 2.7811547275046495e-06,
+      "loss": 0.3074,
+      "step": 9450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.184520680380728,
+      "learning_rate": 2.780145856524127e-06,
+      "loss": 0.3418,
+      "step": 9451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.764119846589398,
+      "learning_rate": 2.7791370980950967e-06,
+      "loss": 0.2918,
+      "step": 9452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.972155197932548,
+      "learning_rate": 2.778128452268698e-06,
+      "loss": 0.2806,
+      "step": 9453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.515213203065745,
+      "learning_rate": 2.7771199190960806e-06,
+      "loss": 0.2808,
+      "step": 9454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.2553116829295,
+      "learning_rate": 2.7761114986283722e-06,
+      "loss": 0.3228,
+      "step": 9455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.408954084111396,
+      "learning_rate": 2.7751031909167046e-06,
+      "loss": 0.3056,
+      "step": 9456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8817801725307937,
+      "learning_rate": 2.7740949960122014e-06,
+      "loss": 0.3378,
+      "step": 9457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3446560136946113,
+      "learning_rate": 2.77308691396598e-06,
+      "loss": 0.2977,
+      "step": 9458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.041195381306621,
+      "learning_rate": 2.772078944829154e-06,
+      "loss": 0.3099,
+      "step": 9459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.393815315689763,
+      "learning_rate": 2.7710710886528246e-06,
+      "loss": 0.2752,
+      "step": 9460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9056628402221603,
+      "learning_rate": 2.7700633454880992e-06,
+      "loss": 0.2987,
+      "step": 9461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.36119090789837427,
+      "learning_rate": 2.7690557153860677e-06,
+      "loss": 0.1341,
+      "step": 9462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.681881592022924,
+      "learning_rate": 2.76804819839782e-06,
+      "loss": 0.3046,
+      "step": 9463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.511112019486092,
+      "learning_rate": 2.7670407945744403e-06,
+      "loss": 0.3398,
+      "step": 9464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.539506674505616,
+      "learning_rate": 2.7660335039670083e-06,
+      "loss": 0.3265,
+      "step": 9465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0740933952910665,
+      "learning_rate": 2.76502632662659e-06,
+      "loss": 0.2661,
+      "step": 9466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2959987430618709,
+      "learning_rate": 2.764019262604258e-06,
+      "loss": 0.1125,
+      "step": 9467
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 9.16693846235666,
+      "learning_rate": 2.7630123119510676e-06,
+      "loss": 0.3205,
+      "step": 9468
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.996705473826685,
+      "learning_rate": 2.7620054747180757e-06,
+      "loss": 0.3418,
+      "step": 9469
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5816512115671353,
+      "learning_rate": 2.7609987509563318e-06,
+      "loss": 0.2976,
+      "step": 9470
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.4665771931640466,
+      "learning_rate": 2.759992140716877e-06,
+      "loss": 0.2988,
+      "step": 9471
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6287943800197695,
+      "learning_rate": 2.7589856440507523e-06,
+      "loss": 0.3416,
+      "step": 9472
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.19853830802725,
+      "learning_rate": 2.757979261008983e-06,
+      "loss": 0.3278,
+      "step": 9473
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.967800107158975,
+      "learning_rate": 2.7569729916426024e-06,
+      "loss": 0.324,
+      "step": 9474
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.356716546212952,
+      "learning_rate": 2.7559668360026253e-06,
+      "loss": 0.326,
+      "step": 9475
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.308177155617027,
+      "learning_rate": 2.7549607941400677e-06,
+      "loss": 0.3372,
+      "step": 9476
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32067351177990466,
+      "learning_rate": 2.7539548661059373e-06,
+      "loss": 0.1218,
+      "step": 9477
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4425445316132497,
+      "learning_rate": 2.752949051951238e-06,
+      "loss": 0.3108,
+      "step": 9478
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3464832107873366,
+      "learning_rate": 2.7519433517269665e-06,
+      "loss": 0.3013,
+      "step": 9479
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3245760770398936,
+      "learning_rate": 2.7509377654841165e-06,
+      "loss": 0.3051,
+      "step": 9480
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9475414330342344,
+      "learning_rate": 2.749932293273668e-06,
+      "loss": 0.3258,
+      "step": 9481
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.49610962481824,
+      "learning_rate": 2.7489269351466037e-06,
+      "loss": 0.3303,
+      "step": 9482
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.62908649521545,
+      "learning_rate": 2.7479216911538974e-06,
+      "loss": 0.3344,
+      "step": 9483
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.989475525900691,
+      "learning_rate": 2.746916561346517e-06,
+      "loss": 0.3076,
+      "step": 9484
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1808533498247153,
+      "learning_rate": 2.745911545775427e-06,
+      "loss": 0.3334,
+      "step": 9485
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.595099926826492,
+      "learning_rate": 2.7449066444915774e-06,
+      "loss": 0.3254,
+      "step": 9486
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.529329730947197,
+      "learning_rate": 2.7439018575459275e-06,
+      "loss": 0.2875,
+      "step": 9487
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5548338355009004,
+      "learning_rate": 2.742897184989414e-06,
+      "loss": 0.3211,
+      "step": 9488
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.293809829172334,
+      "learning_rate": 2.741892626872984e-06,
+      "loss": 0.3081,
+      "step": 9489
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.282874065853074,
+      "learning_rate": 2.740888183247565e-06,
+      "loss": 0.327,
+      "step": 9490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5164439868140804,
+      "learning_rate": 2.739883854164086e-06,
+      "loss": 0.3276,
+      "step": 9491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.537223183879322,
+      "learning_rate": 2.7388796396734696e-06,
+      "loss": 0.317,
+      "step": 9492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.772403560365089,
+      "learning_rate": 2.73787553982663e-06,
+      "loss": 0.3373,
+      "step": 9493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 7.169064877540515,
+      "learning_rate": 2.7368715546744813e-06,
+      "loss": 0.3095,
+      "step": 9494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 25.826208691157618,
+      "learning_rate": 2.735867684267922e-06,
+      "loss": 0.3112,
+      "step": 9495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.594369825660882,
+      "learning_rate": 2.734863928657854e-06,
+      "loss": 0.3319,
+      "step": 9496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.526990906475138,
+      "learning_rate": 2.7338602878951692e-06,
+      "loss": 0.3204,
+      "step": 9497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.426255534854227,
+      "learning_rate": 2.732856762030756e-06,
+      "loss": 0.3064,
+      "step": 9498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7677873036291136,
+      "learning_rate": 2.73185335111549e-06,
+      "loss": 0.3133,
+      "step": 9499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8890960199809896,
+      "learning_rate": 2.7308500552002544e-06,
+      "loss": 0.2922,
+      "step": 9500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.341449915280752,
+      "learning_rate": 2.72984687433591e-06,
+      "loss": 0.3139,
+      "step": 9501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.20108153103745,
+      "learning_rate": 2.7288438085733288e-06,
+      "loss": 0.2798,
+      "step": 9502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1047771765306056,
+      "learning_rate": 2.727840857963362e-06,
+      "loss": 0.3517,
+      "step": 9503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8786046667625733,
+      "learning_rate": 2.7268380225568635e-06,
+      "loss": 0.3249,
+      "step": 9504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5195732542706155,
+      "learning_rate": 2.7258353024046803e-06,
+      "loss": 0.3191,
+      "step": 9505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6414895686007416,
+      "learning_rate": 2.72483269755765e-06,
+      "loss": 0.3064,
+      "step": 9506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.62664184883513,
+      "learning_rate": 2.723830208066611e-06,
+      "loss": 0.2995,
+      "step": 9507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1973075913054423,
+      "learning_rate": 2.7228278339823876e-06,
+      "loss": 0.3272,
+      "step": 9508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.276243677394521,
+      "learning_rate": 2.7218255753558043e-06,
+      "loss": 0.3018,
+      "step": 9509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.7355679517197995,
+      "learning_rate": 2.720823432237677e-06,
+      "loss": 0.3346,
+      "step": 9510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.1036671038305466,
+      "learning_rate": 2.719821404678817e-06,
+      "loss": 0.3131,
+      "step": 9511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.664242341190207,
+      "learning_rate": 2.7188194927300294e-06,
+      "loss": 0.3223,
+      "step": 9512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29612154561755266,
+      "learning_rate": 2.7178176964421155e-06,
+      "loss": 0.1112,
+      "step": 9513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.345378608854423,
+      "learning_rate": 2.716816015865862e-06,
+      "loss": 0.3017,
+      "step": 9514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 12.568055445294924,
+      "learning_rate": 2.7158144510520645e-06,
+      "loss": 0.3328,
+      "step": 9515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.760583215749231,
+      "learning_rate": 2.714813002051497e-06,
+      "loss": 0.3459,
+      "step": 9516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6671474361500174,
+      "learning_rate": 2.7138116689149423e-06,
+      "loss": 0.3135,
+      "step": 9517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.6590110592199783,
+      "learning_rate": 2.712810451693165e-06,
+      "loss": 0.3037,
+      "step": 9518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3204200247904536,
+      "learning_rate": 2.711809350436931e-06,
+      "loss": 0.3291,
+      "step": 9519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.033326483254201,
+      "learning_rate": 2.710808365197e-06,
+      "loss": 0.3694,
+      "step": 9520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.622726319518599,
+      "learning_rate": 2.7098074960241184e-06,
+      "loss": 0.3144,
+      "step": 9521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6916295888953137,
+      "learning_rate": 2.7088067429690403e-06,
+      "loss": 0.2874,
+      "step": 9522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.509214060742286,
+      "learning_rate": 2.7078061060825002e-06,
+      "loss": 0.2978,
+      "step": 9523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.316367196837478,
+      "learning_rate": 2.706805585415234e-06,
+      "loss": 0.3322,
+      "step": 9524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.445460870980694,
+      "learning_rate": 2.705805181017972e-06,
+      "loss": 0.3351,
+      "step": 9525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.62650830356485,
+      "learning_rate": 2.704804892941435e-06,
+      "loss": 0.3269,
+      "step": 9526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.764361746082227,
+      "learning_rate": 2.70380472123634e-06,
+      "loss": 0.3306,
+      "step": 9527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.515597140287391,
+      "learning_rate": 2.702804665953401e-06,
+      "loss": 0.3057,
+      "step": 9528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5525856939825755,
+      "learning_rate": 2.7018047271433163e-06,
+      "loss": 0.289,
+      "step": 9529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.758037379621471,
+      "learning_rate": 2.700804904856793e-06,
+      "loss": 0.2933,
+      "step": 9530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.8149425648813966,
+      "learning_rate": 2.6998051991445185e-06,
+      "loss": 0.3198,
+      "step": 9531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.0317023389438047,
+      "learning_rate": 2.698805610057182e-06,
+      "loss": 0.2988,
+      "step": 9532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 11.888684653600318,
+      "learning_rate": 2.6978061376454666e-06,
+      "loss": 0.331,
+      "step": 9533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9019291060755483,
+      "learning_rate": 2.6968067819600425e-06,
+      "loss": 0.3254,
+      "step": 9534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.962756966245736,
+      "learning_rate": 2.695807543051586e-06,
+      "loss": 0.3045,
+      "step": 9535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.953332116063025,
+      "learning_rate": 2.6948084209707566e-06,
+      "loss": 0.3161,
+      "step": 9536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.844744657380617,
+      "learning_rate": 2.693809415768212e-06,
+      "loss": 0.3151,
+      "step": 9537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.8776759888420225,
+      "learning_rate": 2.6928105274946046e-06,
+      "loss": 0.3461,
+      "step": 9538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.860171391385849,
+      "learning_rate": 2.6918117562005813e-06,
+      "loss": 0.3321,
+      "step": 9539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.596840199720954,
+      "learning_rate": 2.6908131019367804e-06,
+      "loss": 0.3353,
+      "step": 9540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.175259301873414,
+      "learning_rate": 2.6898145647538394e-06,
+      "loss": 0.3016,
+      "step": 9541
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.020173186591557,
+      "learning_rate": 2.6888161447023785e-06,
+      "loss": 0.3289,
+      "step": 9542
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.508717265914187,
+      "learning_rate": 2.68781784183303e-06,
+      "loss": 0.3149,
+      "step": 9543
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.327155586751364,
+      "learning_rate": 2.6868196561964024e-06,
+      "loss": 0.2731,
+      "step": 9544
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.3081484478995165,
+      "learning_rate": 2.685821587843108e-06,
+      "loss": 0.0997,
+      "step": 9545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.510293859807918,
+      "learning_rate": 2.6848236368237513e-06,
+      "loss": 0.3137,
+      "step": 9546
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 6.3470095734428575,
+      "learning_rate": 2.6838258031889313e-06,
+      "loss": 0.3313,
+      "step": 9547
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.4327064795005247,
+      "learning_rate": 2.682828086989242e-06,
+      "loss": 0.2618,
+      "step": 9548
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.7846764538890407,
+      "learning_rate": 2.681830488275263e-06,
+      "loss": 0.3063,
+      "step": 9549
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.3974437620336495,
+      "learning_rate": 2.680833007097584e-06,
+      "loss": 0.3138,
+      "step": 9550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 10.163529160699245,
+      "learning_rate": 2.679835643506773e-06,
+      "loss": 0.2944,
+      "step": 9551
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.811132687687412,
+      "learning_rate": 2.678838397553399e-06,
+      "loss": 0.3201,
+      "step": 9552
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9960940074208433,
+      "learning_rate": 2.6778412692880273e-06,
+      "loss": 0.3551,
+      "step": 9553
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.621397619714048,
+      "learning_rate": 2.676844258761213e-06,
+      "loss": 0.3121,
+      "step": 9554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.891925006912613,
+      "learning_rate": 2.6758473660235072e-06,
+      "loss": 0.2946,
+      "step": 9555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 4.969934032748195,
+      "learning_rate": 2.6748505911254558e-06,
+      "loss": 0.3133,
+      "step": 9556
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.9829268889270404,
+      "learning_rate": 2.673853934117594e-06,
+      "loss": 0.3103,
+      "step": 9557
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 8.485342670182165,
+      "learning_rate": 2.6728573950504565e-06,
+      "loss": 0.3293,
+      "step": 9558
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.312303223237459,
+      "learning_rate": 2.6718609739745694e-06,
+      "loss": 0.3297,
+      "step": 9559
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.29874489243358193,
+      "learning_rate": 2.6708646709404544e-06,
+      "loss": 0.1103,
+      "step": 9560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.085840175497757,
+      "learning_rate": 2.669868485998628e-06,
+      "loss": 0.3004,
+      "step": 9561
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.2971241489324356,
+      "learning_rate": 2.668872419199593e-06,
+      "loss": 0.1217,
+      "step": 9562
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.854620911501526,
+      "learning_rate": 2.6678764705938598e-06,
+      "loss": 0.3036,
+      "step": 9563
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.2339354600779,
+      "learning_rate": 2.666880640231917e-06,
+      "loss": 0.3017,
+      "step": 9564
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32547735363084174,
+      "learning_rate": 2.665884928164264e-06,
+      "loss": 0.1081,
+      "step": 9565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30855603236908036,
+      "learning_rate": 2.6648893344413797e-06,
+      "loss": 0.1167,
+      "step": 9566
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.463540854758993,
+      "learning_rate": 2.663893859113744e-06,
+      "loss": 0.3232,
+      "step": 9567
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.15780373479535,
+      "learning_rate": 2.662898502231831e-06,
+      "loss": 0.3289,
+      "step": 9568
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9799611915090765,
+      "learning_rate": 2.6619032638461063e-06,
+      "loss": 0.3088,
+      "step": 9569
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7810774229593314,
+      "learning_rate": 2.6609081440070335e-06,
+      "loss": 0.3499,
+      "step": 9570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.710093587615786,
+      "learning_rate": 2.6599131427650628e-06,
+      "loss": 0.325,
+      "step": 9571
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.2548822137414515,
+      "learning_rate": 2.6589182601706443e-06,
+      "loss": 0.3254,
+      "step": 9572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.378819388436158,
+      "learning_rate": 2.6579234962742228e-06,
+      "loss": 0.3503,
+      "step": 9573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9116854999963153,
+      "learning_rate": 2.6569288511262347e-06,
+      "loss": 0.285,
+      "step": 9574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.181445113690517,
+      "learning_rate": 2.6559343247771064e-06,
+      "loss": 0.3099,
+      "step": 9575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0071300020639957,
+      "learning_rate": 2.6549399172772704e-06,
+      "loss": 0.3298,
+      "step": 9576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.8524257756076645,
+      "learning_rate": 2.653945628677136e-06,
+      "loss": 0.3462,
+      "step": 9577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.539729065377166,
+      "learning_rate": 2.6529514590271254e-06,
+      "loss": 0.306,
+      "step": 9578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.3011589627569419,
+      "learning_rate": 2.6519574083776376e-06,
+      "loss": 0.1177,
+      "step": 9579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.53480482753977,
+      "learning_rate": 2.650963476779077e-06,
+      "loss": 0.3049,
+      "step": 9580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.180016641032587,
+      "learning_rate": 2.649969664281837e-06,
+      "loss": 0.2932,
+      "step": 9581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.222734634269396,
+      "learning_rate": 2.6489759709363062e-06,
+      "loss": 0.2903,
+      "step": 9582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.683264571331533,
+      "learning_rate": 2.64798239679287e-06,
+      "loss": 0.3032,
+      "step": 9583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.59253054152205,
+      "learning_rate": 2.6469889419018985e-06,
+      "loss": 0.2986,
+      "step": 9584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.125939303047692,
+      "learning_rate": 2.645995606313766e-06,
+      "loss": 0.3061,
+      "step": 9585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.413712700146547,
+      "learning_rate": 2.6450023900788355e-06,
+      "loss": 0.3423,
+      "step": 9586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7068079158615324,
+      "learning_rate": 2.644009293247467e-06,
+      "loss": 0.3219,
+      "step": 9587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.0854471313425336,
+      "learning_rate": 2.6430163158700116e-06,
+      "loss": 0.3271,
+      "step": 9588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.891028491496948,
+      "learning_rate": 2.6420234579968173e-06,
+      "loss": 0.325,
+      "step": 9589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.131303635755496,
+      "learning_rate": 2.6410307196782192e-06,
+      "loss": 0.2983,
+      "step": 9590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.195604415347174,
+      "learning_rate": 2.6400381009645578e-06,
+      "loss": 0.2912,
+      "step": 9591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.2753670805161894,
+      "learning_rate": 2.639045601906154e-06,
+      "loss": 0.3185,
+      "step": 9592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.053041753826495,
+      "learning_rate": 2.6380532225533367e-06,
+      "loss": 0.2896,
+      "step": 9593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.5741845991806,
+      "learning_rate": 2.637060962956417e-06,
+      "loss": 0.3065,
+      "step": 9594
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.541811011415241,
+      "learning_rate": 2.636068823165706e-06,
+      "loss": 0.3045,
+      "step": 9595
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.271866133700993,
+      "learning_rate": 2.63507680323151e-06,
+      "loss": 0.3466,
+      "step": 9596
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.993613880213312,
+      "learning_rate": 2.6340849032041204e-06,
+      "loss": 0.3234,
+      "step": 9597
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.317352803060798,
+      "learning_rate": 2.6330931231338354e-06,
+      "loss": 0.3354,
+      "step": 9598
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.6451441668851965,
+      "learning_rate": 2.632101463070936e-06,
+      "loss": 0.3112,
+      "step": 9599
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 23.191178935898797,
+      "learning_rate": 2.6311099230657033e-06,
+      "loss": 0.321,
+      "step": 9600
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8755144788573532,
+      "learning_rate": 2.6301185031684095e-06,
+      "loss": 0.3204,
+      "step": 9601
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.010657249994953,
+      "learning_rate": 2.629127203429323e-06,
+      "loss": 0.3118,
+      "step": 9602
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32146033473131114,
+      "learning_rate": 2.6281360238987043e-06,
+      "loss": 0.1016,
+      "step": 9603
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.574927684662231,
+      "learning_rate": 2.6271449646268106e-06,
+      "loss": 0.3588,
+      "step": 9604
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.983329214015736,
+      "learning_rate": 2.6261540256638844e-06,
+      "loss": 0.3196,
+      "step": 9605
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.147710727559559,
+      "learning_rate": 2.625163207060177e-06,
+      "loss": 0.3181,
+      "step": 9606
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.8861913773969072,
+      "learning_rate": 2.624172508865919e-06,
+      "loss": 0.3454,
+      "step": 9607
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4219997680951786,
+      "learning_rate": 2.623181931131342e-06,
+      "loss": 0.3258,
+      "step": 9608
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4875438046668146,
+      "learning_rate": 2.6221914739066746e-06,
+      "loss": 0.3233,
+      "step": 9609
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.154418694759657,
+      "learning_rate": 2.6212011372421263e-06,
+      "loss": 0.302,
+      "step": 9610
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.779890761365387,
+      "learning_rate": 2.6202109211879202e-06,
+      "loss": 0.3149,
+      "step": 9611
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1296646693555688,
+      "learning_rate": 2.6192208257942542e-06,
+      "loss": 0.2957,
+      "step": 9612
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.323596772963549,
+      "learning_rate": 2.618230851111332e-06,
+      "loss": 0.3,
+      "step": 9613
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.077122505938071,
+      "learning_rate": 2.6172409971893462e-06,
+      "loss": 0.2933,
+      "step": 9614
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.7893713037759977,
+      "learning_rate": 2.6162512640784853e-06,
+      "loss": 0.3485,
+      "step": 9615
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.304936827816909,
+      "learning_rate": 2.6152616518289307e-06,
+      "loss": 0.1151,
+      "step": 9616
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.579366591126614,
+      "learning_rate": 2.6142721604908606e-06,
+      "loss": 0.3062,
+      "step": 9617
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.359916575745985,
+      "learning_rate": 2.6132827901144376e-06,
+      "loss": 0.3264,
+      "step": 9618
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.326190436184673,
+      "learning_rate": 2.6122935407498328e-06,
+      "loss": 0.3336,
+      "step": 9619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.522633199287842,
+      "learning_rate": 2.6113044124471976e-06,
+      "loss": 0.3173,
+      "step": 9620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.9225527779288285,
+      "learning_rate": 2.610315405256685e-06,
+      "loss": 0.3046,
+      "step": 9621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 11.018318048495606,
+      "learning_rate": 2.6093265192284417e-06,
+      "loss": 0.3541,
+      "step": 9622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.894212497350154,
+      "learning_rate": 2.6083377544126008e-06,
+      "loss": 0.318,
+      "step": 9623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.27531873931987677,
+      "learning_rate": 2.607349110859303e-06,
+      "loss": 0.097,
+      "step": 9624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.85878978118103,
+      "learning_rate": 2.606360588618666e-06,
+      "loss": 0.2979,
+      "step": 9625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32185147250887036,
+      "learning_rate": 2.6053721877408176e-06,
+      "loss": 0.1168,
+      "step": 9626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.858109003765076,
+      "learning_rate": 2.6043839082758677e-06,
+      "loss": 0.2937,
+      "step": 9627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.8077710200571238,
+      "learning_rate": 2.603395750273925e-06,
+      "loss": 0.308,
+      "step": 9628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5463262649317784,
+      "learning_rate": 2.6024077137850913e-06,
+      "loss": 0.3272,
+      "step": 9629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.2902167566400394,
+      "learning_rate": 2.6014197988594625e-06,
+      "loss": 0.117,
+      "step": 9630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.679473315879005,
+      "learning_rate": 2.600432005547128e-06,
+      "loss": 0.317,
+      "step": 9631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.9595108558836833,
+      "learning_rate": 2.5994443338981732e-06,
+      "loss": 0.3226,
+      "step": 9632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.029070543575271,
+      "learning_rate": 2.5984567839626705e-06,
+      "loss": 0.3637,
+      "step": 9633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6316011022738515,
+      "learning_rate": 2.5974693557906943e-06,
+      "loss": 0.3236,
+      "step": 9634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.712923290789859,
+      "learning_rate": 2.5964820494323083e-06,
+      "loss": 0.332,
+      "step": 9635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6112234057608807,
+      "learning_rate": 2.595494864937571e-06,
+      "loss": 0.323,
+      "step": 9636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.042521922446506,
+      "learning_rate": 2.5945078023565374e-06,
+      "loss": 0.3299,
+      "step": 9637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.905801821449942,
+      "learning_rate": 2.5935208617392484e-06,
+      "loss": 0.3092,
+      "step": 9638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.760103709449825,
+      "learning_rate": 2.59253404313575e-06,
+      "loss": 0.2937,
+      "step": 9639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4366133927907163,
+      "learning_rate": 2.5915473465960695e-06,
+      "loss": 0.3154,
+      "step": 9640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.550236323782613,
+      "learning_rate": 2.590560772170242e-06,
+      "loss": 0.321,
+      "step": 9641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.296450740519927,
+      "learning_rate": 2.589574319908284e-06,
+      "loss": 0.2853,
+      "step": 9642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.169644706080187,
+      "learning_rate": 2.588587989860212e-06,
+      "loss": 0.314,
+      "step": 9643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.352055709198351,
+      "learning_rate": 2.587601782076035e-06,
+      "loss": 0.3231,
+      "step": 9644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.891687447272585,
+      "learning_rate": 2.5866156966057583e-06,
+      "loss": 0.3222,
+      "step": 9645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.938856143135854,
+      "learning_rate": 2.5856297334993742e-06,
+      "loss": 0.3181,
+      "step": 9646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.485383597142924,
+      "learning_rate": 2.584643892806876e-06,
+      "loss": 0.3738,
+      "step": 9647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.606565860690551,
+      "learning_rate": 2.5836581745782474e-06,
+      "loss": 0.3337,
+      "step": 9648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.434422577878519,
+      "learning_rate": 2.5826725788634655e-06,
+      "loss": 0.2981,
+      "step": 9649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.324536276684203,
+      "learning_rate": 2.581687105712506e-06,
+      "loss": 0.1106,
+      "step": 9650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.963529821965673,
+      "learning_rate": 2.5807017551753276e-06,
+      "loss": 0.3103,
+      "step": 9651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.926988814060645,
+      "learning_rate": 2.579716527301898e-06,
+      "loss": 0.3096,
+      "step": 9652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.759000613902355,
+      "learning_rate": 2.578731422142162e-06,
+      "loss": 0.3491,
+      "step": 9653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.244582498909404,
+      "learning_rate": 2.5777464397460763e-06,
+      "loss": 0.3066,
+      "step": 9654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.735613331864551,
+      "learning_rate": 2.5767615801635727e-06,
+      "loss": 0.3497,
+      "step": 9655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.507655007876336,
+      "learning_rate": 2.5757768434445897e-06,
+      "loss": 0.2893,
+      "step": 9656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.052626186173379,
+      "learning_rate": 2.5747922296390554e-06,
+      "loss": 0.3084,
+      "step": 9657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4752991861892295,
+      "learning_rate": 2.5738077387968917e-06,
+      "loss": 0.3042,
+      "step": 9658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.807993201361247,
+      "learning_rate": 2.572823370968016e-06,
+      "loss": 0.3039,
+      "step": 9659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 8.218164851100157,
+      "learning_rate": 2.571839126202335e-06,
+      "loss": 0.358,
+      "step": 9660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6374772020246433,
+      "learning_rate": 2.5708550045497537e-06,
+      "loss": 0.2889,
+      "step": 9661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 29.241387458480723,
+      "learning_rate": 2.5698710060601684e-06,
+      "loss": 0.311,
+      "step": 9662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31593251438226116,
+      "learning_rate": 2.5688871307834714e-06,
+      "loss": 0.1228,
+      "step": 9663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.55657756413134,
+      "learning_rate": 2.5679033787695457e-06,
+      "loss": 0.287,
+      "step": 9664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.334413079750512,
+      "learning_rate": 2.566919750068273e-06,
+      "loss": 0.3275,
+      "step": 9665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.660080519489243,
+      "learning_rate": 2.565936244729519e-06,
+      "loss": 0.3136,
+      "step": 9666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.986470039974341,
+      "learning_rate": 2.5649528628031574e-06,
+      "loss": 0.3205,
+      "step": 9667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.085637044768749,
+      "learning_rate": 2.5639696043390417e-06,
+      "loss": 0.3016,
+      "step": 9668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6637170026091432,
+      "learning_rate": 2.5629864693870278e-06,
+      "loss": 0.3332,
+      "step": 9669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.559555955063525,
+      "learning_rate": 2.562003457996963e-06,
+      "loss": 0.3098,
+      "step": 9670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5073792794738927,
+      "learning_rate": 2.5610205702186874e-06,
+      "loss": 0.3309,
+      "step": 9671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.004522018019234,
+      "learning_rate": 2.5600378061020376e-06,
+      "loss": 0.3265,
+      "step": 9672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 9.226184630660853,
+      "learning_rate": 2.5590551656968367e-06,
+      "loss": 0.3165,
+      "step": 9673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.49257287866142,
+      "learning_rate": 2.558072649052913e-06,
+      "loss": 0.2912,
+      "step": 9674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.351976564052761,
+      "learning_rate": 2.557090256220079e-06,
+      "loss": 0.2843,
+      "step": 9675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.7908078338616664,
+      "learning_rate": 2.5561079872481436e-06,
+      "loss": 0.3007,
+      "step": 9676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.272714418121907,
+      "learning_rate": 2.555125842186912e-06,
+      "loss": 0.3457,
+      "step": 9677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.191882879802433,
+      "learning_rate": 2.554143821086179e-06,
+      "loss": 0.3115,
+      "step": 9678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4884623007373357,
+      "learning_rate": 2.5531619239957372e-06,
+      "loss": 0.3045,
+      "step": 9679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.53753856320537,
+      "learning_rate": 2.552180150965372e-06,
+      "loss": 0.3251,
+      "step": 9680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6003174529277446,
+      "learning_rate": 2.5511985020448556e-06,
+      "loss": 0.3247,
+      "step": 9681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.793477477638662,
+      "learning_rate": 2.5502169772839676e-06,
+      "loss": 0.308,
+      "step": 9682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.487814401316804,
+      "learning_rate": 2.549235576732467e-06,
+      "loss": 0.2898,
+      "step": 9683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.942109563717581,
+      "learning_rate": 2.548254300440117e-06,
+      "loss": 0.322,
+      "step": 9684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6830663813145352,
+      "learning_rate": 2.5472731484566704e-06,
+      "loss": 0.3158,
+      "step": 9685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.046791866622984,
+      "learning_rate": 2.546292120831868e-06,
+      "loss": 0.328,
+      "step": 9686
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.987642666298201,
+      "learning_rate": 2.5453112176154587e-06,
+      "loss": 0.2991,
+      "step": 9687
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9944745781886333,
+      "learning_rate": 2.5443304388571703e-06,
+      "loss": 0.3029,
+      "step": 9688
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 10.021037476178424,
+      "learning_rate": 2.5433497846067323e-06,
+      "loss": 0.3104,
+      "step": 9689
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.4008174048873085,
+      "learning_rate": 2.542369254913867e-06,
+      "loss": 0.2873,
+      "step": 9690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.3065462133570227,
+      "learning_rate": 2.541388849828287e-06,
+      "loss": 0.2958,
+      "step": 9691
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 7.261978960659708,
+      "learning_rate": 2.5404085693997033e-06,
+      "loss": 0.3,
+      "step": 9692
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.976271678654847,
+      "learning_rate": 2.5394284136778193e-06,
+      "loss": 0.3364,
+      "step": 9693
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.866747532897228,
+      "learning_rate": 2.538448382712325e-06,
+      "loss": 0.3061,
+      "step": 9694
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.942525914459321,
+      "learning_rate": 2.537468476552919e-06,
+      "loss": 0.2792,
+      "step": 9695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.027166154606705,
+      "learning_rate": 2.5364886952492775e-06,
+      "loss": 0.3479,
+      "step": 9696
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3650077893724224,
+      "learning_rate": 2.5355090388510806e-06,
+      "loss": 0.3013,
+      "step": 9697
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.421292786690001,
+      "learning_rate": 2.534529507408e-06,
+      "loss": 0.3103,
+      "step": 9698
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.079157372310181,
+      "learning_rate": 2.533550100969695e-06,
+      "loss": 0.3209,
+      "step": 9699
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.30608441566676237,
+      "learning_rate": 2.532570819585831e-06,
+      "loss": 0.1147,
+      "step": 9700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.112325164439081,
+      "learning_rate": 2.531591663306052e-06,
+      "loss": 0.3469,
+      "step": 9701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.797793301821986,
+      "learning_rate": 2.530612632180012e-06,
+      "loss": 0.2917,
+      "step": 9702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.1524577953716975,
+      "learning_rate": 2.5296337262573435e-06,
+      "loss": 0.32,
+      "step": 9703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.939732553411809,
+      "learning_rate": 2.5286549455876803e-06,
+      "loss": 0.3085,
+      "step": 9704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 5.469434965062053,
+      "learning_rate": 2.5276762902206497e-06,
+      "loss": 0.3126,
+      "step": 9705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.000570782073836,
+      "learning_rate": 2.5266977602058724e-06,
+      "loss": 0.3024,
+      "step": 9706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1834849910020893,
+      "learning_rate": 2.5257193555929606e-06,
+      "loss": 0.3017,
+      "step": 9707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.716652263443567,
+      "learning_rate": 2.524741076431524e-06,
+      "loss": 0.3281,
+      "step": 9708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 6.19500355622441,
+      "learning_rate": 2.52376292277116e-06,
+      "loss": 0.2937,
+      "step": 9709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.072116125128024,
+      "learning_rate": 2.522784894661464e-06,
+      "loss": 0.2911,
+      "step": 9710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2086862239156893,
+      "learning_rate": 2.521806992152026e-06,
+      "loss": 0.296,
+      "step": 9711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.28087711070632,
+      "learning_rate": 2.520829215292426e-06,
+      "loss": 0.329,
+      "step": 9712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7252491723761683,
+      "learning_rate": 2.519851564132243e-06,
+      "loss": 0.3252,
+      "step": 9713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.115062992316499,
+      "learning_rate": 2.5188740387210387e-06,
+      "loss": 0.2922,
+      "step": 9714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 16.447789117494683,
+      "learning_rate": 2.517896639108385e-06,
+      "loss": 0.3203,
+      "step": 9715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.562121071646715,
+      "learning_rate": 2.516919365343831e-06,
+      "loss": 0.2693,
+      "step": 9716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8051638656819224,
+      "learning_rate": 2.5159422174769287e-06,
+      "loss": 0.3337,
+      "step": 9717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9781128655341744,
+      "learning_rate": 2.514965195557223e-06,
+      "loss": 0.3173,
+      "step": 9718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.082199679348381,
+      "learning_rate": 2.5139882996342502e-06,
+      "loss": 0.3044,
+      "step": 9719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.68309215887198,
+      "learning_rate": 2.5130115297575406e-06,
+      "loss": 0.3061,
+      "step": 9720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.541335774893,
+      "learning_rate": 2.512034885976621e-06,
+      "loss": 0.3382,
+      "step": 9721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6137244737226166,
+      "learning_rate": 2.5110583683410063e-06,
+      "loss": 0.3061,
+      "step": 9722
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.338580800049515,
+      "learning_rate": 2.510081976900208e-06,
+      "loss": 0.3095,
+      "step": 9723
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.6094913231087915,
+      "learning_rate": 2.509105711703734e-06,
+      "loss": 0.2842,
+      "step": 9724
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.469742809829257,
+      "learning_rate": 2.508129572801081e-06,
+      "loss": 0.3036,
+      "step": 9725
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.167336631267342,
+      "learning_rate": 2.5071535602417436e-06,
+      "loss": 0.3062,
+      "step": 9726
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7860531129280828,
+      "learning_rate": 2.5061776740752025e-06,
+      "loss": 0.3132,
+      "step": 9727
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.239955715880409,
+      "learning_rate": 2.5052019143509454e-06,
+      "loss": 0.2806,
+      "step": 9728
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.847468473654232,
+      "learning_rate": 2.5042262811184364e-06,
+      "loss": 0.3298,
+      "step": 9729
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.022757700759578,
+      "learning_rate": 2.5032507744271517e-06,
+      "loss": 0.2719,
+      "step": 9730
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.786162015994556,
+      "learning_rate": 2.5022753943265444e-06,
+      "loss": 0.3181,
+      "step": 9731
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2926276197750293,
+      "learning_rate": 2.501300140866071e-06,
+      "loss": 0.119,
+      "step": 9732
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.9835891780867314,
+      "learning_rate": 2.5003250140951785e-06,
+      "loss": 0.3471,
+      "step": 9733
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.9666730753158603,
+      "learning_rate": 2.499350014063308e-06,
+      "loss": 0.3067,
+      "step": 9734
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.961463241125282,
+      "learning_rate": 2.498375140819897e-06,
+      "loss": 0.3456,
+      "step": 9735
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.784348381857445,
+      "learning_rate": 2.4974003944143693e-06,
+      "loss": 0.3384,
+      "step": 9736
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.130370392205016,
+      "learning_rate": 2.4964257748961486e-06,
+      "loss": 0.3101,
+      "step": 9737
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.2952330826395285,
+      "learning_rate": 2.4954512823146505e-06,
+      "loss": 0.1105,
+      "step": 9738
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.006688712792072,
+      "learning_rate": 2.4944769167192827e-06,
+      "loss": 0.3083,
+      "step": 9739
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.314800879607847,
+      "learning_rate": 2.4935026781594494e-06,
+      "loss": 0.3051,
+      "step": 9740
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3126206408367318,
+      "learning_rate": 2.492528566684548e-06,
+      "loss": 0.1107,
+      "step": 9741
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.829869929322281,
+      "learning_rate": 2.4915545823439616e-06,
+      "loss": 0.3155,
+      "step": 9742
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 17.634541096457202,
+      "learning_rate": 2.490580725187082e-06,
+      "loss": 0.3252,
+      "step": 9743
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.187389064806702,
+      "learning_rate": 2.489606995263279e-06,
+      "loss": 0.3162,
+      "step": 9744
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.678815101659538,
+      "learning_rate": 2.4886333926219265e-06,
+      "loss": 0.3048,
+      "step": 9745
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.468564706050544,
+      "learning_rate": 2.4876599173123863e-06,
+      "loss": 0.3056,
+      "step": 9746
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.037848417818649,
+      "learning_rate": 2.486686569384017e-06,
+      "loss": 0.3385,
+      "step": 9747
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5354915946755794,
+      "learning_rate": 2.4857133488861713e-06,
+      "loss": 0.3023,
+      "step": 9748
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.127494285524199,
+      "learning_rate": 2.4847402558681877e-06,
+      "loss": 0.3324,
+      "step": 9749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8031173420515194,
+      "learning_rate": 2.4837672903794114e-06,
+      "loss": 0.334,
+      "step": 9750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.30329514909815225,
+      "learning_rate": 2.482794452469169e-06,
+      "loss": 0.1103,
+      "step": 9751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.933441146873369,
+      "learning_rate": 2.481821742186787e-06,
+      "loss": 0.3242,
+      "step": 9752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3094444593799127,
+      "learning_rate": 2.4808491595815836e-06,
+      "loss": 0.1187,
+      "step": 9753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.082085050191276,
+      "learning_rate": 2.4798767047028734e-06,
+      "loss": 0.2969,
+      "step": 9754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8510456245911033,
+      "learning_rate": 2.4789043775999553e-06,
+      "loss": 0.3239,
+      "step": 9755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.640165523739006,
+      "learning_rate": 2.4779321783221377e-06,
+      "loss": 0.293,
+      "step": 9756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.866851207480071,
+      "learning_rate": 2.476960106918706e-06,
+      "loss": 0.2909,
+      "step": 9757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0632819341651882,
+      "learning_rate": 2.4759881634389495e-06,
+      "loss": 0.3463,
+      "step": 9758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.600375330451103,
+      "learning_rate": 2.4750163479321463e-06,
+      "loss": 0.2705,
+      "step": 9759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5905875420145295,
+      "learning_rate": 2.4740446604475715e-06,
+      "loss": 0.3638,
+      "step": 9760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.9582276963786445,
+      "learning_rate": 2.4730731010344923e-06,
+      "loss": 0.3498,
+      "step": 9761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.60591046691144,
+      "learning_rate": 2.472101669742163e-06,
+      "loss": 0.3359,
+      "step": 9762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.883365015353513,
+      "learning_rate": 2.471130366619846e-06,
+      "loss": 0.3265,
+      "step": 9763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.532920434620883,
+      "learning_rate": 2.470159191716782e-06,
+      "loss": 0.3273,
+      "step": 9764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 11.540993917060424,
+      "learning_rate": 2.469188145082214e-06,
+      "loss": 0.3465,
+      "step": 9765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.308110681592268,
+      "learning_rate": 2.4682172267653764e-06,
+      "loss": 0.3183,
+      "step": 9766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.99513067139171,
+      "learning_rate": 2.467246436815497e-06,
+      "loss": 0.3157,
+      "step": 9767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.191631851232714,
+      "learning_rate": 2.4662757752817955e-06,
+      "loss": 0.3216,
+      "step": 9768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.56779077890773,
+      "learning_rate": 2.4653052422134905e-06,
+      "loss": 0.3307,
+      "step": 9769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.309806881131754,
+      "learning_rate": 2.464334837659783e-06,
+      "loss": 0.3182,
+      "step": 9770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28424168097772956,
+      "learning_rate": 2.4633645616698833e-06,
+      "loss": 0.1023,
+      "step": 9771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.793787193694763,
+      "learning_rate": 2.4623944142929796e-06,
+      "loss": 0.3067,
+      "step": 9772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.436081669471871,
+      "learning_rate": 2.4614243955782634e-06,
+      "loss": 0.3173,
+      "step": 9773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.547543433270583,
+      "learning_rate": 2.4604545055749186e-06,
+      "loss": 0.3272,
+      "step": 9774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.521834387463096,
+      "learning_rate": 2.4594847443321145e-06,
+      "loss": 0.2988,
+      "step": 9775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 14.50972324552115,
+      "learning_rate": 2.4585151118990286e-06,
+      "loss": 0.3207,
+      "step": 9776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7929080653860616,
+      "learning_rate": 2.4575456083248143e-06,
+      "loss": 0.3212,
+      "step": 9777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.415467990070221,
+      "learning_rate": 2.4565762336586376e-06,
+      "loss": 0.3006,
+      "step": 9778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.27092479158795,
+      "learning_rate": 2.45560698794964e-06,
+      "loss": 0.3094,
+      "step": 9779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.004477674918878,
+      "learning_rate": 2.4546378712469665e-06,
+      "loss": 0.2841,
+      "step": 9780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 24.87535894161355,
+      "learning_rate": 2.4536688835997546e-06,
+      "loss": 0.3132,
+      "step": 9781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.29991999163974,
+      "learning_rate": 2.452700025057133e-06,
+      "loss": 0.3184,
+      "step": 9782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.829022431089421,
+      "learning_rate": 2.4517312956682267e-06,
+      "loss": 0.2944,
+      "step": 9783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.766309113109215,
+      "learning_rate": 2.4507626954821517e-06,
+      "loss": 0.3619,
+      "step": 9784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.856209546801325,
+      "learning_rate": 2.449794224548017e-06,
+      "loss": 0.3335,
+      "step": 9785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.616532942162429,
+      "learning_rate": 2.448825882914926e-06,
+      "loss": 0.3263,
+      "step": 9786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.182087329042603,
+      "learning_rate": 2.4478576706319767e-06,
+      "loss": 0.2855,
+      "step": 9787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.023923815534025,
+      "learning_rate": 2.4468895877482594e-06,
+      "loss": 0.2851,
+      "step": 9788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8808022986753334,
+      "learning_rate": 2.4459216343128606e-06,
+      "loss": 0.3084,
+      "step": 9789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.28401416017350395,
+      "learning_rate": 2.444953810374851e-06,
+      "loss": 0.1169,
+      "step": 9790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.665009418308671,
+      "learning_rate": 2.4439861159833095e-06,
+      "loss": 0.3071,
+      "step": 9791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.274090930650879,
+      "learning_rate": 2.4430185511872944e-06,
+      "loss": 0.3159,
+      "step": 9792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 7.092106710239432,
+      "learning_rate": 2.442051116035865e-06,
+      "loss": 0.2777,
+      "step": 9793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4506272643556652,
+      "learning_rate": 2.441083810578074e-06,
+      "loss": 0.3593,
+      "step": 9794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.064299075311795,
+      "learning_rate": 2.4401166348629637e-06,
+      "loss": 0.3329,
+      "step": 9795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.303235119285244,
+      "learning_rate": 2.4391495889395734e-06,
+      "loss": 0.2802,
+      "step": 9796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.124323841629497,
+      "learning_rate": 2.4381826728569357e-06,
+      "loss": 0.3152,
+      "step": 9797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 37.95831838031016,
+      "learning_rate": 2.437215886664072e-06,
+      "loss": 0.3563,
+      "step": 9798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.628181176290804,
+      "learning_rate": 2.4362492304100026e-06,
+      "loss": 0.3213,
+      "step": 9799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.2745150633948885,
+      "learning_rate": 2.4352827041437387e-06,
+      "loss": 0.3324,
+      "step": 9800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.841028678438828,
+      "learning_rate": 2.434316307914285e-06,
+      "loss": 0.309,
+      "step": 9801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.612869247930414,
+      "learning_rate": 2.433350041770643e-06,
+      "loss": 0.3104,
+      "step": 9802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.863785604268871,
+      "learning_rate": 2.432383905761797e-06,
+      "loss": 0.3236,
+      "step": 9803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.852696365910958,
+      "learning_rate": 2.4314178999367417e-06,
+      "loss": 0.2967,
+      "step": 9804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.481318025129254,
+      "learning_rate": 2.4304520243444467e-06,
+      "loss": 0.3075,
+      "step": 9805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.204648614240278,
+      "learning_rate": 2.429486279033892e-06,
+      "loss": 0.2873,
+      "step": 9806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.481056934290699,
+      "learning_rate": 2.4285206640540383e-06,
+      "loss": 0.316,
+      "step": 9807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6785551976631408,
+      "learning_rate": 2.4275551794538445e-06,
+      "loss": 0.3377,
+      "step": 9808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.312022987060448,
+      "learning_rate": 2.4265898252822644e-06,
+      "loss": 0.31,
+      "step": 9809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.992299501807163,
+      "learning_rate": 2.4256246015882427e-06,
+      "loss": 0.3199,
+      "step": 9810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.270846539900942,
+      "learning_rate": 2.4246595084207208e-06,
+      "loss": 0.3121,
+      "step": 9811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.917701774611199,
+      "learning_rate": 2.4236945458286267e-06,
+      "loss": 0.31,
+      "step": 9812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 13.880383080563158,
+      "learning_rate": 2.4227297138608885e-06,
+      "loss": 0.3445,
+      "step": 9813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.812810377178918,
+      "learning_rate": 2.421765012566425e-06,
+      "loss": 0.3137,
+      "step": 9814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7347204799911866,
+      "learning_rate": 2.420800441994149e-06,
+      "loss": 0.2994,
+      "step": 9815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8691386219259045,
+      "learning_rate": 2.4198360021929656e-06,
+      "loss": 0.3065,
+      "step": 9816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.291216741842667,
+      "learning_rate": 2.418871693211777e-06,
+      "loss": 0.3233,
+      "step": 9817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.313129890397622,
+      "learning_rate": 2.4179075150994692e-06,
+      "loss": 0.2986,
+      "step": 9818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.170847510880369,
+      "learning_rate": 2.416943467904936e-06,
+      "loss": 0.3385,
+      "step": 9819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0182247315116855,
+      "learning_rate": 2.415979551677051e-06,
+      "loss": 0.3053,
+      "step": 9820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.281249415581183,
+      "learning_rate": 2.4150157664646895e-06,
+      "loss": 0.266,
+      "step": 9821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.112227681477838,
+      "learning_rate": 2.4140521123167165e-06,
+      "loss": 0.3452,
+      "step": 9822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.230990220735155,
+      "learning_rate": 2.4130885892819923e-06,
+      "loss": 0.2971,
+      "step": 9823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.54543349745705,
+      "learning_rate": 2.4121251974093706e-06,
+      "loss": 0.2937,
+      "step": 9824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.250681224914005,
+      "learning_rate": 2.411161936747694e-06,
+      "loss": 0.3016,
+      "step": 9825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.21507507419552,
+      "learning_rate": 2.4101988073458044e-06,
+      "loss": 0.3075,
+      "step": 9826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.931834009647606,
+      "learning_rate": 2.4092358092525335e-06,
+      "loss": 0.3186,
+      "step": 9827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.342720649055565,
+      "learning_rate": 2.4082729425167085e-06,
+      "loss": 0.2835,
+      "step": 9828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.248738450056744,
+      "learning_rate": 2.4073102071871486e-06,
+      "loss": 0.3402,
+      "step": 9829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.12286430262011,
+      "learning_rate": 2.4063476033126675e-06,
+      "loss": 0.332,
+      "step": 9830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.554894616592643,
+      "learning_rate": 2.4053851309420674e-06,
+      "loss": 0.2965,
+      "step": 9831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.8465655211955654,
+      "learning_rate": 2.404422790124154e-06,
+      "loss": 0.319,
+      "step": 9832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 10.077020729410211,
+      "learning_rate": 2.403460580907715e-06,
+      "loss": 0.3223,
+      "step": 9833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.8071050608942265,
+      "learning_rate": 2.4024985033415377e-06,
+      "loss": 0.2966,
+      "step": 9834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.5976505015163194,
+      "learning_rate": 2.4015365574744026e-06,
+      "loss": 0.2827,
+      "step": 9835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.611188879821226,
+      "learning_rate": 2.400574743355082e-06,
+      "loss": 0.304,
+      "step": 9836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.048481494690369,
+      "learning_rate": 2.3996130610323437e-06,
+      "loss": 0.3628,
+      "step": 9837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.412806467477738,
+      "learning_rate": 2.3986515105549418e-06,
+      "loss": 0.3427,
+      "step": 9838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1771182747596347,
+      "learning_rate": 2.3976900919716362e-06,
+      "loss": 0.2801,
+      "step": 9839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.134902906195198,
+      "learning_rate": 2.396728805331167e-06,
+      "loss": 0.343,
+      "step": 9840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4656605899755792,
+      "learning_rate": 2.3957676506822763e-06,
+      "loss": 0.3257,
+      "step": 9841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.913458790698541,
+      "learning_rate": 2.3948066280736964e-06,
+      "loss": 0.3535,
+      "step": 9842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.3103497541878575,
+      "learning_rate": 2.3938457375541536e-06,
+      "loss": 0.3167,
+      "step": 9843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.286336545836424,
+      "learning_rate": 2.3928849791723664e-06,
+      "loss": 0.3596,
+      "step": 9844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.056355508441052,
+      "learning_rate": 2.3919243529770493e-06,
+      "loss": 0.2781,
+      "step": 9845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.827659321398494,
+      "learning_rate": 2.390963859016905e-06,
+      "loss": 0.351,
+      "step": 9846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0111981732952535,
+      "learning_rate": 2.3900034973406337e-06,
+      "loss": 0.3231,
+      "step": 9847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 6.623364140767176,
+      "learning_rate": 2.3890432679969292e-06,
+      "loss": 0.2624,
+      "step": 9848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 8.544258814884227,
+      "learning_rate": 2.388083171034476e-06,
+      "loss": 0.328,
+      "step": 9849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31342350931857954,
+      "learning_rate": 2.3871232065019556e-06,
+      "loss": 0.0975,
+      "step": 9850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.9821371135194696,
+      "learning_rate": 2.3861633744480343e-06,
+      "loss": 0.3372,
+      "step": 9851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.617607918783898,
+      "learning_rate": 2.385203674921386e-06,
+      "loss": 0.3618,
+      "step": 9852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 5.0841373912589924,
+      "learning_rate": 2.3842441079706608e-06,
+      "loss": 0.2877,
+      "step": 9853
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.947454683681673,
+      "learning_rate": 2.38328467364452e-06,
+      "loss": 0.3092,
+      "step": 9854
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 17.92375937358608,
+      "learning_rate": 2.382325371991601e-06,
+      "loss": 0.2804,
+      "step": 9855
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30043997026996816,
+      "learning_rate": 2.3813662030605473e-06,
+      "loss": 0.1115,
+      "step": 9856
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3318941540144604,
+      "learning_rate": 2.3804071668999882e-06,
+      "loss": 0.3573,
+      "step": 9857
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6895642145808303,
+      "learning_rate": 2.3794482635585508e-06,
+      "loss": 0.3067,
+      "step": 9858
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4652771863638656,
+      "learning_rate": 2.3784894930848533e-06,
+      "loss": 0.3033,
+      "step": 9859
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.588362157143752,
+      "learning_rate": 2.3775308555275083e-06,
+      "loss": 0.2962,
+      "step": 9860
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.718708063486481,
+      "learning_rate": 2.3765723509351186e-06,
+      "loss": 0.3278,
+      "step": 9861
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.479998487657687,
+      "learning_rate": 2.3756139793562834e-06,
+      "loss": 0.3556,
+      "step": 9862
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.899431354201198,
+      "learning_rate": 2.3746557408395954e-06,
+      "loss": 0.2938,
+      "step": 9863
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.186294934574065,
+      "learning_rate": 2.3736976354336354e-06,
+      "loss": 0.3243,
+      "step": 9864
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.022938845897626,
+      "learning_rate": 2.372739663186988e-06,
+      "loss": 0.3105,
+      "step": 9865
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.159315619553366,
+      "learning_rate": 2.3717818241482177e-06,
+      "loss": 0.345,
+      "step": 9866
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8113903874313246,
+      "learning_rate": 2.370824118365896e-06,
+      "loss": 0.3526,
+      "step": 9867
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.6518737828472863,
+      "learning_rate": 2.3698665458885745e-06,
+      "loss": 0.2874,
+      "step": 9868
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.915632116210502,
+      "learning_rate": 2.3689091067648073e-06,
+      "loss": 0.3527,
+      "step": 9869
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.790913774200584,
+      "learning_rate": 2.367951801043138e-06,
+      "loss": 0.3074,
+      "step": 9870
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.28951770413401,
+      "learning_rate": 2.366994628772104e-06,
+      "loss": 0.1184,
+      "step": 9871
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.229179376712513,
+      "learning_rate": 2.3660375900002364e-06,
+      "loss": 0.3479,
+      "step": 9872
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6731456029642726,
+      "learning_rate": 2.365080684776061e-06,
+      "loss": 0.3183,
+      "step": 9873
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.4062772476236187,
+      "learning_rate": 2.3641239131480908e-06,
+      "loss": 0.3136,
+      "step": 9874
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.477743148662733,
+      "learning_rate": 2.3631672751648386e-06,
+      "loss": 0.3061,
+      "step": 9875
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3538808679042034,
+      "learning_rate": 2.3622107708748083e-06,
+      "loss": 0.3025,
+      "step": 9876
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.7355203222172717,
+      "learning_rate": 2.3612544003264964e-06,
+      "loss": 0.346,
+      "step": 9877
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8808047632056515,
+      "learning_rate": 2.3602981635683946e-06,
+      "loss": 0.2755,
+      "step": 9878
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7566397403176013,
+      "learning_rate": 2.3593420606489807e-06,
+      "loss": 0.3194,
+      "step": 9879
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.350778544069897,
+      "learning_rate": 2.358386091616739e-06,
+      "loss": 0.2963,
+      "step": 9880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29026195583924164,
+      "learning_rate": 2.3574302565201317e-06,
+      "loss": 0.1058,
+      "step": 9881
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29953818087567813,
+      "learning_rate": 2.356474555407629e-06,
+      "loss": 0.1142,
+      "step": 9882
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6217574467525,
+      "learning_rate": 2.3555189883276815e-06,
+      "loss": 0.2997,
+      "step": 9883
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2954289773382943,
+      "learning_rate": 2.3545635553287406e-06,
+      "loss": 0.1034,
+      "step": 9884
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.415659907086615,
+      "learning_rate": 2.3536082564592487e-06,
+      "loss": 0.3111,
+      "step": 9885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 14.711799853517773,
+      "learning_rate": 2.3526530917676415e-06,
+      "loss": 0.3104,
+      "step": 9886
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.169915149756204,
+      "learning_rate": 2.35169806130235e-06,
+      "loss": 0.3165,
+      "step": 9887
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9679517024535396,
+      "learning_rate": 2.350743165111793e-06,
+      "loss": 0.3099,
+      "step": 9888
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.546259810331012,
+      "learning_rate": 2.3497884032443863e-06,
+      "loss": 0.3093,
+      "step": 9889
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.379478849380449,
+      "learning_rate": 2.34883377574854e-06,
+      "loss": 0.3155,
+      "step": 9890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7562748643455866,
+      "learning_rate": 2.3478792826726547e-06,
+      "loss": 0.3221,
+      "step": 9891
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5576880239498854,
+      "learning_rate": 2.3469249240651264e-06,
+      "loss": 0.2838,
+      "step": 9892
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.049609896041736,
+      "learning_rate": 2.3459706999743443e-06,
+      "loss": 0.3045,
+      "step": 9893
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.464121104782959,
+      "learning_rate": 2.345016610448684e-06,
+      "loss": 0.3263,
+      "step": 9894
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 11.136506548080531,
+      "learning_rate": 2.3440626555365275e-06,
+      "loss": 0.2873,
+      "step": 9895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6166853524646005,
+      "learning_rate": 2.343108835286237e-06,
+      "loss": 0.3027,
+      "step": 9896
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.366674835309581,
+      "learning_rate": 2.3421551497461747e-06,
+      "loss": 0.3044,
+      "step": 9897
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5481640978747224,
+      "learning_rate": 2.3412015989646954e-06,
+      "loss": 0.3117,
+      "step": 9898
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0455553819153254,
+      "learning_rate": 2.3402481829901457e-06,
+      "loss": 0.3235,
+      "step": 9899
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 9.066218546955644,
+      "learning_rate": 2.339294901870868e-06,
+      "loss": 0.3015,
+      "step": 9900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.469791135962637,
+      "learning_rate": 2.3383417556551907e-06,
+      "loss": 0.3386,
+      "step": 9901
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8360866210746516,
+      "learning_rate": 2.3373887443914445e-06,
+      "loss": 0.3011,
+      "step": 9902
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.471612889644321,
+      "learning_rate": 2.3364358681279476e-06,
+      "loss": 0.324,
+      "step": 9903
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 43.8085199356114,
+      "learning_rate": 2.3354831269130133e-06,
+      "loss": 0.3054,
+      "step": 9904
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.516631768236627,
+      "learning_rate": 2.334530520794948e-06,
+      "loss": 0.311,
+      "step": 9905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.293376826948801,
+      "learning_rate": 2.333578049822052e-06,
+      "loss": 0.104,
+      "step": 9906
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.502315572962333,
+      "learning_rate": 2.3326257140426136e-06,
+      "loss": 0.3473,
+      "step": 9907
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.628943510391688,
+      "learning_rate": 2.331673513504924e-06,
+      "loss": 0.3452,
+      "step": 9908
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.30063104313275923,
+      "learning_rate": 2.3307214482572577e-06,
+      "loss": 0.1193,
+      "step": 9909
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.265870836359712,
+      "learning_rate": 2.3297695183478873e-06,
+      "loss": 0.3163,
+      "step": 9910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.396728123133855,
+      "learning_rate": 2.328817723825078e-06,
+      "loss": 0.2936,
+      "step": 9911
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 42.90308245446748,
+      "learning_rate": 2.3278660647370884e-06,
+      "loss": 0.3341,
+      "step": 9912
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9828127129212114,
+      "learning_rate": 2.3269145411321713e-06,
+      "loss": 0.3156,
+      "step": 9913
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.136016341950253,
+      "learning_rate": 2.325963153058565e-06,
+      "loss": 0.3365,
+      "step": 9914
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.885376172330094,
+      "learning_rate": 2.325011900564515e-06,
+      "loss": 0.284,
+      "step": 9915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7780428734145244,
+      "learning_rate": 2.324060783698246e-06,
+      "loss": 0.3546,
+      "step": 9916
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.468044912465325,
+      "learning_rate": 2.323109802507984e-06,
+      "loss": 0.3599,
+      "step": 9917
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.671188417401176,
+      "learning_rate": 2.322158957041945e-06,
+      "loss": 0.2876,
+      "step": 9918
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.708598411158746,
+      "learning_rate": 2.32120824734834e-06,
+      "loss": 0.3113,
+      "step": 9919
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0790040145013524,
+      "learning_rate": 2.320257673475371e-06,
+      "loss": 0.2963,
+      "step": 9920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9158191387615497,
+      "learning_rate": 2.3193072354712375e-06,
+      "loss": 0.2773,
+      "step": 9921
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.720134488360226,
+      "learning_rate": 2.318356933384123e-06,
+      "loss": 0.3121,
+      "step": 9922
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.29544769641510976,
+      "learning_rate": 2.317406767262214e-06,
+      "loss": 0.1079,
+      "step": 9923
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.369516732625531,
+      "learning_rate": 2.3164567371536847e-06,
+      "loss": 0.2985,
+      "step": 9924
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.433471707315444,
+      "learning_rate": 2.3155068431067033e-06,
+      "loss": 0.3126,
+      "step": 9925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.7011850027178035,
+      "learning_rate": 2.314557085169435e-06,
+      "loss": 0.3297,
+      "step": 9926
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7946191976560444,
+      "learning_rate": 2.3136074633900276e-06,
+      "loss": 0.281,
+      "step": 9927
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.217531102727077,
+      "learning_rate": 2.312657977816637e-06,
+      "loss": 0.3329,
+      "step": 9928
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.0742621793773255,
+      "learning_rate": 2.3117086284973966e-06,
+      "loss": 0.3398,
+      "step": 9929
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.729318522871333,
+      "learning_rate": 2.310759415480448e-06,
+      "loss": 0.3294,
+      "step": 9930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3657986661256807,
+      "learning_rate": 2.309810338813913e-06,
+      "loss": 0.347,
+      "step": 9931
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 12.971722814480835,
+      "learning_rate": 2.308861398545914e-06,
+      "loss": 0.3241,
+      "step": 9932
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.580207863484804,
+      "learning_rate": 2.3079125947245635e-06,
+      "loss": 0.3399,
+      "step": 9933
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.153039035457806,
+      "learning_rate": 2.3069639273979706e-06,
+      "loss": 0.3279,
+      "step": 9934
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2944324854842546,
+      "learning_rate": 2.30601539661423e-06,
+      "loss": 0.3147,
+      "step": 9935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.044585946784609,
+      "learning_rate": 2.305067002421438e-06,
+      "loss": 0.3183,
+      "step": 9936
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6756582023812214,
+      "learning_rate": 2.3041187448676784e-06,
+      "loss": 0.3085,
+      "step": 9937
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.340560382541923,
+      "learning_rate": 2.303170624001031e-06,
+      "loss": 0.3534,
+      "step": 9938
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.942902836716157,
+      "learning_rate": 2.302222639869569e-06,
+      "loss": 0.3763,
+      "step": 9939
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.6275560236672795,
+      "learning_rate": 2.301274792521352e-06,
+      "loss": 0.3312,
+      "step": 9940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8513118670360993,
+      "learning_rate": 2.3003270820044452e-06,
+      "loss": 0.303,
+      "step": 9941
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.568071516506583,
+      "learning_rate": 2.299379508366893e-06,
+      "loss": 0.3286,
+      "step": 9942
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.448965947422247,
+      "learning_rate": 2.2984320716567456e-06,
+      "loss": 0.3337,
+      "step": 9943
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3948460522701494,
+      "learning_rate": 2.297484771922035e-06,
+      "loss": 0.2913,
+      "step": 9944
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.801357206285919,
+      "learning_rate": 2.2965376092107934e-06,
+      "loss": 0.331,
+      "step": 9945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 16.716701480056326,
+      "learning_rate": 2.295590583571044e-06,
+      "loss": 0.2942,
+      "step": 9946
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.608935538506227,
+      "learning_rate": 2.294643695050803e-06,
+      "loss": 0.3422,
+      "step": 9947
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.892958868930794,
+      "learning_rate": 2.2936969436980807e-06,
+      "loss": 0.3123,
+      "step": 9948
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.792113119940163,
+      "learning_rate": 2.2927503295608794e-06,
+      "loss": 0.3147,
+      "step": 9949
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.93076960977601,
+      "learning_rate": 2.291803852687192e-06,
+      "loss": 0.3594,
+      "step": 9950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.342211005194602,
+      "learning_rate": 2.290857513125009e-06,
+      "loss": 0.2976,
+      "step": 9951
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.87902719401291,
+      "learning_rate": 2.2899113109223113e-06,
+      "loss": 0.3458,
+      "step": 9952
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.850791009539893,
+      "learning_rate": 2.288965246127074e-06,
+      "loss": 0.3137,
+      "step": 9953
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.9553007646881357,
+      "learning_rate": 2.288019318787266e-06,
+      "loss": 0.3208,
+      "step": 9954
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.104418975923112,
+      "learning_rate": 2.287073528950842e-06,
+      "loss": 0.3074,
+      "step": 9955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.8619296701957437,
+      "learning_rate": 2.2861278766657644e-06,
+      "loss": 0.3137,
+      "step": 9956
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.670565122336746,
+      "learning_rate": 2.285182361979972e-06,
+      "loss": 0.3103,
+      "step": 9957
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8853766561333978,
+      "learning_rate": 2.284236984941411e-06,
+      "loss": 0.3211,
+      "step": 9958
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.0426738131960356,
+      "learning_rate": 2.2832917455980085e-06,
+      "loss": 0.2854,
+      "step": 9959
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.598993092833161,
+      "learning_rate": 2.282346643997693e-06,
+      "loss": 0.3259,
+      "step": 9960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.9355359975520665,
+      "learning_rate": 2.281401680188383e-06,
+      "loss": 0.3043,
+      "step": 9961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.633250884919315,
+      "learning_rate": 2.2804568542179907e-06,
+      "loss": 0.3394,
+      "step": 9962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.114407362624485,
+      "learning_rate": 2.279512166134421e-06,
+      "loss": 0.3288,
+      "step": 9963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.982142715448967,
+      "learning_rate": 2.2785676159855695e-06,
+      "loss": 0.3078,
+      "step": 9964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.777605586649782,
+      "learning_rate": 2.2776232038193286e-06,
+      "loss": 0.3253,
+      "step": 9965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.7033966688324,
+      "learning_rate": 2.2766789296835817e-06,
+      "loss": 0.2986,
+      "step": 9966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2930912015589303,
+      "learning_rate": 2.275734793626206e-06,
+      "loss": 0.1098,
+      "step": 9967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.3776157260277926,
+      "learning_rate": 2.274790795695071e-06,
+      "loss": 0.2932,
+      "step": 9968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.990023732312215,
+      "learning_rate": 2.273846935938041e-06,
+      "loss": 0.28,
+      "step": 9969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.602814184603692,
+      "learning_rate": 2.2729032144029665e-06,
+      "loss": 0.2966,
+      "step": 9970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.932388583440395,
+      "learning_rate": 2.271959631137704e-06,
+      "loss": 0.3271,
+      "step": 9971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3068846392198168,
+      "learning_rate": 2.27101618619009e-06,
+      "loss": 0.1092,
+      "step": 9972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.1983056829905054,
+      "learning_rate": 2.2700728796079597e-06,
+      "loss": 0.3047,
+      "step": 9973
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.970499294531322,
+      "learning_rate": 2.2691297114391416e-06,
+      "loss": 0.337,
+      "step": 9974
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9533597711741724,
+      "learning_rate": 2.268186681731456e-06,
+      "loss": 0.3062,
+      "step": 9975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.39050106988412,
+      "learning_rate": 2.26724379053272e-06,
+      "loss": 0.3398,
+      "step": 9976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.671022539910009,
+      "learning_rate": 2.266301037890735e-06,
+      "loss": 0.2969,
+      "step": 9977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.588272229988401,
+      "learning_rate": 2.2653584238533027e-06,
+      "loss": 0.3261,
+      "step": 9978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.086392240953325,
+      "learning_rate": 2.264415948468216e-06,
+      "loss": 0.3108,
+      "step": 9979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.7251420645125055,
+      "learning_rate": 2.26347361178326e-06,
+      "loss": 0.3267,
+      "step": 9980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.381449403161727,
+      "learning_rate": 2.2625314138462146e-06,
+      "loss": 0.2785,
+      "step": 9981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.117077098761206,
+      "learning_rate": 2.261589354704851e-06,
+      "loss": 0.3124,
+      "step": 9982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.323311867888316,
+      "learning_rate": 2.26064743440693e-06,
+      "loss": 0.3352,
+      "step": 9983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.424666416521347,
+      "learning_rate": 2.259705653000216e-06,
+      "loss": 0.3581,
+      "step": 9984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.684405257783487,
+      "learning_rate": 2.2587640105324534e-06,
+      "loss": 0.3243,
+      "step": 9985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5101092593224537,
+      "learning_rate": 2.257822507051387e-06,
+      "loss": 0.2847,
+      "step": 9986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.2107723254855514,
+      "learning_rate": 2.256881142604754e-06,
+      "loss": 0.3193,
+      "step": 9987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.098360157801278,
+      "learning_rate": 2.2559399172402835e-06,
+      "loss": 0.3457,
+      "step": 9988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.302509714682465,
+      "learning_rate": 2.2549988310056994e-06,
+      "loss": 0.1117,
+      "step": 9989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.056277349373591,
+      "learning_rate": 2.2540578839487103e-06,
+      "loss": 0.2829,
+      "step": 9990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.7974930477507565,
+      "learning_rate": 2.253117076117034e-06,
+      "loss": 0.3144,
+      "step": 9991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 6.728506268162217,
+      "learning_rate": 2.2521764075583636e-06,
+      "loss": 0.3341,
+      "step": 9992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 4.4926550737369375,
+      "learning_rate": 2.251235878320396e-06,
+      "loss": 0.3337,
+      "step": 9993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.8281289311362934,
+      "learning_rate": 2.250295488450819e-06,
+      "loss": 0.3109,
+      "step": 9994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5235587946593814,
+      "learning_rate": 2.249355237997311e-06,
+      "loss": 0.3356,
+      "step": 9995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9093500509830705,
+      "learning_rate": 2.248415127007545e-06,
+      "loss": 0.2831,
+      "step": 9996
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4428508363136867,
+      "learning_rate": 2.247475155529189e-06,
+      "loss": 0.3278,
+      "step": 9997
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.533615004735008,
+      "learning_rate": 2.2465353236098975e-06,
+      "loss": 0.3251,
+      "step": 9998
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2924273199086733,
+      "learning_rate": 2.2455956312973243e-06,
+      "loss": 0.2856,
+      "step": 9999
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1771016539478936,
+      "learning_rate": 2.2446560786391135e-06,
+      "loss": 0.287,
+      "step": 10000
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.775661473261746,
+      "learning_rate": 2.243716665682903e-06,
+      "loss": 0.3238,
+      "step": 10001
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.234474938639163,
+      "learning_rate": 2.242777392476324e-06,
+      "loss": 0.2994,
+      "step": 10002
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.730654197775563,
+      "learning_rate": 2.241838259066995e-06,
+      "loss": 0.3163,
+      "step": 10003
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.340487687913973,
+      "learning_rate": 2.2408992655025398e-06,
+      "loss": 0.2817,
+      "step": 10004
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.5814702096344675,
+      "learning_rate": 2.239960411830559e-06,
+      "loss": 0.2959,
+      "step": 10005
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.04971395319675,
+      "learning_rate": 2.239021698098663e-06,
+      "loss": 0.3204,
+      "step": 10006
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.778568662228681,
+      "learning_rate": 2.2380831243544405e-06,
+      "loss": 0.3231,
+      "step": 10007
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.23265162049591,
+      "learning_rate": 2.237144690645481e-06,
+      "loss": 0.3094,
+      "step": 10008
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.99197525591693,
+      "learning_rate": 2.236206397019365e-06,
+      "loss": 0.2977,
+      "step": 10009
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.132742554919447,
+      "learning_rate": 2.235268243523669e-06,
+      "loss": 0.3309,
+      "step": 10010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 9.600541479967166,
+      "learning_rate": 2.2343302302059545e-06,
+      "loss": 0.2901,
+      "step": 10011
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.054193572956136,
+      "learning_rate": 2.2333923571137834e-06,
+      "loss": 0.3255,
+      "step": 10012
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 12.995381166665187,
+      "learning_rate": 2.2324546242947077e-06,
+      "loss": 0.3277,
+      "step": 10013
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.762036815103104,
+      "learning_rate": 2.2315170317962726e-06,
+      "loss": 0.3503,
+      "step": 10014
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.119260710114464,
+      "learning_rate": 2.2305795796660178e-06,
+      "loss": 0.3155,
+      "step": 10015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.0818312279056155,
+      "learning_rate": 2.2296422679514686e-06,
+      "loss": 0.316,
+      "step": 10016
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 25.247482030682484,
+      "learning_rate": 2.228705096700157e-06,
+      "loss": 0.3073,
+      "step": 10017
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.596272224639373,
+      "learning_rate": 2.2277680659595907e-06,
+      "loss": 0.331,
+      "step": 10018
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2275060529860315,
+      "learning_rate": 2.2268311757772875e-06,
+      "loss": 0.2958,
+      "step": 10019
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.43908510061113,
+      "learning_rate": 2.2258944262007447e-06,
+      "loss": 0.3439,
+      "step": 10020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.39175499703799,
+      "learning_rate": 2.224957817277459e-06,
+      "loss": 0.3198,
+      "step": 10021
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.679724533485766,
+      "learning_rate": 2.224021349054918e-06,
+      "loss": 0.3192,
+      "step": 10022
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.522144762677971,
+      "learning_rate": 2.223085021580603e-06,
+      "loss": 0.2803,
+      "step": 10023
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.052981269176039,
+      "learning_rate": 2.2221488349019903e-06,
+      "loss": 0.3365,
+      "step": 10024
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.286379726760665,
+      "learning_rate": 2.2212127890665426e-06,
+      "loss": 0.3197,
+      "step": 10025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.281006411076795,
+      "learning_rate": 2.220276884121721e-06,
+      "loss": 0.3016,
+      "step": 10026
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.202285099976143,
+      "learning_rate": 2.2193411201149777e-06,
+      "loss": 0.3093,
+      "step": 10027
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 49.53127125628199,
+      "learning_rate": 2.218405497093759e-06,
+      "loss": 0.3157,
+      "step": 10028
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.002631186131412,
+      "learning_rate": 2.2174700151055027e-06,
+      "loss": 0.299,
+      "step": 10029
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.194356280162214,
+      "learning_rate": 2.2165346741976412e-06,
+      "loss": 0.311,
+      "step": 10030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.350584639236097,
+      "learning_rate": 2.2155994744175936e-06,
+      "loss": 0.2995,
+      "step": 10031
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.754623324920453,
+      "learning_rate": 2.2146644158127827e-06,
+      "loss": 0.3517,
+      "step": 10032
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.990839493403373,
+      "learning_rate": 2.2137294984306114e-06,
+      "loss": 0.296,
+      "step": 10033
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.09379505496164,
+      "learning_rate": 2.21279472231849e-06,
+      "loss": 0.3285,
+      "step": 10034
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6352240372224323,
+      "learning_rate": 2.2118600875238066e-06,
+      "loss": 0.3151,
+      "step": 10035
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.073478671393045,
+      "learning_rate": 2.2109255940939526e-06,
+      "loss": 0.3401,
+      "step": 10036
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.531673007961829,
+      "learning_rate": 2.2099912420763086e-06,
+      "loss": 0.2858,
+      "step": 10037
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3104050593691191,
+      "learning_rate": 2.209057031518248e-06,
+      "loss": 0.1106,
+      "step": 10038
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6404614995429943,
+      "learning_rate": 2.2081229624671393e-06,
+      "loss": 0.3426,
+      "step": 10039
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.19733683250797,
+      "learning_rate": 2.2071890349703377e-06,
+      "loss": 0.317,
+      "step": 10040
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.206737853457733,
+      "learning_rate": 2.206255249075198e-06,
+      "loss": 0.3118,
+      "step": 10041
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4216146080305117,
+      "learning_rate": 2.2053216048290643e-06,
+      "loss": 0.3216,
+      "step": 10042
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 16.56803403231285,
+      "learning_rate": 2.204388102279277e-06,
+      "loss": 0.3354,
+      "step": 10043
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.171230089573393,
+      "learning_rate": 2.2034547414731617e-06,
+      "loss": 0.3102,
+      "step": 10044
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.167097491409967,
+      "learning_rate": 2.2025215224580475e-06,
+      "loss": 0.2913,
+      "step": 10045
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 17.178772682054692,
+      "learning_rate": 2.201588445281245e-06,
+      "loss": 0.3397,
+      "step": 10046
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.472417673868305,
+      "learning_rate": 2.2006555099900695e-06,
+      "loss": 0.3042,
+      "step": 10047
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.434754975883486,
+      "learning_rate": 2.199722716631818e-06,
+      "loss": 0.3083,
+      "step": 10048
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.2748991887349925,
+      "learning_rate": 2.198790065253787e-06,
+      "loss": 0.3274,
+      "step": 10049
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.2140979160228955,
+      "learning_rate": 2.1978575559032633e-06,
+      "loss": 0.3263,
+      "step": 10050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4134894167094756,
+      "learning_rate": 2.1969251886275277e-06,
+      "loss": 0.311,
+      "step": 10051
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.218026112187507,
+      "learning_rate": 2.1959929634738554e-06,
+      "loss": 0.3032,
+      "step": 10052
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.248459759323351,
+      "learning_rate": 2.195060880489508e-06,
+      "loss": 0.3381,
+      "step": 10053
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.1052819170377655,
+      "learning_rate": 2.1941289397217465e-06,
+      "loss": 0.3019,
+      "step": 10054
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6359210389537697,
+      "learning_rate": 2.1931971412178214e-06,
+      "loss": 0.3398,
+      "step": 10055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.35292927031854,
+      "learning_rate": 2.192265485024978e-06,
+      "loss": 0.2776,
+      "step": 10056
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.458556505841336,
+      "learning_rate": 2.1913339711904534e-06,
+      "loss": 0.3198,
+      "step": 10057
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.721720188581841,
+      "learning_rate": 2.190402599761478e-06,
+      "loss": 0.2947,
+      "step": 10058
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.680712852287985,
+      "learning_rate": 2.1894713707852703e-06,
+      "loss": 0.3261,
+      "step": 10059
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.576343647174617,
+      "learning_rate": 2.188540284309052e-06,
+      "loss": 0.3003,
+      "step": 10060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.331252182950351,
+      "learning_rate": 2.1876093403800264e-06,
+      "loss": 0.3135,
+      "step": 10061
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.760405447829886,
+      "learning_rate": 2.186678539045396e-06,
+      "loss": 0.3491,
+      "step": 10062
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.721131233974312,
+      "learning_rate": 2.185747880352354e-06,
+      "loss": 0.3542,
+      "step": 10063
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.948739283928732,
+      "learning_rate": 2.1848173643480875e-06,
+      "loss": 0.289,
+      "step": 10064
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2682255246700045,
+      "learning_rate": 2.183886991079777e-06,
+      "loss": 0.2969,
+      "step": 10065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.438456113828773,
+      "learning_rate": 2.1829567605945896e-06,
+      "loss": 0.3049,
+      "step": 10066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5219297617471765,
+      "learning_rate": 2.1820266729396965e-06,
+      "loss": 0.305,
+      "step": 10067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.139517504629948,
+      "learning_rate": 2.181096728162251e-06,
+      "loss": 0.3104,
+      "step": 10068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.906096441867182,
+      "learning_rate": 2.1801669263094033e-06,
+      "loss": 0.3116,
+      "step": 10069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.122068087708328,
+      "learning_rate": 2.1792372674282984e-06,
+      "loss": 0.3031,
+      "step": 10070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.253079610188941,
+      "learning_rate": 2.1783077515660707e-06,
+      "loss": 0.3305,
+      "step": 10071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.139973511091454,
+      "learning_rate": 2.1773783787698494e-06,
+      "loss": 0.3099,
+      "step": 10072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.528670889044001,
+      "learning_rate": 2.1764491490867573e-06,
+      "loss": 0.2776,
+      "step": 10073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.624393884314614,
+      "learning_rate": 2.1755200625639045e-06,
+      "loss": 0.3389,
+      "step": 10074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 10.9039103645061,
+      "learning_rate": 2.1745911192484004e-06,
+      "loss": 0.2954,
+      "step": 10075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.449252162965667,
+      "learning_rate": 2.1736623191873428e-06,
+      "loss": 0.2872,
+      "step": 10076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.439009668884511,
+      "learning_rate": 2.172733662427826e-06,
+      "loss": 0.3194,
+      "step": 10077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 19.92832135748034,
+      "learning_rate": 2.1718051490169356e-06,
+      "loss": 0.3312,
+      "step": 10078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.677457352003496,
+      "learning_rate": 2.1708767790017436e-06,
+      "loss": 0.3149,
+      "step": 10079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.306600635375911,
+      "learning_rate": 2.1699485524293285e-06,
+      "loss": 0.316,
+      "step": 10080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28384215795975287,
+      "learning_rate": 2.169020469346747e-06,
+      "loss": 0.1154,
+      "step": 10081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2762604955426107,
+      "learning_rate": 2.1680925298010576e-06,
+      "loss": 0.3214,
+      "step": 10082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.1708666356515876,
+      "learning_rate": 2.167164733839309e-06,
+      "loss": 0.2897,
+      "step": 10083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.432902728546014,
+      "learning_rate": 2.166237081508541e-06,
+      "loss": 0.3454,
+      "step": 10084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.28817953762249887,
+      "learning_rate": 2.1653095728557894e-06,
+      "loss": 0.104,
+      "step": 10085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.307058732172016,
+      "learning_rate": 2.164382207928082e-06,
+      "loss": 0.2923,
+      "step": 10086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 14.065879437687208,
+      "learning_rate": 2.163454986772434e-06,
+      "loss": 0.328,
+      "step": 10087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.427228499286934,
+      "learning_rate": 2.162527909435861e-06,
+      "loss": 0.2715,
+      "step": 10088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.125028384802459,
+      "learning_rate": 2.161600975965366e-06,
+      "loss": 0.2991,
+      "step": 10089
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.800298951976026,
+      "learning_rate": 2.1606741864079477e-06,
+      "loss": 0.321,
+      "step": 10090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.6815787688266832,
+      "learning_rate": 2.159747540810598e-06,
+      "loss": 0.3174,
+      "step": 10091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 7.507355295377151,
+      "learning_rate": 2.158821039220295e-06,
+      "loss": 0.3352,
+      "step": 10092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.351032836426251,
+      "learning_rate": 2.1578946816840195e-06,
+      "loss": 0.3481,
+      "step": 10093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.650225749575495,
+      "learning_rate": 2.1569684682487353e-06,
+      "loss": 0.323,
+      "step": 10094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 22.2889637513303,
+      "learning_rate": 2.1560423989614087e-06,
+      "loss": 0.2907,
+      "step": 10095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.831417729855997,
+      "learning_rate": 2.1551164738689896e-06,
+      "loss": 0.2889,
+      "step": 10096
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.2958331701399125,
+      "learning_rate": 2.1541906930184247e-06,
+      "loss": 0.1101,
+      "step": 10097
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2946170564027986,
+      "learning_rate": 2.153265056456654e-06,
+      "loss": 0.3245,
+      "step": 10098
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.7970907016080275,
+      "learning_rate": 2.15233956423061e-06,
+      "loss": 0.2999,
+      "step": 10099
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1633070082686485,
+      "learning_rate": 2.1514142163872175e-06,
+      "loss": 0.3193,
+      "step": 10100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.305803633489911,
+      "learning_rate": 2.1504890129733907e-06,
+      "loss": 0.2611,
+      "step": 10101
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0132547071663325,
+      "learning_rate": 2.1495639540360413e-06,
+      "loss": 0.3286,
+      "step": 10102
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2987411294050526,
+      "learning_rate": 2.1486390396220723e-06,
+      "loss": 0.331,
+      "step": 10103
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7023340572316155,
+      "learning_rate": 2.147714269778378e-06,
+      "loss": 0.2881,
+      "step": 10104
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7900658245721575,
+      "learning_rate": 2.1467896445518467e-06,
+      "loss": 0.301,
+      "step": 10105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.011425379622647,
+      "learning_rate": 2.1458651639893608e-06,
+      "loss": 0.2882,
+      "step": 10106
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3275857552655586,
+      "learning_rate": 2.144940828137788e-06,
+      "loss": 0.3308,
+      "step": 10107
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.074430856445256,
+      "learning_rate": 2.144016637044002e-06,
+      "loss": 0.3352,
+      "step": 10108
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.869913841293683,
+      "learning_rate": 2.1430925907548528e-06,
+      "loss": 0.3185,
+      "step": 10109
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.363268656614288,
+      "learning_rate": 2.1421686893171993e-06,
+      "loss": 0.3511,
+      "step": 10110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.7264313283340615,
+      "learning_rate": 2.1412449327778806e-06,
+      "loss": 0.3274,
+      "step": 10111
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.9513626981089773,
+      "learning_rate": 2.1403213211837344e-06,
+      "loss": 0.3166,
+      "step": 10112
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 8.047204087041075,
+      "learning_rate": 2.1393978545815896e-06,
+      "loss": 0.3129,
+      "step": 10113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.767775613429867,
+      "learning_rate": 2.1384745330182686e-06,
+      "loss": 0.2974,
+      "step": 10114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7003474892890256,
+      "learning_rate": 2.137551356540588e-06,
+      "loss": 0.3145,
+      "step": 10115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4895886146632633,
+      "learning_rate": 2.13662832519535e-06,
+      "loss": 0.3217,
+      "step": 10116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.687544612811198,
+      "learning_rate": 2.135705439029357e-06,
+      "loss": 0.3269,
+      "step": 10117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.167035140209814,
+      "learning_rate": 2.134782698089401e-06,
+      "loss": 0.3548,
+      "step": 10118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.288761807484519,
+      "learning_rate": 2.1338601024222687e-06,
+      "loss": 0.3331,
+      "step": 10119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.976764052168235,
+      "learning_rate": 2.132937652074733e-06,
+      "loss": 0.3311,
+      "step": 10120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3083948551553735,
+      "learning_rate": 2.1320153470935704e-06,
+      "loss": 0.1135,
+      "step": 10121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.956321284399753,
+      "learning_rate": 2.1310931875255375e-06,
+      "loss": 0.2969,
+      "step": 10122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.316724623849598,
+      "learning_rate": 2.130171173417397e-06,
+      "loss": 0.3129,
+      "step": 10123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.7191619099857656,
+      "learning_rate": 2.129249304815891e-06,
+      "loss": 0.3049,
+      "step": 10124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.280477850303732,
+      "learning_rate": 2.1283275817677617e-06,
+      "loss": 0.2933,
+      "step": 10125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.8374127635935427,
+      "learning_rate": 2.127406004319744e-06,
+      "loss": 0.3438,
+      "step": 10126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.565797604438706,
+      "learning_rate": 2.1264845725185623e-06,
+      "loss": 0.3204,
+      "step": 10127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.3286127423763507,
+      "learning_rate": 2.125563286410938e-06,
+      "loss": 0.3514,
+      "step": 10128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.030298826667512,
+      "learning_rate": 2.124642146043579e-06,
+      "loss": 0.3304,
+      "step": 10129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.927430463432454,
+      "learning_rate": 2.12372115146319e-06,
+      "loss": 0.3053,
+      "step": 10130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.4689985514283332,
+      "learning_rate": 2.1228003027164678e-06,
+      "loss": 0.3394,
+      "step": 10131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.6916599800059284,
+      "learning_rate": 2.121879599850102e-06,
+      "loss": 0.3177,
+      "step": 10132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2956542638758193,
+      "learning_rate": 2.1209590429107734e-06,
+      "loss": 0.2935,
+      "step": 10133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.098463559627677,
+      "learning_rate": 2.120038631945159e-06,
+      "loss": 0.3262,
+      "step": 10134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.303494244948899,
+      "learning_rate": 2.1191183669999187e-06,
+      "loss": 0.2916,
+      "step": 10135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.897406402311237,
+      "learning_rate": 2.118198248121721e-06,
+      "loss": 0.3017,
+      "step": 10136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 6.478046218910735,
+      "learning_rate": 2.1172782753572117e-06,
+      "loss": 0.3175,
+      "step": 10137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.591865099826546,
+      "learning_rate": 2.116358448753037e-06,
+      "loss": 0.2964,
+      "step": 10138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3001683057960041,
+      "learning_rate": 2.115438768355835e-06,
+      "loss": 0.112,
+      "step": 10139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 5.206755560242226,
+      "learning_rate": 2.1145192342122338e-06,
+      "loss": 0.3052,
+      "step": 10140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 33.81298865954182,
+      "learning_rate": 2.1135998463688596e-06,
+      "loss": 0.3182,
+      "step": 10141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.389664412303838,
+      "learning_rate": 2.1126806048723208e-06,
+      "loss": 0.3131,
+      "step": 10142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.995826018774752,
+      "learning_rate": 2.111761509769233e-06,
+      "loss": 0.3413,
+      "step": 10143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.279961429101715,
+      "learning_rate": 2.11084256110619e-06,
+      "loss": 0.3276,
+      "step": 10144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.427429905975643,
+      "learning_rate": 2.1099237589297867e-06,
+      "loss": 0.3083,
+      "step": 10145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.151390818057225,
+      "learning_rate": 2.109005103286609e-06,
+      "loss": 0.3229,
+      "step": 10146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.31318080370101453,
+      "learning_rate": 2.108086594223234e-06,
+      "loss": 0.1229,
+      "step": 10147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.141749537932232,
+      "learning_rate": 2.107168231786231e-06,
+      "loss": 0.3291,
+      "step": 10148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.040231298722051,
+      "learning_rate": 2.1062500160221677e-06,
+      "loss": 0.325,
+      "step": 10149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9712376954614856,
+      "learning_rate": 2.1053319469775942e-06,
+      "loss": 0.3104,
+      "step": 10150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.296719955400982,
+      "learning_rate": 2.1044140246990607e-06,
+      "loss": 0.3484,
+      "step": 10151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.474224389026386,
+      "learning_rate": 2.103496249233108e-06,
+      "loss": 0.3235,
+      "step": 10152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.336323265684127,
+      "learning_rate": 2.102578620626269e-06,
+      "loss": 0.3179,
+      "step": 10153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.329881867657955,
+      "learning_rate": 2.1016611389250713e-06,
+      "loss": 0.3495,
+      "step": 10154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30163980487322956,
+      "learning_rate": 2.100743804176028e-06,
+      "loss": 0.1093,
+      "step": 10155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2777098796992786,
+      "learning_rate": 2.099826616425658e-06,
+      "loss": 0.1128,
+      "step": 10156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.3773967627169466,
+      "learning_rate": 2.098909575720458e-06,
+      "loss": 0.369,
+      "step": 10157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5186927655863594,
+      "learning_rate": 2.097992682106927e-06,
+      "loss": 0.296,
+      "step": 10158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9880352878432466,
+      "learning_rate": 2.0970759356315523e-06,
+      "loss": 0.2978,
+      "step": 10159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 10.7344922033813,
+      "learning_rate": 2.0961593363408154e-06,
+      "loss": 0.3389,
+      "step": 10160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.197654905667418,
+      "learning_rate": 2.095242884281191e-06,
+      "loss": 0.3197,
+      "step": 10161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3281998652792792,
+      "learning_rate": 2.094326579499145e-06,
+      "loss": 0.1177,
+      "step": 10162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 14.07871860250462,
+      "learning_rate": 2.0934104220411337e-06,
+      "loss": 0.2623,
+      "step": 10163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.308507684034427,
+      "learning_rate": 2.0924944119536105e-06,
+      "loss": 0.316,
+      "step": 10164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.3574693620080245,
+      "learning_rate": 2.0915785492830187e-06,
+      "loss": 0.3184,
+      "step": 10165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30289095969396546,
+      "learning_rate": 2.090662834075794e-06,
+      "loss": 0.1206,
+      "step": 10166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.550950483210305,
+      "learning_rate": 2.0897472663783673e-06,
+      "loss": 0.3602,
+      "step": 10167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.974897180520297,
+      "learning_rate": 2.0888318462371547e-06,
+      "loss": 0.3144,
+      "step": 10168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.258538010531391,
+      "learning_rate": 2.087916573698578e-06,
+      "loss": 0.3196,
+      "step": 10169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.443885963644803,
+      "learning_rate": 2.0870014488090345e-06,
+      "loss": 0.328,
+      "step": 10170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.489962375624599,
+      "learning_rate": 2.0860864716149315e-06,
+      "loss": 0.3492,
+      "step": 10171
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.595369620751759,
+      "learning_rate": 2.085171642162655e-06,
+      "loss": 0.2928,
+      "step": 10172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9981974533052544,
+      "learning_rate": 2.0842569604985897e-06,
+      "loss": 0.3307,
+      "step": 10173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1262334445478044,
+      "learning_rate": 2.0833424266691132e-06,
+      "loss": 0.328,
+      "step": 10174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4962527236030376,
+      "learning_rate": 2.082428040720593e-06,
+      "loss": 0.3339,
+      "step": 10175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.084890358042361,
+      "learning_rate": 2.081513802699394e-06,
+      "loss": 0.3463,
+      "step": 10176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.100798321271977,
+      "learning_rate": 2.080599712651865e-06,
+      "loss": 0.3316,
+      "step": 10177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.812695018091039,
+      "learning_rate": 2.079685770624355e-06,
+      "loss": 0.3119,
+      "step": 10178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.765817802751447,
+      "learning_rate": 2.0787719766632025e-06,
+      "loss": 0.3073,
+      "step": 10179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.606307023092818,
+      "learning_rate": 2.077858330814739e-06,
+      "loss": 0.3251,
+      "step": 10180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.754352744179956,
+      "learning_rate": 2.076944833125288e-06,
+      "loss": 0.3286,
+      "step": 10181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.472528866592684,
+      "learning_rate": 2.0760314836411684e-06,
+      "loss": 0.3293,
+      "step": 10182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.352585099802675,
+      "learning_rate": 2.075118282408683e-06,
+      "loss": 0.3183,
+      "step": 10183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0726029240557873,
+      "learning_rate": 2.0742052294741405e-06,
+      "loss": 0.3119,
+      "step": 10184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2520417131270984,
+      "learning_rate": 2.073292324883827e-06,
+      "loss": 0.2983,
+      "step": 10185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.771636649690903,
+      "learning_rate": 2.072379568684037e-06,
+      "loss": 0.3233,
+      "step": 10186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.3741657933766405,
+      "learning_rate": 2.0714669609210437e-06,
+      "loss": 0.3079,
+      "step": 10187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.123708682007859,
+      "learning_rate": 2.070554501641119e-06,
+      "loss": 0.2965,
+      "step": 10188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.255222303058492,
+      "learning_rate": 2.0696421908905294e-06,
+      "loss": 0.2813,
+      "step": 10189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.9950381387282405,
+      "learning_rate": 2.068730028715527e-06,
+      "loss": 0.3143,
+      "step": 10190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.404448146677773,
+      "learning_rate": 2.067818015162363e-06,
+      "loss": 0.3112,
+      "step": 10191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.337704178937599,
+      "learning_rate": 2.0669061502772776e-06,
+      "loss": 0.2802,
+      "step": 10192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4668215558012307,
+      "learning_rate": 2.065994434106504e-06,
+      "loss": 0.3134,
+      "step": 10193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3139531129625105,
+      "learning_rate": 2.06508286669627e-06,
+      "loss": 0.1043,
+      "step": 10194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.96282100351778,
+      "learning_rate": 2.064171448092794e-06,
+      "loss": 0.3108,
+      "step": 10195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4265101403273484,
+      "learning_rate": 2.063260178342282e-06,
+      "loss": 0.3351,
+      "step": 10196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.5408242700395345,
+      "learning_rate": 2.062349057490946e-06,
+      "loss": 0.3347,
+      "step": 10197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.133867287535144,
+      "learning_rate": 2.0614380855849727e-06,
+      "loss": 0.2993,
+      "step": 10198
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.047161059367696,
+      "learning_rate": 2.0605272626705587e-06,
+      "loss": 0.3127,
+      "step": 10199
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.489279207639553,
+      "learning_rate": 2.059616588793879e-06,
+      "loss": 0.2855,
+      "step": 10200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.928479406447039,
+      "learning_rate": 2.058706064001108e-06,
+      "loss": 0.2769,
+      "step": 10201
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.846287843161873,
+      "learning_rate": 2.0577956883384126e-06,
+      "loss": 0.3279,
+      "step": 10202
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.344071379231892,
+      "learning_rate": 2.05688546185195e-06,
+      "loss": 0.3045,
+      "step": 10203
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.540159009120508,
+      "learning_rate": 2.055975384587872e-06,
+      "loss": 0.327,
+      "step": 10204
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5433571856583015,
+      "learning_rate": 2.0550654565923196e-06,
+      "loss": 0.283,
+      "step": 10205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.194313906591654,
+      "learning_rate": 2.054155677911428e-06,
+      "loss": 0.3689,
+      "step": 10206
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.345258366374407,
+      "learning_rate": 2.0532460485913263e-06,
+      "loss": 0.2922,
+      "step": 10207
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.082838802923101,
+      "learning_rate": 2.0523365686781345e-06,
+      "loss": 0.3498,
+      "step": 10208
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 8.3131732046637,
+      "learning_rate": 2.0514272382179643e-06,
+      "loss": 0.302,
+      "step": 10209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8362585422956808,
+      "learning_rate": 2.050518057256924e-06,
+      "loss": 0.2951,
+      "step": 10210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2210659442649323,
+      "learning_rate": 2.0496090258411044e-06,
+      "loss": 0.2981,
+      "step": 10211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.326956754688793,
+      "learning_rate": 2.048700144016603e-06,
+      "loss": 0.3322,
+      "step": 10212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4043490919499466,
+      "learning_rate": 2.0477914118294976e-06,
+      "loss": 0.3287,
+      "step": 10213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33439908046042677,
+      "learning_rate": 2.0468828293258634e-06,
+      "loss": 0.1143,
+      "step": 10214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.952605722273948,
+      "learning_rate": 2.0459743965517685e-06,
+      "loss": 0.3536,
+      "step": 10215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.29397972184184695,
+      "learning_rate": 2.0450661135532717e-06,
+      "loss": 0.1138,
+      "step": 10216
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.356326845593911,
+      "learning_rate": 2.0441579803764273e-06,
+      "loss": 0.3146,
+      "step": 10217
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9656130314407148,
+      "learning_rate": 2.043249997067274e-06,
+      "loss": 0.3285,
+      "step": 10218
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3110311993056633,
+      "learning_rate": 2.042342163671856e-06,
+      "loss": 0.12,
+      "step": 10219
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.3366435658075275,
+      "learning_rate": 2.041434480236196e-06,
+      "loss": 0.3462,
+      "step": 10220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.057238851082445,
+      "learning_rate": 2.0405269468063194e-06,
+      "loss": 0.333,
+      "step": 10221
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.631009033732871,
+      "learning_rate": 2.0396195634282385e-06,
+      "loss": 0.3338,
+      "step": 10222
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.38304242624552,
+      "learning_rate": 2.0387123301479594e-06,
+      "loss": 0.3207,
+      "step": 10223
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4538855306497,
+      "learning_rate": 2.0378052470114822e-06,
+      "loss": 0.312,
+      "step": 10224
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.107520380804592,
+      "learning_rate": 2.0368983140647992e-06,
+      "loss": 0.2997,
+      "step": 10225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.5664497842280785,
+      "learning_rate": 2.03599153135389e-06,
+      "loss": 0.2969,
+      "step": 10226
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.714512404125313,
+      "learning_rate": 2.035084898924732e-06,
+      "loss": 0.3022,
+      "step": 10227
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.345761426778988,
+      "learning_rate": 2.034178416823294e-06,
+      "loss": 0.2979,
+      "step": 10228
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.902491124291846,
+      "learning_rate": 2.033272085095537e-06,
+      "loss": 0.3062,
+      "step": 10229
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1067315490556506,
+      "learning_rate": 2.0323659037874157e-06,
+      "loss": 0.2756,
+      "step": 10230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8431436625036963,
+      "learning_rate": 2.0314598729448687e-06,
+      "loss": 0.3329,
+      "step": 10231
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 13.302407499718083,
+      "learning_rate": 2.0305539926138427e-06,
+      "loss": 0.3278,
+      "step": 10232
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.552287992309089,
+      "learning_rate": 2.029648262840262e-06,
+      "loss": 0.3215,
+      "step": 10233
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.623148388001417,
+      "learning_rate": 2.02874268367005e-06,
+      "loss": 0.3253,
+      "step": 10234
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.4035829351652804,
+      "learning_rate": 2.0278372551491227e-06,
+      "loss": 0.2883,
+      "step": 10235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.033665716204164,
+      "learning_rate": 2.026931977323387e-06,
+      "loss": 0.3378,
+      "step": 10236
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.4142148083421615,
+      "learning_rate": 2.026026850238742e-06,
+      "loss": 0.2977,
+      "step": 10237
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9842179191260705,
+      "learning_rate": 2.0251218739410815e-06,
+      "loss": 0.3355,
+      "step": 10238
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9853067521655694,
+      "learning_rate": 2.0242170484762867e-06,
+      "loss": 0.3569,
+      "step": 10239
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0291575587425092,
+      "learning_rate": 2.0233123738902355e-06,
+      "loss": 0.321,
+      "step": 10240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6122123413037195,
+      "learning_rate": 2.0224078502287974e-06,
+      "loss": 0.332,
+      "step": 10241
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6992272596706575,
+      "learning_rate": 2.0215034775378336e-06,
+      "loss": 0.3089,
+      "step": 10242
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.868370959699352,
+      "learning_rate": 2.020599255863199e-06,
+      "loss": 0.2899,
+      "step": 10243
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2791835281045316,
+      "learning_rate": 2.019695185250735e-06,
+      "loss": 0.0975,
+      "step": 10244
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.61064295314822,
+      "learning_rate": 2.0187912657462873e-06,
+      "loss": 0.3028,
+      "step": 10245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.47996927974397,
+      "learning_rate": 2.0178874973956784e-06,
+      "loss": 0.3274,
+      "step": 10246
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8569863484143676,
+      "learning_rate": 2.0169838802447395e-06,
+      "loss": 0.309,
+      "step": 10247
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.133962304759528,
+      "learning_rate": 2.0160804143392803e-06,
+      "loss": 0.3102,
+      "step": 10248
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8599308716312017,
+      "learning_rate": 2.0151770997251104e-06,
+      "loss": 0.3498,
+      "step": 10249
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.044058598324313,
+      "learning_rate": 2.014273936448029e-06,
+      "loss": 0.3125,
+      "step": 10250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.780542952341524,
+      "learning_rate": 2.013370924553829e-06,
+      "loss": 0.3379,
+      "step": 10251
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.040705090163023,
+      "learning_rate": 2.0124680640882977e-06,
+      "loss": 0.3246,
+      "step": 10252
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.2919583563173294,
+      "learning_rate": 2.0115653550972073e-06,
+      "loss": 0.1071,
+      "step": 10253
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.325964680333649,
+      "learning_rate": 2.0106627976263297e-06,
+      "loss": 0.3156,
+      "step": 10254
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5797003554344036,
+      "learning_rate": 2.0097603917214267e-06,
+      "loss": 0.2738,
+      "step": 10255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 18.4873042337706,
+      "learning_rate": 2.0088581374282514e-06,
+      "loss": 0.3071,
+      "step": 10256
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.30910566555019126,
+      "learning_rate": 2.007956034792551e-06,
+      "loss": 0.1143,
+      "step": 10257
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2536579771379546,
+      "learning_rate": 2.0070540838600648e-06,
+      "loss": 0.3186,
+      "step": 10258
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.536868210500403,
+      "learning_rate": 2.006152284676519e-06,
+      "loss": 0.2982,
+      "step": 10259
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.077513769314076,
+      "learning_rate": 2.0052506372876444e-06,
+      "loss": 0.3109,
+      "step": 10260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9975714078807045,
+      "learning_rate": 2.0043491417391505e-06,
+      "loss": 0.331,
+      "step": 10261
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8818381746188835,
+      "learning_rate": 2.003447798076747e-06,
+      "loss": 0.2943,
+      "step": 10262
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.776446536715423,
+      "learning_rate": 2.0025466063461343e-06,
+      "loss": 0.3207,
+      "step": 10263
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.206981120478375,
+      "learning_rate": 2.001645566593004e-06,
+      "loss": 0.3421,
+      "step": 10264
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 7.833196147160578,
+      "learning_rate": 2.0007446788630434e-06,
+      "loss": 0.3396,
+      "step": 10265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.21745115249125,
+      "learning_rate": 1.999843943201926e-06,
+      "loss": 0.3306,
+      "step": 10266
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.396377138331904,
+      "learning_rate": 1.998943359655322e-06,
+      "loss": 0.299,
+      "step": 10267
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.830584895088993,
+      "learning_rate": 1.998042928268894e-06,
+      "loss": 0.2966,
+      "step": 10268
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6761649845830617,
+      "learning_rate": 1.997142649088295e-06,
+      "loss": 0.3218,
+      "step": 10269
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.81067849377369,
+      "learning_rate": 1.996242522159172e-06,
+      "loss": 0.3237,
+      "step": 10270
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.596869718220781,
+      "learning_rate": 1.995342547527164e-06,
+      "loss": 0.3128,
+      "step": 10271
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.455543347990166,
+      "learning_rate": 1.994442725237898e-06,
+      "loss": 0.2975,
+      "step": 10272
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.889654336210766,
+      "learning_rate": 1.9935430553370023e-06,
+      "loss": 0.3038,
+      "step": 10273
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.8094826047867008,
+      "learning_rate": 1.992643537870086e-06,
+      "loss": 0.3059,
+      "step": 10274
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6918840123418795,
+      "learning_rate": 1.9917441728827634e-06,
+      "loss": 0.3196,
+      "step": 10275
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.5763110438300676,
+      "learning_rate": 1.99084496042063e-06,
+      "loss": 0.3323,
+      "step": 10276
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 9.814485943277669,
+      "learning_rate": 1.989945900529278e-06,
+      "loss": 0.3099,
+      "step": 10277
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3164838540190764,
+      "learning_rate": 1.9890469932542946e-06,
+      "loss": 0.1231,
+      "step": 10278
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.7943351946244817,
+      "learning_rate": 1.9881482386412505e-06,
+      "loss": 0.3075,
+      "step": 10279
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.423147716362978,
+      "learning_rate": 1.987249636735722e-06,
+      "loss": 0.2921,
+      "step": 10280
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.3826001242771304,
+      "learning_rate": 1.986351187583264e-06,
+      "loss": 0.3285,
+      "step": 10281
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.2309333158556246,
+      "learning_rate": 1.9854528912294335e-06,
+      "loss": 0.3033,
+      "step": 10282
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.900823582744782,
+      "learning_rate": 1.984554747719774e-06,
+      "loss": 0.3136,
+      "step": 10283
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.6670095263554767,
+      "learning_rate": 1.983656757099825e-06,
+      "loss": 0.3158,
+      "step": 10284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.788646870699407,
+      "learning_rate": 1.9827589194151152e-06,
+      "loss": 0.3667,
+      "step": 10285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.952043848114277,
+      "learning_rate": 1.98186123471117e-06,
+      "loss": 0.2973,
+      "step": 10286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28851652174512626,
+      "learning_rate": 1.9809637030334975e-06,
+      "loss": 0.1116,
+      "step": 10287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.029254194755174,
+      "learning_rate": 1.980066324427613e-06,
+      "loss": 0.3043,
+      "step": 10288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5629763039975058,
+      "learning_rate": 1.979169098939009e-06,
+      "loss": 0.324,
+      "step": 10289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.10226329394097,
+      "learning_rate": 1.9782720266131796e-06,
+      "loss": 0.3148,
+      "step": 10290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.388937398963052,
+      "learning_rate": 1.977375107495607e-06,
+      "loss": 0.3028,
+      "step": 10291
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.806468814359428,
+      "learning_rate": 1.9764783416317677e-06,
+      "loss": 0.2927,
+      "step": 10292
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.872463887964478,
+      "learning_rate": 1.975581729067132e-06,
+      "loss": 0.3111,
+      "step": 10293
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.295787689602721,
+      "learning_rate": 1.9746852698471542e-06,
+      "loss": 0.3276,
+      "step": 10294
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.808282483279977,
+      "learning_rate": 1.973788964017294e-06,
+      "loss": 0.3178,
+      "step": 10295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.11323191948283,
+      "learning_rate": 1.9728928116229907e-06,
+      "loss": 0.3242,
+      "step": 10296
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.71263084080346,
+      "learning_rate": 1.971996812709683e-06,
+      "loss": 0.3344,
+      "step": 10297
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.1910437408339005,
+      "learning_rate": 1.9711009673228e-06,
+      "loss": 0.2816,
+      "step": 10298
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.343555615962785,
+      "learning_rate": 1.970205275507765e-06,
+      "loss": 0.312,
+      "step": 10299
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.891704916572504,
+      "learning_rate": 1.9693097373099857e-06,
+      "loss": 0.3001,
+      "step": 10300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.391234664039273,
+      "learning_rate": 1.9684143527748755e-06,
+      "loss": 0.3291,
+      "step": 10301
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33049536770585963,
+      "learning_rate": 1.9675191219478273e-06,
+      "loss": 0.1227,
+      "step": 10302
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.31123282747688275,
+      "learning_rate": 1.9666240448742324e-06,
+      "loss": 0.1047,
+      "step": 10303
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.280466130484516,
+      "learning_rate": 1.965729121599473e-06,
+      "loss": 0.3097,
+      "step": 10304
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2934016290150228,
+      "learning_rate": 1.9648343521689245e-06,
+      "loss": 0.1162,
+      "step": 10305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.93299813307445,
+      "learning_rate": 1.9639397366279543e-06,
+      "loss": 0.3238,
+      "step": 10306
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6086369285150823,
+      "learning_rate": 1.9630452750219175e-06,
+      "loss": 0.3321,
+      "step": 10307
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9384600343402685,
+      "learning_rate": 1.9621509673961714e-06,
+      "loss": 0.3592,
+      "step": 10308
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.257122381950855,
+      "learning_rate": 1.961256813796054e-06,
+      "loss": 0.3357,
+      "step": 10309
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.3675857427935405,
+      "learning_rate": 1.9603628142669034e-06,
+      "loss": 0.3435,
+      "step": 10310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8352756125551184,
+      "learning_rate": 1.959468968854047e-06,
+      "loss": 0.3313,
+      "step": 10311
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.2988702072846421,
+      "learning_rate": 1.958575277602804e-06,
+      "loss": 0.1112,
+      "step": 10312
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.9569099610956915,
+      "learning_rate": 1.9576817405584875e-06,
+      "loss": 0.32,
+      "step": 10313
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.980063971184108,
+      "learning_rate": 1.9567883577664027e-06,
+      "loss": 0.2822,
+      "step": 10314
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 16.8829891373223,
+      "learning_rate": 1.9558951292718436e-06,
+      "loss": 0.3099,
+      "step": 10315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.247091279558814,
+      "learning_rate": 1.9550020551201e-06,
+      "loss": 0.3043,
+      "step": 10316
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9229935986881865,
+      "learning_rate": 1.9541091353564525e-06,
+      "loss": 0.2676,
+      "step": 10317
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.76367123292301,
+      "learning_rate": 1.9532163700261734e-06,
+      "loss": 0.318,
+      "step": 10318
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.336926843042544,
+      "learning_rate": 1.952323759174532e-06,
+      "loss": 0.3221,
+      "step": 10319
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.943353151031913,
+      "learning_rate": 1.9514313028467783e-06,
+      "loss": 0.3435,
+      "step": 10320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.125549689861913,
+      "learning_rate": 1.95053900108817e-06,
+      "loss": 0.3292,
+      "step": 10321
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 11.475676787920857,
+      "learning_rate": 1.9496468539439403e-06,
+      "loss": 0.2978,
+      "step": 10322
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.388317327923101,
+      "learning_rate": 1.948754861459331e-06,
+      "loss": 0.301,
+      "step": 10323
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.859881123199775,
+      "learning_rate": 1.9478630236795633e-06,
+      "loss": 0.2819,
+      "step": 10324
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.38795470588599,
+      "learning_rate": 1.9469713406498563e-06,
+      "loss": 0.3526,
+      "step": 10325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1176654439460343,
+      "learning_rate": 1.9460798124154205e-06,
+      "loss": 0.2989,
+      "step": 10326
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4440472030275786,
+      "learning_rate": 1.945188439021459e-06,
+      "loss": 0.3418,
+      "step": 10327
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.389699241751033,
+      "learning_rate": 1.9442972205131667e-06,
+      "loss": 0.3209,
+      "step": 10328
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.102757725781651,
+      "learning_rate": 1.9434061569357277e-06,
+      "loss": 0.3342,
+      "step": 10329
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.348782802312145,
+      "learning_rate": 1.942515248334323e-06,
+      "loss": 0.2875,
+      "step": 10330
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.498297452860342,
+      "learning_rate": 1.941624494754123e-06,
+      "loss": 0.3223,
+      "step": 10331
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.833252811704349,
+      "learning_rate": 1.940733896240291e-06,
+      "loss": 0.3613,
+      "step": 10332
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.348849405315414,
+      "learning_rate": 1.939843452837982e-06,
+      "loss": 0.2947,
+      "step": 10333
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.008843309741145,
+      "learning_rate": 1.9389531645923466e-06,
+      "loss": 0.2881,
+      "step": 10334
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.687641371587951,
+      "learning_rate": 1.9380630315485177e-06,
+      "loss": 0.3089,
+      "step": 10335
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.102947279323508,
+      "learning_rate": 1.9371730537516344e-06,
+      "loss": 0.2856,
+      "step": 10336
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1973185345466217,
+      "learning_rate": 1.936283231246815e-06,
+      "loss": 0.3295,
+      "step": 10337
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8105728869227526,
+      "learning_rate": 1.9353935640791778e-06,
+      "loss": 0.3197,
+      "step": 10338
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6614229143897137,
+      "learning_rate": 1.9345040522938304e-06,
+      "loss": 0.337,
+      "step": 10339
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.477962575174155,
+      "learning_rate": 1.933614695935873e-06,
+      "loss": 0.3025,
+      "step": 10340
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.692371298032495,
+      "learning_rate": 1.9327254950504006e-06,
+      "loss": 0.3065,
+      "step": 10341
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.740992681229088,
+      "learning_rate": 1.931836449682493e-06,
+      "loss": 0.3477,
+      "step": 10342
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28300988492448553,
+      "learning_rate": 1.9309475598772287e-06,
+      "loss": 0.0915,
+      "step": 10343
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.505328890933215,
+      "learning_rate": 1.9300588256796766e-06,
+      "loss": 0.3081,
+      "step": 10344
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.213957201125714,
+      "learning_rate": 1.929170247134898e-06,
+      "loss": 0.3144,
+      "step": 10345
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.693310753252629,
+      "learning_rate": 1.928281824287944e-06,
+      "loss": 0.2931,
+      "step": 10346
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.1731834958007084,
+      "learning_rate": 1.9273935571838635e-06,
+      "loss": 0.3048,
+      "step": 10347
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9533859385356456,
+      "learning_rate": 1.926505445867688e-06,
+      "loss": 0.288,
+      "step": 10348
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.281259910199151,
+      "learning_rate": 1.925617490384452e-06,
+      "loss": 0.3448,
+      "step": 10349
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.5430912978012734,
+      "learning_rate": 1.924729690779172e-06,
+      "loss": 0.329,
+      "step": 10350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.0930164741665593,
+      "learning_rate": 1.9238420470968665e-06,
+      "loss": 0.2933,
+      "step": 10351
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9424945180898683,
+      "learning_rate": 1.9229545593825367e-06,
+      "loss": 0.3098,
+      "step": 10352
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7305715697370117,
+      "learning_rate": 1.9220672276811818e-06,
+      "loss": 0.2986,
+      "step": 10353
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3775365543302236,
+      "learning_rate": 1.9211800520377933e-06,
+      "loss": 0.313,
+      "step": 10354
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 31.94615554993429,
+      "learning_rate": 1.920293032497348e-06,
+      "loss": 0.2857,
+      "step": 10355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.515180069705385,
+      "learning_rate": 1.919406169104826e-06,
+      "loss": 0.3333,
+      "step": 10356
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.782383998694498,
+      "learning_rate": 1.9185194619051882e-06,
+      "loss": 0.295,
+      "step": 10357
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0946243835109764,
+      "learning_rate": 1.9176329109433946e-06,
+      "loss": 0.3058,
+      "step": 10358
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.34185809736441,
+      "learning_rate": 1.916746516264395e-06,
+      "loss": 0.3123,
+      "step": 10359
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.802694936491634,
+      "learning_rate": 1.915860277913132e-06,
+      "loss": 0.3036,
+      "step": 10360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.2995335421567,
+      "learning_rate": 1.91497419593454e-06,
+      "loss": 0.3518,
+      "step": 10361
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.553549774474493,
+      "learning_rate": 1.914088270373546e-06,
+      "loss": 0.3308,
+      "step": 10362
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.679878870011003,
+      "learning_rate": 1.9132025012750636e-06,
+      "loss": 0.3387,
+      "step": 10363
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.799577363611897,
+      "learning_rate": 1.912316888684011e-06,
+      "loss": 0.3032,
+      "step": 10364
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7214697584526744,
+      "learning_rate": 1.9114314326452842e-06,
+      "loss": 0.3143,
+      "step": 10365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.369301491542975,
+      "learning_rate": 1.9105461332037806e-06,
+      "loss": 0.3309,
+      "step": 10366
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.285834472758758,
+      "learning_rate": 1.909660990404388e-06,
+      "loss": 0.2788,
+      "step": 10367
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9078765434140714,
+      "learning_rate": 1.9087760042919808e-06,
+      "loss": 0.3349,
+      "step": 10368
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.35347811634933,
+      "learning_rate": 1.907891174911435e-06,
+      "loss": 0.3379,
+      "step": 10369
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.519766710412642,
+      "learning_rate": 1.907006502307609e-06,
+      "loss": 0.2891,
+      "step": 10370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 12.238144130091028,
+      "learning_rate": 1.9061219865253599e-06,
+      "loss": 0.3026,
+      "step": 10371
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 9.247916886990232,
+      "learning_rate": 1.9052376276095342e-06,
+      "loss": 0.3145,
+      "step": 10372
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7895603314665287,
+      "learning_rate": 1.9043534256049707e-06,
+      "loss": 0.2981,
+      "step": 10373
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.586777311809731,
+      "learning_rate": 1.9034693805565002e-06,
+      "loss": 0.308,
+      "step": 10374
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.065747060539042,
+      "learning_rate": 1.9025854925089481e-06,
+      "loss": 0.3219,
+      "step": 10375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9396727317089084,
+      "learning_rate": 1.9017017615071242e-06,
+      "loss": 0.3265,
+      "step": 10376
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.635771435380974,
+      "learning_rate": 1.900818187595842e-06,
+      "loss": 0.311,
+      "step": 10377
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.114495514147153,
+      "learning_rate": 1.8999347708198956e-06,
+      "loss": 0.3262,
+      "step": 10378
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.033495349805739,
+      "learning_rate": 1.8990515112240783e-06,
+      "loss": 0.3237,
+      "step": 10379
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3613018507082737,
+      "learning_rate": 1.8981684088531727e-06,
+      "loss": 0.3075,
+      "step": 10380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28850312789811244,
+      "learning_rate": 1.8972854637519545e-06,
+      "loss": 0.1057,
+      "step": 10381
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.923210204124926,
+      "learning_rate": 1.8964026759651926e-06,
+      "loss": 0.3015,
+      "step": 10382
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3093967798060318,
+      "learning_rate": 1.8955200455376411e-06,
+      "loss": 0.3232,
+      "step": 10383
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.3812437489223734,
+      "learning_rate": 1.8946375725140581e-06,
+      "loss": 0.3368,
+      "step": 10384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.655484955064764,
+      "learning_rate": 1.893755256939182e-06,
+      "loss": 0.3516,
+      "step": 10385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3218557052916604,
+      "learning_rate": 1.8928730988577498e-06,
+      "loss": 0.3398,
+      "step": 10386
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.937910104663914,
+      "learning_rate": 1.8919910983144885e-06,
+      "loss": 0.2793,
+      "step": 10387
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.29309644206106905,
+      "learning_rate": 1.8911092553541172e-06,
+      "loss": 0.1174,
+      "step": 10388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.112602512410603,
+      "learning_rate": 1.8902275700213484e-06,
+      "loss": 0.3018,
+      "step": 10389
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 8.50300865167836,
+      "learning_rate": 1.8893460423608866e-06,
+      "loss": 0.3016,
+      "step": 10390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4035423233502047,
+      "learning_rate": 1.8884646724174233e-06,
+      "loss": 0.2997,
+      "step": 10391
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.956361875770918,
+      "learning_rate": 1.8875834602356485e-06,
+      "loss": 0.3243,
+      "step": 10392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.778076398361758,
+      "learning_rate": 1.8867024058602412e-06,
+      "loss": 0.3229,
+      "step": 10393
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.522872483096856,
+      "learning_rate": 1.8858215093358723e-06,
+      "loss": 0.3192,
+      "step": 10394
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.706644138661751,
+      "learning_rate": 1.8849407707072075e-06,
+      "loss": 0.3112,
+      "step": 10395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.414064098765038,
+      "learning_rate": 1.884060190018897e-06,
+      "loss": 0.3338,
+      "step": 10396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.415877470267815,
+      "learning_rate": 1.883179767315595e-06,
+      "loss": 0.2985,
+      "step": 10397
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 7.307211427068227,
+      "learning_rate": 1.882299502641934e-06,
+      "loss": 0.2969,
+      "step": 10398
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7950071118884265,
+      "learning_rate": 1.881419396042552e-06,
+      "loss": 0.2926,
+      "step": 10399
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.697125466775474,
+      "learning_rate": 1.8805394475620674e-06,
+      "loss": 0.3104,
+      "step": 10400
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6186607181661397,
+      "learning_rate": 1.879659657245097e-06,
+      "loss": 0.3384,
+      "step": 10401
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.9641427626466688,
+      "learning_rate": 1.8787800251362486e-06,
+      "loss": 0.3469,
+      "step": 10402
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.547095927578076,
+      "learning_rate": 1.8779005512801212e-06,
+      "loss": 0.3149,
+      "step": 10403
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.7501016664337454,
+      "learning_rate": 1.8770212357213075e-06,
+      "loss": 0.3202,
+      "step": 10404
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 19.70796387301939,
+      "learning_rate": 1.8761420785043877e-06,
+      "loss": 0.3072,
+      "step": 10405
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.2209157747021835,
+      "learning_rate": 1.875263079673939e-06,
+      "loss": 0.3401,
+      "step": 10406
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.025272824914547,
+      "learning_rate": 1.8743842392745275e-06,
+      "loss": 0.3388,
+      "step": 10407
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.250492099139742,
+      "learning_rate": 1.873505557350715e-06,
+      "loss": 0.299,
+      "step": 10408
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.363929571042551,
+      "learning_rate": 1.872627033947047e-06,
+      "loss": 0.3318,
+      "step": 10409
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.540892001016186,
+      "learning_rate": 1.8717486691080738e-06,
+      "loss": 0.3319,
+      "step": 10410
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.027404765242512,
+      "learning_rate": 1.8708704628783236e-06,
+      "loss": 0.3107,
+      "step": 10411
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 13.072441198282347,
+      "learning_rate": 1.8699924153023296e-06,
+      "loss": 0.3156,
+      "step": 10412
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.061827915490803,
+      "learning_rate": 1.8691145264246064e-06,
+      "loss": 0.2882,
+      "step": 10413
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.846154038658359,
+      "learning_rate": 1.8682367962896658e-06,
+      "loss": 0.2785,
+      "step": 10414
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 10.493482719064728,
+      "learning_rate": 1.8673592249420114e-06,
+      "loss": 0.3215,
+      "step": 10415
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.289552524205187,
+      "learning_rate": 1.8664818124261375e-06,
+      "loss": 0.3023,
+      "step": 10416
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 6.483988157416612,
+      "learning_rate": 1.865604558786533e-06,
+      "loss": 0.3482,
+      "step": 10417
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.792169672700622,
+      "learning_rate": 1.8647274640676726e-06,
+      "loss": 0.339,
+      "step": 10418
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.200520141161151,
+      "learning_rate": 1.8638505283140296e-06,
+      "loss": 0.3174,
+      "step": 10419
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.8681630177990383,
+      "learning_rate": 1.8629737515700652e-06,
+      "loss": 0.3471,
+      "step": 10420
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.971052479381477,
+      "learning_rate": 1.8620971338802345e-06,
+      "loss": 0.3385,
+      "step": 10421
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3428175209323823,
+      "learning_rate": 1.861220675288985e-06,
+      "loss": 0.3389,
+      "step": 10422
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.175347798724783,
+      "learning_rate": 1.8603443758407553e-06,
+      "loss": 0.3266,
+      "step": 10423
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.28958746653995077,
+      "learning_rate": 1.8594682355799709e-06,
+      "loss": 0.1042,
+      "step": 10424
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.904097463927608,
+      "learning_rate": 1.8585922545510616e-06,
+      "loss": 0.2807,
+      "step": 10425
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.3617447113322396,
+      "learning_rate": 1.8577164327984343e-06,
+      "loss": 0.3049,
+      "step": 10426
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 4.257055846161416,
+      "learning_rate": 1.8568407703665015e-06,
+      "loss": 0.3107,
+      "step": 10427
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.4980449722040112,
+      "learning_rate": 1.8559652672996563e-06,
+      "loss": 0.3394,
+      "step": 10428
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.35771508992013,
+      "learning_rate": 1.8550899236422915e-06,
+      "loss": 0.3007,
+      "step": 10429
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4090375276441462,
+      "learning_rate": 1.8542147394387889e-06,
+      "loss": 0.3251,
+      "step": 10430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.039015093246316,
+      "learning_rate": 1.853339714733518e-06,
+      "loss": 0.3164,
+      "step": 10431
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.7008909127419605,
+      "learning_rate": 1.8524648495708514e-06,
+      "loss": 0.3095,
+      "step": 10432
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.199192352915818,
+      "learning_rate": 1.8515901439951412e-06,
+      "loss": 0.368,
+      "step": 10433
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.601333014232937,
+      "learning_rate": 1.8507155980507385e-06,
+      "loss": 0.3543,
+      "step": 10434
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7063855033564894,
+      "learning_rate": 1.8498412117819852e-06,
+      "loss": 0.3052,
+      "step": 10435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 14.714418734196714,
+      "learning_rate": 1.8489669852332148e-06,
+      "loss": 0.3085,
+      "step": 10436
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.686399693140075,
+      "learning_rate": 1.8480929184487518e-06,
+      "loss": 0.3135,
+      "step": 10437
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.3541364382146925,
+      "learning_rate": 1.8472190114729155e-06,
+      "loss": 0.3187,
+      "step": 10438
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.3407898532550195,
+      "learning_rate": 1.8463452643500096e-06,
+      "loss": 0.3142,
+      "step": 10439
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.495725002476835,
+      "learning_rate": 1.8454716771243415e-06,
+      "loss": 0.3001,
+      "step": 10440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.9345523418080415,
+      "learning_rate": 1.8445982498401993e-06,
+      "loss": 0.329,
+      "step": 10441
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.004191833709295,
+      "learning_rate": 1.8437249825418696e-06,
+      "loss": 0.2818,
+      "step": 10442
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.373375831440158,
+      "learning_rate": 1.8428518752736302e-06,
+      "loss": 0.2834,
+      "step": 10443
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.136059830498882,
+      "learning_rate": 1.841978928079745e-06,
+      "loss": 0.3568,
+      "step": 10444
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.6532921223113295,
+      "learning_rate": 1.8411061410044806e-06,
+      "loss": 0.2993,
+      "step": 10445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5587253004219397,
+      "learning_rate": 1.8402335140920841e-06,
+      "loss": 0.3397,
+      "step": 10446
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.993504980810485,
+      "learning_rate": 1.8393610473868018e-06,
+      "loss": 0.3165,
+      "step": 10447
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.261397665332212,
+      "learning_rate": 1.8384887409328688e-06,
+      "loss": 0.2894,
+      "step": 10448
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.856079966284449,
+      "learning_rate": 1.8376165947745138e-06,
+      "loss": 0.3077,
+      "step": 10449
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.618279356911632,
+      "learning_rate": 1.8367446089559565e-06,
+      "loss": 0.3047,
+      "step": 10450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.1164811750938686,
+      "learning_rate": 1.8358727835214101e-06,
+      "loss": 0.2948,
+      "step": 10451
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7890258430774386,
+      "learning_rate": 1.835001118515073e-06,
+      "loss": 0.3064,
+      "step": 10452
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.901314402586828,
+      "learning_rate": 1.8341296139811465e-06,
+      "loss": 0.3296,
+      "step": 10453
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.930357903121783,
+      "learning_rate": 1.833258269963814e-06,
+      "loss": 0.2837,
+      "step": 10454
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.2878736743047505,
+      "learning_rate": 1.8323870865072552e-06,
+      "loss": 0.1161,
+      "step": 10455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.684155746926985,
+      "learning_rate": 1.8315160636556434e-06,
+      "loss": 0.2989,
+      "step": 10456
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.714665042996784,
+      "learning_rate": 1.8306452014531362e-06,
+      "loss": 0.3369,
+      "step": 10457
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.195362608782338,
+      "learning_rate": 1.829774499943895e-06,
+      "loss": 0.3051,
+      "step": 10458
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.11503157231769,
+      "learning_rate": 1.8289039591720587e-06,
+      "loss": 0.3342,
+      "step": 10459
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.385074284872863,
+      "learning_rate": 1.8280335791817733e-06,
+      "loss": 0.2837,
+      "step": 10460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 21.565075335293407,
+      "learning_rate": 1.827163360017164e-06,
+      "loss": 0.3025,
+      "step": 10461
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.035973711907385,
+      "learning_rate": 1.8262933017223539e-06,
+      "loss": 0.3151,
+      "step": 10462
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.407465214546509,
+      "learning_rate": 1.825423404341457e-06,
+      "loss": 0.297,
+      "step": 10463
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.553584285869205,
+      "learning_rate": 1.8245536679185793e-06,
+      "loss": 0.3179,
+      "step": 10464
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.182148051755198,
+      "learning_rate": 1.823684092497819e-06,
+      "loss": 0.3346,
+      "step": 10465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.011201725513136,
+      "learning_rate": 1.8228146781232653e-06,
+      "loss": 0.3143,
+      "step": 10466
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.189868516249514,
+      "learning_rate": 1.8219454248389979e-06,
+      "loss": 0.2904,
+      "step": 10467
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.954206184369342,
+      "learning_rate": 1.82107633268909e-06,
+      "loss": 0.29,
+      "step": 10468
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4076109806652757,
+      "learning_rate": 1.820207401717608e-06,
+      "loss": 0.2949,
+      "step": 10469
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.025186995116017,
+      "learning_rate": 1.8193386319686073e-06,
+      "loss": 0.3132,
+      "step": 10470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.664099756389299,
+      "learning_rate": 1.8184700234861391e-06,
+      "loss": 0.2956,
+      "step": 10471
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.364636771439738,
+      "learning_rate": 1.8176015763142386e-06,
+      "loss": 0.3196,
+      "step": 10472
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4662663241800153,
+      "learning_rate": 1.8167332904969436e-06,
+      "loss": 0.29,
+      "step": 10473
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.4695065450584766,
+      "learning_rate": 1.8158651660782727e-06,
+      "loss": 0.3387,
+      "step": 10474
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.988913039469973,
+      "learning_rate": 1.8149972031022484e-06,
+      "loss": 0.3155,
+      "step": 10475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.390157486140574,
+      "learning_rate": 1.814129401612872e-06,
+      "loss": 0.2998,
+      "step": 10476
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6972623075152797,
+      "learning_rate": 1.813261761654146e-06,
+      "loss": 0.276,
+      "step": 10477
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.575495013339372,
+      "learning_rate": 1.812394283270061e-06,
+      "loss": 0.3083,
+      "step": 10478
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9065230929655224,
+      "learning_rate": 1.8115269665046014e-06,
+      "loss": 0.3354,
+      "step": 10479
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7623582151211843,
+      "learning_rate": 1.8106598114017398e-06,
+      "loss": 0.3222,
+      "step": 10480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.8440252448545005,
+      "learning_rate": 1.8097928180054437e-06,
+      "loss": 0.3061,
+      "step": 10481
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9336855713185255,
+      "learning_rate": 1.8089259863596713e-06,
+      "loss": 0.3163,
+      "step": 10482
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.562746287236664,
+      "learning_rate": 1.8080593165083743e-06,
+      "loss": 0.333,
+      "step": 10483
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.9341423081087576,
+      "learning_rate": 1.8071928084954948e-06,
+      "loss": 0.3108,
+      "step": 10484
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.454953802424039,
+      "learning_rate": 1.8063264623649628e-06,
+      "loss": 0.3263,
+      "step": 10485
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.553903947369211,
+      "learning_rate": 1.80546027816071e-06,
+      "loss": 0.2888,
+      "step": 10486
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7218886933295,
+      "learning_rate": 1.8045942559266471e-06,
+      "loss": 0.3505,
+      "step": 10487
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.791170561417732,
+      "learning_rate": 1.803728395706691e-06,
+      "loss": 0.3021,
+      "step": 10488
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.4646359815674415,
+      "learning_rate": 1.802862697544736e-06,
+      "loss": 0.3202,
+      "step": 10489
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.245293956254104,
+      "learning_rate": 1.8019971614846782e-06,
+      "loss": 0.319,
+      "step": 10490
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8320401678342435,
+      "learning_rate": 1.8011317875704015e-06,
+      "loss": 0.3203,
+      "step": 10491
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 15.692838898569537,
+      "learning_rate": 1.8002665758457822e-06,
+      "loss": 0.2978,
+      "step": 10492
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.122394277371904,
+      "learning_rate": 1.7994015263546905e-06,
+      "loss": 0.2978,
+      "step": 10493
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.308469788422025,
+      "learning_rate": 1.798536639140982e-06,
+      "loss": 0.3294,
+      "step": 10494
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.742028245295386,
+      "learning_rate": 1.7976719142485111e-06,
+      "loss": 0.3145,
+      "step": 10495
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.917281451057853,
+      "learning_rate": 1.796807351721121e-06,
+      "loss": 0.3132,
+      "step": 10496
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3217503723997001,
+      "learning_rate": 1.7959429516026467e-06,
+      "loss": 0.1142,
+      "step": 10497
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.765008425354903,
+      "learning_rate": 1.7950787139369152e-06,
+      "loss": 0.3453,
+      "step": 10498
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5640471743810127,
+      "learning_rate": 1.7942146387677473e-06,
+      "loss": 0.315,
+      "step": 10499
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.074557972090344,
+      "learning_rate": 1.7933507261389477e-06,
+      "loss": 0.3146,
+      "step": 10500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.796544715238582,
+      "learning_rate": 1.7924869760943264e-06,
+      "loss": 0.3458,
+      "step": 10501
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.164383622164497,
+      "learning_rate": 1.7916233886776717e-06,
+      "loss": 0.2976,
+      "step": 10502
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.289981732434985,
+      "learning_rate": 1.790759963932771e-06,
+      "loss": 0.3078,
+      "step": 10503
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.409208792194172,
+      "learning_rate": 1.7898967019034026e-06,
+      "loss": 0.3518,
+      "step": 10504
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.643324202289811,
+      "learning_rate": 1.7890336026333349e-06,
+      "loss": 0.3399,
+      "step": 10505
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9467203403328424,
+      "learning_rate": 1.7881706661663312e-06,
+      "loss": 0.3311,
+      "step": 10506
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.6757324027410667,
+      "learning_rate": 1.7873078925461385e-06,
+      "loss": 0.3542,
+      "step": 10507
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 11.507155326392486,
+      "learning_rate": 1.786445281816509e-06,
+      "loss": 0.3463,
+      "step": 10508
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.8688614762850775,
+      "learning_rate": 1.7855828340211734e-06,
+      "loss": 0.3272,
+      "step": 10509
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.5436114152845475,
+      "learning_rate": 1.7847205492038617e-06,
+      "loss": 0.3263,
+      "step": 10510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.57392926032479,
+      "learning_rate": 1.7838584274082932e-06,
+      "loss": 0.3241,
+      "step": 10511
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3057430757612166,
+      "learning_rate": 1.7829964686781793e-06,
+      "loss": 0.1121,
+      "step": 10512
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7657845859723382,
+      "learning_rate": 1.7821346730572236e-06,
+      "loss": 0.3026,
+      "step": 10513
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.831547819832643,
+      "learning_rate": 1.7812730405891227e-06,
+      "loss": 0.3129,
+      "step": 10514
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.734308720253229,
+      "learning_rate": 1.780411571317558e-06,
+      "loss": 0.3108,
+      "step": 10515
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.5572977403410224,
+      "learning_rate": 1.7795502652862152e-06,
+      "loss": 0.348,
+      "step": 10516
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.7911765001891355,
+      "learning_rate": 1.7786891225387582e-06,
+      "loss": 0.2735,
+      "step": 10517
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.007976907325197,
+      "learning_rate": 1.7778281431188515e-06,
+      "loss": 0.3156,
+      "step": 10518
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.308024535624497,
+      "learning_rate": 1.7769673270701498e-06,
+      "loss": 0.2961,
+      "step": 10519
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.065072106191642,
+      "learning_rate": 1.7761066744362936e-06,
+      "loss": 0.3343,
+      "step": 10520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 9.132230698195603,
+      "learning_rate": 1.7752461852609271e-06,
+      "loss": 0.2968,
+      "step": 10521
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.581363606949313,
+      "learning_rate": 1.774385859587673e-06,
+      "loss": 0.3396,
+      "step": 10522
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.288753441657392,
+      "learning_rate": 1.7735256974601538e-06,
+      "loss": 0.3224,
+      "step": 10523
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.2413021122261965,
+      "learning_rate": 1.7726656989219815e-06,
+      "loss": 0.3042,
+      "step": 10524
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.88373933049799,
+      "learning_rate": 1.77180586401676e-06,
+      "loss": 0.3059,
+      "step": 10525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.857557130003466,
+      "learning_rate": 1.7709461927880855e-06,
+      "loss": 0.3085,
+      "step": 10526
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.4946761086900375,
+      "learning_rate": 1.7700866852795457e-06,
+      "loss": 0.3333,
+      "step": 10527
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.216743727366197,
+      "learning_rate": 1.769227341534715e-06,
+      "loss": 0.3271,
+      "step": 10528
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.641630394153822,
+      "learning_rate": 1.768368161597171e-06,
+      "loss": 0.2807,
+      "step": 10529
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.705913031626951,
+      "learning_rate": 1.7675091455104704e-06,
+      "loss": 0.3226,
+      "step": 10530
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.79997199476943,
+      "learning_rate": 1.7666502933181695e-06,
+      "loss": 0.316,
+      "step": 10531
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.9728318864741667,
+      "learning_rate": 1.7657916050638157e-06,
+      "loss": 0.3251,
+      "step": 10532
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.116741350996393,
+      "learning_rate": 1.7649330807909405e-06,
+      "loss": 0.3559,
+      "step": 10533
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.257089988755028,
+      "learning_rate": 1.7640747205430809e-06,
+      "loss": 0.2904,
+      "step": 10534
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.938560639200215,
+      "learning_rate": 1.7632165243637495e-06,
+      "loss": 0.3231,
+      "step": 10535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.0382651614493885,
+      "learning_rate": 1.7623584922964665e-06,
+      "loss": 0.3126,
+      "step": 10536
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.396580062735448,
+      "learning_rate": 1.761500624384731e-06,
+      "loss": 0.2996,
+      "step": 10537
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.405749828439021,
+      "learning_rate": 1.76064292067204e-06,
+      "loss": 0.3145,
+      "step": 10538
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.027494362115121,
+      "learning_rate": 1.759785381201881e-06,
+      "loss": 0.3041,
+      "step": 10539
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.709784302731968,
+      "learning_rate": 1.7589280060177332e-06,
+      "loss": 0.2983,
+      "step": 10540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.31551064288102526,
+      "learning_rate": 1.7580707951630682e-06,
+      "loss": 0.1098,
+      "step": 10541
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.6119920170280615,
+      "learning_rate": 1.757213748681349e-06,
+      "loss": 0.3065,
+      "step": 10542
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.052858752575616,
+      "learning_rate": 1.7563568666160264e-06,
+      "loss": 0.3168,
+      "step": 10543
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.287887578974892,
+      "learning_rate": 1.755500149010549e-06,
+      "loss": 0.3191,
+      "step": 10544
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.112074394094158,
+      "learning_rate": 1.754643595908353e-06,
+      "loss": 0.3072,
+      "step": 10545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.721513882503015,
+      "learning_rate": 1.7537872073528683e-06,
+      "loss": 0.2812,
+      "step": 10546
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.7028800914132605,
+      "learning_rate": 1.7529309833875174e-06,
+      "loss": 0.3335,
+      "step": 10547
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.337350438313519,
+      "learning_rate": 1.7520749240557073e-06,
+      "loss": 0.3238,
+      "step": 10548
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 13.34125412225827,
+      "learning_rate": 1.7512190294008492e-06,
+      "loss": 0.3432,
+      "step": 10549
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.670307699010747,
+      "learning_rate": 1.7503632994663316e-06,
+      "loss": 0.2825,
+      "step": 10550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.513354111373852,
+      "learning_rate": 1.749507734295549e-06,
+      "loss": 0.3258,
+      "step": 10551
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.233307883769882,
+      "learning_rate": 1.7486523339318757e-06,
+      "loss": 0.3182,
+      "step": 10552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.550621295731476,
+      "learning_rate": 1.7477970984186832e-06,
+      "loss": 0.3222,
+      "step": 10553
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.3718513822325296,
+      "learning_rate": 1.7469420277993342e-06,
+      "loss": 0.3269,
+      "step": 10554
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.036150537572388,
+      "learning_rate": 1.7460871221171849e-06,
+      "loss": 0.2883,
+      "step": 10555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 5.5639602399711485,
+      "learning_rate": 1.7452323814155769e-06,
+      "loss": 0.3034,
+      "step": 10556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.462989751228814,
+      "learning_rate": 1.7443778057378492e-06,
+      "loss": 0.3139,
+      "step": 10557
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 57.35426744750466,
+      "learning_rate": 1.7435233951273306e-06,
+      "loss": 0.3261,
+      "step": 10558
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.590532978683563,
+      "learning_rate": 1.7426691496273423e-06,
+      "loss": 0.3574,
+      "step": 10559
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.653298255023084,
+      "learning_rate": 1.7418150692811969e-06,
+      "loss": 0.302,
+      "step": 10560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.13980099120181,
+      "learning_rate": 1.7409611541321942e-06,
+      "loss": 0.3569,
+      "step": 10561
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.869578047145659,
+      "learning_rate": 1.740107404223636e-06,
+      "loss": 0.3129,
+      "step": 10562
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.451288737867251,
+      "learning_rate": 1.7392538195988024e-06,
+      "loss": 0.2821,
+      "step": 10563
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.660211163965008,
+      "learning_rate": 1.7384004003009785e-06,
+      "loss": 0.3367,
+      "step": 10564
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.337800875072026,
+      "learning_rate": 1.73754714637343e-06,
+      "loss": 0.3073,
+      "step": 10565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 10.492351351485592,
+      "learning_rate": 1.73669405785942e-06,
+      "loss": 0.325,
+      "step": 10566
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.02717761817882,
+      "learning_rate": 1.7358411348022026e-06,
+      "loss": 0.3332,
+      "step": 10567
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 8.016190158188072,
+      "learning_rate": 1.734988377245022e-06,
+      "loss": 0.2913,
+      "step": 10568
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.58375395186292,
+      "learning_rate": 1.7341357852311175e-06,
+      "loss": 0.3098,
+      "step": 10569
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 6.204593356402102,
+      "learning_rate": 1.7332833588037135e-06,
+      "loss": 0.3358,
+      "step": 10570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 7.251783523966844,
+      "learning_rate": 1.732431098006031e-06,
+      "loss": 0.3241,
+      "step": 10571
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.450248564464547,
+      "learning_rate": 1.7315790028812829e-06,
+      "loss": 0.3528,
+      "step": 10572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.724055428190715,
+      "learning_rate": 1.7307270734726711e-06,
+      "loss": 0.2963,
+      "step": 10573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.243368570367459,
+      "learning_rate": 1.7298753098233906e-06,
+      "loss": 0.3006,
+      "step": 10574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.08054576012365,
+      "learning_rate": 1.72902371197663e-06,
+      "loss": 0.3003,
+      "step": 10575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.559507275317814,
+      "learning_rate": 1.7281722799755613e-06,
+      "loss": 0.3368,
+      "step": 10576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4635464240607208,
+      "learning_rate": 1.727321013863361e-06,
+      "loss": 0.3427,
+      "step": 10577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7780076730807246,
+      "learning_rate": 1.7264699136831854e-06,
+      "loss": 0.2777,
+      "step": 10578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.087852313371457,
+      "learning_rate": 1.7256189794781886e-06,
+      "loss": 0.299,
+      "step": 10579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4313926411481668,
+      "learning_rate": 1.7247682112915148e-06,
+      "loss": 0.3264,
+      "step": 10580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.972524105441143,
+      "learning_rate": 1.7239176091662996e-06,
+      "loss": 0.3177,
+      "step": 10581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.810445889855938,
+      "learning_rate": 1.7230671731456722e-06,
+      "loss": 0.3238,
+      "step": 10582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.272667205908752,
+      "learning_rate": 1.722216903272747e-06,
+      "loss": 0.295,
+      "step": 10583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.907399691257148,
+      "learning_rate": 1.7213667995906413e-06,
+      "loss": 0.3176,
+      "step": 10584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.191737431353304,
+      "learning_rate": 1.720516862142451e-06,
+      "loss": 0.3285,
+      "step": 10585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.895954019433363,
+      "learning_rate": 1.719667090971273e-06,
+      "loss": 0.3373,
+      "step": 10586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.6025364764010717,
+      "learning_rate": 1.7188174861201917e-06,
+      "loss": 0.2899,
+      "step": 10587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.196943381180313,
+      "learning_rate": 1.7179680476322857e-06,
+      "loss": 0.3469,
+      "step": 10588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.602659422410397,
+      "learning_rate": 1.7171187755506185e-06,
+      "loss": 0.3088,
+      "step": 10589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.634896251903001,
+      "learning_rate": 1.7162696699182567e-06,
+      "loss": 0.3077,
+      "step": 10590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1568431643448704,
+      "learning_rate": 1.7154207307782466e-06,
+      "loss": 0.2986,
+      "step": 10591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.298211374417399,
+      "learning_rate": 1.7145719581736337e-06,
+      "loss": 0.3062,
+      "step": 10592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.124113011632788,
+      "learning_rate": 1.7137233521474517e-06,
+      "loss": 0.3179,
+      "step": 10593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.513977043221937,
+      "learning_rate": 1.7128749127427274e-06,
+      "loss": 0.323,
+      "step": 10594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.142475767494577,
+      "learning_rate": 1.7120266400024798e-06,
+      "loss": 0.266,
+      "step": 10595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1318208060645474,
+      "learning_rate": 1.7111785339697134e-06,
+      "loss": 0.2795,
+      "step": 10596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.424643951716954,
+      "learning_rate": 1.710330594687436e-06,
+      "loss": 0.308,
+      "step": 10597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.504420183874065,
+      "learning_rate": 1.7094828221986348e-06,
+      "loss": 0.3087,
+      "step": 10598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.257048199732994,
+      "learning_rate": 1.7086352165462948e-06,
+      "loss": 0.3435,
+      "step": 10599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.816877412676108,
+      "learning_rate": 1.7077877777733925e-06,
+      "loss": 0.34,
+      "step": 10600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.891023620734343,
+      "learning_rate": 1.7069405059228945e-06,
+      "loss": 0.3209,
+      "step": 10601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.691518651814136,
+      "learning_rate": 1.706093401037759e-06,
+      "loss": 0.3473,
+      "step": 10602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.548159998221023,
+      "learning_rate": 1.7052464631609378e-06,
+      "loss": 0.2942,
+      "step": 10603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.104571541936625,
+      "learning_rate": 1.7043996923353677e-06,
+      "loss": 0.285,
+      "step": 10604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.947796197046866,
+      "learning_rate": 1.703553088603989e-06,
+      "loss": 0.288,
+      "step": 10605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8652186000352216,
+      "learning_rate": 1.7027066520097208e-06,
+      "loss": 0.305,
+      "step": 10606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2992531918799873,
+      "learning_rate": 1.7018603825954805e-06,
+      "loss": 0.1142,
+      "step": 10607
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.273749711453987,
+      "learning_rate": 1.7010142804041785e-06,
+      "loss": 0.345,
+      "step": 10608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.334526246806545,
+      "learning_rate": 1.7001683454787082e-06,
+      "loss": 0.3089,
+      "step": 10609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.021055964580396,
+      "learning_rate": 1.6993225778619677e-06,
+      "loss": 0.3403,
+      "step": 10610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.096354495365,
+      "learning_rate": 1.6984769775968325e-06,
+      "loss": 0.3277,
+      "step": 10611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2819455154125181,
+      "learning_rate": 1.6976315447261827e-06,
+      "loss": 0.1198,
+      "step": 10612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 13.707525063062743,
+      "learning_rate": 1.6967862792928785e-06,
+      "loss": 0.3406,
+      "step": 10613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4053497029974125,
+      "learning_rate": 1.6959411813397791e-06,
+      "loss": 0.3134,
+      "step": 10614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.947278046228455,
+      "learning_rate": 1.6950962509097325e-06,
+      "loss": 0.2966,
+      "step": 10615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.04840554356012,
+      "learning_rate": 1.694251488045578e-06,
+      "loss": 0.3176,
+      "step": 10616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 8.024412988768601,
+      "learning_rate": 1.6934068927901477e-06,
+      "loss": 0.2936,
+      "step": 10617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1885727285599423,
+      "learning_rate": 1.6925624651862655e-06,
+      "loss": 0.3135,
+      "step": 10618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.101967346664549,
+      "learning_rate": 1.6917182052767427e-06,
+      "loss": 0.3086,
+      "step": 10619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.799345911695074,
+      "learning_rate": 1.690874113104387e-06,
+      "loss": 0.3104,
+      "step": 10620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.273632388348232,
+      "learning_rate": 1.6900301887119957e-06,
+      "loss": 0.3237,
+      "step": 10621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.746239679979197,
+      "learning_rate": 1.6891864321423574e-06,
+      "loss": 0.3122,
+      "step": 10622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.31203309408454794,
+      "learning_rate": 1.6883428434382537e-06,
+      "loss": 0.1126,
+      "step": 10623
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.970181501724367,
+      "learning_rate": 1.6874994226424518e-06,
+      "loss": 0.3324,
+      "step": 10624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.681891446878625,
+      "learning_rate": 1.6866561697977219e-06,
+      "loss": 0.3312,
+      "step": 10625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.267055493195171,
+      "learning_rate": 1.6858130849468135e-06,
+      "loss": 0.2962,
+      "step": 10626
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.14866721788529,
+      "learning_rate": 1.6849701681324748e-06,
+      "loss": 0.3192,
+      "step": 10627
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.3856976072399725,
+      "learning_rate": 1.6841274193974433e-06,
+      "loss": 0.3381,
+      "step": 10628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.756815894238762,
+      "learning_rate": 1.6832848387844481e-06,
+      "loss": 0.3675,
+      "step": 10629
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.985472092346321,
+      "learning_rate": 1.68244242633621e-06,
+      "loss": 0.328,
+      "step": 10630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.661783121771224,
+      "learning_rate": 1.6816001820954426e-06,
+      "loss": 0.3061,
+      "step": 10631
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.183453278219762,
+      "learning_rate": 1.6807581061048467e-06,
+      "loss": 0.3176,
+      "step": 10632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.706297116206154,
+      "learning_rate": 1.6799161984071189e-06,
+      "loss": 0.3274,
+      "step": 10633
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.014645433786052,
+      "learning_rate": 1.6790744590449454e-06,
+      "loss": 0.3168,
+      "step": 10634
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.9685839051764473,
+      "learning_rate": 1.678232888061005e-06,
+      "loss": 0.3492,
+      "step": 10635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.456787860265964,
+      "learning_rate": 1.6773914854979683e-06,
+      "loss": 0.3094,
+      "step": 10636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.035725809497279,
+      "learning_rate": 1.6765502513984911e-06,
+      "loss": 0.3138,
+      "step": 10637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5893062462673617,
+      "learning_rate": 1.6757091858052332e-06,
+      "loss": 0.3179,
+      "step": 10638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.991592701868656,
+      "learning_rate": 1.6748682887608308e-06,
+      "loss": 0.2896,
+      "step": 10639
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.334135532314408,
+      "learning_rate": 1.674027560307927e-06,
+      "loss": 0.2959,
+      "step": 10640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1178656248623873,
+      "learning_rate": 1.6731870004891427e-06,
+      "loss": 0.3023,
+      "step": 10641
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.5566049606714656,
+      "learning_rate": 1.672346609347098e-06,
+      "loss": 0.2898,
+      "step": 10642
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.249357923004523,
+      "learning_rate": 1.671506386924403e-06,
+      "loss": 0.2921,
+      "step": 10643
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7291450963887898,
+      "learning_rate": 1.6706663332636586e-06,
+      "loss": 0.355,
+      "step": 10644
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.0112288811453025,
+      "learning_rate": 1.669826448407459e-06,
+      "loss": 0.3126,
+      "step": 10645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.494114542020498,
+      "learning_rate": 1.6689867323983843e-06,
+      "loss": 0.3379,
+      "step": 10646
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5601079590915368,
+      "learning_rate": 1.6681471852790126e-06,
+      "loss": 0.3144,
+      "step": 10647
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.706103944496336,
+      "learning_rate": 1.6673078070919109e-06,
+      "loss": 0.2939,
+      "step": 10648
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.7694735670876565,
+      "learning_rate": 1.6664685978796363e-06,
+      "loss": 0.3104,
+      "step": 10649
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.023074367633568,
+      "learning_rate": 1.6656295576847393e-06,
+      "loss": 0.3074,
+      "step": 10650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.220959333276316,
+      "learning_rate": 1.6647906865497625e-06,
+      "loss": 0.3327,
+      "step": 10651
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.773814301822905,
+      "learning_rate": 1.663951984517234e-06,
+      "loss": 0.2989,
+      "step": 10652
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4934204254060552,
+      "learning_rate": 1.663113451629685e-06,
+      "loss": 0.3014,
+      "step": 10653
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.329083567932972,
+      "learning_rate": 1.662275087929624e-06,
+      "loss": 0.2989,
+      "step": 10654
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.767309736448474,
+      "learning_rate": 1.6614368934595609e-06,
+      "loss": 0.3251,
+      "step": 10655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4487094763402983,
+      "learning_rate": 1.6605988682619944e-06,
+      "loss": 0.2873,
+      "step": 10656
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.665708721550492,
+      "learning_rate": 1.6597610123794134e-06,
+      "loss": 0.3334,
+      "step": 10657
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.547339856087588,
+      "learning_rate": 1.6589233258543008e-06,
+      "loss": 0.2674,
+      "step": 10658
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.9025953809583775,
+      "learning_rate": 1.658085808729124e-06,
+      "loss": 0.3117,
+      "step": 10659
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 28.334674957130957,
+      "learning_rate": 1.6572484610463546e-06,
+      "loss": 0.2941,
+      "step": 10660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.798099411567872,
+      "learning_rate": 1.656411282848442e-06,
+      "loss": 0.2932,
+      "step": 10661
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.52793004541175,
+      "learning_rate": 1.6555742741778347e-06,
+      "loss": 0.3159,
+      "step": 10662
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.891717332950957,
+      "learning_rate": 1.6547374350769713e-06,
+      "loss": 0.2917,
+      "step": 10663
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.087241149337934,
+      "learning_rate": 1.6539007655882833e-06,
+      "loss": 0.2838,
+      "step": 10664
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.663955489933615,
+      "learning_rate": 1.6530642657541856e-06,
+      "loss": 0.3197,
+      "step": 10665
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.570631176144225,
+      "learning_rate": 1.6522279356170983e-06,
+      "loss": 0.3331,
+      "step": 10666
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.9327611436547745,
+      "learning_rate": 1.6513917752194198e-06,
+      "loss": 0.3321,
+      "step": 10667
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.6340460671566226,
+      "learning_rate": 1.6505557846035464e-06,
+      "loss": 0.3145,
+      "step": 10668
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 21.22504583557889,
+      "learning_rate": 1.6497199638118656e-06,
+      "loss": 0.2611,
+      "step": 10669
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.8568357696191224,
+      "learning_rate": 1.6488843128867548e-06,
+      "loss": 0.3092,
+      "step": 10670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.424220262775636,
+      "learning_rate": 1.6480488318705857e-06,
+      "loss": 0.3091,
+      "step": 10671
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1480801183983003,
+      "learning_rate": 1.6472135208057128e-06,
+      "loss": 0.2957,
+      "step": 10672
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5916162235478906,
+      "learning_rate": 1.6463783797344962e-06,
+      "loss": 0.2921,
+      "step": 10673
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.581155272907804,
+      "learning_rate": 1.6455434086992732e-06,
+      "loss": 0.3142,
+      "step": 10674
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.248758340734047,
+      "learning_rate": 1.6447086077423806e-06,
+      "loss": 0.2709,
+      "step": 10675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.0906241450205836,
+      "learning_rate": 1.6438739769061458e-06,
+      "loss": 0.2971,
+      "step": 10676
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.195761996956426,
+      "learning_rate": 1.643039516232885e-06,
+      "loss": 0.3086,
+      "step": 10677
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 11.442667691903877,
+      "learning_rate": 1.642205225764908e-06,
+      "loss": 0.295,
+      "step": 10678
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.617910454566712,
+      "learning_rate": 1.6413711055445164e-06,
+      "loss": 0.3202,
+      "step": 10679
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.370899824644687,
+      "learning_rate": 1.640537155613997e-06,
+      "loss": 0.3323,
+      "step": 10680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.035704166383665,
+      "learning_rate": 1.6397033760156395e-06,
+      "loss": 0.3152,
+      "step": 10681
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.7386786596500534,
+      "learning_rate": 1.6388697667917142e-06,
+      "loss": 0.3028,
+      "step": 10682
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.122653934672854,
+      "learning_rate": 1.6380363279844873e-06,
+      "loss": 0.3277,
+      "step": 10683
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.2352888403666595,
+      "learning_rate": 1.6372030596362183e-06,
+      "loss": 0.3274,
+      "step": 10684
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2057120370220935,
+      "learning_rate": 1.636369961789151e-06,
+      "loss": 0.2934,
+      "step": 10685
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.916639553085376,
+      "learning_rate": 1.6355370344855315e-06,
+      "loss": 0.3108,
+      "step": 10686
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.248261721681635,
+      "learning_rate": 1.6347042777675842e-06,
+      "loss": 0.309,
+      "step": 10687
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8069252173263264,
+      "learning_rate": 1.6338716916775394e-06,
+      "loss": 0.2759,
+      "step": 10688
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7314620147239945,
+      "learning_rate": 1.6330392762576048e-06,
+      "loss": 0.3144,
+      "step": 10689
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.098522446913228,
+      "learning_rate": 1.6322070315499882e-06,
+      "loss": 0.3236,
+      "step": 10690
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.69549930210126,
+      "learning_rate": 1.6313749575968858e-06,
+      "loss": 0.3149,
+      "step": 10691
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 12.57897346121537,
+      "learning_rate": 1.6305430544404855e-06,
+      "loss": 0.2979,
+      "step": 10692
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.28575552155953,
+      "learning_rate": 1.6297113221229672e-06,
+      "loss": 0.3248,
+      "step": 10693
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4013499555228917,
+      "learning_rate": 1.628879760686503e-06,
+      "loss": 0.2898,
+      "step": 10694
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.7855506357532223,
+      "learning_rate": 1.6280483701732513e-06,
+      "loss": 0.3023,
+      "step": 10695
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1460801715141393,
+      "learning_rate": 1.627217150625367e-06,
+      "loss": 0.2985,
+      "step": 10696
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 45.831497760319294,
+      "learning_rate": 1.626386102084997e-06,
+      "loss": 0.3521,
+      "step": 10697
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.5250824826493297,
+      "learning_rate": 1.6255552245942713e-06,
+      "loss": 0.2955,
+      "step": 10698
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 14.665669532019978,
+      "learning_rate": 1.6247245181953253e-06,
+      "loss": 0.2738,
+      "step": 10699
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.1197372599784052,
+      "learning_rate": 1.6238939829302696e-06,
+      "loss": 0.2985,
+      "step": 10700
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.8166501886261464,
+      "learning_rate": 1.6230636188412218e-06,
+      "loss": 0.3232,
+      "step": 10701
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.9851208066433195,
+      "learning_rate": 1.6222334259702776e-06,
+      "loss": 0.3341,
+      "step": 10702
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.791062804532311,
+      "learning_rate": 1.6214034043595317e-06,
+      "loss": 0.2808,
+      "step": 10703
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.463580377244825,
+      "learning_rate": 1.6205735540510676e-06,
+      "loss": 0.3187,
+      "step": 10704
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.43171725060221,
+      "learning_rate": 1.619743875086961e-06,
+      "loss": 0.3322,
+      "step": 10705
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.438575593546221,
+      "learning_rate": 1.6189143675092783e-06,
+      "loss": 0.3177,
+      "step": 10706
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.805825554618875,
+      "learning_rate": 1.6180850313600788e-06,
+      "loss": 0.2873,
+      "step": 10707
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 6.1363417279272,
+      "learning_rate": 1.617255866681408e-06,
+      "loss": 0.3116,
+      "step": 10708
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.2387338093284574,
+      "learning_rate": 1.6164268735153093e-06,
+      "loss": 0.2893,
+      "step": 10709
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.035252312943682,
+      "learning_rate": 1.6155980519038134e-06,
+      "loss": 0.3041,
+      "step": 10710
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.2759225046355426,
+      "learning_rate": 1.6147694018889431e-06,
+      "loss": 0.1055,
+      "step": 10711
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.244470513541704,
+      "learning_rate": 1.6139409235127156e-06,
+      "loss": 0.2741,
+      "step": 10712
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 7.191784141162951,
+      "learning_rate": 1.6131126168171307e-06,
+      "loss": 0.3384,
+      "step": 10713
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.477512766907342,
+      "learning_rate": 1.6122844818441924e-06,
+      "loss": 0.3044,
+      "step": 10714
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 4.359429907733366,
+      "learning_rate": 1.6114565186358826e-06,
+      "loss": 0.3222,
+      "step": 10715
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.019774499570603,
+      "learning_rate": 1.610628727234187e-06,
+      "loss": 0.3032,
+      "step": 10716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.465629659064195,
+      "learning_rate": 1.609801107681071e-06,
+      "loss": 0.3586,
+      "step": 10717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.867562432853363,
+      "learning_rate": 1.6089736600184995e-06,
+      "loss": 0.3043,
+      "step": 10718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.856071586575248,
+      "learning_rate": 1.6081463842884253e-06,
+      "loss": 0.3011,
+      "step": 10719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.693687765664374,
+      "learning_rate": 1.6073192805327936e-06,
+      "loss": 0.3191,
+      "step": 10720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.040583916310937,
+      "learning_rate": 1.6064923487935413e-06,
+      "loss": 0.2914,
+      "step": 10721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.124109608058877,
+      "learning_rate": 1.6056655891125922e-06,
+      "loss": 0.2861,
+      "step": 10722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.600389547016114,
+      "learning_rate": 1.6048390015318676e-06,
+      "loss": 0.3256,
+      "step": 10723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.254093265082026,
+      "learning_rate": 1.6040125860932765e-06,
+      "loss": 0.3048,
+      "step": 10724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8397413055988308,
+      "learning_rate": 1.6031863428387201e-06,
+      "loss": 0.3502,
+      "step": 10725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.966164612070627,
+      "learning_rate": 1.6023602718100912e-06,
+      "loss": 0.3182,
+      "step": 10726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.075438040345859,
+      "learning_rate": 1.6015343730492743e-06,
+      "loss": 0.2753,
+      "step": 10727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.68141790136336,
+      "learning_rate": 1.6007086465981397e-06,
+      "loss": 0.3258,
+      "step": 10728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.918693981611562,
+      "learning_rate": 1.5998830924985604e-06,
+      "loss": 0.2812,
+      "step": 10729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.867583283195368,
+      "learning_rate": 1.5990577107923887e-06,
+      "loss": 0.3314,
+      "step": 10730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.021451479525655,
+      "learning_rate": 1.598232501521475e-06,
+      "loss": 0.3382,
+      "step": 10731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.516546764319625,
+      "learning_rate": 1.5974074647276588e-06,
+      "loss": 0.3622,
+      "step": 10732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6575921679352,
+      "learning_rate": 1.5965826004527712e-06,
+      "loss": 0.3438,
+      "step": 10733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3589281190313667,
+      "learning_rate": 1.5957579087386371e-06,
+      "loss": 0.2983,
+      "step": 10734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.092448114836084,
+      "learning_rate": 1.5949333896270664e-06,
+      "loss": 0.2839,
+      "step": 10735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.146396267672737,
+      "learning_rate": 1.5941090431598654e-06,
+      "loss": 0.3252,
+      "step": 10736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.093479083256424,
+      "learning_rate": 1.5932848693788312e-06,
+      "loss": 0.3184,
+      "step": 10737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.671972681274842,
+      "learning_rate": 1.5924608683257504e-06,
+      "loss": 0.3469,
+      "step": 10738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.445493506290543,
+      "learning_rate": 1.5916370400424014e-06,
+      "loss": 0.3189,
+      "step": 10739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.213450652880184,
+      "learning_rate": 1.5908133845705565e-06,
+      "loss": 0.3413,
+      "step": 10740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6333186816115073,
+      "learning_rate": 1.5899899019519722e-06,
+      "loss": 0.3557,
+      "step": 10741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3005567109594841,
+      "learning_rate": 1.5891665922284061e-06,
+      "loss": 0.1081,
+      "step": 10742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2262350518261895,
+      "learning_rate": 1.5883434554415984e-06,
+      "loss": 0.3292,
+      "step": 10743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.409180984827956,
+      "learning_rate": 1.5875204916332843e-06,
+      "loss": 0.2964,
+      "step": 10744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.351120074589271,
+      "learning_rate": 1.5866977008451906e-06,
+      "loss": 0.2964,
+      "step": 10745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.436906046260896,
+      "learning_rate": 1.5858750831190345e-06,
+      "loss": 0.358,
+      "step": 10746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.545703766357559,
+      "learning_rate": 1.5850526384965259e-06,
+      "loss": 0.3383,
+      "step": 10747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9524410769634812,
+      "learning_rate": 1.5842303670193604e-06,
+      "loss": 0.3384,
+      "step": 10748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.096364523214711,
+      "learning_rate": 1.5834082687292347e-06,
+      "loss": 0.2997,
+      "step": 10749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0292060105294007,
+      "learning_rate": 1.5825863436678262e-06,
+      "loss": 0.3225,
+      "step": 10750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.084456165209167,
+      "learning_rate": 1.5817645918768098e-06,
+      "loss": 0.2841,
+      "step": 10751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.528644771705652,
+      "learning_rate": 1.580943013397851e-06,
+      "loss": 0.3048,
+      "step": 10752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8312931329536286,
+      "learning_rate": 1.580121608272605e-06,
+      "loss": 0.3429,
+      "step": 10753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5532573302592447,
+      "learning_rate": 1.579300376542719e-06,
+      "loss": 0.3183,
+      "step": 10754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.192822350046899,
+      "learning_rate": 1.5784793182498325e-06,
+      "loss": 0.3163,
+      "step": 10755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.810918455232482,
+      "learning_rate": 1.5776584334355721e-06,
+      "loss": 0.2847,
+      "step": 10756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4741529130306996,
+      "learning_rate": 1.5768377221415603e-06,
+      "loss": 0.3434,
+      "step": 10757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.4097161346832867,
+      "learning_rate": 1.5760171844094086e-06,
+      "loss": 0.3148,
+      "step": 10758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.821214125990964,
+      "learning_rate": 1.57519682028072e-06,
+      "loss": 0.2865,
+      "step": 10759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 15.958426291092723,
+      "learning_rate": 1.5743766297970903e-06,
+      "loss": 0.3065,
+      "step": 10760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.655562851328944,
+      "learning_rate": 1.5735566130001007e-06,
+      "loss": 0.3422,
+      "step": 10761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0448184760767156,
+      "learning_rate": 1.572736769931334e-06,
+      "loss": 0.2922,
+      "step": 10762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.319305294810199,
+      "learning_rate": 1.5719171006323513e-06,
+      "loss": 0.3127,
+      "step": 10763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.1023295590950415,
+      "learning_rate": 1.5710976051447185e-06,
+      "loss": 0.3096,
+      "step": 10764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1980659526955275,
+      "learning_rate": 1.5702782835099807e-06,
+      "loss": 0.3636,
+      "step": 10765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.559898533569583,
+      "learning_rate": 1.569459135769681e-06,
+      "loss": 0.2875,
+      "step": 10766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.021888739891441,
+      "learning_rate": 1.5686401619653524e-06,
+      "loss": 0.3188,
+      "step": 10767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.497911610178419,
+      "learning_rate": 1.567821362138518e-06,
+      "loss": 0.2939,
+      "step": 10768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.571860561521987,
+      "learning_rate": 1.5670027363306928e-06,
+      "loss": 0.3226,
+      "step": 10769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.02581718942726,
+      "learning_rate": 1.5661842845833851e-06,
+      "loss": 0.3304,
+      "step": 10770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.3717661674928,
+      "learning_rate": 1.5653660069380888e-06,
+      "loss": 0.3234,
+      "step": 10771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.433135846311785,
+      "learning_rate": 1.5645479034362938e-06,
+      "loss": 0.3226,
+      "step": 10772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.579909350965067,
+      "learning_rate": 1.5637299741194817e-06,
+      "loss": 0.2888,
+      "step": 10773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.699636934529329,
+      "learning_rate": 1.5629122190291179e-06,
+      "loss": 0.3016,
+      "step": 10774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.36382716014353,
+      "learning_rate": 1.562094638206671e-06,
+      "loss": 0.3202,
+      "step": 10775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.950648600589396,
+      "learning_rate": 1.5612772316935881e-06,
+      "loss": 0.3371,
+      "step": 10776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.184583375766741,
+      "learning_rate": 1.5604599995313196e-06,
+      "loss": 0.305,
+      "step": 10777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.0942138660450293,
+      "learning_rate": 1.5596429417612964e-06,
+      "loss": 0.3076,
+      "step": 10778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.191087230870885,
+      "learning_rate": 1.5588260584249466e-06,
+      "loss": 0.3061,
+      "step": 10779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8019125042591884,
+      "learning_rate": 1.5580093495636878e-06,
+      "loss": 0.3233,
+      "step": 10780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.948281462165301,
+      "learning_rate": 1.5571928152189286e-06,
+      "loss": 0.3378,
+      "step": 10781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.9175622345188,
+      "learning_rate": 1.5563764554320703e-06,
+      "loss": 0.29,
+      "step": 10782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.481959590543529,
+      "learning_rate": 1.555560270244505e-06,
+      "loss": 0.3162,
+      "step": 10783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.654173234811763,
+      "learning_rate": 1.5547442596976115e-06,
+      "loss": 0.2999,
+      "step": 10784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.1367356413957,
+      "learning_rate": 1.5539284238327657e-06,
+      "loss": 0.2765,
+      "step": 10785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.095407402626068,
+      "learning_rate": 1.5531127626913317e-06,
+      "loss": 0.3163,
+      "step": 10786
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.8566081473425857,
+      "learning_rate": 1.5522972763146653e-06,
+      "loss": 0.3256,
+      "step": 10787
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.102448303216285,
+      "learning_rate": 1.551481964744116e-06,
+      "loss": 0.3288,
+      "step": 10788
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9814463935734214,
+      "learning_rate": 1.5506668280210157e-06,
+      "loss": 0.3109,
+      "step": 10789
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29241964820040367,
+      "learning_rate": 1.5498518661867007e-06,
+      "loss": 0.1085,
+      "step": 10790
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.509828671466418,
+      "learning_rate": 1.5490370792824855e-06,
+      "loss": 0.2978,
+      "step": 10791
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.1892390817939957,
+      "learning_rate": 1.548222467349687e-06,
+      "loss": 0.3238,
+      "step": 10792
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.055198122234162,
+      "learning_rate": 1.5474080304296036e-06,
+      "loss": 0.3262,
+      "step": 10793
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.488722413166543,
+      "learning_rate": 1.5465937685635308e-06,
+      "loss": 0.2924,
+      "step": 10794
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.29423763140456993,
+      "learning_rate": 1.545779681792753e-06,
+      "loss": 0.1136,
+      "step": 10795
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.23455478841306,
+      "learning_rate": 1.5449657701585457e-06,
+      "loss": 0.3364,
+      "step": 10796
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.594601651049113,
+      "learning_rate": 1.5441520337021786e-06,
+      "loss": 0.3202,
+      "step": 10797
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.999096345199169,
+      "learning_rate": 1.543338472464907e-06,
+      "loss": 0.3323,
+      "step": 10798
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.079576806033168,
+      "learning_rate": 1.542525086487981e-06,
+      "loss": 0.274,
+      "step": 10799
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.737260916498374,
+      "learning_rate": 1.5417118758126408e-06,
+      "loss": 0.2911,
+      "step": 10800
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.253055547259506,
+      "learning_rate": 1.540898840480119e-06,
+      "loss": 0.3108,
+      "step": 10801
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9723369149121095,
+      "learning_rate": 1.5400859805316375e-06,
+      "loss": 0.3171,
+      "step": 10802
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 8.567869522320189,
+      "learning_rate": 1.5392732960084122e-06,
+      "loss": 0.3155,
+      "step": 10803
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.654495773509039,
+      "learning_rate": 1.5384607869516426e-06,
+      "loss": 0.3362,
+      "step": 10804
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.720284165359905,
+      "learning_rate": 1.5376484534025315e-06,
+      "loss": 0.2949,
+      "step": 10805
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.017527357679902,
+      "learning_rate": 1.5368362954022613e-06,
+      "loss": 0.2843,
+      "step": 10806
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9005785920534524,
+      "learning_rate": 1.5360243129920116e-06,
+      "loss": 0.2936,
+      "step": 10807
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.383086749496424,
+      "learning_rate": 1.535212506212952e-06,
+      "loss": 0.313,
+      "step": 10808
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.7227314669681153,
+      "learning_rate": 1.5344008751062423e-06,
+      "loss": 0.3008,
+      "step": 10809
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2792574855576253,
+      "learning_rate": 1.5335894197130358e-06,
+      "loss": 0.1003,
+      "step": 10810
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.9120808849608495,
+      "learning_rate": 1.5327781400744718e-06,
+      "loss": 0.3249,
+      "step": 10811
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.1378807840459,
+      "learning_rate": 1.5319670362316863e-06,
+      "loss": 0.292,
+      "step": 10812
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.0798156923464655,
+      "learning_rate": 1.531156108225803e-06,
+      "loss": 0.3297,
+      "step": 10813
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.6246312646823835,
+      "learning_rate": 1.5303453560979386e-06,
+      "loss": 0.3286,
+      "step": 10814
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.246301985984442,
+      "learning_rate": 1.5295347798891997e-06,
+      "loss": 0.2789,
+      "step": 10815
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.142094717420603,
+      "learning_rate": 1.5287243796406852e-06,
+      "loss": 0.366,
+      "step": 10816
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.317135893727318,
+      "learning_rate": 1.5279141553934807e-06,
+      "loss": 0.3092,
+      "step": 10817
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.381270476619891,
+      "learning_rate": 1.5271041071886716e-06,
+      "loss": 0.3567,
+      "step": 10818
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.2997531750759155,
+      "learning_rate": 1.526294235067325e-06,
+      "loss": 0.3064,
+      "step": 10819
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9188403265985823,
+      "learning_rate": 1.5254845390705047e-06,
+      "loss": 0.3116,
+      "step": 10820
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.575786786998518,
+      "learning_rate": 1.5246750192392634e-06,
+      "loss": 0.3109,
+      "step": 10821
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 11.38035089526377,
+      "learning_rate": 1.523865675614647e-06,
+      "loss": 0.2963,
+      "step": 10822
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.410262577904079,
+      "learning_rate": 1.5230565082376909e-06,
+      "loss": 0.3194,
+      "step": 10823
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 20.71350232441843,
+      "learning_rate": 1.5222475171494184e-06,
+      "loss": 0.3157,
+      "step": 10824
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.050891366956142,
+      "learning_rate": 1.5214387023908523e-06,
+      "loss": 0.2738,
+      "step": 10825
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.18687530329946,
+      "learning_rate": 1.520630064002997e-06,
+      "loss": 0.3133,
+      "step": 10826
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.423594285849336,
+      "learning_rate": 1.5198216020268535e-06,
+      "loss": 0.3344,
+      "step": 10827
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 10.283487817706682,
+      "learning_rate": 1.5190133165034137e-06,
+      "loss": 0.2993,
+      "step": 10828
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.585830748666644,
+      "learning_rate": 1.5182052074736581e-06,
+      "loss": 0.3287,
+      "step": 10829
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.252399639062085,
+      "learning_rate": 1.5173972749785604e-06,
+      "loss": 0.3119,
+      "step": 10830
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.797311507671084,
+      "learning_rate": 1.5165895190590856e-06,
+      "loss": 0.3196,
+      "step": 10831
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6882673302278253,
+      "learning_rate": 1.5157819397561863e-06,
+      "loss": 0.2924,
+      "step": 10832
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.231107791611745,
+      "learning_rate": 1.5149745371108099e-06,
+      "loss": 0.3076,
+      "step": 10833
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.709015161048756,
+      "learning_rate": 1.5141673111638928e-06,
+      "loss": 0.3626,
+      "step": 10834
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.594190774121724,
+      "learning_rate": 1.5133602619563637e-06,
+      "loss": 0.3028,
+      "step": 10835
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.3210887024287965,
+      "learning_rate": 1.5125533895291433e-06,
+      "loss": 0.3399,
+      "step": 10836
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.712934791957292,
+      "learning_rate": 1.5117466939231373e-06,
+      "loss": 0.2916,
+      "step": 10837
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.468198278106476,
+      "learning_rate": 1.5109401751792524e-06,
+      "loss": 0.3144,
+      "step": 10838
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8253703971384336,
+      "learning_rate": 1.5101338333383759e-06,
+      "loss": 0.3061,
+      "step": 10839
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.6194724767092636,
+      "learning_rate": 1.509327668441396e-06,
+      "loss": 0.3288,
+      "step": 10840
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.839508265682215,
+      "learning_rate": 1.5085216805291835e-06,
+      "loss": 0.311,
+      "step": 10841
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2796493898507754,
+      "learning_rate": 1.5077158696426048e-06,
+      "loss": 0.1079,
+      "step": 10842
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8368217317019764,
+      "learning_rate": 1.5069102358225157e-06,
+      "loss": 0.2721,
+      "step": 10843
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.415478219537598,
+      "learning_rate": 1.506104779109766e-06,
+      "loss": 0.2883,
+      "step": 10844
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.221600015076044,
+      "learning_rate": 1.5052994995451915e-06,
+      "loss": 0.3004,
+      "step": 10845
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.2502284114902,
+      "learning_rate": 1.5044943971696218e-06,
+      "loss": 0.334,
+      "step": 10846
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.7838588187371123,
+      "learning_rate": 1.503689472023878e-06,
+      "loss": 0.3222,
+      "step": 10847
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2972113835359371,
+      "learning_rate": 1.5028847241487715e-06,
+      "loss": 0.1116,
+      "step": 10848
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.5236079612705,
+      "learning_rate": 1.5020801535851072e-06,
+      "loss": 0.3516,
+      "step": 10849
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.2952469455797262,
+      "learning_rate": 1.5012757603736723e-06,
+      "loss": 0.1106,
+      "step": 10850
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.94478791529771,
+      "learning_rate": 1.5004715445552593e-06,
+      "loss": 0.3084,
+      "step": 10851
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.2168213073174217,
+      "learning_rate": 1.4996675061706355e-06,
+      "loss": 0.3168,
+      "step": 10852
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.804424217242654,
+      "learning_rate": 1.498863645260576e-06,
+      "loss": 0.2971,
+      "step": 10853
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.924594625362166,
+      "learning_rate": 1.4980599618658314e-06,
+      "loss": 0.3144,
+      "step": 10854
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.203610558824387,
+      "learning_rate": 1.4972564560271536e-06,
+      "loss": 0.3277,
+      "step": 10855
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 9.808699820064023,
+      "learning_rate": 1.4964531277852805e-06,
+      "loss": 0.3385,
+      "step": 10856
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.477174936119575,
+      "learning_rate": 1.4956499771809434e-06,
+      "loss": 0.3027,
+      "step": 10857
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.95700215977975,
+      "learning_rate": 1.4948470042548647e-06,
+      "loss": 0.3572,
+      "step": 10858
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 7.076693762481167,
+      "learning_rate": 1.494044209047757e-06,
+      "loss": 0.2829,
+      "step": 10859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.1948322548045605,
+      "learning_rate": 1.4932415916003212e-06,
+      "loss": 0.3386,
+      "step": 10860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.636747854811857,
+      "learning_rate": 1.4924391519532533e-06,
+      "loss": 0.3291,
+      "step": 10861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9894497744957254,
+      "learning_rate": 1.491636890147239e-06,
+      "loss": 0.2969,
+      "step": 10862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.269494842693577,
+      "learning_rate": 1.490834806222954e-06,
+      "loss": 0.3421,
+      "step": 10863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3112527542197117,
+      "learning_rate": 1.4900329002210684e-06,
+      "loss": 0.3036,
+      "step": 10864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.369412330034279,
+      "learning_rate": 1.489231172182235e-06,
+      "loss": 0.31,
+      "step": 10865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.590696025372441,
+      "learning_rate": 1.48842962214711e-06,
+      "loss": 0.2976,
+      "step": 10866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.580295190622213,
+      "learning_rate": 1.487628250156327e-06,
+      "loss": 0.2829,
+      "step": 10867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.163716562259725,
+      "learning_rate": 1.4868270562505244e-06,
+      "loss": 0.3297,
+      "step": 10868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.966826948466531,
+      "learning_rate": 1.4860260404703187e-06,
+      "loss": 0.2946,
+      "step": 10869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9486075660020434,
+      "learning_rate": 1.4852252028563252e-06,
+      "loss": 0.3243,
+      "step": 10870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.149562498763534,
+      "learning_rate": 1.4844245434491482e-06,
+      "loss": 0.3217,
+      "step": 10871
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.004443187737325,
+      "learning_rate": 1.4836240622893832e-06,
+      "loss": 0.3082,
+      "step": 10872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1465846223795704,
+      "learning_rate": 1.4828237594176176e-06,
+      "loss": 0.3195,
+      "step": 10873
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 20.236106438285123,
+      "learning_rate": 1.4820236348744255e-06,
+      "loss": 0.3561,
+      "step": 10874
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.38831550370324,
+      "learning_rate": 1.4812236887003762e-06,
+      "loss": 0.3083,
+      "step": 10875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2876888461372653,
+      "learning_rate": 1.4804239209360288e-06,
+      "loss": 0.2945,
+      "step": 10876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.339862056733858,
+      "learning_rate": 1.4796243316219338e-06,
+      "loss": 0.333,
+      "step": 10877
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.858806590170895,
+      "learning_rate": 1.4788249207986317e-06,
+      "loss": 0.3062,
+      "step": 10878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.69537921030639,
+      "learning_rate": 1.4780256885066558e-06,
+      "loss": 0.286,
+      "step": 10879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.765729530700074,
+      "learning_rate": 1.477226634786525e-06,
+      "loss": 0.3365,
+      "step": 10880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3250795150929612,
+      "learning_rate": 1.4764277596787586e-06,
+      "loss": 0.1087,
+      "step": 10881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7846116750502703,
+      "learning_rate": 1.4756290632238562e-06,
+      "loss": 0.2942,
+      "step": 10882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.3569484609762785,
+      "learning_rate": 1.4748305454623163e-06,
+      "loss": 0.3077,
+      "step": 10883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8695767354617745,
+      "learning_rate": 1.4740322064346252e-06,
+      "loss": 0.2649,
+      "step": 10884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2972774333929326,
+      "learning_rate": 1.4732340461812595e-06,
+      "loss": 0.1259,
+      "step": 10885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.0376071212385103,
+      "learning_rate": 1.4724360647426906e-06,
+      "loss": 0.2941,
+      "step": 10886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.712116190989187,
+      "learning_rate": 1.4716382621593734e-06,
+      "loss": 0.2826,
+      "step": 10887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4275912429368405,
+      "learning_rate": 1.4708406384717605e-06,
+      "loss": 0.3057,
+      "step": 10888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.154803973252404,
+      "learning_rate": 1.4700431937202935e-06,
+      "loss": 0.2974,
+      "step": 10889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.482811624900534,
+      "learning_rate": 1.469245927945404e-06,
+      "loss": 0.3215,
+      "step": 10890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.146666027747522,
+      "learning_rate": 1.4684488411875158e-06,
+      "loss": 0.36,
+      "step": 10891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.447607968176764,
+      "learning_rate": 1.4676519334870443e-06,
+      "loss": 0.344,
+      "step": 10892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 37.27232661451137,
+      "learning_rate": 1.4668552048843888e-06,
+      "loss": 0.3131,
+      "step": 10893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.157410434309029,
+      "learning_rate": 1.4660586554199529e-06,
+      "loss": 0.3405,
+      "step": 10894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30603633641002104,
+      "learning_rate": 1.4652622851341176e-06,
+      "loss": 0.1221,
+      "step": 10895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7352221087048494,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.3329,
+      "step": 10896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8592465420208675,
+      "learning_rate": 1.4636700822597572e-06,
+      "loss": 0.3219,
+      "step": 10897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30627043969149065,
+      "learning_rate": 1.46287424975196e-06,
+      "loss": 0.1166,
+      "step": 10898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.30974303649042445,
+      "learning_rate": 1.4620785965842232e-06,
+      "loss": 0.1124,
+      "step": 10899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.177860731672618,
+      "learning_rate": 1.4612831227968838e-06,
+      "loss": 0.2822,
+      "step": 10900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.229213749596525,
+      "learning_rate": 1.4604878284302797e-06,
+      "loss": 0.29,
+      "step": 10901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1585606943076763,
+      "learning_rate": 1.4596927135247307e-06,
+      "loss": 0.3135,
+      "step": 10902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1475604066883123,
+      "learning_rate": 1.458897778120551e-06,
+      "loss": 0.3116,
+      "step": 10903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.2391030067541955,
+      "learning_rate": 1.4581030222580467e-06,
+      "loss": 0.298,
+      "step": 10904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2814863972378854,
+      "learning_rate": 1.4573084459775127e-06,
+      "loss": 0.3051,
+      "step": 10905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.335397895145249,
+      "learning_rate": 1.456514049319236e-06,
+      "loss": 0.2912,
+      "step": 10906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 26.330287610190595,
+      "learning_rate": 1.4557198323234967e-06,
+      "loss": 0.2971,
+      "step": 10907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3139700987361248,
+      "learning_rate": 1.454925795030559e-06,
+      "loss": 0.1203,
+      "step": 10908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.76360758892413,
+      "learning_rate": 1.4541319374806845e-06,
+      "loss": 0.3123,
+      "step": 10909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.672806181392574,
+      "learning_rate": 1.4533382597141232e-06,
+      "loss": 0.332,
+      "step": 10910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.639516569375234,
+      "learning_rate": 1.4525447617711175e-06,
+      "loss": 0.3292,
+      "step": 10911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28690273746718103,
+      "learning_rate": 1.4517514436918995e-06,
+      "loss": 0.1117,
+      "step": 10912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.032691108205207,
+      "learning_rate": 1.4509583055166882e-06,
+      "loss": 0.3601,
+      "step": 10913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.381102286036428,
+      "learning_rate": 1.4501653472857037e-06,
+      "loss": 0.2966,
+      "step": 10914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5771730064482385,
+      "learning_rate": 1.4493725690391458e-06,
+      "loss": 0.3437,
+      "step": 10915
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.26724730747348,
+      "learning_rate": 1.4485799708172116e-06,
+      "loss": 0.3286,
+      "step": 10916
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.433737084472745,
+      "learning_rate": 1.4477875526600882e-06,
+      "loss": 0.3122,
+      "step": 10917
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 31.601593346938536,
+      "learning_rate": 1.4469953146079524e-06,
+      "loss": 0.3353,
+      "step": 10918
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.520294050628755,
+      "learning_rate": 1.4462032567009727e-06,
+      "loss": 0.3142,
+      "step": 10919
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.510973586624825,
+      "learning_rate": 1.4454113789793096e-06,
+      "loss": 0.3479,
+      "step": 10920
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.012888249777679,
+      "learning_rate": 1.4446196814831093e-06,
+      "loss": 0.3143,
+      "step": 10921
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.678866108489528,
+      "learning_rate": 1.4438281642525153e-06,
+      "loss": 0.3128,
+      "step": 10922
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4416952002029193,
+      "learning_rate": 1.443036827327659e-06,
+      "loss": 0.3082,
+      "step": 10923
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 36.50665856016072,
+      "learning_rate": 1.4422456707486621e-06,
+      "loss": 0.3205,
+      "step": 10924
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.29743660512183634,
+      "learning_rate": 1.4414546945556408e-06,
+      "loss": 0.1162,
+      "step": 10925
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9264963752756015,
+      "learning_rate": 1.4406638987886935e-06,
+      "loss": 0.332,
+      "step": 10926
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6850247747014353,
+      "learning_rate": 1.439873283487922e-06,
+      "loss": 0.2836,
+      "step": 10927
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.048652132814266,
+      "learning_rate": 1.439082848693406e-06,
+      "loss": 0.3442,
+      "step": 10928
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.555546907264317,
+      "learning_rate": 1.4382925944452286e-06,
+      "loss": 0.3066,
+      "step": 10929
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 8.497536857793603,
+      "learning_rate": 1.4375025207834526e-06,
+      "loss": 0.3153,
+      "step": 10930
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.992072043144169,
+      "learning_rate": 1.436712627748138e-06,
+      "loss": 0.3038,
+      "step": 10931
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.210331788484875,
+      "learning_rate": 1.435922915379334e-06,
+      "loss": 0.3177,
+      "step": 10932
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.137654676220779,
+      "learning_rate": 1.4351333837170817e-06,
+      "loss": 0.3202,
+      "step": 10933
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.667516655507252,
+      "learning_rate": 1.4343440328014124e-06,
+      "loss": 0.3216,
+      "step": 10934
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.517528211648713,
+      "learning_rate": 1.4335548626723456e-06,
+      "loss": 0.3194,
+      "step": 10935
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2978278128966471,
+      "learning_rate": 1.4327658733698947e-06,
+      "loss": 0.1154,
+      "step": 10936
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.171936006181255,
+      "learning_rate": 1.4319770649340641e-06,
+      "loss": 0.3244,
+      "step": 10937
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.685846399131721,
+      "learning_rate": 1.431188437404848e-06,
+      "loss": 0.3256,
+      "step": 10938
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.068349977139641,
+      "learning_rate": 1.430399990822231e-06,
+      "loss": 0.3054,
+      "step": 10939
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.4470095898400483,
+      "learning_rate": 1.429611725226191e-06,
+      "loss": 0.3224,
+      "step": 10940
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.446963526905142,
+      "learning_rate": 1.4288236406566896e-06,
+      "loss": 0.312,
+      "step": 10941
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.337791297563564,
+      "learning_rate": 1.4280357371536919e-06,
+      "loss": 0.3266,
+      "step": 10942
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.198469617701781,
+      "learning_rate": 1.4272480147571388e-06,
+      "loss": 0.2884,
+      "step": 10943
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.580675093874458,
+      "learning_rate": 1.4264604735069764e-06,
+      "loss": 0.318,
+      "step": 10944
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.538805854353697,
+      "learning_rate": 1.4256731134431307e-06,
+      "loss": 0.3309,
+      "step": 10945
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.976541364987216,
+      "learning_rate": 1.4248859346055228e-06,
+      "loss": 0.3336,
+      "step": 10946
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.2806414433904365,
+      "learning_rate": 1.4240989370340652e-06,
+      "loss": 0.31,
+      "step": 10947
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9431468192555648,
+      "learning_rate": 1.42331212076866e-06,
+      "loss": 0.3182,
+      "step": 10948
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.28164803590737986,
+      "learning_rate": 1.4225254858492028e-06,
+      "loss": 0.1073,
+      "step": 10949
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.937398591889817,
+      "learning_rate": 1.4217390323155738e-06,
+      "loss": 0.319,
+      "step": 10950
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 11.519519522333534,
+      "learning_rate": 1.4209527602076505e-06,
+      "loss": 0.3114,
+      "step": 10951
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.302974104497702,
+      "learning_rate": 1.4201666695652976e-06,
+      "loss": 0.3078,
+      "step": 10952
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.89346169156433,
+      "learning_rate": 1.4193807604283731e-06,
+      "loss": 0.2827,
+      "step": 10953
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.143873947420875,
+      "learning_rate": 1.418595032836721e-06,
+      "loss": 0.3255,
+      "step": 10954
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.148880691625544,
+      "learning_rate": 1.4178094868301839e-06,
+      "loss": 0.3269,
+      "step": 10955
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.347643715809147,
+      "learning_rate": 1.4170241224485859e-06,
+      "loss": 0.3384,
+      "step": 10956
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.254841009380523,
+      "learning_rate": 1.4162389397317522e-06,
+      "loss": 0.327,
+      "step": 10957
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.653714608526398,
+      "learning_rate": 1.4154539387194882e-06,
+      "loss": 0.3527,
+      "step": 10958
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.880057260486679,
+      "learning_rate": 1.4146691194515982e-06,
+      "loss": 0.3108,
+      "step": 10959
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.879557235259022,
+      "learning_rate": 1.4138844819678726e-06,
+      "loss": 0.2891,
+      "step": 10960
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.9079330342242247,
+      "learning_rate": 1.4131000263080957e-06,
+      "loss": 0.2947,
+      "step": 10961
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1373923051498895,
+      "learning_rate": 1.4123157525120418e-06,
+      "loss": 0.3102,
+      "step": 10962
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.224334869487143,
+      "learning_rate": 1.4115316606194723e-06,
+      "loss": 0.314,
+      "step": 10963
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.265150300048879,
+      "learning_rate": 1.410747750670144e-06,
+      "loss": 0.2877,
+      "step": 10964
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.8677045187010135,
+      "learning_rate": 1.409964022703803e-06,
+      "loss": 0.3477,
+      "step": 10965
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.064926422536909,
+      "learning_rate": 1.4091804767601863e-06,
+      "loss": 0.3081,
+      "step": 10966
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1929921180093,
+      "learning_rate": 1.4083971128790208e-06,
+      "loss": 0.2975,
+      "step": 10967
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.184836927067349,
+      "learning_rate": 1.4076139311000264e-06,
+      "loss": 0.3173,
+      "step": 10968
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.809655466165637,
+      "learning_rate": 1.4068309314629074e-06,
+      "loss": 0.3034,
+      "step": 10969
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 15.62743453657758,
+      "learning_rate": 1.4060481140073707e-06,
+      "loss": 0.3347,
+      "step": 10970
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.598465388153946,
+      "learning_rate": 1.4052654787731012e-06,
+      "loss": 0.3195,
+      "step": 10971
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.4491558479082505,
+      "learning_rate": 1.404483025799782e-06,
+      "loss": 0.3194,
+      "step": 10972
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 17.69553788184633,
+      "learning_rate": 1.403700755127086e-06,
+      "loss": 0.3234,
+      "step": 10973
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.971692116281172,
+      "learning_rate": 1.4029186667946748e-06,
+      "loss": 0.3221,
+      "step": 10974
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.119933441864078,
+      "learning_rate": 1.4021367608422043e-06,
+      "loss": 0.3071,
+      "step": 10975
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8430691943651127,
+      "learning_rate": 1.4013550373093139e-06,
+      "loss": 0.3153,
+      "step": 10976
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.7931982641784225,
+      "learning_rate": 1.4005734962356454e-06,
+      "loss": 0.3336,
+      "step": 10977
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.9902150574217874,
+      "learning_rate": 1.3997921376608192e-06,
+      "loss": 0.3064,
+      "step": 10978
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5128096815054106,
+      "learning_rate": 1.399010961624454e-06,
+      "loss": 0.3057,
+      "step": 10979
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.5721917436562154,
+      "learning_rate": 1.3982299681661577e-06,
+      "loss": 0.3133,
+      "step": 10980
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8008315953755947,
+      "learning_rate": 1.3974491573255271e-06,
+      "loss": 0.3251,
+      "step": 10981
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.8320238048861777,
+      "learning_rate": 1.3966685291421516e-06,
+      "loss": 0.3324,
+      "step": 10982
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.65321012284076,
+      "learning_rate": 1.3958880836556127e-06,
+      "loss": 0.3201,
+      "step": 10983
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8973020874706767,
+      "learning_rate": 1.395107820905477e-06,
+      "loss": 0.3099,
+      "step": 10984
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.5753808837142875,
+      "learning_rate": 1.3943277409313078e-06,
+      "loss": 0.2938,
+      "step": 10985
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.5793940304198495,
+      "learning_rate": 1.3935478437726557e-06,
+      "loss": 0.3131,
+      "step": 10986
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.171144321767439,
+      "learning_rate": 1.3927681294690643e-06,
+      "loss": 0.3011,
+      "step": 10987
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6795428859361543,
+      "learning_rate": 1.3919885980600683e-06,
+      "loss": 0.3057,
+      "step": 10988
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.79775179260116,
+      "learning_rate": 1.3912092495851864e-06,
+      "loss": 0.2877,
+      "step": 10989
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.894013718053051,
+      "learning_rate": 1.3904300840839396e-06,
+      "loss": 0.2822,
+      "step": 10990
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.053085542966629,
+      "learning_rate": 1.3896511015958292e-06,
+      "loss": 0.323,
+      "step": 10991
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2921635523346584,
+      "learning_rate": 1.388872302160353e-06,
+      "loss": 0.1178,
+      "step": 10992
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.6429541075821787,
+      "learning_rate": 1.388093685816997e-06,
+      "loss": 0.3393,
+      "step": 10993
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3003834880870318,
+      "learning_rate": 1.3873152526052386e-06,
+      "loss": 0.1049,
+      "step": 10994
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.988953057459803,
+      "learning_rate": 1.386537002564547e-06,
+      "loss": 0.3089,
+      "step": 10995
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.179925615872963,
+      "learning_rate": 1.3857589357343826e-06,
+      "loss": 0.333,
+      "step": 10996
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 6.759614417984015,
+      "learning_rate": 1.3849810521541917e-06,
+      "loss": 0.3174,
+      "step": 10997
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 4.366180545351714,
+      "learning_rate": 1.3842033518634163e-06,
+      "loss": 0.2871,
+      "step": 10998
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7069251687948888,
+      "learning_rate": 1.3834258349014873e-06,
+      "loss": 0.3167,
+      "step": 10999
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.2707786260432457,
+      "learning_rate": 1.3826485013078268e-06,
+      "loss": 0.2941,
+      "step": 11000
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 9.707732499204846,
+      "learning_rate": 1.381871351121849e-06,
+      "loss": 0.323,
+      "step": 11001
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.325382368293443,
+      "learning_rate": 1.3810943843829522e-06,
+      "loss": 0.3074,
+      "step": 11002
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.5341843090840594,
+      "learning_rate": 1.3803176011305368e-06,
+      "loss": 0.3249,
+      "step": 11003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.144917976870369,
+      "learning_rate": 1.3795410014039806e-06,
+      "loss": 0.2805,
+      "step": 11004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.392708714636178,
+      "learning_rate": 1.3787645852426663e-06,
+      "loss": 0.333,
+      "step": 11005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.682699140127983,
+      "learning_rate": 1.3779883526859538e-06,
+      "loss": 0.3396,
+      "step": 11006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9786651979750194,
+      "learning_rate": 1.3772123037732028e-06,
+      "loss": 0.3135,
+      "step": 11007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.964279669968499,
+      "learning_rate": 1.3764364385437595e-06,
+      "loss": 0.3624,
+      "step": 11008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.41497769480999,
+      "learning_rate": 1.3756607570369623e-06,
+      "loss": 0.3409,
+      "step": 11009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.0776827626968855,
+      "learning_rate": 1.3748852592921418e-06,
+      "loss": 0.3263,
+      "step": 11010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.755056667135741,
+      "learning_rate": 1.3741099453486145e-06,
+      "loss": 0.3154,
+      "step": 11011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9851619429077876,
+      "learning_rate": 1.373334815245691e-06,
+      "loss": 0.2928,
+      "step": 11012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.892235119142359,
+      "learning_rate": 1.3725598690226738e-06,
+      "loss": 0.2955,
+      "step": 11013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31806065070146067,
+      "learning_rate": 1.3717851067188526e-06,
+      "loss": 0.1166,
+      "step": 11014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6812448284945054,
+      "learning_rate": 1.371010528373511e-06,
+      "loss": 0.3163,
+      "step": 11015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1949486428024683,
+      "learning_rate": 1.3702361340259224e-06,
+      "loss": 0.3198,
+      "step": 11016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3971256160691388,
+      "learning_rate": 1.3694619237153462e-06,
+      "loss": 0.3346,
+      "step": 11017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.204826115502747,
+      "learning_rate": 1.3686878974810425e-06,
+      "loss": 0.3053,
+      "step": 11018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.851376636814677,
+      "learning_rate": 1.3679140553622505e-06,
+      "loss": 0.2769,
+      "step": 11019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1875727188044363,
+      "learning_rate": 1.3671403973982106e-06,
+      "loss": 0.3427,
+      "step": 11020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7751409219312646,
+      "learning_rate": 1.3663669236281458e-06,
+      "loss": 0.3247,
+      "step": 11021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.190060164089563,
+      "learning_rate": 1.3655936340912734e-06,
+      "loss": 0.3272,
+      "step": 11022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.558262423032122,
+      "learning_rate": 1.364820528826803e-06,
+      "loss": 0.3247,
+      "step": 11023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.129870957279547,
+      "learning_rate": 1.3640476078739296e-06,
+      "loss": 0.3078,
+      "step": 11024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.575431089784856,
+      "learning_rate": 1.363274871271843e-06,
+      "loss": 0.3286,
+      "step": 11025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.6994406847462225,
+      "learning_rate": 1.3625023190597236e-06,
+      "loss": 0.344,
+      "step": 11026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.903450056304924,
+      "learning_rate": 1.3617299512767406e-06,
+      "loss": 0.3127,
+      "step": 11027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7374082270319415,
+      "learning_rate": 1.3609577679620551e-06,
+      "loss": 0.3216,
+      "step": 11028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.299480597682647,
+      "learning_rate": 1.36018576915482e-06,
+      "loss": 0.298,
+      "step": 11029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.462077044994786,
+      "learning_rate": 1.3594139548941732e-06,
+      "loss": 0.3238,
+      "step": 11030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.626196315165288,
+      "learning_rate": 1.3586423252192527e-06,
+      "loss": 0.3324,
+      "step": 11031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.890679072839392,
+      "learning_rate": 1.357870880169176e-06,
+      "loss": 0.2861,
+      "step": 11032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.949289483203707,
+      "learning_rate": 1.3570996197830632e-06,
+      "loss": 0.3201,
+      "step": 11033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.1283647777798205,
+      "learning_rate": 1.3563285441000145e-06,
+      "loss": 0.3042,
+      "step": 11034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2620057834554532,
+      "learning_rate": 1.3555576531591264e-06,
+      "loss": 0.3026,
+      "step": 11035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.472800397465463,
+      "learning_rate": 1.3547869469994851e-06,
+      "loss": 0.3072,
+      "step": 11036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.761647790661294,
+      "learning_rate": 1.3540164256601674e-06,
+      "loss": 0.2953,
+      "step": 11037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2003298094807024,
+      "learning_rate": 1.3532460891802408e-06,
+      "loss": 0.3026,
+      "step": 11038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2816419232529013,
+      "learning_rate": 1.3524759375987606e-06,
+      "loss": 0.3061,
+      "step": 11039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.121500322526954,
+      "learning_rate": 1.351705970954777e-06,
+      "loss": 0.2786,
+      "step": 11040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.23895815426017,
+      "learning_rate": 1.3509361892873285e-06,
+      "loss": 0.3263,
+      "step": 11041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.541713713262006,
+      "learning_rate": 1.3501665926354451e-06,
+      "loss": 0.2951,
+      "step": 11042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.34459361920180825,
+      "learning_rate": 1.3493971810381474e-06,
+      "loss": 0.1123,
+      "step": 11043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.6048727995913685,
+      "learning_rate": 1.3486279545344465e-06,
+      "loss": 0.3147,
+      "step": 11044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.209319047139193,
+      "learning_rate": 1.3478589131633407e-06,
+      "loss": 0.3345,
+      "step": 11045
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.28244733250470144,
+      "learning_rate": 1.3470900569638273e-06,
+      "loss": 0.1038,
+      "step": 11046
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.277441728465137,
+      "learning_rate": 1.3463213859748853e-06,
+      "loss": 0.3353,
+      "step": 11047
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.732079109284317,
+      "learning_rate": 1.3455529002354884e-06,
+      "loss": 0.2981,
+      "step": 11048
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.746681510413397,
+      "learning_rate": 1.344784599784601e-06,
+      "loss": 0.2994,
+      "step": 11049
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.197464849334233,
+      "learning_rate": 1.344016484661178e-06,
+      "loss": 0.3006,
+      "step": 11050
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.888380530900801,
+      "learning_rate": 1.3432485549041658e-06,
+      "loss": 0.31,
+      "step": 11051
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 8.784891646337053,
+      "learning_rate": 1.3424808105524956e-06,
+      "loss": 0.3091,
+      "step": 11052
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5910479196727776,
+      "learning_rate": 1.3417132516450993e-06,
+      "loss": 0.2812,
+      "step": 11053
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6949962739463107,
+      "learning_rate": 1.34094587822089e-06,
+      "loss": 0.3386,
+      "step": 11054
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.653762840296952,
+      "learning_rate": 1.3401786903187763e-06,
+      "loss": 0.3112,
+      "step": 11055
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 65.06260006457305,
+      "learning_rate": 1.339411687977657e-06,
+      "loss": 0.3078,
+      "step": 11056
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.066966093687354,
+      "learning_rate": 1.3386448712364198e-06,
+      "loss": 0.2971,
+      "step": 11057
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4823037949539972,
+      "learning_rate": 1.337878240133944e-06,
+      "loss": 0.3058,
+      "step": 11058
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5900902897635514,
+      "learning_rate": 1.3371117947091022e-06,
+      "loss": 0.2937,
+      "step": 11059
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.014092533309338,
+      "learning_rate": 1.336345535000751e-06,
+      "loss": 0.3005,
+      "step": 11060
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.5819230431377815,
+      "learning_rate": 1.335579461047743e-06,
+      "loss": 0.3378,
+      "step": 11061
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.64010531099029,
+      "learning_rate": 1.3348135728889194e-06,
+      "loss": 0.2874,
+      "step": 11062
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.43648933008729,
+      "learning_rate": 1.3340478705631137e-06,
+      "loss": 0.3171,
+      "step": 11063
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.822327763101432,
+      "learning_rate": 1.3332823541091488e-06,
+      "loss": 0.3114,
+      "step": 11064
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.686767835191618,
+      "learning_rate": 1.3325170235658347e-06,
+      "loss": 0.3098,
+      "step": 11065
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.779532399713947,
+      "learning_rate": 1.3317518789719803e-06,
+      "loss": 0.3308,
+      "step": 11066
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.7941964462641695,
+      "learning_rate": 1.3309869203663767e-06,
+      "loss": 0.3323,
+      "step": 11067
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9197563810936553,
+      "learning_rate": 1.3302221477878092e-06,
+      "loss": 0.2871,
+      "step": 11068
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.658145110141079,
+      "learning_rate": 1.329457561275055e-06,
+      "loss": 0.3357,
+      "step": 11069
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.51790849767036,
+      "learning_rate": 1.328693160866879e-06,
+      "loss": 0.3325,
+      "step": 11070
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.211553433762996,
+      "learning_rate": 1.3279289466020384e-06,
+      "loss": 0.2969,
+      "step": 11071
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6282299120295938,
+      "learning_rate": 1.327164918519282e-06,
+      "loss": 0.286,
+      "step": 11072
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3213604012181204,
+      "learning_rate": 1.3264010766573454e-06,
+      "loss": 0.315,
+      "step": 11073
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.49755814857768,
+      "learning_rate": 1.3256374210549578e-06,
+      "loss": 0.3529,
+      "step": 11074
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.042203091400098,
+      "learning_rate": 1.324873951750838e-06,
+      "loss": 0.2816,
+      "step": 11075
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.802931135845875,
+      "learning_rate": 1.3241106687836964e-06,
+      "loss": 0.3537,
+      "step": 11076
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.260820994135072,
+      "learning_rate": 1.323347572192234e-06,
+      "loss": 0.3025,
+      "step": 11077
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.347801019985807,
+      "learning_rate": 1.3225846620151372e-06,
+      "loss": 0.3304,
+      "step": 11078
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.765481822839771,
+      "learning_rate": 1.321821938291093e-06,
+      "loss": 0.3449,
+      "step": 11079
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.234119665964835,
+      "learning_rate": 1.321059401058768e-06,
+      "loss": 0.305,
+      "step": 11080
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.058711896250564,
+      "learning_rate": 1.3202970503568291e-06,
+      "loss": 0.2993,
+      "step": 11081
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6093157772843947,
+      "learning_rate": 1.319534886223926e-06,
+      "loss": 0.3078,
+      "step": 11082
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.2714291186298965,
+      "learning_rate": 1.3187729086987028e-06,
+      "loss": 0.3272,
+      "step": 11083
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.7096534789311075,
+      "learning_rate": 1.3180111178197942e-06,
+      "loss": 0.3146,
+      "step": 11084
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.316334277020809,
+      "learning_rate": 1.3172495136258235e-06,
+      "loss": 0.282,
+      "step": 11085
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2710381129345816,
+      "learning_rate": 1.3164880961554083e-06,
+      "loss": 0.322,
+      "step": 11086
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0788282362274897,
+      "learning_rate": 1.3157268654471505e-06,
+      "loss": 0.2938,
+      "step": 11087
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.0296920253673205,
+      "learning_rate": 1.3149658215396478e-06,
+      "loss": 0.2989,
+      "step": 11088
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.056807126858441,
+      "learning_rate": 1.3142049644714872e-06,
+      "loss": 0.3057,
+      "step": 11089
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1857588303174635,
+      "learning_rate": 1.3134442942812454e-06,
+      "loss": 0.307,
+      "step": 11090
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.728932490852323,
+      "learning_rate": 1.3126838110074897e-06,
+      "loss": 0.3068,
+      "step": 11091
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.25012198517327,
+      "learning_rate": 1.311923514688781e-06,
+      "loss": 0.3052,
+      "step": 11092
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.987525979218215,
+      "learning_rate": 1.3111634053636623e-06,
+      "loss": 0.3252,
+      "step": 11093
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.93393918972608,
+      "learning_rate": 1.3104034830706792e-06,
+      "loss": 0.3095,
+      "step": 11094
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.648796608607133,
+      "learning_rate": 1.3096437478483553e-06,
+      "loss": 0.2888,
+      "step": 11095
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4550018935262803,
+      "learning_rate": 1.3088841997352174e-06,
+      "loss": 0.3337,
+      "step": 11096
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.866855125995864,
+      "learning_rate": 1.3081248387697714e-06,
+      "loss": 0.2947,
+      "step": 11097
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.650094862273486,
+      "learning_rate": 1.3073656649905203e-06,
+      "loss": 0.3288,
+      "step": 11098
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.374324197268242,
+      "learning_rate": 1.3066066784359566e-06,
+      "loss": 0.33,
+      "step": 11099
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.915515756818401,
+      "learning_rate": 1.3058478791445606e-06,
+      "loss": 0.3407,
+      "step": 11100
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.008475142148949,
+      "learning_rate": 1.305089267154806e-06,
+      "loss": 0.2857,
+      "step": 11101
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8346400073348517,
+      "learning_rate": 1.3043308425051559e-06,
+      "loss": 0.3336,
+      "step": 11102
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6436350173264542,
+      "learning_rate": 1.3035726052340648e-06,
+      "loss": 0.3179,
+      "step": 11103
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.463965675789288,
+      "learning_rate": 1.3028145553799764e-06,
+      "loss": 0.303,
+      "step": 11104
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.1708435866951783,
+      "learning_rate": 1.3020566929813273e-06,
+      "loss": 0.3008,
+      "step": 11105
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9376142456752974,
+      "learning_rate": 1.3012990180765379e-06,
+      "loss": 0.3359,
+      "step": 11106
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.078131527766416,
+      "learning_rate": 1.3005415307040309e-06,
+      "loss": 0.3399,
+      "step": 11107
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.973819092859138,
+      "learning_rate": 1.2997842309022057e-06,
+      "loss": 0.3222,
+      "step": 11108
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.6964761494794263,
+      "learning_rate": 1.299027118709465e-06,
+      "loss": 0.3062,
+      "step": 11109
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.227519689246099,
+      "learning_rate": 1.2982701941641923e-06,
+      "loss": 0.3052,
+      "step": 11110
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 10.906494176480843,
+      "learning_rate": 1.2975134573047666e-06,
+      "loss": 0.3199,
+      "step": 11111
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.478428199212671,
+      "learning_rate": 1.296756908169557e-06,
+      "loss": 0.3593,
+      "step": 11112
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.439144766127752,
+      "learning_rate": 1.2960005467969183e-06,
+      "loss": 0.3183,
+      "step": 11113
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.325813920144812,
+      "learning_rate": 1.2952443732252058e-06,
+      "loss": 0.3092,
+      "step": 11114
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.679943126165942,
+      "learning_rate": 1.2944883874927538e-06,
+      "loss": 0.2864,
+      "step": 11115
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.655825788929674,
+      "learning_rate": 1.293732589637895e-06,
+      "loss": 0.2976,
+      "step": 11116
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2712659862911924,
+      "learning_rate": 1.2929769796989493e-06,
+      "loss": 0.3138,
+      "step": 11117
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 12.512366783248195,
+      "learning_rate": 1.2922215577142284e-06,
+      "loss": 0.2835,
+      "step": 11118
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.8764932291208485,
+      "learning_rate": 1.2914663237220337e-06,
+      "loss": 0.2952,
+      "step": 11119
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3475974167436067,
+      "learning_rate": 1.290711277760658e-06,
+      "loss": 0.3176,
+      "step": 11120
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2680152670369824,
+      "learning_rate": 1.28995641986838e-06,
+      "loss": 0.2768,
+      "step": 11121
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.4080065981607652,
+      "learning_rate": 1.2892017500834786e-06,
+      "loss": 0.2849,
+      "step": 11122
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3686178026331266,
+      "learning_rate": 1.2884472684442122e-06,
+      "loss": 0.3289,
+      "step": 11123
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 9.428795981242892,
+      "learning_rate": 1.2876929749888374e-06,
+      "loss": 0.3079,
+      "step": 11124
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9391620088569472,
+      "learning_rate": 1.2869388697555968e-06,
+      "loss": 0.3255,
+      "step": 11125
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 22.70143848770302,
+      "learning_rate": 1.286184952782727e-06,
+      "loss": 0.3193,
+      "step": 11126
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 7.737271710132668,
+      "learning_rate": 1.2854312241084532e-06,
+      "loss": 0.3251,
+      "step": 11127
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.5409582426139585,
+      "learning_rate": 1.2846776837709874e-06,
+      "loss": 0.287,
+      "step": 11128
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.00653928385292,
+      "learning_rate": 1.2839243318085414e-06,
+      "loss": 0.3282,
+      "step": 11129
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.83802093739515,
+      "learning_rate": 1.2831711682593078e-06,
+      "loss": 0.3026,
+      "step": 11130
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.330438903950983,
+      "learning_rate": 1.2824181931614748e-06,
+      "loss": 0.3341,
+      "step": 11131
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.700740518765343,
+      "learning_rate": 1.2816654065532203e-06,
+      "loss": 0.2925,
+      "step": 11132
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.9382333703874894,
+      "learning_rate": 1.280912808472713e-06,
+      "loss": 0.3183,
+      "step": 11133
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.026275671527609,
+      "learning_rate": 1.2801603989581073e-06,
+      "loss": 0.3095,
+      "step": 11134
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.2848587553615358,
+      "learning_rate": 1.2794081780475581e-06,
+      "loss": 0.1047,
+      "step": 11135
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.6426965295604035,
+      "learning_rate": 1.2786561457791996e-06,
+      "loss": 0.3256,
+      "step": 11136
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.165850949888617,
+      "learning_rate": 1.2779043021911635e-06,
+      "loss": 0.2947,
+      "step": 11137
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.553202224086223,
+      "learning_rate": 1.27715264732157e-06,
+      "loss": 0.3205,
+      "step": 11138
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.6132061529295,
+      "learning_rate": 1.2764011812085293e-06,
+      "loss": 0.338,
+      "step": 11139
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.9988855962853,
+      "learning_rate": 1.2756499038901437e-06,
+      "loss": 0.2943,
+      "step": 11140
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 14.11894018790952,
+      "learning_rate": 1.2748988154045005e-06,
+      "loss": 0.3182,
+      "step": 11141
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.084007130979638,
+      "learning_rate": 1.274147915789688e-06,
+      "loss": 0.3074,
+      "step": 11142
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.364251431010953,
+      "learning_rate": 1.273397205083773e-06,
+      "loss": 0.3106,
+      "step": 11143
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 5.008489838904686,
+      "learning_rate": 1.2726466833248201e-06,
+      "loss": 0.3213,
+      "step": 11144
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.768007668499933,
+      "learning_rate": 1.2718963505508824e-06,
+      "loss": 0.3168,
+      "step": 11145
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.490546583594456,
+      "learning_rate": 1.2711462068000035e-06,
+      "loss": 0.2973,
+      "step": 11146
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 15.383593640654368,
+      "learning_rate": 1.270396252110217e-06,
+      "loss": 0.3061,
+      "step": 11147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.916170179255611,
+      "learning_rate": 1.2696464865195496e-06,
+      "loss": 0.3271,
+      "step": 11148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.041226480038677,
+      "learning_rate": 1.268896910066012e-06,
+      "loss": 0.3609,
+      "step": 11149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.123397823178727,
+      "learning_rate": 1.2681475227876117e-06,
+      "loss": 0.3083,
+      "step": 11150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.225715906903679,
+      "learning_rate": 1.2673983247223438e-06,
+      "loss": 0.3101,
+      "step": 11151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.288130554136752,
+      "learning_rate": 1.2666493159081944e-06,
+      "loss": 0.3189,
+      "step": 11152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4994164917123594,
+      "learning_rate": 1.2659004963831417e-06,
+      "loss": 0.2946,
+      "step": 11153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.899420108266816,
+      "learning_rate": 1.2651518661851475e-06,
+      "loss": 0.3122,
+      "step": 11154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.211863502432555,
+      "learning_rate": 1.2644034253521753e-06,
+      "loss": 0.2814,
+      "step": 11155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5484722780091347,
+      "learning_rate": 1.2636551739221665e-06,
+      "loss": 0.3244,
+      "step": 11156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.0098125144600525,
+      "learning_rate": 1.2629071119330655e-06,
+      "loss": 0.3106,
+      "step": 11157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.02120385273089,
+      "learning_rate": 1.2621592394227954e-06,
+      "loss": 0.2976,
+      "step": 11158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.141282345174481,
+      "learning_rate": 1.2614115564292779e-06,
+      "loss": 0.3204,
+      "step": 11159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.592099515201742,
+      "learning_rate": 1.2606640629904205e-06,
+      "loss": 0.3208,
+      "step": 11160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.28248218622791416,
+      "learning_rate": 1.2599167591441236e-06,
+      "loss": 0.1019,
+      "step": 11161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.811105755798463,
+      "learning_rate": 1.2591696449282786e-06,
+      "loss": 0.3304,
+      "step": 11162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.479955776327625,
+      "learning_rate": 1.258422720380763e-06,
+      "loss": 0.2844,
+      "step": 11163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9269722617219545,
+      "learning_rate": 1.2576759855394489e-06,
+      "loss": 0.3169,
+      "step": 11164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.042669418517816,
+      "learning_rate": 1.2569294404421972e-06,
+      "loss": 0.3135,
+      "step": 11165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.37142200343465015,
+      "learning_rate": 1.2561830851268598e-06,
+      "loss": 0.1086,
+      "step": 11166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.679036531206564,
+      "learning_rate": 1.2554369196312783e-06,
+      "loss": 0.3035,
+      "step": 11167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.089689353748739,
+      "learning_rate": 1.2546909439932858e-06,
+      "loss": 0.2651,
+      "step": 11168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.871998462415747,
+      "learning_rate": 1.253945158250701e-06,
+      "loss": 0.2932,
+      "step": 11169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.6368324016010054,
+      "learning_rate": 1.2531995624413434e-06,
+      "loss": 0.2881,
+      "step": 11170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.359106987688069,
+      "learning_rate": 1.2524541566030112e-06,
+      "loss": 0.305,
+      "step": 11171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.149929079313224,
+      "learning_rate": 1.2517089407735e-06,
+      "loss": 0.2749,
+      "step": 11172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.299763307213501,
+      "learning_rate": 1.2509639149905933e-06,
+      "loss": 0.3376,
+      "step": 11173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 52.37793729240282,
+      "learning_rate": 1.250219079292066e-06,
+      "loss": 0.3433,
+      "step": 11174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.848971754152898,
+      "learning_rate": 1.249474433715684e-06,
+      "loss": 0.318,
+      "step": 11175
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.063072818324455,
+      "learning_rate": 1.2487299782992001e-06,
+      "loss": 0.3195,
+      "step": 11176
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7546194701954323,
+      "learning_rate": 1.2479857130803607e-06,
+      "loss": 0.2997,
+      "step": 11177
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.187429857774741,
+      "learning_rate": 1.2472416380969026e-06,
+      "loss": 0.3059,
+      "step": 11178
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.29141933779020013,
+      "learning_rate": 1.2464977533865507e-06,
+      "loss": 0.1123,
+      "step": 11179
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.502863741010121,
+      "learning_rate": 1.2457540589870227e-06,
+      "loss": 0.2784,
+      "step": 11180
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.515264770680638,
+      "learning_rate": 1.2450105549360264e-06,
+      "loss": 0.3188,
+      "step": 11181
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.67896608427281,
+      "learning_rate": 1.2442672412712553e-06,
+      "loss": 0.3231,
+      "step": 11182
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.142341057912999,
+      "learning_rate": 1.243524118030402e-06,
+      "loss": 0.2813,
+      "step": 11183
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.919259880515476,
+      "learning_rate": 1.2427811852511396e-06,
+      "loss": 0.342,
+      "step": 11184
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.6483824255836,
+      "learning_rate": 1.2420384429711413e-06,
+      "loss": 0.3376,
+      "step": 11185
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.3937470059690105,
+      "learning_rate": 1.241295891228062e-06,
+      "loss": 0.2818,
+      "step": 11186
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.60087368859756,
+      "learning_rate": 1.2405535300595522e-06,
+      "loss": 0.3263,
+      "step": 11187
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5250478467132096,
+      "learning_rate": 1.239811359503253e-06,
+      "loss": 0.291,
+      "step": 11188
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.749903464289114,
+      "learning_rate": 1.2390693795967884e-06,
+      "loss": 0.3265,
+      "step": 11189
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3635794026269275,
+      "learning_rate": 1.2383275903777858e-06,
+      "loss": 0.2986,
+      "step": 11190
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5852213263026154,
+      "learning_rate": 1.23758599188385e-06,
+      "loss": 0.3269,
+      "step": 11191
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4168767133022784,
+      "learning_rate": 1.236844584152584e-06,
+      "loss": 0.3589,
+      "step": 11192
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.552144898282294,
+      "learning_rate": 1.2361033672215783e-06,
+      "loss": 0.3224,
+      "step": 11193
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.817302515205617,
+      "learning_rate": 1.2353623411284144e-06,
+      "loss": 0.3351,
+      "step": 11194
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 43.16813359030028,
+      "learning_rate": 1.2346215059106637e-06,
+      "loss": 0.338,
+      "step": 11195
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3397387607592908,
+      "learning_rate": 1.23388086160589e-06,
+      "loss": 0.3029,
+      "step": 11196
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.860368460119416,
+      "learning_rate": 1.2331404082516414e-06,
+      "loss": 0.2495,
+      "step": 11197
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.454710294281055,
+      "learning_rate": 1.2324001458854656e-06,
+      "loss": 0.3531,
+      "step": 11198
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2851507658632206,
+      "learning_rate": 1.2316600745448914e-06,
+      "loss": 0.1039,
+      "step": 11199
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8817139223913033,
+      "learning_rate": 1.2309201942674442e-06,
+      "loss": 0.2883,
+      "step": 11200
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.559341314695477,
+      "learning_rate": 1.2301805050906368e-06,
+      "loss": 0.341,
+      "step": 11201
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 25.71512560187791,
+      "learning_rate": 1.229441007051973e-06,
+      "loss": 0.2666,
+      "step": 11202
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.7303833749145574,
+      "learning_rate": 1.228701700188949e-06,
+      "loss": 0.3027,
+      "step": 11203
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.535221391047903,
+      "learning_rate": 1.2279625845390447e-06,
+      "loss": 0.3139,
+      "step": 11204
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9701261326228536,
+      "learning_rate": 1.2272236601397409e-06,
+      "loss": 0.3149,
+      "step": 11205
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.884721389401604,
+      "learning_rate": 1.2264849270284972e-06,
+      "loss": 0.3018,
+      "step": 11206
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.249714744173824,
+      "learning_rate": 1.2257463852427725e-06,
+      "loss": 0.299,
+      "step": 11207
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.532083469005002,
+      "learning_rate": 1.2250080348200105e-06,
+      "loss": 0.3221,
+      "step": 11208
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.092062885837227,
+      "learning_rate": 1.2242698757976496e-06,
+      "loss": 0.3361,
+      "step": 11209
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.317504719582757,
+      "learning_rate": 1.223531908213112e-06,
+      "loss": 0.3015,
+      "step": 11210
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2984601569994157,
+      "learning_rate": 1.2227941321038194e-06,
+      "loss": 0.1092,
+      "step": 11211
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2869643810684867,
+      "learning_rate": 1.2220565475071745e-06,
+      "loss": 0.1101,
+      "step": 11212
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.130281335545948,
+      "learning_rate": 1.2213191544605762e-06,
+      "loss": 0.2885,
+      "step": 11213
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4151024275358974,
+      "learning_rate": 1.2205819530014119e-06,
+      "loss": 0.3378,
+      "step": 11214
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.341440836916618,
+      "learning_rate": 1.2198449431670589e-06,
+      "loss": 0.3219,
+      "step": 11215
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.972266845388886,
+      "learning_rate": 1.2191081249948871e-06,
+      "loss": 0.3122,
+      "step": 11216
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.4883334697337847,
+      "learning_rate": 1.2183714985222507e-06,
+      "loss": 0.316,
+      "step": 11217
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.73062808976705,
+      "learning_rate": 1.2176350637865032e-06,
+      "loss": 0.3117,
+      "step": 11218
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9162922338911215,
+      "learning_rate": 1.21689882082498e-06,
+      "loss": 0.2967,
+      "step": 11219
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2915162545546786,
+      "learning_rate": 1.2161627696750112e-06,
+      "loss": 0.1126,
+      "step": 11220
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.615839307101408,
+      "learning_rate": 1.2154269103739164e-06,
+      "loss": 0.3111,
+      "step": 11221
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.756438027287479,
+      "learning_rate": 1.2146912429590045e-06,
+      "loss": 0.3173,
+      "step": 11222
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.733528763973156,
+      "learning_rate": 1.2139557674675773e-06,
+      "loss": 0.3173,
+      "step": 11223
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.767503783300545,
+      "learning_rate": 1.2132204839369243e-06,
+      "loss": 0.3111,
+      "step": 11224
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.811341312867231,
+      "learning_rate": 1.212485392404325e-06,
+      "loss": 0.3407,
+      "step": 11225
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.45745001414365,
+      "learning_rate": 1.21175049290705e-06,
+      "loss": 0.3383,
+      "step": 11226
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.958992490923388,
+      "learning_rate": 1.2110157854823617e-06,
+      "loss": 0.2936,
+      "step": 11227
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2918395754309663,
+      "learning_rate": 1.2102812701675104e-06,
+      "loss": 0.1089,
+      "step": 11228
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.26268078179791,
+      "learning_rate": 1.2095469469997395e-06,
+      "loss": 0.3042,
+      "step": 11229
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8353105086685266,
+      "learning_rate": 1.2088128160162765e-06,
+      "loss": 0.2924,
+      "step": 11230
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.285684193620308,
+      "learning_rate": 1.2080788772543494e-06,
+      "loss": 0.3193,
+      "step": 11231
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.32496337075430587,
+      "learning_rate": 1.2073451307511642e-06,
+      "loss": 0.1314,
+      "step": 11232
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 12.183175786149592,
+      "learning_rate": 1.2066115765439295e-06,
+      "loss": 0.3101,
+      "step": 11233
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.576087324614189,
+      "learning_rate": 1.205878214669834e-06,
+      "loss": 0.3057,
+      "step": 11234
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.424578231662367,
+      "learning_rate": 1.2051450451660618e-06,
+      "loss": 0.3022,
+      "step": 11235
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3020407028939699,
+      "learning_rate": 1.2044120680697862e-06,
+      "loss": 0.1148,
+      "step": 11236
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.851635972277392,
+      "learning_rate": 1.2036792834181716e-06,
+      "loss": 0.3095,
+      "step": 11237
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.114339721188008,
+      "learning_rate": 1.2029466912483717e-06,
+      "loss": 0.3012,
+      "step": 11238
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.041367120787563,
+      "learning_rate": 1.2022142915975294e-06,
+      "loss": 0.3285,
+      "step": 11239
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.623947242118053,
+      "learning_rate": 1.2014820845027791e-06,
+      "loss": 0.3177,
+      "step": 11240
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.468015142101128,
+      "learning_rate": 1.200750070001246e-06,
+      "loss": 0.3134,
+      "step": 11241
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.292750800114045,
+      "learning_rate": 1.2000182481300455e-06,
+      "loss": 0.3269,
+      "step": 11242
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.289022856890359,
+      "learning_rate": 1.199286618926279e-06,
+      "loss": 0.3192,
+      "step": 11243
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 8.99838176683705,
+      "learning_rate": 1.198555182427048e-06,
+      "loss": 0.295,
+      "step": 11244
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.630812675682412,
+      "learning_rate": 1.1978239386694307e-06,
+      "loss": 0.3122,
+      "step": 11245
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9804682462395307,
+      "learning_rate": 1.1970928876905096e-06,
+      "loss": 0.3478,
+      "step": 11246
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.328466497601953,
+      "learning_rate": 1.1963620295273459e-06,
+      "loss": 0.3411,
+      "step": 11247
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2890673520890768,
+      "learning_rate": 1.1956313642169974e-06,
+      "loss": 0.1126,
+      "step": 11248
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9685523930011084,
+      "learning_rate": 1.1949008917965105e-06,
+      "loss": 0.3268,
+      "step": 11249
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.129204484480842,
+      "learning_rate": 1.1941706123029213e-06,
+      "loss": 0.2974,
+      "step": 11250
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 9.315209679698482,
+      "learning_rate": 1.1934405257732579e-06,
+      "loss": 0.3157,
+      "step": 11251
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2714255483648813,
+      "learning_rate": 1.1927106322445347e-06,
+      "loss": 0.1001,
+      "step": 11252
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.8398308475849223,
+      "learning_rate": 1.1919809317537612e-06,
+      "loss": 0.319,
+      "step": 11253
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 13.064802530116909,
+      "learning_rate": 1.191251424337933e-06,
+      "loss": 0.3183,
+      "step": 11254
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.1182645185987194,
+      "learning_rate": 1.1905221100340397e-06,
+      "loss": 0.2759,
+      "step": 11255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.709911189545743,
+      "learning_rate": 1.1897929888790578e-06,
+      "loss": 0.2918,
+      "step": 11256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2711935920899564,
+      "learning_rate": 1.189064060909957e-06,
+      "loss": 0.3198,
+      "step": 11257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.2929484433794243,
+      "learning_rate": 1.1883353261636916e-06,
+      "loss": 0.3208,
+      "step": 11258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.326596928575784,
+      "learning_rate": 1.1876067846772155e-06,
+      "loss": 0.3563,
+      "step": 11259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.140139880688251,
+      "learning_rate": 1.1868784364874619e-06,
+      "loss": 0.3426,
+      "step": 11260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.870454419565292,
+      "learning_rate": 1.1861502816313652e-06,
+      "loss": 0.2975,
+      "step": 11261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.144333071058944,
+      "learning_rate": 1.1854223201458398e-06,
+      "loss": 0.3067,
+      "step": 11262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.728997729056664,
+      "learning_rate": 1.1846945520677972e-06,
+      "loss": 0.3088,
+      "step": 11263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 10.68794848371835,
+      "learning_rate": 1.1839669774341378e-06,
+      "loss": 0.3019,
+      "step": 11264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.672339986867618,
+      "learning_rate": 1.1832395962817473e-06,
+      "loss": 0.2712,
+      "step": 11265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.625199630437911,
+      "learning_rate": 1.1825124086475103e-06,
+      "loss": 0.366,
+      "step": 11266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3326474552044254,
+      "learning_rate": 1.1817854145682933e-06,
+      "loss": 0.3024,
+      "step": 11267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.705016263797846,
+      "learning_rate": 1.1810586140809582e-06,
+      "loss": 0.2793,
+      "step": 11268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.220145572415427,
+      "learning_rate": 1.1803320072223545e-06,
+      "loss": 0.2795,
+      "step": 11269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.1728819920386515,
+      "learning_rate": 1.1796055940293233e-06,
+      "loss": 0.313,
+      "step": 11270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.440516040495137,
+      "learning_rate": 1.1788793745386956e-06,
+      "loss": 0.328,
+      "step": 11271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.578213518514852,
+      "learning_rate": 1.1781533487872932e-06,
+      "loss": 0.303,
+      "step": 11272
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.22863987170897,
+      "learning_rate": 1.1774275168119232e-06,
+      "loss": 0.2904,
+      "step": 11273
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.999880409879227,
+      "learning_rate": 1.1767018786493921e-06,
+      "loss": 0.2956,
+      "step": 11274
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.615500397717607,
+      "learning_rate": 1.1759764343364882e-06,
+      "loss": 0.2941,
+      "step": 11275
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9881510722350266,
+      "learning_rate": 1.175251183909994e-06,
+      "loss": 0.343,
+      "step": 11276
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.282744137190041,
+      "learning_rate": 1.1745261274066816e-06,
+      "loss": 0.328,
+      "step": 11277
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.43059717843737,
+      "learning_rate": 1.1738012648633106e-06,
+      "loss": 0.323,
+      "step": 11278
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.3406803540093732,
+      "learning_rate": 1.1730765963166375e-06,
+      "loss": 0.3222,
+      "step": 11279
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.720062561910904,
+      "learning_rate": 1.1723521218034006e-06,
+      "loss": 0.3177,
+      "step": 11280
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.208058528158086,
+      "learning_rate": 1.1716278413603338e-06,
+      "loss": 0.3259,
+      "step": 11281
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.068848016500961,
+      "learning_rate": 1.1709037550241604e-06,
+      "loss": 0.2738,
+      "step": 11282
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 21.926966588003655,
+      "learning_rate": 1.170179862831592e-06,
+      "loss": 0.2894,
+      "step": 11283
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.044979002568594,
+      "learning_rate": 1.1694561648193326e-06,
+      "loss": 0.3313,
+      "step": 11284
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.390388894948781,
+      "learning_rate": 1.1687326610240758e-06,
+      "loss": 0.3156,
+      "step": 11285
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.099461611999541,
+      "learning_rate": 1.168009351482501e-06,
+      "loss": 0.3154,
+      "step": 11286
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 7.657646306791034,
+      "learning_rate": 1.167286236231287e-06,
+      "loss": 0.288,
+      "step": 11287
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.5164563431820777,
+      "learning_rate": 1.1665633153070937e-06,
+      "loss": 0.29,
+      "step": 11288
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.2828550324605537,
+      "learning_rate": 1.1658405887465757e-06,
+      "loss": 0.1062,
+      "step": 11289
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 6.107091081699184,
+      "learning_rate": 1.1651180565863767e-06,
+      "loss": 0.2859,
+      "step": 11290
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.061082028086048,
+      "learning_rate": 1.164395718863131e-06,
+      "loss": 0.3116,
+      "step": 11291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.156355336369214,
+      "learning_rate": 1.1636735756134642e-06,
+      "loss": 0.3069,
+      "step": 11292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.851635466622274,
+      "learning_rate": 1.1629516268739866e-06,
+      "loss": 0.3128,
+      "step": 11293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.202210603455869,
+      "learning_rate": 1.162229872681307e-06,
+      "loss": 0.3093,
+      "step": 11294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.136740966567996,
+      "learning_rate": 1.1615083130720166e-06,
+      "loss": 0.3363,
+      "step": 11295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3325679491717947,
+      "learning_rate": 1.1607869480827016e-06,
+      "loss": 0.1095,
+      "step": 11296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.881593025120685,
+      "learning_rate": 1.1600657777499369e-06,
+      "loss": 0.2849,
+      "step": 11297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3097903037203293,
+      "learning_rate": 1.1593448021102866e-06,
+      "loss": 0.2999,
+      "step": 11298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.874999300907032,
+      "learning_rate": 1.1586240212003057e-06,
+      "loss": 0.2785,
+      "step": 11299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4386396197588462,
+      "learning_rate": 1.157903435056542e-06,
+      "loss": 0.3176,
+      "step": 11300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8258434579272826,
+      "learning_rate": 1.1571830437155274e-06,
+      "loss": 0.3007,
+      "step": 11301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.424424901825066,
+      "learning_rate": 1.1564628472137884e-06,
+      "loss": 0.3161,
+      "step": 11302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.0890183833383835,
+      "learning_rate": 1.1557428455878405e-06,
+      "loss": 0.3407,
+      "step": 11303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.584595100931177,
+      "learning_rate": 1.1550230388741902e-06,
+      "loss": 0.3149,
+      "step": 11304
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7167732153033795,
+      "learning_rate": 1.1543034271093335e-06,
+      "loss": 0.2985,
+      "step": 11305
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28442707659407634,
+      "learning_rate": 1.153584010329753e-06,
+      "loss": 0.1019,
+      "step": 11306
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3351973683758599,
+      "learning_rate": 1.1528647885719302e-06,
+      "loss": 0.117,
+      "step": 11307
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2990817370489761,
+      "learning_rate": 1.1521457618723254e-06,
+      "loss": 0.1038,
+      "step": 11308
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.180620735360256,
+      "learning_rate": 1.1514269302674002e-06,
+      "loss": 0.3131,
+      "step": 11309
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.979153408286217,
+      "learning_rate": 1.150708293793597e-06,
+      "loss": 0.2925,
+      "step": 11310
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1966347057818285,
+      "learning_rate": 1.1499898524873542e-06,
+      "loss": 0.3072,
+      "step": 11311
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.555599311032655,
+      "learning_rate": 1.1492716063850973e-06,
+      "loss": 0.2999,
+      "step": 11312
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3527536979022567,
+      "learning_rate": 1.148553555523244e-06,
+      "loss": 0.3111,
+      "step": 11313
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.250928014466894,
+      "learning_rate": 1.1478356999382017e-06,
+      "loss": 0.3163,
+      "step": 11314
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3844852106364036,
+      "learning_rate": 1.147118039666365e-06,
+      "loss": 0.2955,
+      "step": 11315
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.159245767558844,
+      "learning_rate": 1.1464005747441215e-06,
+      "loss": 0.2814,
+      "step": 11316
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.332268385700682,
+      "learning_rate": 1.1456833052078492e-06,
+      "loss": 0.3121,
+      "step": 11317
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.489013116761907,
+      "learning_rate": 1.1449662310939159e-06,
+      "loss": 0.3438,
+      "step": 11318
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.207021225822313,
+      "learning_rate": 1.144249352438675e-06,
+      "loss": 0.3059,
+      "step": 11319
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.747474729900812,
+      "learning_rate": 1.1435326692784788e-06,
+      "loss": 0.3161,
+      "step": 11320
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7847665262338612,
+      "learning_rate": 1.1428161816496603e-06,
+      "loss": 0.3084,
+      "step": 11321
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.586021803918064,
+      "learning_rate": 1.1420998895885515e-06,
+      "loss": 0.3054,
+      "step": 11322
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6278986109835087,
+      "learning_rate": 1.1413837931314669e-06,
+      "loss": 0.3041,
+      "step": 11323
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.108103796389431,
+      "learning_rate": 1.1406678923147142e-06,
+      "loss": 0.2778,
+      "step": 11324
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6949373800357392,
+      "learning_rate": 1.1399521871745923e-06,
+      "loss": 0.2965,
+      "step": 11325
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.792758833945904,
+      "learning_rate": 1.1392366777473884e-06,
+      "loss": 0.3071,
+      "step": 11326
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.189211833399106,
+      "learning_rate": 1.1385213640693821e-06,
+      "loss": 0.3202,
+      "step": 11327
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.294790818381229,
+      "learning_rate": 1.137806246176838e-06,
+      "loss": 0.3435,
+      "step": 11328
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.052675839887932,
+      "learning_rate": 1.137091324106016e-06,
+      "loss": 0.2829,
+      "step": 11329
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.356547053796238,
+      "learning_rate": 1.1363765978931646e-06,
+      "loss": 0.3058,
+      "step": 11330
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6275674093239925,
+      "learning_rate": 1.1356620675745216e-06,
+      "loss": 0.3171,
+      "step": 11331
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.601040867751448,
+      "learning_rate": 1.134947733186315e-06,
+      "loss": 0.3147,
+      "step": 11332
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.919050842477509,
+      "learning_rate": 1.1342335947647653e-06,
+      "loss": 0.3072,
+      "step": 11333
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.3344961743442045,
+      "learning_rate": 1.1335196523460756e-06,
+      "loss": 0.2991,
+      "step": 11334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.390689183700898,
+      "learning_rate": 1.132805905966451e-06,
+      "loss": 0.306,
+      "step": 11335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1038968597027075,
+      "learning_rate": 1.1320923556620738e-06,
+      "loss": 0.3133,
+      "step": 11336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5572088786756892,
+      "learning_rate": 1.1313790014691284e-06,
+      "loss": 0.3391,
+      "step": 11337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.537940870891999,
+      "learning_rate": 1.1306658434237787e-06,
+      "loss": 0.3163,
+      "step": 11338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6537539266170755,
+      "learning_rate": 1.1299528815621858e-06,
+      "loss": 0.3255,
+      "step": 11339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.34459996190502923,
+      "learning_rate": 1.1292401159204985e-06,
+      "loss": 0.1006,
+      "step": 11340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3350368439223693,
+      "learning_rate": 1.1285275465348527e-06,
+      "loss": 0.122,
+      "step": 11341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6679932722319055,
+      "learning_rate": 1.1278151734413822e-06,
+      "loss": 0.3057,
+      "step": 11342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.376391992732698,
+      "learning_rate": 1.1271029966762021e-06,
+      "loss": 0.2951,
+      "step": 11343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.707793028637803,
+      "learning_rate": 1.1263910162754222e-06,
+      "loss": 0.3243,
+      "step": 11344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.349151092427887,
+      "learning_rate": 1.1256792322751414e-06,
+      "loss": 0.3223,
+      "step": 11345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.4548705158847985,
+      "learning_rate": 1.1249676447114494e-06,
+      "loss": 0.3004,
+      "step": 11346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2796442448144403,
+      "learning_rate": 1.1242562536204243e-06,
+      "loss": 0.1109,
+      "step": 11347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.2940481693350687,
+      "learning_rate": 1.1235450590381374e-06,
+      "loss": 0.3197,
+      "step": 11348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.957268997888687,
+      "learning_rate": 1.1228340610006434e-06,
+      "loss": 0.3372,
+      "step": 11349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2895034951979068,
+      "learning_rate": 1.1221232595439974e-06,
+      "loss": 0.112,
+      "step": 11350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.482053206468831,
+      "learning_rate": 1.1214126547042337e-06,
+      "loss": 0.3086,
+      "step": 11351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.822473279744039,
+      "learning_rate": 1.1207022465173828e-06,
+      "loss": 0.2845,
+      "step": 11352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 11.095045928543332,
+      "learning_rate": 1.1199920350194666e-06,
+      "loss": 0.3404,
+      "step": 11353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.394674259683042,
+      "learning_rate": 1.1192820202464887e-06,
+      "loss": 0.3335,
+      "step": 11354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.422083377527473,
+      "learning_rate": 1.1185722022344553e-06,
+      "loss": 0.3071,
+      "step": 11355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.080202887591934,
+      "learning_rate": 1.1178625810193505e-06,
+      "loss": 0.281,
+      "step": 11356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.378447698808015,
+      "learning_rate": 1.1171531566371557e-06,
+      "loss": 0.3119,
+      "step": 11357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7809827160779252,
+      "learning_rate": 1.1164439291238398e-06,
+      "loss": 0.3032,
+      "step": 11358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.29876639892729634,
+      "learning_rate": 1.1157348985153631e-06,
+      "loss": 0.106,
+      "step": 11359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30901912039925633,
+      "learning_rate": 1.1150260648476742e-06,
+      "loss": 0.1023,
+      "step": 11360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.023614503694159,
+      "learning_rate": 1.1143174281567138e-06,
+      "loss": 0.2997,
+      "step": 11361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.335110055064265,
+      "learning_rate": 1.1136089884784074e-06,
+      "loss": 0.2872,
+      "step": 11362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.6684110725679546,
+      "learning_rate": 1.1129007458486807e-06,
+      "loss": 0.3149,
+      "step": 11363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.082451521893021,
+      "learning_rate": 1.1121927003034378e-06,
+      "loss": 0.3029,
+      "step": 11364
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.284510055297986,
+      "learning_rate": 1.1114848518785803e-06,
+      "loss": 0.3308,
+      "step": 11365
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.588553250377081,
+      "learning_rate": 1.1107772006099987e-06,
+      "loss": 0.3302,
+      "step": 11366
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0164427616032374,
+      "learning_rate": 1.110069746533569e-06,
+      "loss": 0.305,
+      "step": 11367
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.105768442885513,
+      "learning_rate": 1.1093624896851657e-06,
+      "loss": 0.3228,
+      "step": 11368
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.11174915190196,
+      "learning_rate": 1.1086554301006425e-06,
+      "loss": 0.306,
+      "step": 11369
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.666199371863643,
+      "learning_rate": 1.1079485678158546e-06,
+      "loss": 0.3187,
+      "step": 11370
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 10.880795768828426,
+      "learning_rate": 1.1072419028666377e-06,
+      "loss": 0.3285,
+      "step": 11371
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.703175862428405,
+      "learning_rate": 1.106535435288822e-06,
+      "loss": 0.3,
+      "step": 11372
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 13.49845620065891,
+      "learning_rate": 1.1058291651182273e-06,
+      "loss": 0.3184,
+      "step": 11373
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3106103143083305,
+      "learning_rate": 1.1051230923906636e-06,
+      "loss": 0.3203,
+      "step": 11374
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.3181147511342735,
+      "learning_rate": 1.1044172171419293e-06,
+      "loss": 0.3332,
+      "step": 11375
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 25.399001714442875,
+      "learning_rate": 1.1037115394078162e-06,
+      "loss": 0.3184,
+      "step": 11376
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 9.414347521897428,
+      "learning_rate": 1.1030060592240999e-06,
+      "loss": 0.3178,
+      "step": 11377
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 8.434651864484891,
+      "learning_rate": 1.102300776626552e-06,
+      "loss": 0.2722,
+      "step": 11378
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.676280957851967,
+      "learning_rate": 1.1015956916509318e-06,
+      "loss": 0.2935,
+      "step": 11379
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.571250321884696,
+      "learning_rate": 1.1008908043329886e-06,
+      "loss": 0.3308,
+      "step": 11380
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.2653783358131925,
+      "learning_rate": 1.1001861147084625e-06,
+      "loss": 0.2907,
+      "step": 11381
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.228327380088267,
+      "learning_rate": 1.0994816228130795e-06,
+      "loss": 0.3028,
+      "step": 11382
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.171398277016998,
+      "learning_rate": 1.0987773286825642e-06,
+      "loss": 0.2942,
+      "step": 11383
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.856861897374272,
+      "learning_rate": 1.09807323235262e-06,
+      "loss": 0.2873,
+      "step": 11384
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1258918478636737,
+      "learning_rate": 1.0973693338589519e-06,
+      "loss": 0.3381,
+      "step": 11385
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6309905262508986,
+      "learning_rate": 1.096665633237245e-06,
+      "loss": 0.2997,
+      "step": 11386
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6691283488511606,
+      "learning_rate": 1.0959621305231799e-06,
+      "loss": 0.2945,
+      "step": 11387
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8150968293720133,
+      "learning_rate": 1.0952588257524255e-06,
+      "loss": 0.2914,
+      "step": 11388
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.812925147212097,
+      "learning_rate": 1.0945557189606427e-06,
+      "loss": 0.3101,
+      "step": 11389
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.235002257995327,
+      "learning_rate": 1.0938528101834772e-06,
+      "loss": 0.3013,
+      "step": 11390
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.256934215359167,
+      "learning_rate": 1.0931500994565697e-06,
+      "loss": 0.2923,
+      "step": 11391
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6487472885188827,
+      "learning_rate": 1.0924475868155493e-06,
+      "loss": 0.3209,
+      "step": 11392
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.801166837694886,
+      "learning_rate": 1.0917452722960352e-06,
+      "loss": 0.2752,
+      "step": 11393
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.063260650943964,
+      "learning_rate": 1.0910431559336372e-06,
+      "loss": 0.3404,
+      "step": 11394
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.9580179808572926,
+      "learning_rate": 1.0903412377639493e-06,
+      "loss": 0.3104,
+      "step": 11395
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 18.355045134899974,
+      "learning_rate": 1.0896395178225677e-06,
+      "loss": 0.2973,
+      "step": 11396
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6489073596363215,
+      "learning_rate": 1.0889379961450642e-06,
+      "loss": 0.3181,
+      "step": 11397
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 12.513833880216911,
+      "learning_rate": 1.088236672767013e-06,
+      "loss": 0.2946,
+      "step": 11398
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.227527930098849,
+      "learning_rate": 1.087535547723969e-06,
+      "loss": 0.374,
+      "step": 11399
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.421267106555349,
+      "learning_rate": 1.0868346210514825e-06,
+      "loss": 0.2642,
+      "step": 11400
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.28082258369163554,
+      "learning_rate": 1.086133892785091e-06,
+      "loss": 0.1073,
+      "step": 11401
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.545449448820383,
+      "learning_rate": 1.0854333629603248e-06,
+      "loss": 0.2925,
+      "step": 11402
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.109686095040652,
+      "learning_rate": 1.0847330316127013e-06,
+      "loss": 0.3449,
+      "step": 11403
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.0573458511219287,
+      "learning_rate": 1.0840328987777283e-06,
+      "loss": 0.3244,
+      "step": 11404
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9712597257433178,
+      "learning_rate": 1.0833329644909035e-06,
+      "loss": 0.3054,
+      "step": 11405
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.492647886621743,
+      "learning_rate": 1.0826332287877168e-06,
+      "loss": 0.3193,
+      "step": 11406
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.30551868882969924,
+      "learning_rate": 1.081933691703645e-06,
+      "loss": 0.115,
+      "step": 11407
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.116289405869235,
+      "learning_rate": 1.081234353274157e-06,
+      "loss": 0.2813,
+      "step": 11408
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.500483034753759,
+      "learning_rate": 1.080535213534712e-06,
+      "loss": 0.277,
+      "step": 11409
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.462889811159573,
+      "learning_rate": 1.079836272520754e-06,
+      "loss": 0.2932,
+      "step": 11410
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.910132261694512,
+      "learning_rate": 1.0791375302677255e-06,
+      "loss": 0.31,
+      "step": 11411
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6753000690672977,
+      "learning_rate": 1.0784389868110513e-06,
+      "loss": 0.3038,
+      "step": 11412
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.1994552273441044,
+      "learning_rate": 1.0777406421861492e-06,
+      "loss": 0.318,
+      "step": 11413
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3993725143442615,
+      "learning_rate": 1.0770424964284276e-06,
+      "loss": 0.2948,
+      "step": 11414
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.2247079241804295,
+      "learning_rate": 1.0763445495732843e-06,
+      "loss": 0.3247,
+      "step": 11415
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.624745924690897,
+      "learning_rate": 1.0756468016561079e-06,
+      "loss": 0.3269,
+      "step": 11416
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.253684185874686,
+      "learning_rate": 1.0749492527122713e-06,
+      "loss": 0.3061,
+      "step": 11417
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.637784429660397,
+      "learning_rate": 1.074251902777147e-06,
+      "loss": 0.3316,
+      "step": 11418
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7941553405618715,
+      "learning_rate": 1.0735547518860888e-06,
+      "loss": 0.3168,
+      "step": 11419
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.695909040752955,
+      "learning_rate": 1.072857800074445e-06,
+      "loss": 0.3118,
+      "step": 11420
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.437890139008782,
+      "learning_rate": 1.0721610473775517e-06,
+      "loss": 0.3118,
+      "step": 11421
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.869279430522249,
+      "learning_rate": 1.0714644938307372e-06,
+      "loss": 0.2676,
+      "step": 11422
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.1433471548020675,
+      "learning_rate": 1.0707681394693176e-06,
+      "loss": 0.3118,
+      "step": 11423
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.607383747629235,
+      "learning_rate": 1.070071984328601e-06,
+      "loss": 0.3301,
+      "step": 11424
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.7060716772341826,
+      "learning_rate": 1.0693760284438797e-06,
+      "loss": 0.3108,
+      "step": 11425
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.324562320652588,
+      "learning_rate": 1.0686802718504458e-06,
+      "loss": 0.3065,
+      "step": 11426
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3062270267342115,
+      "learning_rate": 1.067984714583572e-06,
+      "loss": 0.2965,
+      "step": 11427
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3024485165965887,
+      "learning_rate": 1.0672893566785253e-06,
+      "loss": 0.1024,
+      "step": 11428
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.143774946164367,
+      "learning_rate": 1.0665941981705634e-06,
+      "loss": 0.3208,
+      "step": 11429
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.78999723559272,
+      "learning_rate": 1.065899239094929e-06,
+      "loss": 0.2724,
+      "step": 11430
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.9391383768615644,
+      "learning_rate": 1.065204479486862e-06,
+      "loss": 0.311,
+      "step": 11431
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.863070063489409,
+      "learning_rate": 1.064509919381586e-06,
+      "loss": 0.3053,
+      "step": 11432
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.5831433094192087,
+      "learning_rate": 1.0638155588143167e-06,
+      "loss": 0.3047,
+      "step": 11433
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.2936057258636268,
+      "learning_rate": 1.0631213978202604e-06,
+      "loss": 0.0985,
+      "step": 11434
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.6789021286523442,
+      "learning_rate": 1.0624274364346122e-06,
+      "loss": 0.3126,
+      "step": 11435
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 62.421113209175275,
+      "learning_rate": 1.0617336746925571e-06,
+      "loss": 0.2995,
+      "step": 11436
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.284164418213308,
+      "learning_rate": 1.061040112629273e-06,
+      "loss": 0.3189,
+      "step": 11437
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3893993974678733,
+      "learning_rate": 1.0603467502799198e-06,
+      "loss": 0.3106,
+      "step": 11438
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.160934366578991,
+      "learning_rate": 1.0596535876796583e-06,
+      "loss": 0.2823,
+      "step": 11439
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.691208444126816,
+      "learning_rate": 1.0589606248636291e-06,
+      "loss": 0.3198,
+      "step": 11440
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.114072516531486,
+      "learning_rate": 1.058267861866969e-06,
+      "loss": 0.3053,
+      "step": 11441
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.621319598054393,
+      "learning_rate": 1.0575752987248034e-06,
+      "loss": 0.3508,
+      "step": 11442
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6147903714065963,
+      "learning_rate": 1.0568829354722427e-06,
+      "loss": 0.3067,
+      "step": 11443
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.360898579737109,
+      "learning_rate": 1.0561907721443964e-06,
+      "loss": 0.3173,
+      "step": 11444
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.639439243616901,
+      "learning_rate": 1.055498808776354e-06,
+      "loss": 0.3226,
+      "step": 11445
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.2105703492013795,
+      "learning_rate": 1.0548070454032045e-06,
+      "loss": 0.3035,
+      "step": 11446
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.04334706096826,
+      "learning_rate": 1.0541154820600174e-06,
+      "loss": 0.3164,
+      "step": 11447
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.206935803908317,
+      "learning_rate": 1.0534241187818588e-06,
+      "loss": 0.3025,
+      "step": 11448
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.187262601671817,
+      "learning_rate": 1.052732955603782e-06,
+      "loss": 0.331,
+      "step": 11449
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.722425804773873,
+      "learning_rate": 1.05204199256083e-06,
+      "loss": 0.2789,
+      "step": 11450
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7443966643644515,
+      "learning_rate": 1.0513512296880362e-06,
+      "loss": 0.2875,
+      "step": 11451
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.640012141749643,
+      "learning_rate": 1.050660667020426e-06,
+      "loss": 0.3127,
+      "step": 11452
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3630428584350742,
+      "learning_rate": 1.0499703045930088e-06,
+      "loss": 0.2855,
+      "step": 11453
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.355305223273439,
+      "learning_rate": 1.0492801424407894e-06,
+      "loss": 0.3213,
+      "step": 11454
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.616709724659639,
+      "learning_rate": 1.0485901805987604e-06,
+      "loss": 0.2724,
+      "step": 11455
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5099518199461164,
+      "learning_rate": 1.0479004191019043e-06,
+      "loss": 0.3054,
+      "step": 11456
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.427821111561163,
+      "learning_rate": 1.0472108579851948e-06,
+      "loss": 0.3262,
+      "step": 11457
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.347660169642433,
+      "learning_rate": 1.0465214972835902e-06,
+      "loss": 0.3083,
+      "step": 11458
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0083262401070052,
+      "learning_rate": 1.0458323370320484e-06,
+      "loss": 0.3362,
+      "step": 11459
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7597624062524044,
+      "learning_rate": 1.0451433772655068e-06,
+      "loss": 0.3213,
+      "step": 11460
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.537532616685377,
+      "learning_rate": 1.0444546180188986e-06,
+      "loss": 0.3705,
+      "step": 11461
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.427632314087793,
+      "learning_rate": 1.0437660593271454e-06,
+      "loss": 0.3117,
+      "step": 11462
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 29.391867563590456,
+      "learning_rate": 1.0430777012251586e-06,
+      "loss": 0.3324,
+      "step": 11463
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.837838690054331,
+      "learning_rate": 1.0423895437478393e-06,
+      "loss": 0.3072,
+      "step": 11464
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.681583739832827,
+      "learning_rate": 1.0417015869300807e-06,
+      "loss": 0.3326,
+      "step": 11465
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30853494473022736,
+      "learning_rate": 1.0410138308067608e-06,
+      "loss": 0.1106,
+      "step": 11466
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3546645533620763,
+      "learning_rate": 1.0403262754127508e-06,
+      "loss": 0.1237,
+      "step": 11467
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.028740185767819,
+      "learning_rate": 1.0396389207829122e-06,
+      "loss": 0.3015,
+      "step": 11468
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.471448317295474,
+      "learning_rate": 1.0389517669520954e-06,
+      "loss": 0.2748,
+      "step": 11469
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5054403824393257,
+      "learning_rate": 1.038264813955141e-06,
+      "loss": 0.3306,
+      "step": 11470
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0616682369636186,
+      "learning_rate": 1.037578061826876e-06,
+      "loss": 0.3248,
+      "step": 11471
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.10113799272256,
+      "learning_rate": 1.0368915106021255e-06,
+      "loss": 0.3151,
+      "step": 11472
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.283977050298275,
+      "learning_rate": 1.0362051603156936e-06,
+      "loss": 0.312,
+      "step": 11473
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9968785796365314,
+      "learning_rate": 1.035519011002385e-06,
+      "loss": 0.3112,
+      "step": 11474
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.891905241237328,
+      "learning_rate": 1.034833062696985e-06,
+      "loss": 0.3104,
+      "step": 11475
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.71040737593561,
+      "learning_rate": 1.034147315434274e-06,
+      "loss": 0.2975,
+      "step": 11476
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.2964284760503754,
+      "learning_rate": 1.0334617692490205e-06,
+      "loss": 0.11,
+      "step": 11477
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.66661481413364,
+      "learning_rate": 1.0327764241759848e-06,
+      "loss": 0.3368,
+      "step": 11478
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4871045429292318,
+      "learning_rate": 1.032091280249915e-06,
+      "loss": 0.3292,
+      "step": 11479
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.688009474508224,
+      "learning_rate": 1.0314063375055472e-06,
+      "loss": 0.3277,
+      "step": 11480
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.832865621991346,
+      "learning_rate": 1.0307215959776111e-06,
+      "loss": 0.3086,
+      "step": 11481
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.037622994845634,
+      "learning_rate": 1.0300370557008248e-06,
+      "loss": 0.3003,
+      "step": 11482
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 17.93731829246575,
+      "learning_rate": 1.0293527167098954e-06,
+      "loss": 0.3142,
+      "step": 11483
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.897839297814308,
+      "learning_rate": 1.0286685790395211e-06,
+      "loss": 0.32,
+      "step": 11484
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.099067203970907,
+      "learning_rate": 1.0279846427243901e-06,
+      "loss": 0.2948,
+      "step": 11485
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.096331480765508,
+      "learning_rate": 1.0273009077991746e-06,
+      "loss": 0.3244,
+      "step": 11486
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.295504951867523,
+      "learning_rate": 1.0266173742985486e-06,
+      "loss": 0.102,
+      "step": 11487
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8741192553876704,
+      "learning_rate": 1.0259340422571635e-06,
+      "loss": 0.3301,
+      "step": 11488
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3606890518315358,
+      "learning_rate": 1.0252509117096676e-06,
+      "loss": 0.3205,
+      "step": 11489
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.3254088714756325,
+      "learning_rate": 1.0245679826906967e-06,
+      "loss": 0.2695,
+      "step": 11490
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3797350320688295,
+      "learning_rate": 1.0238852552348772e-06,
+      "loss": 0.3053,
+      "step": 11491
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.549033039328853,
+      "learning_rate": 1.023202729376826e-06,
+      "loss": 0.3129,
+      "step": 11492
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.236136855063349,
+      "learning_rate": 1.0225204051511445e-06,
+      "loss": 0.2962,
+      "step": 11493
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.335628924126996,
+      "learning_rate": 1.021838282592434e-06,
+      "loss": 0.3252,
+      "step": 11494
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.750965041782733,
+      "learning_rate": 1.0211563617352748e-06,
+      "loss": 0.3069,
+      "step": 11495
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.863148652781699,
+      "learning_rate": 1.0204746426142437e-06,
+      "loss": 0.3044,
+      "step": 11496
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.518876716390625,
+      "learning_rate": 1.0197931252639053e-06,
+      "loss": 0.3123,
+      "step": 11497
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9781619615313666,
+      "learning_rate": 1.019111809718815e-06,
+      "loss": 0.3023,
+      "step": 11498
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8908265150042705,
+      "learning_rate": 1.0184306960135132e-06,
+      "loss": 0.2743,
+      "step": 11499
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.786124749638766,
+      "learning_rate": 1.0177497841825396e-06,
+      "loss": 0.342,
+      "step": 11500
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.825959391366538,
+      "learning_rate": 1.0170690742604133e-06,
+      "loss": 0.3018,
+      "step": 11501
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.797191157948074,
+      "learning_rate": 1.0163885662816496e-06,
+      "loss": 0.3087,
+      "step": 11502
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.411525245975824,
+      "learning_rate": 1.0157082602807516e-06,
+      "loss": 0.3122,
+      "step": 11503
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 51.71106530023641,
+      "learning_rate": 1.015028156292212e-06,
+      "loss": 0.2814,
+      "step": 11504
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.1959587690335765,
+      "learning_rate": 1.014348254350515e-06,
+      "loss": 0.3493,
+      "step": 11505
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4318345896179814,
+      "learning_rate": 1.0136685544901299e-06,
+      "loss": 0.3532,
+      "step": 11506
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.842362977269109,
+      "learning_rate": 1.012989056745523e-06,
+      "loss": 0.3086,
+      "step": 11507
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.41907796428397,
+      "learning_rate": 1.0123097611511435e-06,
+      "loss": 0.3054,
+      "step": 11508
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.293418950720074,
+      "learning_rate": 1.011630667741434e-06,
+      "loss": 0.313,
+      "step": 11509
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.140465207664112,
+      "learning_rate": 1.0109517765508259e-06,
+      "loss": 0.3481,
+      "step": 11510
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.60021511512961,
+      "learning_rate": 1.0102730876137406e-06,
+      "loss": 0.3459,
+      "step": 11511
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4176956694608807,
+      "learning_rate": 1.0095946009645896e-06,
+      "loss": 0.3181,
+      "step": 11512
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.026987776299547,
+      "learning_rate": 1.0089163166377742e-06,
+      "loss": 0.3374,
+      "step": 11513
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.289349726558382,
+      "learning_rate": 1.008238234667681e-06,
+      "loss": 0.2844,
+      "step": 11514
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.221230009945828,
+      "learning_rate": 1.0075603550886964e-06,
+      "loss": 0.3283,
+      "step": 11515
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2797022752715437,
+      "learning_rate": 1.006882677935186e-06,
+      "loss": 0.3077,
+      "step": 11516
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.248440983780271,
+      "learning_rate": 1.0062052032415104e-06,
+      "loss": 0.3088,
+      "step": 11517
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8438976514720027,
+      "learning_rate": 1.0055279310420208e-06,
+      "loss": 0.3263,
+      "step": 11518
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.80817841888214,
+      "learning_rate": 1.0048508613710517e-06,
+      "loss": 0.2998,
+      "step": 11519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.038617290867975,
+      "learning_rate": 1.0041739942629387e-06,
+      "loss": 0.3175,
+      "step": 11520
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.889849228663778,
+      "learning_rate": 1.003497329751994e-06,
+      "loss": 0.3167,
+      "step": 11521
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.2464181097612705,
+      "learning_rate": 1.0028208678725315e-06,
+      "loss": 0.2686,
+      "step": 11522
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6533972933428505,
+      "learning_rate": 1.002144608658846e-06,
+      "loss": 0.3392,
+      "step": 11523
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.934491551074859,
+      "learning_rate": 1.0014685521452256e-06,
+      "loss": 0.3052,
+      "step": 11524
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.290863513049103,
+      "learning_rate": 1.000792698365949e-06,
+      "loss": 0.3226,
+      "step": 11525
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.682043835731607,
+      "learning_rate": 1.0001170473552823e-06,
+      "loss": 0.3214,
+      "step": 11526
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.596810287561436,
+      "learning_rate": 9.994415991474832e-07,
+      "loss": 0.3208,
+      "step": 11527
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.181288323189729,
+      "learning_rate": 9.987663537767994e-07,
+      "loss": 0.325,
+      "step": 11528
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3874263952378985,
+      "learning_rate": 9.980913112774653e-07,
+      "loss": 0.2923,
+      "step": 11529
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.39580402227652,
+      "learning_rate": 9.974164716837082e-07,
+      "loss": 0.2888,
+      "step": 11530
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.3248413133482515,
+      "learning_rate": 9.96741835029743e-07,
+      "loss": 0.3219,
+      "step": 11531
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9353737790740757,
+      "learning_rate": 9.960674013497762e-07,
+      "loss": 0.3171,
+      "step": 11532
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.15427249728498,
+      "learning_rate": 9.95393170678004e-07,
+      "loss": 0.3272,
+      "step": 11533
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 16.39610105459704,
+      "learning_rate": 9.947191430486081e-07,
+      "loss": 0.3125,
+      "step": 11534
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.360861333879012,
+      "learning_rate": 9.940453184957671e-07,
+      "loss": 0.3056,
+      "step": 11535
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.995321155558662,
+      "learning_rate": 9.933716970536428e-07,
+      "loss": 0.3239,
+      "step": 11536
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4085245820905605,
+      "learning_rate": 9.9269827875639e-07,
+      "loss": 0.3175,
+      "step": 11537
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.913736966654049,
+      "learning_rate": 9.920250636381529e-07,
+      "loss": 0.3375,
+      "step": 11538
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.497943704227452,
+      "learning_rate": 9.913520517330638e-07,
+      "loss": 0.3168,
+      "step": 11539
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4252693779771066,
+      "learning_rate": 9.906792430752477e-07,
+      "loss": 0.3435,
+      "step": 11540
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.0521816389175283,
+      "learning_rate": 9.90006637698817e-07,
+      "loss": 0.313,
+      "step": 11541
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.916226476943896,
+      "learning_rate": 9.893342356378728e-07,
+      "loss": 0.3039,
+      "step": 11542
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.398711756319213,
+      "learning_rate": 9.886620369265082e-07,
+      "loss": 0.3055,
+      "step": 11543
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.993314127155125,
+      "learning_rate": 9.879900415988058e-07,
+      "loss": 0.2903,
+      "step": 11544
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 11.137802549915815,
+      "learning_rate": 9.87318249688836e-07,
+      "loss": 0.3012,
+      "step": 11545
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.6310661862797198,
+      "learning_rate": 9.86646661230663e-07,
+      "loss": 0.325,
+      "step": 11546
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.6996882758460865,
+      "learning_rate": 9.859752762583324e-07,
+      "loss": 0.335,
+      "step": 11547
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.419220151412088,
+      "learning_rate": 9.853040948058912e-07,
+      "loss": 0.3127,
+      "step": 11548
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8823422493871047,
+      "learning_rate": 9.846331169073636e-07,
+      "loss": 0.3326,
+      "step": 11549
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8468805493850624,
+      "learning_rate": 9.83962342596776e-07,
+      "loss": 0.3258,
+      "step": 11550
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.677825140478026,
+      "learning_rate": 9.832917719081335e-07,
+      "loss": 0.3009,
+      "step": 11551
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.5086114232118595,
+      "learning_rate": 9.826214048754368e-07,
+      "loss": 0.3059,
+      "step": 11552
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.3949563454597875,
+      "learning_rate": 9.819512415326754e-07,
+      "loss": 0.3482,
+      "step": 11553
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.944367993421723,
+      "learning_rate": 9.812812819138274e-07,
+      "loss": 0.3013,
+      "step": 11554
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 8.278376581086597,
+      "learning_rate": 9.806115260528632e-07,
+      "loss": 0.2923,
+      "step": 11555
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.30793292870231614,
+      "learning_rate": 9.79941973983738e-07,
+      "loss": 0.1006,
+      "step": 11556
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.345545964756256,
+      "learning_rate": 9.792726257404006e-07,
+      "loss": 0.3051,
+      "step": 11557
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4563554173928592,
+      "learning_rate": 9.78603481356789e-07,
+      "loss": 0.3367,
+      "step": 11558
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.7076076024459885,
+      "learning_rate": 9.779345408668295e-07,
+      "loss": 0.3206,
+      "step": 11559
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.029872011016469,
+      "learning_rate": 9.772658043044397e-07,
+      "loss": 0.2987,
+      "step": 11560
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.031092596729829,
+      "learning_rate": 9.765972717035272e-07,
+      "loss": 0.2995,
+      "step": 11561
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.831385565968861,
+      "learning_rate": 9.75928943097984e-07,
+      "loss": 0.2862,
+      "step": 11562
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2820088426715333,
+      "learning_rate": 9.752608185217005e-07,
+      "loss": 0.2917,
+      "step": 11563
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.0933308982127325,
+      "learning_rate": 9.74592898008549e-07,
+      "loss": 0.3136,
+      "step": 11564
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.807621473804722,
+      "learning_rate": 9.739251815923957e-07,
+      "loss": 0.3449,
+      "step": 11565
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.4925825972695685,
+      "learning_rate": 9.732576693070945e-07,
+      "loss": 0.3299,
+      "step": 11566
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.773826544697552,
+      "learning_rate": 9.725903611864907e-07,
+      "loss": 0.3307,
+      "step": 11567
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.8542775626343335,
+      "learning_rate": 9.719232572644189e-07,
+      "loss": 0.3167,
+      "step": 11568
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.404694424867003,
+      "learning_rate": 9.712563575747007e-07,
+      "loss": 0.2906,
+      "step": 11569
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.4190613329095183,
+      "learning_rate": 9.705896621511508e-07,
+      "loss": 0.3276,
+      "step": 11570
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8527148562697544,
+      "learning_rate": 9.699231710275715e-07,
+      "loss": 0.3056,
+      "step": 11571
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2850843436635326,
+      "learning_rate": 9.692568842377552e-07,
+      "loss": 0.2937,
+      "step": 11572
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9987550402837644,
+      "learning_rate": 9.68590801815485e-07,
+      "loss": 0.313,
+      "step": 11573
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 6.39046131462795,
+      "learning_rate": 9.679249237945332e-07,
+      "loss": 0.2919,
+      "step": 11574
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.5868176410958115,
+      "learning_rate": 9.672592502086581e-07,
+      "loss": 0.2983,
+      "step": 11575
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.632689055754061,
+      "learning_rate": 9.66593781091616e-07,
+      "loss": 0.3307,
+      "step": 11576
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.312374299022882,
+      "learning_rate": 9.659285164771426e-07,
+      "loss": 0.2897,
+      "step": 11577
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.096737034457407,
+      "learning_rate": 9.65263456398971e-07,
+      "loss": 0.3144,
+      "step": 11578
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.9075227319463215,
+      "learning_rate": 9.645986008908204e-07,
+      "loss": 0.2945,
+      "step": 11579
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 17.336063757467308,
+      "learning_rate": 9.639339499864008e-07,
+      "loss": 0.3298,
+      "step": 11580
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4578970188045854,
+      "learning_rate": 9.632695037194128e-07,
+      "loss": 0.3479,
+      "step": 11581
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.502453823603,
+      "learning_rate": 9.62605262123541e-07,
+      "loss": 0.337,
+      "step": 11582
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.584905807772009,
+      "learning_rate": 9.619412252324688e-07,
+      "loss": 0.3304,
+      "step": 11583
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3292191716655086,
+      "learning_rate": 9.61277393079862e-07,
+      "loss": 0.2947,
+      "step": 11584
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.3171288405356524,
+      "learning_rate": 9.606137656993776e-07,
+      "loss": 0.1141,
+      "step": 11585
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.6140450529605745,
+      "learning_rate": 9.599503431246643e-07,
+      "loss": 0.3234,
+      "step": 11586
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.368669207291378,
+      "learning_rate": 9.592871253893587e-07,
+      "loss": 0.2957,
+      "step": 11587
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.400946357225676,
+      "learning_rate": 9.586241125270879e-07,
+      "loss": 0.3206,
+      "step": 11588
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.678865679338457,
+      "learning_rate": 9.579613045714686e-07,
+      "loss": 0.3172,
+      "step": 11589
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.529150872241995,
+      "learning_rate": 9.572987015561046e-07,
+      "loss": 0.3289,
+      "step": 11590
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8315409370255646,
+      "learning_rate": 9.56636303514592e-07,
+      "loss": 0.3295,
+      "step": 11591
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.6922792771830055,
+      "learning_rate": 9.559741104805164e-07,
+      "loss": 0.3295,
+      "step": 11592
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.0448921024660764,
+      "learning_rate": 9.553121224874528e-07,
+      "loss": 0.2932,
+      "step": 11593
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.469720921858887,
+      "learning_rate": 9.546503395689661e-07,
+      "loss": 0.2968,
+      "step": 11594
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.564121148332235,
+      "learning_rate": 9.539887617586058e-07,
+      "loss": 0.3243,
+      "step": 11595
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.311034090320815,
+      "learning_rate": 9.533273890899214e-07,
+      "loss": 0.3189,
+      "step": 11596
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.429444610415489,
+      "learning_rate": 9.526662215964405e-07,
+      "loss": 0.3388,
+      "step": 11597
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.145732786512728,
+      "learning_rate": 9.520052593116913e-07,
+      "loss": 0.315,
+      "step": 11598
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.759882143135349,
+      "learning_rate": 9.513445022691814e-07,
+      "loss": 0.2676,
+      "step": 11599
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.496018866968185,
+      "learning_rate": 9.506839505024146e-07,
+      "loss": 0.3088,
+      "step": 11600
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.719406755433305,
+      "learning_rate": 9.500236040448818e-07,
+      "loss": 0.3025,
+      "step": 11601
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.972858613394026,
+      "learning_rate": 9.493634629300646e-07,
+      "loss": 0.3318,
+      "step": 11602
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.299057498995527,
+      "learning_rate": 9.487035271914329e-07,
+      "loss": 0.298,
+      "step": 11603
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.098142966088677,
+      "learning_rate": 9.48043796862449e-07,
+      "loss": 0.3268,
+      "step": 11604
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.29802978920236095,
+      "learning_rate": 9.473842719765591e-07,
+      "loss": 0.1055,
+      "step": 11605
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 13.075698278546664,
+      "learning_rate": 9.467249525672046e-07,
+      "loss": 0.3212,
+      "step": 11606
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.649933053296057,
+      "learning_rate": 9.460658386678157e-07,
+      "loss": 0.28,
+      "step": 11607
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 35.287246719641004,
+      "learning_rate": 9.454069303118068e-07,
+      "loss": 0.2866,
+      "step": 11608
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.059479131352207,
+      "learning_rate": 9.447482275325914e-07,
+      "loss": 0.3175,
+      "step": 11609
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.5399276736371315,
+      "learning_rate": 9.440897303635616e-07,
+      "loss": 0.3076,
+      "step": 11610
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2627129322637866,
+      "learning_rate": 9.434314388381105e-07,
+      "loss": 0.1051,
+      "step": 11611
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7865804324769536,
+      "learning_rate": 9.427733529896105e-07,
+      "loss": 0.2978,
+      "step": 11612
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6324081393912735,
+      "learning_rate": 9.421154728514292e-07,
+      "loss": 0.3442,
+      "step": 11613
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.035118747401255,
+      "learning_rate": 9.414577984569228e-07,
+      "loss": 0.3228,
+      "step": 11614
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.07182773206815,
+      "learning_rate": 9.408003298394375e-07,
+      "loss": 0.3257,
+      "step": 11615
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.115616587756332,
+      "learning_rate": 9.40143067032307e-07,
+      "loss": 0.3002,
+      "step": 11616
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.826148511669765,
+      "learning_rate": 9.394860100688585e-07,
+      "loss": 0.2817,
+      "step": 11617
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30712091085138965,
+      "learning_rate": 9.388291589824033e-07,
+      "loss": 0.1017,
+      "step": 11618
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.576954152212917,
+      "learning_rate": 9.381725138062459e-07,
+      "loss": 0.3563,
+      "step": 11619
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.159049663581641,
+      "learning_rate": 9.375160745736806e-07,
+      "loss": 0.3134,
+      "step": 11620
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.960662801545923,
+      "learning_rate": 9.368598413179891e-07,
+      "loss": 0.3308,
+      "step": 11621
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5436068205446087,
+      "learning_rate": 9.362038140724466e-07,
+      "loss": 0.3498,
+      "step": 11622
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 12.386654190006483,
+      "learning_rate": 9.355479928703104e-07,
+      "loss": 0.3173,
+      "step": 11623
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.755934179230204,
+      "learning_rate": 9.348923777448371e-07,
+      "loss": 0.3071,
+      "step": 11624
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.988087378727934,
+      "learning_rate": 9.342369687292635e-07,
+      "loss": 0.2676,
+      "step": 11625
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.382443368040869,
+      "learning_rate": 9.335817658568241e-07,
+      "loss": 0.3195,
+      "step": 11626
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31057206291821143,
+      "learning_rate": 9.329267691607369e-07,
+      "loss": 0.1074,
+      "step": 11627
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9808185117805808,
+      "learning_rate": 9.322719786742118e-07,
+      "loss": 0.3479,
+      "step": 11628
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.480979260467842,
+      "learning_rate": 9.31617394430448e-07,
+      "loss": 0.3148,
+      "step": 11629
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.199336320862547,
+      "learning_rate": 9.309630164626355e-07,
+      "loss": 0.2936,
+      "step": 11630
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.147876702252836,
+      "learning_rate": 9.303088448039527e-07,
+      "loss": 0.301,
+      "step": 11631
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.0020584457824695,
+      "learning_rate": 9.296548794875659e-07,
+      "loss": 0.3167,
+      "step": 11632
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.30413683493873567,
+      "learning_rate": 9.290011205466337e-07,
+      "loss": 0.1101,
+      "step": 11633
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8729127058505255,
+      "learning_rate": 9.283475680143028e-07,
+      "loss": 0.3002,
+      "step": 11634
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.7508563645253847,
+      "learning_rate": 9.276942219237095e-07,
+      "loss": 0.2951,
+      "step": 11635
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9017693431577927,
+      "learning_rate": 9.270410823079806e-07,
+      "loss": 0.3046,
+      "step": 11636
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.6846629901693824,
+      "learning_rate": 9.263881492002325e-07,
+      "loss": 0.3229,
+      "step": 11637
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.206780151279534,
+      "learning_rate": 9.257354226335668e-07,
+      "loss": 0.3567,
+      "step": 11638
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3258709315108335,
+      "learning_rate": 9.250829026410835e-07,
+      "loss": 0.2968,
+      "step": 11639
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.955368424438126,
+      "learning_rate": 9.24430589255862e-07,
+      "loss": 0.3078,
+      "step": 11640
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.539876728171227,
+      "learning_rate": 9.237784825109786e-07,
+      "loss": 0.3066,
+      "step": 11641
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 11.072707793383845,
+      "learning_rate": 9.231265824394953e-07,
+      "loss": 0.3759,
+      "step": 11642
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.889154285427212,
+      "learning_rate": 9.224748890744656e-07,
+      "loss": 0.294,
+      "step": 11643
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7866532889266575,
+      "learning_rate": 9.218234024489331e-07,
+      "loss": 0.3012,
+      "step": 11644
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 14.22135764865298,
+      "learning_rate": 9.211721225959269e-07,
+      "loss": 0.3096,
+      "step": 11645
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.0762284784838965,
+      "learning_rate": 9.205210495484695e-07,
+      "loss": 0.2873,
+      "step": 11646
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0273921718620547,
+      "learning_rate": 9.198701833395718e-07,
+      "loss": 0.3123,
+      "step": 11647
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2506055751806473,
+      "learning_rate": 9.19219524002234e-07,
+      "loss": 0.3286,
+      "step": 11648
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.078488028663749,
+      "learning_rate": 9.185690715694462e-07,
+      "loss": 0.2987,
+      "step": 11649
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.4493704498809,
+      "learning_rate": 9.179188260741889e-07,
+      "loss": 0.3376,
+      "step": 11650
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.610656739955752,
+      "learning_rate": 9.172687875494273e-07,
+      "loss": 0.2906,
+      "step": 11651
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7064377917952305,
+      "learning_rate": 9.166189560281247e-07,
+      "loss": 0.3355,
+      "step": 11652
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9391004882484344,
+      "learning_rate": 9.159693315432255e-07,
+      "loss": 0.297,
+      "step": 11653
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.629783137060242,
+      "learning_rate": 9.153199141276681e-07,
+      "loss": 0.2854,
+      "step": 11654
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.4461161079342775,
+      "learning_rate": 9.146707038143787e-07,
+      "loss": 0.3061,
+      "step": 11655
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2881483597164176,
+      "learning_rate": 9.140217006362751e-07,
+      "loss": 0.3081,
+      "step": 11656
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.743456369536228,
+      "learning_rate": 9.133729046262635e-07,
+      "loss": 0.2919,
+      "step": 11657
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5141602598657395,
+      "learning_rate": 9.127243158172361e-07,
+      "loss": 0.2857,
+      "step": 11658
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.5777948494474865,
+      "learning_rate": 9.120759342420821e-07,
+      "loss": 0.3294,
+      "step": 11659
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.074938435608064,
+      "learning_rate": 9.11427759933673e-07,
+      "loss": 0.3393,
+      "step": 11660
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.714400493102677,
+      "learning_rate": 9.107797929248729e-07,
+      "loss": 0.2909,
+      "step": 11661
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.838926590001597,
+      "learning_rate": 9.101320332485358e-07,
+      "loss": 0.2816,
+      "step": 11662
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.647415301242839,
+      "learning_rate": 9.094844809375042e-07,
+      "loss": 0.3071,
+      "step": 11663
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.090173196733619,
+      "learning_rate": 9.088371360246107e-07,
+      "loss": 0.331,
+      "step": 11664
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.966056415113204,
+      "learning_rate": 9.081899985426784e-07,
+      "loss": 0.3284,
+      "step": 11665
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.28663559509912045,
+      "learning_rate": 9.075430685245162e-07,
+      "loss": 0.0965,
+      "step": 11666
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.109897402607669,
+      "learning_rate": 9.068963460029257e-07,
+      "loss": 0.2686,
+      "step": 11667
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.657544434546185,
+      "learning_rate": 9.062498310106971e-07,
+      "loss": 0.2889,
+      "step": 11668
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.348279293741983,
+      "learning_rate": 9.056035235806104e-07,
+      "loss": 0.3066,
+      "step": 11669
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.43374503719196,
+      "learning_rate": 9.049574237454362e-07,
+      "loss": 0.3377,
+      "step": 11670
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.433323482143647,
+      "learning_rate": 9.043115315379292e-07,
+      "loss": 0.3193,
+      "step": 11671
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.556361156687401,
+      "learning_rate": 9.036658469908427e-07,
+      "loss": 0.343,
+      "step": 11672
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5696641784604046,
+      "learning_rate": 9.030203701369095e-07,
+      "loss": 0.3297,
+      "step": 11673
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.05510479445233,
+      "learning_rate": 9.02375101008861e-07,
+      "loss": 0.3437,
+      "step": 11674
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.724098709395985,
+      "learning_rate": 9.017300396394108e-07,
+      "loss": 0.303,
+      "step": 11675
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.769794868200761,
+      "learning_rate": 9.010851860612657e-07,
+      "loss": 0.3445,
+      "step": 11676
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.000457737183579,
+      "learning_rate": 9.004405403071214e-07,
+      "loss": 0.2792,
+      "step": 11677
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.448875373301414,
+      "learning_rate": 8.997961024096641e-07,
+      "loss": 0.3206,
+      "step": 11678
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.013065819019608,
+      "learning_rate": 8.991518724015657e-07,
+      "loss": 0.3176,
+      "step": 11679
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.320207621892896,
+      "learning_rate": 8.985078503154914e-07,
+      "loss": 0.3189,
+      "step": 11680
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.121364150630025,
+      "learning_rate": 8.978640361840946e-07,
+      "loss": 0.325,
+      "step": 11681
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.854982698573061,
+      "learning_rate": 8.972204300400173e-07,
+      "loss": 0.301,
+      "step": 11682
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.616797685849901,
+      "learning_rate": 8.96577031915894e-07,
+      "loss": 0.2991,
+      "step": 11683
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.386981363722114,
+      "learning_rate": 8.959338418443425e-07,
+      "loss": 0.3216,
+      "step": 11684
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.33695413566792,
+      "learning_rate": 8.952908598579785e-07,
+      "loss": 0.3053,
+      "step": 11685
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.649935190355417,
+      "learning_rate": 8.946480859893975e-07,
+      "loss": 0.2952,
+      "step": 11686
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2973103793004622,
+      "learning_rate": 8.940055202711955e-07,
+      "loss": 0.1122,
+      "step": 11687
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5669393822426083,
+      "learning_rate": 8.933631627359473e-07,
+      "loss": 0.2964,
+      "step": 11688
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.138555925616713,
+      "learning_rate": 8.927210134162235e-07,
+      "loss": 0.331,
+      "step": 11689
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.128394729090195,
+      "learning_rate": 8.920790723445821e-07,
+      "loss": 0.2904,
+      "step": 11690
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.395366063525171,
+      "learning_rate": 8.914373395535719e-07,
+      "loss": 0.271,
+      "step": 11691
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.736899404592742,
+      "learning_rate": 8.90795815075729e-07,
+      "loss": 0.3115,
+      "step": 11692
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.725049488920038,
+      "learning_rate": 8.901544989435818e-07,
+      "loss": 0.3116,
+      "step": 11693
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.7379566953912504,
+      "learning_rate": 8.895133911896442e-07,
+      "loss": 0.3874,
+      "step": 11694
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.112113617844396,
+      "learning_rate": 8.888724918464226e-07,
+      "loss": 0.3455,
+      "step": 11695
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2758854351602266,
+      "learning_rate": 8.882318009464124e-07,
+      "loss": 0.1067,
+      "step": 11696
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.2770008078503703,
+      "learning_rate": 8.87591318522098e-07,
+      "loss": 0.302,
+      "step": 11697
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.03977918511395,
+      "learning_rate": 8.869510446059543e-07,
+      "loss": 0.2993,
+      "step": 11698
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8575942290223137,
+      "learning_rate": 8.863109792304408e-07,
+      "loss": 0.3183,
+      "step": 11699
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.144497503115823,
+      "learning_rate": 8.856711224280157e-07,
+      "loss": 0.3057,
+      "step": 11700
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.867679378318175,
+      "learning_rate": 8.85031474231115e-07,
+      "loss": 0.2957,
+      "step": 11701
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5991204714129648,
+      "learning_rate": 8.843920346721768e-07,
+      "loss": 0.2664,
+      "step": 11702
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.3912390378027863,
+      "learning_rate": 8.837528037836174e-07,
+      "loss": 0.3232,
+      "step": 11703
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.962626001932721,
+      "learning_rate": 8.831137815978485e-07,
+      "loss": 0.3072,
+      "step": 11704
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.361748554159868,
+      "learning_rate": 8.824749681472705e-07,
+      "loss": 0.3262,
+      "step": 11705
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.8999334535413794,
+      "learning_rate": 8.818363634642718e-07,
+      "loss": 0.304,
+      "step": 11706
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 10.09431767178268,
+      "learning_rate": 8.811979675812332e-07,
+      "loss": 0.2999,
+      "step": 11707
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.653512058745021,
+      "learning_rate": 8.805597805305194e-07,
+      "loss": 0.3279,
+      "step": 11708
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.2901791074707402,
+      "learning_rate": 8.7992180234449e-07,
+      "loss": 0.109,
+      "step": 11709
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.3211924276871825,
+      "learning_rate": 8.792840330554908e-07,
+      "loss": 0.3232,
+      "step": 11710
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.418937526375692,
+      "learning_rate": 8.786464726958593e-07,
+      "loss": 0.2925,
+      "step": 11711
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5879034576576987,
+      "learning_rate": 8.780091212979208e-07,
+      "loss": 0.2812,
+      "step": 11712
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.301722714970447,
+      "learning_rate": 8.77371978893991e-07,
+      "loss": 0.3174,
+      "step": 11713
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.884709641561306,
+      "learning_rate": 8.76735045516372e-07,
+      "loss": 0.343,
+      "step": 11714
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.31759614944713277,
+      "learning_rate": 8.760983211973618e-07,
+      "loss": 0.1208,
+      "step": 11715
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.898354178767807,
+      "learning_rate": 8.754618059692393e-07,
+      "loss": 0.3096,
+      "step": 11716
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.015074250438885,
+      "learning_rate": 8.748254998642803e-07,
+      "loss": 0.3018,
+      "step": 11717
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4021000519724667,
+      "learning_rate": 8.741894029147452e-07,
+      "loss": 0.354,
+      "step": 11718
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 9.197990164265692,
+      "learning_rate": 8.735535151528868e-07,
+      "loss": 0.3673,
+      "step": 11719
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 7.234500804914004,
+      "learning_rate": 8.729178366109464e-07,
+      "loss": 0.3229,
+      "step": 11720
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.716580864466856,
+      "learning_rate": 8.722823673211522e-07,
+      "loss": 0.3548,
+      "step": 11721
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 8.514287793743158,
+      "learning_rate": 8.716471073157251e-07,
+      "loss": 0.2991,
+      "step": 11722
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3773000576398156,
+      "learning_rate": 8.710120566268737e-07,
+      "loss": 0.2935,
+      "step": 11723
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.205708620915206,
+      "learning_rate": 8.703772152867973e-07,
+      "loss": 0.3227,
+      "step": 11724
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9115602893369745,
+      "learning_rate": 8.697425833276835e-07,
+      "loss": 0.2909,
+      "step": 11725
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.251843552941919,
+      "learning_rate": 8.691081607817103e-07,
+      "loss": 0.3146,
+      "step": 11726
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.2926830206237385,
+      "learning_rate": 8.684739476810411e-07,
+      "loss": 0.3048,
+      "step": 11727
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3602126495963405,
+      "learning_rate": 8.678399440578367e-07,
+      "loss": 0.3298,
+      "step": 11728
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3135619890372423,
+      "learning_rate": 8.672061499442386e-07,
+      "loss": 0.3008,
+      "step": 11729
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.616210768040346,
+      "learning_rate": 8.665725653723828e-07,
+      "loss": 0.2886,
+      "step": 11730
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.637451786373551,
+      "learning_rate": 8.659391903743935e-07,
+      "loss": 0.325,
+      "step": 11731
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6282715394242304,
+      "learning_rate": 8.653060249823847e-07,
+      "loss": 0.3393,
+      "step": 11732
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6460155186367422,
+      "learning_rate": 8.646730692284599e-07,
+      "loss": 0.3056,
+      "step": 11733
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.09665201993418,
+      "learning_rate": 8.640403231447075e-07,
+      "loss": 0.3371,
+      "step": 11734
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.331907819122727,
+      "learning_rate": 8.634077867632145e-07,
+      "loss": 0.3159,
+      "step": 11735
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.4214158251570113,
+      "learning_rate": 8.627754601160488e-07,
+      "loss": 0.316,
+      "step": 11736
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.018029247612475,
+      "learning_rate": 8.621433432352705e-07,
+      "loss": 0.3114,
+      "step": 11737
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.643499901005241,
+      "learning_rate": 8.615114361529303e-07,
+      "loss": 0.3129,
+      "step": 11738
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.719325213906516,
+      "learning_rate": 8.60879738901067e-07,
+      "loss": 0.2735,
+      "step": 11739
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.712446210646124,
+      "learning_rate": 8.602482515117089e-07,
+      "loss": 0.3,
+      "step": 11740
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7800911585513104,
+      "learning_rate": 8.596169740168758e-07,
+      "loss": 0.306,
+      "step": 11741
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.889076545837684,
+      "learning_rate": 8.589859064485711e-07,
+      "loss": 0.2911,
+      "step": 11742
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.520205310471321,
+      "learning_rate": 8.583550488387938e-07,
+      "loss": 0.3291,
+      "step": 11743
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.893633434855234,
+      "learning_rate": 8.577244012195291e-07,
+      "loss": 0.3135,
+      "step": 11744
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.17715608323108,
+      "learning_rate": 8.570939636227521e-07,
+      "loss": 0.3095,
+      "step": 11745
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.434544037688996,
+      "learning_rate": 8.56463736080429e-07,
+      "loss": 0.3093,
+      "step": 11746
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 58.85467826577525,
+      "learning_rate": 8.558337186245102e-07,
+      "loss": 0.3184,
+      "step": 11747
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.549139726979424,
+      "learning_rate": 8.552039112869431e-07,
+      "loss": 0.3213,
+      "step": 11748
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.18618497822044,
+      "learning_rate": 8.545743140996565e-07,
+      "loss": 0.3462,
+      "step": 11749
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.657267750800833,
+      "learning_rate": 8.53944927094576e-07,
+      "loss": 0.316,
+      "step": 11750
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.2529816878426105,
+      "learning_rate": 8.533157503036105e-07,
+      "loss": 0.3213,
+      "step": 11751
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.295250033608374,
+      "learning_rate": 8.526867837586605e-07,
+      "loss": 0.3041,
+      "step": 11752
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1486390786453904,
+      "learning_rate": 8.520580274916174e-07,
+      "loss": 0.2832,
+      "step": 11753
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.703966526021286,
+      "learning_rate": 8.51429481534361e-07,
+      "loss": 0.2808,
+      "step": 11754
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.565045767047056,
+      "learning_rate": 8.508011459187576e-07,
+      "loss": 0.3468,
+      "step": 11755
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9378941511834955,
+      "learning_rate": 8.501730206766661e-07,
+      "loss": 0.3012,
+      "step": 11756
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8014642864687,
+      "learning_rate": 8.495451058399345e-07,
+      "loss": 0.2751,
+      "step": 11757
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7219905465250354,
+      "learning_rate": 8.48917401440399e-07,
+      "loss": 0.2882,
+      "step": 11758
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3038698344507147,
+      "learning_rate": 8.482899075098872e-07,
+      "loss": 0.099,
+      "step": 11759
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 14.725368778708066,
+      "learning_rate": 8.476626240802099e-07,
+      "loss": 0.3342,
+      "step": 11760
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9532521146972424,
+      "learning_rate": 8.47035551183178e-07,
+      "loss": 0.3119,
+      "step": 11761
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.966780494995053,
+      "learning_rate": 8.464086888505796e-07,
+      "loss": 0.3157,
+      "step": 11762
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.143259967627889,
+      "learning_rate": 8.457820371142028e-07,
+      "loss": 0.306,
+      "step": 11763
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.802893290514554,
+      "learning_rate": 8.451555960058173e-07,
+      "loss": 0.2909,
+      "step": 11764
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.024886505165491,
+      "learning_rate": 8.445293655571851e-07,
+      "loss": 0.3162,
+      "step": 11765
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.01659595810273,
+      "learning_rate": 8.43903345800059e-07,
+      "loss": 0.3037,
+      "step": 11766
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31323454964267533,
+      "learning_rate": 8.432775367661783e-07,
+      "loss": 0.1041,
+      "step": 11767
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.166222101257623,
+      "learning_rate": 8.426519384872733e-07,
+      "loss": 0.2912,
+      "step": 11768
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.477837314637301,
+      "learning_rate": 8.420265509950648e-07,
+      "loss": 0.3153,
+      "step": 11769
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.120633511031707,
+      "learning_rate": 8.41401374321259e-07,
+      "loss": 0.3266,
+      "step": 11770
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.048133512202363,
+      "learning_rate": 8.40776408497554e-07,
+      "loss": 0.2662,
+      "step": 11771
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.593241303359269,
+      "learning_rate": 8.401516535556375e-07,
+      "loss": 0.3015,
+      "step": 11772
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7073796656454356,
+      "learning_rate": 8.395271095271856e-07,
+      "loss": 0.3307,
+      "step": 11773
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.729113614264014,
+      "learning_rate": 8.389027764438662e-07,
+      "loss": 0.3245,
+      "step": 11774
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.275938205674888,
+      "learning_rate": 8.3827865433733e-07,
+      "loss": 0.3414,
+      "step": 11775
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7322681772588955,
+      "learning_rate": 8.376547432392262e-07,
+      "loss": 0.3287,
+      "step": 11776
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.0106691720362475,
+      "learning_rate": 8.37031043181184e-07,
+      "loss": 0.3143,
+      "step": 11777
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.773952290220527,
+      "learning_rate": 8.364075541948307e-07,
+      "loss": 0.3403,
+      "step": 11778
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3135946282224535,
+      "learning_rate": 8.357842763117752e-07,
+      "loss": 0.1146,
+      "step": 11779
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2779697773529956,
+      "learning_rate": 8.351612095636202e-07,
+      "loss": 0.1063,
+      "step": 11780
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2998404956795111,
+      "learning_rate": 8.345383539819568e-07,
+      "loss": 0.1097,
+      "step": 11781
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.142710281989739,
+      "learning_rate": 8.339157095983652e-07,
+      "loss": 0.3303,
+      "step": 11782
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.872637580217152,
+      "learning_rate": 8.332932764444152e-07,
+      "loss": 0.2907,
+      "step": 11783
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.640363957529052,
+      "learning_rate": 8.326710545516642e-07,
+      "loss": 0.3104,
+      "step": 11784
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.519802803122927,
+      "learning_rate": 8.32049043951661e-07,
+      "loss": 0.3195,
+      "step": 11785
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.761721000842759,
+      "learning_rate": 8.314272446759425e-07,
+      "loss": 0.2946,
+      "step": 11786
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.149041918772097,
+      "learning_rate": 8.308056567560374e-07,
+      "loss": 0.3085,
+      "step": 11787
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5751973258738885,
+      "learning_rate": 8.301842802234572e-07,
+      "loss": 0.3101,
+      "step": 11788
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 35.92689268518967,
+      "learning_rate": 8.295631151097116e-07,
+      "loss": 0.3022,
+      "step": 11789
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.979630922061625,
+      "learning_rate": 8.289421614462911e-07,
+      "loss": 0.3234,
+      "step": 11790
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 24.36859084928235,
+      "learning_rate": 8.283214192646843e-07,
+      "loss": 0.3074,
+      "step": 11791
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.555872957133879,
+      "learning_rate": 8.277008885963594e-07,
+      "loss": 0.2984,
+      "step": 11792
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.180006451795596,
+      "learning_rate": 8.270805694727813e-07,
+      "loss": 0.3072,
+      "step": 11793
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1049233511035434,
+      "learning_rate": 8.264604619254002e-07,
+      "loss": 0.3578,
+      "step": 11794
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.680304651466561,
+      "learning_rate": 8.258405659856583e-07,
+      "loss": 0.3034,
+      "step": 11795
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.328108694680757,
+      "learning_rate": 8.252208816849855e-07,
+      "loss": 0.3062,
+      "step": 11796
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8229665638923516,
+      "learning_rate": 8.246014090547999e-07,
+      "loss": 0.3246,
+      "step": 11797
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.999779688886573,
+      "learning_rate": 8.239821481265103e-07,
+      "loss": 0.2815,
+      "step": 11798
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3799083920000217,
+      "learning_rate": 8.233630989315156e-07,
+      "loss": 0.3007,
+      "step": 11799
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.273742845644951,
+      "learning_rate": 8.227442615012026e-07,
+      "loss": 0.3071,
+      "step": 11800
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.112290381849995,
+      "learning_rate": 8.221256358669472e-07,
+      "loss": 0.3038,
+      "step": 11801
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5632806900411973,
+      "learning_rate": 8.215072220601167e-07,
+      "loss": 0.3061,
+      "step": 11802
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.339981921239123,
+      "learning_rate": 8.208890201120628e-07,
+      "loss": 0.3303,
+      "step": 11803
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.098043079919627,
+      "learning_rate": 8.202710300541339e-07,
+      "loss": 0.2972,
+      "step": 11804
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.169184698213742,
+      "learning_rate": 8.196532519176604e-07,
+      "loss": 0.3261,
+      "step": 11805
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.9031155657149785,
+      "learning_rate": 8.190356857339655e-07,
+      "loss": 0.3305,
+      "step": 11806
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.2743275891013255,
+      "learning_rate": 8.184183315343613e-07,
+      "loss": 0.294,
+      "step": 11807
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.40955510712984,
+      "learning_rate": 8.178011893501498e-07,
+      "loss": 0.3026,
+      "step": 11808
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.820742278470294,
+      "learning_rate": 8.171842592126223e-07,
+      "loss": 0.3464,
+      "step": 11809
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2830946225515407,
+      "learning_rate": 8.165675411530543e-07,
+      "loss": 0.109,
+      "step": 11810
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.9978517411924406,
+      "learning_rate": 8.159510352027205e-07,
+      "loss": 0.3176,
+      "step": 11811
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.80311252063245,
+      "learning_rate": 8.153347413928747e-07,
+      "loss": 0.3008,
+      "step": 11812
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.964307011103821,
+      "learning_rate": 8.147186597547657e-07,
+      "loss": 0.2955,
+      "step": 11813
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1665050940471207,
+      "learning_rate": 8.141027903196308e-07,
+      "loss": 0.3126,
+      "step": 11814
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.817758615117734,
+      "learning_rate": 8.134871331186955e-07,
+      "loss": 0.3087,
+      "step": 11815
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.237119546733668,
+      "learning_rate": 8.128716881831745e-07,
+      "loss": 0.3204,
+      "step": 11816
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.937961901951175,
+      "learning_rate": 8.122564555442741e-07,
+      "loss": 0.31,
+      "step": 11817
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.663797351303694,
+      "learning_rate": 8.116414352331858e-07,
+      "loss": 0.2971,
+      "step": 11818
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.98226023448084,
+      "learning_rate": 8.110266272810929e-07,
+      "loss": 0.3186,
+      "step": 11819
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.0219033111502407,
+      "learning_rate": 8.104120317191677e-07,
+      "loss": 0.2831,
+      "step": 11820
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.95089240108938,
+      "learning_rate": 8.097976485785719e-07,
+      "loss": 0.3204,
+      "step": 11821
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2319327262010105,
+      "learning_rate": 8.091834778904573e-07,
+      "loss": 0.334,
+      "step": 11822
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 18.428011468242648,
+      "learning_rate": 8.085695196859594e-07,
+      "loss": 0.3139,
+      "step": 11823
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.968813006837595,
+      "learning_rate": 8.079557739962129e-07,
+      "loss": 0.3073,
+      "step": 11824
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.564805450119666,
+      "learning_rate": 8.073422408523324e-07,
+      "loss": 0.3212,
+      "step": 11825
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.96995969576303,
+      "learning_rate": 8.067289202854256e-07,
+      "loss": 0.3191,
+      "step": 11826
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8961261308817123,
+      "learning_rate": 8.061158123265899e-07,
+      "loss": 0.2863,
+      "step": 11827
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.126981731717114,
+      "learning_rate": 8.055029170069112e-07,
+      "loss": 0.2609,
+      "step": 11828
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3147788344796734,
+      "learning_rate": 8.048902343574649e-07,
+      "loss": 0.3263,
+      "step": 11829
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.886664586291623,
+      "learning_rate": 8.042777644093164e-07,
+      "loss": 0.3033,
+      "step": 11830
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.2311803938543178,
+      "learning_rate": 8.036655071935162e-07,
+      "loss": 0.3286,
+      "step": 11831
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.096979015413067,
+      "learning_rate": 8.030534627411091e-07,
+      "loss": 0.33,
+      "step": 11832
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.877200279026921,
+      "learning_rate": 8.024416310831268e-07,
+      "loss": 0.3029,
+      "step": 11833
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.127931261360391,
+      "learning_rate": 8.018300122505907e-07,
+      "loss": 0.3056,
+      "step": 11834
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.45939579793291,
+      "learning_rate": 8.012186062745126e-07,
+      "loss": 0.325,
+      "step": 11835
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.342832372195975,
+      "learning_rate": 8.006074131858876e-07,
+      "loss": 0.3054,
+      "step": 11836
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.236366845304119,
+      "learning_rate": 7.999964330157101e-07,
+      "loss": 0.3203,
+      "step": 11837
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.58934870357346,
+      "learning_rate": 7.993856657949533e-07,
+      "loss": 0.2821,
+      "step": 11838
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.605790205416122,
+      "learning_rate": 7.987751115545889e-07,
+      "loss": 0.3286,
+      "step": 11839
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.020336321776039,
+      "learning_rate": 7.981647703255702e-07,
+      "loss": 0.3048,
+      "step": 11840
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.697409371115942,
+      "learning_rate": 7.975546421388436e-07,
+      "loss": 0.3066,
+      "step": 11841
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 9.06826846822221,
+      "learning_rate": 7.969447270253444e-07,
+      "loss": 0.3101,
+      "step": 11842
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5950771559767234,
+      "learning_rate": 7.963350250159962e-07,
+      "loss": 0.2891,
+      "step": 11843
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2851707824098209,
+      "learning_rate": 7.957255361417137e-07,
+      "loss": 0.1036,
+      "step": 11844
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3382288087376404,
+      "learning_rate": 7.951162604333967e-07,
+      "loss": 0.2994,
+      "step": 11845
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30541256631416414,
+      "learning_rate": 7.945071979219382e-07,
+      "loss": 0.1015,
+      "step": 11846
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.30010716066742654,
+      "learning_rate": 7.93898348638219e-07,
+      "loss": 0.1297,
+      "step": 11847
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.2899181089701862,
+      "learning_rate": 7.93289712613109e-07,
+      "loss": 0.1121,
+      "step": 11848
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.752021314001168,
+      "learning_rate": 7.926812898774677e-07,
+      "loss": 0.2919,
+      "step": 11849
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 8.904158560579342,
+      "learning_rate": 7.92073080462144e-07,
+      "loss": 0.3345,
+      "step": 11850
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.360055427874073,
+      "learning_rate": 7.914650843979726e-07,
+      "loss": 0.3057,
+      "step": 11851
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.402862159849658,
+      "learning_rate": 7.90857301715785e-07,
+      "loss": 0.2591,
+      "step": 11852
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.730770733335287,
+      "learning_rate": 7.902497324463915e-07,
+      "loss": 0.3688,
+      "step": 11853
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.952699257507841,
+      "learning_rate": 7.896423766206029e-07,
+      "loss": 0.2962,
+      "step": 11854
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.837540864971671,
+      "learning_rate": 7.890352342692092e-07,
+      "loss": 0.2818,
+      "step": 11855
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.5754741557541525,
+      "learning_rate": 7.884283054229958e-07,
+      "loss": 0.3095,
+      "step": 11856
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 30.555796838335276,
+      "learning_rate": 7.878215901127345e-07,
+      "loss": 0.3073,
+      "step": 11857
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1865191980587793,
+      "learning_rate": 7.872150883691876e-07,
+      "loss": 0.2667,
+      "step": 11858
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 13.853379932049624,
+      "learning_rate": 7.866088002231076e-07,
+      "loss": 0.2893,
+      "step": 11859
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.451315355532291,
+      "learning_rate": 7.860027257052311e-07,
+      "loss": 0.32,
+      "step": 11860
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.271802577121537,
+      "learning_rate": 7.853968648462901e-07,
+      "loss": 0.2773,
+      "step": 11861
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 7.506293664848721,
+      "learning_rate": 7.847912176770018e-07,
+      "loss": 0.3118,
+      "step": 11862
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.541854209020566,
+      "learning_rate": 7.841857842280753e-07,
+      "loss": 0.2532,
+      "step": 11863
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.6209175326708514,
+      "learning_rate": 7.835805645302047e-07,
+      "loss": 0.3007,
+      "step": 11864
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.7324894552443406,
+      "learning_rate": 7.829755586140797e-07,
+      "loss": 0.3011,
+      "step": 11865
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.548935372996712,
+      "learning_rate": 7.823707665103714e-07,
+      "loss": 0.3087,
+      "step": 11866
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.506457275973846,
+      "learning_rate": 7.817661882497484e-07,
+      "loss": 0.327,
+      "step": 11867
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2993133804326513,
+      "learning_rate": 7.811618238628604e-07,
+      "loss": 0.3394,
+      "step": 11868
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5389793722527556,
+      "learning_rate": 7.805576733803522e-07,
+      "loss": 0.3015,
+      "step": 11869
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4318319726238444,
+      "learning_rate": 7.799537368328541e-07,
+      "loss": 0.3224,
+      "step": 11870
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.186441350259544,
+      "learning_rate": 7.79350014250988e-07,
+      "loss": 0.2781,
+      "step": 11871
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.582574096867779,
+      "learning_rate": 7.787465056653653e-07,
+      "loss": 0.3272,
+      "step": 11872
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.289440568211513,
+      "learning_rate": 7.781432111065823e-07,
+      "loss": 0.3615,
+      "step": 11873
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8331227465093227,
+      "learning_rate": 7.775401306052283e-07,
+      "loss": 0.2993,
+      "step": 11874
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.812020645417306,
+      "learning_rate": 7.769372641918815e-07,
+      "loss": 0.3036,
+      "step": 11875
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.1471260642556205,
+      "learning_rate": 7.763346118971082e-07,
+      "loss": 0.3081,
+      "step": 11876
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.227394851044112,
+      "learning_rate": 7.757321737514645e-07,
+      "loss": 0.305,
+      "step": 11877
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.00380005388356,
+      "learning_rate": 7.751299497854964e-07,
+      "loss": 0.2762,
+      "step": 11878
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.710167241457327,
+      "learning_rate": 7.74527940029734e-07,
+      "loss": 0.3243,
+      "step": 11879
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.878816069843269,
+      "learning_rate": 7.739261445147056e-07,
+      "loss": 0.3062,
+      "step": 11880
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.30236186001546045,
+      "learning_rate": 7.733245632709202e-07,
+      "loss": 0.1201,
+      "step": 11881
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.0448322114164315,
+      "learning_rate": 7.7272319632888e-07,
+      "loss": 0.3128,
+      "step": 11882
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.051295342217346,
+      "learning_rate": 7.721220437190762e-07,
+      "loss": 0.2697,
+      "step": 11883
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.416305093894169,
+      "learning_rate": 7.715211054719879e-07,
+      "loss": 0.2964,
+      "step": 11884
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5614553782371585,
+      "learning_rate": 7.70920381618086e-07,
+      "loss": 0.297,
+      "step": 11885
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5649310461329904,
+      "learning_rate": 7.703198721878241e-07,
+      "loss": 0.3422,
+      "step": 11886
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.572792727674521,
+      "learning_rate": 7.697195772116545e-07,
+      "loss": 0.3301,
+      "step": 11887
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.34108660031317,
+      "learning_rate": 7.691194967200099e-07,
+      "loss": 0.2931,
+      "step": 11888
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.632166015915517,
+      "learning_rate": 7.685196307433168e-07,
+      "loss": 0.2916,
+      "step": 11889
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9844149448660984,
+      "learning_rate": 7.679199793119901e-07,
+      "loss": 0.2991,
+      "step": 11890
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.99694012435066,
+      "learning_rate": 7.673205424564334e-07,
+      "loss": 0.3177,
+      "step": 11891
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.552491911530154,
+      "learning_rate": 7.667213202070389e-07,
+      "loss": 0.2619,
+      "step": 11892
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.900545307349801,
+      "learning_rate": 7.661223125941897e-07,
+      "loss": 0.2851,
+      "step": 11893
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.284469968685873,
+      "learning_rate": 7.655235196482552e-07,
+      "loss": 0.3231,
+      "step": 11894
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0020423847967193,
+      "learning_rate": 7.649249413995958e-07,
+      "loss": 0.3049,
+      "step": 11895
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7147099659006213,
+      "learning_rate": 7.643265778785613e-07,
+      "loss": 0.3013,
+      "step": 11896
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.751843043379334,
+      "learning_rate": 7.637284291154906e-07,
+      "loss": 0.3435,
+      "step": 11897
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.880009170428828,
+      "learning_rate": 7.631304951407109e-07,
+      "loss": 0.3103,
+      "step": 11898
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5725564264728904,
+      "learning_rate": 7.625327759845364e-07,
+      "loss": 0.2942,
+      "step": 11899
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.52539651098941,
+      "learning_rate": 7.619352716772765e-07,
+      "loss": 0.3386,
+      "step": 11900
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.815078066590267,
+      "learning_rate": 7.613379822492234e-07,
+      "loss": 0.2901,
+      "step": 11901
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.614946798611108,
+      "learning_rate": 7.607409077306621e-07,
+      "loss": 0.32,
+      "step": 11902
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.64998833298746,
+      "learning_rate": 7.601440481518651e-07,
+      "loss": 0.3214,
+      "step": 11903
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2124575452438795,
+      "learning_rate": 7.595474035430944e-07,
+      "loss": 0.3001,
+      "step": 11904
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.749770279079868,
+      "learning_rate": 7.589509739346018e-07,
+      "loss": 0.3161,
+      "step": 11905
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.563385643039112,
+      "learning_rate": 7.583547593566282e-07,
+      "loss": 0.2897,
+      "step": 11906
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8864671090801877,
+      "learning_rate": 7.577587598394009e-07,
+      "loss": 0.3176,
+      "step": 11907
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.785402243842661,
+      "learning_rate": 7.571629754131399e-07,
+      "loss": 0.3127,
+      "step": 11908
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.326267380726621,
+      "learning_rate": 7.565674061080519e-07,
+      "loss": 0.2742,
+      "step": 11909
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9244958864033097,
+      "learning_rate": 7.559720519543345e-07,
+      "loss": 0.3077,
+      "step": 11910
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.03010206767288,
+      "learning_rate": 7.553769129821742e-07,
+      "loss": 0.3146,
+      "step": 11911
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.769506431997061,
+      "learning_rate": 7.547819892217423e-07,
+      "loss": 0.2656,
+      "step": 11912
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.816658095583148,
+      "learning_rate": 7.541872807032074e-07,
+      "loss": 0.2904,
+      "step": 11913
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.313317820829761,
+      "learning_rate": 7.535927874567184e-07,
+      "loss": 0.3057,
+      "step": 11914
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4388415300058077,
+      "learning_rate": 7.529985095124215e-07,
+      "loss": 0.3008,
+      "step": 11915
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8122708500903504,
+      "learning_rate": 7.524044469004443e-07,
+      "loss": 0.3614,
+      "step": 11916
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.559655422076748,
+      "learning_rate": 7.518105996509085e-07,
+      "loss": 0.3238,
+      "step": 11917
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.60091092543179,
+      "learning_rate": 7.512169677939235e-07,
+      "loss": 0.31,
+      "step": 11918
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.701586415669686,
+      "learning_rate": 7.506235513595878e-07,
+      "loss": 0.3122,
+      "step": 11919
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.750098488891184,
+      "learning_rate": 7.500303503779898e-07,
+      "loss": 0.3132,
+      "step": 11920
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.336100531445632,
+      "learning_rate": 7.494373648792042e-07,
+      "loss": 0.3292,
+      "step": 11921
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.483814259468198,
+      "learning_rate": 7.488445948932977e-07,
+      "loss": 0.3262,
+      "step": 11922
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.041295467421003,
+      "learning_rate": 7.482520404503246e-07,
+      "loss": 0.3198,
+      "step": 11923
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.345730156640729,
+      "learning_rate": 7.476597015803288e-07,
+      "loss": 0.3527,
+      "step": 11924
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.94913562181283,
+      "learning_rate": 7.470675783133441e-07,
+      "loss": 0.3155,
+      "step": 11925
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.758986901467148,
+      "learning_rate": 7.464756706793924e-07,
+      "loss": 0.297,
+      "step": 11926
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.851657047919043,
+      "learning_rate": 7.458839787084821e-07,
+      "loss": 0.2865,
+      "step": 11927
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2295084577701205,
+      "learning_rate": 7.452925024306174e-07,
+      "loss": 0.299,
+      "step": 11928
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.8778476971852225,
+      "learning_rate": 7.447012418757831e-07,
+      "loss": 0.3168,
+      "step": 11929
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.016877772671976,
+      "learning_rate": 7.441101970739622e-07,
+      "loss": 0.3041,
+      "step": 11930
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.987959983732929,
+      "learning_rate": 7.435193680551178e-07,
+      "loss": 0.3308,
+      "step": 11931
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.003881148111166,
+      "learning_rate": 7.429287548492082e-07,
+      "loss": 0.2885,
+      "step": 11932
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.755392578314098,
+      "learning_rate": 7.423383574861792e-07,
+      "loss": 0.298,
+      "step": 11933
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.628369182787973,
+      "learning_rate": 7.417481759959639e-07,
+      "loss": 0.3212,
+      "step": 11934
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2971386039249858,
+      "learning_rate": 7.411582104084858e-07,
+      "loss": 0.1182,
+      "step": 11935
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.887792795019684,
+      "learning_rate": 7.405684607536584e-07,
+      "loss": 0.324,
+      "step": 11936
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8001987010922065,
+      "learning_rate": 7.399789270613828e-07,
+      "loss": 0.3228,
+      "step": 11937
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.9410781604853384,
+      "learning_rate": 7.393896093615499e-07,
+      "loss": 0.3217,
+      "step": 11938
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.029249237908793,
+      "learning_rate": 7.388005076840405e-07,
+      "loss": 0.3137,
+      "step": 11939
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.149094695005897,
+      "learning_rate": 7.382116220587199e-07,
+      "loss": 0.3438,
+      "step": 11940
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.678231236411989,
+      "learning_rate": 7.376229525154499e-07,
+      "loss": 0.3306,
+      "step": 11941
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6518168339539416,
+      "learning_rate": 7.370344990840738e-07,
+      "loss": 0.3029,
+      "step": 11942
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8424838224927944,
+      "learning_rate": 7.364462617944318e-07,
+      "loss": 0.3145,
+      "step": 11943
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.2906696067858519,
+      "learning_rate": 7.358582406763448e-07,
+      "loss": 0.1052,
+      "step": 11944
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7556731193406936,
+      "learning_rate": 7.35270435759628e-07,
+      "loss": 0.2777,
+      "step": 11945
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.208633075825948,
+      "learning_rate": 7.346828470740852e-07,
+      "loss": 0.2973,
+      "step": 11946
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 13.008926802454805,
+      "learning_rate": 7.340954746495077e-07,
+      "loss": 0.298,
+      "step": 11947
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.138423052384404,
+      "learning_rate": 7.335083185156777e-07,
+      "loss": 0.2919,
+      "step": 11948
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2866164970005385,
+      "learning_rate": 7.329213787023632e-07,
+      "loss": 0.3046,
+      "step": 11949
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.036401308984353,
+      "learning_rate": 7.323346552393251e-07,
+      "loss": 0.3037,
+      "step": 11950
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.648150724182742,
+      "learning_rate": 7.317481481563105e-07,
+      "loss": 0.2907,
+      "step": 11951
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.521198255747968,
+      "learning_rate": 7.31161857483057e-07,
+      "loss": 0.3008,
+      "step": 11952
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.96851351517162,
+      "learning_rate": 7.305757832492916e-07,
+      "loss": 0.3143,
+      "step": 11953
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.006478291003928,
+      "learning_rate": 7.299899254847292e-07,
+      "loss": 0.2812,
+      "step": 11954
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.702471330044935,
+      "learning_rate": 7.294042842190719e-07,
+      "loss": 0.3384,
+      "step": 11955
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.958483149221776,
+      "learning_rate": 7.288188594820173e-07,
+      "loss": 0.3167,
+      "step": 11956
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1779701291997506,
+      "learning_rate": 7.28233651303244e-07,
+      "loss": 0.2932,
+      "step": 11957
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8165685145891337,
+      "learning_rate": 7.276486597124244e-07,
+      "loss": 0.321,
+      "step": 11958
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.093870044996208,
+      "learning_rate": 7.270638847392192e-07,
+      "loss": 0.3064,
+      "step": 11959
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 15.915922924483441,
+      "learning_rate": 7.264793264132775e-07,
+      "loss": 0.3368,
+      "step": 11960
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.277716862645978,
+      "learning_rate": 7.258949847642393e-07,
+      "loss": 0.3064,
+      "step": 11961
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6757953881067,
+      "learning_rate": 7.253108598217284e-07,
+      "loss": 0.3033,
+      "step": 11962
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.4914312641930896,
+      "learning_rate": 7.247269516153655e-07,
+      "loss": 0.3602,
+      "step": 11963
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.067360473324861,
+      "learning_rate": 7.241432601747527e-07,
+      "loss": 0.3355,
+      "step": 11964
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.541141075783142,
+      "learning_rate": 7.235597855294857e-07,
+      "loss": 0.3201,
+      "step": 11965
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0171624716857615,
+      "learning_rate": 7.229765277091477e-07,
+      "loss": 0.28,
+      "step": 11966
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.966968768020328,
+      "learning_rate": 7.223934867433114e-07,
+      "loss": 0.3114,
+      "step": 11967
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.062436536052987,
+      "learning_rate": 7.218106626615384e-07,
+      "loss": 0.3057,
+      "step": 11968
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.28112465297116,
+      "learning_rate": 7.212280554933804e-07,
+      "loss": 0.3165,
+      "step": 11969
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 11.259472328807865,
+      "learning_rate": 7.206456652683736e-07,
+      "loss": 0.3083,
+      "step": 11970
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.494730537926596,
+      "learning_rate": 7.200634920160488e-07,
+      "loss": 0.3087,
+      "step": 11971
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.4224272453077145,
+      "learning_rate": 7.19481535765923e-07,
+      "loss": 0.2911,
+      "step": 11972
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5756840081097745,
+      "learning_rate": 7.188997965475025e-07,
+      "loss": 0.2908,
+      "step": 11973
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.399553930018975,
+      "learning_rate": 7.183182743902839e-07,
+      "loss": 0.3078,
+      "step": 11974
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2130839569460043,
+      "learning_rate": 7.177369693237485e-07,
+      "loss": 0.2958,
+      "step": 11975
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.005463117383523,
+      "learning_rate": 7.171558813773744e-07,
+      "loss": 0.3111,
+      "step": 11976
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.7565381744351067,
+      "learning_rate": 7.165750105806202e-07,
+      "loss": 0.2918,
+      "step": 11977
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.6490233943953,
+      "learning_rate": 7.159943569629385e-07,
+      "loss": 0.2771,
+      "step": 11978
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.083191040924307,
+      "learning_rate": 7.154139205537702e-07,
+      "loss": 0.3263,
+      "step": 11979
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.8078986525219474,
+      "learning_rate": 7.148337013825441e-07,
+      "loss": 0.3177,
+      "step": 11980
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.732948558881704,
+      "learning_rate": 7.142536994786786e-07,
+      "loss": 0.2799,
+      "step": 11981
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9319468745205635,
+      "learning_rate": 7.136739148715832e-07,
+      "loss": 0.316,
+      "step": 11982
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.9679717440588234,
+      "learning_rate": 7.13094347590651e-07,
+      "loss": 0.3162,
+      "step": 11983
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 22.407516343469805,
+      "learning_rate": 7.125149976652684e-07,
+      "loss": 0.3121,
+      "step": 11984
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.6119024026107205,
+      "learning_rate": 7.119358651248104e-07,
+      "loss": 0.2966,
+      "step": 11985
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.572265378560853,
+      "learning_rate": 7.113569499986401e-07,
+      "loss": 0.3058,
+      "step": 11986
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7529187080798048,
+      "learning_rate": 7.1077825231611e-07,
+      "loss": 0.322,
+      "step": 11987
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.595332149236414,
+      "learning_rate": 7.101997721065595e-07,
+      "loss": 0.2888,
+      "step": 11988
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2055147309125687,
+      "learning_rate": 7.09621509399322e-07,
+      "loss": 0.349,
+      "step": 11989
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.621917173266694,
+      "learning_rate": 7.090434642237132e-07,
+      "loss": 0.3379,
+      "step": 11990
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 10.991649118241256,
+      "learning_rate": 7.084656366090448e-07,
+      "loss": 0.329,
+      "step": 11991
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.909001301681539,
+      "learning_rate": 7.078880265846117e-07,
+      "loss": 0.2937,
+      "step": 11992
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3367032664825884,
+      "learning_rate": 7.073106341796998e-07,
+      "loss": 0.2845,
+      "step": 11993
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.94439957496149,
+      "learning_rate": 7.067334594235853e-07,
+      "loss": 0.3239,
+      "step": 11994
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.393538321139282,
+      "learning_rate": 7.061565023455313e-07,
+      "loss": 0.3116,
+      "step": 11995
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.8672113475184084,
+      "learning_rate": 7.055797629747929e-07,
+      "loss": 0.3038,
+      "step": 11996
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 6.933045014697639,
+      "learning_rate": 7.050032413406088e-07,
+      "loss": 0.2823,
+      "step": 11997
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.125323513357213,
+      "learning_rate": 7.044269374722118e-07,
+      "loss": 0.3117,
+      "step": 11998
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3798851433672645,
+      "learning_rate": 7.038508513988213e-07,
+      "loss": 0.3147,
+      "step": 11999
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.019812820079353,
+      "learning_rate": 7.032749831496466e-07,
+      "loss": 0.3236,
+      "step": 12000
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0360554067832313,
+      "learning_rate": 7.02699332753885e-07,
+      "loss": 0.2958,
+      "step": 12001
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.3585017942073447,
+      "learning_rate": 7.021239002407249e-07,
+      "loss": 0.2938,
+      "step": 12002
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5353374822360073,
+      "learning_rate": 7.015486856393378e-07,
+      "loss": 0.2955,
+      "step": 12003
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.791703413572335,
+      "learning_rate": 7.009736889788937e-07,
+      "loss": 0.3118,
+      "step": 12004
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 8.1099861408718,
+      "learning_rate": 7.003989102885423e-07,
+      "loss": 0.3001,
+      "step": 12005
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.451059680228504,
+      "learning_rate": 6.998243495974278e-07,
+      "loss": 0.3043,
+      "step": 12006
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.495756346829054,
+      "learning_rate": 6.992500069346808e-07,
+      "loss": 0.3221,
+      "step": 12007
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.179367134043835,
+      "learning_rate": 6.986758823294226e-07,
+      "loss": 0.3095,
+      "step": 12008
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28176275537042594,
+      "learning_rate": 6.981019758107632e-07,
+      "loss": 0.1016,
+      "step": 12009
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.514085678106478,
+      "learning_rate": 6.975282874077994e-07,
+      "loss": 0.2831,
+      "step": 12010
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.042161065405335,
+      "learning_rate": 6.969548171496182e-07,
+      "loss": 0.3479,
+      "step": 12011
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.738057908038139,
+      "learning_rate": 6.963815650652972e-07,
+      "loss": 0.2737,
+      "step": 12012
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.866982662722566,
+      "learning_rate": 6.958085311839008e-07,
+      "loss": 0.3156,
+      "step": 12013
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.527874635202948,
+      "learning_rate": 6.952357155344835e-07,
+      "loss": 0.3343,
+      "step": 12014
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.09671811965632,
+      "learning_rate": 6.946631181460889e-07,
+      "loss": 0.3315,
+      "step": 12015
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.541116024498059,
+      "learning_rate": 6.940907390477458e-07,
+      "loss": 0.3381,
+      "step": 12016
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.135306088375876,
+      "learning_rate": 6.9351857826848e-07,
+      "loss": 0.3282,
+      "step": 12017
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7361635167274923,
+      "learning_rate": 6.929466358372966e-07,
+      "loss": 0.3331,
+      "step": 12018
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.0623017606459335,
+      "learning_rate": 6.92374911783198e-07,
+      "loss": 0.3308,
+      "step": 12019
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4258891076087923,
+      "learning_rate": 6.918034061351697e-07,
+      "loss": 0.2985,
+      "step": 12020
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.97977106752399,
+      "learning_rate": 6.912321189221888e-07,
+      "loss": 0.2753,
+      "step": 12021
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.604741544009696,
+      "learning_rate": 6.906610501732225e-07,
+      "loss": 0.2858,
+      "step": 12022
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.37027288492828,
+      "learning_rate": 6.90090199917221e-07,
+      "loss": 0.3176,
+      "step": 12023
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9832899349809883,
+      "learning_rate": 6.895195681831324e-07,
+      "loss": 0.2912,
+      "step": 12024
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.916796072792595,
+      "learning_rate": 6.889491549998862e-07,
+      "loss": 0.335,
+      "step": 12025
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6880751626114896,
+      "learning_rate": 6.883789603964042e-07,
+      "loss": 0.3543,
+      "step": 12026
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4836961318124557,
+      "learning_rate": 6.878089844015962e-07,
+      "loss": 0.3254,
+      "step": 12027
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.251277856097513,
+      "learning_rate": 6.872392270443624e-07,
+      "loss": 0.3299,
+      "step": 12028
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.100800245333517,
+      "learning_rate": 6.866696883535895e-07,
+      "loss": 0.3151,
+      "step": 12029
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.662324071164728,
+      "learning_rate": 6.861003683581563e-07,
+      "loss": 0.3063,
+      "step": 12030
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.422261935368765,
+      "learning_rate": 6.855312670869252e-07,
+      "loss": 0.3124,
+      "step": 12031
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.673076975864218,
+      "learning_rate": 6.849623845687547e-07,
+      "loss": 0.3086,
+      "step": 12032
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0045650521561034,
+      "learning_rate": 6.843937208324852e-07,
+      "loss": 0.2998,
+      "step": 12033
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.490929529177271,
+      "learning_rate": 6.838252759069513e-07,
+      "loss": 0.2805,
+      "step": 12034
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.972130077022578,
+      "learning_rate": 6.832570498209728e-07,
+      "loss": 0.3162,
+      "step": 12035
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2120901272061904,
+      "learning_rate": 6.826890426033617e-07,
+      "loss": 0.333,
+      "step": 12036
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.364880073340907,
+      "learning_rate": 6.82121254282917e-07,
+      "loss": 0.284,
+      "step": 12037
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6730382864413404,
+      "learning_rate": 6.815536848884247e-07,
+      "loss": 0.3219,
+      "step": 12038
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9756261242993625,
+      "learning_rate": 6.80986334448665e-07,
+      "loss": 0.3194,
+      "step": 12039
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.517417404068392,
+      "learning_rate": 6.804192029924011e-07,
+      "loss": 0.3278,
+      "step": 12040
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.595737905947653,
+      "learning_rate": 6.798522905483884e-07,
+      "loss": 0.2954,
+      "step": 12041
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5493177497214394,
+      "learning_rate": 6.792855971453716e-07,
+      "loss": 0.2822,
+      "step": 12042
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3688241213232275,
+      "learning_rate": 6.787191228120843e-07,
+      "loss": 0.3298,
+      "step": 12043
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7936362616183477,
+      "learning_rate": 6.781528675772437e-07,
+      "loss": 0.3128,
+      "step": 12044
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.985984399040915,
+      "learning_rate": 6.77586831469565e-07,
+      "loss": 0.3295,
+      "step": 12045
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.396670122376385,
+      "learning_rate": 6.770210145177447e-07,
+      "loss": 0.3261,
+      "step": 12046
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4722362666847886,
+      "learning_rate": 6.764554167504717e-07,
+      "loss": 0.2792,
+      "step": 12047
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.340751453466207,
+      "learning_rate": 6.758900381964228e-07,
+      "loss": 0.2831,
+      "step": 12048
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.602859112083484,
+      "learning_rate": 6.753248788842642e-07,
+      "loss": 0.3382,
+      "step": 12049
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7884422822784996,
+      "learning_rate": 6.747599388426523e-07,
+      "loss": 0.3125,
+      "step": 12050
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3939274110970046,
+      "learning_rate": 6.741952181002265e-07,
+      "loss": 0.2872,
+      "step": 12051
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 13.912011869068387,
+      "learning_rate": 6.736307166856243e-07,
+      "loss": 0.3467,
+      "step": 12052
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29487690684757084,
+      "learning_rate": 6.73066434627464e-07,
+      "loss": 0.106,
+      "step": 12053
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.524781381084745,
+      "learning_rate": 6.725023719543571e-07,
+      "loss": 0.3028,
+      "step": 12054
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.269312602479103,
+      "learning_rate": 6.719385286949026e-07,
+      "loss": 0.311,
+      "step": 12055
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.486370201829837,
+      "learning_rate": 6.713749048776885e-07,
+      "loss": 0.3234,
+      "step": 12056
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.823469297738695,
+      "learning_rate": 6.708115005312921e-07,
+      "loss": 0.3268,
+      "step": 12057
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.828390123519651,
+      "learning_rate": 6.702483156842798e-07,
+      "loss": 0.3102,
+      "step": 12058
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 9.655168468575953,
+      "learning_rate": 6.696853503652045e-07,
+      "loss": 0.3357,
+      "step": 12059
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9395919431111763,
+      "learning_rate": 6.691226046026106e-07,
+      "loss": 0.3252,
+      "step": 12060
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5915936160117137,
+      "learning_rate": 6.685600784250313e-07,
+      "loss": 0.313,
+      "step": 12061
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.818517896022799,
+      "learning_rate": 6.67997771860987e-07,
+      "loss": 0.2792,
+      "step": 12062
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.538553574584318,
+      "learning_rate": 6.67435684938989e-07,
+      "loss": 0.2791,
+      "step": 12063
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.991237554259082,
+      "learning_rate": 6.668738176875339e-07,
+      "loss": 0.2756,
+      "step": 12064
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9993846321359947,
+      "learning_rate": 6.663121701351138e-07,
+      "loss": 0.2998,
+      "step": 12065
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5501703355663037,
+      "learning_rate": 6.657507423102005e-07,
+      "loss": 0.303,
+      "step": 12066
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.828414258366519,
+      "learning_rate": 6.651895342412639e-07,
+      "loss": 0.3248,
+      "step": 12067
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.967134933435662,
+      "learning_rate": 6.646285459567559e-07,
+      "loss": 0.3043,
+      "step": 12068
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.050570615030864,
+      "learning_rate": 6.640677774851207e-07,
+      "loss": 0.3097,
+      "step": 12069
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2639624217183574,
+      "learning_rate": 6.635072288547905e-07,
+      "loss": 0.3165,
+      "step": 12070
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.20853391209256,
+      "learning_rate": 6.629469000941863e-07,
+      "loss": 0.2975,
+      "step": 12071
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.989386031563707,
+      "learning_rate": 6.623867912317189e-07,
+      "loss": 0.3178,
+      "step": 12072
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.270273506452496,
+      "learning_rate": 6.618269022957846e-07,
+      "loss": 0.3155,
+      "step": 12073
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.234627287207794,
+      "learning_rate": 6.61267233314773e-07,
+      "loss": 0.3308,
+      "step": 12074
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8102160664947706,
+      "learning_rate": 6.6070778431706e-07,
+      "loss": 0.3212,
+      "step": 12075
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.200562762108069,
+      "learning_rate": 6.60148555331011e-07,
+      "loss": 0.3064,
+      "step": 12076
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4367288785243164,
+      "learning_rate": 6.595895463849805e-07,
+      "loss": 0.3145,
+      "step": 12077
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8184952727903063,
+      "learning_rate": 6.590307575073118e-07,
+      "loss": 0.2814,
+      "step": 12078
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.575603236405616,
+      "learning_rate": 6.584721887263334e-07,
+      "loss": 0.3296,
+      "step": 12079
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.9967071120369373,
+      "learning_rate": 6.579138400703716e-07,
+      "loss": 0.314,
+      "step": 12080
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.708720918485679,
+      "learning_rate": 6.573557115677315e-07,
+      "loss": 0.3359,
+      "step": 12081
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.3070885141965194,
+      "learning_rate": 6.567978032467126e-07,
+      "loss": 0.2926,
+      "step": 12082
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.737852453444905,
+      "learning_rate": 6.562401151356024e-07,
+      "loss": 0.2822,
+      "step": 12083
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.1321618323607945,
+      "learning_rate": 6.556826472626771e-07,
+      "loss": 0.3119,
+      "step": 12084
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3129456605272833,
+      "learning_rate": 6.551253996562018e-07,
+      "loss": 0.1069,
+      "step": 12085
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8744819191291544,
+      "learning_rate": 6.54568372344429e-07,
+      "loss": 0.3296,
+      "step": 12086
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.603709235374268,
+      "learning_rate": 6.540115653556017e-07,
+      "loss": 0.3356,
+      "step": 12087
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.727881228357336,
+      "learning_rate": 6.534549787179517e-07,
+      "loss": 0.2848,
+      "step": 12088
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.960642746565579,
+      "learning_rate": 6.528986124596986e-07,
+      "loss": 0.3121,
+      "step": 12089
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.424211751178686,
+      "learning_rate": 6.523424666090516e-07,
+      "loss": 0.3321,
+      "step": 12090
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.577897588264472,
+      "learning_rate": 6.517865411942098e-07,
+      "loss": 0.3505,
+      "step": 12091
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9481296388199967,
+      "learning_rate": 6.512308362433568e-07,
+      "loss": 0.2995,
+      "step": 12092
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.068193205491163,
+      "learning_rate": 6.506753517846714e-07,
+      "loss": 0.3072,
+      "step": 12093
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.1795359207818015,
+      "learning_rate": 6.501200878463143e-07,
+      "loss": 0.3185,
+      "step": 12094
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2012087924582255,
+      "learning_rate": 6.495650444564433e-07,
+      "loss": 0.3281,
+      "step": 12095
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.978185100297637,
+      "learning_rate": 6.490102216431964e-07,
+      "loss": 0.2996,
+      "step": 12096
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.571038639894814,
+      "learning_rate": 6.484556194347058e-07,
+      "loss": 0.3104,
+      "step": 12097
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.057155985218277,
+      "learning_rate": 6.479012378590915e-07,
+      "loss": 0.3042,
+      "step": 12098
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.375590602369272,
+      "learning_rate": 6.473470769444595e-07,
+      "loss": 0.2782,
+      "step": 12099
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.384773183132006,
+      "learning_rate": 6.467931367189106e-07,
+      "loss": 0.3203,
+      "step": 12100
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.0538319482495817,
+      "learning_rate": 6.462394172105279e-07,
+      "loss": 0.3219,
+      "step": 12101
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7270343479787043,
+      "learning_rate": 6.456859184473874e-07,
+      "loss": 0.312,
+      "step": 12102
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.374771354851423,
+      "learning_rate": 6.451326404575525e-07,
+      "loss": 0.2897,
+      "step": 12103
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.168404099447328,
+      "learning_rate": 6.44579583269076e-07,
+      "loss": 0.2911,
+      "step": 12104
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 11.8090551651812,
+      "learning_rate": 6.440267469099981e-07,
+      "loss": 0.3144,
+      "step": 12105
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.149407508021103,
+      "learning_rate": 6.434741314083509e-07,
+      "loss": 0.3042,
+      "step": 12106
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.9821713810071504,
+      "learning_rate": 6.429217367921498e-07,
+      "loss": 0.3121,
+      "step": 12107
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.271079873782492,
+      "learning_rate": 6.423695630894061e-07,
+      "loss": 0.3246,
+      "step": 12108
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.574706038567372,
+      "learning_rate": 6.418176103281137e-07,
+      "loss": 0.3296,
+      "step": 12109
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.788540643899643,
+      "learning_rate": 6.412658785362586e-07,
+      "loss": 0.3101,
+      "step": 12110
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.165550497746831,
+      "learning_rate": 6.40714367741816e-07,
+      "loss": 0.323,
+      "step": 12111
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.8974500964928236,
+      "learning_rate": 6.401630779727453e-07,
+      "loss": 0.2913,
+      "step": 12112
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8330168303901186,
+      "learning_rate": 6.39612009257003e-07,
+      "loss": 0.3011,
+      "step": 12113
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5715463310900137,
+      "learning_rate": 6.39061161622525e-07,
+      "loss": 0.2966,
+      "step": 12114
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.2853532979607091,
+      "learning_rate": 6.385105350972432e-07,
+      "loss": 0.1059,
+      "step": 12115
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4696283741778573,
+      "learning_rate": 6.379601297090743e-07,
+      "loss": 0.2987,
+      "step": 12116
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.101535492671578,
+      "learning_rate": 6.374099454859256e-07,
+      "loss": 0.2923,
+      "step": 12117
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 8.618853969240636,
+      "learning_rate": 6.368599824556926e-07,
+      "loss": 0.3012,
+      "step": 12118
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.29122817725465017,
+      "learning_rate": 6.363102406462607e-07,
+      "loss": 0.1165,
+      "step": 12119
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6889509709253274,
+      "learning_rate": 6.357607200854999e-07,
+      "loss": 0.313,
+      "step": 12120
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.02724058694654,
+      "learning_rate": 6.352114208012761e-07,
+      "loss": 0.2872,
+      "step": 12121
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.670590904172956,
+      "learning_rate": 6.346623428214366e-07,
+      "loss": 0.2916,
+      "step": 12122
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.68750145189773,
+      "learning_rate": 6.341134861738229e-07,
+      "loss": 0.311,
+      "step": 12123
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.547391660201276,
+      "learning_rate": 6.335648508862624e-07,
+      "loss": 0.3092,
+      "step": 12124
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.1290984279788665,
+      "learning_rate": 6.330164369865721e-07,
+      "loss": 0.2938,
+      "step": 12125
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.313717524981235,
+      "learning_rate": 6.324682445025592e-07,
+      "loss": 0.3033,
+      "step": 12126
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.223376711634295,
+      "learning_rate": 6.31920273462015e-07,
+      "loss": 0.3341,
+      "step": 12127
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 27.48760442279471,
+      "learning_rate": 6.313725238927271e-07,
+      "loss": 0.3039,
+      "step": 12128
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.500966013245234,
+      "learning_rate": 6.308249958224638e-07,
+      "loss": 0.306,
+      "step": 12129
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.562380410256402,
+      "learning_rate": 6.302776892789886e-07,
+      "loss": 0.3352,
+      "step": 12130
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.728114292510857,
+      "learning_rate": 6.297306042900492e-07,
+      "loss": 0.2796,
+      "step": 12131
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4854430765944158,
+      "learning_rate": 6.291837408833856e-07,
+      "loss": 0.289,
+      "step": 12132
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.047988136422664,
+      "learning_rate": 6.286370990867241e-07,
+      "loss": 0.3355,
+      "step": 12133
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.89975496476558,
+      "learning_rate": 6.280906789277819e-07,
+      "loss": 0.2803,
+      "step": 12134
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.7708029057415895,
+      "learning_rate": 6.275444804342617e-07,
+      "loss": 0.2983,
+      "step": 12135
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.2035531786292593,
+      "learning_rate": 6.269985036338577e-07,
+      "loss": 0.2906,
+      "step": 12136
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.478002546675196,
+      "learning_rate": 6.264527485542527e-07,
+      "loss": 0.3054,
+      "step": 12137
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.001486786572064,
+      "learning_rate": 6.259072152231177e-07,
+      "loss": 0.2974,
+      "step": 12138
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 17.395151687719853,
+      "learning_rate": 6.253619036681124e-07,
+      "loss": 0.3122,
+      "step": 12139
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.017897318931592,
+      "learning_rate": 6.248168139168837e-07,
+      "loss": 0.3008,
+      "step": 12140
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.204493062221419,
+      "learning_rate": 6.242719459970714e-07,
+      "loss": 0.2958,
+      "step": 12141
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.797115296816443,
+      "learning_rate": 6.237272999362987e-07,
+      "loss": 0.2996,
+      "step": 12142
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.800912212109017,
+      "learning_rate": 6.231828757621838e-07,
+      "loss": 0.3046,
+      "step": 12143
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 7.279951438710894,
+      "learning_rate": 6.226386735023271e-07,
+      "loss": 0.2909,
+      "step": 12144
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.765541885920113,
+      "learning_rate": 6.220946931843225e-07,
+      "loss": 0.3151,
+      "step": 12145
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 6.856835452204934,
+      "learning_rate": 6.215509348357502e-07,
+      "loss": 0.2922,
+      "step": 12146
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.8125609686264528,
+      "learning_rate": 6.210073984841802e-07,
+      "loss": 0.3036,
+      "step": 12147
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.161225426140937,
+      "learning_rate": 6.204640841571724e-07,
+      "loss": 0.2862,
+      "step": 12148
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.888587444618446,
+      "learning_rate": 6.199209918822719e-07,
+      "loss": 0.3303,
+      "step": 12149
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 10.387195227484197,
+      "learning_rate": 6.193781216870154e-07,
+      "loss": 0.3135,
+      "step": 12150
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.011963363452035,
+      "learning_rate": 6.188354735989272e-07,
+      "loss": 0.3233,
+      "step": 12151
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.302903912041171,
+      "learning_rate": 6.182930476455229e-07,
+      "loss": 0.3266,
+      "step": 12152
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.647921390423324,
+      "learning_rate": 6.177508438543006e-07,
+      "loss": 0.3171,
+      "step": 12153
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.6255975931331745,
+      "learning_rate": 6.172088622527561e-07,
+      "loss": 0.2965,
+      "step": 12154
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.070605015321315,
+      "learning_rate": 6.16667102868364e-07,
+      "loss": 0.2661,
+      "step": 12155
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.664798038629186,
+      "learning_rate": 6.16125565728598e-07,
+      "loss": 0.3029,
+      "step": 12156
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.323768152792268,
+      "learning_rate": 6.155842508609111e-07,
+      "loss": 0.312,
+      "step": 12157
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.205300484605825,
+      "learning_rate": 6.150431582927508e-07,
+      "loss": 0.308,
+      "step": 12158
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.964743627917466,
+      "learning_rate": 6.145022880515516e-07,
+      "loss": 0.3281,
+      "step": 12159
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.652072091444713,
+      "learning_rate": 6.139616401647364e-07,
+      "loss": 0.3022,
+      "step": 12160
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.30290683461372375,
+      "learning_rate": 6.13421214659719e-07,
+      "loss": 0.1151,
+      "step": 12161
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.614423320886937,
+      "learning_rate": 6.128810115638972e-07,
+      "loss": 0.3199,
+      "step": 12162
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 18.057483077093686,
+      "learning_rate": 6.123410309046623e-07,
+      "loss": 0.3022,
+      "step": 12163
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.174960237271364,
+      "learning_rate": 6.11801272709392e-07,
+      "loss": 0.3229,
+      "step": 12164
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.137566218909551,
+      "learning_rate": 6.112617370054535e-07,
+      "loss": 0.325,
+      "step": 12165
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6983863362876668,
+      "learning_rate": 6.107224238202025e-07,
+      "loss": 0.3047,
+      "step": 12166
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.909873715812201,
+      "learning_rate": 6.101833331809843e-07,
+      "loss": 0.3146,
+      "step": 12167
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.472466664401281,
+      "learning_rate": 6.096444651151284e-07,
+      "loss": 0.3147,
+      "step": 12168
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4920864139309717,
+      "learning_rate": 6.091058196499616e-07,
+      "loss": 0.2989,
+      "step": 12169
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.600651523199611,
+      "learning_rate": 6.085673968127897e-07,
+      "loss": 0.3049,
+      "step": 12170
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.774567671311733,
+      "learning_rate": 6.080291966309165e-07,
+      "loss": 0.3133,
+      "step": 12171
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7186124727215804,
+      "learning_rate": 6.074912191316262e-07,
+      "loss": 0.3135,
+      "step": 12172
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8852411337051214,
+      "learning_rate": 6.06953464342197e-07,
+      "loss": 0.2843,
+      "step": 12173
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.536344509445409,
+      "learning_rate": 6.064159322898955e-07,
+      "loss": 0.2865,
+      "step": 12174
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.670412236264602,
+      "learning_rate": 6.058786230019714e-07,
+      "loss": 0.3291,
+      "step": 12175
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.408108378507274,
+      "learning_rate": 6.053415365056731e-07,
+      "loss": 0.3151,
+      "step": 12176
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.279800168479291,
+      "learning_rate": 6.048046728282281e-07,
+      "loss": 0.3348,
+      "step": 12177
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3901945537205713,
+      "learning_rate": 6.04268031996858e-07,
+      "loss": 0.314,
+      "step": 12178
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.999255964652558,
+      "learning_rate": 6.037316140387717e-07,
+      "loss": 0.2992,
+      "step": 12179
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2850647605715378,
+      "learning_rate": 6.031954189811662e-07,
+      "loss": 0.1101,
+      "step": 12180
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.934919555062836,
+      "learning_rate": 6.026594468512281e-07,
+      "loss": 0.3441,
+      "step": 12181
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.181534712847959,
+      "learning_rate": 6.021236976761336e-07,
+      "loss": 0.2993,
+      "step": 12182
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.615937537810575,
+      "learning_rate": 6.015881714830429e-07,
+      "loss": 0.2935,
+      "step": 12183
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.52281835997026,
+      "learning_rate": 6.010528682991129e-07,
+      "loss": 0.3178,
+      "step": 12184
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.157913637432179,
+      "learning_rate": 6.005177881514817e-07,
+      "loss": 0.312,
+      "step": 12185
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.369085432576123,
+      "learning_rate": 5.999829310672794e-07,
+      "loss": 0.2644,
+      "step": 12186
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.410187183855646,
+      "learning_rate": 5.994482970736259e-07,
+      "loss": 0.3262,
+      "step": 12187
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 23.866486841251525,
+      "learning_rate": 5.989138861976246e-07,
+      "loss": 0.307,
+      "step": 12188
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.382129899058032,
+      "learning_rate": 5.983796984663765e-07,
+      "loss": 0.3255,
+      "step": 12189
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.126185236493419,
+      "learning_rate": 5.978457339069621e-07,
+      "loss": 0.3305,
+      "step": 12190
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.031280448951855,
+      "learning_rate": 5.97311992546456e-07,
+      "loss": 0.273,
+      "step": 12191
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2894288949368049,
+      "learning_rate": 5.967784744119204e-07,
+      "loss": 0.0989,
+      "step": 12192
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7778259410064967,
+      "learning_rate": 5.962451795304052e-07,
+      "loss": 0.3081,
+      "step": 12193
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.174687246102484,
+      "learning_rate": 5.957121079289497e-07,
+      "loss": 0.3143,
+      "step": 12194
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7300643004926433,
+      "learning_rate": 5.951792596345834e-07,
+      "loss": 0.3308,
+      "step": 12195
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.9906746905878085,
+      "learning_rate": 5.946466346743191e-07,
+      "loss": 0.3345,
+      "step": 12196
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.480424518500916,
+      "learning_rate": 5.941142330751665e-07,
+      "loss": 0.3556,
+      "step": 12197
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 10.164559674579786,
+      "learning_rate": 5.935820548641169e-07,
+      "loss": 0.3245,
+      "step": 12198
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.45557715943195,
+      "learning_rate": 5.930501000681527e-07,
+      "loss": 0.3209,
+      "step": 12199
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.327855835310053,
+      "learning_rate": 5.925183687142477e-07,
+      "loss": 0.308,
+      "step": 12200
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3479257355818155,
+      "learning_rate": 5.919868608293572e-07,
+      "loss": 0.2931,
+      "step": 12201
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2970669156771057,
+      "learning_rate": 5.914555764404356e-07,
+      "loss": 0.1102,
+      "step": 12202
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.044727497846081,
+      "learning_rate": 5.909245155744148e-07,
+      "loss": 0.3047,
+      "step": 12203
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.025019710072123,
+      "learning_rate": 5.903936782582253e-07,
+      "loss": 0.3206,
+      "step": 12204
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9584487104412314,
+      "learning_rate": 5.898630645187786e-07,
+      "loss": 0.3127,
+      "step": 12205
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.168766710290918,
+      "learning_rate": 5.893326743829797e-07,
+      "loss": 0.2969,
+      "step": 12206
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.127312943479799,
+      "learning_rate": 5.888025078777193e-07,
+      "loss": 0.3631,
+      "step": 12207
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.749182001868517,
+      "learning_rate": 5.882725650298787e-07,
+      "loss": 0.3353,
+      "step": 12208
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.23695973332353,
+      "learning_rate": 5.87742845866327e-07,
+      "loss": 0.3157,
+      "step": 12209
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2220940869393506,
+      "learning_rate": 5.872133504139232e-07,
+      "loss": 0.2579,
+      "step": 12210
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.016843630869282,
+      "learning_rate": 5.866840786995121e-07,
+      "loss": 0.3124,
+      "step": 12211
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.521903788010013,
+      "learning_rate": 5.861550307499298e-07,
+      "loss": 0.3058,
+      "step": 12212
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.587406986859432,
+      "learning_rate": 5.856262065919998e-07,
+      "loss": 0.3062,
+      "step": 12213
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.569722755591097,
+      "learning_rate": 5.850976062525354e-07,
+      "loss": 0.3259,
+      "step": 12214
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9037149671607065,
+      "learning_rate": 5.845692297583378e-07,
+      "loss": 0.31,
+      "step": 12215
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.989724590512634,
+      "learning_rate": 5.840410771361948e-07,
+      "loss": 0.3538,
+      "step": 12216
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.442233754105701,
+      "learning_rate": 5.835131484128886e-07,
+      "loss": 0.3111,
+      "step": 12217
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.302455588545568,
+      "learning_rate": 5.829854436151821e-07,
+      "loss": 0.2784,
+      "step": 12218
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.298970011781028,
+      "learning_rate": 5.824579627698351e-07,
+      "loss": 0.2888,
+      "step": 12219
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3012879131171192,
+      "learning_rate": 5.819307059035894e-07,
+      "loss": 0.1104,
+      "step": 12220
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.3915689689846475,
+      "learning_rate": 5.814036730431793e-07,
+      "loss": 0.3062,
+      "step": 12221
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.777794659721486,
+      "learning_rate": 5.808768642153257e-07,
+      "loss": 0.2963,
+      "step": 12222
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.7828314995534646,
+      "learning_rate": 5.803502794467402e-07,
+      "loss": 0.3282,
+      "step": 12223
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.2846836191193242,
+      "learning_rate": 5.798239187641208e-07,
+      "loss": 0.1048,
+      "step": 12224
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4634197676493055,
+      "learning_rate": 5.792977821941548e-07,
+      "loss": 0.3183,
+      "step": 12225
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.28566696889354637,
+      "learning_rate": 5.787718697635187e-07,
+      "loss": 0.1045,
+      "step": 12226
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.044880320420006,
+      "learning_rate": 5.782461814988782e-07,
+      "loss": 0.3044,
+      "step": 12227
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.971455406979077,
+      "learning_rate": 5.777207174268873e-07,
+      "loss": 0.3315,
+      "step": 12228
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.084848793452472,
+      "learning_rate": 5.771954775741851e-07,
+      "loss": 0.314,
+      "step": 12229
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.446049289752407,
+      "learning_rate": 5.766704619674068e-07,
+      "loss": 0.2934,
+      "step": 12230
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.2417948347589447,
+      "learning_rate": 5.761456706331676e-07,
+      "loss": 0.2872,
+      "step": 12231
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0768458475775953,
+      "learning_rate": 5.756211035980791e-07,
+      "loss": 0.2982,
+      "step": 12232
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.282636339510386,
+      "learning_rate": 5.750967608887359e-07,
+      "loss": 0.2833,
+      "step": 12233
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.936012776363485,
+      "learning_rate": 5.745726425317244e-07,
+      "loss": 0.3804,
+      "step": 12234
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.2707799297023765,
+      "learning_rate": 5.740487485536178e-07,
+      "loss": 0.3038,
+      "step": 12235
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.789436724837847,
+      "learning_rate": 5.735250789809787e-07,
+      "loss": 0.2928,
+      "step": 12236
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.229933648944233,
+      "learning_rate": 5.730016338403599e-07,
+      "loss": 0.2908,
+      "step": 12237
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.44147319555181,
+      "learning_rate": 5.724784131582983e-07,
+      "loss": 0.3049,
+      "step": 12238
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 8.632046081786603,
+      "learning_rate": 5.719554169613245e-07,
+      "loss": 0.3082,
+      "step": 12239
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.758870609254971,
+      "learning_rate": 5.71432645275955e-07,
+      "loss": 0.3041,
+      "step": 12240
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.173953270605839,
+      "learning_rate": 5.709100981286952e-07,
+      "loss": 0.3275,
+      "step": 12241
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.780132379209636,
+      "learning_rate": 5.703877755460402e-07,
+      "loss": 0.3049,
+      "step": 12242
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.348737597950869,
+      "learning_rate": 5.698656775544726e-07,
+      "loss": 0.3199,
+      "step": 12243
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 7.152262517951291,
+      "learning_rate": 5.693438041804622e-07,
+      "loss": 0.3144,
+      "step": 12244
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.020529742328518,
+      "learning_rate": 5.688221554504725e-07,
+      "loss": 0.2991,
+      "step": 12245
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.600315906024421,
+      "learning_rate": 5.683007313909495e-07,
+      "loss": 0.3265,
+      "step": 12246
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4530144484645713,
+      "learning_rate": 5.677795320283308e-07,
+      "loss": 0.2914,
+      "step": 12247
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.338737960692442,
+      "learning_rate": 5.672585573890438e-07,
+      "loss": 0.311,
+      "step": 12248
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.407350361010243,
+      "learning_rate": 5.667378074995017e-07,
+      "loss": 0.2706,
+      "step": 12249
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.01341619844948,
+      "learning_rate": 5.662172823861089e-07,
+      "loss": 0.3063,
+      "step": 12250
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.9567201656700535,
+      "learning_rate": 5.656969820752551e-07,
+      "loss": 0.342,
+      "step": 12251
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4107087310802435,
+      "learning_rate": 5.651769065933238e-07,
+      "loss": 0.3387,
+      "step": 12252
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.161463805790278,
+      "learning_rate": 5.646570559666808e-07,
+      "loss": 0.3061,
+      "step": 12253
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1602900504764038,
+      "learning_rate": 5.641374302216856e-07,
+      "loss": 0.3005,
+      "step": 12254
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.5166514928039705,
+      "learning_rate": 5.636180293846833e-07,
+      "loss": 0.3412,
+      "step": 12255
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.27682526977461525,
+      "learning_rate": 5.630988534820097e-07,
+      "loss": 0.1012,
+      "step": 12256
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.527719213043,
+      "learning_rate": 5.625799025399875e-07,
+      "loss": 0.3212,
+      "step": 12257
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.1667400596302375,
+      "learning_rate": 5.620611765849299e-07,
+      "loss": 0.2792,
+      "step": 12258
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.6329670330001766,
+      "learning_rate": 5.615426756431342e-07,
+      "loss": 0.3088,
+      "step": 12259
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 44.65337372434193,
+      "learning_rate": 5.610243997408932e-07,
+      "loss": 0.3123,
+      "step": 12260
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 13.15779856048496,
+      "learning_rate": 5.605063489044827e-07,
+      "loss": 0.2955,
+      "step": 12261
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8459251990800047,
+      "learning_rate": 5.599885231601693e-07,
+      "loss": 0.2737,
+      "step": 12262
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.000193654169188,
+      "learning_rate": 5.594709225342094e-07,
+      "loss": 0.3163,
+      "step": 12263
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.442891011399117,
+      "learning_rate": 5.589535470528424e-07,
+      "loss": 0.3124,
+      "step": 12264
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.207493361908439,
+      "learning_rate": 5.584363967423056e-07,
+      "loss": 0.3075,
+      "step": 12265
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4944500790669872,
+      "learning_rate": 5.579194716288161e-07,
+      "loss": 0.3052,
+      "step": 12266
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.962150550814525,
+      "learning_rate": 5.574027717385844e-07,
+      "loss": 0.3057,
+      "step": 12267
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.9759669406814,
+      "learning_rate": 5.568862970978078e-07,
+      "loss": 0.2821,
+      "step": 12268
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5566856329022327,
+      "learning_rate": 5.563700477326733e-07,
+      "loss": 0.2874,
+      "step": 12269
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.348780837624702,
+      "learning_rate": 5.558540236693555e-07,
+      "loss": 0.3089,
+      "step": 12270
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.645251015623493,
+      "learning_rate": 5.553382249340188e-07,
+      "loss": 0.3167,
+      "step": 12271
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.6906300911528485,
+      "learning_rate": 5.548226515528133e-07,
+      "loss": 0.3183,
+      "step": 12272
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.219460892064709,
+      "learning_rate": 5.54307303551882e-07,
+      "loss": 0.3101,
+      "step": 12273
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.347658486515645,
+      "learning_rate": 5.537921809573526e-07,
+      "loss": 0.3208,
+      "step": 12274
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.225785366439998,
+      "learning_rate": 5.532772837953437e-07,
+      "loss": 0.3051,
+      "step": 12275
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.100306347277218,
+      "learning_rate": 5.527626120919627e-07,
+      "loss": 0.3047,
+      "step": 12276
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.498393603778524,
+      "learning_rate": 5.522481658733014e-07,
+      "loss": 0.2988,
+      "step": 12277
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0567352461285835,
+      "learning_rate": 5.517339451654469e-07,
+      "loss": 0.3464,
+      "step": 12278
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.796833111773066,
+      "learning_rate": 5.512199499944682e-07,
+      "loss": 0.3095,
+      "step": 12279
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0047291018493607,
+      "learning_rate": 5.507061803864294e-07,
+      "loss": 0.3167,
+      "step": 12280
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.291752995623387,
+      "learning_rate": 5.501926363673777e-07,
+      "loss": 0.3428,
+      "step": 12281
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.411454631653404,
+      "learning_rate": 5.496793179633503e-07,
+      "loss": 0.3084,
+      "step": 12282
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.18046798310176,
+      "learning_rate": 5.491662252003743e-07,
+      "loss": 0.2813,
+      "step": 12283
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.7392346687336344,
+      "learning_rate": 5.486533581044657e-07,
+      "loss": 0.3184,
+      "step": 12284
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 24.072171768542347,
+      "learning_rate": 5.481407167016262e-07,
+      "loss": 0.3113,
+      "step": 12285
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.800158107211954,
+      "learning_rate": 5.476283010178502e-07,
+      "loss": 0.2968,
+      "step": 12286
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.052834840140873,
+      "learning_rate": 5.471161110791156e-07,
+      "loss": 0.2943,
+      "step": 12287
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.453451320971913,
+      "learning_rate": 5.466041469113925e-07,
+      "loss": 0.3099,
+      "step": 12288
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 12.220170717788653,
+      "learning_rate": 5.460924085406394e-07,
+      "loss": 0.3183,
+      "step": 12289
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.275554800935374,
+      "learning_rate": 5.455808959928017e-07,
+      "loss": 0.3349,
+      "step": 12290
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.1443018706867036,
+      "learning_rate": 5.450696092938157e-07,
+      "loss": 0.2749,
+      "step": 12291
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.812420102418216,
+      "learning_rate": 5.445585484696014e-07,
+      "loss": 0.2771,
+      "step": 12292
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 11.176380274760461,
+      "learning_rate": 5.440477135460748e-07,
+      "loss": 0.315,
+      "step": 12293
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.802321191953725,
+      "learning_rate": 5.435371045491328e-07,
+      "loss": 0.2975,
+      "step": 12294
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.151645628295845,
+      "learning_rate": 5.43026721504668e-07,
+      "loss": 0.3359,
+      "step": 12295
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 5.071063211107018,
+      "learning_rate": 5.42516564438555e-07,
+      "loss": 0.3186,
+      "step": 12296
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.725620781324766,
+      "learning_rate": 5.42006633376661e-07,
+      "loss": 0.3163,
+      "step": 12297
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.358046824658599,
+      "learning_rate": 5.414969283448401e-07,
+      "loss": 0.321,
+      "step": 12298
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3868282385909754,
+      "learning_rate": 5.409874493689372e-07,
+      "loss": 0.3202,
+      "step": 12299
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6760185357832755,
+      "learning_rate": 5.404781964747818e-07,
+      "loss": 0.3325,
+      "step": 12300
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.643697813194763,
+      "learning_rate": 5.399691696881948e-07,
+      "loss": 0.3185,
+      "step": 12301
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.2605993661994095,
+      "learning_rate": 5.394603690349853e-07,
+      "loss": 0.2999,
+      "step": 12302
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1198828025585965,
+      "learning_rate": 5.389517945409506e-07,
+      "loss": 0.3594,
+      "step": 12303
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.444933089947647,
+      "learning_rate": 5.384434462318778e-07,
+      "loss": 0.2921,
+      "step": 12304
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.170847469513967,
+      "learning_rate": 5.379353241335372e-07,
+      "loss": 0.3021,
+      "step": 12305
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.102494960227173,
+      "learning_rate": 5.374274282716974e-07,
+      "loss": 0.3049,
+      "step": 12306
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3402593766753346,
+      "learning_rate": 5.369197586721042e-07,
+      "loss": 0.2995,
+      "step": 12307
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.841114117080338,
+      "learning_rate": 5.364123153605021e-07,
+      "loss": 0.3057,
+      "step": 12308
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.048440482603959,
+      "learning_rate": 5.359050983626174e-07,
+      "loss": 0.2923,
+      "step": 12309
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.711346036619448,
+      "learning_rate": 5.35398107704167e-07,
+      "loss": 0.3188,
+      "step": 12310
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.1163736327856135,
+      "learning_rate": 5.348913434108571e-07,
+      "loss": 0.3209,
+      "step": 12311
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.88574887901271,
+      "learning_rate": 5.343848055083817e-07,
+      "loss": 0.315,
+      "step": 12312
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.978669756202239,
+      "learning_rate": 5.338784940224239e-07,
+      "loss": 0.3097,
+      "step": 12313
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9911521685279925,
+      "learning_rate": 5.333724089786535e-07,
+      "loss": 0.307,
+      "step": 12314
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4112565206508823,
+      "learning_rate": 5.328665504027308e-07,
+      "loss": 0.2942,
+      "step": 12315
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.51145952932062,
+      "learning_rate": 5.323609183203038e-07,
+      "loss": 0.3336,
+      "step": 12316
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.095017970665209,
+      "learning_rate": 5.318555127570097e-07,
+      "loss": 0.3432,
+      "step": 12317
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 10.646137513954828,
+      "learning_rate": 5.313503337384729e-07,
+      "loss": 0.3159,
+      "step": 12318
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.294832244138837,
+      "learning_rate": 5.308453812903086e-07,
+      "loss": 0.2955,
+      "step": 12319
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 17.725294239790514,
+      "learning_rate": 5.303406554381157e-07,
+      "loss": 0.2954,
+      "step": 12320
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.803746199018343,
+      "learning_rate": 5.298361562074888e-07,
+      "loss": 0.3491,
+      "step": 12321
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.388820638269831,
+      "learning_rate": 5.293318836240047e-07,
+      "loss": 0.3153,
+      "step": 12322
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 8.280847760143217,
+      "learning_rate": 5.288278377132322e-07,
+      "loss": 0.2979,
+      "step": 12323
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.224467937317036,
+      "learning_rate": 5.283240185007266e-07,
+      "loss": 0.304,
+      "step": 12324
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0957267706707126,
+      "learning_rate": 5.278204260120334e-07,
+      "loss": 0.334,
+      "step": 12325
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.642564179595499,
+      "learning_rate": 5.273170602726868e-07,
+      "loss": 0.2619,
+      "step": 12326
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.212800077707894,
+      "learning_rate": 5.268139213082057e-07,
+      "loss": 0.2799,
+      "step": 12327
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.057354411350249,
+      "learning_rate": 5.263110091441037e-07,
+      "loss": 0.3303,
+      "step": 12328
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.1554888380485426,
+      "learning_rate": 5.258083238058764e-07,
+      "loss": 0.3278,
+      "step": 12329
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.952523076937615,
+      "learning_rate": 5.253058653190135e-07,
+      "loss": 0.3386,
+      "step": 12330
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.338589799305227,
+      "learning_rate": 5.24803633708989e-07,
+      "loss": 0.2844,
+      "step": 12331
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6055306198788775,
+      "learning_rate": 5.243016290012693e-07,
+      "loss": 0.3043,
+      "step": 12332
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.600432159286337,
+      "learning_rate": 5.237998512213038e-07,
+      "loss": 0.3112,
+      "step": 12333
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 232.15154205699315,
+      "learning_rate": 5.23298300394538e-07,
+      "loss": 0.3133,
+      "step": 12334
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4957884834533104,
+      "learning_rate": 5.22796976546398e-07,
+      "loss": 0.302,
+      "step": 12335
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.518281637206723,
+      "learning_rate": 5.222958797023036e-07,
+      "loss": 0.2791,
+      "step": 12336
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.090563798502178,
+      "learning_rate": 5.217950098876612e-07,
+      "loss": 0.2863,
+      "step": 12337
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9876499258558566,
+      "learning_rate": 5.212943671278664e-07,
+      "loss": 0.3167,
+      "step": 12338
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8932538679506274,
+      "learning_rate": 5.207939514483034e-07,
+      "loss": 0.307,
+      "step": 12339
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.685194071893129,
+      "learning_rate": 5.202937628743416e-07,
+      "loss": 0.345,
+      "step": 12340
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.243734912021126,
+      "learning_rate": 5.197938014313453e-07,
+      "loss": 0.3247,
+      "step": 12341
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.4461514264141595,
+      "learning_rate": 5.192940671446617e-07,
+      "loss": 0.2978,
+      "step": 12342
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9388952483704314,
+      "learning_rate": 5.187945600396282e-07,
+      "loss": 0.2908,
+      "step": 12343
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9099088235757495,
+      "learning_rate": 5.182952801415714e-07,
+      "loss": 0.2652,
+      "step": 12344
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.104749346418942,
+      "learning_rate": 5.177962274758058e-07,
+      "loss": 0.3086,
+      "step": 12345
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3292374335148245,
+      "learning_rate": 5.172974020676353e-07,
+      "loss": 0.2793,
+      "step": 12346
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.721155633292649,
+      "learning_rate": 5.16798803942351e-07,
+      "loss": 0.326,
+      "step": 12347
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.790778336466659,
+      "learning_rate": 5.163004331252308e-07,
+      "loss": 0.3181,
+      "step": 12348
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3806868452639725,
+      "learning_rate": 5.158022896415471e-07,
+      "loss": 0.3092,
+      "step": 12349
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.585337141808203,
+      "learning_rate": 5.153043735165536e-07,
+      "loss": 0.3316,
+      "step": 12350
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.794121834811668,
+      "learning_rate": 5.148066847754973e-07,
+      "loss": 0.2795,
+      "step": 12351
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8934736367732086,
+      "learning_rate": 5.143092234436125e-07,
+      "loss": 0.3308,
+      "step": 12352
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.348922711186456,
+      "learning_rate": 5.138119895461191e-07,
+      "loss": 0.3014,
+      "step": 12353
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7709003323772885,
+      "learning_rate": 5.133149831082318e-07,
+      "loss": 0.3298,
+      "step": 12354
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 151.51743489784283,
+      "learning_rate": 5.128182041551455e-07,
+      "loss": 0.3234,
+      "step": 12355
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.798661118013118,
+      "learning_rate": 5.123216527120528e-07,
+      "loss": 0.3007,
+      "step": 12356
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.7229516682248365,
+      "learning_rate": 5.118253288041258e-07,
+      "loss": 0.2867,
+      "step": 12357
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5548060955836953,
+      "learning_rate": 5.113292324565316e-07,
+      "loss": 0.3128,
+      "step": 12358
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.878345661719356,
+      "learning_rate": 5.108333636944223e-07,
+      "loss": 0.3077,
+      "step": 12359
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.101956579676861,
+      "learning_rate": 5.103377225429396e-07,
+      "loss": 0.273,
+      "step": 12360
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7372588208703155,
+      "learning_rate": 5.098423090272142e-07,
+      "loss": 0.2748,
+      "step": 12361
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.9289449217975285,
+      "learning_rate": 5.093471231723652e-07,
+      "loss": 0.3203,
+      "step": 12362
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8377779839655477,
+      "learning_rate": 5.088521650034977e-07,
+      "loss": 0.2821,
+      "step": 12363
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.995120287311216,
+      "learning_rate": 5.083574345457087e-07,
+      "loss": 0.2912,
+      "step": 12364
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.313065007471975,
+      "learning_rate": 5.078629318240814e-07,
+      "loss": 0.1154,
+      "step": 12365
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9530972070808574,
+      "learning_rate": 5.073686568636882e-07,
+      "loss": 0.3245,
+      "step": 12366
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.917467222407692,
+      "learning_rate": 5.068746096895915e-07,
+      "loss": 0.29,
+      "step": 12367
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.135854465184579,
+      "learning_rate": 5.06380790326837e-07,
+      "loss": 0.3519,
+      "step": 12368
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.155066777915182,
+      "learning_rate": 5.058871988004665e-07,
+      "loss": 0.2985,
+      "step": 12369
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.488778342374724,
+      "learning_rate": 5.053938351355037e-07,
+      "loss": 0.3062,
+      "step": 12370
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.6936191446641224,
+      "learning_rate": 5.049006993569638e-07,
+      "loss": 0.3212,
+      "step": 12371
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.536125885788964,
+      "learning_rate": 5.044077914898498e-07,
+      "loss": 0.2998,
+      "step": 12372
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.595649302541909,
+      "learning_rate": 5.039151115591534e-07,
+      "loss": 0.3114,
+      "step": 12373
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.344252669386334,
+      "learning_rate": 5.034226595898545e-07,
+      "loss": 0.3137,
+      "step": 12374
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.631361718372862,
+      "learning_rate": 5.029304356069226e-07,
+      "loss": 0.3024,
+      "step": 12375
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27406879403015394,
+      "learning_rate": 5.024384396353121e-07,
+      "loss": 0.1088,
+      "step": 12376
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.295178853639483,
+      "learning_rate": 5.019466716999699e-07,
+      "loss": 0.0858,
+      "step": 12377
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.62387277002458,
+      "learning_rate": 5.014551318258299e-07,
+      "loss": 0.278,
+      "step": 12378
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.973621985388923,
+      "learning_rate": 5.00963820037813e-07,
+      "loss": 0.3086,
+      "step": 12379
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.9037948498414505,
+      "learning_rate": 5.004727363608319e-07,
+      "loss": 0.2862,
+      "step": 12380
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.1234541613392235,
+      "learning_rate": 4.999818808197821e-07,
+      "loss": 0.3064,
+      "step": 12381
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.237830610892893,
+      "learning_rate": 4.994912534395558e-07,
+      "loss": 0.3196,
+      "step": 12382
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.27503681087082626,
+      "learning_rate": 4.990008542450237e-07,
+      "loss": 0.1034,
+      "step": 12383
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.0961113466848635,
+      "learning_rate": 4.985106832610553e-07,
+      "loss": 0.307,
+      "step": 12384
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.06151937149282,
+      "learning_rate": 4.980207405124993e-07,
+      "loss": 0.2877,
+      "step": 12385
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.045241295129897,
+      "learning_rate": 4.975310260241989e-07,
+      "loss": 0.326,
+      "step": 12386
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.194170286179292,
+      "learning_rate": 4.970415398209826e-07,
+      "loss": 0.2992,
+      "step": 12387
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8202889087378975,
+      "learning_rate": 4.965522819276692e-07,
+      "loss": 0.2926,
+      "step": 12388
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6836639147994643,
+      "learning_rate": 4.960632523690661e-07,
+      "loss": 0.3321,
+      "step": 12389
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.447628189240636,
+      "learning_rate": 4.955744511699662e-07,
+      "loss": 0.3299,
+      "step": 12390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.555537427184273,
+      "learning_rate": 4.950858783551532e-07,
+      "loss": 0.3027,
+      "step": 12391
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.719635411856872,
+      "learning_rate": 4.945975339493997e-07,
+      "loss": 0.329,
+      "step": 12392
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.2936435419329175,
+      "learning_rate": 4.941094179774652e-07,
+      "loss": 0.334,
+      "step": 12393
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.187225630984146,
+      "learning_rate": 4.936215304640984e-07,
+      "loss": 0.3149,
+      "step": 12394
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9709457565460093,
+      "learning_rate": 4.931338714340372e-07,
+      "loss": 0.3045,
+      "step": 12395
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.528420172285021,
+      "learning_rate": 4.926464409120047e-07,
+      "loss": 0.3191,
+      "step": 12396
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.095759438850193,
+      "learning_rate": 4.921592389227175e-07,
+      "loss": 0.2978,
+      "step": 12397
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4441990682336074,
+      "learning_rate": 4.916722654908757e-07,
+      "loss": 0.3104,
+      "step": 12398
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.701229443512838,
+      "learning_rate": 4.911855206411703e-07,
+      "loss": 0.2975,
+      "step": 12399
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.36073517544944,
+      "learning_rate": 4.906990043982813e-07,
+      "loss": 0.3124,
+      "step": 12400
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 11.84260061980028,
+      "learning_rate": 4.902127167868748e-07,
+      "loss": 0.2982,
+      "step": 12401
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.4204887308620995,
+      "learning_rate": 4.897266578316085e-07,
+      "loss": 0.3181,
+      "step": 12402
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4672750065528244,
+      "learning_rate": 4.892408275571236e-07,
+      "loss": 0.3023,
+      "step": 12403
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.323400136558717,
+      "learning_rate": 4.887552259880568e-07,
+      "loss": 0.3139,
+      "step": 12404
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6463333414853936,
+      "learning_rate": 4.88269853149026e-07,
+      "loss": 0.2868,
+      "step": 12405
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.4283598683634544,
+      "learning_rate": 4.877847090646421e-07,
+      "loss": 0.2592,
+      "step": 12406
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.543469877256777,
+      "learning_rate": 4.87299793759502e-07,
+      "loss": 0.3209,
+      "step": 12407
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.9052139256340674,
+      "learning_rate": 4.86815107258194e-07,
+      "loss": 0.337,
+      "step": 12408
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.167702824967463,
+      "learning_rate": 4.863306495852893e-07,
+      "loss": 0.3237,
+      "step": 12409
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.360556426947719,
+      "learning_rate": 4.858464207653552e-07,
+      "loss": 0.2918,
+      "step": 12410
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.351422904216459,
+      "learning_rate": 4.853624208229396e-07,
+      "loss": 0.3188,
+      "step": 12411
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.9226220513704115,
+      "learning_rate": 4.848786497825835e-07,
+      "loss": 0.3271,
+      "step": 12412
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.116725477749844,
+      "learning_rate": 4.843951076688158e-07,
+      "loss": 0.3061,
+      "step": 12413
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.97922014266397,
+      "learning_rate": 4.839117945061527e-07,
+      "loss": 0.3301,
+      "step": 12414
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.96401461922907,
+      "learning_rate": 4.834287103191005e-07,
+      "loss": 0.348,
+      "step": 12415
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.932837978266194,
+      "learning_rate": 4.829458551321492e-07,
+      "loss": 0.3275,
+      "step": 12416
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3383060190903655,
+      "learning_rate": 4.82463228969785e-07,
+      "loss": 0.2933,
+      "step": 12417
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.23332763313512,
+      "learning_rate": 4.819808318564745e-07,
+      "loss": 0.3159,
+      "step": 12418
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.455817479406277,
+      "learning_rate": 4.814986638166786e-07,
+      "loss": 0.2786,
+      "step": 12419
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2989370626892085,
+      "learning_rate": 4.810167248748427e-07,
+      "loss": 0.3219,
+      "step": 12420
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.505394389231803,
+      "learning_rate": 4.805350150554028e-07,
+      "loss": 0.3508,
+      "step": 12421
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.8754894950762,
+      "learning_rate": 4.800535343827834e-07,
+      "loss": 0.3267,
+      "step": 12422
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5507991151169525,
+      "learning_rate": 4.795722828813964e-07,
+      "loss": 0.2969,
+      "step": 12423
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.738646583646405,
+      "learning_rate": 4.790912605756404e-07,
+      "loss": 0.2903,
+      "step": 12424
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 9.049220861723501,
+      "learning_rate": 4.786104674899073e-07,
+      "loss": 0.3587,
+      "step": 12425
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.417610403638684,
+      "learning_rate": 4.78129903648572e-07,
+      "loss": 0.309,
+      "step": 12426
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8942443476423656,
+      "learning_rate": 4.776495690760008e-07,
+      "loss": 0.3006,
+      "step": 12427
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.3535953873234012,
+      "learning_rate": 4.771694637965491e-07,
+      "loss": 0.2751,
+      "step": 12428
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 13.99464168656142,
+      "learning_rate": 4.766895878345562e-07,
+      "loss": 0.2701,
+      "step": 12429
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.023553879331856,
+      "learning_rate": 4.762099412143567e-07,
+      "loss": 0.2826,
+      "step": 12430
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.29282798540551475,
+      "learning_rate": 4.7573052396026556e-07,
+      "loss": 0.1101,
+      "step": 12431
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 5.468856353431337,
+      "learning_rate": 4.752513360965949e-07,
+      "loss": 0.3483,
+      "step": 12432
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.406699047307421,
+      "learning_rate": 4.7477237764763705e-07,
+      "loss": 0.3139,
+      "step": 12433
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.8399349957752533,
+      "learning_rate": 4.742936486376776e-07,
+      "loss": 0.3309,
+      "step": 12434
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8927647681969826,
+      "learning_rate": 4.7381514909098867e-07,
+      "loss": 0.3103,
+      "step": 12435
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.098984638962523,
+      "learning_rate": 4.733368790318321e-07,
+      "loss": 0.3246,
+      "step": 12436
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.7863127490116293,
+      "learning_rate": 4.7285883848445624e-07,
+      "loss": 0.3056,
+      "step": 12437
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.205056460156732,
+      "learning_rate": 4.723810274731e-07,
+      "loss": 0.3194,
+      "step": 12438
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 12.918424099032665,
+      "learning_rate": 4.7190344602198844e-07,
+      "loss": 0.3303,
+      "step": 12439
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 6.046322337511577,
+      "learning_rate": 4.7142609415533615e-07,
+      "loss": 0.3407,
+      "step": 12440
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.158468347501741,
+      "learning_rate": 4.7094897189734643e-07,
+      "loss": 0.3022,
+      "step": 12441
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 7.536317231123982,
+      "learning_rate": 4.704720792722084e-07,
+      "loss": 0.3079,
+      "step": 12442
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.454324550717646,
+      "learning_rate": 4.699954163041054e-07,
+      "loss": 0.3107,
+      "step": 12443
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.80667658255354,
+      "learning_rate": 4.695189830172009e-07,
+      "loss": 0.2888,
+      "step": 12444
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 13.683878911889835,
+      "learning_rate": 4.6904277943565556e-07,
+      "loss": 0.2839,
+      "step": 12445
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.2410475732889,
+      "learning_rate": 4.685668055836101e-07,
+      "loss": 0.3084,
+      "step": 12446
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 19.74203115597599,
+      "learning_rate": 4.680910614851997e-07,
+      "loss": 0.3504,
+      "step": 12447
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.687792112857756,
+      "learning_rate": 4.676155471645444e-07,
+      "loss": 0.2842,
+      "step": 12448
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.08206931773265,
+      "learning_rate": 4.67140262645755e-07,
+      "loss": 0.322,
+      "step": 12449
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 10.26406908391244,
+      "learning_rate": 4.666652079529282e-07,
+      "loss": 0.3212,
+      "step": 12450
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6205968957174686,
+      "learning_rate": 4.6619038311015275e-07,
+      "loss": 0.2856,
+      "step": 12451
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.140799970950118,
+      "learning_rate": 4.6571578814149975e-07,
+      "loss": 0.3263,
+      "step": 12452
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.958863559507586,
+      "learning_rate": 4.652414230710345e-07,
+      "loss": 0.3009,
+      "step": 12453
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.869401557804293,
+      "learning_rate": 4.6476728792280767e-07,
+      "loss": 0.2942,
+      "step": 12454
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.2796639754711648,
+      "learning_rate": 4.642933827208596e-07,
+      "loss": 0.1021,
+      "step": 12455
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7705029431651798,
+      "learning_rate": 4.638197074892181e-07,
+      "loss": 0.3226,
+      "step": 12456
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6216018779998644,
+      "learning_rate": 4.6334626225189804e-07,
+      "loss": 0.3049,
+      "step": 12457
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.135208452378314,
+      "learning_rate": 4.628730470329068e-07,
+      "loss": 0.3314,
+      "step": 12458
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.682364023608705,
+      "learning_rate": 4.624000618562341e-07,
+      "loss": 0.338,
+      "step": 12459
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.83581184062625,
+      "learning_rate": 4.6192730674586574e-07,
+      "loss": 0.2957,
+      "step": 12460
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.5381342206872555,
+      "learning_rate": 4.614547817257675e-07,
+      "loss": 0.2956,
+      "step": 12461
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.729831571049877,
+      "learning_rate": 4.609824868198992e-07,
+      "loss": 0.358,
+      "step": 12462
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.620568558064506,
+      "learning_rate": 4.605104220522072e-07,
+      "loss": 0.3127,
+      "step": 12463
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.988586196628232,
+      "learning_rate": 4.6003858744662564e-07,
+      "loss": 0.2931,
+      "step": 12464
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0380632304166753,
+      "learning_rate": 4.595669830270788e-07,
+      "loss": 0.2996,
+      "step": 12465
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.723510251770805,
+      "learning_rate": 4.590956088174764e-07,
+      "loss": 0.3021,
+      "step": 12466
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.69043964683814,
+      "learning_rate": 4.586244648417187e-07,
+      "loss": 0.2943,
+      "step": 12467
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4841180697442606,
+      "learning_rate": 4.58153551123694e-07,
+      "loss": 0.3233,
+      "step": 12468
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4029088676322714,
+      "learning_rate": 4.576828676872791e-07,
+      "loss": 0.333,
+      "step": 12469
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.019877838501292,
+      "learning_rate": 4.5721241455633783e-07,
+      "loss": 0.3039,
+      "step": 12470
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6042530132419093,
+      "learning_rate": 4.5674219175472436e-07,
+      "loss": 0.308,
+      "step": 12471
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8752964715366915,
+      "learning_rate": 4.562721993062774e-07,
+      "loss": 0.3048,
+      "step": 12472
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.430787119082277,
+      "learning_rate": 4.558024372348302e-07,
+      "loss": 0.3122,
+      "step": 12473
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.248192211706744,
+      "learning_rate": 4.5533290556419805e-07,
+      "loss": 0.3053,
+      "step": 12474
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.108078004468135,
+      "learning_rate": 4.5486360431818865e-07,
+      "loss": 0.364,
+      "step": 12475
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.700041166398023,
+      "learning_rate": 4.5439453352059573e-07,
+      "loss": 0.2614,
+      "step": 12476
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.931956265619945,
+      "learning_rate": 4.5392569319520243e-07,
+      "loss": 0.2848,
+      "step": 12477
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.591541187275094,
+      "learning_rate": 4.5345708336578143e-07,
+      "loss": 0.3222,
+      "step": 12478
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.376605262091536,
+      "learning_rate": 4.529887040560899e-07,
+      "loss": 0.3051,
+      "step": 12479
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.343290192205773,
+      "learning_rate": 4.5252055528987647e-07,
+      "loss": 0.3201,
+      "step": 12480
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.534833886198081,
+      "learning_rate": 4.5205263709087844e-07,
+      "loss": 0.3464,
+      "step": 12481
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8798026778295442,
+      "learning_rate": 4.515849494828189e-07,
+      "loss": 0.3362,
+      "step": 12482
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.192440501637021,
+      "learning_rate": 4.511174924894113e-07,
+      "loss": 0.3371,
+      "step": 12483
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.3992188648309964,
+      "learning_rate": 4.5065026613435757e-07,
+      "loss": 0.3306,
+      "step": 12484
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0288710800081056,
+      "learning_rate": 4.501832704413445e-07,
+      "loss": 0.3179,
+      "step": 12485
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.965020592466695,
+      "learning_rate": 4.49716505434053e-07,
+      "loss": 0.2783,
+      "step": 12486
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6256881223784334,
+      "learning_rate": 4.4924997113614697e-07,
+      "loss": 0.2637,
+      "step": 12487
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.422459326017076,
+      "learning_rate": 4.4878366757128146e-07,
+      "loss": 0.3105,
+      "step": 12488
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8696950232341862,
+      "learning_rate": 4.483175947630991e-07,
+      "loss": 0.3055,
+      "step": 12489
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.161996205813564,
+      "learning_rate": 4.478517527352305e-07,
+      "loss": 0.2973,
+      "step": 12490
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.830515884947901,
+      "learning_rate": 4.473861415112957e-07,
+      "loss": 0.3274,
+      "step": 12491
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.959337575608403,
+      "learning_rate": 4.469207611149001e-07,
+      "loss": 0.3105,
+      "step": 12492
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.1627977817132433,
+      "learning_rate": 4.464556115696428e-07,
+      "loss": 0.3157,
+      "step": 12493
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.230694688043305,
+      "learning_rate": 4.459906928991048e-07,
+      "loss": 0.3063,
+      "step": 12494
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9875942064475534,
+      "learning_rate": 4.455260051268601e-07,
+      "loss": 0.2861,
+      "step": 12495
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.3913500633675344,
+      "learning_rate": 4.4506154827646917e-07,
+      "loss": 0.2928,
+      "step": 12496
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.283457229903859,
+      "learning_rate": 4.4459732237148056e-07,
+      "loss": 0.2926,
+      "step": 12497
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5595568700562508,
+      "learning_rate": 4.4413332743543205e-07,
+      "loss": 0.2787,
+      "step": 12498
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4170857527095673,
+      "learning_rate": 4.4366956349185033e-07,
+      "loss": 0.2933,
+      "step": 12499
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27848388099445104,
+      "learning_rate": 4.432060305642466e-07,
+      "loss": 0.1019,
+      "step": 12500
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.013084454045531,
+      "learning_rate": 4.4274272867612436e-07,
+      "loss": 0.2733,
+      "step": 12501
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.793936482586344,
+      "learning_rate": 4.422796578509742e-07,
+      "loss": 0.3562,
+      "step": 12502
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.998597698082575,
+      "learning_rate": 4.418168181122751e-07,
+      "loss": 0.2745,
+      "step": 12503
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.028049226472558,
+      "learning_rate": 4.4135420948349386e-07,
+      "loss": 0.3442,
+      "step": 12504
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 9.442396418308281,
+      "learning_rate": 4.4089183198808394e-07,
+      "loss": 0.3002,
+      "step": 12505
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.778869349246261,
+      "learning_rate": 4.4042968564949216e-07,
+      "loss": 0.3237,
+      "step": 12506
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 25.55124944040849,
+      "learning_rate": 4.39967770491147e-07,
+      "loss": 0.307,
+      "step": 12507
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.20049853400556,
+      "learning_rate": 4.3950608653647254e-07,
+      "loss": 0.3067,
+      "step": 12508
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5269700085617948,
+      "learning_rate": 4.390446338088733e-07,
+      "loss": 0.3223,
+      "step": 12509
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.986797978373684,
+      "learning_rate": 4.3858341233174785e-07,
+      "loss": 0.2988,
+      "step": 12510
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.418363155815065,
+      "learning_rate": 4.3812242212848077e-07,
+      "loss": 0.3074,
+      "step": 12511
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0493072365275027,
+      "learning_rate": 4.3766166322244505e-07,
+      "loss": 0.2747,
+      "step": 12512
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.304002970670604,
+      "learning_rate": 4.372011356370026e-07,
+      "loss": 0.3298,
+      "step": 12513
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.405738234809579,
+      "learning_rate": 4.3674083939550415e-07,
+      "loss": 0.2989,
+      "step": 12514
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.455134762842309,
+      "learning_rate": 4.3628077452128545e-07,
+      "loss": 0.3344,
+      "step": 12515
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2721432106116457,
+      "learning_rate": 4.3582094103767403e-07,
+      "loss": 0.2905,
+      "step": 12516
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.14291210909875,
+      "learning_rate": 4.353613389679856e-07,
+      "loss": 0.2808,
+      "step": 12517
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0397289702150387,
+      "learning_rate": 4.349019683355199e-07,
+      "loss": 0.2839,
+      "step": 12518
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.9456303321191206,
+      "learning_rate": 4.344428291635716e-07,
+      "loss": 0.3484,
+      "step": 12519
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.31276956449195487,
+      "learning_rate": 4.3398392147541657e-07,
+      "loss": 0.1183,
+      "step": 12520
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.772127247059982,
+      "learning_rate": 4.335252452943256e-07,
+      "loss": 0.3064,
+      "step": 12521
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28976502070231336,
+      "learning_rate": 4.330668006435529e-07,
+      "loss": 0.1017,
+      "step": 12522
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.27264281506126636,
+      "learning_rate": 4.3260858754634315e-07,
+      "loss": 0.107,
+      "step": 12523
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.452735057592599,
+      "learning_rate": 4.321506060259284e-07,
+      "loss": 0.2843,
+      "step": 12524
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4929362569317375,
+      "learning_rate": 4.3169285610552893e-07,
+      "loss": 0.3276,
+      "step": 12525
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.704627425220844,
+      "learning_rate": 4.31235337808355e-07,
+      "loss": 0.2778,
+      "step": 12526
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.7636651037321207,
+      "learning_rate": 4.3077805115760374e-07,
+      "loss": 0.3314,
+      "step": 12527
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.540014168257258,
+      "learning_rate": 4.3032099617645874e-07,
+      "loss": 0.2959,
+      "step": 12528
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.191140622110423,
+      "learning_rate": 4.298641728880948e-07,
+      "loss": 0.3293,
+      "step": 12529
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.097018075087714,
+      "learning_rate": 4.294075813156745e-07,
+      "loss": 0.3224,
+      "step": 12530
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.278195513938046,
+      "learning_rate": 4.289512214823466e-07,
+      "loss": 0.3343,
+      "step": 12531
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.693527734189172,
+      "learning_rate": 4.2849509341125196e-07,
+      "loss": 0.274,
+      "step": 12532
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.456398643916114,
+      "learning_rate": 4.280391971255138e-07,
+      "loss": 0.3139,
+      "step": 12533
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.123631750972317,
+      "learning_rate": 4.2758353264825027e-07,
+      "loss": 0.3059,
+      "step": 12534
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.837328923127375,
+      "learning_rate": 4.271281000025618e-07,
+      "loss": 0.3019,
+      "step": 12535
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.0350589729918003,
+      "learning_rate": 4.266728992115432e-07,
+      "loss": 0.2953,
+      "step": 12536
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.554055037188873,
+      "learning_rate": 4.262179302982711e-07,
+      "loss": 0.2944,
+      "step": 12537
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.796113236490628,
+      "learning_rate": 4.257631932858147e-07,
+      "loss": 0.2915,
+      "step": 12538
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.066327411158921,
+      "learning_rate": 4.2530868819723013e-07,
+      "loss": 0.3013,
+      "step": 12539
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.29214964726134235,
+      "learning_rate": 4.248544150555617e-07,
+      "loss": 0.1123,
+      "step": 12540
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.773957626639652,
+      "learning_rate": 4.2440037388384314e-07,
+      "loss": 0.3082,
+      "step": 12541
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5951655897009336,
+      "learning_rate": 4.239465647050928e-07,
+      "loss": 0.2882,
+      "step": 12542
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.314280692206255,
+      "learning_rate": 4.234929875423216e-07,
+      "loss": 0.3006,
+      "step": 12543
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.336526569952717,
+      "learning_rate": 4.230396424185268e-07,
+      "loss": 0.3095,
+      "step": 12544
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.104468938334688,
+      "learning_rate": 4.225865293566933e-07,
+      "loss": 0.315,
+      "step": 12545
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.748809333077421,
+      "learning_rate": 4.221336483797955e-07,
+      "loss": 0.334,
+      "step": 12546
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.830895712363851,
+      "learning_rate": 4.216809995107962e-07,
+      "loss": 0.3069,
+      "step": 12547
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6025456084260705,
+      "learning_rate": 4.2122858277264314e-07,
+      "loss": 0.3246,
+      "step": 12548
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.265825457154651,
+      "learning_rate": 4.2077639818827843e-07,
+      "loss": 0.2985,
+      "step": 12549
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.306249100935355,
+      "learning_rate": 4.20324445780626e-07,
+      "loss": 0.3127,
+      "step": 12550
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.710826983544654,
+      "learning_rate": 4.198727255726015e-07,
+      "loss": 0.2821,
+      "step": 12551
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.323619742451093,
+      "learning_rate": 4.1942123758710873e-07,
+      "loss": 0.3377,
+      "step": 12552
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.28760186556532014,
+      "learning_rate": 4.1896998184703886e-07,
+      "loss": 0.1111,
+      "step": 12553
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.362616016865052,
+      "learning_rate": 4.185189583752719e-07,
+      "loss": 0.3219,
+      "step": 12554
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.111098808069561,
+      "learning_rate": 4.180681671946746e-07,
+      "loss": 0.2888,
+      "step": 12555
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 6.8351336855192235,
+      "learning_rate": 4.1761760832810415e-07,
+      "loss": 0.2898,
+      "step": 12556
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8886447524252294,
+      "learning_rate": 4.1716728179840405e-07,
+      "loss": 0.3306,
+      "step": 12557
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.9178962428585358,
+      "learning_rate": 4.167171876284076e-07,
+      "loss": 0.3242,
+      "step": 12558
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.903663617539226,
+      "learning_rate": 4.162673258409355e-07,
+      "loss": 0.2937,
+      "step": 12559
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.369933422932025,
+      "learning_rate": 4.1581769645879675e-07,
+      "loss": 0.306,
+      "step": 12560
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 22.396191281629065,
+      "learning_rate": 4.1536829950478697e-07,
+      "loss": 0.3069,
+      "step": 12561
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.6097568628299155,
+      "learning_rate": 4.1491913500169454e-07,
+      "loss": 0.3123,
+      "step": 12562
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8682220667282388,
+      "learning_rate": 4.1447020297229077e-07,
+      "loss": 0.3122,
+      "step": 12563
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.503876947257102,
+      "learning_rate": 4.14021503439338e-07,
+      "loss": 0.3106,
+      "step": 12564
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.980106422690282,
+      "learning_rate": 4.135730364255869e-07,
+      "loss": 0.3288,
+      "step": 12565
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.030481533020783,
+      "learning_rate": 4.131248019537748e-07,
+      "loss": 0.3188,
+      "step": 12566
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 7.3112402449999845,
+      "learning_rate": 4.126768000466297e-07,
+      "loss": 0.3105,
+      "step": 12567
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 11.021188933172606,
+      "learning_rate": 4.122290307268634e-07,
+      "loss": 0.3385,
+      "step": 12568
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.615713485590502,
+      "learning_rate": 4.117814940171827e-07,
+      "loss": 0.2841,
+      "step": 12569
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.341341264036053,
+      "learning_rate": 4.113341899402756e-07,
+      "loss": 0.3209,
+      "step": 12570
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.006152841890986,
+      "learning_rate": 4.1088711851882237e-07,
+      "loss": 0.2833,
+      "step": 12571
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.181523309840292,
+      "learning_rate": 4.104402797754908e-07,
+      "loss": 0.3467,
+      "step": 12572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.985533576020627,
+      "learning_rate": 4.099936737329363e-07,
+      "loss": 0.3103,
+      "step": 12573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.176939008297152,
+      "learning_rate": 4.0954730041380287e-07,
+      "loss": 0.2941,
+      "step": 12574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.5890479112095974,
+      "learning_rate": 4.091011598407235e-07,
+      "loss": 0.2938,
+      "step": 12575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.670535691789138,
+      "learning_rate": 4.0865525203631626e-07,
+      "loss": 0.2822,
+      "step": 12576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.951279307397814,
+      "learning_rate": 4.0820957702319195e-07,
+      "loss": 0.3229,
+      "step": 12577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.085629131140648,
+      "learning_rate": 4.0776413482394585e-07,
+      "loss": 0.3638,
+      "step": 12578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.782296280898751,
+      "learning_rate": 4.0731892546116316e-07,
+      "loss": 0.3311,
+      "step": 12579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.253459846895884,
+      "learning_rate": 4.0687394895741803e-07,
+      "loss": 0.3074,
+      "step": 12580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.2101876430077136,
+      "learning_rate": 4.0642920533526977e-07,
+      "loss": 0.299,
+      "step": 12581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 12.46433727232035,
+      "learning_rate": 4.059846946172702e-07,
+      "loss": 0.333,
+      "step": 12582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.886430014391516,
+      "learning_rate": 4.0554041682595415e-07,
+      "loss": 0.2971,
+      "step": 12583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.149106560189883,
+      "learning_rate": 4.0509637198385075e-07,
+      "loss": 0.2697,
+      "step": 12584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.44433351516485,
+      "learning_rate": 4.0465256011347153e-07,
+      "loss": 0.3081,
+      "step": 12585
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.104242939220991,
+      "learning_rate": 4.042089812373201e-07,
+      "loss": 0.2999,
+      "step": 12586
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0582315023811866,
+      "learning_rate": 4.037656353778857e-07,
+      "loss": 0.3127,
+      "step": 12587
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6383818691360466,
+      "learning_rate": 4.033225225576487e-07,
+      "loss": 0.3252,
+      "step": 12588
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.618947948251666,
+      "learning_rate": 4.028796427990739e-07,
+      "loss": 0.3281,
+      "step": 12589
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3077025119229355,
+      "learning_rate": 4.024369961246177e-07,
+      "loss": 0.3148,
+      "step": 12590
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.282169763230969,
+      "learning_rate": 4.019945825567223e-07,
+      "loss": 0.2919,
+      "step": 12591
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 171.03636007180384,
+      "learning_rate": 4.0155240211781966e-07,
+      "loss": 0.3305,
+      "step": 12592
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.1380951465899045,
+      "learning_rate": 4.0111045483032965e-07,
+      "loss": 0.3052,
+      "step": 12593
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.2674157349729205,
+      "learning_rate": 4.006687407166582e-07,
+      "loss": 0.3173,
+      "step": 12594
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8283507085700585,
+      "learning_rate": 4.002272597992035e-07,
+      "loss": 0.3009,
+      "step": 12595
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.797646773424156,
+      "learning_rate": 3.997860121003477e-07,
+      "loss": 0.324,
+      "step": 12596
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.156863872697717,
+      "learning_rate": 3.9934499764246517e-07,
+      "loss": 0.3105,
+      "step": 12597
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4648647136924096,
+      "learning_rate": 3.98904216447914e-07,
+      "loss": 0.3226,
+      "step": 12598
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.194171794905401,
+      "learning_rate": 3.9846366853904414e-07,
+      "loss": 0.3527,
+      "step": 12599
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 10.419307673817872,
+      "learning_rate": 3.980233539381917e-07,
+      "loss": 0.278,
+      "step": 12600
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9090281392594,
+      "learning_rate": 3.9758327266768203e-07,
+      "loss": 0.3283,
+      "step": 12601
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26950257909667563,
+      "learning_rate": 3.971434247498279e-07,
+      "loss": 0.1018,
+      "step": 12602
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8904381310034544,
+      "learning_rate": 3.9670381020693206e-07,
+      "loss": 0.3224,
+      "step": 12603
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.844583741262518,
+      "learning_rate": 3.962644290612816e-07,
+      "loss": 0.3098,
+      "step": 12604
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9547330408538612,
+      "learning_rate": 3.9582528133515484e-07,
+      "loss": 0.3199,
+      "step": 12605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.6417363391680455,
+      "learning_rate": 3.953863670508179e-07,
+      "loss": 0.3123,
+      "step": 12606
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3151742500619474,
+      "learning_rate": 3.949476862305246e-07,
+      "loss": 0.2855,
+      "step": 12607
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.973555528939438,
+      "learning_rate": 3.9450923889651825e-07,
+      "loss": 0.2648,
+      "step": 12608
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.914653056516696,
+      "learning_rate": 3.9407102507102556e-07,
+      "loss": 0.3354,
+      "step": 12609
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.736007309468463,
+      "learning_rate": 3.9363304477626985e-07,
+      "loss": 0.2769,
+      "step": 12610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.33824021775874,
+      "learning_rate": 3.931952980344528e-07,
+      "loss": 0.2914,
+      "step": 12611
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.059848210268941,
+      "learning_rate": 3.927577848677733e-07,
+      "loss": 0.3345,
+      "step": 12612
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.536036027111298,
+      "learning_rate": 3.923205052984114e-07,
+      "loss": 0.3296,
+      "step": 12613
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 9.78993556323269,
+      "learning_rate": 3.9188345934853933e-07,
+      "loss": 0.3345,
+      "step": 12614
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.628346332384427,
+      "learning_rate": 3.9144664704031553e-07,
+      "loss": 0.3161,
+      "step": 12615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.743202074719135,
+      "learning_rate": 3.910100683958884e-07,
+      "loss": 0.3388,
+      "step": 12616
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.403779603580971,
+      "learning_rate": 3.905737234373935e-07,
+      "loss": 0.2845,
+      "step": 12617
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.100230301507666,
+      "learning_rate": 3.9013761218695324e-07,
+      "loss": 0.2881,
+      "step": 12618
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9618013184838614,
+      "learning_rate": 3.897017346666798e-07,
+      "loss": 0.3333,
+      "step": 12619
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.935179972525382,
+      "learning_rate": 3.8926609089867395e-07,
+      "loss": 0.3111,
+      "step": 12620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.076492755240671,
+      "learning_rate": 3.888306809050224e-07,
+      "loss": 0.3343,
+      "step": 12621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.224990777704877,
+      "learning_rate": 3.883955047078031e-07,
+      "loss": 0.3179,
+      "step": 12622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9380610400907394,
+      "learning_rate": 3.8796056232908e-07,
+      "loss": 0.3093,
+      "step": 12623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.557298937345109,
+      "learning_rate": 3.875258537909032e-07,
+      "loss": 0.306,
+      "step": 12624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.456669117663048,
+      "learning_rate": 3.870913791153175e-07,
+      "loss": 0.3216,
+      "step": 12625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.034012174911063,
+      "learning_rate": 3.866571383243484e-07,
+      "loss": 0.3256,
+      "step": 12626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.799340238525715,
+      "learning_rate": 3.8622313144001445e-07,
+      "loss": 0.3057,
+      "step": 12627
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.261108183295192,
+      "learning_rate": 3.857893584843203e-07,
+      "loss": 0.3475,
+      "step": 12628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8627007291975075,
+      "learning_rate": 3.8535581947925895e-07,
+      "loss": 0.2936,
+      "step": 12629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.1596527523749005,
+      "learning_rate": 3.8492251444681326e-07,
+      "loss": 0.3166,
+      "step": 12630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2730789090323187,
+      "learning_rate": 3.844894434089508e-07,
+      "loss": 0.3089,
+      "step": 12631
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.5410806796094723,
+      "learning_rate": 3.8405660638763e-07,
+      "loss": 0.3102,
+      "step": 12632
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7971332325926817,
+      "learning_rate": 3.836240034047967e-07,
+      "loss": 0.3009,
+      "step": 12633
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.259468030167489,
+      "learning_rate": 3.831916344823844e-07,
+      "loss": 0.2916,
+      "step": 12634
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.4245364718421625,
+      "learning_rate": 3.827594996423162e-07,
+      "loss": 0.2924,
+      "step": 12635
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.611399938810528,
+      "learning_rate": 3.8232759890650173e-07,
+      "loss": 0.2904,
+      "step": 12636
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3867253199030225,
+      "learning_rate": 3.8189593229683794e-07,
+      "loss": 0.296,
+      "step": 12637
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.272752185585896,
+      "learning_rate": 3.814644998352141e-07,
+      "loss": 0.283,
+      "step": 12638
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9668511818796084,
+      "learning_rate": 3.810333015435025e-07,
+      "loss": 0.3127,
+      "step": 12639
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.942110026098278,
+      "learning_rate": 3.8060233744356634e-07,
+      "loss": 0.3209,
+      "step": 12640
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.26961584914802006,
+      "learning_rate": 3.80171607557257e-07,
+      "loss": 0.0994,
+      "step": 12641
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.6965391040810003,
+      "learning_rate": 3.797411119064132e-07,
+      "loss": 0.2927,
+      "step": 12642
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.712743877971254,
+      "learning_rate": 3.7931085051286234e-07,
+      "loss": 0.2988,
+      "step": 12643
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.101394269656224,
+      "learning_rate": 3.788808233984176e-07,
+      "loss": 0.3221,
+      "step": 12644
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9702606579268367,
+      "learning_rate": 3.7845103058488597e-07,
+      "loss": 0.3029,
+      "step": 12645
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.43260399919221,
+      "learning_rate": 3.780214720940556e-07,
+      "loss": 0.3295,
+      "step": 12646
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.270724128905567,
+      "learning_rate": 3.7759214794770735e-07,
+      "loss": 0.3206,
+      "step": 12647
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7015104900003366,
+      "learning_rate": 3.771630581676089e-07,
+      "loss": 0.27,
+      "step": 12648
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2948713218606298,
+      "learning_rate": 3.767342027755161e-07,
+      "loss": 0.1056,
+      "step": 12649
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.2903987594095189,
+      "learning_rate": 3.763055817931727e-07,
+      "loss": 0.11,
+      "step": 12650
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4184113372253537,
+      "learning_rate": 3.7587719524231124e-07,
+      "loss": 0.2937,
+      "step": 12651
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2714600211370066,
+      "learning_rate": 3.754490431446506e-07,
+      "loss": 0.2979,
+      "step": 12652
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3981368554078157,
+      "learning_rate": 3.750211255219005e-07,
+      "loss": 0.3023,
+      "step": 12653
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2580921698742693,
+      "learning_rate": 3.745934423957559e-07,
+      "loss": 0.2841,
+      "step": 12654
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.050095083977508,
+      "learning_rate": 3.741659937879022e-07,
+      "loss": 0.2999,
+      "step": 12655
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.30989207264759805,
+      "learning_rate": 3.737387797200126e-07,
+      "loss": 0.112,
+      "step": 12656
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.449554093046274,
+      "learning_rate": 3.7331180021374534e-07,
+      "loss": 0.3035,
+      "step": 12657
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 8.659164235279968,
+      "learning_rate": 3.728850552907526e-07,
+      "loss": 0.3088,
+      "step": 12658
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.918523315102038,
+      "learning_rate": 3.724585449726692e-07,
+      "loss": 0.3282,
+      "step": 12659
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.6976364807426396,
+      "learning_rate": 3.720322692811196e-07,
+      "loss": 0.3043,
+      "step": 12660
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3882318227983306,
+      "learning_rate": 3.7160622823771864e-07,
+      "loss": 0.2643,
+      "step": 12661
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7012282359129944,
+      "learning_rate": 3.711804218640658e-07,
+      "loss": 0.3208,
+      "step": 12662
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1787445820999123,
+      "learning_rate": 3.7075485018175207e-07,
+      "loss": 0.3052,
+      "step": 12663
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8932359828225667,
+      "learning_rate": 3.7032951321235464e-07,
+      "loss": 0.31,
+      "step": 12664
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 12.74768338780166,
+      "learning_rate": 3.6990441097743743e-07,
+      "loss": 0.3101,
+      "step": 12665
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.539379872778637,
+      "learning_rate": 3.694795434985554e-07,
+      "loss": 0.3099,
+      "step": 12666
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.990027978375501,
+      "learning_rate": 3.690549107972502e-07,
+      "loss": 0.3013,
+      "step": 12667
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8017364054522202,
+      "learning_rate": 3.6863051289505124e-07,
+      "loss": 0.2791,
+      "step": 12668
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.4941975326918038,
+      "learning_rate": 3.6820634981347745e-07,
+      "loss": 0.3435,
+      "step": 12669
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.0429472038868495,
+      "learning_rate": 3.677824215740322e-07,
+      "loss": 0.3037,
+      "step": 12670
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.938292182829095,
+      "learning_rate": 3.673587281982133e-07,
+      "loss": 0.277,
+      "step": 12671
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.156679129342306,
+      "learning_rate": 3.669352697074996e-07,
+      "loss": 0.3088,
+      "step": 12672
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.632227088163149,
+      "learning_rate": 3.6651204612336397e-07,
+      "loss": 0.3293,
+      "step": 12673
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.947657449982887,
+      "learning_rate": 3.660890574672632e-07,
+      "loss": 0.2798,
+      "step": 12674
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.836821185365947,
+      "learning_rate": 3.6566630376064395e-07,
+      "loss": 0.3181,
+      "step": 12675
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.530059635795168,
+      "learning_rate": 3.6524378502494075e-07,
+      "loss": 0.3071,
+      "step": 12676
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.646600723752187,
+      "learning_rate": 3.6482150128157646e-07,
+      "loss": 0.3284,
+      "step": 12677
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.961819635067437,
+      "learning_rate": 3.6439945255196286e-07,
+      "loss": 0.3129,
+      "step": 12678
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0983840522821895,
+      "learning_rate": 3.6397763885749726e-07,
+      "loss": 0.3274,
+      "step": 12679
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8946523387138043,
+      "learning_rate": 3.6355606021956646e-07,
+      "loss": 0.3246,
+      "step": 12680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9415774537073243,
+      "learning_rate": 3.631347166595456e-07,
+      "loss": 0.2684,
+      "step": 12681
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.521859708692814,
+      "learning_rate": 3.6271360819879864e-07,
+      "loss": 0.2915,
+      "step": 12682
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.541362043413118,
+      "learning_rate": 3.622927348586758e-07,
+      "loss": 0.2828,
+      "step": 12683
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.858921872809615,
+      "learning_rate": 3.618720966605177e-07,
+      "loss": 0.2825,
+      "step": 12684
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 7.2433537391923215,
+      "learning_rate": 3.6145169362564844e-07,
+      "loss": 0.3325,
+      "step": 12685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.45410179177747,
+      "learning_rate": 3.610315257753877e-07,
+      "loss": 0.3356,
+      "step": 12686
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.927168818964064,
+      "learning_rate": 3.6061159313103447e-07,
+      "loss": 0.3166,
+      "step": 12687
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1403171736081332,
+      "learning_rate": 3.6019189571388444e-07,
+      "loss": 0.2903,
+      "step": 12688
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.595974955962459,
+      "learning_rate": 3.597724335452141e-07,
+      "loss": 0.2985,
+      "step": 12689
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.287284771375765,
+      "learning_rate": 3.593532066462929e-07,
+      "loss": 0.1126,
+      "step": 12690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.405016992904074,
+      "learning_rate": 3.5893421503837513e-07,
+      "loss": 0.3043,
+      "step": 12691
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.725435833972566,
+      "learning_rate": 3.5851545874270586e-07,
+      "loss": 0.3102,
+      "step": 12692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.07266040223268,
+      "learning_rate": 3.580969377805171e-07,
+      "loss": 0.3246,
+      "step": 12693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.333212509593804,
+      "learning_rate": 3.576786521730269e-07,
+      "loss": 0.3204,
+      "step": 12694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28282325698946026,
+      "learning_rate": 3.5726060194144485e-07,
+      "loss": 0.1068,
+      "step": 12695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.764120034780988,
+      "learning_rate": 3.568427871069668e-07,
+      "loss": 0.3424,
+      "step": 12696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7779727019725167,
+      "learning_rate": 3.564252076907776e-07,
+      "loss": 0.3304,
+      "step": 12697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2817903673387687,
+      "learning_rate": 3.560078637140463e-07,
+      "loss": 0.3139,
+      "step": 12698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.156716367321163,
+      "learning_rate": 3.5559075519793763e-07,
+      "loss": 0.2943,
+      "step": 12699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.873126586152884,
+      "learning_rate": 3.551738821635958e-07,
+      "loss": 0.2975,
+      "step": 12700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.397610620001267,
+      "learning_rate": 3.547572446321612e-07,
+      "loss": 0.2954,
+      "step": 12701
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.535068523186751,
+      "learning_rate": 3.543408426247552e-07,
+      "loss": 0.3128,
+      "step": 12702
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 13.528949995473445,
+      "learning_rate": 3.5392467616249094e-07,
+      "loss": 0.3363,
+      "step": 12703
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.839669805243032,
+      "learning_rate": 3.5350874526646925e-07,
+      "loss": 0.2791,
+      "step": 12704
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.7573919400795166,
+      "learning_rate": 3.530930499577795e-07,
+      "loss": 0.3067,
+      "step": 12705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.918730711833265,
+      "learning_rate": 3.52677590257498e-07,
+      "loss": 0.3077,
+      "step": 12706
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.293808387590978,
+      "learning_rate": 3.522623661866881e-07,
+      "loss": 0.3191,
+      "step": 12707
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.693530708229626,
+      "learning_rate": 3.5184737776640386e-07,
+      "loss": 0.3106,
+      "step": 12708
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 11.225033336155258,
+      "learning_rate": 3.5143262501768585e-07,
+      "loss": 0.3454,
+      "step": 12709
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.8573293179812156,
+      "learning_rate": 3.5101810796156323e-07,
+      "loss": 0.3119,
+      "step": 12710
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.870645464391077,
+      "learning_rate": 3.50603826619052e-07,
+      "loss": 0.2931,
+      "step": 12711
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 16.61328630303698,
+      "learning_rate": 3.501897810111593e-07,
+      "loss": 0.3121,
+      "step": 12712
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.9593075687204884,
+      "learning_rate": 3.497759711588744e-07,
+      "loss": 0.3349,
+      "step": 12713
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.26192735630213,
+      "learning_rate": 3.493623970831828e-07,
+      "loss": 0.3009,
+      "step": 12714
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.013533563588173,
+      "learning_rate": 3.489490588050498e-07,
+      "loss": 0.3233,
+      "step": 12715
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9762168434068244,
+      "learning_rate": 3.4853595634543493e-07,
+      "loss": 0.2799,
+      "step": 12716
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.28470917919091526,
+      "learning_rate": 3.481230897252824e-07,
+      "loss": 0.1074,
+      "step": 12717
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.62730757207048,
+      "learning_rate": 3.4771045896552555e-07,
+      "loss": 0.3043,
+      "step": 12718
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.5000504076457695,
+      "learning_rate": 3.472980640870871e-07,
+      "loss": 0.2925,
+      "step": 12719
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.245017050637801,
+      "learning_rate": 3.4688590511087304e-07,
+      "loss": 0.3065,
+      "step": 12720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.1769907711330014,
+      "learning_rate": 3.464739820577845e-07,
+      "loss": 0.2799,
+      "step": 12721
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.871913375650927,
+      "learning_rate": 3.4606229494870414e-07,
+      "loss": 0.3329,
+      "step": 12722
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 21.389158477411815,
+      "learning_rate": 3.45650843804507e-07,
+      "loss": 0.3227,
+      "step": 12723
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.483377632176951,
+      "learning_rate": 3.4523962864605354e-07,
+      "loss": 0.3412,
+      "step": 12724
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.96685512073729,
+      "learning_rate": 3.448286494941938e-07,
+      "loss": 0.3399,
+      "step": 12725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.3778324101849604,
+      "learning_rate": 3.444179063697656e-07,
+      "loss": 0.3008,
+      "step": 12726
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.476144470577204,
+      "learning_rate": 3.440073992935944e-07,
+      "loss": 0.3195,
+      "step": 12727
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.149167917230166,
+      "learning_rate": 3.4359712828649306e-07,
+      "loss": 0.291,
+      "step": 12728
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 5.275359936888931,
+      "learning_rate": 3.4318709336926384e-07,
+      "loss": 0.3032,
+      "step": 12729
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.801408530538427,
+      "learning_rate": 3.4277729456269614e-07,
+      "loss": 0.3076,
+      "step": 12730
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6274467739724607,
+      "learning_rate": 3.423677318875679e-07,
+      "loss": 0.3183,
+      "step": 12731
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.440100355339497,
+      "learning_rate": 3.4195840536464574e-07,
+      "loss": 0.3458,
+      "step": 12732
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.03009905253099,
+      "learning_rate": 3.415493150146809e-07,
+      "loss": 0.3022,
+      "step": 12733
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.755132837050582,
+      "learning_rate": 3.4114046085841847e-07,
+      "loss": 0.3198,
+      "step": 12734
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.829455616465171,
+      "learning_rate": 3.407318429165857e-07,
+      "loss": 0.339,
+      "step": 12735
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.535966791332874,
+      "learning_rate": 3.40323461209901e-07,
+      "loss": 0.3676,
+      "step": 12736
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8496333993471983,
+      "learning_rate": 3.399153157590701e-07,
+      "loss": 0.3033,
+      "step": 12737
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8824608168656956,
+      "learning_rate": 3.3950740658478754e-07,
+      "loss": 0.2902,
+      "step": 12738
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.592983927978706,
+      "learning_rate": 3.390997337077351e-07,
+      "loss": 0.3234,
+      "step": 12739
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.927570612082081,
+      "learning_rate": 3.3869229714858344e-07,
+      "loss": 0.2994,
+      "step": 12740
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.032437714570997,
+      "learning_rate": 3.3828509692798827e-07,
+      "loss": 0.3078,
+      "step": 12741
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.887443936788069,
+      "learning_rate": 3.37878133066597e-07,
+      "loss": 0.304,
+      "step": 12742
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.150174264485211,
+      "learning_rate": 3.374714055850437e-07,
+      "loss": 0.3074,
+      "step": 12743
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.56461047580076,
+      "learning_rate": 3.370649145039495e-07,
+      "loss": 0.2873,
+      "step": 12744
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4669411451417935,
+      "learning_rate": 3.366586598439264e-07,
+      "loss": 0.3266,
+      "step": 12745
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.338245129681688,
+      "learning_rate": 3.3625264162556904e-07,
+      "loss": 0.2803,
+      "step": 12746
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3116255327598817,
+      "learning_rate": 3.358468598694664e-07,
+      "loss": 0.1056,
+      "step": 12747
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.717939499949455,
+      "learning_rate": 3.3544131459619043e-07,
+      "loss": 0.2661,
+      "step": 12748
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.218193803413958,
+      "learning_rate": 3.350360058263058e-07,
+      "loss": 0.3038,
+      "step": 12749
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 69.09400954281284,
+      "learning_rate": 3.346309335803605e-07,
+      "loss": 0.3371,
+      "step": 12750
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.024740624894926,
+      "learning_rate": 3.342260978788925e-07,
+      "loss": 0.3048,
+      "step": 12751
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7526590716048345,
+      "learning_rate": 3.338214987424282e-07,
+      "loss": 0.2966,
+      "step": 12752
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 33.09960035736967,
+      "learning_rate": 3.334171361914823e-07,
+      "loss": 0.3084,
+      "step": 12753
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.940739674042875,
+      "learning_rate": 3.330130102465573e-07,
+      "loss": 0.2824,
+      "step": 12754
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.7516155888696465,
+      "learning_rate": 3.326091209281412e-07,
+      "loss": 0.2799,
+      "step": 12755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.316892461329892,
+      "learning_rate": 3.322054682567133e-07,
+      "loss": 0.3458,
+      "step": 12756
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2860741857488134,
+      "learning_rate": 3.318020522527399e-07,
+      "loss": 0.3034,
+      "step": 12757
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.30498815463181334,
+      "learning_rate": 3.313988729366752e-07,
+      "loss": 0.114,
+      "step": 12758
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.051450224648325,
+      "learning_rate": 3.309959303289606e-07,
+      "loss": 0.313,
+      "step": 12759
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.637635786272789,
+      "learning_rate": 3.3059322445002706e-07,
+      "loss": 0.3172,
+      "step": 12760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.716460909539705,
+      "learning_rate": 3.3019075532029043e-07,
+      "loss": 0.3112,
+      "step": 12761
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.540996033968547,
+      "learning_rate": 3.297885229601605e-07,
+      "loss": 0.3261,
+      "step": 12762
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.441693207796321,
+      "learning_rate": 3.293865273900276e-07,
+      "loss": 0.3128,
+      "step": 12763
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.752329513028563,
+      "learning_rate": 3.2898476863027664e-07,
+      "loss": 0.3245,
+      "step": 12764
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.707754917850271,
+      "learning_rate": 3.2858324670127617e-07,
+      "loss": 0.313,
+      "step": 12765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5517745356902712,
+      "learning_rate": 3.2818196162338447e-07,
+      "loss": 0.3096,
+      "step": 12766
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.058432051868207,
+      "learning_rate": 3.277809134169485e-07,
+      "loss": 0.318,
+      "step": 12767
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.10574965497393,
+      "learning_rate": 3.273801021023004e-07,
+      "loss": 0.3102,
+      "step": 12768
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.515038126867511,
+      "learning_rate": 3.2697952769976336e-07,
+      "loss": 0.2837,
+      "step": 12769
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.56919440173947,
+      "learning_rate": 3.2657919022964723e-07,
+      "loss": 0.3182,
+      "step": 12770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.350982343693829,
+      "learning_rate": 3.261790897122502e-07,
+      "loss": 0.292,
+      "step": 12771
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.537797314667408,
+      "learning_rate": 3.2577922616785826e-07,
+      "loss": 0.3238,
+      "step": 12772
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 18.31183029102732,
+      "learning_rate": 3.253795996167458e-07,
+      "loss": 0.2916,
+      "step": 12773
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.987964063992267,
+      "learning_rate": 3.249802100791721e-07,
+      "loss": 0.3347,
+      "step": 12774
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.6176971385381296,
+      "learning_rate": 3.245810575753916e-07,
+      "loss": 0.3066,
+      "step": 12775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.3562316481963315,
+      "learning_rate": 3.2418214212563746e-07,
+      "loss": 0.2793,
+      "step": 12776
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.166989394640893,
+      "learning_rate": 3.237834637501397e-07,
+      "loss": 0.3088,
+      "step": 12777
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.2341198542723175,
+      "learning_rate": 3.2338502246910986e-07,
+      "loss": 0.3127,
+      "step": 12778
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28817227692013264,
+      "learning_rate": 3.229868183027501e-07,
+      "loss": 0.1115,
+      "step": 12779
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.282367447347158,
+      "learning_rate": 3.225888512712505e-07,
+      "loss": 0.2721,
+      "step": 12780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.848692696219458,
+      "learning_rate": 3.221911213947887e-07,
+      "loss": 0.3325,
+      "step": 12781
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.616932275459207,
+      "learning_rate": 3.2179362869353136e-07,
+      "loss": 0.3073,
+      "step": 12782
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.42740490625443,
+      "learning_rate": 3.2139637318763075e-07,
+      "loss": 0.29,
+      "step": 12783
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5357862447131767,
+      "learning_rate": 3.20999354897229e-07,
+      "loss": 0.3309,
+      "step": 12784
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.332992897161478,
+      "learning_rate": 3.206025738424562e-07,
+      "loss": 0.3276,
+      "step": 12785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2756007055301676,
+      "learning_rate": 3.202060300434301e-07,
+      "loss": 0.1009,
+      "step": 12786
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.326279484694752,
+      "learning_rate": 3.198097235202563e-07,
+      "loss": 0.318,
+      "step": 12787
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.678898845242527,
+      "learning_rate": 3.194136542930293e-07,
+      "loss": 0.3097,
+      "step": 12788
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.001441699859349,
+      "learning_rate": 3.190178223818274e-07,
+      "loss": 0.3341,
+      "step": 12789
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.5103924341816715,
+      "learning_rate": 3.1862222780672467e-07,
+      "loss": 0.303,
+      "step": 12790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.536294144359234,
+      "learning_rate": 3.182268705877756e-07,
+      "loss": 0.3199,
+      "step": 12791
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2079406898199476,
+      "learning_rate": 3.178317507450257e-07,
+      "loss": 0.3239,
+      "step": 12792
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0262606664228344,
+      "learning_rate": 3.1743686829851017e-07,
+      "loss": 0.3016,
+      "step": 12793
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.364479730855363,
+      "learning_rate": 3.170422232682485e-07,
+      "loss": 0.3238,
+      "step": 12794
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.397505704332269,
+      "learning_rate": 3.166478156742525e-07,
+      "loss": 0.3002,
+      "step": 12795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4213267141753194,
+      "learning_rate": 3.1625364553651615e-07,
+      "loss": 0.2859,
+      "step": 12796
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6398446467076626,
+      "learning_rate": 3.158597128750279e-07,
+      "loss": 0.2964,
+      "step": 12797
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.869664333959736,
+      "learning_rate": 3.1546601770975904e-07,
+      "loss": 0.3124,
+      "step": 12798
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.928404153992974,
+      "learning_rate": 3.150725600606719e-07,
+      "loss": 0.3387,
+      "step": 12799
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 14.717210955490293,
+      "learning_rate": 3.146793399477144e-07,
+      "loss": 0.318,
+      "step": 12800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3253473308348706,
+      "learning_rate": 3.1428635739082504e-07,
+      "loss": 0.1246,
+      "step": 12801
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.818152112728011,
+      "learning_rate": 3.138936124099279e-07,
+      "loss": 0.2716,
+      "step": 12802
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.573831246586453,
+      "learning_rate": 3.135011050249376e-07,
+      "loss": 0.3292,
+      "step": 12803
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.989160299579741,
+      "learning_rate": 3.1310883525575264e-07,
+      "loss": 0.2997,
+      "step": 12804
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.328998398914418,
+      "learning_rate": 3.127168031222633e-07,
+      "loss": 0.3311,
+      "step": 12805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 13.443418829324061,
+      "learning_rate": 3.123250086443469e-07,
+      "loss": 0.3383,
+      "step": 12806
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7699959137731196,
+      "learning_rate": 3.119334518418676e-07,
+      "loss": 0.3273,
+      "step": 12807
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4950053569050357,
+      "learning_rate": 3.1154213273467906e-07,
+      "loss": 0.273,
+      "step": 12808
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.144808266183467,
+      "learning_rate": 3.1115105134261967e-07,
+      "loss": 0.3069,
+      "step": 12809
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.508193319110271,
+      "learning_rate": 3.1076020768552205e-07,
+      "loss": 0.3087,
+      "step": 12810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4876103457442538,
+      "learning_rate": 3.103696017831992e-07,
+      "loss": 0.2914,
+      "step": 12811
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 8.133764350073148,
+      "learning_rate": 3.0997923365545744e-07,
+      "loss": 0.3368,
+      "step": 12812
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.481821917475852,
+      "learning_rate": 3.095891033220888e-07,
+      "loss": 0.2967,
+      "step": 12813
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2839168982187531,
+      "learning_rate": 3.091992108028735e-07,
+      "loss": 0.1099,
+      "step": 12814
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8031045903535867,
+      "learning_rate": 3.088095561175808e-07,
+      "loss": 0.2964,
+      "step": 12815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2745296913924774,
+      "learning_rate": 3.0842013928596757e-07,
+      "loss": 0.1035,
+      "step": 12816
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.901337091463197,
+      "learning_rate": 3.080309603277759e-07,
+      "loss": 0.3138,
+      "step": 12817
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.633003610880978,
+      "learning_rate": 3.076420192627394e-07,
+      "loss": 0.2927,
+      "step": 12818
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4845620748019166,
+      "learning_rate": 3.0725331611057784e-07,
+      "loss": 0.305,
+      "step": 12819
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.2857684378263309,
+      "learning_rate": 3.0686485089099994e-07,
+      "loss": 0.0969,
+      "step": 12820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.272688621792583,
+      "learning_rate": 3.0647662362370213e-07,
+      "loss": 0.3109,
+      "step": 12821
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.126199219507577,
+      "learning_rate": 3.060886343283659e-07,
+      "loss": 0.2817,
+      "step": 12822
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.387749998699523,
+      "learning_rate": 3.057008830246666e-07,
+      "loss": 0.3114,
+      "step": 12823
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.7464162376366543,
+      "learning_rate": 3.053133697322608e-07,
+      "loss": 0.305,
+      "step": 12824
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.0747312374404565,
+      "learning_rate": 3.0492609447079937e-07,
+      "loss": 0.3343,
+      "step": 12825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5937974761294553,
+      "learning_rate": 3.045390572599155e-07,
+      "loss": 0.2777,
+      "step": 12826
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.020471035384902,
+      "learning_rate": 3.041522581192341e-07,
+      "loss": 0.2644,
+      "step": 12827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4207134429043986,
+      "learning_rate": 3.0376569706836613e-07,
+      "loss": 0.3118,
+      "step": 12828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 10.739432618472518,
+      "learning_rate": 3.0337937412691144e-07,
+      "loss": 0.3056,
+      "step": 12829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8882632535331614,
+      "learning_rate": 3.0299328931445824e-07,
+      "loss": 0.3081,
+      "step": 12830
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.730004423723425,
+      "learning_rate": 3.026074426505804e-07,
+      "loss": 0.3252,
+      "step": 12831
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.0194145987255023,
+      "learning_rate": 3.022218341548422e-07,
+      "loss": 0.3111,
+      "step": 12832
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.057985306783537,
+      "learning_rate": 3.018364638467941e-07,
+      "loss": 0.3061,
+      "step": 12833
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.036697156560057,
+      "learning_rate": 3.014513317459755e-07,
+      "loss": 0.3078,
+      "step": 12834
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6538376766096015,
+      "learning_rate": 3.0106643787191356e-07,
+      "loss": 0.3083,
+      "step": 12835
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.326602833618025,
+      "learning_rate": 3.0068178224412434e-07,
+      "loss": 0.3345,
+      "step": 12836
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.586741232133588,
+      "learning_rate": 3.002973648821078e-07,
+      "loss": 0.315,
+      "step": 12837
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.675521405506844,
+      "learning_rate": 2.9991318580535835e-07,
+      "loss": 0.2982,
+      "step": 12838
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.29419348358982467,
+      "learning_rate": 2.995292450333509e-07,
+      "loss": 0.1167,
+      "step": 12839
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.602517351726532,
+      "learning_rate": 2.9914554258555605e-07,
+      "loss": 0.2807,
+      "step": 12840
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.90605907139544,
+      "learning_rate": 2.9876207848142603e-07,
+      "loss": 0.3485,
+      "step": 12841
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3378560209289097,
+      "learning_rate": 2.98378852740403e-07,
+      "loss": 0.2819,
+      "step": 12842
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.153742540569421,
+      "learning_rate": 2.9799586538191925e-07,
+      "loss": 0.3042,
+      "step": 12843
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.515927438745548,
+      "learning_rate": 2.976131164253915e-07,
+      "loss": 0.2803,
+      "step": 12844
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.285058708660904,
+      "learning_rate": 2.972306058902258e-07,
+      "loss": 0.1004,
+      "step": 12845
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.271687110015551,
+      "learning_rate": 2.968483337958172e-07,
+      "loss": 0.1043,
+      "step": 12846
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 9.74097837819943,
+      "learning_rate": 2.96466300161547e-07,
+      "loss": 0.315,
+      "step": 12847
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.5532588494563795,
+      "learning_rate": 2.9608450500678566e-07,
+      "loss": 0.2988,
+      "step": 12848
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.731312985772003,
+      "learning_rate": 2.9570294835089164e-07,
+      "loss": 0.3144,
+      "step": 12849
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6943333823913513,
+      "learning_rate": 2.953216302132089e-07,
+      "loss": 0.3208,
+      "step": 12850
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.291233368744039,
+      "learning_rate": 2.9494055061307313e-07,
+      "loss": 0.2923,
+      "step": 12851
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.041828570837863,
+      "learning_rate": 2.9455970956980384e-07,
+      "loss": 0.2804,
+      "step": 12852
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.5210361236226495,
+      "learning_rate": 2.941791071027128e-07,
+      "loss": 0.3274,
+      "step": 12853
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.489074786625486,
+      "learning_rate": 2.9379874323109567e-07,
+      "loss": 0.2914,
+      "step": 12854
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.559533938287786,
+      "learning_rate": 2.934186179742382e-07,
+      "loss": 0.2899,
+      "step": 12855
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.057389694149861,
+      "learning_rate": 2.9303873135141437e-07,
+      "loss": 0.3072,
+      "step": 12856
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.307934208444396,
+      "learning_rate": 2.926590833818832e-07,
+      "loss": 0.3107,
+      "step": 12857
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.036444654426802,
+      "learning_rate": 2.9227967408489653e-07,
+      "loss": 0.2653,
+      "step": 12858
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.8027844588012165,
+      "learning_rate": 2.9190050347968904e-07,
+      "loss": 0.3018,
+      "step": 12859
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.3100528104603146,
+      "learning_rate": 2.915215715854858e-07,
+      "loss": 0.2967,
+      "step": 12860
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8871620335869044,
+      "learning_rate": 2.9114287842150033e-07,
+      "loss": 0.2994,
+      "step": 12861
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.983009640658557,
+      "learning_rate": 2.9076442400693296e-07,
+      "loss": 0.3079,
+      "step": 12862
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.569876832399175,
+      "learning_rate": 2.903862083609715e-07,
+      "loss": 0.31,
+      "step": 12863
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.28769318373707703,
+      "learning_rate": 2.9000823150279355e-07,
+      "loss": 0.1171,
+      "step": 12864
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 5.870511193558494,
+      "learning_rate": 2.896304934515615e-07,
+      "loss": 0.2925,
+      "step": 12865
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.229872981812998,
+      "learning_rate": 2.8925299422642994e-07,
+      "loss": 0.3011,
+      "step": 12866
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.017931357949598,
+      "learning_rate": 2.88875733846537e-07,
+      "loss": 0.3086,
+      "step": 12867
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.4845647430502247,
+      "learning_rate": 2.884987123310112e-07,
+      "loss": 0.3255,
+      "step": 12868
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.88816602533088,
+      "learning_rate": 2.881219296989679e-07,
+      "loss": 0.3124,
+      "step": 12869
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6377468087803604,
+      "learning_rate": 2.8774538596951175e-07,
+      "loss": 0.2991,
+      "step": 12870
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.559060062699298,
+      "learning_rate": 2.8736908116173465e-07,
+      "loss": 0.2864,
+      "step": 12871
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.584937368073234,
+      "learning_rate": 2.8699301529471313e-07,
+      "loss": 0.3162,
+      "step": 12872
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 6.673534404185959,
+      "learning_rate": 2.866171883875191e-07,
+      "loss": 0.3475,
+      "step": 12873
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31547056913363,
+      "learning_rate": 2.8624160045920444e-07,
+      "loss": 0.1186,
+      "step": 12874
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.134260929164716,
+      "learning_rate": 2.85866251528813e-07,
+      "loss": 0.3337,
+      "step": 12875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.531054052899506,
+      "learning_rate": 2.8549114161537606e-07,
+      "loss": 0.3228,
+      "step": 12876
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.154689739700359,
+      "learning_rate": 2.851162707379135e-07,
+      "loss": 0.2951,
+      "step": 12877
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.722729580235037,
+      "learning_rate": 2.847416389154295e-07,
+      "loss": 0.2966,
+      "step": 12878
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.4498349221829745,
+      "learning_rate": 2.8436724616692226e-07,
+      "loss": 0.3059,
+      "step": 12879
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.336250460286745,
+      "learning_rate": 2.839930925113715e-07,
+      "loss": 0.2977,
+      "step": 12880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.16404102610247,
+      "learning_rate": 2.836191779677483e-07,
+      "loss": 0.3736,
+      "step": 12881
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.067525179050098,
+      "learning_rate": 2.832455025550113e-07,
+      "loss": 0.3137,
+      "step": 12882
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3006316721044988,
+      "learning_rate": 2.828720662921064e-07,
+      "loss": 0.0986,
+      "step": 12883
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.810845737959252,
+      "learning_rate": 2.824988691979691e-07,
+      "loss": 0.2761,
+      "step": 12884
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.689056502052749,
+      "learning_rate": 2.8212591129151813e-07,
+      "loss": 0.3048,
+      "step": 12885
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1748672478036855,
+      "learning_rate": 2.817531925916667e-07,
+      "loss": 0.2834,
+      "step": 12886
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 117.78012428245573,
+      "learning_rate": 2.813807131173102e-07,
+      "loss": 0.3162,
+      "step": 12887
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.794599568490809,
+      "learning_rate": 2.810084728873347e-07,
+      "loss": 0.3158,
+      "step": 12888
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3055440552069987,
+      "learning_rate": 2.806364719206139e-07,
+      "loss": 0.1075,
+      "step": 12889
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.001357821896015,
+      "learning_rate": 2.80264710236009e-07,
+      "loss": 0.279,
+      "step": 12890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 28.560442521585035,
+      "learning_rate": 2.7989318785236917e-07,
+      "loss": 0.3066,
+      "step": 12891
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7172462173836447,
+      "learning_rate": 2.795219047885317e-07,
+      "loss": 0.3388,
+      "step": 12892
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.82159001216248,
+      "learning_rate": 2.791508610633209e-07,
+      "loss": 0.3283,
+      "step": 12893
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7989352754767167,
+      "learning_rate": 2.787800566955495e-07,
+      "loss": 0.323,
+      "step": 12894
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.700327287396821,
+      "learning_rate": 2.784094917040181e-07,
+      "loss": 0.305,
+      "step": 12895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.925520272964103,
+      "learning_rate": 2.780391661075155e-07,
+      "loss": 0.3416,
+      "step": 12896
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29051093737687167,
+      "learning_rate": 2.7766907992481886e-07,
+      "loss": 0.1058,
+      "step": 12897
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1640224150465746,
+      "learning_rate": 2.772992331746893e-07,
+      "loss": 0.2802,
+      "step": 12898
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.74795915654219,
+      "learning_rate": 2.76929625875883e-07,
+      "loss": 0.3111,
+      "step": 12899
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.737473882210747,
+      "learning_rate": 2.76560258047136e-07,
+      "loss": 0.3092,
+      "step": 12900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9797301608821396,
+      "learning_rate": 2.7619112970717956e-07,
+      "loss": 0.3255,
+      "step": 12901
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.960900451292681,
+      "learning_rate": 2.7582224087472687e-07,
+      "loss": 0.333,
+      "step": 12902
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.991335431688276,
+      "learning_rate": 2.754535915684819e-07,
+      "loss": 0.3049,
+      "step": 12903
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.508587835800671,
+      "learning_rate": 2.750851818071365e-07,
+      "loss": 0.3176,
+      "step": 12904
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.28125779430279113,
+      "learning_rate": 2.747170116093695e-07,
+      "loss": 0.1068,
+      "step": 12905
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9849625946338914,
+      "learning_rate": 2.743490809938487e-07,
+      "loss": 0.2803,
+      "step": 12906
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.841208879721427,
+      "learning_rate": 2.739813899792276e-07,
+      "loss": 0.2903,
+      "step": 12907
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 9.453834037187601,
+      "learning_rate": 2.736139385841496e-07,
+      "loss": 0.3236,
+      "step": 12908
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2835602126668975,
+      "learning_rate": 2.732467268272454e-07,
+      "loss": 0.1082,
+      "step": 12909
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.599513300943425,
+      "learning_rate": 2.7287975472713337e-07,
+      "loss": 0.3138,
+      "step": 12910
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.773678092173736,
+      "learning_rate": 2.725130223024197e-07,
+      "loss": 0.3066,
+      "step": 12911
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6892441790027486,
+      "learning_rate": 2.721465295716996e-07,
+      "loss": 0.2905,
+      "step": 12912
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.755207233502956,
+      "learning_rate": 2.7178027655355267e-07,
+      "loss": 0.3156,
+      "step": 12913
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.627799244710059,
+      "learning_rate": 2.714142632665517e-07,
+      "loss": 0.3409,
+      "step": 12914
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.262081959351703,
+      "learning_rate": 2.71048489729252e-07,
+      "loss": 0.3272,
+      "step": 12915
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5722596455247237,
+      "learning_rate": 2.706829559601998e-07,
+      "loss": 0.32,
+      "step": 12916
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8410414634480268,
+      "learning_rate": 2.7031766197792906e-07,
+      "loss": 0.3008,
+      "step": 12917
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.3262234713881424,
+      "learning_rate": 2.699526078009601e-07,
+      "loss": 0.3049,
+      "step": 12918
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2860009584529409,
+      "learning_rate": 2.695877934478036e-07,
+      "loss": 0.1215,
+      "step": 12919
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.6027454299244535,
+      "learning_rate": 2.6922321893695423e-07,
+      "loss": 0.3103,
+      "step": 12920
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.349357295392686,
+      "learning_rate": 2.6885888428689775e-07,
+      "loss": 0.2946,
+      "step": 12921
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.899795850126031,
+      "learning_rate": 2.6849478951610665e-07,
+      "loss": 0.2615,
+      "step": 12922
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.240485458488187,
+      "learning_rate": 2.6813093464304164e-07,
+      "loss": 0.3225,
+      "step": 12923
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31033127374532327,
+      "learning_rate": 2.677673196861502e-07,
+      "loss": 0.1118,
+      "step": 12924
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.59168757574774,
+      "learning_rate": 2.674039446638704e-07,
+      "loss": 0.2805,
+      "step": 12925
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7556268039211256,
+      "learning_rate": 2.6704080959462307e-07,
+      "loss": 0.3121,
+      "step": 12926
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.666409920016757,
+      "learning_rate": 2.666779144968229e-07,
+      "loss": 0.296,
+      "step": 12927
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.910130037115362,
+      "learning_rate": 2.663152593888668e-07,
+      "loss": 0.3406,
+      "step": 12928
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.077610443976661,
+      "learning_rate": 2.6595284428914514e-07,
+      "loss": 0.3137,
+      "step": 12929
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6472533824598026,
+      "learning_rate": 2.655906692160304e-07,
+      "loss": 0.3139,
+      "step": 12930
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.322130992452532,
+      "learning_rate": 2.652287341878873e-07,
+      "loss": 0.2914,
+      "step": 12931
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.5399094961663105,
+      "learning_rate": 2.648670392230668e-07,
+      "loss": 0.2952,
+      "step": 12932
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.163971629440206,
+      "learning_rate": 2.6450558433990527e-07,
+      "loss": 0.3014,
+      "step": 12933
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.7948964510428405,
+      "learning_rate": 2.6414436955673305e-07,
+      "loss": 0.3067,
+      "step": 12934
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.5634902658099517,
+      "learning_rate": 2.6378339489186166e-07,
+      "loss": 0.3299,
+      "step": 12935
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2448256807120086,
+      "learning_rate": 2.634226603635942e-07,
+      "loss": 0.2984,
+      "step": 12936
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.714589671000769,
+      "learning_rate": 2.6306216599022105e-07,
+      "loss": 0.332,
+      "step": 12937
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.310013993198288,
+      "learning_rate": 2.6270191179001925e-07,
+      "loss": 0.3224,
+      "step": 12938
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.710181220034342,
+      "learning_rate": 2.623418977812553e-07,
+      "loss": 0.3138,
+      "step": 12939
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.808642655095605,
+      "learning_rate": 2.619821239821829e-07,
+      "loss": 0.2994,
+      "step": 12940
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.276050133343112,
+      "learning_rate": 2.6162259041104143e-07,
+      "loss": 0.3201,
+      "step": 12941
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.867629933791118,
+      "learning_rate": 2.612632970860629e-07,
+      "loss": 0.3141,
+      "step": 12942
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.545149915086065,
+      "learning_rate": 2.6090424402546224e-07,
+      "loss": 0.3267,
+      "step": 12943
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.354703491086456,
+      "learning_rate": 2.605454312474448e-07,
+      "loss": 0.33,
+      "step": 12944
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.076209690304832,
+      "learning_rate": 2.6018685877020334e-07,
+      "loss": 0.3382,
+      "step": 12945
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.818529053470522,
+      "learning_rate": 2.598285266119177e-07,
+      "loss": 0.2653,
+      "step": 12946
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.021635489922409,
+      "learning_rate": 2.5947043479075726e-07,
+      "loss": 0.3341,
+      "step": 12947
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.609182968867612,
+      "learning_rate": 2.591125833248764e-07,
+      "loss": 0.3163,
+      "step": 12948
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1188685848090922,
+      "learning_rate": 2.5875497223242065e-07,
+      "loss": 0.3183,
+      "step": 12949
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.270743698568477,
+      "learning_rate": 2.583976015315209e-07,
+      "loss": 0.295,
+      "step": 12950
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2985296865765302,
+      "learning_rate": 2.5804047124029617e-07,
+      "loss": 0.1088,
+      "step": 12951
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.943442361339316,
+      "learning_rate": 2.576835813768547e-07,
+      "loss": 0.3069,
+      "step": 12952
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.290642689940122,
+      "learning_rate": 2.5732693195929136e-07,
+      "loss": 0.3119,
+      "step": 12953
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.1183411880423635,
+      "learning_rate": 2.5697052300568793e-07,
+      "loss": 0.3253,
+      "step": 12954
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2927798415650936,
+      "learning_rate": 2.566143545341171e-07,
+      "loss": 0.1077,
+      "step": 12955
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5196858460164027,
+      "learning_rate": 2.5625842656263555e-07,
+      "loss": 0.2885,
+      "step": 12956
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 14.863726886075742,
+      "learning_rate": 2.5590273910929106e-07,
+      "loss": 0.2909,
+      "step": 12957
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 10.03685782721624,
+      "learning_rate": 2.555472921921165e-07,
+      "loss": 0.2845,
+      "step": 12958
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.2048000516411586,
+      "learning_rate": 2.5519208582913455e-07,
+      "loss": 0.3165,
+      "step": 12959
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.048020811297139,
+      "learning_rate": 2.5483712003835535e-07,
+      "loss": 0.3217,
+      "step": 12960
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8689052289545773,
+      "learning_rate": 2.54482394837775e-07,
+      "loss": 0.3106,
+      "step": 12961
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.272543397008135,
+      "learning_rate": 2.5412791024538084e-07,
+      "loss": 0.3288,
+      "step": 12962
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3954922233247795,
+      "learning_rate": 2.5377366627914403e-07,
+      "loss": 0.2832,
+      "step": 12963
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.37102481303285,
+      "learning_rate": 2.534196629570268e-07,
+      "loss": 0.2951,
+      "step": 12964
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.320824262233772,
+      "learning_rate": 2.530659002969776e-07,
+      "loss": 0.2986,
+      "step": 12965
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.767911974139008,
+      "learning_rate": 2.527123783169327e-07,
+      "loss": 0.3146,
+      "step": 12966
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.776056285014746,
+      "learning_rate": 2.523590970348166e-07,
+      "loss": 0.2709,
+      "step": 12967
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 12.768570068539537,
+      "learning_rate": 2.5200605646854224e-07,
+      "loss": 0.3055,
+      "step": 12968
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3028181126555385,
+      "learning_rate": 2.5165325663600804e-07,
+      "loss": 0.3225,
+      "step": 12969
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2965327161852268,
+      "learning_rate": 2.5130069755510243e-07,
+      "loss": 0.1171,
+      "step": 12970
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2906411261857911,
+      "learning_rate": 2.509483792437012e-07,
+      "loss": 0.1083,
+      "step": 12971
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8420237246319,
+      "learning_rate": 2.5059630171966774e-07,
+      "loss": 0.2938,
+      "step": 12972
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.176552537078575,
+      "learning_rate": 2.5024446500085285e-07,
+      "loss": 0.3021,
+      "step": 12973
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.403256984027711,
+      "learning_rate": 2.4989286910509445e-07,
+      "loss": 0.3172,
+      "step": 12974
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.205656250800737,
+      "learning_rate": 2.495415140502216e-07,
+      "loss": 0.3043,
+      "step": 12975
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.733778578730952,
+      "learning_rate": 2.4919039985404626e-07,
+      "loss": 0.3672,
+      "step": 12976
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.7550019406215394,
+      "learning_rate": 2.4883952653437294e-07,
+      "loss": 0.32,
+      "step": 12977
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.872117420758294,
+      "learning_rate": 2.4848889410898967e-07,
+      "loss": 0.2944,
+      "step": 12978
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.655901611723282,
+      "learning_rate": 2.481385025956756e-07,
+      "loss": 0.3266,
+      "step": 12979
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.570385069737852,
+      "learning_rate": 2.477883520121954e-07,
+      "loss": 0.3017,
+      "step": 12980
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.806991157557392,
+      "learning_rate": 2.4743844237630377e-07,
+      "loss": 0.3521,
+      "step": 12981
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.29718786847595224,
+      "learning_rate": 2.4708877370574147e-07,
+      "loss": 0.1053,
+      "step": 12982
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4520554252771403,
+      "learning_rate": 2.467393460182366e-07,
+      "loss": 0.3013,
+      "step": 12983
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.233945937558014,
+      "learning_rate": 2.46390159331506e-07,
+      "loss": 0.2837,
+      "step": 12984
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.406994131966409,
+      "learning_rate": 2.4604121366325507e-07,
+      "loss": 0.2858,
+      "step": 12985
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.146775816652226,
+      "learning_rate": 2.4569250903117623e-07,
+      "loss": 0.3026,
+      "step": 12986
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 8.561151648233597,
+      "learning_rate": 2.4534404545294765e-07,
+      "loss": 0.3187,
+      "step": 12987
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.997727487472638,
+      "learning_rate": 2.449958229462401e-07,
+      "loss": 0.3132,
+      "step": 12988
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9899556332269808,
+      "learning_rate": 2.4464784152870623e-07,
+      "loss": 0.3203,
+      "step": 12989
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.098884262795474,
+      "learning_rate": 2.4430010121799243e-07,
+      "loss": 0.3107,
+      "step": 12990
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.174394105354343,
+      "learning_rate": 2.4395260203172787e-07,
+      "loss": 0.2941,
+      "step": 12991
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.451027047860295,
+      "learning_rate": 2.436053439875319e-07,
+      "loss": 0.2958,
+      "step": 12992
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.042013526307348,
+      "learning_rate": 2.43258327103012e-07,
+      "loss": 0.2802,
+      "step": 12993
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.961720217913052,
+      "learning_rate": 2.429115513957614e-07,
+      "loss": 0.3309,
+      "step": 12994
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.83727297684772,
+      "learning_rate": 2.425650168833643e-07,
+      "loss": 0.311,
+      "step": 12995
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.725387259810226,
+      "learning_rate": 2.4221872358338895e-07,
+      "loss": 0.3328,
+      "step": 12996
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.076720715947862,
+      "learning_rate": 2.4187267151339343e-07,
+      "loss": 0.3317,
+      "step": 12997
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.521285075922062,
+      "learning_rate": 2.415268606909238e-07,
+      "loss": 0.3029,
+      "step": 12998
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.2982920215360387,
+      "learning_rate": 2.4118129113351374e-07,
+      "loss": 0.3118,
+      "step": 12999
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.096845939355942,
+      "learning_rate": 2.408359628586837e-07,
+      "loss": 0.3176,
+      "step": 13000
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 6.098074457366071,
+      "learning_rate": 2.404908758839436e-07,
+      "loss": 0.3464,
+      "step": 13001
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 7.683453272283689,
+      "learning_rate": 2.401460302267883e-07,
+      "loss": 0.3163,
+      "step": 13002
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.706806327039697,
+      "learning_rate": 2.398014259047049e-07,
+      "loss": 0.2686,
+      "step": 13003
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 13.040852109736452,
+      "learning_rate": 2.3945706293516226e-07,
+      "loss": 0.2999,
+      "step": 13004
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.858504946831815,
+      "learning_rate": 2.3911294133562357e-07,
+      "loss": 0.3224,
+      "step": 13005
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.2887096416376093,
+      "learning_rate": 2.3876906112353435e-07,
+      "loss": 0.1105,
+      "step": 13006
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.802325469073244,
+      "learning_rate": 2.3842542231633125e-07,
+      "loss": 0.3061,
+      "step": 13007
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.116994737783494,
+      "learning_rate": 2.380820249314375e-07,
+      "loss": 0.3204,
+      "step": 13008
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.246740533559138,
+      "learning_rate": 2.3773886898626198e-07,
+      "loss": 0.3054,
+      "step": 13009
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.401933683849155,
+      "learning_rate": 2.3739595449820684e-07,
+      "loss": 0.3245,
+      "step": 13010
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.3540602521526774,
+      "learning_rate": 2.3705328148465657e-07,
+      "loss": 0.3355,
+      "step": 13011
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.3436596650599615,
+      "learning_rate": 2.3671084996298555e-07,
+      "loss": 0.2928,
+      "step": 13012
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 11.508256887912934,
+      "learning_rate": 2.3636865995055601e-07,
+      "loss": 0.3417,
+      "step": 13013
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.079127446160343,
+      "learning_rate": 2.36026711464718e-07,
+      "loss": 0.3384,
+      "step": 13014
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.8828389067262092,
+      "learning_rate": 2.3568500452280874e-07,
+      "loss": 0.3317,
+      "step": 13015
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.451360218031044,
+      "learning_rate": 2.3534353914215435e-07,
+      "loss": 0.284,
+      "step": 13016
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.9984710830338943,
+      "learning_rate": 2.35002315340066e-07,
+      "loss": 0.3515,
+      "step": 13017
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.1093054155389455,
+      "learning_rate": 2.3466133313384653e-07,
+      "loss": 0.3157,
+      "step": 13018
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.61167917328179,
+      "learning_rate": 2.343205925407832e-07,
+      "loss": 0.2879,
+      "step": 13019
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.273535308055047,
+      "learning_rate": 2.3398009357815277e-07,
+      "loss": 0.3572,
+      "step": 13020
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 26.384619498903003,
+      "learning_rate": 2.3363983626322029e-07,
+      "loss": 0.2756,
+      "step": 13021
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 27.60711357015699,
+      "learning_rate": 2.3329982061323474e-07,
+      "loss": 0.2766,
+      "step": 13022
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.701818016177006,
+      "learning_rate": 2.3296004664543902e-07,
+      "loss": 0.2923,
+      "step": 13023
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.0602676995219475,
+      "learning_rate": 2.3262051437705768e-07,
+      "loss": 0.2863,
+      "step": 13024
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.043635976968524,
+      "learning_rate": 2.3228122382530748e-07,
+      "loss": 0.2883,
+      "step": 13025
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7834086165135483,
+      "learning_rate": 2.3194217500739026e-07,
+      "loss": 0.2996,
+      "step": 13026
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.368661799156949,
+      "learning_rate": 2.316033679404972e-07,
+      "loss": 0.3157,
+      "step": 13027
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2916886871747415,
+      "learning_rate": 2.3126480264180573e-07,
+      "loss": 0.1109,
+      "step": 13028
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.358889843953424,
+      "learning_rate": 2.3092647912848376e-07,
+      "loss": 0.2825,
+      "step": 13029
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.042898984656624,
+      "learning_rate": 2.3058839741768146e-07,
+      "loss": 0.3045,
+      "step": 13030
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4614915735227236,
+      "learning_rate": 2.3025055752654458e-07,
+      "loss": 0.2974,
+      "step": 13031
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.39158450113103,
+      "learning_rate": 2.299129594721994e-07,
+      "loss": 0.298,
+      "step": 13032
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9915711142214505,
+      "learning_rate": 2.2957560327176386e-07,
+      "loss": 0.3086,
+      "step": 13033
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.091690632088228,
+      "learning_rate": 2.2923848894234268e-07,
+      "loss": 0.3162,
+      "step": 13034
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5278701451693983,
+      "learning_rate": 2.2890161650102772e-07,
+      "loss": 0.3211,
+      "step": 13035
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.461047736866523,
+      "learning_rate": 2.2856498596490085e-07,
+      "loss": 0.3082,
+      "step": 13036
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.659986515592511,
+      "learning_rate": 2.282285973510273e-07,
+      "loss": 0.2739,
+      "step": 13037
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.18266300436317,
+      "learning_rate": 2.2789245067646516e-07,
+      "loss": 0.3433,
+      "step": 13038
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.712252012861612,
+      "learning_rate": 2.2755654595825628e-07,
+      "loss": 0.325,
+      "step": 13039
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.434499616033657,
+      "learning_rate": 2.272208832134326e-07,
+      "loss": 0.3155,
+      "step": 13040
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.657198210783513,
+      "learning_rate": 2.2688546245901223e-07,
+      "loss": 0.331,
+      "step": 13041
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9399231946784705,
+      "learning_rate": 2.265502837120026e-07,
+      "loss": 0.2922,
+      "step": 13042
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.503417529028623,
+      "learning_rate": 2.262153469893974e-07,
+      "loss": 0.3239,
+      "step": 13043
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28271747335018643,
+      "learning_rate": 2.258806523081797e-07,
+      "loss": 0.1108,
+      "step": 13044
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0181533361546293,
+      "learning_rate": 2.25546199685317e-07,
+      "loss": 0.3165,
+      "step": 13045
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.0924134028885435,
+      "learning_rate": 2.252119891377691e-07,
+      "loss": 0.3033,
+      "step": 13046
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.319384321512743,
+      "learning_rate": 2.2487802068247965e-07,
+      "loss": 0.2907,
+      "step": 13047
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.929856627712086,
+      "learning_rate": 2.2454429433638292e-07,
+      "loss": 0.3256,
+      "step": 13048
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.71880907572806,
+      "learning_rate": 2.2421081011639868e-07,
+      "loss": 0.3289,
+      "step": 13049
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.166804812394233,
+      "learning_rate": 2.2387756803943506e-07,
+      "loss": 0.3371,
+      "step": 13050
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.293803487862505,
+      "learning_rate": 2.2354456812238912e-07,
+      "loss": 0.2911,
+      "step": 13051
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.133391999696051,
+      "learning_rate": 2.2321181038214347e-07,
+      "loss": 0.3138,
+      "step": 13052
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2666124185695113,
+      "learning_rate": 2.2287929483557124e-07,
+      "loss": 0.0952,
+      "step": 13053
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.329922071290403,
+      "learning_rate": 2.2254702149953068e-07,
+      "loss": 0.3049,
+      "step": 13054
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.2725873844232,
+      "learning_rate": 2.222149903908688e-07,
+      "loss": 0.3115,
+      "step": 13055
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.007192468095328,
+      "learning_rate": 2.218832015264205e-07,
+      "loss": 0.3159,
+      "step": 13056
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9081198329916553,
+      "learning_rate": 2.215516549230079e-07,
+      "loss": 0.2937,
+      "step": 13057
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29856338846776154,
+      "learning_rate": 2.2122035059744195e-07,
+      "loss": 0.1221,
+      "step": 13058
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3175000179899587,
+      "learning_rate": 2.208892885665198e-07,
+      "loss": 0.2976,
+      "step": 13059
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.21714400316549,
+      "learning_rate": 2.2055846884702692e-07,
+      "loss": 0.3259,
+      "step": 13060
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.566907022037275,
+      "learning_rate": 2.2022789145573653e-07,
+      "loss": 0.2954,
+      "step": 13061
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7658453175701685,
+      "learning_rate": 2.1989755640941079e-07,
+      "loss": 0.2929,
+      "step": 13062
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.300651733385389,
+      "learning_rate": 2.195674637247963e-07,
+      "loss": 0.3324,
+      "step": 13063
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.138884050226412,
+      "learning_rate": 2.1923761341863247e-07,
+      "loss": 0.3227,
+      "step": 13064
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.611115797582356,
+      "learning_rate": 2.189080055076398e-07,
+      "loss": 0.2826,
+      "step": 13065
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.46600095109837,
+      "learning_rate": 2.1857864000853378e-07,
+      "loss": 0.2901,
+      "step": 13066
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.30811459582990025,
+      "learning_rate": 2.1824951693801167e-07,
+      "loss": 0.1113,
+      "step": 13067
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2771937958650565,
+      "learning_rate": 2.1792063631276116e-07,
+      "loss": 0.3376,
+      "step": 13068
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.130374580141662,
+      "learning_rate": 2.175919981494573e-07,
+      "loss": 0.2627,
+      "step": 13069
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.217687636213898,
+      "learning_rate": 2.1726360246476287e-07,
+      "loss": 0.2828,
+      "step": 13070
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3045052542584763,
+      "learning_rate": 2.1693544927532895e-07,
+      "loss": 0.1184,
+      "step": 13071
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.622092708527263,
+      "learning_rate": 2.1660753859779225e-07,
+      "loss": 0.324,
+      "step": 13072
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.523129626916043,
+      "learning_rate": 2.1627987044877953e-07,
+      "loss": 0.3051,
+      "step": 13073
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.0523839889297095,
+      "learning_rate": 2.1595244484490353e-07,
+      "loss": 0.3368,
+      "step": 13074
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.148059215830521,
+      "learning_rate": 2.15625261802766e-07,
+      "loss": 0.313,
+      "step": 13075
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.802963682934303,
+      "learning_rate": 2.152983213389559e-07,
+      "loss": 0.2832,
+      "step": 13076
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.2166682619529,
+      "learning_rate": 2.1497162347004996e-07,
+      "loss": 0.2942,
+      "step": 13077
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.751860127123058,
+      "learning_rate": 2.1464516821261107e-07,
+      "loss": 0.3028,
+      "step": 13078
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9659441580173294,
+      "learning_rate": 2.1431895558319372e-07,
+      "loss": 0.3377,
+      "step": 13079
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.222600447967332,
+      "learning_rate": 2.139929855983347e-07,
+      "loss": 0.3283,
+      "step": 13080
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.974204437840145,
+      "learning_rate": 2.1366725827456468e-07,
+      "loss": 0.2859,
+      "step": 13081
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.7544534956235145,
+      "learning_rate": 2.1334177362839604e-07,
+      "loss": 0.3085,
+      "step": 13082
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3304275430740487,
+      "learning_rate": 2.1301653167633273e-07,
+      "loss": 0.1052,
+      "step": 13083
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9201322927066955,
+      "learning_rate": 2.126915324348655e-07,
+      "loss": 0.3013,
+      "step": 13084
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2867467927261779,
+      "learning_rate": 2.1236677592047117e-07,
+      "loss": 0.1118,
+      "step": 13085
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.137464572106731,
+      "learning_rate": 2.120422621496171e-07,
+      "loss": 0.2936,
+      "step": 13086
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.5868777476180025,
+      "learning_rate": 2.1171799113875624e-07,
+      "loss": 0.2927,
+      "step": 13087
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.088899014568934,
+      "learning_rate": 2.113939629043299e-07,
+      "loss": 0.3271,
+      "step": 13088
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4556350446638646,
+      "learning_rate": 2.1107017746276658e-07,
+      "loss": 0.3341,
+      "step": 13089
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.327014611597163,
+      "learning_rate": 2.1074663483048375e-07,
+      "loss": 0.332,
+      "step": 13090
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.920189802713978,
+      "learning_rate": 2.104233350238849e-07,
+      "loss": 0.2839,
+      "step": 13091
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7759447374465154,
+      "learning_rate": 2.1010027805936307e-07,
+      "loss": 0.3188,
+      "step": 13092
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.633658791292508,
+      "learning_rate": 2.0977746395329625e-07,
+      "loss": 0.3088,
+      "step": 13093
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.704877581232703,
+      "learning_rate": 2.094548927220541e-07,
+      "loss": 0.3046,
+      "step": 13094
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.338876905307156,
+      "learning_rate": 2.0913256438199026e-07,
+      "loss": 0.2961,
+      "step": 13095
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.769410196877171,
+      "learning_rate": 2.0881047894944716e-07,
+      "loss": 0.3053,
+      "step": 13096
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.6845979297012312,
+      "learning_rate": 2.0848863644075678e-07,
+      "loss": 0.3159,
+      "step": 13097
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.28487130113817244,
+      "learning_rate": 2.0816703687223494e-07,
+      "loss": 0.1004,
+      "step": 13098
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7712900883272433,
+      "learning_rate": 2.0784568026019025e-07,
+      "loss": 0.2902,
+      "step": 13099
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.777255024339449,
+      "learning_rate": 2.0752456662091357e-07,
+      "loss": 0.326,
+      "step": 13100
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.2292457981669638,
+      "learning_rate": 2.0720369597068802e-07,
+      "loss": 0.2724,
+      "step": 13101
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.852788217573254,
+      "learning_rate": 2.068830683257811e-07,
+      "loss": 0.3108,
+      "step": 13102
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3007781675309049,
+      "learning_rate": 2.065626837024498e-07,
+      "loss": 0.1112,
+      "step": 13103
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.395867711293913,
+      "learning_rate": 2.0624254211693894e-07,
+      "loss": 0.2978,
+      "step": 13104
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.276780151434739,
+      "learning_rate": 2.0592264358547999e-07,
+      "loss": 0.3048,
+      "step": 13105
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.41046119239121,
+      "learning_rate": 2.0560298812429158e-07,
+      "loss": 0.3001,
+      "step": 13106
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 8.47214515994881,
+      "learning_rate": 2.0528357574958246e-07,
+      "loss": 0.3354,
+      "step": 13107
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 13.712718092083453,
+      "learning_rate": 2.049644064775469e-07,
+      "loss": 0.3075,
+      "step": 13108
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.254332685412388,
+      "learning_rate": 2.0464548032436694e-07,
+      "loss": 0.324,
+      "step": 13109
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.809665668067321,
+      "learning_rate": 2.0432679730621406e-07,
+      "loss": 0.3021,
+      "step": 13110
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 10.932577452402954,
+      "learning_rate": 2.040083574392443e-07,
+      "loss": 0.3325,
+      "step": 13111
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.297828047452645,
+      "learning_rate": 2.036901607396058e-07,
+      "loss": 0.2982,
+      "step": 13112
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.9065086461399385,
+      "learning_rate": 2.03372207223429e-07,
+      "loss": 0.3046,
+      "step": 13113
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.679359873489994,
+      "learning_rate": 2.0305449690683709e-07,
+      "loss": 0.2908,
+      "step": 13114
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.542260165562717,
+      "learning_rate": 2.027370298059378e-07,
+      "loss": 0.3237,
+      "step": 13115
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8395665731077644,
+      "learning_rate": 2.0241980593682764e-07,
+      "loss": 0.2587,
+      "step": 13116
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4193747434556268,
+      "learning_rate": 2.0210282531559044e-07,
+      "loss": 0.282,
+      "step": 13117
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.924942909018676,
+      "learning_rate": 2.0178608795829724e-07,
+      "loss": 0.2983,
+      "step": 13118
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.052836734003057,
+      "learning_rate": 2.0146959388100852e-07,
+      "loss": 0.2945,
+      "step": 13119
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.8330666268543965,
+      "learning_rate": 2.0115334309977085e-07,
+      "loss": 0.2864,
+      "step": 13120
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 12.354205773615583,
+      "learning_rate": 2.0083733563061813e-07,
+      "loss": 0.3081,
+      "step": 13121
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.3615205422491607,
+      "learning_rate": 2.0052157148957253e-07,
+      "loss": 0.3148,
+      "step": 13122
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5902337691795734,
+      "learning_rate": 2.0020605069264508e-07,
+      "loss": 0.3136,
+      "step": 13123
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.32214789550677,
+      "learning_rate": 1.998907732558325e-07,
+      "loss": 0.2828,
+      "step": 13124
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.568819301843481,
+      "learning_rate": 1.9957573919512142e-07,
+      "loss": 0.2801,
+      "step": 13125
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.024238451669534,
+      "learning_rate": 1.9926094852648238e-07,
+      "loss": 0.3391,
+      "step": 13126
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.7293534400533366,
+      "learning_rate": 1.9894640126587818e-07,
+      "loss": 0.3241,
+      "step": 13127
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.906613619790935,
+      "learning_rate": 1.9863209742925548e-07,
+      "loss": 0.3065,
+      "step": 13128
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.4388016735837583,
+      "learning_rate": 1.983180370325516e-07,
+      "loss": 0.2905,
+      "step": 13129
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.384733803616991,
+      "learning_rate": 1.9800422009168873e-07,
+      "loss": 0.3122,
+      "step": 13130
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2827906852308143,
+      "learning_rate": 1.976906466225792e-07,
+      "loss": 0.1036,
+      "step": 13131
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.611547678223231,
+      "learning_rate": 1.9737731664112137e-07,
+      "loss": 0.3114,
+      "step": 13132
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.414061828352721,
+      "learning_rate": 1.9706423016320198e-07,
+      "loss": 0.3192,
+      "step": 13133
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.336157969502666,
+      "learning_rate": 1.9675138720469445e-07,
+      "loss": 0.3065,
+      "step": 13134
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 11.41934614733115,
+      "learning_rate": 1.9643878778146164e-07,
+      "loss": 0.3079,
+      "step": 13135
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1747433744304043,
+      "learning_rate": 1.96126431909352e-07,
+      "loss": 0.296,
+      "step": 13136
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.7184938576428865,
+      "learning_rate": 1.9581431960420337e-07,
+      "loss": 0.3351,
+      "step": 13137
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.9978681487266097,
+      "learning_rate": 1.9550245088184094e-07,
+      "loss": 0.2992,
+      "step": 13138
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.209167843574298,
+      "learning_rate": 1.9519082575807534e-07,
+      "loss": 0.3082,
+      "step": 13139
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 9.879019493290725,
+      "learning_rate": 1.9487944424870952e-07,
+      "loss": 0.2991,
+      "step": 13140
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.579048837056524,
+      "learning_rate": 1.9456830636952805e-07,
+      "loss": 0.3212,
+      "step": 13141
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.396562567317352,
+      "learning_rate": 1.9425741213630889e-07,
+      "loss": 0.2784,
+      "step": 13142
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.946937882774631,
+      "learning_rate": 1.9394676156481328e-07,
+      "loss": 0.3284,
+      "step": 13143
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.656311278879156,
+      "learning_rate": 1.9363635467079312e-07,
+      "loss": 0.3222,
+      "step": 13144
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.860733887138433,
+      "learning_rate": 1.9332619146998576e-07,
+      "loss": 0.3292,
+      "step": 13145
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.0642586661465065,
+      "learning_rate": 1.9301627197811756e-07,
+      "loss": 0.3106,
+      "step": 13146
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.622919835258127,
+      "learning_rate": 1.9270659621090315e-07,
+      "loss": 0.3321,
+      "step": 13147
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.199024796471642,
+      "learning_rate": 1.923971641840422e-07,
+      "loss": 0.286,
+      "step": 13148
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.028174395885989,
+      "learning_rate": 1.9208797591322382e-07,
+      "loss": 0.3055,
+      "step": 13149
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.159293464334485,
+      "learning_rate": 1.917790314141249e-07,
+      "loss": 0.2845,
+      "step": 13150
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.745798696646892,
+      "learning_rate": 1.9147033070241016e-07,
+      "loss": 0.2825,
+      "step": 13151
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.07114291282158,
+      "learning_rate": 1.9116187379373043e-07,
+      "loss": 0.282,
+      "step": 13152
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.866033147249958,
+      "learning_rate": 1.908536607037259e-07,
+      "loss": 0.3087,
+      "step": 13153
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.348659944151343,
+      "learning_rate": 1.905456914480225e-07,
+      "loss": 0.2753,
+      "step": 13154
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.808308485301513,
+      "learning_rate": 1.902379660422371e-07,
+      "loss": 0.3164,
+      "step": 13155
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 5.525549767276539,
+      "learning_rate": 1.899304845019695e-07,
+      "loss": 0.3019,
+      "step": 13156
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.7792744302430785,
+      "learning_rate": 1.8962324684281165e-07,
+      "loss": 0.2806,
+      "step": 13157
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.4936136640952045,
+      "learning_rate": 1.8931625308033995e-07,
+      "loss": 0.2939,
+      "step": 13158
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 6.015858742982455,
+      "learning_rate": 1.8900950323012034e-07,
+      "loss": 0.333,
+      "step": 13159
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 7.349129052944838,
+      "learning_rate": 1.8870299730770646e-07,
+      "loss": 0.3028,
+      "step": 13160
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.9176142684318838,
+      "learning_rate": 1.883967353286359e-07,
+      "loss": 0.3486,
+      "step": 13161
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.824399509808124,
+      "learning_rate": 1.8809071730844063e-07,
+      "loss": 0.3035,
+      "step": 13162
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.523807356855859,
+      "learning_rate": 1.8778494326263386e-07,
+      "loss": 0.3198,
+      "step": 13163
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.022750994891177,
+      "learning_rate": 1.8747941320671982e-07,
+      "loss": 0.2762,
+      "step": 13164
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.512984993621683,
+      "learning_rate": 1.8717412715618942e-07,
+      "loss": 0.3342,
+      "step": 13165
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.317070591142976,
+      "learning_rate": 1.8686908512652146e-07,
+      "loss": 0.3101,
+      "step": 13166
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2005777274625404,
+      "learning_rate": 1.8656428713318242e-07,
+      "loss": 0.3052,
+      "step": 13167
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.550737873321031,
+      "learning_rate": 1.8625973319162605e-07,
+      "loss": 0.3006,
+      "step": 13168
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2273787034755927,
+      "learning_rate": 1.8595542331729333e-07,
+      "loss": 0.3109,
+      "step": 13169
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1269997260784024,
+      "learning_rate": 1.856513575256147e-07,
+      "loss": 0.3375,
+      "step": 13170
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.791995822195411,
+      "learning_rate": 1.853475358320056e-07,
+      "loss": 0.3076,
+      "step": 13171
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.909752173531739,
+      "learning_rate": 1.850439582518715e-07,
+      "loss": 0.3367,
+      "step": 13172
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.161356146160505,
+      "learning_rate": 1.8474062480060452e-07,
+      "loss": 0.3368,
+      "step": 13173
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3171340368314772,
+      "learning_rate": 1.8443753549358233e-07,
+      "loss": 0.3101,
+      "step": 13174
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.293170974616497,
+      "learning_rate": 1.8413469034617547e-07,
+      "loss": 0.2952,
+      "step": 13175
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9422072306010736,
+      "learning_rate": 1.838320893737361e-07,
+      "loss": 0.3012,
+      "step": 13176
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2904314172616761,
+      "learning_rate": 1.83529732591608e-07,
+      "loss": 0.1084,
+      "step": 13177
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.048452141377199,
+      "learning_rate": 1.8322762001512118e-07,
+      "loss": 0.344,
+      "step": 13178
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.676791191459291,
+      "learning_rate": 1.8292575165959337e-07,
+      "loss": 0.2721,
+      "step": 13179
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.732211307103113,
+      "learning_rate": 1.826241275403301e-07,
+      "loss": 0.3027,
+      "step": 13180
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.587603062924012,
+      "learning_rate": 1.823227476726247e-07,
+      "loss": 0.295,
+      "step": 13181
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.939873584792414,
+      "learning_rate": 1.8202161207175606e-07,
+      "loss": 0.3059,
+      "step": 13182
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 13.028648920890204,
+      "learning_rate": 1.8172072075299475e-07,
+      "loss": 0.3086,
+      "step": 13183
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.997721408075111,
+      "learning_rate": 1.8142007373159521e-07,
+      "loss": 0.2906,
+      "step": 13184
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.494387820968015,
+      "learning_rate": 1.8111967102280082e-07,
+      "loss": 0.2791,
+      "step": 13185
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.918633281543091,
+      "learning_rate": 1.8081951264184383e-07,
+      "loss": 0.3176,
+      "step": 13186
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.510997272317052,
+      "learning_rate": 1.8051959860394098e-07,
+      "loss": 0.2997,
+      "step": 13187
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.564681198083134,
+      "learning_rate": 1.8021992892430118e-07,
+      "loss": 0.2966,
+      "step": 13188
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.664483386104206,
+      "learning_rate": 1.7992050361811563e-07,
+      "loss": 0.3254,
+      "step": 13189
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2896922329398951,
+      "learning_rate": 1.7962132270056832e-07,
+      "loss": 0.111,
+      "step": 13190
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.3988835692148855,
+      "learning_rate": 1.793223861868265e-07,
+      "loss": 0.2872,
+      "step": 13191
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.143638839579418,
+      "learning_rate": 1.790236940920481e-07,
+      "loss": 0.2963,
+      "step": 13192
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.855863590056731,
+      "learning_rate": 1.7872524643137657e-07,
+      "loss": 0.3132,
+      "step": 13193
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.194827564573365,
+      "learning_rate": 1.7842704321994419e-07,
+      "loss": 0.3371,
+      "step": 13194
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 28.594954004303947,
+      "learning_rate": 1.7812908447287113e-07,
+      "loss": 0.3113,
+      "step": 13195
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.434317610734488,
+      "learning_rate": 1.7783137020526419e-07,
+      "loss": 0.292,
+      "step": 13196
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.606754864867924,
+      "learning_rate": 1.7753390043221742e-07,
+      "loss": 0.3001,
+      "step": 13197
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.0613017995247676,
+      "learning_rate": 1.7723667516881372e-07,
+      "loss": 0.2916,
+      "step": 13198
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.481345387699385,
+      "learning_rate": 1.7693969443012327e-07,
+      "loss": 0.3137,
+      "step": 13199
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.325696336512415,
+      "learning_rate": 1.7664295823120347e-07,
+      "loss": 0.3193,
+      "step": 13200
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2980011179940361,
+      "learning_rate": 1.763464665871001e-07,
+      "loss": 0.0884,
+      "step": 13201
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.513844562702155,
+      "learning_rate": 1.760502195128444e-07,
+      "loss": 0.276,
+      "step": 13202
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.16040129010435,
+      "learning_rate": 1.757542170234583e-07,
+      "loss": 0.288,
+      "step": 13203
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.32508236199476,
+      "learning_rate": 1.7545845913394921e-07,
+      "loss": 0.345,
+      "step": 13204
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.2879361324089285,
+      "learning_rate": 1.7516294585931236e-07,
+      "loss": 0.3026,
+      "step": 13205
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.1271882038299585,
+      "learning_rate": 1.7486767721453134e-07,
+      "loss": 0.3074,
+      "step": 13206
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2930274528669052,
+      "learning_rate": 1.745726532145764e-07,
+      "loss": 0.2695,
+      "step": 13207
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2952057838150193,
+      "learning_rate": 1.7427787387440665e-07,
+      "loss": 0.1156,
+      "step": 13208
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 33.88763709537668,
+      "learning_rate": 1.7398333920896847e-07,
+      "loss": 0.276,
+      "step": 13209
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.0572358858769855,
+      "learning_rate": 1.7368904923319384e-07,
+      "loss": 0.3101,
+      "step": 13210
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6477093475434335,
+      "learning_rate": 1.733950039620047e-07,
+      "loss": 0.3069,
+      "step": 13211
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.006059156704779,
+      "learning_rate": 1.7310120341031023e-07,
+      "loss": 0.3014,
+      "step": 13212
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8128146649649586,
+      "learning_rate": 1.728076475930057e-07,
+      "loss": 0.2996,
+      "step": 13213
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.677001125825262,
+      "learning_rate": 1.7251433652497708e-07,
+      "loss": 0.3287,
+      "step": 13214
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.913748296294978,
+      "learning_rate": 1.7222127022109293e-07,
+      "loss": 0.3045,
+      "step": 13215
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.1222411330931825,
+      "learning_rate": 1.7192844869621472e-07,
+      "loss": 0.2836,
+      "step": 13216
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.972726288368197,
+      "learning_rate": 1.7163587196518782e-07,
+      "loss": 0.3039,
+      "step": 13217
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.381813000591067,
+      "learning_rate": 1.7134354004284759e-07,
+      "loss": 0.2951,
+      "step": 13218
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.688973682598761,
+      "learning_rate": 1.7105145294401548e-07,
+      "loss": 0.2711,
+      "step": 13219
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.698915314028734,
+      "learning_rate": 1.7075961068350023e-07,
+      "loss": 0.2952,
+      "step": 13220
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2455241312398724,
+      "learning_rate": 1.7046801327609995e-07,
+      "loss": 0.3218,
+      "step": 13221
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0419854845094823,
+      "learning_rate": 1.7017666073659845e-07,
+      "loss": 0.3147,
+      "step": 13222
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 26.029980362916397,
+      "learning_rate": 1.6988555307976883e-07,
+      "loss": 0.3006,
+      "step": 13223
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.4759635804553115,
+      "learning_rate": 1.695946903203699e-07,
+      "loss": 0.3007,
+      "step": 13224
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.4315099883492834,
+      "learning_rate": 1.6930407247314984e-07,
+      "loss": 0.3019,
+      "step": 13225
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.27433227023294,
+      "learning_rate": 1.6901369955284297e-07,
+      "loss": 0.3355,
+      "step": 13226
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.8398125205431075,
+      "learning_rate": 1.6872357157417195e-07,
+      "loss": 0.3673,
+      "step": 13227
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3029221020189759,
+      "learning_rate": 1.6843368855184728e-07,
+      "loss": 0.1156,
+      "step": 13228
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.672941259195663,
+      "learning_rate": 1.6814405050056714e-07,
+      "loss": 0.3166,
+      "step": 13229
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.6300922103651767,
+      "learning_rate": 1.6785465743501539e-07,
+      "loss": 0.3561,
+      "step": 13230
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.714164490772477,
+      "learning_rate": 1.6756550936986638e-07,
+      "loss": 0.3176,
+      "step": 13231
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.4695842824390484,
+      "learning_rate": 1.6727660631977894e-07,
+      "loss": 0.3282,
+      "step": 13232
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9294391852599047,
+      "learning_rate": 1.6698794829940247e-07,
+      "loss": 0.2806,
+      "step": 13233
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.093069972434479,
+      "learning_rate": 1.666995353233719e-07,
+      "loss": 0.3027,
+      "step": 13234
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.0680245437444995,
+      "learning_rate": 1.664113674063106e-07,
+      "loss": 0.2687,
+      "step": 13235
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.775691693174923,
+      "learning_rate": 1.6612344456283013e-07,
+      "loss": 0.3471,
+      "step": 13236
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.53230097770478,
+      "learning_rate": 1.6583576680752667e-07,
+      "loss": 0.2856,
+      "step": 13237
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.593906848811701,
+      "learning_rate": 1.655483341549885e-07,
+      "loss": 0.3009,
+      "step": 13238
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.29926922866459055,
+      "learning_rate": 1.6526114661978786e-07,
+      "loss": 0.1054,
+      "step": 13239
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3562805478486686,
+      "learning_rate": 1.6497420421648535e-07,
+      "loss": 0.3221,
+      "step": 13240
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.56246657705914,
+      "learning_rate": 1.6468750695963042e-07,
+      "loss": 0.311,
+      "step": 13241
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7310126097305716,
+      "learning_rate": 1.6440105486375978e-07,
+      "loss": 0.269,
+      "step": 13242
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1724849416717658,
+      "learning_rate": 1.6411484794339516e-07,
+      "loss": 0.3039,
+      "step": 13243
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.797594740601411,
+      "learning_rate": 1.6382888621304992e-07,
+      "loss": 0.3083,
+      "step": 13244
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.375818982918554,
+      "learning_rate": 1.6354316968722195e-07,
+      "loss": 0.2841,
+      "step": 13245
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.924829979789864,
+      "learning_rate": 1.6325769838039795e-07,
+      "loss": 0.2708,
+      "step": 13246
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.096326843639377,
+      "learning_rate": 1.6297247230705193e-07,
+      "loss": 0.2972,
+      "step": 13247
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.594363749540564,
+      "learning_rate": 1.6268749148164563e-07,
+      "loss": 0.3136,
+      "step": 13248
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.06332423879746,
+      "learning_rate": 1.6240275591862808e-07,
+      "loss": 0.3113,
+      "step": 13249
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.7919247239015146,
+      "learning_rate": 1.621182656324355e-07,
+      "loss": 0.2973,
+      "step": 13250
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.6957800429393455,
+      "learning_rate": 1.61834020637493e-07,
+      "loss": 0.3206,
+      "step": 13251
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.469624496562684,
+      "learning_rate": 1.615500209482118e-07,
+      "loss": 0.3235,
+      "step": 13252
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.083327545664178,
+      "learning_rate": 1.6126626657899215e-07,
+      "loss": 0.3011,
+      "step": 13253
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.867314609598927,
+      "learning_rate": 1.6098275754421965e-07,
+      "loss": 0.3116,
+      "step": 13254
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.855076513401049,
+      "learning_rate": 1.606994938582701e-07,
+      "loss": 0.3076,
+      "step": 13255
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.533903798683842,
+      "learning_rate": 1.6041647553550477e-07,
+      "loss": 0.2962,
+      "step": 13256
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.158799687535214,
+      "learning_rate": 1.601337025902744e-07,
+      "loss": 0.3577,
+      "step": 13257
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.990083633874562,
+      "learning_rate": 1.5985117503691417e-07,
+      "loss": 0.3148,
+      "step": 13258
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.545522361691908,
+      "learning_rate": 1.5956889288975152e-07,
+      "loss": 0.3409,
+      "step": 13259
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.537863510151903,
+      "learning_rate": 1.592868561630967e-07,
+      "loss": 0.2903,
+      "step": 13260
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.374453619074431,
+      "learning_rate": 1.5900506487125045e-07,
+      "loss": 0.2962,
+      "step": 13261
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.28602195277443404,
+      "learning_rate": 1.587235190285008e-07,
+      "loss": 0.107,
+      "step": 13262
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.490840149269125,
+      "learning_rate": 1.5844221864912024e-07,
+      "loss": 0.3002,
+      "step": 13263
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.231628683016916,
+      "learning_rate": 1.5816116374737456e-07,
+      "loss": 0.3113,
+      "step": 13264
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7690341226184563,
+      "learning_rate": 1.5788035433751126e-07,
+      "loss": 0.2912,
+      "step": 13265
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 16.39957308941967,
+      "learning_rate": 1.5759979043377006e-07,
+      "loss": 0.3054,
+      "step": 13266
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.973711080400989,
+      "learning_rate": 1.5731947205037456e-07,
+      "loss": 0.2972,
+      "step": 13267
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8555132835400463,
+      "learning_rate": 1.5703939920153844e-07,
+      "loss": 0.3344,
+      "step": 13268
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9022074701259553,
+      "learning_rate": 1.5675957190146195e-07,
+      "loss": 0.294,
+      "step": 13269
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.31415638250901,
+      "learning_rate": 1.5647999016433212e-07,
+      "loss": 0.2881,
+      "step": 13270
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.745441043727933,
+      "learning_rate": 1.5620065400432537e-07,
+      "loss": 0.2791,
+      "step": 13271
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.36434733113395,
+      "learning_rate": 1.5592156343560484e-07,
+      "loss": 0.312,
+      "step": 13272
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.476932144302252,
+      "learning_rate": 1.556427184723197e-07,
+      "loss": 0.3042,
+      "step": 13273
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.7112639983914595,
+      "learning_rate": 1.5536411912860871e-07,
+      "loss": 0.3522,
+      "step": 13274
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.041918325321143,
+      "learning_rate": 1.5508576541859776e-07,
+      "loss": 0.3188,
+      "step": 13275
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2931945357108116,
+      "learning_rate": 1.548076573564e-07,
+      "loss": 0.0999,
+      "step": 13276
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.812503867580872,
+      "learning_rate": 1.5452979495611588e-07,
+      "loss": 0.2891,
+      "step": 13277
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3671312145849344,
+      "learning_rate": 1.5425217823183293e-07,
+      "loss": 0.3201,
+      "step": 13278
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 10.60632422585398,
+      "learning_rate": 1.5397480719762882e-07,
+      "loss": 0.3363,
+      "step": 13279
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.6950480561366374,
+      "learning_rate": 1.536976818675645e-07,
+      "loss": 0.3432,
+      "step": 13280
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.88665699231129,
+      "learning_rate": 1.5342080225569268e-07,
+      "loss": 0.3053,
+      "step": 13281
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.370592835665104,
+      "learning_rate": 1.5314416837605096e-07,
+      "loss": 0.2754,
+      "step": 13282
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.882406961307943,
+      "learning_rate": 1.5286778024266534e-07,
+      "loss": 0.2876,
+      "step": 13283
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.8816352786697776,
+      "learning_rate": 1.5259163786954912e-07,
+      "loss": 0.292,
+      "step": 13284
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.71275259311327,
+      "learning_rate": 1.5231574127070436e-07,
+      "loss": 0.2993,
+      "step": 13285
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.089662645939413,
+      "learning_rate": 1.5204009046011882e-07,
+      "loss": 0.3108,
+      "step": 13286
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.257776217003585,
+      "learning_rate": 1.517646854517679e-07,
+      "loss": 0.2882,
+      "step": 13287
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.305628617313628,
+      "learning_rate": 1.5148952625961665e-07,
+      "loss": 0.3109,
+      "step": 13288
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.8645433928518615,
+      "learning_rate": 1.5121461289761497e-07,
+      "loss": 0.2859,
+      "step": 13289
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5426058365481974,
+      "learning_rate": 1.5093994537970336e-07,
+      "loss": 0.3296,
+      "step": 13290
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.208902748687126,
+      "learning_rate": 1.5066552371980513e-07,
+      "loss": 0.3015,
+      "step": 13291
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.24463264381249,
+      "learning_rate": 1.5039134793183752e-07,
+      "loss": 0.3127,
+      "step": 13292
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 5.621383409772935,
+      "learning_rate": 1.501174180296988e-07,
+      "loss": 0.2934,
+      "step": 13293
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 28.59296192555215,
+      "learning_rate": 1.4984373402728014e-07,
+      "loss": 0.318,
+      "step": 13294
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3163249962231256,
+      "learning_rate": 1.4957029593845652e-07,
+      "loss": 0.3498,
+      "step": 13295
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 9.580600938497007,
+      "learning_rate": 1.492971037770924e-07,
+      "loss": 0.3059,
+      "step": 13296
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 8.194715069595933,
+      "learning_rate": 1.4902415755703892e-07,
+      "loss": 0.3184,
+      "step": 13297
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.9913901089601778,
+      "learning_rate": 1.4875145729213557e-07,
+      "loss": 0.3124,
+      "step": 13298
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.145185758255215,
+      "learning_rate": 1.484790029962091e-07,
+      "loss": 0.3233,
+      "step": 13299
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3148377466822105,
+      "learning_rate": 1.482067946830723e-07,
+      "loss": 0.2993,
+      "step": 13300
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.2948834799525288,
+      "learning_rate": 1.4793483236652694e-07,
+      "loss": 0.1098,
+      "step": 13301
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.829248350220736,
+      "learning_rate": 1.476631160603631e-07,
+      "loss": 0.3236,
+      "step": 13302
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.3568043112128194,
+      "learning_rate": 1.4739164577835696e-07,
+      "loss": 0.2881,
+      "step": 13303
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2920806550222403,
+      "learning_rate": 1.4712042153427196e-07,
+      "loss": 0.3293,
+      "step": 13304
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 6.202387344095366,
+      "learning_rate": 1.4684944334186158e-07,
+      "loss": 0.3288,
+      "step": 13305
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.957819422476048,
+      "learning_rate": 1.4657871121486255e-07,
+      "loss": 0.2875,
+      "step": 13306
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.558439605397821,
+      "learning_rate": 1.4630822516700338e-07,
+      "loss": 0.3432,
+      "step": 13307
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.472478673745626,
+      "learning_rate": 1.4603798521199808e-07,
+      "loss": 0.285,
+      "step": 13308
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6903161632645376,
+      "learning_rate": 1.4576799136354737e-07,
+      "loss": 0.3023,
+      "step": 13309
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.782889072486279,
+      "learning_rate": 1.4549824363534137e-07,
+      "loss": 0.3188,
+      "step": 13310
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.712161248866709,
+      "learning_rate": 1.4522874204105696e-07,
+      "loss": 0.3285,
+      "step": 13311
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.106927131205561,
+      "learning_rate": 1.4495948659435932e-07,
+      "loss": 0.3206,
+      "step": 13312
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7226280952934996,
+      "learning_rate": 1.4469047730889806e-07,
+      "loss": 0.3072,
+      "step": 13313
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.567375166554731,
+      "learning_rate": 1.4442171419831396e-07,
+      "loss": 0.3188,
+      "step": 13314
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3857064802566468,
+      "learning_rate": 1.4415319727623333e-07,
+      "loss": 0.3101,
+      "step": 13315
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.199292885963717,
+      "learning_rate": 1.438849265562714e-07,
+      "loss": 0.3278,
+      "step": 13316
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.869818301628255,
+      "learning_rate": 1.436169020520295e-07,
+      "loss": 0.2996,
+      "step": 13317
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3358273009362702,
+      "learning_rate": 1.4334912377709732e-07,
+      "loss": 0.3238,
+      "step": 13318
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.217144401525592,
+      "learning_rate": 1.4308159174505064e-07,
+      "loss": 0.2951,
+      "step": 13319
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.272016107705226,
+      "learning_rate": 1.4281430596945644e-07,
+      "loss": 0.3232,
+      "step": 13320
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.675499490132814,
+      "learning_rate": 1.425472664638644e-07,
+      "loss": 0.266,
+      "step": 13321
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.904085702260201,
+      "learning_rate": 1.4228047324181483e-07,
+      "loss": 0.2918,
+      "step": 13322
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.869998408501722,
+      "learning_rate": 1.4201392631683465e-07,
+      "loss": 0.3335,
+      "step": 13323
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.992189249841176,
+      "learning_rate": 1.417476257024386e-07,
+      "loss": 0.2937,
+      "step": 13324
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.826609445930353,
+      "learning_rate": 1.414815714121287e-07,
+      "loss": 0.3137,
+      "step": 13325
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.308603468366639,
+      "learning_rate": 1.4121576345939359e-07,
+      "loss": 0.3147,
+      "step": 13326
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.565190060384741,
+      "learning_rate": 1.409502018577119e-07,
+      "loss": 0.3253,
+      "step": 13327
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.8306169690162,
+      "learning_rate": 1.4068488662054735e-07,
+      "loss": 0.3088,
+      "step": 13328
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3190054147224357,
+      "learning_rate": 1.404198177613514e-07,
+      "loss": 0.2789,
+      "step": 13329
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.9728239332085105,
+      "learning_rate": 1.4015499529356434e-07,
+      "loss": 0.3265,
+      "step": 13330
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.474251019949806,
+      "learning_rate": 1.398904192306133e-07,
+      "loss": 0.3054,
+      "step": 13331
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.411586539631309,
+      "learning_rate": 1.396260895859125e-07,
+      "loss": 0.2972,
+      "step": 13332
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.254153988503194,
+      "learning_rate": 1.3936200637286512e-07,
+      "loss": 0.3108,
+      "step": 13333
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.753299821116188,
+      "learning_rate": 1.3909816960485879e-07,
+      "loss": 0.326,
+      "step": 13334
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.571526591003044,
+      "learning_rate": 1.388345792952722e-07,
+      "loss": 0.2869,
+      "step": 13335
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8964508946041825,
+      "learning_rate": 1.385712354574692e-07,
+      "loss": 0.2907,
+      "step": 13336
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.099002066918979,
+      "learning_rate": 1.3830813810480238e-07,
+      "loss": 0.3276,
+      "step": 13337
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.759515000152514,
+      "learning_rate": 1.3804528725061105e-07,
+      "loss": 0.3263,
+      "step": 13338
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.235017775243373,
+      "learning_rate": 1.3778268290822183e-07,
+      "loss": 0.2877,
+      "step": 13339
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.943287095848149,
+      "learning_rate": 1.3752032509095127e-07,
+      "loss": 0.3099,
+      "step": 13340
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.802574741398075,
+      "learning_rate": 1.372582138120987e-07,
+      "loss": 0.2915,
+      "step": 13341
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.392805163215486,
+      "learning_rate": 1.3699634908495628e-07,
+      "loss": 0.3151,
+      "step": 13342
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.236549129777739,
+      "learning_rate": 1.3673473092279953e-07,
+      "loss": 0.317,
+      "step": 13343
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9411997974065445,
+      "learning_rate": 1.364733593388934e-07,
+      "loss": 0.2935,
+      "step": 13344
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.083477974426949,
+      "learning_rate": 1.3621223434649056e-07,
+      "loss": 0.3157,
+      "step": 13345
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.7705032260571256,
+      "learning_rate": 1.359513559588299e-07,
+      "loss": 0.314,
+      "step": 13346
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6771366603049804,
+      "learning_rate": 1.3569072418913864e-07,
+      "loss": 0.3333,
+      "step": 13347
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7204342581030367,
+      "learning_rate": 1.3543033905063285e-07,
+      "loss": 0.2764,
+      "step": 13348
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.093916720033543,
+      "learning_rate": 1.3517020055651197e-07,
+      "loss": 0.321,
+      "step": 13349
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.744967101918294,
+      "learning_rate": 1.349103087199677e-07,
+      "loss": 0.3234,
+      "step": 13350
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 51.2916187464552,
+      "learning_rate": 1.3465066355417665e-07,
+      "loss": 0.3171,
+      "step": 13351
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.489726107875927,
+      "learning_rate": 1.3439126507230227e-07,
+      "loss": 0.2813,
+      "step": 13352
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.055405118815473,
+      "learning_rate": 1.3413211328749843e-07,
+      "loss": 0.3507,
+      "step": 13353
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.899379775088493,
+      "learning_rate": 1.3387320821290296e-07,
+      "loss": 0.3299,
+      "step": 13354
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.253364018907684,
+      "learning_rate": 1.336145498616448e-07,
+      "loss": 0.3147,
+      "step": 13355
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.381889820813872,
+      "learning_rate": 1.333561382468368e-07,
+      "loss": 0.2976,
+      "step": 13356
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.951320103717082,
+      "learning_rate": 1.330979733815818e-07,
+      "loss": 0.3512,
+      "step": 13357
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 20.403300849764584,
+      "learning_rate": 1.3284005527896938e-07,
+      "loss": 0.3312,
+      "step": 13358
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.095521257516264,
+      "learning_rate": 1.3258238395207568e-07,
+      "loss": 0.2917,
+      "step": 13359
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.9053786028752633,
+      "learning_rate": 1.323249594139664e-07,
+      "loss": 0.3012,
+      "step": 13360
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.54054327693186,
+      "learning_rate": 1.3206778167769384e-07,
+      "loss": 0.3351,
+      "step": 13361
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.557104726557291,
+      "learning_rate": 1.3181085075629542e-07,
+      "loss": 0.3141,
+      "step": 13362
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28533661027491214,
+      "learning_rate": 1.3155416666280009e-07,
+      "loss": 0.1042,
+      "step": 13363
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.590462201937844,
+      "learning_rate": 1.312977294102208e-07,
+      "loss": 0.3152,
+      "step": 13364
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.957262451493164,
+      "learning_rate": 1.3104153901156048e-07,
+      "loss": 0.3144,
+      "step": 13365
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.089187312404761,
+      "learning_rate": 1.3078559547980928e-07,
+      "loss": 0.3024,
+      "step": 13366
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.00988682845414,
+      "learning_rate": 1.3052989882794132e-07,
+      "loss": 0.3471,
+      "step": 13367
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.23774632758326,
+      "learning_rate": 1.302744490689245e-07,
+      "loss": 0.3117,
+      "step": 13368
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.549105138460088,
+      "learning_rate": 1.3001924621570738e-07,
+      "loss": 0.3271,
+      "step": 13369
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1106243974020935,
+      "learning_rate": 1.297642902812324e-07,
+      "loss": 0.3025,
+      "step": 13370
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.626727489154577,
+      "learning_rate": 1.2950958127842417e-07,
+      "loss": 0.3179,
+      "step": 13371
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.8839495765332614,
+      "learning_rate": 1.2925511922019795e-07,
+      "loss": 0.3046,
+      "step": 13372
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0402207697598223,
+      "learning_rate": 1.290009041194551e-07,
+      "loss": 0.2873,
+      "step": 13373
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4820591420561904,
+      "learning_rate": 1.2874693598908472e-07,
+      "loss": 0.3298,
+      "step": 13374
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.454509326526886,
+      "learning_rate": 1.2849321484196485e-07,
+      "loss": 0.2899,
+      "step": 13375
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.596119799250261,
+      "learning_rate": 1.2823974069095802e-07,
+      "loss": 0.3212,
+      "step": 13376
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.066875138463878,
+      "learning_rate": 1.2798651354891666e-07,
+      "loss": 0.313,
+      "step": 13377
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.610202073667747,
+      "learning_rate": 1.2773353342868e-07,
+      "loss": 0.3067,
+      "step": 13378
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.038383053026297,
+      "learning_rate": 1.2748080034307496e-07,
+      "loss": 0.3261,
+      "step": 13379
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.118630471907385,
+      "learning_rate": 1.2722831430491522e-07,
+      "loss": 0.2931,
+      "step": 13380
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 8.476326282987467,
+      "learning_rate": 1.2697607532700273e-07,
+      "loss": 0.3311,
+      "step": 13381
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.108567945614358,
+      "learning_rate": 1.2672408342212562e-07,
+      "loss": 0.3112,
+      "step": 13382
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.135804536522592,
+      "learning_rate": 1.2647233860306197e-07,
+      "loss": 0.3122,
+      "step": 13383
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 26.556035164980674,
+      "learning_rate": 1.2622084088257435e-07,
+      "loss": 0.3082,
+      "step": 13384
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8140221993796564,
+      "learning_rate": 1.2596959027341483e-07,
+      "loss": 0.3312,
+      "step": 13385
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2782500130055348,
+      "learning_rate": 1.257185867883226e-07,
+      "loss": 0.3425,
+      "step": 13386
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.676553096378092,
+      "learning_rate": 1.254678304400242e-07,
+      "loss": 0.318,
+      "step": 13387
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.1463814189845625,
+      "learning_rate": 1.2521732124123332e-07,
+      "loss": 0.3249,
+      "step": 13388
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.012751252679478,
+      "learning_rate": 1.249670592046509e-07,
+      "loss": 0.3092,
+      "step": 13389
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3591284289464696,
+      "learning_rate": 1.2471704434296572e-07,
+      "loss": 0.325,
+      "step": 13390
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.515837755227504,
+      "learning_rate": 1.2446727666885482e-07,
+      "loss": 0.3135,
+      "step": 13391
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0686881842632103,
+      "learning_rate": 1.24217756194982e-07,
+      "loss": 0.2803,
+      "step": 13392
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2878946101630517,
+      "learning_rate": 1.2396848293399765e-07,
+      "loss": 0.1101,
+      "step": 13393
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.897151743364789,
+      "learning_rate": 1.237194568985417e-07,
+      "loss": 0.3001,
+      "step": 13394
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 13.501642685394923,
+      "learning_rate": 1.2347067810123847e-07,
+      "loss": 0.2903,
+      "step": 13395
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.92395828336602,
+      "learning_rate": 1.2322214655470344e-07,
+      "loss": 0.3174,
+      "step": 13396
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.422298399226547,
+      "learning_rate": 1.229738622715365e-07,
+      "loss": 0.3163,
+      "step": 13397
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1095249403218337,
+      "learning_rate": 1.2272582526432707e-07,
+      "loss": 0.3005,
+      "step": 13398
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.615791783496316,
+      "learning_rate": 1.2247803554565062e-07,
+      "loss": 0.3272,
+      "step": 13399
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.999347919807207,
+      "learning_rate": 1.2223049312807045e-07,
+      "loss": 0.3217,
+      "step": 13400
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.7438739399478256,
+      "learning_rate": 1.219831980241387e-07,
+      "loss": 0.3215,
+      "step": 13401
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.315269984534887,
+      "learning_rate": 1.2173615024639152e-07,
+      "loss": 0.3179,
+      "step": 13402
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.86252346100933,
+      "learning_rate": 1.2148934980735772e-07,
+      "loss": 0.3009,
+      "step": 13403
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.859302957051167,
+      "learning_rate": 1.2124279671954787e-07,
+      "loss": 0.3122,
+      "step": 13404
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.030451679891118,
+      "learning_rate": 1.2099649099546474e-07,
+      "loss": 0.3587,
+      "step": 13405
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.62142680915101,
+      "learning_rate": 1.2075043264759501e-07,
+      "loss": 0.2944,
+      "step": 13406
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.692386242708826,
+      "learning_rate": 1.2050462168841592e-07,
+      "loss": 0.3103,
+      "step": 13407
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.839270659285587,
+      "learning_rate": 1.2025905813038917e-07,
+      "loss": 0.3232,
+      "step": 13408
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.597027627547658,
+      "learning_rate": 1.2001374198596704e-07,
+      "loss": 0.3384,
+      "step": 13409
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2902189387305207,
+      "learning_rate": 1.1976867326758623e-07,
+      "loss": 0.114,
+      "step": 13410
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 34.19577718068576,
+      "learning_rate": 1.1952385198767236e-07,
+      "loss": 0.3105,
+      "step": 13411
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8998742425040813,
+      "learning_rate": 1.192792781586388e-07,
+      "loss": 0.2996,
+      "step": 13412
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.861191425536502,
+      "learning_rate": 1.190349517928857e-07,
+      "loss": 0.3043,
+      "step": 13413
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.733406987508299,
+      "learning_rate": 1.18790872902802e-07,
+      "loss": 0.3117,
+      "step": 13414
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.837303123680345,
+      "learning_rate": 1.1854704150076057e-07,
+      "loss": 0.346,
+      "step": 13415
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.703905102750715,
+      "learning_rate": 1.1830345759912709e-07,
+      "loss": 0.3017,
+      "step": 13416
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.6404943344993446,
+      "learning_rate": 1.1806012121025002e-07,
+      "loss": 0.2928,
+      "step": 13417
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.9557701724867687,
+      "learning_rate": 1.178170323464678e-07,
+      "loss": 0.3168,
+      "step": 13418
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.2076992843651455,
+      "learning_rate": 1.1757419102010504e-07,
+      "loss": 0.3098,
+      "step": 13419
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.4002991058918033,
+      "learning_rate": 1.1733159724347465e-07,
+      "loss": 0.2781,
+      "step": 13420
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.5887498608060024,
+      "learning_rate": 1.170892510288768e-07,
+      "loss": 0.3152,
+      "step": 13421
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.938947887170542,
+      "learning_rate": 1.1684715238859946e-07,
+      "loss": 0.3177,
+      "step": 13422
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28331540311520625,
+      "learning_rate": 1.1660530133491554e-07,
+      "loss": 0.1039,
+      "step": 13423
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2453808771367276,
+      "learning_rate": 1.1636369788008973e-07,
+      "loss": 0.3171,
+      "step": 13424
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.037586434772675,
+      "learning_rate": 1.1612234203637052e-07,
+      "loss": 0.3316,
+      "step": 13425
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.2482627498295904,
+      "learning_rate": 1.1588123381599536e-07,
+      "loss": 0.3282,
+      "step": 13426
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.698009021804729,
+      "learning_rate": 1.1564037323119004e-07,
+      "loss": 0.3501,
+      "step": 13427
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.987305329224217,
+      "learning_rate": 1.1539976029416422e-07,
+      "loss": 0.2836,
+      "step": 13428
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.28966031025728795,
+      "learning_rate": 1.1515939501712093e-07,
+      "loss": 0.1129,
+      "step": 13429
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.224207060002517,
+      "learning_rate": 1.1491927741224373e-07,
+      "loss": 0.2911,
+      "step": 13430
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.371641581322296,
+      "learning_rate": 1.1467940749171014e-07,
+      "loss": 0.2933,
+      "step": 13431
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.722564535513696,
+      "learning_rate": 1.1443978526768041e-07,
+      "loss": 0.342,
+      "step": 13432
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.569669181247247,
+      "learning_rate": 1.1420041075230426e-07,
+      "loss": 0.2662,
+      "step": 13433
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.832840165722866,
+      "learning_rate": 1.1396128395771866e-07,
+      "loss": 0.2939,
+      "step": 13434
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.467972081144105,
+      "learning_rate": 1.1372240489604725e-07,
+      "loss": 0.2961,
+      "step": 13435
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.3853831737223152,
+      "learning_rate": 1.1348377357940255e-07,
+      "loss": 0.2931,
+      "step": 13436
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0436971769203955,
+      "learning_rate": 1.1324539001988322e-07,
+      "loss": 0.3097,
+      "step": 13437
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.734263376218562,
+      "learning_rate": 1.1300725422957626e-07,
+      "loss": 0.3151,
+      "step": 13438
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.730079637122186,
+      "learning_rate": 1.1276936622055479e-07,
+      "loss": 0.344,
+      "step": 13439
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.314781268703587,
+      "learning_rate": 1.1253172600488083e-07,
+      "loss": 0.2965,
+      "step": 13440
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.8148408966062206,
+      "learning_rate": 1.1229433359460362e-07,
+      "loss": 0.2888,
+      "step": 13441
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.6545578272853607,
+      "learning_rate": 1.1205718900175966e-07,
+      "loss": 0.2781,
+      "step": 13442
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.068000108744821,
+      "learning_rate": 1.1182029223837098e-07,
+      "loss": 0.3203,
+      "step": 13443
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.632402612854646,
+      "learning_rate": 1.1158364331645077e-07,
+      "loss": 0.2768,
+      "step": 13444
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 5.5587179978598575,
+      "learning_rate": 1.113472422479961e-07,
+      "loss": 0.2937,
+      "step": 13445
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.947910512719906,
+      "learning_rate": 1.1111108904499458e-07,
+      "loss": 0.3214,
+      "step": 13446
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.8786814237203875,
+      "learning_rate": 1.108751837194183e-07,
+      "loss": 0.3075,
+      "step": 13447
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 16.03050632973342,
+      "learning_rate": 1.1063952628322883e-07,
+      "loss": 0.3079,
+      "step": 13448
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.006654732464064,
+      "learning_rate": 1.1040411674837493e-07,
+      "loss": 0.2931,
+      "step": 13449
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.6636317047831,
+      "learning_rate": 1.1016895512679148e-07,
+      "loss": 0.3219,
+      "step": 13450
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.700978328575264,
+      "learning_rate": 1.0993404143040232e-07,
+      "loss": 0.3335,
+      "step": 13451
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.701118034983995,
+      "learning_rate": 1.0969937567111788e-07,
+      "loss": 0.2931,
+      "step": 13452
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.06113128444149,
+      "learning_rate": 1.0946495786083589e-07,
+      "loss": 0.3235,
+      "step": 13453
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30056459963657745,
+      "learning_rate": 1.0923078801144238e-07,
+      "loss": 0.1164,
+      "step": 13454
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.385500464723265,
+      "learning_rate": 1.0899686613480953e-07,
+      "loss": 0.2803,
+      "step": 13455
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 8.495594895562592,
+      "learning_rate": 1.0876319224279896e-07,
+      "loss": 0.303,
+      "step": 13456
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.402836689714466,
+      "learning_rate": 1.0852976634725786e-07,
+      "loss": 0.3311,
+      "step": 13457
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3815147592737977,
+      "learning_rate": 1.0829658846002066e-07,
+      "loss": 0.3381,
+      "step": 13458
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.937676538305121,
+      "learning_rate": 1.0806365859291124e-07,
+      "loss": 0.3125,
+      "step": 13459
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.252268530891597,
+      "learning_rate": 1.0783097675773846e-07,
+      "loss": 0.3214,
+      "step": 13460
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 13.689246276384532,
+      "learning_rate": 1.0759854296630068e-07,
+      "loss": 0.3022,
+      "step": 13461
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.680734248206823,
+      "learning_rate": 1.0736635723038291e-07,
+      "loss": 0.2844,
+      "step": 13462
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.918340998218797,
+      "learning_rate": 1.071344195617563e-07,
+      "loss": 0.3359,
+      "step": 13463
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8702196203898693,
+      "learning_rate": 1.0690272997218254e-07,
+      "loss": 0.3095,
+      "step": 13464
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9767250901297877,
+      "learning_rate": 1.0667128847340669e-07,
+      "loss": 0.295,
+      "step": 13465
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.79485605468046,
+      "learning_rate": 1.064400950771649e-07,
+      "loss": 0.3079,
+      "step": 13466
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.297429165179451,
+      "learning_rate": 1.0620914979517782e-07,
+      "loss": 0.3061,
+      "step": 13467
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0367289500827583,
+      "learning_rate": 1.0597845263915608e-07,
+      "loss": 0.2875,
+      "step": 13468
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.098951892406894,
+      "learning_rate": 1.057480036207964e-07,
+      "loss": 0.3231,
+      "step": 13469
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.982652815732765,
+      "learning_rate": 1.0551780275178281e-07,
+      "loss": 0.3159,
+      "step": 13470
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.550352044896055,
+      "learning_rate": 1.0528785004378595e-07,
+      "loss": 0.3269,
+      "step": 13471
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7251490940953476,
+      "learning_rate": 1.0505814550846705e-07,
+      "loss": 0.2882,
+      "step": 13472
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.6994163722826405,
+      "learning_rate": 1.0482868915747124e-07,
+      "loss": 0.2979,
+      "step": 13473
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.756624450498723,
+      "learning_rate": 1.0459948100243255e-07,
+      "loss": 0.3215,
+      "step": 13474
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9947827418342716,
+      "learning_rate": 1.0437052105497225e-07,
+      "loss": 0.297,
+      "step": 13475
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.910785503890256,
+      "learning_rate": 1.0414180932669992e-07,
+      "loss": 0.3145,
+      "step": 13476
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.050111410716679,
+      "learning_rate": 1.0391334582921131e-07,
+      "loss": 0.2909,
+      "step": 13477
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9213140598589993,
+      "learning_rate": 1.0368513057408935e-07,
+      "loss": 0.3236,
+      "step": 13478
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3162960471045632,
+      "learning_rate": 1.0345716357290592e-07,
+      "loss": 0.1139,
+      "step": 13479
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.688832518553303,
+      "learning_rate": 1.0322944483721897e-07,
+      "loss": 0.2883,
+      "step": 13480
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8629221777866842,
+      "learning_rate": 1.0300197437857485e-07,
+      "loss": 0.299,
+      "step": 13481
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3064944339044744,
+      "learning_rate": 1.0277475220850652e-07,
+      "loss": 0.2936,
+      "step": 13482
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9443800284552744,
+      "learning_rate": 1.0254777833853424e-07,
+      "loss": 0.2825,
+      "step": 13483
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.3018516630629104,
+      "learning_rate": 1.0232105278016658e-07,
+      "loss": 0.3368,
+      "step": 13484
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 30.025945504343014,
+      "learning_rate": 1.0209457554489932e-07,
+      "loss": 0.3139,
+      "step": 13485
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.690538127103168,
+      "learning_rate": 1.0186834664421496e-07,
+      "loss": 0.3098,
+      "step": 13486
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.51002566415154,
+      "learning_rate": 1.0164236608958322e-07,
+      "loss": 0.2953,
+      "step": 13487
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5150980623700048,
+      "learning_rate": 1.014166338924627e-07,
+      "loss": 0.3112,
+      "step": 13488
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.3647906269816295,
+      "learning_rate": 1.0119115006429814e-07,
+      "loss": 0.3235,
+      "step": 13489
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.381376723463217,
+      "learning_rate": 1.009659146165226e-07,
+      "loss": 0.3127,
+      "step": 13490
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.516823210330767,
+      "learning_rate": 1.007409275605542e-07,
+      "loss": 0.3404,
+      "step": 13491
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.818135631660708,
+      "learning_rate": 1.0051618890780324e-07,
+      "loss": 0.2902,
+      "step": 13492
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7734796834675257,
+      "learning_rate": 1.002916986696617e-07,
+      "loss": 0.3118,
+      "step": 13493
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.160104575724657,
+      "learning_rate": 1.0006745685751329e-07,
+      "loss": 0.3205,
+      "step": 13494
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.929283952969882,
+      "learning_rate": 9.984346348272721e-08,
+      "loss": 0.327,
+      "step": 13495
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0857478790472754,
+      "learning_rate": 9.961971855665998e-08,
+      "loss": 0.2955,
+      "step": 13496
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.168273955627125,
+      "learning_rate": 9.939622209065691e-08,
+      "loss": 0.2945,
+      "step": 13497
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.146188323492624,
+      "learning_rate": 9.917297409604898e-08,
+      "loss": 0.2963,
+      "step": 13498
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.188732184235971,
+      "learning_rate": 9.894997458415545e-08,
+      "loss": 0.3375,
+      "step": 13499
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5011010348522986,
+      "learning_rate": 9.872722356628284e-08,
+      "loss": 0.3034,
+      "step": 13500
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3205625857416658,
+      "learning_rate": 9.850472105372544e-08,
+      "loss": 0.0965,
+      "step": 13501
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.860307201434692,
+      "learning_rate": 9.828246705776478e-08,
+      "loss": 0.3029,
+      "step": 13502
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 12.27373208327377,
+      "learning_rate": 9.806046158966909e-08,
+      "loss": 0.2834,
+      "step": 13503
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.433559624080251,
+      "learning_rate": 9.783870466069433e-08,
+      "loss": 0.2819,
+      "step": 13504
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.242101544968569,
+      "learning_rate": 9.761719628208488e-08,
+      "loss": 0.3039,
+      "step": 13505
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.28710644977516486,
+      "learning_rate": 9.739593646507061e-08,
+      "loss": 0.0941,
+      "step": 13506
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 15.864587088070056,
+      "learning_rate": 9.717492522087147e-08,
+      "loss": 0.3246,
+      "step": 13507
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.327393939772559,
+      "learning_rate": 9.695416256069235e-08,
+      "loss": 0.3162,
+      "step": 13508
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.130662601488473,
+      "learning_rate": 9.673364849572542e-08,
+      "loss": 0.2873,
+      "step": 13509
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.787440758436856,
+      "learning_rate": 9.651338303715285e-08,
+      "loss": 0.2888,
+      "step": 13510
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.9816878484056275,
+      "learning_rate": 9.629336619614183e-08,
+      "loss": 0.2783,
+      "step": 13511
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.528783312752312,
+      "learning_rate": 9.607359798384785e-08,
+      "loss": 0.278,
+      "step": 13512
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.357444934718946,
+      "learning_rate": 9.58540784114137e-08,
+      "loss": 0.3478,
+      "step": 13513
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.299664139925101,
+      "learning_rate": 9.563480748996879e-08,
+      "loss": 0.2841,
+      "step": 13514
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.576727706083959,
+      "learning_rate": 9.5415785230632e-08,
+      "loss": 0.31,
+      "step": 13515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.298112581168226,
+      "learning_rate": 9.519701164450667e-08,
+      "loss": 0.3125,
+      "step": 13516
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.857695743302299,
+      "learning_rate": 9.497848674268617e-08,
+      "loss": 0.2714,
+      "step": 13517
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.297604094324004,
+      "learning_rate": 9.476021053625051e-08,
+      "loss": 0.3436,
+      "step": 13518
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5039491016998543,
+      "learning_rate": 9.454218303626528e-08,
+      "loss": 0.2872,
+      "step": 13519
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.883450330696581,
+      "learning_rate": 9.432440425378664e-08,
+      "loss": 0.3012,
+      "step": 13520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.728986998079578,
+      "learning_rate": 9.410687419985465e-08,
+      "loss": 0.294,
+      "step": 13521
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.331016766212187,
+      "learning_rate": 9.388959288550048e-08,
+      "loss": 0.2778,
+      "step": 13522
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8947647725798262,
+      "learning_rate": 9.367256032173977e-08,
+      "loss": 0.2907,
+      "step": 13523
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.430974590321036,
+      "learning_rate": 9.345577651957593e-08,
+      "loss": 0.2894,
+      "step": 13524
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7293405943742037,
+      "learning_rate": 9.323924149000186e-08,
+      "loss": 0.2774,
+      "step": 13525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.616728600888937,
+      "learning_rate": 9.30229552439954e-08,
+      "loss": 0.29,
+      "step": 13526
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 14.579588001818,
+      "learning_rate": 9.280691779252338e-08,
+      "loss": 0.3015,
+      "step": 13527
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.4733545782983772,
+      "learning_rate": 9.259112914653812e-08,
+      "loss": 0.3137,
+      "step": 13528
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.5139317573166777,
+      "learning_rate": 9.237558931698199e-08,
+      "loss": 0.3042,
+      "step": 13529
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.3551882001768885,
+      "learning_rate": 9.216029831478235e-08,
+      "loss": 0.2927,
+      "step": 13530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.262588464823303,
+      "learning_rate": 9.194525615085548e-08,
+      "loss": 0.3227,
+      "step": 13531
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.060331922775514,
+      "learning_rate": 9.173046283610376e-08,
+      "loss": 0.3003,
+      "step": 13532
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.801361440193044,
+      "learning_rate": 9.15159183814196e-08,
+      "loss": 0.3123,
+      "step": 13533
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.519670012375085,
+      "learning_rate": 9.130162279767817e-08,
+      "loss": 0.3254,
+      "step": 13534
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 11.38696771480201,
+      "learning_rate": 9.108757609574692e-08,
+      "loss": 0.3299,
+      "step": 13535
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.5251445752478405,
+      "learning_rate": 9.087377828647714e-08,
+      "loss": 0.3184,
+      "step": 13536
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.566543494650932,
+      "learning_rate": 9.066022938070961e-08,
+      "loss": 0.3044,
+      "step": 13537
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.038714018861402,
+      "learning_rate": 9.044692938927125e-08,
+      "loss": 0.3025,
+      "step": 13538
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.453325822223876,
+      "learning_rate": 9.023387832297726e-08,
+      "loss": 0.3172,
+      "step": 13539
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.2468708459509115,
+      "learning_rate": 9.002107619263012e-08,
+      "loss": 0.3204,
+      "step": 13540
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1927185092751658,
+      "learning_rate": 8.980852300901788e-08,
+      "loss": 0.2896,
+      "step": 13541
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9219847831563297,
+      "learning_rate": 8.959621878291913e-08,
+      "loss": 0.288,
+      "step": 13542
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.046824602287951,
+      "learning_rate": 8.938416352509694e-08,
+      "loss": 0.335,
+      "step": 13543
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.556655471761548,
+      "learning_rate": 8.91723572463038e-08,
+      "loss": 0.2923,
+      "step": 13544
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.946260199445824,
+      "learning_rate": 8.896079995727836e-08,
+      "loss": 0.3427,
+      "step": 13545
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.2715431783729407,
+      "learning_rate": 8.874949166874758e-08,
+      "loss": 0.2921,
+      "step": 13546
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7544433289303627,
+      "learning_rate": 8.853843239142345e-08,
+      "loss": 0.3022,
+      "step": 13547
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.225815288044782,
+      "learning_rate": 8.83276221360102e-08,
+      "loss": 0.2891,
+      "step": 13548
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.535856802222379,
+      "learning_rate": 8.811706091319372e-08,
+      "loss": 0.2695,
+      "step": 13549
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.5359021698161595,
+      "learning_rate": 8.790674873365102e-08,
+      "loss": 0.3496,
+      "step": 13550
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.8612662333159475,
+      "learning_rate": 8.769668560804522e-08,
+      "loss": 0.3031,
+      "step": 13551
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.059694303764805,
+      "learning_rate": 8.748687154702673e-08,
+      "loss": 0.3336,
+      "step": 13552
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.203422054688488,
+      "learning_rate": 8.727730656123479e-08,
+      "loss": 0.2855,
+      "step": 13553
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7712077323078494,
+      "learning_rate": 8.706799066129257e-08,
+      "loss": 0.3017,
+      "step": 13554
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.9972130485884736,
+      "learning_rate": 8.685892385781547e-08,
+      "loss": 0.2882,
+      "step": 13555
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.923085053680287,
+      "learning_rate": 8.665010616140168e-08,
+      "loss": 0.3049,
+      "step": 13556
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.1859991859821215,
+      "learning_rate": 8.64415375826394e-08,
+      "loss": 0.325,
+      "step": 13557
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0033626329216054,
+      "learning_rate": 8.623321813210351e-08,
+      "loss": 0.3062,
+      "step": 13558
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.407443918503865,
+      "learning_rate": 8.602514782035609e-08,
+      "loss": 0.3195,
+      "step": 13559
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.0998020455704376,
+      "learning_rate": 8.581732665794651e-08,
+      "loss": 0.3334,
+      "step": 13560
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.6547504856585835,
+      "learning_rate": 8.560975465541299e-08,
+      "loss": 0.3419,
+      "step": 13561
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 10.041242585377576,
+      "learning_rate": 8.540243182327878e-08,
+      "loss": 0.3293,
+      "step": 13562
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8499628825552583,
+      "learning_rate": 8.519535817205605e-08,
+      "loss": 0.2951,
+      "step": 13563
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.827926280069857,
+      "learning_rate": 8.498853371224303e-08,
+      "loss": 0.3286,
+      "step": 13564
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7201429258514107,
+      "learning_rate": 8.478195845432746e-08,
+      "loss": 0.3077,
+      "step": 13565
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.7185023962685446,
+      "learning_rate": 8.457563240878263e-08,
+      "loss": 0.306,
+      "step": 13566
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.54256416762012,
+      "learning_rate": 8.436955558606908e-08,
+      "loss": 0.3182,
+      "step": 13567
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.173127388914217,
+      "learning_rate": 8.416372799663674e-08,
+      "loss": 0.2728,
+      "step": 13568
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6196794831758443,
+      "learning_rate": 8.395814965092008e-08,
+      "loss": 0.334,
+      "step": 13569
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.174769580387658,
+      "learning_rate": 8.375282055934298e-08,
+      "loss": 0.3356,
+      "step": 13570
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.1668572013533005,
+      "learning_rate": 8.354774073231598e-08,
+      "loss": 0.2865,
+      "step": 13571
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.013882114009154,
+      "learning_rate": 8.3342910180238e-08,
+      "loss": 0.3525,
+      "step": 13572
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.358863625223251,
+      "learning_rate": 8.313832891349294e-08,
+      "loss": 0.3603,
+      "step": 13573
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.6452106710072694,
+      "learning_rate": 8.293399694245475e-08,
+      "loss": 0.3285,
+      "step": 13574
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.027438074622674,
+      "learning_rate": 8.272991427748233e-08,
+      "loss": 0.2917,
+      "step": 13575
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.809659279882443,
+      "learning_rate": 8.252608092892411e-08,
+      "loss": 0.2952,
+      "step": 13576
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.441359445807635,
+      "learning_rate": 8.232249690711403e-08,
+      "loss": 0.2887,
+      "step": 13577
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.807661047390232,
+      "learning_rate": 8.211916222237492e-08,
+      "loss": 0.2964,
+      "step": 13578
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.703312684733363,
+      "learning_rate": 8.191607688501691e-08,
+      "loss": 0.2976,
+      "step": 13579
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.6384465321381105,
+      "learning_rate": 8.17132409053345e-08,
+      "loss": 0.3111,
+      "step": 13580
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.581245936934161,
+      "learning_rate": 8.151065429361505e-08,
+      "loss": 0.292,
+      "step": 13581
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.784013104775974,
+      "learning_rate": 8.130831706012753e-08,
+      "loss": 0.3101,
+      "step": 13582
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.612750331361978,
+      "learning_rate": 8.110622921513267e-08,
+      "loss": 0.3206,
+      "step": 13583
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.8976704060365597,
+      "learning_rate": 8.090439076887557e-08,
+      "loss": 0.3059,
+      "step": 13584
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.2994075801536379,
+      "learning_rate": 8.070280173159028e-08,
+      "loss": 0.1154,
+      "step": 13585
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.31075595173428616,
+      "learning_rate": 8.050146211349807e-08,
+      "loss": 0.1144,
+      "step": 13586
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.419801730991874,
+      "learning_rate": 8.030037192480744e-08,
+      "loss": 0.3047,
+      "step": 13587
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.404753047806495,
+      "learning_rate": 8.009953117571356e-08,
+      "loss": 0.3264,
+      "step": 13588
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.384639404097485,
+      "learning_rate": 7.98989398763994e-08,
+      "loss": 0.3338,
+      "step": 13589
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.009756336075089,
+      "learning_rate": 7.969859803703628e-08,
+      "loss": 0.32,
+      "step": 13590
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.101299691548934,
+      "learning_rate": 7.949850566778106e-08,
+      "loss": 0.2951,
+      "step": 13591
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.1757397169271835,
+      "learning_rate": 7.929866277877896e-08,
+      "loss": 0.2859,
+      "step": 13592
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.904073535476174,
+      "learning_rate": 7.909906938016298e-08,
+      "loss": 0.3248,
+      "step": 13593
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7372326997563166,
+      "learning_rate": 7.889972548205282e-08,
+      "loss": 0.3059,
+      "step": 13594
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2064589876225447,
+      "learning_rate": 7.870063109455484e-08,
+      "loss": 0.3426,
+      "step": 13595
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.921846879517999,
+      "learning_rate": 7.850178622776483e-08,
+      "loss": 0.3111,
+      "step": 13596
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9613052557072908,
+      "learning_rate": 7.830319089176309e-08,
+      "loss": 0.3119,
+      "step": 13597
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.948034369953311,
+      "learning_rate": 7.810484509662097e-08,
+      "loss": 0.3125,
+      "step": 13598
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.0630219447621965,
+      "learning_rate": 7.790674885239325e-08,
+      "loss": 0.312,
+      "step": 13599
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.9436931002548485,
+      "learning_rate": 7.770890216912463e-08,
+      "loss": 0.3231,
+      "step": 13600
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.150669960782374,
+      "learning_rate": 7.751130505684602e-08,
+      "loss": 0.3196,
+      "step": 13601
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.179996049921463,
+      "learning_rate": 7.731395752557657e-08,
+      "loss": 0.3118,
+      "step": 13602
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.183531331078974,
+      "learning_rate": 7.711685958532222e-08,
+      "loss": 0.3358,
+      "step": 13603
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.31480907079769227,
+      "learning_rate": 7.692001124607496e-08,
+      "loss": 0.1143,
+      "step": 13604
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.108069848089618,
+      "learning_rate": 7.672341251781679e-08,
+      "loss": 0.3181,
+      "step": 13605
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.50808248965273,
+      "learning_rate": 7.65270634105153e-08,
+      "loss": 0.3028,
+      "step": 13606
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.119906488170155,
+      "learning_rate": 7.633096393412586e-08,
+      "loss": 0.3349,
+      "step": 13607
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.699171366530543,
+      "learning_rate": 7.613511409859109e-08,
+      "loss": 0.2948,
+      "step": 13608
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.46050680820258,
+      "learning_rate": 7.593951391384136e-08,
+      "loss": 0.3171,
+      "step": 13609
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.433571844095798,
+      "learning_rate": 7.574416338979263e-08,
+      "loss": 0.3187,
+      "step": 13610
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.682004249183376,
+      "learning_rate": 7.554906253635197e-08,
+      "loss": 0.3101,
+      "step": 13611
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.116680893453472,
+      "learning_rate": 7.535421136340925e-08,
+      "loss": 0.3146,
+      "step": 13612
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.132219686400695,
+      "learning_rate": 7.515960988084492e-08,
+      "loss": 0.2873,
+      "step": 13613
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6760051568096284,
+      "learning_rate": 7.49652580985255e-08,
+      "loss": 0.2775,
+      "step": 13614
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3245883368480365,
+      "learning_rate": 7.477115602630536e-08,
+      "loss": 0.289,
+      "step": 13615
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2698033693034138,
+      "learning_rate": 7.45773036740255e-08,
+      "loss": 0.0986,
+      "step": 13616
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.04070157674387,
+      "learning_rate": 7.438370105151472e-08,
+      "loss": 0.311,
+      "step": 13617
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.175040178218148,
+      "learning_rate": 7.419034816858905e-08,
+      "loss": 0.315,
+      "step": 13618
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.9669453932226753,
+      "learning_rate": 7.399724503505178e-08,
+      "loss": 0.3143,
+      "step": 13619
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.482614584293628,
+      "learning_rate": 7.380439166069398e-08,
+      "loss": 0.3147,
+      "step": 13620
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3562653989698905,
+      "learning_rate": 7.36117880552939e-08,
+      "loss": 0.303,
+      "step": 13621
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4390886685890223,
+      "learning_rate": 7.341943422861653e-08,
+      "loss": 0.31,
+      "step": 13622
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.730685424535696,
+      "learning_rate": 7.322733019041461e-08,
+      "loss": 0.3121,
+      "step": 13623
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.161334316911656,
+      "learning_rate": 7.303547595042926e-08,
+      "loss": 0.2858,
+      "step": 13624
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.962967424976585,
+      "learning_rate": 7.284387151838601e-08,
+      "loss": 0.2949,
+      "step": 13625
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.56636195945276,
+      "learning_rate": 7.265251690400155e-08,
+      "loss": 0.3013,
+      "step": 13626
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.042129635599273,
+      "learning_rate": 7.246141211697644e-08,
+      "loss": 0.2934,
+      "step": 13627
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.732532202929178,
+      "learning_rate": 7.227055716700126e-08,
+      "loss": 0.3114,
+      "step": 13628
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.856635365338065,
+      "learning_rate": 7.207995206375273e-08,
+      "loss": 0.316,
+      "step": 13629
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.642674244327672,
+      "learning_rate": 7.188959681689367e-08,
+      "loss": 0.3104,
+      "step": 13630
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.5061034261077575,
+      "learning_rate": 7.169949143607691e-08,
+      "loss": 0.3036,
+      "step": 13631
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6164421272433698,
+      "learning_rate": 7.150963593094029e-08,
+      "loss": 0.3073,
+      "step": 13632
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.676331013713108,
+      "learning_rate": 7.132003031111057e-08,
+      "loss": 0.2936,
+      "step": 13633
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.767574202426953,
+      "learning_rate": 7.113067458620115e-08,
+      "loss": 0.3116,
+      "step": 13634
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.705603608360114,
+      "learning_rate": 7.094156876581216e-08,
+      "loss": 0.3216,
+      "step": 13635
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.88533570697224,
+      "learning_rate": 7.0752712859532e-08,
+      "loss": 0.2994,
+      "step": 13636
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3619959127801327,
+      "learning_rate": 7.056410687693638e-08,
+      "loss": 0.3336,
+      "step": 13637
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.817770441539374,
+      "learning_rate": 7.037575082758763e-08,
+      "loss": 0.372,
+      "step": 13638
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.941934740887071,
+      "learning_rate": 7.01876447210359e-08,
+      "loss": 0.3152,
+      "step": 13639
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3080236655928504,
+      "learning_rate": 6.99997885668191e-08,
+      "loss": 0.1157,
+      "step": 13640
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.468710389609204,
+      "learning_rate": 6.981218237446075e-08,
+      "loss": 0.3368,
+      "step": 13641
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.803035198470223,
+      "learning_rate": 6.962482615347487e-08,
+      "loss": 0.302,
+      "step": 13642
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6554519368330203,
+      "learning_rate": 6.943771991335834e-08,
+      "loss": 0.291,
+      "step": 13643
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.620658798096081,
+      "learning_rate": 6.925086366359967e-08,
+      "loss": 0.2967,
+      "step": 13644
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.615055312603303,
+      "learning_rate": 6.90642574136724e-08,
+      "loss": 0.3059,
+      "step": 13645
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2792651885885082,
+      "learning_rate": 6.887790117303783e-08,
+      "loss": 0.1083,
+      "step": 13646
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.276645356278094,
+      "learning_rate": 6.869179495114453e-08,
+      "loss": 0.2843,
+      "step": 13647
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.433370838986629,
+      "learning_rate": 6.850593875742827e-08,
+      "loss": 0.2851,
+      "step": 13648
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 19.901173150066896,
+      "learning_rate": 6.832033260131321e-08,
+      "loss": 0.3662,
+      "step": 13649
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.435886129612901,
+      "learning_rate": 6.813497649220958e-08,
+      "loss": 0.2872,
+      "step": 13650
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.953425499982533,
+      "learning_rate": 6.794987043951484e-08,
+      "loss": 0.31,
+      "step": 13651
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.974784091327411,
+      "learning_rate": 6.776501445261486e-08,
+      "loss": 0.3256,
+      "step": 13652
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.061034725146379,
+      "learning_rate": 6.75804085408821e-08,
+      "loss": 0.3241,
+      "step": 13653
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.2258572509611705,
+      "learning_rate": 6.739605271367633e-08,
+      "loss": 0.3164,
+      "step": 13654
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7356524472702692,
+      "learning_rate": 6.721194698034505e-08,
+      "loss": 0.2908,
+      "step": 13655
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.9420790460075885,
+      "learning_rate": 6.702809135022193e-08,
+      "loss": 0.3193,
+      "step": 13656
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.54994361617675,
+      "learning_rate": 6.684448583263059e-08,
+      "loss": 0.2992,
+      "step": 13657
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1920414983026175,
+      "learning_rate": 6.66611304368786e-08,
+      "loss": 0.3199,
+      "step": 13658
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2052303151242145,
+      "learning_rate": 6.647802517226353e-08,
+      "loss": 0.3377,
+      "step": 13659
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3021425185517535,
+      "learning_rate": 6.629517004806907e-08,
+      "loss": 0.111,
+      "step": 13660
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.464330524309898,
+      "learning_rate": 6.611256507356555e-08,
+      "loss": 0.2837,
+      "step": 13661
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.494847960074281,
+      "learning_rate": 6.593021025801227e-08,
+      "loss": 0.3408,
+      "step": 13662
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.20405159725256,
+      "learning_rate": 6.574810561065458e-08,
+      "loss": 0.3365,
+      "step": 13663
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.289960029103102,
+      "learning_rate": 6.556625114072623e-08,
+      "loss": 0.2985,
+      "step": 13664
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1976274812393624,
+      "learning_rate": 6.538464685744706e-08,
+      "loss": 0.3073,
+      "step": 13665
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6239790502039595,
+      "learning_rate": 6.520329277002469e-08,
+      "loss": 0.2864,
+      "step": 13666
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.595932006627122,
+      "learning_rate": 6.502218888765455e-08,
+      "loss": 0.3012,
+      "step": 13667
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 7.746686157480269,
+      "learning_rate": 6.484133521951874e-08,
+      "loss": 0.3485,
+      "step": 13668
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.8147180281513355,
+      "learning_rate": 6.466073177478715e-08,
+      "loss": 0.286,
+      "step": 13669
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.274496189548122,
+      "learning_rate": 6.448037856261691e-08,
+      "loss": 0.2961,
+      "step": 13670
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.440165976655025,
+      "learning_rate": 6.430027559215179e-08,
+      "loss": 0.2686,
+      "step": 13671
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.296767967203582,
+      "learning_rate": 6.412042287252396e-08,
+      "loss": 0.2935,
+      "step": 13672
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.338846254801119,
+      "learning_rate": 6.394082041285165e-08,
+      "loss": 0.3154,
+      "step": 13673
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.133358408211945,
+      "learning_rate": 6.37614682222426e-08,
+      "loss": 0.2955,
+      "step": 13674
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 12.02781019483211,
+      "learning_rate": 6.358236630978843e-08,
+      "loss": 0.3102,
+      "step": 13675
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.792353917707368,
+      "learning_rate": 6.340351468457129e-08,
+      "loss": 0.2461,
+      "step": 13676
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5852343262676807,
+      "learning_rate": 6.322491335565894e-08,
+      "loss": 0.2958,
+      "step": 13677
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.132674352198881,
+      "learning_rate": 6.304656233210694e-08,
+      "loss": 0.3053,
+      "step": 13678
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.9205528687811,
+      "learning_rate": 6.286846162295746e-08,
+      "loss": 0.2984,
+      "step": 13679
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.325840227820054,
+      "learning_rate": 6.269061123724163e-08,
+      "loss": 0.3285,
+      "step": 13680
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3186278433370098,
+      "learning_rate": 6.251301118397612e-08,
+      "loss": 0.3446,
+      "step": 13681
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.2097291540255037,
+      "learning_rate": 6.233566147216597e-08,
+      "loss": 0.3324,
+      "step": 13682
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.722594315121973,
+      "learning_rate": 6.215856211080396e-08,
+      "loss": 0.3216,
+      "step": 13683
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 15.221627832849434,
+      "learning_rate": 6.198171310886735e-08,
+      "loss": 0.3133,
+      "step": 13684
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.03914888166661,
+      "learning_rate": 6.180511447532512e-08,
+      "loss": 0.3051,
+      "step": 13685
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.766227803097198,
+      "learning_rate": 6.162876621912895e-08,
+      "loss": 0.3443,
+      "step": 13686
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6892909294458067,
+      "learning_rate": 6.145266834922225e-08,
+      "loss": 0.3096,
+      "step": 13687
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.920418325039952,
+      "learning_rate": 6.12768208745318e-08,
+      "loss": 0.3026,
+      "step": 13688
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.9680123366763507,
+      "learning_rate": 6.110122380397488e-08,
+      "loss": 0.3013,
+      "step": 13689
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 17.714642998907124,
+      "learning_rate": 6.092587714645326e-08,
+      "loss": 0.3034,
+      "step": 13690
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.465564781116877,
+      "learning_rate": 6.075078091085873e-08,
+      "loss": 0.3117,
+      "step": 13691
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 8.261452140112947,
+      "learning_rate": 6.057593510606863e-08,
+      "loss": 0.3079,
+      "step": 13692
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.896995190556451,
+      "learning_rate": 6.040133974094753e-08,
+      "loss": 0.2924,
+      "step": 13693
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.6566474248288445,
+      "learning_rate": 6.02269948243478e-08,
+      "loss": 0.3141,
+      "step": 13694
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.994353440905585,
+      "learning_rate": 6.005290036510958e-08,
+      "loss": 0.3245,
+      "step": 13695
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.893479195787794,
+      "learning_rate": 5.987905637206026e-08,
+      "loss": 0.3163,
+      "step": 13696
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.467595734361199,
+      "learning_rate": 5.97054628540128e-08,
+      "loss": 0.3107,
+      "step": 13697
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.29653051705965405,
+      "learning_rate": 5.953211981977014e-08,
+      "loss": 0.1096,
+      "step": 13698
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.253995775798903,
+      "learning_rate": 5.935902727811971e-08,
+      "loss": 0.3091,
+      "step": 13699
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.975710303541065,
+      "learning_rate": 5.918618523783948e-08,
+      "loss": 0.3183,
+      "step": 13700
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.042889617844205,
+      "learning_rate": 5.901359370769133e-08,
+      "loss": 0.3055,
+      "step": 13701
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.998221148211926,
+      "learning_rate": 5.8841252696426596e-08,
+      "loss": 0.3071,
+      "step": 13702
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7579828411711675,
+      "learning_rate": 5.866916221278385e-08,
+      "loss": 0.29,
+      "step": 13703
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.949048442495093,
+      "learning_rate": 5.849732226548777e-08,
+      "loss": 0.2798,
+      "step": 13704
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.551968639861352,
+      "learning_rate": 5.832573286325138e-08,
+      "loss": 0.3519,
+      "step": 13705
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.4826094961874072,
+      "learning_rate": 5.815439401477385e-08,
+      "loss": 0.3041,
+      "step": 13706
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.619996683213039,
+      "learning_rate": 5.7983305728743776e-08,
+      "loss": 0.3349,
+      "step": 13707
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.697532975687461,
+      "learning_rate": 5.781246801383478e-08,
+      "loss": 0.3243,
+      "step": 13708
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.284844369347476,
+      "learning_rate": 5.764188087870937e-08,
+      "loss": 0.3152,
+      "step": 13709
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.7991695935732452,
+      "learning_rate": 5.747154433201563e-08,
+      "loss": 0.3165,
+      "step": 13710
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 6.320309579991056,
+      "learning_rate": 5.73014583823911e-08,
+      "loss": 0.3188,
+      "step": 13711
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 9.371153131725073,
+      "learning_rate": 5.713162303845887e-08,
+      "loss": 0.2757,
+      "step": 13712
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.933459957244122,
+      "learning_rate": 5.6962038308830406e-08,
+      "loss": 0.2874,
+      "step": 13713
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.263628283824263,
+      "learning_rate": 5.679270420210381e-08,
+      "loss": 0.3131,
+      "step": 13714
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.337241868018285,
+      "learning_rate": 5.6623620726864445e-08,
+      "loss": 0.3011,
+      "step": 13715
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 10.555854585676009,
+      "learning_rate": 5.645478789168546e-08,
+      "loss": 0.2955,
+      "step": 13716
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.623035550578989,
+      "learning_rate": 5.628620570512666e-08,
+      "loss": 0.3313,
+      "step": 13717
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.064198020976704,
+      "learning_rate": 5.611787417573678e-08,
+      "loss": 0.3258,
+      "step": 13718
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.615332979397881,
+      "learning_rate": 5.594979331204842e-08,
+      "loss": 0.2791,
+      "step": 13719
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.284969610811751,
+      "learning_rate": 5.5781963122586434e-08,
+      "loss": 0.2543,
+      "step": 13720
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 5.59427561613101,
+      "learning_rate": 5.5614383615857913e-08,
+      "loss": 0.3486,
+      "step": 13721
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 21.9356291532451,
+      "learning_rate": 5.544705480035995e-08,
+      "loss": 0.2929,
+      "step": 13722
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.431785428646337,
+      "learning_rate": 5.527997668457741e-08,
+      "loss": 0.3384,
+      "step": 13723
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2837723275475145,
+      "learning_rate": 5.511314927698075e-08,
+      "loss": 0.1067,
+      "step": 13724
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3376448842441286,
+      "learning_rate": 5.4946572586028756e-08,
+      "loss": 0.3362,
+      "step": 13725
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.818588717850064,
+      "learning_rate": 5.47802466201669e-08,
+      "loss": 0.2896,
+      "step": 13726
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8061444055564557,
+      "learning_rate": 5.4614171387828984e-08,
+      "loss": 0.3121,
+      "step": 13727
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.62626020468793,
+      "learning_rate": 5.444834689743439e-08,
+      "loss": 0.2976,
+      "step": 13728
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.8209887572251646,
+      "learning_rate": 5.428277315739139e-08,
+      "loss": 0.3244,
+      "step": 13729
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.939677427425365,
+      "learning_rate": 5.411745017609493e-08,
+      "loss": 0.3334,
+      "step": 13730
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8964702206583572,
+      "learning_rate": 5.395237796192776e-08,
+      "loss": 0.3313,
+      "step": 13731
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2707555044194732,
+      "learning_rate": 5.378755652325818e-08,
+      "loss": 0.1009,
+      "step": 13732
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.588067416090744,
+      "learning_rate": 5.3622985868443943e-08,
+      "loss": 0.312,
+      "step": 13733
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.743010643742568,
+      "learning_rate": 5.345866600582838e-08,
+      "loss": 0.3035,
+      "step": 13734
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.389246029819833,
+      "learning_rate": 5.329459694374373e-08,
+      "loss": 0.3062,
+      "step": 13735
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.881151457098457,
+      "learning_rate": 5.313077869050831e-08,
+      "loss": 0.3153,
+      "step": 13736
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.762961895410733,
+      "learning_rate": 5.2967211254428274e-08,
+      "loss": 0.3095,
+      "step": 13737
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.4432006956626475,
+      "learning_rate": 5.2803894643795875e-08,
+      "loss": 0.3258,
+      "step": 13738
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.403918345454993,
+      "learning_rate": 5.264082886689281e-08,
+      "loss": 0.2978,
+      "step": 13739
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.63351520317241,
+      "learning_rate": 5.247801393198637e-08,
+      "loss": 0.2887,
+      "step": 13740
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.639091185480821,
+      "learning_rate": 5.231544984733161e-08,
+      "loss": 0.2858,
+      "step": 13741
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.11002251077923,
+      "learning_rate": 5.215313662117139e-08,
+      "loss": 0.3131,
+      "step": 13742
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.048118086629208,
+      "learning_rate": 5.199107426173411e-08,
+      "loss": 0.283,
+      "step": 13743
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.819330865821942,
+      "learning_rate": 5.182926277723821e-08,
+      "loss": 0.3274,
+      "step": 13744
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.85033623664372,
+      "learning_rate": 5.166770217588657e-08,
+      "loss": 0.3151,
+      "step": 13745
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.798034081240878,
+      "learning_rate": 5.1506392465871526e-08,
+      "loss": 0.3123,
+      "step": 13746
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.951645184047853,
+      "learning_rate": 5.134533365537153e-08,
+      "loss": 0.2623,
+      "step": 13747
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5167180131233087,
+      "learning_rate": 5.1184525752552836e-08,
+      "loss": 0.3154,
+      "step": 13748
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.022993695896652,
+      "learning_rate": 5.1023968765568366e-08,
+      "loss": 0.3045,
+      "step": 13749
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2843810480011605,
+      "learning_rate": 5.086366270255882e-08,
+      "loss": 0.0957,
+      "step": 13750
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.537137693853043,
+      "learning_rate": 5.070360757165216e-08,
+      "loss": 0.3295,
+      "step": 13751
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8391745075813044,
+      "learning_rate": 5.05438033809641e-08,
+      "loss": 0.3038,
+      "step": 13752
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.827726077235039,
+      "learning_rate": 5.038425013859649e-08,
+      "loss": 0.278,
+      "step": 13753
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.153457177703631,
+      "learning_rate": 5.022494785263843e-08,
+      "loss": 0.3054,
+      "step": 13754
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2092168243092303,
+      "learning_rate": 5.006589653116789e-08,
+      "loss": 0.32,
+      "step": 13755
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.097699931701748,
+      "learning_rate": 4.9907096182248425e-08,
+      "loss": 0.2988,
+      "step": 13756
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9174104895470596,
+      "learning_rate": 4.974854681393193e-08,
+      "loss": 0.3033,
+      "step": 13757
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4762863904808006,
+      "learning_rate": 4.959024843425753e-08,
+      "loss": 0.3344,
+      "step": 13758
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.214224070852307,
+      "learning_rate": 4.9432201051251016e-08,
+      "loss": 0.3145,
+      "step": 13759
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.634478097782098,
+      "learning_rate": 4.927440467292488e-08,
+      "loss": 0.2455,
+      "step": 13760
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.56968435675422,
+      "learning_rate": 4.9116859307281604e-08,
+      "loss": 0.3268,
+      "step": 13761
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.89936891875479,
+      "learning_rate": 4.895956496230703e-08,
+      "loss": 0.3351,
+      "step": 13762
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.474753787368449,
+      "learning_rate": 4.8802521645978097e-08,
+      "loss": 0.2884,
+      "step": 13763
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9971660482458473,
+      "learning_rate": 4.864572936625567e-08,
+      "loss": 0.3043,
+      "step": 13764
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.105009496125181,
+      "learning_rate": 4.84891881310906e-08,
+      "loss": 0.3134,
+      "step": 13765
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7945518135466867,
+      "learning_rate": 4.833289794841989e-08,
+      "loss": 0.3377,
+      "step": 13766
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.458754228064812,
+      "learning_rate": 4.817685882616663e-08,
+      "loss": 0.2804,
+      "step": 13767
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.581121548241648,
+      "learning_rate": 4.802107077224394e-08,
+      "loss": 0.3103,
+      "step": 13768
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.665370818809862,
+      "learning_rate": 4.786553379454939e-08,
+      "loss": 0.3295,
+      "step": 13769
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4138555000154422,
+      "learning_rate": 4.7710247900969455e-08,
+      "loss": 0.3206,
+      "step": 13770
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.572943066427863,
+      "learning_rate": 4.7555213099377826e-08,
+      "loss": 0.3242,
+      "step": 13771
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.39005403306221,
+      "learning_rate": 4.740042939763434e-08,
+      "loss": 0.3014,
+      "step": 13772
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.894619116293819,
+      "learning_rate": 4.724589680358771e-08,
+      "loss": 0.3058,
+      "step": 13773
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 11.365613357456947,
+      "learning_rate": 4.7091615325072785e-08,
+      "loss": 0.3171,
+      "step": 13774
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.778179828918447,
+      "learning_rate": 4.693758496991108e-08,
+      "loss": 0.3049,
+      "step": 13775
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4448805952220005,
+      "learning_rate": 4.678380574591357e-08,
+      "loss": 0.3136,
+      "step": 13776
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 24.16020389428168,
+      "learning_rate": 4.66302776608768e-08,
+      "loss": 0.2885,
+      "step": 13777
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.011141855395352,
+      "learning_rate": 4.6477000722584544e-08,
+      "loss": 0.253,
+      "step": 13778
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.138468341351215,
+      "learning_rate": 4.632397493880891e-08,
+      "loss": 0.3112,
+      "step": 13779
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5303115266338363,
+      "learning_rate": 4.617120031730815e-08,
+      "loss": 0.3045,
+      "step": 13780
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8793486423705428,
+      "learning_rate": 4.601867686582884e-08,
+      "loss": 0.2931,
+      "step": 13781
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.406812632694951,
+      "learning_rate": 4.586640459210312e-08,
+      "loss": 0.3178,
+      "step": 13782
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7032062719778596,
+      "learning_rate": 4.5714383503853156e-08,
+      "loss": 0.3246,
+      "step": 13783
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.60043668564124,
+      "learning_rate": 4.5562613608785e-08,
+      "loss": 0.2932,
+      "step": 13784
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7060280882601853,
+      "learning_rate": 4.541109491459528e-08,
+      "loss": 0.3119,
+      "step": 13785
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.626949179596629,
+      "learning_rate": 4.5259827428965063e-08,
+      "loss": 0.3004,
+      "step": 13786
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.953287567842031,
+      "learning_rate": 4.510881115956545e-08,
+      "loss": 0.327,
+      "step": 13787
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.128146963243642,
+      "learning_rate": 4.495804611405141e-08,
+      "loss": 0.3105,
+      "step": 13788
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7262056991460124,
+      "learning_rate": 4.480753230006851e-08,
+      "loss": 0.2994,
+      "step": 13789
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.519815015662616,
+      "learning_rate": 4.465726972524731e-08,
+      "loss": 0.3309,
+      "step": 13790
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.074323700754882,
+      "learning_rate": 4.4507258397207265e-08,
+      "loss": 0.3332,
+      "step": 13791
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.31339526389216354,
+      "learning_rate": 4.43574983235534e-08,
+      "loss": 0.1168,
+      "step": 13792
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.669670440092013,
+      "learning_rate": 4.4207989511879104e-08,
+      "loss": 0.2784,
+      "step": 13793
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.922105738870182,
+      "learning_rate": 4.4058731969766064e-08,
+      "loss": 0.3062,
+      "step": 13794
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.969683518656397,
+      "learning_rate": 4.390972570477936e-08,
+      "loss": 0.2989,
+      "step": 13795
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.5253534073337494,
+      "learning_rate": 4.376097072447682e-08,
+      "loss": 0.2533,
+      "step": 13796
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.307611080708577,
+      "learning_rate": 4.361246703639854e-08,
+      "loss": 0.3189,
+      "step": 13797
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 16.514274368488422,
+      "learning_rate": 4.346421464807515e-08,
+      "loss": 0.3123,
+      "step": 13798
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.44524944093962,
+      "learning_rate": 4.331621356702231e-08,
+      "loss": 0.3053,
+      "step": 13799
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.62918557332487,
+      "learning_rate": 4.316846380074513e-08,
+      "loss": 0.3518,
+      "step": 13800
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.955338015360236,
+      "learning_rate": 4.3020965356734276e-08,
+      "loss": 0.285,
+      "step": 13801
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8552054478176605,
+      "learning_rate": 4.287371824246877e-08,
+      "loss": 0.306,
+      "step": 13802
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 17.669708462576573,
+      "learning_rate": 4.2726722465413205e-08,
+      "loss": 0.287,
+      "step": 13803
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.996046180098315,
+      "learning_rate": 4.257997803302161e-08,
+      "loss": 0.31,
+      "step": 13804
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8886048738153662,
+      "learning_rate": 4.243348495273414e-08,
+      "loss": 0.2914,
+      "step": 13805
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2022848992609427,
+      "learning_rate": 4.228724323197819e-08,
+      "loss": 0.2912,
+      "step": 13806
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2470796437010905,
+      "learning_rate": 4.214125287816839e-08,
+      "loss": 0.3332,
+      "step": 13807
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.235517171254143,
+      "learning_rate": 4.19955138987066e-08,
+      "loss": 0.3165,
+      "step": 13808
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.1139361582506915,
+      "learning_rate": 4.185002630098245e-08,
+      "loss": 0.2889,
+      "step": 13809
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.7711874795473355,
+      "learning_rate": 4.170479009237227e-08,
+      "loss": 0.3043,
+      "step": 13810
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.363508020199083,
+      "learning_rate": 4.155980528024017e-08,
+      "loss": 0.3098,
+      "step": 13811
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8113934210901155,
+      "learning_rate": 4.1415071871936944e-08,
+      "loss": 0.3112,
+      "step": 13812
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9185086714716757,
+      "learning_rate": 4.1270589874801146e-08,
+      "loss": 0.3042,
+      "step": 13813
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.033639735735782,
+      "learning_rate": 4.112635929615805e-08,
+      "loss": 0.3381,
+      "step": 13814
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.984191969040364,
+      "learning_rate": 4.098238014332068e-08,
+      "loss": 0.3248,
+      "step": 13815
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28956836918373746,
+      "learning_rate": 4.083865242358876e-08,
+      "loss": 0.1121,
+      "step": 13816
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.714428870249346,
+      "learning_rate": 4.069517614425034e-08,
+      "loss": 0.3257,
+      "step": 13817
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 9.013462448171616,
+      "learning_rate": 4.055195131257905e-08,
+      "loss": 0.3239,
+      "step": 13818
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3102171618829075,
+      "learning_rate": 4.040897793583687e-08,
+      "loss": 0.1018,
+      "step": 13819
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.837910524468858,
+      "learning_rate": 4.0266256021273544e-08,
+      "loss": 0.3156,
+      "step": 13820
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.070583943544163,
+      "learning_rate": 4.01237855761244e-08,
+      "loss": 0.3395,
+      "step": 13821
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.2725841991008326,
+      "learning_rate": 3.998156660761421e-08,
+      "loss": 0.0997,
+      "step": 13822
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4822367142222928,
+      "learning_rate": 3.983959912295277e-08,
+      "loss": 0.3241,
+      "step": 13823
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5310125204553193,
+      "learning_rate": 3.9697883129338756e-08,
+      "loss": 0.2823,
+      "step": 13824
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8270098337079426,
+      "learning_rate": 3.9556418633956986e-08,
+      "loss": 0.3092,
+      "step": 13825
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.941973402870271,
+      "learning_rate": 3.94152056439806e-08,
+      "loss": 0.2953,
+      "step": 13826
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9575468615854485,
+      "learning_rate": 3.9274244166568887e-08,
+      "loss": 0.3461,
+      "step": 13827
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 12.561214623315243,
+      "learning_rate": 3.91335342088689e-08,
+      "loss": 0.3105,
+      "step": 13828
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4205974959195484,
+      "learning_rate": 3.8993075778015476e-08,
+      "loss": 0.2878,
+      "step": 13829
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.537417188377312,
+      "learning_rate": 3.8852868881129605e-08,
+      "loss": 0.2837,
+      "step": 13830
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.209533881963719,
+      "learning_rate": 3.8712913525320026e-08,
+      "loss": 0.3009,
+      "step": 13831
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.590213001130556,
+      "learning_rate": 3.857320971768386e-08,
+      "loss": 0.3151,
+      "step": 13832
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.569904521330388,
+      "learning_rate": 3.8433757465302644e-08,
+      "loss": 0.2881,
+      "step": 13833
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.875183067891422,
+      "learning_rate": 3.829455677524852e-08,
+      "loss": 0.3104,
+      "step": 13834
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.8074055170231516,
+      "learning_rate": 3.8155607654578596e-08,
+      "loss": 0.3444,
+      "step": 13835
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1343920894960515,
+      "learning_rate": 3.8016910110337814e-08,
+      "loss": 0.3288,
+      "step": 13836
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.344060492394654,
+      "learning_rate": 3.7878464149558866e-08,
+      "loss": 0.303,
+      "step": 13837
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 10.646386036184746,
+      "learning_rate": 3.7740269779260043e-08,
+      "loss": 0.3313,
+      "step": 13838
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4232011769799224,
+      "learning_rate": 3.7602327006450166e-08,
+      "loss": 0.3377,
+      "step": 13839
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.1135119743337856,
+      "learning_rate": 3.746463583812143e-08,
+      "loss": 0.3414,
+      "step": 13840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3771092921213195,
+      "learning_rate": 3.732719628125603e-08,
+      "loss": 0.3154,
+      "step": 13841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.367173951611056,
+      "learning_rate": 3.719000834282227e-08,
+      "loss": 0.2915,
+      "step": 13842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9152140099715607,
+      "learning_rate": 3.705307202977515e-08,
+      "loss": 0.29,
+      "step": 13843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.220896099787453,
+      "learning_rate": 3.691638734905856e-08,
+      "loss": 0.2991,
+      "step": 13844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.196645517594772,
+      "learning_rate": 3.6779954307602504e-08,
+      "loss": 0.2876,
+      "step": 13845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3039759148117762,
+      "learning_rate": 3.664377291232424e-08,
+      "loss": 0.1129,
+      "step": 13846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7477774145384743,
+      "learning_rate": 3.650784317012879e-08,
+      "loss": 0.3258,
+      "step": 13847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.3864352163377824,
+      "learning_rate": 3.6372165087907865e-08,
+      "loss": 0.2955,
+      "step": 13848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.661536057615202,
+      "learning_rate": 3.6236738672540406e-08,
+      "loss": 0.2928,
+      "step": 13849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9535627182165323,
+      "learning_rate": 3.61015639308937e-08,
+      "loss": 0.3063,
+      "step": 13850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.666777729316247,
+      "learning_rate": 3.5966640869820044e-08,
+      "loss": 0.3286,
+      "step": 13851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.865288232454819,
+      "learning_rate": 3.5831969496161744e-08,
+      "loss": 0.3196,
+      "step": 13852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.193897608460178,
+      "learning_rate": 3.5697549816745555e-08,
+      "loss": 0.2964,
+      "step": 13853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.610506570753382,
+      "learning_rate": 3.556338183838825e-08,
+      "loss": 0.3062,
+      "step": 13854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.682954551540335,
+      "learning_rate": 3.542946556789162e-08,
+      "loss": 0.2953,
+      "step": 13855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.492946495772791,
+      "learning_rate": 3.529580101204466e-08,
+      "loss": 0.3085,
+      "step": 13856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.7910977518011286,
+      "learning_rate": 3.516238817762641e-08,
+      "loss": 0.334,
+      "step": 13857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.8494626476085285,
+      "learning_rate": 3.502922707140033e-08,
+      "loss": 0.3109,
+      "step": 13858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.981320517823536,
+      "learning_rate": 3.48963177001177e-08,
+      "loss": 0.2831,
+      "step": 13859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.954318661520524,
+      "learning_rate": 3.476366007051701e-08,
+      "loss": 0.302,
+      "step": 13860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.64198612162714,
+      "learning_rate": 3.4631254189325114e-08,
+      "loss": 0.2881,
+      "step": 13861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.421462524826004,
+      "learning_rate": 3.449910006325497e-08,
+      "loss": 0.3137,
+      "step": 13862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.0246021763247075,
+      "learning_rate": 3.4367197699007315e-08,
+      "loss": 0.2842,
+      "step": 13863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 7.062443467246728,
+      "learning_rate": 3.423554710326904e-08,
+      "loss": 0.3187,
+      "step": 13864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.587487160918221,
+      "learning_rate": 3.410414828271646e-08,
+      "loss": 0.3091,
+      "step": 13865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.399789357477564,
+      "learning_rate": 3.397300124401093e-08,
+      "loss": 0.3109,
+      "step": 13866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.649600617024299,
+      "learning_rate": 3.384210599380155e-08,
+      "loss": 0.2908,
+      "step": 13867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.793190532328277,
+      "learning_rate": 3.371146253872582e-08,
+      "loss": 0.2981,
+      "step": 13868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.9530102783447605,
+      "learning_rate": 3.358107088540674e-08,
+      "loss": 0.3252,
+      "step": 13869
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.789595926785676,
+      "learning_rate": 3.345093104045682e-08,
+      "loss": 0.3116,
+      "step": 13870
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.014058750028286,
+      "learning_rate": 3.332104301047301e-08,
+      "loss": 0.3626,
+      "step": 13871
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.759103169226932,
+      "learning_rate": 3.3191406802041693e-08,
+      "loss": 0.3348,
+      "step": 13872
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.179000927327979,
+      "learning_rate": 3.30620224217354e-08,
+      "loss": 0.2837,
+      "step": 13873
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.127365022367514,
+      "learning_rate": 3.293288987611443e-08,
+      "loss": 0.2885,
+      "step": 13874
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 8.749248727361834,
+      "learning_rate": 3.280400917172577e-08,
+      "loss": 0.3199,
+      "step": 13875
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.067038107344072,
+      "learning_rate": 3.2675380315104755e-08,
+      "loss": 0.3022,
+      "step": 13876
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.28870535788306406,
+      "learning_rate": 3.2547003312772274e-08,
+      "loss": 0.1105,
+      "step": 13877
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.265684914550726,
+      "learning_rate": 3.2418878171237565e-08,
+      "loss": 0.3371,
+      "step": 13878
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.30112130601771997,
+      "learning_rate": 3.22910048969971e-08,
+      "loss": 0.116,
+      "step": 13879
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 4.006777389535607,
+      "learning_rate": 3.216338349653403e-08,
+      "loss": 0.286,
+      "step": 13880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.250320038874585,
+      "learning_rate": 3.203601397631928e-08,
+      "loss": 0.2886,
+      "step": 13881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.859757684652465,
+      "learning_rate": 3.190889634281047e-08,
+      "loss": 0.3151,
+      "step": 13882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 11.949233954825598,
+      "learning_rate": 3.1782030602452994e-08,
+      "loss": 0.3558,
+      "step": 13883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9955247379270773,
+      "learning_rate": 3.1655416761678934e-08,
+      "loss": 0.312,
+      "step": 13884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.539408153524983,
+      "learning_rate": 3.152905482690871e-08,
+      "loss": 0.2953,
+      "step": 13885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7305968339907687,
+      "learning_rate": 3.1402944804547755e-08,
+      "loss": 0.3075,
+      "step": 13886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.387596629760473,
+      "learning_rate": 3.12770867009915e-08,
+      "loss": 0.3286,
+      "step": 13887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.683812391913039,
+      "learning_rate": 3.11514805226204e-08,
+      "loss": 0.316,
+      "step": 13888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.041106798762883,
+      "learning_rate": 3.102612627580326e-08,
+      "loss": 0.3511,
+      "step": 13889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.384031997660671,
+      "learning_rate": 3.0901023966896094e-08,
+      "loss": 0.3276,
+      "step": 13890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.179312464699883,
+      "learning_rate": 3.077617360224106e-08,
+      "loss": 0.3305,
+      "step": 13891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.8261070125302545,
+      "learning_rate": 3.065157518816919e-08,
+      "loss": 0.3211,
+      "step": 13892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6858821344920742,
+      "learning_rate": 3.052722873099767e-08,
+      "loss": 0.2946,
+      "step": 13893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.9361089119199844,
+      "learning_rate": 3.040313423703034e-08,
+      "loss": 0.2819,
+      "step": 13894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.76230938055214,
+      "learning_rate": 3.02792917125605e-08,
+      "loss": 0.3261,
+      "step": 13895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.050650722536608,
+      "learning_rate": 3.0155701163865905e-08,
+      "loss": 0.3533,
+      "step": 13896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.374214466417181,
+      "learning_rate": 3.00323625972132e-08,
+      "loss": 0.32,
+      "step": 13897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.829497498137761,
+      "learning_rate": 2.990927601885685e-08,
+      "loss": 0.3423,
+      "step": 13898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4473846578768628,
+      "learning_rate": 2.9786441435036285e-08,
+      "loss": 0.2918,
+      "step": 13899
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.14726272556726,
+      "learning_rate": 2.9663858851980974e-08,
+      "loss": 0.2961,
+      "step": 13900
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.209312312918458,
+      "learning_rate": 2.9541528275904842e-08,
+      "loss": 0.3154,
+      "step": 13901
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.337563408950165,
+      "learning_rate": 2.941944971301014e-08,
+      "loss": 0.3136,
+      "step": 13902
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.895739918297007,
+      "learning_rate": 2.929762316948803e-08,
+      "loss": 0.2556,
+      "step": 13903
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.061612163816037,
+      "learning_rate": 2.9176048651513578e-08,
+      "loss": 0.3249,
+      "step": 13904
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.471045197972228,
+      "learning_rate": 2.9054726165252955e-08,
+      "loss": 0.3339,
+      "step": 13905
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.471216647343342,
+      "learning_rate": 2.893365571685569e-08,
+      "loss": 0.3284,
+      "step": 13906
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.854476163649412,
+      "learning_rate": 2.8812837312460762e-08,
+      "loss": 0.3132,
+      "step": 13907
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.29024851562762954,
+      "learning_rate": 2.8692270958193825e-08,
+      "loss": 0.1097,
+      "step": 13908
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2242682160011857,
+      "learning_rate": 2.8571956660168875e-08,
+      "loss": 0.2733,
+      "step": 13909
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6010441909365145,
+      "learning_rate": 2.8451894424484928e-08,
+      "loss": 0.3125,
+      "step": 13910
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.338230487300206,
+      "learning_rate": 2.8332084257229886e-08,
+      "loss": 0.2982,
+      "step": 13911
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.30044776073142554,
+      "learning_rate": 2.8212526164478337e-08,
+      "loss": 0.1192,
+      "step": 13912
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2912991686501365,
+      "learning_rate": 2.8093220152292656e-08,
+      "loss": 0.2904,
+      "step": 13913
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.791251282134385,
+      "learning_rate": 2.797416622672078e-08,
+      "loss": 0.2738,
+      "step": 13914
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8582598774910877,
+      "learning_rate": 2.78553643938001e-08,
+      "loss": 0.3281,
+      "step": 13915
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.180296917810731,
+      "learning_rate": 2.7736814659553023e-08,
+      "loss": 0.29,
+      "step": 13916
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8748159436022003,
+      "learning_rate": 2.761851702999141e-08,
+      "loss": 0.2929,
+      "step": 13917
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.720984964259859,
+      "learning_rate": 2.750047151111268e-08,
+      "loss": 0.3181,
+      "step": 13918
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.078584891694278,
+      "learning_rate": 2.738267810890205e-08,
+      "loss": 0.3339,
+      "step": 13919
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.424033959859537,
+      "learning_rate": 2.726513682933196e-08,
+      "loss": 0.3098,
+      "step": 13920
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.911940279668147,
+      "learning_rate": 2.7147847678362092e-08,
+      "loss": 0.2765,
+      "step": 13921
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 16.564909380761797,
+      "learning_rate": 2.70308106619388e-08,
+      "loss": 0.3183,
+      "step": 13922
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2550975104535267,
+      "learning_rate": 2.691402578599678e-08,
+      "loss": 0.3072,
+      "step": 13923
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.562649977543185,
+      "learning_rate": 2.6797493056456847e-08,
+      "loss": 0.3107,
+      "step": 13924
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8058190998602157,
+      "learning_rate": 2.668121247922706e-08,
+      "loss": 0.3177,
+      "step": 13925
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.823013802188273,
+      "learning_rate": 2.6565184060204365e-08,
+      "loss": 0.2916,
+      "step": 13926
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5660460361648827,
+      "learning_rate": 2.6449407805270168e-08,
+      "loss": 0.3036,
+      "step": 13927
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.66205770297347,
+      "learning_rate": 2.633388372029644e-08,
+      "loss": 0.3392,
+      "step": 13928
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.947159769713214,
+      "learning_rate": 2.6218611811138493e-08,
+      "loss": 0.3079,
+      "step": 13929
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.747379445307649,
+      "learning_rate": 2.610359208364166e-08,
+      "loss": 0.2979,
+      "step": 13930
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.890686519569811,
+      "learning_rate": 2.598882454363849e-08,
+      "loss": 0.3179,
+      "step": 13931
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6273486936260797,
+      "learning_rate": 2.5874309196946555e-08,
+      "loss": 0.3108,
+      "step": 13932
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.2883346239106235,
+      "learning_rate": 2.5760046049373433e-08,
+      "loss": 0.2906,
+      "step": 13933
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.659359734677824,
+      "learning_rate": 2.5646035106711154e-08,
+      "loss": 0.3159,
+      "step": 13934
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.315100932338582,
+      "learning_rate": 2.5532276374741205e-08,
+      "loss": 0.3129,
+      "step": 13935
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.0161905334554735,
+      "learning_rate": 2.5418769859231194e-08,
+      "loss": 0.3106,
+      "step": 13936
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.6197134096547074,
+      "learning_rate": 2.5305515565936523e-08,
+      "loss": 0.3131,
+      "step": 13937
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.892032500669607,
+      "learning_rate": 2.51925135005987e-08,
+      "loss": 0.2846,
+      "step": 13938
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7506520450708707,
+      "learning_rate": 2.50797636689476e-08,
+      "loss": 0.3014,
+      "step": 13939
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.28420334444053946,
+      "learning_rate": 2.4967266076699194e-08,
+      "loss": 0.1029,
+      "step": 13940
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.178200283237656,
+      "learning_rate": 2.485502072955892e-08,
+      "loss": 0.3329,
+      "step": 13941
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.187524069132161,
+      "learning_rate": 2.4743027633216676e-08,
+      "loss": 0.3244,
+      "step": 13942
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9937009649423127,
+      "learning_rate": 2.4631286793350695e-08,
+      "loss": 0.3074,
+      "step": 13943
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.920996718351238,
+      "learning_rate": 2.4519798215627555e-08,
+      "loss": 0.3233,
+      "step": 13944
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1213682850909876,
+      "learning_rate": 2.440856190569829e-08,
+      "loss": 0.2983,
+      "step": 13945
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3099643850124749,
+      "learning_rate": 2.4297577869203946e-08,
+      "loss": 0.1081,
+      "step": 13946
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.583787922998748,
+      "learning_rate": 2.418684611177169e-08,
+      "loss": 0.3314,
+      "step": 13947
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.7627250480626655,
+      "learning_rate": 2.4076366639015914e-08,
+      "loss": 0.2947,
+      "step": 13948
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.995054975380003,
+      "learning_rate": 2.3966139456537142e-08,
+      "loss": 0.2976,
+      "step": 13949
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.970815405312538,
+      "learning_rate": 2.3856164569925346e-08,
+      "loss": 0.3045,
+      "step": 13950
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.82209928588685,
+      "learning_rate": 2.3746441984756062e-08,
+      "loss": 0.3141,
+      "step": 13951
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.101936709077862,
+      "learning_rate": 2.3636971706592627e-08,
+      "loss": 0.3009,
+      "step": 13952
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.533156190637907,
+      "learning_rate": 2.3527753740984483e-08,
+      "loss": 0.2946,
+      "step": 13953
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5197820081196887,
+      "learning_rate": 2.341878809347109e-08,
+      "loss": 0.3124,
+      "step": 13954
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.4003684061362067,
+      "learning_rate": 2.3310074769575254e-08,
+      "loss": 0.3363,
+      "step": 13955
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.076065643017603,
+      "learning_rate": 2.320161377481034e-08,
+      "loss": 0.2971,
+      "step": 13956
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.420591849590092,
+      "learning_rate": 2.309340511467473e-08,
+      "loss": 0.3137,
+      "step": 13957
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.591130594853399,
+      "learning_rate": 2.298544879465514e-08,
+      "loss": 0.3001,
+      "step": 13958
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.911117885337938,
+      "learning_rate": 2.2877744820225535e-08,
+      "loss": 0.2967,
+      "step": 13959
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.778613822912459,
+      "learning_rate": 2.2770293196845982e-08,
+      "loss": 0.3137,
+      "step": 13960
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8752498432935845,
+      "learning_rate": 2.2663093929966018e-08,
+      "loss": 0.2886,
+      "step": 13961
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.712923271836172,
+      "learning_rate": 2.2556147025018515e-08,
+      "loss": 0.3283,
+      "step": 13962
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0722608944145646,
+      "learning_rate": 2.2449452487428024e-08,
+      "loss": 0.303,
+      "step": 13963
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.635498852917713,
+      "learning_rate": 2.2343010322603553e-08,
+      "loss": 0.2941,
+      "step": 13964
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8481414787042345,
+      "learning_rate": 2.2236820535941338e-08,
+      "loss": 0.3192,
+      "step": 13965
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8963855068656637,
+      "learning_rate": 2.213088313282652e-08,
+      "loss": 0.3298,
+      "step": 13966
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.3276428852285225,
+      "learning_rate": 2.2025198118629245e-08,
+      "loss": 0.2973,
+      "step": 13967
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.620029097726035,
+      "learning_rate": 2.1919765498708556e-08,
+      "loss": 0.3013,
+      "step": 13968
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.194302028577483,
+      "learning_rate": 2.1814585278410185e-08,
+      "loss": 0.3349,
+      "step": 13969
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.664147074622605,
+      "learning_rate": 2.1709657463066523e-08,
+      "loss": 0.3188,
+      "step": 13970
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.410107522955597,
+      "learning_rate": 2.1604982057998326e-08,
+      "loss": 0.314,
+      "step": 13971
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.674747820128977,
+      "learning_rate": 2.1500559068513005e-08,
+      "loss": 0.3135,
+      "step": 13972
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.186887268236912,
+      "learning_rate": 2.1396388499903552e-08,
+      "loss": 0.3196,
+      "step": 13973
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.177894701269502,
+      "learning_rate": 2.1292470357452967e-08,
+      "loss": 0.313,
+      "step": 13974
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.920538822973648,
+      "learning_rate": 2.1188804646429806e-08,
+      "loss": 0.301,
+      "step": 13975
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.250039192142762,
+      "learning_rate": 2.1085391372090425e-08,
+      "loss": 0.3269,
+      "step": 13976
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.923271346907432,
+      "learning_rate": 2.098223053967785e-08,
+      "loss": 0.3125,
+      "step": 13977
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.386436304212422,
+      "learning_rate": 2.0879322154422343e-08,
+      "loss": 0.2887,
+      "step": 13978
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.321766127489191,
+      "learning_rate": 2.077666622154195e-08,
+      "loss": 0.308,
+      "step": 13979
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.894386158680187,
+      "learning_rate": 2.067426274624085e-08,
+      "loss": 0.3184,
+      "step": 13980
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.931828609747054,
+      "learning_rate": 2.0572111733712098e-08,
+      "loss": 0.3091,
+      "step": 13981
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.423988775115069,
+      "learning_rate": 2.0470213189134892e-08,
+      "loss": 0.3533,
+      "step": 13982
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.304953426122307,
+      "learning_rate": 2.036856711767454e-08,
+      "loss": 0.296,
+      "step": 13983
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1877138690694404,
+      "learning_rate": 2.0267173524485816e-08,
+      "loss": 0.3209,
+      "step": 13984
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7487474913873062,
+      "learning_rate": 2.01660324147096e-08,
+      "loss": 0.311,
+      "step": 13985
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 9.115737807360318,
+      "learning_rate": 2.0065143793473463e-08,
+      "loss": 0.2775,
+      "step": 13986
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.07851704935793,
+      "learning_rate": 1.9964507665893306e-08,
+      "loss": 0.2913,
+      "step": 13987
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.581700715223975,
+      "learning_rate": 1.986412403707061e-08,
+      "loss": 0.3111,
+      "step": 13988
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.8636119780954554,
+      "learning_rate": 1.9763992912096296e-08,
+      "loss": 0.3373,
+      "step": 13989
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5224756395939534,
+      "learning_rate": 1.9664114296045754e-08,
+      "loss": 0.2904,
+      "step": 13990
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.892302077834206,
+      "learning_rate": 1.9564488193984378e-08,
+      "loss": 0.2962,
+      "step": 13991
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.915577779450171,
+      "learning_rate": 1.946511461096312e-08,
+      "loss": 0.3064,
+      "step": 13992
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.163443420868553,
+      "learning_rate": 1.9365993552019624e-08,
+      "loss": 0.3328,
+      "step": 13993
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.603894005419899,
+      "learning_rate": 1.9267125022180977e-08,
+      "loss": 0.2997,
+      "step": 13994
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.009590475799026,
+      "learning_rate": 1.9168509026458725e-08,
+      "loss": 0.2845,
+      "step": 13995
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.546066225313461,
+      "learning_rate": 1.9070145569853315e-08,
+      "loss": 0.3283,
+      "step": 13996
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.489858543557187,
+      "learning_rate": 1.8972034657352423e-08,
+      "loss": 0.3178,
+      "step": 13997
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.770047804497135,
+      "learning_rate": 1.8874176293929847e-08,
+      "loss": 0.3199,
+      "step": 13998
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.759920907051597,
+      "learning_rate": 1.877657048454773e-08,
+      "loss": 0.2896,
+      "step": 13999
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.66177600453376,
+      "learning_rate": 1.8679217234154335e-08,
+      "loss": 0.3324,
+      "step": 14000
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.235373100871775,
+      "learning_rate": 1.8582116547686267e-08,
+      "loss": 0.3101,
+      "step": 14001
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.5741446438918647,
+      "learning_rate": 1.8485268430066817e-08,
+      "loss": 0.304,
+      "step": 14002
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.904582666827751,
+      "learning_rate": 1.8388672886205384e-08,
+      "loss": 0.3191,
+      "step": 14003
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.88441396661476,
+      "learning_rate": 1.8292329921001385e-08,
+      "loss": 0.3492,
+      "step": 14004
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.2496122157459824,
+      "learning_rate": 1.8196239539337578e-08,
+      "loss": 0.3208,
+      "step": 14005
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.0073984522637045,
+      "learning_rate": 1.810040174608729e-08,
+      "loss": 0.2517,
+      "step": 14006
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6930693347193575,
+      "learning_rate": 1.8004816546109417e-08,
+      "loss": 0.2924,
+      "step": 14007
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.390014551709129,
+      "learning_rate": 1.7909483944250072e-08,
+      "loss": 0.3227,
+      "step": 14008
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.7328301189546105,
+      "learning_rate": 1.7814403945343174e-08,
+      "loss": 0.3169,
+      "step": 14009
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9502670547814396,
+      "learning_rate": 1.7719576554209306e-08,
+      "loss": 0.3256,
+      "step": 14010
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 6.197737622295253,
+      "learning_rate": 1.762500177565629e-08,
+      "loss": 0.3067,
+      "step": 14011
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.27980222381136516,
+      "learning_rate": 1.7530679614479185e-08,
+      "loss": 0.1007,
+      "step": 14012
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.2952539710575981,
+      "learning_rate": 1.7436610075460826e-08,
+      "loss": 0.1085,
+      "step": 14013
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 7.723224029435702,
+      "learning_rate": 1.734279316337073e-08,
+      "loss": 0.3216,
+      "step": 14014
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.479928039663396,
+      "learning_rate": 1.7249228882965095e-08,
+      "loss": 0.3028,
+      "step": 14015
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.094508188545538,
+      "learning_rate": 1.7155917238987906e-08,
+      "loss": 0.3398,
+      "step": 14016
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.254269177061795,
+      "learning_rate": 1.706285823617093e-08,
+      "loss": 0.3288,
+      "step": 14017
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.068170993621607,
+      "learning_rate": 1.6970051879232063e-08,
+      "loss": 0.2881,
+      "step": 14018
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 25.270811640276843,
+      "learning_rate": 1.6877498172876426e-08,
+      "loss": 0.3205,
+      "step": 14019
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.847254210208823,
+      "learning_rate": 1.6785197121797492e-08,
+      "loss": 0.3309,
+      "step": 14020
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.710639781489966,
+      "learning_rate": 1.669314873067429e-08,
+      "loss": 0.313,
+      "step": 14021
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.592545251829984,
+      "learning_rate": 1.6601353004174202e-08,
+      "loss": 0.3194,
+      "step": 14022
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.6471832829214654,
+      "learning_rate": 1.6509809946951282e-08,
+      "loss": 0.3156,
+      "step": 14023
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 12.844706681778057,
+      "learning_rate": 1.641851956364793e-08,
+      "loss": 0.3244,
+      "step": 14024
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2923264270868536,
+      "learning_rate": 1.6327481858891547e-08,
+      "loss": 0.1114,
+      "step": 14025
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.404439637565545,
+      "learning_rate": 1.623669683729845e-08,
+      "loss": 0.3363,
+      "step": 14026
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.714391158413538,
+      "learning_rate": 1.6146164503471616e-08,
+      "loss": 0.3253,
+      "step": 14027
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 15.896398631294183,
+      "learning_rate": 1.605588486200127e-08,
+      "loss": 0.3092,
+      "step": 14028
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.018478582178994,
+      "learning_rate": 1.5965857917464855e-08,
+      "loss": 0.3133,
+      "step": 14029
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.4162409877240423,
+      "learning_rate": 1.58760836744265e-08,
+      "loss": 0.2919,
+      "step": 14030
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3765226978410396,
+      "learning_rate": 1.5786562137438676e-08,
+      "loss": 0.2934,
+      "step": 14031
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.081090206213098,
+      "learning_rate": 1.5697293311039973e-08,
+      "loss": 0.3205,
+      "step": 14032
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.414720681082278,
+      "learning_rate": 1.5608277199756216e-08,
+      "loss": 0.3247,
+      "step": 14033
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3970556049835205,
+      "learning_rate": 1.5519513808101016e-08,
+      "loss": 0.3176,
+      "step": 14034
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.289652803857216,
+      "learning_rate": 1.5431003140574662e-08,
+      "loss": 0.3025,
+      "step": 14035
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8147968770114313,
+      "learning_rate": 1.5342745201665232e-08,
+      "loss": 0.3243,
+      "step": 14036
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 13.474100581178668,
+      "learning_rate": 1.5254739995847477e-08,
+      "loss": 0.2833,
+      "step": 14037
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.826472766292481,
+      "learning_rate": 1.5166987527583387e-08,
+      "loss": 0.3058,
+      "step": 14038
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.1689902641977605,
+      "learning_rate": 1.5079487801322178e-08,
+      "loss": 0.3092,
+      "step": 14039
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.782960154862105,
+      "learning_rate": 1.4992240821499748e-08,
+      "loss": 0.3126,
+      "step": 14040
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.368281062200082,
+      "learning_rate": 1.4905246592540888e-08,
+      "loss": 0.3269,
+      "step": 14041
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.943701433180165,
+      "learning_rate": 1.4818505118855408e-08,
+      "loss": 0.2983,
+      "step": 14042
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.767800429383791,
+      "learning_rate": 1.4732016404842009e-08,
+      "loss": 0.2957,
+      "step": 14043
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.421171866899793,
+      "learning_rate": 1.4645780454885517e-08,
+      "loss": 0.3237,
+      "step": 14044
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.889086831997563,
+      "learning_rate": 1.4559797273357435e-08,
+      "loss": 0.2961,
+      "step": 14045
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6621427399107636,
+      "learning_rate": 1.4474066864618718e-08,
+      "loss": 0.3033,
+      "step": 14046
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.411482318657003,
+      "learning_rate": 1.4388589233015338e-08,
+      "loss": 0.2963,
+      "step": 14047
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 16.81960523233349,
+      "learning_rate": 1.4303364382881601e-08,
+      "loss": 0.3248,
+      "step": 14048
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.319728342983963,
+      "learning_rate": 1.4218392318538499e-08,
+      "loss": 0.3177,
+      "step": 14049
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.170118703072978,
+      "learning_rate": 1.4133673044293695e-08,
+      "loss": 0.3014,
+      "step": 14050
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.44491716774122,
+      "learning_rate": 1.4049206564443197e-08,
+      "loss": 0.3075,
+      "step": 14051
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.614954751803119,
+      "learning_rate": 1.396499288326969e-08,
+      "loss": 0.276,
+      "step": 14052
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.897948201920385,
+      "learning_rate": 1.3881032005042538e-08,
+      "loss": 0.3184,
+      "step": 14053
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2837594407014428,
+      "learning_rate": 1.3797323934019447e-08,
+      "loss": 0.3027,
+      "step": 14054
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9064666580807565,
+      "learning_rate": 1.3713868674443687e-08,
+      "loss": 0.2772,
+      "step": 14055
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.902927904590583,
+      "learning_rate": 1.3630666230547429e-08,
+      "loss": 0.3235,
+      "step": 14056
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.301423173717159,
+      "learning_rate": 1.3547716606548967e-08,
+      "loss": 0.3288,
+      "step": 14057
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.403052729411858,
+      "learning_rate": 1.3465019806653822e-08,
+      "loss": 0.3298,
+      "step": 14058
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.500051656168662,
+      "learning_rate": 1.338257583505531e-08,
+      "loss": 0.2788,
+      "step": 14059
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6793870919637506,
+      "learning_rate": 1.3300384695932866e-08,
+      "loss": 0.2819,
+      "step": 14060
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6579625300094483,
+      "learning_rate": 1.321844639345482e-08,
+      "loss": 0.2848,
+      "step": 14061
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.536518779322217,
+      "learning_rate": 1.3136760931774517e-08,
+      "loss": 0.3155,
+      "step": 14062
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 10.082400678183173,
+      "learning_rate": 1.3055328315034198e-08,
+      "loss": 0.3073,
+      "step": 14063
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.938569797078911,
+      "learning_rate": 1.2974148547362231e-08,
+      "loss": 0.3116,
+      "step": 14064
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.181410060261659,
+      "learning_rate": 1.2893221632875319e-08,
+      "loss": 0.3072,
+      "step": 14065
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.673338092830721,
+      "learning_rate": 1.2812547575676293e-08,
+      "loss": 0.2809,
+      "step": 14066
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.0111448149624005,
+      "learning_rate": 1.2732126379855214e-08,
+      "loss": 0.3109,
+      "step": 14067
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.2042385323025,
+      "learning_rate": 1.2651958049489377e-08,
+      "loss": 0.3104,
+      "step": 14068
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.7381176414884023,
+      "learning_rate": 1.2572042588644417e-08,
+      "loss": 0.3393,
+      "step": 14069
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.746585220667191,
+      "learning_rate": 1.2492380001372095e-08,
+      "loss": 0.308,
+      "step": 14070
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2961330226221732,
+      "learning_rate": 1.2412970291710847e-08,
+      "loss": 0.3156,
+      "step": 14071
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.160989693681481,
+      "learning_rate": 1.2333813463686895e-08,
+      "loss": 0.329,
+      "step": 14072
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3299682085092974,
+      "learning_rate": 1.2254909521314251e-08,
+      "loss": 0.3266,
+      "step": 14073
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7279693997638836,
+      "learning_rate": 1.2176258468593605e-08,
+      "loss": 0.3111,
+      "step": 14074
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.3241735508026675,
+      "learning_rate": 1.2097860309511767e-08,
+      "loss": 0.2956,
+      "step": 14075
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5336219429487485,
+      "learning_rate": 1.2019715048045e-08,
+      "loss": 0.3045,
+      "step": 14076
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.129071880923211,
+      "learning_rate": 1.1941822688154026e-08,
+      "loss": 0.3151,
+      "step": 14077
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8144400017233586,
+      "learning_rate": 1.1864183233789018e-08,
+      "loss": 0.2902,
+      "step": 14078
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1212962834144418,
+      "learning_rate": 1.1786796688886826e-08,
+      "loss": 0.2826,
+      "step": 14079
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.914299749465431,
+      "learning_rate": 1.1709663057370424e-08,
+      "loss": 0.3426,
+      "step": 14080
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.801373519726839,
+      "learning_rate": 1.1632782343150572e-08,
+      "loss": 0.3226,
+      "step": 14081
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 20.557907333803982,
+      "learning_rate": 1.155615455012582e-08,
+      "loss": 0.3314,
+      "step": 14082
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2088506592218864,
+      "learning_rate": 1.1479779682180836e-08,
+      "loss": 0.3628,
+      "step": 14083
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.695776178372375,
+      "learning_rate": 1.1403657743188634e-08,
+      "loss": 0.3095,
+      "step": 14084
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6504256280132794,
+      "learning_rate": 1.132778873700835e-08,
+      "loss": 0.3175,
+      "step": 14085
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.369991649400301,
+      "learning_rate": 1.1252172667486904e-08,
+      "loss": 0.3212,
+      "step": 14086
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.5596387286966817,
+      "learning_rate": 1.11768095384579e-08,
+      "loss": 0.304,
+      "step": 14087
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.7467141505101496,
+      "learning_rate": 1.1101699353742723e-08,
+      "loss": 0.2884,
+      "step": 14088
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1062416413354788,
+      "learning_rate": 1.1026842117148884e-08,
+      "loss": 0.3137,
+      "step": 14089
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.60397074585437,
+      "learning_rate": 1.0952237832472789e-08,
+      "loss": 0.3121,
+      "step": 14090
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.29517984538022,
+      "learning_rate": 1.087788650349697e-08,
+      "loss": 0.3088,
+      "step": 14091
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.521646870654504,
+      "learning_rate": 1.0803788133990634e-08,
+      "loss": 0.2801,
+      "step": 14092
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.386657271226152,
+      "learning_rate": 1.0729942727711329e-08,
+      "loss": 0.2796,
+      "step": 14093
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.538534602709648,
+      "learning_rate": 1.0656350288402172e-08,
+      "loss": 0.3161,
+      "step": 14094
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.455995575336327,
+      "learning_rate": 1.0583010819795735e-08,
+      "loss": 0.2861,
+      "step": 14095
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.8846013479173997,
+      "learning_rate": 1.05099243256096e-08,
+      "loss": 0.2862,
+      "step": 14096
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6822984724044514,
+      "learning_rate": 1.0437090809549688e-08,
+      "loss": 0.2784,
+      "step": 14097
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.71048638915302,
+      "learning_rate": 1.0364510275309158e-08,
+      "loss": 0.303,
+      "step": 14098
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.118324038270304,
+      "learning_rate": 1.029218272656729e-08,
+      "loss": 0.2915,
+      "step": 14099
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.031028392637312,
+      "learning_rate": 1.0220108166991705e-08,
+      "loss": 0.2999,
+      "step": 14100
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2783177359744394,
+      "learning_rate": 1.01482866002367e-08,
+      "loss": 0.1098,
+      "step": 14101
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 30.416406515121096,
+      "learning_rate": 1.007671802994381e-08,
+      "loss": 0.3142,
+      "step": 14102
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.018310758789497,
+      "learning_rate": 1.0005402459741798e-08,
+      "loss": 0.2943,
+      "step": 14103
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.934231538577718,
+      "learning_rate": 9.934339893246104e-09,
+      "loss": 0.3485,
+      "step": 14104
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.205326536921696,
+      "learning_rate": 9.863530334059956e-09,
+      "loss": 0.3098,
+      "step": 14105
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0279027702572234,
+      "learning_rate": 9.792973785774373e-09,
+      "loss": 0.3147,
+      "step": 14106
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9670246432674423,
+      "learning_rate": 9.722670251965383e-09,
+      "loss": 0.323,
+      "step": 14107
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.942130410998033,
+      "learning_rate": 9.65261973619791e-09,
+      "loss": 0.3018,
+      "step": 14108
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 32.73445352836255,
+      "learning_rate": 9.58282224202467e-09,
+      "loss": 0.3116,
+      "step": 14109
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.448002972077152,
+      "learning_rate": 9.513277772982832e-09,
+      "loss": 0.3492,
+      "step": 14110
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0055341233897535,
+      "learning_rate": 9.44398633260013e-09,
+      "loss": 0.2699,
+      "step": 14111
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2757717774955537,
+      "learning_rate": 9.374947924388755e-09,
+      "loss": 0.0985,
+      "step": 14112
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.545727631020051,
+      "learning_rate": 9.306162551849796e-09,
+      "loss": 0.2758,
+      "step": 14113
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.2861035372752008,
+      "learning_rate": 9.237630218469907e-09,
+      "loss": 0.1114,
+      "step": 14114
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8895790732536426,
+      "learning_rate": 9.169350927724641e-09,
+      "loss": 0.3342,
+      "step": 14115
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2054362277666826,
+      "learning_rate": 9.101324683075674e-09,
+      "loss": 0.3169,
+      "step": 14116
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.6093479696605417,
+      "learning_rate": 9.033551487971915e-09,
+      "loss": 0.2996,
+      "step": 14117
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.614818892497251,
+      "learning_rate": 8.966031345850056e-09,
+      "loss": 0.2916,
+      "step": 14118
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.051048304253541,
+      "learning_rate": 8.898764260132919e-09,
+      "loss": 0.3004,
+      "step": 14119
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.088981361879477,
+      "learning_rate": 8.831750234231107e-09,
+      "loss": 0.3128,
+      "step": 14120
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.74235620920348,
+      "learning_rate": 8.764989271543013e-09,
+      "loss": 0.2776,
+      "step": 14121
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.509214819867798,
+      "learning_rate": 8.69848137545315e-09,
+      "loss": 0.3023,
+      "step": 14122
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.038409260526576,
+      "learning_rate": 8.632226549333821e-09,
+      "loss": 0.3086,
+      "step": 14123
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.734802966146891,
+      "learning_rate": 8.566224796544009e-09,
+      "loss": 0.2916,
+      "step": 14124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 17.433098148297876,
+      "learning_rate": 8.500476120429923e-09,
+      "loss": 0.3085,
+      "step": 14125
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 6.847594361255532,
+      "learning_rate": 8.434980524326119e-09,
+      "loss": 0.3353,
+      "step": 14126
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.747771429862945,
+      "learning_rate": 8.369738011552164e-09,
+      "loss": 0.3344,
+      "step": 14127
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.941658123481776,
+      "learning_rate": 8.304748585417077e-09,
+      "loss": 0.3337,
+      "step": 14128
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.083805673954181,
+      "learning_rate": 8.240012249215445e-09,
+      "loss": 0.3062,
+      "step": 14129
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.013610350741637,
+      "learning_rate": 8.175529006229089e-09,
+      "loss": 0.3025,
+      "step": 14130
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.2676635257817095,
+      "learning_rate": 8.111298859728722e-09,
+      "loss": 0.2918,
+      "step": 14131
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 12.807494329340901,
+      "learning_rate": 8.047321812969521e-09,
+      "loss": 0.3074,
+      "step": 14132
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.773969266723606,
+      "learning_rate": 7.983597869196113e-09,
+      "loss": 0.3194,
+      "step": 14133
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.637477085314018,
+      "learning_rate": 7.920127031639802e-09,
+      "loss": 0.2848,
+      "step": 14134
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1920202548709335,
+      "learning_rate": 7.856909303517458e-09,
+      "loss": 0.3157,
+      "step": 14135
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.507793810578193,
+      "learning_rate": 7.793944688035404e-09,
+      "loss": 0.3163,
+      "step": 14136
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 14.637920547788827,
+      "learning_rate": 7.731233188385534e-09,
+      "loss": 0.3102,
+      "step": 14137
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31286117552098314,
+      "learning_rate": 7.668774807747526e-09,
+      "loss": 0.1145,
+      "step": 14138
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.638698648891778,
+      "learning_rate": 7.60656954928829e-09,
+      "loss": 0.3377,
+      "step": 14139
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 7.1474285333010705,
+      "learning_rate": 7.54461741616197e-09,
+      "loss": 0.3205,
+      "step": 14140
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.184938550188741,
+      "learning_rate": 7.482918411509388e-09,
+      "loss": 0.3182,
+      "step": 14141
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9114539373225807,
+      "learning_rate": 7.421472538459151e-09,
+      "loss": 0.3163,
+      "step": 14142
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0227150728741954,
+      "learning_rate": 7.360279800125991e-09,
+      "loss": 0.303,
+      "step": 14143
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.423229067378653,
+      "learning_rate": 7.299340199613536e-09,
+      "loss": 0.3078,
+      "step": 14144
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 53.77858036969553,
+      "learning_rate": 7.238653740010426e-09,
+      "loss": 0.3292,
+      "step": 14145
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 8.593440401410122,
+      "learning_rate": 7.1782204243947554e-09,
+      "loss": 0.3372,
+      "step": 14146
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.567763002059681,
+      "learning_rate": 7.118040255829628e-09,
+      "loss": 0.3518,
+      "step": 14147
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.75446129050678,
+      "learning_rate": 7.058113237367048e-09,
+      "loss": 0.3474,
+      "step": 14148
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31126406244008553,
+      "learning_rate": 6.998439372044585e-09,
+      "loss": 0.1186,
+      "step": 14149
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 18.07688119618388,
+      "learning_rate": 6.9390186628887076e-09,
+      "loss": 0.3024,
+      "step": 14150
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.883251323605561,
+      "learning_rate": 6.879851112912006e-09,
+      "loss": 0.2805,
+      "step": 14151
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1418595632459723,
+      "learning_rate": 6.820936725113747e-09,
+      "loss": 0.3158,
+      "step": 14152
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.578224873548522,
+      "learning_rate": 6.762275502482097e-09,
+      "loss": 0.3141,
+      "step": 14153
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.6864046217997,
+      "learning_rate": 6.703867447990231e-09,
+      "loss": 0.329,
+      "step": 14154
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.9954436496972776,
+      "learning_rate": 6.6457125646002265e-09,
+      "loss": 0.319,
+      "step": 14155
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3353662366983654,
+      "learning_rate": 6.587810855260279e-09,
+      "loss": 0.3217,
+      "step": 14156
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1193451317307836,
+      "learning_rate": 6.53016232290693e-09,
+      "loss": 0.3561,
+      "step": 14157
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.650114593976981,
+      "learning_rate": 6.472766970462285e-09,
+      "loss": 0.3138,
+      "step": 14158
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.4125407140982,
+      "learning_rate": 6.4156248008362396e-09,
+      "loss": 0.3073,
+      "step": 14159
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.142289082587251,
+      "learning_rate": 6.3587358169264755e-09,
+      "loss": 0.3015,
+      "step": 14160
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.578277562334128,
+      "learning_rate": 6.3021000216179074e-09,
+      "loss": 0.2911,
+      "step": 14161
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.129087045672575,
+      "learning_rate": 6.245717417781017e-09,
+      "loss": 0.3129,
+      "step": 14162
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.856641624139882,
+      "learning_rate": 6.189588008275182e-09,
+      "loss": 0.3183,
+      "step": 14163
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.293661636785202,
+      "learning_rate": 6.133711795946462e-09,
+      "loss": 0.2934,
+      "step": 14164
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.092619523585973,
+      "learning_rate": 6.078088783627034e-09,
+      "loss": 0.3231,
+      "step": 14165
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 5.014025060815671,
+      "learning_rate": 6.022718974137976e-09,
+      "loss": 0.2829,
+      "step": 14166
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.31218518388579786,
+      "learning_rate": 5.967602370286485e-09,
+      "loss": 0.1076,
+      "step": 14167
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 9.789042163588443,
+      "learning_rate": 5.912738974866439e-09,
+      "loss": 0.2774,
+      "step": 14168
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.10052623332806,
+      "learning_rate": 5.858128790660611e-09,
+      "loss": 0.2841,
+      "step": 14169
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 11.033620794577208,
+      "learning_rate": 5.803771820436788e-09,
+      "loss": 0.303,
+      "step": 14170
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.957642259290475,
+      "learning_rate": 5.7496680669516524e-09,
+      "loss": 0.2975,
+      "step": 14171
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8924353088719665,
+      "learning_rate": 5.695817532948567e-09,
+      "loss": 0.3073,
+      "step": 14172
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7925674446099134,
+      "learning_rate": 5.642220221157013e-09,
+      "loss": 0.3169,
+      "step": 14173
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.323650622198277,
+      "learning_rate": 5.588876134295374e-09,
+      "loss": 0.0991,
+      "step": 14174
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.680139621020046,
+      "learning_rate": 5.535785275067595e-09,
+      "loss": 0.3512,
+      "step": 14175
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.757544344213651,
+      "learning_rate": 5.48294764616597e-09,
+      "loss": 0.3308,
+      "step": 14176
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.789756524231623,
+      "learning_rate": 5.430363250269466e-09,
+      "loss": 0.3069,
+      "step": 14177
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.146461414821093,
+      "learning_rate": 5.378032090044283e-09,
+      "loss": 0.2774,
+      "step": 14178
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.797604640819419,
+      "learning_rate": 5.325954168143299e-09,
+      "loss": 0.3355,
+      "step": 14179
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.84835963113392,
+      "learning_rate": 5.27412948720718e-09,
+      "loss": 0.3125,
+      "step": 14180
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3805001340538587,
+      "learning_rate": 5.222558049863824e-09,
+      "loss": 0.3057,
+      "step": 14181
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8774018397026424,
+      "learning_rate": 5.171239858727806e-09,
+      "loss": 0.2854,
+      "step": 14182
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.995136133319812,
+      "learning_rate": 5.120174916400933e-09,
+      "loss": 0.3367,
+      "step": 14183
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30021041325172626,
+      "learning_rate": 5.069363225472801e-09,
+      "loss": 0.1076,
+      "step": 14184
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.415766782696194,
+      "learning_rate": 5.018804788519127e-09,
+      "loss": 0.302,
+      "step": 14185
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.30368492679798836,
+      "learning_rate": 4.968499608103417e-09,
+      "loss": 0.1096,
+      "step": 14186
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.754943550621998,
+      "learning_rate": 4.918447686776406e-09,
+      "loss": 0.3266,
+      "step": 14187
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8899621525666275,
+      "learning_rate": 4.868649027075512e-09,
+      "loss": 0.3109,
+      "step": 14188
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.635631314471875,
+      "learning_rate": 4.819103631526489e-09,
+      "loss": 0.3436,
+      "step": 14189
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4685699825945444,
+      "learning_rate": 4.769811502640664e-09,
+      "loss": 0.2954,
+      "step": 14190
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9804229156645676,
+      "learning_rate": 4.720772642917149e-09,
+      "loss": 0.2741,
+      "step": 14191
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7749583633118924,
+      "learning_rate": 4.671987054842842e-09,
+      "loss": 0.3379,
+      "step": 14192
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.700554032213322,
+      "learning_rate": 4.623454740890765e-09,
+      "loss": 0.3104,
+      "step": 14193
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.597027056648388,
+      "learning_rate": 4.575175703522283e-09,
+      "loss": 0.2762,
+      "step": 14194
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.572147918585962,
+      "learning_rate": 4.527149945184328e-09,
+      "loss": 0.2876,
+      "step": 14195
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.410595655178808,
+      "learning_rate": 4.4793774683127286e-09,
+      "loss": 0.3047,
+      "step": 14196
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.8475325449774007,
+      "learning_rate": 4.431858275329437e-09,
+      "loss": 0.3092,
+      "step": 14197
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3900335536065636,
+      "learning_rate": 4.384592368643637e-09,
+      "loss": 0.2801,
+      "step": 14198
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.156421012290179,
+      "learning_rate": 4.337579750651744e-09,
+      "loss": 0.2995,
+      "step": 14199
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2975133285950831,
+      "learning_rate": 4.290820423737963e-09,
+      "loss": 0.122,
+      "step": 14200
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.843985371619224,
+      "learning_rate": 4.244314390272619e-09,
+      "loss": 0.3209,
+      "step": 14201
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.72800556974167,
+      "learning_rate": 4.198061652613272e-09,
+      "loss": 0.2952,
+      "step": 14202
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.06526244083059,
+      "learning_rate": 4.152062213105823e-09,
+      "loss": 0.2988,
+      "step": 14203
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.862534363576964,
+      "learning_rate": 4.10631607408174e-09,
+      "loss": 0.2907,
+      "step": 14204
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 12.070317355375163,
+      "learning_rate": 4.060823237861389e-09,
+      "loss": 0.2942,
+      "step": 14205
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.208677415251463,
+      "learning_rate": 4.015583706750703e-09,
+      "loss": 0.2809,
+      "step": 14206
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.485644726312977,
+      "learning_rate": 3.970597483043958e-09,
+      "loss": 0.334,
+      "step": 14207
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.230541147886433,
+      "learning_rate": 3.925864569021554e-09,
+      "loss": 0.2909,
+      "step": 14208
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.9827692075133045,
+      "learning_rate": 3.881384966951673e-09,
+      "loss": 0.3006,
+      "step": 14209
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.738513417528931,
+      "learning_rate": 3.837158679089181e-09,
+      "loss": 0.2956,
+      "step": 14210
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.86240316252685,
+      "learning_rate": 3.7931857076772826e-09,
+      "loss": 0.2971,
+      "step": 14211
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.88461366668116,
+      "learning_rate": 3.749466054944751e-09,
+      "loss": 0.3225,
+      "step": 14212
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6746840446098306,
+      "learning_rate": 3.7059997231086996e-09,
+      "loss": 0.2745,
+      "step": 14213
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.541737983924339,
+      "learning_rate": 3.6627867143729233e-09,
+      "loss": 0.3216,
+      "step": 14214
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.551361775628899,
+      "learning_rate": 3.6198270309278916e-09,
+      "loss": 0.3486,
+      "step": 14215
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.31715651892722113,
+      "learning_rate": 3.577120674952972e-09,
+      "loss": 0.1003,
+      "step": 14216
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.090262661741617,
+      "learning_rate": 3.534667648611989e-09,
+      "loss": 0.298,
+      "step": 14217
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.3366384191744958,
+      "learning_rate": 3.4924679540582207e-09,
+      "loss": 0.2977,
+      "step": 14218
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.827013294403161,
+      "learning_rate": 3.4505215934316217e-09,
+      "loss": 0.2951,
+      "step": 14219
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.517951305551984,
+      "learning_rate": 3.408828568858269e-09,
+      "loss": 0.3032,
+      "step": 14220
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 8.0552136036802,
+      "learning_rate": 3.3673888824520275e-09,
+      "loss": 0.3054,
+      "step": 14221
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2940064188485669,
+      "learning_rate": 3.326202536314549e-09,
+      "loss": 0.1047,
+      "step": 14222
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9631990999908635,
+      "learning_rate": 3.2852695325336082e-09,
+      "loss": 0.3162,
+      "step": 14223
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.173399517163131,
+      "learning_rate": 3.244589873185322e-09,
+      "loss": 0.2827,
+      "step": 14224
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.598628234407879,
+      "learning_rate": 3.2041635603313747e-09,
+      "loss": 0.3221,
+      "step": 14225
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.221127081000923,
+      "learning_rate": 3.1639905960217933e-09,
+      "loss": 0.3088,
+      "step": 14226
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 14.720046214187322,
+      "learning_rate": 3.1240709822932815e-09,
+      "loss": 0.2947,
+      "step": 14227
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.630881430737811,
+      "learning_rate": 3.084404721170331e-09,
+      "loss": 0.3147,
+      "step": 14228
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0526945667868715,
+      "learning_rate": 3.044991814663556e-09,
+      "loss": 0.3139,
+      "step": 14229
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.6188753756588925,
+      "learning_rate": 3.005832264771358e-09,
+      "loss": 0.3454,
+      "step": 14230
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.370556108645116,
+      "learning_rate": 2.9669260734799256e-09,
+      "loss": 0.315,
+      "step": 14231
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2806854685817911,
+      "learning_rate": 2.928273242761015e-09,
+      "loss": 0.1069,
+      "step": 14232
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.27736285748412,
+      "learning_rate": 2.889873774574725e-09,
+      "loss": 0.3311,
+      "step": 14233
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.199656975640958,
+      "learning_rate": 2.851727670867832e-09,
+      "loss": 0.3042,
+      "step": 14234
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.535172964055156,
+      "learning_rate": 2.8138349335748995e-09,
+      "loss": 0.3191,
+      "step": 14235
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.175304425760248,
+      "learning_rate": 2.776195564616613e-09,
+      "loss": 0.3033,
+      "step": 14236
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.166738990217897,
+      "learning_rate": 2.7388095659020008e-09,
+      "loss": 0.2753,
+      "step": 14237
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5833413226974593,
+      "learning_rate": 2.7016769393262144e-09,
+      "loss": 0.3125,
+      "step": 14238
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.1787444507266622,
+      "learning_rate": 2.6647976867716365e-09,
+      "loss": 0.3116,
+      "step": 14239
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.699891851804397,
+      "learning_rate": 2.6281718101089927e-09,
+      "loss": 0.3063,
+      "step": 14240
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 150.17179542854225,
+      "learning_rate": 2.591799311194021e-09,
+      "loss": 0.2943,
+      "step": 14241
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4530411162862427,
+      "learning_rate": 2.555680191872467e-09,
+      "loss": 0.2986,
+      "step": 14242
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.064123105860693,
+      "learning_rate": 2.5198144539739787e-09,
+      "loss": 0.2958,
+      "step": 14243
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.635068944799661,
+      "learning_rate": 2.4842020993187667e-09,
+      "loss": 0.3289,
+      "step": 14244
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.038960401368524,
+      "learning_rate": 2.4488431297109426e-09,
+      "loss": 0.306,
+      "step": 14245
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.625903785519005,
+      "learning_rate": 2.4137375469440727e-09,
+      "loss": 0.318,
+      "step": 14246
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.789123079975259,
+      "learning_rate": 2.378885352797844e-09,
+      "loss": 0.3127,
+      "step": 14247
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.152582877088818,
+      "learning_rate": 2.3442865490397315e-09,
+      "loss": 0.3163,
+      "step": 14248
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.0081838970548085,
+      "learning_rate": 2.309941137423333e-09,
+      "loss": 0.3309,
+      "step": 14249
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9608821741933067,
+      "learning_rate": 2.2758491196905873e-09,
+      "loss": 0.2656,
+      "step": 14250
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.266120271802105,
+      "learning_rate": 2.2420104975695576e-09,
+      "loss": 0.3497,
+      "step": 14251
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5395954262059335,
+      "learning_rate": 2.2084252727766487e-09,
+      "loss": 0.3023,
+      "step": 14252
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.229483997330917,
+      "learning_rate": 2.1750934470138317e-09,
+      "loss": 0.2964,
+      "step": 14253
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3019146886526768,
+      "learning_rate": 2.1420150219714218e-09,
+      "loss": 0.1194,
+      "step": 14254
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 17.739659380039026,
+      "learning_rate": 2.109189999326966e-09,
+      "loss": 0.3108,
+      "step": 14255
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.05383066620802,
+      "learning_rate": 2.076618380744133e-09,
+      "loss": 0.3223,
+      "step": 14256
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.567010295298696,
+      "learning_rate": 2.044300167874935e-09,
+      "loss": 0.2862,
+      "step": 14257
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.7059444945614315,
+      "learning_rate": 2.0122353623580616e-09,
+      "loss": 0.2896,
+      "step": 14258
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.181247354370517,
+      "learning_rate": 1.9804239658183234e-09,
+      "loss": 0.3217,
+      "step": 14259
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.293963453520122,
+      "learning_rate": 1.94886597986943e-09,
+      "loss": 0.3033,
+      "step": 14260
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.812335706657012,
+      "learning_rate": 1.9175614061117677e-09,
+      "loss": 0.3181,
+      "step": 14261
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 10.26918546839208,
+      "learning_rate": 1.88651024613129e-09,
+      "loss": 0.3127,
+      "step": 14262
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2179340589951955,
+      "learning_rate": 1.8557125015034038e-09,
+      "loss": 0.2868,
+      "step": 14263
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.542071775571474,
+      "learning_rate": 1.8251681737896375e-09,
+      "loss": 0.2886,
+      "step": 14264
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.346351980212771,
+      "learning_rate": 1.7948772645381973e-09,
+      "loss": 0.3112,
+      "step": 14265
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.9107911972176206,
+      "learning_rate": 1.7648397752850766e-09,
+      "loss": 0.321,
+      "step": 14266
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.754836285790147,
+      "learning_rate": 1.7350557075529461e-09,
+      "loss": 0.3372,
+      "step": 14267
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.833644691602219,
+      "learning_rate": 1.7055250628522647e-09,
+      "loss": 0.3209,
+      "step": 14268
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.278102120327444,
+      "learning_rate": 1.6762478426807228e-09,
+      "loss": 0.3023,
+      "step": 14269
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.998239173079515,
+      "learning_rate": 1.6472240485215785e-09,
+      "loss": 0.2814,
+      "step": 14270
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.925109251335467,
+      "learning_rate": 1.6184536818475426e-09,
+      "loss": 0.2983,
+      "step": 14271
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2640518846513165,
+      "learning_rate": 1.5899367441168934e-09,
+      "loss": 0.3057,
+      "step": 14272
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.272231042463132,
+      "learning_rate": 1.5616732367751408e-09,
+      "loss": 0.3052,
+      "step": 14273
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.192462888335343,
+      "learning_rate": 1.5336631612555831e-09,
+      "loss": 0.2959,
+      "step": 14274
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.28204561401379097,
+      "learning_rate": 1.5059065189787502e-09,
+      "loss": 0.0998,
+      "step": 14275
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.4802987094791544,
+      "learning_rate": 1.4784033113518504e-09,
+      "loss": 0.2785,
+      "step": 14276
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.4544967232114026,
+      "learning_rate": 1.4511535397687682e-09,
+      "loss": 0.3487,
+      "step": 14277
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.582338341192454,
+      "learning_rate": 1.4241572056117315e-09,
+      "loss": 0.307,
+      "step": 14278
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.46132955858862,
+      "learning_rate": 1.39741431024909e-09,
+      "loss": 0.2694,
+      "step": 14279
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.277531884202145,
+      "learning_rate": 1.3709248550375365e-09,
+      "loss": 0.2958,
+      "step": 14280
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.714378479907026,
+      "learning_rate": 1.3446888413193304e-09,
+      "loss": 0.3091,
+      "step": 14281
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.0150123294387505,
+      "learning_rate": 1.318706270424519e-09,
+      "loss": 0.2748,
+      "step": 14282
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8766646084081473,
+      "learning_rate": 1.2929771436714921e-09,
+      "loss": 0.2878,
+      "step": 14283
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5715603816384975,
+      "learning_rate": 1.2675014623636518e-09,
+      "loss": 0.3434,
+      "step": 14284
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.819325598448268,
+      "learning_rate": 1.2422792277938523e-09,
+      "loss": 0.2936,
+      "step": 14285
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.6460628621976925,
+      "learning_rate": 1.2173104412399606e-09,
+      "loss": 0.3183,
+      "step": 14286
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.39513167960856,
+      "learning_rate": 1.192595103968186e-09,
+      "loss": 0.3309,
+      "step": 14287
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.1011171442782794,
+      "learning_rate": 1.1681332172319704e-09,
+      "loss": 0.3009,
+      "step": 14288
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.060180795872777,
+      "learning_rate": 1.1439247822714327e-09,
+      "loss": 0.3293,
+      "step": 14289
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7258569267520047,
+      "learning_rate": 1.1199698003133697e-09,
+      "loss": 0.3049,
+      "step": 14290
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 7.829747237607752,
+      "learning_rate": 1.0962682725734752e-09,
+      "loss": 0.3637,
+      "step": 14291
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 9.15142771586164,
+      "learning_rate": 1.0728202002524556e-09,
+      "loss": 0.2922,
+      "step": 14292
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.833605536107829,
+      "learning_rate": 1.049625584539915e-09,
+      "loss": 0.3143,
+      "step": 14293
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.677970658533137,
+      "learning_rate": 1.0266844266115794e-09,
+      "loss": 0.3193,
+      "step": 14294
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.875177991479585,
+      "learning_rate": 1.0039967276304075e-09,
+      "loss": 0.334,
+      "step": 14295
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.116772015556219,
+      "learning_rate": 9.815624887471454e-10,
+      "loss": 0.3266,
+      "step": 14296
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.121563089907779,
+      "learning_rate": 9.593817110992166e-10,
+      "loss": 0.289,
+      "step": 14297
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.065108177752184,
+      "learning_rate": 9.37454395810722e-10,
+      "loss": 0.3421,
+      "step": 14298
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.6257544583414636,
+      "learning_rate": 9.157805439935497e-10,
+      "loss": 0.2933,
+      "step": 14299
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.59209787099285,
+      "learning_rate": 8.943601567473759e-10,
+      "loss": 0.297,
+      "step": 14300
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.262717729820372,
+      "learning_rate": 8.731932351574435e-10,
+      "loss": 0.3211,
+      "step": 14301
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.331096373082132,
+      "learning_rate": 8.52279780296783e-10,
+      "loss": 0.2935,
+      "step": 14302
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.5961485611591266,
+      "learning_rate": 8.316197932267677e-10,
+      "loss": 0.3071,
+      "step": 14303
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.9349896603430565,
+      "learning_rate": 8.11213274994338e-10,
+      "loss": 0.2887,
+      "step": 14304
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.122213026634429,
+      "learning_rate": 7.910602266336664e-10,
+      "loss": 0.2742,
+      "step": 14305
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.688890873730926,
+      "learning_rate": 7.711606491672685e-10,
+      "loss": 0.3092,
+      "step": 14306
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.921250015643738,
+      "learning_rate": 7.515145436037819e-10,
+      "loss": 0.307,
+      "step": 14307
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.764104972464086,
+      "learning_rate": 7.321219109396316e-10,
+      "loss": 0.3039,
+      "step": 14308
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.825394645099777,
+      "learning_rate": 7.129827521573651e-10,
+      "loss": 0.305,
+      "step": 14309
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 4.490107869823988,
+      "learning_rate": 6.940970682284276e-10,
+      "loss": 0.2971,
+      "step": 14310
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 5.341263129646459,
+      "learning_rate": 6.754648601092761e-10,
+      "loss": 0.2747,
+      "step": 14311
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 6.227516948037512,
+      "learning_rate": 6.570861287452656e-10,
+      "loss": 0.3075,
+      "step": 14312
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.031138732723173,
+      "learning_rate": 6.389608750678733e-10,
+      "loss": 0.3222,
+      "step": 14313
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.770483095386821,
+      "learning_rate": 6.210890999969187e-10,
+      "loss": 0.3488,
+      "step": 14314
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.277762213778937,
+      "learning_rate": 6.034708044372339e-10,
+      "loss": 0.3266,
+      "step": 14315
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6391952890212425,
+      "learning_rate": 5.861059892831033e-10,
+      "loss": 0.3274,
+      "step": 14316
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.154570499823468,
+      "learning_rate": 5.689946554143788e-10,
+      "loss": 0.2938,
+      "step": 14317
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.737803164736359,
+      "learning_rate": 5.521368036992547e-10,
+      "loss": 0.3051,
+      "step": 14318
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.019698713930417,
+      "learning_rate": 5.355324349914926e-10,
+      "loss": 0.3159,
+      "step": 14319
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 50.509330338989564,
+      "learning_rate": 5.191815501343067e-10,
+      "loss": 0.3287,
+      "step": 14320
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.636211123010218,
+      "learning_rate": 5.030841499559235e-10,
+      "loss": 0.3389,
+      "step": 14321
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.419408348450087,
+      "learning_rate": 4.87240235272357e-10,
+      "loss": 0.3571,
+      "step": 14322
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.81691320435266,
+      "learning_rate": 4.716498068868536e-10,
+      "loss": 0.3125,
+      "step": 14323
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.276466955895566,
+      "learning_rate": 4.5631286559044694e-10,
+      "loss": 0.3654,
+      "step": 14324
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.468789384993602,
+      "learning_rate": 4.4122941216084846e-10,
+      "loss": 0.2862,
+      "step": 14325
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.028619275797311,
+      "learning_rate": 4.2639944736189156e-10,
+      "loss": 0.2901,
+      "step": 14326
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.082076616749459,
+      "learning_rate": 4.1182297194630736e-10,
+      "loss": 0.2916,
+      "step": 14327
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.755394650191386,
+      "learning_rate": 3.9749998665294944e-10,
+      "loss": 0.313,
+      "step": 14328
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.420910153866123,
+      "learning_rate": 3.834304922084586e-10,
+      "loss": 0.2906,
+      "step": 14329
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.354764602586362,
+      "learning_rate": 3.696144893250431e-10,
+      "loss": 0.3003,
+      "step": 14330
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.2454321218565845,
+      "learning_rate": 3.560519787043637e-10,
+      "loss": 0.2849,
+      "step": 14331
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.597825613982327,
+      "learning_rate": 3.4274296103309347e-10,
+      "loss": 0.311,
+      "step": 14332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.084610578142356,
+      "learning_rate": 3.2968743698680307e-10,
+      "loss": 0.2803,
+      "step": 14333
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.564181635374928,
+      "learning_rate": 3.1688540722718544e-10,
+      "loss": 0.3299,
+      "step": 14334
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.0293445812006947,
+      "learning_rate": 3.043368724031659e-10,
+      "loss": 0.293,
+      "step": 14335
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.904010830168639,
+      "learning_rate": 2.920418331514574e-10,
+      "loss": 0.3075,
+      "step": 14336
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.582747492548153,
+      "learning_rate": 2.8000029009489505e-10,
+      "loss": 0.3235,
+      "step": 14337
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.35130027469535,
+      "learning_rate": 2.682122438441015e-10,
+      "loss": 0.3218,
+      "step": 14338
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.622359555634369,
+      "learning_rate": 2.5667769499693184e-10,
+      "loss": 0.3386,
+      "step": 14339
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5400870281033847,
+      "learning_rate": 2.453966441379185e-10,
+      "loss": 0.3113,
+      "step": 14340
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.006376422807234,
+      "learning_rate": 2.3436909183993663e-10,
+      "loss": 0.3035,
+      "step": 14341
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.2954885501692726,
+      "learning_rate": 2.2359503866087318e-10,
+      "loss": 0.1085,
+      "step": 14342
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.2528219406521135,
+      "learning_rate": 2.1307448514806816e-10,
+      "loss": 0.3141,
+      "step": 14343
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.271664500451337,
+      "learning_rate": 2.0280743183387353e-10,
+      "loss": 0.3405,
+      "step": 14344
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6939934825738696,
+      "learning_rate": 1.9279387923953895e-10,
+      "loss": 0.2692,
+      "step": 14345
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2074862400112103,
+      "learning_rate": 1.830338278729915e-10,
+      "loss": 0.317,
+      "step": 14346
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.725869329432076,
+      "learning_rate": 1.7352727822883552e-10,
+      "loss": 0.3462,
+      "step": 14347
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.181866194382427,
+      "learning_rate": 1.642742307883527e-10,
+      "loss": 0.2863,
+      "step": 14348
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.972954356909312,
+      "learning_rate": 1.5527468602172246e-10,
+      "loss": 0.3284,
+      "step": 14349
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.3651938485729485,
+      "learning_rate": 1.465286443852465e-10,
+      "loss": 0.3492,
+      "step": 14350
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.6453921591430145,
+      "learning_rate": 1.3803610632134867e-10,
+      "loss": 0.2984,
+      "step": 14351
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.547895609979644,
+      "learning_rate": 1.2979707226135063e-10,
+      "loss": 0.2959,
+      "step": 14352
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.409949460573909,
+      "learning_rate": 1.2181154262325135e-10,
+      "loss": 0.2938,
+      "step": 14353
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.062252702215091,
+      "learning_rate": 1.1407951781117199e-10,
+      "loss": 0.2851,
+      "step": 14354
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 8.80952339201089,
+      "learning_rate": 1.0660099821757641e-10,
+      "loss": 0.2885,
+      "step": 14355
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1090095878659856,
+      "learning_rate": 9.937598422216088e-11,
+      "loss": 0.27,
+      "step": 14356
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.251908608098135,
+      "learning_rate": 9.240447619018878e-11,
+      "loss": 0.314,
+      "step": 14357
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.4397252781246497,
+      "learning_rate": 8.568647447526612e-11,
+      "loss": 0.3404,
+      "step": 14358
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.925335254377784,
+      "learning_rate": 7.922197941878651e-11,
+      "loss": 0.3162,
+      "step": 14359
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.357288551628367,
+      "learning_rate": 7.301099134826572e-11,
+      "loss": 0.3376,
+      "step": 14360
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.995317981970801,
+      "learning_rate": 6.705351057845199e-11,
+      "loss": 0.3329,
+      "step": 14361
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.513686166236453,
+      "learning_rate": 6.134953741132599e-11,
+      "loss": 0.3248,
+      "step": 14362
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.559555594051942,
+      "learning_rate": 5.58990721361008e-11,
+      "loss": 0.3039,
+      "step": 14363
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.889790788238704,
+      "learning_rate": 5.0702115028666845e-11,
+      "loss": 0.3425,
+      "step": 14364
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.837504263417373,
+      "learning_rate": 4.575866635381232e-11,
+      "loss": 0.2983,
+      "step": 14365
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.293822121950429,
+      "learning_rate": 4.1068726361337405e-11,
+      "loss": 0.2896,
+      "step": 14366
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.346682595696159,
+      "learning_rate": 3.6632295288829834e-11,
+      "loss": 0.3186,
+      "step": 14367
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26758615454094614,
+      "learning_rate": 3.244937336166487e-11,
+      "loss": 0.1079,
+      "step": 14368
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.62241480745851,
+      "learning_rate": 2.851996079189512e-11,
+      "loss": 0.2981,
+      "step": 14369
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.896180917362173,
+      "learning_rate": 2.4844057778250496e-11,
+      "loss": 0.3284,
+      "step": 14370
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.448922407672576,
+      "learning_rate": 2.1421664507803586e-11,
+      "loss": 0.3383,
+      "step": 14371
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1375223306571782,
+      "learning_rate": 1.8252781153749176e-11,
+      "loss": 0.319,
+      "step": 14372
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5709561040494657,
+      "learning_rate": 1.53374078765145e-11,
+      "loss": 0.3131,
+      "step": 14373
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.876867750581348,
+      "learning_rate": 1.2675544824314323e-11,
+      "loss": 0.3032,
+      "step": 14374
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.8552495853896485,
+      "learning_rate": 1.0267192132040749e-11,
+      "loss": 0.3086,
+      "step": 14375
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8826415281207476,
+      "learning_rate": 8.112349921263196e-12,
+      "loss": 0.3279,
+      "step": 14376
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.329868020187146,
+      "learning_rate": 6.211018302448857e-12,
+      "loss": 0.2962,
+      "step": 14377
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.711017234229576,
+      "learning_rate": 4.563197370521799e-12,
+      "loss": 0.3133,
+      "step": 14378
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.2589409315753657,
+      "learning_rate": 3.1688872104140846e-12,
+      "loss": 0.3145,
+      "step": 14379
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8696483882734274,
+      "learning_rate": 2.0280878915146517e-12,
+      "loss": 0.2789,
+      "step": 14380
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8770321226133575,
+      "learning_rate": 1.1407994726653215e-12,
+      "loss": 0.3098,
+      "step": 14381
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.8156093304918155,
+      "learning_rate": 5.070219988301261e-13,
+      "loss": 0.3148,
+      "step": 14382
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.7713925914546427,
+      "learning_rate": 1.2675550109531033e-13,
+      "loss": 0.2988,
+      "step": 14383
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0004280345094478,
+      "learning_rate": 0.0,
+      "loss": 0.1653,
+      "step": 14384
+    },
+    {
+      "epoch": 1.0,
+      "step": 14384,
+      "total_flos": 2.334980292512973e+16,
+      "train_loss": 0.123261748941551,
+      "train_runtime": 249048.0743,
+      "train_samples_per_second": 14.786,
+      "train_steps_per_second": 0.058
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 14384,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 2.334980292512973e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}