diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,89161 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 12733,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 45.35453070047731,
+      "learning_rate": 2.617801047120419e-08,
+      "loss": 2.0919,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 46.13214000486192,
+      "learning_rate": 5.235602094240838e-08,
+      "loss": 2.1655,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.256185844438896,
+      "learning_rate": 7.853403141361257e-08,
+      "loss": 2.1959,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.691128378686303,
+      "learning_rate": 1.0471204188481677e-07,
+      "loss": 0.8643,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 50.55032367486194,
+      "learning_rate": 1.3089005235602095e-07,
+      "loss": 2.2682,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.215536200795114,
+      "learning_rate": 1.5706806282722514e-07,
+      "loss": 2.1322,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 49.32478506639732,
+      "learning_rate": 1.8324607329842932e-07,
+      "loss": 2.2113,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 48.727801154955124,
+      "learning_rate": 2.0942408376963353e-07,
+      "loss": 2.2509,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 46.213727291809505,
+      "learning_rate": 2.3560209424083772e-07,
+      "loss": 2.268,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 46.9165533638362,
+      "learning_rate": 2.617801047120419e-07,
+      "loss": 2.136,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 47.513601900674225,
+      "learning_rate": 2.879581151832461e-07,
+      "loss": 2.1864,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 48.15897615141773,
+      "learning_rate": 3.1413612565445027e-07,
+      "loss": 2.1267,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 44.9728150083751,
+      "learning_rate": 3.403141361256545e-07,
+      "loss": 2.175,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 38.31471245525897,
+      "learning_rate": 3.6649214659685864e-07,
+      "loss": 2.0458,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.517769751470475,
+      "learning_rate": 3.926701570680629e-07,
+      "loss": 2.1302,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.03509277918861,
+      "learning_rate": 4.1884816753926706e-07,
+      "loss": 2.1028,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 35.90078278812605,
+      "learning_rate": 4.4502617801047125e-07,
+      "loss": 2.0215,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 36.6697157725411,
+      "learning_rate": 4.7120418848167543e-07,
+      "loss": 2.0382,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 25.06941191683323,
+      "learning_rate": 4.973821989528796e-07,
+      "loss": 1.8211,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 26.23231811919509,
+      "learning_rate": 5.235602094240838e-07,
+      "loss": 1.7987,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 23.06337189948916,
+      "learning_rate": 5.49738219895288e-07,
+      "loss": 1.6787,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.361321476963756,
+      "learning_rate": 5.759162303664922e-07,
+      "loss": 1.705,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.549608835731224,
+      "learning_rate": 6.020942408376964e-07,
+      "loss": 1.7105,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 21.021308074943473,
+      "learning_rate": 6.282722513089005e-07,
+      "loss": 1.6903,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 19.64468927723245,
+      "learning_rate": 6.544502617801048e-07,
+      "loss": 1.599,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.842243155236734,
+      "learning_rate": 6.80628272251309e-07,
+      "loss": 1.3371,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.007324116297383,
+      "learning_rate": 7.068062827225131e-07,
+      "loss": 1.2012,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.699921616286352,
+      "learning_rate": 7.329842931937173e-07,
+      "loss": 1.1503,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1694482285209635,
+      "learning_rate": 7.591623036649215e-07,
+      "loss": 0.8114,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.473830646625572,
+      "learning_rate": 7.853403141361258e-07,
+      "loss": 1.1225,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.1434185968423534,
+      "learning_rate": 8.115183246073299e-07,
+      "loss": 0.818,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.733538116772037,
+      "learning_rate": 8.376963350785341e-07,
+      "loss": 1.1269,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.034106740920896,
+      "learning_rate": 8.638743455497383e-07,
+      "loss": 1.0395,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8819983017729784,
+      "learning_rate": 8.900523560209425e-07,
+      "loss": 0.8208,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.565638901619074,
+      "learning_rate": 9.162303664921466e-07,
+      "loss": 1.0407,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.942797329758625,
+      "learning_rate": 9.424083769633509e-07,
+      "loss": 0.9814,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.204111590726132,
+      "learning_rate": 9.685863874345552e-07,
+      "loss": 0.9773,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.489121510119625,
+      "learning_rate": 9.947643979057591e-07,
+      "loss": 1.0226,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.9381644636937665,
+      "learning_rate": 1.0209424083769635e-06,
+      "loss": 0.9893,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.7947487895876972,
+      "learning_rate": 1.0471204188481676e-06,
+      "loss": 0.9293,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.813517039648087,
+      "learning_rate": 1.0732984293193717e-06,
+      "loss": 0.9161,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.661896737719922,
+      "learning_rate": 1.099476439790576e-06,
+      "loss": 0.9508,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7547648016055883,
+      "learning_rate": 1.1256544502617802e-06,
+      "loss": 0.844,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7526840475439416,
+      "learning_rate": 1.1518324607329843e-06,
+      "loss": 0.8643,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.653828512334836,
+      "learning_rate": 1.1780104712041885e-06,
+      "loss": 0.8822,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.10466377297939,
+      "learning_rate": 1.2041884816753928e-06,
+      "loss": 0.9145,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5673009092667383,
+      "learning_rate": 1.230366492146597e-06,
+      "loss": 0.8749,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.4764666653184566,
+      "learning_rate": 1.256544502617801e-06,
+      "loss": 0.8176,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.6013297216851865,
+      "learning_rate": 1.2827225130890052e-06,
+      "loss": 0.7885,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.29556227978767,
+      "learning_rate": 1.3089005235602096e-06,
+      "loss": 0.8863,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.5164595777894039,
+      "learning_rate": 1.3350785340314137e-06,
+      "loss": 0.7677,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.531990875446019,
+      "learning_rate": 1.361256544502618e-06,
+      "loss": 0.7966,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.575988250397976,
+      "learning_rate": 1.3874345549738222e-06,
+      "loss": 0.7997,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.3833393515964518,
+      "learning_rate": 1.4136125654450263e-06,
+      "loss": 0.7431,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.399781372632845,
+      "learning_rate": 1.4397905759162306e-06,
+      "loss": 0.7796,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.333718947214041,
+      "learning_rate": 1.4659685863874346e-06,
+      "loss": 0.7743,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.7429944804838637,
+      "learning_rate": 1.4921465968586387e-06,
+      "loss": 0.7777,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2671153309569032,
+      "learning_rate": 1.518324607329843e-06,
+      "loss": 0.7201,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.0254278124287546,
+      "learning_rate": 1.5445026178010472e-06,
+      "loss": 0.7478,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1766331642515933,
+      "learning_rate": 1.5706806282722515e-06,
+      "loss": 0.7655,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.2767456118987313,
+      "learning_rate": 1.5968586387434556e-06,
+      "loss": 0.755,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8077649837417682,
+      "learning_rate": 1.6230366492146598e-06,
+      "loss": 0.8457,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.8535791279862603,
+      "learning_rate": 1.6492146596858641e-06,
+      "loss": 0.7367,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8543924838381893,
+      "learning_rate": 1.6753926701570683e-06,
+      "loss": 0.8021,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1524070641273987,
+      "learning_rate": 1.7015706806282726e-06,
+      "loss": 0.801,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9329178296807856,
+      "learning_rate": 1.7277486910994765e-06,
+      "loss": 0.775,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.957473044576722,
+      "learning_rate": 1.7539267015706806e-06,
+      "loss": 0.7681,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1310805128679333,
+      "learning_rate": 1.780104712041885e-06,
+      "loss": 0.7694,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8353382725906293,
+      "learning_rate": 1.8062827225130891e-06,
+      "loss": 0.7818,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1697942415397358,
+      "learning_rate": 1.8324607329842933e-06,
+      "loss": 0.7907,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9731505521769248,
+      "learning_rate": 1.8586387434554976e-06,
+      "loss": 0.7442,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1352160467709305,
+      "learning_rate": 1.8848167539267017e-06,
+      "loss": 0.7367,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.206590331680779,
+      "learning_rate": 1.910994764397906e-06,
+      "loss": 0.7124,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8992891746324881,
+      "learning_rate": 1.9371727748691104e-06,
+      "loss": 0.722,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.118054922158826,
+      "learning_rate": 1.9633507853403143e-06,
+      "loss": 0.786,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1194976413857725,
+      "learning_rate": 1.9895287958115183e-06,
+      "loss": 0.7665,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2500958372978688,
+      "learning_rate": 2.0157068062827226e-06,
+      "loss": 0.7261,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9164092845719047,
+      "learning_rate": 2.041884816753927e-06,
+      "loss": 0.7227,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.027027307166561,
+      "learning_rate": 2.068062827225131e-06,
+      "loss": 0.7995,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6736602072120665,
+      "learning_rate": 2.094240837696335e-06,
+      "loss": 0.7129,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9629688978851059,
+      "learning_rate": 2.1204188481675396e-06,
+      "loss": 0.7553,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.083917113258539,
+      "learning_rate": 2.1465968586387435e-06,
+      "loss": 0.7382,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9663694967419423,
+      "learning_rate": 2.172774869109948e-06,
+      "loss": 0.7549,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8805571070809088,
+      "learning_rate": 2.198952879581152e-06,
+      "loss": 0.7458,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9954794692218607,
+      "learning_rate": 2.2251308900523565e-06,
+      "loss": 0.6968,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9157681391283916,
+      "learning_rate": 2.2513089005235604e-06,
+      "loss": 0.7518,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.872223663895253,
+      "learning_rate": 2.2774869109947643e-06,
+      "loss": 0.6708,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.969999877582168,
+      "learning_rate": 2.3036649214659687e-06,
+      "loss": 0.6439,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.858958355519399,
+      "learning_rate": 2.329842931937173e-06,
+      "loss": 0.7436,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3418565405365506,
+      "learning_rate": 2.356020942408377e-06,
+      "loss": 0.7478,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7936193337983721,
+      "learning_rate": 2.3821989528795813e-06,
+      "loss": 0.7125,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.892150276071557,
+      "learning_rate": 2.4083769633507856e-06,
+      "loss": 0.7338,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.097308265278116,
+      "learning_rate": 2.43455497382199e-06,
+      "loss": 0.7129,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.072006523639674,
+      "learning_rate": 2.460732984293194e-06,
+      "loss": 0.671,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0785238995907087,
+      "learning_rate": 2.4869109947643982e-06,
+      "loss": 0.7266,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8997777764978196,
+      "learning_rate": 2.513089005235602e-06,
+      "loss": 0.6566,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.014901719738401,
+      "learning_rate": 2.5392670157068065e-06,
+      "loss": 0.7186,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.370922928969181,
+      "learning_rate": 2.5654450261780104e-06,
+      "loss": 0.658,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7807101146304956,
+      "learning_rate": 2.591623036649215e-06,
+      "loss": 0.6483,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8367424093743598,
+      "learning_rate": 2.617801047120419e-06,
+      "loss": 0.7254,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.926337759300584,
+      "learning_rate": 2.643979057591623e-06,
+      "loss": 0.7583,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9956710385213385,
+      "learning_rate": 2.6701570680628274e-06,
+      "loss": 0.6249,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.791609885701657,
+      "learning_rate": 2.6963350785340313e-06,
+      "loss": 0.6928,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9518302967935883,
+      "learning_rate": 2.722513089005236e-06,
+      "loss": 0.678,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.039064240931749,
+      "learning_rate": 2.74869109947644e-06,
+      "loss": 0.7056,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3678913250403104,
+      "learning_rate": 2.7748691099476443e-06,
+      "loss": 0.6715,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9267736176367727,
+      "learning_rate": 2.8010471204188483e-06,
+      "loss": 0.7046,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8636531425355595,
+      "learning_rate": 2.8272251308900526e-06,
+      "loss": 0.701,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0604628063700496,
+      "learning_rate": 2.853403141361257e-06,
+      "loss": 0.7115,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7930819856262614,
+      "learning_rate": 2.8795811518324613e-06,
+      "loss": 0.6732,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.105418395736594,
+      "learning_rate": 2.905759162303665e-06,
+      "loss": 0.6431,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8189305880342908,
+      "learning_rate": 2.931937172774869e-06,
+      "loss": 0.6296,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8235003667751857,
+      "learning_rate": 2.9581151832460735e-06,
+      "loss": 0.7217,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2766271969490917,
+      "learning_rate": 2.9842931937172774e-06,
+      "loss": 0.6682,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8820752427366934,
+      "learning_rate": 3.010471204188482e-06,
+      "loss": 0.6738,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2754682331303964,
+      "learning_rate": 3.036649214659686e-06,
+      "loss": 0.7329,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8743072977229744,
+      "learning_rate": 3.0628272251308904e-06,
+      "loss": 0.6589,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1694655485962713,
+      "learning_rate": 3.0890052356020943e-06,
+      "loss": 0.6579,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0199711618861205,
+      "learning_rate": 3.115183246073299e-06,
+      "loss": 0.6807,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1292929227860204,
+      "learning_rate": 3.141361256544503e-06,
+      "loss": 0.6724,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1393759446029326,
+      "learning_rate": 3.167539267015707e-06,
+      "loss": 0.6673,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3856643323755993,
+      "learning_rate": 3.1937172774869113e-06,
+      "loss": 0.6203,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.561140903379722,
+      "learning_rate": 3.219895287958115e-06,
+      "loss": 0.6751,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7662570854834847,
+      "learning_rate": 3.2460732984293196e-06,
+      "loss": 0.6614,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8324500732709519,
+      "learning_rate": 3.2722513089005235e-06,
+      "loss": 0.6513,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7679113291895352,
+      "learning_rate": 3.2984293193717282e-06,
+      "loss": 0.6594,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.951173192966468,
+      "learning_rate": 3.324607329842932e-06,
+      "loss": 0.6229,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.333074081486533,
+      "learning_rate": 3.3507853403141365e-06,
+      "loss": 0.595,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.881689358329475,
+      "learning_rate": 3.3769633507853404e-06,
+      "loss": 0.6513,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.070975299362807,
+      "learning_rate": 3.403141361256545e-06,
+      "loss": 0.614,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.206312931997375,
+      "learning_rate": 3.429319371727749e-06,
+      "loss": 0.6907,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7629747669733618,
+      "learning_rate": 3.455497382198953e-06,
+      "loss": 0.6602,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9321217548759493,
+      "learning_rate": 3.4816753926701574e-06,
+      "loss": 0.6575,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9808782435471375,
+      "learning_rate": 3.5078534031413613e-06,
+      "loss": 0.6389,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.131088425682013,
+      "learning_rate": 3.534031413612566e-06,
+      "loss": 0.6231,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.028134267184075,
+      "learning_rate": 3.56020942408377e-06,
+      "loss": 0.6508,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.138081828542439,
+      "learning_rate": 3.5863874345549743e-06,
+      "loss": 0.6456,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.716613124461599,
+      "learning_rate": 3.6125654450261782e-06,
+      "loss": 0.6142,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.023720585340836,
+      "learning_rate": 3.6387434554973826e-06,
+      "loss": 0.6671,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 10.194023941680866,
+      "learning_rate": 3.6649214659685865e-06,
+      "loss": 0.6451,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9972966870883009,
+      "learning_rate": 3.6910994764397904e-06,
+      "loss": 0.6319,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0212654702091317,
+      "learning_rate": 3.717277486910995e-06,
+      "loss": 0.6509,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9433165276775327,
+      "learning_rate": 3.743455497382199e-06,
+      "loss": 0.6993,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1309731395963945,
+      "learning_rate": 3.7696335078534035e-06,
+      "loss": 0.7071,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8921011279008414,
+      "learning_rate": 3.7958115183246074e-06,
+      "loss": 0.709,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6366436090625625,
+      "learning_rate": 3.821989528795812e-06,
+      "loss": 0.6248,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.857571469328871,
+      "learning_rate": 3.848167539267016e-06,
+      "loss": 0.6901,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7736744769520083,
+      "learning_rate": 3.874345549738221e-06,
+      "loss": 0.6689,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2598614397167878,
+      "learning_rate": 3.900523560209425e-06,
+      "loss": 0.644,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7893197071307048,
+      "learning_rate": 3.926701570680629e-06,
+      "loss": 0.6462,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6474949981268692,
+      "learning_rate": 3.952879581151833e-06,
+      "loss": 0.6007,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.999142759485342,
+      "learning_rate": 3.9790575916230365e-06,
+      "loss": 0.6089,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.3139743873530603,
+      "learning_rate": 4.005235602094241e-06,
+      "loss": 0.6537,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8846663774879808,
+      "learning_rate": 4.031413612565445e-06,
+      "loss": 0.6639,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.734985030227232,
+      "learning_rate": 4.05759162303665e-06,
+      "loss": 0.6616,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5920569033949876,
+      "learning_rate": 4.083769633507854e-06,
+      "loss": 0.6046,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6491730373484021,
+      "learning_rate": 4.109947643979058e-06,
+      "loss": 0.5964,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.586006489836427,
+      "learning_rate": 4.136125654450262e-06,
+      "loss": 0.5764,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.10847911947796,
+      "learning_rate": 4.1623036649214665e-06,
+      "loss": 0.6442,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7830649184861866,
+      "learning_rate": 4.18848167539267e-06,
+      "loss": 0.5902,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5904815875916594,
+      "learning_rate": 4.214659685863874e-06,
+      "loss": 0.6195,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6920124196984998,
+      "learning_rate": 4.240837696335079e-06,
+      "loss": 0.6312,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8741205082859589,
+      "learning_rate": 4.267015706806283e-06,
+      "loss": 0.6279,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4532328049185614,
+      "learning_rate": 4.293193717277487e-06,
+      "loss": 0.6587,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7725115371560787,
+      "learning_rate": 4.319371727748692e-06,
+      "loss": 0.5951,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6146636137372306,
+      "learning_rate": 4.345549738219896e-06,
+      "loss": 0.6257,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2136877777522312,
+      "learning_rate": 4.3717277486910996e-06,
+      "loss": 0.6117,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 5.268881685131859,
+      "learning_rate": 4.397905759162304e-06,
+      "loss": 0.6686,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.696900463368965,
+      "learning_rate": 4.424083769633508e-06,
+      "loss": 0.7198,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7059533677803898,
+      "learning_rate": 4.450261780104713e-06,
+      "loss": 0.5859,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.44624084635017,
+      "learning_rate": 4.476439790575917e-06,
+      "loss": 0.6901,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2909153639175632,
+      "learning_rate": 4.502617801047121e-06,
+      "loss": 0.6767,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7291815189262174,
+      "learning_rate": 4.528795811518325e-06,
+      "loss": 0.6618,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9162523229709274,
+      "learning_rate": 4.554973821989529e-06,
+      "loss": 0.6349,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.469418735305691,
+      "learning_rate": 4.5811518324607335e-06,
+      "loss": 0.5924,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6602749230145328,
+      "learning_rate": 4.607329842931937e-06,
+      "loss": 0.6785,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0646182832653475,
+      "learning_rate": 4.633507853403142e-06,
+      "loss": 0.6659,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.888560880998048,
+      "learning_rate": 4.659685863874346e-06,
+      "loss": 0.6238,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2806725046124194,
+      "learning_rate": 4.68586387434555e-06,
+      "loss": 0.7167,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.2385032657843675,
+      "learning_rate": 4.712041884816754e-06,
+      "loss": 0.6622,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.705931243861708,
+      "learning_rate": 4.738219895287958e-06,
+      "loss": 0.6014,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.741836552376092,
+      "learning_rate": 4.764397905759163e-06,
+      "loss": 0.6766,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1663446335535017,
+      "learning_rate": 4.7905759162303665e-06,
+      "loss": 0.5637,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.2821381374598895,
+      "learning_rate": 4.816753926701571e-06,
+      "loss": 0.5777,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6867896701604612,
+      "learning_rate": 4.842931937172775e-06,
+      "loss": 0.672,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7084110913182926,
+      "learning_rate": 4.86910994764398e-06,
+      "loss": 0.6199,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.37314978140354,
+      "learning_rate": 4.895287958115184e-06,
+      "loss": 0.6857,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.91825279061721,
+      "learning_rate": 4.921465968586388e-06,
+      "loss": 0.5904,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0749995402993258,
+      "learning_rate": 4.947643979057592e-06,
+      "loss": 0.6254,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.5100045738950123,
+      "learning_rate": 4.9738219895287965e-06,
+      "loss": 0.611,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7371943140092654,
+      "learning_rate": 5e-06,
+      "loss": 0.537,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4153437448622332,
+      "learning_rate": 5.026178010471204e-06,
+      "loss": 0.6771,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6508481797020853,
+      "learning_rate": 5.052356020942408e-06,
+      "loss": 0.6298,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7882164194005299,
+      "learning_rate": 5.078534031413613e-06,
+      "loss": 0.6204,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9525946335786866,
+      "learning_rate": 5.104712041884817e-06,
+      "loss": 0.5357,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.957509212074309,
+      "learning_rate": 5.130890052356021e-06,
+      "loss": 0.6606,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0018704189443026,
+      "learning_rate": 5.157068062827225e-06,
+      "loss": 0.613,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.14760935743133,
+      "learning_rate": 5.18324607329843e-06,
+      "loss": 0.6183,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9544711320036594,
+      "learning_rate": 5.209424083769634e-06,
+      "loss": 0.6021,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4912528084584915,
+      "learning_rate": 5.235602094240838e-06,
+      "loss": 0.5545,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9315115703051489,
+      "learning_rate": 5.261780104712042e-06,
+      "loss": 0.6131,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7251473282650897,
+      "learning_rate": 5.287958115183246e-06,
+      "loss": 0.6768,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.693954234080208,
+      "learning_rate": 5.314136125654451e-06,
+      "loss": 0.6137,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7756818197281292,
+      "learning_rate": 5.340314136125655e-06,
+      "loss": 0.5806,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 13.078500588517276,
+      "learning_rate": 5.366492146596859e-06,
+      "loss": 0.6225,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.021719702575313,
+      "learning_rate": 5.392670157068063e-06,
+      "loss": 0.6309,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7220006778453985,
+      "learning_rate": 5.418848167539268e-06,
+      "loss": 0.6425,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.8813082062793094,
+      "learning_rate": 5.445026178010472e-06,
+      "loss": 0.6765,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7665225877764177,
+      "learning_rate": 5.471204188481676e-06,
+      "loss": 0.5653,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2776931345002365,
+      "learning_rate": 5.49738219895288e-06,
+      "loss": 0.5794,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8439308712146631,
+      "learning_rate": 5.523560209424085e-06,
+      "loss": 0.6702,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.695772975501985,
+      "learning_rate": 5.549738219895289e-06,
+      "loss": 0.5512,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0913982050412083,
+      "learning_rate": 5.575916230366493e-06,
+      "loss": 0.6256,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1507122257592375,
+      "learning_rate": 5.6020942408376965e-06,
+      "loss": 0.5894,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.45220638799841,
+      "learning_rate": 5.6282722513089e-06,
+      "loss": 0.6211,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7144943444245992,
+      "learning_rate": 5.654450261780105e-06,
+      "loss": 0.6629,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9247248310532166,
+      "learning_rate": 5.680628272251309e-06,
+      "loss": 0.6102,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1066955967685352,
+      "learning_rate": 5.706806282722514e-06,
+      "loss": 0.6287,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1029979108896297,
+      "learning_rate": 5.732984293193718e-06,
+      "loss": 0.6234,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.894030012191344,
+      "learning_rate": 5.7591623036649226e-06,
+      "loss": 0.591,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.073173112916301,
+      "learning_rate": 5.7853403141361265e-06,
+      "loss": 0.6153,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7002274519542699,
+      "learning_rate": 5.81151832460733e-06,
+      "loss": 0.6592,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7930017934151798,
+      "learning_rate": 5.837696335078534e-06,
+      "loss": 0.6153,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.475914951564157,
+      "learning_rate": 5.863874345549738e-06,
+      "loss": 0.5637,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.741749162713132,
+      "learning_rate": 5.890052356020943e-06,
+      "loss": 0.62,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7759720431129993,
+      "learning_rate": 5.916230366492147e-06,
+      "loss": 0.6693,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7343394617034111,
+      "learning_rate": 5.942408376963351e-06,
+      "loss": 0.6367,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.793166057087866,
+      "learning_rate": 5.968586387434555e-06,
+      "loss": 0.5872,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.4545737450862313,
+      "learning_rate": 5.99476439790576e-06,
+      "loss": 0.5685,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.4463504497293354,
+      "learning_rate": 6.020942408376964e-06,
+      "loss": 0.6142,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.382576115967975,
+      "learning_rate": 6.047120418848168e-06,
+      "loss": 0.6581,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7305290249804952,
+      "learning_rate": 6.073298429319372e-06,
+      "loss": 0.6467,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8490308149622259,
+      "learning_rate": 6.099476439790576e-06,
+      "loss": 0.6647,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1960310676592774,
+      "learning_rate": 6.125654450261781e-06,
+      "loss": 0.6218,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7694383713956765,
+      "learning_rate": 6.151832460732985e-06,
+      "loss": 0.5273,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6739184255103511,
+      "learning_rate": 6.178010471204189e-06,
+      "loss": 0.6381,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0293083758240877,
+      "learning_rate": 6.204188481675393e-06,
+      "loss": 0.5907,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.566980184009032,
+      "learning_rate": 6.230366492146598e-06,
+      "loss": 0.5775,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.419682987980466,
+      "learning_rate": 6.256544502617802e-06,
+      "loss": 0.6018,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7971667174279409,
+      "learning_rate": 6.282722513089006e-06,
+      "loss": 0.6207,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8222602544518638,
+      "learning_rate": 6.30890052356021e-06,
+      "loss": 0.5495,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9655782418817094,
+      "learning_rate": 6.335078534031414e-06,
+      "loss": 0.5693,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6491065496522279,
+      "learning_rate": 6.361256544502619e-06,
+      "loss": 0.66,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.7489167017238296,
+      "learning_rate": 6.3874345549738226e-06,
+      "loss": 0.5976,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7879946736689729,
+      "learning_rate": 6.4136125654450265e-06,
+      "loss": 0.6461,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8619264996832411,
+      "learning_rate": 6.43979057591623e-06,
+      "loss": 0.6521,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9410176997917161,
+      "learning_rate": 6.465968586387435e-06,
+      "loss": 0.611,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7008009225139062,
+      "learning_rate": 6.492146596858639e-06,
+      "loss": 0.626,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.189246698862531,
+      "learning_rate": 6.518324607329843e-06,
+      "loss": 0.5695,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.53575826182709,
+      "learning_rate": 6.544502617801047e-06,
+      "loss": 0.6037,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7095931613050814,
+      "learning_rate": 6.5706806282722526e-06,
+      "loss": 0.6386,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9137628661884745,
+      "learning_rate": 6.5968586387434565e-06,
+      "loss": 0.5649,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9677194691877633,
+      "learning_rate": 6.62303664921466e-06,
+      "loss": 0.6404,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8663511932921686,
+      "learning_rate": 6.649214659685864e-06,
+      "loss": 0.5249,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0004844468630747,
+      "learning_rate": 6.675392670157068e-06,
+      "loss": 0.6263,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.099160308105679,
+      "learning_rate": 6.701570680628273e-06,
+      "loss": 0.6105,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.392738953440951,
+      "learning_rate": 6.727748691099477e-06,
+      "loss": 0.5954,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0355276124794885,
+      "learning_rate": 6.753926701570681e-06,
+      "loss": 0.6671,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.03458385969778,
+      "learning_rate": 6.780104712041885e-06,
+      "loss": 0.5604,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.8883597384286643,
+      "learning_rate": 6.80628272251309e-06,
+      "loss": 0.5639,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.866747787672583,
+      "learning_rate": 6.832460732984294e-06,
+      "loss": 0.6616,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.903986982294793,
+      "learning_rate": 6.858638743455498e-06,
+      "loss": 0.5886,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.073994782101355,
+      "learning_rate": 6.884816753926702e-06,
+      "loss": 0.6262,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.753531478359543,
+      "learning_rate": 6.910994764397906e-06,
+      "loss": 0.6491,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9269080468807167,
+      "learning_rate": 6.937172774869111e-06,
+      "loss": 0.6587,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8591959193729306,
+      "learning_rate": 6.963350785340315e-06,
+      "loss": 0.575,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7600350309134827,
+      "learning_rate": 6.989528795811519e-06,
+      "loss": 0.6398,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7602001182676164,
+      "learning_rate": 7.015706806282723e-06,
+      "loss": 0.6146,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8551764949302686,
+      "learning_rate": 7.041884816753927e-06,
+      "loss": 0.6504,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.0024281431622826,
+      "learning_rate": 7.068062827225132e-06,
+      "loss": 0.6621,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2210609238450276,
+      "learning_rate": 7.094240837696336e-06,
+      "loss": 0.552,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8397689913412292,
+      "learning_rate": 7.12041884816754e-06,
+      "loss": 0.6015,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.218150729643147,
+      "learning_rate": 7.146596858638744e-06,
+      "loss": 0.6113,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1596853885393013,
+      "learning_rate": 7.172774869109949e-06,
+      "loss": 0.636,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.668983011893593,
+      "learning_rate": 7.1989528795811526e-06,
+      "loss": 0.5872,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5893830057802592,
+      "learning_rate": 7.2251308900523565e-06,
+      "loss": 0.5862,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.5747322565354605,
+      "learning_rate": 7.25130890052356e-06,
+      "loss": 0.6201,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8829488599479214,
+      "learning_rate": 7.277486910994765e-06,
+      "loss": 0.5457,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.336956899504294,
+      "learning_rate": 7.303664921465969e-06,
+      "loss": 0.6038,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7291076352632406,
+      "learning_rate": 7.329842931937173e-06,
+      "loss": 0.5971,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9847124671832421,
+      "learning_rate": 7.356020942408377e-06,
+      "loss": 0.6058,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.798704849334657,
+      "learning_rate": 7.382198952879581e-06,
+      "loss": 0.6312,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7331834478419776,
+      "learning_rate": 7.4083769633507865e-06,
+      "loss": 0.6406,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.564801529458803,
+      "learning_rate": 7.43455497382199e-06,
+      "loss": 0.5365,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7463328102533797,
+      "learning_rate": 7.460732984293194e-06,
+      "loss": 0.5818,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.9650635818556155,
+      "learning_rate": 7.486910994764398e-06,
+      "loss": 0.577,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.970520728459688,
+      "learning_rate": 7.513089005235603e-06,
+      "loss": 0.5818,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8540747283645477,
+      "learning_rate": 7.539267015706807e-06,
+      "loss": 0.6753,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1081833360174698,
+      "learning_rate": 7.565445026178011e-06,
+      "loss": 0.5891,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0117508764811607,
+      "learning_rate": 7.591623036649215e-06,
+      "loss": 0.6295,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6970549536269707,
+      "learning_rate": 7.61780104712042e-06,
+      "loss": 0.6755,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0062637312225844,
+      "learning_rate": 7.643979057591624e-06,
+      "loss": 0.5686,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0362096093387865,
+      "learning_rate": 7.670157068062828e-06,
+      "loss": 0.5673,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8131146598326235,
+      "learning_rate": 7.696335078534032e-06,
+      "loss": 0.6225,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.40117781779116,
+      "learning_rate": 7.722513089005236e-06,
+      "loss": 0.6092,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.942941447243737,
+      "learning_rate": 7.748691099476442e-06,
+      "loss": 0.5567,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.500181078589944,
+      "learning_rate": 7.774869109947646e-06,
+      "loss": 0.591,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2251963124773035,
+      "learning_rate": 7.80104712041885e-06,
+      "loss": 0.6031,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0529066305570653,
+      "learning_rate": 7.827225130890053e-06,
+      "loss": 0.6128,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 3.090935404900086,
+      "learning_rate": 7.853403141361257e-06,
+      "loss": 0.5996,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9555027063866162,
+      "learning_rate": 7.879581151832461e-06,
+      "loss": 0.605,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0087905370864692,
+      "learning_rate": 7.905759162303665e-06,
+      "loss": 0.6811,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7043236552760113,
+      "learning_rate": 7.931937172774869e-06,
+      "loss": 0.6454,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7342801484676242,
+      "learning_rate": 7.958115183246073e-06,
+      "loss": 0.5983,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.065925420547359,
+      "learning_rate": 7.984293193717279e-06,
+      "loss": 0.6141,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.3265166806831226,
+      "learning_rate": 8.010471204188483e-06,
+      "loss": 0.575,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1327194566548098,
+      "learning_rate": 8.036649214659686e-06,
+      "loss": 0.6017,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.70623402273943,
+      "learning_rate": 8.06282722513089e-06,
+      "loss": 0.6197,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1456416045219573,
+      "learning_rate": 8.089005235602096e-06,
+      "loss": 0.5736,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.1804800221288447,
+      "learning_rate": 8.1151832460733e-06,
+      "loss": 0.5494,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4974476877105807,
+      "learning_rate": 8.141361256544504e-06,
+      "loss": 0.6817,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6592394252262372,
+      "learning_rate": 8.167539267015708e-06,
+      "loss": 0.5578,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.7337570555432884,
+      "learning_rate": 8.193717277486912e-06,
+      "loss": 0.6749,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.760211895414078,
+      "learning_rate": 8.219895287958116e-06,
+      "loss": 0.5212,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9209381045760696,
+      "learning_rate": 8.24607329842932e-06,
+      "loss": 0.6555,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8331572579003792,
+      "learning_rate": 8.272251308900523e-06,
+      "loss": 0.5658,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.2183719228535876,
+      "learning_rate": 8.298429319371727e-06,
+      "loss": 0.538,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.303546757300317,
+      "learning_rate": 8.324607329842933e-06,
+      "loss": 0.575,
+      "step": 318
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.4077826408323855,
+      "learning_rate": 8.350785340314137e-06,
+      "loss": 0.6142,
+      "step": 319
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7345594508031987,
+      "learning_rate": 8.37696335078534e-06,
+      "loss": 0.6065,
+      "step": 320
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9304800395881956,
+      "learning_rate": 8.403141361256545e-06,
+      "loss": 0.5618,
+      "step": 321
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.5968679781572175,
+      "learning_rate": 8.429319371727749e-06,
+      "loss": 0.5563,
+      "step": 322
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.006032645539263,
+      "learning_rate": 8.455497382198954e-06,
+      "loss": 0.536,
+      "step": 323
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0156420776757384,
+      "learning_rate": 8.481675392670158e-06,
+      "loss": 0.5946,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3700939632147118,
+      "learning_rate": 8.507853403141362e-06,
+      "loss": 0.6774,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7536817452775255,
+      "learning_rate": 8.534031413612566e-06,
+      "loss": 0.5735,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.7216317155309775,
+      "learning_rate": 8.56020942408377e-06,
+      "loss": 0.601,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8906173526863174,
+      "learning_rate": 8.586387434554974e-06,
+      "loss": 0.644,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7774582105780772,
+      "learning_rate": 8.612565445026178e-06,
+      "loss": 0.5895,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.633526331686594,
+      "learning_rate": 8.638743455497383e-06,
+      "loss": 0.587,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9678423695374159,
+      "learning_rate": 8.664921465968587e-06,
+      "loss": 0.5704,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8876313906812718,
+      "learning_rate": 8.691099476439791e-06,
+      "loss": 0.6116,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.796173396005943,
+      "learning_rate": 8.717277486910995e-06,
+      "loss": 0.6238,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.836606460764124,
+      "learning_rate": 8.743455497382199e-06,
+      "loss": 0.5695,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9758437638543772,
+      "learning_rate": 8.769633507853403e-06,
+      "loss": 0.6644,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.560046653526461,
+      "learning_rate": 8.795811518324609e-06,
+      "loss": 0.5174,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8161507847683459,
+      "learning_rate": 8.821989528795813e-06,
+      "loss": 0.6484,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 4.276855610138708,
+      "learning_rate": 8.848167539267016e-06,
+      "loss": 0.5425,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9185312861963857,
+      "learning_rate": 8.87434554973822e-06,
+      "loss": 0.5879,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8551708136535057,
+      "learning_rate": 8.900523560209426e-06,
+      "loss": 0.5835,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8561936713146543,
+      "learning_rate": 8.92670157068063e-06,
+      "loss": 0.5626,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.034340329326399,
+      "learning_rate": 8.952879581151834e-06,
+      "loss": 0.6255,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9914305572347764,
+      "learning_rate": 8.979057591623038e-06,
+      "loss": 0.5705,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.6099638033601917,
+      "learning_rate": 9.005235602094242e-06,
+      "loss": 0.6908,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0050492959636452,
+      "learning_rate": 9.031413612565446e-06,
+      "loss": 0.5644,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1737043159832683,
+      "learning_rate": 9.05759162303665e-06,
+      "loss": 0.5373,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0136487689349023,
+      "learning_rate": 9.083769633507853e-06,
+      "loss": 0.5364,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.002767248285591,
+      "learning_rate": 9.109947643979057e-06,
+      "loss": 0.5913,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.542235705571547,
+      "learning_rate": 9.136125654450263e-06,
+      "loss": 0.563,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1967348149858994,
+      "learning_rate": 9.162303664921467e-06,
+      "loss": 0.5827,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3565261729014044,
+      "learning_rate": 9.18848167539267e-06,
+      "loss": 0.6061,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6375159468393599,
+      "learning_rate": 9.214659685863875e-06,
+      "loss": 0.5575,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6534083574327783,
+      "learning_rate": 9.240837696335079e-06,
+      "loss": 0.585,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.331897606240619,
+      "learning_rate": 9.267015706806284e-06,
+      "loss": 0.6553,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8143977084759415,
+      "learning_rate": 9.293193717277488e-06,
+      "loss": 0.5698,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.106723226375449,
+      "learning_rate": 9.319371727748692e-06,
+      "loss": 0.5792,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.832056961026013,
+      "learning_rate": 9.345549738219896e-06,
+      "loss": 0.6394,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9451383298091462,
+      "learning_rate": 9.3717277486911e-06,
+      "loss": 0.634,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9579650730860603,
+      "learning_rate": 9.397905759162304e-06,
+      "loss": 0.5958,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7314640239068833,
+      "learning_rate": 9.424083769633508e-06,
+      "loss": 0.588,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.054137365716144,
+      "learning_rate": 9.450261780104712e-06,
+      "loss": 0.6334,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0649400917957876,
+      "learning_rate": 9.476439790575916e-06,
+      "loss": 0.6584,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0675860262051775,
+      "learning_rate": 9.502617801047121e-06,
+      "loss": 0.5991,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8552529365043544,
+      "learning_rate": 9.528795811518325e-06,
+      "loss": 0.601,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.442538081278847,
+      "learning_rate": 9.554973821989529e-06,
+      "loss": 0.5867,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.773274313259584,
+      "learning_rate": 9.581151832460733e-06,
+      "loss": 0.5683,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.899424417954219,
+      "learning_rate": 9.607329842931939e-06,
+      "loss": 0.5813,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8323019750385396,
+      "learning_rate": 9.633507853403143e-06,
+      "loss": 0.5988,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.275918024752598,
+      "learning_rate": 9.659685863874346e-06,
+      "loss": 0.5755,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8062118446256774,
+      "learning_rate": 9.68586387434555e-06,
+      "loss": 0.531,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.5191468664125214,
+      "learning_rate": 9.712041884816756e-06,
+      "loss": 0.569,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.3224071179207275,
+      "learning_rate": 9.73821989528796e-06,
+      "loss": 0.6412,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8788417268521862,
+      "learning_rate": 9.764397905759164e-06,
+      "loss": 0.6379,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0602449040659305,
+      "learning_rate": 9.790575916230368e-06,
+      "loss": 0.5902,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.292427056233333,
+      "learning_rate": 9.816753926701572e-06,
+      "loss": 0.5581,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.013468798676717,
+      "learning_rate": 9.842931937172776e-06,
+      "loss": 0.5499,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8980067534882563,
+      "learning_rate": 9.86910994764398e-06,
+      "loss": 0.5419,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6982130201462555,
+      "learning_rate": 9.895287958115183e-06,
+      "loss": 0.5613,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8559293426571322,
+      "learning_rate": 9.921465968586387e-06,
+      "loss": 0.6029,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6357801187513554,
+      "learning_rate": 9.947643979057593e-06,
+      "loss": 0.552,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.153690087246258,
+      "learning_rate": 9.973821989528797e-06,
+      "loss": 0.5801,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9724041677639783,
+      "learning_rate": 1e-05,
+      "loss": 0.5646,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5976335935631814,
+      "learning_rate": 9.999999838253271e-06,
+      "loss": 0.5556,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6554892049292345,
+      "learning_rate": 9.999999353013093e-06,
+      "loss": 0.5392,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6887486379737568,
+      "learning_rate": 9.999998544279496e-06,
+      "loss": 0.641,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7935812618030595,
+      "learning_rate": 9.999997412052538e-06,
+      "loss": 0.6678,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.872627525790363,
+      "learning_rate": 9.999995956332285e-06,
+      "loss": 0.6358,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8964959524243654,
+      "learning_rate": 9.999994177118834e-06,
+      "loss": 0.5797,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.605241461550817,
+      "learning_rate": 9.9999920744123e-06,
+      "loss": 0.6252,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0935016939752162,
+      "learning_rate": 9.999989648212823e-06,
+      "loss": 0.5725,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9617880813371835,
+      "learning_rate": 9.999986898520556e-06,
+      "loss": 0.5966,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.864885226261083,
+      "learning_rate": 9.999983825335676e-06,
+      "loss": 0.5576,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8079353702078496,
+      "learning_rate": 9.999980428658383e-06,
+      "loss": 0.5538,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8796330905646186,
+      "learning_rate": 9.999976708488898e-06,
+      "loss": 0.5572,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.781206713302281,
+      "learning_rate": 9.99997266482746e-06,
+      "loss": 0.5743,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7108591015121346,
+      "learning_rate": 9.999968297674332e-06,
+      "loss": 0.6134,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 8.211612047212949,
+      "learning_rate": 9.999963607029795e-06,
+      "loss": 0.6146,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6435930713590208,
+      "learning_rate": 9.999958592894155e-06,
+      "loss": 0.5371,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9552178045133713,
+      "learning_rate": 9.999953255267733e-06,
+      "loss": 0.6656,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.726505569911282,
+      "learning_rate": 9.999947594150877e-06,
+      "loss": 0.5598,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.6077916866358,
+      "learning_rate": 9.999941609543953e-06,
+      "loss": 0.5931,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8457085357305854,
+      "learning_rate": 9.999935301447348e-06,
+      "loss": 0.5685,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.8357313863810902,
+      "learning_rate": 9.999928669861467e-06,
+      "loss": 0.593,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6322761612349168,
+      "learning_rate": 9.999921714786745e-06,
+      "loss": 0.6372,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7770033749615561,
+      "learning_rate": 9.999914436223627e-06,
+      "loss": 0.5455,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7008374043823205,
+      "learning_rate": 9.999906834172585e-06,
+      "loss": 0.6531,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6963471980218585,
+      "learning_rate": 9.999898908634113e-06,
+      "loss": 0.6485,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.153031692888861,
+      "learning_rate": 9.999890659608722e-06,
+      "loss": 0.521,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7396940173231474,
+      "learning_rate": 9.999882087096945e-06,
+      "loss": 0.6012,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.3235988170966353,
+      "learning_rate": 9.99987319109934e-06,
+      "loss": 0.5793,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7559347263093723,
+      "learning_rate": 9.999863971616479e-06,
+      "loss": 0.6397,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.354326281569878,
+      "learning_rate": 9.999854428648958e-06,
+      "loss": 0.5783,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5214335964831733,
+      "learning_rate": 9.999844562197398e-06,
+      "loss": 0.5089,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9200327170650573,
+      "learning_rate": 9.999834372262435e-06,
+      "loss": 0.5989,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7711204625051933,
+      "learning_rate": 9.999823858844728e-06,
+      "loss": 0.6315,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1729131572733382,
+      "learning_rate": 9.99981302194496e-06,
+      "loss": 0.567,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.390677763196411,
+      "learning_rate": 9.999801861563828e-06,
+      "loss": 0.6078,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9439383485691675,
+      "learning_rate": 9.999790377702057e-06,
+      "loss": 0.5127,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9192322067186274,
+      "learning_rate": 9.999778570360387e-06,
+      "loss": 0.6163,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9973243230444233,
+      "learning_rate": 9.999766439539588e-06,
+      "loss": 0.5951,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2213161668187524,
+      "learning_rate": 9.999753985240439e-06,
+      "loss": 0.5763,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9057432027906143,
+      "learning_rate": 9.999741207463747e-06,
+      "loss": 0.5866,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1312535311870664,
+      "learning_rate": 9.99972810621034e-06,
+      "loss": 0.5238,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9074537651287446,
+      "learning_rate": 9.999714681481064e-06,
+      "loss": 0.6493,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0890623629097855,
+      "learning_rate": 9.99970093327679e-06,
+      "loss": 0.5281,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6919691513639264,
+      "learning_rate": 9.999686861598406e-06,
+      "loss": 0.6272,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0637737850327458,
+      "learning_rate": 9.999672466446821e-06,
+      "loss": 0.5835,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7076696289292165,
+      "learning_rate": 9.999657747822969e-06,
+      "loss": 0.6441,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1453231980105363,
+      "learning_rate": 9.9996427057278e-06,
+      "loss": 0.5449,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.129405678119133,
+      "learning_rate": 9.99962734016229e-06,
+      "loss": 0.5779,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8297615493005956,
+      "learning_rate": 9.99961165112743e-06,
+      "loss": 0.6076,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9121379632034627,
+      "learning_rate": 9.999595638624236e-06,
+      "loss": 0.5209,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5795173555559119,
+      "learning_rate": 9.999579302653746e-06,
+      "loss": 0.568,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 13.785597597336825,
+      "learning_rate": 9.999562643217016e-06,
+      "loss": 0.5585,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6807090497651151,
+      "learning_rate": 9.999545660315121e-06,
+      "loss": 0.5426,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7456440571424378,
+      "learning_rate": 9.999528353949163e-06,
+      "loss": 0.6478,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.603526470949391,
+      "learning_rate": 9.999510724120261e-06,
+      "loss": 0.5677,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.9361182104948327,
+      "learning_rate": 9.999492770829555e-06,
+      "loss": 0.5616,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.815414035525676,
+      "learning_rate": 9.999474494078208e-06,
+      "loss": 0.6006,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0405395842239304,
+      "learning_rate": 9.9994558938674e-06,
+      "loss": 0.5745,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.706757137122106,
+      "learning_rate": 9.999436970198336e-06,
+      "loss": 0.5764,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8583675570380447,
+      "learning_rate": 9.99941772307224e-06,
+      "loss": 0.5528,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8480658688812206,
+      "learning_rate": 9.999398152490358e-06,
+      "loss": 0.5683,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.997398455386773,
+      "learning_rate": 9.999378258453955e-06,
+      "loss": 0.5687,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.9253154859269572,
+      "learning_rate": 9.999358040964317e-06,
+      "loss": 0.5228,
+      "step": 445
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6508927814790604,
+      "learning_rate": 9.999337500022757e-06,
+      "loss": 0.5457,
+      "step": 446
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1139981650045523,
+      "learning_rate": 9.9993166356306e-06,
+      "loss": 0.6168,
+      "step": 447
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.245122019332452,
+      "learning_rate": 9.999295447789194e-06,
+      "loss": 0.6761,
+      "step": 448
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5726851446782861,
+      "learning_rate": 9.999273936499915e-06,
+      "loss": 0.5886,
+      "step": 449
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.456112672830984,
+      "learning_rate": 9.999252101764152e-06,
+      "loss": 0.6735,
+      "step": 450
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.080850576872273,
+      "learning_rate": 9.999229943583318e-06,
+      "loss": 0.5938,
+      "step": 451
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5544416740106515,
+      "learning_rate": 9.999207461958845e-06,
+      "loss": 0.6004,
+      "step": 452
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8551083229636015,
+      "learning_rate": 9.999184656892191e-06,
+      "loss": 0.6267,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7219032684934263,
+      "learning_rate": 9.999161528384828e-06,
+      "loss": 0.5357,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2101149334707166,
+      "learning_rate": 9.999138076438253e-06,
+      "loss": 0.5981,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7591576753116576,
+      "learning_rate": 9.999114301053985e-06,
+      "loss": 0.5965,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3301228266112384,
+      "learning_rate": 9.999090202233563e-06,
+      "loss": 0.583,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.558391230528593,
+      "learning_rate": 9.999065779978543e-06,
+      "loss": 0.4999,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.778056637994504,
+      "learning_rate": 9.999041034290507e-06,
+      "loss": 0.6149,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.2902358325736727,
+      "learning_rate": 9.999015965171055e-06,
+      "loss": 0.5639,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9289983315039514,
+      "learning_rate": 9.998990572621809e-06,
+      "loss": 0.5459,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.570344736020921,
+      "learning_rate": 9.998964856644415e-06,
+      "loss": 0.5418,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9601195646587074,
+      "learning_rate": 9.998938817240533e-06,
+      "loss": 0.6423,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6633424508024957,
+      "learning_rate": 9.99891245441185e-06,
+      "loss": 0.499,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.8421338389685755,
+      "learning_rate": 9.998885768160067e-06,
+      "loss": 0.5657,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.321384680394112,
+      "learning_rate": 9.998858758486918e-06,
+      "loss": 0.517,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.612696332012091,
+      "learning_rate": 9.998831425394144e-06,
+      "loss": 0.5491,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5909574982404084,
+      "learning_rate": 9.998803768883519e-06,
+      "loss": 0.5557,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8864328103442396,
+      "learning_rate": 9.998775788956828e-06,
+      "loss": 0.5444,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7971065471224329,
+      "learning_rate": 9.998747485615881e-06,
+      "loss": 0.6487,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.6202011135411327,
+      "learning_rate": 9.998718858862512e-06,
+      "loss": 0.5903,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8928062540213153,
+      "learning_rate": 9.998689908698572e-06,
+      "loss": 0.5279,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6895509093866368,
+      "learning_rate": 9.998660635125934e-06,
+      "loss": 0.6332,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.461861206664755,
+      "learning_rate": 9.998631038146492e-06,
+      "loss": 0.5533,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.983845693162468,
+      "learning_rate": 9.998601117762161e-06,
+      "loss": 0.5756,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9070793571873468,
+      "learning_rate": 9.998570873974877e-06,
+      "loss": 0.513,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 7.209137369994001,
+      "learning_rate": 9.998540306786596e-06,
+      "loss": 0.4909,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7112339911445603,
+      "learning_rate": 9.998509416199295e-06,
+      "loss": 0.6164,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.782360710129199,
+      "learning_rate": 9.998478202214977e-06,
+      "loss": 0.4776,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6440856649670651,
+      "learning_rate": 9.998446664835655e-06,
+      "loss": 0.6212,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.118211648327851,
+      "learning_rate": 9.998414804063375e-06,
+      "loss": 0.5997,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6327547481832834,
+      "learning_rate": 9.998382619900193e-06,
+      "loss": 0.5396,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7440362417683903,
+      "learning_rate": 9.998350112348196e-06,
+      "loss": 0.5936,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7363223395339917,
+      "learning_rate": 9.998317281409484e-06,
+      "loss": 0.5861,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9837355583495033,
+      "learning_rate": 9.998284127086184e-06,
+      "loss": 0.5785,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7799500037688217,
+      "learning_rate": 9.998250649380439e-06,
+      "loss": 0.6473,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9169313668320596,
+      "learning_rate": 9.998216848294415e-06,
+      "loss": 0.6517,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8009002370620928,
+      "learning_rate": 9.9981827238303e-06,
+      "loss": 0.5299,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7200891797745994,
+      "learning_rate": 9.998148275990303e-06,
+      "loss": 0.4847,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3169202787398895,
+      "learning_rate": 9.99811350477665e-06,
+      "loss": 0.5774,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8033974469246268,
+      "learning_rate": 9.99807841019159e-06,
+      "loss": 0.5664,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.882752067879728,
+      "learning_rate": 9.998042992237396e-06,
+      "loss": 0.5511,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6013244167514782,
+      "learning_rate": 9.998007250916357e-06,
+      "loss": 0.5369,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7481318379539572,
+      "learning_rate": 9.99797118623079e-06,
+      "loss": 0.6292,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6969545092110656,
+      "learning_rate": 9.997934798183025e-06,
+      "loss": 0.6444,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5947288176684786,
+      "learning_rate": 9.997898086775414e-06,
+      "loss": 0.5805,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8251212638886976,
+      "learning_rate": 9.997861052010338e-06,
+      "loss": 0.5603,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9720668309181737,
+      "learning_rate": 9.997823693890187e-06,
+      "loss": 0.513,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5832798899210685,
+      "learning_rate": 9.997786012417384e-06,
+      "loss": 0.5385,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4751794966006897,
+      "learning_rate": 9.997748007594362e-06,
+      "loss": 0.5409,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8079973951163184,
+      "learning_rate": 9.997709679423581e-06,
+      "loss": 0.5685,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9278705521432908,
+      "learning_rate": 9.997671027907525e-06,
+      "loss": 0.5606,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.750442916986704,
+      "learning_rate": 9.997632053048689e-06,
+      "loss": 0.587,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9769647270035005,
+      "learning_rate": 9.997592754849596e-06,
+      "loss": 0.5079,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5269182077926,
+      "learning_rate": 9.997553133312791e-06,
+      "loss": 0.4898,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7854812898086454,
+      "learning_rate": 9.997513188440835e-06,
+      "loss": 0.5567,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9364317471238655,
+      "learning_rate": 9.997472920236313e-06,
+      "loss": 0.5901,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8234122103097035,
+      "learning_rate": 9.997432328701833e-06,
+      "loss": 0.5775,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3331251800788033,
+      "learning_rate": 9.997391413840016e-06,
+      "loss": 0.5736,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6943457714915588,
+      "learning_rate": 9.997350175653513e-06,
+      "loss": 0.6057,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0169004390224894,
+      "learning_rate": 9.997308614144991e-06,
+      "loss": 0.6382,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.906197079848998,
+      "learning_rate": 9.997266729317138e-06,
+      "loss": 0.5265,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9935889874911294,
+      "learning_rate": 9.997224521172668e-06,
+      "loss": 0.5756,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6039594678295857,
+      "learning_rate": 9.997181989714305e-06,
+      "loss": 0.6301,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0744146872246585,
+      "learning_rate": 9.997139134944806e-06,
+      "loss": 0.6191,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9948630254105333,
+      "learning_rate": 9.997095956866943e-06,
+      "loss": 0.5708,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 5.684738977016506,
+      "learning_rate": 9.997052455483507e-06,
+      "loss": 0.621,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1592779535081883,
+      "learning_rate": 9.997008630797314e-06,
+      "loss": 0.5327,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.5927090466906564,
+      "learning_rate": 9.9969644828112e-06,
+      "loss": 0.6172,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.006732288470622,
+      "learning_rate": 9.996920011528022e-06,
+      "loss": 0.599,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.3293150369569178,
+      "learning_rate": 9.996875216950655e-06,
+      "loss": 0.5903,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8781733322577094,
+      "learning_rate": 9.996830099081998e-06,
+      "loss": 0.5702,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0783582531549163,
+      "learning_rate": 9.99678465792497e-06,
+      "loss": 0.535,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7179385542204724,
+      "learning_rate": 9.996738893482512e-06,
+      "loss": 0.5615,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.881319869854783,
+      "learning_rate": 9.996692805757584e-06,
+      "loss": 0.5954,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8457513024622156,
+      "learning_rate": 9.996646394753167e-06,
+      "loss": 0.5948,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.785525702792255,
+      "learning_rate": 9.996599660472266e-06,
+      "loss": 0.6063,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.969284203328923,
+      "learning_rate": 9.996552602917902e-06,
+      "loss": 0.5627,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8793133100589599,
+      "learning_rate": 9.996505222093123e-06,
+      "loss": 0.5171,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.049709958133192,
+      "learning_rate": 9.99645751800099e-06,
+      "loss": 0.5886,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7297266235629103,
+      "learning_rate": 9.996409490644593e-06,
+      "loss": 0.5558,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.060280551842997,
+      "learning_rate": 9.996361140027038e-06,
+      "loss": 0.5915,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6787659365945304,
+      "learning_rate": 9.996312466151452e-06,
+      "loss": 0.5484,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7646445520989786,
+      "learning_rate": 9.996263469020988e-06,
+      "loss": 0.5747,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4909817124771623,
+      "learning_rate": 9.996214148638811e-06,
+      "loss": 0.5836,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.4306265798346596,
+      "learning_rate": 9.996164505008117e-06,
+      "loss": 0.5647,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1263405062083787,
+      "learning_rate": 9.996114538132114e-06,
+      "loss": 0.5668,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0345490407764717,
+      "learning_rate": 9.996064248014036e-06,
+      "loss": 0.6117,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.0367369087259948,
+      "learning_rate": 9.996013634657136e-06,
+      "loss": 0.5639,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5990843421878174,
+      "learning_rate": 9.995962698064692e-06,
+      "loss": 0.5131,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3732024214638143,
+      "learning_rate": 9.995911438239995e-06,
+      "loss": 0.6699,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5947259957861935,
+      "learning_rate": 9.995859855186363e-06,
+      "loss": 0.4764,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.642717213862172,
+      "learning_rate": 9.995807948907134e-06,
+      "loss": 0.4893,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6425898448409968,
+      "learning_rate": 9.995755719405667e-06,
+      "loss": 0.5773,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.032220639511612,
+      "learning_rate": 9.99570316668534e-06,
+      "loss": 0.5967,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.4773074771051986,
+      "learning_rate": 9.995650290749553e-06,
+      "loss": 0.6133,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6605611505082294,
+      "learning_rate": 9.995597091601727e-06,
+      "loss": 0.5106,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6518753356913105,
+      "learning_rate": 9.995543569245304e-06,
+      "loss": 0.637,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4718792844526425,
+      "learning_rate": 9.995489723683748e-06,
+      "loss": 0.595,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8623012018187929,
+      "learning_rate": 9.995435554920544e-06,
+      "loss": 0.6184,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.968637488689114,
+      "learning_rate": 9.99538106295919e-06,
+      "loss": 0.625,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.950245196131321,
+      "learning_rate": 9.99532624780322e-06,
+      "loss": 0.5535,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9149983646849986,
+      "learning_rate": 9.995271109456172e-06,
+      "loss": 0.5666,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.274088694453281,
+      "learning_rate": 9.995215647921622e-06,
+      "loss": 0.5415,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.73987440564278,
+      "learning_rate": 9.995159863203152e-06,
+      "loss": 0.5245,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6573855149718795,
+      "learning_rate": 9.995103755304373e-06,
+      "loss": 0.6269,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9237948443021977,
+      "learning_rate": 9.995047324228917e-06,
+      "loss": 0.5921,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7347863889733812,
+      "learning_rate": 9.994990569980433e-06,
+      "loss": 0.4868,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7652694773820714,
+      "learning_rate": 9.994933492562593e-06,
+      "loss": 0.526,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7276782294284835,
+      "learning_rate": 9.994876091979092e-06,
+      "loss": 0.6039,
+      "step": 560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.5026831704780053,
+      "learning_rate": 9.994818368233639e-06,
+      "loss": 0.5584,
+      "step": 561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6758481212343759,
+      "learning_rate": 9.994760321329972e-06,
+      "loss": 0.582,
+      "step": 562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.851377994974264,
+      "learning_rate": 9.994701951271848e-06,
+      "loss": 0.613,
+      "step": 563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5502575172031208,
+      "learning_rate": 9.99464325806304e-06,
+      "loss": 0.5582,
+      "step": 564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.142598862605693,
+      "learning_rate": 9.994584241707349e-06,
+      "loss": 0.5432,
+      "step": 565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.1344114937122054,
+      "learning_rate": 9.994524902208588e-06,
+      "loss": 0.5304,
+      "step": 566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5010984184079867,
+      "learning_rate": 9.994465239570602e-06,
+      "loss": 0.554,
+      "step": 567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.824975232800367,
+      "learning_rate": 9.994405253797248e-06,
+      "loss": 0.5822,
+      "step": 568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 3.888750146785224,
+      "learning_rate": 9.994344944892409e-06,
+      "loss": 0.5507,
+      "step": 569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.982323767876106,
+      "learning_rate": 9.994284312859982e-06,
+      "loss": 0.5559,
+      "step": 570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.9975489872465646,
+      "learning_rate": 9.994223357703897e-06,
+      "loss": 0.6041,
+      "step": 571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5830085624445525,
+      "learning_rate": 9.994162079428093e-06,
+      "loss": 0.5848,
+      "step": 572
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7045779016520832,
+      "learning_rate": 9.994100478036535e-06,
+      "loss": 0.6087,
+      "step": 573
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8054863232173629,
+      "learning_rate": 9.994038553533209e-06,
+      "loss": 0.627,
+      "step": 574
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7823204476550272,
+      "learning_rate": 9.993976305922121e-06,
+      "loss": 0.6026,
+      "step": 575
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9304874880786358,
+      "learning_rate": 9.993913735207302e-06,
+      "loss": 0.5707,
+      "step": 576
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6657479767116596,
+      "learning_rate": 9.993850841392793e-06,
+      "loss": 0.6016,
+      "step": 577
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7438489220590868,
+      "learning_rate": 9.99378762448267e-06,
+      "loss": 0.6276,
+      "step": 578
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7827656143735742,
+      "learning_rate": 9.993724084481021e-06,
+      "loss": 0.4939,
+      "step": 579
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6711051316659882,
+      "learning_rate": 9.993660221391954e-06,
+      "loss": 0.5347,
+      "step": 580
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.5199185534698825,
+      "learning_rate": 9.993596035219606e-06,
+      "loss": 0.5315,
+      "step": 581
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7312584987888469,
+      "learning_rate": 9.993531525968126e-06,
+      "loss": 0.6152,
+      "step": 582
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5469038874697256,
+      "learning_rate": 9.993466693641686e-06,
+      "loss": 0.5269,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.03759891966535,
+      "learning_rate": 9.993401538244485e-06,
+      "loss": 0.4825,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9040702805980985,
+      "learning_rate": 9.993336059780738e-06,
+      "loss": 0.5958,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 10.318347704451378,
+      "learning_rate": 9.99327025825468e-06,
+      "loss": 0.5795,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6898040743323544,
+      "learning_rate": 9.993204133670566e-06,
+      "loss": 0.6006,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.615720637024177,
+      "learning_rate": 9.993137686032677e-06,
+      "loss": 0.5788,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6646898782066699,
+      "learning_rate": 9.993070915345313e-06,
+      "loss": 0.6296,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7478658287285007,
+      "learning_rate": 9.993003821612793e-06,
+      "loss": 0.5327,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5733117889891142,
+      "learning_rate": 9.992936404839455e-06,
+      "loss": 0.5835,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.978359391145377,
+      "learning_rate": 9.992868665029665e-06,
+      "loss": 0.5571,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 3.0671812455877396,
+      "learning_rate": 9.992800602187801e-06,
+      "loss": 0.5279,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6376525722911022,
+      "learning_rate": 9.992732216318274e-06,
+      "loss": 0.5348,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.654361914860124,
+      "learning_rate": 9.992663507425501e-06,
+      "loss": 0.6317,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6811524265556017,
+      "learning_rate": 9.99259447551393e-06,
+      "loss": 0.5359,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.6561149190153057,
+      "learning_rate": 9.99252512058803e-06,
+      "loss": 0.5287,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6110961145389977,
+      "learning_rate": 9.992455442652283e-06,
+      "loss": 0.6205,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5951273628797897,
+      "learning_rate": 9.9923854417112e-06,
+      "loss": 0.5888,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6104740172196699,
+      "learning_rate": 9.992315117769311e-06,
+      "loss": 0.5414,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3726273660888617,
+      "learning_rate": 9.992244470831164e-06,
+      "loss": 0.5423,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.580580612329893,
+      "learning_rate": 9.992173500901333e-06,
+      "loss": 0.5519,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5206276318798455,
+      "learning_rate": 9.992102207984404e-06,
+      "loss": 0.5819,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7683649186744994,
+      "learning_rate": 9.992030592084994e-06,
+      "loss": 0.6138,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1938631226806624,
+      "learning_rate": 9.991958653207733e-06,
+      "loss": 0.5584,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5323436888772248,
+      "learning_rate": 9.99188639135728e-06,
+      "loss": 0.5663,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.788553160569277,
+      "learning_rate": 9.991813806538308e-06,
+      "loss": 0.5529,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6655266890921836,
+      "learning_rate": 9.99174089875551e-06,
+      "loss": 0.5899,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6514508990007981,
+      "learning_rate": 9.991667668013609e-06,
+      "loss": 0.5053,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6414595295986534,
+      "learning_rate": 9.991594114317338e-06,
+      "loss": 0.5956,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9054910132221903,
+      "learning_rate": 9.991520237671457e-06,
+      "loss": 0.5416,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.731454544089483,
+      "learning_rate": 9.991446038080748e-06,
+      "loss": 0.6269,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6591049700940341,
+      "learning_rate": 9.99137151555001e-06,
+      "loss": 0.5599,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6925409216542935,
+      "learning_rate": 9.991296670084062e-06,
+      "loss": 0.5663,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7397869738967933,
+      "learning_rate": 9.991221501687751e-06,
+      "loss": 0.5605,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.485879958708546,
+      "learning_rate": 9.99114601036594e-06,
+      "loss": 0.5217,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5878441087476907,
+      "learning_rate": 9.991070196123507e-06,
+      "loss": 0.5732,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4155310116666375,
+      "learning_rate": 9.990994058965363e-06,
+      "loss": 0.5444,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5860267849373817,
+      "learning_rate": 9.990917598896435e-06,
+      "loss": 0.5275,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4901511178261713,
+      "learning_rate": 9.990840815921665e-06,
+      "loss": 0.5657,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.537497034554049,
+      "learning_rate": 9.990763710046024e-06,
+      "loss": 0.5353,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.514538939347452,
+      "learning_rate": 9.990686281274498e-06,
+      "loss": 0.5114,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0841471116524226,
+      "learning_rate": 9.9906085296121e-06,
+      "loss": 0.5508,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.8151218457919573,
+      "learning_rate": 9.990530455063857e-06,
+      "loss": 0.6077,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.616315591965677,
+      "learning_rate": 9.990452057634823e-06,
+      "loss": 0.5962,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0880724421909727,
+      "learning_rate": 9.99037333733007e-06,
+      "loss": 0.5463,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7263399225229703,
+      "learning_rate": 9.990294294154688e-06,
+      "loss": 0.5858,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5850760926265768,
+      "learning_rate": 9.990214928113795e-06,
+      "loss": 0.5781,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.582352121211005,
+      "learning_rate": 9.990135239212525e-06,
+      "loss": 0.5396,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6401119170026337,
+      "learning_rate": 9.990055227456032e-06,
+      "loss": 0.6795,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6893689436990103,
+      "learning_rate": 9.989974892849493e-06,
+      "loss": 0.5403,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.639760767908914,
+      "learning_rate": 9.989894235398106e-06,
+      "loss": 0.5752,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6783589780746055,
+      "learning_rate": 9.98981325510709e-06,
+      "loss": 0.5871,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7562068784219536,
+      "learning_rate": 9.989731951981685e-06,
+      "loss": 0.6009,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7364692267775057,
+      "learning_rate": 9.989650326027149e-06,
+      "loss": 0.537,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3960999324602983,
+      "learning_rate": 9.989568377248763e-06,
+      "loss": 0.5847,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6914464114149388,
+      "learning_rate": 9.989486105651834e-06,
+      "loss": 0.5498,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8140557813433928,
+      "learning_rate": 9.989403511241678e-06,
+      "loss": 0.5872,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7729738346576522,
+      "learning_rate": 9.989320594023641e-06,
+      "loss": 0.6296,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5894564205044843,
+      "learning_rate": 9.98923735400309e-06,
+      "loss": 0.5373,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.749295350456021,
+      "learning_rate": 9.98915379118541e-06,
+      "loss": 0.564,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0152103607782212,
+      "learning_rate": 9.989069905576005e-06,
+      "loss": 0.5217,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7465331989766465,
+      "learning_rate": 9.988985697180302e-06,
+      "loss": 0.61,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0524545411264117,
+      "learning_rate": 9.988901166003754e-06,
+      "loss": 0.6092,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6936581437838388,
+      "learning_rate": 9.988816312051826e-06,
+      "loss": 0.6388,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.981466425626048,
+      "learning_rate": 9.988731135330008e-06,
+      "loss": 0.5948,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4908283554319752,
+      "learning_rate": 9.988645635843811e-06,
+      "loss": 0.4917,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8155960162954827,
+      "learning_rate": 9.988559813598769e-06,
+      "loss": 0.5442,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6106436421850783,
+      "learning_rate": 9.988473668600431e-06,
+      "loss": 0.5451,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.149973094569598,
+      "learning_rate": 9.988387200854373e-06,
+      "loss": 0.5675,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6671964338293705,
+      "learning_rate": 9.98830041036619e-06,
+      "loss": 0.544,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.746848370094641,
+      "learning_rate": 9.988213297141495e-06,
+      "loss": 0.5601,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7662086274810078,
+      "learning_rate": 9.988125861185924e-06,
+      "loss": 0.4977,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.543500942379991,
+      "learning_rate": 9.988038102505138e-06,
+      "loss": 0.5043,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.64246197690511,
+      "learning_rate": 9.987950021104808e-06,
+      "loss": 0.5572,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9491055885273412,
+      "learning_rate": 9.98786161699064e-06,
+      "loss": 0.5596,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7995570194812731,
+      "learning_rate": 9.98777289016835e-06,
+      "loss": 0.5701,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.9162968582522972,
+      "learning_rate": 9.987683840643679e-06,
+      "loss": 0.5333,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4986684799096002,
+      "learning_rate": 9.987594468422385e-06,
+      "loss": 0.5326,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7281353638113948,
+      "learning_rate": 9.987504773510257e-06,
+      "loss": 0.6053,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6344904333642738,
+      "learning_rate": 9.987414755913094e-06,
+      "loss": 0.4935,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.933292452356171,
+      "learning_rate": 9.987324415636718e-06,
+      "loss": 0.5729,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7914235554236044,
+      "learning_rate": 9.98723375268698e-06,
+      "loss": 0.6318,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.368714590547771,
+      "learning_rate": 9.98714276706974e-06,
+      "loss": 0.5287,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9735632981986477,
+      "learning_rate": 9.987051458790889e-06,
+      "loss": 0.5005,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5160112049572383,
+      "learning_rate": 9.98695982785633e-06,
+      "loss": 0.6272,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.579649141714724,
+      "learning_rate": 9.986867874271996e-06,
+      "loss": 0.5465,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1989897198407973,
+      "learning_rate": 9.986775598043834e-06,
+      "loss": 0.5511,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6256322233560254,
+      "learning_rate": 9.986682999177813e-06,
+      "loss": 0.5317,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7126356655723156,
+      "learning_rate": 9.986590077679927e-06,
+      "loss": 0.5709,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.955461437916781,
+      "learning_rate": 9.986496833556185e-06,
+      "loss": 0.5155,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4895207567443318,
+      "learning_rate": 9.98640326681262e-06,
+      "loss": 0.5894,
+      "step": 672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7472953365230885,
+      "learning_rate": 9.98630937745529e-06,
+      "loss": 0.5876,
+      "step": 673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6475174166949256,
+      "learning_rate": 9.986215165490264e-06,
+      "loss": 0.5266,
+      "step": 674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7141868521510457,
+      "learning_rate": 9.98612063092364e-06,
+      "loss": 0.5519,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 42.853834822125805,
+      "learning_rate": 9.986025773761533e-06,
+      "loss": 0.5155,
+      "step": 676
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.082537131381896,
+      "learning_rate": 9.985930594010083e-06,
+      "loss": 0.5543,
+      "step": 677
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5225198949648842,
+      "learning_rate": 9.985835091675444e-06,
+      "loss": 0.5232,
+      "step": 678
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6803135638601516,
+      "learning_rate": 9.985739266763797e-06,
+      "loss": 0.5877,
+      "step": 679
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0559300998059715,
+      "learning_rate": 9.985643119281344e-06,
+      "loss": 0.6349,
+      "step": 680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8267682851929372,
+      "learning_rate": 9.9855466492343e-06,
+      "loss": 0.5484,
+      "step": 681
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6796824374810231,
+      "learning_rate": 9.98544985662891e-06,
+      "loss": 0.6461,
+      "step": 682
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7403251176751362,
+      "learning_rate": 9.985352741471439e-06,
+      "loss": 0.6164,
+      "step": 683
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5508084183842126,
+      "learning_rate": 9.985255303768164e-06,
+      "loss": 0.5708,
+      "step": 684
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7474671328171805,
+      "learning_rate": 9.985157543525394e-06,
+      "loss": 0.6007,
+      "step": 685
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1095791709808855,
+      "learning_rate": 9.985059460749453e-06,
+      "loss": 0.5955,
+      "step": 686
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5541538599601818,
+      "learning_rate": 9.984961055446685e-06,
+      "loss": 0.6355,
+      "step": 687
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4152587868611695,
+      "learning_rate": 9.984862327623458e-06,
+      "loss": 0.5109,
+      "step": 688
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.781653951768241,
+      "learning_rate": 9.984763277286161e-06,
+      "loss": 0.5307,
+      "step": 689
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.1900541088304304,
+      "learning_rate": 9.984663904441198e-06,
+      "loss": 0.5883,
+      "step": 690
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4947068369586485,
+      "learning_rate": 9.984564209095002e-06,
+      "loss": 0.6574,
+      "step": 691
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0568687830164945,
+      "learning_rate": 9.984464191254024e-06,
+      "loss": 0.5079,
+      "step": 692
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8592171098737291,
+      "learning_rate": 9.984363850924733e-06,
+      "loss": 0.5149,
+      "step": 693
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4174768356190603,
+      "learning_rate": 9.98426318811362e-06,
+      "loss": 0.4992,
+      "step": 694
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3358526555338235,
+      "learning_rate": 9.984162202827199e-06,
+      "loss": 0.5565,
+      "step": 695
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.571414051814849,
+      "learning_rate": 9.984060895072003e-06,
+      "loss": 0.6,
+      "step": 696
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5627449619313605,
+      "learning_rate": 9.98395926485459e-06,
+      "loss": 0.5652,
+      "step": 697
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6945946002972168,
+      "learning_rate": 9.983857312181528e-06,
+      "loss": 0.5221,
+      "step": 698
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5645184257198779,
+      "learning_rate": 9.983755037059422e-06,
+      "loss": 0.5436,
+      "step": 699
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.709418681398146,
+      "learning_rate": 9.983652439494882e-06,
+      "loss": 0.5555,
+      "step": 700
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0497617900514182,
+      "learning_rate": 9.983549519494549e-06,
+      "loss": 0.5679,
+      "step": 701
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6868186412445652,
+      "learning_rate": 9.983446277065083e-06,
+      "loss": 0.5734,
+      "step": 702
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6743149331404321,
+      "learning_rate": 9.98334271221316e-06,
+      "loss": 0.5609,
+      "step": 703
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7812128529881837,
+      "learning_rate": 9.983238824945483e-06,
+      "loss": 0.5723,
+      "step": 704
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6914989472807405,
+      "learning_rate": 9.983134615268774e-06,
+      "loss": 0.5276,
+      "step": 705
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9038375159668738,
+      "learning_rate": 9.983030083189773e-06,
+      "loss": 0.6145,
+      "step": 706
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6030771139020674,
+      "learning_rate": 9.982925228715244e-06,
+      "loss": 0.6108,
+      "step": 707
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6154291239510121,
+      "learning_rate": 9.982820051851972e-06,
+      "loss": 0.5479,
+      "step": 708
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.658916379870633,
+      "learning_rate": 9.98271455260676e-06,
+      "loss": 0.499,
+      "step": 709
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0560607094342087,
+      "learning_rate": 9.982608730986437e-06,
+      "loss": 0.5828,
+      "step": 710
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.92355124818166,
+      "learning_rate": 9.982502586997846e-06,
+      "loss": 0.5297,
+      "step": 711
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7581998270043666,
+      "learning_rate": 9.982396120647855e-06,
+      "loss": 0.539,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8064850218074654,
+      "learning_rate": 9.982289331943353e-06,
+      "loss": 0.6105,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9950779674889938,
+      "learning_rate": 9.982182220891247e-06,
+      "loss": 0.5653,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9160067653399984,
+      "learning_rate": 9.982074787498472e-06,
+      "loss": 0.5355,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9059522111894736,
+      "learning_rate": 9.981967031771974e-06,
+      "loss": 0.5741,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.801979600586274,
+      "learning_rate": 9.981858953718728e-06,
+      "loss": 0.5713,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1209981091794274,
+      "learning_rate": 9.981750553345721e-06,
+      "loss": 0.515,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0448020072984674,
+      "learning_rate": 9.981641830659973e-06,
+      "loss": 0.5355,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7455436440863386,
+      "learning_rate": 9.981532785668516e-06,
+      "loss": 0.5064,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6925400323926265,
+      "learning_rate": 9.981423418378403e-06,
+      "loss": 0.6155,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.701973946700673,
+      "learning_rate": 9.98131372879671e-06,
+      "loss": 0.606,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5844382716238123,
+      "learning_rate": 9.981203716930538e-06,
+      "loss": 0.5409,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7799743786390323,
+      "learning_rate": 9.981093382787002e-06,
+      "loss": 0.5154,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1690764214661664,
+      "learning_rate": 9.980982726373238e-06,
+      "loss": 0.6018,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.053480245351553,
+      "learning_rate": 9.980871747696408e-06,
+      "loss": 0.5493,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.954289388156097,
+      "learning_rate": 9.980760446763693e-06,
+      "loss": 0.5959,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9904545251670165,
+      "learning_rate": 9.980648823582291e-06,
+      "loss": 0.5441,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5105164865614267,
+      "learning_rate": 9.980536878159427e-06,
+      "loss": 0.5475,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7845429346033919,
+      "learning_rate": 9.980424610502342e-06,
+      "loss": 0.5404,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.835663999661278,
+      "learning_rate": 9.9803120206183e-06,
+      "loss": 0.65,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1707415611262446,
+      "learning_rate": 9.980199108514584e-06,
+      "loss": 0.5608,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7935927538206793,
+      "learning_rate": 9.980085874198502e-06,
+      "loss": 0.6181,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.561644773014807,
+      "learning_rate": 9.979972317677377e-06,
+      "loss": 0.5715,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0055433769015885,
+      "learning_rate": 9.97985843895856e-06,
+      "loss": 0.478,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6267540376363032,
+      "learning_rate": 9.979744238049415e-06,
+      "loss": 0.5291,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.951456473765339,
+      "learning_rate": 9.979629714957334e-06,
+      "loss": 0.6164,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8752697401666278,
+      "learning_rate": 9.979514869689722e-06,
+      "loss": 0.5404,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6245028973500744,
+      "learning_rate": 9.979399702254014e-06,
+      "loss": 0.5654,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5736047017997292,
+      "learning_rate": 9.979284212657658e-06,
+      "loss": 0.5437,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.6196004369659276,
+      "learning_rate": 9.979168400908126e-06,
+      "loss": 0.5429,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1417918247001153,
+      "learning_rate": 9.979052267012914e-06,
+      "loss": 0.5479,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4178030775106332,
+      "learning_rate": 9.978935810979533e-06,
+      "loss": 0.5579,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5953066823336068,
+      "learning_rate": 9.978819032815519e-06,
+      "loss": 0.5785,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8650087318715263,
+      "learning_rate": 9.978701932528425e-06,
+      "loss": 0.5388,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5486865519521378,
+      "learning_rate": 9.97858451012583e-06,
+      "loss": 0.5812,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.531871293831441,
+      "learning_rate": 9.97846676561533e-06,
+      "loss": 0.5458,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5201953954497902,
+      "learning_rate": 9.978348699004545e-06,
+      "loss": 0.5092,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.693273511506159,
+      "learning_rate": 9.978230310301109e-06,
+      "loss": 0.5574,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5816583900922208,
+      "learning_rate": 9.978111599512685e-06,
+      "loss": 0.5782,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4556411247802536,
+      "learning_rate": 9.977992566646952e-06,
+      "loss": 0.5797,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.1673046067611432,
+      "learning_rate": 9.977873211711612e-06,
+      "loss": 0.5325,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9299906705132726,
+      "learning_rate": 9.977753534714389e-06,
+      "loss": 0.6113,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.848851679145008,
+      "learning_rate": 9.977633535663022e-06,
+      "loss": 0.5334,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.109332384398877,
+      "learning_rate": 9.977513214565278e-06,
+      "loss": 0.5236,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.684981400874114,
+      "learning_rate": 9.977392571428942e-06,
+      "loss": 0.5614,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7803211661689158,
+      "learning_rate": 9.977271606261814e-06,
+      "loss": 0.6041,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8787862460695115,
+      "learning_rate": 9.977150319071727e-06,
+      "loss": 0.5675,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6195226252697599,
+      "learning_rate": 9.977028709866523e-06,
+      "loss": 0.5208,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0772050145709864,
+      "learning_rate": 9.976906778654074e-06,
+      "loss": 0.5649,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4158696211470787,
+      "learning_rate": 9.976784525442268e-06,
+      "loss": 0.5593,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7502360519045748,
+      "learning_rate": 9.97666195023901e-06,
+      "loss": 0.6044,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4283400601973875,
+      "learning_rate": 9.976539053052237e-06,
+      "loss": 0.5229,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4764638956236051,
+      "learning_rate": 9.976415833889897e-06,
+      "loss": 0.5334,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0859069582570218,
+      "learning_rate": 9.976292292759963e-06,
+      "loss": 0.5395,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5203338827933996,
+      "learning_rate": 9.976168429670426e-06,
+      "loss": 0.5849,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6775133181138706,
+      "learning_rate": 9.976044244629301e-06,
+      "loss": 0.5404,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5952308103512223,
+      "learning_rate": 9.975919737644625e-06,
+      "loss": 0.5697,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5129362326307954,
+      "learning_rate": 9.97579490872445e-06,
+      "loss": 0.5385,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6688181805939526,
+      "learning_rate": 9.975669757876853e-06,
+      "loss": 0.5229,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3711266094196919,
+      "learning_rate": 9.975544285109933e-06,
+      "loss": 0.5373,
+      "step": 771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8880072358254525,
+      "learning_rate": 9.975418490431806e-06,
+      "loss": 0.5803,
+      "step": 772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0819201866747257,
+      "learning_rate": 9.975292373850611e-06,
+      "loss": 0.5256,
+      "step": 773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6854259455741785,
+      "learning_rate": 9.97516593537451e-06,
+      "loss": 0.5829,
+      "step": 774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.821712367167311,
+      "learning_rate": 9.975039175011678e-06,
+      "loss": 0.6448,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.034348543153588,
+      "learning_rate": 9.97491209277032e-06,
+      "loss": 0.544,
+      "step": 776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4021850039764745,
+      "learning_rate": 9.974784688658661e-06,
+      "loss": 0.5352,
+      "step": 777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6813570671633562,
+      "learning_rate": 9.974656962684936e-06,
+      "loss": 0.5999,
+      "step": 778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4561371347973227,
+      "learning_rate": 9.974528914857417e-06,
+      "loss": 0.5059,
+      "step": 779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5416131533424684,
+      "learning_rate": 9.974400545184382e-06,
+      "loss": 0.5174,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6876817090353196,
+      "learning_rate": 9.97427185367414e-06,
+      "loss": 0.5978,
+      "step": 781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0869340382487938,
+      "learning_rate": 9.974142840335017e-06,
+      "loss": 0.5731,
+      "step": 782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6547970253362836,
+      "learning_rate": 9.974013505175358e-06,
+      "loss": 0.5751,
+      "step": 783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6394056968397406,
+      "learning_rate": 9.973883848203535e-06,
+      "loss": 0.5344,
+      "step": 784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6531189421145551,
+      "learning_rate": 9.973753869427928e-06,
+      "loss": 0.5443,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7556512131704725,
+      "learning_rate": 9.973623568856956e-06,
+      "loss": 0.5148,
+      "step": 786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9169417621087852,
+      "learning_rate": 9.973492946499045e-06,
+      "loss": 0.5918,
+      "step": 787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4570709613454904,
+      "learning_rate": 9.973362002362647e-06,
+      "loss": 0.5373,
+      "step": 788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5311428640142923,
+      "learning_rate": 9.973230736456232e-06,
+      "loss": 0.5683,
+      "step": 789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.398557127650882,
+      "learning_rate": 9.973099148788295e-06,
+      "loss": 0.4895,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6256153234822469,
+      "learning_rate": 9.97296723936735e-06,
+      "loss": 0.6335,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7729082398853595,
+      "learning_rate": 9.972835008201928e-06,
+      "loss": 0.5294,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6396473568274677,
+      "learning_rate": 9.972702455300586e-06,
+      "loss": 0.5396,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5249789217901535,
+      "learning_rate": 9.972569580671902e-06,
+      "loss": 0.5774,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.594759032543909,
+      "learning_rate": 9.97243638432447e-06,
+      "loss": 0.5309,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.461584294536784,
+      "learning_rate": 9.97230286626691e-06,
+      "loss": 0.5632,
+      "step": 796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4955351507068555,
+      "learning_rate": 9.972169026507858e-06,
+      "loss": 0.5515,
+      "step": 797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5529748120600473,
+      "learning_rate": 9.972034865055974e-06,
+      "loss": 0.5405,
+      "step": 798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.4473463039722567,
+      "learning_rate": 9.97190038191994e-06,
+      "loss": 0.5749,
+      "step": 799
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.457597800019409,
+      "learning_rate": 9.971765577108452e-06,
+      "loss": 0.5262,
+      "step": 800
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6492051417181997,
+      "learning_rate": 9.971630450630239e-06,
+      "loss": 0.6388,
+      "step": 801
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7491249427264517,
+      "learning_rate": 9.971495002494038e-06,
+      "loss": 0.5415,
+      "step": 802
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6221351057051563,
+      "learning_rate": 9.971359232708613e-06,
+      "loss": 0.579,
+      "step": 803
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.76924559842447,
+      "learning_rate": 9.971223141282751e-06,
+      "loss": 0.5339,
+      "step": 804
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.222345346043504,
+      "learning_rate": 9.971086728225252e-06,
+      "loss": 0.5591,
+      "step": 805
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4504455435367212,
+      "learning_rate": 9.970949993544946e-06,
+      "loss": 0.5064,
+      "step": 806
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7439488363520979,
+      "learning_rate": 9.97081293725068e-06,
+      "loss": 0.5294,
+      "step": 807
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9886027020808867,
+      "learning_rate": 9.970675559351318e-06,
+      "loss": 0.5344,
+      "step": 808
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6060333858180997,
+      "learning_rate": 9.97053785985575e-06,
+      "loss": 0.5378,
+      "step": 809
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.9118784895312244,
+      "learning_rate": 9.970399838772884e-06,
+      "loss": 0.5349,
+      "step": 810
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.658800341863543,
+      "learning_rate": 9.970261496111653e-06,
+      "loss": 0.5952,
+      "step": 811
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6027283855454784,
+      "learning_rate": 9.970122831881003e-06,
+      "loss": 0.5334,
+      "step": 812
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8124737776309374,
+      "learning_rate": 9.969983846089908e-06,
+      "loss": 0.5393,
+      "step": 813
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.566665964156739,
+      "learning_rate": 9.96984453874736e-06,
+      "loss": 0.5302,
+      "step": 814
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6052932456772644,
+      "learning_rate": 9.969704909862371e-06,
+      "loss": 0.59,
+      "step": 815
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8317701541844458,
+      "learning_rate": 9.969564959443976e-06,
+      "loss": 0.5675,
+      "step": 816
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8985976469269226,
+      "learning_rate": 9.96942468750123e-06,
+      "loss": 0.576,
+      "step": 817
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6173788735255878,
+      "learning_rate": 9.969284094043206e-06,
+      "loss": 0.6021,
+      "step": 818
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.769458974596789,
+      "learning_rate": 9.969143179079001e-06,
+      "loss": 0.518,
+      "step": 819
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6675758149395968,
+      "learning_rate": 9.969001942617735e-06,
+      "loss": 0.5989,
+      "step": 820
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.319624859889963,
+      "learning_rate": 9.968860384668543e-06,
+      "loss": 0.5086,
+      "step": 821
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.702108816104618,
+      "learning_rate": 9.968718505240583e-06,
+      "loss": 0.6366,
+      "step": 822
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6139169083984537,
+      "learning_rate": 9.968576304343037e-06,
+      "loss": 0.4558,
+      "step": 823
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.876970871841841,
+      "learning_rate": 9.968433781985102e-06,
+      "loss": 0.5122,
+      "step": 824
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6221913466613788,
+      "learning_rate": 9.968290938176005e-06,
+      "loss": 0.619,
+      "step": 825
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6744159877239608,
+      "learning_rate": 9.968147772924978e-06,
+      "loss": 0.5227,
+      "step": 826
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.648872827924901,
+      "learning_rate": 9.968004286241292e-06,
+      "loss": 0.5928,
+      "step": 827
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8105188149966776,
+      "learning_rate": 9.96786047813423e-06,
+      "loss": 0.5876,
+      "step": 828
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6192993763887075,
+      "learning_rate": 9.96771634861309e-06,
+      "loss": 0.6072,
+      "step": 829
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8377241392408228,
+      "learning_rate": 9.967571897687203e-06,
+      "loss": 0.5574,
+      "step": 830
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5074218455849433,
+      "learning_rate": 9.967427125365912e-06,
+      "loss": 0.5624,
+      "step": 831
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.005189790639872,
+      "learning_rate": 9.967282031658584e-06,
+      "loss": 0.5323,
+      "step": 832
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3825511337688314,
+      "learning_rate": 9.967136616574606e-06,
+      "loss": 0.6102,
+      "step": 833
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.592670773526046,
+      "learning_rate": 9.966990880123388e-06,
+      "loss": 0.5577,
+      "step": 834
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.931299454364683,
+      "learning_rate": 9.966844822314357e-06,
+      "loss": 0.5836,
+      "step": 835
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.793291077425198,
+      "learning_rate": 9.966698443156964e-06,
+      "loss": 0.5411,
+      "step": 836
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6495658389731604,
+      "learning_rate": 9.96655174266068e-06,
+      "loss": 0.4965,
+      "step": 837
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5146596647644432,
+      "learning_rate": 9.966404720834992e-06,
+      "loss": 0.5759,
+      "step": 838
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7260735505923195,
+      "learning_rate": 9.966257377689418e-06,
+      "loss": 0.5808,
+      "step": 839
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5124861393632685,
+      "learning_rate": 9.966109713233487e-06,
+      "loss": 0.537,
+      "step": 840
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9507934332500567,
+      "learning_rate": 9.965961727476756e-06,
+      "loss": 0.5306,
+      "step": 841
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6767990291840955,
+      "learning_rate": 9.965813420428796e-06,
+      "loss": 0.6145,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8805452824431472,
+      "learning_rate": 9.965664792099203e-06,
+      "loss": 0.5342,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.596668191985389,
+      "learning_rate": 9.965515842497596e-06,
+      "loss": 0.5203,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8156872051069486,
+      "learning_rate": 9.965366571633608e-06,
+      "loss": 0.5711,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5583898042883626,
+      "learning_rate": 9.965216979516899e-06,
+      "loss": 0.519,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6736508843975682,
+      "learning_rate": 9.965067066157148e-06,
+      "loss": 0.5449,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8870776883869451,
+      "learning_rate": 9.964916831564051e-06,
+      "loss": 0.5474,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7116516860532125,
+      "learning_rate": 9.964766275747331e-06,
+      "loss": 0.5496,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.3252707831884516,
+      "learning_rate": 9.964615398716727e-06,
+      "loss": 0.585,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7580207760405641,
+      "learning_rate": 9.964464200482001e-06,
+      "loss": 0.5908,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 6.180810716710363,
+      "learning_rate": 9.964312681052936e-06,
+      "loss": 0.5385,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5551309645028908,
+      "learning_rate": 9.964160840439335e-06,
+      "loss": 0.5864,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4278084825241908,
+      "learning_rate": 9.964008678651022e-06,
+      "loss": 0.468,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7554944044530334,
+      "learning_rate": 9.96385619569784e-06,
+      "loss": 0.5185,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6771038462422185,
+      "learning_rate": 9.963703391589656e-06,
+      "loss": 0.6201,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 16.006908628016113,
+      "learning_rate": 9.963550266336356e-06,
+      "loss": 0.558,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4464863924809186,
+      "learning_rate": 9.963396819947848e-06,
+      "loss": 0.5225,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6597158664820468,
+      "learning_rate": 9.963243052434057e-06,
+      "loss": 0.5871,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6316903864196992,
+      "learning_rate": 9.963088963804935e-06,
+      "loss": 0.5929,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5871133559724693,
+      "learning_rate": 9.962934554070448e-06,
+      "loss": 0.5901,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9400071273088348,
+      "learning_rate": 9.962779823240588e-06,
+      "loss": 0.5702,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1042290678101203,
+      "learning_rate": 9.962624771325367e-06,
+      "loss": 0.5027,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7921002432875013,
+      "learning_rate": 9.962469398334813e-06,
+      "loss": 0.5356,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.581178866886812,
+      "learning_rate": 9.962313704278981e-06,
+      "loss": 0.5682,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4627098442648818,
+      "learning_rate": 9.962157689167946e-06,
+      "loss": 0.519,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.329483671930689,
+      "learning_rate": 9.962001353011797e-06,
+      "loss": 0.5804,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9277917768002495,
+      "learning_rate": 9.961844695820653e-06,
+      "loss": 0.5498,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5793188293390557,
+      "learning_rate": 9.961687717604649e-06,
+      "loss": 0.5078,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5227241596239927,
+      "learning_rate": 9.96153041837394e-06,
+      "loss": 0.5561,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.139967604556099,
+      "learning_rate": 9.961372798138701e-06,
+      "loss": 0.571,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0375554771390405,
+      "learning_rate": 9.961214856909135e-06,
+      "loss": 0.5316,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.776611150892853,
+      "learning_rate": 9.961056594695457e-06,
+      "loss": 0.5406,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8643510783800241,
+      "learning_rate": 9.960898011507908e-06,
+      "loss": 0.4994,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4742256604281796,
+      "learning_rate": 9.960739107356745e-06,
+      "loss": 0.5004,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8596194888817468,
+      "learning_rate": 9.960579882252252e-06,
+      "loss": 0.5481,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0318071537157785,
+      "learning_rate": 9.960420336204733e-06,
+      "loss": 0.6024,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.615335217042538,
+      "learning_rate": 9.960260469224503e-06,
+      "loss": 0.5532,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8856124442468722,
+      "learning_rate": 9.960100281321912e-06,
+      "loss": 0.5571,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0684094625935234,
+      "learning_rate": 9.95993977250732e-06,
+      "loss": 0.6024,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6864868336485753,
+      "learning_rate": 9.959778942791113e-06,
+      "loss": 0.6173,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.624615411700884,
+      "learning_rate": 9.959617792183698e-06,
+      "loss": 0.5256,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.661687689859257,
+      "learning_rate": 9.9594563206955e-06,
+      "loss": 0.53,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6005222420008236,
+      "learning_rate": 9.959294528336965e-06,
+      "loss": 0.5358,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7525159286083739,
+      "learning_rate": 9.959132415118562e-06,
+      "loss": 0.5955,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7306692929271674,
+      "learning_rate": 9.958969981050779e-06,
+      "loss": 0.5785,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6232752862988116,
+      "learning_rate": 9.958807226144125e-06,
+      "loss": 0.6083,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5767741495613001,
+      "learning_rate": 9.958644150409131e-06,
+      "loss": 0.4971,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8666960383117122,
+      "learning_rate": 9.958480753856348e-06,
+      "loss": 0.5328,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7144573238298764,
+      "learning_rate": 9.958317036496345e-06,
+      "loss": 0.6039,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6783983513904499,
+      "learning_rate": 9.958152998339716e-06,
+      "loss": 0.5868,
+      "step": 891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6489078723090188,
+      "learning_rate": 9.957988639397075e-06,
+      "loss": 0.6145,
+      "step": 892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8975812443867086,
+      "learning_rate": 9.957823959679054e-06,
+      "loss": 0.5699,
+      "step": 893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.611594307359781,
+      "learning_rate": 9.95765895919631e-06,
+      "loss": 0.5479,
+      "step": 894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6522817354250684,
+      "learning_rate": 9.957493637959515e-06,
+      "loss": 0.5958,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8905114408309076,
+      "learning_rate": 9.957327995979369e-06,
+      "loss": 0.5345,
+      "step": 896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.467240298569885,
+      "learning_rate": 9.957162033266585e-06,
+      "loss": 0.5515,
+      "step": 897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6186076252613966,
+      "learning_rate": 9.956995749831901e-06,
+      "loss": 0.5259,
+      "step": 898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0446014227661657,
+      "learning_rate": 9.95682914568608e-06,
+      "loss": 0.5454,
+      "step": 899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6688268405948952,
+      "learning_rate": 9.956662220839894e-06,
+      "loss": 0.5535,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7128713805204633,
+      "learning_rate": 9.95649497530415e-06,
+      "loss": 0.6102,
+      "step": 901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.965374807342687,
+      "learning_rate": 9.956327409089661e-06,
+      "loss": 0.5953,
+      "step": 902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.831792007341404,
+      "learning_rate": 9.956159522207277e-06,
+      "loss": 0.5614,
+      "step": 903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5565606664474412,
+      "learning_rate": 9.955991314667852e-06,
+      "loss": 0.5447,
+      "step": 904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6026980401707079,
+      "learning_rate": 9.955822786482273e-06,
+      "loss": 0.5993,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1705195564671724,
+      "learning_rate": 9.955653937661442e-06,
+      "loss": 0.5342,
+      "step": 906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4770746287985959,
+      "learning_rate": 9.955484768216285e-06,
+      "loss": 0.5131,
+      "step": 907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.81230631547397,
+      "learning_rate": 9.955315278157746e-06,
+      "loss": 0.542,
+      "step": 908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1367991332148866,
+      "learning_rate": 9.955145467496791e-06,
+      "loss": 0.5119,
+      "step": 909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8304261390708023,
+      "learning_rate": 9.954975336244406e-06,
+      "loss": 0.568,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9517494614890825,
+      "learning_rate": 9.954804884411599e-06,
+      "loss": 0.5645,
+      "step": 911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.1276485398244995,
+      "learning_rate": 9.954634112009398e-06,
+      "loss": 0.5643,
+      "step": 912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7617142690377036,
+      "learning_rate": 9.954463019048851e-06,
+      "loss": 0.6276,
+      "step": 913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7217248337831912,
+      "learning_rate": 9.954291605541026e-06,
+      "loss": 0.6072,
+      "step": 914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.501394892380727,
+      "learning_rate": 9.95411987149702e-06,
+      "loss": 0.4689,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7729234428555742,
+      "learning_rate": 9.953947816927934e-06,
+      "loss": 0.5204,
+      "step": 916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.185362223885419,
+      "learning_rate": 9.953775441844909e-06,
+      "loss": 0.521,
+      "step": 917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6079038028760768,
+      "learning_rate": 9.95360274625909e-06,
+      "loss": 0.564,
+      "step": 918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7400941128552325,
+      "learning_rate": 9.953429730181653e-06,
+      "loss": 0.5973,
+      "step": 919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.6726644201146255,
+      "learning_rate": 9.953256393623796e-06,
+      "loss": 0.5313,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8589076341892388,
+      "learning_rate": 9.953082736596728e-06,
+      "loss": 0.5492,
+      "step": 921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.259483192284386,
+      "learning_rate": 9.952908759111686e-06,
+      "loss": 0.5289,
+      "step": 922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6767203705815568,
+      "learning_rate": 9.952734461179928e-06,
+      "loss": 0.5795,
+      "step": 923
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0406854436985276,
+      "learning_rate": 9.952559842812727e-06,
+      "loss": 0.6299,
+      "step": 924
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6748003861195493,
+      "learning_rate": 9.952384904021384e-06,
+      "loss": 0.5819,
+      "step": 925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4822729001771697,
+      "learning_rate": 9.952209644817217e-06,
+      "loss": 0.5239,
+      "step": 926
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5286538811513033,
+      "learning_rate": 9.952034065211564e-06,
+      "loss": 0.5427,
+      "step": 927
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.933443222515339,
+      "learning_rate": 9.951858165215784e-06,
+      "loss": 0.5397,
+      "step": 928
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.8541424133754627,
+      "learning_rate": 9.951681944841261e-06,
+      "loss": 0.6041,
+      "step": 929
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6823530273326184,
+      "learning_rate": 9.95150540409939e-06,
+      "loss": 0.5614,
+      "step": 930
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9767037463874199,
+      "learning_rate": 9.9513285430016e-06,
+      "loss": 0.6268,
+      "step": 931
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7526610700649616,
+      "learning_rate": 9.95115136155933e-06,
+      "loss": 0.5317,
+      "step": 932
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6740726398040634,
+      "learning_rate": 9.950973859784044e-06,
+      "loss": 0.6091,
+      "step": 933
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.729628226005561,
+      "learning_rate": 9.950796037687224e-06,
+      "loss": 0.494,
+      "step": 934
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.498242238639019,
+      "learning_rate": 9.950617895280378e-06,
+      "loss": 0.5191,
+      "step": 935
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8158851160510832,
+      "learning_rate": 9.950439432575029e-06,
+      "loss": 0.593,
+      "step": 936
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.713610419521136,
+      "learning_rate": 9.950260649582727e-06,
+      "loss": 0.6038,
+      "step": 937
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6149588645914572,
+      "learning_rate": 9.950081546315036e-06,
+      "loss": 0.4825,
+      "step": 938
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9144448777184389,
+      "learning_rate": 9.949902122783543e-06,
+      "loss": 0.4915,
+      "step": 939
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.4699324692377584,
+      "learning_rate": 9.94972237899986e-06,
+      "loss": 0.5352,
+      "step": 940
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7020448788853424,
+      "learning_rate": 9.949542314975614e-06,
+      "loss": 0.5009,
+      "step": 941
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.2273994812048663,
+      "learning_rate": 9.949361930722454e-06,
+      "loss": 0.5665,
+      "step": 942
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9062644487692846,
+      "learning_rate": 9.949181226252052e-06,
+      "loss": 0.5697,
+      "step": 943
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.397948146906995,
+      "learning_rate": 9.949000201576099e-06,
+      "loss": 0.5531,
+      "step": 944
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7691874044800316,
+      "learning_rate": 9.948818856706307e-06,
+      "loss": 0.5847,
+      "step": 945
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6817228788006264,
+      "learning_rate": 9.948637191654409e-06,
+      "loss": 0.5453,
+      "step": 946
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7114398697803828,
+      "learning_rate": 9.948455206432158e-06,
+      "loss": 0.6103,
+      "step": 947
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.9617675799843879,
+      "learning_rate": 9.94827290105133e-06,
+      "loss": 0.5433,
+      "step": 948
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6086781572250257,
+      "learning_rate": 9.948090275523715e-06,
+      "loss": 0.5977,
+      "step": 949
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0295545831625303,
+      "learning_rate": 9.947907329861134e-06,
+      "loss": 0.5381,
+      "step": 950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 2.0137677234649014,
+      "learning_rate": 9.947724064075424e-06,
+      "loss": 0.5281,
+      "step": 951
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8715057144270584,
+      "learning_rate": 9.947540478178437e-06,
+      "loss": 0.5564,
+      "step": 952
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6468170055125908,
+      "learning_rate": 9.947356572182054e-06,
+      "loss": 0.5265,
+      "step": 953
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5432641442620518,
+      "learning_rate": 9.94717234609817e-06,
+      "loss": 0.5669,
+      "step": 954
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.490222437757155,
+      "learning_rate": 9.946987799938709e-06,
+      "loss": 0.5394,
+      "step": 955
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8879803402680078,
+      "learning_rate": 9.94680293371561e-06,
+      "loss": 0.6243,
+      "step": 956
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7762821660739267,
+      "learning_rate": 9.946617747440831e-06,
+      "loss": 0.5436,
+      "step": 957
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8238243745218685,
+      "learning_rate": 9.946432241126356e-06,
+      "loss": 0.4932,
+      "step": 958
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7498322789663598,
+      "learning_rate": 9.946246414784185e-06,
+      "loss": 0.6133,
+      "step": 959
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.694567505767062,
+      "learning_rate": 9.94606026842634e-06,
+      "loss": 0.5302,
+      "step": 960
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3462244393563485,
+      "learning_rate": 9.945873802064868e-06,
+      "loss": 0.5256,
+      "step": 961
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7490947370116301,
+      "learning_rate": 9.94568701571183e-06,
+      "loss": 0.595,
+      "step": 962
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7079264990411434,
+      "learning_rate": 9.945499909379313e-06,
+      "loss": 0.4981,
+      "step": 963
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9685712106007467,
+      "learning_rate": 9.94531248307942e-06,
+      "loss": 0.5312,
+      "step": 964
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.1674323740287,
+      "learning_rate": 9.945124736824279e-06,
+      "loss": 0.5468,
+      "step": 965
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.754151031059945,
+      "learning_rate": 9.944936670626037e-06,
+      "loss": 0.5988,
+      "step": 966
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9532201134233702,
+      "learning_rate": 9.944748284496862e-06,
+      "loss": 0.5391,
+      "step": 967
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4879688469545818,
+      "learning_rate": 9.94455957844894e-06,
+      "loss": 0.5002,
+      "step": 968
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.363012334286684,
+      "learning_rate": 9.944370552494483e-06,
+      "loss": 0.5906,
+      "step": 969
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0825769972367367,
+      "learning_rate": 9.944181206645717e-06,
+      "loss": 0.5484,
+      "step": 970
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.2140134698920337,
+      "learning_rate": 9.943991540914895e-06,
+      "loss": 0.5803,
+      "step": 971
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.518754069989875,
+      "learning_rate": 9.94380155531429e-06,
+      "loss": 0.4938,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8423085440283506,
+      "learning_rate": 9.94361124985619e-06,
+      "loss": 0.5254,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8612029622640518,
+      "learning_rate": 9.943420624552912e-06,
+      "loss": 0.5604,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.1594880861182837,
+      "learning_rate": 9.943229679416781e-06,
+      "loss": 0.5699,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5165155064905773,
+      "learning_rate": 9.94303841446016e-06,
+      "loss": 0.5495,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.770356362868865,
+      "learning_rate": 9.942846829695421e-06,
+      "loss": 0.5334,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7861187482712447,
+      "learning_rate": 9.942654925134956e-06,
+      "loss": 0.6102,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5461562958293462,
+      "learning_rate": 9.942462700791184e-06,
+      "loss": 0.5183,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5126002590025025,
+      "learning_rate": 9.94227015667654e-06,
+      "loss": 0.5805,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5423802641756543,
+      "learning_rate": 9.942077292803482e-06,
+      "loss": 0.4713,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.4954707255336817,
+      "learning_rate": 9.94188410918449e-06,
+      "loss": 0.5057,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6072627022730294,
+      "learning_rate": 9.94169060583206e-06,
+      "loss": 0.5806,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.516686615141718,
+      "learning_rate": 9.94149678275871e-06,
+      "loss": 0.503,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.488974357923619,
+      "learning_rate": 9.941302639976986e-06,
+      "loss": 0.5455,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7606502347050086,
+      "learning_rate": 9.941108177499443e-06,
+      "loss": 0.59,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7371916913477436,
+      "learning_rate": 9.940913395338666e-06,
+      "loss": 0.5393,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7164326666715825,
+      "learning_rate": 9.940718293507256e-06,
+      "loss": 0.585,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.706301106054119,
+      "learning_rate": 9.940522872017835e-06,
+      "loss": 0.5801,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5350077699723788,
+      "learning_rate": 9.940327130883047e-06,
+      "loss": 0.5097,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8882076258928164,
+      "learning_rate": 9.940131070115556e-06,
+      "loss": 0.5071,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5351914381895546,
+      "learning_rate": 9.939934689728049e-06,
+      "loss": 0.5284,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5247635231434422,
+      "learning_rate": 9.939737989733229e-06,
+      "loss": 0.5585,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.798598776582196,
+      "learning_rate": 9.939540970143822e-06,
+      "loss": 0.5123,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6452957848803892,
+      "learning_rate": 9.939343630972577e-06,
+      "loss": 0.5933,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8192408615607587,
+      "learning_rate": 9.93914597223226e-06,
+      "loss": 0.5794,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7808330491704183,
+      "learning_rate": 9.93894799393566e-06,
+      "loss": 0.5296,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9023417591585585,
+      "learning_rate": 9.938749696095587e-06,
+      "loss": 0.524,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8152193989706902,
+      "learning_rate": 9.938551078724868e-06,
+      "loss": 0.5667,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 13.572818665937177,
+      "learning_rate": 9.938352141836353e-06,
+      "loss": 0.6009,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5814671479256166,
+      "learning_rate": 9.938152885442918e-06,
+      "loss": 0.5659,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.333974539644457,
+      "learning_rate": 9.937953309557449e-06,
+      "loss": 0.503,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4972653038883414,
+      "learning_rate": 9.937753414192862e-06,
+      "loss": 0.4959,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6770736242011542,
+      "learning_rate": 9.937553199362087e-06,
+      "loss": 0.6006,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8935001269940408,
+      "learning_rate": 9.937352665078079e-06,
+      "loss": 0.5692,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7698208612485533,
+      "learning_rate": 9.937151811353813e-06,
+      "loss": 0.4887,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6250383997315418,
+      "learning_rate": 9.936950638202285e-06,
+      "loss": 0.5898,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.705044983447187,
+      "learning_rate": 9.936749145636507e-06,
+      "loss": 0.5411,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7714644906235089,
+      "learning_rate": 9.936547333669518e-06,
+      "loss": 0.5398,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6274548418574036,
+      "learning_rate": 9.936345202314375e-06,
+      "loss": 0.5914,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6620052754199686,
+      "learning_rate": 9.936142751584155e-06,
+      "loss": 0.5855,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5046377889573184,
+      "learning_rate": 9.935939981491956e-06,
+      "loss": 0.5198,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.748746586721077,
+      "learning_rate": 9.935736892050896e-06,
+      "loss": 0.5316,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.730735977778147,
+      "learning_rate": 9.935533483274116e-06,
+      "loss": 0.523,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6197527040727683,
+      "learning_rate": 9.935329755174778e-06,
+      "loss": 0.5767,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5736824816812927,
+      "learning_rate": 9.93512570776606e-06,
+      "loss": 0.5403,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7339996444043406,
+      "learning_rate": 9.934921341061163e-06,
+      "loss": 0.5459,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5993971027053606,
+      "learning_rate": 9.934716655073313e-06,
+      "loss": 0.5483,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6188171864925958,
+      "learning_rate": 9.93451164981575e-06,
+      "loss": 0.5725,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6627933713150337,
+      "learning_rate": 9.934306325301738e-06,
+      "loss": 0.5895,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3699299823882745,
+      "learning_rate": 9.934100681544565e-06,
+      "loss": 0.543,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6120803371109612,
+      "learning_rate": 9.93389471855753e-06,
+      "loss": 0.5759,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6643023868542766,
+      "learning_rate": 9.93368843635396e-06,
+      "loss": 0.5167,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9480069936880335,
+      "learning_rate": 9.933481834947202e-06,
+      "loss": 0.5837,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6510826524848268,
+      "learning_rate": 9.933274914350626e-06,
+      "loss": 0.5808,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6457080260430015,
+      "learning_rate": 9.933067674577615e-06,
+      "loss": 0.5836,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7591246229301212,
+      "learning_rate": 9.932860115641578e-06,
+      "loss": 0.5797,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5784890758539505,
+      "learning_rate": 9.932652237555944e-06,
+      "loss": 0.5733,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.564394300783467,
+      "learning_rate": 9.932444040334164e-06,
+      "loss": 0.4743,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.730185369519864,
+      "learning_rate": 9.932235523989708e-06,
+      "loss": 0.5221,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.565744584361003,
+      "learning_rate": 9.932026688536064e-06,
+      "loss": 0.5422,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4576745608130557,
+      "learning_rate": 9.931817533986746e-06,
+      "loss": 0.5837,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3256309374259265,
+      "learning_rate": 9.931608060355285e-06,
+      "loss": 0.5568,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6115044490699204,
+      "learning_rate": 9.931398267655234e-06,
+      "loss": 0.5436,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.615101172123175,
+      "learning_rate": 9.931188155900166e-06,
+      "loss": 0.5512,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.264488167745833,
+      "learning_rate": 9.930977725103677e-06,
+      "loss": 0.5564,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9128849680919735,
+      "learning_rate": 9.930766975279377e-06,
+      "loss": 0.5419,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.63106770573509,
+      "learning_rate": 9.930555906440904e-06,
+      "loss": 0.5486,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5462764376684095,
+      "learning_rate": 9.930344518601915e-06,
+      "loss": 0.5502,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0242733585929527,
+      "learning_rate": 9.930132811776086e-06,
+      "loss": 0.5811,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.5013362333595275,
+      "learning_rate": 9.929920785977113e-06,
+      "loss": 0.542,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.3139349306950163,
+      "learning_rate": 9.929708441218713e-06,
+      "loss": 0.5228,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.666985306124112,
+      "learning_rate": 9.929495777514627e-06,
+      "loss": 0.5512,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7912828430283418,
+      "learning_rate": 9.929282794878612e-06,
+      "loss": 0.5435,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.410374826009012,
+      "learning_rate": 9.92906949332445e-06,
+      "loss": 0.5212,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.929853020598953,
+      "learning_rate": 9.928855872865939e-06,
+      "loss": 0.5749,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9628600701257146,
+      "learning_rate": 9.9286419335169e-06,
+      "loss": 0.6122,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6109600879356432,
+      "learning_rate": 9.928427675291176e-06,
+      "loss": 0.59,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5068988521060063,
+      "learning_rate": 9.928213098202628e-06,
+      "loss": 0.5009,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.46949249489788,
+      "learning_rate": 9.92799820226514e-06,
+      "loss": 0.5564,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6146957557987986,
+      "learning_rate": 9.927782987492615e-06,
+      "loss": 0.5256,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5439560808845414,
+      "learning_rate": 9.927567453898976e-06,
+      "loss": 0.5295,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 3.1587152677864214,
+      "learning_rate": 9.92735160149817e-06,
+      "loss": 0.5235,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.058990840543154,
+      "learning_rate": 9.927135430304163e-06,
+      "loss": 0.514,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6311768782813831,
+      "learning_rate": 9.926918940330937e-06,
+      "loss": 0.514,
+      "step": 1055
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.510142259114802,
+      "learning_rate": 9.926702131592499e-06,
+      "loss": 0.5425,
+      "step": 1056
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8900908226528865,
+      "learning_rate": 9.92648500410288e-06,
+      "loss": 0.4955,
+      "step": 1057
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9029963982653995,
+      "learning_rate": 9.926267557876126e-06,
+      "loss": 0.5404,
+      "step": 1058
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7082840035148038,
+      "learning_rate": 9.926049792926306e-06,
+      "loss": 0.582,
+      "step": 1059
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.63287933963858,
+      "learning_rate": 9.925831709267508e-06,
+      "loss": 0.6011,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6044060767130703,
+      "learning_rate": 9.925613306913841e-06,
+      "loss": 0.5781,
+      "step": 1061
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.9724293579053236,
+      "learning_rate": 9.925394585879437e-06,
+      "loss": 0.623,
+      "step": 1062
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.6565961273592924,
+      "learning_rate": 9.925175546178446e-06,
+      "loss": 0.5364,
+      "step": 1063
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.291512772489767,
+      "learning_rate": 9.924956187825042e-06,
+      "loss": 0.6025,
+      "step": 1064
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.161368285830542,
+      "learning_rate": 9.924736510833414e-06,
+      "loss": 0.5129,
+      "step": 1065
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.12218620517083,
+      "learning_rate": 9.924516515217777e-06,
+      "loss": 0.5816,
+      "step": 1066
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6374160170409087,
+      "learning_rate": 9.924296200992363e-06,
+      "loss": 0.5581,
+      "step": 1067
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6925732716441213,
+      "learning_rate": 9.924075568171426e-06,
+      "loss": 0.518,
+      "step": 1068
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.46094466049942,
+      "learning_rate": 9.923854616769242e-06,
+      "loss": 0.5098,
+      "step": 1069
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.0717882703457695,
+      "learning_rate": 9.923633346800106e-06,
+      "loss": 0.5407,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8481453871417897,
+      "learning_rate": 9.923411758278332e-06,
+      "loss": 0.52,
+      "step": 1071
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.2527526706634229,
+      "learning_rate": 9.923189851218259e-06,
+      "loss": 0.4865,
+      "step": 1072
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8806352326109435,
+      "learning_rate": 9.922967625634242e-06,
+      "loss": 0.5239,
+      "step": 1073
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 2.115634407199955,
+      "learning_rate": 9.92274508154066e-06,
+      "loss": 0.4786,
+      "step": 1074
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8050603884343028,
+      "learning_rate": 9.92252221895191e-06,
+      "loss": 0.5277,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.224550206429174,
+      "learning_rate": 9.922299037882414e-06,
+      "loss": 0.606,
+      "step": 1076
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8670289813816698,
+      "learning_rate": 9.922075538346609e-06,
+      "loss": 0.5474,
+      "step": 1077
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8822859173716924,
+      "learning_rate": 9.921851720358955e-06,
+      "loss": 0.562,
+      "step": 1078
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5965102812637362,
+      "learning_rate": 9.921627583933933e-06,
+      "loss": 0.5324,
+      "step": 1079
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6361916028300054,
+      "learning_rate": 9.921403129086045e-06,
+      "loss": 0.5393,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.7839271079939816,
+      "learning_rate": 9.921178355829814e-06,
+      "loss": 0.5129,
+      "step": 1081
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.417877062497937,
+      "learning_rate": 9.92095326417978e-06,
+      "loss": 0.5705,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4910706289462825,
+      "learning_rate": 9.920727854150506e-06,
+      "loss": 0.5297,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8269739747028728,
+      "learning_rate": 9.920502125756578e-06,
+      "loss": 0.5774,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7114254833253686,
+      "learning_rate": 9.920276079012599e-06,
+      "loss": 0.5932,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6518328394775394,
+      "learning_rate": 9.920049713933194e-06,
+      "loss": 0.5203,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6603079275322299,
+      "learning_rate": 9.91982303053301e-06,
+      "loss": 0.5622,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.015170251934802,
+      "learning_rate": 9.91959602882671e-06,
+      "loss": 0.5836,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0624073194477806,
+      "learning_rate": 9.919368708828984e-06,
+      "loss": 0.6064,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8912433537277038,
+      "learning_rate": 9.919141070554536e-06,
+      "loss": 0.5725,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6990275821998273,
+      "learning_rate": 9.918913114018098e-06,
+      "loss": 0.5509,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5317338871178732,
+      "learning_rate": 9.918684839234417e-06,
+      "loss": 0.5311,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.75601592082343,
+      "learning_rate": 9.918456246218257e-06,
+      "loss": 0.553,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.41188752375783,
+      "learning_rate": 9.918227334984415e-06,
+      "loss": 0.5342,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6735406371518926,
+      "learning_rate": 9.9179981055477e-06,
+      "loss": 0.5491,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6717981258329313,
+      "learning_rate": 9.917768557922938e-06,
+      "loss": 0.5913,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6377027054064629,
+      "learning_rate": 9.917538692124985e-06,
+      "loss": 0.6031,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.415486373513553,
+      "learning_rate": 9.917308508168712e-06,
+      "loss": 0.5209,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4596525792172892,
+      "learning_rate": 9.91707800606901e-06,
+      "loss": 0.5589,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4543368451562873,
+      "learning_rate": 9.916847185840794e-06,
+      "loss": 0.5676,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5111023543429314,
+      "learning_rate": 9.916616047498997e-06,
+      "loss": 0.5702,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1472788063836186,
+      "learning_rate": 9.916384591058572e-06,
+      "loss": 0.5149,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5943060132923663,
+      "learning_rate": 9.916152816534498e-06,
+      "loss": 0.5309,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.819489143091455,
+      "learning_rate": 9.915920723941766e-06,
+      "loss": 0.5937,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9050453179069587,
+      "learning_rate": 9.915688313295394e-06,
+      "loss": 0.6036,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8463856618307843,
+      "learning_rate": 9.915455584610421e-06,
+      "loss": 0.5552,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7220375909189171,
+      "learning_rate": 9.915222537901901e-06,
+      "loss": 0.5955,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6414912405376276,
+      "learning_rate": 9.91498917318491e-06,
+      "loss": 0.5723,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5847141620042753,
+      "learning_rate": 9.914755490474552e-06,
+      "loss": 0.4816,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5016225084197317,
+      "learning_rate": 9.914521489785942e-06,
+      "loss": 0.5772,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.420942215296239,
+      "learning_rate": 9.91428717113422e-06,
+      "loss": 0.5954,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8200109828718756,
+      "learning_rate": 9.914052534534549e-06,
+      "loss": 0.5476,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.707095152757118,
+      "learning_rate": 9.913817580002105e-06,
+      "loss": 0.5576,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6163665457304273,
+      "learning_rate": 9.913582307552093e-06,
+      "loss": 0.5664,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9017239011073102,
+      "learning_rate": 9.913346717199732e-06,
+      "loss": 0.552,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.644169730197281,
+      "learning_rate": 9.913110808960267e-06,
+      "loss": 0.5806,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.3921308909827137,
+      "learning_rate": 9.912874582848962e-06,
+      "loss": 0.5274,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.2225075993943535,
+      "learning_rate": 9.912638038881095e-06,
+      "loss": 0.5494,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.778695859631835,
+      "learning_rate": 9.912401177071975e-06,
+      "loss": 0.574,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4082362118868081,
+      "learning_rate": 9.912163997436924e-06,
+      "loss": 0.5081,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.570397398568426,
+      "learning_rate": 9.911926499991289e-06,
+      "loss": 0.4997,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6477497021227163,
+      "learning_rate": 9.911688684750434e-06,
+      "loss": 0.491,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.084614691566177,
+      "learning_rate": 9.911450551729748e-06,
+      "loss": 0.5064,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7261631412640468,
+      "learning_rate": 9.911212100944635e-06,
+      "loss": 0.5378,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7702709946863495,
+      "learning_rate": 9.910973332410525e-06,
+      "loss": 0.5366,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.425931522713767,
+      "learning_rate": 9.910734246142863e-06,
+      "loss": 0.6441,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1138034968221866,
+      "learning_rate": 9.910494842157122e-06,
+      "loss": 0.6148,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7728583640950348,
+      "learning_rate": 9.910255120468786e-06,
+      "loss": 0.579,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9569603924074712,
+      "learning_rate": 9.910015081093368e-06,
+      "loss": 0.5742,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.977089611745831,
+      "learning_rate": 9.909774724046398e-06,
+      "loss": 0.5104,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.018412895390878,
+      "learning_rate": 9.909534049343425e-06,
+      "loss": 0.549,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.681252942241913,
+      "learning_rate": 9.909293057000023e-06,
+      "loss": 0.5117,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.779568602619412,
+      "learning_rate": 9.909051747031779e-06,
+      "loss": 0.5848,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8169564653913255,
+      "learning_rate": 9.908810119454314e-06,
+      "loss": 0.6466,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9603714693645404,
+      "learning_rate": 9.90856817428325e-06,
+      "loss": 0.534,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7850962623196969,
+      "learning_rate": 9.908325911534252e-06,
+      "loss": 0.5557,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0202431046437184,
+      "learning_rate": 9.908083331222984e-06,
+      "loss": 0.5159,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3433708111638265,
+      "learning_rate": 9.907840433365147e-06,
+      "loss": 0.6575,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8939310334819288,
+      "learning_rate": 9.907597217976455e-06,
+      "loss": 0.5019,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5784776243039382,
+      "learning_rate": 9.90735368507264e-06,
+      "loss": 0.4981,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.124867962676086,
+      "learning_rate": 9.907109834669465e-06,
+      "loss": 0.4753,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8039609130892484,
+      "learning_rate": 9.9068656667827e-06,
+      "loss": 0.5625,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5044615506449317,
+      "learning_rate": 9.906621181428147e-06,
+      "loss": 0.4979,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8432820805843526,
+      "learning_rate": 9.906376378621622e-06,
+      "loss": 0.5249,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0730345028416575,
+      "learning_rate": 9.906131258378962e-06,
+      "loss": 0.5799,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4493145163681083,
+      "learning_rate": 9.90588582071603e-06,
+      "loss": 0.6159,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.409714049828381,
+      "learning_rate": 9.905640065648702e-06,
+      "loss": 0.4979,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5676881416873902,
+      "learning_rate": 9.90539399319288e-06,
+      "loss": 0.5713,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7848371650749066,
+      "learning_rate": 9.905147603364482e-06,
+      "loss": 0.5889,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8876131380260615,
+      "learning_rate": 9.904900896179453e-06,
+      "loss": 0.605,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.829473423719891,
+      "learning_rate": 9.90465387165375e-06,
+      "loss": 0.457,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8076063906281903,
+      "learning_rate": 9.904406529803358e-06,
+      "loss": 0.5512,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7666782174284146,
+      "learning_rate": 9.90415887064428e-06,
+      "loss": 0.5702,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7877188920761047,
+      "learning_rate": 9.903910894192539e-06,
+      "loss": 0.5167,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.742603751782403,
+      "learning_rate": 9.903662600464178e-06,
+      "loss": 0.5379,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2913534156314952,
+      "learning_rate": 9.90341398947526e-06,
+      "loss": 0.6054,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2253206110079993,
+      "learning_rate": 9.903165061241874e-06,
+      "loss": 0.5382,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6770581552745216,
+      "learning_rate": 9.90291581578012e-06,
+      "loss": 0.5075,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9240297676423572,
+      "learning_rate": 9.90266625310613e-06,
+      "loss": 0.5482,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7813365269240671,
+      "learning_rate": 9.902416373236045e-06,
+      "loss": 0.5696,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4683662296453355,
+      "learning_rate": 9.902166176186033e-06,
+      "loss": 0.506,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.676338790144313,
+      "learning_rate": 9.901915661972283e-06,
+      "loss": 0.5263,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8655221044721126,
+      "learning_rate": 9.901664830611001e-06,
+      "loss": 0.5655,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.2925530866968598,
+      "learning_rate": 9.90141368211842e-06,
+      "loss": 0.5282,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7252102077236804,
+      "learning_rate": 9.901162216510782e-06,
+      "loss": 0.516,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5643643831272696,
+      "learning_rate": 9.900910433804362e-06,
+      "loss": 0.5535,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8168170881362014,
+      "learning_rate": 9.900658334015447e-06,
+      "loss": 0.54,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.476310146007673,
+      "learning_rate": 9.90040591716035e-06,
+      "loss": 0.5048,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.921811761595845,
+      "learning_rate": 9.900153183255399e-06,
+      "loss": 0.5389,
+      "step": 1169
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4756422899686497,
+      "learning_rate": 9.899900132316948e-06,
+      "loss": 0.5052,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.802787367686697,
+      "learning_rate": 9.89964676436137e-06,
+      "loss": 0.5695,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6473922553722975,
+      "learning_rate": 9.899393079405052e-06,
+      "loss": 0.509,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.5675827943889926,
+      "learning_rate": 9.899139077464415e-06,
+      "loss": 0.5399,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7172682241451236,
+      "learning_rate": 9.898884758555886e-06,
+      "loss": 0.5703,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0029849527498562,
+      "learning_rate": 9.898630122695923e-06,
+      "loss": 0.5917,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6910533251645322,
+      "learning_rate": 9.898375169901e-06,
+      "loss": 0.5799,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.194931525722795,
+      "learning_rate": 9.89811990018761e-06,
+      "loss": 0.5571,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3894673980832237,
+      "learning_rate": 9.89786431357227e-06,
+      "loss": 0.5292,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8268981789954277,
+      "learning_rate": 9.897608410071516e-06,
+      "loss": 0.5117,
+      "step": 1179
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4192589651751089,
+      "learning_rate": 9.897352189701905e-06,
+      "loss": 0.496,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6878047301647991,
+      "learning_rate": 9.897095652480015e-06,
+      "loss": 0.5488,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6771007886010452,
+      "learning_rate": 9.896838798422442e-06,
+      "loss": 0.4723,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.769328655821698,
+      "learning_rate": 9.896581627545803e-06,
+      "loss": 0.507,
+      "step": 1183
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3991607416495504,
+      "learning_rate": 9.89632413986674e-06,
+      "loss": 0.4844,
+      "step": 1184
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7840591629159244,
+      "learning_rate": 9.89606633540191e-06,
+      "loss": 0.5406,
+      "step": 1185
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6645683450276998,
+      "learning_rate": 9.895808214167993e-06,
+      "loss": 0.6108,
+      "step": 1186
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.468709595267863,
+      "learning_rate": 9.895549776181688e-06,
+      "loss": 0.5086,
+      "step": 1187
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5995715842468021,
+      "learning_rate": 9.895291021459717e-06,
+      "loss": 0.5424,
+      "step": 1188
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.942887596448376,
+      "learning_rate": 9.89503195001882e-06,
+      "loss": 0.5499,
+      "step": 1189
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.0231796313865105,
+      "learning_rate": 9.894772561875762e-06,
+      "loss": 0.5459,
+      "step": 1190
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3985537761117157,
+      "learning_rate": 9.89451285704732e-06,
+      "loss": 0.5304,
+      "step": 1191
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.2671654084436923,
+      "learning_rate": 9.8942528355503e-06,
+      "loss": 0.6083,
+      "step": 1192
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4890698610583495,
+      "learning_rate": 9.893992497401525e-06,
+      "loss": 0.4915,
+      "step": 1193
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.179985887695711,
+      "learning_rate": 9.893731842617837e-06,
+      "loss": 0.5406,
+      "step": 1194
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7280838176852462,
+      "learning_rate": 9.8934708712161e-06,
+      "loss": 0.5482,
+      "step": 1195
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.009635792352556,
+      "learning_rate": 9.8932095832132e-06,
+      "loss": 0.5418,
+      "step": 1196
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.8085803562648344,
+      "learning_rate": 9.89294797862604e-06,
+      "loss": 0.5237,
+      "step": 1197
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.9624385187839906,
+      "learning_rate": 9.892686057471546e-06,
+      "loss": 0.5413,
+      "step": 1198
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3496477164764153,
+      "learning_rate": 9.892423819766665e-06,
+      "loss": 0.5749,
+      "step": 1199
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5307520083689794,
+      "learning_rate": 9.892161265528364e-06,
+      "loss": 0.5132,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 3.0645863515855494,
+      "learning_rate": 9.891898394773627e-06,
+      "loss": 0.5128,
+      "step": 1201
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8214767289810571,
+      "learning_rate": 9.891635207519465e-06,
+      "loss": 0.5872,
+      "step": 1202
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7236140506117072,
+      "learning_rate": 9.891371703782903e-06,
+      "loss": 0.487,
+      "step": 1203
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3945748066681585,
+      "learning_rate": 9.891107883580991e-06,
+      "loss": 0.4747,
+      "step": 1204
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.797244860512192,
+      "learning_rate": 9.890843746930798e-06,
+      "loss": 0.5223,
+      "step": 1205
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7982323161894755,
+      "learning_rate": 9.890579293849411e-06,
+      "loss": 0.6009,
+      "step": 1206
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.7127512640065596,
+      "learning_rate": 9.890314524353943e-06,
+      "loss": 0.5318,
+      "step": 1207
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6932316447311612,
+      "learning_rate": 9.890049438461522e-06,
+      "loss": 0.5865,
+      "step": 1208
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6083723596524213,
+      "learning_rate": 9.889784036189299e-06,
+      "loss": 0.5334,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4163273722969887,
+      "learning_rate": 9.889518317554446e-06,
+      "loss": 0.4844,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6544059924690359,
+      "learning_rate": 9.889252282574154e-06,
+      "loss": 0.586,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8249012799504838,
+      "learning_rate": 9.888985931265636e-06,
+      "loss": 0.5191,
+      "step": 1212
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.437528382478115,
+      "learning_rate": 9.888719263646123e-06,
+      "loss": 0.5303,
+      "step": 1213
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7918008831203407,
+      "learning_rate": 9.888452279732869e-06,
+      "loss": 0.5279,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8608842217306174,
+      "learning_rate": 9.888184979543147e-06,
+      "loss": 0.5879,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6838552323971444,
+      "learning_rate": 9.887917363094251e-06,
+      "loss": 0.5707,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7540621116587611,
+      "learning_rate": 9.887649430403496e-06,
+      "loss": 0.5425,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9588091992275638,
+      "learning_rate": 9.887381181488218e-06,
+      "loss": 0.4948,
+      "step": 1218
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6194388994458946,
+      "learning_rate": 9.88711261636577e-06,
+      "loss": 0.556,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.756001502283283,
+      "learning_rate": 9.886843735053529e-06,
+      "loss": 0.5522,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6867062350229141,
+      "learning_rate": 9.88657453756889e-06,
+      "loss": 0.5829,
+      "step": 1221
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0274529200836353,
+      "learning_rate": 9.886305023929275e-06,
+      "loss": 0.5188,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.008126032463789,
+      "learning_rate": 9.886035194152112e-06,
+      "loss": 0.5163,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.471211758607614,
+      "learning_rate": 9.885765048254868e-06,
+      "loss": 0.5069,
+      "step": 1224
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6639189530910646,
+      "learning_rate": 9.885494586255014e-06,
+      "loss": 0.6089,
+      "step": 1225
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4257934648675714,
+      "learning_rate": 9.885223808170052e-06,
+      "loss": 0.5742,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5721368863947638,
+      "learning_rate": 9.8849527140175e-06,
+      "loss": 0.4974,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5520222587908474,
+      "learning_rate": 9.884681303814897e-06,
+      "loss": 0.5552,
+      "step": 1228
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7262446028307725,
+      "learning_rate": 9.884409577579806e-06,
+      "loss": 0.4991,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5708671441469854,
+      "learning_rate": 9.884137535329804e-06,
+      "loss": 0.5775,
+      "step": 1230
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.952333054597408,
+      "learning_rate": 9.88386517708249e-06,
+      "loss": 0.5548,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6833378299804076,
+      "learning_rate": 9.88359250285549e-06,
+      "loss": 0.5882,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.520028970153725,
+      "learning_rate": 9.883319512666445e-06,
+      "loss": 0.518,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6256232805447575,
+      "learning_rate": 9.883046206533013e-06,
+      "loss": 0.5202,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3382413282029975,
+      "learning_rate": 9.88277258447288e-06,
+      "loss": 0.5141,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7050831348380444,
+      "learning_rate": 9.88249864650375e-06,
+      "loss": 0.564,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.615472673312182,
+      "learning_rate": 9.882224392643345e-06,
+      "loss": 0.494,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0195638334355883,
+      "learning_rate": 9.881949822909407e-06,
+      "loss": 0.5014,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4666438591452124,
+      "learning_rate": 9.881674937319701e-06,
+      "loss": 0.5253,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.72783375265924,
+      "learning_rate": 9.881399735892015e-06,
+      "loss": 0.5927,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7029540999984278,
+      "learning_rate": 9.881124218644149e-06,
+      "loss": 0.5219,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.612080122698668,
+      "learning_rate": 9.880848385593933e-06,
+      "loss": 0.4798,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0919602926902234,
+      "learning_rate": 9.880572236759212e-06,
+      "loss": 0.4861,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6938457806945626,
+      "learning_rate": 9.880295772157851e-06,
+      "loss": 0.5135,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5911468138287221,
+      "learning_rate": 9.880018991807738e-06,
+      "loss": 0.5113,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.661098829903792,
+      "learning_rate": 9.87974189572678e-06,
+      "loss": 0.5255,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6772855822065804,
+      "learning_rate": 9.879464483932903e-06,
+      "loss": 0.605,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6571045413883058,
+      "learning_rate": 9.87918675644406e-06,
+      "loss": 0.5996,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5724475114954535,
+      "learning_rate": 9.878908713278215e-06,
+      "loss": 0.5825,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6097375050400653,
+      "learning_rate": 9.87863035445336e-06,
+      "loss": 0.5275,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9651590123796776,
+      "learning_rate": 9.8783516799875e-06,
+      "loss": 0.5599,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.708592695708444,
+      "learning_rate": 9.878072689898672e-06,
+      "loss": 0.5862,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.889189301901958,
+      "learning_rate": 9.87779338420492e-06,
+      "loss": 0.5906,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9150809881336093,
+      "learning_rate": 9.877513762924318e-06,
+      "loss": 0.5375,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.563942624315454,
+      "learning_rate": 9.877233826074956e-06,
+      "loss": 0.5229,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.885316798708101,
+      "learning_rate": 9.876953573674946e-06,
+      "loss": 0.535,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6721612372487031,
+      "learning_rate": 9.876673005742417e-06,
+      "loss": 0.5169,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7369683493765,
+      "learning_rate": 9.876392122295526e-06,
+      "loss": 0.5679,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6703920825731133,
+      "learning_rate": 9.876110923352445e-06,
+      "loss": 0.5365,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.793162907991587,
+      "learning_rate": 9.875829408931363e-06,
+      "loss": 0.5848,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.829510897559637,
+      "learning_rate": 9.875547579050499e-06,
+      "loss": 0.5124,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4748272604271628,
+      "learning_rate": 9.875265433728085e-06,
+      "loss": 0.5415,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0439648587007704,
+      "learning_rate": 9.874982972982374e-06,
+      "loss": 0.512,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6352923371075194,
+      "learning_rate": 9.874700196831641e-06,
+      "loss": 0.4964,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.736061006334386,
+      "learning_rate": 9.874417105294184e-06,
+      "loss": 0.5157,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5881587037058288,
+      "learning_rate": 9.874133698388316e-06,
+      "loss": 0.5706,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8358642890634616,
+      "learning_rate": 9.873849976132375e-06,
+      "loss": 0.5549,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.57107991887304,
+      "learning_rate": 9.873565938544714e-06,
+      "loss": 0.5568,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6572499251655388,
+      "learning_rate": 9.873281585643715e-06,
+      "loss": 0.6017,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.826058026132881,
+      "learning_rate": 9.87299691744777e-06,
+      "loss": 0.5527,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.763590780015478,
+      "learning_rate": 9.8727119339753e-06,
+      "loss": 0.528,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.812536053303485,
+      "learning_rate": 9.872426635244742e-06,
+      "loss": 0.5591,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6504726498169892,
+      "learning_rate": 9.872141021274554e-06,
+      "loss": 0.592,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 8.379879107627136,
+      "learning_rate": 9.871855092083218e-06,
+      "loss": 0.568,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7727368237544894,
+      "learning_rate": 9.871568847689228e-06,
+      "loss": 0.4708,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0473632667752346,
+      "learning_rate": 9.871282288111109e-06,
+      "loss": 0.5511,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6548522612466148,
+      "learning_rate": 9.870995413367397e-06,
+      "loss": 0.5473,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.694968127640105,
+      "learning_rate": 9.870708223476653e-06,
+      "loss": 0.486,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.5146314248321153,
+      "learning_rate": 9.870420718457458e-06,
+      "loss": 0.5625,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6934324925723094,
+      "learning_rate": 9.870132898328417e-06,
+      "loss": 0.528,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6060753301742932,
+      "learning_rate": 9.869844763108146e-06,
+      "loss": 0.5657,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9161951237878325,
+      "learning_rate": 9.869556312815289e-06,
+      "loss": 0.515,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5818907560780635,
+      "learning_rate": 9.86926754746851e-06,
+      "loss": 0.5839,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.314610828847146,
+      "learning_rate": 9.868978467086491e-06,
+      "loss": 0.5163,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.0328770026992995,
+      "learning_rate": 9.868689071687933e-06,
+      "loss": 0.4814,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6747212853550181,
+      "learning_rate": 9.868399361291562e-06,
+      "loss": 0.5957,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.1030080651097802,
+      "learning_rate": 9.86810933591612e-06,
+      "loss": 0.569,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6495845441356625,
+      "learning_rate": 9.867818995580374e-06,
+      "loss": 0.5093,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.584903648441113,
+      "learning_rate": 9.867528340303106e-06,
+      "loss": 0.4968,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4761169676700796,
+      "learning_rate": 9.867237370103122e-06,
+      "loss": 0.5339,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5423299442713723,
+      "learning_rate": 9.866946084999248e-06,
+      "loss": 0.5407,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.652278681218033,
+      "learning_rate": 9.866654485010328e-06,
+      "loss": 0.5696,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5666281947299256,
+      "learning_rate": 9.86636257015523e-06,
+      "loss": 0.5754,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.700553659731398,
+      "learning_rate": 9.866070340452838e-06,
+      "loss": 0.5546,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7590005795153083,
+      "learning_rate": 9.865777795922063e-06,
+      "loss": 0.526,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7584781049415148,
+      "learning_rate": 9.865484936581828e-06,
+      "loss": 0.5194,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6470975460098836,
+      "learning_rate": 9.865191762451084e-06,
+      "loss": 0.5433,
+      "step": 1297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6680342145032583,
+      "learning_rate": 9.864898273548795e-06,
+      "loss": 0.5717,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.768194547229024,
+      "learning_rate": 9.864604469893955e-06,
+      "loss": 0.4997,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2541545361701116,
+      "learning_rate": 9.864310351505567e-06,
+      "loss": 0.5424,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9917846693599401,
+      "learning_rate": 9.864015918402663e-06,
+      "loss": 0.5349,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.074978244726442,
+      "learning_rate": 9.863721170604292e-06,
+      "loss": 0.5732,
+      "step": 1302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9285982627310976,
+      "learning_rate": 9.863426108129526e-06,
+      "loss": 0.4947,
+      "step": 1303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6210590885819158,
+      "learning_rate": 9.86313073099745e-06,
+      "loss": 0.5657,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.757925430682471,
+      "learning_rate": 9.862835039227179e-06,
+      "loss": 0.5,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.737712200477531,
+      "learning_rate": 9.862539032837842e-06,
+      "loss": 0.5606,
+      "step": 1306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6431584145872844,
+      "learning_rate": 9.862242711848591e-06,
+      "loss": 0.584,
+      "step": 1307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5422463729744336,
+      "learning_rate": 9.861946076278597e-06,
+      "loss": 0.5563,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.127514497102055,
+      "learning_rate": 9.861649126147051e-06,
+      "loss": 0.5697,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6040292275629819,
+      "learning_rate": 9.861351861473168e-06,
+      "loss": 0.5708,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 37.84452125351232,
+      "learning_rate": 9.861054282276176e-06,
+      "loss": 0.5082,
+      "step": 1311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.533607621046888,
+      "learning_rate": 9.860756388575335e-06,
+      "loss": 0.5099,
+      "step": 1312
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4770269210060845,
+      "learning_rate": 9.860458180389913e-06,
+      "loss": 0.524,
+      "step": 1313
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8705863343213631,
+      "learning_rate": 9.860159657739204e-06,
+      "loss": 0.5324,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7295674462802433,
+      "learning_rate": 9.859860820642524e-06,
+      "loss": 0.5313,
+      "step": 1315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6626274426325387,
+      "learning_rate": 9.859561669119206e-06,
+      "loss": 0.5068,
+      "step": 1316
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.180259681304325,
+      "learning_rate": 9.859262203188605e-06,
+      "loss": 0.5379,
+      "step": 1317
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.838881916381452,
+      "learning_rate": 9.858962422870095e-06,
+      "loss": 0.5616,
+      "step": 1318
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.0229298248697094,
+      "learning_rate": 9.858662328183075e-06,
+      "loss": 0.5023,
+      "step": 1319
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6338025867418218,
+      "learning_rate": 9.858361919146958e-06,
+      "loss": 0.571,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.837196256955751,
+      "learning_rate": 9.85806119578118e-06,
+      "loss": 0.5511,
+      "step": 1321
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3725510214590027,
+      "learning_rate": 9.857760158105196e-06,
+      "loss": 0.4682,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9023426772230936,
+      "learning_rate": 9.857458806138486e-06,
+      "loss": 0.4947,
+      "step": 1323
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.8491900263042647,
+      "learning_rate": 9.857157139900546e-06,
+      "loss": 0.489,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.565264169737493,
+      "learning_rate": 9.856855159410892e-06,
+      "loss": 0.5655,
+      "step": 1325
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 7.1432198868615,
+      "learning_rate": 9.856552864689061e-06,
+      "loss": 0.5175,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6060281783253083,
+      "learning_rate": 9.856250255754616e-06,
+      "loss": 0.5691,
+      "step": 1327
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.117209006777162,
+      "learning_rate": 9.855947332627131e-06,
+      "loss": 0.5733,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9329718723564508,
+      "learning_rate": 9.855644095326207e-06,
+      "loss": 0.527,
+      "step": 1329
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8594030365564438,
+      "learning_rate": 9.855340543871461e-06,
+      "loss": 0.5384,
+      "step": 1330
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7186728683031587,
+      "learning_rate": 9.855036678282534e-06,
+      "loss": 0.5536,
+      "step": 1331
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.599736993199715,
+      "learning_rate": 9.854732498579085e-06,
+      "loss": 0.4908,
+      "step": 1332
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.505437838624874,
+      "learning_rate": 9.854428004780795e-06,
+      "loss": 0.5173,
+      "step": 1333
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2822588775643706,
+      "learning_rate": 9.85412319690736e-06,
+      "loss": 0.503,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8235519936042275,
+      "learning_rate": 9.853818074978507e-06,
+      "loss": 0.5061,
+      "step": 1335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.9654980749160764,
+      "learning_rate": 9.853512639013974e-06,
+      "loss": 0.5024,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7867603827032181,
+      "learning_rate": 9.853206889033522e-06,
+      "loss": 0.5159,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.230542976966494,
+      "learning_rate": 9.852900825056932e-06,
+      "loss": 0.5201,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.388001971391122,
+      "learning_rate": 9.85259444710401e-06,
+      "loss": 0.5214,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6091559680675073,
+      "learning_rate": 9.852287755194572e-06,
+      "loss": 0.5839,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6666323337098522,
+      "learning_rate": 9.851980749348466e-06,
+      "loss": 0.5824,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6965746567540188,
+      "learning_rate": 9.851673429585551e-06,
+      "loss": 0.5384,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6053502340558741,
+      "learning_rate": 9.851365795925713e-06,
+      "loss": 0.5696,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.459681709532209,
+      "learning_rate": 9.851057848388854e-06,
+      "loss": 0.5484,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.83036064579155,
+      "learning_rate": 9.850749586994899e-06,
+      "loss": 0.5722,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5877544683178362,
+      "learning_rate": 9.85044101176379e-06,
+      "loss": 0.4946,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7115011474218138,
+      "learning_rate": 9.850132122715494e-06,
+      "loss": 0.5395,
+      "step": 1347
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 4.23868157720752,
+      "learning_rate": 9.849822919869993e-06,
+      "loss": 0.4991,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4160330432570516,
+      "learning_rate": 9.849513403247295e-06,
+      "loss": 0.5295,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2615378012818037,
+      "learning_rate": 9.849203572867422e-06,
+      "loss": 0.4873,
+      "step": 1350
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6556965043702019,
+      "learning_rate": 9.848893428750423e-06,
+      "loss": 0.5227,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.28201500431767,
+      "learning_rate": 9.848582970916362e-06,
+      "loss": 0.5869,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.982023100234855,
+      "learning_rate": 9.848272199385325e-06,
+      "loss": 0.5402,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6239353227185154,
+      "learning_rate": 9.84796111417742e-06,
+      "loss": 0.5482,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7995960136382251,
+      "learning_rate": 9.847649715312772e-06,
+      "loss": 0.5769,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6412381285309934,
+      "learning_rate": 9.847338002811528e-06,
+      "loss": 0.5374,
+      "step": 1356
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6558805936658236,
+      "learning_rate": 9.847025976693858e-06,
+      "loss": 0.584,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4622303712083156,
+      "learning_rate": 9.846713636979948e-06,
+      "loss": 0.5423,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5979879056073257,
+      "learning_rate": 9.846400983690005e-06,
+      "loss": 0.5787,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5215494068616249,
+      "learning_rate": 9.846088016844259e-06,
+      "loss": 0.5392,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8862063087748073,
+      "learning_rate": 9.845774736462957e-06,
+      "loss": 0.5773,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8672282340012472,
+      "learning_rate": 9.84546114256637e-06,
+      "loss": 0.5243,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9708495825706198,
+      "learning_rate": 9.845147235174783e-06,
+      "loss": 0.4748,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.095536283810096,
+      "learning_rate": 9.844833014308512e-06,
+      "loss": 0.544,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4483113251656274,
+      "learning_rate": 9.84451847998788e-06,
+      "loss": 0.5238,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6255499065890018,
+      "learning_rate": 9.84420363223324e-06,
+      "loss": 0.5313,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.832733133591127,
+      "learning_rate": 9.843888471064962e-06,
+      "loss": 0.5125,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6817054967578888,
+      "learning_rate": 9.843572996503439e-06,
+      "loss": 0.4931,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.669300786995267,
+      "learning_rate": 9.843257208569077e-06,
+      "loss": 0.5306,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5601871404054013,
+      "learning_rate": 9.842941107282309e-06,
+      "loss": 0.5733,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.878293945492033,
+      "learning_rate": 9.842624692663587e-06,
+      "loss": 0.5434,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.691183331546977,
+      "learning_rate": 9.842307964733384e-06,
+      "loss": 0.523,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7804431247497463,
+      "learning_rate": 9.841990923512188e-06,
+      "loss": 0.5588,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 5.5651239457282236,
+      "learning_rate": 9.841673569020515e-06,
+      "loss": 0.5561,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4980770357622069,
+      "learning_rate": 9.841355901278894e-06,
+      "loss": 0.5309,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7779967076940275,
+      "learning_rate": 9.841037920307883e-06,
+      "loss": 0.498,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4770660297817095,
+      "learning_rate": 9.840719626128048e-06,
+      "loss": 0.512,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.851319568812458,
+      "learning_rate": 9.840401018759987e-06,
+      "loss": 0.4876,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8227859046318933,
+      "learning_rate": 9.840082098224312e-06,
+      "loss": 0.4991,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.770611266843305,
+      "learning_rate": 9.839762864541657e-06,
+      "loss": 0.5275,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.822972233353695,
+      "learning_rate": 9.839443317732676e-06,
+      "loss": 0.5862,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8583608706680923,
+      "learning_rate": 9.839123457818042e-06,
+      "loss": 0.5376,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.005421427183514,
+      "learning_rate": 9.838803284818452e-06,
+      "loss": 0.5685,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6530182181391837,
+      "learning_rate": 9.838482798754619e-06,
+      "loss": 0.589,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6388839999237954,
+      "learning_rate": 9.838161999647278e-06,
+      "loss": 0.5163,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7269629148847347,
+      "learning_rate": 9.837840887517185e-06,
+      "loss": 0.5362,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6699267780384877,
+      "learning_rate": 9.837519462385115e-06,
+      "loss": 0.5688,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4802324891003922,
+      "learning_rate": 9.837197724271864e-06,
+      "loss": 0.502,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.906861470965291,
+      "learning_rate": 9.836875673198248e-06,
+      "loss": 0.5198,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.776398276648744,
+      "learning_rate": 9.836553309185105e-06,
+      "loss": 0.508,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6020112677575478,
+      "learning_rate": 9.836230632253288e-06,
+      "loss": 0.5536,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0532990833310403,
+      "learning_rate": 9.835907642423676e-06,
+      "loss": 0.5457,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6988009258101529,
+      "learning_rate": 9.835584339717165e-06,
+      "loss": 0.5247,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6355331440428171,
+      "learning_rate": 9.835260724154677e-06,
+      "loss": 0.5169,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2626647379952245,
+      "learning_rate": 9.83493679575714e-06,
+      "loss": 0.5228,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6725802558174745,
+      "learning_rate": 9.83461255454552e-06,
+      "loss": 0.5744,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.922426019633825,
+      "learning_rate": 9.834288000540792e-06,
+      "loss": 0.5862,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7868788910892452,
+      "learning_rate": 9.833963133763954e-06,
+      "loss": 0.5953,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.940312551674799,
+      "learning_rate": 9.833637954236025e-06,
+      "loss": 0.5533,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.0341150882818635,
+      "learning_rate": 9.833312461978043e-06,
+      "loss": 0.5292,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.523013759060074,
+      "learning_rate": 9.83298665701107e-06,
+      "loss": 0.5686,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4448239388296553,
+      "learning_rate": 9.83266053935618e-06,
+      "loss": 0.4888,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7905533028888934,
+      "learning_rate": 9.832334109034476e-06,
+      "loss": 0.4767,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5698400468407063,
+      "learning_rate": 9.832007366067078e-06,
+      "loss": 0.5302,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.707683195713622,
+      "learning_rate": 9.831680310475122e-06,
+      "loss": 0.5215,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.267256289365699,
+      "learning_rate": 9.831352942279772e-06,
+      "loss": 0.5186,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4546914619928246,
+      "learning_rate": 9.831025261502206e-06,
+      "loss": 0.5275,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.679855944284893,
+      "learning_rate": 9.830697268163625e-06,
+      "loss": 0.5655,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8752637294371775,
+      "learning_rate": 9.830368962285252e-06,
+      "loss": 0.544,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6730391483689642,
+      "learning_rate": 9.830040343888324e-06,
+      "loss": 0.5096,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6602892914724103,
+      "learning_rate": 9.829711412994105e-06,
+      "loss": 0.5402,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7564187138079868,
+      "learning_rate": 9.829382169623876e-06,
+      "loss": 0.4417,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6734626202539157,
+      "learning_rate": 9.829052613798938e-06,
+      "loss": 0.5882,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4511343352318309,
+      "learning_rate": 9.828722745540614e-06,
+      "loss": 0.5274,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6923609140431217,
+      "learning_rate": 9.828392564870244e-06,
+      "loss": 0.5572,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.44571988217081,
+      "learning_rate": 9.82806207180919e-06,
+      "loss": 0.5008,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.941799350571069,
+      "learning_rate": 9.827731266378839e-06,
+      "loss": 0.5574,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.080911393359229,
+      "learning_rate": 9.827400148600587e-06,
+      "loss": 0.5531,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6235865438465549,
+      "learning_rate": 9.827068718495864e-06,
+      "loss": 0.5209,
+      "step": 1419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6557147659369517,
+      "learning_rate": 9.826736976086108e-06,
+      "loss": 0.5548,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8011044262229758,
+      "learning_rate": 9.826404921392783e-06,
+      "loss": 0.5512,
+      "step": 1421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.1399277896554283,
+      "learning_rate": 9.826072554437373e-06,
+      "loss": 0.5159,
+      "step": 1422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.609804852868493,
+      "learning_rate": 9.825739875241385e-06,
+      "loss": 0.4594,
+      "step": 1423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6221689631727851,
+      "learning_rate": 9.825406883826338e-06,
+      "loss": 0.5925,
+      "step": 1424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 3.5946377852738878,
+      "learning_rate": 9.825073580213777e-06,
+      "loss": 0.5129,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.017252465409614,
+      "learning_rate": 9.824739964425269e-06,
+      "loss": 0.533,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9745787734988633,
+      "learning_rate": 9.824406036482395e-06,
+      "loss": 0.549,
+      "step": 1427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.2246304226320923,
+      "learning_rate": 9.824071796406764e-06,
+      "loss": 0.5745,
+      "step": 1428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6496788207316133,
+      "learning_rate": 9.823737244219997e-06,
+      "loss": 0.5107,
+      "step": 1429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4978005909024623,
+      "learning_rate": 9.82340237994374e-06,
+      "loss": 0.5626,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5886551158230016,
+      "learning_rate": 9.82306720359966e-06,
+      "loss": 0.4978,
+      "step": 1431
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4178806615443416,
+      "learning_rate": 9.822731715209442e-06,
+      "loss": 0.4672,
+      "step": 1432
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9708377515536957,
+      "learning_rate": 9.82239591479479e-06,
+      "loss": 0.5258,
+      "step": 1433
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6477282936181471,
+      "learning_rate": 9.82205980237743e-06,
+      "loss": 0.5831,
+      "step": 1434
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4094772562134705,
+      "learning_rate": 9.821723377979109e-06,
+      "loss": 0.4855,
+      "step": 1435
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.93180016626713,
+      "learning_rate": 9.821386641621593e-06,
+      "loss": 0.5367,
+      "step": 1436
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6316281356435898,
+      "learning_rate": 9.82104959332667e-06,
+      "loss": 0.5313,
+      "step": 1437
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5754686944405991,
+      "learning_rate": 9.820712233116142e-06,
+      "loss": 0.5328,
+      "step": 1438
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.856940447247618,
+      "learning_rate": 9.820374561011843e-06,
+      "loss": 0.4817,
+      "step": 1439
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5849232468849792,
+      "learning_rate": 9.820036577035613e-06,
+      "loss": 0.5035,
+      "step": 1440
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.256538708948276,
+      "learning_rate": 9.819698281209322e-06,
+      "loss": 0.5361,
+      "step": 1441
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5878994442457026,
+      "learning_rate": 9.819359673554858e-06,
+      "loss": 0.5591,
+      "step": 1442
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6536165730068886,
+      "learning_rate": 9.819020754094127e-06,
+      "loss": 0.5494,
+      "step": 1443
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.369498155797225,
+      "learning_rate": 9.818681522849058e-06,
+      "loss": 0.5282,
+      "step": 1444
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5715907393419204,
+      "learning_rate": 9.818341979841597e-06,
+      "loss": 0.5188,
+      "step": 1445
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.973900466108539,
+      "learning_rate": 9.818002125093714e-06,
+      "loss": 0.5452,
+      "step": 1446
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5295581024970804,
+      "learning_rate": 9.817661958627396e-06,
+      "loss": 0.4874,
+      "step": 1447
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6108620827341058,
+      "learning_rate": 9.817321480464653e-06,
+      "loss": 0.599,
+      "step": 1448
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.4224038972422504,
+      "learning_rate": 9.816980690627509e-06,
+      "loss": 0.5385,
+      "step": 1449
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6470157292810779,
+      "learning_rate": 9.816639589138017e-06,
+      "loss": 0.5633,
+      "step": 1450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4256876525283169,
+      "learning_rate": 9.816298176018245e-06,
+      "loss": 0.5498,
+      "step": 1451
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6546136677673913,
+      "learning_rate": 9.815956451290281e-06,
+      "loss": 0.5021,
+      "step": 1452
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6340100529066786,
+      "learning_rate": 9.815614414976235e-06,
+      "loss": 0.575,
+      "step": 1453
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 2.157416781368045,
+      "learning_rate": 9.815272067098236e-06,
+      "loss": 0.5307,
+      "step": 1454
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.9071490390100114,
+      "learning_rate": 9.814929407678434e-06,
+      "loss": 0.5218,
+      "step": 1455
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6957023678573544,
+      "learning_rate": 9.814586436738998e-06,
+      "loss": 0.5384,
+      "step": 1456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4580621009037396,
+      "learning_rate": 9.814243154302116e-06,
+      "loss": 0.5077,
+      "step": 1457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.597096166441479,
+      "learning_rate": 9.813899560390001e-06,
+      "loss": 0.5819,
+      "step": 1458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7378799313400277,
+      "learning_rate": 9.813555655024882e-06,
+      "loss": 0.5448,
+      "step": 1459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.7045672054254442,
+      "learning_rate": 9.813211438229009e-06,
+      "loss": 0.5219,
+      "step": 1460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4727926327197707,
+      "learning_rate": 9.812866910024652e-06,
+      "loss": 0.537,
+      "step": 1461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.466531885303182,
+      "learning_rate": 9.812522070434103e-06,
+      "loss": 0.4731,
+      "step": 1462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6654557048017115,
+      "learning_rate": 9.81217691947967e-06,
+      "loss": 0.5876,
+      "step": 1463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.397770804910777,
+      "learning_rate": 9.811831457183687e-06,
+      "loss": 0.5189,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.803115288219963,
+      "learning_rate": 9.811485683568502e-06,
+      "loss": 0.5279,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4601816883687113,
+      "learning_rate": 9.81113959865649e-06,
+      "loss": 0.5603,
+      "step": 1466
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8989026487522842,
+      "learning_rate": 9.810793202470036e-06,
+      "loss": 0.5531,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6289229039008521,
+      "learning_rate": 9.810446495031558e-06,
+      "loss": 0.5558,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.067336153759269,
+      "learning_rate": 9.810099476363484e-06,
+      "loss": 0.5586,
+      "step": 1469
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.056891487080746,
+      "learning_rate": 9.809752146488266e-06,
+      "loss": 0.4961,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6564264985537602,
+      "learning_rate": 9.809404505428375e-06,
+      "loss": 0.5802,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5638423977694067,
+      "learning_rate": 9.809056553206303e-06,
+      "loss": 0.5657,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 6.224399909497267,
+      "learning_rate": 9.808708289844566e-06,
+      "loss": 0.532,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6126993278595447,
+      "learning_rate": 9.808359715365693e-06,
+      "loss": 0.4909,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9035483311416836,
+      "learning_rate": 9.808010829792236e-06,
+      "loss": 0.5504,
+      "step": 1475
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8259158149478862,
+      "learning_rate": 9.807661633146768e-06,
+      "loss": 0.5872,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6213409351363248,
+      "learning_rate": 9.807312125451881e-06,
+      "loss": 0.5873,
+      "step": 1477
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.542885500047365,
+      "learning_rate": 9.806962306730187e-06,
+      "loss": 0.5357,
+      "step": 1478
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7538979240062165,
+      "learning_rate": 9.806612177004324e-06,
+      "loss": 0.5147,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.57243733898596,
+      "learning_rate": 9.806261736296938e-06,
+      "loss": 0.4887,
+      "step": 1480
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3839805352205314,
+      "learning_rate": 9.805910984630705e-06,
+      "loss": 0.5363,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6680489543720407,
+      "learning_rate": 9.805559922028319e-06,
+      "loss": 0.532,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.614143424048914,
+      "learning_rate": 9.805208548512493e-06,
+      "loss": 0.495,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.086016035569516,
+      "learning_rate": 9.804856864105958e-06,
+      "loss": 0.4799,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6966086411337373,
+      "learning_rate": 9.80450486883147e-06,
+      "loss": 0.5701,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3565860804155265,
+      "learning_rate": 9.804152562711804e-06,
+      "loss": 0.5562,
+      "step": 1486
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6776347346331966,
+      "learning_rate": 9.80379994576975e-06,
+      "loss": 0.5594,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9966643843278638,
+      "learning_rate": 9.803447018028124e-06,
+      "loss": 0.5533,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.163997501723208,
+      "learning_rate": 9.80309377950976e-06,
+      "loss": 0.52,
+      "step": 1489
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3915427807023326,
+      "learning_rate": 9.802740230237512e-06,
+      "loss": 0.5062,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6175670277563174,
+      "learning_rate": 9.802386370234254e-06,
+      "loss": 0.5082,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9827328281711285,
+      "learning_rate": 9.80203219952288e-06,
+      "loss": 0.5425,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5512848654967046,
+      "learning_rate": 9.801677718126303e-06,
+      "loss": 0.4952,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6499644823428672,
+      "learning_rate": 9.801322926067461e-06,
+      "loss": 0.5491,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4826993428053694,
+      "learning_rate": 9.800967823369304e-06,
+      "loss": 0.5007,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5673435981039556,
+      "learning_rate": 9.800612410054813e-06,
+      "loss": 0.4842,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4156769917693852,
+      "learning_rate": 9.800256686146977e-06,
+      "loss": 0.4813,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.447942557281987,
+      "learning_rate": 9.799900651668812e-06,
+      "loss": 0.5137,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7012942444692674,
+      "learning_rate": 9.799544306643356e-06,
+      "loss": 0.569,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.520176049387108,
+      "learning_rate": 9.79918765109366e-06,
+      "loss": 0.5114,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6843540325782766,
+      "learning_rate": 9.798830685042801e-06,
+      "loss": 0.5352,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0591510550059278,
+      "learning_rate": 9.798473408513876e-06,
+      "loss": 0.5064,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.420343296350935,
+      "learning_rate": 9.79811582153e-06,
+      "loss": 0.5234,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.48144342230223,
+      "learning_rate": 9.797757924114304e-06,
+      "loss": 0.5221,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.2374424438602327,
+      "learning_rate": 9.797399716289947e-06,
+      "loss": 0.5163,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.611288380411681,
+      "learning_rate": 9.797041198080105e-06,
+      "loss": 0.5551,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7291779697126157,
+      "learning_rate": 9.796682369507972e-06,
+      "loss": 0.5988,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6869981498164651,
+      "learning_rate": 9.796323230596766e-06,
+      "loss": 0.582,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5072725421642132,
+      "learning_rate": 9.79596378136972e-06,
+      "loss": 0.5291,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.69148184300472,
+      "learning_rate": 9.795604021850092e-06,
+      "loss": 0.4704,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7123627989850265,
+      "learning_rate": 9.795243952061158e-06,
+      "loss": 0.5889,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.890178310026196,
+      "learning_rate": 9.79488357202621e-06,
+      "loss": 0.5609,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3162541572302286,
+      "learning_rate": 9.794522881768572e-06,
+      "loss": 0.4732,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.702095102320251,
+      "learning_rate": 9.794161881311573e-06,
+      "loss": 0.526,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3811814105546294,
+      "learning_rate": 9.793800570678571e-06,
+      "loss": 0.4861,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.590785241525209,
+      "learning_rate": 9.793438949892945e-06,
+      "loss": 0.5666,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8844246494596113,
+      "learning_rate": 9.793077018978088e-06,
+      "loss": 0.5628,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5263312497308046,
+      "learning_rate": 9.792714777957419e-06,
+      "loss": 0.5582,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4830971862725775,
+      "learning_rate": 9.792352226854373e-06,
+      "loss": 0.5092,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2238405662846028,
+      "learning_rate": 9.791989365692408e-06,
+      "loss": 0.4651,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4932476673251032,
+      "learning_rate": 9.791626194494998e-06,
+      "loss": 0.5224,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.598361710120726,
+      "learning_rate": 9.791262713285644e-06,
+      "loss": 0.5366,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.690852240103195,
+      "learning_rate": 9.790898922087857e-06,
+      "loss": 0.5084,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4988239699623018,
+      "learning_rate": 9.79053482092518e-06,
+      "loss": 0.467,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.494002245915824,
+      "learning_rate": 9.790170409821165e-06,
+      "loss": 0.485,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2676960726270852,
+      "learning_rate": 9.789805688799393e-06,
+      "loss": 0.6222,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9210009322427357,
+      "learning_rate": 9.789440657883456e-06,
+      "loss": 0.553,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.894456412323396,
+      "learning_rate": 9.789075317096974e-06,
+      "loss": 0.5112,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.70628108519553,
+      "learning_rate": 9.788709666463585e-06,
+      "loss": 0.5754,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7528622705318033,
+      "learning_rate": 9.788343706006946e-06,
+      "loss": 0.5289,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4592595251949918,
+      "learning_rate": 9.787977435750732e-06,
+      "loss": 0.5532,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3513536932020114,
+      "learning_rate": 9.787610855718642e-06,
+      "loss": 0.5162,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5857925588106039,
+      "learning_rate": 9.787243965934392e-06,
+      "loss": 0.5595,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.484735905113311,
+      "learning_rate": 9.78687676642172e-06,
+      "loss": 0.6419,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.258203678294199,
+      "learning_rate": 9.786509257204383e-06,
+      "loss": 0.6182,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6780516009544602,
+      "learning_rate": 9.786141438306158e-06,
+      "loss": 0.5504,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.0999507819127587,
+      "learning_rate": 9.785773309750845e-06,
+      "loss": 0.5363,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.1956560557985756,
+      "learning_rate": 9.785404871562258e-06,
+      "loss": 0.5049,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7152970081106214,
+      "learning_rate": 9.785036123764235e-06,
+      "loss": 0.5392,
+      "step": 1539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.891569954455362,
+      "learning_rate": 9.784667066380637e-06,
+      "loss": 0.5582,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5335814055798747,
+      "learning_rate": 9.784297699435336e-06,
+      "loss": 0.5322,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6996267991293694,
+      "learning_rate": 9.783928022952236e-06,
+      "loss": 0.5292,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6319139637576137,
+      "learning_rate": 9.783558036955249e-06,
+      "loss": 0.5297,
+      "step": 1543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6416554708943427,
+      "learning_rate": 9.783187741468314e-06,
+      "loss": 0.5417,
+      "step": 1544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.862631740632092,
+      "learning_rate": 9.78281713651539e-06,
+      "loss": 0.5714,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5145195376960159,
+      "learning_rate": 9.782446222120454e-06,
+      "loss": 0.5793,
+      "step": 1546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7042888159825762,
+      "learning_rate": 9.782074998307505e-06,
+      "loss": 0.5346,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6858550078197532,
+      "learning_rate": 9.781703465100556e-06,
+      "loss": 0.5321,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3126190482958917,
+      "learning_rate": 9.781331622523651e-06,
+      "loss": 0.4458,
+      "step": 1549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3359489271363565,
+      "learning_rate": 9.780959470600844e-06,
+      "loss": 0.6358,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6392377825953799,
+      "learning_rate": 9.780587009356214e-06,
+      "loss": 0.5296,
+      "step": 1551
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.782425389193383,
+      "learning_rate": 9.780214238813858e-06,
+      "loss": 0.5062,
+      "step": 1552
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8528416557986036,
+      "learning_rate": 9.779841158997894e-06,
+      "loss": 0.5809,
+      "step": 1553
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6752901568743175,
+      "learning_rate": 9.77946776993246e-06,
+      "loss": 0.5431,
+      "step": 1554
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5932064850137686,
+      "learning_rate": 9.779094071641712e-06,
+      "loss": 0.546,
+      "step": 1555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.570049069081433,
+      "learning_rate": 9.778720064149832e-06,
+      "loss": 0.4393,
+      "step": 1556
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7192342294319541,
+      "learning_rate": 9.778345747481013e-06,
+      "loss": 0.5768,
+      "step": 1557
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7340970681645025,
+      "learning_rate": 9.777971121659477e-06,
+      "loss": 0.4999,
+      "step": 1558
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.604312995989312,
+      "learning_rate": 9.777596186709457e-06,
+      "loss": 0.4728,
+      "step": 1559
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.464654577981638,
+      "learning_rate": 9.777220942655217e-06,
+      "loss": 0.5394,
+      "step": 1560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.3416511751395794,
+      "learning_rate": 9.776845389521029e-06,
+      "loss": 0.534,
+      "step": 1561
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.72288666086215,
+      "learning_rate": 9.776469527331194e-06,
+      "loss": 0.5788,
+      "step": 1562
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3807501628675511,
+      "learning_rate": 9.776093356110027e-06,
+      "loss": 0.5289,
+      "step": 1563
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7069391500052442,
+      "learning_rate": 9.77571687588187e-06,
+      "loss": 0.5787,
+      "step": 1564
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6171416699039798,
+      "learning_rate": 9.77534008667108e-06,
+      "loss": 0.5065,
+      "step": 1565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2639808361605234,
+      "learning_rate": 9.774962988502032e-06,
+      "loss": 0.4682,
+      "step": 1566
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7201672652512507,
+      "learning_rate": 9.774585581399122e-06,
+      "loss": 0.53,
+      "step": 1567
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.451863257985592,
+      "learning_rate": 9.774207865386775e-06,
+      "loss": 0.5423,
+      "step": 1568
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7379154152765588,
+      "learning_rate": 9.773829840489423e-06,
+      "loss": 0.5157,
+      "step": 1569
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6664559563873693,
+      "learning_rate": 9.773451506731526e-06,
+      "loss": 0.5696,
+      "step": 1570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7447898060919843,
+      "learning_rate": 9.773072864137561e-06,
+      "loss": 0.5707,
+      "step": 1571
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5378793325476334,
+      "learning_rate": 9.772693912732022e-06,
+      "loss": 0.5003,
+      "step": 1572
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.0609274320201045,
+      "learning_rate": 9.772314652539434e-06,
+      "loss": 0.5596,
+      "step": 1573
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.800248890787103,
+      "learning_rate": 9.771935083584331e-06,
+      "loss": 0.5566,
+      "step": 1574
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7318886983275132,
+      "learning_rate": 9.77155520589127e-06,
+      "loss": 0.5524,
+      "step": 1575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3584178721111977,
+      "learning_rate": 9.771175019484828e-06,
+      "loss": 0.4872,
+      "step": 1576
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4705481906399265,
+      "learning_rate": 9.770794524389605e-06,
+      "loss": 0.4981,
+      "step": 1577
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.7649880341716853,
+      "learning_rate": 9.770413720630218e-06,
+      "loss": 0.5223,
+      "step": 1578
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.053151650178062,
+      "learning_rate": 9.770032608231302e-06,
+      "loss": 0.4928,
+      "step": 1579
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1040535860312617,
+      "learning_rate": 9.769651187217517e-06,
+      "loss": 0.514,
+      "step": 1580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5991185596135733,
+      "learning_rate": 9.76926945761354e-06,
+      "loss": 0.5427,
+      "step": 1581
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.8899517653224276,
+      "learning_rate": 9.768887419444066e-06,
+      "loss": 0.5116,
+      "step": 1582
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5359368569470062,
+      "learning_rate": 9.768505072733816e-06,
+      "loss": 0.5249,
+      "step": 1583
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5176740708586691,
+      "learning_rate": 9.768122417507526e-06,
+      "loss": 0.5637,
+      "step": 1584
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6580980378745395,
+      "learning_rate": 9.767739453789954e-06,
+      "loss": 0.5516,
+      "step": 1585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 2.4070421840452294,
+      "learning_rate": 9.767356181605877e-06,
+      "loss": 0.5389,
+      "step": 1586
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5297105888816123,
+      "learning_rate": 9.766972600980088e-06,
+      "loss": 0.52,
+      "step": 1587
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6301945130075601,
+      "learning_rate": 9.76658871193741e-06,
+      "loss": 0.5546,
+      "step": 1588
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 4.068753370664915,
+      "learning_rate": 9.766204514502677e-06,
+      "loss": 0.5058,
+      "step": 1589
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.55996614446939,
+      "learning_rate": 9.765820008700746e-06,
+      "loss": 0.5777,
+      "step": 1590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5117961121032426,
+      "learning_rate": 9.765435194556497e-06,
+      "loss": 0.5309,
+      "step": 1591
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5515926901217598,
+      "learning_rate": 9.765050072094824e-06,
+      "loss": 0.4924,
+      "step": 1592
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6230129898930237,
+      "learning_rate": 9.764664641340645e-06,
+      "loss": 0.5591,
+      "step": 1593
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3891450348864518,
+      "learning_rate": 9.764278902318897e-06,
+      "loss": 0.5327,
+      "step": 1594
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6914096374222076,
+      "learning_rate": 9.763892855054535e-06,
+      "loss": 0.5191,
+      "step": 1595
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5977521762519895,
+      "learning_rate": 9.76350649957254e-06,
+      "loss": 0.4915,
+      "step": 1596
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.551567802635987,
+      "learning_rate": 9.763119835897903e-06,
+      "loss": 0.4902,
+      "step": 1597
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4238094491514561,
+      "learning_rate": 9.762732864055644e-06,
+      "loss": 0.5109,
+      "step": 1598
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.583506751194209,
+      "learning_rate": 9.7623455840708e-06,
+      "loss": 0.5598,
+      "step": 1599
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7185491932134798,
+      "learning_rate": 9.761957995968427e-06,
+      "loss": 0.5086,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.531833488777546,
+      "learning_rate": 9.761570099773601e-06,
+      "loss": 0.5028,
+      "step": 1601
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5789329340551965,
+      "learning_rate": 9.761181895511418e-06,
+      "loss": 0.5805,
+      "step": 1602
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5464500569078543,
+      "learning_rate": 9.760793383206995e-06,
+      "loss": 0.5178,
+      "step": 1603
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.865437475135024,
+      "learning_rate": 9.760404562885467e-06,
+      "loss": 0.5655,
+      "step": 1604
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.479885642920405,
+      "learning_rate": 9.760015434571992e-06,
+      "loss": 0.4512,
+      "step": 1605
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.589882726282577,
+      "learning_rate": 9.759625998291745e-06,
+      "loss": 0.5767,
+      "step": 1606
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.649224239675404,
+      "learning_rate": 9.759236254069923e-06,
+      "loss": 0.5771,
+      "step": 1607
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3534319033551103,
+      "learning_rate": 9.75884620193174e-06,
+      "loss": 0.5043,
+      "step": 1608
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6492649823796712,
+      "learning_rate": 9.758455841902435e-06,
+      "loss": 0.5968,
+      "step": 1609
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.050652369526952,
+      "learning_rate": 9.75806517400726e-06,
+      "loss": 0.4906,
+      "step": 1610
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6898348586662564,
+      "learning_rate": 9.757674198271494e-06,
+      "loss": 0.5393,
+      "step": 1611
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 10.950546050284869,
+      "learning_rate": 9.757282914720429e-06,
+      "loss": 0.5186,
+      "step": 1612
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.658259001884661,
+      "learning_rate": 9.756891323379385e-06,
+      "loss": 0.5675,
+      "step": 1613
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4683383249373028,
+      "learning_rate": 9.756499424273695e-06,
+      "loss": 0.5672,
+      "step": 1614
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7569992686688976,
+      "learning_rate": 9.756107217428713e-06,
+      "loss": 0.5053,
+      "step": 1615
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5152816350257374,
+      "learning_rate": 9.755714702869817e-06,
+      "loss": 0.5586,
+      "step": 1616
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7800974300880035,
+      "learning_rate": 9.7553218806224e-06,
+      "loss": 0.5472,
+      "step": 1617
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.3777375272487453,
+      "learning_rate": 9.75492875071188e-06,
+      "loss": 0.5608,
+      "step": 1618
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.871321126321876,
+      "learning_rate": 9.754535313163688e-06,
+      "loss": 0.538,
+      "step": 1619
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7459032930878007,
+      "learning_rate": 9.754141568003282e-06,
+      "loss": 0.5106,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7127039115769913,
+      "learning_rate": 9.753747515256135e-06,
+      "loss": 0.5688,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6860423589996247,
+      "learning_rate": 9.753353154947744e-06,
+      "loss": 0.5598,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.340843642522983,
+      "learning_rate": 9.752958487103621e-06,
+      "loss": 0.5754,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3617610359453556,
+      "learning_rate": 9.752563511749301e-06,
+      "loss": 0.4916,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4916774479155517,
+      "learning_rate": 9.75216822891034e-06,
+      "loss": 0.5035,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7754750251274392,
+      "learning_rate": 9.75177263861231e-06,
+      "loss": 0.5415,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4215838028483774,
+      "learning_rate": 9.751376740880807e-06,
+      "loss": 0.5383,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.647273687588268,
+      "learning_rate": 9.750980535741445e-06,
+      "loss": 0.5558,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5184696858048425,
+      "learning_rate": 9.750584023219857e-06,
+      "loss": 0.4795,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4657025657978797,
+      "learning_rate": 9.750187203341697e-06,
+      "loss": 0.4889,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4033560792850248,
+      "learning_rate": 9.74979007613264e-06,
+      "loss": 0.4899,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5971544079302799,
+      "learning_rate": 9.749392641618376e-06,
+      "loss": 0.5529,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.7817058638774235,
+      "learning_rate": 9.748994899824623e-06,
+      "loss": 0.4927,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.70157270407114,
+      "learning_rate": 9.748596850777112e-06,
+      "loss": 0.4958,
+      "step": 1634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.623236364196928,
+      "learning_rate": 9.748198494501598e-06,
+      "loss": 0.5723,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4282010727427261,
+      "learning_rate": 9.74779983102385e-06,
+      "loss": 0.4841,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6383678381789328,
+      "learning_rate": 9.747400860369667e-06,
+      "loss": 0.4932,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 5.733737644793054,
+      "learning_rate": 9.747001582564858e-06,
+      "loss": 0.5181,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6618091467452387,
+      "learning_rate": 9.746601997635255e-06,
+      "loss": 0.5042,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.215757011602407,
+      "learning_rate": 9.746202105606713e-06,
+      "loss": 0.4993,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.683547464034928,
+      "learning_rate": 9.745801906505104e-06,
+      "loss": 0.5589,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.5425950127757986,
+      "learning_rate": 9.745401400356319e-06,
+      "loss": 0.5371,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7426373165516795,
+      "learning_rate": 9.74500058718627e-06,
+      "loss": 0.5945,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8710461669524259,
+      "learning_rate": 9.744599467020893e-06,
+      "loss": 0.5184,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9968527080851093,
+      "learning_rate": 9.744198039886136e-06,
+      "loss": 0.534,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6916076943544632,
+      "learning_rate": 9.743796305807971e-06,
+      "loss": 0.5661,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.993616041570552,
+      "learning_rate": 9.743394264812392e-06,
+      "loss": 0.5291,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6813934367960476,
+      "learning_rate": 9.742991916925409e-06,
+      "loss": 0.5734,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1973323136025145,
+      "learning_rate": 9.742589262173054e-06,
+      "loss": 0.5313,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.428243363523825,
+      "learning_rate": 9.742186300581378e-06,
+      "loss": 0.4944,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8069885197915512,
+      "learning_rate": 9.741783032176451e-06,
+      "loss": 0.5313,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.628547012963086,
+      "learning_rate": 9.741379456984364e-06,
+      "loss": 0.5044,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.559235470937039,
+      "learning_rate": 9.740975575031229e-06,
+      "loss": 0.5404,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2910206029676095,
+      "learning_rate": 9.740571386343178e-06,
+      "loss": 0.5437,
+      "step": 1654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6307642700341504,
+      "learning_rate": 9.740166890946358e-06,
+      "loss": 0.5584,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8314192314848383,
+      "learning_rate": 9.739762088866942e-06,
+      "loss": 0.5256,
+      "step": 1656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7527729257928064,
+      "learning_rate": 9.739356980131119e-06,
+      "loss": 0.6005,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6251643308377584,
+      "learning_rate": 9.7389515647651e-06,
+      "loss": 0.5172,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6195906157000807,
+      "learning_rate": 9.738545842795112e-06,
+      "loss": 0.5753,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3892737766682495,
+      "learning_rate": 9.73813981424741e-06,
+      "loss": 0.5002,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5042822326110639,
+      "learning_rate": 9.737733479148256e-06,
+      "loss": 0.5298,
+      "step": 1661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5228980471688347,
+      "learning_rate": 9.737326837523949e-06,
+      "loss": 0.5113,
+      "step": 1662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.443048134714524,
+      "learning_rate": 9.736919889400788e-06,
+      "loss": 0.5425,
+      "step": 1663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8434939849135352,
+      "learning_rate": 9.736512634805109e-06,
+      "loss": 0.4786,
+      "step": 1664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9788812456309413,
+      "learning_rate": 9.736105073763258e-06,
+      "loss": 0.52,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5371055258407116,
+      "learning_rate": 9.735697206301604e-06,
+      "loss": 0.514,
+      "step": 1666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6480894276309557,
+      "learning_rate": 9.735289032446537e-06,
+      "loss": 0.5054,
+      "step": 1667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.043350919660803,
+      "learning_rate": 9.734880552224464e-06,
+      "loss": 0.5175,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4658362287119135,
+      "learning_rate": 9.734471765661813e-06,
+      "loss": 0.4716,
+      "step": 1669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.444677822537247,
+      "learning_rate": 9.734062672785032e-06,
+      "loss": 0.5784,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6467824819419734,
+      "learning_rate": 9.73365327362059e-06,
+      "loss": 0.5454,
+      "step": 1671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4507805809126435,
+      "learning_rate": 9.733243568194972e-06,
+      "loss": 0.4862,
+      "step": 1672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6197653597491384,
+      "learning_rate": 9.732833556534688e-06,
+      "loss": 0.5086,
+      "step": 1673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6182002536547113,
+      "learning_rate": 9.732423238666264e-06,
+      "loss": 0.4992,
+      "step": 1674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6541744262119211,
+      "learning_rate": 9.732012614616247e-06,
+      "loss": 0.5102,
+      "step": 1675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.17576379338614,
+      "learning_rate": 9.731601684411205e-06,
+      "loss": 0.5393,
+      "step": 1676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8763999630970931,
+      "learning_rate": 9.731190448077722e-06,
+      "loss": 0.5179,
+      "step": 1677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1088263287279316,
+      "learning_rate": 9.730778905642408e-06,
+      "loss": 0.5521,
+      "step": 1678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6366864440725426,
+      "learning_rate": 9.730367057131887e-06,
+      "loss": 0.547,
+      "step": 1679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8517200118684531,
+      "learning_rate": 9.729954902572805e-06,
+      "loss": 0.5616,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4409162741326125,
+      "learning_rate": 9.729542441991828e-06,
+      "loss": 0.528,
+      "step": 1681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7691234026832334,
+      "learning_rate": 9.729129675415643e-06,
+      "loss": 0.5191,
+      "step": 1682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4350535279287002,
+      "learning_rate": 9.728716602870953e-06,
+      "loss": 0.477,
+      "step": 1683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6638235108855273,
+      "learning_rate": 9.728303224384486e-06,
+      "loss": 0.5295,
+      "step": 1684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8198356234309545,
+      "learning_rate": 9.727889539982986e-06,
+      "loss": 0.5254,
+      "step": 1685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7229016790969366,
+      "learning_rate": 9.727475549693217e-06,
+      "loss": 0.5742,
+      "step": 1686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6968211014118626,
+      "learning_rate": 9.727061253541964e-06,
+      "loss": 0.5818,
+      "step": 1687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.504046662667056,
+      "learning_rate": 9.726646651556032e-06,
+      "loss": 0.4868,
+      "step": 1688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4992271466730613,
+      "learning_rate": 9.726231743762245e-06,
+      "loss": 0.5544,
+      "step": 1689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5621181356574096,
+      "learning_rate": 9.725816530187446e-06,
+      "loss": 0.4675,
+      "step": 1690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.144746641944118,
+      "learning_rate": 9.725401010858501e-06,
+      "loss": 0.5404,
+      "step": 1691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2655213078327483,
+      "learning_rate": 9.724985185802291e-06,
+      "loss": 0.51,
+      "step": 1692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7092326385085878,
+      "learning_rate": 9.724569055045722e-06,
+      "loss": 0.576,
+      "step": 1693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5784509140903653,
+      "learning_rate": 9.724152618615715e-06,
+      "loss": 0.4963,
+      "step": 1694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3711484540038723,
+      "learning_rate": 9.723735876539213e-06,
+      "loss": 0.4967,
+      "step": 1695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4070060480961435,
+      "learning_rate": 9.72331882884318e-06,
+      "loss": 0.5051,
+      "step": 1696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.486967023482276,
+      "learning_rate": 9.722901475554597e-06,
+      "loss": 0.5691,
+      "step": 1697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.1970759035084924,
+      "learning_rate": 9.722483816700468e-06,
+      "loss": 0.5674,
+      "step": 1698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.495059754616997,
+      "learning_rate": 9.722065852307812e-06,
+      "loss": 0.521,
+      "step": 1699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7021353370053836,
+      "learning_rate": 9.721647582403673e-06,
+      "loss": 0.479,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8126720414937005,
+      "learning_rate": 9.721229007015114e-06,
+      "loss": 0.4991,
+      "step": 1701
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.289951646733298,
+      "learning_rate": 9.720810126169212e-06,
+      "loss": 0.4633,
+      "step": 1702
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.018439649116375,
+      "learning_rate": 9.720390939893071e-06,
+      "loss": 0.5013,
+      "step": 1703
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7761033324332052,
+      "learning_rate": 9.719971448213812e-06,
+      "loss": 0.5696,
+      "step": 1704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6628740905106985,
+      "learning_rate": 9.719551651158576e-06,
+      "loss": 0.5112,
+      "step": 1705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3501049593861383,
+      "learning_rate": 9.71913154875452e-06,
+      "loss": 0.4853,
+      "step": 1706
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6259209015622624,
+      "learning_rate": 9.718711141028827e-06,
+      "loss": 0.4906,
+      "step": 1707
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7160406777842405,
+      "learning_rate": 9.718290428008695e-06,
+      "loss": 0.571,
+      "step": 1708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6560760909528205,
+      "learning_rate": 9.717869409721345e-06,
+      "loss": 0.5737,
+      "step": 1709
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.6455879426816984,
+      "learning_rate": 9.717448086194018e-06,
+      "loss": 0.5468,
+      "step": 1710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.485525117055862,
+      "learning_rate": 9.717026457453968e-06,
+      "loss": 0.4967,
+      "step": 1711
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.9130633761540319,
+      "learning_rate": 9.716604523528478e-06,
+      "loss": 0.5555,
+      "step": 1712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.607519234331676,
+      "learning_rate": 9.716182284444846e-06,
+      "loss": 0.5433,
+      "step": 1713
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.509645606515361,
+      "learning_rate": 9.71575974023039e-06,
+      "loss": 0.4744,
+      "step": 1714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8539093451507088,
+      "learning_rate": 9.715336890912447e-06,
+      "loss": 0.5959,
+      "step": 1715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6230768787145395,
+      "learning_rate": 9.714913736518376e-06,
+      "loss": 0.5584,
+      "step": 1716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.814448100723574,
+      "learning_rate": 9.714490277075554e-06,
+      "loss": 0.5398,
+      "step": 1717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.2400617170260593,
+      "learning_rate": 9.714066512611378e-06,
+      "loss": 0.5521,
+      "step": 1718
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7399657565531563,
+      "learning_rate": 9.713642443153267e-06,
+      "loss": 0.5473,
+      "step": 1719
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.295790179389231,
+      "learning_rate": 9.713218068728655e-06,
+      "loss": 0.544,
+      "step": 1720
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6719966236345472,
+      "learning_rate": 9.712793389364998e-06,
+      "loss": 0.5621,
+      "step": 1721
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7187655563687367,
+      "learning_rate": 9.712368405089776e-06,
+      "loss": 0.5141,
+      "step": 1722
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.591317768009486,
+      "learning_rate": 9.711943115930483e-06,
+      "loss": 0.5609,
+      "step": 1723
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5034687695364906,
+      "learning_rate": 9.711517521914633e-06,
+      "loss": 0.5263,
+      "step": 1724
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7542595062014372,
+      "learning_rate": 9.711091623069763e-06,
+      "loss": 0.5567,
+      "step": 1725
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3868572193886552,
+      "learning_rate": 9.710665419423428e-06,
+      "loss": 0.5032,
+      "step": 1726
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.517994675967119,
+      "learning_rate": 9.710238911003203e-06,
+      "loss": 0.5306,
+      "step": 1727
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6439478699038876,
+      "learning_rate": 9.709812097836682e-06,
+      "loss": 0.5595,
+      "step": 1728
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5608526248234098,
+      "learning_rate": 9.70938497995148e-06,
+      "loss": 0.603,
+      "step": 1729
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5234860858168624,
+      "learning_rate": 9.708957557375229e-06,
+      "loss": 0.5461,
+      "step": 1730
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5251944177207828,
+      "learning_rate": 9.708529830135585e-06,
+      "loss": 0.5419,
+      "step": 1731
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7510216372451706,
+      "learning_rate": 9.708101798260221e-06,
+      "loss": 0.4901,
+      "step": 1732
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4545860711076086,
+      "learning_rate": 9.70767346177683e-06,
+      "loss": 0.5765,
+      "step": 1733
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.103546008279845,
+      "learning_rate": 9.707244820713122e-06,
+      "loss": 0.467,
+      "step": 1734
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.097954618516181,
+      "learning_rate": 9.706815875096834e-06,
+      "loss": 0.4974,
+      "step": 1735
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6078551717133867,
+      "learning_rate": 9.706386624955715e-06,
+      "loss": 0.5655,
+      "step": 1736
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6116546954667679,
+      "learning_rate": 9.70595707031754e-06,
+      "loss": 0.5532,
+      "step": 1737
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7544291143770019,
+      "learning_rate": 9.705527211210097e-06,
+      "loss": 0.5078,
+      "step": 1738
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6343169675210163,
+      "learning_rate": 9.7050970476612e-06,
+      "loss": 0.53,
+      "step": 1739
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7707060006987545,
+      "learning_rate": 9.704666579698679e-06,
+      "loss": 0.5993,
+      "step": 1740
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.221264562357025,
+      "learning_rate": 9.704235807350384e-06,
+      "loss": 0.5403,
+      "step": 1741
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7023598032292442,
+      "learning_rate": 9.703804730644186e-06,
+      "loss": 0.5215,
+      "step": 1742
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5284298066779207,
+      "learning_rate": 9.703373349607976e-06,
+      "loss": 0.526,
+      "step": 1743
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7815986493787258,
+      "learning_rate": 9.702941664269663e-06,
+      "loss": 0.5673,
+      "step": 1744
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6962268846826394,
+      "learning_rate": 9.702509674657177e-06,
+      "loss": 0.5439,
+      "step": 1745
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.649285391945932,
+      "learning_rate": 9.702077380798466e-06,
+      "loss": 0.5313,
+      "step": 1746
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6952489953791672,
+      "learning_rate": 9.7016447827215e-06,
+      "loss": 0.5146,
+      "step": 1747
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5842574928253417,
+      "learning_rate": 9.701211880454267e-06,
+      "loss": 0.5463,
+      "step": 1748
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6485267459369294,
+      "learning_rate": 9.700778674024776e-06,
+      "loss": 0.5816,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4417921738354134,
+      "learning_rate": 9.700345163461054e-06,
+      "loss": 0.5162,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6974699900256511,
+      "learning_rate": 9.699911348791146e-06,
+      "loss": 0.5892,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.568076211135904,
+      "learning_rate": 9.699477230043125e-06,
+      "loss": 0.4811,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9146312233358298,
+      "learning_rate": 9.699042807245073e-06,
+      "loss": 0.5082,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5387474057364796,
+      "learning_rate": 9.6986080804251e-06,
+      "loss": 0.549,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.506488077391304,
+      "learning_rate": 9.69817304961133e-06,
+      "loss": 0.5125,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.474289422948104,
+      "learning_rate": 9.69773771483191e-06,
+      "loss": 0.5744,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6083775393796964,
+      "learning_rate": 9.697302076115006e-06,
+      "loss": 0.5818,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.445486177874537,
+      "learning_rate": 9.696866133488802e-06,
+      "loss": 0.5365,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8135264462638443,
+      "learning_rate": 9.696429886981501e-06,
+      "loss": 0.532,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.506348090128629,
+      "learning_rate": 9.695993336621335e-06,
+      "loss": 0.5205,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5890776274123277,
+      "learning_rate": 9.695556482436538e-06,
+      "loss": 0.5526,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3364539061876732,
+      "learning_rate": 9.695119324455383e-06,
+      "loss": 0.4913,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6228452283152321,
+      "learning_rate": 9.694681862706149e-06,
+      "loss": 0.5115,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4209032406888533,
+      "learning_rate": 9.694244097217137e-06,
+      "loss": 0.4981,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.218839443350974,
+      "learning_rate": 9.693806028016675e-06,
+      "loss": 0.5174,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4384728816344388,
+      "learning_rate": 9.693367655133104e-06,
+      "loss": 0.5554,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6574116358365748,
+      "learning_rate": 9.692928978594784e-06,
+      "loss": 0.5946,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6928596953735693,
+      "learning_rate": 9.692489998430099e-06,
+      "loss": 0.52,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5451116674639809,
+      "learning_rate": 9.692050714667449e-06,
+      "loss": 0.5271,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.413142128196924,
+      "learning_rate": 9.691611127335256e-06,
+      "loss": 0.5062,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2580287383446627,
+      "learning_rate": 9.691171236461962e-06,
+      "loss": 0.4976,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7451358815987252,
+      "learning_rate": 9.690731042076023e-06,
+      "loss": 0.4973,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.402984116910305,
+      "learning_rate": 9.690290544205925e-06,
+      "loss": 0.5332,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6379001757246756,
+      "learning_rate": 9.689849742880162e-06,
+      "loss": 0.5174,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8315793849617958,
+      "learning_rate": 9.689408638127256e-06,
+      "loss": 0.5736,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6815743445542826,
+      "learning_rate": 9.688967229975747e-06,
+      "loss": 0.5126,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3325690608260463,
+      "learning_rate": 9.68852551845419e-06,
+      "loss": 0.4944,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6743833026042148,
+      "learning_rate": 9.688083503591167e-06,
+      "loss": 0.5852,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3520939543412167,
+      "learning_rate": 9.687641185415272e-06,
+      "loss": 0.553,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.904512495150474,
+      "learning_rate": 9.687198563955128e-06,
+      "loss": 0.5639,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.419072885115324,
+      "learning_rate": 9.686755639239367e-06,
+      "loss": 0.488,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.507648038097745,
+      "learning_rate": 9.686312411296646e-06,
+      "loss": 0.5399,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.808924917015108,
+      "learning_rate": 9.685868880155644e-06,
+      "loss": 0.4984,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6341159395632192,
+      "learning_rate": 9.685425045845054e-06,
+      "loss": 0.523,
+      "step": 1784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6377263366117027,
+      "learning_rate": 9.684980908393593e-06,
+      "loss": 0.5174,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5521653709967353,
+      "learning_rate": 9.684536467829997e-06,
+      "loss": 0.5508,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.353641114514894,
+      "learning_rate": 9.68409172418302e-06,
+      "loss": 0.5158,
+      "step": 1787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5174633587328563,
+      "learning_rate": 9.683646677481435e-06,
+      "loss": 0.4577,
+      "step": 1788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.783497306524081,
+      "learning_rate": 9.683201327754037e-06,
+      "loss": 0.5673,
+      "step": 1789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6455861903110509,
+      "learning_rate": 9.68275567502964e-06,
+      "loss": 0.5819,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.602629966736427,
+      "learning_rate": 9.682309719337076e-06,
+      "loss": 0.5809,
+      "step": 1791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.107392470822457,
+      "learning_rate": 9.6818634607052e-06,
+      "loss": 0.5502,
+      "step": 1792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5439475114638852,
+      "learning_rate": 9.681416899162879e-06,
+      "loss": 0.485,
+      "step": 1793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5476511995777875,
+      "learning_rate": 9.680970034739012e-06,
+      "loss": 0.4377,
+      "step": 1794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4375572363615428,
+      "learning_rate": 9.680522867462506e-06,
+      "loss": 0.5175,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.722444077610072,
+      "learning_rate": 9.680075397362294e-06,
+      "loss": 0.5436,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4942882073227948,
+      "learning_rate": 9.679627624467324e-06,
+      "loss": 0.4859,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.873996190761962,
+      "learning_rate": 9.679179548806571e-06,
+      "loss": 0.4955,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3917977480194774,
+      "learning_rate": 9.678731170409022e-06,
+      "loss": 0.5276,
+      "step": 1799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4523396733310785,
+      "learning_rate": 9.678282489303687e-06,
+      "loss": 0.4823,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9098835177132896,
+      "learning_rate": 9.677833505519594e-06,
+      "loss": 0.4891,
+      "step": 1801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7755070604298627,
+      "learning_rate": 9.677384219085793e-06,
+      "loss": 0.574,
+      "step": 1802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8463259455598227,
+      "learning_rate": 9.676934630031351e-06,
+      "loss": 0.4954,
+      "step": 1803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4463426929136307,
+      "learning_rate": 9.676484738385357e-06,
+      "loss": 0.5914,
+      "step": 1804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.403698398562429,
+      "learning_rate": 9.676034544176918e-06,
+      "loss": 0.5039,
+      "step": 1805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8265208015992391,
+      "learning_rate": 9.675584047435162e-06,
+      "loss": 0.5021,
+      "step": 1806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6203410613003286,
+      "learning_rate": 9.675133248189234e-06,
+      "loss": 0.5592,
+      "step": 1807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.781399633281742,
+      "learning_rate": 9.6746821464683e-06,
+      "loss": 0.5121,
+      "step": 1808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7643649387698164,
+      "learning_rate": 9.674230742301547e-06,
+      "loss": 0.5087,
+      "step": 1809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6596655233887727,
+      "learning_rate": 9.67377903571818e-06,
+      "loss": 0.5707,
+      "step": 1810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 4.870373659473655,
+      "learning_rate": 9.673327026747423e-06,
+      "loss": 0.5315,
+      "step": 1811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5966611860805129,
+      "learning_rate": 9.67287471541852e-06,
+      "loss": 0.5124,
+      "step": 1812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.196601475411892,
+      "learning_rate": 9.672422101760737e-06,
+      "loss": 0.5309,
+      "step": 1813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.564087241551201,
+      "learning_rate": 9.671969185803357e-06,
+      "loss": 0.5736,
+      "step": 1814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4698674510702414,
+      "learning_rate": 9.671515967575681e-06,
+      "loss": 0.5621,
+      "step": 1815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.617615539542415,
+      "learning_rate": 9.671062447107033e-06,
+      "loss": 0.5295,
+      "step": 1816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6045583464834527,
+      "learning_rate": 9.670608624426757e-06,
+      "loss": 0.5006,
+      "step": 1817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8979323463742641,
+      "learning_rate": 9.67015449956421e-06,
+      "loss": 0.5301,
+      "step": 1818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4508348847270582,
+      "learning_rate": 9.669700072548778e-06,
+      "loss": 0.5333,
+      "step": 1819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6346115860202617,
+      "learning_rate": 9.66924534340986e-06,
+      "loss": 0.5097,
+      "step": 1820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4982474966609727,
+      "learning_rate": 9.668790312176875e-06,
+      "loss": 0.5244,
+      "step": 1821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8411165317620186,
+      "learning_rate": 9.668334978879265e-06,
+      "loss": 0.5455,
+      "step": 1822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4292594581633755,
+      "learning_rate": 9.66787934354649e-06,
+      "loss": 0.486,
+      "step": 1823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3065997663129374,
+      "learning_rate": 9.667423406208027e-06,
+      "loss": 0.5443,
+      "step": 1824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.579094046218995,
+      "learning_rate": 9.666967166893375e-06,
+      "loss": 0.4972,
+      "step": 1825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6574470320859422,
+      "learning_rate": 9.666510625632054e-06,
+      "loss": 0.5723,
+      "step": 1826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.296926797692981,
+      "learning_rate": 9.666053782453599e-06,
+      "loss": 0.5368,
+      "step": 1827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5627458943990271,
+      "learning_rate": 9.665596637387568e-06,
+      "loss": 0.4995,
+      "step": 1828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4805219479583542,
+      "learning_rate": 9.665139190463539e-06,
+      "loss": 0.5388,
+      "step": 1829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.570146199347442,
+      "learning_rate": 9.664681441711106e-06,
+      "loss": 0.5328,
+      "step": 1830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.202864327517874,
+      "learning_rate": 9.664223391159885e-06,
+      "loss": 0.5047,
+      "step": 1831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.019827046463568,
+      "learning_rate": 9.663765038839515e-06,
+      "loss": 0.4973,
+      "step": 1832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5224290163554257,
+      "learning_rate": 9.663306384779647e-06,
+      "loss": 0.5057,
+      "step": 1833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7678254113514829,
+      "learning_rate": 9.662847429009955e-06,
+      "loss": 0.5446,
+      "step": 1834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6582509691217369,
+      "learning_rate": 9.662388171560135e-06,
+      "loss": 0.5609,
+      "step": 1835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5087165859730982,
+      "learning_rate": 9.661928612459898e-06,
+      "loss": 0.5605,
+      "step": 1836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5580049614146727,
+      "learning_rate": 9.66146875173898e-06,
+      "loss": 0.5353,
+      "step": 1837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5455597557591704,
+      "learning_rate": 9.661008589427133e-06,
+      "loss": 0.5554,
+      "step": 1838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.9736831663080014,
+      "learning_rate": 9.660548125554125e-06,
+      "loss": 0.5289,
+      "step": 1839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5548122729123355,
+      "learning_rate": 9.66008736014975e-06,
+      "loss": 0.5676,
+      "step": 1840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.106114288539542,
+      "learning_rate": 9.65962629324382e-06,
+      "loss": 0.4542,
+      "step": 1841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.2494402280394934,
+      "learning_rate": 9.659164924866165e-06,
+      "loss": 0.5294,
+      "step": 1842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 10.906144039147925,
+      "learning_rate": 9.658703255046634e-06,
+      "loss": 0.5444,
+      "step": 1843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.3440469796227137,
+      "learning_rate": 9.658241283815096e-06,
+      "loss": 0.5518,
+      "step": 1844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.8820609313495038,
+      "learning_rate": 9.657779011201442e-06,
+      "loss": 0.5177,
+      "step": 1845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4163341712307647,
+      "learning_rate": 9.657316437235577e-06,
+      "loss": 0.4951,
+      "step": 1846
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6997179901426247,
+      "learning_rate": 9.656853561947433e-06,
+      "loss": 0.5618,
+      "step": 1847
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6447091796466468,
+      "learning_rate": 9.656390385366956e-06,
+      "loss": 0.5594,
+      "step": 1848
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8346317439532254,
+      "learning_rate": 9.65592690752411e-06,
+      "loss": 0.5087,
+      "step": 1849
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.9181109510977046,
+      "learning_rate": 9.655463128448885e-06,
+      "loss": 0.547,
+      "step": 1850
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8643152815418083,
+      "learning_rate": 9.654999048171286e-06,
+      "loss": 0.5484,
+      "step": 1851
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9999572742707186,
+      "learning_rate": 9.654534666721337e-06,
+      "loss": 0.5462,
+      "step": 1852
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4143696273948014,
+      "learning_rate": 9.654069984129084e-06,
+      "loss": 0.4896,
+      "step": 1853
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4719939420212298,
+      "learning_rate": 9.653605000424593e-06,
+      "loss": 0.556,
+      "step": 1854
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7752739389840464,
+      "learning_rate": 9.653139715637945e-06,
+      "loss": 0.5599,
+      "step": 1855
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8079952365426812,
+      "learning_rate": 9.652674129799245e-06,
+      "loss": 0.478,
+      "step": 1856
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6783707506550835,
+      "learning_rate": 9.652208242938615e-06,
+      "loss": 0.56,
+      "step": 1857
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4963456869203255,
+      "learning_rate": 9.651742055086198e-06,
+      "loss": 0.5311,
+      "step": 1858
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4906937955192159,
+      "learning_rate": 9.651275566272154e-06,
+      "loss": 0.4762,
+      "step": 1859
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7476428707420315,
+      "learning_rate": 9.650808776526666e-06,
+      "loss": 0.5148,
+      "step": 1860
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4726969791063416,
+      "learning_rate": 9.650341685879933e-06,
+      "loss": 0.498,
+      "step": 1861
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.58376727646574,
+      "learning_rate": 9.649874294362178e-06,
+      "loss": 0.5365,
+      "step": 1862
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8777151059392998,
+      "learning_rate": 9.64940660200364e-06,
+      "loss": 0.4922,
+      "step": 1863
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4920573924232605,
+      "learning_rate": 9.648938608834574e-06,
+      "loss": 0.4857,
+      "step": 1864
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9082385730671085,
+      "learning_rate": 9.648470314885264e-06,
+      "loss": 0.5907,
+      "step": 1865
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3755917367638923,
+      "learning_rate": 9.648001720186005e-06,
+      "loss": 0.5187,
+      "step": 1866
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4400922067591324,
+      "learning_rate": 9.647532824767115e-06,
+      "loss": 0.5144,
+      "step": 1867
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.506349325690889,
+      "learning_rate": 9.64706362865893e-06,
+      "loss": 0.567,
+      "step": 1868
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2426292166831137,
+      "learning_rate": 9.646594131891809e-06,
+      "loss": 0.5444,
+      "step": 1869
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.127422761446468,
+      "learning_rate": 9.646124334496126e-06,
+      "loss": 0.5225,
+      "step": 1870
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.996991372549424,
+      "learning_rate": 9.645654236502276e-06,
+      "loss": 0.5146,
+      "step": 1871
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7446835832740144,
+      "learning_rate": 9.645183837940674e-06,
+      "loss": 0.5436,
+      "step": 1872
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6248085400054184,
+      "learning_rate": 9.644713138841754e-06,
+      "loss": 0.5246,
+      "step": 1873
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.143086252612302,
+      "learning_rate": 9.64424213923597e-06,
+      "loss": 0.554,
+      "step": 1874
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0521624980551545,
+      "learning_rate": 9.643770839153796e-06,
+      "loss": 0.4867,
+      "step": 1875
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7585777300523453,
+      "learning_rate": 9.643299238625722e-06,
+      "loss": 0.5434,
+      "step": 1876
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.572531173966726,
+      "learning_rate": 9.642827337682264e-06,
+      "loss": 0.4877,
+      "step": 1877
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9706030696387773,
+      "learning_rate": 9.642355136353949e-06,
+      "loss": 0.5629,
+      "step": 1878
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6352121891277558,
+      "learning_rate": 9.641882634671329e-06,
+      "loss": 0.582,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 11.31733320602329,
+      "learning_rate": 9.641409832664977e-06,
+      "loss": 0.5134,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.289160771882736,
+      "learning_rate": 9.640936730365478e-06,
+      "loss": 0.5148,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.372506564675461,
+      "learning_rate": 9.640463327803445e-06,
+      "loss": 0.4932,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6655094649274812,
+      "learning_rate": 9.639989625009505e-06,
+      "loss": 0.5715,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5348168732024798,
+      "learning_rate": 9.639515622014305e-06,
+      "loss": 0.5457,
+      "step": 1884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5944701019559275,
+      "learning_rate": 9.639041318848517e-06,
+      "loss": 0.5453,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6506860711249078,
+      "learning_rate": 9.63856671554282e-06,
+      "loss": 0.5561,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3484934661566084,
+      "learning_rate": 9.638091812127928e-06,
+      "loss": 0.5228,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5517653798172206,
+      "learning_rate": 9.637616608634561e-06,
+      "loss": 0.528,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8089568202222732,
+      "learning_rate": 9.637141105093468e-06,
+      "loss": 0.5303,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6377038927365635,
+      "learning_rate": 9.636665301535411e-06,
+      "loss": 0.5534,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6489053543987315,
+      "learning_rate": 9.636189197991172e-06,
+      "loss": 0.5283,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5429584733055357,
+      "learning_rate": 9.63571279449156e-06,
+      "loss": 0.4832,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6363844468220257,
+      "learning_rate": 9.635236091067394e-06,
+      "loss": 0.5376,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3671053455566085,
+      "learning_rate": 9.634759087749514e-06,
+      "loss": 0.5744,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5160486796831134,
+      "learning_rate": 9.634281784568787e-06,
+      "loss": 0.5327,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5884070971612532,
+      "learning_rate": 9.63380418155609e-06,
+      "loss": 0.5062,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4566344502319915,
+      "learning_rate": 9.633326278742325e-06,
+      "loss": 0.5759,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8860146630487553,
+      "learning_rate": 9.63284807615841e-06,
+      "loss": 0.4751,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6132877472106457,
+      "learning_rate": 9.632369573835284e-06,
+      "loss": 0.5391,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6238791780200794,
+      "learning_rate": 9.631890771803909e-06,
+      "loss": 0.4599,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4486617776656527,
+      "learning_rate": 9.631411670095258e-06,
+      "loss": 0.4979,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4058252331531722,
+      "learning_rate": 9.630932268740332e-06,
+      "loss": 0.5338,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.04346017356555,
+      "learning_rate": 9.630452567770144e-06,
+      "loss": 0.5607,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4687463979665778,
+      "learning_rate": 9.629972567215734e-06,
+      "loss": 0.4801,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9388915894152863,
+      "learning_rate": 9.629492267108157e-06,
+      "loss": 0.5682,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7339000176460159,
+      "learning_rate": 9.629011667478484e-06,
+      "loss": 0.575,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8281390656291485,
+      "learning_rate": 9.628530768357813e-06,
+      "loss": 0.4948,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4653270075321028,
+      "learning_rate": 9.628049569777257e-06,
+      "loss": 0.516,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6224407398210768,
+      "learning_rate": 9.627568071767946e-06,
+      "loss": 0.5551,
+      "step": 1909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6673207861708482,
+      "learning_rate": 9.627086274361036e-06,
+      "loss": 0.5694,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7113701655987905,
+      "learning_rate": 9.626604177587696e-06,
+      "loss": 0.5301,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0581051329711286,
+      "learning_rate": 9.626121781479118e-06,
+      "loss": 0.5805,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.605840579928456,
+      "learning_rate": 9.625639086066511e-06,
+      "loss": 0.5098,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6848669735443895,
+      "learning_rate": 9.62515609138111e-06,
+      "loss": 0.511,
+      "step": 1914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9044271557992667,
+      "learning_rate": 9.624672797454157e-06,
+      "loss": 0.5195,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1602164265870334,
+      "learning_rate": 9.624189204316925e-06,
+      "loss": 0.5972,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7654281950420349,
+      "learning_rate": 9.623705312000701e-06,
+      "loss": 0.5526,
+      "step": 1917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8515363200676294,
+      "learning_rate": 9.623221120536792e-06,
+      "loss": 0.5297,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6872852435569502,
+      "learning_rate": 9.622736629956525e-06,
+      "loss": 0.5608,
+      "step": 1919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6258595650686778,
+      "learning_rate": 9.622251840291245e-06,
+      "loss": 0.5096,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9314525783056422,
+      "learning_rate": 9.621766751572317e-06,
+      "loss": 0.5598,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5698478227833776,
+      "learning_rate": 9.621281363831126e-06,
+      "loss": 0.4863,
+      "step": 1922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.4269589832401026,
+      "learning_rate": 9.620795677099077e-06,
+      "loss": 0.5225,
+      "step": 1923
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6314829794840122,
+      "learning_rate": 9.620309691407592e-06,
+      "loss": 0.5318,
+      "step": 1924
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5194310597065492,
+      "learning_rate": 9.619823406788114e-06,
+      "loss": 0.4951,
+      "step": 1925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7614795016610035,
+      "learning_rate": 9.619336823272106e-06,
+      "loss": 0.5522,
+      "step": 1926
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.805454404965748,
+      "learning_rate": 9.618849940891046e-06,
+      "loss": 0.5686,
+      "step": 1927
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 7.453542619324918,
+      "learning_rate": 9.618362759676439e-06,
+      "loss": 0.532,
+      "step": 1928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6613239910171699,
+      "learning_rate": 9.617875279659803e-06,
+      "loss": 0.5931,
+      "step": 1929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0617400527585406,
+      "learning_rate": 9.617387500872677e-06,
+      "loss": 0.4785,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5004758332995602,
+      "learning_rate": 9.61689942334662e-06,
+      "loss": 0.4767,
+      "step": 1931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.593341760170481,
+      "learning_rate": 9.61641104711321e-06,
+      "loss": 0.5452,
+      "step": 1932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7383773839174395,
+      "learning_rate": 9.615922372204043e-06,
+      "loss": 0.4588,
+      "step": 1933
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 4.182898045049694,
+      "learning_rate": 9.61543339865074e-06,
+      "loss": 0.5215,
+      "step": 1934
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8015117616952703,
+      "learning_rate": 9.61494412648493e-06,
+      "loss": 0.5234,
+      "step": 1935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.044893856415567,
+      "learning_rate": 9.614454555738274e-06,
+      "loss": 0.4816,
+      "step": 1936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.714073234740042,
+      "learning_rate": 9.613964686442446e-06,
+      "loss": 0.4554,
+      "step": 1937
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9920259924299915,
+      "learning_rate": 9.613474518629138e-06,
+      "loss": 0.4789,
+      "step": 1938
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.574747097966089,
+      "learning_rate": 9.612984052330064e-06,
+      "loss": 0.5208,
+      "step": 1939
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7855558288129156,
+      "learning_rate": 9.612493287576955e-06,
+      "loss": 0.4755,
+      "step": 1940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.3198880221077043,
+      "learning_rate": 9.612002224401565e-06,
+      "loss": 0.5327,
+      "step": 1941
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6722870389170643,
+      "learning_rate": 9.611510862835663e-06,
+      "loss": 0.5601,
+      "step": 1942
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8104386319619111,
+      "learning_rate": 9.611019202911042e-06,
+      "loss": 0.5795,
+      "step": 1943
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7094562742464343,
+      "learning_rate": 9.610527244659512e-06,
+      "loss": 0.5423,
+      "step": 1944
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7855967811002127,
+      "learning_rate": 9.610034988112897e-06,
+      "loss": 0.5371,
+      "step": 1945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.1226150119770044,
+      "learning_rate": 9.609542433303052e-06,
+      "loss": 0.5262,
+      "step": 1946
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4103001112012876,
+      "learning_rate": 9.60904958026184e-06,
+      "loss": 0.4759,
+      "step": 1947
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.394735922551564,
+      "learning_rate": 9.60855642902115e-06,
+      "loss": 0.5074,
+      "step": 1948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.499935072642003,
+      "learning_rate": 9.608062979612889e-06,
+      "loss": 0.535,
+      "step": 1949
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2507857701161162,
+      "learning_rate": 9.60756923206898e-06,
+      "loss": 0.4995,
+      "step": 1950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.331283292668649,
+      "learning_rate": 9.60707518642137e-06,
+      "loss": 0.5007,
+      "step": 1951
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0591413706909243,
+      "learning_rate": 9.606580842702021e-06,
+      "loss": 0.566,
+      "step": 1952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6593037543731923,
+      "learning_rate": 9.60608620094292e-06,
+      "loss": 0.5879,
+      "step": 1953
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3845228695305747,
+      "learning_rate": 9.605591261176064e-06,
+      "loss": 0.5514,
+      "step": 1954
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.0294834405918096,
+      "learning_rate": 9.60509602343348e-06,
+      "loss": 0.5637,
+      "step": 1955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9209999454199727,
+      "learning_rate": 9.604600487747207e-06,
+      "loss": 0.4848,
+      "step": 1956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9446447548083243,
+      "learning_rate": 9.604104654149306e-06,
+      "loss": 0.4826,
+      "step": 1957
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6369386012313103,
+      "learning_rate": 9.603608522671859e-06,
+      "loss": 0.4992,
+      "step": 1958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6522098268516108,
+      "learning_rate": 9.60311209334696e-06,
+      "loss": 0.518,
+      "step": 1959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.546673319598247,
+      "learning_rate": 9.60261536620673e-06,
+      "loss": 0.5042,
+      "step": 1960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5901752631891816,
+      "learning_rate": 9.60211834128331e-06,
+      "loss": 0.5301,
+      "step": 1961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.25079203756869,
+      "learning_rate": 9.60162101860885e-06,
+      "loss": 0.564,
+      "step": 1962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.6233735065992376,
+      "learning_rate": 9.601123398215533e-06,
+      "loss": 0.5147,
+      "step": 1963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.729986969012851,
+      "learning_rate": 9.60062548013555e-06,
+      "loss": 0.5146,
+      "step": 1964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.939563466461983,
+      "learning_rate": 9.600127264401116e-06,
+      "loss": 0.4707,
+      "step": 1965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5696199034258749,
+      "learning_rate": 9.599628751044467e-06,
+      "loss": 0.4531,
+      "step": 1966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6266967564033181,
+      "learning_rate": 9.599129940097853e-06,
+      "loss": 0.5678,
+      "step": 1967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4883283697831413,
+      "learning_rate": 9.59863083159355e-06,
+      "loss": 0.4779,
+      "step": 1968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7246248807352447,
+      "learning_rate": 9.598131425563847e-06,
+      "loss": 0.5043,
+      "step": 1969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.9943019252348066,
+      "learning_rate": 9.597631722041056e-06,
+      "loss": 0.5569,
+      "step": 1970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 5.3855774758865245,
+      "learning_rate": 9.597131721057508e-06,
+      "loss": 0.4608,
+      "step": 1971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7884628353667116,
+      "learning_rate": 9.59663142264555e-06,
+      "loss": 0.5328,
+      "step": 1972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3979294782491385,
+      "learning_rate": 9.596130826837552e-06,
+      "loss": 0.4577,
+      "step": 1973
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5994632476954805,
+      "learning_rate": 9.595629933665904e-06,
+      "loss": 0.5735,
+      "step": 1974
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.006160878390288,
+      "learning_rate": 9.595128743163009e-06,
+      "loss": 0.504,
+      "step": 1975
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6419994751213296,
+      "learning_rate": 9.594627255361298e-06,
+      "loss": 0.5468,
+      "step": 1976
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6296557913145661,
+      "learning_rate": 9.594125470293211e-06,
+      "loss": 0.5525,
+      "step": 1977
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.5450590996954117,
+      "learning_rate": 9.593623387991218e-06,
+      "loss": 0.5273,
+      "step": 1978
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6651739325971893,
+      "learning_rate": 9.593121008487801e-06,
+      "loss": 0.5585,
+      "step": 1979
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6740180466062824,
+      "learning_rate": 9.592618331815461e-06,
+      "loss": 0.4974,
+      "step": 1980
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9609641355773166,
+      "learning_rate": 9.592115358006725e-06,
+      "loss": 0.5442,
+      "step": 1981
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9008689839362534,
+      "learning_rate": 9.591612087094133e-06,
+      "loss": 0.501,
+      "step": 1982
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.056617713083659,
+      "learning_rate": 9.591108519110246e-06,
+      "loss": 0.5661,
+      "step": 1983
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5508514459532974,
+      "learning_rate": 9.590604654087641e-06,
+      "loss": 0.5264,
+      "step": 1984
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.028596966399857,
+      "learning_rate": 9.590100492058922e-06,
+      "loss": 0.4966,
+      "step": 1985
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4063992116273172,
+      "learning_rate": 9.589596033056705e-06,
+      "loss": 0.4863,
+      "step": 1986
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8797127213032618,
+      "learning_rate": 9.589091277113628e-06,
+      "loss": 0.4624,
+      "step": 1987
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.837964870874461,
+      "learning_rate": 9.58858622426235e-06,
+      "loss": 0.5578,
+      "step": 1988
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.879246426021428,
+      "learning_rate": 9.588080874535547e-06,
+      "loss": 0.5101,
+      "step": 1989
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.394393583494497,
+      "learning_rate": 9.587575227965912e-06,
+      "loss": 0.4926,
+      "step": 1990
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.076549783347219,
+      "learning_rate": 9.587069284586162e-06,
+      "loss": 0.4853,
+      "step": 1991
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.6326883649225374,
+      "learning_rate": 9.586563044429027e-06,
+      "loss": 0.4353,
+      "step": 1992
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6078603902071085,
+      "learning_rate": 9.586056507527266e-06,
+      "loss": 0.5476,
+      "step": 1993
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9144519827767175,
+      "learning_rate": 9.585549673913647e-06,
+      "loss": 0.5289,
+      "step": 1994
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7868042808192288,
+      "learning_rate": 9.585042543620964e-06,
+      "loss": 0.4653,
+      "step": 1995
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6101988739600026,
+      "learning_rate": 9.584535116682024e-06,
+      "loss": 0.525,
+      "step": 1996
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.549959156834061,
+      "learning_rate": 9.584027393129662e-06,
+      "loss": 0.533,
+      "step": 1997
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7298063326063522,
+      "learning_rate": 9.583519372996723e-06,
+      "loss": 0.5169,
+      "step": 1998
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.960533985153442,
+      "learning_rate": 9.583011056316076e-06,
+      "loss": 0.5052,
+      "step": 1999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.042457507059685,
+      "learning_rate": 9.58250244312061e-06,
+      "loss": 0.5087,
+      "step": 2000
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5805061347807925,
+      "learning_rate": 9.58199353344323e-06,
+      "loss": 0.4964,
+      "step": 2001
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6816361660868795,
+      "learning_rate": 9.581484327316862e-06,
+      "loss": 0.5632,
+      "step": 2002
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.515994614064876,
+      "learning_rate": 9.580974824774452e-06,
+      "loss": 0.4676,
+      "step": 2003
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4702558803539987,
+      "learning_rate": 9.580465025848964e-06,
+      "loss": 0.4679,
+      "step": 2004
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8866135071526426,
+      "learning_rate": 9.57995493057338e-06,
+      "loss": 0.4649,
+      "step": 2005
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.919991825031542,
+      "learning_rate": 9.579444538980704e-06,
+      "loss": 0.5436,
+      "step": 2006
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.460924148102141,
+      "learning_rate": 9.578933851103955e-06,
+      "loss": 0.5057,
+      "step": 2007
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7917148710566269,
+      "learning_rate": 9.578422866976179e-06,
+      "loss": 0.5454,
+      "step": 2008
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 4.681393069998754,
+      "learning_rate": 9.57791158663043e-06,
+      "loss": 0.5182,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6882539508153274,
+      "learning_rate": 9.577400010099791e-06,
+      "loss": 0.5774,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3935347021610167,
+      "learning_rate": 9.576888137417359e-06,
+      "loss": 0.5082,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4379004557713304,
+      "learning_rate": 9.576375968616253e-06,
+      "loss": 0.4561,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.218044137839161,
+      "learning_rate": 9.575863503729607e-06,
+      "loss": 0.456,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.624595576545622,
+      "learning_rate": 9.57535074279058e-06,
+      "loss": 0.5016,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.734987647831434,
+      "learning_rate": 9.574837685832345e-06,
+      "loss": 0.5444,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7904434658644846,
+      "learning_rate": 9.574324332888097e-06,
+      "loss": 0.5167,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7145130938309323,
+      "learning_rate": 9.573810683991047e-06,
+      "loss": 0.5051,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8812049928002021,
+      "learning_rate": 9.573296739174429e-06,
+      "loss": 0.4792,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9692378478446724,
+      "learning_rate": 9.572782498471495e-06,
+      "loss": 0.5338,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4889322853218172,
+      "learning_rate": 9.572267961915516e-06,
+      "loss": 0.4761,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4912677996970003,
+      "learning_rate": 9.57175312953978e-06,
+      "loss": 0.5376,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5246908260779142,
+      "learning_rate": 9.571238001377597e-06,
+      "loss": 0.5382,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6524357683208102,
+      "learning_rate": 9.570722577462298e-06,
+      "loss": 0.5437,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7788487919012534,
+      "learning_rate": 9.570206857827223e-06,
+      "loss": 0.5225,
+      "step": 2024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5108515525735133,
+      "learning_rate": 9.569690842505746e-06,
+      "loss": 0.5489,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.377732465657477,
+      "learning_rate": 9.569174531531249e-06,
+      "loss": 0.4267,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7378627558321154,
+      "learning_rate": 9.568657924937137e-06,
+      "loss": 0.5348,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6228995537471274,
+      "learning_rate": 9.568141022756832e-06,
+      "loss": 0.5384,
+      "step": 2028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.576654087130885,
+      "learning_rate": 9.567623825023781e-06,
+      "loss": 0.536,
+      "step": 2029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.629275234919402,
+      "learning_rate": 9.567106331771443e-06,
+      "loss": 0.5845,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4489677124859823,
+      "learning_rate": 9.5665885430333e-06,
+      "loss": 0.5789,
+      "step": 2031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5999177822298869,
+      "learning_rate": 9.566070458842851e-06,
+      "loss": 0.5491,
+      "step": 2032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6790315655769896,
+      "learning_rate": 9.565552079233617e-06,
+      "loss": 0.5351,
+      "step": 2033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6190922991049055,
+      "learning_rate": 9.565033404239136e-06,
+      "loss": 0.5708,
+      "step": 2034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5129711452020425,
+      "learning_rate": 9.564514433892967e-06,
+      "loss": 0.4599,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6428463009973222,
+      "learning_rate": 9.563995168228685e-06,
+      "loss": 0.5761,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.284363873913847,
+      "learning_rate": 9.563475607279884e-06,
+      "loss": 0.5158,
+      "step": 2037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 5.059535215244837,
+      "learning_rate": 9.562955751080183e-06,
+      "loss": 0.5723,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3301787019320954,
+      "learning_rate": 9.562435599663213e-06,
+      "loss": 0.554,
+      "step": 2039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.2049616738862214,
+      "learning_rate": 9.561915153062628e-06,
+      "loss": 0.5496,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.574171426614344,
+      "learning_rate": 9.561394411312101e-06,
+      "loss": 0.497,
+      "step": 2041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4155571045947233,
+      "learning_rate": 9.560873374445324e-06,
+      "loss": 0.5016,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.600973038158765,
+      "learning_rate": 9.560352042496003e-06,
+      "loss": 0.5106,
+      "step": 2043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5482967993591092,
+      "learning_rate": 9.559830415497874e-06,
+      "loss": 0.4883,
+      "step": 2044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7819318872980026,
+      "learning_rate": 9.559308493484678e-06,
+      "loss": 0.5706,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.611778446364573,
+      "learning_rate": 9.558786276490188e-06,
+      "loss": 0.5555,
+      "step": 2046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7421506099276582,
+      "learning_rate": 9.558263764548191e-06,
+      "loss": 0.5503,
+      "step": 2047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7444687343288983,
+      "learning_rate": 9.55774095769249e-06,
+      "loss": 0.5206,
+      "step": 2048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.70686489024873,
+      "learning_rate": 9.557217855956912e-06,
+      "loss": 0.5403,
+      "step": 2049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4782358401591296,
+      "learning_rate": 9.556694459375301e-06,
+      "loss": 0.5485,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6466517853359657,
+      "learning_rate": 9.55617076798152e-06,
+      "loss": 0.5234,
+      "step": 2051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9353923381414608,
+      "learning_rate": 9.555646781809448e-06,
+      "loss": 0.57,
+      "step": 2052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5197688339756326,
+      "learning_rate": 9.55512250089299e-06,
+      "loss": 0.5045,
+      "step": 2053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5903060322717537,
+      "learning_rate": 9.554597925266066e-06,
+      "loss": 0.5285,
+      "step": 2054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4782117646404644,
+      "learning_rate": 9.554073054962614e-06,
+      "loss": 0.4806,
+      "step": 2055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7885308177218775,
+      "learning_rate": 9.553547890016592e-06,
+      "loss": 0.5314,
+      "step": 2056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.938781855049388,
+      "learning_rate": 9.553022430461978e-06,
+      "loss": 0.5321,
+      "step": 2057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6054817513725239,
+      "learning_rate": 9.55249667633277e-06,
+      "loss": 0.5586,
+      "step": 2058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5267750804127047,
+      "learning_rate": 9.551970627662982e-06,
+      "loss": 0.5115,
+      "step": 2059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.9415949031960835,
+      "learning_rate": 9.551444284486649e-06,
+      "loss": 0.5267,
+      "step": 2060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5682888938008787,
+      "learning_rate": 9.550917646837825e-06,
+      "loss": 0.5515,
+      "step": 2061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.570009752796976,
+      "learning_rate": 9.550390714750582e-06,
+      "loss": 0.5627,
+      "step": 2062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9566511229347934,
+      "learning_rate": 9.549863488259012e-06,
+      "loss": 0.5641,
+      "step": 2063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0482280967175064,
+      "learning_rate": 9.549335967397228e-06,
+      "loss": 0.4968,
+      "step": 2064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9914595867227152,
+      "learning_rate": 9.548808152199358e-06,
+      "loss": 0.5124,
+      "step": 2065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9232968396869008,
+      "learning_rate": 9.548280042699551e-06,
+      "loss": 0.5378,
+      "step": 2066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.527159329196347,
+      "learning_rate": 9.547751638931975e-06,
+      "loss": 0.5189,
+      "step": 2067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7525773965788447,
+      "learning_rate": 9.547222940930816e-06,
+      "loss": 0.5453,
+      "step": 2068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6609644855259922,
+      "learning_rate": 9.546693948730281e-06,
+      "loss": 0.5638,
+      "step": 2069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8268757786279337,
+      "learning_rate": 9.546164662364596e-06,
+      "loss": 0.5305,
+      "step": 2070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6033676829929758,
+      "learning_rate": 9.545635081868006e-06,
+      "loss": 0.5666,
+      "step": 2071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7333667889017632,
+      "learning_rate": 9.545105207274773e-06,
+      "loss": 0.5519,
+      "step": 2072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5616028321094193,
+      "learning_rate": 9.544575038619176e-06,
+      "loss": 0.5437,
+      "step": 2073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6059526506702847,
+      "learning_rate": 9.54404457593552e-06,
+      "loss": 0.5433,
+      "step": 2074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7962628720685776,
+      "learning_rate": 9.543513819258125e-06,
+      "loss": 0.4803,
+      "step": 2075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7038094571996882,
+      "learning_rate": 9.54298276862133e-06,
+      "loss": 0.502,
+      "step": 2076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.400098088362526,
+      "learning_rate": 9.542451424059491e-06,
+      "loss": 0.4995,
+      "step": 2077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4428748234460604,
+      "learning_rate": 9.541919785606988e-06,
+      "loss": 0.4739,
+      "step": 2078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6203634553538203,
+      "learning_rate": 9.541387853298217e-06,
+      "loss": 0.5493,
+      "step": 2079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6044456136186844,
+      "learning_rate": 9.540855627167593e-06,
+      "loss": 0.5641,
+      "step": 2080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.8488272095627583,
+      "learning_rate": 9.540323107249549e-06,
+      "loss": 0.4509,
+      "step": 2081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8092851040192266,
+      "learning_rate": 9.53979029357854e-06,
+      "loss": 0.5494,
+      "step": 2082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.483299768318366,
+      "learning_rate": 9.539257186189039e-06,
+      "loss": 0.4873,
+      "step": 2083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.806621437048565,
+      "learning_rate": 9.538723785115534e-06,
+      "loss": 0.5327,
+      "step": 2084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.9642216730941904,
+      "learning_rate": 9.538190090392538e-06,
+      "loss": 0.5058,
+      "step": 2085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8153278047222474,
+      "learning_rate": 9.537656102054582e-06,
+      "loss": 0.5233,
+      "step": 2086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.415882449847442,
+      "learning_rate": 9.537121820136208e-06,
+      "loss": 0.5148,
+      "step": 2087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6481448236743124,
+      "learning_rate": 9.536587244671991e-06,
+      "loss": 0.5725,
+      "step": 2088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.912606776206628,
+      "learning_rate": 9.536052375696514e-06,
+      "loss": 0.5396,
+      "step": 2089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6407457116264279,
+      "learning_rate": 9.53551721324438e-06,
+      "loss": 0.5593,
+      "step": 2090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6193963001497935,
+      "learning_rate": 9.534981757350218e-06,
+      "loss": 0.517,
+      "step": 2091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7685896087585327,
+      "learning_rate": 9.534446008048667e-06,
+      "loss": 0.4717,
+      "step": 2092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4447835250096623,
+      "learning_rate": 9.53390996537439e-06,
+      "loss": 0.5443,
+      "step": 2093
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8330164234658446,
+      "learning_rate": 9.533373629362072e-06,
+      "loss": 0.5221,
+      "step": 2094
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5040478392684282,
+      "learning_rate": 9.532837000046408e-06,
+      "loss": 0.5363,
+      "step": 2095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.931012572435919,
+      "learning_rate": 9.532300077462122e-06,
+      "loss": 0.5781,
+      "step": 2096
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5215064601989499,
+      "learning_rate": 9.531762861643949e-06,
+      "loss": 0.53,
+      "step": 2097
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5904653867858773,
+      "learning_rate": 9.531225352626648e-06,
+      "loss": 0.5171,
+      "step": 2098
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 2.0280920271209886,
+      "learning_rate": 9.530687550444994e-06,
+      "loss": 0.5807,
+      "step": 2099
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.8054079400137233,
+      "learning_rate": 9.530149455133782e-06,
+      "loss": 0.5169,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6801987212415377,
+      "learning_rate": 9.529611066727827e-06,
+      "loss": 0.4622,
+      "step": 2101
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7607987420941,
+      "learning_rate": 9.529072385261959e-06,
+      "loss": 0.488,
+      "step": 2102
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7983203641623786,
+      "learning_rate": 9.528533410771034e-06,
+      "loss": 0.5122,
+      "step": 2103
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.701013361806068,
+      "learning_rate": 9.527994143289922e-06,
+      "loss": 0.4882,
+      "step": 2104
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8116597531730263,
+      "learning_rate": 9.527454582853512e-06,
+      "loss": 0.471,
+      "step": 2105
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5779240881862402,
+      "learning_rate": 9.526914729496714e-06,
+      "loss": 0.5004,
+      "step": 2106
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5187493734507571,
+      "learning_rate": 9.526374583254454e-06,
+      "loss": 0.4675,
+      "step": 2107
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3876179761631933,
+      "learning_rate": 9.525834144161681e-06,
+      "loss": 0.5093,
+      "step": 2108
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.007192540896147,
+      "learning_rate": 9.525293412253357e-06,
+      "loss": 0.4598,
+      "step": 2109
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9486770376135167,
+      "learning_rate": 9.524752387564471e-06,
+      "loss": 0.5033,
+      "step": 2110
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7501211150585835,
+      "learning_rate": 9.524211070130023e-06,
+      "loss": 0.5667,
+      "step": 2111
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9392259572779216,
+      "learning_rate": 9.523669459985039e-06,
+      "loss": 0.5638,
+      "step": 2112
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4828847468376445,
+      "learning_rate": 9.523127557164558e-06,
+      "loss": 0.5033,
+      "step": 2113
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6065252681880944,
+      "learning_rate": 9.52258536170364e-06,
+      "loss": 0.5779,
+      "step": 2114
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6630596262414732,
+      "learning_rate": 9.522042873637366e-06,
+      "loss": 0.5046,
+      "step": 2115
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.668916647465811,
+      "learning_rate": 9.521500093000833e-06,
+      "loss": 0.5286,
+      "step": 2116
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2125270477711876,
+      "learning_rate": 9.52095701982916e-06,
+      "loss": 0.4843,
+      "step": 2117
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.254204404113744,
+      "learning_rate": 9.520413654157482e-06,
+      "loss": 0.5155,
+      "step": 2118
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.694252567211013,
+      "learning_rate": 9.519869996020954e-06,
+      "loss": 0.5321,
+      "step": 2119
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.832320394486869,
+      "learning_rate": 9.519326045454749e-06,
+      "loss": 0.4912,
+      "step": 2120
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9471570857269564,
+      "learning_rate": 9.518781802494062e-06,
+      "loss": 0.5609,
+      "step": 2121
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.539677382065425,
+      "learning_rate": 9.518237267174103e-06,
+      "loss": 0.5267,
+      "step": 2122
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4490923506115598,
+      "learning_rate": 9.517692439530104e-06,
+      "loss": 0.476,
+      "step": 2123
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.629027182551497,
+      "learning_rate": 9.517147319597313e-06,
+      "loss": 0.5571,
+      "step": 2124
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7154338341355222,
+      "learning_rate": 9.516601907411e-06,
+      "loss": 0.5832,
+      "step": 2125
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.729014737154734,
+      "learning_rate": 9.516056203006453e-06,
+      "loss": 0.5147,
+      "step": 2126
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3980640712067298,
+      "learning_rate": 9.515510206418976e-06,
+      "loss": 0.4696,
+      "step": 2127
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7317943594479357,
+      "learning_rate": 9.514963917683898e-06,
+      "loss": 0.476,
+      "step": 2128
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4403536605349305,
+      "learning_rate": 9.51441733683656e-06,
+      "loss": 0.5104,
+      "step": 2129
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3785960974255924,
+      "learning_rate": 9.513870463912324e-06,
+      "loss": 0.4891,
+      "step": 2130
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9198488849643011,
+      "learning_rate": 9.513323298946576e-06,
+      "loss": 0.5551,
+      "step": 2131
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7859998219101592,
+      "learning_rate": 9.512775841974714e-06,
+      "loss": 0.5344,
+      "step": 2132
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.63740456365635,
+      "learning_rate": 9.512228093032156e-06,
+      "loss": 0.5761,
+      "step": 2133
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.979243443044541,
+      "learning_rate": 9.511680052154345e-06,
+      "loss": 0.5093,
+      "step": 2134
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3713806588894197,
+      "learning_rate": 9.511131719376738e-06,
+      "loss": 0.5428,
+      "step": 2135
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9608286744608714,
+      "learning_rate": 9.51058309473481e-06,
+      "loss": 0.4815,
+      "step": 2136
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.756086181219677,
+      "learning_rate": 9.510034178264054e-06,
+      "loss": 0.5526,
+      "step": 2137
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8848825413799138,
+      "learning_rate": 9.509484969999989e-06,
+      "loss": 0.5635,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4996951542792805,
+      "learning_rate": 9.508935469978144e-06,
+      "loss": 0.4685,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.9139132477413123,
+      "learning_rate": 9.508385678234073e-06,
+      "loss": 0.5109,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7231226992964058,
+      "learning_rate": 9.507835594803346e-06,
+      "loss": 0.5697,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6828284685680502,
+      "learning_rate": 9.507285219721553e-06,
+      "loss": 0.5662,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.955667760172087,
+      "learning_rate": 9.506734553024304e-06,
+      "loss": 0.5404,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6149749604204677,
+      "learning_rate": 9.506183594747222e-06,
+      "loss": 0.5597,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.371312231628323,
+      "learning_rate": 9.505632344925958e-06,
+      "loss": 0.5172,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.971079035340859,
+      "learning_rate": 9.505080803596176e-06,
+      "loss": 0.4778,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.542722165045504,
+      "learning_rate": 9.50452897079356e-06,
+      "loss": 0.5361,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 5.291693875618912,
+      "learning_rate": 9.503976846553811e-06,
+      "loss": 0.5631,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9558789705837276,
+      "learning_rate": 9.503424430912652e-06,
+      "loss": 0.5266,
+      "step": 2149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.025466277741234,
+      "learning_rate": 9.502871723905825e-06,
+      "loss": 0.5999,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9666408709868786,
+      "learning_rate": 9.502318725569085e-06,
+      "loss": 0.5126,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.666263990859537,
+      "learning_rate": 9.501765435938216e-06,
+      "loss": 0.5128,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7368694923333615,
+      "learning_rate": 9.501211855049011e-06,
+      "loss": 0.5582,
+      "step": 2153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8082447152472192,
+      "learning_rate": 9.500657982937287e-06,
+      "loss": 0.4883,
+      "step": 2154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8936319297779098,
+      "learning_rate": 9.500103819638879e-06,
+      "loss": 0.5669,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9644663181200546,
+      "learning_rate": 9.499549365189643e-06,
+      "loss": 0.5797,
+      "step": 2156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6765265508455378,
+      "learning_rate": 9.498994619625447e-06,
+      "loss": 0.545,
+      "step": 2157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.796049908870172,
+      "learning_rate": 9.498439582982185e-06,
+      "loss": 0.4534,
+      "step": 2158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6460298601515715,
+      "learning_rate": 9.497884255295766e-06,
+      "loss": 0.5469,
+      "step": 2159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.610860572271665,
+      "learning_rate": 9.49732863660212e-06,
+      "loss": 0.4775,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8344738717138687,
+      "learning_rate": 9.496772726937196e-06,
+      "loss": 0.5538,
+      "step": 2161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.439132973223709,
+      "learning_rate": 9.496216526336957e-06,
+      "loss": 0.4746,
+      "step": 2162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7786435972033484,
+      "learning_rate": 9.495660034837393e-06,
+      "loss": 0.5623,
+      "step": 2163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5698008427662702,
+      "learning_rate": 9.495103252474504e-06,
+      "loss": 0.5073,
+      "step": 2164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4909887886223667,
+      "learning_rate": 9.494546179284314e-06,
+      "loss": 0.4869,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5667567774250105,
+      "learning_rate": 9.493988815302867e-06,
+      "loss": 0.5447,
+      "step": 2166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6436980784916715,
+      "learning_rate": 9.493431160566225e-06,
+      "loss": 0.4821,
+      "step": 2167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.590159498895422,
+      "learning_rate": 9.492873215110461e-06,
+      "loss": 0.5056,
+      "step": 2168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6764983946149048,
+      "learning_rate": 9.49231497897168e-06,
+      "loss": 0.5068,
+      "step": 2169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.628110584941614,
+      "learning_rate": 9.491756452185995e-06,
+      "loss": 0.5524,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7202964215007375,
+      "learning_rate": 9.491197634789544e-06,
+      "loss": 0.4811,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7719591304936537,
+      "learning_rate": 9.490638526818482e-06,
+      "loss": 0.5336,
+      "step": 2172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6335562008335842,
+      "learning_rate": 9.49007912830898e-06,
+      "loss": 0.5013,
+      "step": 2173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.4384627257461466,
+      "learning_rate": 9.489519439297234e-06,
+      "loss": 0.5701,
+      "step": 2174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5538572384087763,
+      "learning_rate": 9.488959459819452e-06,
+      "loss": 0.4505,
+      "step": 2175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.013405254450756,
+      "learning_rate": 9.488399189911866e-06,
+      "loss": 0.5855,
+      "step": 2176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.539930182953416,
+      "learning_rate": 9.487838629610725e-06,
+      "loss": 0.5002,
+      "step": 2177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8617361968328119,
+      "learning_rate": 9.487277778952293e-06,
+      "loss": 0.486,
+      "step": 2178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 4.132060974960517,
+      "learning_rate": 9.486716637972862e-06,
+      "loss": 0.5837,
+      "step": 2179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7430838648860986,
+      "learning_rate": 9.486155206708732e-06,
+      "loss": 0.5591,
+      "step": 2180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.715512990093565,
+      "learning_rate": 9.485593485196227e-06,
+      "loss": 0.5042,
+      "step": 2181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5721527506048059,
+      "learning_rate": 9.485031473471693e-06,
+      "loss": 0.4908,
+      "step": 2182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.6696885337679284,
+      "learning_rate": 9.484469171571491e-06,
+      "loss": 0.5093,
+      "step": 2183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 3.250448576673862,
+      "learning_rate": 9.483906579531998e-06,
+      "loss": 0.5087,
+      "step": 2184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6823222839716344,
+      "learning_rate": 9.483343697389617e-06,
+      "loss": 0.5696,
+      "step": 2185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.685245524775517,
+      "learning_rate": 9.482780525180763e-06,
+      "loss": 0.537,
+      "step": 2186
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.207196749810209,
+      "learning_rate": 9.482217062941872e-06,
+      "loss": 0.5517,
+      "step": 2187
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.545300585877083,
+      "learning_rate": 9.481653310709402e-06,
+      "loss": 0.5324,
+      "step": 2188
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.535925527956117,
+      "learning_rate": 9.481089268519825e-06,
+      "loss": 0.4974,
+      "step": 2189
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6625140261892406,
+      "learning_rate": 9.480524936409634e-06,
+      "loss": 0.5228,
+      "step": 2190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9596707494165937,
+      "learning_rate": 9.479960314415341e-06,
+      "loss": 0.5392,
+      "step": 2191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6415917752005628,
+      "learning_rate": 9.479395402573476e-06,
+      "loss": 0.5564,
+      "step": 2192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.209430408861574,
+      "learning_rate": 9.47883020092059e-06,
+      "loss": 0.4596,
+      "step": 2193
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4880507942805108,
+      "learning_rate": 9.478264709493249e-06,
+      "loss": 0.5139,
+      "step": 2194
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5931903712120157,
+      "learning_rate": 9.477698928328036e-06,
+      "loss": 0.5265,
+      "step": 2195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0880721146531847,
+      "learning_rate": 9.477132857461563e-06,
+      "loss": 0.5657,
+      "step": 2196
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8169151361905544,
+      "learning_rate": 9.476566496930451e-06,
+      "loss": 0.5304,
+      "step": 2197
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3868525052029015,
+      "learning_rate": 9.47599984677134e-06,
+      "loss": 0.5653,
+      "step": 2198
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.3511931780378257,
+      "learning_rate": 9.475432907020896e-06,
+      "loss": 0.5451,
+      "step": 2199
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4832665092620316,
+      "learning_rate": 9.474865677715798e-06,
+      "loss": 0.4904,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5152439314768429,
+      "learning_rate": 9.474298158892745e-06,
+      "loss": 0.4939,
+      "step": 2201
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6793842839059967,
+      "learning_rate": 9.473730350588452e-06,
+      "loss": 0.5504,
+      "step": 2202
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.755852820715602,
+      "learning_rate": 9.47316225283966e-06,
+      "loss": 0.5135,
+      "step": 2203
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5133950253550443,
+      "learning_rate": 9.472593865683123e-06,
+      "loss": 0.535,
+      "step": 2204
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7667686447617437,
+      "learning_rate": 9.47202518915561e-06,
+      "loss": 0.5206,
+      "step": 2205
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.1064077042558353,
+      "learning_rate": 9.47145622329392e-06,
+      "loss": 0.4994,
+      "step": 2206
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.689717774744424,
+      "learning_rate": 9.47088696813486e-06,
+      "loss": 0.5019,
+      "step": 2207
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6981276524475272,
+      "learning_rate": 9.470317423715263e-06,
+      "loss": 0.5795,
+      "step": 2208
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.2714049819003086,
+      "learning_rate": 9.469747590071976e-06,
+      "loss": 0.5373,
+      "step": 2209
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.076811823304265,
+      "learning_rate": 9.469177467241867e-06,
+      "loss": 0.5172,
+      "step": 2210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8150532772877375,
+      "learning_rate": 9.468607055261822e-06,
+      "loss": 0.4869,
+      "step": 2211
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0898307624149535,
+      "learning_rate": 9.468036354168748e-06,
+      "loss": 0.5526,
+      "step": 2212
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.886619357714772,
+      "learning_rate": 9.467465363999566e-06,
+      "loss": 0.4988,
+      "step": 2213
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.4782408804663034,
+      "learning_rate": 9.466894084791218e-06,
+      "loss": 0.5105,
+      "step": 2214
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7335450606919611,
+      "learning_rate": 9.466322516580668e-06,
+      "loss": 0.5366,
+      "step": 2215
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.753197706197562,
+      "learning_rate": 9.465750659404894e-06,
+      "loss": 0.5353,
+      "step": 2216
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7451030108602756,
+      "learning_rate": 9.465178513300892e-06,
+      "loss": 0.501,
+      "step": 2217
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0408632587346482,
+      "learning_rate": 9.464606078305683e-06,
+      "loss": 0.5093,
+      "step": 2218
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8795835710619089,
+      "learning_rate": 9.4640333544563e-06,
+      "loss": 0.5329,
+      "step": 2219
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7439632386830464,
+      "learning_rate": 9.463460341789799e-06,
+      "loss": 0.4925,
+      "step": 2220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6178824467668695,
+      "learning_rate": 9.462887040343254e-06,
+      "loss": 0.5473,
+      "step": 2221
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5467664061824875,
+      "learning_rate": 9.462313450153754e-06,
+      "loss": 0.507,
+      "step": 2222
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.8949400220430728,
+      "learning_rate": 9.461739571258413e-06,
+      "loss": 0.5455,
+      "step": 2223
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5390597003227833,
+      "learning_rate": 9.461165403694357e-06,
+      "loss": 0.5307,
+      "step": 2224
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.0636367595569562,
+      "learning_rate": 9.460590947498736e-06,
+      "loss": 0.5532,
+      "step": 2225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4664087207213812,
+      "learning_rate": 9.460016202708714e-06,
+      "loss": 0.5334,
+      "step": 2226
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8014685693556765,
+      "learning_rate": 9.45944116936148e-06,
+      "loss": 0.4824,
+      "step": 2227
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5358028986672532,
+      "learning_rate": 9.458865847494236e-06,
+      "loss": 0.5024,
+      "step": 2228
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6573628026759983,
+      "learning_rate": 9.458290237144205e-06,
+      "loss": 0.5842,
+      "step": 2229
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6602263633637209,
+      "learning_rate": 9.457714338348626e-06,
+      "loss": 0.5159,
+      "step": 2230
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6139024681438195,
+      "learning_rate": 9.457138151144764e-06,
+      "loss": 0.5792,
+      "step": 2231
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5945183555884275,
+      "learning_rate": 9.456561675569892e-06,
+      "loss": 0.5278,
+      "step": 2232
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6033556554984202,
+      "learning_rate": 9.455984911661309e-06,
+      "loss": 0.5485,
+      "step": 2233
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5674993633382198,
+      "learning_rate": 9.455407859456332e-06,
+      "loss": 0.5783,
+      "step": 2234
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.11408179360972,
+      "learning_rate": 9.454830518992294e-06,
+      "loss": 0.5537,
+      "step": 2235
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4763932610312303,
+      "learning_rate": 9.454252890306552e-06,
+      "loss": 0.5417,
+      "step": 2236
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6117372283316223,
+      "learning_rate": 9.453674973436473e-06,
+      "loss": 0.5166,
+      "step": 2237
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.678477108921375,
+      "learning_rate": 9.45309676841945e-06,
+      "loss": 0.4941,
+      "step": 2238
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4799043555394564,
+      "learning_rate": 9.45251827529289e-06,
+      "loss": 0.4846,
+      "step": 2239
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6135290399646561,
+      "learning_rate": 9.451939494094224e-06,
+      "loss": 0.5539,
+      "step": 2240
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7848469738025845,
+      "learning_rate": 9.451360424860894e-06,
+      "loss": 0.4955,
+      "step": 2241
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5525750642375514,
+      "learning_rate": 9.45078106763037e-06,
+      "loss": 0.5581,
+      "step": 2242
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5881333343918638,
+      "learning_rate": 9.450201422440133e-06,
+      "loss": 0.4898,
+      "step": 2243
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0301774323972652,
+      "learning_rate": 9.449621489327685e-06,
+      "loss": 0.4648,
+      "step": 2244
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8841544844523594,
+      "learning_rate": 9.449041268330549e-06,
+      "loss": 0.5193,
+      "step": 2245
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4764900810287394,
+      "learning_rate": 9.44846075948626e-06,
+      "loss": 0.5027,
+      "step": 2246
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5954800081112893,
+      "learning_rate": 9.44787996283238e-06,
+      "loss": 0.5478,
+      "step": 2247
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.531125130556519,
+      "learning_rate": 9.447298878406485e-06,
+      "loss": 0.5076,
+      "step": 2248
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5958142527672674,
+      "learning_rate": 9.446717506246172e-06,
+      "loss": 0.5129,
+      "step": 2249
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8909909750083016,
+      "learning_rate": 9.446135846389053e-06,
+      "loss": 0.5618,
+      "step": 2250
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0486469940678385,
+      "learning_rate": 9.44555389887276e-06,
+      "loss": 0.5214,
+      "step": 2251
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6477507168927046,
+      "learning_rate": 9.444971663734947e-06,
+      "loss": 0.5282,
+      "step": 2252
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4901228179774741,
+      "learning_rate": 9.44438914101328e-06,
+      "loss": 0.4987,
+      "step": 2253
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7435292091271193,
+      "learning_rate": 9.443806330745452e-06,
+      "loss": 0.5298,
+      "step": 2254
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6021033278007386,
+      "learning_rate": 9.443223232969166e-06,
+      "loss": 0.483,
+      "step": 2255
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.731336969626263,
+      "learning_rate": 9.442639847722148e-06,
+      "loss": 0.5087,
+      "step": 2256
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5747625880693306,
+      "learning_rate": 9.442056175042144e-06,
+      "loss": 0.4914,
+      "step": 2257
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.782548503408499,
+      "learning_rate": 9.441472214966917e-06,
+      "loss": 0.5115,
+      "step": 2258
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.520027621733375,
+      "learning_rate": 9.440887967534249e-06,
+      "loss": 0.4215,
+      "step": 2259
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4367831231971078,
+      "learning_rate": 9.440303432781938e-06,
+      "loss": 0.5158,
+      "step": 2260
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8834995644565726,
+      "learning_rate": 9.439718610747804e-06,
+      "loss": 0.5263,
+      "step": 2261
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6061854911063451,
+      "learning_rate": 9.439133501469684e-06,
+      "loss": 0.5204,
+      "step": 2262
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.459743685437855,
+      "learning_rate": 9.438548104985433e-06,
+      "loss": 0.4606,
+      "step": 2263
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0991237285913638,
+      "learning_rate": 9.437962421332926e-06,
+      "loss": 0.5086,
+      "step": 2264
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7677020064549238,
+      "learning_rate": 9.437376450550055e-06,
+      "loss": 0.5572,
+      "step": 2265
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.033569799019311,
+      "learning_rate": 9.436790192674734e-06,
+      "loss": 0.5466,
+      "step": 2266
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.12638574212309,
+      "learning_rate": 9.43620364774489e-06,
+      "loss": 0.5222,
+      "step": 2267
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6705963302741609,
+      "learning_rate": 9.435616815798476e-06,
+      "loss": 0.5523,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2791309706239464,
+      "learning_rate": 9.435029696873455e-06,
+      "loss": 0.5978,
+      "step": 2269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.634397078943637,
+      "learning_rate": 9.434442291007812e-06,
+      "loss": 0.5036,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5920390987773815,
+      "learning_rate": 9.433854598239557e-06,
+      "loss": 0.5565,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.4070029869294918,
+      "learning_rate": 9.433266618606708e-06,
+      "loss": 0.4633,
+      "step": 2272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.41621220579687,
+      "learning_rate": 9.432678352147309e-06,
+      "loss": 0.5047,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4620047371608027,
+      "learning_rate": 9.43208979889942e-06,
+      "loss": 0.5282,
+      "step": 2274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.812494745292715,
+      "learning_rate": 9.431500958901118e-06,
+      "loss": 0.5393,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.978913246603885,
+      "learning_rate": 9.4309118321905e-06,
+      "loss": 0.529,
+      "step": 2276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4983438275082541,
+      "learning_rate": 9.430322418805683e-06,
+      "loss": 0.492,
+      "step": 2277
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4951832867201973,
+      "learning_rate": 9.429732718784803e-06,
+      "loss": 0.5144,
+      "step": 2278
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7883941178800622,
+      "learning_rate": 9.42914273216601e-06,
+      "loss": 0.5328,
+      "step": 2279
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.647159648946074,
+      "learning_rate": 9.428552458987476e-06,
+      "loss": 0.5458,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7109976036822474,
+      "learning_rate": 9.427961899287393e-06,
+      "loss": 0.475,
+      "step": 2281
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.421955507918678,
+      "learning_rate": 9.427371053103964e-06,
+      "loss": 0.477,
+      "step": 2282
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.733695639419623,
+      "learning_rate": 9.426779920475424e-06,
+      "loss": 0.5275,
+      "step": 2283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5560713998579256,
+      "learning_rate": 9.42618850144001e-06,
+      "loss": 0.499,
+      "step": 2284
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.712873691592354,
+      "learning_rate": 9.425596796035991e-06,
+      "loss": 0.5177,
+      "step": 2285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5795401756221055,
+      "learning_rate": 9.425004804301651e-06,
+      "loss": 0.5025,
+      "step": 2286
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6292684447911882,
+      "learning_rate": 9.424412526275287e-06,
+      "loss": 0.5459,
+      "step": 2287
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4913136948225565,
+      "learning_rate": 9.423819961995222e-06,
+      "loss": 0.5503,
+      "step": 2288
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.544311050129815,
+      "learning_rate": 9.42322711149979e-06,
+      "loss": 0.5331,
+      "step": 2289
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5922629329754747,
+      "learning_rate": 9.422633974827351e-06,
+      "loss": 0.4649,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2765294820326885,
+      "learning_rate": 9.422040552016281e-06,
+      "loss": 0.5001,
+      "step": 2291
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5271345926361055,
+      "learning_rate": 9.421446843104971e-06,
+      "loss": 0.525,
+      "step": 2292
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.075259498546482,
+      "learning_rate": 9.420852848131834e-06,
+      "loss": 0.5178,
+      "step": 2293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9203221083847481,
+      "learning_rate": 9.420258567135301e-06,
+      "loss": 0.4975,
+      "step": 2294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5971047418818387,
+      "learning_rate": 9.419664000153822e-06,
+      "loss": 0.5732,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5730136796337478,
+      "learning_rate": 9.41906914722586e-06,
+      "loss": 0.5372,
+      "step": 2296
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.103266285482752,
+      "learning_rate": 9.418474008389909e-06,
+      "loss": 0.5174,
+      "step": 2297
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6108121294380026,
+      "learning_rate": 9.41787858368447e-06,
+      "loss": 0.56,
+      "step": 2298
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4329732799630057,
+      "learning_rate": 9.417282873148064e-06,
+      "loss": 0.4745,
+      "step": 2299
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8874811943930814,
+      "learning_rate": 9.416686876819236e-06,
+      "loss": 0.5307,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4378028787099404,
+      "learning_rate": 9.416090594736544e-06,
+      "loss": 0.475,
+      "step": 2301
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.3691513629188687,
+      "learning_rate": 9.415494026938567e-06,
+      "loss": 0.5209,
+      "step": 2302
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7221289050405197,
+      "learning_rate": 9.414897173463903e-06,
+      "loss": 0.5221,
+      "step": 2303
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5929202454879234,
+      "learning_rate": 9.414300034351168e-06,
+      "loss": 0.4972,
+      "step": 2304
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6109479125422732,
+      "learning_rate": 9.413702609638996e-06,
+      "loss": 0.5677,
+      "step": 2305
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6363986099004775,
+      "learning_rate": 9.413104899366037e-06,
+      "loss": 0.5724,
+      "step": 2306
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.391726569386818,
+      "learning_rate": 9.412506903570967e-06,
+      "loss": 0.5378,
+      "step": 2307
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.989213820339274,
+      "learning_rate": 9.41190862229247e-06,
+      "loss": 0.5336,
+      "step": 2308
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6785683709464219,
+      "learning_rate": 9.411310055569256e-06,
+      "loss": 0.5,
+      "step": 2309
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6382645087715891,
+      "learning_rate": 9.410711203440054e-06,
+      "loss": 0.5829,
+      "step": 2310
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6535352912465338,
+      "learning_rate": 9.410112065943607e-06,
+      "loss": 0.5324,
+      "step": 2311
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.080201920354252,
+      "learning_rate": 9.409512643118677e-06,
+      "loss": 0.5298,
+      "step": 2312
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.227892563087289,
+      "learning_rate": 9.40891293500405e-06,
+      "loss": 0.5825,
+      "step": 2313
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7525007702316326,
+      "learning_rate": 9.408312941638522e-06,
+      "loss": 0.4959,
+      "step": 2314
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7172000295472807,
+      "learning_rate": 9.407712663060912e-06,
+      "loss": 0.5381,
+      "step": 2315
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3769873144991938,
+      "learning_rate": 9.40711209931006e-06,
+      "loss": 0.4477,
+      "step": 2316
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9541325857621255,
+      "learning_rate": 9.40651125042482e-06,
+      "loss": 0.472,
+      "step": 2317
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5255693045981917,
+      "learning_rate": 9.405910116444069e-06,
+      "loss": 0.5291,
+      "step": 2318
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4968811711312802,
+      "learning_rate": 9.405308697406696e-06,
+      "loss": 0.5866,
+      "step": 2319
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5268983974770414,
+      "learning_rate": 9.404706993351612e-06,
+      "loss": 0.5451,
+      "step": 2320
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.948464827515323,
+      "learning_rate": 9.404105004317748e-06,
+      "loss": 0.5353,
+      "step": 2321
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0058347411167414,
+      "learning_rate": 9.403502730344051e-06,
+      "loss": 0.5327,
+      "step": 2322
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.614822854056582,
+      "learning_rate": 9.402900171469488e-06,
+      "loss": 0.5125,
+      "step": 2323
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4282983644306353,
+      "learning_rate": 9.402297327733046e-06,
+      "loss": 0.5486,
+      "step": 2324
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.663568770889312,
+      "learning_rate": 9.401694199173723e-06,
+      "loss": 0.5623,
+      "step": 2325
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.001983857730837,
+      "learning_rate": 9.401090785830544e-06,
+      "loss": 0.579,
+      "step": 2326
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6503426383795254,
+      "learning_rate": 9.40048708774255e-06,
+      "loss": 0.4753,
+      "step": 2327
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6427289067911675,
+      "learning_rate": 9.399883104948796e-06,
+      "loss": 0.4748,
+      "step": 2328
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.804868454992703,
+      "learning_rate": 9.399278837488361e-06,
+      "loss": 0.5749,
+      "step": 2329
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.613403289304872,
+      "learning_rate": 9.398674285400343e-06,
+      "loss": 0.5042,
+      "step": 2330
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6462151024874334,
+      "learning_rate": 9.39806944872385e-06,
+      "loss": 0.5336,
+      "step": 2331
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4372057471315656,
+      "learning_rate": 9.397464327498017e-06,
+      "loss": 0.4723,
+      "step": 2332
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.0397146824410655,
+      "learning_rate": 9.396858921761997e-06,
+      "loss": 0.5209,
+      "step": 2333
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.163052873395621,
+      "learning_rate": 9.396253231554955e-06,
+      "loss": 0.4878,
+      "step": 2334
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.273409355808137,
+      "learning_rate": 9.39564725691608e-06,
+      "loss": 0.4835,
+      "step": 2335
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.8436663511430758,
+      "learning_rate": 9.395040997884577e-06,
+      "loss": 0.4967,
+      "step": 2336
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9421779480408548,
+      "learning_rate": 9.394434454499672e-06,
+      "loss": 0.5519,
+      "step": 2337
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7358446237705316,
+      "learning_rate": 9.393827626800607e-06,
+      "loss": 0.5457,
+      "step": 2338
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6700770126555389,
+      "learning_rate": 9.39322051482664e-06,
+      "loss": 0.4648,
+      "step": 2339
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4697413164756665,
+      "learning_rate": 9.392613118617055e-06,
+      "loss": 0.5048,
+      "step": 2340
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6671480675763035,
+      "learning_rate": 9.392005438211146e-06,
+      "loss": 0.5176,
+      "step": 2341
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5875622720488974,
+      "learning_rate": 9.391397473648232e-06,
+      "loss": 0.5256,
+      "step": 2342
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7115004482474931,
+      "learning_rate": 9.390789224967646e-06,
+      "loss": 0.5122,
+      "step": 2343
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7742578424242894,
+      "learning_rate": 9.39018069220874e-06,
+      "loss": 0.5419,
+      "step": 2344
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.152852308694212,
+      "learning_rate": 9.389571875410887e-06,
+      "loss": 0.5359,
+      "step": 2345
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5067936452570176,
+      "learning_rate": 9.388962774613476e-06,
+      "loss": 0.4673,
+      "step": 2346
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9685809243174113,
+      "learning_rate": 9.388353389855914e-06,
+      "loss": 0.4706,
+      "step": 2347
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6114036773698448,
+      "learning_rate": 9.387743721177628e-06,
+      "loss": 0.5156,
+      "step": 2348
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.9090615151170502,
+      "learning_rate": 9.387133768618065e-06,
+      "loss": 0.489,
+      "step": 2349
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2684801172775457,
+      "learning_rate": 9.386523532216686e-06,
+      "loss": 0.556,
+      "step": 2350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7651268602140282,
+      "learning_rate": 9.385913012012972e-06,
+      "loss": 0.5176,
+      "step": 2351
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3892053106044289,
+      "learning_rate": 9.385302208046423e-06,
+      "loss": 0.4673,
+      "step": 2352
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.5423652308794584,
+      "learning_rate": 9.384691120356559e-06,
+      "loss": 0.4989,
+      "step": 2353
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.783193599489322,
+      "learning_rate": 9.384079748982915e-06,
+      "loss": 0.5681,
+      "step": 2354
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4606282353341635,
+      "learning_rate": 9.383468093965046e-06,
+      "loss": 0.4939,
+      "step": 2355
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3812283129012592,
+      "learning_rate": 9.382856155342528e-06,
+      "loss": 0.537,
+      "step": 2356
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4507625015503391,
+      "learning_rate": 9.382243933154947e-06,
+      "loss": 0.5429,
+      "step": 2357
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.938593077995465,
+      "learning_rate": 9.381631427441917e-06,
+      "loss": 0.5186,
+      "step": 2358
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4418782447928233,
+      "learning_rate": 9.381018638243066e-06,
+      "loss": 0.5648,
+      "step": 2359
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.746298084827724,
+      "learning_rate": 9.38040556559804e-06,
+      "loss": 0.5089,
+      "step": 2360
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7254748018946469,
+      "learning_rate": 9.379792209546506e-06,
+      "loss": 0.4985,
+      "step": 2361
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5121379859610484,
+      "learning_rate": 9.379178570128144e-06,
+      "loss": 0.4862,
+      "step": 2362
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5273655850392251,
+      "learning_rate": 9.378564647382657e-06,
+      "loss": 0.4773,
+      "step": 2363
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4928443462884877,
+      "learning_rate": 9.377950441349765e-06,
+      "loss": 0.5315,
+      "step": 2364
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6954159335832535,
+      "learning_rate": 9.377335952069207e-06,
+      "loss": 0.5536,
+      "step": 2365
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.753455783569322,
+      "learning_rate": 9.37672117958074e-06,
+      "loss": 0.4251,
+      "step": 2366
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3533838293796452,
+      "learning_rate": 9.376106123924135e-06,
+      "loss": 0.5111,
+      "step": 2367
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5963480316635105,
+      "learning_rate": 9.375490785139192e-06,
+      "loss": 0.5606,
+      "step": 2368
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5838545600315282,
+      "learning_rate": 9.374875163265718e-06,
+      "loss": 0.5421,
+      "step": 2369
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3471547695430006,
+      "learning_rate": 9.374259258343543e-06,
+      "loss": 0.498,
+      "step": 2370
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.120084112071938,
+      "learning_rate": 9.373643070412516e-06,
+      "loss": 0.4499,
+      "step": 2371
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6153495042905446,
+      "learning_rate": 9.373026599512504e-06,
+      "loss": 0.4955,
+      "step": 2372
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5706488942444603,
+      "learning_rate": 9.372409845683395e-06,
+      "loss": 0.4992,
+      "step": 2373
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3773229266923068,
+      "learning_rate": 9.371792808965086e-06,
+      "loss": 0.473,
+      "step": 2374
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5106853887897262,
+      "learning_rate": 9.371175489397501e-06,
+      "loss": 0.5012,
+      "step": 2375
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 14.224879493818861,
+      "learning_rate": 9.370557887020579e-06,
+      "loss": 0.5281,
+      "step": 2376
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7067366198261651,
+      "learning_rate": 9.369940001874282e-06,
+      "loss": 0.5714,
+      "step": 2377
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4186155039940087,
+      "learning_rate": 9.369321833998582e-06,
+      "loss": 0.4996,
+      "step": 2378
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6582392131451884,
+      "learning_rate": 9.368703383433476e-06,
+      "loss": 0.5299,
+      "step": 2379
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4304662679294322,
+      "learning_rate": 9.368084650218974e-06,
+      "loss": 0.4917,
+      "step": 2380
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8664515510210562,
+      "learning_rate": 9.367465634395111e-06,
+      "loss": 0.576,
+      "step": 2381
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.897548791650952,
+      "learning_rate": 9.366846336001935e-06,
+      "loss": 0.4985,
+      "step": 2382
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4597056332333282,
+      "learning_rate": 9.366226755079513e-06,
+      "loss": 0.5351,
+      "step": 2383
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5183534120009308,
+      "learning_rate": 9.365606891667932e-06,
+      "loss": 0.4534,
+      "step": 2384
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.004775628570737,
+      "learning_rate": 9.364986745807295e-06,
+      "loss": 0.5288,
+      "step": 2385
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5505787120960475,
+      "learning_rate": 9.364366317537727e-06,
+      "loss": 0.5588,
+      "step": 2386
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5966758135598842,
+      "learning_rate": 9.363745606899365e-06,
+      "loss": 0.535,
+      "step": 2387
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9483306329129215,
+      "learning_rate": 9.363124613932374e-06,
+      "loss": 0.5171,
+      "step": 2388
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3932276354836652,
+      "learning_rate": 9.362503338676927e-06,
+      "loss": 0.4992,
+      "step": 2389
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8327744324453519,
+      "learning_rate": 9.36188178117322e-06,
+      "loss": 0.4908,
+      "step": 2390
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7902601310897783,
+      "learning_rate": 9.361259941461469e-06,
+      "loss": 0.4967,
+      "step": 2391
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6598261179331797,
+      "learning_rate": 9.360637819581906e-06,
+      "loss": 0.4928,
+      "step": 2392
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.638804222146243,
+      "learning_rate": 9.360015415574779e-06,
+      "loss": 0.4621,
+      "step": 2393
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6848555409083126,
+      "learning_rate": 9.359392729480358e-06,
+      "loss": 0.533,
+      "step": 2394
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3704329017697905,
+      "learning_rate": 9.35876976133893e-06,
+      "loss": 0.5313,
+      "step": 2395
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3399479966682195,
+      "learning_rate": 9.358146511190803e-06,
+      "loss": 0.5031,
+      "step": 2396
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.729187889222948,
+      "learning_rate": 9.357522979076295e-06,
+      "loss": 0.5071,
+      "step": 2397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6830775472357482,
+      "learning_rate": 9.356899165035751e-06,
+      "loss": 0.5285,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7711701959265302,
+      "learning_rate": 9.35627506910953e-06,
+      "loss": 0.5078,
+      "step": 2399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4198755177166493,
+      "learning_rate": 9.355650691338013e-06,
+      "loss": 0.4736,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6243408343941619,
+      "learning_rate": 9.355026031761593e-06,
+      "loss": 0.5317,
+      "step": 2401
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8431603246106645,
+      "learning_rate": 9.354401090420687e-06,
+      "loss": 0.4836,
+      "step": 2402
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.440158501619492,
+      "learning_rate": 9.353775867355724e-06,
+      "loss": 0.5762,
+      "step": 2403
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9309948547716995,
+      "learning_rate": 9.35315036260716e-06,
+      "loss": 0.4504,
+      "step": 2404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 3.684308469533794,
+      "learning_rate": 9.35252457621546e-06,
+      "loss": 0.5281,
+      "step": 2405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.3030591476308606,
+      "learning_rate": 9.351898508221116e-06,
+      "loss": 0.5302,
+      "step": 2406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.822537922753776,
+      "learning_rate": 9.351272158664631e-06,
+      "loss": 0.5299,
+      "step": 2407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5898885623482772,
+      "learning_rate": 9.35064552758653e-06,
+      "loss": 0.5011,
+      "step": 2408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9840236663253659,
+      "learning_rate": 9.350018615027353e-06,
+      "loss": 0.5354,
+      "step": 2409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.0493243032414052,
+      "learning_rate": 9.349391421027663e-06,
+      "loss": 0.5387,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.2296397282355986,
+      "learning_rate": 9.34876394562804e-06,
+      "loss": 0.5384,
+      "step": 2411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.136592604878245,
+      "learning_rate": 9.348136188869075e-06,
+      "loss": 0.4543,
+      "step": 2412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.148440053748222,
+      "learning_rate": 9.347508150791389e-06,
+      "loss": 0.5459,
+      "step": 2413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4285905857377035,
+      "learning_rate": 9.346879831435611e-06,
+      "loss": 0.4948,
+      "step": 2414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6384862729910865,
+      "learning_rate": 9.346251230842396e-06,
+      "loss": 0.5301,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6422996790529345,
+      "learning_rate": 9.345622349052412e-06,
+      "loss": 0.5193,
+      "step": 2416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4651279940690312,
+      "learning_rate": 9.344993186106346e-06,
+      "loss": 0.4933,
+      "step": 2417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5328998184140752,
+      "learning_rate": 9.344363742044904e-06,
+      "loss": 0.5384,
+      "step": 2418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4079090765844,
+      "learning_rate": 9.343734016908814e-06,
+      "loss": 0.464,
+      "step": 2419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7404645029611585,
+      "learning_rate": 9.343104010738812e-06,
+      "loss": 0.5273,
+      "step": 2420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.5842554817997296,
+      "learning_rate": 9.342473723575664e-06,
+      "loss": 0.5023,
+      "step": 2421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6554813354428914,
+      "learning_rate": 9.341843155460146e-06,
+      "loss": 0.5433,
+      "step": 2422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.1110350884669673,
+      "learning_rate": 9.341212306433057e-06,
+      "loss": 0.5458,
+      "step": 2423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6454758135191926,
+      "learning_rate": 9.34058117653521e-06,
+      "loss": 0.5535,
+      "step": 2424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6687962762838733,
+      "learning_rate": 9.339949765807439e-06,
+      "loss": 0.508,
+      "step": 2425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5668257223259372,
+      "learning_rate": 9.339318074290595e-06,
+      "loss": 0.5447,
+      "step": 2426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.394214298203269,
+      "learning_rate": 9.338686102025548e-06,
+      "loss": 0.5093,
+      "step": 2427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.767121032019649,
+      "learning_rate": 9.338053849053186e-06,
+      "loss": 0.4809,
+      "step": 2428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4683517615102588,
+      "learning_rate": 9.337421315414416e-06,
+      "loss": 0.5146,
+      "step": 2429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6643181930892917,
+      "learning_rate": 9.336788501150161e-06,
+      "loss": 0.5625,
+      "step": 2430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5348951349170583,
+      "learning_rate": 9.336155406301361e-06,
+      "loss": 0.5173,
+      "step": 2431
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7708267284197963,
+      "learning_rate": 9.335522030908982e-06,
+      "loss": 0.513,
+      "step": 2432
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6124539344850676,
+      "learning_rate": 9.334888375013995e-06,
+      "loss": 0.5835,
+      "step": 2433
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.4186792900560805,
+      "learning_rate": 9.334254438657403e-06,
+      "loss": 0.5055,
+      "step": 2434
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.671986033143908,
+      "learning_rate": 9.333620221880219e-06,
+      "loss": 0.5439,
+      "step": 2435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5517674192339705,
+      "learning_rate": 9.332985724723476e-06,
+      "loss": 0.4861,
+      "step": 2436
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 5.227719009303436,
+      "learning_rate": 9.332350947228224e-06,
+      "loss": 0.5209,
+      "step": 2437
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6886748770591715,
+      "learning_rate": 9.331715889435534e-06,
+      "loss": 0.4749,
+      "step": 2438
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6185730144995171,
+      "learning_rate": 9.331080551386493e-06,
+      "loss": 0.5313,
+      "step": 2439
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7354136140177496,
+      "learning_rate": 9.330444933122205e-06,
+      "loss": 0.5454,
+      "step": 2440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6391524412056271,
+      "learning_rate": 9.329809034683795e-06,
+      "loss": 0.5735,
+      "step": 2441
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.190755305730543,
+      "learning_rate": 9.329172856112406e-06,
+      "loss": 0.4783,
+      "step": 2442
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.656855985917903,
+      "learning_rate": 9.328536397449193e-06,
+      "loss": 0.5249,
+      "step": 2443
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.550455312991144,
+      "learning_rate": 9.32789965873534e-06,
+      "loss": 0.492,
+      "step": 2444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5902494872542919,
+      "learning_rate": 9.327262640012041e-06,
+      "loss": 0.5591,
+      "step": 2445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7895144006906403,
+      "learning_rate": 9.326625341320508e-06,
+      "loss": 0.5025,
+      "step": 2446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.570714191295494,
+      "learning_rate": 9.325987762701977e-06,
+      "loss": 0.5225,
+      "step": 2447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5798050223001514,
+      "learning_rate": 9.325349904197696e-06,
+      "loss": 0.5525,
+      "step": 2448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.499626433734733,
+      "learning_rate": 9.324711765848935e-06,
+      "loss": 0.5539,
+      "step": 2449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5712541147592944,
+      "learning_rate": 9.32407334769698e-06,
+      "loss": 0.5359,
+      "step": 2450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.630495720904165,
+      "learning_rate": 9.323434649783136e-06,
+      "loss": 0.5351,
+      "step": 2451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.757508956740128,
+      "learning_rate": 9.322795672148726e-06,
+      "loss": 0.545,
+      "step": 2452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5130141942761337,
+      "learning_rate": 9.32215641483509e-06,
+      "loss": 0.5308,
+      "step": 2453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3705792829527919,
+      "learning_rate": 9.321516877883589e-06,
+      "loss": 0.5457,
+      "step": 2454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5307834166203997,
+      "learning_rate": 9.3208770613356e-06,
+      "loss": 0.5481,
+      "step": 2455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.468553393915982,
+      "learning_rate": 9.320236965232517e-06,
+      "loss": 0.4389,
+      "step": 2456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8949551194414926,
+      "learning_rate": 9.319596589615754e-06,
+      "loss": 0.5405,
+      "step": 2457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6320690158287077,
+      "learning_rate": 9.318955934526741e-06,
+      "loss": 0.5178,
+      "step": 2458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6657187442572023,
+      "learning_rate": 9.318315000006932e-06,
+      "loss": 0.4835,
+      "step": 2459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.540101786281009,
+      "learning_rate": 9.31767378609779e-06,
+      "loss": 0.4919,
+      "step": 2460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5558528654271906,
+      "learning_rate": 9.317032292840801e-06,
+      "loss": 0.4456,
+      "step": 2461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6525624945164196,
+      "learning_rate": 9.31639052027747e-06,
+      "loss": 0.5403,
+      "step": 2462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6007247902437685,
+      "learning_rate": 9.315748468449321e-06,
+      "loss": 0.493,
+      "step": 2463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7631675309675205,
+      "learning_rate": 9.31510613739789e-06,
+      "loss": 0.5036,
+      "step": 2464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.11609819270451,
+      "learning_rate": 9.314463527164738e-06,
+      "loss": 0.5287,
+      "step": 2465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4240439363926427,
+      "learning_rate": 9.313820637791439e-06,
+      "loss": 0.4841,
+      "step": 2466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.7016774256736709,
+      "learning_rate": 9.313177469319587e-06,
+      "loss": 0.5141,
+      "step": 2467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.210374064022001,
+      "learning_rate": 9.312534021790797e-06,
+      "loss": 0.509,
+      "step": 2468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6771910072252202,
+      "learning_rate": 9.311890295246697e-06,
+      "loss": 0.5219,
+      "step": 2469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.9800599048509315,
+      "learning_rate": 9.311246289728935e-06,
+      "loss": 0.5337,
+      "step": 2470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6303729007875951,
+      "learning_rate": 9.310602005279178e-06,
+      "loss": 0.5563,
+      "step": 2471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.626216340514345,
+      "learning_rate": 9.30995744193911e-06,
+      "loss": 0.5549,
+      "step": 2472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6889596614687363,
+      "learning_rate": 9.309312599750434e-06,
+      "loss": 0.477,
+      "step": 2473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5392073025154502,
+      "learning_rate": 9.30866747875487e-06,
+      "loss": 0.5628,
+      "step": 2474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.8983043942678033,
+      "learning_rate": 9.308022078994157e-06,
+      "loss": 0.547,
+      "step": 2475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6595414968974405,
+      "learning_rate": 9.30737640051005e-06,
+      "loss": 0.4459,
+      "step": 2476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.6606609265892773,
+      "learning_rate": 9.306730443344325e-06,
+      "loss": 0.4734,
+      "step": 2477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 6.23533866193931,
+      "learning_rate": 9.306084207538774e-06,
+      "loss": 0.4923,
+      "step": 2478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5969663660976549,
+      "learning_rate": 9.30543769313521e-06,
+      "loss": 0.5523,
+      "step": 2479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3910886054155382,
+      "learning_rate": 9.304790900175456e-06,
+      "loss": 0.4919,
+      "step": 2480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.9029888699624555,
+      "learning_rate": 9.304143828701364e-06,
+      "loss": 0.4977,
+      "step": 2481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5810444568903859,
+      "learning_rate": 9.303496478754796e-06,
+      "loss": 0.4797,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.654818180036754,
+      "learning_rate": 9.302848850377638e-06,
+      "loss": 0.5063,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6797601482034494,
+      "learning_rate": 9.302200943611785e-06,
+      "loss": 0.5028,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6007829122669868,
+      "learning_rate": 9.30155275849916e-06,
+      "loss": 0.4663,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 7.267745616077354,
+      "learning_rate": 9.3009042950817e-06,
+      "loss": 0.4726,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0530444658076292,
+      "learning_rate": 9.300255553401356e-06,
+      "loss": 0.5429,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6253270479975034,
+      "learning_rate": 9.299606533500105e-06,
+      "loss": 0.528,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6176674114601628,
+      "learning_rate": 9.298957235419937e-06,
+      "loss": 0.4399,
+      "step": 2489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.406172440790163,
+      "learning_rate": 9.298307659202857e-06,
+      "loss": 0.4955,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.614710522080688,
+      "learning_rate": 9.297657804890896e-06,
+      "loss": 0.5173,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6062436197129677,
+      "learning_rate": 9.297007672526096e-06,
+      "loss": 0.5325,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6055856823890806,
+      "learning_rate": 9.296357262150521e-06,
+      "loss": 0.5292,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6161825695671473,
+      "learning_rate": 9.29570657380625e-06,
+      "loss": 0.5499,
+      "step": 2494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4657040336651872,
+      "learning_rate": 9.295055607535386e-06,
+      "loss": 0.5325,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7486182267440138,
+      "learning_rate": 9.294404363380043e-06,
+      "loss": 0.4927,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9540755169643915,
+      "learning_rate": 9.293752841382353e-06,
+      "loss": 0.5486,
+      "step": 2497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.32659223365593,
+      "learning_rate": 9.293101041584473e-06,
+      "loss": 0.4909,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7841093403961563,
+      "learning_rate": 9.29244896402857e-06,
+      "loss": 0.4878,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7737783868126311,
+      "learning_rate": 9.291796608756835e-06,
+      "loss": 0.5513,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8393602016369437,
+      "learning_rate": 9.291143975811475e-06,
+      "loss": 0.4754,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 9.048277555393728,
+      "learning_rate": 9.290491065234712e-06,
+      "loss": 0.5592,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9606004419976215,
+      "learning_rate": 9.28983787706879e-06,
+      "loss": 0.4742,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8231670429851545,
+      "learning_rate": 9.28918441135597e-06,
+      "loss": 0.474,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4205568905836603,
+      "learning_rate": 9.28853066813853e-06,
+      "loss": 0.5187,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4942019717814887,
+      "learning_rate": 9.287876647458762e-06,
+      "loss": 0.4089,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5951477055645675,
+      "learning_rate": 9.287222349358988e-06,
+      "loss": 0.5443,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3752145727834633,
+      "learning_rate": 9.286567773881535e-06,
+      "loss": 0.4949,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5879767423925902,
+      "learning_rate": 9.285912921068755e-06,
+      "loss": 0.5701,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5933361261628092,
+      "learning_rate": 9.285257790963015e-06,
+      "loss": 0.5917,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5649841477380809,
+      "learning_rate": 9.284602383606702e-06,
+      "loss": 0.5433,
+      "step": 2511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8203588827743857,
+      "learning_rate": 9.28394669904222e-06,
+      "loss": 0.5236,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.340695603287489,
+      "learning_rate": 9.283290737311991e-06,
+      "loss": 0.5305,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.588381817738051,
+      "learning_rate": 9.282634498458453e-06,
+      "loss": 0.4739,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8698348215943588,
+      "learning_rate": 9.281977982524067e-06,
+      "loss": 0.4776,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.679640183032855,
+      "learning_rate": 9.281321189551306e-06,
+      "loss": 0.511,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6468875542388393,
+      "learning_rate": 9.280664119582665e-06,
+      "loss": 0.5393,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4475882392335921,
+      "learning_rate": 9.280006772660657e-06,
+      "loss": 0.4942,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9339275119658914,
+      "learning_rate": 9.279349148827807e-06,
+      "loss": 0.5213,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5335389226721083,
+      "learning_rate": 9.278691248126667e-06,
+      "loss": 0.5293,
+      "step": 2520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.092041907356189,
+      "learning_rate": 9.2780330705998e-06,
+      "loss": 0.4949,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3842815744965236,
+      "learning_rate": 9.27737461628979e-06,
+      "loss": 0.4974,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6121262746450775,
+      "learning_rate": 9.276715885239237e-06,
+      "loss": 0.5873,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.803224198328928,
+      "learning_rate": 9.276056877490762e-06,
+      "loss": 0.4998,
+      "step": 2524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4863087609035737,
+      "learning_rate": 9.275397593087e-06,
+      "loss": 0.4883,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6783808296900016,
+      "learning_rate": 9.274738032070607e-06,
+      "loss": 0.5123,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6223974418873989,
+      "learning_rate": 9.274078194484255e-06,
+      "loss": 0.5464,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9510064564816119,
+      "learning_rate": 9.273418080370636e-06,
+      "loss": 0.4932,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0600775317553364,
+      "learning_rate": 9.272757689772456e-06,
+      "loss": 0.5406,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7128511944441709,
+      "learning_rate": 9.272097022732444e-06,
+      "loss": 0.5235,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5460169837490706,
+      "learning_rate": 9.271436079293344e-06,
+      "loss": 0.5459,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6183378288200972,
+      "learning_rate": 9.270774859497915e-06,
+      "loss": 0.5433,
+      "step": 2532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5872751369969683,
+      "learning_rate": 9.27011336338894e-06,
+      "loss": 0.5626,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5679339505670329,
+      "learning_rate": 9.269451591009217e-06,
+      "loss": 0.4554,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.991732840440832,
+      "learning_rate": 9.268789542401561e-06,
+      "loss": 0.5408,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.73457195070717,
+      "learning_rate": 9.268127217608808e-06,
+      "loss": 0.4462,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8327039989671163,
+      "learning_rate": 9.267464616673805e-06,
+      "loss": 0.4865,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.985879379158774,
+      "learning_rate": 9.266801739639424e-06,
+      "loss": 0.5724,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7102053522381389,
+      "learning_rate": 9.266138586548554e-06,
+      "loss": 0.5787,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4799858646820447,
+      "learning_rate": 9.265475157444097e-06,
+      "loss": 0.5164,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3337443628689458,
+      "learning_rate": 9.264811452368975e-06,
+      "loss": 0.5302,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6258498233847551,
+      "learning_rate": 9.264147471366136e-06,
+      "loss": 0.4423,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.6790662259144122,
+      "learning_rate": 9.263483214478531e-06,
+      "loss": 0.542,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.865466279276663,
+      "learning_rate": 9.262818681749138e-06,
+      "loss": 0.5146,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.480442880007,
+      "learning_rate": 9.262153873220955e-06,
+      "loss": 0.4526,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5326639430014113,
+      "learning_rate": 9.26148878893699e-06,
+      "loss": 0.538,
+      "step": 2546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6816182543409122,
+      "learning_rate": 9.260823428940277e-06,
+      "loss": 0.5573,
+      "step": 2547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8268978002230043,
+      "learning_rate": 9.260157793273862e-06,
+      "loss": 0.5385,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6602251701839639,
+      "learning_rate": 9.259491881980809e-06,
+      "loss": 0.5556,
+      "step": 2549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5987614770527108,
+      "learning_rate": 9.258825695104205e-06,
+      "loss": 0.5598,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4379072168870373,
+      "learning_rate": 9.258159232687149e-06,
+      "loss": 0.4772,
+      "step": 2551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.57991378450244,
+      "learning_rate": 9.257492494772762e-06,
+      "loss": 0.524,
+      "step": 2552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7976902036337339,
+      "learning_rate": 9.256825481404178e-06,
+      "loss": 0.4985,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6121422433939019,
+      "learning_rate": 9.256158192624555e-06,
+      "loss": 0.4677,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4926517261373795,
+      "learning_rate": 9.255490628477067e-06,
+      "loss": 0.5255,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1801020978369503,
+      "learning_rate": 9.254822789004899e-06,
+      "loss": 0.5262,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.097957425772567,
+      "learning_rate": 9.254154674251263e-06,
+      "loss": 0.5258,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.068429361392017,
+      "learning_rate": 9.253486284259387e-06,
+      "loss": 0.4973,
+      "step": 2558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8888508398945307,
+      "learning_rate": 9.25281761907251e-06,
+      "loss": 0.5152,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4118273213892327,
+      "learning_rate": 9.2521486787339e-06,
+      "loss": 0.5072,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7943462190316661,
+      "learning_rate": 9.251479463286829e-06,
+      "loss": 0.5075,
+      "step": 2561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6498719599748597,
+      "learning_rate": 9.2508099727746e-06,
+      "loss": 0.5113,
+      "step": 2562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7305537559497464,
+      "learning_rate": 9.250140207240525e-06,
+      "loss": 0.4727,
+      "step": 2563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6276739073958881,
+      "learning_rate": 9.24947016672794e-06,
+      "loss": 0.5056,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5310894309638283,
+      "learning_rate": 9.248799851280195e-06,
+      "loss": 0.5033,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5276382086348612,
+      "learning_rate": 9.248129260940657e-06,
+      "loss": 0.5027,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4954389820212917,
+      "learning_rate": 9.247458395752713e-06,
+      "loss": 0.4618,
+      "step": 2567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.642816702718338,
+      "learning_rate": 9.246787255759768e-06,
+      "loss": 0.4734,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.398967441979739,
+      "learning_rate": 9.246115841005241e-06,
+      "loss": 0.5238,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5380909002930145,
+      "learning_rate": 9.245444151532575e-06,
+      "loss": 0.5081,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5555556811824625,
+      "learning_rate": 9.244772187385226e-06,
+      "loss": 0.4619,
+      "step": 2571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6978667120079489,
+      "learning_rate": 9.244099948606669e-06,
+      "loss": 0.555,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.0710441404831075,
+      "learning_rate": 9.243427435240398e-06,
+      "loss": 0.4772,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3824172940693416,
+      "learning_rate": 9.242754647329924e-06,
+      "loss": 0.5205,
+      "step": 2574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7750099356874,
+      "learning_rate": 9.242081584918772e-06,
+      "loss": 0.5172,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3630676982965495,
+      "learning_rate": 9.241408248050491e-06,
+      "loss": 0.4725,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.573708674396772,
+      "learning_rate": 9.240734636768647e-06,
+      "loss": 0.5406,
+      "step": 2577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6110943153048605,
+      "learning_rate": 9.240060751116819e-06,
+      "loss": 0.5264,
+      "step": 2578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4390467854313878,
+      "learning_rate": 9.239386591138605e-06,
+      "loss": 0.4962,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7378517985498094,
+      "learning_rate": 9.238712156877627e-06,
+      "loss": 0.4845,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8812462856411312,
+      "learning_rate": 9.238037448377517e-06,
+      "loss": 0.5486,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.660538384616219,
+      "learning_rate": 9.237362465681928e-06,
+      "loss": 0.4713,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5949164790812576,
+      "learning_rate": 9.23668720883453e-06,
+      "loss": 0.5433,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5859573460698153,
+      "learning_rate": 9.23601167787901e-06,
+      "loss": 0.5117,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8402757180454896,
+      "learning_rate": 9.235335872859079e-06,
+      "loss": 0.4873,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5518089395718657,
+      "learning_rate": 9.234659793818456e-06,
+      "loss": 0.5378,
+      "step": 2586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5813244685276074,
+      "learning_rate": 9.233983440800884e-06,
+      "loss": 0.5483,
+      "step": 2587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5221433601624796,
+      "learning_rate": 9.233306813850123e-06,
+      "loss": 0.4393,
+      "step": 2588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.93521833768943,
+      "learning_rate": 9.232629913009947e-06,
+      "loss": 0.5471,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6199135857665948,
+      "learning_rate": 9.231952738324155e-06,
+      "loss": 0.5729,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8326261686383545,
+      "learning_rate": 9.231275289836556e-06,
+      "loss": 0.4834,
+      "step": 2591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7260511513849952,
+      "learning_rate": 9.23059756759098e-06,
+      "loss": 0.6006,
+      "step": 2592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.4009975147116807,
+      "learning_rate": 9.229919571631277e-06,
+      "loss": 0.4439,
+      "step": 2593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4694791811679881,
+      "learning_rate": 9.22924130200131e-06,
+      "loss": 0.484,
+      "step": 2594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6585233955609227,
+      "learning_rate": 9.228562758744966e-06,
+      "loss": 0.5497,
+      "step": 2595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.1561782682471633,
+      "learning_rate": 9.22788394190614e-06,
+      "loss": 0.5229,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.065327285837434,
+      "learning_rate": 9.227204851528756e-06,
+      "loss": 0.5119,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 4.011982820206326,
+      "learning_rate": 9.226525487656745e-06,
+      "loss": 0.5296,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.220457601911713,
+      "learning_rate": 9.225845850334067e-06,
+      "loss": 0.4894,
+      "step": 2599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8276780151470424,
+      "learning_rate": 9.225165939604689e-06,
+      "loss": 0.4883,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8629313945801276,
+      "learning_rate": 9.224485755512603e-06,
+      "loss": 0.4745,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4712248178272749,
+      "learning_rate": 9.223805298101813e-06,
+      "loss": 0.5082,
+      "step": 2602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.917503838721372,
+      "learning_rate": 9.223124567416349e-06,
+      "loss": 0.5395,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6606170470467999,
+      "learning_rate": 9.222443563500248e-06,
+      "loss": 0.5668,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8118028218894473,
+      "learning_rate": 9.22176228639757e-06,
+      "loss": 0.4796,
+      "step": 2605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8756072147160745,
+      "learning_rate": 9.221080736152397e-06,
+      "loss": 0.4822,
+      "step": 2606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.012381992087522,
+      "learning_rate": 9.220398912808823e-06,
+      "loss": 0.5071,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4838835123750598,
+      "learning_rate": 9.21971681641096e-06,
+      "loss": 0.5107,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.8666858660166257,
+      "learning_rate": 9.219034447002938e-06,
+      "loss": 0.5052,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5885004897932643,
+      "learning_rate": 9.218351804628906e-06,
+      "loss": 0.5622,
+      "step": 2610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7403238918840045,
+      "learning_rate": 9.217668889333033e-06,
+      "loss": 0.495,
+      "step": 2611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.5392095736681655,
+      "learning_rate": 9.216985701159497e-06,
+      "loss": 0.5121,
+      "step": 2612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.31300080116189,
+      "learning_rate": 9.216302240152506e-06,
+      "loss": 0.5483,
+      "step": 2613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9237314492746174,
+      "learning_rate": 9.215618506356273e-06,
+      "loss": 0.4837,
+      "step": 2614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7398613529320874,
+      "learning_rate": 9.214934499815038e-06,
+      "loss": 0.5132,
+      "step": 2615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8622030621325105,
+      "learning_rate": 9.214250220573057e-06,
+      "loss": 0.5502,
+      "step": 2616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7818598423759586,
+      "learning_rate": 9.213565668674597e-06,
+      "loss": 0.5339,
+      "step": 2617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.030518708224634,
+      "learning_rate": 9.212880844163952e-06,
+      "loss": 0.5462,
+      "step": 2618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6570427018992091,
+      "learning_rate": 9.212195747085425e-06,
+      "loss": 0.5346,
+      "step": 2619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5854610627979908,
+      "learning_rate": 9.211510377483345e-06,
+      "loss": 0.5346,
+      "step": 2620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6011879553156652,
+      "learning_rate": 9.210824735402052e-06,
+      "loss": 0.5276,
+      "step": 2621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8500449147535236,
+      "learning_rate": 9.21013882088591e-06,
+      "loss": 0.5922,
+      "step": 2622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.321270718208347,
+      "learning_rate": 9.209452633979293e-06,
+      "loss": 0.5002,
+      "step": 2623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6083260414587445,
+      "learning_rate": 9.208766174726594e-06,
+      "loss": 0.5464,
+      "step": 2624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3858755828936427,
+      "learning_rate": 9.208079443172232e-06,
+      "loss": 0.5084,
+      "step": 2625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.775026930353286,
+      "learning_rate": 9.207392439360634e-06,
+      "loss": 0.5032,
+      "step": 2626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5694391932366076,
+      "learning_rate": 9.20670516333625e-06,
+      "loss": 0.5058,
+      "step": 2627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6264255863149715,
+      "learning_rate": 9.206017615143544e-06,
+      "loss": 0.5751,
+      "step": 2628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6552076236781883,
+      "learning_rate": 9.205329794827e-06,
+      "loss": 0.5209,
+      "step": 2629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.725858608673693,
+      "learning_rate": 9.20464170243112e-06,
+      "loss": 0.4889,
+      "step": 2630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8118171306412376,
+      "learning_rate": 9.203953338000424e-06,
+      "loss": 0.5131,
+      "step": 2631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.297217581535255,
+      "learning_rate": 9.203264701579444e-06,
+      "loss": 0.5254,
+      "step": 2632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6282178712020734,
+      "learning_rate": 9.202575793212739e-06,
+      "loss": 0.5685,
+      "step": 2633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6048953725096347,
+      "learning_rate": 9.201886612944875e-06,
+      "loss": 0.484,
+      "step": 2634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8829288553880361,
+      "learning_rate": 9.201197160820445e-06,
+      "loss": 0.572,
+      "step": 2635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8990023761283759,
+      "learning_rate": 9.200507436884055e-06,
+      "loss": 0.5204,
+      "step": 2636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.835438830770884,
+      "learning_rate": 9.199817441180329e-06,
+      "loss": 0.5297,
+      "step": 2637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.796807368142079,
+      "learning_rate": 9.199127173753908e-06,
+      "loss": 0.4793,
+      "step": 2638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8520063494133083,
+      "learning_rate": 9.198436634649453e-06,
+      "loss": 0.5273,
+      "step": 2639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7000947854569692,
+      "learning_rate": 9.197745823911638e-06,
+      "loss": 0.5507,
+      "step": 2640
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.278051545801207,
+      "learning_rate": 9.197054741585161e-06,
+      "loss": 0.4909,
+      "step": 2641
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.474755397950277,
+      "learning_rate": 9.196363387714733e-06,
+      "loss": 0.5151,
+      "step": 2642
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8360007140114516,
+      "learning_rate": 9.195671762345082e-06,
+      "loss": 0.4631,
+      "step": 2643
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5864162456388085,
+      "learning_rate": 9.194979865520956e-06,
+      "loss": 0.5235,
+      "step": 2644
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5853259569073647,
+      "learning_rate": 9.194287697287123e-06,
+      "loss": 0.5113,
+      "step": 2645
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7747919258325278,
+      "learning_rate": 9.193595257688362e-06,
+      "loss": 0.5335,
+      "step": 2646
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.3209317294533025,
+      "learning_rate": 9.192902546769473e-06,
+      "loss": 0.517,
+      "step": 2647
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9839506573733687,
+      "learning_rate": 9.192209564575274e-06,
+      "loss": 0.5106,
+      "step": 2648
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.4361810476651993,
+      "learning_rate": 9.191516311150601e-06,
+      "loss": 0.5082,
+      "step": 2649
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.0726254547801757,
+      "learning_rate": 9.190822786540306e-06,
+      "loss": 0.5065,
+      "step": 2650
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6077059152148249,
+      "learning_rate": 9.190128990789258e-06,
+      "loss": 0.5422,
+      "step": 2651
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8114339182981618,
+      "learning_rate": 9.189434923942346e-06,
+      "loss": 0.5075,
+      "step": 2652
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8784217987715583,
+      "learning_rate": 9.188740586044476e-06,
+      "loss": 0.4795,
+      "step": 2653
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.1459750111937637,
+      "learning_rate": 9.188045977140568e-06,
+      "loss": 0.5061,
+      "step": 2654
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6000653411872513,
+      "learning_rate": 9.187351097275565e-06,
+      "loss": 0.5509,
+      "step": 2655
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4373767981387768,
+      "learning_rate": 9.186655946494422e-06,
+      "loss": 0.4865,
+      "step": 2656
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.642961230293047,
+      "learning_rate": 9.18596052484212e-06,
+      "loss": 0.5375,
+      "step": 2657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.040502167874546,
+      "learning_rate": 9.185264832363644e-06,
+      "loss": 0.5377,
+      "step": 2658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.8655359507918545,
+      "learning_rate": 9.18456886910401e-06,
+      "loss": 0.4938,
+      "step": 2659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6731186915516718,
+      "learning_rate": 9.183872635108243e-06,
+      "loss": 0.4885,
+      "step": 2660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5839653290679983,
+      "learning_rate": 9.183176130421391e-06,
+      "loss": 0.5122,
+      "step": 2661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8613414786445373,
+      "learning_rate": 9.182479355088515e-06,
+      "loss": 0.4835,
+      "step": 2662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5878412410589576,
+      "learning_rate": 9.181782309154698e-06,
+      "loss": 0.5664,
+      "step": 2663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6270804420032785,
+      "learning_rate": 9.181084992665035e-06,
+      "loss": 0.5568,
+      "step": 2664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7128413011190065,
+      "learning_rate": 9.180387405664643e-06,
+      "loss": 0.5648,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7788814464471132,
+      "learning_rate": 9.179689548198654e-06,
+      "loss": 0.4359,
+      "step": 2666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 3.3713750878665585,
+      "learning_rate": 9.178991420312218e-06,
+      "loss": 0.5151,
+      "step": 2667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.834602217949142,
+      "learning_rate": 9.178293022050505e-06,
+      "loss": 0.5209,
+      "step": 2668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.9152390144370137,
+      "learning_rate": 9.177594353458699e-06,
+      "loss": 0.5273,
+      "step": 2669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5338652214961708,
+      "learning_rate": 9.176895414582002e-06,
+      "loss": 0.5365,
+      "step": 2670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 16.084834506187832,
+      "learning_rate": 9.176196205465637e-06,
+      "loss": 0.4901,
+      "step": 2671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7973729427597513,
+      "learning_rate": 9.175496726154842e-06,
+      "loss": 0.5234,
+      "step": 2672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.1752119354252497,
+      "learning_rate": 9.174796976694868e-06,
+      "loss": 0.5064,
+      "step": 2673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7698503409138169,
+      "learning_rate": 9.174096957130993e-06,
+      "loss": 0.4477,
+      "step": 2674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6802586053719029,
+      "learning_rate": 9.173396667508505e-06,
+      "loss": 0.5067,
+      "step": 2675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.053429485940372,
+      "learning_rate": 9.172696107872712e-06,
+      "loss": 0.5106,
+      "step": 2676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6502500640872055,
+      "learning_rate": 9.171995278268939e-06,
+      "loss": 0.5369,
+      "step": 2677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.543070132249157,
+      "learning_rate": 9.17129417874253e-06,
+      "loss": 0.4535,
+      "step": 2678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6957905248512428,
+      "learning_rate": 9.170592809338844e-06,
+      "loss": 0.5138,
+      "step": 2679
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5711582636892383,
+      "learning_rate": 9.16989117010326e-06,
+      "loss": 0.4802,
+      "step": 2680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4962519854554293,
+      "learning_rate": 9.16918926108117e-06,
+      "loss": 0.4666,
+      "step": 2681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7085743621251734,
+      "learning_rate": 9.168487082317989e-06,
+      "loss": 0.529,
+      "step": 2682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8192377952965582,
+      "learning_rate": 9.167784633859149e-06,
+      "loss": 0.5216,
+      "step": 2683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5419651943714792,
+      "learning_rate": 9.167081915750093e-06,
+      "loss": 0.4485,
+      "step": 2684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7550200479246778,
+      "learning_rate": 9.166378928036291e-06,
+      "loss": 0.4828,
+      "step": 2685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8834211718544254,
+      "learning_rate": 9.165675670763222e-06,
+      "loss": 0.559,
+      "step": 2686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6507865163112143,
+      "learning_rate": 9.164972143976383e-06,
+      "loss": 0.5897,
+      "step": 2687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7538085201763504,
+      "learning_rate": 9.1642683477213e-06,
+      "loss": 0.5441,
+      "step": 2688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.2846535308997873,
+      "learning_rate": 9.163564282043497e-06,
+      "loss": 0.5461,
+      "step": 2689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6724783933260177,
+      "learning_rate": 9.162859946988537e-06,
+      "loss": 0.4819,
+      "step": 2690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.554488112960826,
+      "learning_rate": 9.16215534260198e-06,
+      "loss": 0.5051,
+      "step": 2691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4943071276313928,
+      "learning_rate": 9.161450468929419e-06,
+      "loss": 0.5257,
+      "step": 2692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.003264144237427,
+      "learning_rate": 9.160745326016456e-06,
+      "loss": 0.5201,
+      "step": 2693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7296444154409523,
+      "learning_rate": 9.160039913908715e-06,
+      "loss": 0.5197,
+      "step": 2694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7439491132642566,
+      "learning_rate": 9.159334232651831e-06,
+      "loss": 0.5256,
+      "step": 2695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6920103158440429,
+      "learning_rate": 9.158628282291464e-06,
+      "loss": 0.5406,
+      "step": 2696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8889711448984638,
+      "learning_rate": 9.157922062873288e-06,
+      "loss": 0.5084,
+      "step": 2697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5719399611565419,
+      "learning_rate": 9.157215574442993e-06,
+      "loss": 0.5326,
+      "step": 2698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3754722344938513,
+      "learning_rate": 9.156508817046288e-06,
+      "loss": 0.4966,
+      "step": 2699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8140733422251627,
+      "learning_rate": 9.155801790728903e-06,
+      "loss": 0.4954,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5750676488263426,
+      "learning_rate": 9.155094495536575e-06,
+      "loss": 0.5267,
+      "step": 2701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.023837455811502,
+      "learning_rate": 9.154386931515072e-06,
+      "loss": 0.4711,
+      "step": 2702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6268180113303401,
+      "learning_rate": 9.153679098710166e-06,
+      "loss": 0.5516,
+      "step": 2703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8020853976196107,
+      "learning_rate": 9.152970997167657e-06,
+      "loss": 0.5012,
+      "step": 2704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6754609690048694,
+      "learning_rate": 9.152262626933358e-06,
+      "loss": 0.5154,
+      "step": 2705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8602159651867618,
+      "learning_rate": 9.1515539880531e-06,
+      "loss": 0.5412,
+      "step": 2706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6309716429159686,
+      "learning_rate": 9.150845080572727e-06,
+      "loss": 0.5477,
+      "step": 2707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7858605343605,
+      "learning_rate": 9.150135904538109e-06,
+      "loss": 0.5222,
+      "step": 2708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.8487718808929,
+      "learning_rate": 9.149426459995127e-06,
+      "loss": 0.5122,
+      "step": 2709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5288921335413317,
+      "learning_rate": 9.14871674698968e-06,
+      "loss": 0.5025,
+      "step": 2710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.581631766301348,
+      "learning_rate": 9.148006765567688e-06,
+      "loss": 0.4923,
+      "step": 2711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6107427639309868,
+      "learning_rate": 9.147296515775084e-06,
+      "loss": 0.539,
+      "step": 2712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4914724967806177,
+      "learning_rate": 9.146585997657822e-06,
+      "loss": 0.4528,
+      "step": 2713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5971958208802981,
+      "learning_rate": 9.145875211261867e-06,
+      "loss": 0.5436,
+      "step": 2714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4165825097750098,
+      "learning_rate": 9.145164156633212e-06,
+      "loss": 0.5173,
+      "step": 2715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3428359361705824,
+      "learning_rate": 9.14445283381786e-06,
+      "loss": 0.5029,
+      "step": 2716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.553750673348457,
+      "learning_rate": 9.14374124286183e-06,
+      "loss": 0.5324,
+      "step": 2717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6022061754528902,
+      "learning_rate": 9.14302938381116e-06,
+      "loss": 0.5303,
+      "step": 2718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 7.525054985575711,
+      "learning_rate": 9.14231725671191e-06,
+      "loss": 0.5363,
+      "step": 2719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6709165060391473,
+      "learning_rate": 9.141604861610154e-06,
+      "loss": 0.5282,
+      "step": 2720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7081744244554102,
+      "learning_rate": 9.14089219855198e-06,
+      "loss": 0.5171,
+      "step": 2721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5867454105420833,
+      "learning_rate": 9.140179267583497e-06,
+      "loss": 0.5073,
+      "step": 2722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.12732315255267,
+      "learning_rate": 9.139466068750833e-06,
+      "loss": 0.4696,
+      "step": 2723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4861968279556428,
+      "learning_rate": 9.138752602100128e-06,
+      "loss": 0.4583,
+      "step": 2724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6114100694847304,
+      "learning_rate": 9.138038867677546e-06,
+      "loss": 0.5207,
+      "step": 2725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6601654992669101,
+      "learning_rate": 9.13732486552926e-06,
+      "loss": 0.5555,
+      "step": 2726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.245988943235142,
+      "learning_rate": 9.136610595701469e-06,
+      "loss": 0.4761,
+      "step": 2727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6916528815085077,
+      "learning_rate": 9.135896058240384e-06,
+      "loss": 0.4832,
+      "step": 2728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.207989604914988,
+      "learning_rate": 9.135181253192234e-06,
+      "loss": 0.5656,
+      "step": 2729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.65048945075191,
+      "learning_rate": 9.134466180603265e-06,
+      "loss": 0.584,
+      "step": 2730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.527197700457935,
+      "learning_rate": 9.133750840519744e-06,
+      "loss": 0.5185,
+      "step": 2731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.7389115375135906,
+      "learning_rate": 9.133035232987952e-06,
+      "loss": 0.486,
+      "step": 2732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.57240820413734,
+      "learning_rate": 9.132319358054185e-06,
+      "loss": 0.4845,
+      "step": 2733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3285407603205157,
+      "learning_rate": 9.131603215764764e-06,
+      "loss": 0.5002,
+      "step": 2734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5468452930722145,
+      "learning_rate": 9.130886806166018e-06,
+      "loss": 0.5214,
+      "step": 2735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.66274584206512,
+      "learning_rate": 9.130170129304298e-06,
+      "loss": 0.5273,
+      "step": 2736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5466066573961323,
+      "learning_rate": 9.129453185225976e-06,
+      "loss": 0.5325,
+      "step": 2737
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9020722524579603,
+      "learning_rate": 9.128735973977433e-06,
+      "loss": 0.5198,
+      "step": 2738
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6679319670437078,
+      "learning_rate": 9.128018495605077e-06,
+      "loss": 0.5216,
+      "step": 2739
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4968362996665217,
+      "learning_rate": 9.12730075015532e-06,
+      "loss": 0.466,
+      "step": 2740
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.832538910354158,
+      "learning_rate": 9.126582737674608e-06,
+      "loss": 0.5265,
+      "step": 2741
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7385854139924435,
+      "learning_rate": 9.125864458209388e-06,
+      "loss": 0.557,
+      "step": 2742
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3899856196366254,
+      "learning_rate": 9.125145911806138e-06,
+      "loss": 0.4677,
+      "step": 2743
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6927475904140197,
+      "learning_rate": 9.12442709851134e-06,
+      "loss": 0.5095,
+      "step": 2744
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4477363947627442,
+      "learning_rate": 9.123708018371507e-06,
+      "loss": 0.5195,
+      "step": 2745
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4310694276339349,
+      "learning_rate": 9.122988671433159e-06,
+      "loss": 0.5118,
+      "step": 2746
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3569402909371706,
+      "learning_rate": 9.122269057742837e-06,
+      "loss": 0.5268,
+      "step": 2747
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2417188825959897,
+      "learning_rate": 9.1215491773471e-06,
+      "loss": 0.4802,
+      "step": 2748
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7207911456566134,
+      "learning_rate": 9.120829030292522e-06,
+      "loss": 0.5579,
+      "step": 2749
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.475787218501387,
+      "learning_rate": 9.120108616625697e-06,
+      "loss": 0.5156,
+      "step": 2750
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3375673133641435,
+      "learning_rate": 9.119387936393235e-06,
+      "loss": 0.5218,
+      "step": 2751
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7538318315433967,
+      "learning_rate": 9.11866698964176e-06,
+      "loss": 0.5254,
+      "step": 2752
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6137294144374963,
+      "learning_rate": 9.11794577641792e-06,
+      "loss": 0.4819,
+      "step": 2753
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.433009280461519,
+      "learning_rate": 9.117224296768376e-06,
+      "loss": 0.4931,
+      "step": 2754
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.45518271958093,
+      "learning_rate": 9.116502550739808e-06,
+      "loss": 0.5259,
+      "step": 2755
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5384352163901207,
+      "learning_rate": 9.115780538378907e-06,
+      "loss": 0.4537,
+      "step": 2756
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6472177529971626,
+      "learning_rate": 9.11505825973239e-06,
+      "loss": 0.4885,
+      "step": 2757
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5080377918369028,
+      "learning_rate": 9.114335714846987e-06,
+      "loss": 0.5816,
+      "step": 2758
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6143389665606054,
+      "learning_rate": 9.113612903769445e-06,
+      "loss": 0.5021,
+      "step": 2759
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8551384046735047,
+      "learning_rate": 9.112889826546529e-06,
+      "loss": 0.5173,
+      "step": 2760
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2720958246592944,
+      "learning_rate": 9.112166483225023e-06,
+      "loss": 0.4857,
+      "step": 2761
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5713366834254272,
+      "learning_rate": 9.111442873851724e-06,
+      "loss": 0.485,
+      "step": 2762
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.681878826862694,
+      "learning_rate": 9.11071899847345e-06,
+      "loss": 0.4714,
+      "step": 2763
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2072634415872,
+      "learning_rate": 9.109994857137032e-06,
+      "loss": 0.5744,
+      "step": 2764
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5866836267675368,
+      "learning_rate": 9.109270449889326e-06,
+      "loss": 0.5113,
+      "step": 2765
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6979777898958301,
+      "learning_rate": 9.108545776777196e-06,
+      "loss": 0.5246,
+      "step": 2766
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.306011405924336,
+      "learning_rate": 9.10782083784753e-06,
+      "loss": 0.5007,
+      "step": 2767
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7426384002770077,
+      "learning_rate": 9.107095633147229e-06,
+      "loss": 0.5251,
+      "step": 2768
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6225278803895051,
+      "learning_rate": 9.106370162723214e-06,
+      "loss": 0.5537,
+      "step": 2769
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6099372723394942,
+      "learning_rate": 9.10564442662242e-06,
+      "loss": 0.5433,
+      "step": 2770
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3911305350894354,
+      "learning_rate": 9.104918424891803e-06,
+      "loss": 0.5599,
+      "step": 2771
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.6749314632751924,
+      "learning_rate": 9.104192157578335e-06,
+      "loss": 0.5221,
+      "step": 2772
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6866236974367717,
+      "learning_rate": 9.103465624729002e-06,
+      "loss": 0.4573,
+      "step": 2773
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1723896324424787,
+      "learning_rate": 9.102738826390811e-06,
+      "loss": 0.4878,
+      "step": 2774
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9114811437032293,
+      "learning_rate": 9.102011762610785e-06,
+      "loss": 0.5717,
+      "step": 2775
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4113126482763392,
+      "learning_rate": 9.101284433435965e-06,
+      "loss": 0.5227,
+      "step": 2776
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7202827793717557,
+      "learning_rate": 9.100556838913407e-06,
+      "loss": 0.5451,
+      "step": 2777
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.312020549281303,
+      "learning_rate": 9.099828979090183e-06,
+      "loss": 0.504,
+      "step": 2778
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.382271834865188,
+      "learning_rate": 9.099100854013392e-06,
+      "loss": 0.4568,
+      "step": 2779
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6484539465120087,
+      "learning_rate": 9.098372463730135e-06,
+      "loss": 0.5143,
+      "step": 2780
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4628791695061754,
+      "learning_rate": 9.097643808287541e-06,
+      "loss": 0.4723,
+      "step": 2781
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8010748515146753,
+      "learning_rate": 9.096914887732756e-06,
+      "loss": 0.5093,
+      "step": 2782
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6832296187403024,
+      "learning_rate": 9.096185702112933e-06,
+      "loss": 0.5413,
+      "step": 2783
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5124258537433442,
+      "learning_rate": 9.095456251475257e-06,
+      "loss": 0.5428,
+      "step": 2784
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4823417602332192,
+      "learning_rate": 9.094726535866918e-06,
+      "loss": 0.4974,
+      "step": 2785
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6253719640746063,
+      "learning_rate": 9.093996555335128e-06,
+      "loss": 0.5395,
+      "step": 2786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6370834986373386,
+      "learning_rate": 9.093266309927116e-06,
+      "loss": 0.4977,
+      "step": 2787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.3536790357965214,
+      "learning_rate": 9.092535799690128e-06,
+      "loss": 0.4736,
+      "step": 2788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6419275994965008,
+      "learning_rate": 9.091805024671429e-06,
+      "loss": 0.5132,
+      "step": 2789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6530221631295237,
+      "learning_rate": 9.091073984918298e-06,
+      "loss": 0.5344,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3600676332693231,
+      "learning_rate": 9.090342680478031e-06,
+      "loss": 0.4836,
+      "step": 2791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6018024125547977,
+      "learning_rate": 9.089611111397943e-06,
+      "loss": 0.5492,
+      "step": 2792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5414973589670579,
+      "learning_rate": 9.088879277725367e-06,
+      "loss": 0.5501,
+      "step": 2793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.762059875585048,
+      "learning_rate": 9.088147179507651e-06,
+      "loss": 0.5045,
+      "step": 2794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3776620741200691,
+      "learning_rate": 9.087414816792159e-06,
+      "loss": 0.4508,
+      "step": 2795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5834146098261641,
+      "learning_rate": 9.086682189626277e-06,
+      "loss": 0.5243,
+      "step": 2796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5293276498184103,
+      "learning_rate": 9.085949298057402e-06,
+      "loss": 0.5135,
+      "step": 2797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5794505395275915,
+      "learning_rate": 9.085216142132953e-06,
+      "loss": 0.4728,
+      "step": 2798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6522462735443423,
+      "learning_rate": 9.084482721900363e-06,
+      "loss": 0.5408,
+      "step": 2799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.440492453833728,
+      "learning_rate": 9.083749037407086e-06,
+      "loss": 0.49,
+      "step": 2800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5893867109776132,
+      "learning_rate": 9.083015088700588e-06,
+      "loss": 0.5277,
+      "step": 2801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.666155076999855,
+      "learning_rate": 9.082280875828354e-06,
+      "loss": 0.534,
+      "step": 2802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6406019427495705,
+      "learning_rate": 9.081546398837888e-06,
+      "loss": 0.5693,
+      "step": 2803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6976584664007963,
+      "learning_rate": 9.08081165777671e-06,
+      "loss": 0.5563,
+      "step": 2804
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6429484806417651,
+      "learning_rate": 9.080076652692355e-06,
+      "loss": 0.5181,
+      "step": 2805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4863492027077427,
+      "learning_rate": 9.079341383632379e-06,
+      "loss": 0.5061,
+      "step": 2806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6626208028678764,
+      "learning_rate": 9.07860585064435e-06,
+      "loss": 0.5028,
+      "step": 2807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4515077155567888,
+      "learning_rate": 9.07787005377586e-06,
+      "loss": 0.5271,
+      "step": 2808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8112282684826926,
+      "learning_rate": 9.07713399307451e-06,
+      "loss": 0.5738,
+      "step": 2809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5821108254790223,
+      "learning_rate": 9.076397668587927e-06,
+      "loss": 0.5444,
+      "step": 2810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4284033581599325,
+      "learning_rate": 9.075661080363745e-06,
+      "loss": 0.4605,
+      "step": 2811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6832464233956461,
+      "learning_rate": 9.074924228449625e-06,
+      "loss": 0.4385,
+      "step": 2812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9701456139963462,
+      "learning_rate": 9.074187112893235e-06,
+      "loss": 0.5065,
+      "step": 2813
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7096055600463889,
+      "learning_rate": 9.073449733742271e-06,
+      "loss": 0.5171,
+      "step": 2814
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4879151210266681,
+      "learning_rate": 9.072712091044437e-06,
+      "loss": 0.3914,
+      "step": 2815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7243821409432285,
+      "learning_rate": 9.071974184847459e-06,
+      "loss": 0.5233,
+      "step": 2816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7309908316396845,
+      "learning_rate": 9.071236015199077e-06,
+      "loss": 0.5418,
+      "step": 2817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3725950230404538,
+      "learning_rate": 9.070497582147051e-06,
+      "loss": 0.5371,
+      "step": 2818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6086772680201136,
+      "learning_rate": 9.069758885739157e-06,
+      "loss": 0.4876,
+      "step": 2819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5945002385263806,
+      "learning_rate": 9.069019926023189e-06,
+      "loss": 0.5384,
+      "step": 2820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4179480905817499,
+      "learning_rate": 9.068280703046951e-06,
+      "loss": 0.4793,
+      "step": 2821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6994833367032034,
+      "learning_rate": 9.067541216858276e-06,
+      "loss": 0.507,
+      "step": 2822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4678100262733464,
+      "learning_rate": 9.066801467505004e-06,
+      "loss": 0.4944,
+      "step": 2823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7350869892905907,
+      "learning_rate": 9.066061455034996e-06,
+      "loss": 0.5141,
+      "step": 2824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5420637972086488,
+      "learning_rate": 9.065321179496134e-06,
+      "loss": 0.4947,
+      "step": 2825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9724487521108094,
+      "learning_rate": 9.064580640936307e-06,
+      "loss": 0.4753,
+      "step": 2826
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7642551680786225,
+      "learning_rate": 9.063839839403431e-06,
+      "loss": 0.5185,
+      "step": 2827
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5441391355314873,
+      "learning_rate": 9.063098774945433e-06,
+      "loss": 0.4697,
+      "step": 2828
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.681699786924045,
+      "learning_rate": 9.06235744761026e-06,
+      "loss": 0.493,
+      "step": 2829
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3830097143050537,
+      "learning_rate": 9.061615857445875e-06,
+      "loss": 0.5076,
+      "step": 2830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5945140648278653,
+      "learning_rate": 9.060874004500256e-06,
+      "loss": 0.4948,
+      "step": 2831
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4693440368477615,
+      "learning_rate": 9.060131888821402e-06,
+      "loss": 0.4809,
+      "step": 2832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.4632975431155053,
+      "learning_rate": 9.059389510457326e-06,
+      "loss": 0.4838,
+      "step": 2833
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5982332725747161,
+      "learning_rate": 9.058646869456058e-06,
+      "loss": 0.5429,
+      "step": 2834
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3919668128167277,
+      "learning_rate": 9.057903965865649e-06,
+      "loss": 0.5183,
+      "step": 2835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5115741831961291,
+      "learning_rate": 9.057160799734159e-06,
+      "loss": 0.5021,
+      "step": 2836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6264061339354404,
+      "learning_rate": 9.056417371109674e-06,
+      "loss": 0.5642,
+      "step": 2837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1285779893030767,
+      "learning_rate": 9.055673680040293e-06,
+      "loss": 0.5562,
+      "step": 2838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3036097615341762,
+      "learning_rate": 9.054929726574128e-06,
+      "loss": 0.5069,
+      "step": 2839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.089050739402883,
+      "learning_rate": 9.054185510759317e-06,
+      "loss": 0.493,
+      "step": 2840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4317091534646331,
+      "learning_rate": 9.053441032644005e-06,
+      "loss": 0.5251,
+      "step": 2841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7529296572257094,
+      "learning_rate": 9.052696292276362e-06,
+      "loss": 0.4284,
+      "step": 2842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3521139688492267,
+      "learning_rate": 9.051951289704568e-06,
+      "loss": 0.4772,
+      "step": 2843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7215488872856497,
+      "learning_rate": 9.051206024976829e-06,
+      "loss": 0.5369,
+      "step": 2844
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5208064307109832,
+      "learning_rate": 9.050460498141358e-06,
+      "loss": 0.5162,
+      "step": 2845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.59466320606524,
+      "learning_rate": 9.049714709246392e-06,
+      "loss": 0.5307,
+      "step": 2846
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3903674029947377,
+      "learning_rate": 9.048968658340183e-06,
+      "loss": 0.461,
+      "step": 2847
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.704528050067276,
+      "learning_rate": 9.048222345470996e-06,
+      "loss": 0.4746,
+      "step": 2848
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2776992690118854,
+      "learning_rate": 9.047475770687123e-06,
+      "loss": 0.5108,
+      "step": 2849
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6687377615283843,
+      "learning_rate": 9.046728934036857e-06,
+      "loss": 0.5632,
+      "step": 2850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.7149521824080989,
+      "learning_rate": 9.045981835568527e-06,
+      "loss": 0.5414,
+      "step": 2851
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.2744692611388375,
+      "learning_rate": 9.045234475330464e-06,
+      "loss": 0.5032,
+      "step": 2852
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.597800986924832,
+      "learning_rate": 9.044486853371022e-06,
+      "loss": 0.5342,
+      "step": 2853
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3459891160538193,
+      "learning_rate": 9.043738969738572e-06,
+      "loss": 0.5375,
+      "step": 2854
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.545951589508904,
+      "learning_rate": 9.042990824481499e-06,
+      "loss": 0.4352,
+      "step": 2855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4034211739231588,
+      "learning_rate": 9.04224241764821e-06,
+      "loss": 0.5653,
+      "step": 2856
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5685069722431875,
+      "learning_rate": 9.041493749287125e-06,
+      "loss": 0.5095,
+      "step": 2857
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.11569744710073,
+      "learning_rate": 9.04074481944668e-06,
+      "loss": 0.476,
+      "step": 2858
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.93506020171291,
+      "learning_rate": 9.03999562817533e-06,
+      "loss": 0.5252,
+      "step": 2859
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.220709694908683,
+      "learning_rate": 9.03924617552155e-06,
+      "loss": 0.4888,
+      "step": 2860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6238570606132527,
+      "learning_rate": 9.038496461533825e-06,
+      "loss": 0.5296,
+      "step": 2861
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.641906368429085,
+      "learning_rate": 9.037746486260664e-06,
+      "loss": 0.5381,
+      "step": 2862
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.8105920847981534,
+      "learning_rate": 9.036996249750588e-06,
+      "loss": 0.5279,
+      "step": 2863
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6864143595600605,
+      "learning_rate": 9.036245752052132e-06,
+      "loss": 0.4535,
+      "step": 2864
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6539356115962756,
+      "learning_rate": 9.035494993213858e-06,
+      "loss": 0.5294,
+      "step": 2865
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4343781648388971,
+      "learning_rate": 9.034743973284337e-06,
+      "loss": 0.5181,
+      "step": 2866
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4776321455317365,
+      "learning_rate": 9.03399269231216e-06,
+      "loss": 0.4911,
+      "step": 2867
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4289199939080706,
+      "learning_rate": 9.03324115034593e-06,
+      "loss": 0.5033,
+      "step": 2868
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7724262912160629,
+      "learning_rate": 9.032489347434277e-06,
+      "loss": 0.5386,
+      "step": 2869
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5657577330742882,
+      "learning_rate": 9.031737283625836e-06,
+      "loss": 0.5464,
+      "step": 2870
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7436895582300712,
+      "learning_rate": 9.030984958969268e-06,
+      "loss": 0.5449,
+      "step": 2871
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.124191405709207,
+      "learning_rate": 9.030232373513245e-06,
+      "loss": 0.5112,
+      "step": 2872
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6282859595040777,
+      "learning_rate": 9.029479527306461e-06,
+      "loss": 0.5483,
+      "step": 2873
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.454302204484329,
+      "learning_rate": 9.028726420397624e-06,
+      "loss": 0.5421,
+      "step": 2874
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5097872185938759,
+      "learning_rate": 9.027973052835456e-06,
+      "loss": 0.4727,
+      "step": 2875
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5660091707479298,
+      "learning_rate": 9.0272194246687e-06,
+      "loss": 0.4165,
+      "step": 2876
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.541005863309527,
+      "learning_rate": 9.026465535946118e-06,
+      "loss": 0.5534,
+      "step": 2877
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.142136977845894,
+      "learning_rate": 9.025711386716481e-06,
+      "loss": 0.4883,
+      "step": 2878
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6658377872974515,
+      "learning_rate": 9.024956977028585e-06,
+      "loss": 0.4715,
+      "step": 2879
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4720771029130917,
+      "learning_rate": 9.024202306931236e-06,
+      "loss": 0.5362,
+      "step": 2880
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5792433885212707,
+      "learning_rate": 9.023447376473264e-06,
+      "loss": 0.4849,
+      "step": 2881
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3899423807451092,
+      "learning_rate": 9.022692185703509e-06,
+      "loss": 0.5003,
+      "step": 2882
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7607924823257483,
+      "learning_rate": 9.021936734670833e-06,
+      "loss": 0.5295,
+      "step": 2883
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4556657198388738,
+      "learning_rate": 9.02118102342411e-06,
+      "loss": 0.5483,
+      "step": 2884
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4909160131099273,
+      "learning_rate": 9.020425052012237e-06,
+      "loss": 0.5419,
+      "step": 2885
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0453172855147144,
+      "learning_rate": 9.019668820484123e-06,
+      "loss": 0.508,
+      "step": 2886
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6576818076751387,
+      "learning_rate": 9.018912328888691e-06,
+      "loss": 0.5205,
+      "step": 2887
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.632333043865491,
+      "learning_rate": 9.018155577274891e-06,
+      "loss": 0.52,
+      "step": 2888
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4295342656778467,
+      "learning_rate": 9.017398565691681e-06,
+      "loss": 0.4993,
+      "step": 2889
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7491171432443455,
+      "learning_rate": 9.01664129418804e-06,
+      "loss": 0.5264,
+      "step": 2890
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5224450032887107,
+      "learning_rate": 9.015883762812962e-06,
+      "loss": 0.5241,
+      "step": 2891
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5918272408452991,
+      "learning_rate": 9.015125971615459e-06,
+      "loss": 0.5332,
+      "step": 2892
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1659958412664304,
+      "learning_rate": 9.014367920644555e-06,
+      "loss": 0.5321,
+      "step": 2893
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7440380719766047,
+      "learning_rate": 9.0136096099493e-06,
+      "loss": 0.5025,
+      "step": 2894
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6324977814773578,
+      "learning_rate": 9.012851039578754e-06,
+      "loss": 0.5203,
+      "step": 2895
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3312526164378193,
+      "learning_rate": 9.012092209581993e-06,
+      "loss": 0.4961,
+      "step": 2896
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8789883484291892,
+      "learning_rate": 9.011333120008117e-06,
+      "loss": 0.4962,
+      "step": 2897
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9555911357403164,
+      "learning_rate": 9.010573770906235e-06,
+      "loss": 0.504,
+      "step": 2898
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.507846986930788,
+      "learning_rate": 9.009814162325475e-06,
+      "loss": 0.493,
+      "step": 2899
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6539060493772393,
+      "learning_rate": 9.009054294314985e-06,
+      "loss": 0.4869,
+      "step": 2900
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5556406957322377,
+      "learning_rate": 9.008294166923927e-06,
+      "loss": 0.4958,
+      "step": 2901
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5758554372564502,
+      "learning_rate": 9.00753378020148e-06,
+      "loss": 0.4741,
+      "step": 2902
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9010188327482946,
+      "learning_rate": 9.00677313419684e-06,
+      "loss": 0.5748,
+      "step": 2903
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8777674113446061,
+      "learning_rate": 9.00601222895922e-06,
+      "loss": 0.5207,
+      "step": 2904
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7647453387347591,
+      "learning_rate": 9.005251064537848e-06,
+      "loss": 0.5057,
+      "step": 2905
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6119672434298182,
+      "learning_rate": 9.004489640981973e-06,
+      "loss": 0.4968,
+      "step": 2906
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.502966750421292,
+      "learning_rate": 9.003727958340856e-06,
+      "loss": 0.5159,
+      "step": 2907
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.408235278989794,
+      "learning_rate": 9.00296601666378e-06,
+      "loss": 0.4842,
+      "step": 2908
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4346957404124154,
+      "learning_rate": 9.002203816000035e-06,
+      "loss": 0.4848,
+      "step": 2909
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2620684598175245,
+      "learning_rate": 9.001441356398942e-06,
+      "loss": 0.5583,
+      "step": 2910
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2756562163083214,
+      "learning_rate": 9.000678637909825e-06,
+      "loss": 0.4583,
+      "step": 2911
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.4008095359476687,
+      "learning_rate": 8.999915660582037e-06,
+      "loss": 0.5168,
+      "step": 2912
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9789433094176034,
+      "learning_rate": 8.999152424464936e-06,
+      "loss": 0.536,
+      "step": 2913
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8715275234546433,
+      "learning_rate": 8.998388929607905e-06,
+      "loss": 0.518,
+      "step": 2914
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.367149063709105,
+      "learning_rate": 8.99762517606034e-06,
+      "loss": 0.4899,
+      "step": 2915
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9146217653856598,
+      "learning_rate": 8.996861163871658e-06,
+      "loss": 0.5517,
+      "step": 2916
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6458253638372093,
+      "learning_rate": 8.996096893091285e-06,
+      "loss": 0.5124,
+      "step": 2917
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7997800559279415,
+      "learning_rate": 8.995332363768671e-06,
+      "loss": 0.4993,
+      "step": 2918
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5517983308219325,
+      "learning_rate": 8.994567575953281e-06,
+      "loss": 0.4857,
+      "step": 2919
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7227828254481616,
+      "learning_rate": 8.993802529694594e-06,
+      "loss": 0.4875,
+      "step": 2920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9496139081332147,
+      "learning_rate": 8.993037225042107e-06,
+      "loss": 0.4711,
+      "step": 2921
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7083956237898862,
+      "learning_rate": 8.992271662045335e-06,
+      "loss": 0.55,
+      "step": 2922
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.062223356295321,
+      "learning_rate": 8.99150584075381e-06,
+      "loss": 0.4845,
+      "step": 2923
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.215008892440812,
+      "learning_rate": 8.990739761217078e-06,
+      "loss": 0.5112,
+      "step": 2924
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.496851018519653,
+      "learning_rate": 8.989973423484703e-06,
+      "loss": 0.5623,
+      "step": 2925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4126782867075052,
+      "learning_rate": 8.989206827606269e-06,
+      "loss": 0.4927,
+      "step": 2926
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7585448414127276,
+      "learning_rate": 8.988439973631371e-06,
+      "loss": 0.4981,
+      "step": 2927
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4429615325228062,
+      "learning_rate": 8.987672861609624e-06,
+      "loss": 0.4725,
+      "step": 2928
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6524667605260932,
+      "learning_rate": 8.986905491590659e-06,
+      "loss": 0.4503,
+      "step": 2929
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.026708868261442,
+      "learning_rate": 8.986137863624125e-06,
+      "loss": 0.5129,
+      "step": 2930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.566714723275509,
+      "learning_rate": 8.985369977759686e-06,
+      "loss": 0.5309,
+      "step": 2931
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.339343421436997,
+      "learning_rate": 8.984601834047022e-06,
+      "loss": 0.4528,
+      "step": 2932
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.319215389446425,
+      "learning_rate": 8.983833432535833e-06,
+      "loss": 0.4408,
+      "step": 2933
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4016226891541588,
+      "learning_rate": 8.98306477327583e-06,
+      "loss": 0.4836,
+      "step": 2934
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 4.458482075773622,
+      "learning_rate": 8.98229585631675e-06,
+      "loss": 0.4834,
+      "step": 2935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.432574781078674,
+      "learning_rate": 8.981526681708336e-06,
+      "loss": 0.4929,
+      "step": 2936
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4965938792972264,
+      "learning_rate": 8.980757249500354e-06,
+      "loss": 0.5196,
+      "step": 2937
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1744069059084414,
+      "learning_rate": 8.979987559742587e-06,
+      "loss": 0.5396,
+      "step": 2938
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4983700945529155,
+      "learning_rate": 8.97921761248483e-06,
+      "loss": 0.5163,
+      "step": 2939
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.0021631525644827,
+      "learning_rate": 8.978447407776898e-06,
+      "loss": 0.4277,
+      "step": 2940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6299889987173506,
+      "learning_rate": 8.977676945668626e-06,
+      "loss": 0.5363,
+      "step": 2941
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 8.714687486946245,
+      "learning_rate": 8.976906226209856e-06,
+      "loss": 0.5304,
+      "step": 2942
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9899206843886754,
+      "learning_rate": 8.976135249450457e-06,
+      "loss": 0.5399,
+      "step": 2943
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3626477637139076,
+      "learning_rate": 8.975364015440308e-06,
+      "loss": 0.5037,
+      "step": 2944
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7064942261108158,
+      "learning_rate": 8.974592524229308e-06,
+      "loss": 0.5659,
+      "step": 2945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 3.182204965088374,
+      "learning_rate": 8.973820775867372e-06,
+      "loss": 0.487,
+      "step": 2946
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6143727409805179,
+      "learning_rate": 8.97304877040443e-06,
+      "loss": 0.5637,
+      "step": 2947
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.3788805637516175,
+      "learning_rate": 8.97227650789043e-06,
+      "loss": 0.4783,
+      "step": 2948
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5494322091186193,
+      "learning_rate": 8.971503988375335e-06,
+      "loss": 0.5201,
+      "step": 2949
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.444057326039935,
+      "learning_rate": 8.970731211909129e-06,
+      "loss": 0.4852,
+      "step": 2950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.911288284899553,
+      "learning_rate": 8.969958178541807e-06,
+      "loss": 0.5258,
+      "step": 2951
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5100572561690182,
+      "learning_rate": 8.969184888323383e-06,
+      "loss": 0.4999,
+      "step": 2952
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3126743767315752,
+      "learning_rate": 8.968411341303892e-06,
+      "loss": 0.526,
+      "step": 2953
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.525898642698643,
+      "learning_rate": 8.967637537533376e-06,
+      "loss": 0.495,
+      "step": 2954
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4456612731284988,
+      "learning_rate": 8.966863477061903e-06,
+      "loss": 0.496,
+      "step": 2955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.596646481271798,
+      "learning_rate": 8.966089159939552e-06,
+      "loss": 0.531,
+      "step": 2956
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4436116533246504,
+      "learning_rate": 8.965314586216421e-06,
+      "loss": 0.4832,
+      "step": 2957
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.692619191648268,
+      "learning_rate": 8.964539755942623e-06,
+      "loss": 0.4841,
+      "step": 2958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.254709168430868,
+      "learning_rate": 8.963764669168289e-06,
+      "loss": 0.5293,
+      "step": 2959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9243015500359633,
+      "learning_rate": 8.962989325943566e-06,
+      "loss": 0.4826,
+      "step": 2960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7831459984966213,
+      "learning_rate": 8.962213726318619e-06,
+      "loss": 0.5548,
+      "step": 2961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6708816241947577,
+      "learning_rate": 8.961437870343626e-06,
+      "loss": 0.5295,
+      "step": 2962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7050027458258565,
+      "learning_rate": 8.960661758068784e-06,
+      "loss": 0.5147,
+      "step": 2963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9510175835763244,
+      "learning_rate": 8.959885389544309e-06,
+      "loss": 0.4719,
+      "step": 2964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3616403640203567,
+      "learning_rate": 8.95910876482043e-06,
+      "loss": 0.4837,
+      "step": 2965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1973680633989634,
+      "learning_rate": 8.958331883947394e-06,
+      "loss": 0.4772,
+      "step": 2966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.90835177558252,
+      "learning_rate": 8.95755474697546e-06,
+      "loss": 0.4957,
+      "step": 2967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1293848871189254,
+      "learning_rate": 8.956777353954913e-06,
+      "loss": 0.5165,
+      "step": 2968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6288944396527829,
+      "learning_rate": 8.955999704936048e-06,
+      "loss": 0.5527,
+      "step": 2969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.7808352663163705,
+      "learning_rate": 8.955221799969175e-06,
+      "loss": 0.5215,
+      "step": 2970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7316313795696343,
+      "learning_rate": 8.954443639104627e-06,
+      "loss": 0.4577,
+      "step": 2971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.78196437812799,
+      "learning_rate": 8.953665222392749e-06,
+      "loss": 0.5178,
+      "step": 2972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.1785705724047957,
+      "learning_rate": 8.952886549883903e-06,
+      "loss": 0.4842,
+      "step": 2973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6199279242639198,
+      "learning_rate": 8.952107621628467e-06,
+      "loss": 0.5553,
+      "step": 2974
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4784759421501155,
+      "learning_rate": 8.951328437676838e-06,
+      "loss": 0.5147,
+      "step": 2975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.6102532278792783,
+      "learning_rate": 8.95054899807943e-06,
+      "loss": 0.5012,
+      "step": 2976
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5174904290569684,
+      "learning_rate": 8.949769302886668e-06,
+      "loss": 0.5092,
+      "step": 2977
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7058770722273573,
+      "learning_rate": 8.948989352149e-06,
+      "loss": 0.4989,
+      "step": 2978
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6347562946065516,
+      "learning_rate": 8.948209145916887e-06,
+      "loss": 0.5355,
+      "step": 2979
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.163904549459621,
+      "learning_rate": 8.947428684240806e-06,
+      "loss": 0.5093,
+      "step": 2980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7219955383657253,
+      "learning_rate": 8.946647967171254e-06,
+      "loss": 0.5675,
+      "step": 2981
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6072410792437541,
+      "learning_rate": 8.945866994758741e-06,
+      "loss": 0.5143,
+      "step": 2982
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4144966452187229,
+      "learning_rate": 8.945085767053795e-06,
+      "loss": 0.5257,
+      "step": 2983
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2775274421513896,
+      "learning_rate": 8.944304284106962e-06,
+      "loss": 0.544,
+      "step": 2984
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5597746369937876,
+      "learning_rate": 8.9435225459688e-06,
+      "loss": 0.543,
+      "step": 2985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8733101885218788,
+      "learning_rate": 8.942740552689889e-06,
+      "loss": 0.5263,
+      "step": 2986
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 15.492842926976435,
+      "learning_rate": 8.941958304320822e-06,
+      "loss": 0.497,
+      "step": 2987
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6270224114243583,
+      "learning_rate": 8.941175800912208e-06,
+      "loss": 0.5283,
+      "step": 2988
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.463116484662335,
+      "learning_rate": 8.940393042514677e-06,
+      "loss": 0.4839,
+      "step": 2989
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.076617583798195,
+      "learning_rate": 8.93961002917887e-06,
+      "loss": 0.4493,
+      "step": 2990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9022759452459503,
+      "learning_rate": 8.938826760955448e-06,
+      "loss": 0.5055,
+      "step": 2991
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4830167986237655,
+      "learning_rate": 8.938043237895088e-06,
+      "loss": 0.4793,
+      "step": 2992
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2882167507977618,
+      "learning_rate": 8.93725946004848e-06,
+      "loss": 0.5053,
+      "step": 2993
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0489027367616477,
+      "learning_rate": 8.936475427466337e-06,
+      "loss": 0.475,
+      "step": 2994
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7106351183014727,
+      "learning_rate": 8.935691140199384e-06,
+      "loss": 0.4982,
+      "step": 2995
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5194274829594947,
+      "learning_rate": 8.934906598298362e-06,
+      "loss": 0.519,
+      "step": 2996
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.3120612924456894,
+      "learning_rate": 8.934121801814031e-06,
+      "loss": 0.4814,
+      "step": 2997
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5880876238398923,
+      "learning_rate": 8.933336750797167e-06,
+      "loss": 0.5339,
+      "step": 2998
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.4282950857027745,
+      "learning_rate": 8.932551445298557e-06,
+      "loss": 0.4914,
+      "step": 2999
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3462086628833854,
+      "learning_rate": 8.931765885369015e-06,
+      "loss": 0.4673,
+      "step": 3000
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.586298512310245,
+      "learning_rate": 8.930980071059364e-06,
+      "loss": 0.5219,
+      "step": 3001
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7248005822900319,
+      "learning_rate": 8.930194002420444e-06,
+      "loss": 0.5072,
+      "step": 3002
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8035215228246972,
+      "learning_rate": 8.929407679503116e-06,
+      "loss": 0.5165,
+      "step": 3003
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4122584097906916,
+      "learning_rate": 8.928621102358248e-06,
+      "loss": 0.4661,
+      "step": 3004
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6308846017127305,
+      "learning_rate": 8.927834271036736e-06,
+      "loss": 0.5113,
+      "step": 3005
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.473294475759395,
+      "learning_rate": 8.927047185589484e-06,
+      "loss": 0.5185,
+      "step": 3006
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0268388645750672,
+      "learning_rate": 8.926259846067417e-06,
+      "loss": 0.5465,
+      "step": 3007
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8867958940032006,
+      "learning_rate": 8.925472252521473e-06,
+      "loss": 0.5201,
+      "step": 3008
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6498501779556012,
+      "learning_rate": 8.924684405002611e-06,
+      "loss": 0.5606,
+      "step": 3009
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7589893741212537,
+      "learning_rate": 8.9238963035618e-06,
+      "loss": 0.559,
+      "step": 3010
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5745117020075853,
+      "learning_rate": 8.923107948250034e-06,
+      "loss": 0.5285,
+      "step": 3011
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5677130676196426,
+      "learning_rate": 8.922319339118314e-06,
+      "loss": 0.5534,
+      "step": 3012
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6395081549772318,
+      "learning_rate": 8.921530476217664e-06,
+      "loss": 0.5548,
+      "step": 3013
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5797526429901023,
+      "learning_rate": 8.920741359599121e-06,
+      "loss": 0.4914,
+      "step": 3014
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.573935855623021,
+      "learning_rate": 8.919951989313744e-06,
+      "loss": 0.4837,
+      "step": 3015
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.7279511052330405,
+      "learning_rate": 8.919162365412599e-06,
+      "loss": 0.5043,
+      "step": 3016
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5770046852717214,
+      "learning_rate": 8.918372487946778e-06,
+      "loss": 0.526,
+      "step": 3017
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.631091561852185,
+      "learning_rate": 8.91758235696738e-06,
+      "loss": 0.558,
+      "step": 3018
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3805439043420007,
+      "learning_rate": 8.91679197252553e-06,
+      "loss": 0.5068,
+      "step": 3019
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4903873138654185,
+      "learning_rate": 8.916001334672364e-06,
+      "loss": 0.5263,
+      "step": 3020
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3237904314031874,
+      "learning_rate": 8.915210443459032e-06,
+      "loss": 0.4857,
+      "step": 3021
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5360733782311171,
+      "learning_rate": 8.914419298936707e-06,
+      "loss": 0.409,
+      "step": 3022
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9466591946127414,
+      "learning_rate": 8.913627901156575e-06,
+      "loss": 0.5492,
+      "step": 3023
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7015465361947841,
+      "learning_rate": 8.912836250169836e-06,
+      "loss": 0.5398,
+      "step": 3024
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.027560627948112,
+      "learning_rate": 8.912044346027713e-06,
+      "loss": 0.5216,
+      "step": 3025
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.563701008560731,
+      "learning_rate": 8.911252188781436e-06,
+      "loss": 0.4956,
+      "step": 3026
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3813243487670344,
+      "learning_rate": 8.910459778482259e-06,
+      "loss": 0.443,
+      "step": 3027
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8919595423101965,
+      "learning_rate": 8.909667115181451e-06,
+      "loss": 0.4951,
+      "step": 3028
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3593557255606197,
+      "learning_rate": 8.908874198930295e-06,
+      "loss": 0.5014,
+      "step": 3029
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4126988804293257,
+      "learning_rate": 8.90808102978009e-06,
+      "loss": 0.4787,
+      "step": 3030
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.294230115784155,
+      "learning_rate": 8.907287607782155e-06,
+      "loss": 0.4903,
+      "step": 3031
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.18623492178922,
+      "learning_rate": 8.906493932987826e-06,
+      "loss": 0.4897,
+      "step": 3032
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.611104192218659,
+      "learning_rate": 8.905700005448448e-06,
+      "loss": 0.5437,
+      "step": 3033
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7810415694281176,
+      "learning_rate": 8.904905825215388e-06,
+      "loss": 0.5327,
+      "step": 3034
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.558021654524866,
+      "learning_rate": 8.904111392340032e-06,
+      "loss": 0.4348,
+      "step": 3035
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6221079372879845,
+      "learning_rate": 8.903316706873774e-06,
+      "loss": 0.5292,
+      "step": 3036
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4770482458714866,
+      "learning_rate": 8.902521768868031e-06,
+      "loss": 0.5241,
+      "step": 3037
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.943158151224759,
+      "learning_rate": 8.901726578374236e-06,
+      "loss": 0.4686,
+      "step": 3038
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5914701802293288,
+      "learning_rate": 8.900931135443836e-06,
+      "loss": 0.5206,
+      "step": 3039
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7305636577727892,
+      "learning_rate": 8.900135440128293e-06,
+      "loss": 0.4806,
+      "step": 3040
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1038927682171797,
+      "learning_rate": 8.899339492479089e-06,
+      "loss": 0.564,
+      "step": 3041
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5551081292089663,
+      "learning_rate": 8.898543292547722e-06,
+      "loss": 0.566,
+      "step": 3042
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5618309861926594,
+      "learning_rate": 8.897746840385702e-06,
+      "loss": 0.521,
+      "step": 3043
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.717537312785067,
+      "learning_rate": 8.896950136044562e-06,
+      "loss": 0.5539,
+      "step": 3044
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.469427370536025,
+      "learning_rate": 8.896153179575846e-06,
+      "loss": 0.5036,
+      "step": 3045
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7534648726901858,
+      "learning_rate": 8.895355971031115e-06,
+      "loss": 0.5236,
+      "step": 3046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9419696136579172,
+      "learning_rate": 8.89455851046195e-06,
+      "loss": 0.4723,
+      "step": 3047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4934485368523365,
+      "learning_rate": 8.893760797919944e-06,
+      "loss": 0.478,
+      "step": 3048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1092954186614357,
+      "learning_rate": 8.892962833456707e-06,
+      "loss": 0.5428,
+      "step": 3049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4721703253080265,
+      "learning_rate": 8.892164617123868e-06,
+      "loss": 0.4657,
+      "step": 3050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6062260899251004,
+      "learning_rate": 8.891366148973068e-06,
+      "loss": 0.4943,
+      "step": 3051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6322899350853253,
+      "learning_rate": 8.890567429055971e-06,
+      "loss": 0.5577,
+      "step": 3052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4255870652846119,
+      "learning_rate": 8.889768457424251e-06,
+      "loss": 0.5044,
+      "step": 3053
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9267448471780224,
+      "learning_rate": 8.8889692341296e-06,
+      "loss": 0.4607,
+      "step": 3054
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.1795041577052054,
+      "learning_rate": 8.888169759223724e-06,
+      "loss": 0.4737,
+      "step": 3055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.27235228077261,
+      "learning_rate": 8.887370032758354e-06,
+      "loss": 0.4158,
+      "step": 3056
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.293772124217995,
+      "learning_rate": 8.886570054785229e-06,
+      "loss": 0.4545,
+      "step": 3057
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7006109566993444,
+      "learning_rate": 8.885769825356103e-06,
+      "loss": 0.5323,
+      "step": 3058
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.552584860623139,
+      "learning_rate": 8.884969344522754e-06,
+      "loss": 0.5008,
+      "step": 3059
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6032096740093893,
+      "learning_rate": 8.88416861233697e-06,
+      "loss": 0.5167,
+      "step": 3060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4262951209898767,
+      "learning_rate": 8.883367628850559e-06,
+      "loss": 0.4652,
+      "step": 3061
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5277024225253921,
+      "learning_rate": 8.882566394115342e-06,
+      "loss": 0.4611,
+      "step": 3062
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8187640992830203,
+      "learning_rate": 8.881764908183158e-06,
+      "loss": 0.4959,
+      "step": 3063
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7174432185726323,
+      "learning_rate": 8.88096317110586e-06,
+      "loss": 0.4993,
+      "step": 3064
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5487884835373051,
+      "learning_rate": 8.880161182935325e-06,
+      "loss": 0.4942,
+      "step": 3065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3239023171077364,
+      "learning_rate": 8.879358943723437e-06,
+      "loss": 0.5076,
+      "step": 3066
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5179132662349666,
+      "learning_rate": 8.8785564535221e-06,
+      "loss": 0.4897,
+      "step": 3067
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7359474439774842,
+      "learning_rate": 8.877753712383233e-06,
+      "loss": 0.5694,
+      "step": 3068
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.512267998587302,
+      "learning_rate": 8.876950720358775e-06,
+      "loss": 0.5669,
+      "step": 3069
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5229756802237184,
+      "learning_rate": 8.876147477500677e-06,
+      "loss": 0.4775,
+      "step": 3070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5655739320285087,
+      "learning_rate": 8.875343983860909e-06,
+      "loss": 0.492,
+      "step": 3071
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4279273924498868,
+      "learning_rate": 8.874540239491451e-06,
+      "loss": 0.5207,
+      "step": 3072
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4554563971275536,
+      "learning_rate": 8.873736244444311e-06,
+      "loss": 0.5252,
+      "step": 3073
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.1976722123173635,
+      "learning_rate": 8.872931998771503e-06,
+      "loss": 0.5435,
+      "step": 3074
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 4.829944928659648,
+      "learning_rate": 8.87212750252506e-06,
+      "loss": 0.4896,
+      "step": 3075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4530745990655476,
+      "learning_rate": 8.871322755757035e-06,
+      "loss": 0.5021,
+      "step": 3076
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6407994643413658,
+      "learning_rate": 8.87051775851949e-06,
+      "loss": 0.5332,
+      "step": 3077
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.752968774268347,
+      "learning_rate": 8.86971251086451e-06,
+      "loss": 0.5502,
+      "step": 3078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.9166724487872593,
+      "learning_rate": 8.868907012844194e-06,
+      "loss": 0.5259,
+      "step": 3079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3852884581115712,
+      "learning_rate": 8.868101264510654e-06,
+      "loss": 0.5006,
+      "step": 3080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5369636900006798,
+      "learning_rate": 8.867295265916023e-06,
+      "loss": 0.5076,
+      "step": 3081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.051539483656026,
+      "learning_rate": 8.866489017112448e-06,
+      "loss": 0.5207,
+      "step": 3082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5526077143090808,
+      "learning_rate": 8.86568251815209e-06,
+      "loss": 0.5364,
+      "step": 3083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5285930571826203,
+      "learning_rate": 8.864875769087131e-06,
+      "loss": 0.4797,
+      "step": 3084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.416139148730127,
+      "learning_rate": 8.864068769969766e-06,
+      "loss": 0.5016,
+      "step": 3085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5366863759112341,
+      "learning_rate": 8.863261520852205e-06,
+      "loss": 0.5016,
+      "step": 3086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6097807415807062,
+      "learning_rate": 8.862454021786678e-06,
+      "loss": 0.5412,
+      "step": 3087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3168312456560078,
+      "learning_rate": 8.861646272825429e-06,
+      "loss": 0.4689,
+      "step": 3088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4699890118151353,
+      "learning_rate": 8.860838274020717e-06,
+      "loss": 0.5165,
+      "step": 3089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.514821563746873,
+      "learning_rate": 8.860030025424819e-06,
+      "loss": 0.4826,
+      "step": 3090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5031885803853393,
+      "learning_rate": 8.85922152709003e-06,
+      "loss": 0.4759,
+      "step": 3091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4441384732414912,
+      "learning_rate": 8.858412779068654e-06,
+      "loss": 0.4427,
+      "step": 3092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.774647452200091,
+      "learning_rate": 8.857603781413021e-06,
+      "loss": 0.5301,
+      "step": 3093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5542537272717182,
+      "learning_rate": 8.856794534175468e-06,
+      "loss": 0.5343,
+      "step": 3094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4553466995449236,
+      "learning_rate": 8.855985037408355e-06,
+      "loss": 0.4763,
+      "step": 3095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7004284650344321,
+      "learning_rate": 8.855175291164055e-06,
+      "loss": 0.4687,
+      "step": 3096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.512420646989729,
+      "learning_rate": 8.854365295494956e-06,
+      "loss": 0.5527,
+      "step": 3097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.448911627850262,
+      "learning_rate": 8.853555050453465e-06,
+      "loss": 0.4444,
+      "step": 3098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4076264983860287,
+      "learning_rate": 8.852744556092002e-06,
+      "loss": 0.509,
+      "step": 3099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.268669756593638,
+      "learning_rate": 8.851933812463008e-06,
+      "loss": 0.4879,
+      "step": 3100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.349123948896521,
+      "learning_rate": 8.851122819618933e-06,
+      "loss": 0.502,
+      "step": 3101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6626940076930226,
+      "learning_rate": 8.85031157761225e-06,
+      "loss": 0.5764,
+      "step": 3102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3723081163137278,
+      "learning_rate": 8.849500086495446e-06,
+      "loss": 0.4865,
+      "step": 3103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5138858463693807,
+      "learning_rate": 8.84868834632102e-06,
+      "loss": 0.5187,
+      "step": 3104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.0433399833446817,
+      "learning_rate": 8.847876357141496e-06,
+      "loss": 0.5008,
+      "step": 3105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.8331221720906892,
+      "learning_rate": 8.847064119009405e-06,
+      "loss": 0.5745,
+      "step": 3106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7511896801048878,
+      "learning_rate": 8.846251631977295e-06,
+      "loss": 0.5266,
+      "step": 3107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4466839052672444,
+      "learning_rate": 8.845438896097738e-06,
+      "loss": 0.4793,
+      "step": 3108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.2681686985736493,
+      "learning_rate": 8.844625911423315e-06,
+      "loss": 0.5385,
+      "step": 3109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7535888439970788,
+      "learning_rate": 8.843812678006624e-06,
+      "loss": 0.4582,
+      "step": 3110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6580119491188257,
+      "learning_rate": 8.842999195900283e-06,
+      "loss": 0.5354,
+      "step": 3111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.339098954009609,
+      "learning_rate": 8.842185465156919e-06,
+      "loss": 0.4935,
+      "step": 3112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7020675605517868,
+      "learning_rate": 8.841371485829183e-06,
+      "loss": 0.5185,
+      "step": 3113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6753045389856148,
+      "learning_rate": 8.840557257969736e-06,
+      "loss": 0.5386,
+      "step": 3114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.749086134653893,
+      "learning_rate": 8.83974278163126e-06,
+      "loss": 0.5324,
+      "step": 3115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.365155516834661,
+      "learning_rate": 8.838928056866447e-06,
+      "loss": 0.5456,
+      "step": 3116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4510773906155148,
+      "learning_rate": 8.838113083728012e-06,
+      "loss": 0.563,
+      "step": 3117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.290808827295803,
+      "learning_rate": 8.837297862268682e-06,
+      "loss": 0.4781,
+      "step": 3118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.626076260297186,
+      "learning_rate": 8.836482392541199e-06,
+      "loss": 0.4993,
+      "step": 3119
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8673669963616522,
+      "learning_rate": 8.835666674598325e-06,
+      "loss": 0.4477,
+      "step": 3120
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7910083470834832,
+      "learning_rate": 8.834850708492834e-06,
+      "loss": 0.4685,
+      "step": 3121
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8399016600547733,
+      "learning_rate": 8.83403449427752e-06,
+      "loss": 0.549,
+      "step": 3122
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6847530645609963,
+      "learning_rate": 8.833218032005187e-06,
+      "loss": 0.4842,
+      "step": 3123
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8826394963854247,
+      "learning_rate": 8.832401321728663e-06,
+      "loss": 0.4647,
+      "step": 3124
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7610924767354525,
+      "learning_rate": 8.831584363500787e-06,
+      "loss": 0.5497,
+      "step": 3125
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0676635122026816,
+      "learning_rate": 8.830767157374415e-06,
+      "loss": 0.4936,
+      "step": 3126
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9387568015925745,
+      "learning_rate": 8.829949703402421e-06,
+      "loss": 0.4648,
+      "step": 3127
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7376162664313684,
+      "learning_rate": 8.82913200163769e-06,
+      "loss": 0.4809,
+      "step": 3128
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6642056388818773,
+      "learning_rate": 8.828314052133126e-06,
+      "loss": 0.4788,
+      "step": 3129
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5203849793825879,
+      "learning_rate": 8.827495854941654e-06,
+      "loss": 0.5479,
+      "step": 3130
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8140951860702474,
+      "learning_rate": 8.826677410116206e-06,
+      "loss": 0.5058,
+      "step": 3131
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7610510568990287,
+      "learning_rate": 8.825858717709734e-06,
+      "loss": 0.4938,
+      "step": 3132
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3969329516371154,
+      "learning_rate": 8.82503977777521e-06,
+      "loss": 0.4832,
+      "step": 3133
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7659137507976672,
+      "learning_rate": 8.824220590365616e-06,
+      "loss": 0.5557,
+      "step": 3134
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4229605554275584,
+      "learning_rate": 8.82340115553395e-06,
+      "loss": 0.5249,
+      "step": 3135
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.533844320104774,
+      "learning_rate": 8.822581473333233e-06,
+      "loss": 0.4495,
+      "step": 3136
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.902563259383878,
+      "learning_rate": 8.821761543816493e-06,
+      "loss": 0.5091,
+      "step": 3137
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5093162421573296,
+      "learning_rate": 8.820941367036784e-06,
+      "loss": 0.465,
+      "step": 3138
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9304759091326718,
+      "learning_rate": 8.820120943047166e-06,
+      "loss": 0.5007,
+      "step": 3139
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2361609085672767,
+      "learning_rate": 8.819300271900719e-06,
+      "loss": 0.4687,
+      "step": 3140
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8083382332595386,
+      "learning_rate": 8.818479353650543e-06,
+      "loss": 0.5002,
+      "step": 3141
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0860660948885523,
+      "learning_rate": 8.817658188349745e-06,
+      "loss": 0.5135,
+      "step": 3142
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7132751180378571,
+      "learning_rate": 8.816836776051458e-06,
+      "loss": 0.4742,
+      "step": 3143
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8744501727453178,
+      "learning_rate": 8.816015116808824e-06,
+      "loss": 0.5578,
+      "step": 3144
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5570388588450488,
+      "learning_rate": 8.815193210675004e-06,
+      "loss": 0.521,
+      "step": 3145
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7695743333176919,
+      "learning_rate": 8.814371057703175e-06,
+      "loss": 0.497,
+      "step": 3146
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7321807668042545,
+      "learning_rate": 8.813548657946527e-06,
+      "loss": 0.5667,
+      "step": 3147
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5875410576121596,
+      "learning_rate": 8.812726011458271e-06,
+      "loss": 0.5387,
+      "step": 3148
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4771052746617401,
+      "learning_rate": 8.81190311829163e-06,
+      "loss": 0.4709,
+      "step": 3149
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8513356688037073,
+      "learning_rate": 8.811079978499842e-06,
+      "loss": 0.4647,
+      "step": 3150
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6843832918379211,
+      "learning_rate": 8.810256592136167e-06,
+      "loss": 0.5568,
+      "step": 3151
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.8728502062768015,
+      "learning_rate": 8.809432959253872e-06,
+      "loss": 0.5063,
+      "step": 3152
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3963022169318688,
+      "learning_rate": 8.80860907990625e-06,
+      "loss": 0.4604,
+      "step": 3153
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7736925305679498,
+      "learning_rate": 8.807784954146603e-06,
+      "loss": 0.4754,
+      "step": 3154
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0135890232706406,
+      "learning_rate": 8.80696058202825e-06,
+      "loss": 0.5439,
+      "step": 3155
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9332997922609096,
+      "learning_rate": 8.806135963604528e-06,
+      "loss": 0.4876,
+      "step": 3156
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7667936677673874,
+      "learning_rate": 8.805311098928786e-06,
+      "loss": 0.5294,
+      "step": 3157
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0882298186371524,
+      "learning_rate": 8.804485988054396e-06,
+      "loss": 0.5645,
+      "step": 3158
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5148688458297657,
+      "learning_rate": 8.80366063103474e-06,
+      "loss": 0.5245,
+      "step": 3159
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9781505062163962,
+      "learning_rate": 8.802835027923216e-06,
+      "loss": 0.5607,
+      "step": 3160
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.9892915221923826,
+      "learning_rate": 8.80200917877324e-06,
+      "loss": 0.473,
+      "step": 3161
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 7.830992108623623,
+      "learning_rate": 8.801183083638246e-06,
+      "loss": 0.5277,
+      "step": 3162
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5575430448215832,
+      "learning_rate": 8.800356742571677e-06,
+      "loss": 0.4697,
+      "step": 3163
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5257775414500692,
+      "learning_rate": 8.799530155626998e-06,
+      "loss": 0.5312,
+      "step": 3164
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5054808946340248,
+      "learning_rate": 8.79870332285769e-06,
+      "loss": 0.4893,
+      "step": 3165
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7468684242112201,
+      "learning_rate": 8.797876244317245e-06,
+      "loss": 0.5553,
+      "step": 3166
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5333496316874167,
+      "learning_rate": 8.797048920059176e-06,
+      "loss": 0.5021,
+      "step": 3167
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5846806974005915,
+      "learning_rate": 8.796221350137008e-06,
+      "loss": 0.5398,
+      "step": 3168
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5575635790752691,
+      "learning_rate": 8.795393534604287e-06,
+      "loss": 0.5495,
+      "step": 3169
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7613725761019852,
+      "learning_rate": 8.794565473514567e-06,
+      "loss": 0.4925,
+      "step": 3170
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.804247965213342,
+      "learning_rate": 8.793737166921425e-06,
+      "loss": 0.503,
+      "step": 3171
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6884671644973206,
+      "learning_rate": 8.792908614878452e-06,
+      "loss": 0.5569,
+      "step": 3172
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8276788543739844,
+      "learning_rate": 8.792079817439254e-06,
+      "loss": 0.539,
+      "step": 3173
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.508319866248653,
+      "learning_rate": 8.791250774657451e-06,
+      "loss": 0.5033,
+      "step": 3174
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5903042871739241,
+      "learning_rate": 8.790421486586683e-06,
+      "loss": 0.5093,
+      "step": 3175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0328027879885724,
+      "learning_rate": 8.789591953280603e-06,
+      "loss": 0.5077,
+      "step": 3176
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.411172637050618,
+      "learning_rate": 8.788762174792881e-06,
+      "loss": 0.4961,
+      "step": 3177
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6755275725927786,
+      "learning_rate": 8.787932151177202e-06,
+      "loss": 0.5175,
+      "step": 3178
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.267851735282927,
+      "learning_rate": 8.78710188248727e-06,
+      "loss": 0.5038,
+      "step": 3179
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6633436385046281,
+      "learning_rate": 8.7862713687768e-06,
+      "loss": 0.5494,
+      "step": 3180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.580064818721464,
+      "learning_rate": 8.785440610099524e-06,
+      "loss": 0.5117,
+      "step": 3181
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6871990608212786,
+      "learning_rate": 8.784609606509194e-06,
+      "loss": 0.5067,
+      "step": 3182
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5325275845605506,
+      "learning_rate": 8.783778358059572e-06,
+      "loss": 0.5218,
+      "step": 3183
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.500977311959191,
+      "learning_rate": 8.78294686480444e-06,
+      "loss": 0.5051,
+      "step": 3184
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8644052704952736,
+      "learning_rate": 8.782115126797596e-06,
+      "loss": 0.5185,
+      "step": 3185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7437842477055276,
+      "learning_rate": 8.78128314409285e-06,
+      "loss": 0.4834,
+      "step": 3186
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7712264902130028,
+      "learning_rate": 8.780450916744031e-06,
+      "loss": 0.4954,
+      "step": 3187
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.9195142499829734,
+      "learning_rate": 8.779618444804982e-06,
+      "loss": 0.4514,
+      "step": 3188
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7447346492184025,
+      "learning_rate": 8.778785728329566e-06,
+      "loss": 0.5433,
+      "step": 3189
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6649538228897592,
+      "learning_rate": 8.777952767371657e-06,
+      "loss": 0.5583,
+      "step": 3190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.3016235933557017,
+      "learning_rate": 8.777119561985145e-06,
+      "loss": 0.4531,
+      "step": 3191
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7996035092874412,
+      "learning_rate": 8.77628611222394e-06,
+      "loss": 0.5191,
+      "step": 3192
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4926075695635426,
+      "learning_rate": 8.775452418141961e-06,
+      "loss": 0.4726,
+      "step": 3193
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6793602704633813,
+      "learning_rate": 8.774618479793151e-06,
+      "loss": 0.5407,
+      "step": 3194
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0835045075478518,
+      "learning_rate": 8.773784297231463e-06,
+      "loss": 0.4688,
+      "step": 3195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.424885529519649,
+      "learning_rate": 8.772949870510867e-06,
+      "loss": 0.5112,
+      "step": 3196
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.361354891643077,
+      "learning_rate": 8.772115199685352e-06,
+      "loss": 0.4576,
+      "step": 3197
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.724132644475796,
+      "learning_rate": 8.771280284808917e-06,
+      "loss": 0.5238,
+      "step": 3198
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7404022362769855,
+      "learning_rate": 8.770445125935578e-06,
+      "loss": 0.5397,
+      "step": 3199
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6447434530001412,
+      "learning_rate": 8.769609723119375e-06,
+      "loss": 0.548,
+      "step": 3200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4599295704522952,
+      "learning_rate": 8.768774076414354e-06,
+      "loss": 0.4834,
+      "step": 3201
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6013134169360055,
+      "learning_rate": 8.76793818587458e-06,
+      "loss": 0.4621,
+      "step": 3202
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8481091119131194,
+      "learning_rate": 8.767102051554135e-06,
+      "loss": 0.5225,
+      "step": 3203
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6465609485685071,
+      "learning_rate": 8.766265673507115e-06,
+      "loss": 0.519,
+      "step": 3204
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.413333271809522,
+      "learning_rate": 8.765429051787632e-06,
+      "loss": 0.4871,
+      "step": 3205
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.77399736349863,
+      "learning_rate": 8.764592186449816e-06,
+      "loss": 0.4449,
+      "step": 3206
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.645696450137369,
+      "learning_rate": 8.76375507754781e-06,
+      "loss": 0.5292,
+      "step": 3207
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.076529834975819,
+      "learning_rate": 8.762917725135774e-06,
+      "loss": 0.5211,
+      "step": 3208
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.698390638238923,
+      "learning_rate": 8.762080129267884e-06,
+      "loss": 0.5324,
+      "step": 3209
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5026611251242934,
+      "learning_rate": 8.761242289998331e-06,
+      "loss": 0.4814,
+      "step": 3210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7292116106622832,
+      "learning_rate": 8.76040420738132e-06,
+      "loss": 0.4996,
+      "step": 3211
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6149842253112265,
+      "learning_rate": 8.75956588147108e-06,
+      "loss": 0.4268,
+      "step": 3212
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.358671104247102,
+      "learning_rate": 8.758727312321843e-06,
+      "loss": 0.5247,
+      "step": 3213
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6200749324635986,
+      "learning_rate": 8.757888499987867e-06,
+      "loss": 0.5221,
+      "step": 3214
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7562451300752853,
+      "learning_rate": 8.75704944452342e-06,
+      "loss": 0.5353,
+      "step": 3215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.523703295661214,
+      "learning_rate": 8.75621014598279e-06,
+      "loss": 0.4522,
+      "step": 3216
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.6403191588425283,
+      "learning_rate": 8.755370604420275e-06,
+      "loss": 0.5001,
+      "step": 3217
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.8222004723604748,
+      "learning_rate": 8.754530819890198e-06,
+      "loss": 0.4893,
+      "step": 3218
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5640569929709862,
+      "learning_rate": 8.753690792446885e-06,
+      "loss": 0.4924,
+      "step": 3219
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.386318405315565,
+      "learning_rate": 8.752850522144689e-06,
+      "loss": 0.4846,
+      "step": 3220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7216518695006389,
+      "learning_rate": 8.752010009037975e-06,
+      "loss": 0.4654,
+      "step": 3221
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5999725224290333,
+      "learning_rate": 8.751169253181121e-06,
+      "loss": 0.4966,
+      "step": 3222
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.8484397076851256,
+      "learning_rate": 8.750328254628524e-06,
+      "loss": 0.4631,
+      "step": 3223
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6382728476558481,
+      "learning_rate": 8.749487013434594e-06,
+      "loss": 0.5595,
+      "step": 3224
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6363968018135961,
+      "learning_rate": 8.74864552965376e-06,
+      "loss": 0.4936,
+      "step": 3225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6285678400003176,
+      "learning_rate": 8.747803803340463e-06,
+      "loss": 0.5568,
+      "step": 3226
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.251318348222844,
+      "learning_rate": 8.746961834549163e-06,
+      "loss": 0.5065,
+      "step": 3227
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.644789333543007,
+      "learning_rate": 8.746119623334335e-06,
+      "loss": 0.5445,
+      "step": 3228
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3160445098555669,
+      "learning_rate": 8.745277169750467e-06,
+      "loss": 0.443,
+      "step": 3229
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5362129657004773,
+      "learning_rate": 8.744434473852066e-06,
+      "loss": 0.5371,
+      "step": 3230
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.669278323870098,
+      "learning_rate": 8.743591535693652e-06,
+      "loss": 0.5061,
+      "step": 3231
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6953628617773973,
+      "learning_rate": 8.742748355329764e-06,
+      "loss": 0.4661,
+      "step": 3232
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3445470792409033,
+      "learning_rate": 8.741904932814953e-06,
+      "loss": 0.4895,
+      "step": 3233
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.003000671299978,
+      "learning_rate": 8.741061268203787e-06,
+      "loss": 0.5377,
+      "step": 3234
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4669265619273744,
+      "learning_rate": 8.740217361550853e-06,
+      "loss": 0.4834,
+      "step": 3235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4673369310699333,
+      "learning_rate": 8.739373212910746e-06,
+      "loss": 0.5194,
+      "step": 3236
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 3.0255072270781818,
+      "learning_rate": 8.738528822338086e-06,
+      "loss": 0.5364,
+      "step": 3237
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6034722141597749,
+      "learning_rate": 8.737684189887501e-06,
+      "loss": 0.5431,
+      "step": 3238
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.7187732665533946,
+      "learning_rate": 8.736839315613638e-06,
+      "loss": 0.5219,
+      "step": 3239
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.194592268510745,
+      "learning_rate": 8.73599419957116e-06,
+      "loss": 0.4634,
+      "step": 3240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3821733391194253,
+      "learning_rate": 8.735148841814745e-06,
+      "loss": 0.5176,
+      "step": 3241
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.067743053267743,
+      "learning_rate": 8.734303242399086e-06,
+      "loss": 0.4699,
+      "step": 3242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.653625425643597,
+      "learning_rate": 8.733457401378893e-06,
+      "loss": 0.4582,
+      "step": 3243
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.850761426068509,
+      "learning_rate": 8.732611318808888e-06,
+      "loss": 0.4891,
+      "step": 3244
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5457580884012472,
+      "learning_rate": 8.731764994743814e-06,
+      "loss": 0.5367,
+      "step": 3245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6879385865555376,
+      "learning_rate": 8.730918429238429e-06,
+      "loss": 0.4985,
+      "step": 3246
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.629580502344068,
+      "learning_rate": 8.7300716223475e-06,
+      "loss": 0.5087,
+      "step": 3247
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5559450640212475,
+      "learning_rate": 8.729224574125818e-06,
+      "loss": 0.4897,
+      "step": 3248
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5759081461595328,
+      "learning_rate": 8.728377284628183e-06,
+      "loss": 0.4913,
+      "step": 3249
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4414065206820996,
+      "learning_rate": 8.727529753909417e-06,
+      "loss": 0.4893,
+      "step": 3250
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5000103084689527,
+      "learning_rate": 8.72668198202435e-06,
+      "loss": 0.4987,
+      "step": 3251
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6514408429427263,
+      "learning_rate": 8.725833969027835e-06,
+      "loss": 0.5395,
+      "step": 3252
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7633915136207532,
+      "learning_rate": 8.724985714974735e-06,
+      "loss": 0.489,
+      "step": 3253
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6115957997969865,
+      "learning_rate": 8.724137219919932e-06,
+      "loss": 0.51,
+      "step": 3254
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6810536140975803,
+      "learning_rate": 8.723288483918324e-06,
+      "loss": 0.5143,
+      "step": 3255
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4177634087851634,
+      "learning_rate": 8.722439507024823e-06,
+      "loss": 0.5046,
+      "step": 3256
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3166413699116888,
+      "learning_rate": 8.721590289294353e-06,
+      "loss": 0.5099,
+      "step": 3257
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.955926209630783,
+      "learning_rate": 8.720740830781862e-06,
+      "loss": 0.5019,
+      "step": 3258
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5304960229849456,
+      "learning_rate": 8.719891131542304e-06,
+      "loss": 0.5122,
+      "step": 3259
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6329933326511215,
+      "learning_rate": 8.719041191630657e-06,
+      "loss": 0.5359,
+      "step": 3260
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9471169594043491,
+      "learning_rate": 8.718191011101911e-06,
+      "loss": 0.5373,
+      "step": 3261
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8252809784576949,
+      "learning_rate": 8.71734059001107e-06,
+      "loss": 0.5353,
+      "step": 3262
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.453997269518261,
+      "learning_rate": 8.716489928413154e-06,
+      "loss": 0.4698,
+      "step": 3263
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2199214494286372,
+      "learning_rate": 8.715639026363204e-06,
+      "loss": 0.4867,
+      "step": 3264
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.576161409055691,
+      "learning_rate": 8.71478788391627e-06,
+      "loss": 0.523,
+      "step": 3265
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.850861240333354,
+      "learning_rate": 8.713936501127417e-06,
+      "loss": 0.5402,
+      "step": 3266
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8810382680260067,
+      "learning_rate": 8.713084878051732e-06,
+      "loss": 0.5424,
+      "step": 3267
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.05436447099998,
+      "learning_rate": 8.712233014744312e-06,
+      "loss": 0.4559,
+      "step": 3268
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.605676107103658,
+      "learning_rate": 8.711380911260274e-06,
+      "loss": 0.4992,
+      "step": 3269
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.590019514033087,
+      "learning_rate": 8.710528567654743e-06,
+      "loss": 0.4857,
+      "step": 3270
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6607119507240535,
+      "learning_rate": 8.709675983982871e-06,
+      "loss": 0.5601,
+      "step": 3271
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.29371474544447,
+      "learning_rate": 8.708823160299815e-06,
+      "loss": 0.5174,
+      "step": 3272
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8046736155722014,
+      "learning_rate": 8.70797009666075e-06,
+      "loss": 0.525,
+      "step": 3273
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.474239005668495,
+      "learning_rate": 8.707116793120873e-06,
+      "loss": 0.4464,
+      "step": 3274
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3882866349497234,
+      "learning_rate": 8.706263249735386e-06,
+      "loss": 0.5017,
+      "step": 3275
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6740474734258126,
+      "learning_rate": 8.70540946655952e-06,
+      "loss": 0.5239,
+      "step": 3276
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.596804058037078,
+      "learning_rate": 8.704555443648505e-06,
+      "loss": 0.5289,
+      "step": 3277
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5539151439924102,
+      "learning_rate": 8.7037011810576e-06,
+      "loss": 0.5074,
+      "step": 3278
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7434101723847306,
+      "learning_rate": 8.702846678842074e-06,
+      "loss": 0.5344,
+      "step": 3279
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.1180408446355363,
+      "learning_rate": 8.701991937057211e-06,
+      "loss": 0.5061,
+      "step": 3280
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8048402485957895,
+      "learning_rate": 8.701136955758312e-06,
+      "loss": 0.5147,
+      "step": 3281
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 4.171740514451025,
+      "learning_rate": 8.700281735000695e-06,
+      "loss": 0.4934,
+      "step": 3282
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6132874747052792,
+      "learning_rate": 8.69942627483969e-06,
+      "loss": 0.522,
+      "step": 3283
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.502744531882189,
+      "learning_rate": 8.698570575330644e-06,
+      "loss": 0.4656,
+      "step": 3284
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9950113678664435,
+      "learning_rate": 8.69771463652892e-06,
+      "loss": 0.5367,
+      "step": 3285
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2878335049985021,
+      "learning_rate": 8.696858458489898e-06,
+      "loss": 0.4522,
+      "step": 3286
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5912814947163051,
+      "learning_rate": 8.696002041268966e-06,
+      "loss": 0.55,
+      "step": 3287
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6716647123639256,
+      "learning_rate": 8.695145384921542e-06,
+      "loss": 0.5471,
+      "step": 3288
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3367109825298962,
+      "learning_rate": 8.694288489503042e-06,
+      "loss": 0.4993,
+      "step": 3289
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6054248328073868,
+      "learning_rate": 8.69343135506891e-06,
+      "loss": 0.5828,
+      "step": 3290
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5257874945760699,
+      "learning_rate": 8.6925739816746e-06,
+      "loss": 0.5205,
+      "step": 3291
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.361738914723251,
+      "learning_rate": 8.691716369375587e-06,
+      "loss": 0.4405,
+      "step": 3292
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.026740763953979,
+      "learning_rate": 8.690858518227353e-06,
+      "loss": 0.5302,
+      "step": 3293
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4678756295040023,
+      "learning_rate": 8.6900004282854e-06,
+      "loss": 0.4763,
+      "step": 3294
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.725647756449732,
+      "learning_rate": 8.689142099605245e-06,
+      "loss": 0.5136,
+      "step": 3295
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5277729043077464,
+      "learning_rate": 8.688283532242425e-06,
+      "loss": 0.3995,
+      "step": 3296
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5927902338038629,
+      "learning_rate": 8.687424726252485e-06,
+      "loss": 0.5528,
+      "step": 3297
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8701808459539466,
+      "learning_rate": 8.686565681690988e-06,
+      "loss": 0.531,
+      "step": 3298
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6743002856732259,
+      "learning_rate": 8.685706398613514e-06,
+      "loss": 0.5056,
+      "step": 3299
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5225811032038654,
+      "learning_rate": 8.684846877075659e-06,
+      "loss": 0.4669,
+      "step": 3300
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.711775799003001,
+      "learning_rate": 8.68398711713303e-06,
+      "loss": 0.555,
+      "step": 3301
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6350364104433937,
+      "learning_rate": 8.683127118841254e-06,
+      "loss": 0.4619,
+      "step": 3302
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.156567580785746,
+      "learning_rate": 8.682266882255972e-06,
+      "loss": 0.4562,
+      "step": 3303
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.620475054651976,
+      "learning_rate": 8.68140640743284e-06,
+      "loss": 0.5479,
+      "step": 3304
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7374599886474034,
+      "learning_rate": 8.680545694427528e-06,
+      "loss": 0.4963,
+      "step": 3305
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.606717356991121,
+      "learning_rate": 8.679684743295725e-06,
+      "loss": 0.5434,
+      "step": 3306
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4544686314428859,
+      "learning_rate": 8.678823554093132e-06,
+      "loss": 0.4459,
+      "step": 3307
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3149254028289086,
+      "learning_rate": 8.67796212687547e-06,
+      "loss": 0.465,
+      "step": 3308
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.843479598401039,
+      "learning_rate": 8.677100461698466e-06,
+      "loss": 0.5351,
+      "step": 3309
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3859516061523927,
+      "learning_rate": 8.676238558617875e-06,
+      "loss": 0.4493,
+      "step": 3310
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6351923970050173,
+      "learning_rate": 8.675376417689459e-06,
+      "loss": 0.5427,
+      "step": 3311
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6128089524835199,
+      "learning_rate": 8.674514038968996e-06,
+      "loss": 0.4923,
+      "step": 3312
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.310511921178792,
+      "learning_rate": 8.673651422512281e-06,
+      "loss": 0.5573,
+      "step": 3313
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4606585598881572,
+      "learning_rate": 8.672788568375125e-06,
+      "loss": 0.5078,
+      "step": 3314
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.54212534972613,
+      "learning_rate": 8.671925476613353e-06,
+      "loss": 0.5029,
+      "step": 3315
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9622402443042726,
+      "learning_rate": 8.671062147282807e-06,
+      "loss": 0.4937,
+      "step": 3316
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5698044735318659,
+      "learning_rate": 8.670198580439342e-06,
+      "loss": 0.5179,
+      "step": 3317
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5463224803747988,
+      "learning_rate": 8.669334776138829e-06,
+      "loss": 0.4754,
+      "step": 3318
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3254102150144311,
+      "learning_rate": 8.668470734437157e-06,
+      "loss": 0.4855,
+      "step": 3319
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.559140754757652,
+      "learning_rate": 8.667606455390226e-06,
+      "loss": 0.5137,
+      "step": 3320
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.3147781211969822,
+      "learning_rate": 8.666741939053958e-06,
+      "loss": 0.4807,
+      "step": 3321
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.572447897466725,
+      "learning_rate": 8.665877185484281e-06,
+      "loss": 0.5015,
+      "step": 3322
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4744094706933022,
+      "learning_rate": 8.665012194737147e-06,
+      "loss": 0.468,
+      "step": 3323
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5737793676016821,
+      "learning_rate": 8.664146966868519e-06,
+      "loss": 0.5422,
+      "step": 3324
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5925443018244743,
+      "learning_rate": 8.663281501934372e-06,
+      "loss": 0.4818,
+      "step": 3325
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4223252091858785,
+      "learning_rate": 8.662415799990707e-06,
+      "loss": 0.4704,
+      "step": 3326
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5623611588489181,
+      "learning_rate": 8.66154986109353e-06,
+      "loss": 0.5465,
+      "step": 3327
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6089261590280686,
+      "learning_rate": 8.660683685298867e-06,
+      "loss": 0.5636,
+      "step": 3328
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4209618999564035,
+      "learning_rate": 8.65981727266276e-06,
+      "loss": 0.5333,
+      "step": 3329
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5574860279318512,
+      "learning_rate": 8.658950623241262e-06,
+      "loss": 0.5413,
+      "step": 3330
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5558097303755274,
+      "learning_rate": 8.658083737090444e-06,
+      "loss": 0.4755,
+      "step": 3331
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3059030384001609,
+      "learning_rate": 8.657216614266394e-06,
+      "loss": 0.5341,
+      "step": 3332
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6125236336984035,
+      "learning_rate": 8.656349254825213e-06,
+      "loss": 0.4849,
+      "step": 3333
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5683101046807325,
+      "learning_rate": 8.65548165882302e-06,
+      "loss": 0.5498,
+      "step": 3334
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3988168744481584,
+      "learning_rate": 8.654613826315943e-06,
+      "loss": 0.4569,
+      "step": 3335
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2314396177863982,
+      "learning_rate": 8.653745757360135e-06,
+      "loss": 0.5176,
+      "step": 3336
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.56285090287388,
+      "learning_rate": 8.652877452011755e-06,
+      "loss": 0.5162,
+      "step": 3337
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.457067885340824,
+      "learning_rate": 8.652008910326983e-06,
+      "loss": 0.5274,
+      "step": 3338
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9775372910758946,
+      "learning_rate": 8.65114013236201e-06,
+      "loss": 0.4748,
+      "step": 3339
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5899952973360124,
+      "learning_rate": 8.650271118173048e-06,
+      "loss": 0.5172,
+      "step": 3340
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4831999555815314,
+      "learning_rate": 8.64940186781632e-06,
+      "loss": 0.5177,
+      "step": 3341
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0813248369817536,
+      "learning_rate": 8.648532381348066e-06,
+      "loss": 0.4394,
+      "step": 3342
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8501094633745563,
+      "learning_rate": 8.647662658824538e-06,
+      "loss": 0.5609,
+      "step": 3343
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.9964853560539324,
+      "learning_rate": 8.646792700302009e-06,
+      "loss": 0.5001,
+      "step": 3344
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.3075778928664605,
+      "learning_rate": 8.645922505836763e-06,
+      "loss": 0.4586,
+      "step": 3345
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9759538492898014,
+      "learning_rate": 8.6450520754851e-06,
+      "loss": 0.4753,
+      "step": 3346
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.8609803044445266,
+      "learning_rate": 8.644181409303336e-06,
+      "loss": 0.4754,
+      "step": 3347
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5824436515225349,
+      "learning_rate": 8.643310507347802e-06,
+      "loss": 0.4629,
+      "step": 3348
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4148299513700913,
+      "learning_rate": 8.642439369674845e-06,
+      "loss": 0.4575,
+      "step": 3349
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.8928965743500823,
+      "learning_rate": 8.641567996340824e-06,
+      "loss": 0.503,
+      "step": 3350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5631734093884255,
+      "learning_rate": 8.640696387402119e-06,
+      "loss": 0.494,
+      "step": 3351
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.360224022047198,
+      "learning_rate": 8.639824542915118e-06,
+      "loss": 0.4929,
+      "step": 3352
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5762223259071542,
+      "learning_rate": 8.638952462936232e-06,
+      "loss": 0.4965,
+      "step": 3353
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6690362841576996,
+      "learning_rate": 8.638080147521884e-06,
+      "loss": 0.5716,
+      "step": 3354
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7903314196851994,
+      "learning_rate": 8.637207596728508e-06,
+      "loss": 0.4674,
+      "step": 3355
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 3.088566694134135,
+      "learning_rate": 8.636334810612557e-06,
+      "loss": 0.5136,
+      "step": 3356
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0379502343558458,
+      "learning_rate": 8.635461789230501e-06,
+      "loss": 0.4708,
+      "step": 3357
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7487518801585136,
+      "learning_rate": 8.634588532638826e-06,
+      "loss": 0.4998,
+      "step": 3358
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6634527485066162,
+      "learning_rate": 8.633715040894023e-06,
+      "loss": 0.4617,
+      "step": 3359
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4499574126399932,
+      "learning_rate": 8.632841314052614e-06,
+      "loss": 0.5024,
+      "step": 3360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5038158478733046,
+      "learning_rate": 8.631967352171125e-06,
+      "loss": 0.5279,
+      "step": 3361
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6920843307996423,
+      "learning_rate": 8.631093155306097e-06,
+      "loss": 0.5303,
+      "step": 3362
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2854719067976281,
+      "learning_rate": 8.630218723514092e-06,
+      "loss": 0.4987,
+      "step": 3363
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.632168597003605,
+      "learning_rate": 8.629344056851687e-06,
+      "loss": 0.4821,
+      "step": 3364
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9018018946605562,
+      "learning_rate": 8.628469155375467e-06,
+      "loss": 0.45,
+      "step": 3365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.040628688809857,
+      "learning_rate": 8.627594019142039e-06,
+      "loss": 0.5008,
+      "step": 3366
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.853851364784639,
+      "learning_rate": 8.626718648208025e-06,
+      "loss": 0.5249,
+      "step": 3367
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4682428625440593,
+      "learning_rate": 8.625843042630058e-06,
+      "loss": 0.4879,
+      "step": 3368
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4162731744946866,
+      "learning_rate": 8.624967202464789e-06,
+      "loss": 0.4614,
+      "step": 3369
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7673872685161327,
+      "learning_rate": 8.624091127768884e-06,
+      "loss": 0.4706,
+      "step": 3370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.0122375189824626,
+      "learning_rate": 8.623214818599024e-06,
+      "loss": 0.4831,
+      "step": 3371
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5586610834189096,
+      "learning_rate": 8.622338275011907e-06,
+      "loss": 0.526,
+      "step": 3372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.637162716249673,
+      "learning_rate": 8.621461497064241e-06,
+      "loss": 0.5126,
+      "step": 3373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7119784291758082,
+      "learning_rate": 8.620584484812753e-06,
+      "loss": 0.4291,
+      "step": 3374
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3360263829998105,
+      "learning_rate": 8.619707238314186e-06,
+      "loss": 0.4423,
+      "step": 3375
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4398874554887273,
+      "learning_rate": 8.618829757625295e-06,
+      "loss": 0.4524,
+      "step": 3376
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6591620093181925,
+      "learning_rate": 8.617952042802853e-06,
+      "loss": 0.466,
+      "step": 3377
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5802421722380953,
+      "learning_rate": 8.617074093903646e-06,
+      "loss": 0.5542,
+      "step": 3378
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4277828447085816,
+      "learning_rate": 8.616195910984477e-06,
+      "loss": 0.5168,
+      "step": 3379
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5708225290722235,
+      "learning_rate": 8.615317494102164e-06,
+      "loss": 0.4659,
+      "step": 3380
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7368543749934282,
+      "learning_rate": 8.614438843313538e-06,
+      "loss": 0.5221,
+      "step": 3381
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6170659242108831,
+      "learning_rate": 8.613559958675448e-06,
+      "loss": 0.5126,
+      "step": 3382
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0461309388225177,
+      "learning_rate": 8.612680840244756e-06,
+      "loss": 0.4733,
+      "step": 3383
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5025158314178289,
+      "learning_rate": 8.611801488078337e-06,
+      "loss": 0.548,
+      "step": 3384
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4660126561541762,
+      "learning_rate": 8.61092190223309e-06,
+      "loss": 0.4597,
+      "step": 3385
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3272809697306307,
+      "learning_rate": 8.610042082765917e-06,
+      "loss": 0.5066,
+      "step": 3386
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1500451240413545,
+      "learning_rate": 8.609162029733745e-06,
+      "loss": 0.5163,
+      "step": 3387
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.551537554343829,
+      "learning_rate": 8.608281743193511e-06,
+      "loss": 0.5176,
+      "step": 3388
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.214506125127254,
+      "learning_rate": 8.607401223202168e-06,
+      "loss": 0.5297,
+      "step": 3389
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3729506952670276,
+      "learning_rate": 8.606520469816686e-06,
+      "loss": 0.4776,
+      "step": 3390
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6672889002072822,
+      "learning_rate": 8.605639483094046e-06,
+      "loss": 0.5324,
+      "step": 3391
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.428277935905827,
+      "learning_rate": 8.604758263091248e-06,
+      "loss": 0.4697,
+      "step": 3392
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5867874160773533,
+      "learning_rate": 8.603876809865308e-06,
+      "loss": 0.5192,
+      "step": 3393
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.9519093028723848,
+      "learning_rate": 8.602995123473252e-06,
+      "loss": 0.4651,
+      "step": 3394
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7290439037133947,
+      "learning_rate": 8.602113203972124e-06,
+      "loss": 0.5088,
+      "step": 3395
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6604500598235477,
+      "learning_rate": 8.601231051418984e-06,
+      "loss": 0.4979,
+      "step": 3396
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.902113332736501,
+      "learning_rate": 8.600348665870907e-06,
+      "loss": 0.4779,
+      "step": 3397
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8990372612130764,
+      "learning_rate": 8.599466047384981e-06,
+      "loss": 0.5493,
+      "step": 3398
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6093362917532382,
+      "learning_rate": 8.59858319601831e-06,
+      "loss": 0.5285,
+      "step": 3399
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.5925181383728004,
+      "learning_rate": 8.597700111828013e-06,
+      "loss": 0.5368,
+      "step": 3400
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0464333240802937,
+      "learning_rate": 8.596816794871226e-06,
+      "loss": 0.4615,
+      "step": 3401
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.496038259774819,
+      "learning_rate": 8.595933245205097e-06,
+      "loss": 0.5162,
+      "step": 3402
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6301056904045803,
+      "learning_rate": 8.595049462886793e-06,
+      "loss": 0.5421,
+      "step": 3403
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3537039073986628,
+      "learning_rate": 8.594165447973489e-06,
+      "loss": 0.4274,
+      "step": 3404
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4196602530127012,
+      "learning_rate": 8.593281200522383e-06,
+      "loss": 0.4199,
+      "step": 3405
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.338662609894142,
+      "learning_rate": 8.592396720590686e-06,
+      "loss": 0.5216,
+      "step": 3406
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8143335477961813,
+      "learning_rate": 8.591512008235618e-06,
+      "loss": 0.4796,
+      "step": 3407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5739591118269822,
+      "learning_rate": 8.590627063514423e-06,
+      "loss": 0.4716,
+      "step": 3408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6092519362794887,
+      "learning_rate": 8.589741886484353e-06,
+      "loss": 0.5341,
+      "step": 3409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.79803200226372,
+      "learning_rate": 8.58885647720268e-06,
+      "loss": 0.5172,
+      "step": 3410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.537787121283042,
+      "learning_rate": 8.587970835726686e-06,
+      "loss": 0.5115,
+      "step": 3411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8829377719950688,
+      "learning_rate": 8.587084962113675e-06,
+      "loss": 0.4658,
+      "step": 3412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5864625990007081,
+      "learning_rate": 8.58619885642096e-06,
+      "loss": 0.5287,
+      "step": 3413
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.503080233308453,
+      "learning_rate": 8.585312518705867e-06,
+      "loss": 0.4819,
+      "step": 3414
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6320269067837785,
+      "learning_rate": 8.584425949025745e-06,
+      "loss": 0.5016,
+      "step": 3415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6228090259418075,
+      "learning_rate": 8.583539147437955e-06,
+      "loss": 0.4552,
+      "step": 3416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.522898879856291,
+      "learning_rate": 8.58265211399987e-06,
+      "loss": 0.5469,
+      "step": 3417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7537577765786239,
+      "learning_rate": 8.581764848768878e-06,
+      "loss": 0.585,
+      "step": 3418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7892521641738275,
+      "learning_rate": 8.580877351802385e-06,
+      "loss": 0.4875,
+      "step": 3419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6481390121889066,
+      "learning_rate": 8.579989623157813e-06,
+      "loss": 0.4471,
+      "step": 3420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6274441155600545,
+      "learning_rate": 8.579101662892597e-06,
+      "loss": 0.5045,
+      "step": 3421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8432354931210932,
+      "learning_rate": 8.578213471064184e-06,
+      "loss": 0.4739,
+      "step": 3422
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3120040824502734,
+      "learning_rate": 8.577325047730042e-06,
+      "loss": 0.4858,
+      "step": 3423
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5782400480818928,
+      "learning_rate": 8.576436392947647e-06,
+      "loss": 0.5321,
+      "step": 3424
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7700043481258967,
+      "learning_rate": 8.575547506774498e-06,
+      "loss": 0.4671,
+      "step": 3425
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3612692648329476,
+      "learning_rate": 8.574658389268102e-06,
+      "loss": 0.5095,
+      "step": 3426
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6673306555114653,
+      "learning_rate": 8.573769040485984e-06,
+      "loss": 0.5422,
+      "step": 3427
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.02520097025789,
+      "learning_rate": 8.572879460485684e-06,
+      "loss": 0.4889,
+      "step": 3428
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7105565641077909,
+      "learning_rate": 8.571989649324756e-06,
+      "loss": 0.454,
+      "step": 3429
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.692577742408476,
+      "learning_rate": 8.571099607060772e-06,
+      "loss": 0.5264,
+      "step": 3430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4097702946714268,
+      "learning_rate": 8.570209333751314e-06,
+      "loss": 0.5346,
+      "step": 3431
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8243493143063017,
+      "learning_rate": 8.569318829453983e-06,
+      "loss": 0.5361,
+      "step": 3432
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5749127129389127,
+      "learning_rate": 8.568428094226394e-06,
+      "loss": 0.549,
+      "step": 3433
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.7597873761827763,
+      "learning_rate": 8.567537128126173e-06,
+      "loss": 0.5259,
+      "step": 3434
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5294892748791478,
+      "learning_rate": 8.566645931210968e-06,
+      "loss": 0.5227,
+      "step": 3435
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.1399663766329424,
+      "learning_rate": 8.565754503538439e-06,
+      "loss": 0.4902,
+      "step": 3436
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4966191786860732,
+      "learning_rate": 8.564862845166255e-06,
+      "loss": 0.4682,
+      "step": 3437
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3491070142745738,
+      "learning_rate": 8.56397095615211e-06,
+      "loss": 0.4548,
+      "step": 3438
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.0830471332374536,
+      "learning_rate": 8.563078836553707e-06,
+      "loss": 0.4876,
+      "step": 3439
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7089210568201886,
+      "learning_rate": 8.562186486428762e-06,
+      "loss": 0.5138,
+      "step": 3440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4759646961717325,
+      "learning_rate": 8.561293905835013e-06,
+      "loss": 0.4832,
+      "step": 3441
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3650574549663443,
+      "learning_rate": 8.560401094830207e-06,
+      "loss": 0.5138,
+      "step": 3442
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5442107272449699,
+      "learning_rate": 8.559508053472108e-06,
+      "loss": 0.5,
+      "step": 3443
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.703855504966165,
+      "learning_rate": 8.558614781818493e-06,
+      "loss": 0.4669,
+      "step": 3444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.648494388774955,
+      "learning_rate": 8.557721279927157e-06,
+      "loss": 0.5152,
+      "step": 3445
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.481186336780409,
+      "learning_rate": 8.55682754785591e-06,
+      "loss": 0.4628,
+      "step": 3446
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7798652409788283,
+      "learning_rate": 8.555933585662572e-06,
+      "loss": 0.4172,
+      "step": 3447
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6996518725215578,
+      "learning_rate": 8.555039393404985e-06,
+      "loss": 0.5579,
+      "step": 3448
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4615321263119314,
+      "learning_rate": 8.554144971140998e-06,
+      "loss": 0.4335,
+      "step": 3449
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.251720493892175,
+      "learning_rate": 8.55325031892848e-06,
+      "loss": 0.4778,
+      "step": 3450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5947884910918764,
+      "learning_rate": 8.552355436825316e-06,
+      "loss": 0.5415,
+      "step": 3451
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8274859239957015,
+      "learning_rate": 8.551460324889402e-06,
+      "loss": 0.4991,
+      "step": 3452
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.611402691481396,
+      "learning_rate": 8.55056498317865e-06,
+      "loss": 0.4536,
+      "step": 3453
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.564327029483822,
+      "learning_rate": 8.549669411750988e-06,
+      "loss": 0.4617,
+      "step": 3454
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5594684701759445,
+      "learning_rate": 8.548773610664361e-06,
+      "loss": 0.491,
+      "step": 3455
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5490750688444723,
+      "learning_rate": 8.547877579976722e-06,
+      "loss": 0.52,
+      "step": 3456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6966686510651647,
+      "learning_rate": 8.546981319746046e-06,
+      "loss": 0.5338,
+      "step": 3457
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.713725992603186,
+      "learning_rate": 8.546084830030317e-06,
+      "loss": 0.4598,
+      "step": 3458
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6150268713324031,
+      "learning_rate": 8.545188110887539e-06,
+      "loss": 0.5053,
+      "step": 3459
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5345439453332597,
+      "learning_rate": 8.544291162375729e-06,
+      "loss": 0.5192,
+      "step": 3460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.354579180065015,
+      "learning_rate": 8.543393984552916e-06,
+      "loss": 0.4442,
+      "step": 3461
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.23243592720789,
+      "learning_rate": 8.542496577477149e-06,
+      "loss": 0.4598,
+      "step": 3462
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5274012221906568,
+      "learning_rate": 8.541598941206486e-06,
+      "loss": 0.5108,
+      "step": 3463
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7360261842061206,
+      "learning_rate": 8.540701075799005e-06,
+      "loss": 0.4803,
+      "step": 3464
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6934249516311686,
+      "learning_rate": 8.539802981312795e-06,
+      "loss": 0.5458,
+      "step": 3465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3574359297032537,
+      "learning_rate": 8.538904657805965e-06,
+      "loss": 0.509,
+      "step": 3466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.100629773586588,
+      "learning_rate": 8.538006105336631e-06,
+      "loss": 0.5159,
+      "step": 3467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4285321684010133,
+      "learning_rate": 8.537107323962932e-06,
+      "loss": 0.4773,
+      "step": 3468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4802576286819313,
+      "learning_rate": 8.536208313743016e-06,
+      "loss": 0.5039,
+      "step": 3469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.081440882178319,
+      "learning_rate": 8.535309074735047e-06,
+      "loss": 0.4882,
+      "step": 3470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6176467017207867,
+      "learning_rate": 8.534409606997207e-06,
+      "loss": 0.4805,
+      "step": 3471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7660708097358375,
+      "learning_rate": 8.53350991058769e-06,
+      "loss": 0.4956,
+      "step": 3472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2831561890726368,
+      "learning_rate": 8.532609985564701e-06,
+      "loss": 0.4917,
+      "step": 3473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5781511020199106,
+      "learning_rate": 8.531709831986469e-06,
+      "loss": 0.4811,
+      "step": 3474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4593111363582796,
+      "learning_rate": 8.53080944991123e-06,
+      "loss": 0.4479,
+      "step": 3475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4232914925711544,
+      "learning_rate": 8.52990883939724e-06,
+      "loss": 0.458,
+      "step": 3476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9057724807387423,
+      "learning_rate": 8.529008000502766e-06,
+      "loss": 0.5447,
+      "step": 3477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.743136439665139,
+      "learning_rate": 8.528106933286089e-06,
+      "loss": 0.5301,
+      "step": 3478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.537862652631991,
+      "learning_rate": 8.527205637805508e-06,
+      "loss": 0.5601,
+      "step": 3479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.4148624643922876,
+      "learning_rate": 8.526304114119339e-06,
+      "loss": 0.5009,
+      "step": 3480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6608835214722222,
+      "learning_rate": 8.525402362285905e-06,
+      "loss": 0.5365,
+      "step": 3481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5078079039446,
+      "learning_rate": 8.524500382363552e-06,
+      "loss": 0.4686,
+      "step": 3482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6206931386970211,
+      "learning_rate": 8.523598174410633e-06,
+      "loss": 0.5012,
+      "step": 3483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.465400663347099,
+      "learning_rate": 8.522695738485523e-06,
+      "loss": 0.5581,
+      "step": 3484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3296633385338978,
+      "learning_rate": 8.521793074646604e-06,
+      "loss": 0.4785,
+      "step": 3485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0689859196842466,
+      "learning_rate": 8.520890182952284e-06,
+      "loss": 0.5494,
+      "step": 3486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.60204598116505,
+      "learning_rate": 8.519987063460973e-06,
+      "loss": 0.5232,
+      "step": 3487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.258823055899541,
+      "learning_rate": 8.519083716231105e-06,
+      "loss": 0.4628,
+      "step": 3488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 3.1386437106876524,
+      "learning_rate": 8.518180141321121e-06,
+      "loss": 0.5062,
+      "step": 3489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8585312654123055,
+      "learning_rate": 8.517276338789489e-06,
+      "loss": 0.5448,
+      "step": 3490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.586787706011846,
+      "learning_rate": 8.516372308694677e-06,
+      "loss": 0.495,
+      "step": 3491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8550070518681759,
+      "learning_rate": 8.515468051095175e-06,
+      "loss": 0.5316,
+      "step": 3492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7538440038665466,
+      "learning_rate": 8.51456356604949e-06,
+      "loss": 0.5226,
+      "step": 3493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7535593369925833,
+      "learning_rate": 8.51365885361614e-06,
+      "loss": 0.5137,
+      "step": 3494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6673528282688843,
+      "learning_rate": 8.512753913853658e-06,
+      "loss": 0.5373,
+      "step": 3495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6757778939099565,
+      "learning_rate": 8.511848746820593e-06,
+      "loss": 0.5291,
+      "step": 3496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.878406014535827,
+      "learning_rate": 8.510943352575509e-06,
+      "loss": 0.5273,
+      "step": 3497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6404802688089888,
+      "learning_rate": 8.510037731176983e-06,
+      "loss": 0.5027,
+      "step": 3498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6603182051462595,
+      "learning_rate": 8.509131882683606e-06,
+      "loss": 0.536,
+      "step": 3499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8001080671366464,
+      "learning_rate": 8.50822580715399e-06,
+      "loss": 0.5278,
+      "step": 3500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6856738625758936,
+      "learning_rate": 8.50731950464675e-06,
+      "loss": 0.4439,
+      "step": 3501
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7091377257495401,
+      "learning_rate": 8.506412975220528e-06,
+      "loss": 0.4755,
+      "step": 3502
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7821481878106913,
+      "learning_rate": 8.505506218933975e-06,
+      "loss": 0.5099,
+      "step": 3503
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4891233117372127,
+      "learning_rate": 8.504599235845752e-06,
+      "loss": 0.4575,
+      "step": 3504
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.868290029113564,
+      "learning_rate": 8.503692026014544e-06,
+      "loss": 0.5007,
+      "step": 3505
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6012568877311155,
+      "learning_rate": 8.502784589499046e-06,
+      "loss": 0.5204,
+      "step": 3506
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7552268232766387,
+      "learning_rate": 8.501876926357967e-06,
+      "loss": 0.5059,
+      "step": 3507
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6799784752327345,
+      "learning_rate": 8.500969036650034e-06,
+      "loss": 0.5267,
+      "step": 3508
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.295192757248136,
+      "learning_rate": 8.50006092043398e-06,
+      "loss": 0.4716,
+      "step": 3509
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6552627331885357,
+      "learning_rate": 8.499152577768566e-06,
+      "loss": 0.4958,
+      "step": 3510
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5350352822429176,
+      "learning_rate": 8.498244008712557e-06,
+      "loss": 0.46,
+      "step": 3511
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.821835902574183,
+      "learning_rate": 8.497335213324738e-06,
+      "loss": 0.5048,
+      "step": 3512
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5317020506007402,
+      "learning_rate": 8.496426191663905e-06,
+      "loss": 0.4856,
+      "step": 3513
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4918654086961218,
+      "learning_rate": 8.495516943788871e-06,
+      "loss": 0.4673,
+      "step": 3514
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.4274132032077538,
+      "learning_rate": 8.494607469758463e-06,
+      "loss": 0.528,
+      "step": 3515
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4739577834237916,
+      "learning_rate": 8.493697769631524e-06,
+      "loss": 0.5008,
+      "step": 3516
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5457108714457204,
+      "learning_rate": 8.49278784346691e-06,
+      "loss": 0.5497,
+      "step": 3517
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3456723386650964,
+      "learning_rate": 8.49187769132349e-06,
+      "loss": 0.449,
+      "step": 3518
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3806135820535863,
+      "learning_rate": 8.490967313260152e-06,
+      "loss": 0.4417,
+      "step": 3519
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.68446419143438,
+      "learning_rate": 8.490056709335797e-06,
+      "loss": 0.48,
+      "step": 3520
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 3.1048098672420754,
+      "learning_rate": 8.489145879609338e-06,
+      "loss": 0.503,
+      "step": 3521
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5995198924446654,
+      "learning_rate": 8.488234824139702e-06,
+      "loss": 0.4826,
+      "step": 3522
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.435998340285681,
+      "learning_rate": 8.48732354298584e-06,
+      "loss": 0.4914,
+      "step": 3523
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4155928686510237,
+      "learning_rate": 8.486412036206706e-06,
+      "loss": 0.4645,
+      "step": 3524
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4267204486689717,
+      "learning_rate": 8.485500303861273e-06,
+      "loss": 0.4812,
+      "step": 3525
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.443856573254523,
+      "learning_rate": 8.48458834600853e-06,
+      "loss": 0.493,
+      "step": 3526
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.777438449343234,
+      "learning_rate": 8.48367616270748e-06,
+      "loss": 0.5108,
+      "step": 3527
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6475435757119186,
+      "learning_rate": 8.482763754017139e-06,
+      "loss": 0.53,
+      "step": 3528
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.351917881907467,
+      "learning_rate": 8.481851119996539e-06,
+      "loss": 0.4909,
+      "step": 3529
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5278940188659935,
+      "learning_rate": 8.480938260704726e-06,
+      "loss": 0.4982,
+      "step": 3530
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 4.073438423814597,
+      "learning_rate": 8.480025176200763e-06,
+      "loss": 0.4855,
+      "step": 3531
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7445848600246636,
+      "learning_rate": 8.479111866543721e-06,
+      "loss": 0.5028,
+      "step": 3532
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8181362255757585,
+      "learning_rate": 8.478198331792694e-06,
+      "loss": 0.558,
+      "step": 3533
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4535009695487249,
+      "learning_rate": 8.477284572006786e-06,
+      "loss": 0.4959,
+      "step": 3534
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4926737882913486,
+      "learning_rate": 8.476370587245114e-06,
+      "loss": 0.4862,
+      "step": 3535
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6684025680871163,
+      "learning_rate": 8.475456377566813e-06,
+      "loss": 0.5432,
+      "step": 3536
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5798026142982897,
+      "learning_rate": 8.474541943031029e-06,
+      "loss": 0.5199,
+      "step": 3537
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3192915533053526,
+      "learning_rate": 8.47362728369693e-06,
+      "loss": 0.4334,
+      "step": 3538
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.3255908843484505,
+      "learning_rate": 8.472712399623686e-06,
+      "loss": 0.4786,
+      "step": 3539
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4751602488406883,
+      "learning_rate": 8.471797290870497e-06,
+      "loss": 0.5017,
+      "step": 3540
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.7480428416570915,
+      "learning_rate": 8.470881957496559e-06,
+      "loss": 0.4833,
+      "step": 3541
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 6.056886866342889,
+      "learning_rate": 8.469966399561102e-06,
+      "loss": 0.5036,
+      "step": 3542
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7825901255222387,
+      "learning_rate": 8.46905061712336e-06,
+      "loss": 0.5584,
+      "step": 3543
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.859025861444017,
+      "learning_rate": 8.46813461024258e-06,
+      "loss": 0.4797,
+      "step": 3544
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6106268913973407,
+      "learning_rate": 8.467218378978028e-06,
+      "loss": 0.4446,
+      "step": 3545
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.8228460113244127,
+      "learning_rate": 8.466301923388981e-06,
+      "loss": 0.4717,
+      "step": 3546
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.735035631212725,
+      "learning_rate": 8.465385243534735e-06,
+      "loss": 0.4518,
+      "step": 3547
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5826214244177539,
+      "learning_rate": 8.464468339474596e-06,
+      "loss": 0.5217,
+      "step": 3548
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3887343216738572,
+      "learning_rate": 8.463551211267888e-06,
+      "loss": 0.5174,
+      "step": 3549
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7914155208098235,
+      "learning_rate": 8.462633858973948e-06,
+      "loss": 0.473,
+      "step": 3550
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3641775461124308,
+      "learning_rate": 8.461716282652127e-06,
+      "loss": 0.4774,
+      "step": 3551
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6515559426233766,
+      "learning_rate": 8.460798482361792e-06,
+      "loss": 0.5333,
+      "step": 3552
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4193496124092668,
+      "learning_rate": 8.459880458162322e-06,
+      "loss": 0.5079,
+      "step": 3553
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5314299743510118,
+      "learning_rate": 8.45896221011311e-06,
+      "loss": 0.5175,
+      "step": 3554
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7408926352892873,
+      "learning_rate": 8.458043738273572e-06,
+      "loss": 0.5003,
+      "step": 3555
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3346030736207017,
+      "learning_rate": 8.457125042703124e-06,
+      "loss": 0.4664,
+      "step": 3556
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.678751819972822,
+      "learning_rate": 8.45620612346121e-06,
+      "loss": 0.5013,
+      "step": 3557
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6570189391176575,
+      "learning_rate": 8.455286980607282e-06,
+      "loss": 0.4734,
+      "step": 3558
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3768948006700044,
+      "learning_rate": 8.454367614200805e-06,
+      "loss": 0.4619,
+      "step": 3559
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6859896118371359,
+      "learning_rate": 8.453448024301262e-06,
+      "loss": 0.5476,
+      "step": 3560
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6494732187888806,
+      "learning_rate": 8.45252821096815e-06,
+      "loss": 0.5421,
+      "step": 3561
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6054544767721078,
+      "learning_rate": 8.45160817426098e-06,
+      "loss": 0.5314,
+      "step": 3562
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6439177271886503,
+      "learning_rate": 8.450687914239275e-06,
+      "loss": 0.521,
+      "step": 3563
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.643460096864342,
+      "learning_rate": 8.449767430962577e-06,
+      "loss": 0.5235,
+      "step": 3564
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7664559954007528,
+      "learning_rate": 8.448846724490438e-06,
+      "loss": 0.5396,
+      "step": 3565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.646243594182818,
+      "learning_rate": 8.44792579488243e-06,
+      "loss": 0.5239,
+      "step": 3566
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7128601074280965,
+      "learning_rate": 8.44700464219813e-06,
+      "loss": 0.4825,
+      "step": 3567
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.438059198246418,
+      "learning_rate": 8.446083266497142e-06,
+      "loss": 0.4409,
+      "step": 3568
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.752887591465036,
+      "learning_rate": 8.445161667839075e-06,
+      "loss": 0.4985,
+      "step": 3569
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.411418947943841,
+      "learning_rate": 8.444239846283553e-06,
+      "loss": 0.5236,
+      "step": 3570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.616626438960821,
+      "learning_rate": 8.443317801890219e-06,
+      "loss": 0.4858,
+      "step": 3571
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.491130309075099,
+      "learning_rate": 8.442395534718729e-06,
+      "loss": 0.4665,
+      "step": 3572
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.106509568799857,
+      "learning_rate": 8.44147304482875e-06,
+      "loss": 0.4756,
+      "step": 3573
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4264216274682489,
+      "learning_rate": 8.440550332279965e-06,
+      "loss": 0.4429,
+      "step": 3574
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3626845341637992,
+      "learning_rate": 8.439627397132075e-06,
+      "loss": 0.5021,
+      "step": 3575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5767618520775728,
+      "learning_rate": 8.438704239444795e-06,
+      "loss": 0.4652,
+      "step": 3576
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4846372811702195,
+      "learning_rate": 8.437780859277846e-06,
+      "loss": 0.4692,
+      "step": 3577
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.452022770824617,
+      "learning_rate": 8.436857256690974e-06,
+      "loss": 0.5224,
+      "step": 3578
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6081258294461187,
+      "learning_rate": 8.435933431743935e-06,
+      "loss": 0.5447,
+      "step": 3579
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6767102130019608,
+      "learning_rate": 8.435009384496496e-06,
+      "loss": 0.5245,
+      "step": 3580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.9187670411482056,
+      "learning_rate": 8.434085115008443e-06,
+      "loss": 0.4945,
+      "step": 3581
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4444595754548155,
+      "learning_rate": 8.433160623339575e-06,
+      "loss": 0.4912,
+      "step": 3582
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 9.650099163614648,
+      "learning_rate": 8.432235909549706e-06,
+      "loss": 0.5117,
+      "step": 3583
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2664314748880683,
+      "learning_rate": 8.431310973698665e-06,
+      "loss": 0.502,
+      "step": 3584
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6266398943723601,
+      "learning_rate": 8.43038581584629e-06,
+      "loss": 0.523,
+      "step": 3585
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.325519955965469,
+      "learning_rate": 8.429460436052443e-06,
+      "loss": 0.4827,
+      "step": 3586
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6798428840807524,
+      "learning_rate": 8.428534834376993e-06,
+      "loss": 0.5074,
+      "step": 3587
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.354008458249079,
+      "learning_rate": 8.427609010879823e-06,
+      "loss": 0.4918,
+      "step": 3588
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5927332098688273,
+      "learning_rate": 8.426682965620832e-06,
+      "loss": 0.5471,
+      "step": 3589
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3996701652895742,
+      "learning_rate": 8.425756698659937e-06,
+      "loss": 0.5002,
+      "step": 3590
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6368440826370525,
+      "learning_rate": 8.424830210057066e-06,
+      "loss": 0.5113,
+      "step": 3591
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4847076839752928,
+      "learning_rate": 8.42390349987216e-06,
+      "loss": 0.4728,
+      "step": 3592
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6278271222290683,
+      "learning_rate": 8.422976568165176e-06,
+      "loss": 0.4866,
+      "step": 3593
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7368030981109381,
+      "learning_rate": 8.422049414996087e-06,
+      "loss": 0.4791,
+      "step": 3594
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 8.255774396991459,
+      "learning_rate": 8.421122040424876e-06,
+      "loss": 0.4982,
+      "step": 3595
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5335374335815604,
+      "learning_rate": 8.420194444511545e-06,
+      "loss": 0.495,
+      "step": 3596
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5998144187920822,
+      "learning_rate": 8.419266627316109e-06,
+      "loss": 0.5537,
+      "step": 3597
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4180678312098747,
+      "learning_rate": 8.418338588898594e-06,
+      "loss": 0.4827,
+      "step": 3598
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4130962905531788,
+      "learning_rate": 8.417410329319043e-06,
+      "loss": 0.5016,
+      "step": 3599
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5608017635613922,
+      "learning_rate": 8.416481848637515e-06,
+      "loss": 0.5209,
+      "step": 3600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5568996716218513,
+      "learning_rate": 8.415553146914081e-06,
+      "loss": 0.5327,
+      "step": 3601
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.455557141402252,
+      "learning_rate": 8.414624224208828e-06,
+      "loss": 0.4659,
+      "step": 3602
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5745702764272025,
+      "learning_rate": 8.413695080581851e-06,
+      "loss": 0.4604,
+      "step": 3603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.668732852064739,
+      "learning_rate": 8.412765716093273e-06,
+      "loss": 0.5274,
+      "step": 3604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3305580848365866,
+      "learning_rate": 8.411836130803212e-06,
+      "loss": 0.4715,
+      "step": 3605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5926423959064373,
+      "learning_rate": 8.41090632477182e-06,
+      "loss": 0.5242,
+      "step": 3606
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3083595333188167,
+      "learning_rate": 8.409976298059251e-06,
+      "loss": 0.4807,
+      "step": 3607
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3071979716957935,
+      "learning_rate": 8.409046050725675e-06,
+      "loss": 0.455,
+      "step": 3608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4367513481825063,
+      "learning_rate": 8.408115582831278e-06,
+      "loss": 0.4911,
+      "step": 3609
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6271851818348064,
+      "learning_rate": 8.407184894436263e-06,
+      "loss": 0.537,
+      "step": 3610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5840238330626331,
+      "learning_rate": 8.406253985600843e-06,
+      "loss": 0.5285,
+      "step": 3611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5009676087361101,
+      "learning_rate": 8.405322856385246e-06,
+      "loss": 0.4817,
+      "step": 3612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5135178153963298,
+      "learning_rate": 8.404391506849714e-06,
+      "loss": 0.4865,
+      "step": 3613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.378625146542494,
+      "learning_rate": 8.403459937054504e-06,
+      "loss": 0.5105,
+      "step": 3614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3172797320275331,
+      "learning_rate": 8.40252814705989e-06,
+      "loss": 0.5182,
+      "step": 3615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5075756972089034,
+      "learning_rate": 8.401596136926156e-06,
+      "loss": 0.4732,
+      "step": 3616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4831869707440288,
+      "learning_rate": 8.400663906713599e-06,
+      "loss": 0.4662,
+      "step": 3617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5312659921644027,
+      "learning_rate": 8.39973145648254e-06,
+      "loss": 0.5196,
+      "step": 3618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4662583621163991,
+      "learning_rate": 8.3987987862933e-06,
+      "loss": 0.5217,
+      "step": 3619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3412122834367557,
+      "learning_rate": 8.397865896206226e-06,
+      "loss": 0.5044,
+      "step": 3620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.356033907806469,
+      "learning_rate": 8.396932786281674e-06,
+      "loss": 0.5699,
+      "step": 3621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3912361012916707,
+      "learning_rate": 8.395999456580015e-06,
+      "loss": 0.5185,
+      "step": 3622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2778690381500915,
+      "learning_rate": 8.395065907161632e-06,
+      "loss": 0.4698,
+      "step": 3623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.749647622379653,
+      "learning_rate": 8.394132138086925e-06,
+      "loss": 0.5237,
+      "step": 3624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6833563563685068,
+      "learning_rate": 8.393198149416311e-06,
+      "loss": 0.5363,
+      "step": 3625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.425835200719499,
+      "learning_rate": 8.392263941210217e-06,
+      "loss": 0.5174,
+      "step": 3626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6020891866968889,
+      "learning_rate": 8.39132951352908e-06,
+      "loss": 0.5236,
+      "step": 3627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6096526898170944,
+      "learning_rate": 8.390394866433362e-06,
+      "loss": 0.4736,
+      "step": 3628
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.339035364894106,
+      "learning_rate": 8.38945999998353e-06,
+      "loss": 0.5202,
+      "step": 3629
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3730360375920494,
+      "learning_rate": 8.388524914240073e-06,
+      "loss": 0.4885,
+      "step": 3630
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0538202914352133,
+      "learning_rate": 8.387589609263483e-06,
+      "loss": 0.4812,
+      "step": 3631
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9050672358915803,
+      "learning_rate": 8.386654085114278e-06,
+      "loss": 0.5672,
+      "step": 3632
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4460819114537384,
+      "learning_rate": 8.385718341852985e-06,
+      "loss": 0.5277,
+      "step": 3633
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5245473951355866,
+      "learning_rate": 8.384782379540145e-06,
+      "loss": 0.5122,
+      "step": 3634
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4189005824995047,
+      "learning_rate": 8.383846198236311e-06,
+      "loss": 0.4784,
+      "step": 3635
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4181678193995244,
+      "learning_rate": 8.382909798002056e-06,
+      "loss": 0.4887,
+      "step": 3636
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.486439996632292,
+      "learning_rate": 8.381973178897961e-06,
+      "loss": 0.5246,
+      "step": 3637
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.695807785165534,
+      "learning_rate": 8.381036340984628e-06,
+      "loss": 0.5181,
+      "step": 3638
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.1976558446208188,
+      "learning_rate": 8.380099284322665e-06,
+      "loss": 0.4543,
+      "step": 3639
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4127796115145583,
+      "learning_rate": 8.379162008972698e-06,
+      "loss": 0.4863,
+      "step": 3640
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4012040675919155,
+      "learning_rate": 8.378224514995372e-06,
+      "loss": 0.5236,
+      "step": 3641
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8020871533457317,
+      "learning_rate": 8.37728680245134e-06,
+      "loss": 0.5477,
+      "step": 3642
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2999157109260366,
+      "learning_rate": 8.376348871401268e-06,
+      "loss": 0.5062,
+      "step": 3643
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.664901543565906,
+      "learning_rate": 8.37541072190584e-06,
+      "loss": 0.4927,
+      "step": 3644
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4800927882483752,
+      "learning_rate": 8.374472354025756e-06,
+      "loss": 0.5381,
+      "step": 3645
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.0990127583477545,
+      "learning_rate": 8.373533767821725e-06,
+      "loss": 0.5116,
+      "step": 3646
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7631627674661245,
+      "learning_rate": 8.372594963354473e-06,
+      "loss": 0.4683,
+      "step": 3647
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4616009640933676,
+      "learning_rate": 8.371655940684737e-06,
+      "loss": 0.4995,
+      "step": 3648
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.509840774863518,
+      "learning_rate": 8.370716699873273e-06,
+      "loss": 0.4553,
+      "step": 3649
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.55416593727066,
+      "learning_rate": 8.369777240980848e-06,
+      "loss": 0.4944,
+      "step": 3650
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6333501333635299,
+      "learning_rate": 8.368837564068243e-06,
+      "loss": 0.5128,
+      "step": 3651
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4277752004247202,
+      "learning_rate": 8.367897669196256e-06,
+      "loss": 0.5478,
+      "step": 3652
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.560428823301475,
+      "learning_rate": 8.366957556425694e-06,
+      "loss": 0.4718,
+      "step": 3653
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6383039856130129,
+      "learning_rate": 8.366017225817382e-06,
+      "loss": 0.5447,
+      "step": 3654
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6394021376648567,
+      "learning_rate": 8.365076677432162e-06,
+      "loss": 0.559,
+      "step": 3655
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5771203653166764,
+      "learning_rate": 8.36413591133088e-06,
+      "loss": 0.5407,
+      "step": 3656
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6686323733832875,
+      "learning_rate": 8.363194927574407e-06,
+      "loss": 0.5108,
+      "step": 3657
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.435826133856278,
+      "learning_rate": 8.362253726223622e-06,
+      "loss": 0.4819,
+      "step": 3658
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3510474858697588,
+      "learning_rate": 8.361312307339419e-06,
+      "loss": 0.5396,
+      "step": 3659
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4687808067060752,
+      "learning_rate": 8.360370670982706e-06,
+      "loss": 0.4597,
+      "step": 3660
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.359503248306933,
+      "learning_rate": 8.359428817214408e-06,
+      "loss": 0.4444,
+      "step": 3661
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5341989240544043,
+      "learning_rate": 8.358486746095458e-06,
+      "loss": 0.4941,
+      "step": 3662
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5049758966319133,
+      "learning_rate": 8.35754445768681e-06,
+      "loss": 0.5512,
+      "step": 3663
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5272427359393728,
+      "learning_rate": 8.356601952049429e-06,
+      "loss": 0.4939,
+      "step": 3664
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.213701047324028,
+      "learning_rate": 8.35565922924429e-06,
+      "loss": 0.4178,
+      "step": 3665
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5123748759675288,
+      "learning_rate": 8.354716289332393e-06,
+      "loss": 0.4761,
+      "step": 3666
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.9578774684331033,
+      "learning_rate": 8.353773132374737e-06,
+      "loss": 0.4775,
+      "step": 3667
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6699576718790616,
+      "learning_rate": 8.352829758432347e-06,
+      "loss": 0.5018,
+      "step": 3668
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6568876379253346,
+      "learning_rate": 8.351886167566258e-06,
+      "loss": 0.5358,
+      "step": 3669
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.373874163611425,
+      "learning_rate": 8.350942359837519e-06,
+      "loss": 0.5139,
+      "step": 3670
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.574280488067774,
+      "learning_rate": 8.349998335307194e-06,
+      "loss": 0.514,
+      "step": 3671
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8896079019950032,
+      "learning_rate": 8.349054094036358e-06,
+      "loss": 0.5124,
+      "step": 3672
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7579451809697362,
+      "learning_rate": 8.348109636086104e-06,
+      "loss": 0.4579,
+      "step": 3673
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3339654044984615,
+      "learning_rate": 8.347164961517536e-06,
+      "loss": 0.4612,
+      "step": 3674
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.234452034880042,
+      "learning_rate": 8.346220070391773e-06,
+      "loss": 0.5284,
+      "step": 3675
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.0521907168035736,
+      "learning_rate": 8.345274962769947e-06,
+      "loss": 0.5818,
+      "step": 3676
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2870402296543073,
+      "learning_rate": 8.34432963871321e-06,
+      "loss": 0.4791,
+      "step": 3677
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.600590234802461,
+      "learning_rate": 8.34338409828272e-06,
+      "loss": 0.5055,
+      "step": 3678
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7591572777409372,
+      "learning_rate": 8.342438341539651e-06,
+      "loss": 0.4582,
+      "step": 3679
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6863892118210946,
+      "learning_rate": 8.341492368545195e-06,
+      "loss": 0.4537,
+      "step": 3680
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5182792774301714,
+      "learning_rate": 8.340546179360555e-06,
+      "loss": 0.4878,
+      "step": 3681
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3898958747068006,
+      "learning_rate": 8.339599774046945e-06,
+      "loss": 0.4807,
+      "step": 3682
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.158201053523248,
+      "learning_rate": 8.3386531526656e-06,
+      "loss": 0.474,
+      "step": 3683
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2782419474140292,
+      "learning_rate": 8.337706315277763e-06,
+      "loss": 0.5269,
+      "step": 3684
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8343384399913252,
+      "learning_rate": 8.336759261944694e-06,
+      "loss": 0.4806,
+      "step": 3685
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8976073722291327,
+      "learning_rate": 8.335811992727665e-06,
+      "loss": 0.5382,
+      "step": 3686
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.745789822859232,
+      "learning_rate": 8.334864507687964e-06,
+      "loss": 0.5362,
+      "step": 3687
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.095025528556599,
+      "learning_rate": 8.333916806886893e-06,
+      "loss": 0.4743,
+      "step": 3688
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9150248256766138,
+      "learning_rate": 8.332968890385764e-06,
+      "loss": 0.4742,
+      "step": 3689
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6154573575864144,
+      "learning_rate": 8.332020758245909e-06,
+      "loss": 0.4682,
+      "step": 3690
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5999771527864105,
+      "learning_rate": 8.331072410528669e-06,
+      "loss": 0.5291,
+      "step": 3691
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6168834821015051,
+      "learning_rate": 8.330123847295403e-06,
+      "loss": 0.5064,
+      "step": 3692
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6991481769379488,
+      "learning_rate": 8.32917506860748e-06,
+      "loss": 0.4843,
+      "step": 3693
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5929590193488814,
+      "learning_rate": 8.328226074526284e-06,
+      "loss": 0.4981,
+      "step": 3694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.76714040876322,
+      "learning_rate": 8.327276865113216e-06,
+      "loss": 0.5124,
+      "step": 3695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5525842591015449,
+      "learning_rate": 8.326327440429688e-06,
+      "loss": 0.4611,
+      "step": 3696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7311441835789767,
+      "learning_rate": 8.325377800537124e-06,
+      "loss": 0.482,
+      "step": 3697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4681202042929036,
+      "learning_rate": 8.32442794549697e-06,
+      "loss": 0.4654,
+      "step": 3698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 6.03116144345462,
+      "learning_rate": 8.323477875370673e-06,
+      "loss": 0.4903,
+      "step": 3699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3568805539139934,
+      "learning_rate": 8.322527590219708e-06,
+      "loss": 0.4957,
+      "step": 3700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.239083802536056,
+      "learning_rate": 8.321577090105551e-06,
+      "loss": 0.517,
+      "step": 3701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.508132178364029,
+      "learning_rate": 8.320626375089705e-06,
+      "loss": 0.5023,
+      "step": 3702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6680491274277014,
+      "learning_rate": 8.319675445233673e-06,
+      "loss": 0.4861,
+      "step": 3703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8665834755448074,
+      "learning_rate": 8.318724300598984e-06,
+      "loss": 0.4787,
+      "step": 3704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4859202594869967,
+      "learning_rate": 8.317772941247175e-06,
+      "loss": 0.4481,
+      "step": 3705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8304996951403574,
+      "learning_rate": 8.316821367239797e-06,
+      "loss": 0.446,
+      "step": 3706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8880781423600487,
+      "learning_rate": 8.315869578638413e-06,
+      "loss": 0.5209,
+      "step": 3707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2970785410497607,
+      "learning_rate": 8.314917575504607e-06,
+      "loss": 0.4917,
+      "step": 3708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.7830153900326597,
+      "learning_rate": 8.313965357899972e-06,
+      "loss": 0.5072,
+      "step": 3709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.140628315428497,
+      "learning_rate": 8.31301292588611e-06,
+      "loss": 0.5014,
+      "step": 3710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9698084138189735,
+      "learning_rate": 8.312060279524647e-06,
+      "loss": 0.5245,
+      "step": 3711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5618078101630656,
+      "learning_rate": 8.311107418877216e-06,
+      "loss": 0.5316,
+      "step": 3712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8519049290637395,
+      "learning_rate": 8.310154344005468e-06,
+      "loss": 0.4636,
+      "step": 3713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8962938675342438,
+      "learning_rate": 8.309201054971064e-06,
+      "loss": 0.4429,
+      "step": 3714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6473207862588837,
+      "learning_rate": 8.30824755183568e-06,
+      "loss": 0.5175,
+      "step": 3715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4466294537403426,
+      "learning_rate": 8.307293834661008e-06,
+      "loss": 0.468,
+      "step": 3716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4365003490148947,
+      "learning_rate": 8.30633990350875e-06,
+      "loss": 0.4991,
+      "step": 3717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.786382553864655,
+      "learning_rate": 8.305385758440627e-06,
+      "loss": 0.4943,
+      "step": 3718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3940101036012826,
+      "learning_rate": 8.30443139951837e-06,
+      "loss": 0.5393,
+      "step": 3719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8638661185479024,
+      "learning_rate": 8.30347682680372e-06,
+      "loss": 0.5263,
+      "step": 3720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.06328229021623,
+      "learning_rate": 8.302522040358446e-06,
+      "loss": 0.488,
+      "step": 3721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2047078109725002,
+      "learning_rate": 8.301567040244312e-06,
+      "loss": 0.4732,
+      "step": 3722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8093265826488936,
+      "learning_rate": 8.30061182652311e-06,
+      "loss": 0.5469,
+      "step": 3723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5135087958002031,
+      "learning_rate": 8.299656399256644e-06,
+      "loss": 0.496,
+      "step": 3724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5689711344266022,
+      "learning_rate": 8.298700758506722e-06,
+      "loss": 0.4415,
+      "step": 3725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6836482587843618,
+      "learning_rate": 8.297744904335179e-06,
+      "loss": 0.4973,
+      "step": 3726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.770427420309001,
+      "learning_rate": 8.296788836803853e-06,
+      "loss": 0.5624,
+      "step": 3727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4087254943549288,
+      "learning_rate": 8.295832555974602e-06,
+      "loss": 0.4677,
+      "step": 3728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.532908804421045,
+      "learning_rate": 8.294876061909296e-06,
+      "loss": 0.4526,
+      "step": 3729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6111580101885888,
+      "learning_rate": 8.29391935466982e-06,
+      "loss": 0.5354,
+      "step": 3730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 4.279204277074283,
+      "learning_rate": 8.292962434318071e-06,
+      "loss": 0.5904,
+      "step": 3731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6063369680500129,
+      "learning_rate": 8.292005300915957e-06,
+      "loss": 0.5449,
+      "step": 3732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3822382912939843,
+      "learning_rate": 8.29104795452541e-06,
+      "loss": 0.4256,
+      "step": 3733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3601984443549668,
+      "learning_rate": 8.290090395208363e-06,
+      "loss": 0.4985,
+      "step": 3734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.9526006432895975,
+      "learning_rate": 8.289132623026774e-06,
+      "loss": 0.5362,
+      "step": 3735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4957540809941587,
+      "learning_rate": 8.288174638042606e-06,
+      "loss": 0.4975,
+      "step": 3736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2348007329995847,
+      "learning_rate": 8.28721644031784e-06,
+      "loss": 0.4282,
+      "step": 3737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7233954770576565,
+      "learning_rate": 8.286258029914472e-06,
+      "loss": 0.5387,
+      "step": 3738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4780060156273445,
+      "learning_rate": 8.285299406894506e-06,
+      "loss": 0.5248,
+      "step": 3739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2489293535804054,
+      "learning_rate": 8.284340571319968e-06,
+      "loss": 0.4692,
+      "step": 3740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4638442473295616,
+      "learning_rate": 8.28338152325289e-06,
+      "loss": 0.5394,
+      "step": 3741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.510621021386326,
+      "learning_rate": 8.282422262755323e-06,
+      "loss": 0.4906,
+      "step": 3742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3725085462111033,
+      "learning_rate": 8.281462789889331e-06,
+      "loss": 0.4916,
+      "step": 3743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4767098502512201,
+      "learning_rate": 8.280503104716989e-06,
+      "loss": 0.5191,
+      "step": 3744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6113840721718989,
+      "learning_rate": 8.279543207300384e-06,
+      "loss": 0.5396,
+      "step": 3745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5724818544146906,
+      "learning_rate": 8.278583097701626e-06,
+      "loss": 0.5254,
+      "step": 3746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3047635487215181,
+      "learning_rate": 8.277622775982831e-06,
+      "loss": 0.4541,
+      "step": 3747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.249100614443897,
+      "learning_rate": 8.276662242206131e-06,
+      "loss": 0.4665,
+      "step": 3748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5966880575870631,
+      "learning_rate": 8.27570149643367e-06,
+      "loss": 0.5176,
+      "step": 3749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6579532880021566,
+      "learning_rate": 8.274740538727606e-06,
+      "loss": 0.5118,
+      "step": 3750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6090880241849397,
+      "learning_rate": 8.273779369150114e-06,
+      "loss": 0.5142,
+      "step": 3751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3907504051977013,
+      "learning_rate": 8.27281798776338e-06,
+      "loss": 0.457,
+      "step": 3752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.988991808112168,
+      "learning_rate": 8.271856394629603e-06,
+      "loss": 0.4843,
+      "step": 3753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3476339810405968,
+      "learning_rate": 8.270894589810997e-06,
+      "loss": 0.4554,
+      "step": 3754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.556395210254575,
+      "learning_rate": 8.26993257336979e-06,
+      "loss": 0.5011,
+      "step": 3755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5095410980026644,
+      "learning_rate": 8.268970345368222e-06,
+      "loss": 0.5031,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3991110806372482,
+      "learning_rate": 8.26800790586855e-06,
+      "loss": 0.4787,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4213444400208373,
+      "learning_rate": 8.267045254933043e-06,
+      "loss": 0.5181,
+      "step": 3758
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5847043684014035,
+      "learning_rate": 8.26608239262398e-06,
+      "loss": 0.5145,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6460130119621552,
+      "learning_rate": 8.26511931900366e-06,
+      "loss": 0.5166,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9161026640669243,
+      "learning_rate": 8.26415603413439e-06,
+      "loss": 0.4704,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.181486940120883,
+      "learning_rate": 8.263192538078495e-06,
+      "loss": 0.5099,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5252421694424716,
+      "learning_rate": 8.262228830898313e-06,
+      "loss": 0.4897,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3924525306911006,
+      "learning_rate": 8.261264912656191e-06,
+      "loss": 0.4848,
+      "step": 3764
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3705173756501405,
+      "learning_rate": 8.260300783414498e-06,
+      "loss": 0.4635,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6007149022555054,
+      "learning_rate": 8.259336443235608e-06,
+      "loss": 0.5372,
+      "step": 3766
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8816265028123444,
+      "learning_rate": 8.258371892181912e-06,
+      "loss": 0.4786,
+      "step": 3767
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4464214161199063,
+      "learning_rate": 8.25740713031582e-06,
+      "loss": 0.4723,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.111059252394764,
+      "learning_rate": 8.256442157699746e-06,
+      "loss": 0.4803,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5886211899008102,
+      "learning_rate": 8.255476974396126e-06,
+      "loss": 0.5027,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3881769996000595,
+      "learning_rate": 8.254511580467403e-06,
+      "loss": 0.4448,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9474478219553628,
+      "learning_rate": 8.253545975976039e-06,
+      "loss": 0.5684,
+      "step": 3772
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5616487683665813,
+      "learning_rate": 8.252580160984505e-06,
+      "loss": 0.5032,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3924499096867569,
+      "learning_rate": 8.251614135555291e-06,
+      "loss": 0.5118,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5941882152869444,
+      "learning_rate": 8.250647899750896e-06,
+      "loss": 0.5288,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4327270530742207,
+      "learning_rate": 8.249681453633834e-06,
+      "loss": 0.4992,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.59535382951488,
+      "learning_rate": 8.248714797266632e-06,
+      "loss": 0.4722,
+      "step": 3777
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7006317203430288,
+      "learning_rate": 8.247747930711834e-06,
+      "loss": 0.5019,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.320923812529501,
+      "learning_rate": 8.246780854031993e-06,
+      "loss": 0.5386,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.708374086807389,
+      "learning_rate": 8.245813567289678e-06,
+      "loss": 0.508,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2748509242495234,
+      "learning_rate": 8.244846070547472e-06,
+      "loss": 0.4967,
+      "step": 3781
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4124715004601487,
+      "learning_rate": 8.24387836386797e-06,
+      "loss": 0.5045,
+      "step": 3782
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6094962518506664,
+      "learning_rate": 8.24291044731378e-06,
+      "loss": 0.5258,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.268824318363105,
+      "learning_rate": 8.241942320947527e-06,
+      "loss": 0.5042,
+      "step": 3784
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6998645074533534,
+      "learning_rate": 8.240973984831847e-06,
+      "loss": 0.5022,
+      "step": 3785
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3304236613377396,
+      "learning_rate": 8.24000543902939e-06,
+      "loss": 0.4572,
+      "step": 3786
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.655992244791374,
+      "learning_rate": 8.23903668360282e-06,
+      "loss": 0.5231,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5843155878717685,
+      "learning_rate": 8.238067718614812e-06,
+      "loss": 0.4962,
+      "step": 3788
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3791953889625859,
+      "learning_rate": 8.237098544128059e-06,
+      "loss": 0.4686,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4605822000879984,
+      "learning_rate": 8.236129160205265e-06,
+      "loss": 0.5382,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3555206016776558,
+      "learning_rate": 8.23515956690915e-06,
+      "loss": 0.5165,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5246824662747311,
+      "learning_rate": 8.234189764302441e-06,
+      "loss": 0.4814,
+      "step": 3792
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5885871312530404,
+      "learning_rate": 8.233219752447886e-06,
+      "loss": 0.4791,
+      "step": 3793
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4607450874107129,
+      "learning_rate": 8.232249531408244e-06,
+      "loss": 0.4835,
+      "step": 3794
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4320910540166778,
+      "learning_rate": 8.231279101246283e-06,
+      "loss": 0.4812,
+      "step": 3795
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3632417252689426,
+      "learning_rate": 8.230308462024792e-06,
+      "loss": 0.4878,
+      "step": 3796
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0369757419737966,
+      "learning_rate": 8.229337613806568e-06,
+      "loss": 0.5293,
+      "step": 3797
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4924730456657083,
+      "learning_rate": 8.228366556654426e-06,
+      "loss": 0.5249,
+      "step": 3798
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.672692632030733,
+      "learning_rate": 8.227395290631192e-06,
+      "loss": 0.5533,
+      "step": 3799
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.526336730335463,
+      "learning_rate": 8.226423815799704e-06,
+      "loss": 0.503,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 9.41381302333368,
+      "learning_rate": 8.225452132222815e-06,
+      "loss": 0.4421,
+      "step": 3801
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6025387503640498,
+      "learning_rate": 8.224480239963393e-06,
+      "loss": 0.4629,
+      "step": 3802
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.509576899710408,
+      "learning_rate": 8.223508139084318e-06,
+      "loss": 0.4644,
+      "step": 3803
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4973308307578845,
+      "learning_rate": 8.222535829648482e-06,
+      "loss": 0.4966,
+      "step": 3804
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.388516501116106,
+      "learning_rate": 8.221563311718794e-06,
+      "loss": 0.4445,
+      "step": 3805
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5934194223292506,
+      "learning_rate": 8.220590585358176e-06,
+      "loss": 0.4744,
+      "step": 3806
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5622187147702429,
+      "learning_rate": 8.21961765062956e-06,
+      "loss": 0.5317,
+      "step": 3807
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.51210706115019,
+      "learning_rate": 8.218644507595891e-06,
+      "loss": 0.5006,
+      "step": 3808
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6351103976185537,
+      "learning_rate": 8.217671156320134e-06,
+      "loss": 0.5509,
+      "step": 3809
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5668474153828773,
+      "learning_rate": 8.216697596865263e-06,
+      "loss": 0.5321,
+      "step": 3810
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3913310698362227,
+      "learning_rate": 8.215723829294264e-06,
+      "loss": 0.5438,
+      "step": 3811
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3266997389244424,
+      "learning_rate": 8.214749853670142e-06,
+      "loss": 0.4606,
+      "step": 3812
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5666684610189263,
+      "learning_rate": 8.213775670055908e-06,
+      "loss": 0.5304,
+      "step": 3813
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8080086422815622,
+      "learning_rate": 8.212801278514596e-06,
+      "loss": 0.5083,
+      "step": 3814
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6893975463257926,
+      "learning_rate": 8.21182667910924e-06,
+      "loss": 0.5402,
+      "step": 3815
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.480307303637885,
+      "learning_rate": 8.2108518719029e-06,
+      "loss": 0.498,
+      "step": 3816
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6052594748805153,
+      "learning_rate": 8.209876856958645e-06,
+      "loss": 0.4719,
+      "step": 3817
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.368660774550835,
+      "learning_rate": 8.208901634339557e-06,
+      "loss": 0.5061,
+      "step": 3818
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.437724570073628,
+      "learning_rate": 8.207926204108732e-06,
+      "loss": 0.4285,
+      "step": 3819
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.1267028157725814,
+      "learning_rate": 8.206950566329276e-06,
+      "loss": 0.4836,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.76688823916607,
+      "learning_rate": 8.205974721064314e-06,
+      "loss": 0.4896,
+      "step": 3821
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5985212805077254,
+      "learning_rate": 8.204998668376983e-06,
+      "loss": 0.521,
+      "step": 3822
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7416833290392446,
+      "learning_rate": 8.204022408330428e-06,
+      "loss": 0.5015,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6174538522905797,
+      "learning_rate": 8.203045940987816e-06,
+      "loss": 0.4931,
+      "step": 3824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7357455524837244,
+      "learning_rate": 8.202069266412322e-06,
+      "loss": 0.4829,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9217659697859035,
+      "learning_rate": 8.201092384667135e-06,
+      "loss": 0.504,
+      "step": 3826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.691841591153293,
+      "learning_rate": 8.200115295815458e-06,
+      "loss": 0.525,
+      "step": 3827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7537367368459895,
+      "learning_rate": 8.199137999920507e-06,
+      "loss": 0.4781,
+      "step": 3828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.933735173893854,
+      "learning_rate": 8.198160497045516e-06,
+      "loss": 0.4499,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5030260581820696,
+      "learning_rate": 8.197182787253721e-06,
+      "loss": 0.5263,
+      "step": 3830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4592796583320504,
+      "learning_rate": 8.196204870608384e-06,
+      "loss": 0.5226,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4147993090273425,
+      "learning_rate": 8.19522674717277e-06,
+      "loss": 0.509,
+      "step": 3832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2148906445425809,
+      "learning_rate": 8.19424841701017e-06,
+      "loss": 0.5278,
+      "step": 3833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6730047014014413,
+      "learning_rate": 8.193269880183872e-06,
+      "loss": 0.5004,
+      "step": 3834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4811486396969538,
+      "learning_rate": 8.192291136757191e-06,
+      "loss": 0.4622,
+      "step": 3835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4793967518297202,
+      "learning_rate": 8.19131218679345e-06,
+      "loss": 0.5004,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.711183296022157,
+      "learning_rate": 8.190333030355986e-06,
+      "loss": 0.45,
+      "step": 3837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5955900295373451,
+      "learning_rate": 8.189353667508149e-06,
+      "loss": 0.5226,
+      "step": 3838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.710295995116325,
+      "learning_rate": 8.1883740983133e-06,
+      "loss": 0.4943,
+      "step": 3839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7315456057696605,
+      "learning_rate": 8.187394322834818e-06,
+      "loss": 0.4833,
+      "step": 3840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8113645514012437,
+      "learning_rate": 8.186414341136094e-06,
+      "loss": 0.5675,
+      "step": 3841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3776792620905256,
+      "learning_rate": 8.18543415328053e-06,
+      "loss": 0.4719,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3928472204088052,
+      "learning_rate": 8.184453759331543e-06,
+      "loss": 0.4914,
+      "step": 3843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3411310025732577,
+      "learning_rate": 8.183473159352564e-06,
+      "loss": 0.4991,
+      "step": 3844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9428183978502724,
+      "learning_rate": 8.182492353407036e-06,
+      "loss": 0.4858,
+      "step": 3845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5338599860376683,
+      "learning_rate": 8.181511341558414e-06,
+      "loss": 0.5038,
+      "step": 3846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4223100571598641,
+      "learning_rate": 8.180530123870174e-06,
+      "loss": 0.4891,
+      "step": 3847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4456168866449077,
+      "learning_rate": 8.179548700405793e-06,
+      "loss": 0.4408,
+      "step": 3848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2636840006506356,
+      "learning_rate": 8.17856707122877e-06,
+      "loss": 0.4738,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.436312251953886,
+      "learning_rate": 8.177585236402618e-06,
+      "loss": 0.4989,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.618164094702493,
+      "learning_rate": 8.176603195990856e-06,
+      "loss": 0.4904,
+      "step": 3851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4549503582199514,
+      "learning_rate": 8.175620950057024e-06,
+      "loss": 0.4618,
+      "step": 3852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.8421113703807848,
+      "learning_rate": 8.17463849866467e-06,
+      "loss": 0.4987,
+      "step": 3853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.349902418450739,
+      "learning_rate": 8.173655841877359e-06,
+      "loss": 0.4445,
+      "step": 3854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.103151088494613,
+      "learning_rate": 8.172672979758665e-06,
+      "loss": 0.4626,
+      "step": 3855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 6.3404286271211525,
+      "learning_rate": 8.171689912372181e-06,
+      "loss": 0.4905,
+      "step": 3856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4263196803529008,
+      "learning_rate": 8.170706639781508e-06,
+      "loss": 0.4325,
+      "step": 3857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6711489240848947,
+      "learning_rate": 8.169723162050263e-06,
+      "loss": 0.5194,
+      "step": 3858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7975298763337655,
+      "learning_rate": 8.168739479242075e-06,
+      "loss": 0.5462,
+      "step": 3859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6694766027780488,
+      "learning_rate": 8.167755591420591e-06,
+      "loss": 0.5319,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2017746683977855,
+      "learning_rate": 8.166771498649463e-06,
+      "loss": 0.5188,
+      "step": 3861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5676744889538627,
+      "learning_rate": 8.165787200992362e-06,
+      "loss": 0.4923,
+      "step": 3862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5626233866235272,
+      "learning_rate": 8.164802698512968e-06,
+      "loss": 0.527,
+      "step": 3863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5711819628586476,
+      "learning_rate": 8.163817991274982e-06,
+      "loss": 0.5375,
+      "step": 3864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2545146168243377,
+      "learning_rate": 8.16283307934211e-06,
+      "loss": 0.5475,
+      "step": 3865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5406114236179673,
+      "learning_rate": 8.161847962778076e-06,
+      "loss": 0.4542,
+      "step": 3866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5718046365712872,
+      "learning_rate": 8.160862641646613e-06,
+      "loss": 0.5388,
+      "step": 3867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.2814309788380807,
+      "learning_rate": 8.159877116011474e-06,
+      "loss": 0.4825,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5820226386026355,
+      "learning_rate": 8.158891385936418e-06,
+      "loss": 0.5391,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5797904228126589,
+      "learning_rate": 8.157905451485223e-06,
+      "loss": 0.5475,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6107546435355637,
+      "learning_rate": 8.156919312721676e-06,
+      "loss": 0.4342,
+      "step": 3871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4299452585662809,
+      "learning_rate": 8.155932969709578e-06,
+      "loss": 0.4725,
+      "step": 3872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.343049223885057,
+      "learning_rate": 8.154946422512746e-06,
+      "loss": 0.4741,
+      "step": 3873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5863282563517187,
+      "learning_rate": 8.153959671195007e-06,
+      "loss": 0.5352,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4295445943371536,
+      "learning_rate": 8.152972715820205e-06,
+      "loss": 0.455,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.548616259170812,
+      "learning_rate": 8.15198555645219e-06,
+      "loss": 0.5071,
+      "step": 3876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.9959273704980987,
+      "learning_rate": 8.150998193154837e-06,
+      "loss": 0.4812,
+      "step": 3877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.5623507836521062,
+      "learning_rate": 8.15001062599202e-06,
+      "loss": 0.5118,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3910780237392357,
+      "learning_rate": 8.149022855027637e-06,
+      "loss": 0.4703,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.0730751751257364,
+      "learning_rate": 8.148034880325594e-06,
+      "loss": 0.522,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4561573403086012,
+      "learning_rate": 8.147046701949813e-06,
+      "loss": 0.4857,
+      "step": 3881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.519844974674417,
+      "learning_rate": 8.146058319964227e-06,
+      "loss": 0.5171,
+      "step": 3882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.804916963985851,
+      "learning_rate": 8.145069734432783e-06,
+      "loss": 0.5151,
+      "step": 3883
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.251861913686813,
+      "learning_rate": 8.144080945419442e-06,
+      "loss": 0.4981,
+      "step": 3884
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4806879075437955,
+      "learning_rate": 8.143091952988177e-06,
+      "loss": 0.4983,
+      "step": 3885
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5013385453517676,
+      "learning_rate": 8.142102757202974e-06,
+      "loss": 0.5358,
+      "step": 3886
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8227615191915136,
+      "learning_rate": 8.141113358127832e-06,
+      "loss": 0.466,
+      "step": 3887
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3493649961755132,
+      "learning_rate": 8.140123755826767e-06,
+      "loss": 0.5247,
+      "step": 3888
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.554141936756321,
+      "learning_rate": 8.139133950363801e-06,
+      "loss": 0.4931,
+      "step": 3889
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4506192565023577,
+      "learning_rate": 8.138143941802976e-06,
+      "loss": 0.4022,
+      "step": 3890
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.1867546974283334,
+      "learning_rate": 8.137153730208342e-06,
+      "loss": 0.4707,
+      "step": 3891
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6824446867247185,
+      "learning_rate": 8.136163315643967e-06,
+      "loss": 0.552,
+      "step": 3892
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3110787561600197,
+      "learning_rate": 8.135172698173927e-06,
+      "loss": 0.4846,
+      "step": 3893
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2247973196013973,
+      "learning_rate": 8.134181877862314e-06,
+      "loss": 0.4551,
+      "step": 3894
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.303755765503516,
+      "learning_rate": 8.133190854773236e-06,
+      "loss": 0.5073,
+      "step": 3895
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8036928722802792,
+      "learning_rate": 8.132199628970807e-06,
+      "loss": 0.503,
+      "step": 3896
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3264475732552545,
+      "learning_rate": 8.13120820051916e-06,
+      "loss": 0.4831,
+      "step": 3897
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5463707597181916,
+      "learning_rate": 8.130216569482437e-06,
+      "loss": 0.5166,
+      "step": 3898
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.893728645837672,
+      "learning_rate": 8.129224735924799e-06,
+      "loss": 0.451,
+      "step": 3899
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5684010132426125,
+      "learning_rate": 8.128232699910413e-06,
+      "loss": 0.5292,
+      "step": 3900
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.778562287876687,
+      "learning_rate": 8.127240461503462e-06,
+      "loss": 0.4617,
+      "step": 3901
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.639141589732679,
+      "learning_rate": 8.126248020768147e-06,
+      "loss": 0.4938,
+      "step": 3902
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7457781284647416,
+      "learning_rate": 8.125255377768673e-06,
+      "loss": 0.5333,
+      "step": 3903
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3348458187743864,
+      "learning_rate": 8.124262532569264e-06,
+      "loss": 0.5035,
+      "step": 3904
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.408955329674299,
+      "learning_rate": 8.123269485234158e-06,
+      "loss": 0.4953,
+      "step": 3905
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.711485312422972,
+      "learning_rate": 8.122276235827599e-06,
+      "loss": 0.4783,
+      "step": 3906
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4131855719125865,
+      "learning_rate": 8.121282784413855e-06,
+      "loss": 0.4645,
+      "step": 3907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7676569957637467,
+      "learning_rate": 8.120289131057197e-06,
+      "loss": 0.5185,
+      "step": 3908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.769750240344266,
+      "learning_rate": 8.119295275821915e-06,
+      "loss": 0.4797,
+      "step": 3909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.48676382256224,
+      "learning_rate": 8.118301218772308e-06,
+      "loss": 0.5312,
+      "step": 3910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3260857886124575,
+      "learning_rate": 8.117306959972693e-06,
+      "loss": 0.4967,
+      "step": 3911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4472440142556946,
+      "learning_rate": 8.116312499487394e-06,
+      "loss": 0.506,
+      "step": 3912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6406276101954045,
+      "learning_rate": 8.115317837380753e-06,
+      "loss": 0.5193,
+      "step": 3913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6378677841241669,
+      "learning_rate": 8.11432297371712e-06,
+      "loss": 0.542,
+      "step": 3914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4963393824989704,
+      "learning_rate": 8.113327908560871e-06,
+      "loss": 0.5089,
+      "step": 3915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3355187570639697,
+      "learning_rate": 8.112332641976375e-06,
+      "loss": 0.4929,
+      "step": 3916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.595812373858805,
+      "learning_rate": 8.11133717402803e-06,
+      "loss": 0.4852,
+      "step": 3917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4333971838555157,
+      "learning_rate": 8.110341504780238e-06,
+      "loss": 0.5082,
+      "step": 3918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5660988748728288,
+      "learning_rate": 8.109345634297423e-06,
+      "loss": 0.5408,
+      "step": 3919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.051466671740846,
+      "learning_rate": 8.108349562644011e-06,
+      "loss": 0.4764,
+      "step": 3920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.419459554688049,
+      "learning_rate": 8.10735328988445e-06,
+      "loss": 0.517,
+      "step": 3921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.456020850581799,
+      "learning_rate": 8.106356816083194e-06,
+      "loss": 0.4489,
+      "step": 3922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8954103689059125,
+      "learning_rate": 8.105360141304717e-06,
+      "loss": 0.5308,
+      "step": 3923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5403968314874155,
+      "learning_rate": 8.1043632656135e-06,
+      "loss": 0.5357,
+      "step": 3924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5275513451836822,
+      "learning_rate": 8.10336618907404e-06,
+      "loss": 0.4441,
+      "step": 3925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.340830977691936,
+      "learning_rate": 8.102368911750848e-06,
+      "loss": 0.478,
+      "step": 3926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.818413522030533,
+      "learning_rate": 8.101371433708447e-06,
+      "loss": 0.5276,
+      "step": 3927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.063325256157453,
+      "learning_rate": 8.10037375501137e-06,
+      "loss": 0.4958,
+      "step": 3928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.461162916087229,
+      "learning_rate": 8.09937587572417e-06,
+      "loss": 0.5142,
+      "step": 3929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6089368528920278,
+      "learning_rate": 8.098377795911403e-06,
+      "loss": 0.4468,
+      "step": 3930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.851637291639948,
+      "learning_rate": 8.097379515637645e-06,
+      "loss": 0.4737,
+      "step": 3931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.660925011614903,
+      "learning_rate": 8.096381034967484e-06,
+      "loss": 0.4808,
+      "step": 3932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.137395886194378,
+      "learning_rate": 8.095382353965522e-06,
+      "loss": 0.4199,
+      "step": 3933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.656377532746799,
+      "learning_rate": 8.09438347269637e-06,
+      "loss": 0.4435,
+      "step": 3934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6545612999235908,
+      "learning_rate": 8.093384391224656e-06,
+      "loss": 0.4926,
+      "step": 3935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8533518144415526,
+      "learning_rate": 8.092385109615018e-06,
+      "loss": 0.4658,
+      "step": 3936
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.533442340124059,
+      "learning_rate": 8.09138562793211e-06,
+      "loss": 0.4684,
+      "step": 3937
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5272413559902536,
+      "learning_rate": 8.090385946240596e-06,
+      "loss": 0.5061,
+      "step": 3938
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3728065026439058,
+      "learning_rate": 8.089386064605152e-06,
+      "loss": 0.4373,
+      "step": 3939
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5121683472675844,
+      "learning_rate": 8.088385983090472e-06,
+      "loss": 0.5122,
+      "step": 3940
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.823147283797327,
+      "learning_rate": 8.08738570176126e-06,
+      "loss": 0.4732,
+      "step": 3941
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5196865070597456,
+      "learning_rate": 8.08638522068223e-06,
+      "loss": 0.5343,
+      "step": 3942
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.515294698986381,
+      "learning_rate": 8.085384539918115e-06,
+      "loss": 0.487,
+      "step": 3943
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5931969491253701,
+      "learning_rate": 8.084383659533656e-06,
+      "loss": 0.5443,
+      "step": 3944
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.489658439824186,
+      "learning_rate": 8.083382579593609e-06,
+      "loss": 0.4562,
+      "step": 3945
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5641128499602256,
+      "learning_rate": 8.082381300162742e-06,
+      "loss": 0.515,
+      "step": 3946
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3511330034842506,
+      "learning_rate": 8.081379821305839e-06,
+      "loss": 0.4535,
+      "step": 3947
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4195059085133939,
+      "learning_rate": 8.080378143087691e-06,
+      "loss": 0.4205,
+      "step": 3948
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4411741243397362,
+      "learning_rate": 8.079376265573108e-06,
+      "loss": 0.5073,
+      "step": 3949
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.771786722237853,
+      "learning_rate": 8.078374188826908e-06,
+      "loss": 0.5236,
+      "step": 3950
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4100060925431146,
+      "learning_rate": 8.077371912913925e-06,
+      "loss": 0.4942,
+      "step": 3951
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 6.005872019808927,
+      "learning_rate": 8.076369437899005e-06,
+      "loss": 0.5223,
+      "step": 3952
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5934552333323317,
+      "learning_rate": 8.075366763847008e-06,
+      "loss": 0.5223,
+      "step": 3953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5727084035372247,
+      "learning_rate": 8.074363890822805e-06,
+      "loss": 0.5332,
+      "step": 3954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4440783099165317,
+      "learning_rate": 8.073360818891277e-06,
+      "loss": 0.4684,
+      "step": 3955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 3.617780973445932,
+      "learning_rate": 8.072357548117327e-06,
+      "loss": 0.5132,
+      "step": 3956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.778971819131723,
+      "learning_rate": 8.071354078565861e-06,
+      "loss": 0.4564,
+      "step": 3957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2262988968925628,
+      "learning_rate": 8.070350410301806e-06,
+      "loss": 0.4675,
+      "step": 3958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.336479118949998,
+      "learning_rate": 8.069346543390092e-06,
+      "loss": 0.4853,
+      "step": 3959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4500891391267445,
+      "learning_rate": 8.068342477895676e-06,
+      "loss": 0.575,
+      "step": 3960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4272318250308782,
+      "learning_rate": 8.067338213883514e-06,
+      "loss": 0.4591,
+      "step": 3961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4704389775380629,
+      "learning_rate": 8.066333751418582e-06,
+      "loss": 0.4315,
+      "step": 3962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.563831687537641,
+      "learning_rate": 8.065329090565867e-06,
+      "loss": 0.4702,
+      "step": 3963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.446301365462807,
+      "learning_rate": 8.06432423139037e-06,
+      "loss": 0.506,
+      "step": 3964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9195032468143451,
+      "learning_rate": 8.063319173957106e-06,
+      "loss": 0.5254,
+      "step": 3965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2617077807602677,
+      "learning_rate": 8.062313918331096e-06,
+      "loss": 0.4814,
+      "step": 3966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.752359238963064,
+      "learning_rate": 8.061308464577384e-06,
+      "loss": 0.5358,
+      "step": 3967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3556315228546831,
+      "learning_rate": 8.060302812761019e-06,
+      "loss": 0.4699,
+      "step": 3968
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4702491916225844,
+      "learning_rate": 8.059296962947063e-06,
+      "loss": 0.5372,
+      "step": 3969
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4197051836723409,
+      "learning_rate": 8.058290915200597e-06,
+      "loss": 0.4997,
+      "step": 3970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5095470020502209,
+      "learning_rate": 8.057284669586708e-06,
+      "loss": 0.5256,
+      "step": 3971
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.346212819489792,
+      "learning_rate": 8.056278226170502e-06,
+      "loss": 0.4624,
+      "step": 3972
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7351039199927025,
+      "learning_rate": 8.055271585017093e-06,
+      "loss": 0.4542,
+      "step": 3973
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 4.4795864308101665,
+      "learning_rate": 8.05426474619161e-06,
+      "loss": 0.4662,
+      "step": 3974
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.415041978384465,
+      "learning_rate": 8.053257709759192e-06,
+      "loss": 0.4914,
+      "step": 3975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5959136386357695,
+      "learning_rate": 8.052250475784994e-06,
+      "loss": 0.4959,
+      "step": 3976
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 5.830099206214093,
+      "learning_rate": 8.051243044334183e-06,
+      "loss": 0.5271,
+      "step": 3977
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3714002596255979,
+      "learning_rate": 8.05023541547194e-06,
+      "loss": 0.4819,
+      "step": 3978
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9527696763000346,
+      "learning_rate": 8.049227589263455e-06,
+      "loss": 0.4953,
+      "step": 3979
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2653000413254794,
+      "learning_rate": 8.048219565773933e-06,
+      "loss": 0.4589,
+      "step": 3980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.445510884249527,
+      "learning_rate": 8.047211345068593e-06,
+      "loss": 0.4567,
+      "step": 3981
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6162513895404875,
+      "learning_rate": 8.046202927212666e-06,
+      "loss": 0.4997,
+      "step": 3982
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.924063816676241,
+      "learning_rate": 8.045194312271394e-06,
+      "loss": 0.4363,
+      "step": 3983
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.5403296061144753,
+      "learning_rate": 8.044185500310035e-06,
+      "loss": 0.4993,
+      "step": 3984
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3187906774747356,
+      "learning_rate": 8.043176491393854e-06,
+      "loss": 0.4782,
+      "step": 3985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6352581509118242,
+      "learning_rate": 8.042167285588138e-06,
+      "loss": 0.4893,
+      "step": 3986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3762239664975533,
+      "learning_rate": 8.041157882958175e-06,
+      "loss": 0.4749,
+      "step": 3987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7913327897967334,
+      "learning_rate": 8.040148283569278e-06,
+      "loss": 0.4996,
+      "step": 3988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4140470054494674,
+      "learning_rate": 8.039138487486763e-06,
+      "loss": 0.5091,
+      "step": 3989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6089841660217014,
+      "learning_rate": 8.038128494775963e-06,
+      "loss": 0.5393,
+      "step": 3990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4196991359435396,
+      "learning_rate": 8.037118305502225e-06,
+      "loss": 0.4825,
+      "step": 3991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.3621940982344363,
+      "learning_rate": 8.036107919730905e-06,
+      "loss": 0.5339,
+      "step": 3992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6826065198307762,
+      "learning_rate": 8.035097337527373e-06,
+      "loss": 0.4805,
+      "step": 3993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7624362305606989,
+      "learning_rate": 8.034086558957015e-06,
+      "loss": 0.4695,
+      "step": 3994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4831963561039059,
+      "learning_rate": 8.033075584085226e-06,
+      "loss": 0.4495,
+      "step": 3995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.1797254923129734,
+      "learning_rate": 8.032064412977414e-06,
+      "loss": 0.4771,
+      "step": 3996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4435024241433796,
+      "learning_rate": 8.031053045699001e-06,
+      "loss": 0.4446,
+      "step": 3997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8575762281620587,
+      "learning_rate": 8.03004148231542e-06,
+      "loss": 0.5181,
+      "step": 3998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5553558599418429,
+      "learning_rate": 8.02902972289212e-06,
+      "loss": 0.4651,
+      "step": 3999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7654567060684982,
+      "learning_rate": 8.02801776749456e-06,
+      "loss": 0.5432,
+      "step": 4000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.8135324410278741,
+      "learning_rate": 8.02700561618821e-06,
+      "loss": 0.438,
+      "step": 4001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4933977045571443,
+      "learning_rate": 8.025993269038559e-06,
+      "loss": 0.4316,
+      "step": 4002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5556543705160437,
+      "learning_rate": 8.0249807261111e-06,
+      "loss": 0.5079,
+      "step": 4003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3212649926277829,
+      "learning_rate": 8.023967987471345e-06,
+      "loss": 0.4764,
+      "step": 4004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6185360599909464,
+      "learning_rate": 8.022955053184817e-06,
+      "loss": 0.5178,
+      "step": 4005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2874683793707153,
+      "learning_rate": 8.021941923317052e-06,
+      "loss": 0.4329,
+      "step": 4006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5825161187806117,
+      "learning_rate": 8.0209285979336e-06,
+      "loss": 0.5173,
+      "step": 4007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.2240869329810695,
+      "learning_rate": 8.019915077100017e-06,
+      "loss": 0.5074,
+      "step": 4008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3557927568261527,
+      "learning_rate": 8.018901360881878e-06,
+      "loss": 0.5423,
+      "step": 4009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5797679297783245,
+      "learning_rate": 8.017887449344773e-06,
+      "loss": 0.5255,
+      "step": 4010
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4341721140664487,
+      "learning_rate": 8.016873342554297e-06,
+      "loss": 0.5024,
+      "step": 4011
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3476614938032836,
+      "learning_rate": 8.015859040576061e-06,
+      "loss": 0.4787,
+      "step": 4012
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6856451763070877,
+      "learning_rate": 8.014844543475692e-06,
+      "loss": 0.4893,
+      "step": 4013
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4816445109298362,
+      "learning_rate": 8.013829851318824e-06,
+      "loss": 0.4493,
+      "step": 4014
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4645156622712026,
+      "learning_rate": 8.012814964171108e-06,
+      "loss": 0.543,
+      "step": 4015
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3968633448675236,
+      "learning_rate": 8.011799882098203e-06,
+      "loss": 0.4882,
+      "step": 4016
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4496473218442163,
+      "learning_rate": 8.010784605165788e-06,
+      "loss": 0.4956,
+      "step": 4017
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.677937874553818,
+      "learning_rate": 8.009769133439547e-06,
+      "loss": 0.4907,
+      "step": 4018
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6108335982554403,
+      "learning_rate": 8.00875346698518e-06,
+      "loss": 0.5453,
+      "step": 4019
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.744973975194409,
+      "learning_rate": 8.0077376058684e-06,
+      "loss": 0.4503,
+      "step": 4020
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3769044324550819,
+      "learning_rate": 8.006721550154933e-06,
+      "loss": 0.435,
+      "step": 4021
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4299208038463596,
+      "learning_rate": 8.005705299910511e-06,
+      "loss": 0.4808,
+      "step": 4022
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6779006591836431,
+      "learning_rate": 8.004688855200891e-06,
+      "loss": 0.5359,
+      "step": 4023
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.298146039148745,
+      "learning_rate": 8.003672216091833e-06,
+      "loss": 0.4982,
+      "step": 4024
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3973684522359757,
+      "learning_rate": 8.00265538264911e-06,
+      "loss": 0.4803,
+      "step": 4025
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3337699701269947,
+      "learning_rate": 8.001638354938513e-06,
+      "loss": 0.4555,
+      "step": 4026
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4889425417126576,
+      "learning_rate": 8.00062113302584e-06,
+      "loss": 0.5169,
+      "step": 4027
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3237428997160223,
+      "learning_rate": 7.999603716976905e-06,
+      "loss": 0.4153,
+      "step": 4028
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4300675307179977,
+      "learning_rate": 7.998586106857535e-06,
+      "loss": 0.4315,
+      "step": 4029
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5903321934068957,
+      "learning_rate": 7.997568302733565e-06,
+      "loss": 0.5379,
+      "step": 4030
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.1823573858325938,
+      "learning_rate": 7.99655030467085e-06,
+      "loss": 0.4919,
+      "step": 4031
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.545849113162527,
+      "learning_rate": 7.995532112735246e-06,
+      "loss": 0.4596,
+      "step": 4032
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5379241688724287,
+      "learning_rate": 7.994513726992636e-06,
+      "loss": 0.5337,
+      "step": 4033
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5399306736940116,
+      "learning_rate": 7.993495147508903e-06,
+      "loss": 0.5105,
+      "step": 4034
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6066289331093364,
+      "learning_rate": 7.992476374349951e-06,
+      "loss": 0.4616,
+      "step": 4035
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4813171993532492,
+      "learning_rate": 7.991457407581694e-06,
+      "loss": 0.4653,
+      "step": 4036
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0194807013814557,
+      "learning_rate": 7.990438247270054e-06,
+      "loss": 0.517,
+      "step": 4037
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.787015014738175,
+      "learning_rate": 7.98941889348097e-06,
+      "loss": 0.4636,
+      "step": 4038
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6758355803821703,
+      "learning_rate": 7.988399346280398e-06,
+      "loss": 0.5247,
+      "step": 4039
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4218295312392741,
+      "learning_rate": 7.987379605734296e-06,
+      "loss": 0.5055,
+      "step": 4040
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5629268249757284,
+      "learning_rate": 7.98635967190864e-06,
+      "loss": 0.4817,
+      "step": 4041
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5844282940754906,
+      "learning_rate": 7.985339544869422e-06,
+      "loss": 0.5095,
+      "step": 4042
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6160681149547396,
+      "learning_rate": 7.98431922468264e-06,
+      "loss": 0.5357,
+      "step": 4043
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2086289661480742,
+      "learning_rate": 7.983298711414307e-06,
+      "loss": 0.5146,
+      "step": 4044
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7007609338670977,
+      "learning_rate": 7.982278005130451e-06,
+      "loss": 0.4541,
+      "step": 4045
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5204710622253985,
+      "learning_rate": 7.98125710589711e-06,
+      "loss": 0.4865,
+      "step": 4046
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6962565038551431,
+      "learning_rate": 7.980236013780334e-06,
+      "loss": 0.5283,
+      "step": 4047
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5729092323143142,
+      "learning_rate": 7.979214728846186e-06,
+      "loss": 0.5215,
+      "step": 4048
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4092739003400256,
+      "learning_rate": 7.978193251160743e-06,
+      "loss": 0.4749,
+      "step": 4049
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 3.135822221198174,
+      "learning_rate": 7.977171580790091e-06,
+      "loss": 0.5167,
+      "step": 4050
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.102199759484227,
+      "learning_rate": 7.976149717800331e-06,
+      "loss": 0.4938,
+      "step": 4051
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.543188176397457,
+      "learning_rate": 7.975127662257582e-06,
+      "loss": 0.5097,
+      "step": 4052
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3828036901278216,
+      "learning_rate": 7.974105414227964e-06,
+      "loss": 0.4502,
+      "step": 4053
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6809879685523046,
+      "learning_rate": 7.973082973777615e-06,
+      "loss": 0.4593,
+      "step": 4054
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3343258451484006,
+      "learning_rate": 7.972060340972688e-06,
+      "loss": 0.4858,
+      "step": 4055
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5497732681412701,
+      "learning_rate": 7.971037515879343e-06,
+      "loss": 0.499,
+      "step": 4056
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6176892427428313,
+      "learning_rate": 7.97001449856376e-06,
+      "loss": 0.5055,
+      "step": 4057
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5677191456024067,
+      "learning_rate": 7.968991289092123e-06,
+      "loss": 0.4231,
+      "step": 4058
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.212052602837464,
+      "learning_rate": 7.967967887530631e-06,
+      "loss": 0.4456,
+      "step": 4059
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6576546195564046,
+      "learning_rate": 7.966944293945503e-06,
+      "loss": 0.4715,
+      "step": 4060
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.69966167920983,
+      "learning_rate": 7.965920508402959e-06,
+      "loss": 0.5033,
+      "step": 4061
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3995408532356108,
+      "learning_rate": 7.964896530969237e-06,
+      "loss": 0.5359,
+      "step": 4062
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6905069407766817,
+      "learning_rate": 7.963872361710589e-06,
+      "loss": 0.5157,
+      "step": 4063
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.3803825133647822,
+      "learning_rate": 7.962848000693277e-06,
+      "loss": 0.4781,
+      "step": 4064
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4286237377798023,
+      "learning_rate": 7.961823447983576e-06,
+      "loss": 0.4873,
+      "step": 4065
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8902129789726683,
+      "learning_rate": 7.96079870364777e-06,
+      "loss": 0.519,
+      "step": 4066
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.504021132366671,
+      "learning_rate": 7.959773767752163e-06,
+      "loss": 0.5092,
+      "step": 4067
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5171075836029009,
+      "learning_rate": 7.958748640363065e-06,
+      "loss": 0.436,
+      "step": 4068
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8020376946163292,
+      "learning_rate": 7.9577233215468e-06,
+      "loss": 0.5203,
+      "step": 4069
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3814921560417932,
+      "learning_rate": 7.956697811369704e-06,
+      "loss": 0.4326,
+      "step": 4070
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.629158151940029,
+      "learning_rate": 7.95567210989813e-06,
+      "loss": 0.4273,
+      "step": 4071
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2891253925793804,
+      "learning_rate": 7.954646217198434e-06,
+      "loss": 0.5013,
+      "step": 4072
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6044810417299975,
+      "learning_rate": 7.953620133336995e-06,
+      "loss": 0.4921,
+      "step": 4073
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5442885447453165,
+      "learning_rate": 7.952593858380197e-06,
+      "loss": 0.5251,
+      "step": 4074
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6676853794346922,
+      "learning_rate": 7.951567392394438e-06,
+      "loss": 0.4753,
+      "step": 4075
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5889792752662136,
+      "learning_rate": 7.95054073544613e-06,
+      "loss": 0.5237,
+      "step": 4076
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4807186295379637,
+      "learning_rate": 7.949513887601698e-06,
+      "loss": 0.4915,
+      "step": 4077
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3959521919083293,
+      "learning_rate": 7.948486848927574e-06,
+      "loss": 0.483,
+      "step": 4078
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5266533859434668,
+      "learning_rate": 7.947459619490208e-06,
+      "loss": 0.4874,
+      "step": 4079
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6435249261332197,
+      "learning_rate": 7.946432199356062e-06,
+      "loss": 0.4804,
+      "step": 4080
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4526072170466304,
+      "learning_rate": 7.945404588591605e-06,
+      "loss": 0.4896,
+      "step": 4081
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2328381719751407,
+      "learning_rate": 7.944376787263327e-06,
+      "loss": 0.503,
+      "step": 4082
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.229904715614715,
+      "learning_rate": 7.94334879543772e-06,
+      "loss": 0.4746,
+      "step": 4083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6103177866547165,
+      "learning_rate": 7.942320613181296e-06,
+      "loss": 0.5004,
+      "step": 4084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5857711224671605,
+      "learning_rate": 7.941292240560579e-06,
+      "loss": 0.4634,
+      "step": 4085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.602692541470008,
+      "learning_rate": 7.940263677642102e-06,
+      "loss": 0.4779,
+      "step": 4086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.0757221030325366,
+      "learning_rate": 7.93923492449241e-06,
+      "loss": 0.4905,
+      "step": 4087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3615527604840532,
+      "learning_rate": 7.938205981178065e-06,
+      "loss": 0.4486,
+      "step": 4088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6727827033567302,
+      "learning_rate": 7.937176847765636e-06,
+      "loss": 0.5394,
+      "step": 4089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2473445348295324,
+      "learning_rate": 7.936147524321708e-06,
+      "loss": 0.466,
+      "step": 4090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2816778775383308,
+      "learning_rate": 7.935118010912873e-06,
+      "loss": 0.4993,
+      "step": 4091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.440629597104561,
+      "learning_rate": 7.934088307605745e-06,
+      "loss": 0.512,
+      "step": 4092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.568559114025573,
+      "learning_rate": 7.93305841446694e-06,
+      "loss": 0.539,
+      "step": 4093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9476506654081815,
+      "learning_rate": 7.932028331563095e-06,
+      "loss": 0.4789,
+      "step": 4094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.243029946142988,
+      "learning_rate": 7.93099805896085e-06,
+      "loss": 0.5198,
+      "step": 4095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 11.364859297254679,
+      "learning_rate": 7.929967596726866e-06,
+      "loss": 0.5008,
+      "step": 4096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7778495184017291,
+      "learning_rate": 7.928936944927813e-06,
+      "loss": 0.499,
+      "step": 4097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5856929359383394,
+      "learning_rate": 7.927906103630368e-06,
+      "loss": 0.4702,
+      "step": 4098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7069058079254769,
+      "learning_rate": 7.92687507290123e-06,
+      "loss": 0.5184,
+      "step": 4099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7710637362504889,
+      "learning_rate": 7.925843852807105e-06,
+      "loss": 0.4935,
+      "step": 4100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4560907142158446,
+      "learning_rate": 7.924812443414708e-06,
+      "loss": 0.5715,
+      "step": 4101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3669236718316833,
+      "learning_rate": 7.923780844790772e-06,
+      "loss": 0.4468,
+      "step": 4102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.367754917310993,
+      "learning_rate": 7.922749057002041e-06,
+      "loss": 0.4667,
+      "step": 4103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.403031543273243,
+      "learning_rate": 7.92171708011527e-06,
+      "loss": 0.501,
+      "step": 4104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2707390845414877,
+      "learning_rate": 7.920684914197225e-06,
+      "loss": 0.467,
+      "step": 4105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5582544157719762,
+      "learning_rate": 7.919652559314686e-06,
+      "loss": 0.5217,
+      "step": 4106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4107535433494207,
+      "learning_rate": 7.91862001553445e-06,
+      "loss": 0.5294,
+      "step": 4107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8293405497207536,
+      "learning_rate": 7.917587282923312e-06,
+      "loss": 0.518,
+      "step": 4108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4202769630295764,
+      "learning_rate": 7.916554361548094e-06,
+      "loss": 0.491,
+      "step": 4109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4502692037290117,
+      "learning_rate": 7.915521251475627e-06,
+      "loss": 0.4469,
+      "step": 4110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2694529812182196,
+      "learning_rate": 7.914487952772747e-06,
+      "loss": 0.4815,
+      "step": 4111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.866413259977874,
+      "learning_rate": 7.91345446550631e-06,
+      "loss": 0.4752,
+      "step": 4112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6136470799592084,
+      "learning_rate": 7.91242078974318e-06,
+      "loss": 0.506,
+      "step": 4113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.436527970906703,
+      "learning_rate": 7.911386925550235e-06,
+      "loss": 0.4589,
+      "step": 4114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3313482782794848,
+      "learning_rate": 7.910352872994365e-06,
+      "loss": 0.4992,
+      "step": 4115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3311467914392403,
+      "learning_rate": 7.90931863214247e-06,
+      "loss": 0.4249,
+      "step": 4116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5781367219714921,
+      "learning_rate": 7.908284203061466e-06,
+      "loss": 0.5454,
+      "step": 4117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6552780352709309,
+      "learning_rate": 7.907249585818278e-06,
+      "loss": 0.5042,
+      "step": 4118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.4229778448762045,
+      "learning_rate": 7.906214780479846e-06,
+      "loss": 0.4979,
+      "step": 4119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5765427392624005,
+      "learning_rate": 7.90517978711312e-06,
+      "loss": 0.5154,
+      "step": 4120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6009325451381912,
+      "learning_rate": 7.90414460578506e-06,
+      "loss": 0.5371,
+      "step": 4121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.8050804278266164,
+      "learning_rate": 7.903109236562645e-06,
+      "loss": 0.4676,
+      "step": 4122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.813375924674175,
+      "learning_rate": 7.902073679512859e-06,
+      "loss": 0.4897,
+      "step": 4123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.046627447310638,
+      "learning_rate": 7.901037934702703e-06,
+      "loss": 0.4893,
+      "step": 4124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6664566752430592,
+      "learning_rate": 7.900002002199188e-06,
+      "loss": 0.5201,
+      "step": 4125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.9219164849935226,
+      "learning_rate": 7.898965882069336e-06,
+      "loss": 0.4968,
+      "step": 4126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.125292604415674,
+      "learning_rate": 7.897929574380186e-06,
+      "loss": 0.4609,
+      "step": 4127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 6.824812098885272,
+      "learning_rate": 7.89689307919878e-06,
+      "loss": 0.4861,
+      "step": 4128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6624107478692396,
+      "learning_rate": 7.895856396592183e-06,
+      "loss": 0.4713,
+      "step": 4129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2590290735866265,
+      "learning_rate": 7.894819526627466e-06,
+      "loss": 0.4662,
+      "step": 4130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7029828802828137,
+      "learning_rate": 7.893782469371713e-06,
+      "loss": 0.4792,
+      "step": 4131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4832039434813225,
+      "learning_rate": 7.89274522489202e-06,
+      "loss": 0.5066,
+      "step": 4132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.7773272716599027,
+      "learning_rate": 7.891707793255493e-06,
+      "loss": 0.5121,
+      "step": 4133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3660085678181846,
+      "learning_rate": 7.890670174529255e-06,
+      "loss": 0.4616,
+      "step": 4134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.501142812654778,
+      "learning_rate": 7.88963236878044e-06,
+      "loss": 0.4622,
+      "step": 4135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5636206448109682,
+      "learning_rate": 7.88859437607619e-06,
+      "loss": 0.5202,
+      "step": 4136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6692075127352977,
+      "learning_rate": 7.887556196483663e-06,
+      "loss": 0.4955,
+      "step": 4137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.6988705188011985,
+      "learning_rate": 7.886517830070026e-06,
+      "loss": 0.4837,
+      "step": 4138
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0673187366579593,
+      "learning_rate": 7.885479276902464e-06,
+      "loss": 0.4766,
+      "step": 4139
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4098061237902029,
+      "learning_rate": 7.884440537048163e-06,
+      "loss": 0.4985,
+      "step": 4140
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4924026645001316,
+      "learning_rate": 7.883401610574338e-06,
+      "loss": 0.489,
+      "step": 4141
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4573804922750397,
+      "learning_rate": 7.882362497548197e-06,
+      "loss": 0.4869,
+      "step": 4142
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4517361199423946,
+      "learning_rate": 7.881323198036974e-06,
+      "loss": 0.4436,
+      "step": 4143
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6590676405695403,
+      "learning_rate": 7.88028371210791e-06,
+      "loss": 0.5504,
+      "step": 4144
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8019173328002358,
+      "learning_rate": 7.879244039828256e-06,
+      "loss": 0.531,
+      "step": 4145
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4937530135517516,
+      "learning_rate": 7.87820418126528e-06,
+      "loss": 0.5183,
+      "step": 4146
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5752403055221513,
+      "learning_rate": 7.877164136486259e-06,
+      "loss": 0.4899,
+      "step": 4147
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6243703146721367,
+      "learning_rate": 7.876123905558484e-06,
+      "loss": 0.4902,
+      "step": 4148
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.597020683694799,
+      "learning_rate": 7.875083488549253e-06,
+      "loss": 0.4661,
+      "step": 4149
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.502917546029407,
+      "learning_rate": 7.874042885525881e-06,
+      "loss": 0.5115,
+      "step": 4150
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 7.689924761462897,
+      "learning_rate": 7.873002096555692e-06,
+      "loss": 0.4922,
+      "step": 4151
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.695146824475652,
+      "learning_rate": 7.871961121706029e-06,
+      "loss": 0.4588,
+      "step": 4152
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.70957677741513,
+      "learning_rate": 7.870919961044237e-06,
+      "loss": 0.4678,
+      "step": 4153
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6214926425123655,
+      "learning_rate": 7.86987861463768e-06,
+      "loss": 0.4847,
+      "step": 4154
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.995431595030365,
+      "learning_rate": 7.868837082553731e-06,
+      "loss": 0.4929,
+      "step": 4155
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0198959432674877,
+      "learning_rate": 7.867795364859775e-06,
+      "loss": 0.5065,
+      "step": 4156
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4971321604119767,
+      "learning_rate": 7.86675346162321e-06,
+      "loss": 0.4795,
+      "step": 4157
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2395556283901006,
+      "learning_rate": 7.865711372911447e-06,
+      "loss": 0.5029,
+      "step": 4158
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6957484652427492,
+      "learning_rate": 7.864669098791909e-06,
+      "loss": 0.46,
+      "step": 4159
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4358837636127295,
+      "learning_rate": 7.863626639332025e-06,
+      "loss": 0.5337,
+      "step": 4160
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6916769749781666,
+      "learning_rate": 7.862583994599243e-06,
+      "loss": 0.5001,
+      "step": 4161
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9731599251986733,
+      "learning_rate": 7.861541164661021e-06,
+      "loss": 0.5139,
+      "step": 4162
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4999651072760727,
+      "learning_rate": 7.860498149584833e-06,
+      "loss": 0.4858,
+      "step": 4163
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.387618659334892,
+      "learning_rate": 7.859454949438152e-06,
+      "loss": 0.4541,
+      "step": 4164
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.564276830578624,
+      "learning_rate": 7.85841156428848e-06,
+      "loss": 0.5416,
+      "step": 4165
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4098974690192476,
+      "learning_rate": 7.857367994203318e-06,
+      "loss": 0.4957,
+      "step": 4166
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4561781067742248,
+      "learning_rate": 7.856324239250184e-06,
+      "loss": 0.4762,
+      "step": 4167
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.719403539111808,
+      "learning_rate": 7.855280299496608e-06,
+      "loss": 0.5103,
+      "step": 4168
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6596030629035754,
+      "learning_rate": 7.854236175010133e-06,
+      "loss": 0.4604,
+      "step": 4169
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4979912557762274,
+      "learning_rate": 7.853191865858309e-06,
+      "loss": 0.5129,
+      "step": 4170
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7186507377502518,
+      "learning_rate": 7.852147372108707e-06,
+      "loss": 0.526,
+      "step": 4171
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7678543981418764,
+      "learning_rate": 7.8511026938289e-06,
+      "loss": 0.497,
+      "step": 4172
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.549440712275477,
+      "learning_rate": 7.850057831086477e-06,
+      "loss": 0.5189,
+      "step": 4173
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.350207249731609,
+      "learning_rate": 7.849012783949042e-06,
+      "loss": 0.4733,
+      "step": 4174
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.641963853619446,
+      "learning_rate": 7.847967552484206e-06,
+      "loss": 0.5138,
+      "step": 4175
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6361502759589202,
+      "learning_rate": 7.846922136759595e-06,
+      "loss": 0.548,
+      "step": 4176
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3954468879599369,
+      "learning_rate": 7.845876536842846e-06,
+      "loss": 0.4565,
+      "step": 4177
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4221757639390908,
+      "learning_rate": 7.84483075280161e-06,
+      "loss": 0.4928,
+      "step": 4178
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6486200275215523,
+      "learning_rate": 7.843784784703544e-06,
+      "loss": 0.489,
+      "step": 4179
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6504466518262612,
+      "learning_rate": 7.842738632616322e-06,
+      "loss": 0.512,
+      "step": 4180
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3234827743163595,
+      "learning_rate": 7.841692296607629e-06,
+      "loss": 0.4617,
+      "step": 4181
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4811971666373964,
+      "learning_rate": 7.840645776745165e-06,
+      "loss": 0.4304,
+      "step": 4182
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5150946132739578,
+      "learning_rate": 7.83959907309663e-06,
+      "loss": 0.5101,
+      "step": 4183
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2992113338676683,
+      "learning_rate": 7.838552185729755e-06,
+      "loss": 0.4882,
+      "step": 4184
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4131840329164382,
+      "learning_rate": 7.837505114712262e-06,
+      "loss": 0.5193,
+      "step": 4185
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8245638392446732,
+      "learning_rate": 7.836457860111903e-06,
+      "loss": 0.4864,
+      "step": 4186
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3301221038911126,
+      "learning_rate": 7.835410421996431e-06,
+      "loss": 0.487,
+      "step": 4187
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5461999576983239,
+      "learning_rate": 7.834362800433614e-06,
+      "loss": 0.5376,
+      "step": 4188
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7004356653527166,
+      "learning_rate": 7.83331499549123e-06,
+      "loss": 0.4547,
+      "step": 4189
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6555015208296363,
+      "learning_rate": 7.832267007237072e-06,
+      "loss": 0.5302,
+      "step": 4190
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6593330346118563,
+      "learning_rate": 7.831218835738947e-06,
+      "loss": 0.5242,
+      "step": 4191
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.582080882643477,
+      "learning_rate": 7.830170481064666e-06,
+      "loss": 0.5394,
+      "step": 4192
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6508544385059107,
+      "learning_rate": 7.829121943282055e-06,
+      "loss": 0.5523,
+      "step": 4193
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6742404909920339,
+      "learning_rate": 7.828073222458956e-06,
+      "loss": 0.5545,
+      "step": 4194
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8422896479447675,
+      "learning_rate": 7.827024318663221e-06,
+      "loss": 0.4596,
+      "step": 4195
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5765984288053445,
+      "learning_rate": 7.825975231962708e-06,
+      "loss": 0.4949,
+      "step": 4196
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6570742754139188,
+      "learning_rate": 7.824925962425296e-06,
+      "loss": 0.522,
+      "step": 4197
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7059131431996593,
+      "learning_rate": 7.823876510118868e-06,
+      "loss": 0.4893,
+      "step": 4198
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9774316073254092,
+      "learning_rate": 7.822826875111327e-06,
+      "loss": 0.4876,
+      "step": 4199
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.304397361202217,
+      "learning_rate": 7.821777057470578e-06,
+      "loss": 0.4807,
+      "step": 4200
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6353878235953156,
+      "learning_rate": 7.820727057264545e-06,
+      "loss": 0.5136,
+      "step": 4201
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5906650848994638,
+      "learning_rate": 7.81967687456116e-06,
+      "loss": 0.5059,
+      "step": 4202
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3077828727225753,
+      "learning_rate": 7.818626509428372e-06,
+      "loss": 0.425,
+      "step": 4203
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4211527533835069,
+      "learning_rate": 7.817575961934135e-06,
+      "loss": 0.4787,
+      "step": 4204
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.450364846582341,
+      "learning_rate": 7.816525232146419e-06,
+      "loss": 0.4906,
+      "step": 4205
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6525828027271319,
+      "learning_rate": 7.815474320133204e-06,
+      "loss": 0.5293,
+      "step": 4206
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5980006335119961,
+      "learning_rate": 7.814423225962487e-06,
+      "loss": 0.5143,
+      "step": 4207
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.354456419242298,
+      "learning_rate": 7.813371949702266e-06,
+      "loss": 0.4723,
+      "step": 4208
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4179018384539985,
+      "learning_rate": 7.812320491420562e-06,
+      "loss": 0.5521,
+      "step": 4209
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6035206885314006,
+      "learning_rate": 7.8112688511854e-06,
+      "loss": 0.5421,
+      "step": 4210
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.409198393286427,
+      "learning_rate": 7.81021702906482e-06,
+      "loss": 0.4748,
+      "step": 4211
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3321669454118439,
+      "learning_rate": 7.809165025126876e-06,
+      "loss": 0.4904,
+      "step": 4212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.0578162721387177,
+      "learning_rate": 7.80811283943963e-06,
+      "loss": 0.4991,
+      "step": 4213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4907443018659574,
+      "learning_rate": 7.807060472071156e-06,
+      "loss": 0.5328,
+      "step": 4214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6605616861814372,
+      "learning_rate": 7.80600792308954e-06,
+      "loss": 0.5112,
+      "step": 4215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.365863358456906,
+      "learning_rate": 7.804955192562884e-06,
+      "loss": 0.5211,
+      "step": 4216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.1184132891837684,
+      "learning_rate": 7.803902280559296e-06,
+      "loss": 0.4855,
+      "step": 4217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6669567306168196,
+      "learning_rate": 7.8028491871469e-06,
+      "loss": 0.4572,
+      "step": 4218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9633928821733793,
+      "learning_rate": 7.801795912393826e-06,
+      "loss": 0.5276,
+      "step": 4219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6226621724871384,
+      "learning_rate": 7.800742456368223e-06,
+      "loss": 0.5058,
+      "step": 4220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7335203607416932,
+      "learning_rate": 7.799688819138244e-06,
+      "loss": 0.5392,
+      "step": 4221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3667034946973886,
+      "learning_rate": 7.798635000772063e-06,
+      "loss": 0.4512,
+      "step": 4222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.9457655890353047,
+      "learning_rate": 7.797581001337859e-06,
+      "loss": 0.5087,
+      "step": 4223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.603192431856554,
+      "learning_rate": 7.796526820903824e-06,
+      "loss": 0.482,
+      "step": 4224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5857141457337143,
+      "learning_rate": 7.795472459538163e-06,
+      "loss": 0.4743,
+      "step": 4225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.235978057919025,
+      "learning_rate": 7.794417917309088e-06,
+      "loss": 0.5077,
+      "step": 4226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.498696315947516,
+      "learning_rate": 7.79336319428483e-06,
+      "loss": 0.4989,
+      "step": 4227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3561215025578375,
+      "learning_rate": 7.79230829053363e-06,
+      "loss": 0.491,
+      "step": 4228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5315660631165804,
+      "learning_rate": 7.791253206123734e-06,
+      "loss": 0.4517,
+      "step": 4229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4403277509595056,
+      "learning_rate": 7.790197941123407e-06,
+      "loss": 0.4144,
+      "step": 4230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4247318531650885,
+      "learning_rate": 7.789142495600923e-06,
+      "loss": 0.485,
+      "step": 4231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6847963478019263,
+      "learning_rate": 7.788086869624569e-06,
+      "loss": 0.5396,
+      "step": 4232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4293670010167894,
+      "learning_rate": 7.787031063262643e-06,
+      "loss": 0.5038,
+      "step": 4233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.495505904343361,
+      "learning_rate": 7.785975076583452e-06,
+      "loss": 0.4783,
+      "step": 4234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5435137052644385,
+      "learning_rate": 7.784918909655316e-06,
+      "loss": 0.5084,
+      "step": 4235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.7006228240395367,
+      "learning_rate": 7.783862562546574e-06,
+      "loss": 0.5017,
+      "step": 4236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.359351762023234,
+      "learning_rate": 7.782806035325562e-06,
+      "loss": 0.4868,
+      "step": 4237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6454585815392235,
+      "learning_rate": 7.781749328060642e-06,
+      "loss": 0.5227,
+      "step": 4238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4938601073418323,
+      "learning_rate": 7.780692440820179e-06,
+      "loss": 0.4969,
+      "step": 4239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.191547574108064,
+      "learning_rate": 7.779635373672554e-06,
+      "loss": 0.4857,
+      "step": 4240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.925197255205875,
+      "learning_rate": 7.778578126686154e-06,
+      "loss": 0.4996,
+      "step": 4241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3058125515773447,
+      "learning_rate": 7.777520699929383e-06,
+      "loss": 0.473,
+      "step": 4242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3567887041266828,
+      "learning_rate": 7.776463093470661e-06,
+      "loss": 0.4638,
+      "step": 4243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7039260955090508,
+      "learning_rate": 7.775405307378406e-06,
+      "loss": 0.5267,
+      "step": 4244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8912792826922173,
+      "learning_rate": 7.774347341721058e-06,
+      "loss": 0.4672,
+      "step": 4245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5600752712026751,
+      "learning_rate": 7.773289196567066e-06,
+      "loss": 0.5014,
+      "step": 4246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6699735908437303,
+      "learning_rate": 7.772230871984893e-06,
+      "loss": 0.4986,
+      "step": 4247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8646478285312775,
+      "learning_rate": 7.771172368043008e-06,
+      "loss": 0.4956,
+      "step": 4248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3423309797242764,
+      "learning_rate": 7.770113684809896e-06,
+      "loss": 0.5157,
+      "step": 4249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4599978940784153,
+      "learning_rate": 7.769054822354052e-06,
+      "loss": 0.5383,
+      "step": 4250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.28204164325328,
+      "learning_rate": 7.767995780743985e-06,
+      "loss": 0.4631,
+      "step": 4251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.8765615277182686,
+      "learning_rate": 7.766936560048209e-06,
+      "loss": 0.4678,
+      "step": 4252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.058919150482084,
+      "learning_rate": 7.765877160335258e-06,
+      "loss": 0.4909,
+      "step": 4253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.6900241968797984,
+      "learning_rate": 7.764817581673673e-06,
+      "loss": 0.4773,
+      "step": 4254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8093000996695546,
+      "learning_rate": 7.76375782413201e-06,
+      "loss": 0.5018,
+      "step": 4255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.575340846990629,
+      "learning_rate": 7.762697887778828e-06,
+      "loss": 0.4902,
+      "step": 4256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.3513216223042015,
+      "learning_rate": 7.761637772682709e-06,
+      "loss": 0.4858,
+      "step": 4257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.351264589488454,
+      "learning_rate": 7.760577478912237e-06,
+      "loss": 0.4736,
+      "step": 4258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4763792852312911,
+      "learning_rate": 7.759517006536015e-06,
+      "loss": 0.5045,
+      "step": 4259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.2133795120987636,
+      "learning_rate": 7.758456355622651e-06,
+      "loss": 0.5099,
+      "step": 4260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5133623030142365,
+      "learning_rate": 7.75739552624077e-06,
+      "loss": 0.481,
+      "step": 4261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.35383302772019,
+      "learning_rate": 7.756334518459006e-06,
+      "loss": 0.4511,
+      "step": 4262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4520839797217628,
+      "learning_rate": 7.755273332346004e-06,
+      "loss": 0.4753,
+      "step": 4263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6741063390538984,
+      "learning_rate": 7.75421196797042e-06,
+      "loss": 0.546,
+      "step": 4264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.7485147859764902,
+      "learning_rate": 7.753150425400928e-06,
+      "loss": 0.4776,
+      "step": 4265
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4267801249566783,
+      "learning_rate": 7.752088704706201e-06,
+      "loss": 0.4402,
+      "step": 4266
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6923576111351786,
+      "learning_rate": 7.75102680595494e-06,
+      "loss": 0.5117,
+      "step": 4267
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.153602201020611,
+      "learning_rate": 7.74996472921584e-06,
+      "loss": 0.5003,
+      "step": 4268
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7486212430766905,
+      "learning_rate": 7.74890247455762e-06,
+      "loss": 0.4964,
+      "step": 4269
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.481852831581686,
+      "learning_rate": 7.747840042049006e-06,
+      "loss": 0.4153,
+      "step": 4270
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.602232273034823,
+      "learning_rate": 7.746777431758737e-06,
+      "loss": 0.5101,
+      "step": 4271
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.650932206242909,
+      "learning_rate": 7.74571464375556e-06,
+      "loss": 0.5216,
+      "step": 4272
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.35510304234774,
+      "learning_rate": 7.744651678108238e-06,
+      "loss": 0.4667,
+      "step": 4273
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.973813397319983,
+      "learning_rate": 7.743588534885543e-06,
+      "loss": 0.4249,
+      "step": 4274
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1722369170509006,
+      "learning_rate": 7.742525214156257e-06,
+      "loss": 0.4618,
+      "step": 4275
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3638551695917065,
+      "learning_rate": 7.74146171598918e-06,
+      "loss": 0.4657,
+      "step": 4276
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5544442393026597,
+      "learning_rate": 7.740398040453115e-06,
+      "loss": 0.531,
+      "step": 4277
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.174774090183045,
+      "learning_rate": 7.739334187616883e-06,
+      "loss": 0.4782,
+      "step": 4278
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4825862363176272,
+      "learning_rate": 7.73827015754931e-06,
+      "loss": 0.478,
+      "step": 4279
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7629979481365252,
+      "learning_rate": 7.737205950319242e-06,
+      "loss": 0.4932,
+      "step": 4280
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5951301622477199,
+      "learning_rate": 7.736141565995529e-06,
+      "loss": 0.5347,
+      "step": 4281
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7866128860923791,
+      "learning_rate": 7.735077004647036e-06,
+      "loss": 0.4741,
+      "step": 4282
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.324560280427871,
+      "learning_rate": 7.73401226634264e-06,
+      "loss": 0.4114,
+      "step": 4283
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7554236939088093,
+      "learning_rate": 7.732947351151225e-06,
+      "loss": 0.4902,
+      "step": 4284
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2934189087565242,
+      "learning_rate": 7.731882259141692e-06,
+      "loss": 0.4598,
+      "step": 4285
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7928806488371776,
+      "learning_rate": 7.730816990382951e-06,
+      "loss": 0.5202,
+      "step": 4286
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4284621759652492,
+      "learning_rate": 7.729751544943921e-06,
+      "loss": 0.467,
+      "step": 4287
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5552575056622058,
+      "learning_rate": 7.72868592289354e-06,
+      "loss": 0.4893,
+      "step": 4288
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6302013283580403,
+      "learning_rate": 7.727620124300748e-06,
+      "loss": 0.5084,
+      "step": 4289
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4334367953649194,
+      "learning_rate": 7.726554149234504e-06,
+      "loss": 0.4998,
+      "step": 4290
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3376296537783832,
+      "learning_rate": 7.725487997763772e-06,
+      "loss": 0.4847,
+      "step": 4291
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4780406845783058,
+      "learning_rate": 7.72442166995753e-06,
+      "loss": 0.5186,
+      "step": 4292
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3813080537532703,
+      "learning_rate": 7.723355165884772e-06,
+      "loss": 0.5,
+      "step": 4293
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4369200367525592,
+      "learning_rate": 7.722288485614498e-06,
+      "loss": 0.5019,
+      "step": 4294
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.375635666083701,
+      "learning_rate": 7.721221629215718e-06,
+      "loss": 0.4748,
+      "step": 4295
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5525079445059042,
+      "learning_rate": 7.72015459675746e-06,
+      "loss": 0.5301,
+      "step": 4296
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6051327980844967,
+      "learning_rate": 7.719087388308756e-06,
+      "loss": 0.5236,
+      "step": 4297
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9258708994796352,
+      "learning_rate": 7.718020003938658e-06,
+      "loss": 0.4618,
+      "step": 4298
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0605787144660113,
+      "learning_rate": 7.716952443716218e-06,
+      "loss": 0.4406,
+      "step": 4299
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.891532882614519,
+      "learning_rate": 7.715884707710513e-06,
+      "loss": 0.4702,
+      "step": 4300
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.723559293544447,
+      "learning_rate": 7.714816795990618e-06,
+      "loss": 0.4704,
+      "step": 4301
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8147073341472029,
+      "learning_rate": 7.713748708625627e-06,
+      "loss": 0.4751,
+      "step": 4302
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.577078186747778,
+      "learning_rate": 7.712680445684646e-06,
+      "loss": 0.4652,
+      "step": 4303
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8640570688823983,
+      "learning_rate": 7.711612007236789e-06,
+      "loss": 0.4881,
+      "step": 4304
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4361075905430707,
+      "learning_rate": 7.710543393351182e-06,
+      "loss": 0.4746,
+      "step": 4305
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5909832680837996,
+      "learning_rate": 7.709474604096963e-06,
+      "loss": 0.495,
+      "step": 4306
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6292558805088015,
+      "learning_rate": 7.708405639543281e-06,
+      "loss": 0.5228,
+      "step": 4307
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9821311711865293,
+      "learning_rate": 7.707336499759298e-06,
+      "loss": 0.4601,
+      "step": 4308
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7184752520929847,
+      "learning_rate": 7.706267184814184e-06,
+      "loss": 0.4763,
+      "step": 4309
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5242265105441464,
+      "learning_rate": 7.705197694777124e-06,
+      "loss": 0.4218,
+      "step": 4310
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.587975544495384,
+      "learning_rate": 7.704128029717312e-06,
+      "loss": 0.5029,
+      "step": 4311
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5997077328610512,
+      "learning_rate": 7.703058189703955e-06,
+      "loss": 0.5065,
+      "step": 4312
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5145095031762683,
+      "learning_rate": 7.701988174806268e-06,
+      "loss": 0.4859,
+      "step": 4313
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4428099085336208,
+      "learning_rate": 7.70091798509348e-06,
+      "loss": 0.4434,
+      "step": 4314
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6745222330239091,
+      "learning_rate": 7.699847620634834e-06,
+      "loss": 0.5171,
+      "step": 4315
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.739370528880135,
+      "learning_rate": 7.698777081499578e-06,
+      "loss": 0.4832,
+      "step": 4316
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.964668465668289,
+      "learning_rate": 7.697706367756975e-06,
+      "loss": 0.4848,
+      "step": 4317
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.718421245827606,
+      "learning_rate": 7.696635479476297e-06,
+      "loss": 0.532,
+      "step": 4318
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 4.375451919173206,
+      "learning_rate": 7.695564416726833e-06,
+      "loss": 0.4851,
+      "step": 4319
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.613946048655244,
+      "learning_rate": 7.69449317957788e-06,
+      "loss": 0.533,
+      "step": 4320
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6089080656732121,
+      "learning_rate": 7.693421768098737e-06,
+      "loss": 0.5141,
+      "step": 4321
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.574601259765763,
+      "learning_rate": 7.692350182358734e-06,
+      "loss": 0.5521,
+      "step": 4322
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.341151563807748,
+      "learning_rate": 7.691278422427195e-06,
+      "loss": 0.5249,
+      "step": 4323
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.2117090562204527,
+      "learning_rate": 7.690206488373462e-06,
+      "loss": 0.4793,
+      "step": 4324
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5602181295175142,
+      "learning_rate": 7.689134380266889e-06,
+      "loss": 0.5237,
+      "step": 4325
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0577191995624338,
+      "learning_rate": 7.688062098176839e-06,
+      "loss": 0.4836,
+      "step": 4326
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6869708829235728,
+      "learning_rate": 7.686989642172689e-06,
+      "loss": 0.4894,
+      "step": 4327
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7977093888484492,
+      "learning_rate": 7.685917012323823e-06,
+      "loss": 0.5456,
+      "step": 4328
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8470741965190391,
+      "learning_rate": 7.68484420869964e-06,
+      "loss": 0.5004,
+      "step": 4329
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3816486970602246,
+      "learning_rate": 7.683771231369548e-06,
+      "loss": 0.4755,
+      "step": 4330
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8709481793134477,
+      "learning_rate": 7.68269808040297e-06,
+      "loss": 0.5383,
+      "step": 4331
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7425303780839247,
+      "learning_rate": 7.681624755869334e-06,
+      "loss": 0.506,
+      "step": 4332
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5565652519556754,
+      "learning_rate": 7.680551257838084e-06,
+      "loss": 0.51,
+      "step": 4333
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.351851686327484,
+      "learning_rate": 7.679477586378677e-06,
+      "loss": 0.4737,
+      "step": 4334
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5454763805076641,
+      "learning_rate": 7.678403741560573e-06,
+      "loss": 0.4436,
+      "step": 4335
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.2641535531679686,
+      "learning_rate": 7.677329723453252e-06,
+      "loss": 0.4917,
+      "step": 4336
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7143770970358974,
+      "learning_rate": 7.6762555321262e-06,
+      "loss": 0.5201,
+      "step": 4337
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5615910483231392,
+      "learning_rate": 7.675181167648915e-06,
+      "loss": 0.4738,
+      "step": 4338
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2275472229208917,
+      "learning_rate": 7.674106630090909e-06,
+      "loss": 0.4637,
+      "step": 4339
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5159703107992575,
+      "learning_rate": 7.673031919521702e-06,
+      "loss": 0.4676,
+      "step": 4340
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3755483131325623,
+      "learning_rate": 7.671957036010826e-06,
+      "loss": 0.4324,
+      "step": 4341
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6210919746364294,
+      "learning_rate": 7.670881979627826e-06,
+      "loss": 0.4892,
+      "step": 4342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7309107173625531,
+      "learning_rate": 7.669806750442255e-06,
+      "loss": 0.5331,
+      "step": 4343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5433164307483809,
+      "learning_rate": 7.66873134852368e-06,
+      "loss": 0.4964,
+      "step": 4344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6817753380601144,
+      "learning_rate": 7.667655773941678e-06,
+      "loss": 0.4532,
+      "step": 4345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5650110346154513,
+      "learning_rate": 7.666580026765837e-06,
+      "loss": 0.4849,
+      "step": 4346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5494881443328056,
+      "learning_rate": 7.665504107065758e-06,
+      "loss": 0.5334,
+      "step": 4347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.3017144043290494,
+      "learning_rate": 7.66442801491105e-06,
+      "loss": 0.4896,
+      "step": 4348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5904229633114444,
+      "learning_rate": 7.663351750371332e-06,
+      "loss": 0.4897,
+      "step": 4349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6485757024964992,
+      "learning_rate": 7.662275313516243e-06,
+      "loss": 0.5211,
+      "step": 4350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5231119531252881,
+      "learning_rate": 7.661198704415423e-06,
+      "loss": 0.4615,
+      "step": 4351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.265497318500195,
+      "learning_rate": 7.660121923138528e-06,
+      "loss": 0.4815,
+      "step": 4352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4626762056154226,
+      "learning_rate": 7.659044969755225e-06,
+      "loss": 0.5032,
+      "step": 4353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.0540489265860185,
+      "learning_rate": 7.65796784433519e-06,
+      "loss": 0.5033,
+      "step": 4354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4258538114224182,
+      "learning_rate": 7.656890546948112e-06,
+      "loss": 0.449,
+      "step": 4355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4542109733978075,
+      "learning_rate": 7.655813077663691e-06,
+      "loss": 0.4649,
+      "step": 4356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3480076472239677,
+      "learning_rate": 7.65473543655164e-06,
+      "loss": 0.5194,
+      "step": 4357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3010966863527664,
+      "learning_rate": 7.653657623681679e-06,
+      "loss": 0.4509,
+      "step": 4358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3394330729443062,
+      "learning_rate": 7.652579639123541e-06,
+      "loss": 0.4503,
+      "step": 4359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5489017734953014,
+      "learning_rate": 7.651501482946969e-06,
+      "loss": 0.4735,
+      "step": 4360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6590780554686906,
+      "learning_rate": 7.65042315522172e-06,
+      "loss": 0.5075,
+      "step": 4361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6159152878545825,
+      "learning_rate": 7.649344656017562e-06,
+      "loss": 0.5392,
+      "step": 4362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6473455556863108,
+      "learning_rate": 7.648265985404268e-06,
+      "loss": 0.4576,
+      "step": 4363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5405783571301032,
+      "learning_rate": 7.647187143451631e-06,
+      "loss": 0.5133,
+      "step": 4364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3620234173421468,
+      "learning_rate": 7.646108130229449e-06,
+      "loss": 0.4734,
+      "step": 4365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.779954695652949,
+      "learning_rate": 7.645028945807528e-06,
+      "loss": 0.4713,
+      "step": 4366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2990426732745304,
+      "learning_rate": 7.6439495902557e-06,
+      "loss": 0.4447,
+      "step": 4367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.996841344058395,
+      "learning_rate": 7.64287006364379e-06,
+      "loss": 0.4962,
+      "step": 4368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.614699377740711,
+      "learning_rate": 7.641790366041644e-06,
+      "loss": 0.5171,
+      "step": 4369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.632958533183693,
+      "learning_rate": 7.640710497519117e-06,
+      "loss": 0.5081,
+      "step": 4370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.6109065886675396,
+      "learning_rate": 7.639630458146077e-06,
+      "loss": 0.4367,
+      "step": 4371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.319798317176399,
+      "learning_rate": 7.638550247992397e-06,
+      "loss": 0.4701,
+      "step": 4372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2971717328149732,
+      "learning_rate": 7.637469867127968e-06,
+      "loss": 0.446,
+      "step": 4373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5393790354485144,
+      "learning_rate": 7.63638931562269e-06,
+      "loss": 0.5499,
+      "step": 4374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5021270914245253,
+      "learning_rate": 7.635308593546474e-06,
+      "loss": 0.4658,
+      "step": 4375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8145823166428965,
+      "learning_rate": 7.634227700969235e-06,
+      "loss": 0.5014,
+      "step": 4376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.372617629241495,
+      "learning_rate": 7.633146637960912e-06,
+      "loss": 0.4764,
+      "step": 4377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.261945646447017,
+      "learning_rate": 7.632065404591445e-06,
+      "loss": 0.5023,
+      "step": 4378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.699046807640755,
+      "learning_rate": 7.630984000930792e-06,
+      "loss": 0.4593,
+      "step": 4379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.509644255899814,
+      "learning_rate": 7.629902427048914e-06,
+      "loss": 0.5292,
+      "step": 4380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6937798388492128,
+      "learning_rate": 7.62882068301579e-06,
+      "loss": 0.5485,
+      "step": 4381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8111486310872247,
+      "learning_rate": 7.627738768901406e-06,
+      "loss": 0.4728,
+      "step": 4382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.781695639199614,
+      "learning_rate": 7.626656684775762e-06,
+      "loss": 0.426,
+      "step": 4383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.205270962038038,
+      "learning_rate": 7.625574430708867e-06,
+      "loss": 0.4895,
+      "step": 4384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.8600944675609776,
+      "learning_rate": 7.624492006770739e-06,
+      "loss": 0.4449,
+      "step": 4385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.960252924934926,
+      "learning_rate": 7.623409413031413e-06,
+      "loss": 0.4903,
+      "step": 4386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5070960344259208,
+      "learning_rate": 7.622326649560929e-06,
+      "loss": 0.4497,
+      "step": 4387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2674293616770762,
+      "learning_rate": 7.62124371642934e-06,
+      "loss": 0.5188,
+      "step": 4388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.531908117905589,
+      "learning_rate": 7.620160613706715e-06,
+      "loss": 0.5434,
+      "step": 4389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.6009428398160452,
+      "learning_rate": 7.619077341463123e-06,
+      "loss": 0.4629,
+      "step": 4390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 2.1354423427655678,
+      "learning_rate": 7.617993899768657e-06,
+      "loss": 0.4963,
+      "step": 4391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.1998541801763876,
+      "learning_rate": 7.6169102886934065e-06,
+      "loss": 0.4641,
+      "step": 4392
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6246609202243575,
+      "learning_rate": 7.615826508307485e-06,
+      "loss": 0.5376,
+      "step": 4393
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4523258671770887,
+      "learning_rate": 7.61474255868101e-06,
+      "loss": 0.4746,
+      "step": 4394
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5509905049147423,
+      "learning_rate": 7.613658439884113e-06,
+      "loss": 0.4367,
+      "step": 4395
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6205160095055837,
+      "learning_rate": 7.612574151986934e-06,
+      "loss": 0.4435,
+      "step": 4396
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5473912199290134,
+      "learning_rate": 7.611489695059623e-06,
+      "loss": 0.4881,
+      "step": 4397
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.588197470952627,
+      "learning_rate": 7.610405069172346e-06,
+      "loss": 0.5022,
+      "step": 4398
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4717862304020497,
+      "learning_rate": 7.609320274395276e-06,
+      "loss": 0.419,
+      "step": 4399
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5975036295577221,
+      "learning_rate": 7.608235310798599e-06,
+      "loss": 0.5455,
+      "step": 4400
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.506532787527015,
+      "learning_rate": 7.607150178452507e-06,
+      "loss": 0.5351,
+      "step": 4401
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.034853599576947,
+      "learning_rate": 7.606064877427211e-06,
+      "loss": 0.5134,
+      "step": 4402
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5398705020383172,
+      "learning_rate": 7.604979407792925e-06,
+      "loss": 0.4745,
+      "step": 4403
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7659273627551266,
+      "learning_rate": 7.6038937696198815e-06,
+      "loss": 0.4972,
+      "step": 4404
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6067845163178638,
+      "learning_rate": 7.602807962978316e-06,
+      "loss": 0.4928,
+      "step": 4405
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6939930679058588,
+      "learning_rate": 7.601721987938479e-06,
+      "loss": 0.5049,
+      "step": 4406
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.592387113127101,
+      "learning_rate": 7.600635844570634e-06,
+      "loss": 0.4848,
+      "step": 4407
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5468742431161744,
+      "learning_rate": 7.599549532945052e-06,
+      "loss": 0.5312,
+      "step": 4408
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6067866597447613,
+      "learning_rate": 7.598463053132016e-06,
+      "loss": 0.5409,
+      "step": 4409
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5106847560416075,
+      "learning_rate": 7.597376405201819e-06,
+      "loss": 0.4799,
+      "step": 4410
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3059740672464717,
+      "learning_rate": 7.596289589224766e-06,
+      "loss": 0.3845,
+      "step": 4411
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6144569511324067,
+      "learning_rate": 7.595202605271175e-06,
+      "loss": 0.4878,
+      "step": 4412
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.182215333323812,
+      "learning_rate": 7.594115453411368e-06,
+      "loss": 0.4476,
+      "step": 4413
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6143739893897535,
+      "learning_rate": 7.593028133715686e-06,
+      "loss": 0.5359,
+      "step": 4414
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.707523466158357,
+      "learning_rate": 7.591940646254476e-06,
+      "loss": 0.4703,
+      "step": 4415
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6407080817071937,
+      "learning_rate": 7.5908529910980964e-06,
+      "loss": 0.509,
+      "step": 4416
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.75168225955281,
+      "learning_rate": 7.589765168316918e-06,
+      "loss": 0.4827,
+      "step": 4417
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.597709830879285,
+      "learning_rate": 7.58867717798132e-06,
+      "loss": 0.4636,
+      "step": 4418
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6259871893353737,
+      "learning_rate": 7.5875890201616964e-06,
+      "loss": 0.4817,
+      "step": 4419
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9950193784671602,
+      "learning_rate": 7.586500694928447e-06,
+      "loss": 0.4551,
+      "step": 4420
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.516211618015143,
+      "learning_rate": 7.585412202351987e-06,
+      "loss": 0.479,
+      "step": 4421
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.1026612318751723,
+      "learning_rate": 7.58432354250274e-06,
+      "loss": 0.5381,
+      "step": 4422
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.9350742312526639,
+      "learning_rate": 7.5832347154511386e-06,
+      "loss": 0.5313,
+      "step": 4423
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4000866419093525,
+      "learning_rate": 7.582145721267633e-06,
+      "loss": 0.4565,
+      "step": 4424
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5655442207733838,
+      "learning_rate": 7.581056560022675e-06,
+      "loss": 0.5092,
+      "step": 4425
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.515006977205771,
+      "learning_rate": 7.579967231786736e-06,
+      "loss": 0.4437,
+      "step": 4426
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6110213674022404,
+      "learning_rate": 7.57887773663029e-06,
+      "loss": 0.541,
+      "step": 4427
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.599276691508581,
+      "learning_rate": 7.577788074623829e-06,
+      "loss": 0.4939,
+      "step": 4428
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4529793534262705,
+      "learning_rate": 7.576698245837852e-06,
+      "loss": 0.4514,
+      "step": 4429
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6750135965516877,
+      "learning_rate": 7.575608250342869e-06,
+      "loss": 0.4679,
+      "step": 4430
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3763606760101372,
+      "learning_rate": 7.574518088209401e-06,
+      "loss": 0.4602,
+      "step": 4431
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6092617787608126,
+      "learning_rate": 7.573427759507981e-06,
+      "loss": 0.5321,
+      "step": 4432
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3459612186619383,
+      "learning_rate": 7.572337264309152e-06,
+      "loss": 0.4786,
+      "step": 4433
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2291549546809786,
+      "learning_rate": 7.571246602683465e-06,
+      "loss": 0.4905,
+      "step": 4434
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5159017242696466,
+      "learning_rate": 7.570155774701489e-06,
+      "loss": 0.4809,
+      "step": 4435
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3485380385311292,
+      "learning_rate": 7.569064780433795e-06,
+      "loss": 0.4248,
+      "step": 4436
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.5148950725374664,
+      "learning_rate": 7.567973619950971e-06,
+      "loss": 0.571,
+      "step": 4437
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.845712594157829,
+      "learning_rate": 7.566882293323613e-06,
+      "loss": 0.4992,
+      "step": 4438
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5455301234944434,
+      "learning_rate": 7.565790800622329e-06,
+      "loss": 0.4865,
+      "step": 4439
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8154327258099632,
+      "learning_rate": 7.564699141917736e-06,
+      "loss": 0.5074,
+      "step": 4440
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7065437889458257,
+      "learning_rate": 7.5636073172804645e-06,
+      "loss": 0.5275,
+      "step": 4441
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5511083564884024,
+      "learning_rate": 7.562515326781152e-06,
+      "loss": 0.5138,
+      "step": 4442
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3618501667780516,
+      "learning_rate": 7.561423170490452e-06,
+      "loss": 0.4971,
+      "step": 4443
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5426749343066507,
+      "learning_rate": 7.56033084847902e-06,
+      "loss": 0.5052,
+      "step": 4444
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.487224584720076,
+      "learning_rate": 7.559238360817535e-06,
+      "loss": 0.5055,
+      "step": 4445
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4165232719981535,
+      "learning_rate": 7.558145707576674e-06,
+      "loss": 0.4588,
+      "step": 4446
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5801604714832982,
+      "learning_rate": 7.557052888827133e-06,
+      "loss": 0.5193,
+      "step": 4447
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4379617927182353,
+      "learning_rate": 7.5559599046396145e-06,
+      "loss": 0.4823,
+      "step": 4448
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 5.38571091919479,
+      "learning_rate": 7.5548667550848355e-06,
+      "loss": 0.5037,
+      "step": 4449
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6177756897110601,
+      "learning_rate": 7.5537734402335175e-06,
+      "loss": 0.5409,
+      "step": 4450
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6108201666330376,
+      "learning_rate": 7.552679960156399e-06,
+      "loss": 0.5335,
+      "step": 4451
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.003773633491461,
+      "learning_rate": 7.551586314924227e-06,
+      "loss": 0.5074,
+      "step": 4452
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5208651064661973,
+      "learning_rate": 7.5504925046077596e-06,
+      "loss": 0.4446,
+      "step": 4453
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.356252304142289,
+      "learning_rate": 7.549398529277762e-06,
+      "loss": 0.453,
+      "step": 4454
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6821756659298526,
+      "learning_rate": 7.548304389005014e-06,
+      "loss": 0.4774,
+      "step": 4455
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.044114879586208,
+      "learning_rate": 7.547210083860306e-06,
+      "loss": 0.4785,
+      "step": 4456
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6554998870232182,
+      "learning_rate": 7.54611561391444e-06,
+      "loss": 0.5165,
+      "step": 4457
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5020775187145792,
+      "learning_rate": 7.545020979238223e-06,
+      "loss": 0.4381,
+      "step": 4458
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8668931551165677,
+      "learning_rate": 7.5439261799024764e-06,
+      "loss": 0.4518,
+      "step": 4459
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8737775041547375,
+      "learning_rate": 7.5428312159780345e-06,
+      "loss": 0.4469,
+      "step": 4460
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2292716486198088,
+      "learning_rate": 7.541736087535742e-06,
+      "loss": 0.4918,
+      "step": 4461
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8428146257859725,
+      "learning_rate": 7.540640794646447e-06,
+      "loss": 0.4656,
+      "step": 4462
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.564407579149257,
+      "learning_rate": 7.539545337381016e-06,
+      "loss": 0.5286,
+      "step": 4463
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8797101320893963,
+      "learning_rate": 7.538449715810326e-06,
+      "loss": 0.4974,
+      "step": 4464
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.272480384464611,
+      "learning_rate": 7.537353930005258e-06,
+      "loss": 0.5325,
+      "step": 4465
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.354903940702524,
+      "learning_rate": 7.536257980036711e-06,
+      "loss": 0.5238,
+      "step": 4466
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5462948130690461,
+      "learning_rate": 7.53516186597559e-06,
+      "loss": 0.5064,
+      "step": 4467
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3658738820538754,
+      "learning_rate": 7.534065587892813e-06,
+      "loss": 0.5096,
+      "step": 4468
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4651775932863003,
+      "learning_rate": 7.5329691458593055e-06,
+      "loss": 0.4812,
+      "step": 4469
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6032323310735264,
+      "learning_rate": 7.5318725399460116e-06,
+      "loss": 0.5257,
+      "step": 4470
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4617539525571723,
+      "learning_rate": 7.530775770223874e-06,
+      "loss": 0.4867,
+      "step": 4471
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6629355393212113,
+      "learning_rate": 7.529678836763856e-06,
+      "loss": 0.4741,
+      "step": 4472
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.822646317040909,
+      "learning_rate": 7.528581739636924e-06,
+      "loss": 0.5128,
+      "step": 4473
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.476838310406188,
+      "learning_rate": 7.527484478914065e-06,
+      "loss": 0.4526,
+      "step": 4474
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3590454924793507,
+      "learning_rate": 7.5263870546662646e-06,
+      "loss": 0.4388,
+      "step": 4475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5897290500388133,
+      "learning_rate": 7.525289466964527e-06,
+      "loss": 0.4926,
+      "step": 4476
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6208388993958498,
+      "learning_rate": 7.524191715879863e-06,
+      "loss": 0.5002,
+      "step": 4477
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5680087745681166,
+      "learning_rate": 7.523093801483299e-06,
+      "loss": 0.4851,
+      "step": 4478
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7373850773757729,
+      "learning_rate": 7.521995723845865e-06,
+      "loss": 0.5059,
+      "step": 4479
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8080957698258109,
+      "learning_rate": 7.5208974830386075e-06,
+      "loss": 0.505,
+      "step": 4480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7411347941280448,
+      "learning_rate": 7.51979907913258e-06,
+      "loss": 0.5149,
+      "step": 4481
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8654424290737228,
+      "learning_rate": 7.518700512198851e-06,
+      "loss": 0.5057,
+      "step": 4482
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8711471494456113,
+      "learning_rate": 7.51760178230849e-06,
+      "loss": 0.4655,
+      "step": 4483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7930850778047611,
+      "learning_rate": 7.516502889532591e-06,
+      "loss": 0.4404,
+      "step": 4484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.232414067915649,
+      "learning_rate": 7.5154038339422444e-06,
+      "loss": 0.5064,
+      "step": 4485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5135781376359325,
+      "learning_rate": 7.514304615608561e-06,
+      "loss": 0.4993,
+      "step": 4486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6053257003787963,
+      "learning_rate": 7.5132052346026585e-06,
+      "loss": 0.5224,
+      "step": 4487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.877287040598228,
+      "learning_rate": 7.512105690995663e-06,
+      "loss": 0.5164,
+      "step": 4488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.584063673646794,
+      "learning_rate": 7.511005984858718e-06,
+      "loss": 0.4925,
+      "step": 4489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5118364063314114,
+      "learning_rate": 7.5099061162629695e-06,
+      "loss": 0.4399,
+      "step": 4490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.286386873523664,
+      "learning_rate": 7.508806085279578e-06,
+      "loss": 0.4702,
+      "step": 4491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5322263644398175,
+      "learning_rate": 7.507705891979716e-06,
+      "loss": 0.4479,
+      "step": 4492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5884567306105825,
+      "learning_rate": 7.5066055364345635e-06,
+      "loss": 0.5287,
+      "step": 4493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.136812797393852,
+      "learning_rate": 7.5055050187153114e-06,
+      "loss": 0.4443,
+      "step": 4494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6531259456452758,
+      "learning_rate": 7.504404338893161e-06,
+      "loss": 0.4855,
+      "step": 4495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5765068310108791,
+      "learning_rate": 7.503303497039328e-06,
+      "loss": 0.5363,
+      "step": 4496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0215275519041325,
+      "learning_rate": 7.5022024932250325e-06,
+      "loss": 0.5126,
+      "step": 4497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4193773370525802,
+      "learning_rate": 7.501101327521508e-06,
+      "loss": 0.4856,
+      "step": 4498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8293186785982878,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.4059,
+      "step": 4499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7869349709844538,
+      "learning_rate": 7.498898510731764e-06,
+      "loss": 0.4684,
+      "step": 4500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7956066470240517,
+      "learning_rate": 7.497796859788061e-06,
+      "loss": 0.431,
+      "step": 4501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3924377298140087,
+      "learning_rate": 7.496695047240171e-06,
+      "loss": 0.5482,
+      "step": 4502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.414103993748557,
+      "learning_rate": 7.495593073159375e-06,
+      "loss": 0.4566,
+      "step": 4503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6808589040239943,
+      "learning_rate": 7.494490937616974e-06,
+      "loss": 0.5161,
+      "step": 4504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5916759813193967,
+      "learning_rate": 7.493388640684272e-06,
+      "loss": 0.5371,
+      "step": 4505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5847902985451419,
+      "learning_rate": 7.492286182432587e-06,
+      "loss": 0.5257,
+      "step": 4506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3919912753264105,
+      "learning_rate": 7.4911835629332455e-06,
+      "loss": 0.4415,
+      "step": 4507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6785266860233201,
+      "learning_rate": 7.490080782257586e-06,
+      "loss": 0.4747,
+      "step": 4508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6931148408089607,
+      "learning_rate": 7.488977840476959e-06,
+      "loss": 0.4664,
+      "step": 4509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.0837994284370907,
+      "learning_rate": 7.48787473766272e-06,
+      "loss": 0.4285,
+      "step": 4510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.2407324632378507,
+      "learning_rate": 7.486771473886241e-06,
+      "loss": 0.4773,
+      "step": 4511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.7196108707576485,
+      "learning_rate": 7.4856680492189e-06,
+      "loss": 0.5144,
+      "step": 4512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.699971127645639,
+      "learning_rate": 7.484564463732088e-06,
+      "loss": 0.5035,
+      "step": 4513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3734646529451033,
+      "learning_rate": 7.483460717497206e-06,
+      "loss": 0.3997,
+      "step": 4514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.388657959643578,
+      "learning_rate": 7.482356810585664e-06,
+      "loss": 0.464,
+      "step": 4515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5637001452356531,
+      "learning_rate": 7.481252743068883e-06,
+      "loss": 0.484,
+      "step": 4516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.515987573506066,
+      "learning_rate": 7.480148515018297e-06,
+      "loss": 0.5088,
+      "step": 4517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.568823368420787,
+      "learning_rate": 7.479044126505346e-06,
+      "loss": 0.4498,
+      "step": 4518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.8388535713065177,
+      "learning_rate": 7.477939577601483e-06,
+      "loss": 0.4709,
+      "step": 4519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.4671587391302863,
+      "learning_rate": 7.47683486837817e-06,
+      "loss": 0.4858,
+      "step": 4520
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.254887129395308,
+      "learning_rate": 7.475729998906882e-06,
+      "loss": 0.4248,
+      "step": 4521
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7439106225334517,
+      "learning_rate": 7.474624969259101e-06,
+      "loss": 0.504,
+      "step": 4522
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4869998372794844,
+      "learning_rate": 7.473519779506323e-06,
+      "loss": 0.4224,
+      "step": 4523
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0931860449575193,
+      "learning_rate": 7.4724144297200495e-06,
+      "loss": 0.5338,
+      "step": 4524
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8430415320796543,
+      "learning_rate": 7.471308919971798e-06,
+      "loss": 0.5348,
+      "step": 4525
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.822049588452768,
+      "learning_rate": 7.470203250333091e-06,
+      "loss": 0.4921,
+      "step": 4526
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7237492026059309,
+      "learning_rate": 7.469097420875466e-06,
+      "loss": 0.5263,
+      "step": 4527
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6679548463020846,
+      "learning_rate": 7.4679914316704675e-06,
+      "loss": 0.4919,
+      "step": 4528
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.000580324355168,
+      "learning_rate": 7.4668852827896535e-06,
+      "loss": 0.4792,
+      "step": 4529
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8803825699792125,
+      "learning_rate": 7.465778974304586e-06,
+      "loss": 0.4675,
+      "step": 4530
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0272436272128336,
+      "learning_rate": 7.464672506286847e-06,
+      "loss": 0.4612,
+      "step": 4531
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6209735963804022,
+      "learning_rate": 7.463565878808019e-06,
+      "loss": 0.4874,
+      "step": 4532
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1369553720059353,
+      "learning_rate": 7.462459091939702e-06,
+      "loss": 0.4648,
+      "step": 4533
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4995356013641483,
+      "learning_rate": 7.461352145753504e-06,
+      "loss": 0.5053,
+      "step": 4534
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.529006868246565,
+      "learning_rate": 7.460245040321042e-06,
+      "loss": 0.4226,
+      "step": 4535
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3331856299336633,
+      "learning_rate": 7.459137775713943e-06,
+      "loss": 0.4331,
+      "step": 4536
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.106012461019877,
+      "learning_rate": 7.458030352003848e-06,
+      "loss": 0.4921,
+      "step": 4537
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.615461793405107,
+      "learning_rate": 7.456922769262404e-06,
+      "loss": 0.4952,
+      "step": 4538
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.2671569042011437,
+      "learning_rate": 7.4558150275612705e-06,
+      "loss": 0.447,
+      "step": 4539
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.7794679886603824,
+      "learning_rate": 7.454707126972118e-06,
+      "loss": 0.5211,
+      "step": 4540
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4264673620250976,
+      "learning_rate": 7.453599067566626e-06,
+      "loss": 0.4504,
+      "step": 4541
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3064355993596606,
+      "learning_rate": 7.452490849416484e-06,
+      "loss": 0.5299,
+      "step": 4542
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6467967382094293,
+      "learning_rate": 7.451382472593392e-06,
+      "loss": 0.4448,
+      "step": 4543
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4304162253251227,
+      "learning_rate": 7.4502739371690604e-06,
+      "loss": 0.4742,
+      "step": 4544
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3720667666551278,
+      "learning_rate": 7.449165243215211e-06,
+      "loss": 0.4603,
+      "step": 4545
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7148512221836614,
+      "learning_rate": 7.4480563908035745e-06,
+      "loss": 0.5077,
+      "step": 4546
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5693859683538844,
+      "learning_rate": 7.446947380005889e-06,
+      "loss": 0.4631,
+      "step": 4547
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2663175828028757,
+      "learning_rate": 7.445838210893912e-06,
+      "loss": 0.4664,
+      "step": 4548
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.668593563555438,
+      "learning_rate": 7.444728883539402e-06,
+      "loss": 0.523,
+      "step": 4549
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5325568136069472,
+      "learning_rate": 7.443619398014132e-06,
+      "loss": 0.4953,
+      "step": 4550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.916443328343889,
+      "learning_rate": 7.4425097543898816e-06,
+      "loss": 0.5498,
+      "step": 4551
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7861666358727153,
+      "learning_rate": 7.441399952738449e-06,
+      "loss": 0.4702,
+      "step": 4552
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.359622573070845,
+      "learning_rate": 7.440289993131628e-06,
+      "loss": 0.4928,
+      "step": 4553
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5954992438531257,
+      "learning_rate": 7.439179875641242e-06,
+      "loss": 0.4596,
+      "step": 4554
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3721828389198607,
+      "learning_rate": 7.438069600339105e-06,
+      "loss": 0.4964,
+      "step": 4555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.582390961912261,
+      "learning_rate": 7.436959167297056e-06,
+      "loss": 0.4587,
+      "step": 4556
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3587658636479258,
+      "learning_rate": 7.435848576586936e-06,
+      "loss": 0.4988,
+      "step": 4557
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9264581462937433,
+      "learning_rate": 7.4347378282806e-06,
+      "loss": 0.5069,
+      "step": 4558
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5512499315893842,
+      "learning_rate": 7.43362692244991e-06,
+      "loss": 0.5034,
+      "step": 4559
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3893578835382685,
+      "learning_rate": 7.432515859166745e-06,
+      "loss": 0.5343,
+      "step": 4560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.4991956341952286,
+      "learning_rate": 7.431404638502984e-06,
+      "loss": 0.4385,
+      "step": 4561
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4268899337946406,
+      "learning_rate": 7.430293260530523e-06,
+      "loss": 0.4869,
+      "step": 4562
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6801477901501437,
+      "learning_rate": 7.429181725321268e-06,
+      "loss": 0.5152,
+      "step": 4563
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4539248608675672,
+      "learning_rate": 7.428070032947134e-06,
+      "loss": 0.4718,
+      "step": 4564
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.421164254144659,
+      "learning_rate": 7.426958183480044e-06,
+      "loss": 0.4951,
+      "step": 4565
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7075140960498736,
+      "learning_rate": 7.425846176991936e-06,
+      "loss": 0.477,
+      "step": 4566
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5629106205121988,
+      "learning_rate": 7.424734013554752e-06,
+      "loss": 0.4876,
+      "step": 4567
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6644935581440224,
+      "learning_rate": 7.423621693240449e-06,
+      "loss": 0.5274,
+      "step": 4568
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5890211013645678,
+      "learning_rate": 7.4225092161209945e-06,
+      "loss": 0.541,
+      "step": 4569
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5431425665039833,
+      "learning_rate": 7.421396582268362e-06,
+      "loss": 0.4757,
+      "step": 4570
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5584282424103824,
+      "learning_rate": 7.420283791754538e-06,
+      "loss": 0.4672,
+      "step": 4571
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.872359343935138,
+      "learning_rate": 7.41917084465152e-06,
+      "loss": 0.49,
+      "step": 4572
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7835210803524322,
+      "learning_rate": 7.4180577410313115e-06,
+      "loss": 0.486,
+      "step": 4573
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.436099716267143,
+      "learning_rate": 7.416944480965931e-06,
+      "loss": 0.4529,
+      "step": 4574
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7369250304650892,
+      "learning_rate": 7.4158310645274045e-06,
+      "loss": 0.4541,
+      "step": 4575
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0138837328984263,
+      "learning_rate": 7.414717491787767e-06,
+      "loss": 0.4955,
+      "step": 4576
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.2880692185521965,
+      "learning_rate": 7.413603762819069e-06,
+      "loss": 0.4532,
+      "step": 4577
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9742190975613736,
+      "learning_rate": 7.412489877693363e-06,
+      "loss": 0.4886,
+      "step": 4578
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4206213644313295,
+      "learning_rate": 7.411375836482719e-06,
+      "loss": 0.4629,
+      "step": 4579
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2749168915307603,
+      "learning_rate": 7.410261639259213e-06,
+      "loss": 0.4742,
+      "step": 4580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4701068249232978,
+      "learning_rate": 7.409147286094931e-06,
+      "loss": 0.4686,
+      "step": 4581
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6252098036360981,
+      "learning_rate": 7.4080327770619705e-06,
+      "loss": 0.5067,
+      "step": 4582
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7554462430777722,
+      "learning_rate": 7.406918112232442e-06,
+      "loss": 0.5034,
+      "step": 4583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6617602181855406,
+      "learning_rate": 7.405803291678459e-06,
+      "loss": 0.5188,
+      "step": 4584
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4370230940503752,
+      "learning_rate": 7.40468831547215e-06,
+      "loss": 0.4698,
+      "step": 4585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5064712328118048,
+      "learning_rate": 7.403573183685653e-06,
+      "loss": 0.5004,
+      "step": 4586
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5370013358266668,
+      "learning_rate": 7.402457896391115e-06,
+      "loss": 0.5108,
+      "step": 4587
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5637713562169963,
+      "learning_rate": 7.401342453660694e-06,
+      "loss": 0.5173,
+      "step": 4588
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8776941740260504,
+      "learning_rate": 7.400226855566557e-06,
+      "loss": 0.4815,
+      "step": 4589
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9062465987060218,
+      "learning_rate": 7.399111102180883e-06,
+      "loss": 0.5013,
+      "step": 4590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6104502836362812,
+      "learning_rate": 7.3979951935758596e-06,
+      "loss": 0.5097,
+      "step": 4591
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.346453409697688,
+      "learning_rate": 7.396879129823682e-06,
+      "loss": 0.4811,
+      "step": 4592
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3477548632362835,
+      "learning_rate": 7.395762910996562e-06,
+      "loss": 0.4601,
+      "step": 4593
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8769133626027674,
+      "learning_rate": 7.394646537166716e-06,
+      "loss": 0.4928,
+      "step": 4594
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 3.1914585684698458,
+      "learning_rate": 7.393530008406371e-06,
+      "loss": 0.5043,
+      "step": 4595
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.740525002299161,
+      "learning_rate": 7.392413324787766e-06,
+      "loss": 0.4432,
+      "step": 4596
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9644205910060666,
+      "learning_rate": 7.391296486383149e-06,
+      "loss": 0.5369,
+      "step": 4597
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6060052506183139,
+      "learning_rate": 7.390179493264775e-06,
+      "loss": 0.5191,
+      "step": 4598
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8419163434096446,
+      "learning_rate": 7.389062345504918e-06,
+      "loss": 0.4727,
+      "step": 4599
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3730140966512354,
+      "learning_rate": 7.38794504317585e-06,
+      "loss": 0.5147,
+      "step": 4600
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5917576195321994,
+      "learning_rate": 7.386827586349863e-06,
+      "loss": 0.5509,
+      "step": 4601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8463978005293353,
+      "learning_rate": 7.385709975099253e-06,
+      "loss": 0.4696,
+      "step": 4602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6666029828454636,
+      "learning_rate": 7.384592209496327e-06,
+      "loss": 0.4787,
+      "step": 4603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7721465857892111,
+      "learning_rate": 7.383474289613406e-06,
+      "loss": 0.5318,
+      "step": 4604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5962959036224518,
+      "learning_rate": 7.382356215522817e-06,
+      "loss": 0.4936,
+      "step": 4605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.8137160120811777,
+      "learning_rate": 7.3812379872968965e-06,
+      "loss": 0.5294,
+      "step": 4606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.617643832532217,
+      "learning_rate": 7.380119605007994e-06,
+      "loss": 0.4883,
+      "step": 4607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9790178111365713,
+      "learning_rate": 7.379001068728466e-06,
+      "loss": 0.5106,
+      "step": 4608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 4.312277878219389,
+      "learning_rate": 7.377882378530682e-06,
+      "loss": 0.5312,
+      "step": 4609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.193144227060164,
+      "learning_rate": 7.376763534487017e-06,
+      "loss": 0.4676,
+      "step": 4610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5936775917283627,
+      "learning_rate": 7.3756445366698615e-06,
+      "loss": 0.5072,
+      "step": 4611
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4528012406837663,
+      "learning_rate": 7.374525385151612e-06,
+      "loss": 0.5289,
+      "step": 4612
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.0759791944613757,
+      "learning_rate": 7.373406080004677e-06,
+      "loss": 0.5165,
+      "step": 4613
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5036839731083245,
+      "learning_rate": 7.372286621301472e-06,
+      "loss": 0.4996,
+      "step": 4614
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6647095264121488,
+      "learning_rate": 7.371167009114427e-06,
+      "loss": 0.4789,
+      "step": 4615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.642137836976585,
+      "learning_rate": 7.370047243515979e-06,
+      "loss": 0.4886,
+      "step": 4616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6057604724488086,
+      "learning_rate": 7.368927324578573e-06,
+      "loss": 0.5262,
+      "step": 4617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5737105266684055,
+      "learning_rate": 7.367807252374669e-06,
+      "loss": 0.4702,
+      "step": 4618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6620256473967938,
+      "learning_rate": 7.3666870269767335e-06,
+      "loss": 0.5102,
+      "step": 4619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3210808457947225,
+      "learning_rate": 7.365566648457243e-06,
+      "loss": 0.4883,
+      "step": 4620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.932596130903649,
+      "learning_rate": 7.364446116888685e-06,
+      "loss": 0.5035,
+      "step": 4621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.546071334438265,
+      "learning_rate": 7.363325432343558e-06,
+      "loss": 0.4405,
+      "step": 4622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7904804890626385,
+      "learning_rate": 7.362204594894364e-06,
+      "loss": 0.4818,
+      "step": 4623
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5672802210464944,
+      "learning_rate": 7.3610836046136245e-06,
+      "loss": 0.5096,
+      "step": 4624
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6788645750950209,
+      "learning_rate": 7.359962461573864e-06,
+      "loss": 0.467,
+      "step": 4625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5815507626312333,
+      "learning_rate": 7.35884116584762e-06,
+      "loss": 0.5186,
+      "step": 4626
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5391839136161658,
+      "learning_rate": 7.357719717507438e-06,
+      "loss": 0.5112,
+      "step": 4627
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6568881606067514,
+      "learning_rate": 7.356598116625875e-06,
+      "loss": 0.5229,
+      "step": 4628
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7157083781686024,
+      "learning_rate": 7.355476363275496e-06,
+      "loss": 0.4403,
+      "step": 4629
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7045322490959582,
+      "learning_rate": 7.354354457528879e-06,
+      "loss": 0.4883,
+      "step": 4630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.535848079552201,
+      "learning_rate": 7.353232399458606e-06,
+      "loss": 0.4487,
+      "step": 4631
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5187199415491002,
+      "learning_rate": 7.3521101891372785e-06,
+      "loss": 0.4217,
+      "step": 4632
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5473522073622727,
+      "learning_rate": 7.350987826637496e-06,
+      "loss": 0.5067,
+      "step": 4633
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8035987685377435,
+      "learning_rate": 7.349865312031877e-06,
+      "loss": 0.4822,
+      "step": 4634
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8254647215719237,
+      "learning_rate": 7.348742645393048e-06,
+      "loss": 0.5218,
+      "step": 4635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6523730887136596,
+      "learning_rate": 7.347619826793641e-06,
+      "loss": 0.4448,
+      "step": 4636
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5744395819001273,
+      "learning_rate": 7.346496856306301e-06,
+      "loss": 0.46,
+      "step": 4637
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4661406327294033,
+      "learning_rate": 7.345373734003686e-06,
+      "loss": 0.4935,
+      "step": 4638
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7203413754040668,
+      "learning_rate": 7.344250459958458e-06,
+      "loss": 0.5183,
+      "step": 4639
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.740530995780852,
+      "learning_rate": 7.343127034243291e-06,
+      "loss": 0.4482,
+      "step": 4640
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.314592721387347,
+      "learning_rate": 7.3420034569308714e-06,
+      "loss": 0.4626,
+      "step": 4641
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5976847634236406,
+      "learning_rate": 7.340879728093892e-06,
+      "loss": 0.5143,
+      "step": 4642
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.592952838505204,
+      "learning_rate": 7.339755847805056e-06,
+      "loss": 0.5231,
+      "step": 4643
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5068808662320112,
+      "learning_rate": 7.338631816137078e-06,
+      "loss": 0.5138,
+      "step": 4644
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.094282205836922,
+      "learning_rate": 7.337507633162679e-06,
+      "loss": 0.4593,
+      "step": 4645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.367526506725231,
+      "learning_rate": 7.336383298954595e-06,
+      "loss": 0.4576,
+      "step": 4646
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.1691608634472472,
+      "learning_rate": 7.3352588135855686e-06,
+      "loss": 0.4926,
+      "step": 4647
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2112541387765567,
+      "learning_rate": 7.334134177128351e-06,
+      "loss": 0.4329,
+      "step": 4648
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7094359754251969,
+      "learning_rate": 7.333009389655705e-06,
+      "loss": 0.4716,
+      "step": 4649
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.747094608404503,
+      "learning_rate": 7.331884451240405e-06,
+      "loss": 0.3926,
+      "step": 4650
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.693301252004503,
+      "learning_rate": 7.330759361955231e-06,
+      "loss": 0.5551,
+      "step": 4651
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2724635049790383,
+      "learning_rate": 7.329634121872975e-06,
+      "loss": 0.4346,
+      "step": 4652
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6357116178168966,
+      "learning_rate": 7.328508731066439e-06,
+      "loss": 0.5364,
+      "step": 4653
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.46269166390621,
+      "learning_rate": 7.327383189608433e-06,
+      "loss": 0.4901,
+      "step": 4654
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.588153398959974,
+      "learning_rate": 7.326257497571782e-06,
+      "loss": 0.5033,
+      "step": 4655
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6697534209725695,
+      "learning_rate": 7.325131655029311e-06,
+      "loss": 0.4991,
+      "step": 4656
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.789051368719857,
+      "learning_rate": 7.3240056620538655e-06,
+      "loss": 0.5012,
+      "step": 4657
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5811753435834477,
+      "learning_rate": 7.3228795187182935e-06,
+      "loss": 0.5037,
+      "step": 4658
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6111014758078617,
+      "learning_rate": 7.321753225095456e-06,
+      "loss": 0.5283,
+      "step": 4659
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5626291853319376,
+      "learning_rate": 7.32062678125822e-06,
+      "loss": 0.5122,
+      "step": 4660
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8932933475980906,
+      "learning_rate": 7.319500187279469e-06,
+      "loss": 0.4487,
+      "step": 4661
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6127923986626242,
+      "learning_rate": 7.318373443232088e-06,
+      "loss": 0.5239,
+      "step": 4662
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5705045767094448,
+      "learning_rate": 7.317246549188981e-06,
+      "loss": 0.5302,
+      "step": 4663
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5510543508427849,
+      "learning_rate": 7.316119505223052e-06,
+      "loss": 0.4808,
+      "step": 4664
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4231294492913333,
+      "learning_rate": 7.31499231140722e-06,
+      "loss": 0.5279,
+      "step": 4665
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7582306321020094,
+      "learning_rate": 7.3138649678144155e-06,
+      "loss": 0.5255,
+      "step": 4666
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.242684585752455,
+      "learning_rate": 7.312737474517575e-06,
+      "loss": 0.476,
+      "step": 4667
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6885928532551966,
+      "learning_rate": 7.3116098315896436e-06,
+      "loss": 0.4639,
+      "step": 4668
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2456853204804592,
+      "learning_rate": 7.3104820391035814e-06,
+      "loss": 0.4465,
+      "step": 4669
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1462988778320344,
+      "learning_rate": 7.309354097132352e-06,
+      "loss": 0.5398,
+      "step": 4670
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8372925624109482,
+      "learning_rate": 7.308226005748934e-06,
+      "loss": 0.4976,
+      "step": 4671
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4448518533989736,
+      "learning_rate": 7.307097765026313e-06,
+      "loss": 0.5301,
+      "step": 4672
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.233076634453553,
+      "learning_rate": 7.305969375037486e-06,
+      "loss": 0.4808,
+      "step": 4673
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3905309010745217,
+      "learning_rate": 7.304840835855456e-06,
+      "loss": 0.4619,
+      "step": 4674
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5008392564287607,
+      "learning_rate": 7.303712147553239e-06,
+      "loss": 0.475,
+      "step": 4675
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.580663619157923,
+      "learning_rate": 7.302583310203859e-06,
+      "loss": 0.4783,
+      "step": 4676
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.60986132013244,
+      "learning_rate": 7.301454323880353e-06,
+      "loss": 0.5251,
+      "step": 4677
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5766232508955631,
+      "learning_rate": 7.300325188655762e-06,
+      "loss": 0.5311,
+      "step": 4678
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.63248922056906,
+      "learning_rate": 7.29919590460314e-06,
+      "loss": 0.4626,
+      "step": 4679
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.365804347659826,
+      "learning_rate": 7.298066471795551e-06,
+      "loss": 0.5242,
+      "step": 4680
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.1927090017639683,
+      "learning_rate": 7.296936890306067e-06,
+      "loss": 0.5009,
+      "step": 4681
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.37182540015722,
+      "learning_rate": 7.295807160207771e-06,
+      "loss": 0.4748,
+      "step": 4682
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.914842620531982,
+      "learning_rate": 7.294677281573756e-06,
+      "loss": 0.4925,
+      "step": 4683
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7286300865010455,
+      "learning_rate": 7.293547254477122e-06,
+      "loss": 0.5042,
+      "step": 4684
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4144138954930967,
+      "learning_rate": 7.292417078990982e-06,
+      "loss": 0.4617,
+      "step": 4685
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7640260040014577,
+      "learning_rate": 7.291286755188453e-06,
+      "loss": 0.5248,
+      "step": 4686
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.705506245912563,
+      "learning_rate": 7.290156283142671e-06,
+      "loss": 0.5018,
+      "step": 4687
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9900153301927477,
+      "learning_rate": 7.2890256629267745e-06,
+      "loss": 0.4636,
+      "step": 4688
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6204244068765212,
+      "learning_rate": 7.28789489461391e-06,
+      "loss": 0.4994,
+      "step": 4689
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6324342268063203,
+      "learning_rate": 7.286763978277238e-06,
+      "loss": 0.5194,
+      "step": 4690
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6536256435896809,
+      "learning_rate": 7.2856329139899296e-06,
+      "loss": 0.511,
+      "step": 4691
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9015996167138063,
+      "learning_rate": 7.284501701825162e-06,
+      "loss": 0.497,
+      "step": 4692
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4947597248959605,
+      "learning_rate": 7.283370341856119e-06,
+      "loss": 0.4768,
+      "step": 4693
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4865452842734845,
+      "learning_rate": 7.282238834156006e-06,
+      "loss": 0.4808,
+      "step": 4694
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5054378633856997,
+      "learning_rate": 7.2811071787980245e-06,
+      "loss": 0.4769,
+      "step": 4695
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6115051490176622,
+      "learning_rate": 7.279975375855394e-06,
+      "loss": 0.4704,
+      "step": 4696
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 5.8695037901160205,
+      "learning_rate": 7.278843425401338e-06,
+      "loss": 0.4946,
+      "step": 4697
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9375358684464783,
+      "learning_rate": 7.277711327509094e-06,
+      "loss": 0.462,
+      "step": 4698
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8676380094342842,
+      "learning_rate": 7.276579082251906e-06,
+      "loss": 0.4957,
+      "step": 4699
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7942361489796048,
+      "learning_rate": 7.27544668970303e-06,
+      "loss": 0.458,
+      "step": 4700
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7067408963624273,
+      "learning_rate": 7.2743141499357295e-06,
+      "loss": 0.515,
+      "step": 4701
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3738137346572823,
+      "learning_rate": 7.2731814630232785e-06,
+      "loss": 0.4714,
+      "step": 4702
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7624810386918337,
+      "learning_rate": 7.272048629038961e-06,
+      "loss": 0.4439,
+      "step": 4703
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7887503097455553,
+      "learning_rate": 7.27091564805607e-06,
+      "loss": 0.4486,
+      "step": 4704
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7563305980337875,
+      "learning_rate": 7.269782520147906e-06,
+      "loss": 0.4682,
+      "step": 4705
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9578830878581797,
+      "learning_rate": 7.2686492453877835e-06,
+      "loss": 0.4891,
+      "step": 4706
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4306246809932381,
+      "learning_rate": 7.267515823849021e-06,
+      "loss": 0.4798,
+      "step": 4707
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3130142770195665,
+      "learning_rate": 7.266382255604953e-06,
+      "loss": 0.469,
+      "step": 4708
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0330209398303527,
+      "learning_rate": 7.265248540728915e-06,
+      "loss": 0.4711,
+      "step": 4709
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4413723403377363,
+      "learning_rate": 7.264114679294263e-06,
+      "loss": 0.468,
+      "step": 4710
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.633230928496973,
+      "learning_rate": 7.262980671374351e-06,
+      "loss": 0.5359,
+      "step": 4711
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3920293959348922,
+      "learning_rate": 7.261846517042551e-06,
+      "loss": 0.4707,
+      "step": 4712
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.814928577636905,
+      "learning_rate": 7.26071221637224e-06,
+      "loss": 0.5136,
+      "step": 4713
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4917108149138367,
+      "learning_rate": 7.259577769436806e-06,
+      "loss": 0.5695,
+      "step": 4714
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.773866921124995,
+      "learning_rate": 7.258443176309645e-06,
+      "loss": 0.5175,
+      "step": 4715
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6080369892505022,
+      "learning_rate": 7.257308437064165e-06,
+      "loss": 0.5192,
+      "step": 4716
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9840263076349693,
+      "learning_rate": 7.256173551773783e-06,
+      "loss": 0.4563,
+      "step": 4717
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.496153075933112,
+      "learning_rate": 7.255038520511925e-06,
+      "loss": 0.5158,
+      "step": 4718
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5832219781339507,
+      "learning_rate": 7.253903343352022e-06,
+      "loss": 0.4994,
+      "step": 4719
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.672780103138904,
+      "learning_rate": 7.252768020367523e-06,
+      "loss": 0.4885,
+      "step": 4720
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6815980147872833,
+      "learning_rate": 7.25163255163188e-06,
+      "loss": 0.503,
+      "step": 4721
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8585282963177416,
+      "learning_rate": 7.2504969372185545e-06,
+      "loss": 0.4472,
+      "step": 4722
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5180971806900907,
+      "learning_rate": 7.2493611772010235e-06,
+      "loss": 0.4868,
+      "step": 4723
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.3785406880426674,
+      "learning_rate": 7.2482252716527645e-06,
+      "loss": 0.4713,
+      "step": 4724
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5818437667628589,
+      "learning_rate": 7.2470892206472745e-06,
+      "loss": 0.4527,
+      "step": 4725
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7565609903149626,
+      "learning_rate": 7.245953024258049e-06,
+      "loss": 0.4973,
+      "step": 4726
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.9254773802857275,
+      "learning_rate": 7.244816682558602e-06,
+      "loss": 0.5098,
+      "step": 4727
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5663062738471344,
+      "learning_rate": 7.2436801956224515e-06,
+      "loss": 0.5047,
+      "step": 4728
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.8217040530149746,
+      "learning_rate": 7.242543563523128e-06,
+      "loss": 0.469,
+      "step": 4729
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.933370273578642,
+      "learning_rate": 7.241406786334169e-06,
+      "loss": 0.5443,
+      "step": 4730
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7260818648573564,
+      "learning_rate": 7.240269864129125e-06,
+      "loss": 0.4546,
+      "step": 4731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4425053260266039,
+      "learning_rate": 7.239132796981549e-06,
+      "loss": 0.4787,
+      "step": 4732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2335369793194162,
+      "learning_rate": 7.237995584965012e-06,
+      "loss": 0.4843,
+      "step": 4733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.518058571393306,
+      "learning_rate": 7.2368582281530874e-06,
+      "loss": 0.4405,
+      "step": 4734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5735924443111182,
+      "learning_rate": 7.2357207266193615e-06,
+      "loss": 0.5221,
+      "step": 4735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5775445278768783,
+      "learning_rate": 7.23458308043743e-06,
+      "loss": 0.4745,
+      "step": 4736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4959229900486775,
+      "learning_rate": 7.233445289680896e-06,
+      "loss": 0.4756,
+      "step": 4737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.445940569074372,
+      "learning_rate": 7.232307354423374e-06,
+      "loss": 0.4823,
+      "step": 4738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7181549843338482,
+      "learning_rate": 7.2311692747384844e-06,
+      "loss": 0.4789,
+      "step": 4739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5485213632807011,
+      "learning_rate": 7.230031050699862e-06,
+      "loss": 0.4541,
+      "step": 4740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4601689880243836,
+      "learning_rate": 7.22889268238115e-06,
+      "loss": 0.467,
+      "step": 4741
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.6154505582136087,
+      "learning_rate": 7.227754169855995e-06,
+      "loss": 0.5012,
+      "step": 4742
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0642110931352153,
+      "learning_rate": 7.226615513198061e-06,
+      "loss": 0.4907,
+      "step": 4743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6947524051799303,
+      "learning_rate": 7.225476712481015e-06,
+      "loss": 0.4827,
+      "step": 4744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6002192885577924,
+      "learning_rate": 7.2243377677785375e-06,
+      "loss": 0.4566,
+      "step": 4745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.068949852886102,
+      "learning_rate": 7.2231986791643165e-06,
+      "loss": 0.5054,
+      "step": 4746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.96567998440374,
+      "learning_rate": 7.22205944671205e-06,
+      "loss": 0.5018,
+      "step": 4747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.2318786998201268,
+      "learning_rate": 7.220920070495442e-06,
+      "loss": 0.4569,
+      "step": 4748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 3.0397195478713117,
+      "learning_rate": 7.2197805505882126e-06,
+      "loss": 0.4636,
+      "step": 4749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6227063649748168,
+      "learning_rate": 7.2186408870640835e-06,
+      "loss": 0.5282,
+      "step": 4750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6174938576125719,
+      "learning_rate": 7.217501079996793e-06,
+      "loss": 0.5272,
+      "step": 4751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.1921648849125006,
+      "learning_rate": 7.216361129460082e-06,
+      "loss": 0.4553,
+      "step": 4752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.247493481621767,
+      "learning_rate": 7.215221035527707e-06,
+      "loss": 0.5273,
+      "step": 4753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.4300858176876883,
+      "learning_rate": 7.2140807982734274e-06,
+      "loss": 0.5056,
+      "step": 4754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.58459640088003,
+      "learning_rate": 7.212940417771018e-06,
+      "loss": 0.5025,
+      "step": 4755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5952044885058977,
+      "learning_rate": 7.211799894094258e-06,
+      "loss": 0.5272,
+      "step": 4756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7611399398225924,
+      "learning_rate": 7.2106592273169376e-06,
+      "loss": 0.4701,
+      "step": 4757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4425377151734717,
+      "learning_rate": 7.209518417512858e-06,
+      "loss": 0.5023,
+      "step": 4758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0209301283751606,
+      "learning_rate": 7.208377464755826e-06,
+      "loss": 0.4232,
+      "step": 4759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7073352395793535,
+      "learning_rate": 7.207236369119662e-06,
+      "loss": 0.5717,
+      "step": 4760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5350499596440036,
+      "learning_rate": 7.206095130678192e-06,
+      "loss": 0.5106,
+      "step": 4761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 9.542541609458176,
+      "learning_rate": 7.204953749505252e-06,
+      "loss": 0.4505,
+      "step": 4762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0885494173191135,
+      "learning_rate": 7.2038122256746915e-06,
+      "loss": 0.458,
+      "step": 4763
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.893775557408428,
+      "learning_rate": 7.202670559260359e-06,
+      "loss": 0.4761,
+      "step": 4764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6490055126750536,
+      "learning_rate": 7.2015287503361254e-06,
+      "loss": 0.4904,
+      "step": 4765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.837218531711044,
+      "learning_rate": 7.200386798975863e-06,
+      "loss": 0.4651,
+      "step": 4766
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6892997761287047,
+      "learning_rate": 7.1992447052534495e-06,
+      "loss": 0.5075,
+      "step": 4767
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4618739814609234,
+      "learning_rate": 7.1981024692427835e-06,
+      "loss": 0.481,
+      "step": 4768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6181535962951814,
+      "learning_rate": 7.1969600910177615e-06,
+      "loss": 0.4799,
+      "step": 4769
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.382438497684814,
+      "learning_rate": 7.195817570652297e-06,
+      "loss": 0.4832,
+      "step": 4770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5775223241490135,
+      "learning_rate": 7.194674908220307e-06,
+      "loss": 0.5244,
+      "step": 4771
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5293407628719298,
+      "learning_rate": 7.193532103795723e-06,
+      "loss": 0.4874,
+      "step": 4772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 2.0347785855289833,
+      "learning_rate": 7.1923891574524785e-06,
+      "loss": 0.4983,
+      "step": 4773
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 4.871040264777708,
+      "learning_rate": 7.191246069264526e-06,
+      "loss": 0.4616,
+      "step": 4774
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7134532490523338,
+      "learning_rate": 7.1901028393058185e-06,
+      "loss": 0.4306,
+      "step": 4775
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8216939814528377,
+      "learning_rate": 7.188959467650323e-06,
+      "loss": 0.5158,
+      "step": 4776
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.934450551878168,
+      "learning_rate": 7.187815954372012e-06,
+      "loss": 0.5187,
+      "step": 4777
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.3253752943445383,
+      "learning_rate": 7.186672299544872e-06,
+      "loss": 0.5121,
+      "step": 4778
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.541476990480005,
+      "learning_rate": 7.185528503242894e-06,
+      "loss": 0.5399,
+      "step": 4779
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4478442249182373,
+      "learning_rate": 7.184384565540083e-06,
+      "loss": 0.4331,
+      "step": 4780
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4471939318118097,
+      "learning_rate": 7.1832404865104456e-06,
+      "loss": 0.5462,
+      "step": 4781
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9997887146301945,
+      "learning_rate": 7.182096266228006e-06,
+      "loss": 0.5082,
+      "step": 4782
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6419728978849903,
+      "learning_rate": 7.1809519047667905e-06,
+      "loss": 0.4882,
+      "step": 4783
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8007896785235906,
+      "learning_rate": 7.179807402200842e-06,
+      "loss": 0.5136,
+      "step": 4784
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8438502208463854,
+      "learning_rate": 7.178662758604205e-06,
+      "loss": 0.4861,
+      "step": 4785
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5627457088174193,
+      "learning_rate": 7.1775179740509385e-06,
+      "loss": 0.5245,
+      "step": 4786
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5474543971572006,
+      "learning_rate": 7.176373048615106e-06,
+      "loss": 0.4682,
+      "step": 4787
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 18.04727414902749,
+      "learning_rate": 7.175227982370787e-06,
+      "loss": 0.4773,
+      "step": 4788
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8257081865284526,
+      "learning_rate": 7.174082775392061e-06,
+      "loss": 0.4659,
+      "step": 4789
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5411190225862744,
+      "learning_rate": 7.172937427753025e-06,
+      "loss": 0.5073,
+      "step": 4790
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5990952625617741,
+      "learning_rate": 7.171791939527779e-06,
+      "loss": 0.4928,
+      "step": 4791
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2835301696124328,
+      "learning_rate": 7.170646310790434e-06,
+      "loss": 0.5053,
+      "step": 4792
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.680081901257542,
+      "learning_rate": 7.169500541615115e-06,
+      "loss": 0.477,
+      "step": 4793
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6251597784704082,
+      "learning_rate": 7.168354632075948e-06,
+      "loss": 0.4608,
+      "step": 4794
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5562489837784452,
+      "learning_rate": 7.167208582247072e-06,
+      "loss": 0.4876,
+      "step": 4795
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.9502386103121507,
+      "learning_rate": 7.166062392202637e-06,
+      "loss": 0.5092,
+      "step": 4796
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.010187287820319,
+      "learning_rate": 7.164916062016798e-06,
+      "loss": 0.4915,
+      "step": 4797
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6046809935951256,
+      "learning_rate": 7.163769591763723e-06,
+      "loss": 0.4462,
+      "step": 4798
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3832799618322587,
+      "learning_rate": 7.1626229815175855e-06,
+      "loss": 0.3883,
+      "step": 4799
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.333501992921133,
+      "learning_rate": 7.161476231352569e-06,
+      "loss": 0.4495,
+      "step": 4800
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6268845436130952,
+      "learning_rate": 7.16032934134287e-06,
+      "loss": 0.4943,
+      "step": 4801
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.650064641777069,
+      "learning_rate": 7.159182311562686e-06,
+      "loss": 0.4542,
+      "step": 4802
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6846314137780052,
+      "learning_rate": 7.158035142086234e-06,
+      "loss": 0.4681,
+      "step": 4803
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.9005179214586914,
+      "learning_rate": 7.156887832987729e-06,
+      "loss": 0.5696,
+      "step": 4804
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4852843659572135,
+      "learning_rate": 7.155740384341404e-06,
+      "loss": 0.4521,
+      "step": 4805
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5832144808504138,
+      "learning_rate": 7.154592796221495e-06,
+      "loss": 0.4929,
+      "step": 4806
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5633446584448708,
+      "learning_rate": 7.153445068702252e-06,
+      "loss": 0.5014,
+      "step": 4807
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.563757745098572,
+      "learning_rate": 7.15229720185793e-06,
+      "loss": 0.4516,
+      "step": 4808
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.621305369537592,
+      "learning_rate": 7.1511491957627945e-06,
+      "loss": 0.497,
+      "step": 4809
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.526975849112527,
+      "learning_rate": 7.1500010504911185e-06,
+      "loss": 0.4055,
+      "step": 4810
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.24747164533128,
+      "learning_rate": 7.148852766117189e-06,
+      "loss": 0.4703,
+      "step": 4811
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4960270685763388,
+      "learning_rate": 7.147704342715296e-06,
+      "loss": 0.4623,
+      "step": 4812
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5603263484043122,
+      "learning_rate": 7.146555780359742e-06,
+      "loss": 0.5128,
+      "step": 4813
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6334007579191232,
+      "learning_rate": 7.145407079124835e-06,
+      "loss": 0.5088,
+      "step": 4814
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7452290072680166,
+      "learning_rate": 7.144258239084899e-06,
+      "loss": 0.495,
+      "step": 4815
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4789901181434115,
+      "learning_rate": 7.143109260314259e-06,
+      "loss": 0.4481,
+      "step": 4816
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.940294440961406,
+      "learning_rate": 7.141960142887254e-06,
+      "loss": 0.4561,
+      "step": 4817
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.5760601326236863,
+      "learning_rate": 7.140810886878228e-06,
+      "loss": 0.4819,
+      "step": 4818
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8740071460069843,
+      "learning_rate": 7.13966149236154e-06,
+      "loss": 0.4999,
+      "step": 4819
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2271038742195985,
+      "learning_rate": 7.1385119594115515e-06,
+      "loss": 0.4205,
+      "step": 4820
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8866342324473262,
+      "learning_rate": 7.137362288102638e-06,
+      "loss": 0.4407,
+      "step": 4821
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8296823921926835,
+      "learning_rate": 7.136212478509179e-06,
+      "loss": 0.5038,
+      "step": 4822
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.7668322422326637,
+      "learning_rate": 7.135062530705569e-06,
+      "loss": 0.4907,
+      "step": 4823
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.913982469947087,
+      "learning_rate": 7.133912444766204e-06,
+      "loss": 0.4589,
+      "step": 4824
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.3597901269047026,
+      "learning_rate": 7.132762220765497e-06,
+      "loss": 0.5171,
+      "step": 4825
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3795548796539945,
+      "learning_rate": 7.131611858777863e-06,
+      "loss": 0.4956,
+      "step": 4826
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3193161200408823,
+      "learning_rate": 7.1304613588777315e-06,
+      "loss": 0.4495,
+      "step": 4827
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7026252335249872,
+      "learning_rate": 7.129310721139536e-06,
+      "loss": 0.5246,
+      "step": 4828
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6901845863727984,
+      "learning_rate": 7.128159945637722e-06,
+      "loss": 0.4193,
+      "step": 4829
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5827177853608418,
+      "learning_rate": 7.127009032446744e-06,
+      "loss": 0.5303,
+      "step": 4830
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2866480548892696,
+      "learning_rate": 7.125857981641066e-06,
+      "loss": 0.4727,
+      "step": 4831
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7011752144543928,
+      "learning_rate": 7.1247067932951555e-06,
+      "loss": 0.4251,
+      "step": 4832
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6404625529047874,
+      "learning_rate": 7.123555467483496e-06,
+      "loss": 0.524,
+      "step": 4833
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4976964844189433,
+      "learning_rate": 7.122404004280574e-06,
+      "loss": 0.4778,
+      "step": 4834
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5972232670139491,
+      "learning_rate": 7.121252403760891e-06,
+      "loss": 0.4988,
+      "step": 4835
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2328023132162154,
+      "learning_rate": 7.1201006659989525e-06,
+      "loss": 0.4545,
+      "step": 4836
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6016607389926304,
+      "learning_rate": 7.118948791069273e-06,
+      "loss": 0.5055,
+      "step": 4837
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.826720454821463,
+      "learning_rate": 7.117796779046379e-06,
+      "loss": 0.4753,
+      "step": 4838
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6041532012294353,
+      "learning_rate": 7.116644630004805e-06,
+      "loss": 0.4405,
+      "step": 4839
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5498237787234056,
+      "learning_rate": 7.115492344019091e-06,
+      "loss": 0.5176,
+      "step": 4840
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7361073300515835,
+      "learning_rate": 7.11433992116379e-06,
+      "loss": 0.5187,
+      "step": 4841
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.188239930130391,
+      "learning_rate": 7.113187361513463e-06,
+      "loss": 0.4875,
+      "step": 4842
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.506591375325928,
+      "learning_rate": 7.112034665142677e-06,
+      "loss": 0.44,
+      "step": 4843
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.753477473173662,
+      "learning_rate": 7.110881832126012e-06,
+      "loss": 0.4612,
+      "step": 4844
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3331318780599462,
+      "learning_rate": 7.109728862538054e-06,
+      "loss": 0.4751,
+      "step": 4845
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6212944176183048,
+      "learning_rate": 7.108575756453398e-06,
+      "loss": 0.5153,
+      "step": 4846
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.967168623577125,
+      "learning_rate": 7.107422513946648e-06,
+      "loss": 0.4983,
+      "step": 4847
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.505243696895769,
+      "learning_rate": 7.106269135092419e-06,
+      "loss": 0.4957,
+      "step": 4848
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8629822485856886,
+      "learning_rate": 7.105115619965333e-06,
+      "loss": 0.4579,
+      "step": 4849
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5040150399249184,
+      "learning_rate": 7.10396196864002e-06,
+      "loss": 0.4947,
+      "step": 4850
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8269030564050288,
+      "learning_rate": 7.1028081811911185e-06,
+      "loss": 0.4727,
+      "step": 4851
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.573173475927753,
+      "learning_rate": 7.101654257693279e-06,
+      "loss": 0.5252,
+      "step": 4852
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.912857846292556,
+      "learning_rate": 7.100500198221159e-06,
+      "loss": 0.4342,
+      "step": 4853
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8920831598127514,
+      "learning_rate": 7.099346002849425e-06,
+      "loss": 0.4903,
+      "step": 4854
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4233376215011992,
+      "learning_rate": 7.098191671652747e-06,
+      "loss": 0.5329,
+      "step": 4855
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6720778666982499,
+      "learning_rate": 7.097037204705816e-06,
+      "loss": 0.4724,
+      "step": 4856
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6913831463011904,
+      "learning_rate": 7.095882602083321e-06,
+      "loss": 0.445,
+      "step": 4857
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.419008351806511,
+      "learning_rate": 7.094727863859964e-06,
+      "loss": 0.4933,
+      "step": 4858
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.041095044715757,
+      "learning_rate": 7.093572990110452e-06,
+      "loss": 0.4359,
+      "step": 4859
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7466902650670626,
+      "learning_rate": 7.092417980909508e-06,
+      "loss": 0.4817,
+      "step": 4860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4961182304604086,
+      "learning_rate": 7.091262836331858e-06,
+      "loss": 0.4722,
+      "step": 4861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7217282512976573,
+      "learning_rate": 7.0901075564522385e-06,
+      "loss": 0.5035,
+      "step": 4862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6091739334279902,
+      "learning_rate": 7.088952141345391e-06,
+      "loss": 0.4389,
+      "step": 4863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2911554600974566,
+      "learning_rate": 7.087796591086076e-06,
+      "loss": 0.512,
+      "step": 4864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 7.989701203155342,
+      "learning_rate": 7.086640905749051e-06,
+      "loss": 0.4599,
+      "step": 4865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3138613702983708,
+      "learning_rate": 7.085485085409091e-06,
+      "loss": 0.4569,
+      "step": 4866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.627699335834174,
+      "learning_rate": 7.084329130140972e-06,
+      "loss": 0.5243,
+      "step": 4867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.558303100660825,
+      "learning_rate": 7.083173040019487e-06,
+      "loss": 0.4228,
+      "step": 4868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.1195226362837665,
+      "learning_rate": 7.08201681511943e-06,
+      "loss": 0.4176,
+      "step": 4869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4413023556461042,
+      "learning_rate": 7.080860455515609e-06,
+      "loss": 0.5008,
+      "step": 4870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4560143595922155,
+      "learning_rate": 7.079703961282839e-06,
+      "loss": 0.4527,
+      "step": 4871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.79554529709397,
+      "learning_rate": 7.078547332495942e-06,
+      "loss": 0.4664,
+      "step": 4872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0234189673597034,
+      "learning_rate": 7.077390569229754e-06,
+      "loss": 0.5033,
+      "step": 4873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7639561015200624,
+      "learning_rate": 7.076233671559112e-06,
+      "loss": 0.4258,
+      "step": 4874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 5.1180125315126075,
+      "learning_rate": 7.075076639558868e-06,
+      "loss": 0.4625,
+      "step": 4875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7016107930021975,
+      "learning_rate": 7.073919473303878e-06,
+      "loss": 0.4941,
+      "step": 4876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5795643939678147,
+      "learning_rate": 7.072762172869014e-06,
+      "loss": 0.5151,
+      "step": 4877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.079003508604935,
+      "learning_rate": 7.071604738329148e-06,
+      "loss": 0.4929,
+      "step": 4878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.65421551332718,
+      "learning_rate": 7.0704471697591656e-06,
+      "loss": 0.4375,
+      "step": 4879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6642350307206315,
+      "learning_rate": 7.069289467233959e-06,
+      "loss": 0.51,
+      "step": 4880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4385713241808482,
+      "learning_rate": 7.068131630828432e-06,
+      "loss": 0.5144,
+      "step": 4881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6364476992729464,
+      "learning_rate": 7.066973660617493e-06,
+      "loss": 0.4954,
+      "step": 4882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.647391339356725,
+      "learning_rate": 7.065815556676063e-06,
+      "loss": 0.4559,
+      "step": 4883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5953822832925623,
+      "learning_rate": 7.064657319079068e-06,
+      "loss": 0.5289,
+      "step": 4884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.821152770737596,
+      "learning_rate": 7.063498947901446e-06,
+      "loss": 0.4748,
+      "step": 4885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4681082132647514,
+      "learning_rate": 7.062340443218141e-06,
+      "loss": 0.4291,
+      "step": 4886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7692051834346585,
+      "learning_rate": 7.061181805104107e-06,
+      "loss": 0.4304,
+      "step": 4887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8359531067981039,
+      "learning_rate": 7.060023033634307e-06,
+      "loss": 0.4391,
+      "step": 4888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.8550530975554878,
+      "learning_rate": 7.058864128883711e-06,
+      "loss": 0.5392,
+      "step": 4889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6123365793599493,
+      "learning_rate": 7.0577050909273e-06,
+      "loss": 0.5095,
+      "step": 4890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.593725759136565,
+      "learning_rate": 7.056545919840062e-06,
+      "loss": 0.5162,
+      "step": 4891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.2872937709115053,
+      "learning_rate": 7.055386615696992e-06,
+      "loss": 0.4805,
+      "step": 4892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5970948584744433,
+      "learning_rate": 7.054227178573098e-06,
+      "loss": 0.4979,
+      "step": 4893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.074254067895052,
+      "learning_rate": 7.053067608543392e-06,
+      "loss": 0.458,
+      "step": 4894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.910352809477886,
+      "learning_rate": 7.051907905682898e-06,
+      "loss": 0.4705,
+      "step": 4895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.0055665614568183,
+      "learning_rate": 7.050748070066646e-06,
+      "loss": 0.5051,
+      "step": 4896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.393202642993326,
+      "learning_rate": 7.049588101769675e-06,
+      "loss": 0.5023,
+      "step": 4897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 4.685604545653513,
+      "learning_rate": 7.0484280008670365e-06,
+      "loss": 0.5316,
+      "step": 4898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.2080350043713257,
+      "learning_rate": 7.0472677674337875e-06,
+      "loss": 0.5068,
+      "step": 4899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6182600467843604,
+      "learning_rate": 7.0461074015449906e-06,
+      "loss": 0.4887,
+      "step": 4900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.6754688834906362,
+      "learning_rate": 7.0449469032757224e-06,
+      "loss": 0.4408,
+      "step": 4901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 2.173361702003934,
+      "learning_rate": 7.043786272701063e-06,
+      "loss": 0.4739,
+      "step": 4902
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5653293967148161,
+      "learning_rate": 7.042625509896107e-06,
+      "loss": 0.5038,
+      "step": 4903
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.812792310309564,
+      "learning_rate": 7.041464614935952e-06,
+      "loss": 0.4793,
+      "step": 4904
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.769552483773579,
+      "learning_rate": 7.0403035878957074e-06,
+      "loss": 0.4916,
+      "step": 4905
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5032349139410226,
+      "learning_rate": 7.039142428850489e-06,
+      "loss": 0.4444,
+      "step": 4906
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5519574338753228,
+      "learning_rate": 7.037981137875423e-06,
+      "loss": 0.468,
+      "step": 4907
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5376187289901255,
+      "learning_rate": 7.036819715045644e-06,
+      "loss": 0.518,
+      "step": 4908
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6300988502219944,
+      "learning_rate": 7.035658160436294e-06,
+      "loss": 0.4912,
+      "step": 4909
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.842975930740792,
+      "learning_rate": 7.034496474122523e-06,
+      "loss": 0.4704,
+      "step": 4910
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5617282710368641,
+      "learning_rate": 7.033334656179491e-06,
+      "loss": 0.5221,
+      "step": 4911
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.1337672393430926,
+      "learning_rate": 7.03217270668237e-06,
+      "loss": 0.4271,
+      "step": 4912
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5988202337441175,
+      "learning_rate": 7.031010625706331e-06,
+      "loss": 0.526,
+      "step": 4913
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.8102455719397725,
+      "learning_rate": 7.029848413326561e-06,
+      "loss": 0.4046,
+      "step": 4914
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.569494803861767,
+      "learning_rate": 7.028686069618255e-06,
+      "loss": 0.4939,
+      "step": 4915
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6633172827331617,
+      "learning_rate": 7.027523594656615e-06,
+      "loss": 0.4554,
+      "step": 4916
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5642926104787376,
+      "learning_rate": 7.026360988516848e-06,
+      "loss": 0.5092,
+      "step": 4917
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5261450670091496,
+      "learning_rate": 7.025198251274179e-06,
+      "loss": 0.4174,
+      "step": 4918
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.448012199015533,
+      "learning_rate": 7.024035383003829e-06,
+      "loss": 0.487,
+      "step": 4919
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.537271207249816,
+      "learning_rate": 7.022872383781039e-06,
+      "loss": 0.5191,
+      "step": 4920
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7234610261669818,
+      "learning_rate": 7.021709253681053e-06,
+      "loss": 0.466,
+      "step": 4921
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7796357986476352,
+      "learning_rate": 7.020545992779122e-06,
+      "loss": 0.4971,
+      "step": 4922
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5243852343785267,
+      "learning_rate": 7.019382601150509e-06,
+      "loss": 0.5014,
+      "step": 4923
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.04601509562885,
+      "learning_rate": 7.018219078870484e-06,
+      "loss": 0.5305,
+      "step": 4924
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7641941214772858,
+      "learning_rate": 7.017055426014323e-06,
+      "loss": 0.4978,
+      "step": 4925
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.294434781619565,
+      "learning_rate": 7.015891642657316e-06,
+      "loss": 0.4559,
+      "step": 4926
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5266179607523802,
+      "learning_rate": 7.014727728874757e-06,
+      "loss": 0.5016,
+      "step": 4927
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0349742682609397,
+      "learning_rate": 7.01356368474195e-06,
+      "loss": 0.4782,
+      "step": 4928
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5720871276512539,
+      "learning_rate": 7.012399510334205e-06,
+      "loss": 0.5057,
+      "step": 4929
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5692220507313834,
+      "learning_rate": 7.011235205726845e-06,
+      "loss": 0.5001,
+      "step": 4930
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4774321904303722,
+      "learning_rate": 7.0100707709951965e-06,
+      "loss": 0.4917,
+      "step": 4931
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.6472638458785145,
+      "learning_rate": 7.008906206214601e-06,
+      "loss": 0.5374,
+      "step": 4932
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.535329647149385,
+      "learning_rate": 7.007741511460401e-06,
+      "loss": 0.4825,
+      "step": 4933
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5628796487758504,
+      "learning_rate": 7.006576686807952e-06,
+      "loss": 0.5128,
+      "step": 4934
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5726165167765285,
+      "learning_rate": 7.005411732332615e-06,
+      "loss": 0.5292,
+      "step": 4935
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.216874948484997,
+      "learning_rate": 7.004246648109765e-06,
+      "loss": 0.4737,
+      "step": 4936
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6639112552904567,
+      "learning_rate": 7.003081434214777e-06,
+      "loss": 0.4964,
+      "step": 4937
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.2008482389993125,
+      "learning_rate": 7.001916090723041e-06,
+      "loss": 0.4922,
+      "step": 4938
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5748130365995252,
+      "learning_rate": 7.0007506177099515e-06,
+      "loss": 0.5036,
+      "step": 4939
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.508490163958507,
+      "learning_rate": 6.999585015250916e-06,
+      "loss": 0.4888,
+      "step": 4940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6589962958292217,
+      "learning_rate": 6.998419283421345e-06,
+      "loss": 0.4836,
+      "step": 4941
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4820722143023364,
+      "learning_rate": 6.99725342229666e-06,
+      "loss": 0.5056,
+      "step": 4942
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6344594633301817,
+      "learning_rate": 6.996087431952292e-06,
+      "loss": 0.4868,
+      "step": 4943
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6258585771978618,
+      "learning_rate": 6.994921312463679e-06,
+      "loss": 0.4148,
+      "step": 4944
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7104363721701417,
+      "learning_rate": 6.993755063906266e-06,
+      "loss": 0.5313,
+      "step": 4945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6107469547149282,
+      "learning_rate": 6.992588686355508e-06,
+      "loss": 0.4722,
+      "step": 4946
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5826607330140783,
+      "learning_rate": 6.991422179886871e-06,
+      "loss": 0.5049,
+      "step": 4947
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7770780002171889,
+      "learning_rate": 6.990255544575821e-06,
+      "loss": 0.5097,
+      "step": 4948
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6366698579730952,
+      "learning_rate": 6.989088780497844e-06,
+      "loss": 0.4129,
+      "step": 4949
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 3.089682866935191,
+      "learning_rate": 6.987921887728422e-06,
+      "loss": 0.4679,
+      "step": 4950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6625856514565507,
+      "learning_rate": 6.986754866343056e-06,
+      "loss": 0.4566,
+      "step": 4951
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.751596928925858,
+      "learning_rate": 6.985587716417248e-06,
+      "loss": 0.5104,
+      "step": 4952
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.7713322914392933,
+      "learning_rate": 6.984420438026513e-06,
+      "loss": 0.4836,
+      "step": 4953
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5640731504269751,
+      "learning_rate": 6.983253031246371e-06,
+      "loss": 0.463,
+      "step": 4954
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4377670435686263,
+      "learning_rate": 6.9820854961523545e-06,
+      "loss": 0.4907,
+      "step": 4955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5066571696954416,
+      "learning_rate": 6.980917832819996e-06,
+      "loss": 0.4774,
+      "step": 4956
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.527231053639831,
+      "learning_rate": 6.979750041324849e-06,
+      "loss": 0.4428,
+      "step": 4957
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3284021718067764,
+      "learning_rate": 6.978582121742461e-06,
+      "loss": 0.4766,
+      "step": 4958
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4797033727102322,
+      "learning_rate": 6.9774140741484e-06,
+      "loss": 0.5026,
+      "step": 4959
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6391504308219952,
+      "learning_rate": 6.976245898618234e-06,
+      "loss": 0.5205,
+      "step": 4960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5681791873815025,
+      "learning_rate": 6.975077595227544e-06,
+      "loss": 0.5157,
+      "step": 4961
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6484627628645279,
+      "learning_rate": 6.973909164051916e-06,
+      "loss": 0.4663,
+      "step": 4962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6241573606542122,
+      "learning_rate": 6.9727406051669485e-06,
+      "loss": 0.5233,
+      "step": 4963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6862209364813125,
+      "learning_rate": 6.971571918648245e-06,
+      "loss": 0.4798,
+      "step": 4964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6219436265504572,
+      "learning_rate": 6.970403104571416e-06,
+      "loss": 0.5214,
+      "step": 4965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6918789607352824,
+      "learning_rate": 6.969234163012084e-06,
+      "loss": 0.5242,
+      "step": 4966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6308665631004468,
+      "learning_rate": 6.96806509404588e-06,
+      "loss": 0.5148,
+      "step": 4967
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7037384578755337,
+      "learning_rate": 6.966895897748436e-06,
+      "loss": 0.4893,
+      "step": 4968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.761066223794488,
+      "learning_rate": 6.965726574195403e-06,
+      "loss": 0.5421,
+      "step": 4969
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3874299849755027,
+      "learning_rate": 6.96455712346243e-06,
+      "loss": 0.4649,
+      "step": 4970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9813518125537426,
+      "learning_rate": 6.963387545625183e-06,
+      "loss": 0.4781,
+      "step": 4971
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.4496943273616765,
+      "learning_rate": 6.962217840759329e-06,
+      "loss": 0.5009,
+      "step": 4972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.360374197642996,
+      "learning_rate": 6.961048008940548e-06,
+      "loss": 0.4703,
+      "step": 4973
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6033211934566961,
+      "learning_rate": 6.959878050244526e-06,
+      "loss": 0.5002,
+      "step": 4974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3932231859101576,
+      "learning_rate": 6.958707964746958e-06,
+      "loss": 0.4187,
+      "step": 4975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4570024064795173,
+      "learning_rate": 6.9575377525235464e-06,
+      "loss": 0.4788,
+      "step": 4976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.028896346364392,
+      "learning_rate": 6.956367413650004e-06,
+      "loss": 0.4239,
+      "step": 4977
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0195495783894293,
+      "learning_rate": 6.955196948202047e-06,
+      "loss": 0.4708,
+      "step": 4978
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.615152426391804,
+      "learning_rate": 6.9540263562554085e-06,
+      "loss": 0.4277,
+      "step": 4979
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.645904998245222,
+      "learning_rate": 6.952855637885819e-06,
+      "loss": 0.4369,
+      "step": 4980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7836850273665843,
+      "learning_rate": 6.9516847931690255e-06,
+      "loss": 0.489,
+      "step": 4981
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.776807399496063,
+      "learning_rate": 6.950513822180778e-06,
+      "loss": 0.5456,
+      "step": 4982
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4668380884900674,
+      "learning_rate": 6.9493427249968384e-06,
+      "loss": 0.4615,
+      "step": 4983
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5266440858876407,
+      "learning_rate": 6.948171501692974e-06,
+      "loss": 0.4847,
+      "step": 4984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6462517224193731,
+      "learning_rate": 6.947000152344963e-06,
+      "loss": 0.5371,
+      "step": 4985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7054615723588076,
+      "learning_rate": 6.945828677028588e-06,
+      "loss": 0.4966,
+      "step": 4986
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5794148374695198,
+      "learning_rate": 6.944657075819643e-06,
+      "loss": 0.5106,
+      "step": 4987
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.773895273300592,
+      "learning_rate": 6.943485348793929e-06,
+      "loss": 0.5147,
+      "step": 4988
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6535908386708804,
+      "learning_rate": 6.942313496027255e-06,
+      "loss": 0.482,
+      "step": 4989
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.971117171558713,
+      "learning_rate": 6.941141517595441e-06,
+      "loss": 0.4939,
+      "step": 4990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7068980782744823,
+      "learning_rate": 6.939969413574308e-06,
+      "loss": 0.453,
+      "step": 4991
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.649412592339443,
+      "learning_rate": 6.938797184039694e-06,
+      "loss": 0.4434,
+      "step": 4992
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7180376482864714,
+      "learning_rate": 6.937624829067435e-06,
+      "loss": 0.4605,
+      "step": 4993
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5907463790788607,
+      "learning_rate": 6.936452348733388e-06,
+      "loss": 0.4487,
+      "step": 4994
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6179302612445647,
+      "learning_rate": 6.935279743113404e-06,
+      "loss": 0.4919,
+      "step": 4995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.3374607579106397,
+      "learning_rate": 6.934107012283355e-06,
+      "loss": 0.4261,
+      "step": 4996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0150125914072996,
+      "learning_rate": 6.932934156319111e-06,
+      "loss": 0.4135,
+      "step": 4997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6095539144631388,
+      "learning_rate": 6.931761175296556e-06,
+      "loss": 0.5184,
+      "step": 4998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9333089318539403,
+      "learning_rate": 6.930588069291578e-06,
+      "loss": 0.4979,
+      "step": 4999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5852473633739683,
+      "learning_rate": 6.9294148383800805e-06,
+      "loss": 0.4437,
+      "step": 5000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5861917108381923,
+      "learning_rate": 6.928241482637965e-06,
+      "loss": 0.5119,
+      "step": 5001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.592933565810781,
+      "learning_rate": 6.9270680021411495e-06,
+      "loss": 0.4842,
+      "step": 5002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.495995364739113,
+      "learning_rate": 6.9258943969655545e-06,
+      "loss": 0.515,
+      "step": 5003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8575168537024334,
+      "learning_rate": 6.924720667187112e-06,
+      "loss": 0.4812,
+      "step": 5004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 4.411904756499402,
+      "learning_rate": 6.923546812881759e-06,
+      "loss": 0.42,
+      "step": 5005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5382047958678762,
+      "learning_rate": 6.922372834125445e-06,
+      "loss": 0.5215,
+      "step": 5006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6800668625975574,
+      "learning_rate": 6.921198730994121e-06,
+      "loss": 0.4944,
+      "step": 5007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4541323242411728,
+      "learning_rate": 6.920024503563755e-06,
+      "loss": 0.4905,
+      "step": 5008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6992816469813739,
+      "learning_rate": 6.918850151910313e-06,
+      "loss": 0.472,
+      "step": 5009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5991591128247871,
+      "learning_rate": 6.917675676109777e-06,
+      "loss": 0.5279,
+      "step": 5010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6515537713584012,
+      "learning_rate": 6.916501076238135e-06,
+      "loss": 0.5286,
+      "step": 5011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5982021288327655,
+      "learning_rate": 6.915326352371379e-06,
+      "loss": 0.5203,
+      "step": 5012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.46926604251752,
+      "learning_rate": 6.914151504585513e-06,
+      "loss": 0.5023,
+      "step": 5013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6968138476179577,
+      "learning_rate": 6.9129765329565515e-06,
+      "loss": 0.4554,
+      "step": 5014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6001106094899922,
+      "learning_rate": 6.9118014375605085e-06,
+      "loss": 0.5043,
+      "step": 5015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.7664668180716354,
+      "learning_rate": 6.910626218473414e-06,
+      "loss": 0.4929,
+      "step": 5016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4259177056827834,
+      "learning_rate": 6.909450875771302e-06,
+      "loss": 0.4407,
+      "step": 5017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.0595849998000553,
+      "learning_rate": 6.908275409530216e-06,
+      "loss": 0.4982,
+      "step": 5018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.9938437859447415,
+      "learning_rate": 6.907099819826208e-06,
+      "loss": 0.5125,
+      "step": 5019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6297257268074141,
+      "learning_rate": 6.905924106735337e-06,
+      "loss": 0.5149,
+      "step": 5020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3384022944921539,
+      "learning_rate": 6.9047482703336676e-06,
+      "loss": 0.4811,
+      "step": 5021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5586108762331874,
+      "learning_rate": 6.903572310697279e-06,
+      "loss": 0.5208,
+      "step": 5022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5892949338910995,
+      "learning_rate": 6.902396227902252e-06,
+      "loss": 0.5131,
+      "step": 5023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6068212091860247,
+      "learning_rate": 6.901220022024676e-06,
+      "loss": 0.5122,
+      "step": 5024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4471972566121922,
+      "learning_rate": 6.900043693140653e-06,
+      "loss": 0.4644,
+      "step": 5025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.6808294318402828,
+      "learning_rate": 6.898867241326288e-06,
+      "loss": 0.5291,
+      "step": 5026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6672123741262052,
+      "learning_rate": 6.897690666657697e-06,
+      "loss": 0.5338,
+      "step": 5027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.740739940005463,
+      "learning_rate": 6.896513969211003e-06,
+      "loss": 0.467,
+      "step": 5028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5457317698055332,
+      "learning_rate": 6.8953371490623355e-06,
+      "loss": 0.5082,
+      "step": 5029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5948797384957327,
+      "learning_rate": 6.8941602062878335e-06,
+      "loss": 0.4922,
+      "step": 5030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6854107091560475,
+      "learning_rate": 6.892983140963645e-06,
+      "loss": 0.4846,
+      "step": 5031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9888312521075242,
+      "learning_rate": 6.891805953165921e-06,
+      "loss": 0.4567,
+      "step": 5032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1897044304057816,
+      "learning_rate": 6.8906286429708294e-06,
+      "loss": 0.4709,
+      "step": 5033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.899102860886273,
+      "learning_rate": 6.889451210454536e-06,
+      "loss": 0.4355,
+      "step": 5034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5769336075461506,
+      "learning_rate": 6.888273655693223e-06,
+      "loss": 0.4765,
+      "step": 5035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5473246640734752,
+      "learning_rate": 6.887095978763072e-06,
+      "loss": 0.4925,
+      "step": 5036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.358660447223554,
+      "learning_rate": 6.885918179740283e-06,
+      "loss": 0.4444,
+      "step": 5037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5457213355824386,
+      "learning_rate": 6.884740258701052e-06,
+      "loss": 0.4691,
+      "step": 5038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6208181650465323,
+      "learning_rate": 6.8835622157215944e-06,
+      "loss": 0.4739,
+      "step": 5039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9452381061445998,
+      "learning_rate": 6.882384050878124e-06,
+      "loss": 0.4891,
+      "step": 5040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4196534991024101,
+      "learning_rate": 6.881205764246867e-06,
+      "loss": 0.4824,
+      "step": 5041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9702615180159835,
+      "learning_rate": 6.88002735590406e-06,
+      "loss": 0.5067,
+      "step": 5042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7120684880079208,
+      "learning_rate": 6.878848825925941e-06,
+      "loss": 0.5151,
+      "step": 5043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9020921234282075,
+      "learning_rate": 6.877670174388761e-06,
+      "loss": 0.4351,
+      "step": 5044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.62164328954833,
+      "learning_rate": 6.876491401368778e-06,
+      "loss": 0.5042,
+      "step": 5045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5958142556383952,
+      "learning_rate": 6.875312506942254e-06,
+      "loss": 0.504,
+      "step": 5046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6822996627046238,
+      "learning_rate": 6.874133491185466e-06,
+      "loss": 0.501,
+      "step": 5047
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.484340339136958,
+      "learning_rate": 6.872954354174692e-06,
+      "loss": 0.4374,
+      "step": 5048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6242036702779448,
+      "learning_rate": 6.8717750959862225e-06,
+      "loss": 0.4212,
+      "step": 5049
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4767837839883453,
+      "learning_rate": 6.870595716696352e-06,
+      "loss": 0.4539,
+      "step": 5050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6538931289029237,
+      "learning_rate": 6.869416216381386e-06,
+      "loss": 0.507,
+      "step": 5051
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6434749681529897,
+      "learning_rate": 6.8682365951176355e-06,
+      "loss": 0.5023,
+      "step": 5052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8125848315759476,
+      "learning_rate": 6.867056852981422e-06,
+      "loss": 0.4448,
+      "step": 5053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.567883881181303,
+      "learning_rate": 6.865876990049073e-06,
+      "loss": 0.4841,
+      "step": 5054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.074260794422984,
+      "learning_rate": 6.864697006396922e-06,
+      "loss": 0.4689,
+      "step": 5055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0681580972378173,
+      "learning_rate": 6.863516902101315e-06,
+      "loss": 0.4922,
+      "step": 5056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7263712909998254,
+      "learning_rate": 6.862336677238603e-06,
+      "loss": 0.5126,
+      "step": 5057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.1086741469790637,
+      "learning_rate": 6.861156331885142e-06,
+      "loss": 0.4542,
+      "step": 5058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.990019722348448,
+      "learning_rate": 6.859975866117302e-06,
+      "loss": 0.489,
+      "step": 5059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.835797742563605,
+      "learning_rate": 6.858795280011458e-06,
+      "loss": 0.4146,
+      "step": 5060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.013695076338349,
+      "learning_rate": 6.8576145736439894e-06,
+      "loss": 0.492,
+      "step": 5061
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5825735448989022,
+      "learning_rate": 6.856433747091289e-06,
+      "loss": 0.4669,
+      "step": 5062
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.041998234240429,
+      "learning_rate": 6.8552528004297525e-06,
+      "loss": 0.4556,
+      "step": 5063
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.0204251168558325,
+      "learning_rate": 6.854071733735789e-06,
+      "loss": 0.4285,
+      "step": 5064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3358851395831104,
+      "learning_rate": 6.852890547085808e-06,
+      "loss": 0.5116,
+      "step": 5065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.454788710172959,
+      "learning_rate": 6.851709240556233e-06,
+      "loss": 0.4605,
+      "step": 5066
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3909764108674103,
+      "learning_rate": 6.850527814223491e-06,
+      "loss": 0.4652,
+      "step": 5067
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6852454397915315,
+      "learning_rate": 6.849346268164022e-06,
+      "loss": 0.4992,
+      "step": 5068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7197517532667954,
+      "learning_rate": 6.848164602454268e-06,
+      "loss": 0.479,
+      "step": 5069
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8338719435993904,
+      "learning_rate": 6.846982817170682e-06,
+      "loss": 0.5224,
+      "step": 5070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.042581600498176,
+      "learning_rate": 6.845800912389724e-06,
+      "loss": 0.48,
+      "step": 5071
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2276511238894865,
+      "learning_rate": 6.844618888187861e-06,
+      "loss": 0.4976,
+      "step": 5072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8194930783028291,
+      "learning_rate": 6.8434367446415706e-06,
+      "loss": 0.5325,
+      "step": 5073
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7222285396557782,
+      "learning_rate": 6.8422544818273336e-06,
+      "loss": 0.4397,
+      "step": 5074
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.514281054707239,
+      "learning_rate": 6.841072099821641e-06,
+      "loss": 0.514,
+      "step": 5075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5709793971631403,
+      "learning_rate": 6.839889598700993e-06,
+      "loss": 0.4978,
+      "step": 5076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6375026708296558,
+      "learning_rate": 6.8387069785418935e-06,
+      "loss": 0.539,
+      "step": 5077
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5368866569174415,
+      "learning_rate": 6.837524239420858e-06,
+      "loss": 0.514,
+      "step": 5078
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.055758316298568,
+      "learning_rate": 6.8363413814144085e-06,
+      "loss": 0.4928,
+      "step": 5079
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6996253777790984,
+      "learning_rate": 6.835158404599074e-06,
+      "loss": 0.4372,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4522076917106026,
+      "learning_rate": 6.833975309051391e-06,
+      "loss": 0.4089,
+      "step": 5081
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4830301659964606,
+      "learning_rate": 6.832792094847906e-06,
+      "loss": 0.4518,
+      "step": 5082
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6691377579756101,
+      "learning_rate": 6.8316087620651675e-06,
+      "loss": 0.5209,
+      "step": 5083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7214189735591864,
+      "learning_rate": 6.83042531077974e-06,
+      "loss": 0.5389,
+      "step": 5084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.53216226105628,
+      "learning_rate": 6.829241741068189e-06,
+      "loss": 0.4835,
+      "step": 5085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7268708969373763,
+      "learning_rate": 6.828058053007091e-06,
+      "loss": 0.4778,
+      "step": 5086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5277391432617244,
+      "learning_rate": 6.826874246673027e-06,
+      "loss": 0.4889,
+      "step": 5087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9623051911608813,
+      "learning_rate": 6.825690322142589e-06,
+      "loss": 0.4962,
+      "step": 5088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.597356028652685,
+      "learning_rate": 6.824506279492375e-06,
+      "loss": 0.5142,
+      "step": 5089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8832376881614035,
+      "learning_rate": 6.823322118798994e-06,
+      "loss": 0.4629,
+      "step": 5090
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4613390879183994,
+      "learning_rate": 6.822137840139056e-06,
+      "loss": 0.4577,
+      "step": 5091
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.757873435658759,
+      "learning_rate": 6.820953443589184e-06,
+      "loss": 0.4874,
+      "step": 5092
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5930120215922154,
+      "learning_rate": 6.819768929226004e-06,
+      "loss": 0.5042,
+      "step": 5093
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4284865955310406,
+      "learning_rate": 6.818584297126157e-06,
+      "loss": 0.4927,
+      "step": 5094
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8485465888131762,
+      "learning_rate": 6.817399547366284e-06,
+      "loss": 0.514,
+      "step": 5095
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9476553190637653,
+      "learning_rate": 6.816214680023037e-06,
+      "loss": 0.475,
+      "step": 5096
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.184890202468165,
+      "learning_rate": 6.815029695173079e-06,
+      "loss": 0.482,
+      "step": 5097
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6043139318427282,
+      "learning_rate": 6.813844592893071e-06,
+      "loss": 0.5152,
+      "step": 5098
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.472579417349049,
+      "learning_rate": 6.812659373259693e-06,
+      "loss": 0.4806,
+      "step": 5099
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.67112780929482,
+      "learning_rate": 6.811474036349622e-06,
+      "loss": 0.4964,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9374411363238244,
+      "learning_rate": 6.810288582239553e-06,
+      "loss": 0.4313,
+      "step": 5101
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9371917516419939,
+      "learning_rate": 6.80910301100618e-06,
+      "loss": 0.463,
+      "step": 5102
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.012006334476266,
+      "learning_rate": 6.807917322726208e-06,
+      "loss": 0.4478,
+      "step": 5103
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7502839959317843,
+      "learning_rate": 6.806731517476351e-06,
+      "loss": 0.5153,
+      "step": 5104
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9161902169051828,
+      "learning_rate": 6.805545595333329e-06,
+      "loss": 0.4599,
+      "step": 5105
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6416697170671828,
+      "learning_rate": 6.804359556373868e-06,
+      "loss": 0.4587,
+      "step": 5106
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4945869418050486,
+      "learning_rate": 6.803173400674705e-06,
+      "loss": 0.4337,
+      "step": 5107
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8390498381536393,
+      "learning_rate": 6.8019871283125816e-06,
+      "loss": 0.4392,
+      "step": 5108
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.788199804482733,
+      "learning_rate": 6.800800739364248e-06,
+      "loss": 0.4111,
+      "step": 5109
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.4797201670456888,
+      "learning_rate": 6.799614233906462e-06,
+      "loss": 0.5213,
+      "step": 5110
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7597324596539268,
+      "learning_rate": 6.798427612015991e-06,
+      "loss": 0.4614,
+      "step": 5111
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5580322285437369,
+      "learning_rate": 6.7972408737696025e-06,
+      "loss": 0.4956,
+      "step": 5112
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6019729206448499,
+      "learning_rate": 6.796054019244084e-06,
+      "loss": 0.4699,
+      "step": 5113
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6152465694304937,
+      "learning_rate": 6.794867048516218e-06,
+      "loss": 0.4956,
+      "step": 5114
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.877289048317339,
+      "learning_rate": 6.793679961662804e-06,
+      "loss": 0.4993,
+      "step": 5115
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2614526411451534,
+      "learning_rate": 6.792492758760641e-06,
+      "loss": 0.4604,
+      "step": 5116
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5814892556842117,
+      "learning_rate": 6.7913054398865416e-06,
+      "loss": 0.4687,
+      "step": 5117
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6732843162986006,
+      "learning_rate": 6.790118005117325e-06,
+      "loss": 0.4936,
+      "step": 5118
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.9690469963373776,
+      "learning_rate": 6.788930454529814e-06,
+      "loss": 0.4711,
+      "step": 5119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.585883217429109,
+      "learning_rate": 6.7877427882008426e-06,
+      "loss": 0.527,
+      "step": 5120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.944685675750166,
+      "learning_rate": 6.786555006207253e-06,
+      "loss": 0.4816,
+      "step": 5121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.7840383704541263,
+      "learning_rate": 6.7853671086258896e-06,
+      "loss": 0.4433,
+      "step": 5122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8875140855495454,
+      "learning_rate": 6.78417909553361e-06,
+      "loss": 0.4919,
+      "step": 5123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.816861391423662,
+      "learning_rate": 6.782990967007277e-06,
+      "loss": 0.4539,
+      "step": 5124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6823313936628466,
+      "learning_rate": 6.781802723123762e-06,
+      "loss": 0.4877,
+      "step": 5125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5426109533817866,
+      "learning_rate": 6.780614363959941e-06,
+      "loss": 0.519,
+      "step": 5126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.930082166938152,
+      "learning_rate": 6.779425889592701e-06,
+      "loss": 0.4974,
+      "step": 5127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 4.763846432451691,
+      "learning_rate": 6.778237300098933e-06,
+      "loss": 0.4655,
+      "step": 5128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.372060450554244,
+      "learning_rate": 6.7770485955555374e-06,
+      "loss": 0.4505,
+      "step": 5129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8200617782366366,
+      "learning_rate": 6.775859776039423e-06,
+      "loss": 0.4604,
+      "step": 5130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.835488610331581,
+      "learning_rate": 6.774670841627504e-06,
+      "loss": 0.4099,
+      "step": 5131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6428221814049495,
+      "learning_rate": 6.773481792396703e-06,
+      "loss": 0.4792,
+      "step": 5132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.886474408203313,
+      "learning_rate": 6.77229262842395e-06,
+      "loss": 0.4879,
+      "step": 5133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.372255899616415,
+      "learning_rate": 6.7711033497861826e-06,
+      "loss": 0.4588,
+      "step": 5134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.8092594708960505,
+      "learning_rate": 6.769913956560346e-06,
+      "loss": 0.4789,
+      "step": 5135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3133346123135095,
+      "learning_rate": 6.7687244488233896e-06,
+      "loss": 0.4404,
+      "step": 5136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3790825850808666,
+      "learning_rate": 6.767534826652276e-06,
+      "loss": 0.4431,
+      "step": 5137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5633307773545179,
+      "learning_rate": 6.766345090123973e-06,
+      "loss": 0.4982,
+      "step": 5138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.8218850600908645,
+      "learning_rate": 6.765155239315452e-06,
+      "loss": 0.4782,
+      "step": 5139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.540806276947367,
+      "learning_rate": 6.763965274303697e-06,
+      "loss": 0.4978,
+      "step": 5140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6382920812802584,
+      "learning_rate": 6.762775195165695e-06,
+      "loss": 0.5041,
+      "step": 5141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4138855922406606,
+      "learning_rate": 6.761585001978446e-06,
+      "loss": 0.468,
+      "step": 5142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.5734936592979416,
+      "learning_rate": 6.760394694818949e-06,
+      "loss": 0.47,
+      "step": 5143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5689093762651634,
+      "learning_rate": 6.75920427376422e-06,
+      "loss": 0.4637,
+      "step": 5144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5358992206275358,
+      "learning_rate": 6.758013738891275e-06,
+      "loss": 0.5323,
+      "step": 5145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.6660831689086,
+      "learning_rate": 6.7568230902771415e-06,
+      "loss": 0.4395,
+      "step": 5146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.318270577016534,
+      "learning_rate": 6.755632327998851e-06,
+      "loss": 0.4843,
+      "step": 5147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5369927891902775,
+      "learning_rate": 6.754441452133447e-06,
+      "loss": 0.4713,
+      "step": 5148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.6735658159841027,
+      "learning_rate": 6.753250462757975e-06,
+      "loss": 0.4987,
+      "step": 5149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5844973583263949,
+      "learning_rate": 6.752059359949493e-06,
+      "loss": 0.5098,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.256569937367476,
+      "learning_rate": 6.750868143785062e-06,
+      "loss": 0.4581,
+      "step": 5151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 3.6939315760303884,
+      "learning_rate": 6.749676814341752e-06,
+      "loss": 0.4705,
+      "step": 5152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.396676587127883,
+      "learning_rate": 6.748485371696642e-06,
+      "loss": 0.4804,
+      "step": 5153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.408672423474548,
+      "learning_rate": 6.747293815926814e-06,
+      "loss": 0.4633,
+      "step": 5154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5473062500630678,
+      "learning_rate": 6.746102147109364e-06,
+      "loss": 0.5104,
+      "step": 5155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5383310082158282,
+      "learning_rate": 6.744910365321388e-06,
+      "loss": 0.5001,
+      "step": 5156
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0022295449406244,
+      "learning_rate": 6.7437184706399925e-06,
+      "loss": 0.4638,
+      "step": 5157
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7379963450277462,
+      "learning_rate": 6.742526463142295e-06,
+      "loss": 0.5033,
+      "step": 5158
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.791391071002403,
+      "learning_rate": 6.7413343429054134e-06,
+      "loss": 0.5117,
+      "step": 5159
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9809029555015993,
+      "learning_rate": 6.74014211000648e-06,
+      "loss": 0.4866,
+      "step": 5160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5181030213362454,
+      "learning_rate": 6.738949764522627e-06,
+      "loss": 0.479,
+      "step": 5161
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.683352851769196,
+      "learning_rate": 6.737757306531e-06,
+      "loss": 0.4369,
+      "step": 5162
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5075952887977475,
+      "learning_rate": 6.736564736108747e-06,
+      "loss": 0.4776,
+      "step": 5163
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.728214021905995,
+      "learning_rate": 6.735372053333028e-06,
+      "loss": 0.4914,
+      "step": 5164
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2749612902151166,
+      "learning_rate": 6.734179258281007e-06,
+      "loss": 0.4366,
+      "step": 5165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5759132156446712,
+      "learning_rate": 6.732986351029858e-06,
+      "loss": 0.5033,
+      "step": 5166
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4825462206876123,
+      "learning_rate": 6.731793331656757e-06,
+      "loss": 0.4842,
+      "step": 5167
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5870800851234808,
+      "learning_rate": 6.730600200238895e-06,
+      "loss": 0.4562,
+      "step": 5168
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5816448653373989,
+      "learning_rate": 6.729406956853462e-06,
+      "loss": 0.496,
+      "step": 5169
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3055098524808837,
+      "learning_rate": 6.728213601577664e-06,
+      "loss": 0.4419,
+      "step": 5170
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4489027760800093,
+      "learning_rate": 6.727020134488703e-06,
+      "loss": 0.4981,
+      "step": 5171
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4717592959351993,
+      "learning_rate": 6.7258265556638e-06,
+      "loss": 0.4994,
+      "step": 5172
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6877154486116341,
+      "learning_rate": 6.724632865180178e-06,
+      "loss": 0.5099,
+      "step": 5173
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.524438121444279,
+      "learning_rate": 6.723439063115065e-06,
+      "loss": 0.4928,
+      "step": 5174
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0281202673479983,
+      "learning_rate": 6.722245149545698e-06,
+      "loss": 0.4755,
+      "step": 5175
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.122354332438997,
+      "learning_rate": 6.721051124549325e-06,
+      "loss": 0.437,
+      "step": 5176
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5876003183112888,
+      "learning_rate": 6.719856988203195e-06,
+      "loss": 0.5222,
+      "step": 5177
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.656973280247856,
+      "learning_rate": 6.718662740584566e-06,
+      "loss": 0.5309,
+      "step": 5178
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5107660477620135,
+      "learning_rate": 6.717468381770707e-06,
+      "loss": 0.4771,
+      "step": 5179
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3422963641132817,
+      "learning_rate": 6.71627391183889e-06,
+      "loss": 0.5075,
+      "step": 5180
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8561612201313555,
+      "learning_rate": 6.715079330866397e-06,
+      "loss": 0.5016,
+      "step": 5181
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8582564708903415,
+      "learning_rate": 6.7138846389305146e-06,
+      "loss": 0.4801,
+      "step": 5182
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8099783992796339,
+      "learning_rate": 6.712689836108538e-06,
+      "loss": 0.4694,
+      "step": 5183
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6001959069424734,
+      "learning_rate": 6.711494922477769e-06,
+      "loss": 0.498,
+      "step": 5184
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2461196404636494,
+      "learning_rate": 6.7102998981155186e-06,
+      "loss": 0.4694,
+      "step": 5185
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6101795973735369,
+      "learning_rate": 6.7091047630991015e-06,
+      "loss": 0.4945,
+      "step": 5186
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.013471680506151,
+      "learning_rate": 6.707909517505842e-06,
+      "loss": 0.5189,
+      "step": 5187
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.130140031057444,
+      "learning_rate": 6.7067141614130706e-06,
+      "loss": 0.4961,
+      "step": 5188
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7430828994153171,
+      "learning_rate": 6.705518694898127e-06,
+      "loss": 0.4723,
+      "step": 5189
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6180855723492615,
+      "learning_rate": 6.7043231180383525e-06,
+      "loss": 0.4925,
+      "step": 5190
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5532595581235469,
+      "learning_rate": 6.7031274309111025e-06,
+      "loss": 0.4605,
+      "step": 5191
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6365111908081824,
+      "learning_rate": 6.701931633593737e-06,
+      "loss": 0.5351,
+      "step": 5192
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8057489027771978,
+      "learning_rate": 6.700735726163621e-06,
+      "loss": 0.4959,
+      "step": 5193
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5028535403761758,
+      "learning_rate": 6.6995397086981275e-06,
+      "loss": 0.464,
+      "step": 5194
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7377697380807429,
+      "learning_rate": 6.698343581274639e-06,
+      "loss": 0.4633,
+      "step": 5195
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.909377235500752,
+      "learning_rate": 6.6971473439705415e-06,
+      "loss": 0.5111,
+      "step": 5196
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7022296834058048,
+      "learning_rate": 6.695950996863232e-06,
+      "loss": 0.459,
+      "step": 5197
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4881022404527027,
+      "learning_rate": 6.694754540030111e-06,
+      "loss": 0.4255,
+      "step": 5198
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4060241611870998,
+      "learning_rate": 6.693557973548589e-06,
+      "loss": 0.4563,
+      "step": 5199
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5463512279335623,
+      "learning_rate": 6.69236129749608e-06,
+      "loss": 0.4563,
+      "step": 5200
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9699699615731752,
+      "learning_rate": 6.69116451195001e-06,
+      "loss": 0.4764,
+      "step": 5201
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.597761175524961,
+      "learning_rate": 6.689967616987808e-06,
+      "loss": 0.5245,
+      "step": 5202
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.641397434075564,
+      "learning_rate": 6.688770612686912e-06,
+      "loss": 0.4316,
+      "step": 5203
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7488819032200948,
+      "learning_rate": 6.687573499124766e-06,
+      "loss": 0.4898,
+      "step": 5204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4509353322692586,
+      "learning_rate": 6.686376276378825e-06,
+      "loss": 0.414,
+      "step": 5205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.963802777426502,
+      "learning_rate": 6.685178944526543e-06,
+      "loss": 0.4406,
+      "step": 5206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3961606867075198,
+      "learning_rate": 6.683981503645387e-06,
+      "loss": 0.4023,
+      "step": 5207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0514652783941534,
+      "learning_rate": 6.682783953812832e-06,
+      "loss": 0.4219,
+      "step": 5208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2103422495308553,
+      "learning_rate": 6.681586295106355e-06,
+      "loss": 0.4817,
+      "step": 5209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6063825393558706,
+      "learning_rate": 6.680388527603447e-06,
+      "loss": 0.5349,
+      "step": 5210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9857657937013642,
+      "learning_rate": 6.679190651381597e-06,
+      "loss": 0.4543,
+      "step": 5211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.761141163030882,
+      "learning_rate": 6.67799266651831e-06,
+      "loss": 0.4482,
+      "step": 5212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4166420162947582,
+      "learning_rate": 6.67679457309109e-06,
+      "loss": 0.4591,
+      "step": 5213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5497659793992562,
+      "learning_rate": 6.675596371177457e-06,
+      "loss": 0.5036,
+      "step": 5214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6662252562868136,
+      "learning_rate": 6.674398060854931e-06,
+      "loss": 0.4409,
+      "step": 5215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5831615225596714,
+      "learning_rate": 6.67319964220104e-06,
+      "loss": 0.5103,
+      "step": 5216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.391021813783733,
+      "learning_rate": 6.672001115293321e-06,
+      "loss": 0.438,
+      "step": 5217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.526564763541861,
+      "learning_rate": 6.670802480209318e-06,
+      "loss": 0.514,
+      "step": 5218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5586349988447058,
+      "learning_rate": 6.66960373702658e-06,
+      "loss": 0.4968,
+      "step": 5219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9941590817687156,
+      "learning_rate": 6.668404885822663e-06,
+      "loss": 0.4827,
+      "step": 5220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5054471459826764,
+      "learning_rate": 6.667205926675134e-06,
+      "loss": 0.5138,
+      "step": 5221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.755948983894883,
+      "learning_rate": 6.666006859661562e-06,
+      "loss": 0.4665,
+      "step": 5222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.699843234843785,
+      "learning_rate": 6.6648076848595255e-06,
+      "loss": 0.4854,
+      "step": 5223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.546433115047816,
+      "learning_rate": 6.663608402346611e-06,
+      "loss": 0.5128,
+      "step": 5224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.960047049284606,
+      "learning_rate": 6.662409012200406e-06,
+      "loss": 0.4988,
+      "step": 5225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8118074402567652,
+      "learning_rate": 6.661209514498515e-06,
+      "loss": 0.4745,
+      "step": 5226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 3.479324669206233,
+      "learning_rate": 6.66000990931854e-06,
+      "loss": 0.4756,
+      "step": 5227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.200758377931382,
+      "learning_rate": 6.658810196738097e-06,
+      "loss": 0.4749,
+      "step": 5228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5939484553061328,
+      "learning_rate": 6.657610376834802e-06,
+      "loss": 0.5059,
+      "step": 5229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.736398109432895,
+      "learning_rate": 6.656410449686287e-06,
+      "loss": 0.4595,
+      "step": 5230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7543401167212418,
+      "learning_rate": 6.655210415370181e-06,
+      "loss": 0.4557,
+      "step": 5231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5575651431699798,
+      "learning_rate": 6.654010273964128e-06,
+      "loss": 0.4973,
+      "step": 5232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8723826771564962,
+      "learning_rate": 6.652810025545772e-06,
+      "loss": 0.4687,
+      "step": 5233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3513624426297146,
+      "learning_rate": 6.65160967019277e-06,
+      "loss": 0.4775,
+      "step": 5234
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3761604741248672,
+      "learning_rate": 6.6504092079827835e-06,
+      "loss": 0.4067,
+      "step": 5235
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6210921615464615,
+      "learning_rate": 6.64920863899348e-06,
+      "loss": 0.429,
+      "step": 5236
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5851166777395179,
+      "learning_rate": 6.648007963302534e-06,
+      "loss": 0.4991,
+      "step": 5237
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5427329433971466,
+      "learning_rate": 6.64680718098763e-06,
+      "loss": 0.4966,
+      "step": 5238
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2799983931939845,
+      "learning_rate": 6.645606292126455e-06,
+      "loss": 0.4487,
+      "step": 5239
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0208444968772286,
+      "learning_rate": 6.6444052967967065e-06,
+      "loss": 0.4839,
+      "step": 5240
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5992421661935375,
+      "learning_rate": 6.643204195076085e-06,
+      "loss": 0.5105,
+      "step": 5241
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 6.017277193376594,
+      "learning_rate": 6.642002987042302e-06,
+      "loss": 0.4654,
+      "step": 5242
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5396442439280308,
+      "learning_rate": 6.640801672773075e-06,
+      "loss": 0.501,
+      "step": 5243
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.740682832848836,
+      "learning_rate": 6.639600252346125e-06,
+      "loss": 0.5142,
+      "step": 5244
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7307889081335663,
+      "learning_rate": 6.6383987258391845e-06,
+      "loss": 0.4415,
+      "step": 5245
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5671936834847462,
+      "learning_rate": 6.637197093329989e-06,
+      "loss": 0.5231,
+      "step": 5246
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.922200118862228,
+      "learning_rate": 6.635995354896283e-06,
+      "loss": 0.4734,
+      "step": 5247
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7372505467485393,
+      "learning_rate": 6.634793510615818e-06,
+      "loss": 0.4884,
+      "step": 5248
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.587409568288592,
+      "learning_rate": 6.633591560566353e-06,
+      "loss": 0.5274,
+      "step": 5249
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.063742516692561,
+      "learning_rate": 6.632389504825648e-06,
+      "loss": 0.4967,
+      "step": 5250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5391491113973162,
+      "learning_rate": 6.63118734347148e-06,
+      "loss": 0.4887,
+      "step": 5251
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5779026210640987,
+      "learning_rate": 6.629985076581624e-06,
+      "loss": 0.5211,
+      "step": 5252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3347179266058666,
+      "learning_rate": 6.628782704233866e-06,
+      "loss": 0.463,
+      "step": 5253
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2768839522521973,
+      "learning_rate": 6.627580226505996e-06,
+      "loss": 0.4911,
+      "step": 5254
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5584399625989114,
+      "learning_rate": 6.6263776434758175e-06,
+      "loss": 0.4097,
+      "step": 5255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.411129221149787,
+      "learning_rate": 6.625174955221131e-06,
+      "loss": 0.4974,
+      "step": 5256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.4136155528964847,
+      "learning_rate": 6.6239721618197514e-06,
+      "loss": 0.5212,
+      "step": 5257
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4301267165640956,
+      "learning_rate": 6.622769263349496e-06,
+      "loss": 0.5012,
+      "step": 5258
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.6688177322483075,
+      "learning_rate": 6.6215662598881945e-06,
+      "loss": 0.4833,
+      "step": 5259
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.7555760424823563,
+      "learning_rate": 6.6203631515136755e-06,
+      "loss": 0.4939,
+      "step": 5260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8369715850367871,
+      "learning_rate": 6.619159938303782e-06,
+      "loss": 0.5148,
+      "step": 5261
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9098858334756308,
+      "learning_rate": 6.6179566203363565e-06,
+      "loss": 0.4726,
+      "step": 5262
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.888864449626197,
+      "learning_rate": 6.616753197689256e-06,
+      "loss": 0.4321,
+      "step": 5263
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4541339572193697,
+      "learning_rate": 6.615549670440337e-06,
+      "loss": 0.476,
+      "step": 5264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.411810043478361,
+      "learning_rate": 6.614346038667471e-06,
+      "loss": 0.4726,
+      "step": 5265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5780077938374408,
+      "learning_rate": 6.613142302448525e-06,
+      "loss": 0.4777,
+      "step": 5266
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5359915189862599,
+      "learning_rate": 6.611938461861383e-06,
+      "loss": 0.4676,
+      "step": 5267
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.3098969401085845,
+      "learning_rate": 6.610734516983931e-06,
+      "loss": 0.4704,
+      "step": 5268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9388946370652589,
+      "learning_rate": 6.609530467894064e-06,
+      "loss": 0.5171,
+      "step": 5269
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4387926611369306,
+      "learning_rate": 6.60832631466968e-06,
+      "loss": 0.4756,
+      "step": 5270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5937910201281864,
+      "learning_rate": 6.607122057388687e-06,
+      "loss": 0.4757,
+      "step": 5271
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5187717467205835,
+      "learning_rate": 6.605917696129001e-06,
+      "loss": 0.4349,
+      "step": 5272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.8199663979216534,
+      "learning_rate": 6.60471323096854e-06,
+      "loss": 0.451,
+      "step": 5273
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.2766743801891653,
+      "learning_rate": 6.603508661985232e-06,
+      "loss": 0.4624,
+      "step": 5274
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2386623156256038,
+      "learning_rate": 6.60230398925701e-06,
+      "loss": 0.4799,
+      "step": 5275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.0009454311597774,
+      "learning_rate": 6.601099212861818e-06,
+      "loss": 0.4858,
+      "step": 5276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.463096070489778,
+      "learning_rate": 6.5998943328776e-06,
+      "loss": 0.5172,
+      "step": 5277
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4287524305650479,
+      "learning_rate": 6.598689349382314e-06,
+      "loss": 0.4835,
+      "step": 5278
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6308057430548781,
+      "learning_rate": 6.597484262453916e-06,
+      "loss": 0.5191,
+      "step": 5279
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9207606236648558,
+      "learning_rate": 6.5962790721703775e-06,
+      "loss": 0.5359,
+      "step": 5280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 2.425728682358341,
+      "learning_rate": 6.59507377860967e-06,
+      "loss": 0.5206,
+      "step": 5281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.771678815162885,
+      "learning_rate": 6.593868381849775e-06,
+      "loss": 0.4898,
+      "step": 5282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.933414330745915,
+      "learning_rate": 6.592662881968681e-06,
+      "loss": 0.5507,
+      "step": 5283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.9457655182237008,
+      "learning_rate": 6.591457279044385e-06,
+      "loss": 0.4271,
+      "step": 5284
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1620689322387205,
+      "learning_rate": 6.5902515731548815e-06,
+      "loss": 0.4714,
+      "step": 5285
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6473865175117831,
+      "learning_rate": 6.589045764378184e-06,
+      "loss": 0.5167,
+      "step": 5286
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9778585632012742,
+      "learning_rate": 6.587839852792302e-06,
+      "loss": 0.5053,
+      "step": 5287
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5641387882618306,
+      "learning_rate": 6.586633838475261e-06,
+      "loss": 0.5169,
+      "step": 5288
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7548141676848115,
+      "learning_rate": 6.585427721505085e-06,
+      "loss": 0.4599,
+      "step": 5289
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2719856833380287,
+      "learning_rate": 6.584221501959809e-06,
+      "loss": 0.4932,
+      "step": 5290
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5727632796900416,
+      "learning_rate": 6.583015179917474e-06,
+      "loss": 0.5041,
+      "step": 5291
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.65048949493738,
+      "learning_rate": 6.581808755456128e-06,
+      "loss": 0.5053,
+      "step": 5292
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1164233999770112,
+      "learning_rate": 6.580602228653825e-06,
+      "loss": 0.4655,
+      "step": 5293
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5841200522136643,
+      "learning_rate": 6.579395599588626e-06,
+      "loss": 0.4484,
+      "step": 5294
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3568511714815754,
+      "learning_rate": 6.578188868338598e-06,
+      "loss": 0.5029,
+      "step": 5295
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.437090413765763,
+      "learning_rate": 6.576982034981813e-06,
+      "loss": 0.4519,
+      "step": 5296
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5233526880491428,
+      "learning_rate": 6.5757750995963554e-06,
+      "loss": 0.4801,
+      "step": 5297
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.875895917317297,
+      "learning_rate": 6.574568062260309e-06,
+      "loss": 0.5171,
+      "step": 5298
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5760694706835727,
+      "learning_rate": 6.573360923051769e-06,
+      "loss": 0.4862,
+      "step": 5299
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7687858629149067,
+      "learning_rate": 6.572153682048836e-06,
+      "loss": 0.4831,
+      "step": 5300
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1569142342910905,
+      "learning_rate": 6.570946339329616e-06,
+      "loss": 0.4591,
+      "step": 5301
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.482786752980439,
+      "learning_rate": 6.569738894972224e-06,
+      "loss": 0.405,
+      "step": 5302
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1344542815890306,
+      "learning_rate": 6.568531349054778e-06,
+      "loss": 0.4795,
+      "step": 5303
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6999725417085605,
+      "learning_rate": 6.567323701655404e-06,
+      "loss": 0.5025,
+      "step": 5304
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6614747082405472,
+      "learning_rate": 6.566115952852238e-06,
+      "loss": 0.4711,
+      "step": 5305
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3139225507327146,
+      "learning_rate": 6.56490810272342e-06,
+      "loss": 0.4442,
+      "step": 5306
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5445321222576824,
+      "learning_rate": 6.5637001513470935e-06,
+      "loss": 0.493,
+      "step": 5307
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8299974680926638,
+      "learning_rate": 6.562492098801413e-06,
+      "loss": 0.4764,
+      "step": 5308
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9680165944513737,
+      "learning_rate": 6.561283945164537e-06,
+      "loss": 0.4323,
+      "step": 5309
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6462097817510635,
+      "learning_rate": 6.560075690514633e-06,
+      "loss": 0.4752,
+      "step": 5310
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.1826481540266487,
+      "learning_rate": 6.558867334929872e-06,
+      "loss": 0.4478,
+      "step": 5311
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6852416771914756,
+      "learning_rate": 6.557658878488436e-06,
+      "loss": 0.4224,
+      "step": 5312
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6299310770541685,
+      "learning_rate": 6.556450321268506e-06,
+      "loss": 0.5314,
+      "step": 5313
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.036400608746328,
+      "learning_rate": 6.5552416633482775e-06,
+      "loss": 0.4992,
+      "step": 5314
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1143118095249744,
+      "learning_rate": 6.554032904805946e-06,
+      "loss": 0.4717,
+      "step": 5315
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8518101066286656,
+      "learning_rate": 6.55282404571972e-06,
+      "loss": 0.4588,
+      "step": 5316
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6313454848609727,
+      "learning_rate": 6.55161508616781e-06,
+      "loss": 0.5028,
+      "step": 5317
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.515708328733282,
+      "learning_rate": 6.550406026228432e-06,
+      "loss": 0.4457,
+      "step": 5318
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.263813345051239,
+      "learning_rate": 6.5491968659798145e-06,
+      "loss": 0.5072,
+      "step": 5319
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.876819127007386,
+      "learning_rate": 6.547987605500184e-06,
+      "loss": 0.4766,
+      "step": 5320
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7770001998958938,
+      "learning_rate": 6.546778244867782e-06,
+      "loss": 0.5351,
+      "step": 5321
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6175541167199458,
+      "learning_rate": 6.545568784160851e-06,
+      "loss": 0.5217,
+      "step": 5322
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2617203316927417,
+      "learning_rate": 6.544359223457641e-06,
+      "loss": 0.4491,
+      "step": 5323
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5917743874195696,
+      "learning_rate": 6.5431495628364095e-06,
+      "loss": 0.5413,
+      "step": 5324
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.837365120519526,
+      "learning_rate": 6.541939802375421e-06,
+      "loss": 0.5263,
+      "step": 5325
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5039961241545716,
+      "learning_rate": 6.5407299421529414e-06,
+      "loss": 0.4637,
+      "step": 5326
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7788505944592925,
+      "learning_rate": 6.539519982247254e-06,
+      "loss": 0.4729,
+      "step": 5327
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5125228470553895,
+      "learning_rate": 6.538309922736634e-06,
+      "loss": 0.5053,
+      "step": 5328
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5942237860402042,
+      "learning_rate": 6.537099763699377e-06,
+      "loss": 0.4279,
+      "step": 5329
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.178336226953759,
+      "learning_rate": 6.535889505213775e-06,
+      "loss": 0.4654,
+      "step": 5330
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9946561647837753,
+      "learning_rate": 6.5346791473581325e-06,
+      "loss": 0.4871,
+      "step": 5331
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7384181577954418,
+      "learning_rate": 6.533468690210756e-06,
+      "loss": 0.4669,
+      "step": 5332
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5288381330184222,
+      "learning_rate": 6.532258133849961e-06,
+      "loss": 0.4491,
+      "step": 5333
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2078506868842944,
+      "learning_rate": 6.5310474783540685e-06,
+      "loss": 0.4167,
+      "step": 5334
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8690360608538297,
+      "learning_rate": 6.529836723801408e-06,
+      "loss": 0.513,
+      "step": 5335
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9867303981920907,
+      "learning_rate": 6.528625870270313e-06,
+      "loss": 0.5519,
+      "step": 5336
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6912094305634301,
+      "learning_rate": 6.527414917839124e-06,
+      "loss": 0.5289,
+      "step": 5337
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.2654142858698214,
+      "learning_rate": 6.526203866586186e-06,
+      "loss": 0.4911,
+      "step": 5338
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.000422958227543,
+      "learning_rate": 6.524992716589856e-06,
+      "loss": 0.4646,
+      "step": 5339
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8090904216664347,
+      "learning_rate": 6.523781467928492e-06,
+      "loss": 0.4477,
+      "step": 5340
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7121601333964902,
+      "learning_rate": 6.522570120680461e-06,
+      "loss": 0.4399,
+      "step": 5341
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.246936662024646,
+      "learning_rate": 6.521358674924133e-06,
+      "loss": 0.5028,
+      "step": 5342
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.1021683176989185,
+      "learning_rate": 6.52014713073789e-06,
+      "loss": 0.4877,
+      "step": 5343
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5432548823354566,
+      "learning_rate": 6.518935488200118e-06,
+      "loss": 0.5281,
+      "step": 5344
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.529651949752004,
+      "learning_rate": 6.517723747389205e-06,
+      "loss": 0.511,
+      "step": 5345
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4498331979024621,
+      "learning_rate": 6.5165119083835515e-06,
+      "loss": 0.4922,
+      "step": 5346
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6075341352276735,
+      "learning_rate": 6.51529997126156e-06,
+      "loss": 0.4703,
+      "step": 5347
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6559053541873918,
+      "learning_rate": 6.514087936101645e-06,
+      "loss": 0.5036,
+      "step": 5348
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.070039876123479,
+      "learning_rate": 6.5128758029822194e-06,
+      "loss": 0.4681,
+      "step": 5349
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6360454928297283,
+      "learning_rate": 6.511663571981708e-06,
+      "loss": 0.499,
+      "step": 5350
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6297642183020937,
+      "learning_rate": 6.510451243178543e-06,
+      "loss": 0.5109,
+      "step": 5351
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.779766592004158,
+      "learning_rate": 6.509238816651158e-06,
+      "loss": 0.4191,
+      "step": 5352
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6726633523759864,
+      "learning_rate": 6.508026292477995e-06,
+      "loss": 0.5079,
+      "step": 5353
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.391964634391703,
+      "learning_rate": 6.506813670737504e-06,
+      "loss": 0.473,
+      "step": 5354
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6122790936392446,
+      "learning_rate": 6.505600951508138e-06,
+      "loss": 0.4965,
+      "step": 5355
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2977964913574578,
+      "learning_rate": 6.504388134868363e-06,
+      "loss": 0.4407,
+      "step": 5356
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6650193588750146,
+      "learning_rate": 6.503175220896642e-06,
+      "loss": 0.4293,
+      "step": 5357
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5132659586560648,
+      "learning_rate": 6.501962209671452e-06,
+      "loss": 0.4958,
+      "step": 5358
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6411323926370633,
+      "learning_rate": 6.500749101271271e-06,
+      "loss": 0.4787,
+      "step": 5359
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8139084435376154,
+      "learning_rate": 6.499535895774585e-06,
+      "loss": 0.481,
+      "step": 5360
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.7927433607015058,
+      "learning_rate": 6.498322593259889e-06,
+      "loss": 0.4686,
+      "step": 5361
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.594899966260172,
+      "learning_rate": 6.4971091938056814e-06,
+      "loss": 0.4523,
+      "step": 5362
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6789353114558476,
+      "learning_rate": 6.495895697490468e-06,
+      "loss": 0.5064,
+      "step": 5363
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8416821021905123,
+      "learning_rate": 6.49468210439276e-06,
+      "loss": 0.4841,
+      "step": 5364
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3455532845026175,
+      "learning_rate": 6.4934684145910746e-06,
+      "loss": 0.4255,
+      "step": 5365
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6586766409149718,
+      "learning_rate": 6.492254628163936e-06,
+      "loss": 0.473,
+      "step": 5366
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8194903298709044,
+      "learning_rate": 6.491040745189876e-06,
+      "loss": 0.5127,
+      "step": 5367
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.710576635559154,
+      "learning_rate": 6.489826765747431e-06,
+      "loss": 0.4663,
+      "step": 5368
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7655998802542583,
+      "learning_rate": 6.488612689915142e-06,
+      "loss": 0.4901,
+      "step": 5369
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9827478962400888,
+      "learning_rate": 6.487398517771559e-06,
+      "loss": 0.4103,
+      "step": 5370
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.564186866570126,
+      "learning_rate": 6.486184249395237e-06,
+      "loss": 0.4719,
+      "step": 5371
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.463652212551592,
+      "learning_rate": 6.48496988486474e-06,
+      "loss": 0.4733,
+      "step": 5372
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7569371301315814,
+      "learning_rate": 6.483755424258633e-06,
+      "loss": 0.5145,
+      "step": 5373
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6281449689526085,
+      "learning_rate": 6.482540867655492e-06,
+      "loss": 0.4906,
+      "step": 5374
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.180277332686683,
+      "learning_rate": 6.481326215133895e-06,
+      "loss": 0.4634,
+      "step": 5375
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.64258677886941,
+      "learning_rate": 6.480111466772432e-06,
+      "loss": 0.4356,
+      "step": 5376
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5870049005854123,
+      "learning_rate": 6.478896622649691e-06,
+      "loss": 0.4736,
+      "step": 5377
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3835406955797693,
+      "learning_rate": 6.477681682844274e-06,
+      "loss": 0.5535,
+      "step": 5378
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.3742857725644213,
+      "learning_rate": 6.476466647434785e-06,
+      "loss": 0.4915,
+      "step": 5379
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7701432064110176,
+      "learning_rate": 6.475251516499836e-06,
+      "loss": 0.4429,
+      "step": 5380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.482609484275228,
+      "learning_rate": 6.474036290118042e-06,
+      "loss": 0.4826,
+      "step": 5381
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.613001036561738,
+      "learning_rate": 6.47282096836803e-06,
+      "loss": 0.4615,
+      "step": 5382
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.482729642100069,
+      "learning_rate": 6.471605551328427e-06,
+      "loss": 0.4897,
+      "step": 5383
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8902905676002197,
+      "learning_rate": 6.470390039077871e-06,
+      "loss": 0.4966,
+      "step": 5384
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.5618013963087183,
+      "learning_rate": 6.469174431695002e-06,
+      "loss": 0.4828,
+      "step": 5385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3606008069397435,
+      "learning_rate": 6.46795872925847e-06,
+      "loss": 0.4494,
+      "step": 5386
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4140319006124789,
+      "learning_rate": 6.466742931846927e-06,
+      "loss": 0.4279,
+      "step": 5387
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.887975435962841,
+      "learning_rate": 6.465527039539038e-06,
+      "loss": 0.4491,
+      "step": 5388
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3690906698575989,
+      "learning_rate": 6.464311052413465e-06,
+      "loss": 0.5372,
+      "step": 5389
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.516485037980026,
+      "learning_rate": 6.463094970548881e-06,
+      "loss": 0.4696,
+      "step": 5390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 5.94572036360271,
+      "learning_rate": 6.461878794023968e-06,
+      "loss": 0.4867,
+      "step": 5391
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.570434607802137,
+      "learning_rate": 6.4606625229174096e-06,
+      "loss": 0.453,
+      "step": 5392
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9152710174430152,
+      "learning_rate": 6.459446157307896e-06,
+      "loss": 0.4733,
+      "step": 5393
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.607545940308983,
+      "learning_rate": 6.458229697274125e-06,
+      "loss": 0.4425,
+      "step": 5394
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6382126754714637,
+      "learning_rate": 6.457013142894801e-06,
+      "loss": 0.5102,
+      "step": 5395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.889116308700257,
+      "learning_rate": 6.4557964942486316e-06,
+      "loss": 0.4291,
+      "step": 5396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9242515833245408,
+      "learning_rate": 6.4545797514143346e-06,
+      "loss": 0.4874,
+      "step": 5397
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.759806061622904,
+      "learning_rate": 6.45336291447063e-06,
+      "loss": 0.5048,
+      "step": 5398
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9351823058640119,
+      "learning_rate": 6.452145983496247e-06,
+      "loss": 0.4016,
+      "step": 5399
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 7.693345991693742,
+      "learning_rate": 6.450928958569917e-06,
+      "loss": 0.4443,
+      "step": 5400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 14.334153433934874,
+      "learning_rate": 6.449711839770383e-06,
+      "loss": 0.4628,
+      "step": 5401
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9770433544452197,
+      "learning_rate": 6.448494627176388e-06,
+      "loss": 0.447,
+      "step": 5402
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.9372612270561234,
+      "learning_rate": 6.4472773208666875e-06,
+      "loss": 0.4052,
+      "step": 5403
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7769968364022821,
+      "learning_rate": 6.446059920920036e-06,
+      "loss": 0.441,
+      "step": 5404
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7900742652634223,
+      "learning_rate": 6.444842427415199e-06,
+      "loss": 0.4759,
+      "step": 5405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.627434109668974,
+      "learning_rate": 6.4436248404309466e-06,
+      "loss": 0.4658,
+      "step": 5406
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.7354460892039711,
+      "learning_rate": 6.442407160046057e-06,
+      "loss": 0.5239,
+      "step": 5407
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4531372123604291,
+      "learning_rate": 6.44118938633931e-06,
+      "loss": 0.5075,
+      "step": 5408
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.794227252127523,
+      "learning_rate": 6.439971519389496e-06,
+      "loss": 0.4784,
+      "step": 5409
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.0431258220341673,
+      "learning_rate": 6.438753559275408e-06,
+      "loss": 0.446,
+      "step": 5410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.6328198358822974,
+      "learning_rate": 6.437535506075847e-06,
+      "loss": 0.4754,
+      "step": 5411
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3296987772966862,
+      "learning_rate": 6.43631735986962e-06,
+      "loss": 0.4424,
+      "step": 5412
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6275245300364432,
+      "learning_rate": 6.435099120735537e-06,
+      "loss": 0.4711,
+      "step": 5413
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5364878368741093,
+      "learning_rate": 6.433880788752419e-06,
+      "loss": 0.4527,
+      "step": 5414
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5503886791229595,
+      "learning_rate": 6.43266236399909e-06,
+      "loss": 0.4588,
+      "step": 5415
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9296968599938478,
+      "learning_rate": 6.43144384655438e-06,
+      "loss": 0.5451,
+      "step": 5416
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5525208922518552,
+      "learning_rate": 6.430225236497125e-06,
+      "loss": 0.4384,
+      "step": 5417
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5080603647950124,
+      "learning_rate": 6.429006533906168e-06,
+      "loss": 0.4749,
+      "step": 5418
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9036898854974603,
+      "learning_rate": 6.42778773886036e-06,
+      "loss": 0.5042,
+      "step": 5419
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8785341207331518,
+      "learning_rate": 6.426568851438551e-06,
+      "loss": 0.4628,
+      "step": 5420
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 8.365271394342399,
+      "learning_rate": 6.425349871719604e-06,
+      "loss": 0.4257,
+      "step": 5421
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.195017687174971,
+      "learning_rate": 6.424130799782385e-06,
+      "loss": 0.4394,
+      "step": 5422
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7302862450406908,
+      "learning_rate": 6.422911635705766e-06,
+      "loss": 0.4927,
+      "step": 5423
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.498544750431292,
+      "learning_rate": 6.421692379568626e-06,
+      "loss": 0.4912,
+      "step": 5424
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4687744039618424,
+      "learning_rate": 6.420473031449849e-06,
+      "loss": 0.4344,
+      "step": 5425
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3812623883392923,
+      "learning_rate": 6.419253591428325e-06,
+      "loss": 0.4757,
+      "step": 5426
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5345256421823,
+      "learning_rate": 6.4180340595829495e-06,
+      "loss": 0.4643,
+      "step": 5427
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6472825213474096,
+      "learning_rate": 6.416814435992625e-06,
+      "loss": 0.4671,
+      "step": 5428
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.598756856317711,
+      "learning_rate": 6.415594720736261e-06,
+      "loss": 0.4686,
+      "step": 5429
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5361462315551706,
+      "learning_rate": 6.41437491389277e-06,
+      "loss": 0.4527,
+      "step": 5430
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.151113449173148,
+      "learning_rate": 6.413155015541073e-06,
+      "loss": 0.478,
+      "step": 5431
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4327168674712347,
+      "learning_rate": 6.411935025760094e-06,
+      "loss": 0.4785,
+      "step": 5432
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5969422191599394,
+      "learning_rate": 6.410714944628766e-06,
+      "loss": 0.4838,
+      "step": 5433
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.909595615074427,
+      "learning_rate": 6.4094947722260285e-06,
+      "loss": 0.4704,
+      "step": 5434
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4347264352479394,
+      "learning_rate": 6.4082745086308206e-06,
+      "loss": 0.4811,
+      "step": 5435
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0839511447097134,
+      "learning_rate": 6.407054153922096e-06,
+      "loss": 0.49,
+      "step": 5436
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.721522911306308,
+      "learning_rate": 6.4058337081788066e-06,
+      "loss": 0.5112,
+      "step": 5437
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6242778700245335,
+      "learning_rate": 6.404613171479918e-06,
+      "loss": 0.5322,
+      "step": 5438
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9257484783815182,
+      "learning_rate": 6.403392543904391e-06,
+      "loss": 0.5217,
+      "step": 5439
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7644594287895383,
+      "learning_rate": 6.402171825531205e-06,
+      "loss": 0.4783,
+      "step": 5440
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5909360990160306,
+      "learning_rate": 6.400951016439334e-06,
+      "loss": 0.505,
+      "step": 5441
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6006419925270396,
+      "learning_rate": 6.3997301167077675e-06,
+      "loss": 0.4747,
+      "step": 5442
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.991536195781363,
+      "learning_rate": 6.398509126415492e-06,
+      "loss": 0.4946,
+      "step": 5443
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.631721951424862,
+      "learning_rate": 6.397288045641507e-06,
+      "loss": 0.4884,
+      "step": 5444
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.2515637751974937,
+      "learning_rate": 6.396066874464811e-06,
+      "loss": 0.4945,
+      "step": 5445
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3422218229853164,
+      "learning_rate": 6.3948456129644165e-06,
+      "loss": 0.441,
+      "step": 5446
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.668999912114585,
+      "learning_rate": 6.393624261219335e-06,
+      "loss": 0.4693,
+      "step": 5447
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6932412682056792,
+      "learning_rate": 6.392402819308586e-06,
+      "loss": 0.4271,
+      "step": 5448
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5900655826438592,
+      "learning_rate": 6.391181287311197e-06,
+      "loss": 0.4563,
+      "step": 5449
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6242586916004049,
+      "learning_rate": 6.389959665306198e-06,
+      "loss": 0.5138,
+      "step": 5450
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5915857087937059,
+      "learning_rate": 6.388737953372625e-06,
+      "loss": 0.4899,
+      "step": 5451
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.506342712362207,
+      "learning_rate": 6.3875161515895265e-06,
+      "loss": 0.5074,
+      "step": 5452
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5708623882760382,
+      "learning_rate": 6.3862942600359465e-06,
+      "loss": 0.5133,
+      "step": 5453
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6527977500417101,
+      "learning_rate": 6.385072278790942e-06,
+      "loss": 0.4535,
+      "step": 5454
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1491974926709103,
+      "learning_rate": 6.383850207933573e-06,
+      "loss": 0.4822,
+      "step": 5455
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.792414462685571,
+      "learning_rate": 6.382628047542907e-06,
+      "loss": 0.4792,
+      "step": 5456
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4688810450103786,
+      "learning_rate": 6.381405797698013e-06,
+      "loss": 0.4288,
+      "step": 5457
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8602533658437548,
+      "learning_rate": 6.380183458477972e-06,
+      "loss": 0.4962,
+      "step": 5458
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7216556573498454,
+      "learning_rate": 6.3789610299618654e-06,
+      "loss": 0.4636,
+      "step": 5459
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.26115220474863,
+      "learning_rate": 6.377738512228785e-06,
+      "loss": 0.4467,
+      "step": 5460
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6294761832062299,
+      "learning_rate": 6.376515905357825e-06,
+      "loss": 0.5066,
+      "step": 5461
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5664688453998006,
+      "learning_rate": 6.375293209428087e-06,
+      "loss": 0.4295,
+      "step": 5462
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7730610928029351,
+      "learning_rate": 6.374070424518677e-06,
+      "loss": 0.4521,
+      "step": 5463
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6381401658153492,
+      "learning_rate": 6.372847550708709e-06,
+      "loss": 0.523,
+      "step": 5464
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.169142746871996,
+      "learning_rate": 6.3716245880772985e-06,
+      "loss": 0.4387,
+      "step": 5465
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8986764257053186,
+      "learning_rate": 6.370401536703573e-06,
+      "loss": 0.4232,
+      "step": 5466
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.3144938015891303,
+      "learning_rate": 6.369178396666661e-06,
+      "loss": 0.4498,
+      "step": 5467
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9488229418167975,
+      "learning_rate": 6.3679551680456964e-06,
+      "loss": 0.4498,
+      "step": 5468
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5634560214295914,
+      "learning_rate": 6.366731850919824e-06,
+      "loss": 0.465,
+      "step": 5469
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5923967710324288,
+      "learning_rate": 6.365508445368187e-06,
+      "loss": 0.4976,
+      "step": 5470
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6026706114852998,
+      "learning_rate": 6.364284951469941e-06,
+      "loss": 0.4391,
+      "step": 5471
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.7901685820410713,
+      "learning_rate": 6.363061369304243e-06,
+      "loss": 0.4333,
+      "step": 5472
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8694935282780765,
+      "learning_rate": 6.3618376989502574e-06,
+      "loss": 0.4636,
+      "step": 5473
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3724218773036148,
+      "learning_rate": 6.360613940487152e-06,
+      "loss": 0.4866,
+      "step": 5474
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.8005088425579485,
+      "learning_rate": 6.359390093994109e-06,
+      "loss": 0.4944,
+      "step": 5475
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6907881020604005,
+      "learning_rate": 6.358166159550302e-06,
+      "loss": 0.4613,
+      "step": 5476
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1296381345708353,
+      "learning_rate": 6.356942137234923e-06,
+      "loss": 0.4229,
+      "step": 5477
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8341426144465043,
+      "learning_rate": 6.355718027127161e-06,
+      "loss": 0.4658,
+      "step": 5478
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6540876514616087,
+      "learning_rate": 6.354493829306217e-06,
+      "loss": 0.4672,
+      "step": 5479
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5912348036339827,
+      "learning_rate": 6.353269543851295e-06,
+      "loss": 0.46,
+      "step": 5480
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.8520615174937565,
+      "learning_rate": 6.352045170841603e-06,
+      "loss": 0.4572,
+      "step": 5481
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.5714615522815722,
+      "learning_rate": 6.350820710356357e-06,
+      "loss": 0.4833,
+      "step": 5482
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9609453444653846,
+      "learning_rate": 6.349596162474779e-06,
+      "loss": 0.5097,
+      "step": 5483
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.131953226081147,
+      "learning_rate": 6.348371527276093e-06,
+      "loss": 0.4682,
+      "step": 5484
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5932576661208442,
+      "learning_rate": 6.347146804839535e-06,
+      "loss": 0.4331,
+      "step": 5485
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6130199231462509,
+      "learning_rate": 6.345921995244339e-06,
+      "loss": 0.5061,
+      "step": 5486
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5737971277902503,
+      "learning_rate": 6.344697098569752e-06,
+      "loss": 0.4519,
+      "step": 5487
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7708496825277467,
+      "learning_rate": 6.343472114895022e-06,
+      "loss": 0.4887,
+      "step": 5488
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5435230243512081,
+      "learning_rate": 6.342247044299403e-06,
+      "loss": 0.4187,
+      "step": 5489
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3606083222923795,
+      "learning_rate": 6.341021886862157e-06,
+      "loss": 0.4563,
+      "step": 5490
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.9412986379604802,
+      "learning_rate": 6.339796642662549e-06,
+      "loss": 0.457,
+      "step": 5491
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.4632746095147215,
+      "learning_rate": 6.33857131177985e-06,
+      "loss": 0.5036,
+      "step": 5492
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0265371688800413,
+      "learning_rate": 6.337345894293339e-06,
+      "loss": 0.5231,
+      "step": 5493
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.52936986379853,
+      "learning_rate": 6.3361203902822964e-06,
+      "loss": 0.4755,
+      "step": 5494
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5874212538591792,
+      "learning_rate": 6.334894799826014e-06,
+      "loss": 0.471,
+      "step": 5495
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.86485169696741,
+      "learning_rate": 6.333669123003784e-06,
+      "loss": 0.5112,
+      "step": 5496
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4470090274878336,
+      "learning_rate": 6.332443359894906e-06,
+      "loss": 0.5146,
+      "step": 5497
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.567655033098356,
+      "learning_rate": 6.331217510578687e-06,
+      "loss": 0.478,
+      "step": 5498
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4912396331544506,
+      "learning_rate": 6.329991575134435e-06,
+      "loss": 0.4509,
+      "step": 5499
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.1274600665978647,
+      "learning_rate": 6.328765553641469e-06,
+      "loss": 0.5221,
+      "step": 5500
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.454579301798375,
+      "learning_rate": 6.3275394461791095e-06,
+      "loss": 0.4861,
+      "step": 5501
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.597734300081622,
+      "learning_rate": 6.326313252826685e-06,
+      "loss": 0.5107,
+      "step": 5502
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5312056407054541,
+      "learning_rate": 6.3250869736635265e-06,
+      "loss": 0.5098,
+      "step": 5503
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7702597449802944,
+      "learning_rate": 6.323860608768977e-06,
+      "loss": 0.4367,
+      "step": 5504
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.752765665329628,
+      "learning_rate": 6.322634158222375e-06,
+      "loss": 0.4882,
+      "step": 5505
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.693037343820609,
+      "learning_rate": 6.3214076221030755e-06,
+      "loss": 0.478,
+      "step": 5506
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.599966286276832,
+      "learning_rate": 6.320181000490429e-06,
+      "loss": 0.5129,
+      "step": 5507
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.090312987214145,
+      "learning_rate": 6.318954293463801e-06,
+      "loss": 0.4553,
+      "step": 5508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7979868610930083,
+      "learning_rate": 6.317727501102554e-06,
+      "loss": 0.44,
+      "step": 5509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.569743529972744,
+      "learning_rate": 6.316500623486063e-06,
+      "loss": 0.4455,
+      "step": 5510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.091556601244058,
+      "learning_rate": 6.315273660693702e-06,
+      "loss": 0.4866,
+      "step": 5511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5859257821189616,
+      "learning_rate": 6.3140466128048585e-06,
+      "loss": 0.4985,
+      "step": 5512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5198534750191977,
+      "learning_rate": 6.312819479898915e-06,
+      "loss": 0.4606,
+      "step": 5513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5835576877026925,
+      "learning_rate": 6.311592262055271e-06,
+      "loss": 0.5216,
+      "step": 5514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.411832292277702,
+      "learning_rate": 6.310364959353322e-06,
+      "loss": 0.5213,
+      "step": 5515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7916729425640863,
+      "learning_rate": 6.309137571872476e-06,
+      "loss": 0.4372,
+      "step": 5516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4797800244039416,
+      "learning_rate": 6.307910099692141e-06,
+      "loss": 0.4499,
+      "step": 5517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6047367513266667,
+      "learning_rate": 6.306682542891731e-06,
+      "loss": 0.4963,
+      "step": 5518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6202889265059824,
+      "learning_rate": 6.305454901550672e-06,
+      "loss": 0.4943,
+      "step": 5519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.730061000161206,
+      "learning_rate": 6.304227175748389e-06,
+      "loss": 0.4932,
+      "step": 5520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.601056835929249,
+      "learning_rate": 6.302999365564313e-06,
+      "loss": 0.4446,
+      "step": 5521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 3.32998846036508,
+      "learning_rate": 6.301771471077883e-06,
+      "loss": 0.485,
+      "step": 5522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4177640392760884,
+      "learning_rate": 6.300543492368541e-06,
+      "loss": 0.4429,
+      "step": 5523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 6.0823578752631535,
+      "learning_rate": 6.299315429515738e-06,
+      "loss": 0.4816,
+      "step": 5524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7002113252905469,
+      "learning_rate": 6.298087282598924e-06,
+      "loss": 0.5585,
+      "step": 5525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6217623061662915,
+      "learning_rate": 6.296859051697563e-06,
+      "loss": 0.5216,
+      "step": 5526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.0148967264842996,
+      "learning_rate": 6.2956307368911185e-06,
+      "loss": 0.436,
+      "step": 5527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2032137516498582,
+      "learning_rate": 6.294402338259059e-06,
+      "loss": 0.431,
+      "step": 5528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5681886673243572,
+      "learning_rate": 6.293173855880861e-06,
+      "loss": 0.4569,
+      "step": 5529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.69073980005492,
+      "learning_rate": 6.2919452898360056e-06,
+      "loss": 0.4711,
+      "step": 5530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5698874714475253,
+      "learning_rate": 6.290716640203981e-06,
+      "loss": 0.5097,
+      "step": 5531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.464968671036617,
+      "learning_rate": 6.289487907064279e-06,
+      "loss": 0.427,
+      "step": 5532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.133831155646331,
+      "learning_rate": 6.288259090496396e-06,
+      "loss": 0.5053,
+      "step": 5533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5687086950746946,
+      "learning_rate": 6.2870301905798344e-06,
+      "loss": 0.5137,
+      "step": 5534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4600551344170163,
+      "learning_rate": 6.285801207394103e-06,
+      "loss": 0.4742,
+      "step": 5535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.5395037690645166,
+      "learning_rate": 6.284572141018716e-06,
+      "loss": 0.4886,
+      "step": 5536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.7626111187571571,
+      "learning_rate": 6.283342991533192e-06,
+      "loss": 0.4895,
+      "step": 5537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6659088375724659,
+      "learning_rate": 6.282113759017054e-06,
+      "loss": 0.5237,
+      "step": 5538
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5314847146966124,
+      "learning_rate": 6.280884443549836e-06,
+      "loss": 0.5128,
+      "step": 5539
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8011167226425937,
+      "learning_rate": 6.279655045211068e-06,
+      "loss": 0.4763,
+      "step": 5540
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.087045350487504,
+      "learning_rate": 6.278425564080292e-06,
+      "loss": 0.4836,
+      "step": 5541
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5449633482197765,
+      "learning_rate": 6.277196000237055e-06,
+      "loss": 0.5048,
+      "step": 5542
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3830724000008425,
+      "learning_rate": 6.275966353760908e-06,
+      "loss": 0.4643,
+      "step": 5543
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7798968396591999,
+      "learning_rate": 6.274736624731407e-06,
+      "loss": 0.4834,
+      "step": 5544
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5574783065853035,
+      "learning_rate": 6.273506813228114e-06,
+      "loss": 0.5101,
+      "step": 5545
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.493871500437576,
+      "learning_rate": 6.272276919330595e-06,
+      "loss": 0.4411,
+      "step": 5546
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5955228546102452,
+      "learning_rate": 6.2710469431184265e-06,
+      "loss": 0.4989,
+      "step": 5547
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5654304024634748,
+      "learning_rate": 6.269816884671181e-06,
+      "loss": 0.4941,
+      "step": 5548
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8002561004552389,
+      "learning_rate": 6.268586744068446e-06,
+      "loss": 0.4729,
+      "step": 5549
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3558947544683584,
+      "learning_rate": 6.267356521389806e-06,
+      "loss": 0.422,
+      "step": 5550
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4591873467014147,
+      "learning_rate": 6.266126216714859e-06,
+      "loss": 0.4693,
+      "step": 5551
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5598105293728107,
+      "learning_rate": 6.264895830123199e-06,
+      "loss": 0.5102,
+      "step": 5552
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3436474877353506,
+      "learning_rate": 6.263665361694436e-06,
+      "loss": 0.4909,
+      "step": 5553
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5961351167019608,
+      "learning_rate": 6.2624348115081754e-06,
+      "loss": 0.4741,
+      "step": 5554
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3383688868718662,
+      "learning_rate": 6.261204179644037e-06,
+      "loss": 0.4803,
+      "step": 5555
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8971574240255207,
+      "learning_rate": 6.2599734661816355e-06,
+      "loss": 0.4551,
+      "step": 5556
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4255874149756704,
+      "learning_rate": 6.2587426712006005e-06,
+      "loss": 0.5145,
+      "step": 5557
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0256001218190436,
+      "learning_rate": 6.2575117947805595e-06,
+      "loss": 0.4774,
+      "step": 5558
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6595851502776786,
+      "learning_rate": 6.256280837001153e-06,
+      "loss": 0.465,
+      "step": 5559
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5069803152235348,
+      "learning_rate": 6.255049797942018e-06,
+      "loss": 0.4509,
+      "step": 5560
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9607933876433523,
+      "learning_rate": 6.2538186776828035e-06,
+      "loss": 0.4384,
+      "step": 5561
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.990783689800869,
+      "learning_rate": 6.25258747630316e-06,
+      "loss": 0.5282,
+      "step": 5562
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.435659153315895,
+      "learning_rate": 6.251356193882747e-06,
+      "loss": 0.4459,
+      "step": 5563
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.0793880098222886,
+      "learning_rate": 6.250124830501222e-06,
+      "loss": 0.482,
+      "step": 5564
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6722413374355083,
+      "learning_rate": 6.2488933862382585e-06,
+      "loss": 0.4549,
+      "step": 5565
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6086333705169709,
+      "learning_rate": 6.2476618611735265e-06,
+      "loss": 0.4663,
+      "step": 5566
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5643079363398351,
+      "learning_rate": 6.246430255386704e-06,
+      "loss": 0.5067,
+      "step": 5567
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1312258768988266,
+      "learning_rate": 6.245198568957473e-06,
+      "loss": 0.4543,
+      "step": 5568
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6054135273620439,
+      "learning_rate": 6.243966801965526e-06,
+      "loss": 0.4863,
+      "step": 5569
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5724986623511228,
+      "learning_rate": 6.242734954490552e-06,
+      "loss": 0.5181,
+      "step": 5570
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6538309030428566,
+      "learning_rate": 6.241503026612254e-06,
+      "loss": 0.4581,
+      "step": 5571
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6742973488910453,
+      "learning_rate": 6.240271018410333e-06,
+      "loss": 0.4894,
+      "step": 5572
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.1486432112241935,
+      "learning_rate": 6.2390389299645e-06,
+      "loss": 0.4968,
+      "step": 5573
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.6330177418561878,
+      "learning_rate": 6.23780676135447e-06,
+      "loss": 0.4936,
+      "step": 5574
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.598287161677895,
+      "learning_rate": 6.236574512659961e-06,
+      "loss": 0.5251,
+      "step": 5575
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.024696637329994,
+      "learning_rate": 6.235342183960698e-06,
+      "loss": 0.4368,
+      "step": 5576
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.612976835516847,
+      "learning_rate": 6.234109775336412e-06,
+      "loss": 0.5062,
+      "step": 5577
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7719340937396224,
+      "learning_rate": 6.232877286866838e-06,
+      "loss": 0.4915,
+      "step": 5578
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4505230269332725,
+      "learning_rate": 6.231644718631717e-06,
+      "loss": 0.4767,
+      "step": 5579
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.59935888299376,
+      "learning_rate": 6.230412070710794e-06,
+      "loss": 0.4741,
+      "step": 5580
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0965352194741755,
+      "learning_rate": 6.2291793431838186e-06,
+      "loss": 0.4974,
+      "step": 5581
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9694881442051635,
+      "learning_rate": 6.227946536130549e-06,
+      "loss": 0.5031,
+      "step": 5582
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.4075911070721436,
+      "learning_rate": 6.226713649630744e-06,
+      "loss": 0.4671,
+      "step": 5583
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5846784067395779,
+      "learning_rate": 6.225480683764171e-06,
+      "loss": 0.5013,
+      "step": 5584
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.427659418647081,
+      "learning_rate": 6.2242476386106e-06,
+      "loss": 0.477,
+      "step": 5585
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4426267208821388,
+      "learning_rate": 6.223014514249809e-06,
+      "loss": 0.4643,
+      "step": 5586
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.042579700255808,
+      "learning_rate": 6.221781310761578e-06,
+      "loss": 0.5068,
+      "step": 5587
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5624296818871508,
+      "learning_rate": 6.2205480282256955e-06,
+      "loss": 0.453,
+      "step": 5588
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7295122611340383,
+      "learning_rate": 6.219314666721952e-06,
+      "loss": 0.5127,
+      "step": 5589
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7765466188951584,
+      "learning_rate": 6.218081226330145e-06,
+      "loss": 0.4667,
+      "step": 5590
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.627859906736519,
+      "learning_rate": 6.2168477071300745e-06,
+      "loss": 0.4925,
+      "step": 5591
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.824884442672306,
+      "learning_rate": 6.215614109201551e-06,
+      "loss": 0.4557,
+      "step": 5592
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4211754251162967,
+      "learning_rate": 6.214380432624384e-06,
+      "loss": 0.4919,
+      "step": 5593
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5818456299420639,
+      "learning_rate": 6.213146677478392e-06,
+      "loss": 0.5255,
+      "step": 5594
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 7.900621366762852,
+      "learning_rate": 6.211912843843397e-06,
+      "loss": 0.458,
+      "step": 5595
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.1801447797140567,
+      "learning_rate": 6.210678931799227e-06,
+      "loss": 0.4276,
+      "step": 5596
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.841733810437032,
+      "learning_rate": 6.209444941425712e-06,
+      "loss": 0.5038,
+      "step": 5597
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.450340958504094,
+      "learning_rate": 6.208210872802692e-06,
+      "loss": 0.5164,
+      "step": 5598
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5527146309995212,
+      "learning_rate": 6.20697672601001e-06,
+      "loss": 0.4562,
+      "step": 5599
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5244433387615075,
+      "learning_rate": 6.205742501127513e-06,
+      "loss": 0.5095,
+      "step": 5600
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.554095705199923,
+      "learning_rate": 6.204508198235052e-06,
+      "loss": 0.4919,
+      "step": 5601
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9914185380468825,
+      "learning_rate": 6.203273817412487e-06,
+      "loss": 0.4536,
+      "step": 5602
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.075358483836895,
+      "learning_rate": 6.202039358739681e-06,
+      "loss": 0.483,
+      "step": 5603
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5518661671618242,
+      "learning_rate": 6.2008048222965e-06,
+      "loss": 0.5027,
+      "step": 5604
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7050671177585839,
+      "learning_rate": 6.199570208162819e-06,
+      "loss": 0.4524,
+      "step": 5605
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.768982091393003,
+      "learning_rate": 6.198335516418513e-06,
+      "loss": 0.4587,
+      "step": 5606
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.605151862038958,
+      "learning_rate": 6.197100747143468e-06,
+      "loss": 0.4925,
+      "step": 5607
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0012102797441025,
+      "learning_rate": 6.195865900417569e-06,
+      "loss": 0.4718,
+      "step": 5608
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.818867818488926,
+      "learning_rate": 6.194630976320711e-06,
+      "loss": 0.4518,
+      "step": 5609
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.557986108279998,
+      "learning_rate": 6.193395974932793e-06,
+      "loss": 0.4858,
+      "step": 5610
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0884827335975755,
+      "learning_rate": 6.192160896333717e-06,
+      "loss": 0.5098,
+      "step": 5611
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7627101388728135,
+      "learning_rate": 6.190925740603388e-06,
+      "loss": 0.4615,
+      "step": 5612
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.803152075752051,
+      "learning_rate": 6.189690507821724e-06,
+      "loss": 0.5077,
+      "step": 5613
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4778921436164885,
+      "learning_rate": 6.188455198068639e-06,
+      "loss": 0.4873,
+      "step": 5614
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.802185555367496,
+      "learning_rate": 6.1872198114240575e-06,
+      "loss": 0.4451,
+      "step": 5615
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5697676469216717,
+      "learning_rate": 6.185984347967909e-06,
+      "loss": 0.4551,
+      "step": 5616
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.58149791104693,
+      "learning_rate": 6.184748807780123e-06,
+      "loss": 0.5022,
+      "step": 5617
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4271086567635025,
+      "learning_rate": 6.1835131909406385e-06,
+      "loss": 0.4806,
+      "step": 5618
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7086561351103302,
+      "learning_rate": 6.1822774975294e-06,
+      "loss": 0.4515,
+      "step": 5619
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5843319246388143,
+      "learning_rate": 6.181041727626352e-06,
+      "loss": 0.512,
+      "step": 5620
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9020399473785523,
+      "learning_rate": 6.179805881311452e-06,
+      "loss": 0.4835,
+      "step": 5621
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9639058450004239,
+      "learning_rate": 6.178569958664653e-06,
+      "loss": 0.4693,
+      "step": 5622
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8518338237815346,
+      "learning_rate": 6.177333959765921e-06,
+      "loss": 0.498,
+      "step": 5623
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8055298980715515,
+      "learning_rate": 6.1760978846952205e-06,
+      "loss": 0.4584,
+      "step": 5624
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0207312189718394,
+      "learning_rate": 6.174861733532527e-06,
+      "loss": 0.4628,
+      "step": 5625
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6159370322247352,
+      "learning_rate": 6.173625506357814e-06,
+      "loss": 0.5087,
+      "step": 5626
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.301521068387596,
+      "learning_rate": 6.172389203251068e-06,
+      "loss": 0.5003,
+      "step": 5627
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5679047013106664,
+      "learning_rate": 6.171152824292272e-06,
+      "loss": 0.5059,
+      "step": 5628
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.099372170782018,
+      "learning_rate": 6.169916369561423e-06,
+      "loss": 0.491,
+      "step": 5629
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.857518656450942,
+      "learning_rate": 6.168679839138514e-06,
+      "loss": 0.4708,
+      "step": 5630
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7521114629022725,
+      "learning_rate": 6.167443233103547e-06,
+      "loss": 0.4697,
+      "step": 5631
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.4251136298316553,
+      "learning_rate": 6.166206551536531e-06,
+      "loss": 0.5249,
+      "step": 5632
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6119084010558695,
+      "learning_rate": 6.1649697945174784e-06,
+      "loss": 0.5244,
+      "step": 5633
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.380729841004696,
+      "learning_rate": 6.163732962126402e-06,
+      "loss": 0.4228,
+      "step": 5634
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2870631846189693,
+      "learning_rate": 6.162496054443326e-06,
+      "loss": 0.4822,
+      "step": 5635
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8086724579027718,
+      "learning_rate": 6.161259071548277e-06,
+      "loss": 0.4543,
+      "step": 5636
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5702067965436302,
+      "learning_rate": 6.160022013521284e-06,
+      "loss": 0.5031,
+      "step": 5637
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.208237457511597,
+      "learning_rate": 6.1587848804423846e-06,
+      "loss": 0.4439,
+      "step": 5638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7630773670009363,
+      "learning_rate": 6.15754767239162e-06,
+      "loss": 0.5108,
+      "step": 5639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5461083414637489,
+      "learning_rate": 6.156310389449034e-06,
+      "loss": 0.5062,
+      "step": 5640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0870959559547284,
+      "learning_rate": 6.155073031694679e-06,
+      "loss": 0.4772,
+      "step": 5641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9656273991534357,
+      "learning_rate": 6.153835599208609e-06,
+      "loss": 0.4756,
+      "step": 5642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.548023070604085,
+      "learning_rate": 6.152598092070885e-06,
+      "loss": 0.495,
+      "step": 5643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4233484897278479,
+      "learning_rate": 6.151360510361574e-06,
+      "loss": 0.4642,
+      "step": 5644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.835031001189982,
+      "learning_rate": 6.150122854160743e-06,
+      "loss": 0.4794,
+      "step": 5645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.832944915775355,
+      "learning_rate": 6.148885123548468e-06,
+      "loss": 0.5366,
+      "step": 5646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.6412392321924147,
+      "learning_rate": 6.147647318604829e-06,
+      "loss": 0.4433,
+      "step": 5647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.9409156217546824,
+      "learning_rate": 6.146409439409909e-06,
+      "loss": 0.4454,
+      "step": 5648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5528330949253917,
+      "learning_rate": 6.1451714860437985e-06,
+      "loss": 0.4994,
+      "step": 5649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.649855174091046,
+      "learning_rate": 6.14393345858659e-06,
+      "loss": 0.5155,
+      "step": 5650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2505425072365757,
+      "learning_rate": 6.142695357118384e-06,
+      "loss": 0.5209,
+      "step": 5651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9213216928115613,
+      "learning_rate": 6.141457181719283e-06,
+      "loss": 0.4945,
+      "step": 5652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8144237281481375,
+      "learning_rate": 6.140218932469396e-06,
+      "loss": 0.4912,
+      "step": 5653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.848752946919791,
+      "learning_rate": 6.138980609448835e-06,
+      "loss": 0.4919,
+      "step": 5654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.7193268099339667,
+      "learning_rate": 6.137742212737718e-06,
+      "loss": 0.5184,
+      "step": 5655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.9129333006414508,
+      "learning_rate": 6.136503742416171e-06,
+      "loss": 0.4543,
+      "step": 5656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5631829422991903,
+      "learning_rate": 6.135265198564317e-06,
+      "loss": 0.5286,
+      "step": 5657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2859951386046693,
+      "learning_rate": 6.134026581262289e-06,
+      "loss": 0.4155,
+      "step": 5658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.728400373449617,
+      "learning_rate": 6.132787890590225e-06,
+      "loss": 0.4826,
+      "step": 5659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.0893314374311913,
+      "learning_rate": 6.1315491266282666e-06,
+      "loss": 0.4444,
+      "step": 5660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.3827723442042017,
+      "learning_rate": 6.13031028945656e-06,
+      "loss": 0.5225,
+      "step": 5661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 2.2960147483766415,
+      "learning_rate": 6.129071379155257e-06,
+      "loss": 0.4617,
+      "step": 5662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5878010023473703,
+      "learning_rate": 6.127832395804512e-06,
+      "loss": 0.5166,
+      "step": 5663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 5.678098039696629,
+      "learning_rate": 6.1265933394844855e-06,
+      "loss": 0.4694,
+      "step": 5664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7011261329044907,
+      "learning_rate": 6.125354210275344e-06,
+      "loss": 0.458,
+      "step": 5665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.7514111657476035,
+      "learning_rate": 6.124115008257259e-06,
+      "loss": 0.4932,
+      "step": 5666
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7555776623936834,
+      "learning_rate": 6.1228757335104e-06,
+      "loss": 0.4582,
+      "step": 5667
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.310613942912549,
+      "learning_rate": 6.121636386114954e-06,
+      "loss": 0.4943,
+      "step": 5668
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9377045336420782,
+      "learning_rate": 6.120396966151099e-06,
+      "loss": 0.4797,
+      "step": 5669
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.637232097811697,
+      "learning_rate": 6.119157473699027e-06,
+      "loss": 0.4896,
+      "step": 5670
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.801566334489921,
+      "learning_rate": 6.11791790883893e-06,
+      "loss": 0.5003,
+      "step": 5671
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5908583374350184,
+      "learning_rate": 6.1166782716510065e-06,
+      "loss": 0.5086,
+      "step": 5672
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.468821909172175,
+      "learning_rate": 6.115438562215459e-06,
+      "loss": 0.4578,
+      "step": 5673
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6135798735612559,
+      "learning_rate": 6.114198780612497e-06,
+      "loss": 0.4969,
+      "step": 5674
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2805661945768776,
+      "learning_rate": 6.11295892692233e-06,
+      "loss": 0.4371,
+      "step": 5675
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7102417132814736,
+      "learning_rate": 6.111719001225178e-06,
+      "loss": 0.4804,
+      "step": 5676
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5804944108793355,
+      "learning_rate": 6.110479003601258e-06,
+      "loss": 0.5018,
+      "step": 5677
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5813935176956744,
+      "learning_rate": 6.109238934130802e-06,
+      "loss": 0.5059,
+      "step": 5678
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5740389190923919,
+      "learning_rate": 6.107998792894038e-06,
+      "loss": 0.5099,
+      "step": 5679
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7385142746491293,
+      "learning_rate": 6.106758579971201e-06,
+      "loss": 0.4741,
+      "step": 5680
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6486368805632574,
+      "learning_rate": 6.105518295442531e-06,
+      "loss": 0.4442,
+      "step": 5681
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5623861420798012,
+      "learning_rate": 6.104277939388275e-06,
+      "loss": 0.5035,
+      "step": 5682
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.637097862003304,
+      "learning_rate": 6.103037511888678e-06,
+      "loss": 0.4909,
+      "step": 5683
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.987806739224232,
+      "learning_rate": 6.101797013023999e-06,
+      "loss": 0.4649,
+      "step": 5684
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.548736000432576,
+      "learning_rate": 6.100556442874495e-06,
+      "loss": 0.4993,
+      "step": 5685
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5895862476917514,
+      "learning_rate": 6.099315801520428e-06,
+      "loss": 0.514,
+      "step": 5686
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6140040396279058,
+      "learning_rate": 6.098075089042068e-06,
+      "loss": 0.3912,
+      "step": 5687
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9076130720577347,
+      "learning_rate": 6.096834305519684e-06,
+      "loss": 0.4873,
+      "step": 5688
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.869972422600285,
+      "learning_rate": 6.095593451033557e-06,
+      "loss": 0.5,
+      "step": 5689
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.794705902778675,
+      "learning_rate": 6.094352525663967e-06,
+      "loss": 0.436,
+      "step": 5690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3685303822320125,
+      "learning_rate": 6.0931115294911994e-06,
+      "loss": 0.4863,
+      "step": 5691
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6671922666354537,
+      "learning_rate": 6.091870462595545e-06,
+      "loss": 0.4458,
+      "step": 5692
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5714931401105673,
+      "learning_rate": 6.090629325057302e-06,
+      "loss": 0.5101,
+      "step": 5693
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5413340140346248,
+      "learning_rate": 6.089388116956767e-06,
+      "loss": 0.5047,
+      "step": 5694
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.690536722509592,
+      "learning_rate": 6.088146838374247e-06,
+      "loss": 0.532,
+      "step": 5695
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.4012375008138656,
+      "learning_rate": 6.0869054893900485e-06,
+      "loss": 0.4745,
+      "step": 5696
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8125321997025927,
+      "learning_rate": 6.0856640700844885e-06,
+      "loss": 0.4847,
+      "step": 5697
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5614235174931688,
+      "learning_rate": 6.084422580537882e-06,
+      "loss": 0.5125,
+      "step": 5698
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.492267651310338,
+      "learning_rate": 6.083181020830553e-06,
+      "loss": 0.5031,
+      "step": 5699
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5706014790783737,
+      "learning_rate": 6.081939391042829e-06,
+      "loss": 0.5213,
+      "step": 5700
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4679643714721506,
+      "learning_rate": 6.080697691255043e-06,
+      "loss": 0.5183,
+      "step": 5701
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1385504928518815,
+      "learning_rate": 6.079455921547529e-06,
+      "loss": 0.4805,
+      "step": 5702
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5892377076051167,
+      "learning_rate": 6.07821408200063e-06,
+      "loss": 0.5081,
+      "step": 5703
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4952763533720472,
+      "learning_rate": 6.07697217269469e-06,
+      "loss": 0.421,
+      "step": 5704
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4896629001385853,
+      "learning_rate": 6.075730193710059e-06,
+      "loss": 0.4617,
+      "step": 5705
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.685119616543394,
+      "learning_rate": 6.074488145127091e-06,
+      "loss": 0.4695,
+      "step": 5706
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5276080265530291,
+      "learning_rate": 6.073246027026147e-06,
+      "loss": 0.504,
+      "step": 5707
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8413784587994093,
+      "learning_rate": 6.072003839487588e-06,
+      "loss": 0.4676,
+      "step": 5708
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6453849094167343,
+      "learning_rate": 6.070761582591784e-06,
+      "loss": 0.4066,
+      "step": 5709
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8220154598935012,
+      "learning_rate": 6.069519256419104e-06,
+      "loss": 0.4609,
+      "step": 5710
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9352526671568626,
+      "learning_rate": 6.06827686104993e-06,
+      "loss": 0.4962,
+      "step": 5711
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6158187515564246,
+      "learning_rate": 6.0670343965646385e-06,
+      "loss": 0.4704,
+      "step": 5712
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7937338194347454,
+      "learning_rate": 6.06579186304362e-06,
+      "loss": 0.4004,
+      "step": 5713
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5891895608496196,
+      "learning_rate": 6.064549260567259e-06,
+      "loss": 0.5138,
+      "step": 5714
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7311893032605346,
+      "learning_rate": 6.063306589215956e-06,
+      "loss": 0.5058,
+      "step": 5715
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8248935108612583,
+      "learning_rate": 6.062063849070106e-06,
+      "loss": 0.3959,
+      "step": 5716
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6331680136460127,
+      "learning_rate": 6.0608210402101156e-06,
+      "loss": 0.495,
+      "step": 5717
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 4.302594367416118,
+      "learning_rate": 6.0595781627163906e-06,
+      "loss": 0.4702,
+      "step": 5718
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8406587877177791,
+      "learning_rate": 6.058335216669344e-06,
+      "loss": 0.4401,
+      "step": 5719
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0085716063385486,
+      "learning_rate": 6.057092202149395e-06,
+      "loss": 0.42,
+      "step": 5720
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.7448568654844796,
+      "learning_rate": 6.055849119236963e-06,
+      "loss": 0.4825,
+      "step": 5721
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 7.1349576979723,
+      "learning_rate": 6.0546059680124735e-06,
+      "loss": 0.5271,
+      "step": 5722
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5914155662299729,
+      "learning_rate": 6.053362748556359e-06,
+      "loss": 0.5257,
+      "step": 5723
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.605385006408059,
+      "learning_rate": 6.052119460949051e-06,
+      "loss": 0.5052,
+      "step": 5724
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8822394185843716,
+      "learning_rate": 6.05087610527099e-06,
+      "loss": 0.5202,
+      "step": 5725
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0223771521952125,
+      "learning_rate": 6.049632681602621e-06,
+      "loss": 0.4719,
+      "step": 5726
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5583886230057684,
+      "learning_rate": 6.04838919002439e-06,
+      "loss": 0.4862,
+      "step": 5727
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.948650585882006,
+      "learning_rate": 6.04714563061675e-06,
+      "loss": 0.4927,
+      "step": 5728
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7708249425051292,
+      "learning_rate": 6.045902003460157e-06,
+      "loss": 0.4573,
+      "step": 5729
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.1458237845597736,
+      "learning_rate": 6.044658308635074e-06,
+      "loss": 0.4828,
+      "step": 5730
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6810915776146784,
+      "learning_rate": 6.043414546221963e-06,
+      "loss": 0.485,
+      "step": 5731
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4352756727957299,
+      "learning_rate": 6.042170716301298e-06,
+      "loss": 0.4423,
+      "step": 5732
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.971001893509812,
+      "learning_rate": 6.040926818953548e-06,
+      "loss": 0.4902,
+      "step": 5733
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7169653586652869,
+      "learning_rate": 6.039682854259197e-06,
+      "loss": 0.4191,
+      "step": 5734
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.9488672408935992,
+      "learning_rate": 6.038438822298725e-06,
+      "loss": 0.5054,
+      "step": 5735
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5875973062325169,
+      "learning_rate": 6.037194723152619e-06,
+      "loss": 0.5015,
+      "step": 5736
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.164112220157697,
+      "learning_rate": 6.035950556901371e-06,
+      "loss": 0.4618,
+      "step": 5737
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.682455784345946,
+      "learning_rate": 6.034706323625479e-06,
+      "loss": 0.4737,
+      "step": 5738
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1209283612283025,
+      "learning_rate": 6.03346202340544e-06,
+      "loss": 0.4622,
+      "step": 5739
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1890772648731014,
+      "learning_rate": 6.032217656321761e-06,
+      "loss": 0.4766,
+      "step": 5740
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.1284504151421917,
+      "learning_rate": 6.030973222454949e-06,
+      "loss": 0.4703,
+      "step": 5741
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.343652778061655,
+      "learning_rate": 6.029728721885518e-06,
+      "loss": 0.4845,
+      "step": 5742
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.47550296793369,
+      "learning_rate": 6.028484154693987e-06,
+      "loss": 0.4711,
+      "step": 5743
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.627674144719648,
+      "learning_rate": 6.027239520960875e-06,
+      "loss": 0.4157,
+      "step": 5744
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6220238193866643,
+      "learning_rate": 6.0259948207667095e-06,
+      "loss": 0.4749,
+      "step": 5745
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5849949400029555,
+      "learning_rate": 6.024750054192023e-06,
+      "loss": 0.5117,
+      "step": 5746
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7299428157403334,
+      "learning_rate": 6.023505221317347e-06,
+      "loss": 0.4246,
+      "step": 5747
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6610500413073104,
+      "learning_rate": 6.022260322223224e-06,
+      "loss": 0.4383,
+      "step": 5748
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.7270604454516976,
+      "learning_rate": 6.021015356990194e-06,
+      "loss": 0.4499,
+      "step": 5749
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6603085366683752,
+      "learning_rate": 6.0197703256988075e-06,
+      "loss": 0.4452,
+      "step": 5750
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.151815004081429,
+      "learning_rate": 6.018525228429614e-06,
+      "loss": 0.5395,
+      "step": 5751
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.2779753199704955,
+      "learning_rate": 6.0172800652631706e-06,
+      "loss": 0.5252,
+      "step": 5752
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0007795465214526,
+      "learning_rate": 6.016034836280037e-06,
+      "loss": 0.4379,
+      "step": 5753
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7615038788753057,
+      "learning_rate": 6.0147895415607795e-06,
+      "loss": 0.4453,
+      "step": 5754
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0738005062453295,
+      "learning_rate": 6.013544181185966e-06,
+      "loss": 0.4951,
+      "step": 5755
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0074224481328478,
+      "learning_rate": 6.012298755236169e-06,
+      "loss": 0.4618,
+      "step": 5756
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7868557758287615,
+      "learning_rate": 6.011053263791967e-06,
+      "loss": 0.4597,
+      "step": 5757
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6081913380449693,
+      "learning_rate": 6.009807706933943e-06,
+      "loss": 0.4333,
+      "step": 5758
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.3109878533614023,
+      "learning_rate": 6.008562084742681e-06,
+      "loss": 0.4344,
+      "step": 5759
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5735955320905453,
+      "learning_rate": 6.00731639729877e-06,
+      "loss": 0.4512,
+      "step": 5760
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.117886234101295,
+      "learning_rate": 6.006070644682807e-06,
+      "loss": 0.5504,
+      "step": 5761
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4678537603924338,
+      "learning_rate": 6.004824826975389e-06,
+      "loss": 0.4911,
+      "step": 5762
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.0710004920839196,
+      "learning_rate": 6.0035789442571205e-06,
+      "loss": 0.4465,
+      "step": 5763
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6598112253798396,
+      "learning_rate": 6.002332996608605e-06,
+      "loss": 0.4876,
+      "step": 5764
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.325810267014837,
+      "learning_rate": 6.001086984110457e-06,
+      "loss": 0.4605,
+      "step": 5765
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8099724796625862,
+      "learning_rate": 5.99984090684329e-06,
+      "loss": 0.4487,
+      "step": 5766
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.584535945312912,
+      "learning_rate": 5.998594764887725e-06,
+      "loss": 0.4638,
+      "step": 5767
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.7647056725360064,
+      "learning_rate": 5.997348558324384e-06,
+      "loss": 0.4248,
+      "step": 5768
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 3.0337671979325114,
+      "learning_rate": 5.996102287233898e-06,
+      "loss": 0.4502,
+      "step": 5769
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.4895841480429866,
+      "learning_rate": 5.9948559516968955e-06,
+      "loss": 0.5093,
+      "step": 5770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.493398074058847,
+      "learning_rate": 5.993609551794016e-06,
+      "loss": 0.4554,
+      "step": 5771
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.046162329848687,
+      "learning_rate": 5.992363087605895e-06,
+      "loss": 0.5149,
+      "step": 5772
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5810174083949727,
+      "learning_rate": 5.9911165592131835e-06,
+      "loss": 0.5055,
+      "step": 5773
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8208328462982921,
+      "learning_rate": 5.989869966696525e-06,
+      "loss": 0.4898,
+      "step": 5774
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5252039598590208,
+      "learning_rate": 5.988623310136578e-06,
+      "loss": 0.5007,
+      "step": 5775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5644887687974547,
+      "learning_rate": 5.987376589613992e-06,
+      "loss": 0.5099,
+      "step": 5776
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5611758382408979,
+      "learning_rate": 5.9861298052094354e-06,
+      "loss": 0.5199,
+      "step": 5777
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.858815646612022,
+      "learning_rate": 5.984882957003567e-06,
+      "loss": 0.448,
+      "step": 5778
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5910068157214212,
+      "learning_rate": 5.983636045077062e-06,
+      "loss": 0.4759,
+      "step": 5779
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4403803114139198,
+      "learning_rate": 5.982389069510592e-06,
+      "loss": 0.4608,
+      "step": 5780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.5176631914239365,
+      "learning_rate": 5.981142030384835e-06,
+      "loss": 0.4919,
+      "step": 5781
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6531403102839233,
+      "learning_rate": 5.97989492778047e-06,
+      "loss": 0.4574,
+      "step": 5782
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.701885285227435,
+      "learning_rate": 5.978647761778187e-06,
+      "loss": 0.4707,
+      "step": 5783
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 5.784737868507821,
+      "learning_rate": 5.977400532458673e-06,
+      "loss": 0.5032,
+      "step": 5784
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.25873055910197,
+      "learning_rate": 5.976153239902622e-06,
+      "loss": 0.5111,
+      "step": 5785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.23372903442882,
+      "learning_rate": 5.974905884190735e-06,
+      "loss": 0.4403,
+      "step": 5786
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.658771267047421,
+      "learning_rate": 5.9736584654037125e-06,
+      "loss": 0.5111,
+      "step": 5787
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.014120464521342,
+      "learning_rate": 5.9724109836222595e-06,
+      "loss": 0.509,
+      "step": 5788
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.8811009732055821,
+      "learning_rate": 5.97116343892709e-06,
+      "loss": 0.4625,
+      "step": 5789
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.301368708006228,
+      "learning_rate": 5.969915831398915e-06,
+      "loss": 0.407,
+      "step": 5790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.957903136763112,
+      "learning_rate": 5.968668161118455e-06,
+      "loss": 0.4556,
+      "step": 5791
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6256141682239557,
+      "learning_rate": 5.967420428166432e-06,
+      "loss": 0.5057,
+      "step": 5792
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5909718018368074,
+      "learning_rate": 5.966172632623573e-06,
+      "loss": 0.5086,
+      "step": 5793
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.248574451857549,
+      "learning_rate": 5.9649247745706085e-06,
+      "loss": 0.4518,
+      "step": 5794
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.2609969695639482,
+      "learning_rate": 5.963676854088274e-06,
+      "loss": 0.4651,
+      "step": 5795
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5887988158852362,
+      "learning_rate": 5.962428871257307e-06,
+      "loss": 0.4871,
+      "step": 5796
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4628003113589134,
+      "learning_rate": 5.9611808261584505e-06,
+      "loss": 0.4872,
+      "step": 5797
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.649759791011889,
+      "learning_rate": 5.959932718872453e-06,
+      "loss": 0.4902,
+      "step": 5798
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.774957857779574,
+      "learning_rate": 5.9586845494800635e-06,
+      "loss": 0.4779,
+      "step": 5799
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8795058370938917,
+      "learning_rate": 5.957436318062039e-06,
+      "loss": 0.5101,
+      "step": 5800
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6475431171666449,
+      "learning_rate": 5.956188024699136e-06,
+      "loss": 0.482,
+      "step": 5801
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9402653298123897,
+      "learning_rate": 5.9549396694721185e-06,
+      "loss": 0.4635,
+      "step": 5802
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.077613746180248,
+      "learning_rate": 5.953691252461754e-06,
+      "loss": 0.4788,
+      "step": 5803
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4645388326324116,
+      "learning_rate": 5.952442773748814e-06,
+      "loss": 0.4342,
+      "step": 5804
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7827814670709088,
+      "learning_rate": 5.951194233414072e-06,
+      "loss": 0.4392,
+      "step": 5805
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 11.599889963554114,
+      "learning_rate": 5.949945631538308e-06,
+      "loss": 0.4467,
+      "step": 5806
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0564774717996097,
+      "learning_rate": 5.948696968202302e-06,
+      "loss": 0.4684,
+      "step": 5807
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8378023351677375,
+      "learning_rate": 5.9474482434868455e-06,
+      "loss": 0.4575,
+      "step": 5808
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3015476486609856,
+      "learning_rate": 5.946199457472726e-06,
+      "loss": 0.4994,
+      "step": 5809
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9291286241249834,
+      "learning_rate": 5.9449506102407395e-06,
+      "loss": 0.4256,
+      "step": 5810
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7290201644502774,
+      "learning_rate": 5.943701701871685e-06,
+      "loss": 0.4913,
+      "step": 5811
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7813529206888297,
+      "learning_rate": 5.942452732446365e-06,
+      "loss": 0.4252,
+      "step": 5812
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.508227137901349,
+      "learning_rate": 5.941203702045584e-06,
+      "loss": 0.5078,
+      "step": 5813
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8953557197110757,
+      "learning_rate": 5.939954610750157e-06,
+      "loss": 0.4688,
+      "step": 5814
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.188489622863546,
+      "learning_rate": 5.938705458640895e-06,
+      "loss": 0.4887,
+      "step": 5815
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.8806101190693325,
+      "learning_rate": 5.93745624579862e-06,
+      "loss": 0.4637,
+      "step": 5816
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.135551854779081,
+      "learning_rate": 5.93620697230415e-06,
+      "loss": 0.4196,
+      "step": 5817
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5877604130417955,
+      "learning_rate": 5.934957638238316e-06,
+      "loss": 0.5062,
+      "step": 5818
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.6831847265126774,
+      "learning_rate": 5.933708243681944e-06,
+      "loss": 0.5115,
+      "step": 5819
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4060520112272163,
+      "learning_rate": 5.93245878871587e-06,
+      "loss": 0.4588,
+      "step": 5820
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.305150963124646,
+      "learning_rate": 5.931209273420932e-06,
+      "loss": 0.4679,
+      "step": 5821
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8464200274340548,
+      "learning_rate": 5.929959697877974e-06,
+      "loss": 0.4407,
+      "step": 5822
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.475215356706476,
+      "learning_rate": 5.928710062167837e-06,
+      "loss": 0.5106,
+      "step": 5823
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.105605225564653,
+      "learning_rate": 5.927460366371377e-06,
+      "loss": 0.4918,
+      "step": 5824
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0596178826789147,
+      "learning_rate": 5.926210610569443e-06,
+      "loss": 0.4174,
+      "step": 5825
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0792331550889926,
+      "learning_rate": 5.924960794842894e-06,
+      "loss": 0.4528,
+      "step": 5826
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.744924490311416,
+      "learning_rate": 5.923710919272593e-06,
+      "loss": 0.4131,
+      "step": 5827
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 4.871907315539418,
+      "learning_rate": 5.922460983939403e-06,
+      "loss": 0.5108,
+      "step": 5828
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5535667202844723,
+      "learning_rate": 5.921210988924193e-06,
+      "loss": 0.5034,
+      "step": 5829
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8944184619568136,
+      "learning_rate": 5.919960934307839e-06,
+      "loss": 0.4306,
+      "step": 5830
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7615248959019227,
+      "learning_rate": 5.9187108201712155e-06,
+      "loss": 0.4884,
+      "step": 5831
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4685038738545786,
+      "learning_rate": 5.917460646595202e-06,
+      "loss": 0.4474,
+      "step": 5832
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0704001362419184,
+      "learning_rate": 5.916210413660687e-06,
+      "loss": 0.517,
+      "step": 5833
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0819600663793154,
+      "learning_rate": 5.914960121448556e-06,
+      "loss": 0.5301,
+      "step": 5834
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3578031514753697,
+      "learning_rate": 5.913709770039702e-06,
+      "loss": 0.4648,
+      "step": 5835
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8342873452017652,
+      "learning_rate": 5.912459359515022e-06,
+      "loss": 0.4688,
+      "step": 5836
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5883821569696783,
+      "learning_rate": 5.911208889955413e-06,
+      "loss": 0.4548,
+      "step": 5837
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6189352671434525,
+      "learning_rate": 5.909958361441782e-06,
+      "loss": 0.4862,
+      "step": 5838
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3753168564529217,
+      "learning_rate": 5.9087077740550354e-06,
+      "loss": 0.5519,
+      "step": 5839
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.560424545639952,
+      "learning_rate": 5.907457127876085e-06,
+      "loss": 0.5098,
+      "step": 5840
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5161552311707477,
+      "learning_rate": 5.9062064229858465e-06,
+      "loss": 0.4769,
+      "step": 5841
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.0366954379294806,
+      "learning_rate": 5.904955659465236e-06,
+      "loss": 0.4596,
+      "step": 5842
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5889292786206042,
+      "learning_rate": 5.9037048373951785e-06,
+      "loss": 0.5273,
+      "step": 5843
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4652182913938556,
+      "learning_rate": 5.902453956856601e-06,
+      "loss": 0.4592,
+      "step": 5844
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5794530765775763,
+      "learning_rate": 5.901203017930432e-06,
+      "loss": 0.4952,
+      "step": 5845
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.5782303829548447,
+      "learning_rate": 5.8999520206976065e-06,
+      "loss": 0.454,
+      "step": 5846
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5941284893375037,
+      "learning_rate": 5.898700965239064e-06,
+      "loss": 0.5105,
+      "step": 5847
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0284841697966636,
+      "learning_rate": 5.897449851635743e-06,
+      "loss": 0.4768,
+      "step": 5848
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7926180153985334,
+      "learning_rate": 5.896198679968592e-06,
+      "loss": 0.4427,
+      "step": 5849
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3373848382284095,
+      "learning_rate": 5.894947450318559e-06,
+      "loss": 0.5071,
+      "step": 5850
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.701287282429645,
+      "learning_rate": 5.8936961627665964e-06,
+      "loss": 0.4958,
+      "step": 5851
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6441484297602644,
+      "learning_rate": 5.89244481739366e-06,
+      "loss": 0.4595,
+      "step": 5852
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7045652437646563,
+      "learning_rate": 5.891193414280714e-06,
+      "loss": 0.4775,
+      "step": 5853
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8301237565529862,
+      "learning_rate": 5.889941953508716e-06,
+      "loss": 0.4946,
+      "step": 5854
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.908083974841471,
+      "learning_rate": 5.888690435158642e-06,
+      "loss": 0.5058,
+      "step": 5855
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7230945609893141,
+      "learning_rate": 5.887438859311456e-06,
+      "loss": 0.5011,
+      "step": 5856
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6005786464384618,
+      "learning_rate": 5.886187226048138e-06,
+      "loss": 0.4746,
+      "step": 5857
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5545303461926905,
+      "learning_rate": 5.884935535449665e-06,
+      "loss": 0.4378,
+      "step": 5858
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.338666886282173,
+      "learning_rate": 5.883683787597022e-06,
+      "loss": 0.4155,
+      "step": 5859
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9903876810844285,
+      "learning_rate": 5.882431982571193e-06,
+      "loss": 0.5025,
+      "step": 5860
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0290183277474707,
+      "learning_rate": 5.881180120453171e-06,
+      "loss": 0.4494,
+      "step": 5861
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6745766293342752,
+      "learning_rate": 5.8799282013239456e-06,
+      "loss": 0.5016,
+      "step": 5862
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.27649687448856,
+      "learning_rate": 5.878676225264517e-06,
+      "loss": 0.5101,
+      "step": 5863
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.188612340110292,
+      "learning_rate": 5.877424192355886e-06,
+      "loss": 0.4611,
+      "step": 5864
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.719474048108263,
+      "learning_rate": 5.876172102679058e-06,
+      "loss": 0.453,
+      "step": 5865
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5800762730200366,
+      "learning_rate": 5.8749199563150415e-06,
+      "loss": 0.4837,
+      "step": 5866
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9540804621446486,
+      "learning_rate": 5.873667753344847e-06,
+      "loss": 0.4548,
+      "step": 5867
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7298447287171492,
+      "learning_rate": 5.872415493849494e-06,
+      "loss": 0.4989,
+      "step": 5868
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.5251443249977985,
+      "learning_rate": 5.871163177909998e-06,
+      "loss": 0.4316,
+      "step": 5869
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6477194391251179,
+      "learning_rate": 5.869910805607384e-06,
+      "loss": 0.5328,
+      "step": 5870
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1100128686417925,
+      "learning_rate": 5.8686583770226805e-06,
+      "loss": 0.5086,
+      "step": 5871
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1297005354225123,
+      "learning_rate": 5.867405892236915e-06,
+      "loss": 0.4539,
+      "step": 5872
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6482849869182417,
+      "learning_rate": 5.866153351331123e-06,
+      "loss": 0.4884,
+      "step": 5873
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8791448557362458,
+      "learning_rate": 5.864900754386342e-06,
+      "loss": 0.5082,
+      "step": 5874
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5666265538926212,
+      "learning_rate": 5.863648101483614e-06,
+      "loss": 0.5033,
+      "step": 5875
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3887828814784986,
+      "learning_rate": 5.8623953927039845e-06,
+      "loss": 0.4425,
+      "step": 5876
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6129129590424698,
+      "learning_rate": 5.8611426281285e-06,
+      "loss": 0.5069,
+      "step": 5877
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5861507619302027,
+      "learning_rate": 5.859889807838216e-06,
+      "loss": 0.5193,
+      "step": 5878
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.580961921002182,
+      "learning_rate": 5.858636931914184e-06,
+      "loss": 0.4114,
+      "step": 5879
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4795720861590165,
+      "learning_rate": 5.857384000437466e-06,
+      "loss": 0.4615,
+      "step": 5880
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6182215967166627,
+      "learning_rate": 5.8561310134891246e-06,
+      "loss": 0.4737,
+      "step": 5881
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.003404284900797,
+      "learning_rate": 5.8548779711502275e-06,
+      "loss": 0.4866,
+      "step": 5882
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6406674874483526,
+      "learning_rate": 5.853624873501844e-06,
+      "loss": 0.5045,
+      "step": 5883
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6747791959148048,
+      "learning_rate": 5.8523717206250485e-06,
+      "loss": 0.4609,
+      "step": 5884
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.48182764399353,
+      "learning_rate": 5.8511185126009165e-06,
+      "loss": 0.4513,
+      "step": 5885
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.3013474834564467,
+      "learning_rate": 5.8498652495105315e-06,
+      "loss": 0.4803,
+      "step": 5886
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.9761266099442807,
+      "learning_rate": 5.848611931434975e-06,
+      "loss": 0.4795,
+      "step": 5887
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.265592159894518,
+      "learning_rate": 5.847358558455339e-06,
+      "loss": 0.5079,
+      "step": 5888
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6691885384520395,
+      "learning_rate": 5.84610513065271e-06,
+      "loss": 0.4334,
+      "step": 5889
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5669935443867741,
+      "learning_rate": 5.844851648108188e-06,
+      "loss": 0.4757,
+      "step": 5890
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7369213667952494,
+      "learning_rate": 5.843598110902866e-06,
+      "loss": 0.4559,
+      "step": 5891
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5834001184163092,
+      "learning_rate": 5.842344519117853e-06,
+      "loss": 0.5202,
+      "step": 5892
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5718210239542848,
+      "learning_rate": 5.841090872834249e-06,
+      "loss": 0.5046,
+      "step": 5893
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6549160364089908,
+      "learning_rate": 5.839837172133169e-06,
+      "loss": 0.4875,
+      "step": 5894
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5492828206515015,
+      "learning_rate": 5.838583417095721e-06,
+      "loss": 0.4891,
+      "step": 5895
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8771072769125519,
+      "learning_rate": 5.837329607803024e-06,
+      "loss": 0.4361,
+      "step": 5896
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.086071737548597,
+      "learning_rate": 5.836075744336196e-06,
+      "loss": 0.4442,
+      "step": 5897
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6688669223306436,
+      "learning_rate": 5.834821826776361e-06,
+      "loss": 0.5062,
+      "step": 5898
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5505965613061234,
+      "learning_rate": 5.833567855204646e-06,
+      "loss": 0.5031,
+      "step": 5899
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4575038642462506,
+      "learning_rate": 5.832313829702181e-06,
+      "loss": 0.4732,
+      "step": 5900
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5453788505134876,
+      "learning_rate": 5.8310597503501e-06,
+      "loss": 0.5071,
+      "step": 5901
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.079281880434146,
+      "learning_rate": 5.82980561722954e-06,
+      "loss": 0.4382,
+      "step": 5902
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.8653142660971853,
+      "learning_rate": 5.828551430421642e-06,
+      "loss": 0.4877,
+      "step": 5903
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.0731564170315764,
+      "learning_rate": 5.8272971900075516e-06,
+      "loss": 0.4532,
+      "step": 5904
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 5.356321262044365,
+      "learning_rate": 5.826042896068414e-06,
+      "loss": 0.4857,
+      "step": 5905
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.9722680919458162,
+      "learning_rate": 5.824788548685383e-06,
+      "loss": 0.499,
+      "step": 5906
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5622958669665656,
+      "learning_rate": 5.82353414793961e-06,
+      "loss": 0.4293,
+      "step": 5907
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.684607302907042,
+      "learning_rate": 5.822279693912257e-06,
+      "loss": 0.45,
+      "step": 5908
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.1279044288798445,
+      "learning_rate": 5.821025186684482e-06,
+      "loss": 0.4696,
+      "step": 5909
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.439485712706476,
+      "learning_rate": 5.819770626337451e-06,
+      "loss": 0.4364,
+      "step": 5910
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.46258574072006,
+      "learning_rate": 5.818516012952335e-06,
+      "loss": 0.4123,
+      "step": 5911
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6629937964388954,
+      "learning_rate": 5.817261346610301e-06,
+      "loss": 0.4902,
+      "step": 5912
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6959620808866966,
+      "learning_rate": 5.81600662739253e-06,
+      "loss": 0.4865,
+      "step": 5913
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6255481884661898,
+      "learning_rate": 5.814751855380195e-06,
+      "loss": 0.4597,
+      "step": 5914
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 3.4048018076923885,
+      "learning_rate": 5.813497030654483e-06,
+      "loss": 0.4697,
+      "step": 5915
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 6.775999371655901,
+      "learning_rate": 5.812242153296574e-06,
+      "loss": 0.4516,
+      "step": 5916
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6307908014036668,
+      "learning_rate": 5.810987223387664e-06,
+      "loss": 0.516,
+      "step": 5917
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.6439773845493402,
+      "learning_rate": 5.80973224100894e-06,
+      "loss": 0.4519,
+      "step": 5918
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.7136551366272617,
+      "learning_rate": 5.808477206241599e-06,
+      "loss": 0.4104,
+      "step": 5919
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.213209614489724,
+      "learning_rate": 5.80722211916684e-06,
+      "loss": 0.4363,
+      "step": 5920
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5592161993665855,
+      "learning_rate": 5.805966979865868e-06,
+      "loss": 0.4957,
+      "step": 5921
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0948468695595435,
+      "learning_rate": 5.804711788419885e-06,
+      "loss": 0.5055,
+      "step": 5922
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.531894528417516,
+      "learning_rate": 5.8034565449101025e-06,
+      "loss": 0.4574,
+      "step": 5923
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7642893769557353,
+      "learning_rate": 5.802201249417732e-06,
+      "loss": 0.4951,
+      "step": 5924
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.61287748665114,
+      "learning_rate": 5.8009459020239885e-06,
+      "loss": 0.4701,
+      "step": 5925
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 7.672349005699024,
+      "learning_rate": 5.799690502810096e-06,
+      "loss": 0.4849,
+      "step": 5926
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9783980681204234,
+      "learning_rate": 5.798435051857273e-06,
+      "loss": 0.4968,
+      "step": 5927
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5287259404572018,
+      "learning_rate": 5.797179549246746e-06,
+      "loss": 0.4846,
+      "step": 5928
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5303634349547306,
+      "learning_rate": 5.795923995059747e-06,
+      "loss": 0.4593,
+      "step": 5929
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.494214973666261,
+      "learning_rate": 5.794668389377503e-06,
+      "loss": 0.5137,
+      "step": 5930
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8227509442146623,
+      "learning_rate": 5.793412732281258e-06,
+      "loss": 0.4355,
+      "step": 5931
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8650901351854052,
+      "learning_rate": 5.792157023852244e-06,
+      "loss": 0.4687,
+      "step": 5932
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2814056773803286,
+      "learning_rate": 5.790901264171709e-06,
+      "loss": 0.5157,
+      "step": 5933
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.230403025635125,
+      "learning_rate": 5.789645453320896e-06,
+      "loss": 0.4747,
+      "step": 5934
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7285978324665887,
+      "learning_rate": 5.788389591381055e-06,
+      "loss": 0.4313,
+      "step": 5935
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8364205903394037,
+      "learning_rate": 5.787133678433437e-06,
+      "loss": 0.47,
+      "step": 5936
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6299063429864755,
+      "learning_rate": 5.785877714559303e-06,
+      "loss": 0.4333,
+      "step": 5937
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 10.165181644438015,
+      "learning_rate": 5.7846216998399065e-06,
+      "loss": 0.4852,
+      "step": 5938
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5364922797222298,
+      "learning_rate": 5.783365634356514e-06,
+      "loss": 0.4798,
+      "step": 5939
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8760757305868798,
+      "learning_rate": 5.7821095181903885e-06,
+      "loss": 0.5114,
+      "step": 5940
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.07369167041546,
+      "learning_rate": 5.7808533514228014e-06,
+      "loss": 0.4416,
+      "step": 5941
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6292196711580126,
+      "learning_rate": 5.7795971341350235e-06,
+      "loss": 0.4933,
+      "step": 5942
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3910067975328642,
+      "learning_rate": 5.778340866408331e-06,
+      "loss": 0.4906,
+      "step": 5943
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3985509493240404,
+      "learning_rate": 5.7770845483240015e-06,
+      "loss": 0.4674,
+      "step": 5944
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.942834929969426,
+      "learning_rate": 5.775828179963318e-06,
+      "loss": 0.4643,
+      "step": 5945
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.772759246252017,
+      "learning_rate": 5.7745717614075695e-06,
+      "loss": 0.5548,
+      "step": 5946
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.1086487161254115,
+      "learning_rate": 5.773315292738038e-06,
+      "loss": 0.4232,
+      "step": 5947
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6937888880106402,
+      "learning_rate": 5.77205877403602e-06,
+      "loss": 0.4371,
+      "step": 5948
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.227470297728218,
+      "learning_rate": 5.770802205382809e-06,
+      "loss": 0.5161,
+      "step": 5949
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.3958380194699096,
+      "learning_rate": 5.769545586859704e-06,
+      "loss": 0.4857,
+      "step": 5950
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.513528106049459,
+      "learning_rate": 5.768288918548005e-06,
+      "loss": 0.4596,
+      "step": 5951
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.580650690269417,
+      "learning_rate": 5.76703220052902e-06,
+      "loss": 0.4364,
+      "step": 5952
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6395933368030706,
+      "learning_rate": 5.765775432884053e-06,
+      "loss": 0.4253,
+      "step": 5953
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.643788452978433,
+      "learning_rate": 5.764518615694419e-06,
+      "loss": 0.5019,
+      "step": 5954
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8066383558020616,
+      "learning_rate": 5.76326174904143e-06,
+      "loss": 0.5148,
+      "step": 5955
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5725243410183003,
+      "learning_rate": 5.7620048330064045e-06,
+      "loss": 0.5071,
+      "step": 5956
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.564548069219352,
+      "learning_rate": 5.760747867670663e-06,
+      "loss": 0.4497,
+      "step": 5957
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5178917865002612,
+      "learning_rate": 5.7594908531155305e-06,
+      "loss": 0.4558,
+      "step": 5958
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 12.702301388318126,
+      "learning_rate": 5.7582337894223305e-06,
+      "loss": 0.5076,
+      "step": 5959
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.597597928948306,
+      "learning_rate": 5.756976676672399e-06,
+      "loss": 0.5043,
+      "step": 5960
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5859302154306599,
+      "learning_rate": 5.755719514947064e-06,
+      "loss": 0.503,
+      "step": 5961
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.713056976452306,
+      "learning_rate": 5.754462304327668e-06,
+      "loss": 0.4617,
+      "step": 5962
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.571608808028225,
+      "learning_rate": 5.753205044895547e-06,
+      "loss": 0.4996,
+      "step": 5963
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6667582144887978,
+      "learning_rate": 5.751947736732045e-06,
+      "loss": 0.4845,
+      "step": 5964
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8558307893142236,
+      "learning_rate": 5.750690379918508e-06,
+      "loss": 0.5526,
+      "step": 5965
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0765797703396154,
+      "learning_rate": 5.749432974536285e-06,
+      "loss": 0.4765,
+      "step": 5966
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.34182498614451,
+      "learning_rate": 5.748175520666729e-06,
+      "loss": 0.4294,
+      "step": 5967
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7588767464239168,
+      "learning_rate": 5.746918018391196e-06,
+      "loss": 0.4777,
+      "step": 5968
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1987124651469236,
+      "learning_rate": 5.745660467791045e-06,
+      "loss": 0.4398,
+      "step": 5969
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6259079926310644,
+      "learning_rate": 5.744402868947635e-06,
+      "loss": 0.5044,
+      "step": 5970
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.255531572357372,
+      "learning_rate": 5.743145221942333e-06,
+      "loss": 0.4325,
+      "step": 5971
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.318388191146558,
+      "learning_rate": 5.74188752685651e-06,
+      "loss": 0.5105,
+      "step": 5972
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5472737298567707,
+      "learning_rate": 5.740629783771533e-06,
+      "loss": 0.4963,
+      "step": 5973
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.511725091491056,
+      "learning_rate": 5.739371992768779e-06,
+      "loss": 0.5015,
+      "step": 5974
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.255027601315444,
+      "learning_rate": 5.738114153929622e-06,
+      "loss": 0.4625,
+      "step": 5975
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.606208971416444,
+      "learning_rate": 5.736856267335447e-06,
+      "loss": 0.5026,
+      "step": 5976
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5677110898383697,
+      "learning_rate": 5.735598333067634e-06,
+      "loss": 0.493,
+      "step": 5977
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5046261711409938,
+      "learning_rate": 5.734340351207572e-06,
+      "loss": 0.449,
+      "step": 5978
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.634642446531102,
+      "learning_rate": 5.733082321836649e-06,
+      "loss": 0.4347,
+      "step": 5979
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5352523302536238,
+      "learning_rate": 5.7318242450362594e-06,
+      "loss": 0.4791,
+      "step": 5980
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6656548186386357,
+      "learning_rate": 5.7305661208877995e-06,
+      "loss": 0.402,
+      "step": 5981
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5973999745947913,
+      "learning_rate": 5.729307949472666e-06,
+      "loss": 0.4559,
+      "step": 5982
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0682700701942016,
+      "learning_rate": 5.728049730872262e-06,
+      "loss": 0.489,
+      "step": 5983
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.3387261942253,
+      "learning_rate": 5.7267914651679935e-06,
+      "loss": 0.4133,
+      "step": 5984
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.884499302203894,
+      "learning_rate": 5.725533152441267e-06,
+      "loss": 0.4843,
+      "step": 5985
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.0349422675370645,
+      "learning_rate": 5.724274792773496e-06,
+      "loss": 0.4691,
+      "step": 5986
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8249198288310222,
+      "learning_rate": 5.723016386246093e-06,
+      "loss": 0.4366,
+      "step": 5987
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.786384309504527,
+      "learning_rate": 5.7217579329404745e-06,
+      "loss": 0.4804,
+      "step": 5988
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3222440296848124,
+      "learning_rate": 5.720499432938065e-06,
+      "loss": 0.5139,
+      "step": 5989
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9655359874409943,
+      "learning_rate": 5.719240886320281e-06,
+      "loss": 0.499,
+      "step": 5990
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7512345875837911,
+      "learning_rate": 5.717982293168555e-06,
+      "loss": 0.5165,
+      "step": 5991
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6144634047367286,
+      "learning_rate": 5.716723653564312e-06,
+      "loss": 0.4858,
+      "step": 5992
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5081456494765408,
+      "learning_rate": 5.715464967588987e-06,
+      "loss": 0.473,
+      "step": 5993
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9960947497187504,
+      "learning_rate": 5.714206235324015e-06,
+      "loss": 0.4872,
+      "step": 5994
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5849383844809959,
+      "learning_rate": 5.712947456850834e-06,
+      "loss": 0.5018,
+      "step": 5995
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8764390107526074,
+      "learning_rate": 5.7116886322508845e-06,
+      "loss": 0.4967,
+      "step": 5996
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1808623988109983,
+      "learning_rate": 5.710429761605613e-06,
+      "loss": 0.431,
+      "step": 5997
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7510816402407445,
+      "learning_rate": 5.709170844996463e-06,
+      "loss": 0.5047,
+      "step": 5998
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5818425859911756,
+      "learning_rate": 5.707911882504888e-06,
+      "loss": 0.4847,
+      "step": 5999
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5890593975580446,
+      "learning_rate": 5.706652874212341e-06,
+      "loss": 0.515,
+      "step": 6000
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5295822772699232,
+      "learning_rate": 5.705393820200276e-06,
+      "loss": 0.4489,
+      "step": 6001
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9210435848516876,
+      "learning_rate": 5.704134720550155e-06,
+      "loss": 0.4681,
+      "step": 6002
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.594082442984845,
+      "learning_rate": 5.702875575343438e-06,
+      "loss": 0.4982,
+      "step": 6003
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.251270663811119,
+      "learning_rate": 5.70161638466159e-06,
+      "loss": 0.4951,
+      "step": 6004
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6134939082862363,
+      "learning_rate": 5.700357148586082e-06,
+      "loss": 0.4499,
+      "step": 6005
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.00655544046567,
+      "learning_rate": 5.6990978671983795e-06,
+      "loss": 0.4481,
+      "step": 6006
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6057074508771052,
+      "learning_rate": 5.697838540579963e-06,
+      "loss": 0.4855,
+      "step": 6007
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.670485100193163,
+      "learning_rate": 5.696579168812304e-06,
+      "loss": 0.4607,
+      "step": 6008
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6554599349703256,
+      "learning_rate": 5.6953197519768845e-06,
+      "loss": 0.4721,
+      "step": 6009
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.5312037659960067,
+      "learning_rate": 5.694060290155186e-06,
+      "loss": 0.4452,
+      "step": 6010
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7912825747360135,
+      "learning_rate": 5.692800783428696e-06,
+      "loss": 0.4743,
+      "step": 6011
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 5.044010005804708,
+      "learning_rate": 5.6915412318789e-06,
+      "loss": 0.485,
+      "step": 6012
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4431739806162485,
+      "learning_rate": 5.690281635587291e-06,
+      "loss": 0.5,
+      "step": 6013
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.4282775356192032,
+      "learning_rate": 5.6890219946353645e-06,
+      "loss": 0.4992,
+      "step": 6014
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.7417476308872604,
+      "learning_rate": 5.687762309104615e-06,
+      "loss": 0.4792,
+      "step": 6015
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1528903956823586,
+      "learning_rate": 5.686502579076544e-06,
+      "loss": 0.442,
+      "step": 6016
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4983715341979529,
+      "learning_rate": 5.685242804632655e-06,
+      "loss": 0.4506,
+      "step": 6017
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2836038021960925,
+      "learning_rate": 5.683982985854452e-06,
+      "loss": 0.5072,
+      "step": 6018
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.510402468002663,
+      "learning_rate": 5.682723122823446e-06,
+      "loss": 0.482,
+      "step": 6019
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.85107953551789,
+      "learning_rate": 5.681463215621146e-06,
+      "loss": 0.4981,
+      "step": 6020
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.162470487262685,
+      "learning_rate": 5.680203264329066e-06,
+      "loss": 0.5077,
+      "step": 6021
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.2268190513228987,
+      "learning_rate": 5.678943269028727e-06,
+      "loss": 0.4319,
+      "step": 6022
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.5675505839119994,
+      "learning_rate": 5.677683229801646e-06,
+      "loss": 0.5217,
+      "step": 6023
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7178357654853753,
+      "learning_rate": 5.6764231467293465e-06,
+      "loss": 0.4491,
+      "step": 6024
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6046995496007133,
+      "learning_rate": 5.675163019893354e-06,
+      "loss": 0.4769,
+      "step": 6025
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 4.194152488506448,
+      "learning_rate": 5.673902849375198e-06,
+      "loss": 0.4473,
+      "step": 6026
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8001871807266456,
+      "learning_rate": 5.6726426352564065e-06,
+      "loss": 0.5182,
+      "step": 6027
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9959096874324542,
+      "learning_rate": 5.67138237761852e-06,
+      "loss": 0.4728,
+      "step": 6028
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5764329495487379,
+      "learning_rate": 5.6701220765430694e-06,
+      "loss": 0.4846,
+      "step": 6029
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.1121280222004617,
+      "learning_rate": 5.668861732111599e-06,
+      "loss": 0.4644,
+      "step": 6030
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 6.604238939529686,
+      "learning_rate": 5.667601344405649e-06,
+      "loss": 0.4449,
+      "step": 6031
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.741670168649439,
+      "learning_rate": 5.666340913506766e-06,
+      "loss": 0.4586,
+      "step": 6032
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5380322233169764,
+      "learning_rate": 5.665080439496495e-06,
+      "loss": 0.483,
+      "step": 6033
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.9073685314972835,
+      "learning_rate": 5.663819922456393e-06,
+      "loss": 0.4518,
+      "step": 6034
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8744899268265252,
+      "learning_rate": 5.662559362468007e-06,
+      "loss": 0.4964,
+      "step": 6035
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9140057755041548,
+      "learning_rate": 5.661298759612901e-06,
+      "loss": 0.4472,
+      "step": 6036
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.394471346075338,
+      "learning_rate": 5.6600381139726264e-06,
+      "loss": 0.4904,
+      "step": 6037
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.83579527797309,
+      "learning_rate": 5.6587774256287505e-06,
+      "loss": 0.4636,
+      "step": 6038
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8847597530243525,
+      "learning_rate": 5.657516694662838e-06,
+      "loss": 0.473,
+      "step": 6039
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5636579285514544,
+      "learning_rate": 5.656255921156455e-06,
+      "loss": 0.4866,
+      "step": 6040
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0074486311426494,
+      "learning_rate": 5.654995105191172e-06,
+      "loss": 0.459,
+      "step": 6041
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5902853643738514,
+      "learning_rate": 5.653734246848563e-06,
+      "loss": 0.5036,
+      "step": 6042
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.108414795695215,
+      "learning_rate": 5.652473346210203e-06,
+      "loss": 0.4482,
+      "step": 6043
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4608609173084268,
+      "learning_rate": 5.651212403357672e-06,
+      "loss": 0.477,
+      "step": 6044
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.2100752636639958,
+      "learning_rate": 5.649951418372549e-06,
+      "loss": 0.4747,
+      "step": 6045
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.9105591251298744,
+      "learning_rate": 5.648690391336421e-06,
+      "loss": 0.4484,
+      "step": 6046
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 3.0697762911744095,
+      "learning_rate": 5.647429322330872e-06,
+      "loss": 0.4672,
+      "step": 6047
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5719096662073186,
+      "learning_rate": 5.646168211437494e-06,
+      "loss": 0.5084,
+      "step": 6048
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9114564898503823,
+      "learning_rate": 5.644907058737875e-06,
+      "loss": 0.4351,
+      "step": 6049
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9712835952829533,
+      "learning_rate": 5.643645864313616e-06,
+      "loss": 0.4727,
+      "step": 6050
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8842755280955312,
+      "learning_rate": 5.6423846282463105e-06,
+      "loss": 0.4952,
+      "step": 6051
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.025008643029227,
+      "learning_rate": 5.64112335061756e-06,
+      "loss": 0.4786,
+      "step": 6052
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7560700105767266,
+      "learning_rate": 5.639862031508967e-06,
+      "loss": 0.499,
+      "step": 6053
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.467559970839058,
+      "learning_rate": 5.638600671002139e-06,
+      "loss": 0.4693,
+      "step": 6054
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.932307086996423,
+      "learning_rate": 5.637339269178682e-06,
+      "loss": 0.4792,
+      "step": 6055
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1177779985614262,
+      "learning_rate": 5.6360778261202075e-06,
+      "loss": 0.4893,
+      "step": 6056
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2156268966375663,
+      "learning_rate": 5.6348163419083316e-06,
+      "loss": 0.48,
+      "step": 6057
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5994514785412715,
+      "learning_rate": 5.633554816624667e-06,
+      "loss": 0.5243,
+      "step": 6058
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9005100426541817,
+      "learning_rate": 5.632293250350835e-06,
+      "loss": 0.4703,
+      "step": 6059
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5890460914355664,
+      "learning_rate": 5.631031643168457e-06,
+      "loss": 0.4993,
+      "step": 6060
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7838201035293735,
+      "learning_rate": 5.629769995159157e-06,
+      "loss": 0.4472,
+      "step": 6061
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.437953889305717,
+      "learning_rate": 5.628508306404563e-06,
+      "loss": 0.4935,
+      "step": 6062
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5280163401419353,
+      "learning_rate": 5.627246576986303e-06,
+      "loss": 0.4689,
+      "step": 6063
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6750418086010357,
+      "learning_rate": 5.625984806986009e-06,
+      "loss": 0.4483,
+      "step": 6064
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7533496898342817,
+      "learning_rate": 5.624722996485319e-06,
+      "loss": 0.4326,
+      "step": 6065
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.065431855140561,
+      "learning_rate": 5.623461145565866e-06,
+      "loss": 0.4579,
+      "step": 6066
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1974513610954634,
+      "learning_rate": 5.622199254309295e-06,
+      "loss": 0.4415,
+      "step": 6067
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6402534559665537,
+      "learning_rate": 5.620937322797244e-06,
+      "loss": 0.5222,
+      "step": 6068
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.124427374116349,
+      "learning_rate": 5.619675351111363e-06,
+      "loss": 0.5299,
+      "step": 6069
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8035771347864251,
+      "learning_rate": 5.618413339333295e-06,
+      "loss": 0.5137,
+      "step": 6070
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7843059960627126,
+      "learning_rate": 5.617151287544694e-06,
+      "loss": 0.5129,
+      "step": 6071
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5024451675296469,
+      "learning_rate": 5.615889195827211e-06,
+      "loss": 0.4323,
+      "step": 6072
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5713383423867264,
+      "learning_rate": 5.614627064262504e-06,
+      "loss": 0.4951,
+      "step": 6073
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7226451461293641,
+      "learning_rate": 5.61336489293223e-06,
+      "loss": 0.4621,
+      "step": 6074
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8740923717743874,
+      "learning_rate": 5.612102681918049e-06,
+      "loss": 0.4691,
+      "step": 6075
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.053564382222536,
+      "learning_rate": 5.610840431301625e-06,
+      "loss": 0.4786,
+      "step": 6076
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8827955650804251,
+      "learning_rate": 5.609578141164627e-06,
+      "loss": 0.5018,
+      "step": 6077
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5286405292282421,
+      "learning_rate": 5.608315811588717e-06,
+      "loss": 0.4475,
+      "step": 6078
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.57264488110248,
+      "learning_rate": 5.6070534426555725e-06,
+      "loss": 0.5025,
+      "step": 6079
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3788487483075431,
+      "learning_rate": 5.6057910344468625e-06,
+      "loss": 0.4813,
+      "step": 6080
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6957515654254192,
+      "learning_rate": 5.604528587044266e-06,
+      "loss": 0.4587,
+      "step": 6081
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.813128581815455,
+      "learning_rate": 5.60326610052946e-06,
+      "loss": 0.5202,
+      "step": 6082
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1432462517915516,
+      "learning_rate": 5.6020035749841264e-06,
+      "loss": 0.4449,
+      "step": 6083
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5484477788718938,
+      "learning_rate": 5.600741010489949e-06,
+      "loss": 0.4968,
+      "step": 6084
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.259686135117147,
+      "learning_rate": 5.599478407128616e-06,
+      "loss": 0.5136,
+      "step": 6085
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7588980481346306,
+      "learning_rate": 5.598215764981811e-06,
+      "loss": 0.553,
+      "step": 6086
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8376235175454934,
+      "learning_rate": 5.596953084131231e-06,
+      "loss": 0.472,
+      "step": 6087
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.05214934776534,
+      "learning_rate": 5.595690364658567e-06,
+      "loss": 0.4933,
+      "step": 6088
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6898989348272901,
+      "learning_rate": 5.594427606645516e-06,
+      "loss": 0.5188,
+      "step": 6089
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7363819043376463,
+      "learning_rate": 5.593164810173776e-06,
+      "loss": 0.5444,
+      "step": 6090
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.309081460060708,
+      "learning_rate": 5.591901975325048e-06,
+      "loss": 0.5077,
+      "step": 6091
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.005564923590532,
+      "learning_rate": 5.590639102181038e-06,
+      "loss": 0.4982,
+      "step": 6092
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.088026813605811,
+      "learning_rate": 5.589376190823449e-06,
+      "loss": 0.4466,
+      "step": 6093
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.685541596057917,
+      "learning_rate": 5.5881132413339935e-06,
+      "loss": 0.4526,
+      "step": 6094
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3696106934431707,
+      "learning_rate": 5.586850253794379e-06,
+      "loss": 0.5063,
+      "step": 6095
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5409494168011588,
+      "learning_rate": 5.585587228286321e-06,
+      "loss": 0.5048,
+      "step": 6096
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.535950936429176,
+      "learning_rate": 5.584324164891536e-06,
+      "loss": 0.4939,
+      "step": 6097
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5129600486235135,
+      "learning_rate": 5.583061063691741e-06,
+      "loss": 0.4819,
+      "step": 6098
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8251593039689615,
+      "learning_rate": 5.581797924768658e-06,
+      "loss": 0.5092,
+      "step": 6099
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8343440800905695,
+      "learning_rate": 5.580534748204012e-06,
+      "loss": 0.5272,
+      "step": 6100
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7294355741873155,
+      "learning_rate": 5.579271534079526e-06,
+      "loss": 0.4496,
+      "step": 6101
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.571433664574153,
+      "learning_rate": 5.57800828247693e-06,
+      "loss": 0.5219,
+      "step": 6102
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.496009577053304,
+      "learning_rate": 5.576744993477953e-06,
+      "loss": 0.5009,
+      "step": 6103
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.563879222886923,
+      "learning_rate": 5.575481667164331e-06,
+      "loss": 0.5311,
+      "step": 6104
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.743211654560536,
+      "learning_rate": 5.574218303617797e-06,
+      "loss": 0.4215,
+      "step": 6105
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7257727709458222,
+      "learning_rate": 5.5729549029200895e-06,
+      "loss": 0.4484,
+      "step": 6106
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5658341091370458,
+      "learning_rate": 5.57169146515295e-06,
+      "loss": 0.46,
+      "step": 6107
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6083262941974528,
+      "learning_rate": 5.570427990398121e-06,
+      "loss": 0.4823,
+      "step": 6108
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9879218907773908,
+      "learning_rate": 5.569164478737346e-06,
+      "loss": 0.4875,
+      "step": 6109
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.792675009023016,
+      "learning_rate": 5.567900930252375e-06,
+      "loss": 0.4501,
+      "step": 6110
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.175247302290029,
+      "learning_rate": 5.566637345024956e-06,
+      "loss": 0.5009,
+      "step": 6111
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6160948736418187,
+      "learning_rate": 5.565373723136843e-06,
+      "loss": 0.4995,
+      "step": 6112
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.50250538485351,
+      "learning_rate": 5.564110064669789e-06,
+      "loss": 0.4394,
+      "step": 6113
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9328263524409663,
+      "learning_rate": 5.562846369705552e-06,
+      "loss": 0.4723,
+      "step": 6114
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7346407448550933,
+      "learning_rate": 5.56158263832589e-06,
+      "loss": 0.4732,
+      "step": 6115
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6507434359242423,
+      "learning_rate": 5.5603188706125674e-06,
+      "loss": 0.4427,
+      "step": 6116
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0541310652783973,
+      "learning_rate": 5.559055066647345e-06,
+      "loss": 0.4928,
+      "step": 6117
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5838670408765896,
+      "learning_rate": 5.557791226511994e-06,
+      "loss": 0.4855,
+      "step": 6118
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7481184419553404,
+      "learning_rate": 5.556527350288278e-06,
+      "loss": 0.4409,
+      "step": 6119
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9045567514092938,
+      "learning_rate": 5.5552634380579715e-06,
+      "loss": 0.4393,
+      "step": 6120
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3596702709309434,
+      "learning_rate": 5.553999489902846e-06,
+      "loss": 0.4201,
+      "step": 6121
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4976452060296859,
+      "learning_rate": 5.552735505904679e-06,
+      "loss": 0.4245,
+      "step": 6122
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5989443701499597,
+      "learning_rate": 5.551471486145248e-06,
+      "loss": 0.3932,
+      "step": 6123
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.225451591659345,
+      "learning_rate": 5.550207430706334e-06,
+      "loss": 0.4847,
+      "step": 6124
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.479476377488131,
+      "learning_rate": 5.548943339669718e-06,
+      "loss": 0.461,
+      "step": 6125
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3550681695994307,
+      "learning_rate": 5.5476792131171855e-06,
+      "loss": 0.4742,
+      "step": 6126
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.1845240627455693,
+      "learning_rate": 5.546415051130525e-06,
+      "loss": 0.4939,
+      "step": 6127
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5775564432611615,
+      "learning_rate": 5.545150853791525e-06,
+      "loss": 0.4638,
+      "step": 6128
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.064816875075952,
+      "learning_rate": 5.5438866211819775e-06,
+      "loss": 0.4562,
+      "step": 6129
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.9714440953357892,
+      "learning_rate": 5.542622353383679e-06,
+      "loss": 0.4995,
+      "step": 6130
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 5.793594601267657,
+      "learning_rate": 5.541358050478423e-06,
+      "loss": 0.4376,
+      "step": 6131
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6265429327945393,
+      "learning_rate": 5.540093712548009e-06,
+      "loss": 0.4362,
+      "step": 6132
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.795494843978201,
+      "learning_rate": 5.538829339674239e-06,
+      "loss": 0.4356,
+      "step": 6133
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0128315784153163,
+      "learning_rate": 5.537564931938915e-06,
+      "loss": 0.4721,
+      "step": 6134
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3391124824133613,
+      "learning_rate": 5.536300489423844e-06,
+      "loss": 0.4653,
+      "step": 6135
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.3872751057068453,
+      "learning_rate": 5.535036012210832e-06,
+      "loss": 0.4421,
+      "step": 6136
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.133309984524502,
+      "learning_rate": 5.533771500381691e-06,
+      "loss": 0.4734,
+      "step": 6137
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5810908249839427,
+      "learning_rate": 5.532506954018229e-06,
+      "loss": 0.5068,
+      "step": 6138
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3878516178800802,
+      "learning_rate": 5.531242373202268e-06,
+      "loss": 0.4205,
+      "step": 6139
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.359285839597133,
+      "learning_rate": 5.529977758015616e-06,
+      "loss": 0.408,
+      "step": 6140
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.005221373044829,
+      "learning_rate": 5.5287131085401e-06,
+      "loss": 0.4043,
+      "step": 6141
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 6.241111755241205,
+      "learning_rate": 5.527448424857536e-06,
+      "loss": 0.4846,
+      "step": 6142
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 25.881776510110253,
+      "learning_rate": 5.5261837070497505e-06,
+      "loss": 0.4755,
+      "step": 6143
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7995752265012297,
+      "learning_rate": 5.5249189551985645e-06,
+      "loss": 0.4632,
+      "step": 6144
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5784892367522048,
+      "learning_rate": 5.523654169385813e-06,
+      "loss": 0.4973,
+      "step": 6145
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5810176195797127,
+      "learning_rate": 5.522389349693318e-06,
+      "loss": 0.4935,
+      "step": 6146
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.443109551053779,
+      "learning_rate": 5.521124496202917e-06,
+      "loss": 0.4514,
+      "step": 6147
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7210014668254479,
+      "learning_rate": 5.519859608996443e-06,
+      "loss": 0.4443,
+      "step": 6148
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.620076114313818,
+      "learning_rate": 5.5185946881557314e-06,
+      "loss": 0.4657,
+      "step": 6149
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4573540045081261,
+      "learning_rate": 5.517329733762622e-06,
+      "loss": 0.4906,
+      "step": 6150
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.3256329187963027,
+      "learning_rate": 5.516064745898956e-06,
+      "loss": 0.4589,
+      "step": 6151
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5863399679610443,
+      "learning_rate": 5.514799724646575e-06,
+      "loss": 0.4983,
+      "step": 6152
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.7494621658080103,
+      "learning_rate": 5.513534670087327e-06,
+      "loss": 0.5322,
+      "step": 6153
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2467196423066977,
+      "learning_rate": 5.512269582303057e-06,
+      "loss": 0.5135,
+      "step": 6154
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7220730653502976,
+      "learning_rate": 5.511004461375615e-06,
+      "loss": 0.4497,
+      "step": 6155
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8254372869025794,
+      "learning_rate": 5.509739307386853e-06,
+      "loss": 0.5073,
+      "step": 6156
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6854797003354796,
+      "learning_rate": 5.508474120418626e-06,
+      "loss": 0.4431,
+      "step": 6157
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.361320636776241,
+      "learning_rate": 5.5072089005527865e-06,
+      "loss": 0.4793,
+      "step": 6158
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.751052335608454,
+      "learning_rate": 5.505943647871195e-06,
+      "loss": 0.4291,
+      "step": 6159
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.684360599977664,
+      "learning_rate": 5.5046783624557106e-06,
+      "loss": 0.4584,
+      "step": 6160
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4034955872531034,
+      "learning_rate": 5.503413044388199e-06,
+      "loss": 0.4647,
+      "step": 6161
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4452556883215584,
+      "learning_rate": 5.502147693750518e-06,
+      "loss": 0.4727,
+      "step": 6162
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.7287708993354856,
+      "learning_rate": 5.500882310624542e-06,
+      "loss": 0.4422,
+      "step": 6163
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6317900914395849,
+      "learning_rate": 5.499616895092133e-06,
+      "loss": 0.4519,
+      "step": 6164
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 4.086964907026692,
+      "learning_rate": 5.498351447235166e-06,
+      "loss": 0.4761,
+      "step": 6165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.9960703282046515,
+      "learning_rate": 5.497085967135512e-06,
+      "loss": 0.41,
+      "step": 6166
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.6876454916225092,
+      "learning_rate": 5.495820454875047e-06,
+      "loss": 0.4482,
+      "step": 6167
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4040479888130415,
+      "learning_rate": 5.494554910535646e-06,
+      "loss": 0.4601,
+      "step": 6168
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8704724298215512,
+      "learning_rate": 5.493289334199189e-06,
+      "loss": 0.4651,
+      "step": 6169
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5200863389315538,
+      "learning_rate": 5.4920237259475585e-06,
+      "loss": 0.4362,
+      "step": 6170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.4686944963393813,
+      "learning_rate": 5.490758085862636e-06,
+      "loss": 0.4859,
+      "step": 6171
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 3.1915766630265945,
+      "learning_rate": 5.4894924140263075e-06,
+      "loss": 0.4972,
+      "step": 6172
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.8589595820590077,
+      "learning_rate": 5.488226710520459e-06,
+      "loss": 0.468,
+      "step": 6173
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.544235413154769,
+      "learning_rate": 5.486960975426981e-06,
+      "loss": 0.4493,
+      "step": 6174
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.218455036295482,
+      "learning_rate": 5.485695208827766e-06,
+      "loss": 0.4671,
+      "step": 6175
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5846464270519955,
+      "learning_rate": 5.484429410804707e-06,
+      "loss": 0.4924,
+      "step": 6176
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5468635257090844,
+      "learning_rate": 5.4831635814396976e-06,
+      "loss": 0.4952,
+      "step": 6177
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6223208724422544,
+      "learning_rate": 5.4818977208146375e-06,
+      "loss": 0.4116,
+      "step": 6178
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4215539529799437,
+      "learning_rate": 5.480631829011425e-06,
+      "loss": 0.4358,
+      "step": 6179
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9133772126380435,
+      "learning_rate": 5.479365906111962e-06,
+      "loss": 0.4684,
+      "step": 6180
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5592268299390609,
+      "learning_rate": 5.4780999521981514e-06,
+      "loss": 0.4923,
+      "step": 6181
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5636515448443473,
+      "learning_rate": 5.4768339673519e-06,
+      "loss": 0.5101,
+      "step": 6182
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4302242402018415,
+      "learning_rate": 5.475567951655114e-06,
+      "loss": 0.4302,
+      "step": 6183
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8958875074886061,
+      "learning_rate": 5.474301905189705e-06,
+      "loss": 0.5386,
+      "step": 6184
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.433617858864352,
+      "learning_rate": 5.47303582803758e-06,
+      "loss": 0.4532,
+      "step": 6185
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.503411939861271,
+      "learning_rate": 5.4717697202806586e-06,
+      "loss": 0.4321,
+      "step": 6186
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7729473749048317,
+      "learning_rate": 5.470503582000852e-06,
+      "loss": 0.4557,
+      "step": 6187
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7918641458400406,
+      "learning_rate": 5.469237413280081e-06,
+      "loss": 0.4438,
+      "step": 6188
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.210129788868613,
+      "learning_rate": 5.4679712142002614e-06,
+      "loss": 0.5327,
+      "step": 6189
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.47759080792351,
+      "learning_rate": 5.466704984843317e-06,
+      "loss": 0.4231,
+      "step": 6190
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.219900145400828,
+      "learning_rate": 5.465438725291171e-06,
+      "loss": 0.4879,
+      "step": 6191
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2502232064724494,
+      "learning_rate": 5.464172435625748e-06,
+      "loss": 0.4084,
+      "step": 6192
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.00333794815659,
+      "learning_rate": 5.462906115928977e-06,
+      "loss": 0.5014,
+      "step": 6193
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.701012434027411,
+      "learning_rate": 5.461639766282784e-06,
+      "loss": 0.4549,
+      "step": 6194
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.461459675759506,
+      "learning_rate": 5.460373386769103e-06,
+      "loss": 0.3925,
+      "step": 6195
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.644938329626762,
+      "learning_rate": 5.459106977469865e-06,
+      "loss": 0.4268,
+      "step": 6196
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6098509632255238,
+      "learning_rate": 5.457840538467006e-06,
+      "loss": 0.5228,
+      "step": 6197
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7867680854779369,
+      "learning_rate": 5.456574069842464e-06,
+      "loss": 0.4835,
+      "step": 6198
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.89454600874104,
+      "learning_rate": 5.455307571678177e-06,
+      "loss": 0.4513,
+      "step": 6199
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7775040629700147,
+      "learning_rate": 5.454041044056086e-06,
+      "loss": 0.496,
+      "step": 6200
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.407584426142428,
+      "learning_rate": 5.452774487058132e-06,
+      "loss": 0.4745,
+      "step": 6201
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.688178599007344,
+      "learning_rate": 5.451507900766263e-06,
+      "loss": 0.433,
+      "step": 6202
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.501646445047139,
+      "learning_rate": 5.450241285262422e-06,
+      "loss": 0.4889,
+      "step": 6203
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.995484681314486,
+      "learning_rate": 5.44897464062856e-06,
+      "loss": 0.4942,
+      "step": 6204
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6328232711586017,
+      "learning_rate": 5.447707966946625e-06,
+      "loss": 0.4982,
+      "step": 6205
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4715274991820533,
+      "learning_rate": 5.44644126429857e-06,
+      "loss": 0.4554,
+      "step": 6206
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6494235873816463,
+      "learning_rate": 5.445174532766351e-06,
+      "loss": 0.516,
+      "step": 6207
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7747413492087127,
+      "learning_rate": 5.44390777243192e-06,
+      "loss": 0.5239,
+      "step": 6208
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7380109969520612,
+      "learning_rate": 5.4426409833772375e-06,
+      "loss": 0.4731,
+      "step": 6209
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7542706140547215,
+      "learning_rate": 5.441374165684262e-06,
+      "loss": 0.4313,
+      "step": 6210
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3785219438679248,
+      "learning_rate": 5.440107319434956e-06,
+      "loss": 0.4372,
+      "step": 6211
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.592475919971585,
+      "learning_rate": 5.438840444711282e-06,
+      "loss": 0.4498,
+      "step": 6212
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6069665124641069,
+      "learning_rate": 5.4375735415952065e-06,
+      "loss": 0.5028,
+      "step": 6213
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5355741685479622,
+      "learning_rate": 5.436306610168693e-06,
+      "loss": 0.4859,
+      "step": 6214
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4853113531320452,
+      "learning_rate": 5.4350396505137146e-06,
+      "loss": 0.4278,
+      "step": 6215
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1841063441180957,
+      "learning_rate": 5.43377266271224e-06,
+      "loss": 0.4699,
+      "step": 6216
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.795802384124965,
+      "learning_rate": 5.432505646846241e-06,
+      "loss": 0.4213,
+      "step": 6217
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.500778920689842,
+      "learning_rate": 5.431238602997692e-06,
+      "loss": 0.4968,
+      "step": 6218
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7621174193420872,
+      "learning_rate": 5.429971531248569e-06,
+      "loss": 0.4549,
+      "step": 6219
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7075290169293769,
+      "learning_rate": 5.42870443168085e-06,
+      "loss": 0.4718,
+      "step": 6220
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6382980858710587,
+      "learning_rate": 5.427437304376517e-06,
+      "loss": 0.4944,
+      "step": 6221
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.2399491981815913,
+      "learning_rate": 5.426170149417549e-06,
+      "loss": 0.466,
+      "step": 6222
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.284609725737989,
+      "learning_rate": 5.42490296688593e-06,
+      "loss": 0.4904,
+      "step": 6223
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6676191641008151,
+      "learning_rate": 5.423635756863643e-06,
+      "loss": 0.5118,
+      "step": 6224
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.9806706002057375,
+      "learning_rate": 5.4223685194326795e-06,
+      "loss": 0.4467,
+      "step": 6225
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8270812006639983,
+      "learning_rate": 5.4211012546750244e-06,
+      "loss": 0.4568,
+      "step": 6226
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6956974328598244,
+      "learning_rate": 5.419833962672668e-06,
+      "loss": 0.4927,
+      "step": 6227
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4145221578547051,
+      "learning_rate": 5.418566643507604e-06,
+      "loss": 0.4089,
+      "step": 6228
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7555747064717742,
+      "learning_rate": 5.4172992972618264e-06,
+      "loss": 0.4957,
+      "step": 6229
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6877813514756754,
+      "learning_rate": 5.416031924017327e-06,
+      "loss": 0.4332,
+      "step": 6230
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8889325595611621,
+      "learning_rate": 5.414764523856111e-06,
+      "loss": 0.4602,
+      "step": 6231
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.846671995252927,
+      "learning_rate": 5.41349709686017e-06,
+      "loss": 0.44,
+      "step": 6232
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8721659506671033,
+      "learning_rate": 5.412229643111508e-06,
+      "loss": 0.5248,
+      "step": 6233
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 9.040787732532449,
+      "learning_rate": 5.4109621626921285e-06,
+      "loss": 0.4562,
+      "step": 6234
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6468062117615846,
+      "learning_rate": 5.409694655684035e-06,
+      "loss": 0.4919,
+      "step": 6235
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6358340320674607,
+      "learning_rate": 5.408427122169231e-06,
+      "loss": 0.4942,
+      "step": 6236
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6479255781007034,
+      "learning_rate": 5.407159562229729e-06,
+      "loss": 0.516,
+      "step": 6237
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5313420568833969,
+      "learning_rate": 5.405891975947534e-06,
+      "loss": 0.5059,
+      "step": 6238
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7029536844106739,
+      "learning_rate": 5.404624363404659e-06,
+      "loss": 0.4715,
+      "step": 6239
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1127584456880144,
+      "learning_rate": 5.4033567246831185e-06,
+      "loss": 0.5134,
+      "step": 6240
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8955819473260356,
+      "learning_rate": 5.4020890598649245e-06,
+      "loss": 0.416,
+      "step": 6241
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9736188166879942,
+      "learning_rate": 5.400821369032094e-06,
+      "loss": 0.4535,
+      "step": 6242
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8228085773763956,
+      "learning_rate": 5.399553652266647e-06,
+      "loss": 0.476,
+      "step": 6243
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7750162324333085,
+      "learning_rate": 5.3982859096506e-06,
+      "loss": 0.4963,
+      "step": 6244
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3104342699945752,
+      "learning_rate": 5.397018141265975e-06,
+      "loss": 0.4923,
+      "step": 6245
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6243127274388197,
+      "learning_rate": 5.395750347194798e-06,
+      "loss": 0.5274,
+      "step": 6246
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7970697285433763,
+      "learning_rate": 5.394482527519089e-06,
+      "loss": 0.4669,
+      "step": 6247
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9866467236234258,
+      "learning_rate": 5.393214682320879e-06,
+      "loss": 0.5113,
+      "step": 6248
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.0921703989404383,
+      "learning_rate": 5.391946811682191e-06,
+      "loss": 0.4695,
+      "step": 6249
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4994044141998235,
+      "learning_rate": 5.390678915685058e-06,
+      "loss": 0.4072,
+      "step": 6250
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.045303549511949,
+      "learning_rate": 5.38941099441151e-06,
+      "loss": 0.4973,
+      "step": 6251
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5681921831982807,
+      "learning_rate": 5.38814304794358e-06,
+      "loss": 0.5007,
+      "step": 6252
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5916835522620377,
+      "learning_rate": 5.386875076363302e-06,
+      "loss": 0.5165,
+      "step": 6253
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4773206347372387,
+      "learning_rate": 5.385607079752714e-06,
+      "loss": 0.4473,
+      "step": 6254
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9697999876331795,
+      "learning_rate": 5.384339058193851e-06,
+      "loss": 0.5019,
+      "step": 6255
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5456207097539333,
+      "learning_rate": 5.383071011768755e-06,
+      "loss": 0.5007,
+      "step": 6256
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 5.635064397127404,
+      "learning_rate": 5.381802940559465e-06,
+      "loss": 0.4462,
+      "step": 6257
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5850970705855765,
+      "learning_rate": 5.380534844648025e-06,
+      "loss": 0.4963,
+      "step": 6258
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5522828454025475,
+      "learning_rate": 5.3792667241164775e-06,
+      "loss": 0.5121,
+      "step": 6259
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6494216645929471,
+      "learning_rate": 5.377998579046871e-06,
+      "loss": 0.4607,
+      "step": 6260
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4914620541664583,
+      "learning_rate": 5.376730409521248e-06,
+      "loss": 0.376,
+      "step": 6261
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4667366376594306,
+      "learning_rate": 5.3754622156216615e-06,
+      "loss": 0.4404,
+      "step": 6262
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6930042461225863,
+      "learning_rate": 5.37419399743016e-06,
+      "loss": 0.5298,
+      "step": 6263
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.64698979429338,
+      "learning_rate": 5.372925755028798e-06,
+      "loss": 0.482,
+      "step": 6264
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5874370874880012,
+      "learning_rate": 5.3716574884996255e-06,
+      "loss": 0.5025,
+      "step": 6265
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5929176049622847,
+      "learning_rate": 5.370389197924702e-06,
+      "loss": 0.5022,
+      "step": 6266
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5533408231108035,
+      "learning_rate": 5.36912088338608e-06,
+      "loss": 0.4987,
+      "step": 6267
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7710911789970258,
+      "learning_rate": 5.367852544965821e-06,
+      "loss": 0.5247,
+      "step": 6268
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.613616520710951,
+      "learning_rate": 5.366584182745983e-06,
+      "loss": 0.5103,
+      "step": 6269
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.1463739703508047,
+      "learning_rate": 5.365315796808629e-06,
+      "loss": 0.4596,
+      "step": 6270
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7465574581119085,
+      "learning_rate": 5.364047387235819e-06,
+      "loss": 0.4185,
+      "step": 6271
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8401806458580008,
+      "learning_rate": 5.362778954109621e-06,
+      "loss": 0.4429,
+      "step": 6272
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.4811042826493708,
+      "learning_rate": 5.361510497512098e-06,
+      "loss": 0.4688,
+      "step": 6273
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8947855049732982,
+      "learning_rate": 5.360242017525319e-06,
+      "loss": 0.4724,
+      "step": 6274
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.7403705638165767,
+      "learning_rate": 5.358973514231351e-06,
+      "loss": 0.4532,
+      "step": 6275
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.947961443333871,
+      "learning_rate": 5.35770498771227e-06,
+      "loss": 0.4872,
+      "step": 6276
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5230329892488463,
+      "learning_rate": 5.356436438050141e-06,
+      "loss": 0.4742,
+      "step": 6277
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9063813932018792,
+      "learning_rate": 5.3551678653270434e-06,
+      "loss": 0.4883,
+      "step": 6278
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.6972448311097112,
+      "learning_rate": 5.353899269625047e-06,
+      "loss": 0.5001,
+      "step": 6279
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.686545163769459,
+      "learning_rate": 5.352630651026232e-06,
+      "loss": 0.4903,
+      "step": 6280
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.1578877966916434,
+      "learning_rate": 5.3513620096126766e-06,
+      "loss": 0.4307,
+      "step": 6281
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5824247503932578,
+      "learning_rate": 5.350093345466456e-06,
+      "loss": 0.4784,
+      "step": 6282
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6204861532095239,
+      "learning_rate": 5.348824658669656e-06,
+      "loss": 0.4858,
+      "step": 6283
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.3565999817389724,
+      "learning_rate": 5.347555949304356e-06,
+      "loss": 0.5097,
+      "step": 6284
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4984482317678507,
+      "learning_rate": 5.346287217452641e-06,
+      "loss": 0.4745,
+      "step": 6285
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.784131871401627,
+      "learning_rate": 5.345018463196597e-06,
+      "loss": 0.4749,
+      "step": 6286
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.06381833745769,
+      "learning_rate": 5.343749686618307e-06,
+      "loss": 0.4967,
+      "step": 6287
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.6244322576247214,
+      "learning_rate": 5.342480887799863e-06,
+      "loss": 0.4476,
+      "step": 6288
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5913931340671637,
+      "learning_rate": 5.341212066823356e-06,
+      "loss": 0.5262,
+      "step": 6289
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8231656938250962,
+      "learning_rate": 5.339943223770871e-06,
+      "loss": 0.4963,
+      "step": 6290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.609619946583199,
+      "learning_rate": 5.338674358724507e-06,
+      "loss": 0.4419,
+      "step": 6291
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5044160385517398,
+      "learning_rate": 5.337405471766355e-06,
+      "loss": 0.4284,
+      "step": 6292
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 18.95601643434386,
+      "learning_rate": 5.336136562978509e-06,
+      "loss": 0.4412,
+      "step": 6293
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.7291491239111116,
+      "learning_rate": 5.334867632443067e-06,
+      "loss": 0.5082,
+      "step": 6294
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.9305790703875354,
+      "learning_rate": 5.333598680242129e-06,
+      "loss": 0.5272,
+      "step": 6295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 3.3656206473115815,
+      "learning_rate": 5.3323297064577905e-06,
+      "loss": 0.4488,
+      "step": 6296
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.782919079767497,
+      "learning_rate": 5.331060711172157e-06,
+      "loss": 0.5074,
+      "step": 6297
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.730326236442161,
+      "learning_rate": 5.329791694467326e-06,
+      "loss": 0.4892,
+      "step": 6298
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 2.186419338027681,
+      "learning_rate": 5.328522656425405e-06,
+      "loss": 0.4732,
+      "step": 6299
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6223029868635872,
+      "learning_rate": 5.327253597128497e-06,
+      "loss": 0.4817,
+      "step": 6300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5917530158958798,
+      "learning_rate": 5.325984516658712e-06,
+      "loss": 0.4996,
+      "step": 6301
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3771180427556626,
+      "learning_rate": 5.324715415098154e-06,
+      "loss": 0.3998,
+      "step": 6302
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6652210903019669,
+      "learning_rate": 5.3234462925289335e-06,
+      "loss": 0.4966,
+      "step": 6303
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 4.173669520105727,
+      "learning_rate": 5.32217714903316e-06,
+      "loss": 0.4644,
+      "step": 6304
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3703608074109015,
+      "learning_rate": 5.320907984692948e-06,
+      "loss": 0.4739,
+      "step": 6305
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8522374206600283,
+      "learning_rate": 5.319638799590408e-06,
+      "loss": 0.5635,
+      "step": 6306
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4912772197417128,
+      "learning_rate": 5.318369593807655e-06,
+      "loss": 0.4539,
+      "step": 6307
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8490052938874524,
+      "learning_rate": 5.317100367426808e-06,
+      "loss": 0.4831,
+      "step": 6308
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6436177452084497,
+      "learning_rate": 5.3158311205299794e-06,
+      "loss": 0.4687,
+      "step": 6309
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5753425729614179,
+      "learning_rate": 5.314561853199292e-06,
+      "loss": 0.5186,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0235181584990505,
+      "learning_rate": 5.313292565516864e-06,
+      "loss": 0.4558,
+      "step": 6311
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 8.597075337151079,
+      "learning_rate": 5.312023257564817e-06,
+      "loss": 0.4185,
+      "step": 6312
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9328857002588347,
+      "learning_rate": 5.310753929425273e-06,
+      "loss": 0.4982,
+      "step": 6313
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4816405831123,
+      "learning_rate": 5.309484581180357e-06,
+      "loss": 0.4364,
+      "step": 6314
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7183106406122068,
+      "learning_rate": 5.3082152129121935e-06,
+      "loss": 0.4774,
+      "step": 6315
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.3234240573791825,
+      "learning_rate": 5.306945824702908e-06,
+      "loss": 0.4222,
+      "step": 6316
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4834144575483674,
+      "learning_rate": 5.305676416634628e-06,
+      "loss": 0.4996,
+      "step": 6317
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6884736366387614,
+      "learning_rate": 5.304406988789486e-06,
+      "loss": 0.4818,
+      "step": 6318
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.775642170972731,
+      "learning_rate": 5.30313754124961e-06,
+      "loss": 0.4388,
+      "step": 6319
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6149749825117283,
+      "learning_rate": 5.301868074097132e-06,
+      "loss": 0.4242,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.4294468223332797,
+      "learning_rate": 5.300598587414183e-06,
+      "loss": 0.5201,
+      "step": 6321
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.041359915283597,
+      "learning_rate": 5.299329081282898e-06,
+      "loss": 0.5076,
+      "step": 6322
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.247802370545581,
+      "learning_rate": 5.298059555785414e-06,
+      "loss": 0.4571,
+      "step": 6323
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9882339879234205,
+      "learning_rate": 5.296790011003867e-06,
+      "loss": 0.4774,
+      "step": 6324
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8519505353548662,
+      "learning_rate": 5.2955204470203945e-06,
+      "loss": 0.4039,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9131839228307106,
+      "learning_rate": 5.294250863917136e-06,
+      "loss": 0.5026,
+      "step": 6326
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 5.1052343546269014,
+      "learning_rate": 5.292981261776231e-06,
+      "loss": 0.4592,
+      "step": 6327
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6043179629388822,
+      "learning_rate": 5.291711640679822e-06,
+      "loss": 0.5156,
+      "step": 6328
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7533078625257088,
+      "learning_rate": 5.290442000710051e-06,
+      "loss": 0.4837,
+      "step": 6329
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8504278274861121,
+      "learning_rate": 5.289172341949063e-06,
+      "loss": 0.4767,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.510640757751848,
+      "learning_rate": 5.287902664479002e-06,
+      "loss": 0.4312,
+      "step": 6331
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5991450427357143,
+      "learning_rate": 5.286632968382015e-06,
+      "loss": 0.4717,
+      "step": 6332
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9920646198379657,
+      "learning_rate": 5.28536325374025e-06,
+      "loss": 0.4627,
+      "step": 6333
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6081928394626317,
+      "learning_rate": 5.284093520635857e-06,
+      "loss": 0.484,
+      "step": 6334
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6056460923747577,
+      "learning_rate": 5.282823769150984e-06,
+      "loss": 0.4704,
+      "step": 6335
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8737402831553114,
+      "learning_rate": 5.2815539993677835e-06,
+      "loss": 0.5162,
+      "step": 6336
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7177839424240293,
+      "learning_rate": 5.280284211368407e-06,
+      "loss": 0.4723,
+      "step": 6337
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6226966745235356,
+      "learning_rate": 5.27901440523501e-06,
+      "loss": 0.4923,
+      "step": 6338
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6667447088802565,
+      "learning_rate": 5.277744581049744e-06,
+      "loss": 0.462,
+      "step": 6339
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6313250942295954,
+      "learning_rate": 5.276474738894769e-06,
+      "loss": 0.4908,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3304742436623336,
+      "learning_rate": 5.275204878852239e-06,
+      "loss": 0.4416,
+      "step": 6341
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.6251352719445262,
+      "learning_rate": 5.273935001004313e-06,
+      "loss": 0.4498,
+      "step": 6342
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.71600589789104,
+      "learning_rate": 5.27266510543315e-06,
+      "loss": 0.4649,
+      "step": 6343
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7779323429192808,
+      "learning_rate": 5.271395192220914e-06,
+      "loss": 0.4627,
+      "step": 6344
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3096970466278597,
+      "learning_rate": 5.270125261449761e-06,
+      "loss": 0.4637,
+      "step": 6345
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5935973937058499,
+      "learning_rate": 5.2688553132018595e-06,
+      "loss": 0.4259,
+      "step": 6346
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0346619803263954,
+      "learning_rate": 5.26758534755937e-06,
+      "loss": 0.4734,
+      "step": 6347
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8650269279941372,
+      "learning_rate": 5.26631536460446e-06,
+      "loss": 0.4646,
+      "step": 6348
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7484319902745529,
+      "learning_rate": 5.265045364419291e-06,
+      "loss": 0.5258,
+      "step": 6349
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7993804969421627,
+      "learning_rate": 5.2637753470860365e-06,
+      "loss": 0.4392,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.706865561397999,
+      "learning_rate": 5.2625053126868605e-06,
+      "loss": 0.4516,
+      "step": 6351
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9402787315568721,
+      "learning_rate": 5.261235261303935e-06,
+      "loss": 0.4987,
+      "step": 6352
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6446848818671795,
+      "learning_rate": 5.259965193019431e-06,
+      "loss": 0.4395,
+      "step": 6353
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.454533862690701,
+      "learning_rate": 5.258695107915517e-06,
+      "loss": 0.448,
+      "step": 6354
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1472889889396236,
+      "learning_rate": 5.257425006074368e-06,
+      "loss": 0.4815,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5143769595568661,
+      "learning_rate": 5.2561548875781585e-06,
+      "loss": 0.4208,
+      "step": 6356
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5953522912583749,
+      "learning_rate": 5.254884752509063e-06,
+      "loss": 0.5049,
+      "step": 6357
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9924628357185903,
+      "learning_rate": 5.253614600949256e-06,
+      "loss": 0.4991,
+      "step": 6358
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 7.732888148664725,
+      "learning_rate": 5.252344432980918e-06,
+      "loss": 0.4298,
+      "step": 6359
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1156378815388024,
+      "learning_rate": 5.251074248686226e-06,
+      "loss": 0.4139,
+      "step": 6360
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.035207180116052,
+      "learning_rate": 5.249804048147357e-06,
+      "loss": 0.4529,
+      "step": 6361
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1611716300246155,
+      "learning_rate": 5.2485338314464925e-06,
+      "loss": 0.4468,
+      "step": 6362
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3453832028855723,
+      "learning_rate": 5.2472635986658145e-06,
+      "loss": 0.4204,
+      "step": 6363
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4675127214739616,
+      "learning_rate": 5.245993349887506e-06,
+      "loss": 0.4849,
+      "step": 6364
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5614434986602911,
+      "learning_rate": 5.244723085193749e-06,
+      "loss": 0.5009,
+      "step": 6365
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5588169799401679,
+      "learning_rate": 5.243452804666728e-06,
+      "loss": 0.4658,
+      "step": 6366
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.302343974650844,
+      "learning_rate": 5.24218250838863e-06,
+      "loss": 0.4615,
+      "step": 6367
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4949704828194643,
+      "learning_rate": 5.2409121964416394e-06,
+      "loss": 0.4511,
+      "step": 6368
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.489486550335994,
+      "learning_rate": 5.239641868907946e-06,
+      "loss": 0.4105,
+      "step": 6369
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.762040835435942,
+      "learning_rate": 5.2383715258697364e-06,
+      "loss": 0.463,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6765433871233333,
+      "learning_rate": 5.237101167409202e-06,
+      "loss": 0.442,
+      "step": 6371
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7565350949707235,
+      "learning_rate": 5.2358307936085315e-06,
+      "loss": 0.4858,
+      "step": 6372
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.0581589145637462,
+      "learning_rate": 5.234560404549917e-06,
+      "loss": 0.5044,
+      "step": 6373
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.883796101782298,
+      "learning_rate": 5.233290000315552e-06,
+      "loss": 0.4994,
+      "step": 6374
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9070147395372747,
+      "learning_rate": 5.232019580987628e-06,
+      "loss": 0.5017,
+      "step": 6375
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6127254047567787,
+      "learning_rate": 5.230749146648341e-06,
+      "loss": 0.4878,
+      "step": 6376
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5748707675156586,
+      "learning_rate": 5.2294786973798864e-06,
+      "loss": 0.5051,
+      "step": 6377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.904655159414454,
+      "learning_rate": 5.228208233264459e-06,
+      "loss": 0.4592,
+      "step": 6378
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.1707593129342198,
+      "learning_rate": 5.226937754384259e-06,
+      "loss": 0.499,
+      "step": 6379
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5635030740996013,
+      "learning_rate": 5.225667260821482e-06,
+      "loss": 0.4987,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.2442132335278333,
+      "learning_rate": 5.22439675265833e-06,
+      "loss": 0.4792,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5621225451139222,
+      "learning_rate": 5.223126229977e-06,
+      "loss": 0.5142,
+      "step": 6382
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5431950151514104,
+      "learning_rate": 5.221855692859697e-06,
+      "loss": 0.4895,
+      "step": 6383
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5365914402395804,
+      "learning_rate": 5.2205851413886185e-06,
+      "loss": 0.4081,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.569853713583796,
+      "learning_rate": 5.219314575645972e-06,
+      "loss": 0.512,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6362935225760142,
+      "learning_rate": 5.218043995713958e-06,
+      "loss": 0.5122,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.108327464549863,
+      "learning_rate": 5.216773401674781e-06,
+      "loss": 0.4605,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.451207394963656,
+      "learning_rate": 5.21550279361065e-06,
+      "loss": 0.4897,
+      "step": 6388
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 3.5920953375807123,
+      "learning_rate": 5.214232171603772e-06,
+      "loss": 0.4331,
+      "step": 6389
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6131797343925638,
+      "learning_rate": 5.212961535736351e-06,
+      "loss": 0.4819,
+      "step": 6390
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5832889198961604,
+      "learning_rate": 5.2116908860905976e-06,
+      "loss": 0.4897,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7736041899391994,
+      "learning_rate": 5.2104202227487195e-06,
+      "loss": 0.4706,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.698483160247006,
+      "learning_rate": 5.20914954579293e-06,
+      "loss": 0.467,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8420120044989876,
+      "learning_rate": 5.207878855305438e-06,
+      "loss": 0.471,
+      "step": 6394
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5514768605152538,
+      "learning_rate": 5.206608151368457e-06,
+      "loss": 0.4153,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4756262883700841,
+      "learning_rate": 5.205337434064198e-06,
+      "loss": 0.5061,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6130012676722223,
+      "learning_rate": 5.2040667034748745e-06,
+      "loss": 0.5435,
+      "step": 6397
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4755491510645904,
+      "learning_rate": 5.202795959682704e-06,
+      "loss": 0.4611,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5353269227832131,
+      "learning_rate": 5.201525202769899e-06,
+      "loss": 0.4647,
+      "step": 6399
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7137190570570136,
+      "learning_rate": 5.200254432818677e-06,
+      "loss": 0.5004,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5689974389314436,
+      "learning_rate": 5.198983649911256e-06,
+      "loss": 0.4773,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3641785926603178,
+      "learning_rate": 5.197712854129852e-06,
+      "loss": 0.4132,
+      "step": 6402
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.346223958507651,
+      "learning_rate": 5.196442045556685e-06,
+      "loss": 0.4272,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9931405557139343,
+      "learning_rate": 5.1951712242739775e-06,
+      "loss": 0.4454,
+      "step": 6404
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5612073321204236,
+      "learning_rate": 5.193900390363944e-06,
+      "loss": 0.4926,
+      "step": 6405
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7408529228865668,
+      "learning_rate": 5.192629543908811e-06,
+      "loss": 0.4892,
+      "step": 6406
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.539621728364625,
+      "learning_rate": 5.191358684990796e-06,
+      "loss": 0.4655,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4159810140498088,
+      "learning_rate": 5.190087813692127e-06,
+      "loss": 0.5028,
+      "step": 6408
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7191056669152742,
+      "learning_rate": 5.188816930095023e-06,
+      "loss": 0.492,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7449955515947362,
+      "learning_rate": 5.187546034281712e-06,
+      "loss": 0.4602,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.149520888460154,
+      "learning_rate": 5.186275126334417e-06,
+      "loss": 0.4339,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6984038003079411,
+      "learning_rate": 5.1850042063353655e-06,
+      "loss": 0.5049,
+      "step": 6412
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4921197344135262,
+      "learning_rate": 5.183733274366783e-06,
+      "loss": 0.447,
+      "step": 6413
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4397144311628005,
+      "learning_rate": 5.1824623305109e-06,
+      "loss": 0.4438,
+      "step": 6414
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.548265490614242,
+      "learning_rate": 5.181191374849941e-06,
+      "loss": 0.4441,
+      "step": 6415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5431966626586984,
+      "learning_rate": 5.179920407466138e-06,
+      "loss": 0.5002,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6281042938534527,
+      "learning_rate": 5.17864942844172e-06,
+      "loss": 0.5171,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5831227752072574,
+      "learning_rate": 5.177378437858918e-06,
+      "loss": 0.4534,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.974205123163909,
+      "learning_rate": 5.176107435799962e-06,
+      "loss": 0.4441,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3700131601400525,
+      "learning_rate": 5.174836422347087e-06,
+      "loss": 0.4296,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.9962898303726901,
+      "learning_rate": 5.173565397582522e-06,
+      "loss": 0.5228,
+      "step": 6421
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.7938213727438832,
+      "learning_rate": 5.172294361588504e-06,
+      "loss": 0.4559,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.8380930903094914,
+      "learning_rate": 5.171023314447265e-06,
+      "loss": 0.4561,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5332822122625431,
+      "learning_rate": 5.169752256241043e-06,
+      "loss": 0.4389,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.4603053774310712,
+      "learning_rate": 5.1684811870520715e-06,
+      "loss": 0.4866,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.68945279421647,
+      "learning_rate": 5.167210106962588e-06,
+      "loss": 0.4319,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.976411496691542,
+      "learning_rate": 5.1659390160548285e-06,
+      "loss": 0.4824,
+      "step": 6427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5872765562640447,
+      "learning_rate": 5.164667914411031e-06,
+      "loss": 0.501,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5469721580300385,
+      "learning_rate": 5.163396802113438e-06,
+      "loss": 0.5099,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4528028118496923,
+      "learning_rate": 5.162125679244282e-06,
+      "loss": 0.4447,
+      "step": 6430
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7951980591678836,
+      "learning_rate": 5.16085454588581e-06,
+      "loss": 0.4239,
+      "step": 6431
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.33610109693331,
+      "learning_rate": 5.159583402120256e-06,
+      "loss": 0.5184,
+      "step": 6432
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7105316915530604,
+      "learning_rate": 5.158312248029868e-06,
+      "loss": 0.4635,
+      "step": 6433
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8769433935601085,
+      "learning_rate": 5.157041083696882e-06,
+      "loss": 0.5001,
+      "step": 6434
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.618259329617298,
+      "learning_rate": 5.155769909203543e-06,
+      "loss": 0.4937,
+      "step": 6435
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6962936272300568,
+      "learning_rate": 5.154498724632095e-06,
+      "loss": 0.4814,
+      "step": 6436
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4276255167353245,
+      "learning_rate": 5.153227530064784e-06,
+      "loss": 0.4235,
+      "step": 6437
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5446258708579352,
+      "learning_rate": 5.15195632558385e-06,
+      "loss": 0.5327,
+      "step": 6438
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.023823947320032,
+      "learning_rate": 5.150685111271541e-06,
+      "loss": 0.5068,
+      "step": 6439
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.913079326857015,
+      "learning_rate": 5.1494138872101026e-06,
+      "loss": 0.5524,
+      "step": 6440
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.57045341359687,
+      "learning_rate": 5.1481426534817825e-06,
+      "loss": 0.4509,
+      "step": 6441
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.582410928448665,
+      "learning_rate": 5.146871410168825e-06,
+      "loss": 0.5124,
+      "step": 6442
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8114976265871985,
+      "learning_rate": 5.1456001573534795e-06,
+      "loss": 0.4625,
+      "step": 6443
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1948723183567,
+      "learning_rate": 5.1443288951179946e-06,
+      "loss": 0.4562,
+      "step": 6444
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.2525103819699286,
+      "learning_rate": 5.1430576235446185e-06,
+      "loss": 0.4588,
+      "step": 6445
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.413732211424999,
+      "learning_rate": 5.141786342715601e-06,
+      "loss": 0.4456,
+      "step": 6446
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.41581810389455,
+      "learning_rate": 5.140515052713193e-06,
+      "loss": 0.4358,
+      "step": 6447
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.915289319786108,
+      "learning_rate": 5.139243753619645e-06,
+      "loss": 0.4766,
+      "step": 6448
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2422567335429795,
+      "learning_rate": 5.13797244551721e-06,
+      "loss": 0.4613,
+      "step": 6449
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5608426420138577,
+      "learning_rate": 5.136701128488135e-06,
+      "loss": 0.4621,
+      "step": 6450
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6998418080600932,
+      "learning_rate": 5.1354298026146786e-06,
+      "loss": 0.5164,
+      "step": 6451
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 8.612123966141786,
+      "learning_rate": 5.134158467979089e-06,
+      "loss": 0.4875,
+      "step": 6452
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6548387587265756,
+      "learning_rate": 5.132887124663625e-06,
+      "loss": 0.4522,
+      "step": 6453
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.33475632525976,
+      "learning_rate": 5.131615772750534e-06,
+      "loss": 0.4929,
+      "step": 6454
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.900944760148288,
+      "learning_rate": 5.130344412322078e-06,
+      "loss": 0.483,
+      "step": 6455
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5973543610583176,
+      "learning_rate": 5.129073043460506e-06,
+      "loss": 0.4461,
+      "step": 6456
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.8637051433307064,
+      "learning_rate": 5.127801666248079e-06,
+      "loss": 0.4685,
+      "step": 6457
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9001190043422522,
+      "learning_rate": 5.126530280767051e-06,
+      "loss": 0.4954,
+      "step": 6458
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0386995129008207,
+      "learning_rate": 5.12525888709968e-06,
+      "loss": 0.4712,
+      "step": 6459
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7012475507478124,
+      "learning_rate": 5.123987485328221e-06,
+      "loss": 0.4997,
+      "step": 6460
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.273908563563598,
+      "learning_rate": 5.122716075534936e-06,
+      "loss": 0.4852,
+      "step": 6461
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.7741579037143547,
+      "learning_rate": 5.12144465780208e-06,
+      "loss": 0.4961,
+      "step": 6462
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6428764479979168,
+      "learning_rate": 5.120173232211913e-06,
+      "loss": 0.4864,
+      "step": 6463
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6119801629601755,
+      "learning_rate": 5.118901798846697e-06,
+      "loss": 0.4664,
+      "step": 6464
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2764344647502386,
+      "learning_rate": 5.117630357788689e-06,
+      "loss": 0.3777,
+      "step": 6465
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6930540901717557,
+      "learning_rate": 5.116358909120151e-06,
+      "loss": 0.4381,
+      "step": 6466
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6095589962193397,
+      "learning_rate": 5.115087452923344e-06,
+      "loss": 0.4421,
+      "step": 6467
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.460767995766159,
+      "learning_rate": 5.113815989280528e-06,
+      "loss": 0.4536,
+      "step": 6468
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5636398879812817,
+      "learning_rate": 5.112544518273968e-06,
+      "loss": 0.4839,
+      "step": 6469
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4104955803019765,
+      "learning_rate": 5.111273039985924e-06,
+      "loss": 0.4683,
+      "step": 6470
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.702160225164393,
+      "learning_rate": 5.110001554498661e-06,
+      "loss": 0.4799,
+      "step": 6471
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8112330186615966,
+      "learning_rate": 5.108730061894441e-06,
+      "loss": 0.4455,
+      "step": 6472
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5962363683207306,
+      "learning_rate": 5.107458562255527e-06,
+      "loss": 0.4438,
+      "step": 6473
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.463144416939368,
+      "learning_rate": 5.106187055664187e-06,
+      "loss": 0.5069,
+      "step": 6474
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.677187590392192,
+      "learning_rate": 5.1049155422026805e-06,
+      "loss": 0.5011,
+      "step": 6475
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.494190595061933,
+      "learning_rate": 5.103644021953278e-06,
+      "loss": 0.4931,
+      "step": 6476
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5657685831593735,
+      "learning_rate": 5.102372494998241e-06,
+      "loss": 0.4828,
+      "step": 6477
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3397049285586455,
+      "learning_rate": 5.101100961419839e-06,
+      "loss": 0.4284,
+      "step": 6478
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.562365286627066,
+      "learning_rate": 5.099829421300336e-06,
+      "loss": 0.4782,
+      "step": 6479
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1462171757522883,
+      "learning_rate": 5.098557874722e-06,
+      "loss": 0.4822,
+      "step": 6480
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5686820126035461,
+      "learning_rate": 5.097286321767099e-06,
+      "loss": 0.5108,
+      "step": 6481
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5991714741622418,
+      "learning_rate": 5.0960147625179005e-06,
+      "loss": 0.4967,
+      "step": 6482
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5199891720756461,
+      "learning_rate": 5.094743197056672e-06,
+      "loss": 0.4931,
+      "step": 6483
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5953076041477736,
+      "learning_rate": 5.093471625465682e-06,
+      "loss": 0.4983,
+      "step": 6484
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7694251045227662,
+      "learning_rate": 5.092200047827201e-06,
+      "loss": 0.4731,
+      "step": 6485
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2937997136246253,
+      "learning_rate": 5.090928464223498e-06,
+      "loss": 0.4485,
+      "step": 6486
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.403455163043738,
+      "learning_rate": 5.089656874736841e-06,
+      "loss": 0.4763,
+      "step": 6487
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6380145590804795,
+      "learning_rate": 5.088385279449503e-06,
+      "loss": 0.4553,
+      "step": 6488
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6724149728007274,
+      "learning_rate": 5.087113678443751e-06,
+      "loss": 0.462,
+      "step": 6489
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9532041712429067,
+      "learning_rate": 5.085842071801859e-06,
+      "loss": 0.4554,
+      "step": 6490
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6167674487931547,
+      "learning_rate": 5.084570459606097e-06,
+      "loss": 0.4623,
+      "step": 6491
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8669605234622269,
+      "learning_rate": 5.083298841938738e-06,
+      "loss": 0.4838,
+      "step": 6492
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5640856060635908,
+      "learning_rate": 5.082027218882052e-06,
+      "loss": 0.5006,
+      "step": 6493
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5828311424051065,
+      "learning_rate": 5.080755590518314e-06,
+      "loss": 0.502,
+      "step": 6494
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5314133777730905,
+      "learning_rate": 5.0794839569297915e-06,
+      "loss": 0.4457,
+      "step": 6495
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.479749530438239,
+      "learning_rate": 5.078212318198764e-06,
+      "loss": 0.3963,
+      "step": 6496
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6480121647898287,
+      "learning_rate": 5.0769406744075e-06,
+      "loss": 0.4495,
+      "step": 6497
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.2992146818452595,
+      "learning_rate": 5.075669025638277e-06,
+      "loss": 0.4692,
+      "step": 6498
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4481826041154438,
+      "learning_rate": 5.074397371973365e-06,
+      "loss": 0.484,
+      "step": 6499
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6340666341137968,
+      "learning_rate": 5.073125713495041e-06,
+      "loss": 0.4874,
+      "step": 6500
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4830123884863922,
+      "learning_rate": 5.0718540502855785e-06,
+      "loss": 0.448,
+      "step": 6501
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.52551552511447,
+      "learning_rate": 5.070582382427254e-06,
+      "loss": 0.443,
+      "step": 6502
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8181868208968002,
+      "learning_rate": 5.0693107100023395e-06,
+      "loss": 0.389,
+      "step": 6503
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.49161500487639,
+      "learning_rate": 5.068039033093116e-06,
+      "loss": 0.51,
+      "step": 6504
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5763382188101537,
+      "learning_rate": 5.066767351781854e-06,
+      "loss": 0.487,
+      "step": 6505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 3.04663930982442,
+      "learning_rate": 5.065495666150831e-06,
+      "loss": 0.4222,
+      "step": 6506
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5393219359044084,
+      "learning_rate": 5.0642239762823265e-06,
+      "loss": 0.4766,
+      "step": 6507
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6516923896407305,
+      "learning_rate": 5.062952282258613e-06,
+      "loss": 0.4926,
+      "step": 6508
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.025037515728459,
+      "learning_rate": 5.06168058416197e-06,
+      "loss": 0.4818,
+      "step": 6509
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4653430704681132,
+      "learning_rate": 5.060408882074673e-06,
+      "loss": 0.4469,
+      "step": 6510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8705353443065302,
+      "learning_rate": 5.059137176079002e-06,
+      "loss": 0.5067,
+      "step": 6511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7839725964332016,
+      "learning_rate": 5.057865466257231e-06,
+      "loss": 0.4404,
+      "step": 6512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6739655116644743,
+      "learning_rate": 5.056593752691641e-06,
+      "loss": 0.4476,
+      "step": 6513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3650251966476692,
+      "learning_rate": 5.055322035464508e-06,
+      "loss": 0.455,
+      "step": 6514
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7253294831946853,
+      "learning_rate": 5.0540503146581145e-06,
+      "loss": 0.4202,
+      "step": 6515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5242466372334045,
+      "learning_rate": 5.052778590354735e-06,
+      "loss": 0.4904,
+      "step": 6516
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5871181548313401,
+      "learning_rate": 5.051506862636649e-06,
+      "loss": 0.4845,
+      "step": 6517
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6697088613735322,
+      "learning_rate": 5.050235131586137e-06,
+      "loss": 0.4871,
+      "step": 6518
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.0651968957257587,
+      "learning_rate": 5.048963397285478e-06,
+      "loss": 0.444,
+      "step": 6519
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5361936215777474,
+      "learning_rate": 5.04769165981695e-06,
+      "loss": 0.4847,
+      "step": 6520
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5879318638331181,
+      "learning_rate": 5.046419919262836e-06,
+      "loss": 0.4852,
+      "step": 6521
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2785141861422618,
+      "learning_rate": 5.0451481757054126e-06,
+      "loss": 0.4385,
+      "step": 6522
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7864790265833952,
+      "learning_rate": 5.043876429226962e-06,
+      "loss": 0.4438,
+      "step": 6523
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1180513866463517,
+      "learning_rate": 5.042604679909762e-06,
+      "loss": 0.3892,
+      "step": 6524
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.735681639974712,
+      "learning_rate": 5.041332927836097e-06,
+      "loss": 0.5056,
+      "step": 6525
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7360241007882895,
+      "learning_rate": 5.040061173088245e-06,
+      "loss": 0.4585,
+      "step": 6526
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5672631437033615,
+      "learning_rate": 5.038789415748488e-06,
+      "loss": 0.4153,
+      "step": 6527
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4727150012690737,
+      "learning_rate": 5.037517655899105e-06,
+      "loss": 0.4449,
+      "step": 6528
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3860435287685937,
+      "learning_rate": 5.03624589362238e-06,
+      "loss": 0.4623,
+      "step": 6529
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6981328538886096,
+      "learning_rate": 5.034974129000592e-06,
+      "loss": 0.4377,
+      "step": 6530
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.1816491065009456,
+      "learning_rate": 5.033702362116025e-06,
+      "loss": 0.505,
+      "step": 6531
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7118607001755564,
+      "learning_rate": 5.032430593050959e-06,
+      "loss": 0.4419,
+      "step": 6532
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.618560117825182,
+      "learning_rate": 5.031158821887676e-06,
+      "loss": 0.4744,
+      "step": 6533
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6799788168967031,
+      "learning_rate": 5.029887048708457e-06,
+      "loss": 0.4985,
+      "step": 6534
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 4.38271717859473,
+      "learning_rate": 5.028615273595585e-06,
+      "loss": 0.4783,
+      "step": 6535
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6355921898547582,
+      "learning_rate": 5.027343496631343e-06,
+      "loss": 0.5105,
+      "step": 6536
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5477794782112593,
+      "learning_rate": 5.026071717898012e-06,
+      "loss": 0.4838,
+      "step": 6537
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5834892467927764,
+      "learning_rate": 5.0247999374778755e-06,
+      "loss": 0.4921,
+      "step": 6538
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.9048369715373197,
+      "learning_rate": 5.023528155453217e-06,
+      "loss": 0.4616,
+      "step": 6539
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4381894090616965,
+      "learning_rate": 5.0222563719063155e-06,
+      "loss": 0.4911,
+      "step": 6540
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6092934041167208,
+      "learning_rate": 5.020984586919455e-06,
+      "loss": 0.4746,
+      "step": 6541
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.527099662127723,
+      "learning_rate": 5.019712800574922e-06,
+      "loss": 0.4425,
+      "step": 6542
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6155996733860407,
+      "learning_rate": 5.018441012954994e-06,
+      "loss": 0.4852,
+      "step": 6543
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.064781630285737,
+      "learning_rate": 5.017169224141959e-06,
+      "loss": 0.4702,
+      "step": 6544
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4796614812995288,
+      "learning_rate": 5.015897434218095e-06,
+      "loss": 0.449,
+      "step": 6545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.468740237532509,
+      "learning_rate": 5.014625643265691e-06,
+      "loss": 0.4589,
+      "step": 6546
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7914998376635575,
+      "learning_rate": 5.013353851367023e-06,
+      "loss": 0.5011,
+      "step": 6547
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6147302324372743,
+      "learning_rate": 5.0120820586043815e-06,
+      "loss": 0.4616,
+      "step": 6548
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4625195074328725,
+      "learning_rate": 5.010810265060045e-06,
+      "loss": 0.4513,
+      "step": 6549
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7003502354707902,
+      "learning_rate": 5.0095384708163e-06,
+      "loss": 0.4959,
+      "step": 6550
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7524370562843992,
+      "learning_rate": 5.008266675955428e-06,
+      "loss": 0.4133,
+      "step": 6551
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.688561539101369,
+      "learning_rate": 5.006994880559712e-06,
+      "loss": 0.5215,
+      "step": 6552
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6106223449506112,
+      "learning_rate": 5.005723084711437e-06,
+      "loss": 0.4746,
+      "step": 6553
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4820154619908792,
+      "learning_rate": 5.004451288492886e-06,
+      "loss": 0.4675,
+      "step": 6554
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.246026328559546,
+      "learning_rate": 5.003179491986342e-06,
+      "loss": 0.4857,
+      "step": 6555
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.7613762369657864,
+      "learning_rate": 5.00190769527409e-06,
+      "loss": 0.4103,
+      "step": 6556
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 2.140556331653775,
+      "learning_rate": 5.00063589843841e-06,
+      "loss": 0.4421,
+      "step": 6557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.746998395976713,
+      "learning_rate": 4.999364101561592e-06,
+      "loss": 0.5259,
+      "step": 6558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7602707426218345,
+      "learning_rate": 4.998092304725912e-06,
+      "loss": 0.5022,
+      "step": 6559
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3814016841487906,
+      "learning_rate": 4.9968205080136605e-06,
+      "loss": 0.429,
+      "step": 6560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6379138858433755,
+      "learning_rate": 4.995548711507115e-06,
+      "loss": 0.4855,
+      "step": 6561
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.7931690434332954,
+      "learning_rate": 4.9942769152885655e-06,
+      "loss": 0.4413,
+      "step": 6562
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5984598509535575,
+      "learning_rate": 4.993005119440289e-06,
+      "loss": 0.5105,
+      "step": 6563
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4625184749589843,
+      "learning_rate": 4.991733324044573e-06,
+      "loss": 0.4605,
+      "step": 6564
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.667020001239487,
+      "learning_rate": 4.990461529183701e-06,
+      "loss": 0.4927,
+      "step": 6565
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.672614899182484,
+      "learning_rate": 4.989189734939955e-06,
+      "loss": 0.4448,
+      "step": 6566
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6386465885427992,
+      "learning_rate": 4.987917941395619e-06,
+      "loss": 0.5107,
+      "step": 6567
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7901524897071457,
+      "learning_rate": 4.986646148632977e-06,
+      "loss": 0.4446,
+      "step": 6568
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0106781750327225,
+      "learning_rate": 4.985374356734312e-06,
+      "loss": 0.5412,
+      "step": 6569
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5552134575702795,
+      "learning_rate": 4.984102565781906e-06,
+      "loss": 0.4765,
+      "step": 6570
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7707064918798165,
+      "learning_rate": 4.982830775858044e-06,
+      "loss": 0.4821,
+      "step": 6571
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0863466909489294,
+      "learning_rate": 4.981558987045007e-06,
+      "loss": 0.4865,
+      "step": 6572
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5914179011320913,
+      "learning_rate": 4.98028719942508e-06,
+      "loss": 0.4756,
+      "step": 6573
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8904943073991147,
+      "learning_rate": 4.979015413080546e-06,
+      "loss": 0.4862,
+      "step": 6574
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.697027483572185,
+      "learning_rate": 4.977743628093685e-06,
+      "loss": 0.4622,
+      "step": 6575
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7749033701443462,
+      "learning_rate": 4.976471844546785e-06,
+      "loss": 0.4279,
+      "step": 6576
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6074300144830493,
+      "learning_rate": 4.9752000625221245e-06,
+      "loss": 0.4585,
+      "step": 6577
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.403553716427613,
+      "learning_rate": 4.9739282821019885e-06,
+      "loss": 0.4544,
+      "step": 6578
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6483493864203449,
+      "learning_rate": 4.972656503368658e-06,
+      "loss": 0.4975,
+      "step": 6579
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6865729887753296,
+      "learning_rate": 4.971384726404416e-06,
+      "loss": 0.4474,
+      "step": 6580
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5528975609979347,
+      "learning_rate": 4.970112951291545e-06,
+      "loss": 0.4315,
+      "step": 6581
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.585093527575743,
+      "learning_rate": 4.9688411781123266e-06,
+      "loss": 0.4623,
+      "step": 6582
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3373019867343807,
+      "learning_rate": 4.9675694069490435e-06,
+      "loss": 0.4481,
+      "step": 6583
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5914545680857263,
+      "learning_rate": 4.966297637883977e-06,
+      "loss": 0.5169,
+      "step": 6584
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5629030584735683,
+      "learning_rate": 4.965025870999409e-06,
+      "loss": 0.4918,
+      "step": 6585
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.724908788467732,
+      "learning_rate": 4.963754106377622e-06,
+      "loss": 0.4773,
+      "step": 6586
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6909992653112695,
+      "learning_rate": 4.962482344100897e-06,
+      "loss": 0.4685,
+      "step": 6587
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9737451479676558,
+      "learning_rate": 4.961210584251515e-06,
+      "loss": 0.4931,
+      "step": 6588
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5825480971202149,
+      "learning_rate": 4.959938826911758e-06,
+      "loss": 0.4956,
+      "step": 6589
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7931334949447801,
+      "learning_rate": 4.958667072163904e-06,
+      "loss": 0.5093,
+      "step": 6590
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5849061329745877,
+      "learning_rate": 4.95739532009024e-06,
+      "loss": 0.483,
+      "step": 6591
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9534836672895128,
+      "learning_rate": 4.95612357077304e-06,
+      "loss": 0.4713,
+      "step": 6592
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6364517814415565,
+      "learning_rate": 4.95485182429459e-06,
+      "loss": 0.4564,
+      "step": 6593
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.381835383830577,
+      "learning_rate": 4.953580080737165e-06,
+      "loss": 0.4302,
+      "step": 6594
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.522068476764517,
+      "learning_rate": 4.9523083401830525e-06,
+      "loss": 0.4773,
+      "step": 6595
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5738696836903748,
+      "learning_rate": 4.951036602714523e-06,
+      "loss": 0.51,
+      "step": 6596
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8494017245688474,
+      "learning_rate": 4.949764868413866e-06,
+      "loss": 0.4589,
+      "step": 6597
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4704047226354522,
+      "learning_rate": 4.948493137363352e-06,
+      "loss": 0.4665,
+      "step": 6598
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.754697408068161,
+      "learning_rate": 4.947221409645266e-06,
+      "loss": 0.4425,
+      "step": 6599
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0926232812347916,
+      "learning_rate": 4.945949685341887e-06,
+      "loss": 0.4899,
+      "step": 6600
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7522370551380122,
+      "learning_rate": 4.944677964535491e-06,
+      "loss": 0.4871,
+      "step": 6601
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.759606510750112,
+      "learning_rate": 4.94340624730836e-06,
+      "loss": 0.4566,
+      "step": 6602
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.517509056644059,
+      "learning_rate": 4.94213453374277e-06,
+      "loss": 0.4742,
+      "step": 6603
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5700796353976145,
+      "learning_rate": 4.940862823921001e-06,
+      "loss": 0.4982,
+      "step": 6604
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.396792520871284,
+      "learning_rate": 4.939591117925328e-06,
+      "loss": 0.4608,
+      "step": 6605
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5312569656974562,
+      "learning_rate": 4.938319415838032e-06,
+      "loss": 0.4473,
+      "step": 6606
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.716813521772426,
+      "learning_rate": 4.937047717741388e-06,
+      "loss": 0.4179,
+      "step": 6607
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4851415029821977,
+      "learning_rate": 4.935776023717676e-06,
+      "loss": 0.4562,
+      "step": 6608
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3049576042663205,
+      "learning_rate": 4.93450433384917e-06,
+      "loss": 0.4146,
+      "step": 6609
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8107778157177974,
+      "learning_rate": 4.933232648218146e-06,
+      "loss": 0.4324,
+      "step": 6610
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.155892743859018,
+      "learning_rate": 4.931960966906886e-06,
+      "loss": 0.4465,
+      "step": 6611
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.084578534647461,
+      "learning_rate": 4.93068928999766e-06,
+      "loss": 0.4579,
+      "step": 6612
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5679523033413384,
+      "learning_rate": 4.929417617572748e-06,
+      "loss": 0.4872,
+      "step": 6613
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9941322909236727,
+      "learning_rate": 4.928145949714422e-06,
+      "loss": 0.4781,
+      "step": 6614
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7339470559673689,
+      "learning_rate": 4.9268742865049616e-06,
+      "loss": 0.4949,
+      "step": 6615
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7856585746286044,
+      "learning_rate": 4.925602628026636e-06,
+      "loss": 0.4814,
+      "step": 6616
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5056608638852746,
+      "learning_rate": 4.924330974361727e-06,
+      "loss": 0.5297,
+      "step": 6617
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8359333336880963,
+      "learning_rate": 4.923059325592501e-06,
+      "loss": 0.4741,
+      "step": 6618
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5503684548913055,
+      "learning_rate": 4.921787681801239e-06,
+      "loss": 0.504,
+      "step": 6619
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.651162966840047,
+      "learning_rate": 4.920516043070209e-06,
+      "loss": 0.4996,
+      "step": 6620
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4520006085159736,
+      "learning_rate": 4.919244409481688e-06,
+      "loss": 0.4965,
+      "step": 6621
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.507379351000236,
+      "learning_rate": 4.91797278111795e-06,
+      "loss": 0.4713,
+      "step": 6622
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5232658273501551,
+      "learning_rate": 4.9167011580612626e-06,
+      "loss": 0.4622,
+      "step": 6623
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.2678727128390275,
+      "learning_rate": 4.915429540393904e-06,
+      "loss": 0.4494,
+      "step": 6624
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.595051580143376,
+      "learning_rate": 4.914157928198141e-06,
+      "loss": 0.4981,
+      "step": 6625
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8361111889880855,
+      "learning_rate": 4.912886321556251e-06,
+      "loss": 0.4302,
+      "step": 6626
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1494978728749725,
+      "learning_rate": 4.911614720550498e-06,
+      "loss": 0.5162,
+      "step": 6627
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.338526152909621,
+      "learning_rate": 4.910343125263161e-06,
+      "loss": 0.4824,
+      "step": 6628
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5747473181858885,
+      "learning_rate": 4.909071535776504e-06,
+      "loss": 0.4764,
+      "step": 6629
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8721175945294664,
+      "learning_rate": 4.907799952172801e-06,
+      "loss": 0.4988,
+      "step": 6630
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6734475179245696,
+      "learning_rate": 4.906528374534319e-06,
+      "loss": 0.4748,
+      "step": 6631
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.75169473167023,
+      "learning_rate": 4.905256802943329e-06,
+      "loss": 0.4632,
+      "step": 6632
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5928872313793587,
+      "learning_rate": 4.903985237482101e-06,
+      "loss": 0.4833,
+      "step": 6633
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3230413265437955,
+      "learning_rate": 4.902713678232901e-06,
+      "loss": 0.4837,
+      "step": 6634
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.6438141443289713,
+      "learning_rate": 4.901442125278002e-06,
+      "loss": 0.5444,
+      "step": 6635
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.703821987295174,
+      "learning_rate": 4.900170578699665e-06,
+      "loss": 0.491,
+      "step": 6636
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5532731761245311,
+      "learning_rate": 4.898899038580163e-06,
+      "loss": 0.5038,
+      "step": 6637
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6137113605692694,
+      "learning_rate": 4.897627505001761e-06,
+      "loss": 0.4789,
+      "step": 6638
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.083458359436787,
+      "learning_rate": 4.8963559780467245e-06,
+      "loss": 0.4427,
+      "step": 6639
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 8.735699937683806,
+      "learning_rate": 4.89508445779732e-06,
+      "loss": 0.5132,
+      "step": 6640
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.8038439633313814,
+      "learning_rate": 4.893812944335816e-06,
+      "loss": 0.4564,
+      "step": 6641
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.599346121861896,
+      "learning_rate": 4.892541437744474e-06,
+      "loss": 0.4956,
+      "step": 6642
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6319668672827046,
+      "learning_rate": 4.89126993810556e-06,
+      "loss": 0.4473,
+      "step": 6643
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3807712357452564,
+      "learning_rate": 4.889998445501341e-06,
+      "loss": 0.4615,
+      "step": 6644
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8089269249207602,
+      "learning_rate": 4.888726960014076e-06,
+      "loss": 0.4839,
+      "step": 6645
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6250446315475635,
+      "learning_rate": 4.8874554817260326e-06,
+      "loss": 0.4799,
+      "step": 6646
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.4841089594865604,
+      "learning_rate": 4.886184010719472e-06,
+      "loss": 0.4558,
+      "step": 6647
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.043329935266655,
+      "learning_rate": 4.884912547076658e-06,
+      "loss": 0.4527,
+      "step": 6648
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7329224835715065,
+      "learning_rate": 4.8836410908798494e-06,
+      "loss": 0.4962,
+      "step": 6649
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5786666192420358,
+      "learning_rate": 4.882369642211312e-06,
+      "loss": 0.4526,
+      "step": 6650
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.0449134902508295,
+      "learning_rate": 4.881098201153304e-06,
+      "loss": 0.4538,
+      "step": 6651
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4191880418898544,
+      "learning_rate": 4.8798267677880876e-06,
+      "loss": 0.4679,
+      "step": 6652
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7900797804954813,
+      "learning_rate": 4.878555342197921e-06,
+      "loss": 0.4659,
+      "step": 6653
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4665200419805204,
+      "learning_rate": 4.8772839244650656e-06,
+      "loss": 0.4176,
+      "step": 6654
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6194357665050376,
+      "learning_rate": 4.87601251467178e-06,
+      "loss": 0.4999,
+      "step": 6655
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.126487135138284,
+      "learning_rate": 4.874741112900322e-06,
+      "loss": 0.4715,
+      "step": 6656
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5486977297393273,
+      "learning_rate": 4.87346971923295e-06,
+      "loss": 0.5084,
+      "step": 6657
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.8426484825007232,
+      "learning_rate": 4.8721983337519225e-06,
+      "loss": 0.4898,
+      "step": 6658
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6571438035086183,
+      "learning_rate": 4.870926956539496e-06,
+      "loss": 0.4679,
+      "step": 6659
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9865248860182858,
+      "learning_rate": 4.869655587677924e-06,
+      "loss": 0.4848,
+      "step": 6660
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5541856066603268,
+      "learning_rate": 4.868384227249468e-06,
+      "loss": 0.5084,
+      "step": 6661
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5767249659101897,
+      "learning_rate": 4.867112875336377e-06,
+      "loss": 0.5067,
+      "step": 6662
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5353966167917583,
+      "learning_rate": 4.865841532020913e-06,
+      "loss": 0.4977,
+      "step": 6663
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5283131099066546,
+      "learning_rate": 4.864570197385322e-06,
+      "loss": 0.4931,
+      "step": 6664
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.603006967679426,
+      "learning_rate": 4.863298871511865e-06,
+      "loss": 0.4396,
+      "step": 6665
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.137228529548092,
+      "learning_rate": 4.862027554482792e-06,
+      "loss": 0.492,
+      "step": 6666
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5576536805604777,
+      "learning_rate": 4.860756246380355e-06,
+      "loss": 0.4479,
+      "step": 6667
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5608197759602769,
+      "learning_rate": 4.859484947286807e-06,
+      "loss": 0.4605,
+      "step": 6668
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.046512708853382,
+      "learning_rate": 4.8582136572844e-06,
+      "loss": 0.492,
+      "step": 6669
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.9007808083191864,
+      "learning_rate": 4.856942376455384e-06,
+      "loss": 0.4541,
+      "step": 6670
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7703421058908309,
+      "learning_rate": 4.855671104882007e-06,
+      "loss": 0.4191,
+      "step": 6671
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6791083951423138,
+      "learning_rate": 4.854399842646523e-06,
+      "loss": 0.4354,
+      "step": 6672
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.816002673511426,
+      "learning_rate": 4.853128589831177e-06,
+      "loss": 0.4767,
+      "step": 6673
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6977270770102055,
+      "learning_rate": 4.85185734651822e-06,
+      "loss": 0.4247,
+      "step": 6674
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.3852248613693625,
+      "learning_rate": 4.850586112789898e-06,
+      "loss": 0.4939,
+      "step": 6675
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.7050024404876745,
+      "learning_rate": 4.849314888728461e-06,
+      "loss": 0.4043,
+      "step": 6676
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.250295026271321,
+      "learning_rate": 4.848043674416151e-06,
+      "loss": 0.4919,
+      "step": 6677
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 2.1360740699064604,
+      "learning_rate": 4.846772469935217e-06,
+      "loss": 0.4817,
+      "step": 6678
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6129578544997196,
+      "learning_rate": 4.8455012753679055e-06,
+      "loss": 0.483,
+      "step": 6679
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6085352296148729,
+      "learning_rate": 4.844230090796456e-06,
+      "loss": 0.5079,
+      "step": 6680
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5658977497232243,
+      "learning_rate": 4.84295891630312e-06,
+      "loss": 0.4946,
+      "step": 6681
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.768054008013359,
+      "learning_rate": 4.841687751970135e-06,
+      "loss": 0.4782,
+      "step": 6682
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.9134607265158077,
+      "learning_rate": 4.840416597879745e-06,
+      "loss": 0.5162,
+      "step": 6683
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.6921107203446453,
+      "learning_rate": 4.839145454114192e-06,
+      "loss": 0.4985,
+      "step": 6684
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9679053500275927,
+      "learning_rate": 4.8378743207557185e-06,
+      "loss": 0.4788,
+      "step": 6685
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6901348036170842,
+      "learning_rate": 4.836603197886564e-06,
+      "loss": 0.47,
+      "step": 6686
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.13183847023187,
+      "learning_rate": 4.8353320855889695e-06,
+      "loss": 0.4411,
+      "step": 6687
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5961399061141976,
+      "learning_rate": 4.834060983945173e-06,
+      "loss": 0.4416,
+      "step": 6688
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8722968907181048,
+      "learning_rate": 4.832789893037414e-06,
+      "loss": 0.5138,
+      "step": 6689
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0640403752998626,
+      "learning_rate": 4.83151881294793e-06,
+      "loss": 0.452,
+      "step": 6690
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5473051722122786,
+      "learning_rate": 4.830247743758958e-06,
+      "loss": 0.4393,
+      "step": 6691
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.777382698419964,
+      "learning_rate": 4.828976685552736e-06,
+      "loss": 0.4595,
+      "step": 6692
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8362928620077252,
+      "learning_rate": 4.827705638411498e-06,
+      "loss": 0.46,
+      "step": 6693
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6333266890757783,
+      "learning_rate": 4.8264346024174805e-06,
+      "loss": 0.5163,
+      "step": 6694
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.357337029961531,
+      "learning_rate": 4.8251635776529145e-06,
+      "loss": 0.4597,
+      "step": 6695
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7135765788443145,
+      "learning_rate": 4.823892564200041e-06,
+      "loss": 0.4812,
+      "step": 6696
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6477493336238613,
+      "learning_rate": 4.822621562141083e-06,
+      "loss": 0.5354,
+      "step": 6697
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4076798938440025,
+      "learning_rate": 4.8213505715582825e-06,
+      "loss": 0.432,
+      "step": 6698
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.872817909294625,
+      "learning_rate": 4.820079592533863e-06,
+      "loss": 0.4669,
+      "step": 6699
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7045707242200616,
+      "learning_rate": 4.818808625150059e-06,
+      "loss": 0.4077,
+      "step": 6700
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.3066747084186323,
+      "learning_rate": 4.817537669489102e-06,
+      "loss": 0.4998,
+      "step": 6701
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5719337371379443,
+      "learning_rate": 4.816266725633217e-06,
+      "loss": 0.491,
+      "step": 6702
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5471063264534128,
+      "learning_rate": 4.814995793664635e-06,
+      "loss": 0.4915,
+      "step": 6703
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.863722558240749,
+      "learning_rate": 4.813724873665584e-06,
+      "loss": 0.5057,
+      "step": 6704
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.401620419186506,
+      "learning_rate": 4.81245396571829e-06,
+      "loss": 0.4995,
+      "step": 6705
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1341052732329504,
+      "learning_rate": 4.8111830699049786e-06,
+      "loss": 0.4419,
+      "step": 6706
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.350273591991399,
+      "learning_rate": 4.8099121863078756e-06,
+      "loss": 0.4737,
+      "step": 6707
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.626716249700459,
+      "learning_rate": 4.808641315009205e-06,
+      "loss": 0.4456,
+      "step": 6708
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6789296859407992,
+      "learning_rate": 4.807370456091192e-06,
+      "loss": 0.4653,
+      "step": 6709
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.568284743686291,
+      "learning_rate": 4.8060996096360576e-06,
+      "loss": 0.4059,
+      "step": 6710
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6857332866413042,
+      "learning_rate": 4.804828775726023e-06,
+      "loss": 0.4226,
+      "step": 6711
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5841408113229113,
+      "learning_rate": 4.803557954443316e-06,
+      "loss": 0.4846,
+      "step": 6712
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.381725494190368,
+      "learning_rate": 4.802287145870148e-06,
+      "loss": 0.4608,
+      "step": 6713
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.349057462129503,
+      "learning_rate": 4.801016350088747e-06,
+      "loss": 0.4604,
+      "step": 6714
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5637023360271405,
+      "learning_rate": 4.799745567181323e-06,
+      "loss": 0.5196,
+      "step": 6715
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7164546471348563,
+      "learning_rate": 4.798474797230103e-06,
+      "loss": 0.4291,
+      "step": 6716
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7041170713192688,
+      "learning_rate": 4.797204040317297e-06,
+      "loss": 0.4628,
+      "step": 6717
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1716075974897664,
+      "learning_rate": 4.795933296525126e-06,
+      "loss": 0.4576,
+      "step": 6718
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.005365912475376,
+      "learning_rate": 4.794662565935803e-06,
+      "loss": 0.4331,
+      "step": 6719
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7747751146802806,
+      "learning_rate": 4.793391848631545e-06,
+      "loss": 0.4383,
+      "step": 6720
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.570733955633041,
+      "learning_rate": 4.792121144694563e-06,
+      "loss": 0.4789,
+      "step": 6721
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1695540170070995,
+      "learning_rate": 4.79085045420707e-06,
+      "loss": 0.4865,
+      "step": 6722
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.727169963089276,
+      "learning_rate": 4.789579777251281e-06,
+      "loss": 0.4761,
+      "step": 6723
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7291508018262458,
+      "learning_rate": 4.788309113909403e-06,
+      "loss": 0.4361,
+      "step": 6724
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.705861371318313,
+      "learning_rate": 4.787038464263651e-06,
+      "loss": 0.4223,
+      "step": 6725
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5672587456306586,
+      "learning_rate": 4.785767828396229e-06,
+      "loss": 0.4858,
+      "step": 6726
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.16512750088143,
+      "learning_rate": 4.7844972063893505e-06,
+      "loss": 0.4887,
+      "step": 6727
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 6.97064518061688,
+      "learning_rate": 4.7832265983252195e-06,
+      "loss": 0.4877,
+      "step": 6728
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4704118534320378,
+      "learning_rate": 4.781956004286045e-06,
+      "loss": 0.4352,
+      "step": 6729
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9015981566658091,
+      "learning_rate": 4.7806854243540304e-06,
+      "loss": 0.4831,
+      "step": 6730
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7567716512426574,
+      "learning_rate": 4.779414858611384e-06,
+      "loss": 0.4414,
+      "step": 6731
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8134928907962284,
+      "learning_rate": 4.778144307140305e-06,
+      "loss": 0.4379,
+      "step": 6732
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8554622467275732,
+      "learning_rate": 4.776873770023e-06,
+      "loss": 0.4676,
+      "step": 6733
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5533367181369777,
+      "learning_rate": 4.775603247341671e-06,
+      "loss": 0.465,
+      "step": 6734
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5739806818878227,
+      "learning_rate": 4.774332739178519e-06,
+      "loss": 0.4827,
+      "step": 6735
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4932931040628383,
+      "learning_rate": 4.773062245615742e-06,
+      "loss": 0.4566,
+      "step": 6736
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4523757845468528,
+      "learning_rate": 4.771791766735541e-06,
+      "loss": 0.438,
+      "step": 6737
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6041285269136627,
+      "learning_rate": 4.770521302620116e-06,
+      "loss": 0.5082,
+      "step": 6738
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5909074500097466,
+      "learning_rate": 4.76925085335166e-06,
+      "loss": 0.4633,
+      "step": 6739
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.765590920070977,
+      "learning_rate": 4.7679804190123745e-06,
+      "loss": 0.4354,
+      "step": 6740
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6688630773308473,
+      "learning_rate": 4.76670999968445e-06,
+      "loss": 0.5043,
+      "step": 6741
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.862423802875408,
+      "learning_rate": 4.765439595450085e-06,
+      "loss": 0.4966,
+      "step": 6742
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.497814853240861,
+      "learning_rate": 4.76416920639147e-06,
+      "loss": 0.4505,
+      "step": 6743
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.4166943086215094,
+      "learning_rate": 4.7628988325907985e-06,
+      "loss": 0.4593,
+      "step": 6744
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5090953076732723,
+      "learning_rate": 4.761628474130265e-06,
+      "loss": 0.5156,
+      "step": 6745
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8345607229160001,
+      "learning_rate": 4.760358131092054e-06,
+      "loss": 0.4348,
+      "step": 6746
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6835827591594243,
+      "learning_rate": 4.759087803558362e-06,
+      "loss": 0.4743,
+      "step": 6747
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6058234461878689,
+      "learning_rate": 4.757817491611371e-06,
+      "loss": 0.4258,
+      "step": 6748
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.7275344434298945,
+      "learning_rate": 4.756547195333274e-06,
+      "loss": 0.485,
+      "step": 6749
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7207539024518654,
+      "learning_rate": 4.755276914806252e-06,
+      "loss": 0.4502,
+      "step": 6750
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5110840116608724,
+      "learning_rate": 4.754006650112497e-06,
+      "loss": 0.43,
+      "step": 6751
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.191974552051519,
+      "learning_rate": 4.752736401334186e-06,
+      "loss": 0.455,
+      "step": 6752
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5466207458158461,
+      "learning_rate": 4.751466168553509e-06,
+      "loss": 0.4925,
+      "step": 6753
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.544468075265627,
+      "learning_rate": 4.750195951852645e-06,
+      "loss": 0.5073,
+      "step": 6754
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.567142676446206,
+      "learning_rate": 4.748925751313777e-06,
+      "loss": 0.4525,
+      "step": 6755
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7578513505667428,
+      "learning_rate": 4.7476555670190825e-06,
+      "loss": 0.5002,
+      "step": 6756
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9734864353872705,
+      "learning_rate": 4.746385399050743e-06,
+      "loss": 0.4029,
+      "step": 6757
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.100411063616878,
+      "learning_rate": 4.745115247490939e-06,
+      "loss": 0.4878,
+      "step": 6758
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.031113820668726,
+      "learning_rate": 4.743845112421842e-06,
+      "loss": 0.4416,
+      "step": 6759
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.9038619361443163,
+      "learning_rate": 4.742574993925634e-06,
+      "loss": 0.5338,
+      "step": 6760
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.1726066154631396,
+      "learning_rate": 4.741304892084485e-06,
+      "loss": 0.4818,
+      "step": 6761
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5338824217251135,
+      "learning_rate": 4.7400348069805725e-06,
+      "loss": 0.4042,
+      "step": 6762
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5948232395826915,
+      "learning_rate": 4.738764738696066e-06,
+      "loss": 0.4669,
+      "step": 6763
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4675493739613286,
+      "learning_rate": 4.737494687313142e-06,
+      "loss": 0.4334,
+      "step": 6764
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8959981046707237,
+      "learning_rate": 4.736224652913964e-06,
+      "loss": 0.4581,
+      "step": 6765
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7293635276879877,
+      "learning_rate": 4.734954635580711e-06,
+      "loss": 0.4464,
+      "step": 6766
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6113007672339685,
+      "learning_rate": 4.733684635395543e-06,
+      "loss": 0.5038,
+      "step": 6767
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6181690078500528,
+      "learning_rate": 4.732414652440631e-06,
+      "loss": 0.4889,
+      "step": 6768
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8943666291587404,
+      "learning_rate": 4.731144686798141e-06,
+      "loss": 0.4331,
+      "step": 6769
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7339765916656622,
+      "learning_rate": 4.729874738550238e-06,
+      "loss": 0.471,
+      "step": 6770
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9791596615487284,
+      "learning_rate": 4.728604807779088e-06,
+      "loss": 0.5035,
+      "step": 6771
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6588193877814434,
+      "learning_rate": 4.72733489456685e-06,
+      "loss": 0.47,
+      "step": 6772
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.0322253226925002,
+      "learning_rate": 4.726064998995689e-06,
+      "loss": 0.4633,
+      "step": 6773
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5420673408314831,
+      "learning_rate": 4.724795121147763e-06,
+      "loss": 0.4861,
+      "step": 6774
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2998038760275714,
+      "learning_rate": 4.7235252611052335e-06,
+      "loss": 0.4855,
+      "step": 6775
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.680774961926386,
+      "learning_rate": 4.722255418950257e-06,
+      "loss": 0.4353,
+      "step": 6776
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8501091936577974,
+      "learning_rate": 4.720985594764992e-06,
+      "loss": 0.4797,
+      "step": 6777
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 5.12795171299923,
+      "learning_rate": 4.719715788631594e-06,
+      "loss": 0.5094,
+      "step": 6778
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5327401718329199,
+      "learning_rate": 4.7184460006322165e-06,
+      "loss": 0.4312,
+      "step": 6779
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6151446974365018,
+      "learning_rate": 4.717176230849017e-06,
+      "loss": 0.4769,
+      "step": 6780
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6175670787972947,
+      "learning_rate": 4.715906479364143e-06,
+      "loss": 0.4481,
+      "step": 6781
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3977895363788317,
+      "learning_rate": 4.714636746259752e-06,
+      "loss": 0.3851,
+      "step": 6782
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6031768326040865,
+      "learning_rate": 4.713367031617985e-06,
+      "loss": 0.4549,
+      "step": 6783
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.394703480351404,
+      "learning_rate": 4.712097335521001e-06,
+      "loss": 0.4428,
+      "step": 6784
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8475387504517746,
+      "learning_rate": 4.710827658050938e-06,
+      "loss": 0.4156,
+      "step": 6785
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5977737305318904,
+      "learning_rate": 4.709557999289952e-06,
+      "loss": 0.4891,
+      "step": 6786
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6703112933762194,
+      "learning_rate": 4.708288359320179e-06,
+      "loss": 0.4653,
+      "step": 6787
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6043587101097239,
+      "learning_rate": 4.707018738223771e-06,
+      "loss": 0.4827,
+      "step": 6788
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7868653660431113,
+      "learning_rate": 4.705749136082865e-06,
+      "loss": 0.4292,
+      "step": 6789
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.309115828233961,
+      "learning_rate": 4.704479552979606e-06,
+      "loss": 0.4794,
+      "step": 6790
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5025472626538847,
+      "learning_rate": 4.703209988996134e-06,
+      "loss": 0.4629,
+      "step": 6791
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.342067085574972,
+      "learning_rate": 4.701940444214587e-06,
+      "loss": 0.4643,
+      "step": 6792
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.316857941221925,
+      "learning_rate": 4.700670918717103e-06,
+      "loss": 0.4626,
+      "step": 6793
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8294907434680994,
+      "learning_rate": 4.699401412585819e-06,
+      "loss": 0.4342,
+      "step": 6794
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9137935536850157,
+      "learning_rate": 4.6981319259028715e-06,
+      "loss": 0.4628,
+      "step": 6795
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.602900838490069,
+      "learning_rate": 4.696862458750391e-06,
+      "loss": 0.4987,
+      "step": 6796
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4894904126997455,
+      "learning_rate": 4.695593011210516e-06,
+      "loss": 0.4076,
+      "step": 6797
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7227219366705002,
+      "learning_rate": 4.6943235833653725e-06,
+      "loss": 0.4413,
+      "step": 6798
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8355061977723959,
+      "learning_rate": 4.693054175297095e-06,
+      "loss": 0.4056,
+      "step": 6799
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5692098286245784,
+      "learning_rate": 4.691784787087808e-06,
+      "loss": 0.5025,
+      "step": 6800
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.8870868140879573,
+      "learning_rate": 4.690515418819644e-06,
+      "loss": 0.4681,
+      "step": 6801
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.6186380036414465,
+      "learning_rate": 4.6892460705747275e-06,
+      "loss": 0.4595,
+      "step": 6802
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.9295462236825514,
+      "learning_rate": 4.687976742435184e-06,
+      "loss": 0.4471,
+      "step": 6803
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.357204122165277,
+      "learning_rate": 4.6867074344831375e-06,
+      "loss": 0.4366,
+      "step": 6804
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.7172905622659762,
+      "learning_rate": 4.685438146800709e-06,
+      "loss": 0.4685,
+      "step": 6805
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4973659732329059,
+      "learning_rate": 4.684168879470022e-06,
+      "loss": 0.438,
+      "step": 6806
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.45824204397553,
+      "learning_rate": 4.682899632573195e-06,
+      "loss": 0.4645,
+      "step": 6807
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.654579492802134,
+      "learning_rate": 4.6816304061923465e-06,
+      "loss": 0.482,
+      "step": 6808
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.799216005391002,
+      "learning_rate": 4.680361200409595e-06,
+      "loss": 0.4492,
+      "step": 6809
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 2.6795940314421296,
+      "learning_rate": 4.6790920153070555e-06,
+      "loss": 0.4651,
+      "step": 6810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.5771738285772394,
+      "learning_rate": 4.6778228509668415e-06,
+      "loss": 0.4105,
+      "step": 6811
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5824114305376981,
+      "learning_rate": 4.676553707471068e-06,
+      "loss": 0.5083,
+      "step": 6812
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.431769615061043,
+      "learning_rate": 4.675284584901848e-06,
+      "loss": 0.3939,
+      "step": 6813
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4908612055598287,
+      "learning_rate": 4.674015483341288e-06,
+      "loss": 0.4503,
+      "step": 6814
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8554257559873495,
+      "learning_rate": 4.6727464028715035e-06,
+      "loss": 0.4719,
+      "step": 6815
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.659621956576652,
+      "learning_rate": 4.671477343574595e-06,
+      "loss": 0.4617,
+      "step": 6816
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.666955430664887,
+      "learning_rate": 4.670208305532676e-06,
+      "loss": 0.4894,
+      "step": 6817
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.614622044908237,
+      "learning_rate": 4.668939288827845e-06,
+      "loss": 0.4608,
+      "step": 6818
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.1913143787134808,
+      "learning_rate": 4.667670293542211e-06,
+      "loss": 0.3918,
+      "step": 6819
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.887649836365518,
+      "learning_rate": 4.666401319757873e-06,
+      "loss": 0.4105,
+      "step": 6820
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.673264152790568,
+      "learning_rate": 4.665132367556935e-06,
+      "loss": 0.4874,
+      "step": 6821
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.972038032477824,
+      "learning_rate": 4.6638634370214915e-06,
+      "loss": 0.4641,
+      "step": 6822
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.079730238263787,
+      "learning_rate": 4.662594528233647e-06,
+      "loss": 0.4883,
+      "step": 6823
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9948773702354168,
+      "learning_rate": 4.661325641275494e-06,
+      "loss": 0.4874,
+      "step": 6824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6447950679966556,
+      "learning_rate": 4.660056776229128e-06,
+      "loss": 0.4328,
+      "step": 6825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7685107666624231,
+      "learning_rate": 4.6587879331766465e-06,
+      "loss": 0.4338,
+      "step": 6826
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5384911618441113,
+      "learning_rate": 4.657519112200137e-06,
+      "loss": 0.4633,
+      "step": 6827
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4016491360099923,
+      "learning_rate": 4.656250313381694e-06,
+      "loss": 0.4292,
+      "step": 6828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5500420936321144,
+      "learning_rate": 4.654981536803406e-06,
+      "loss": 0.4382,
+      "step": 6829
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.698424092041151,
+      "learning_rate": 4.653712782547361e-06,
+      "loss": 0.4864,
+      "step": 6830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6562433675665198,
+      "learning_rate": 4.652444050695646e-06,
+      "loss": 0.4612,
+      "step": 6831
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.257581930569672,
+      "learning_rate": 4.651175341330346e-06,
+      "loss": 0.4721,
+      "step": 6832
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9767963757533966,
+      "learning_rate": 4.649906654533545e-06,
+      "loss": 0.4673,
+      "step": 6833
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9312445662347821,
+      "learning_rate": 4.648637990387327e-06,
+      "loss": 0.4224,
+      "step": 6834
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1774070990831613,
+      "learning_rate": 4.6473693489737685e-06,
+      "loss": 0.4589,
+      "step": 6835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.5940896626242633,
+      "learning_rate": 4.646100730374953e-06,
+      "loss": 0.4694,
+      "step": 6836
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5921657602076107,
+      "learning_rate": 4.644832134672958e-06,
+      "loss": 0.498,
+      "step": 6837
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5322538272231689,
+      "learning_rate": 4.643563561949859e-06,
+      "loss": 0.4715,
+      "step": 6838
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1075641352261476,
+      "learning_rate": 4.642295012287732e-06,
+      "loss": 0.4297,
+      "step": 6839
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2623400071224085,
+      "learning_rate": 4.641026485768648e-06,
+      "loss": 0.5033,
+      "step": 6840
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.987621659092553,
+      "learning_rate": 4.639757982474683e-06,
+      "loss": 0.4647,
+      "step": 6841
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9827787146373967,
+      "learning_rate": 4.638489502487904e-06,
+      "loss": 0.4462,
+      "step": 6842
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5497318689175171,
+      "learning_rate": 4.637221045890382e-06,
+      "loss": 0.4816,
+      "step": 6843
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3095292864681283,
+      "learning_rate": 4.635952612764183e-06,
+      "loss": 0.4927,
+      "step": 6844
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.810359376555525,
+      "learning_rate": 4.634684203191374e-06,
+      "loss": 0.4425,
+      "step": 6845
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4270532254772705,
+      "learning_rate": 4.633415817254018e-06,
+      "loss": 0.4729,
+      "step": 6846
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5357933835814878,
+      "learning_rate": 4.63214745503418e-06,
+      "loss": 0.4923,
+      "step": 6847
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.617113838999308,
+      "learning_rate": 4.630879116613921e-06,
+      "loss": 0.528,
+      "step": 6848
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.49302208154492,
+      "learning_rate": 4.629610802075298e-06,
+      "loss": 0.4406,
+      "step": 6849
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5096216863067633,
+      "learning_rate": 4.628342511500375e-06,
+      "loss": 0.4938,
+      "step": 6850
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7557354626260413,
+      "learning_rate": 4.627074244971203e-06,
+      "loss": 0.4386,
+      "step": 6851
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5230727005337314,
+      "learning_rate": 4.625806002569842e-06,
+      "loss": 0.4585,
+      "step": 6852
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5813164470794332,
+      "learning_rate": 4.624537784378339e-06,
+      "loss": 0.4757,
+      "step": 6853
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1158389442094396,
+      "learning_rate": 4.623269590478755e-06,
+      "loss": 0.4528,
+      "step": 6854
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7270364714444189,
+      "learning_rate": 4.622001420953132e-06,
+      "loss": 0.4439,
+      "step": 6855
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.630001069879384,
+      "learning_rate": 4.620733275883525e-06,
+      "loss": 0.4528,
+      "step": 6856
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.1693036318804464,
+      "learning_rate": 4.619465155351976e-06,
+      "loss": 0.4324,
+      "step": 6857
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 4.946226020237933,
+      "learning_rate": 4.618197059440535e-06,
+      "loss": 0.481,
+      "step": 6858
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4822099190238134,
+      "learning_rate": 4.616928988231246e-06,
+      "loss": 0.4689,
+      "step": 6859
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.042278470716708,
+      "learning_rate": 4.6156609418061486e-06,
+      "loss": 0.4639,
+      "step": 6860
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6049835920323006,
+      "learning_rate": 4.614392920247288e-06,
+      "loss": 0.4782,
+      "step": 6861
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.649465280365482,
+      "learning_rate": 4.613124923636698e-06,
+      "loss": 0.5041,
+      "step": 6862
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.264740079993466,
+      "learning_rate": 4.611856952056421e-06,
+      "loss": 0.4572,
+      "step": 6863
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4065453297159005,
+      "learning_rate": 4.610589005588492e-06,
+      "loss": 0.4402,
+      "step": 6864
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7222743484095515,
+      "learning_rate": 4.6093210843149445e-06,
+      "loss": 0.4308,
+      "step": 6865
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5300802907627304,
+      "learning_rate": 4.60805318831781e-06,
+      "loss": 0.48,
+      "step": 6866
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7852265672717624,
+      "learning_rate": 4.606785317679124e-06,
+      "loss": 0.4936,
+      "step": 6867
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5006212431287191,
+      "learning_rate": 4.605517472480912e-06,
+      "loss": 0.4171,
+      "step": 6868
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5679573502048902,
+      "learning_rate": 4.604249652805203e-06,
+      "loss": 0.4791,
+      "step": 6869
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3510129374227096,
+      "learning_rate": 4.602981858734026e-06,
+      "loss": 0.4317,
+      "step": 6870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.008210078038119,
+      "learning_rate": 4.601714090349401e-06,
+      "loss": 0.4798,
+      "step": 6871
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5532850807678098,
+      "learning_rate": 4.600446347733354e-06,
+      "loss": 0.4901,
+      "step": 6872
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.732794627759301,
+      "learning_rate": 4.599178630967906e-06,
+      "loss": 0.4949,
+      "step": 6873
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.546378165576194,
+      "learning_rate": 4.597910940135077e-06,
+      "loss": 0.44,
+      "step": 6874
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.167401987543923,
+      "learning_rate": 4.596643275316882e-06,
+      "loss": 0.4307,
+      "step": 6875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.556657014000904,
+      "learning_rate": 4.595375636595342e-06,
+      "loss": 0.4795,
+      "step": 6876
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5701375459132573,
+      "learning_rate": 4.594108024052468e-06,
+      "loss": 0.5108,
+      "step": 6877
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5439582717198745,
+      "learning_rate": 4.592840437770274e-06,
+      "loss": 0.5143,
+      "step": 6878
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0204214102469957,
+      "learning_rate": 4.59157287783077e-06,
+      "loss": 0.4706,
+      "step": 6879
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.432910553825949,
+      "learning_rate": 4.590305344315967e-06,
+      "loss": 0.4492,
+      "step": 6880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9681775975346065,
+      "learning_rate": 4.589037837307873e-06,
+      "loss": 0.4485,
+      "step": 6881
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5555539969433061,
+      "learning_rate": 4.587770356888493e-06,
+      "loss": 0.495,
+      "step": 6882
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 5.711790421635941,
+      "learning_rate": 4.586502903139832e-06,
+      "loss": 0.4337,
+      "step": 6883
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8138988737524109,
+      "learning_rate": 4.58523547614389e-06,
+      "loss": 0.5258,
+      "step": 6884
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7587329648056718,
+      "learning_rate": 4.583968075982673e-06,
+      "loss": 0.4747,
+      "step": 6885
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3877595104767442,
+      "learning_rate": 4.582700702738175e-06,
+      "loss": 0.4658,
+      "step": 6886
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.9825157930378736,
+      "learning_rate": 4.581433356492398e-06,
+      "loss": 0.4557,
+      "step": 6887
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5748570059722728,
+      "learning_rate": 4.580166037327333e-06,
+      "loss": 0.4792,
+      "step": 6888
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.414922906016234,
+      "learning_rate": 4.578898745324979e-06,
+      "loss": 0.4497,
+      "step": 6889
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.744678109463946,
+      "learning_rate": 4.577631480567321e-06,
+      "loss": 0.4797,
+      "step": 6890
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5570940624037628,
+      "learning_rate": 4.576364243136356e-06,
+      "loss": 0.5089,
+      "step": 6891
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.555639996945321,
+      "learning_rate": 4.575097033114072e-06,
+      "loss": 0.4748,
+      "step": 6892
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8279227444374277,
+      "learning_rate": 4.573829850582452e-06,
+      "loss": 0.5094,
+      "step": 6893
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.39196254048585,
+      "learning_rate": 4.572562695623484e-06,
+      "loss": 0.4737,
+      "step": 6894
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.136720802809768,
+      "learning_rate": 4.57129556831915e-06,
+      "loss": 0.5107,
+      "step": 6895
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.046214988195398,
+      "learning_rate": 4.570028468751433e-06,
+      "loss": 0.4564,
+      "step": 6896
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.3321259841230355,
+      "learning_rate": 4.56876139700231e-06,
+      "loss": 0.4316,
+      "step": 6897
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4726934909524818,
+      "learning_rate": 4.567494353153762e-06,
+      "loss": 0.4212,
+      "step": 6898
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.586878323423012,
+      "learning_rate": 4.566227337287762e-06,
+      "loss": 0.4773,
+      "step": 6899
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.785230054256795,
+      "learning_rate": 4.564960349486287e-06,
+      "loss": 0.4626,
+      "step": 6900
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8403708151543154,
+      "learning_rate": 4.5636933898313074e-06,
+      "loss": 0.5085,
+      "step": 6901
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6771924354772159,
+      "learning_rate": 4.562426458404796e-06,
+      "loss": 0.4834,
+      "step": 6902
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.924181589842709,
+      "learning_rate": 4.561159555288719e-06,
+      "loss": 0.5192,
+      "step": 6903
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4954968456682984,
+      "learning_rate": 4.559892680565044e-06,
+      "loss": 0.5159,
+      "step": 6904
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6186225089464956,
+      "learning_rate": 4.558625834315739e-06,
+      "loss": 0.4858,
+      "step": 6905
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4914130666393217,
+      "learning_rate": 4.557359016622763e-06,
+      "loss": 0.4077,
+      "step": 6906
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4530549492009845,
+      "learning_rate": 4.5560922275680814e-06,
+      "loss": 0.4763,
+      "step": 6907
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.617721556564137,
+      "learning_rate": 4.554825467233651e-06,
+      "loss": 0.4933,
+      "step": 6908
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.681113496045325,
+      "learning_rate": 4.553558735701431e-06,
+      "loss": 0.4465,
+      "step": 6909
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.666711126781402,
+      "learning_rate": 4.5522920330533764e-06,
+      "loss": 0.4675,
+      "step": 6910
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.7042054083469744,
+      "learning_rate": 4.551025359371443e-06,
+      "loss": 0.439,
+      "step": 6911
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.51054173942164,
+      "learning_rate": 4.549758714737579e-06,
+      "loss": 0.4941,
+      "step": 6912
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.2054036541484865,
+      "learning_rate": 4.5484920992337396e-06,
+      "loss": 0.494,
+      "step": 6913
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5786589308505222,
+      "learning_rate": 4.547225512941869e-06,
+      "loss": 0.4735,
+      "step": 6914
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6938909617642963,
+      "learning_rate": 4.545958955943915e-06,
+      "loss": 0.479,
+      "step": 6915
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5828264050227292,
+      "learning_rate": 4.544692428321825e-06,
+      "loss": 0.4711,
+      "step": 6916
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5881486421574957,
+      "learning_rate": 4.543425930157537e-06,
+      "loss": 0.4173,
+      "step": 6917
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3934223903811027,
+      "learning_rate": 4.542159461532995e-06,
+      "loss": 0.5163,
+      "step": 6918
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.281474288352167,
+      "learning_rate": 4.540893022530136e-06,
+      "loss": 0.4359,
+      "step": 6919
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8234038311110423,
+      "learning_rate": 4.5396266132309e-06,
+      "loss": 0.4916,
+      "step": 6920
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8191465599331846,
+      "learning_rate": 4.538360233717217e-06,
+      "loss": 0.4679,
+      "step": 6921
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2887056839190372,
+      "learning_rate": 4.537093884071027e-06,
+      "loss": 0.4704,
+      "step": 6922
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.0184185518797135,
+      "learning_rate": 4.5358275643742525e-06,
+      "loss": 0.451,
+      "step": 6923
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.770487264005411,
+      "learning_rate": 4.5345612747088316e-06,
+      "loss": 0.4341,
+      "step": 6924
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.4927178602069326,
+      "learning_rate": 4.533295015156684e-06,
+      "loss": 0.4655,
+      "step": 6925
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6847378809485796,
+      "learning_rate": 4.532028785799739e-06,
+      "loss": 0.4896,
+      "step": 6926
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4750236358136959,
+      "learning_rate": 4.530762586719921e-06,
+      "loss": 0.4463,
+      "step": 6927
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.556879108868717,
+      "learning_rate": 4.529496417999148e-06,
+      "loss": 0.4485,
+      "step": 6928
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5752369288621986,
+      "learning_rate": 4.528230279719343e-06,
+      "loss": 0.4911,
+      "step": 6929
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.372174342060887,
+      "learning_rate": 4.526964171962421e-06,
+      "loss": 0.425,
+      "step": 6930
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.58420486517682,
+      "learning_rate": 4.525698094810298e-06,
+      "loss": 0.4566,
+      "step": 6931
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6198401184184732,
+      "learning_rate": 4.524432048344888e-06,
+      "loss": 0.4911,
+      "step": 6932
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 2.618657202365816,
+      "learning_rate": 4.523166032648103e-06,
+      "loss": 0.4361,
+      "step": 6933
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8440512826291275,
+      "learning_rate": 4.52190004780185e-06,
+      "loss": 0.4385,
+      "step": 6934
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5382671248848278,
+      "learning_rate": 4.520634093888041e-06,
+      "loss": 0.4789,
+      "step": 6935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6616880070028293,
+      "learning_rate": 4.519368170988577e-06,
+      "loss": 0.4841,
+      "step": 6936
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6681698573778345,
+      "learning_rate": 4.5181022791853625e-06,
+      "loss": 0.4499,
+      "step": 6937
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6266530546604019,
+      "learning_rate": 4.516836418560303e-06,
+      "loss": 0.4645,
+      "step": 6938
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5440259675687322,
+      "learning_rate": 4.515570589195293e-06,
+      "loss": 0.5027,
+      "step": 6939
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.538233418974356,
+      "learning_rate": 4.514304791172235e-06,
+      "loss": 0.4601,
+      "step": 6940
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9546840067155016,
+      "learning_rate": 4.5130390245730186e-06,
+      "loss": 0.5228,
+      "step": 6941
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.691768844924657,
+      "learning_rate": 4.511773289479542e-06,
+      "loss": 0.4666,
+      "step": 6942
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5209016847795411,
+      "learning_rate": 4.510507585973694e-06,
+      "loss": 0.5041,
+      "step": 6943
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 4.005861846814587,
+      "learning_rate": 4.509241914137366e-06,
+      "loss": 0.4226,
+      "step": 6944
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5464427142356353,
+      "learning_rate": 4.507976274052443e-06,
+      "loss": 0.4855,
+      "step": 6945
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8246453506367573,
+      "learning_rate": 4.506710665800813e-06,
+      "loss": 0.4982,
+      "step": 6946
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5243098931004129,
+      "learning_rate": 4.505445089464356e-06,
+      "loss": 0.4512,
+      "step": 6947
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8258556937394714,
+      "learning_rate": 4.504179545124955e-06,
+      "loss": 0.5055,
+      "step": 6948
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.462991577680172,
+      "learning_rate": 4.502914032864489e-06,
+      "loss": 0.4371,
+      "step": 6949
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 7.620715617054399,
+      "learning_rate": 4.501648552764835e-06,
+      "loss": 0.4615,
+      "step": 6950
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.587625363549638,
+      "learning_rate": 4.500383104907868e-06,
+      "loss": 0.4848,
+      "step": 6951
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.1367492692551684,
+      "learning_rate": 4.49911768937546e-06,
+      "loss": 0.4616,
+      "step": 6952
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.946007805748415,
+      "learning_rate": 4.497852306249483e-06,
+      "loss": 0.4956,
+      "step": 6953
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7293702983786041,
+      "learning_rate": 4.496586955611802e-06,
+      "loss": 0.4076,
+      "step": 6954
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5696280959653361,
+      "learning_rate": 4.495321637544291e-06,
+      "loss": 0.4974,
+      "step": 6955
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 16.306800375051598,
+      "learning_rate": 4.494056352128806e-06,
+      "loss": 0.4712,
+      "step": 6956
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3982803426114867,
+      "learning_rate": 4.492791099447217e-06,
+      "loss": 0.4713,
+      "step": 6957
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.004766336615196,
+      "learning_rate": 4.491525879581377e-06,
+      "loss": 0.506,
+      "step": 6958
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.789386428490523,
+      "learning_rate": 4.490260692613148e-06,
+      "loss": 0.4176,
+      "step": 6959
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2389206006700233,
+      "learning_rate": 4.488995538624386e-06,
+      "loss": 0.4833,
+      "step": 6960
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9403510699497735,
+      "learning_rate": 4.487730417696943e-06,
+      "loss": 0.4669,
+      "step": 6961
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6295469606366729,
+      "learning_rate": 4.486465329912674e-06,
+      "loss": 0.4937,
+      "step": 6962
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9905055158988696,
+      "learning_rate": 4.485200275353425e-06,
+      "loss": 0.5231,
+      "step": 6963
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7255830095970441,
+      "learning_rate": 4.483935254101046e-06,
+      "loss": 0.4271,
+      "step": 6964
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5928382705433373,
+      "learning_rate": 4.482670266237379e-06,
+      "loss": 0.4343,
+      "step": 6965
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.235279830406769,
+      "learning_rate": 4.481405311844271e-06,
+      "loss": 0.4133,
+      "step": 6966
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.508964086715466,
+      "learning_rate": 4.480140391003559e-06,
+      "loss": 0.445,
+      "step": 6967
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9155834441610042,
+      "learning_rate": 4.478875503797085e-06,
+      "loss": 0.4114,
+      "step": 6968
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4269378564844795,
+      "learning_rate": 4.477610650306683e-06,
+      "loss": 0.475,
+      "step": 6969
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.939330205162762,
+      "learning_rate": 4.476345830614188e-06,
+      "loss": 0.4399,
+      "step": 6970
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.488488174458895,
+      "learning_rate": 4.475081044801436e-06,
+      "loss": 0.4543,
+      "step": 6971
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5662209261700896,
+      "learning_rate": 4.47381629295025e-06,
+      "loss": 0.505,
+      "step": 6972
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.558954580909388,
+      "learning_rate": 4.472551575142465e-06,
+      "loss": 0.5051,
+      "step": 6973
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3189435923958883,
+      "learning_rate": 4.4712868914599e-06,
+      "loss": 0.4557,
+      "step": 6974
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.819227641448975,
+      "learning_rate": 4.4700222419843845e-06,
+      "loss": 0.4536,
+      "step": 6975
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.381196353971081,
+      "learning_rate": 4.468757626797734e-06,
+      "loss": 0.4577,
+      "step": 6976
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.062259964954439,
+      "learning_rate": 4.467493045981771e-06,
+      "loss": 0.5397,
+      "step": 6977
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7104861853068443,
+      "learning_rate": 4.466228499618311e-06,
+      "loss": 0.4395,
+      "step": 6978
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4806804549512607,
+      "learning_rate": 4.46496398778917e-06,
+      "loss": 0.434,
+      "step": 6979
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7863206435797854,
+      "learning_rate": 4.463699510576158e-06,
+      "loss": 0.5026,
+      "step": 6980
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.6334116227778526,
+      "learning_rate": 4.462435068061087e-06,
+      "loss": 0.4278,
+      "step": 6981
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8767313198898707,
+      "learning_rate": 4.4611706603257626e-06,
+      "loss": 0.4088,
+      "step": 6982
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.014384796636381,
+      "learning_rate": 4.459906287451992e-06,
+      "loss": 0.4658,
+      "step": 6983
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7769979380813947,
+      "learning_rate": 4.458641949521579e-06,
+      "loss": 0.4076,
+      "step": 6984
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5930019757817278,
+      "learning_rate": 4.457377646616322e-06,
+      "loss": 0.4406,
+      "step": 6985
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4434108320394223,
+      "learning_rate": 4.456113378818023e-06,
+      "loss": 0.4821,
+      "step": 6986
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5448205135034024,
+      "learning_rate": 4.454849146208476e-06,
+      "loss": 0.4805,
+      "step": 6987
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.658915191855445,
+      "learning_rate": 4.453584948869477e-06,
+      "loss": 0.4704,
+      "step": 6988
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.1978233164686958,
+      "learning_rate": 4.452320786882817e-06,
+      "loss": 0.4078,
+      "step": 6989
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5753607851255316,
+      "learning_rate": 4.451056660330285e-06,
+      "loss": 0.4612,
+      "step": 6990
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5527879649433624,
+      "learning_rate": 4.449792569293668e-06,
+      "loss": 0.5073,
+      "step": 6991
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7596119004707476,
+      "learning_rate": 4.448528513854754e-06,
+      "loss": 0.4597,
+      "step": 6992
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5887849738727466,
+      "learning_rate": 4.447264494095322e-06,
+      "loss": 0.4883,
+      "step": 6993
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5310426503157457,
+      "learning_rate": 4.446000510097154e-06,
+      "loss": 0.4826,
+      "step": 6994
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.669443625810114,
+      "learning_rate": 4.444736561942029e-06,
+      "loss": 0.4431,
+      "step": 6995
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5326041118385999,
+      "learning_rate": 4.443472649711723e-06,
+      "loss": 0.4654,
+      "step": 6996
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2613484326602404,
+      "learning_rate": 4.442208773488008e-06,
+      "loss": 0.4193,
+      "step": 6997
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7478284875690078,
+      "learning_rate": 4.440944933352655e-06,
+      "loss": 0.4683,
+      "step": 6998
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5501222089889063,
+      "learning_rate": 4.439681129387435e-06,
+      "loss": 0.4901,
+      "step": 6999
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9580923340483176,
+      "learning_rate": 4.438417361674111e-06,
+      "loss": 0.5518,
+      "step": 7000
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8072670405626892,
+      "learning_rate": 4.437153630294451e-06,
+      "loss": 0.4217,
+      "step": 7001
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5984986696400904,
+      "learning_rate": 4.435889935330213e-06,
+      "loss": 0.4898,
+      "step": 7002
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5214374782539432,
+      "learning_rate": 4.4346262768631595e-06,
+      "loss": 0.4878,
+      "step": 7003
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3691561953284572,
+      "learning_rate": 4.433362654975046e-06,
+      "loss": 0.4028,
+      "step": 7004
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9227485148097987,
+      "learning_rate": 4.432099069747625e-06,
+      "loss": 0.4632,
+      "step": 7005
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6774087672073437,
+      "learning_rate": 4.430835521262656e-06,
+      "loss": 0.4643,
+      "step": 7006
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8440707679471362,
+      "learning_rate": 4.42957200960188e-06,
+      "loss": 0.4422,
+      "step": 7007
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8272773348456313,
+      "learning_rate": 4.428308534847052e-06,
+      "loss": 0.4797,
+      "step": 7008
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4642179347199777,
+      "learning_rate": 4.427045097079911e-06,
+      "loss": 0.4412,
+      "step": 7009
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4985959737363224,
+      "learning_rate": 4.4257816963822055e-06,
+      "loss": 0.4294,
+      "step": 7010
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9232557455029031,
+      "learning_rate": 4.424518332835671e-06,
+      "loss": 0.4512,
+      "step": 7011
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.842755491135707,
+      "learning_rate": 4.4232550065220485e-06,
+      "loss": 0.457,
+      "step": 7012
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6432091979194465,
+      "learning_rate": 4.421991717523072e-06,
+      "loss": 0.3988,
+      "step": 7013
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.659266061750174,
+      "learning_rate": 4.420728465920477e-06,
+      "loss": 0.4617,
+      "step": 7014
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.008098713971611,
+      "learning_rate": 4.41946525179599e-06,
+      "loss": 0.5058,
+      "step": 7015
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9006370172418146,
+      "learning_rate": 4.418202075231342e-06,
+      "loss": 0.4986,
+      "step": 7016
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2162076712388235,
+      "learning_rate": 4.4169389363082605e-06,
+      "loss": 0.467,
+      "step": 7017
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5622565699910569,
+      "learning_rate": 4.4156758351084656e-06,
+      "loss": 0.4976,
+      "step": 7018
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3861916469046607,
+      "learning_rate": 4.414412771713681e-06,
+      "loss": 0.4211,
+      "step": 7019
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.913298354307848,
+      "learning_rate": 4.413149746205623e-06,
+      "loss": 0.4593,
+      "step": 7020
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5785481273394253,
+      "learning_rate": 4.411886758666009e-06,
+      "loss": 0.4766,
+      "step": 7021
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5036550785119829,
+      "learning_rate": 4.410623809176552e-06,
+      "loss": 0.4661,
+      "step": 7022
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8089720384128842,
+      "learning_rate": 4.409360897818964e-06,
+      "loss": 0.477,
+      "step": 7023
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4807022139068908,
+      "learning_rate": 4.408098024674953e-06,
+      "loss": 0.4796,
+      "step": 7024
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5555595705544788,
+      "learning_rate": 4.406835189826227e-06,
+      "loss": 0.495,
+      "step": 7025
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.0281205678241614,
+      "learning_rate": 4.405572393354485e-06,
+      "loss": 0.4244,
+      "step": 7026
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.01926758861407,
+      "learning_rate": 4.404309635341434e-06,
+      "loss": 0.4594,
+      "step": 7027
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6257582121771132,
+      "learning_rate": 4.4030469158687696e-06,
+      "loss": 0.4162,
+      "step": 7028
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5818089571998044,
+      "learning_rate": 4.401784235018188e-06,
+      "loss": 0.4838,
+      "step": 7029
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.296009331881997,
+      "learning_rate": 4.400521592871386e-06,
+      "loss": 0.4484,
+      "step": 7030
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6467673236672709,
+      "learning_rate": 4.399258989510052e-06,
+      "loss": 0.4588,
+      "step": 7031
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.0995597367065977,
+      "learning_rate": 4.397996425015875e-06,
+      "loss": 0.4823,
+      "step": 7032
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.920309716629777,
+      "learning_rate": 4.396733899470541e-06,
+      "loss": 0.4552,
+      "step": 7033
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.088402400197526,
+      "learning_rate": 4.395471412955737e-06,
+      "loss": 0.4673,
+      "step": 7034
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4363641129344256,
+      "learning_rate": 4.394208965553139e-06,
+      "loss": 0.4446,
+      "step": 7035
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3770192617406876,
+      "learning_rate": 4.392946557344431e-06,
+      "loss": 0.4505,
+      "step": 7036
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.606612142661513,
+      "learning_rate": 4.391684188411284e-06,
+      "loss": 0.5244,
+      "step": 7037
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6183551326985346,
+      "learning_rate": 4.390421858835374e-06,
+      "loss": 0.4537,
+      "step": 7038
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.573086646102559,
+      "learning_rate": 4.389159568698375e-06,
+      "loss": 0.4092,
+      "step": 7039
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6477005313384241,
+      "learning_rate": 4.387897318081951e-06,
+      "loss": 0.4923,
+      "step": 7040
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 3.236368083172161,
+      "learning_rate": 4.386635107067772e-06,
+      "loss": 0.5005,
+      "step": 7041
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5121322760621845,
+      "learning_rate": 4.385372935737496e-06,
+      "loss": 0.4157,
+      "step": 7042
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6833511404846513,
+      "learning_rate": 4.38411080417279e-06,
+      "loss": 0.4636,
+      "step": 7043
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8877525530703354,
+      "learning_rate": 4.382848712455307e-06,
+      "loss": 0.4625,
+      "step": 7044
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3569894382538616,
+      "learning_rate": 4.381586660666707e-06,
+      "loss": 0.4516,
+      "step": 7045
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4125657630797133,
+      "learning_rate": 4.380324648888638e-06,
+      "loss": 0.4491,
+      "step": 7046
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.673965314766708,
+      "learning_rate": 4.379062677202757e-06,
+      "loss": 0.4686,
+      "step": 7047
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.7372410945554004,
+      "learning_rate": 4.3778007456907065e-06,
+      "loss": 0.5016,
+      "step": 7048
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9089767498205799,
+      "learning_rate": 4.376538854434135e-06,
+      "loss": 0.452,
+      "step": 7049
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4544558627795818,
+      "learning_rate": 4.375277003514683e-06,
+      "loss": 0.4747,
+      "step": 7050
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.2327357536054557,
+      "learning_rate": 4.374015193013992e-06,
+      "loss": 0.4698,
+      "step": 7051
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3678180762884193,
+      "learning_rate": 4.372753423013699e-06,
+      "loss": 0.4496,
+      "step": 7052
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.810492063544269,
+      "learning_rate": 4.3714916935954386e-06,
+      "loss": 0.5069,
+      "step": 7053
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5625807655030113,
+      "learning_rate": 4.370230004840845e-06,
+      "loss": 0.4635,
+      "step": 7054
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.3858176260745867,
+      "learning_rate": 4.368968356831544e-06,
+      "loss": 0.502,
+      "step": 7055
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.877182342981261,
+      "learning_rate": 4.367706749649167e-06,
+      "loss": 0.4614,
+      "step": 7056
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.481665538232791,
+      "learning_rate": 4.366445183375335e-06,
+      "loss": 0.5198,
+      "step": 7057
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.9267904034320824,
+      "learning_rate": 4.365183658091672e-06,
+      "loss": 0.443,
+      "step": 7058
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.8793435474565665,
+      "learning_rate": 4.363922173879794e-06,
+      "loss": 0.4244,
+      "step": 7059
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.469176425870163,
+      "learning_rate": 4.362660730821321e-06,
+      "loss": 0.4528,
+      "step": 7060
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 2.355820290937261,
+      "learning_rate": 4.361399328997862e-06,
+      "loss": 0.4565,
+      "step": 7061
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5947954498565637,
+      "learning_rate": 4.360137968491033e-06,
+      "loss": 0.4273,
+      "step": 7062
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.7571192028754195,
+      "learning_rate": 4.358876649382441e-06,
+      "loss": 0.4377,
+      "step": 7063
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5571334120862022,
+      "learning_rate": 4.35761537175369e-06,
+      "loss": 0.47,
+      "step": 7064
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9585484554976884,
+      "learning_rate": 4.356354135686385e-06,
+      "loss": 0.4779,
+      "step": 7065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.9224194520241955,
+      "learning_rate": 4.355092941262125e-06,
+      "loss": 0.4399,
+      "step": 7066
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3658057771066292,
+      "learning_rate": 4.353831788562509e-06,
+      "loss": 0.4911,
+      "step": 7067
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5592145227770972,
+      "learning_rate": 4.35257067766913e-06,
+      "loss": 0.4757,
+      "step": 7068
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5419519037460963,
+      "learning_rate": 4.351309608663582e-06,
+      "loss": 0.4881,
+      "step": 7069
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6856154431224528,
+      "learning_rate": 4.350048581627452e-06,
+      "loss": 0.4955,
+      "step": 7070
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2851496477520994,
+      "learning_rate": 4.348787596642331e-06,
+      "loss": 0.4363,
+      "step": 7071
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8153696492324172,
+      "learning_rate": 4.347526653789798e-06,
+      "loss": 0.4189,
+      "step": 7072
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3326952606353584,
+      "learning_rate": 4.346265753151438e-06,
+      "loss": 0.4503,
+      "step": 7073
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.459275596930889,
+      "learning_rate": 4.345004894808829e-06,
+      "loss": 0.4884,
+      "step": 7074
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.040050883432156,
+      "learning_rate": 4.343744078843545e-06,
+      "loss": 0.5096,
+      "step": 7075
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4599004057223703,
+      "learning_rate": 4.342483305337164e-06,
+      "loss": 0.4917,
+      "step": 7076
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.098386629632137,
+      "learning_rate": 4.3412225743712494e-06,
+      "loss": 0.4983,
+      "step": 7077
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.810049827702473,
+      "learning_rate": 4.339961886027376e-06,
+      "loss": 0.4596,
+      "step": 7078
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.448124245646561,
+      "learning_rate": 4.338701240387101e-06,
+      "loss": 0.425,
+      "step": 7079
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0835026753430106,
+      "learning_rate": 4.337440637531994e-06,
+      "loss": 0.4683,
+      "step": 7080
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.519222800713956,
+      "learning_rate": 4.3361800775436085e-06,
+      "loss": 0.4533,
+      "step": 7081
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4959999756829805,
+      "learning_rate": 4.334919560503506e-06,
+      "loss": 0.4889,
+      "step": 7082
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4383030012139646,
+      "learning_rate": 4.333659086493236e-06,
+      "loss": 0.483,
+      "step": 7083
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6084560745424255,
+      "learning_rate": 4.332398655594351e-06,
+      "loss": 0.4993,
+      "step": 7084
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7718959094126387,
+      "learning_rate": 4.331138267888402e-06,
+      "loss": 0.4761,
+      "step": 7085
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2973554425799056,
+      "learning_rate": 4.3298779234569305e-06,
+      "loss": 0.5093,
+      "step": 7086
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.636737030073066,
+      "learning_rate": 4.328617622381481e-06,
+      "loss": 0.4771,
+      "step": 7087
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6804304037646827,
+      "learning_rate": 4.327357364743593e-06,
+      "loss": 0.5048,
+      "step": 7088
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5669294703378361,
+      "learning_rate": 4.3260971506248045e-06,
+      "loss": 0.4847,
+      "step": 7089
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.524866791357876,
+      "learning_rate": 4.324836980106648e-06,
+      "loss": 0.4763,
+      "step": 7090
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.027185439879834,
+      "learning_rate": 4.323576853270656e-06,
+      "loss": 0.4378,
+      "step": 7091
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6335679443035642,
+      "learning_rate": 4.322316770198356e-06,
+      "loss": 0.4903,
+      "step": 7092
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5903735576675163,
+      "learning_rate": 4.321056730971275e-06,
+      "loss": 0.4761,
+      "step": 7093
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.646851041760132,
+      "learning_rate": 4.319796735670935e-06,
+      "loss": 0.456,
+      "step": 7094
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7000459523526177,
+      "learning_rate": 4.318536784378855e-06,
+      "loss": 0.4728,
+      "step": 7095
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5931789981034563,
+      "learning_rate": 4.317276877176556e-06,
+      "loss": 0.4354,
+      "step": 7096
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.973054816736263,
+      "learning_rate": 4.316017014145548e-06,
+      "loss": 0.4568,
+      "step": 7097
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.719179731251411,
+      "learning_rate": 4.314757195367347e-06,
+      "loss": 0.4776,
+      "step": 7098
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.721910820249672,
+      "learning_rate": 4.313497420923456e-06,
+      "loss": 0.4755,
+      "step": 7099
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5372784712846987,
+      "learning_rate": 4.312237690895387e-06,
+      "loss": 0.4872,
+      "step": 7100
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7039023374321425,
+      "learning_rate": 4.310978005364637e-06,
+      "loss": 0.4487,
+      "step": 7101
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4958597937073772,
+      "learning_rate": 4.309718364412711e-06,
+      "loss": 0.4449,
+      "step": 7102
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3645944456143195,
+      "learning_rate": 4.308458768121102e-06,
+      "loss": 0.4333,
+      "step": 7103
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6157152481916293,
+      "learning_rate": 4.307199216571307e-06,
+      "loss": 0.4271,
+      "step": 7104
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.320880319183414,
+      "learning_rate": 4.305939709844815e-06,
+      "loss": 0.4651,
+      "step": 7105
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7932893912409602,
+      "learning_rate": 4.304680248023117e-06,
+      "loss": 0.4396,
+      "step": 7106
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6533606693322573,
+      "learning_rate": 4.303420831187698e-06,
+      "loss": 0.4529,
+      "step": 7107
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.538407189673168,
+      "learning_rate": 4.302161459420037e-06,
+      "loss": 0.4801,
+      "step": 7108
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6166162681584009,
+      "learning_rate": 4.300902132801621e-06,
+      "loss": 0.3951,
+      "step": 7109
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.9283161285875248,
+      "learning_rate": 4.299642851413919e-06,
+      "loss": 0.4366,
+      "step": 7110
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.814997773410602,
+      "learning_rate": 4.2983836153384114e-06,
+      "loss": 0.4028,
+      "step": 7111
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8732772018217532,
+      "learning_rate": 4.297124424656563e-06,
+      "loss": 0.4106,
+      "step": 7112
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8022789492333757,
+      "learning_rate": 4.295865279449847e-06,
+      "loss": 0.3994,
+      "step": 7113
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.555915217888963,
+      "learning_rate": 4.2946061797997245e-06,
+      "loss": 0.4437,
+      "step": 7114
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6482567681059899,
+      "learning_rate": 4.293347125787662e-06,
+      "loss": 0.5175,
+      "step": 7115
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.0707033523533798,
+      "learning_rate": 4.292088117495113e-06,
+      "loss": 0.4841,
+      "step": 7116
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6075329077273628,
+      "learning_rate": 4.290829155003538e-06,
+      "loss": 0.4134,
+      "step": 7117
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.3139845191046,
+      "learning_rate": 4.28957023839439e-06,
+      "loss": 0.4303,
+      "step": 7118
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.426227007206049,
+      "learning_rate": 4.288311367749116e-06,
+      "loss": 0.4288,
+      "step": 7119
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.493381934248523,
+      "learning_rate": 4.287052543149167e-06,
+      "loss": 0.4484,
+      "step": 7120
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6787731203515182,
+      "learning_rate": 4.285793764675986e-06,
+      "loss": 0.4967,
+      "step": 7121
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.234573890119122,
+      "learning_rate": 4.284535032411014e-06,
+      "loss": 0.4838,
+      "step": 7122
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5796205847564913,
+      "learning_rate": 4.283276346435689e-06,
+      "loss": 0.4856,
+      "step": 7123
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6946004249080573,
+      "learning_rate": 4.282017706831447e-06,
+      "loss": 0.5004,
+      "step": 7124
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.111828690234118,
+      "learning_rate": 4.28075911367972e-06,
+      "loss": 0.3983,
+      "step": 7125
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.426555629132823,
+      "learning_rate": 4.279500567061938e-06,
+      "loss": 0.495,
+      "step": 7126
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.4244794007682464,
+      "learning_rate": 4.278242067059526e-06,
+      "loss": 0.4493,
+      "step": 7127
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.1633449489178704,
+      "learning_rate": 4.27698361375391e-06,
+      "loss": 0.454,
+      "step": 7128
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5962354948533055,
+      "learning_rate": 4.275725207226505e-06,
+      "loss": 0.5079,
+      "step": 7129
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.837982846000056,
+      "learning_rate": 4.274466847558733e-06,
+      "loss": 0.4961,
+      "step": 7130
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.527886933556026,
+      "learning_rate": 4.273208534832007e-06,
+      "loss": 0.4819,
+      "step": 7131
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.5239023235675115,
+      "learning_rate": 4.271950269127738e-06,
+      "loss": 0.5115,
+      "step": 7132
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5724784306377685,
+      "learning_rate": 4.270692050527335e-06,
+      "loss": 0.4704,
+      "step": 7133
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.949473922851969,
+      "learning_rate": 4.269433879112202e-06,
+      "loss": 0.4463,
+      "step": 7134
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 3.50625765584057,
+      "learning_rate": 4.268175754963741e-06,
+      "loss": 0.4303,
+      "step": 7135
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8915366186292353,
+      "learning_rate": 4.266917678163352e-06,
+      "loss": 0.4944,
+      "step": 7136
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7606990101713447,
+      "learning_rate": 4.26565964879243e-06,
+      "loss": 0.494,
+      "step": 7137
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.32874710581035,
+      "learning_rate": 4.264401666932367e-06,
+      "loss": 0.4527,
+      "step": 7138
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8136147198991983,
+      "learning_rate": 4.2631437326645555e-06,
+      "loss": 0.4568,
+      "step": 7139
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2163663109038967,
+      "learning_rate": 4.261885846070379e-06,
+      "loss": 0.4927,
+      "step": 7140
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5366903757711854,
+      "learning_rate": 4.260628007231223e-06,
+      "loss": 0.4439,
+      "step": 7141
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6128422517537353,
+      "learning_rate": 4.259370216228468e-06,
+      "loss": 0.44,
+      "step": 7142
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5422867976860997,
+      "learning_rate": 4.258112473143491e-06,
+      "loss": 0.4931,
+      "step": 7143
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.732283222605643,
+      "learning_rate": 4.256854778057668e-06,
+      "loss": 0.5267,
+      "step": 7144
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.598296699258677,
+      "learning_rate": 4.2555971310523655e-06,
+      "loss": 0.4724,
+      "step": 7145
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5270406538508988,
+      "learning_rate": 4.254339532208959e-06,
+      "loss": 0.4824,
+      "step": 7146
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.570551625454146,
+      "learning_rate": 4.253081981608805e-06,
+      "loss": 0.4756,
+      "step": 7147
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3306062359476223,
+      "learning_rate": 4.251824479333273e-06,
+      "loss": 0.4501,
+      "step": 7148
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7208407446721028,
+      "learning_rate": 4.250567025463716e-06,
+      "loss": 0.5568,
+      "step": 7149
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5618051072099315,
+      "learning_rate": 4.249309620081495e-06,
+      "loss": 0.4916,
+      "step": 7150
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.041586722877766,
+      "learning_rate": 4.248052263267956e-06,
+      "loss": 0.5151,
+      "step": 7151
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7005550799666431,
+      "learning_rate": 4.246794955104453e-06,
+      "loss": 0.4633,
+      "step": 7152
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3551752685434306,
+      "learning_rate": 4.245537695672333e-06,
+      "loss": 0.4632,
+      "step": 7153
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5549663044760179,
+      "learning_rate": 4.244280485052935e-06,
+      "loss": 0.4941,
+      "step": 7154
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8302813717164201,
+      "learning_rate": 4.2430233233276026e-06,
+      "loss": 0.4849,
+      "step": 7155
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6175272060955515,
+      "learning_rate": 4.2417662105776695e-06,
+      "loss": 0.4571,
+      "step": 7156
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3630483389353352,
+      "learning_rate": 4.240509146884472e-06,
+      "loss": 0.4231,
+      "step": 7157
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.341332577101578,
+      "learning_rate": 4.239252132329338e-06,
+      "loss": 0.4687,
+      "step": 7158
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8412649990272256,
+      "learning_rate": 4.237995166993598e-06,
+      "loss": 0.4495,
+      "step": 7159
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5902555942383094,
+      "learning_rate": 4.236738250958571e-06,
+      "loss": 0.51,
+      "step": 7160
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.53911026827657,
+      "learning_rate": 4.235481384305583e-06,
+      "loss": 0.4452,
+      "step": 7161
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9798016316404117,
+      "learning_rate": 4.234224567115948e-06,
+      "loss": 0.4536,
+      "step": 7162
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.2786078273549752,
+      "learning_rate": 4.2329677994709805e-06,
+      "loss": 0.4736,
+      "step": 7163
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.61142115541771,
+      "learning_rate": 4.231711081451997e-06,
+      "loss": 0.451,
+      "step": 7164
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.629607635428878,
+      "learning_rate": 4.230454413140297e-06,
+      "loss": 0.4419,
+      "step": 7165
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6069100954455453,
+      "learning_rate": 4.229197794617192e-06,
+      "loss": 0.5143,
+      "step": 7166
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9731422168325872,
+      "learning_rate": 4.227941225963981e-06,
+      "loss": 0.467,
+      "step": 7167
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.440688698761986,
+      "learning_rate": 4.226684707261963e-06,
+      "loss": 0.4393,
+      "step": 7168
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5777641425416326,
+      "learning_rate": 4.225428238592433e-06,
+      "loss": 0.4814,
+      "step": 7169
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5786916309508443,
+      "learning_rate": 4.2241718200366824e-06,
+      "loss": 0.4889,
+      "step": 7170
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6835076191670824,
+      "learning_rate": 4.222915451675999e-06,
+      "loss": 0.4227,
+      "step": 7171
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7560958312115467,
+      "learning_rate": 4.221659133591672e-06,
+      "loss": 0.4367,
+      "step": 7172
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.7899144033972494,
+      "learning_rate": 4.220402865864979e-06,
+      "loss": 0.5037,
+      "step": 7173
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4668440822962623,
+      "learning_rate": 4.2191466485772e-06,
+      "loss": 0.528,
+      "step": 7174
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7694986247838935,
+      "learning_rate": 4.217890481809613e-06,
+      "loss": 0.4321,
+      "step": 7175
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6802665156694991,
+      "learning_rate": 4.216634365643488e-06,
+      "loss": 0.4337,
+      "step": 7176
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.290977307441879,
+      "learning_rate": 4.215378300160095e-06,
+      "loss": 0.406,
+      "step": 7177
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6057793011525513,
+      "learning_rate": 4.2141222854407e-06,
+      "loss": 0.4897,
+      "step": 7178
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3561512078015183,
+      "learning_rate": 4.212866321566564e-06,
+      "loss": 0.4598,
+      "step": 7179
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5737466134148299,
+      "learning_rate": 4.2116104086189466e-06,
+      "loss": 0.5004,
+      "step": 7180
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6331235011133518,
+      "learning_rate": 4.2103545466791074e-06,
+      "loss": 0.4741,
+      "step": 7181
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 4.412003853584198,
+      "learning_rate": 4.209098735828293e-06,
+      "loss": 0.4528,
+      "step": 7182
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7275176501531586,
+      "learning_rate": 4.207842976147758e-06,
+      "loss": 0.4694,
+      "step": 7183
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6301747302058787,
+      "learning_rate": 4.206587267718743e-06,
+      "loss": 0.4834,
+      "step": 7184
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.6640047079820801,
+      "learning_rate": 4.205331610622496e-06,
+      "loss": 0.4692,
+      "step": 7185
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.286084659455096,
+      "learning_rate": 4.204076004940255e-06,
+      "loss": 0.4951,
+      "step": 7186
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.546893798102047,
+      "learning_rate": 4.202820450753255e-06,
+      "loss": 0.4712,
+      "step": 7187
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.4878529962976126,
+      "learning_rate": 4.2015649481427285e-06,
+      "loss": 0.4408,
+      "step": 7188
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.164570868973618,
+      "learning_rate": 4.200309497189905e-06,
+      "loss": 0.4744,
+      "step": 7189
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5107459330460842,
+      "learning_rate": 4.199054097976012e-06,
+      "loss": 0.4452,
+      "step": 7190
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4913902984697256,
+      "learning_rate": 4.19779875058227e-06,
+      "loss": 0.441,
+      "step": 7191
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7601764914183795,
+      "learning_rate": 4.1965434550899e-06,
+      "loss": 0.4825,
+      "step": 7192
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5750663937481285,
+      "learning_rate": 4.195288211580118e-06,
+      "loss": 0.4873,
+      "step": 7193
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9728166575549422,
+      "learning_rate": 4.194033020134135e-06,
+      "loss": 0.4849,
+      "step": 7194
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4359490813360012,
+      "learning_rate": 4.192777880833161e-06,
+      "loss": 0.485,
+      "step": 7195
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.367135646028408,
+      "learning_rate": 4.1915227937584015e-06,
+      "loss": 0.431,
+      "step": 7196
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2703325081079995,
+      "learning_rate": 4.190267758991062e-06,
+      "loss": 0.4773,
+      "step": 7197
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.491102761889611,
+      "learning_rate": 4.189012776612337e-06,
+      "loss": 0.4861,
+      "step": 7198
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3692832381163065,
+      "learning_rate": 4.187757846703427e-06,
+      "loss": 0.4279,
+      "step": 7199
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7482550726188164,
+      "learning_rate": 4.186502969345518e-06,
+      "loss": 0.3971,
+      "step": 7200
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.227964683480586,
+      "learning_rate": 4.185248144619806e-06,
+      "loss": 0.4646,
+      "step": 7201
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6120483074020207,
+      "learning_rate": 4.183993372607472e-06,
+      "loss": 0.4614,
+      "step": 7202
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6203608377282328,
+      "learning_rate": 4.1827386533897e-06,
+      "loss": 0.481,
+      "step": 7203
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5741994293492492,
+      "learning_rate": 4.181483987047666e-06,
+      "loss": 0.4657,
+      "step": 7204
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5819616942552157,
+      "learning_rate": 4.18022937366255e-06,
+      "loss": 0.4305,
+      "step": 7205
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9052278979378736,
+      "learning_rate": 4.178974813315519e-06,
+      "loss": 0.4474,
+      "step": 7206
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3999564200050365,
+      "learning_rate": 4.177720306087745e-06,
+      "loss": 0.4733,
+      "step": 7207
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4349190492553616,
+      "learning_rate": 4.176465852060391e-06,
+      "loss": 0.4726,
+      "step": 7208
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5976258482918875,
+      "learning_rate": 4.175211451314619e-06,
+      "loss": 0.4322,
+      "step": 7209
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4175416498897417,
+      "learning_rate": 4.173957103931587e-06,
+      "loss": 0.4666,
+      "step": 7210
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2402160660989288,
+      "learning_rate": 4.17270280999245e-06,
+      "loss": 0.5288,
+      "step": 7211
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0565465602893727,
+      "learning_rate": 4.171448569578359e-06,
+      "loss": 0.4578,
+      "step": 7212
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2889545798991457,
+      "learning_rate": 4.170194382770462e-06,
+      "loss": 0.3994,
+      "step": 7213
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5417510364901649,
+      "learning_rate": 4.168940249649902e-06,
+      "loss": 0.4914,
+      "step": 7214
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6067848747787655,
+      "learning_rate": 4.16768617029782e-06,
+      "loss": 0.483,
+      "step": 7215
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.558924458905801,
+      "learning_rate": 4.166432144795357e-06,
+      "loss": 0.4517,
+      "step": 7216
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3056153487433872,
+      "learning_rate": 4.16517817322364e-06,
+      "loss": 0.4244,
+      "step": 7217
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6203790482880163,
+      "learning_rate": 4.1639242556638075e-06,
+      "loss": 0.4614,
+      "step": 7218
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5866996507146631,
+      "learning_rate": 4.162670392196978e-06,
+      "loss": 0.4761,
+      "step": 7219
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5944572471266427,
+      "learning_rate": 4.161416582904279e-06,
+      "loss": 0.5057,
+      "step": 7220
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8848243412834218,
+      "learning_rate": 4.160162827866832e-06,
+      "loss": 0.5233,
+      "step": 7221
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5436304846627225,
+      "learning_rate": 4.15890912716575e-06,
+      "loss": 0.4648,
+      "step": 7222
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8487228727618306,
+      "learning_rate": 4.157655480882149e-06,
+      "loss": 0.514,
+      "step": 7223
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.62845003223565,
+      "learning_rate": 4.1564018890971345e-06,
+      "loss": 0.4783,
+      "step": 7224
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.326197393556375,
+      "learning_rate": 4.1551483518918156e-06,
+      "loss": 0.3681,
+      "step": 7225
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7320646277997784,
+      "learning_rate": 4.153894869347292e-06,
+      "loss": 0.5074,
+      "step": 7226
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.920768701154002,
+      "learning_rate": 4.152641441544665e-06,
+      "loss": 0.4316,
+      "step": 7227
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6106112115824267,
+      "learning_rate": 4.151388068565026e-06,
+      "loss": 0.4662,
+      "step": 7228
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5536814263242582,
+      "learning_rate": 4.150134750489471e-06,
+      "loss": 0.432,
+      "step": 7229
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.267552671542952,
+      "learning_rate": 4.148881487399085e-06,
+      "loss": 0.4673,
+      "step": 7230
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5251450140465543,
+      "learning_rate": 4.1476282793749515e-06,
+      "loss": 0.5443,
+      "step": 7231
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.570071021459001,
+      "learning_rate": 4.1463751264981574e-06,
+      "loss": 0.4859,
+      "step": 7232
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9499691663920258,
+      "learning_rate": 4.145122028849772e-06,
+      "loss": 0.4509,
+      "step": 7233
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5637653900421415,
+      "learning_rate": 4.143868986510876e-06,
+      "loss": 0.4681,
+      "step": 7234
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3360038123002915,
+      "learning_rate": 4.142615999562534e-06,
+      "loss": 0.4285,
+      "step": 7235
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5607455807744609,
+      "learning_rate": 4.141363068085817e-06,
+      "loss": 0.4696,
+      "step": 7236
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.1869152469017368,
+      "learning_rate": 4.140110192161786e-06,
+      "loss": 0.5326,
+      "step": 7237
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 4.193590935499073,
+      "learning_rate": 4.138857371871501e-06,
+      "loss": 0.4332,
+      "step": 7238
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.079653613191426,
+      "learning_rate": 4.137604607296016e-06,
+      "loss": 0.4867,
+      "step": 7239
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6934470574676732,
+      "learning_rate": 4.1363518985163865e-06,
+      "loss": 0.5073,
+      "step": 7240
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.82202470824787,
+      "learning_rate": 4.1350992456136586e-06,
+      "loss": 0.5266,
+      "step": 7241
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7153789490952156,
+      "learning_rate": 4.133846648668878e-06,
+      "loss": 0.4607,
+      "step": 7242
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5595554969703158,
+      "learning_rate": 4.132594107763087e-06,
+      "loss": 0.4944,
+      "step": 7243
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.185435782620074,
+      "learning_rate": 4.13134162297732e-06,
+      "loss": 0.4775,
+      "step": 7244
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5411136030467448,
+      "learning_rate": 4.130089194392617e-06,
+      "loss": 0.4598,
+      "step": 7245
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9342805837479653,
+      "learning_rate": 4.128836822090003e-06,
+      "loss": 0.4144,
+      "step": 7246
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.3018552170539945,
+      "learning_rate": 4.127584506150508e-06,
+      "loss": 0.3836,
+      "step": 7247
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5652795675424047,
+      "learning_rate": 4.126332246655154e-06,
+      "loss": 0.4889,
+      "step": 7248
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0155641280240064,
+      "learning_rate": 4.125080043684961e-06,
+      "loss": 0.4723,
+      "step": 7249
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.632886268271774,
+      "learning_rate": 4.1238278973209426e-06,
+      "loss": 0.4377,
+      "step": 7250
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.564861750390497,
+      "learning_rate": 4.122575807644116e-06,
+      "loss": 0.5127,
+      "step": 7251
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9712034567633971,
+      "learning_rate": 4.121323774735484e-06,
+      "loss": 0.4741,
+      "step": 7252
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5387590534814723,
+      "learning_rate": 4.120071798676055e-06,
+      "loss": 0.4635,
+      "step": 7253
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.566008860908701,
+      "learning_rate": 4.118819879546832e-06,
+      "loss": 0.4623,
+      "step": 7254
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5599071387118866,
+      "learning_rate": 4.117568017428807e-06,
+      "loss": 0.4865,
+      "step": 7255
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.101934518954288,
+      "learning_rate": 4.116316212402979e-06,
+      "loss": 0.427,
+      "step": 7256
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7828228437772888,
+      "learning_rate": 4.115064464550335e-06,
+      "loss": 0.4653,
+      "step": 7257
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7394057271031385,
+      "learning_rate": 4.113812773951863e-06,
+      "loss": 0.5187,
+      "step": 7258
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8281299954911372,
+      "learning_rate": 4.112561140688545e-06,
+      "loss": 0.4617,
+      "step": 7259
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7795973261585343,
+      "learning_rate": 4.111309564841361e-06,
+      "loss": 0.5197,
+      "step": 7260
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7591658904923522,
+      "learning_rate": 4.1100580464912845e-06,
+      "loss": 0.4864,
+      "step": 7261
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5556895708101333,
+      "learning_rate": 4.108806585719289e-06,
+      "loss": 0.4281,
+      "step": 7262
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5576692292947365,
+      "learning_rate": 4.107555182606341e-06,
+      "loss": 0.4414,
+      "step": 7263
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3711751625582287,
+      "learning_rate": 4.106303837233404e-06,
+      "loss": 0.462,
+      "step": 7264
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9369489098015318,
+      "learning_rate": 4.105052549681443e-06,
+      "loss": 0.4331,
+      "step": 7265
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.14714813316451,
+      "learning_rate": 4.103801320031407e-06,
+      "loss": 0.4718,
+      "step": 7266
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.7741689964401735,
+      "learning_rate": 4.1025501483642575e-06,
+      "loss": 0.4145,
+      "step": 7267
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7222433273465352,
+      "learning_rate": 4.101299034760936e-06,
+      "loss": 0.4111,
+      "step": 7268
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8236979486369775,
+      "learning_rate": 4.100047979302395e-06,
+      "loss": 0.4717,
+      "step": 7269
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6095616761489393,
+      "learning_rate": 4.098796982069568e-06,
+      "loss": 0.4731,
+      "step": 7270
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4960794121352596,
+      "learning_rate": 4.0975460431434006e-06,
+      "loss": 0.4733,
+      "step": 7271
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3625963282404576,
+      "learning_rate": 4.096295162604822e-06,
+      "loss": 0.4282,
+      "step": 7272
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.632710106823323,
+      "learning_rate": 4.095044340534766e-06,
+      "loss": 0.5286,
+      "step": 7273
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5714894172128924,
+      "learning_rate": 4.093793577014156e-06,
+      "loss": 0.4437,
+      "step": 7274
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2254748193516987,
+      "learning_rate": 4.0925428721239166e-06,
+      "loss": 0.4749,
+      "step": 7275
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5507882722707433,
+      "learning_rate": 4.091292225944965e-06,
+      "loss": 0.4624,
+      "step": 7276
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5911538084966974,
+      "learning_rate": 4.090041638558218e-06,
+      "loss": 0.5021,
+      "step": 7277
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5518925948511184,
+      "learning_rate": 4.0887911100445885e-06,
+      "loss": 0.4474,
+      "step": 7278
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.164721432401165,
+      "learning_rate": 4.08754064048498e-06,
+      "loss": 0.4958,
+      "step": 7279
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.568996595220086,
+      "learning_rate": 4.0862902299603e-06,
+      "loss": 0.4322,
+      "step": 7280
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.41196396552437,
+      "learning_rate": 4.0850398785514454e-06,
+      "loss": 0.5127,
+      "step": 7281
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5456935298686468,
+      "learning_rate": 4.083789586339315e-06,
+      "loss": 0.4994,
+      "step": 7282
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6836367250278734,
+      "learning_rate": 4.082539353404799e-06,
+      "loss": 0.4527,
+      "step": 7283
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.683483572937276,
+      "learning_rate": 4.081289179828787e-06,
+      "loss": 0.4481,
+      "step": 7284
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5610040995362815,
+      "learning_rate": 4.080039065692162e-06,
+      "loss": 0.4865,
+      "step": 7285
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.215831774679217,
+      "learning_rate": 4.0787890110758085e-06,
+      "loss": 0.4678,
+      "step": 7286
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9827336352012541,
+      "learning_rate": 4.077539016060598e-06,
+      "loss": 0.4529,
+      "step": 7287
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.725589906867415,
+      "learning_rate": 4.076289080727408e-06,
+      "loss": 0.4962,
+      "step": 7288
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.996218397553376,
+      "learning_rate": 4.075039205157107e-06,
+      "loss": 0.5031,
+      "step": 7289
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.886631401910825,
+      "learning_rate": 4.073789389430557e-06,
+      "loss": 0.4944,
+      "step": 7290
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5410462303540111,
+      "learning_rate": 4.072539633628624e-06,
+      "loss": 0.504,
+      "step": 7291
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5161100302776698,
+      "learning_rate": 4.071289937832163e-06,
+      "loss": 0.3853,
+      "step": 7292
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5175882798239,
+      "learning_rate": 4.0700403021220285e-06,
+      "loss": 0.4794,
+      "step": 7293
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5602935665775335,
+      "learning_rate": 4.0687907265790685e-06,
+      "loss": 0.4728,
+      "step": 7294
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5668635839465165,
+      "learning_rate": 4.067541211284131e-06,
+      "loss": 0.469,
+      "step": 7295
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.552276164470864,
+      "learning_rate": 4.066291756318058e-06,
+      "loss": 0.4492,
+      "step": 7296
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9416798460456859,
+      "learning_rate": 4.065042361761687e-06,
+      "loss": 0.4423,
+      "step": 7297
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8885804351222533,
+      "learning_rate": 4.063793027695851e-06,
+      "loss": 0.4691,
+      "step": 7298
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.562851289644021,
+      "learning_rate": 4.06254375420138e-06,
+      "loss": 0.4952,
+      "step": 7299
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5749357483899877,
+      "learning_rate": 4.0612945413591055e-06,
+      "loss": 0.4379,
+      "step": 7300
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.4771389656381997,
+      "learning_rate": 4.060045389249843e-06,
+      "loss": 0.4289,
+      "step": 7301
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6126208812506262,
+      "learning_rate": 4.058796297954417e-06,
+      "loss": 0.44,
+      "step": 7302
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0144264387376807,
+      "learning_rate": 4.057547267553636e-06,
+      "loss": 0.4565,
+      "step": 7303
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8842739185812187,
+      "learning_rate": 4.056298298128318e-06,
+      "loss": 0.4255,
+      "step": 7304
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5603887924998544,
+      "learning_rate": 4.055049389759261e-06,
+      "loss": 0.5003,
+      "step": 7305
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3477901475875183,
+      "learning_rate": 4.053800542527277e-06,
+      "loss": 0.4605,
+      "step": 7306
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.6678273227006093,
+      "learning_rate": 4.052551756513156e-06,
+      "loss": 0.4717,
+      "step": 7307
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.7958078206603296,
+      "learning_rate": 4.051303031797699e-06,
+      "loss": 0.4259,
+      "step": 7308
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9493410891755092,
+      "learning_rate": 4.050054368461695e-06,
+      "loss": 0.483,
+      "step": 7309
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.59653465259678,
+      "learning_rate": 4.048805766585929e-06,
+      "loss": 0.4676,
+      "step": 7310
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6114293408826292,
+      "learning_rate": 4.047557226251187e-06,
+      "loss": 0.4959,
+      "step": 7311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8201546935081434,
+      "learning_rate": 4.0463087475382464e-06,
+      "loss": 0.4607,
+      "step": 7312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.173987259415849,
+      "learning_rate": 4.045060330527882e-06,
+      "loss": 0.4955,
+      "step": 7313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5549504089358277,
+      "learning_rate": 4.043811975300865e-06,
+      "loss": 0.4701,
+      "step": 7314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.8694713696144272,
+      "learning_rate": 4.0425636819379635e-06,
+      "loss": 0.4317,
+      "step": 7315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.2687829176791734,
+      "learning_rate": 4.041315450519937e-06,
+      "loss": 0.4527,
+      "step": 7316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5552242515860327,
+      "learning_rate": 4.0400672811275495e-06,
+      "loss": 0.5012,
+      "step": 7317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4604753482531572,
+      "learning_rate": 4.038819173841551e-06,
+      "loss": 0.4958,
+      "step": 7318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.6604820135143221,
+      "learning_rate": 4.037571128742696e-06,
+      "loss": 0.4358,
+      "step": 7319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 3.7078393657615045,
+      "learning_rate": 4.036323145911728e-06,
+      "loss": 0.4584,
+      "step": 7320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4537142441522966,
+      "learning_rate": 4.0350752254293914e-06,
+      "loss": 0.4108,
+      "step": 7321
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1624037004743024,
+      "learning_rate": 4.033827367376428e-06,
+      "loss": 0.516,
+      "step": 7322
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6426444613575368,
+      "learning_rate": 4.032579571833568e-06,
+      "loss": 0.4209,
+      "step": 7323
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7029471153652824,
+      "learning_rate": 4.0313318388815465e-06,
+      "loss": 0.5057,
+      "step": 7324
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.991998869316725,
+      "learning_rate": 4.030084168601086e-06,
+      "loss": 0.485,
+      "step": 7325
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.7997079900013677,
+      "learning_rate": 4.028836561072912e-06,
+      "loss": 0.4812,
+      "step": 7326
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.306296173867573,
+      "learning_rate": 4.027589016377741e-06,
+      "loss": 0.4613,
+      "step": 7327
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9150623181305038,
+      "learning_rate": 4.02634153459629e-06,
+      "loss": 0.4577,
+      "step": 7328
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1240719616016066,
+      "learning_rate": 4.025094115809267e-06,
+      "loss": 0.447,
+      "step": 7329
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3433655595156815,
+      "learning_rate": 4.02384676009738e-06,
+      "loss": 0.4225,
+      "step": 7330
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1588471871939694,
+      "learning_rate": 4.0225994675413296e-06,
+      "loss": 0.5034,
+      "step": 7331
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.199053713949702,
+      "learning_rate": 4.021352238221815e-06,
+      "loss": 0.477,
+      "step": 7332
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3233652467101629,
+      "learning_rate": 4.020105072219532e-06,
+      "loss": 0.4709,
+      "step": 7333
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4812575063998095,
+      "learning_rate": 4.018857969615166e-06,
+      "loss": 0.4569,
+      "step": 7334
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.707740175299555,
+      "learning_rate": 4.01761093048941e-06,
+      "loss": 0.4825,
+      "step": 7335
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7652809021015952,
+      "learning_rate": 4.016363954922937e-06,
+      "loss": 0.4721,
+      "step": 7336
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5401002420630759,
+      "learning_rate": 4.015117042996434e-06,
+      "loss": 0.4763,
+      "step": 7337
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6298520674197519,
+      "learning_rate": 4.013870194790566e-06,
+      "loss": 0.4381,
+      "step": 7338
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.394556113438474,
+      "learning_rate": 4.01262341038601e-06,
+      "loss": 0.4272,
+      "step": 7339
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8679451800133462,
+      "learning_rate": 4.011376689863425e-06,
+      "loss": 0.4894,
+      "step": 7340
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5649150563107403,
+      "learning_rate": 4.010130033303477e-06,
+      "loss": 0.508,
+      "step": 7341
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6952363652683937,
+      "learning_rate": 4.008883440786817e-06,
+      "loss": 0.453,
+      "step": 7342
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.806630162561481,
+      "learning_rate": 4.007636912394105e-06,
+      "loss": 0.427,
+      "step": 7343
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5493566111932915,
+      "learning_rate": 4.006390448205987e-06,
+      "loss": 0.4371,
+      "step": 7344
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.534355913169727,
+      "learning_rate": 4.005144048303105e-06,
+      "loss": 0.477,
+      "step": 7345
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5141934498753955,
+      "learning_rate": 4.003897712766104e-06,
+      "loss": 0.4321,
+      "step": 7346
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.1092214984871536,
+      "learning_rate": 4.0026514416756165e-06,
+      "loss": 0.4149,
+      "step": 7347
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5744244289322404,
+      "learning_rate": 4.001405235112277e-06,
+      "loss": 0.4496,
+      "step": 7348
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7818362511447288,
+      "learning_rate": 4.000159093156711e-06,
+      "loss": 0.437,
+      "step": 7349
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4416057643849776,
+      "learning_rate": 3.998913015889545e-06,
+      "loss": 0.4822,
+      "step": 7350
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.06404069743296,
+      "learning_rate": 3.997667003391397e-06,
+      "loss": 0.4986,
+      "step": 7351
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3856399361748397,
+      "learning_rate": 3.996421055742883e-06,
+      "loss": 0.4414,
+      "step": 7352
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9298855227941383,
+      "learning_rate": 3.995175173024612e-06,
+      "loss": 0.3937,
+      "step": 7353
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6264945281713359,
+      "learning_rate": 3.993929355317195e-06,
+      "loss": 0.4737,
+      "step": 7354
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5806961659514345,
+      "learning_rate": 3.992683602701231e-06,
+      "loss": 0.523,
+      "step": 7355
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4466296892700827,
+      "learning_rate": 3.99143791525732e-06,
+      "loss": 0.4494,
+      "step": 7356
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.745918551511774,
+      "learning_rate": 3.990192293066058e-06,
+      "loss": 0.4591,
+      "step": 7357
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0836660489384053,
+      "learning_rate": 3.988946736208032e-06,
+      "loss": 0.4703,
+      "step": 7358
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.736680521513632,
+      "learning_rate": 3.987701244763832e-06,
+      "loss": 0.4602,
+      "step": 7359
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5655580771775796,
+      "learning_rate": 3.986455818814035e-06,
+      "loss": 0.5003,
+      "step": 7360
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9039211818376327,
+      "learning_rate": 3.985210458439223e-06,
+      "loss": 0.4334,
+      "step": 7361
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.3885942210133853,
+      "learning_rate": 3.983965163719965e-06,
+      "loss": 0.4598,
+      "step": 7362
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9334385763128896,
+      "learning_rate": 3.982719934736832e-06,
+      "loss": 0.4344,
+      "step": 7363
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.476164058384664,
+      "learning_rate": 3.981474771570389e-06,
+      "loss": 0.4474,
+      "step": 7364
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5551385182157621,
+      "learning_rate": 3.980229674301195e-06,
+      "loss": 0.4983,
+      "step": 7365
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3820578936824226,
+      "learning_rate": 3.978984643009808e-06,
+      "loss": 0.409,
+      "step": 7366
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.07386479517845,
+      "learning_rate": 3.977739677776777e-06,
+      "loss": 0.4401,
+      "step": 7367
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6337458859869038,
+      "learning_rate": 3.976494778682654e-06,
+      "loss": 0.4322,
+      "step": 7368
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6588109571780973,
+      "learning_rate": 3.975249945807978e-06,
+      "loss": 0.4139,
+      "step": 7369
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8717544611136825,
+      "learning_rate": 3.974005179233292e-06,
+      "loss": 0.453,
+      "step": 7370
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.92536532668072,
+      "learning_rate": 3.972760479039126e-06,
+      "loss": 0.4912,
+      "step": 7371
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.680482447478772,
+      "learning_rate": 3.971515845306017e-06,
+      "loss": 0.4624,
+      "step": 7372
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5547983591456187,
+      "learning_rate": 3.9702712781144826e-06,
+      "loss": 0.4744,
+      "step": 7373
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.99321163940407,
+      "learning_rate": 3.969026777545054e-06,
+      "loss": 0.4546,
+      "step": 7374
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5559816123711208,
+      "learning_rate": 3.96778234367824e-06,
+      "loss": 0.481,
+      "step": 7375
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3959258871192257,
+      "learning_rate": 3.966537976594563e-06,
+      "loss": 0.4817,
+      "step": 7376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5223533166703613,
+      "learning_rate": 3.965293676374523e-06,
+      "loss": 0.4529,
+      "step": 7377
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.094086026677986,
+      "learning_rate": 3.964049443098629e-06,
+      "loss": 0.4648,
+      "step": 7378
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5606546078042821,
+      "learning_rate": 3.962805276847382e-06,
+      "loss": 0.4859,
+      "step": 7379
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5491341187675134,
+      "learning_rate": 3.961561177701276e-06,
+      "loss": 0.4922,
+      "step": 7380
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8849146384387017,
+      "learning_rate": 3.960317145740805e-06,
+      "loss": 0.5117,
+      "step": 7381
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6206771971854343,
+      "learning_rate": 3.959073181046452e-06,
+      "loss": 0.4358,
+      "step": 7382
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.686896264527124,
+      "learning_rate": 3.957829283698705e-06,
+      "loss": 0.437,
+      "step": 7383
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5263655166412793,
+      "learning_rate": 3.956585453778038e-06,
+      "loss": 0.4709,
+      "step": 7384
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7984144835085387,
+      "learning_rate": 3.955341691364929e-06,
+      "loss": 0.4568,
+      "step": 7385
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9101647841574672,
+      "learning_rate": 3.9540979965398445e-06,
+      "loss": 0.4686,
+      "step": 7386
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7023352465794623,
+      "learning_rate": 3.952854369383253e-06,
+      "loss": 0.464,
+      "step": 7387
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5672244174543317,
+      "learning_rate": 3.951610809975613e-06,
+      "loss": 0.4684,
+      "step": 7388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.516838826504019,
+      "learning_rate": 3.950367318397379e-06,
+      "loss": 0.4615,
+      "step": 7389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 5.355924837053419,
+      "learning_rate": 3.9491238947290104e-06,
+      "loss": 0.4884,
+      "step": 7390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8338954509615495,
+      "learning_rate": 3.94788053905095e-06,
+      "loss": 0.4822,
+      "step": 7391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0904145606678006,
+      "learning_rate": 3.946637251443643e-06,
+      "loss": 0.4648,
+      "step": 7392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0547077805103315,
+      "learning_rate": 3.945394031987527e-06,
+      "loss": 0.4547,
+      "step": 7393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6762631939562462,
+      "learning_rate": 3.944150880763039e-06,
+      "loss": 0.472,
+      "step": 7394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5963394429899332,
+      "learning_rate": 3.942907797850606e-06,
+      "loss": 0.4875,
+      "step": 7395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7565173718106364,
+      "learning_rate": 3.941664783330657e-06,
+      "loss": 0.4729,
+      "step": 7396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6259672795588251,
+      "learning_rate": 3.94042183728361e-06,
+      "loss": 0.4935,
+      "step": 7397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5656003893098618,
+      "learning_rate": 3.939178959789887e-06,
+      "loss": 0.4199,
+      "step": 7398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.407504884006315,
+      "learning_rate": 3.937936150929895e-06,
+      "loss": 0.4535,
+      "step": 7399
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4894989002334302,
+      "learning_rate": 3.936693410784045e-06,
+      "loss": 0.4515,
+      "step": 7400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8397592191244374,
+      "learning_rate": 3.935450739432742e-06,
+      "loss": 0.4618,
+      "step": 7401
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5360415190509557,
+      "learning_rate": 3.934208136956383e-06,
+      "loss": 0.4837,
+      "step": 7402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5440287464466432,
+      "learning_rate": 3.932965603435362e-06,
+      "loss": 0.4779,
+      "step": 7403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.705848416704902,
+      "learning_rate": 3.93172313895007e-06,
+      "loss": 0.4531,
+      "step": 7404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.530021351918056,
+      "learning_rate": 3.930480743580897e-06,
+      "loss": 0.5052,
+      "step": 7405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.759331779180002,
+      "learning_rate": 3.929238417408218e-06,
+      "loss": 0.5238,
+      "step": 7406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8854550609759009,
+      "learning_rate": 3.927996160512414e-06,
+      "loss": 0.4621,
+      "step": 7407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5450834459604268,
+      "learning_rate": 3.926753972973854e-06,
+      "loss": 0.4596,
+      "step": 7408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.542596227896359,
+      "learning_rate": 3.925511854872911e-06,
+      "loss": 0.4796,
+      "step": 7409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7769510210967858,
+      "learning_rate": 3.9242698062899426e-06,
+      "loss": 0.4768,
+      "step": 7410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2841289599112335,
+      "learning_rate": 3.923027827305311e-06,
+      "loss": 0.4545,
+      "step": 7411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.476618576837062,
+      "learning_rate": 3.921785917999372e-06,
+      "loss": 0.4665,
+      "step": 7412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4013387474006558,
+      "learning_rate": 3.9205440784524715e-06,
+      "loss": 0.4267,
+      "step": 7413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5322642253464693,
+      "learning_rate": 3.919302308744958e-06,
+      "loss": 0.4807,
+      "step": 7414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.845680264830862,
+      "learning_rate": 3.918060608957172e-06,
+      "loss": 0.46,
+      "step": 7415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0994706228664586,
+      "learning_rate": 3.916818979169448e-06,
+      "loss": 0.4633,
+      "step": 7416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5313591808671198,
+      "learning_rate": 3.91557741946212e-06,
+      "loss": 0.4793,
+      "step": 7417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5704990884260895,
+      "learning_rate": 3.914335929915514e-06,
+      "loss": 0.4725,
+      "step": 7418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 4.378855037092292,
+      "learning_rate": 3.913094510609952e-06,
+      "loss": 0.4749,
+      "step": 7419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.708631827529763,
+      "learning_rate": 3.911853161625756e-06,
+      "loss": 0.4787,
+      "step": 7420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.340800081032865,
+      "learning_rate": 3.9106118830432346e-06,
+      "loss": 0.4954,
+      "step": 7421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7729519923980588,
+      "learning_rate": 3.909370674942699e-06,
+      "loss": 0.5125,
+      "step": 7422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 6.615303762388371,
+      "learning_rate": 3.908129537404456e-06,
+      "loss": 0.4645,
+      "step": 7423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9404998545379093,
+      "learning_rate": 3.906888470508801e-06,
+      "loss": 0.4911,
+      "step": 7424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.634515374564921,
+      "learning_rate": 3.9056474743360345e-06,
+      "loss": 0.5132,
+      "step": 7425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.596996189807429,
+      "learning_rate": 3.904406548966443e-06,
+      "loss": 0.4815,
+      "step": 7426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.688113542860332,
+      "learning_rate": 3.903165694480316e-06,
+      "loss": 0.443,
+      "step": 7427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5774763756783123,
+      "learning_rate": 3.901924910957934e-06,
+      "loss": 0.4195,
+      "step": 7428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5793112422412106,
+      "learning_rate": 3.9006841984795725e-06,
+      "loss": 0.4919,
+      "step": 7429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.031415912084839,
+      "learning_rate": 3.899443557125506e-06,
+      "loss": 0.4767,
+      "step": 7430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5501852732548653,
+      "learning_rate": 3.898202986976002e-06,
+      "loss": 0.444,
+      "step": 7431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5932210997781622,
+      "learning_rate": 3.896962488111323e-06,
+      "loss": 0.5189,
+      "step": 7432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5147729157791983,
+      "learning_rate": 3.895722060611728e-06,
+      "loss": 0.4992,
+      "step": 7433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5939284624807348,
+      "learning_rate": 3.894481704557471e-06,
+      "loss": 0.4616,
+      "step": 7434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2330129640225005,
+      "learning_rate": 3.8932414200288e-06,
+      "loss": 0.4624,
+      "step": 7435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6348468329315848,
+      "learning_rate": 3.892001207105965e-06,
+      "loss": 0.4238,
+      "step": 7436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.8384944408194006,
+      "learning_rate": 3.890761065869199e-06,
+      "loss": 0.4268,
+      "step": 7437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.7689207505901194,
+      "learning_rate": 3.889520996398743e-06,
+      "loss": 0.4561,
+      "step": 7438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.9721861800982003,
+      "learning_rate": 3.888280998774823e-06,
+      "loss": 0.4631,
+      "step": 7439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.351536012948005,
+      "learning_rate": 3.887041073077672e-06,
+      "loss": 0.4422,
+      "step": 7440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.983599362758474,
+      "learning_rate": 3.8858012193875044e-06,
+      "loss": 0.441,
+      "step": 7441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.0298333927005463,
+      "learning_rate": 3.884561437784544e-06,
+      "loss": 0.441,
+      "step": 7442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.878374278875591,
+      "learning_rate": 3.883321728348995e-06,
+      "loss": 0.4953,
+      "step": 7443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.405790356535479,
+      "learning_rate": 3.8820820911610735e-06,
+      "loss": 0.4719,
+      "step": 7444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6118297026644891,
+      "learning_rate": 3.880842526300975e-06,
+      "loss": 0.4844,
+      "step": 7445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 3.142676041920264,
+      "learning_rate": 3.879603033848902e-06,
+      "loss": 0.4014,
+      "step": 7446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5857458353346452,
+      "learning_rate": 3.878363613885048e-06,
+      "loss": 0.435,
+      "step": 7447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.552226407449032,
+      "learning_rate": 3.877124266489599e-06,
+      "loss": 0.4859,
+      "step": 7448
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8333780580199772,
+      "learning_rate": 3.875884991742744e-06,
+      "loss": 0.4587,
+      "step": 7449
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.32552714330887,
+      "learning_rate": 3.874645789724656e-06,
+      "loss": 0.4819,
+      "step": 7450
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3375572939013214,
+      "learning_rate": 3.873406660515516e-06,
+      "loss": 0.4838,
+      "step": 7451
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7782175147692776,
+      "learning_rate": 3.872167604195491e-06,
+      "loss": 0.4877,
+      "step": 7452
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.3954434491440453,
+      "learning_rate": 3.870928620844746e-06,
+      "loss": 0.5154,
+      "step": 7453
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.715273234288327,
+      "learning_rate": 3.869689710543442e-06,
+      "loss": 0.4448,
+      "step": 7454
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.556371677674079,
+      "learning_rate": 3.868450873371735e-06,
+      "loss": 0.4845,
+      "step": 7455
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7847706774020777,
+      "learning_rate": 3.8672121094097765e-06,
+      "loss": 0.4572,
+      "step": 7456
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9577540450012663,
+      "learning_rate": 3.865973418737711e-06,
+      "loss": 0.513,
+      "step": 7457
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7242423797310757,
+      "learning_rate": 3.864734801435686e-06,
+      "loss": 0.4845,
+      "step": 7458
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7153143557195472,
+      "learning_rate": 3.86349625758383e-06,
+      "loss": 0.4653,
+      "step": 7459
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8154174526227398,
+      "learning_rate": 3.862257787262283e-06,
+      "loss": 0.3947,
+      "step": 7460
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6110320094880846,
+      "learning_rate": 3.861019390551165e-06,
+      "loss": 0.4704,
+      "step": 7461
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.159400566809231,
+      "learning_rate": 3.859781067530605e-06,
+      "loss": 0.4998,
+      "step": 7462
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1453356253637037,
+      "learning_rate": 3.858542818280717e-06,
+      "loss": 0.4761,
+      "step": 7463
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4710743701490467,
+      "learning_rate": 3.857304642881617e-06,
+      "loss": 0.4898,
+      "step": 7464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5577366553377581,
+      "learning_rate": 3.85606654141341e-06,
+      "loss": 0.4664,
+      "step": 7465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8171179796873802,
+      "learning_rate": 3.854828513956204e-06,
+      "loss": 0.4188,
+      "step": 7466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5071250447056637,
+      "learning_rate": 3.853590560590093e-06,
+      "loss": 0.5,
+      "step": 7467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5365484925957675,
+      "learning_rate": 3.8523526813951726e-06,
+      "loss": 0.5401,
+      "step": 7468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.688150321189963,
+      "learning_rate": 3.851114876451533e-06,
+      "loss": 0.4435,
+      "step": 7469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7088859417091902,
+      "learning_rate": 3.849877145839258e-06,
+      "loss": 0.486,
+      "step": 7470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8653941969160455,
+      "learning_rate": 3.8486394896384285e-06,
+      "loss": 0.4473,
+      "step": 7471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5045525801628634,
+      "learning_rate": 3.8474019079291156e-06,
+      "loss": 0.4758,
+      "step": 7472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5475195448470077,
+      "learning_rate": 3.846164400791393e-06,
+      "loss": 0.4661,
+      "step": 7473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7219368540708582,
+      "learning_rate": 3.844926968305322e-06,
+      "loss": 0.4266,
+      "step": 7474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5260964229342389,
+      "learning_rate": 3.843689610550969e-06,
+      "loss": 0.4828,
+      "step": 7475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5768506244730163,
+      "learning_rate": 3.842452327608382e-06,
+      "loss": 0.478,
+      "step": 7476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5655196970847991,
+      "learning_rate": 3.841215119557619e-06,
+      "loss": 0.4644,
+      "step": 7477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7569177427503508,
+      "learning_rate": 3.839977986478718e-06,
+      "loss": 0.4411,
+      "step": 7478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 17.69881687539802,
+      "learning_rate": 3.838740928451724e-06,
+      "loss": 0.493,
+      "step": 7479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5822321813869796,
+      "learning_rate": 3.837503945556675e-06,
+      "loss": 0.4742,
+      "step": 7480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.416010237076649,
+      "learning_rate": 3.836267037873598e-06,
+      "loss": 0.4793,
+      "step": 7481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.941841473093685,
+      "learning_rate": 3.835030205482523e-06,
+      "loss": 0.4877,
+      "step": 7482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.4210351094703912,
+      "learning_rate": 3.833793448463469e-06,
+      "loss": 0.4359,
+      "step": 7483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7215623875315207,
+      "learning_rate": 3.8325567668964535e-06,
+      "loss": 0.4537,
+      "step": 7484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6726611613872002,
+      "learning_rate": 3.831320160861488e-06,
+      "loss": 0.4784,
+      "step": 7485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8718943853501355,
+      "learning_rate": 3.830083630438579e-06,
+      "loss": 0.4971,
+      "step": 7486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5493014932106179,
+      "learning_rate": 3.828847175707729e-06,
+      "loss": 0.4819,
+      "step": 7487
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.484999202973541,
+      "learning_rate": 3.827610796748935e-06,
+      "loss": 0.4788,
+      "step": 7488
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8519129174780453,
+      "learning_rate": 3.826374493642187e-06,
+      "loss": 0.4409,
+      "step": 7489
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3662615260466489,
+      "learning_rate": 3.825138266467474e-06,
+      "loss": 0.4509,
+      "step": 7490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.390835430706696,
+      "learning_rate": 3.823902115304781e-06,
+      "loss": 0.5058,
+      "step": 7491
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8684492377471882,
+      "learning_rate": 3.8226660402340796e-06,
+      "loss": 0.4868,
+      "step": 7492
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5466470788148817,
+      "learning_rate": 3.821430041335349e-06,
+      "loss": 0.472,
+      "step": 7493
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5140001405143653,
+      "learning_rate": 3.820194118688549e-06,
+      "loss": 0.4725,
+      "step": 7494
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5908308800759554,
+      "learning_rate": 3.818958272373649e-06,
+      "loss": 0.4996,
+      "step": 7495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0857593220473354,
+      "learning_rate": 3.817722502470601e-06,
+      "loss": 0.4328,
+      "step": 7496
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.552885503155174,
+      "learning_rate": 3.816486809059363e-06,
+      "loss": 0.4837,
+      "step": 7497
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7835564762817329,
+      "learning_rate": 3.8152511922198784e-06,
+      "loss": 0.4516,
+      "step": 7498
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5387163079151308,
+      "learning_rate": 3.8140156520320935e-06,
+      "loss": 0.4789,
+      "step": 7499
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7272360991393496,
+      "learning_rate": 3.812780188575943e-06,
+      "loss": 0.4514,
+      "step": 7500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.6382759784888545,
+      "learning_rate": 3.8115448019313627e-06,
+      "loss": 0.4662,
+      "step": 7501
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8987028427705963,
+      "learning_rate": 3.810309492178278e-06,
+      "loss": 0.417,
+      "step": 7502
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7446791221552287,
+      "learning_rate": 3.809074259396612e-06,
+      "loss": 0.451,
+      "step": 7503
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4492615465546963,
+      "learning_rate": 3.8078391036662853e-06,
+      "loss": 0.4619,
+      "step": 7504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.495003964791977,
+      "learning_rate": 3.806604025067208e-06,
+      "loss": 0.4099,
+      "step": 7505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8006795206588757,
+      "learning_rate": 3.80536902367929e-06,
+      "loss": 0.437,
+      "step": 7506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6269746363166553,
+      "learning_rate": 3.804134099582432e-06,
+      "loss": 0.4218,
+      "step": 7507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5231938858949519,
+      "learning_rate": 3.8028992528565353e-06,
+      "loss": 0.4254,
+      "step": 7508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5450010685127566,
+      "learning_rate": 3.801664483581488e-06,
+      "loss": 0.4586,
+      "step": 7509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.043629432296741,
+      "learning_rate": 3.8004297918371847e-06,
+      "loss": 0.4135,
+      "step": 7510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0453349676474017,
+      "learning_rate": 3.799195177703501e-06,
+      "loss": 0.4397,
+      "step": 7511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9554347083676054,
+      "learning_rate": 3.797960641260322e-06,
+      "loss": 0.4888,
+      "step": 7512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.546082363924803,
+      "learning_rate": 3.7967261825875134e-06,
+      "loss": 0.4887,
+      "step": 7513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.3524054125950715,
+      "learning_rate": 3.795491801764948e-06,
+      "loss": 0.4741,
+      "step": 7514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3866358505416807,
+      "learning_rate": 3.7942574988724884e-06,
+      "loss": 0.4694,
+      "step": 7515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6045931111969536,
+      "learning_rate": 3.793023273989991e-06,
+      "loss": 0.4918,
+      "step": 7516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.614239410821719,
+      "learning_rate": 3.7917891271973086e-06,
+      "loss": 0.443,
+      "step": 7517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.711853004485514,
+      "learning_rate": 3.7905550585742888e-06,
+      "loss": 0.4609,
+      "step": 7518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5234959674040762,
+      "learning_rate": 3.7893210682007752e-06,
+      "loss": 0.4931,
+      "step": 7519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.466008169418806,
+      "learning_rate": 3.7880871561566047e-06,
+      "loss": 0.4933,
+      "step": 7520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.504890924902262,
+      "learning_rate": 3.7868533225216097e-06,
+      "loss": 0.5272,
+      "step": 7521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5312765398108208,
+      "learning_rate": 3.785619567375617e-06,
+      "loss": 0.4217,
+      "step": 7522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5325962467639115,
+      "learning_rate": 3.784385890798451e-06,
+      "loss": 0.5045,
+      "step": 7523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7512442740105427,
+      "learning_rate": 3.7831522928699268e-06,
+      "loss": 0.4251,
+      "step": 7524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6435841437486938,
+      "learning_rate": 3.781918773669856e-06,
+      "loss": 0.4728,
+      "step": 7525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.2660627458672944,
+      "learning_rate": 3.78068533327805e-06,
+      "loss": 0.4717,
+      "step": 7526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5254489562411702,
+      "learning_rate": 3.779451971774305e-06,
+      "loss": 0.4515,
+      "step": 7527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.122201481630115,
+      "learning_rate": 3.7782186892384237e-06,
+      "loss": 0.4786,
+      "step": 7528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6071317332954509,
+      "learning_rate": 3.776985485750192e-06,
+      "loss": 0.4917,
+      "step": 7529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9862601511739864,
+      "learning_rate": 3.7757523613894024e-06,
+      "loss": 0.4167,
+      "step": 7530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8094644472009187,
+      "learning_rate": 3.7745193162358307e-06,
+      "loss": 0.4658,
+      "step": 7531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.699572412339297,
+      "learning_rate": 3.773286350369258e-06,
+      "loss": 0.3809,
+      "step": 7532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.545231109578005,
+      "learning_rate": 3.7720534638694528e-06,
+      "loss": 0.5017,
+      "step": 7533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.608139285228216,
+      "learning_rate": 3.770820656816183e-06,
+      "loss": 0.4317,
+      "step": 7534
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.429044486095302,
+      "learning_rate": 3.7695879292892076e-06,
+      "loss": 0.4415,
+      "step": 7535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.72142634734567,
+      "learning_rate": 3.7683552813682832e-06,
+      "loss": 0.3902,
+      "step": 7536
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5539572115383771,
+      "learning_rate": 3.767122713133163e-06,
+      "loss": 0.4938,
+      "step": 7537
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9593632133811218,
+      "learning_rate": 3.765890224663589e-06,
+      "loss": 0.4668,
+      "step": 7538
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1967879909273456,
+      "learning_rate": 3.7646578160393036e-06,
+      "loss": 0.4349,
+      "step": 7539
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.027903174373566,
+      "learning_rate": 3.763425487340041e-06,
+      "loss": 0.4834,
+      "step": 7540
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1474045819181034,
+      "learning_rate": 3.7621932386455327e-06,
+      "loss": 0.4856,
+      "step": 7541
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.580774870613579,
+      "learning_rate": 3.7609610700355014e-06,
+      "loss": 0.4581,
+      "step": 7542
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.02621066151538,
+      "learning_rate": 3.7597289815896685e-06,
+      "loss": 0.4255,
+      "step": 7543
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.9599562980428238,
+      "learning_rate": 3.7584969733877467e-06,
+      "loss": 0.4301,
+      "step": 7544
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0818454639734445,
+      "learning_rate": 3.7572650455094494e-06,
+      "loss": 0.4945,
+      "step": 7545
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.8782414535697307,
+      "learning_rate": 3.7560331980344756e-06,
+      "loss": 0.4382,
+      "step": 7546
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8792891455135292,
+      "learning_rate": 3.754801431042527e-06,
+      "loss": 0.4489,
+      "step": 7547
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.6534409777082941,
+      "learning_rate": 3.753569744613298e-06,
+      "loss": 0.4199,
+      "step": 7548
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.5917312383605022,
+      "learning_rate": 3.7523381388264747e-06,
+      "loss": 0.4519,
+      "step": 7549
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5565645484542814,
+      "learning_rate": 3.7511066137617424e-06,
+      "loss": 0.4131,
+      "step": 7550
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7810368096934819,
+      "learning_rate": 3.749875169498778e-06,
+      "loss": 0.491,
+      "step": 7551
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.894550441998941,
+      "learning_rate": 3.748643806117256e-06,
+      "loss": 0.467,
+      "step": 7552
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.0001669928032313,
+      "learning_rate": 3.747412523696841e-06,
+      "loss": 0.4612,
+      "step": 7553
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.158579473814672,
+      "learning_rate": 3.746181322317199e-06,
+      "loss": 0.4467,
+      "step": 7554
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.004628338356797,
+      "learning_rate": 3.7449502020579838e-06,
+      "loss": 0.4946,
+      "step": 7555
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.235203119297206,
+      "learning_rate": 3.7437191629988497e-06,
+      "loss": 0.4834,
+      "step": 7556
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7976574735128181,
+      "learning_rate": 3.7424882052194413e-06,
+      "loss": 0.502,
+      "step": 7557
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5763262974118609,
+      "learning_rate": 3.7412573287994e-06,
+      "loss": 0.5,
+      "step": 7558
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.143622612775322,
+      "learning_rate": 3.740026533818366e-06,
+      "loss": 0.4505,
+      "step": 7559
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 3.0564485725564667,
+      "learning_rate": 3.7387958203559637e-06,
+      "loss": 0.4418,
+      "step": 7560
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.153090058794941,
+      "learning_rate": 3.737565188491825e-06,
+      "loss": 0.4672,
+      "step": 7561
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3828542087319728,
+      "learning_rate": 3.7363346383055644e-06,
+      "loss": 0.453,
+      "step": 7562
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.8189087695195227,
+      "learning_rate": 3.735104169876803e-06,
+      "loss": 0.4434,
+      "step": 7563
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7523173795846925,
+      "learning_rate": 3.7338737832851433e-06,
+      "loss": 0.4546,
+      "step": 7564
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.9032817399030804,
+      "learning_rate": 3.7326434786101975e-06,
+      "loss": 0.4607,
+      "step": 7565
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 4.218581318184691,
+      "learning_rate": 3.7314132559315565e-06,
+      "loss": 0.4295,
+      "step": 7566
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.2020124480799166,
+      "learning_rate": 3.730183115328821e-06,
+      "loss": 0.4468,
+      "step": 7567
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5727377795894238,
+      "learning_rate": 3.7289530568815756e-06,
+      "loss": 0.4934,
+      "step": 7568
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.1246769190061277,
+      "learning_rate": 3.727723080669405e-06,
+      "loss": 0.4806,
+      "step": 7569
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7364845315769442,
+      "learning_rate": 3.7264931867718874e-06,
+      "loss": 0.4584,
+      "step": 7570
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.7137726274535838,
+      "learning_rate": 3.725263375268594e-06,
+      "loss": 0.5082,
+      "step": 7571
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5499911796395616,
+      "learning_rate": 3.7240336462390934e-06,
+      "loss": 0.5242,
+      "step": 7572
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5545901741647533,
+      "learning_rate": 3.7228039997629454e-06,
+      "loss": 0.4944,
+      "step": 7573
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5499841977782937,
+      "learning_rate": 3.7215744359197094e-06,
+      "loss": 0.4596,
+      "step": 7574
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.011956204550276,
+      "learning_rate": 3.7203449547889337e-06,
+      "loss": 0.4785,
+      "step": 7575
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 2.620227431304135,
+      "learning_rate": 3.719115556450167e-06,
+      "loss": 0.4439,
+      "step": 7576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.192602117767563,
+      "learning_rate": 3.7178862409829462e-06,
+      "loss": 0.43,
+      "step": 7577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5603052883254783,
+      "learning_rate": 3.7166570084668103e-06,
+      "loss": 0.5065,
+      "step": 7578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8028491058702008,
+      "learning_rate": 3.715427858981285e-06,
+      "loss": 0.4964,
+      "step": 7579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5373297677139584,
+      "learning_rate": 3.7141987926058994e-06,
+      "loss": 0.4731,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6849411897059918,
+      "learning_rate": 3.712969809420167e-06,
+      "loss": 0.4768,
+      "step": 7581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0207992359401916,
+      "learning_rate": 3.711740909503605e-06,
+      "loss": 0.4563,
+      "step": 7582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5842188852604493,
+      "learning_rate": 3.710512092935722e-06,
+      "loss": 0.4106,
+      "step": 7583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.361761753527509,
+      "learning_rate": 3.7092833597960188e-06,
+      "loss": 0.3671,
+      "step": 7584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6195660825297324,
+      "learning_rate": 3.7080547101639953e-06,
+      "loss": 0.4922,
+      "step": 7585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5547700027767101,
+      "learning_rate": 3.70682614411914e-06,
+      "loss": 0.4974,
+      "step": 7586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9321264448281175,
+      "learning_rate": 3.7055976617409434e-06,
+      "loss": 0.484,
+      "step": 7587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5501168051819622,
+      "learning_rate": 3.7043692631088836e-06,
+      "loss": 0.4877,
+      "step": 7588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4935324643422738,
+      "learning_rate": 3.7031409483024383e-06,
+      "loss": 0.45,
+      "step": 7589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.391824485815515,
+      "learning_rate": 3.701912717401076e-06,
+      "loss": 0.457,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8481491332077131,
+      "learning_rate": 3.700684570484264e-06,
+      "loss": 0.4821,
+      "step": 7591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5923686483406625,
+      "learning_rate": 3.69945650763146e-06,
+      "loss": 0.4855,
+      "step": 7592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6077703828433607,
+      "learning_rate": 3.698228528922117e-06,
+      "loss": 0.4407,
+      "step": 7593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6431391995270086,
+      "learning_rate": 3.697000634435689e-06,
+      "loss": 0.4582,
+      "step": 7594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7161481929933124,
+      "learning_rate": 3.695772824251611e-06,
+      "loss": 0.4024,
+      "step": 7595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.273388934066274,
+      "learning_rate": 3.6945450984493292e-06,
+      "loss": 0.4705,
+      "step": 7596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7385253926499884,
+      "learning_rate": 3.6933174571082686e-06,
+      "loss": 0.4484,
+      "step": 7597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4980242811352524,
+      "learning_rate": 3.6920899003078626e-06,
+      "loss": 0.4486,
+      "step": 7598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5376661445788004,
+      "learning_rate": 3.6908624281275252e-06,
+      "loss": 0.4885,
+      "step": 7599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5182183205931765,
+      "learning_rate": 3.68963504064668e-06,
+      "loss": 0.4329,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.040766437166542,
+      "learning_rate": 3.6884077379447303e-06,
+      "loss": 0.4488,
+      "step": 7601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6861908034684918,
+      "learning_rate": 3.687180520101086e-06,
+      "loss": 0.4064,
+      "step": 7602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5336329362762963,
+      "learning_rate": 3.6859533871951436e-06,
+      "loss": 0.4938,
+      "step": 7603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5557794001520622,
+      "learning_rate": 3.684726339306298e-06,
+      "loss": 0.4482,
+      "step": 7604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6401115241273605,
+      "learning_rate": 3.6834993765139388e-06,
+      "loss": 0.4395,
+      "step": 7605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5280644345305249,
+      "learning_rate": 3.682272498897447e-06,
+      "loss": 0.4762,
+      "step": 7606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7972707938066323,
+      "learning_rate": 3.681045706536201e-06,
+      "loss": 0.4807,
+      "step": 7607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.77937635136034,
+      "learning_rate": 3.679818999509572e-06,
+      "loss": 0.4951,
+      "step": 7608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8948050707483497,
+      "learning_rate": 3.678592377896927e-06,
+      "loss": 0.4506,
+      "step": 7609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.04197366815816,
+      "learning_rate": 3.677365841777626e-06,
+      "loss": 0.4523,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.796247275760259,
+      "learning_rate": 3.6761393912310262e-06,
+      "loss": 0.4445,
+      "step": 7611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9844649408118051,
+      "learning_rate": 3.674913026336474e-06,
+      "loss": 0.482,
+      "step": 7612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5337023277196105,
+      "learning_rate": 3.6736867471733173e-06,
+      "loss": 0.4938,
+      "step": 7613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5957026544127116,
+      "learning_rate": 3.6724605538208926e-06,
+      "loss": 0.4912,
+      "step": 7614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3219845331608733,
+      "learning_rate": 3.6712344463585316e-06,
+      "loss": 0.5071,
+      "step": 7615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6563201032923618,
+      "learning_rate": 3.6700084248655653e-06,
+      "loss": 0.4947,
+      "step": 7616
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7591117866407757,
+      "learning_rate": 3.6687824894213136e-06,
+      "loss": 0.4537,
+      "step": 7617
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6005192138344955,
+      "learning_rate": 3.6675566401050944e-06,
+      "loss": 0.4949,
+      "step": 7618
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.641180484234126,
+      "learning_rate": 3.6663308769962168e-06,
+      "loss": 0.4143,
+      "step": 7619
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1313569069820075,
+      "learning_rate": 3.6651052001739874e-06,
+      "loss": 0.4465,
+      "step": 7620
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.189939333902669,
+      "learning_rate": 3.6638796097177044e-06,
+      "loss": 0.4445,
+      "step": 7621
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3100238558012776,
+      "learning_rate": 3.6626541057066634e-06,
+      "loss": 0.4891,
+      "step": 7622
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.471903935293427,
+      "learning_rate": 3.661428688220152e-06,
+      "loss": 0.4575,
+      "step": 7623
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5862996539903595,
+      "learning_rate": 3.660203357337454e-06,
+      "loss": 0.486,
+      "step": 7624
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.918577782731966,
+      "learning_rate": 3.658978113137845e-06,
+      "loss": 0.4822,
+      "step": 7625
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4679381538545826,
+      "learning_rate": 3.6577529557005975e-06,
+      "loss": 0.4385,
+      "step": 7626
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6907090477792461,
+      "learning_rate": 3.6565278851049803e-06,
+      "loss": 0.45,
+      "step": 7627
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6200342886226105,
+      "learning_rate": 3.6553029014302476e-06,
+      "loss": 0.4712,
+      "step": 7628
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.4922113654207108,
+      "learning_rate": 3.6540780047556624e-06,
+      "loss": 0.4301,
+      "step": 7629
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.8690414486335882,
+      "learning_rate": 3.652853195160466e-06,
+      "loss": 0.5125,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.230491976010944,
+      "learning_rate": 3.6516284727239094e-06,
+      "loss": 0.5053,
+      "step": 7631
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4495752858883177,
+      "learning_rate": 3.6504038375252228e-06,
+      "loss": 0.4868,
+      "step": 7632
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.755647708294134,
+      "learning_rate": 3.649179289643646e-06,
+      "loss": 0.4187,
+      "step": 7633
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.645839389034316,
+      "learning_rate": 3.647954829158398e-06,
+      "loss": 0.4644,
+      "step": 7634
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8180972266432045,
+      "learning_rate": 3.646730456148708e-06,
+      "loss": 0.4739,
+      "step": 7635
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9863988228909277,
+      "learning_rate": 3.6455061706937833e-06,
+      "loss": 0.5089,
+      "step": 7636
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6269204093773668,
+      "learning_rate": 3.6442819728728397e-06,
+      "loss": 0.5089,
+      "step": 7637
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.313315457876071,
+      "learning_rate": 3.6430578627650793e-06,
+      "loss": 0.5363,
+      "step": 7638
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5678640422719986,
+      "learning_rate": 3.641833840449699e-06,
+      "loss": 0.5112,
+      "step": 7639
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5569381234824795,
+      "learning_rate": 3.640609906005893e-06,
+      "loss": 0.4471,
+      "step": 7640
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.636436636698173,
+      "learning_rate": 3.6393860595128473e-06,
+      "loss": 0.423,
+      "step": 7641
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7017697145494652,
+      "learning_rate": 3.6381623010497447e-06,
+      "loss": 0.4687,
+      "step": 7642
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5275215880456167,
+      "learning_rate": 3.636938630695759e-06,
+      "loss": 0.4741,
+      "step": 7643
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6531210178449371,
+      "learning_rate": 3.6357150485300607e-06,
+      "loss": 0.4593,
+      "step": 7644
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5547175690167522,
+      "learning_rate": 3.634491554631814e-06,
+      "loss": 0.474,
+      "step": 7645
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5471233003117232,
+      "learning_rate": 3.6332681490801784e-06,
+      "loss": 0.5043,
+      "step": 7646
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6301986354812212,
+      "learning_rate": 3.6320448319543044e-06,
+      "loss": 0.4369,
+      "step": 7647
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5301163825353894,
+      "learning_rate": 3.6308216033333394e-06,
+      "loss": 0.4948,
+      "step": 7648
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7096141323044214,
+      "learning_rate": 3.629598463296429e-06,
+      "loss": 0.5076,
+      "step": 7649
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6477339742066848,
+      "learning_rate": 3.628375411922701e-06,
+      "loss": 0.4778,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7512362293284458,
+      "learning_rate": 3.6271524492912925e-06,
+      "loss": 0.4697,
+      "step": 7651
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.786266437925051,
+      "learning_rate": 3.625929575481323e-06,
+      "loss": 0.5096,
+      "step": 7652
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6167109490259914,
+      "learning_rate": 3.624706790571914e-06,
+      "loss": 0.4651,
+      "step": 7653
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0516247553796467,
+      "learning_rate": 3.6234840946421757e-06,
+      "loss": 0.4268,
+      "step": 7654
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.802407154057507,
+      "learning_rate": 3.6222614877712165e-06,
+      "loss": 0.4699,
+      "step": 7655
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0000480771424844,
+      "learning_rate": 3.6210389700381354e-06,
+      "loss": 0.4672,
+      "step": 7656
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6437022282604529,
+      "learning_rate": 3.619816541522031e-06,
+      "loss": 0.494,
+      "step": 7657
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0221556124662943,
+      "learning_rate": 3.6185942023019887e-06,
+      "loss": 0.4814,
+      "step": 7658
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.556725181814122,
+      "learning_rate": 3.617371952457096e-06,
+      "loss": 0.5127,
+      "step": 7659
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5839996376768188,
+      "learning_rate": 3.616149792066428e-06,
+      "loss": 0.491,
+      "step": 7660
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8591946380951652,
+      "learning_rate": 3.6149277212090583e-06,
+      "loss": 0.5227,
+      "step": 7661
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8961270719862255,
+      "learning_rate": 3.6137057399640548e-06,
+      "loss": 0.4783,
+      "step": 7662
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.348903325352107,
+      "learning_rate": 3.6124838484104734e-06,
+      "loss": 0.4407,
+      "step": 7663
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7047406478812857,
+      "learning_rate": 3.611262046627375e-06,
+      "loss": 0.4864,
+      "step": 7664
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.268825372284758,
+      "learning_rate": 3.6100403346938027e-06,
+      "loss": 0.4386,
+      "step": 7665
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5518052260524733,
+      "learning_rate": 3.6088187126888065e-06,
+      "loss": 0.4873,
+      "step": 7666
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7415478170040337,
+      "learning_rate": 3.607597180691415e-06,
+      "loss": 0.5049,
+      "step": 7667
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.985101611476486,
+      "learning_rate": 3.6063757387806685e-06,
+      "loss": 0.5149,
+      "step": 7668
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8462728255371739,
+      "learning_rate": 3.6051543870355856e-06,
+      "loss": 0.4513,
+      "step": 7669
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.0338340757122033,
+      "learning_rate": 3.6039331255351915e-06,
+      "loss": 0.4564,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.16442278147747,
+      "learning_rate": 3.6027119543584955e-06,
+      "loss": 0.447,
+      "step": 7671
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.725383318224612,
+      "learning_rate": 3.601490873584508e-06,
+      "loss": 0.4166,
+      "step": 7672
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.356614940511579,
+      "learning_rate": 3.6002698832922338e-06,
+      "loss": 0.4269,
+      "step": 7673
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6289791283552701,
+      "learning_rate": 3.5990489835606656e-06,
+      "loss": 0.4447,
+      "step": 7674
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.3757520037540005,
+      "learning_rate": 3.597828174468797e-06,
+      "loss": 0.4668,
+      "step": 7675
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.7843933827797462,
+      "learning_rate": 3.59660745609561e-06,
+      "loss": 0.3889,
+      "step": 7676
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5259484078944399,
+      "learning_rate": 3.5953868285200854e-06,
+      "loss": 0.451,
+      "step": 7677
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.556694859058405,
+      "learning_rate": 3.594166291821194e-06,
+      "loss": 0.4587,
+      "step": 7678
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.5429516423851934,
+      "learning_rate": 3.5929458460779067e-06,
+      "loss": 0.4964,
+      "step": 7679
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1582659769286963,
+      "learning_rate": 3.5917254913691807e-06,
+      "loss": 0.5236,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5376692050729317,
+      "learning_rate": 3.5905052277739744e-06,
+      "loss": 0.4764,
+      "step": 7681
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6952495652352306,
+      "learning_rate": 3.589285055371235e-06,
+      "loss": 0.4539,
+      "step": 7682
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.1483993081415633,
+      "learning_rate": 3.5880649742399058e-06,
+      "loss": 0.4886,
+      "step": 7683
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5271286572373776,
+      "learning_rate": 3.5868449844589295e-06,
+      "loss": 0.4833,
+      "step": 7684
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9792085731753795,
+      "learning_rate": 3.58562508610723e-06,
+      "loss": 0.4368,
+      "step": 7685
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1420917594466338,
+      "learning_rate": 3.5844052792637395e-06,
+      "loss": 0.4507,
+      "step": 7686
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1414340784134756,
+      "learning_rate": 3.5831855640073755e-06,
+      "loss": 0.4712,
+      "step": 7687
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.595631431578242,
+      "learning_rate": 3.581965940417052e-06,
+      "loss": 0.4725,
+      "step": 7688
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8903140052081127,
+      "learning_rate": 3.580746408571677e-06,
+      "loss": 0.5103,
+      "step": 7689
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 3.040744442268335,
+      "learning_rate": 3.579526968550153e-06,
+      "loss": 0.4322,
+      "step": 7690
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.32999625437455,
+      "learning_rate": 3.578307620431375e-06,
+      "loss": 0.4372,
+      "step": 7691
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6215834845809944,
+      "learning_rate": 3.5770883642942357e-06,
+      "loss": 0.4298,
+      "step": 7692
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6835619335881362,
+      "learning_rate": 3.5758692002176163e-06,
+      "loss": 0.4317,
+      "step": 7693
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6404630799702735,
+      "learning_rate": 3.574650128280397e-06,
+      "loss": 0.3905,
+      "step": 7694
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.732495126062656,
+      "learning_rate": 3.57343114856145e-06,
+      "loss": 0.4396,
+      "step": 7695
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7875978818716862,
+      "learning_rate": 3.5722122611396416e-06,
+      "loss": 0.4899,
+      "step": 7696
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.720936110711901,
+      "learning_rate": 3.5709934660938326e-06,
+      "loss": 0.4052,
+      "step": 7697
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.1245678888879795,
+      "learning_rate": 3.569774763502875e-06,
+      "loss": 0.4774,
+      "step": 7698
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.9644602792858576,
+      "learning_rate": 3.5685561534456224e-06,
+      "loss": 0.4137,
+      "step": 7699
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5660455247706425,
+      "learning_rate": 3.567337636000911e-06,
+      "loss": 0.4857,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5728810125649879,
+      "learning_rate": 3.5661192112475835e-06,
+      "loss": 0.4625,
+      "step": 7701
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6465914564422006,
+      "learning_rate": 3.564900879264464e-06,
+      "loss": 0.4365,
+      "step": 7702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6289648893956155,
+      "learning_rate": 3.5636826401303838e-06,
+      "loss": 0.3971,
+      "step": 7703
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8366470678750322,
+      "learning_rate": 3.5624644939241545e-06,
+      "loss": 0.43,
+      "step": 7704
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.464762688304494,
+      "learning_rate": 3.561246440724593e-06,
+      "loss": 0.4787,
+      "step": 7705
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.12442404405517,
+      "learning_rate": 3.560028480610505e-06,
+      "loss": 0.5022,
+      "step": 7706
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1519336350063476,
+      "learning_rate": 3.5588106136606903e-06,
+      "loss": 0.4142,
+      "step": 7707
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2532236207468497,
+      "learning_rate": 3.5575928399539446e-06,
+      "loss": 0.4275,
+      "step": 7708
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0248187526216093,
+      "learning_rate": 3.556375159569054e-06,
+      "loss": 0.5054,
+      "step": 7709
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8092503635355976,
+      "learning_rate": 3.555157572584803e-06,
+      "loss": 0.4334,
+      "step": 7710
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6368092871328486,
+      "learning_rate": 3.553940079079966e-06,
+      "loss": 0.4125,
+      "step": 7711
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.537712873303657,
+      "learning_rate": 3.5527226791333155e-06,
+      "loss": 0.4628,
+      "step": 7712
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.4011925175557485,
+      "learning_rate": 3.5515053728236127e-06,
+      "loss": 0.4663,
+      "step": 7713
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.683257425124835,
+      "learning_rate": 3.550288160229619e-06,
+      "loss": 0.4733,
+      "step": 7714
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7698660117742178,
+      "learning_rate": 3.549071041430084e-06,
+      "loss": 0.4566,
+      "step": 7715
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0152560370037573,
+      "learning_rate": 3.5478540165037534e-06,
+      "loss": 0.4896,
+      "step": 7716
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.245931665025475,
+      "learning_rate": 3.546637085529371e-06,
+      "loss": 0.4544,
+      "step": 7717
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1520176645216127,
+      "learning_rate": 3.545420248585665e-06,
+      "loss": 0.4401,
+      "step": 7718
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5744639864516021,
+      "learning_rate": 3.5442035057513693e-06,
+      "loss": 0.4341,
+      "step": 7719
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7962498897551804,
+      "learning_rate": 3.5429868571051997e-06,
+      "loss": 0.4387,
+      "step": 7720
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9752687255806276,
+      "learning_rate": 3.5417703027258752e-06,
+      "loss": 0.4723,
+      "step": 7721
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6536907384050394,
+      "learning_rate": 3.5405538426921048e-06,
+      "loss": 0.4618,
+      "step": 7722
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.3449518310382738,
+      "learning_rate": 3.539337477082592e-06,
+      "loss": 0.5182,
+      "step": 7723
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.754947429676455,
+      "learning_rate": 3.538121205976033e-06,
+      "loss": 0.487,
+      "step": 7724
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1724998229459946,
+      "learning_rate": 3.5369050294511203e-06,
+      "loss": 0.4067,
+      "step": 7725
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.1806490618679644,
+      "learning_rate": 3.5356889475865374e-06,
+      "loss": 0.4148,
+      "step": 7726
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9271024516365005,
+      "learning_rate": 3.534472960460965e-06,
+      "loss": 0.4665,
+      "step": 7727
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2766448537353607,
+      "learning_rate": 3.5332570681530735e-06,
+      "loss": 0.4711,
+      "step": 7728
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.046013050676859,
+      "learning_rate": 3.5320412707415314e-06,
+      "loss": 0.4624,
+      "step": 7729
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6016884987781782,
+      "learning_rate": 3.5308255683049998e-06,
+      "loss": 0.4813,
+      "step": 7730
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7150542255811525,
+      "learning_rate": 3.5296099609221302e-06,
+      "loss": 0.4255,
+      "step": 7731
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5776614830265858,
+      "learning_rate": 3.528394448671575e-06,
+      "loss": 0.4957,
+      "step": 7732
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.376625592506426,
+      "learning_rate": 3.527179031631972e-06,
+      "loss": 0.4828,
+      "step": 7733
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.006374050378446,
+      "learning_rate": 3.52596370988196e-06,
+      "loss": 0.4355,
+      "step": 7734
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5526348399458957,
+      "learning_rate": 3.5247484835001654e-06,
+      "loss": 0.4468,
+      "step": 7735
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.734668030416912,
+      "learning_rate": 3.5235333525652176e-06,
+      "loss": 0.4393,
+      "step": 7736
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 5.957905405192002,
+      "learning_rate": 3.5223183171557274e-06,
+      "loss": 0.4626,
+      "step": 7737
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.48630398247633,
+      "learning_rate": 3.5211033773503116e-06,
+      "loss": 0.4291,
+      "step": 7738
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8337337694717137,
+      "learning_rate": 3.51988853322757e-06,
+      "loss": 0.4484,
+      "step": 7739
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4987261368824543,
+      "learning_rate": 3.5186737848661044e-06,
+      "loss": 0.4366,
+      "step": 7740
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6530649763027468,
+      "learning_rate": 3.5174591323445097e-06,
+      "loss": 0.4905,
+      "step": 7741
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5850992959957473,
+      "learning_rate": 3.516244575741367e-06,
+      "loss": 0.5069,
+      "step": 7742
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.354566002359784,
+      "learning_rate": 3.5150301151352613e-06,
+      "loss": 0.4495,
+      "step": 7743
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.04868532090511,
+      "learning_rate": 3.513815750604763e-06,
+      "loss": 0.5042,
+      "step": 7744
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8840664840722436,
+      "learning_rate": 3.512601482228443e-06,
+      "loss": 0.4652,
+      "step": 7745
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6978204486549122,
+      "learning_rate": 3.5113873100848594e-06,
+      "loss": 0.4805,
+      "step": 7746
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8603135788949292,
+      "learning_rate": 3.5101732342525716e-06,
+      "loss": 0.4265,
+      "step": 7747
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6535146935256293,
+      "learning_rate": 3.5089592548101247e-06,
+      "loss": 0.415,
+      "step": 7748
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.163528315457827,
+      "learning_rate": 3.5077453718360654e-06,
+      "loss": 0.4559,
+      "step": 7749
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.526713512491908,
+      "learning_rate": 3.5065315854089267e-06,
+      "loss": 0.4709,
+      "step": 7750
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.453791742377531,
+      "learning_rate": 3.50531789560724e-06,
+      "loss": 0.4481,
+      "step": 7751
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9470245829949138,
+      "learning_rate": 3.5041043025095333e-06,
+      "loss": 0.5051,
+      "step": 7752
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7638268250349165,
+      "learning_rate": 3.5028908061943177e-06,
+      "loss": 0.4463,
+      "step": 7753
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.443999839196009,
+      "learning_rate": 3.5016774067401127e-06,
+      "loss": 0.5006,
+      "step": 7754
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.797079983250265,
+      "learning_rate": 3.500464104225415e-06,
+      "loss": 0.4634,
+      "step": 7755
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6851447284671497,
+      "learning_rate": 3.499250898728731e-06,
+      "loss": 0.4115,
+      "step": 7756
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6006573267461806,
+      "learning_rate": 3.4980377903285496e-06,
+      "loss": 0.488,
+      "step": 7757
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4122841529459897,
+      "learning_rate": 3.4968247791033595e-06,
+      "loss": 0.4125,
+      "step": 7758
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7509895624742733,
+      "learning_rate": 3.4956118651316384e-06,
+      "loss": 0.5272,
+      "step": 7759
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5617247527871281,
+      "learning_rate": 3.494399048491862e-06,
+      "loss": 0.467,
+      "step": 7760
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5290189807365688,
+      "learning_rate": 3.4931863292624978e-06,
+      "loss": 0.4239,
+      "step": 7761
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7495206811458648,
+      "learning_rate": 3.4919737075220063e-06,
+      "loss": 0.4544,
+      "step": 7762
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.792238642723388,
+      "learning_rate": 3.490761183348844e-06,
+      "loss": 0.4734,
+      "step": 7763
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.231971424535322,
+      "learning_rate": 3.489548756821458e-06,
+      "loss": 0.4786,
+      "step": 7764
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5810388314843244,
+      "learning_rate": 3.488336428018293e-06,
+      "loss": 0.4624,
+      "step": 7765
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5418697024274549,
+      "learning_rate": 3.4871241970177822e-06,
+      "loss": 0.4978,
+      "step": 7766
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0461562546552328,
+      "learning_rate": 3.485912063898358e-06,
+      "loss": 0.4444,
+      "step": 7767
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.715901377869296,
+      "learning_rate": 3.4847000287384416e-06,
+      "loss": 0.4539,
+      "step": 7768
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.048781475932776,
+      "learning_rate": 3.4834880916164514e-06,
+      "loss": 0.4466,
+      "step": 7769
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2751284556126565,
+      "learning_rate": 3.4822762526107967e-06,
+      "loss": 0.43,
+      "step": 7770
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.545284071763954,
+      "learning_rate": 3.4810645117998855e-06,
+      "loss": 0.4637,
+      "step": 7771
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5639688152372062,
+      "learning_rate": 3.4798528692621105e-06,
+      "loss": 0.4862,
+      "step": 7772
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7629806018659928,
+      "learning_rate": 3.4786413250758665e-06,
+      "loss": 0.4414,
+      "step": 7773
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4877699677539986,
+      "learning_rate": 3.4774298793195407e-06,
+      "loss": 0.5162,
+      "step": 7774
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8013514606503256,
+      "learning_rate": 3.476218532071509e-06,
+      "loss": 0.4667,
+      "step": 7775
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4161527665763827,
+      "learning_rate": 3.4750072834101457e-06,
+      "loss": 0.4967,
+      "step": 7776
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3764002296239217,
+      "learning_rate": 3.473796133413815e-06,
+      "loss": 0.4923,
+      "step": 7777
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0712571933892625,
+      "learning_rate": 3.4725850821608786e-06,
+      "loss": 0.4722,
+      "step": 7778
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5627575763616572,
+      "learning_rate": 3.471374129729689e-06,
+      "loss": 0.4301,
+      "step": 7779
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7541916555877992,
+      "learning_rate": 3.4701632761985937e-06,
+      "loss": 0.4467,
+      "step": 7780
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.157940135597341,
+      "learning_rate": 3.4689525216459323e-06,
+      "loss": 0.4403,
+      "step": 7781
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6141707005086356,
+      "learning_rate": 3.4677418661500416e-06,
+      "loss": 0.4636,
+      "step": 7782
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.0866488598430952,
+      "learning_rate": 3.466531309789246e-06,
+      "loss": 0.4302,
+      "step": 7783
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.0030273116582396,
+      "learning_rate": 3.465320852641868e-06,
+      "loss": 0.4125,
+      "step": 7784
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.394763879770947,
+      "learning_rate": 3.464110494786226e-06,
+      "loss": 0.492,
+      "step": 7785
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5618212109733438,
+      "learning_rate": 3.4629002363006225e-06,
+      "loss": 0.5017,
+      "step": 7786
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8232024912700124,
+      "learning_rate": 3.4616900772633664e-06,
+      "loss": 0.4764,
+      "step": 7787
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5426176265536636,
+      "learning_rate": 3.4604800177527476e-06,
+      "loss": 0.4823,
+      "step": 7788
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8896647741238566,
+      "learning_rate": 3.4592700578470594e-06,
+      "loss": 0.44,
+      "step": 7789
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.4560253565396524,
+      "learning_rate": 3.458060197624581e-06,
+      "loss": 0.4781,
+      "step": 7790
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6343755930901187,
+      "learning_rate": 3.456850437163592e-06,
+      "loss": 0.4429,
+      "step": 7791
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.2050646115609913,
+      "learning_rate": 3.4556407765423604e-06,
+      "loss": 0.5005,
+      "step": 7792
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.759691106928934,
+      "learning_rate": 3.4544312158391515e-06,
+      "loss": 0.4281,
+      "step": 7793
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.123578325195361,
+      "learning_rate": 3.453221755132219e-06,
+      "loss": 0.4433,
+      "step": 7794
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8119399115970658,
+      "learning_rate": 3.4520123944998164e-06,
+      "loss": 0.494,
+      "step": 7795
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5506108452279956,
+      "learning_rate": 3.4508031340201875e-06,
+      "loss": 0.4919,
+      "step": 7796
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.5189440032600667,
+      "learning_rate": 3.449593973771569e-06,
+      "loss": 0.4446,
+      "step": 7797
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.8629101391508636,
+      "learning_rate": 3.448384913832192e-06,
+      "loss": 0.4979,
+      "step": 7798
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7851831975856893,
+      "learning_rate": 3.4471759542802807e-06,
+      "loss": 0.4721,
+      "step": 7799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5750204239526002,
+      "learning_rate": 3.4459670951940555e-06,
+      "loss": 0.4827,
+      "step": 7800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.304558033599131,
+      "learning_rate": 3.4447583366517246e-06,
+      "loss": 0.4497,
+      "step": 7801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7172343639267629,
+      "learning_rate": 3.443549678731496e-06,
+      "loss": 0.4598,
+      "step": 7802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4216467552979597,
+      "learning_rate": 3.4423411215115664e-06,
+      "loss": 0.4548,
+      "step": 7803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8515890439306513,
+      "learning_rate": 3.4411326650701294e-06,
+      "loss": 0.4842,
+      "step": 7804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7606044638048226,
+      "learning_rate": 3.4399243094853675e-06,
+      "loss": 0.455,
+      "step": 7805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5638856299518022,
+      "learning_rate": 3.438716054835465e-06,
+      "loss": 0.4848,
+      "step": 7806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5612651591630189,
+      "learning_rate": 3.437507901198588e-06,
+      "loss": 0.4679,
+      "step": 7807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.66777144876639,
+      "learning_rate": 3.4362998486529077e-06,
+      "loss": 0.4031,
+      "step": 7808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.955437203751179,
+      "learning_rate": 3.4350918972765818e-06,
+      "loss": 0.4433,
+      "step": 7809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5119838559572328,
+      "learning_rate": 3.433884047147762e-06,
+      "loss": 0.46,
+      "step": 7810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5606909319682777,
+      "learning_rate": 3.432676298344597e-06,
+      "loss": 0.4983,
+      "step": 7811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.8030241788304848,
+      "learning_rate": 3.4314686509452245e-06,
+      "loss": 0.4274,
+      "step": 7812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6970846204874446,
+      "learning_rate": 3.430261105027779e-06,
+      "loss": 0.4511,
+      "step": 7813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3881742015479352,
+      "learning_rate": 3.429053660670385e-06,
+      "loss": 0.4914,
+      "step": 7814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.9865297891816587,
+      "learning_rate": 3.4278463179511657e-06,
+      "loss": 0.5169,
+      "step": 7815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 3.260625087334255,
+      "learning_rate": 3.4266390769482316e-06,
+      "loss": 0.497,
+      "step": 7816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5311746370469715,
+      "learning_rate": 3.4254319377396927e-06,
+      "loss": 0.4736,
+      "step": 7817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5376192084856646,
+      "learning_rate": 3.4242249004036466e-06,
+      "loss": 0.484,
+      "step": 7818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7203093064686812,
+      "learning_rate": 3.423017965018186e-06,
+      "loss": 0.4591,
+      "step": 7819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.752747589535994,
+      "learning_rate": 3.4218111316614042e-06,
+      "loss": 0.4442,
+      "step": 7820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6260672806602947,
+      "learning_rate": 3.420604400411374e-06,
+      "loss": 0.3975,
+      "step": 7821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7803915612396886,
+      "learning_rate": 3.4193977713461767e-06,
+      "loss": 0.4526,
+      "step": 7822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 2.1770460797315025,
+      "learning_rate": 3.4181912445438724e-06,
+      "loss": 0.4245,
+      "step": 7823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6507188981396902,
+      "learning_rate": 3.4169848200825286e-06,
+      "loss": 0.4871,
+      "step": 7824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6476040446734745,
+      "learning_rate": 3.415778498040192e-06,
+      "loss": 0.453,
+      "step": 7825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6277957584157905,
+      "learning_rate": 3.4145722784949174e-06,
+      "loss": 0.4546,
+      "step": 7826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.975527833231104,
+      "learning_rate": 3.413366161524741e-06,
+      "loss": 0.5058,
+      "step": 7827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.7755100329824016,
+      "learning_rate": 3.4121601472076993e-06,
+      "loss": 0.4463,
+      "step": 7828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.977926139424973,
+      "learning_rate": 3.4109542356218185e-06,
+      "loss": 0.407,
+      "step": 7829
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.743118723634884,
+      "learning_rate": 3.4097484268451185e-06,
+      "loss": 0.509,
+      "step": 7830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.356843936179444,
+      "learning_rate": 3.4085427209556175e-06,
+      "loss": 0.5006,
+      "step": 7831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 7.865525156171202,
+      "learning_rate": 3.4073371180313188e-06,
+      "loss": 0.4449,
+      "step": 7832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5751941610924689,
+      "learning_rate": 3.4061316181502264e-06,
+      "loss": 0.4839,
+      "step": 7833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5542393068656261,
+      "learning_rate": 3.404926221390332e-06,
+      "loss": 0.4946,
+      "step": 7834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4082272471765855,
+      "learning_rate": 3.403720927829626e-06,
+      "loss": 0.4367,
+      "step": 7835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.180813954898519,
+      "learning_rate": 3.402515737546086e-06,
+      "loss": 0.4259,
+      "step": 7836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.42760682988305,
+      "learning_rate": 3.401310650617689e-06,
+      "loss": 0.4654,
+      "step": 7837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.389878005630295,
+      "learning_rate": 3.4001056671224007e-06,
+      "loss": 0.4532,
+      "step": 7838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9018939273968047,
+      "learning_rate": 3.398900787138184e-06,
+      "loss": 0.4266,
+      "step": 7839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.6220294381130254,
+      "learning_rate": 3.39769601074299e-06,
+      "loss": 0.493,
+      "step": 7840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5580460209290257,
+      "learning_rate": 3.396491338014769e-06,
+      "loss": 0.4951,
+      "step": 7841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7355362742564588,
+      "learning_rate": 3.395286769031462e-06,
+      "loss": 0.4631,
+      "step": 7842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.909112107128007,
+      "learning_rate": 3.3940823038710003e-06,
+      "loss": 0.4539,
+      "step": 7843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.139015606390699,
+      "learning_rate": 3.392877942611314e-06,
+      "loss": 0.5055,
+      "step": 7844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5704977562964734,
+      "learning_rate": 3.391673685330321e-06,
+      "loss": 0.4799,
+      "step": 7845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6495593272371447,
+      "learning_rate": 3.3904695321059384e-06,
+      "loss": 0.4855,
+      "step": 7846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6863620093371725,
+      "learning_rate": 3.3892654830160696e-06,
+      "loss": 0.4663,
+      "step": 7847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5283551907601328,
+      "learning_rate": 3.388061538138619e-06,
+      "loss": 0.475,
+      "step": 7848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5507922938811267,
+      "learning_rate": 3.3868576975514768e-06,
+      "loss": 0.4668,
+      "step": 7849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.081323072902518,
+      "learning_rate": 3.3856539613325322e-06,
+      "loss": 0.5799,
+      "step": 7850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8406038584266533,
+      "learning_rate": 3.384450329559663e-06,
+      "loss": 0.4716,
+      "step": 7851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6744057782957944,
+      "learning_rate": 3.3832468023107444e-06,
+      "loss": 0.4799,
+      "step": 7852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.040029528342024,
+      "learning_rate": 3.3820433796636443e-06,
+      "loss": 0.5024,
+      "step": 7853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9859895638590808,
+      "learning_rate": 3.3808400616962186e-06,
+      "loss": 0.4092,
+      "step": 7854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5933296790648138,
+      "learning_rate": 3.3796368484863253e-06,
+      "loss": 0.4938,
+      "step": 7855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8712467601487086,
+      "learning_rate": 3.3784337401118055e-06,
+      "loss": 0.4442,
+      "step": 7856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5791950781624288,
+      "learning_rate": 3.3772307366505045e-06,
+      "loss": 0.4936,
+      "step": 7857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.531033330104198,
+      "learning_rate": 3.376027838180249e-06,
+      "loss": 0.4727,
+      "step": 7858
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.5337493463100933,
+      "learning_rate": 3.374825044778871e-06,
+      "loss": 0.4834,
+      "step": 7859
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.020548936747588,
+      "learning_rate": 3.3736223565241838e-06,
+      "loss": 0.4177,
+      "step": 7860
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.798624412810488,
+      "learning_rate": 3.3724197734940046e-06,
+      "loss": 0.4519,
+      "step": 7861
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9232709864825064,
+      "learning_rate": 3.3712172957661355e-06,
+      "loss": 0.4383,
+      "step": 7862
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7135987048794163,
+      "learning_rate": 3.3700149234183766e-06,
+      "loss": 0.5133,
+      "step": 7863
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6250844449972464,
+      "learning_rate": 3.368812656528521e-06,
+      "loss": 0.4448,
+      "step": 7864
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9731809579643913,
+      "learning_rate": 3.3676104951743514e-06,
+      "loss": 0.4484,
+      "step": 7865
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5124774234093759,
+      "learning_rate": 3.3664084394336494e-06,
+      "loss": 0.4165,
+      "step": 7866
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5933597967894093,
+      "learning_rate": 3.3652064893841825e-06,
+      "loss": 0.4702,
+      "step": 7867
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2325321504664326,
+      "learning_rate": 3.3640046451037185e-06,
+      "loss": 0.4463,
+      "step": 7868
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6516801948782192,
+      "learning_rate": 3.3628029066700123e-06,
+      "loss": 0.4821,
+      "step": 7869
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5375150685233474,
+      "learning_rate": 3.361601274160817e-06,
+      "loss": 0.4944,
+      "step": 7870
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9222730752437804,
+      "learning_rate": 3.3603997476538765e-06,
+      "loss": 0.4826,
+      "step": 7871
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.067986319574155,
+      "learning_rate": 3.3591983272269272e-06,
+      "loss": 0.4259,
+      "step": 7872
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.1019939910543015,
+      "learning_rate": 3.3579970129576987e-06,
+      "loss": 0.4327,
+      "step": 7873
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5445308673528748,
+      "learning_rate": 3.3567958049239154e-06,
+      "loss": 0.4822,
+      "step": 7874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8008671060584631,
+      "learning_rate": 3.3555947032032956e-06,
+      "loss": 0.4652,
+      "step": 7875
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5415305989542187,
+      "learning_rate": 3.3543937078735457e-06,
+      "loss": 0.4675,
+      "step": 7876
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5604636611272438,
+      "learning_rate": 3.353192819012371e-06,
+      "loss": 0.4277,
+      "step": 7877
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6834465887369972,
+      "learning_rate": 3.3519920366974666e-06,
+      "loss": 0.4596,
+      "step": 7878
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7823758831624599,
+      "learning_rate": 3.3507913610065222e-06,
+      "loss": 0.4299,
+      "step": 7879
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.0523132287458985,
+      "learning_rate": 3.3495907920172178e-06,
+      "loss": 0.4504,
+      "step": 7880
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.742681823469371,
+      "learning_rate": 3.3483903298072317e-06,
+      "loss": 0.4517,
+      "step": 7881
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9376002305963953,
+      "learning_rate": 3.3471899744542296e-06,
+      "loss": 0.4936,
+      "step": 7882
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7434685830135765,
+      "learning_rate": 3.3459897260358753e-06,
+      "loss": 0.4754,
+      "step": 7883
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9164261097584991,
+      "learning_rate": 3.3447895846298207e-06,
+      "loss": 0.469,
+      "step": 7884
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.144855079105593,
+      "learning_rate": 3.3435895503137156e-06,
+      "loss": 0.5029,
+      "step": 7885
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6841579489602816,
+      "learning_rate": 3.3423896231651987e-06,
+      "loss": 0.4398,
+      "step": 7886
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5998234698716693,
+      "learning_rate": 3.341189803261905e-06,
+      "loss": 0.4492,
+      "step": 7887
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.581806227791593,
+      "learning_rate": 3.3399900906814618e-06,
+      "loss": 0.4437,
+      "step": 7888
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9191970825442766,
+      "learning_rate": 3.3387904855014865e-06,
+      "loss": 0.4437,
+      "step": 7889
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4261573668559586,
+      "learning_rate": 3.337590987799596e-06,
+      "loss": 0.4366,
+      "step": 7890
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2741061695266542,
+      "learning_rate": 3.3363915976533913e-06,
+      "loss": 0.4331,
+      "step": 7891
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7849203227975985,
+      "learning_rate": 3.335192315140477e-06,
+      "loss": 0.4371,
+      "step": 7892
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0927305285455735,
+      "learning_rate": 3.333993140338439e-06,
+      "loss": 0.4729,
+      "step": 7893
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 3.4389659112552584,
+      "learning_rate": 3.3327940733248686e-06,
+      "loss": 0.4625,
+      "step": 7894
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 4.581694530153472,
+      "learning_rate": 3.3315951141773372e-06,
+      "loss": 0.4346,
+      "step": 7895
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7276496191562274,
+      "learning_rate": 3.330396262973423e-06,
+      "loss": 0.4297,
+      "step": 7896
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 5.372300333855026,
+      "learning_rate": 3.3291975197906826e-06,
+      "loss": 0.4566,
+      "step": 7897
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.910193767981382,
+      "learning_rate": 3.327998884706679e-06,
+      "loss": 0.4835,
+      "step": 7898
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.442835003729624,
+      "learning_rate": 3.3268003577989604e-06,
+      "loss": 0.4281,
+      "step": 7899
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5549435760047432,
+      "learning_rate": 3.3256019391450696e-06,
+      "loss": 0.4895,
+      "step": 7900
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0587156608800146,
+      "learning_rate": 3.3244036288225434e-06,
+      "loss": 0.4365,
+      "step": 7901
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8609002645804358,
+      "learning_rate": 3.3232054269089098e-06,
+      "loss": 0.4869,
+      "step": 7902
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6449230986259503,
+      "learning_rate": 3.3220073334816928e-06,
+      "loss": 0.4994,
+      "step": 7903
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.374026547723643,
+      "learning_rate": 3.3208093486184044e-06,
+      "loss": 0.4406,
+      "step": 7904
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5552235253148115,
+      "learning_rate": 3.3196114723965563e-06,
+      "loss": 0.4867,
+      "step": 7905
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9794255584994165,
+      "learning_rate": 3.318413704893646e-06,
+      "loss": 0.4768,
+      "step": 7906
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.2349932156156322,
+      "learning_rate": 3.3172160461871707e-06,
+      "loss": 0.4171,
+      "step": 7907
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.574007186470273,
+      "learning_rate": 3.316018496354615e-06,
+      "loss": 0.4843,
+      "step": 7908
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8306596311851202,
+      "learning_rate": 3.3148210554734582e-06,
+      "loss": 0.4141,
+      "step": 7909
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8812035877703819,
+      "learning_rate": 3.3136237236211764e-06,
+      "loss": 0.4153,
+      "step": 7910
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5294741054654777,
+      "learning_rate": 3.312426500875233e-06,
+      "loss": 0.4559,
+      "step": 7911
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6202595309526575,
+      "learning_rate": 3.3112293873130885e-06,
+      "loss": 0.4442,
+      "step": 7912
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0492215835956067,
+      "learning_rate": 3.310032383012193e-06,
+      "loss": 0.457,
+      "step": 7913
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5840852166249517,
+      "learning_rate": 3.3088354880499914e-06,
+      "loss": 0.4918,
+      "step": 7914
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6449017353327084,
+      "learning_rate": 3.307638702503921e-06,
+      "loss": 0.4179,
+      "step": 7915
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.079641154305045,
+      "learning_rate": 3.3064420264514135e-06,
+      "loss": 0.4878,
+      "step": 7916
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7624127495138453,
+      "learning_rate": 3.3052454599698904e-06,
+      "loss": 0.4971,
+      "step": 7917
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5915910402949223,
+      "learning_rate": 3.30404900313677e-06,
+      "loss": 0.509,
+      "step": 7918
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4128039550572555,
+      "learning_rate": 3.3028526560294593e-06,
+      "loss": 0.4692,
+      "step": 7919
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.9023950105550256,
+      "learning_rate": 3.3016564187253623e-06,
+      "loss": 0.4534,
+      "step": 7920
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5172693137827169,
+      "learning_rate": 3.300460291301874e-06,
+      "loss": 0.5055,
+      "step": 7921
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4730317074876444,
+      "learning_rate": 3.2992642738363805e-06,
+      "loss": 0.4213,
+      "step": 7922
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8716503281730743,
+      "learning_rate": 3.298068366406265e-06,
+      "loss": 0.4746,
+      "step": 7923
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5640836734216811,
+      "learning_rate": 3.296872569088897e-06,
+      "loss": 0.4988,
+      "step": 7924
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6717212336333918,
+      "learning_rate": 3.295676881961649e-06,
+      "loss": 0.4428,
+      "step": 7925
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8771564598046564,
+      "learning_rate": 3.294481305101875e-06,
+      "loss": 0.4322,
+      "step": 7926
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6689898712788251,
+      "learning_rate": 3.2932858385869315e-06,
+      "loss": 0.4098,
+      "step": 7927
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.45927235227874,
+      "learning_rate": 3.2920904824941592e-06,
+      "loss": 0.4255,
+      "step": 7928
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5211620052168389,
+      "learning_rate": 3.290895236900902e-06,
+      "loss": 0.4163,
+      "step": 7929
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5390395830382881,
+      "learning_rate": 3.2897001018844827e-06,
+      "loss": 0.4736,
+      "step": 7930
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.0206430615475557,
+      "learning_rate": 3.2885050775222314e-06,
+      "loss": 0.4692,
+      "step": 7931
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4355797423529686,
+      "learning_rate": 3.2873101638914636e-06,
+      "loss": 0.4355,
+      "step": 7932
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7363073947224397,
+      "learning_rate": 3.2861153610694862e-06,
+      "loss": 0.4201,
+      "step": 7933
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9782500730005965,
+      "learning_rate": 3.284920669133604e-06,
+      "loss": 0.399,
+      "step": 7934
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.7056295425714607,
+      "learning_rate": 3.2837260881611105e-06,
+      "loss": 0.476,
+      "step": 7935
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9469819569499474,
+      "learning_rate": 3.2825316182292944e-06,
+      "loss": 0.461,
+      "step": 7936
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4189068331601302,
+      "learning_rate": 3.2813372594154353e-06,
+      "loss": 0.4606,
+      "step": 7937
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7573392982079843,
+      "learning_rate": 3.280143011796808e-06,
+      "loss": 0.4559,
+      "step": 7938
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3430798599108236,
+      "learning_rate": 3.278948875450677e-06,
+      "loss": 0.4122,
+      "step": 7939
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.614049877397535,
+      "learning_rate": 3.2777548504543033e-06,
+      "loss": 0.4883,
+      "step": 7940
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6703984558381462,
+      "learning_rate": 3.276560936884937e-06,
+      "loss": 0.3988,
+      "step": 7941
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.205563079257479,
+      "learning_rate": 3.2753671348198225e-06,
+      "loss": 0.481,
+      "step": 7942
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.922796140964399,
+      "learning_rate": 3.2741734443362007e-06,
+      "loss": 0.4399,
+      "step": 7943
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5353179007416976,
+      "learning_rate": 3.2729798655112966e-06,
+      "loss": 0.4797,
+      "step": 7944
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7840068548268462,
+      "learning_rate": 3.2717863984223384e-06,
+      "loss": 0.4758,
+      "step": 7945
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6810472733731971,
+      "learning_rate": 3.2705930431465383e-06,
+      "loss": 0.421,
+      "step": 7946
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.443972184812371,
+      "learning_rate": 3.269399799761107e-06,
+      "loss": 0.4901,
+      "step": 7947
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8276815718515595,
+      "learning_rate": 3.2682066683432435e-06,
+      "loss": 0.4641,
+      "step": 7948
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9253746051345912,
+      "learning_rate": 3.2670136489701443e-06,
+      "loss": 0.4825,
+      "step": 7949
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.8683854126505743,
+      "learning_rate": 3.2658207417189936e-06,
+      "loss": 0.456,
+      "step": 7950
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.5639043127375922,
+      "learning_rate": 3.264627946666974e-06,
+      "loss": 0.4997,
+      "step": 7951
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5654546389418793,
+      "learning_rate": 3.263435263891255e-06,
+      "loss": 0.4731,
+      "step": 7952
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6031647601100334,
+      "learning_rate": 3.262242693469003e-06,
+      "loss": 0.4737,
+      "step": 7953
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6138786055201226,
+      "learning_rate": 3.261050235477375e-06,
+      "loss": 0.4761,
+      "step": 7954
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6477527683899091,
+      "learning_rate": 3.2598578899935217e-06,
+      "loss": 0.4499,
+      "step": 7955
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.476899503814041,
+      "learning_rate": 3.2586656570945874e-06,
+      "loss": 0.4755,
+      "step": 7956
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5424033250130103,
+      "learning_rate": 3.2574735368577067e-06,
+      "loss": 0.4814,
+      "step": 7957
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.9999283100451462,
+      "learning_rate": 3.256281529360009e-06,
+      "loss": 0.452,
+      "step": 7958
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9318026283154304,
+      "learning_rate": 3.255089634678614e-06,
+      "loss": 0.4355,
+      "step": 7959
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5149372586078518,
+      "learning_rate": 3.2538978528906396e-06,
+      "loss": 0.4025,
+      "step": 7960
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8710454965261167,
+      "learning_rate": 3.252706184073187e-06,
+      "loss": 0.4767,
+      "step": 7961
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.441051841179274,
+      "learning_rate": 3.2515146283033617e-06,
+      "loss": 0.508,
+      "step": 7962
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.296848851092269,
+      "learning_rate": 3.250323185658249e-06,
+      "loss": 0.4957,
+      "step": 7963
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4079413532839486,
+      "learning_rate": 3.249131856214941e-06,
+      "loss": 0.4179,
+      "step": 7964
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5754152520870217,
+      "learning_rate": 3.2479406400505076e-06,
+      "loss": 0.5209,
+      "step": 7965
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5213211066753851,
+      "learning_rate": 3.246749537242024e-06,
+      "loss": 0.4211,
+      "step": 7966
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9193424232770036,
+      "learning_rate": 3.245558547866554e-06,
+      "loss": 0.4788,
+      "step": 7967
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.289393415635444,
+      "learning_rate": 3.2443676720011493e-06,
+      "loss": 0.3686,
+      "step": 7968
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5650251477823203,
+      "learning_rate": 3.2431769097228606e-06,
+      "loss": 0.4949,
+      "step": 7969
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6519145550502334,
+      "learning_rate": 3.241986261108726e-06,
+      "loss": 0.4008,
+      "step": 7970
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.992369713104297,
+      "learning_rate": 3.240795726235782e-06,
+      "loss": 0.411,
+      "step": 7971
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9390246926125874,
+      "learning_rate": 3.2396053051810515e-06,
+      "loss": 0.4997,
+      "step": 7972
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3843742242235517,
+      "learning_rate": 3.238414998021557e-06,
+      "loss": 0.4668,
+      "step": 7973
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.120412494402407,
+      "learning_rate": 3.2372248048343057e-06,
+      "loss": 0.448,
+      "step": 7974
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4194845590102416,
+      "learning_rate": 3.2360347256963052e-06,
+      "loss": 0.4875,
+      "step": 7975
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1248095422017603,
+      "learning_rate": 3.234844760684549e-06,
+      "loss": 0.4697,
+      "step": 7976
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9878960217134987,
+      "learning_rate": 3.233654909876027e-06,
+      "loss": 0.4531,
+      "step": 7977
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4744616690432975,
+      "learning_rate": 3.2324651733477245e-06,
+      "loss": 0.4368,
+      "step": 7978
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7841976390906658,
+      "learning_rate": 3.2312755511766104e-06,
+      "loss": 0.4787,
+      "step": 7979
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.014484077126986,
+      "learning_rate": 3.230086043439656e-06,
+      "loss": 0.4394,
+      "step": 7980
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0449379572200406,
+      "learning_rate": 3.2288966502138187e-06,
+      "loss": 0.4092,
+      "step": 7981
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5871955385110332,
+      "learning_rate": 3.2277073715760508e-06,
+      "loss": 0.4131,
+      "step": 7982
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5920279765919743,
+      "learning_rate": 3.226518207603298e-06,
+      "loss": 0.4848,
+      "step": 7983
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6359311327224952,
+      "learning_rate": 3.2253291583724975e-06,
+      "loss": 0.4059,
+      "step": 7984
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5277722991203128,
+      "learning_rate": 3.2241402239605786e-06,
+      "loss": 0.4797,
+      "step": 7985
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.244064979551664,
+      "learning_rate": 3.2229514044444642e-06,
+      "loss": 0.4659,
+      "step": 7986
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5458962162818889,
+      "learning_rate": 3.2217626999010687e-06,
+      "loss": 0.4857,
+      "step": 7987
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5918979008661037,
+      "learning_rate": 3.2205741104073e-06,
+      "loss": 0.4855,
+      "step": 7988
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6562618828911972,
+      "learning_rate": 3.21938563604006e-06,
+      "loss": 0.4953,
+      "step": 7989
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8316000382083715,
+      "learning_rate": 3.218197276876239e-06,
+      "loss": 0.4632,
+      "step": 7990
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8533171472734944,
+      "learning_rate": 3.217009032992724e-06,
+      "loss": 0.4905,
+      "step": 7991
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5431868150622786,
+      "learning_rate": 3.2158209044663914e-06,
+      "loss": 0.4732,
+      "step": 7992
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6060990819309332,
+      "learning_rate": 3.2146328913741134e-06,
+      "loss": 0.4774,
+      "step": 7993
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.662887753681133,
+      "learning_rate": 3.2134449937927492e-06,
+      "loss": 0.4431,
+      "step": 7994
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8062566351055218,
+      "learning_rate": 3.2122572117991604e-06,
+      "loss": 0.4434,
+      "step": 7995
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4279412890223049,
+      "learning_rate": 3.211069545470187e-06,
+      "loss": 0.4133,
+      "step": 7996
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.1378230697258975,
+      "learning_rate": 3.2098819948826787e-06,
+      "loss": 0.4815,
+      "step": 7997
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8181573222708045,
+      "learning_rate": 3.208694560113459e-06,
+      "loss": 0.4589,
+      "step": 7998
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0958304840534594,
+      "learning_rate": 3.2075072412393598e-06,
+      "loss": 0.4568,
+      "step": 7999
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5821864234010908,
+      "learning_rate": 3.2063200383371985e-06,
+      "loss": 0.4854,
+      "step": 8000
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.391900787662937,
+      "learning_rate": 3.2051329514837827e-06,
+      "loss": 0.4533,
+      "step": 8001
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8159131729587703,
+      "learning_rate": 3.2039459807559183e-06,
+      "loss": 0.4724,
+      "step": 8002
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.009238092162404,
+      "learning_rate": 3.2027591262303974e-06,
+      "loss": 0.4998,
+      "step": 8003
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4252764636384176,
+      "learning_rate": 3.201572387984012e-06,
+      "loss": 0.4433,
+      "step": 8004
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.018465149736098,
+      "learning_rate": 3.200385766093539e-06,
+      "loss": 0.4605,
+      "step": 8005
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5426985878999355,
+      "learning_rate": 3.199199260635754e-06,
+      "loss": 0.4188,
+      "step": 8006
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2330984410304047,
+      "learning_rate": 3.1980128716874205e-06,
+      "loss": 0.46,
+      "step": 8007
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.305929036362998,
+      "learning_rate": 3.1968265993252967e-06,
+      "loss": 0.4606,
+      "step": 8008
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.671252549625628,
+      "learning_rate": 3.1956404436261333e-06,
+      "loss": 0.4148,
+      "step": 8009
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4690923419194715,
+      "learning_rate": 3.1944544046666715e-06,
+      "loss": 0.4506,
+      "step": 8010
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5490230250311785,
+      "learning_rate": 3.1932684825236504e-06,
+      "loss": 0.4588,
+      "step": 8011
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7944900908776351,
+      "learning_rate": 3.1920826772737916e-06,
+      "loss": 0.4948,
+      "step": 8012
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8714352414834001,
+      "learning_rate": 3.1908969889938224e-06,
+      "loss": 0.4797,
+      "step": 8013
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.1637407287182735,
+      "learning_rate": 3.1897114177604483e-06,
+      "loss": 0.4783,
+      "step": 8014
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.692237623732359,
+      "learning_rate": 3.1885259636503785e-06,
+      "loss": 0.4267,
+      "step": 8015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 4.496261233927252,
+      "learning_rate": 3.187340626740309e-06,
+      "loss": 0.4663,
+      "step": 8016
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6115033024535848,
+      "learning_rate": 3.1861554071069305e-06,
+      "loss": 0.3778,
+      "step": 8017
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.5125823933719733,
+      "learning_rate": 3.184970304826923e-06,
+      "loss": 0.4488,
+      "step": 8018
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6241827244099758,
+      "learning_rate": 3.183785319976964e-06,
+      "loss": 0.4583,
+      "step": 8019
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4232988760178529,
+      "learning_rate": 3.182600452633717e-06,
+      "loss": 0.4403,
+      "step": 8020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5565761774801736,
+      "learning_rate": 3.1814157028738447e-06,
+      "loss": 0.4374,
+      "step": 8021
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0081091714522756,
+      "learning_rate": 3.180231070773997e-06,
+      "loss": 0.4652,
+      "step": 8022
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.81122017678456,
+      "learning_rate": 3.1790465564108185e-06,
+      "loss": 0.435,
+      "step": 8023
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9355948786756647,
+      "learning_rate": 3.177862159860946e-06,
+      "loss": 0.5227,
+      "step": 8024
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.341033420041658,
+      "learning_rate": 3.1766778812010074e-06,
+      "loss": 0.4454,
+      "step": 8025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.693327908565767,
+      "learning_rate": 3.175493720507626e-06,
+      "loss": 0.4679,
+      "step": 8026
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6126328170916655,
+      "learning_rate": 3.174309677857412e-06,
+      "loss": 0.4641,
+      "step": 8027
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5479882850527806,
+      "learning_rate": 3.1731257533269753e-06,
+      "loss": 0.4954,
+      "step": 8028
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8022751356140543,
+      "learning_rate": 3.1719419469929106e-06,
+      "loss": 0.4255,
+      "step": 8029
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2481422930274566,
+      "learning_rate": 3.1707582589318135e-06,
+      "loss": 0.489,
+      "step": 8030
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.718744611083982,
+      "learning_rate": 3.1695746892202613e-06,
+      "loss": 0.4121,
+      "step": 8031
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0009974172626177,
+      "learning_rate": 3.1683912379348346e-06,
+      "loss": 0.4638,
+      "step": 8032
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.23184447424556,
+      "learning_rate": 3.1672079051520963e-06,
+      "loss": 0.4266,
+      "step": 8033
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5938644925610839,
+      "learning_rate": 3.16602469094861e-06,
+      "loss": 0.4337,
+      "step": 8034
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9317848396760005,
+      "learning_rate": 3.1648415954009278e-06,
+      "loss": 0.4429,
+      "step": 8035
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.5146475565547917,
+      "learning_rate": 3.1636586185855923e-06,
+      "loss": 0.4407,
+      "step": 8036
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.037298325627997,
+      "learning_rate": 3.1624757605791435e-06,
+      "loss": 0.4924,
+      "step": 8037
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6564079312427806,
+      "learning_rate": 3.161293021458108e-06,
+      "loss": 0.4264,
+      "step": 8038
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5724254414471611,
+      "learning_rate": 3.16011040129901e-06,
+      "loss": 0.4924,
+      "step": 8039
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5626876055955855,
+      "learning_rate": 3.15892790017836e-06,
+      "loss": 0.4699,
+      "step": 8040
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5152428500864998,
+      "learning_rate": 3.157745518172669e-06,
+      "loss": 0.4283,
+      "step": 8041
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.3541681581131493,
+      "learning_rate": 3.1565632553584315e-06,
+      "loss": 0.4758,
+      "step": 8042
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.8182511211798784,
+      "learning_rate": 3.1553811118121404e-06,
+      "loss": 0.4959,
+      "step": 8043
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6595722262963297,
+      "learning_rate": 3.1541990876102775e-06,
+      "loss": 0.4368,
+      "step": 8044
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.916600925056693,
+      "learning_rate": 3.1530171828293177e-06,
+      "loss": 0.476,
+      "step": 8045
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5563422984011408,
+      "learning_rate": 3.151835397545734e-06,
+      "loss": 0.482,
+      "step": 8046
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6898001373334064,
+      "learning_rate": 3.1506537318359785e-06,
+      "loss": 0.4936,
+      "step": 8047
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.379803320583547,
+      "learning_rate": 3.149472185776511e-06,
+      "loss": 0.404,
+      "step": 8048
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.2505078527126456,
+      "learning_rate": 3.1482907594437682e-06,
+      "loss": 0.4348,
+      "step": 8049
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.56534642167107,
+      "learning_rate": 3.147109452914195e-06,
+      "loss": 0.4742,
+      "step": 8050
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5524798132394851,
+      "learning_rate": 3.1459282662642125e-06,
+      "loss": 0.4895,
+      "step": 8051
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9226437449692777,
+      "learning_rate": 3.1447471995702484e-06,
+      "loss": 0.4338,
+      "step": 8052
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7843342537121427,
+      "learning_rate": 3.143566252908712e-06,
+      "loss": 0.4555,
+      "step": 8053
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5389714770665909,
+      "learning_rate": 3.142385426356012e-06,
+      "loss": 0.4771,
+      "step": 8054
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6747314127824016,
+      "learning_rate": 3.1412047199885433e-06,
+      "loss": 0.4711,
+      "step": 8055
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.541902146971272,
+      "learning_rate": 3.1400241338826983e-06,
+      "loss": 0.4951,
+      "step": 8056
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5259650724766551,
+      "learning_rate": 3.1388436681148593e-06,
+      "loss": 0.4333,
+      "step": 8057
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5139520861642124,
+      "learning_rate": 3.137663322761399e-06,
+      "loss": 0.4646,
+      "step": 8058
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5510474395591083,
+      "learning_rate": 3.136483097898687e-06,
+      "loss": 0.4805,
+      "step": 8059
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5545211114698652,
+      "learning_rate": 3.1353029936030794e-06,
+      "loss": 0.5051,
+      "step": 8060
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3427772244684106,
+      "learning_rate": 3.13412300995093e-06,
+      "loss": 0.4181,
+      "step": 8061
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7862540897424357,
+      "learning_rate": 3.13294314701858e-06,
+      "loss": 0.472,
+      "step": 8062
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0012784823422267,
+      "learning_rate": 3.1317634048823665e-06,
+      "loss": 0.4313,
+      "step": 8063
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9447488432758286,
+      "learning_rate": 3.1305837836186147e-06,
+      "loss": 0.4518,
+      "step": 8064
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7511068712778122,
+      "learning_rate": 3.1294042833036508e-06,
+      "loss": 0.4662,
+      "step": 8065
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6246670455096237,
+      "learning_rate": 3.1282249040137784e-06,
+      "loss": 0.4946,
+      "step": 8066
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.9213332752226853,
+      "learning_rate": 3.127045645825308e-06,
+      "loss": 0.505,
+      "step": 8067
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5316323389931839,
+      "learning_rate": 3.1258665088145345e-06,
+      "loss": 0.4271,
+      "step": 8068
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.2221898538928326,
+      "learning_rate": 3.124687493057746e-06,
+      "loss": 0.5094,
+      "step": 8069
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.998998066744559,
+      "learning_rate": 3.123508598631224e-06,
+      "loss": 0.4175,
+      "step": 8070
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6910597003026633,
+      "learning_rate": 3.12232982561124e-06,
+      "loss": 0.4154,
+      "step": 8071
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.50627238938155,
+      "learning_rate": 3.1211511740740607e-06,
+      "loss": 0.513,
+      "step": 8072
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6611569410206988,
+      "learning_rate": 3.1199726440959414e-06,
+      "loss": 0.4707,
+      "step": 8073
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5404070059318264,
+      "learning_rate": 3.1187942357531343e-06,
+      "loss": 0.49,
+      "step": 8074
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0652982116200707,
+      "learning_rate": 3.1176159491218785e-06,
+      "loss": 0.5212,
+      "step": 8075
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5104505509760409,
+      "learning_rate": 3.116437784278408e-06,
+      "loss": 0.4528,
+      "step": 8076
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.255184267332409,
+      "learning_rate": 3.1152597412989493e-06,
+      "loss": 0.4488,
+      "step": 8077
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.6582545600955108,
+      "learning_rate": 3.1140818202597178e-06,
+      "loss": 0.4818,
+      "step": 8078
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.194251705174403,
+      "learning_rate": 3.1129040212369286e-06,
+      "loss": 0.4302,
+      "step": 8079
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.758176609347948,
+      "learning_rate": 3.111726344306778e-06,
+      "loss": 0.4358,
+      "step": 8080
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 6.967028599070529,
+      "learning_rate": 3.1105487895454655e-06,
+      "loss": 0.4407,
+      "step": 8081
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 2.0077199214668977,
+      "learning_rate": 3.109371357029172e-06,
+      "loss": 0.4499,
+      "step": 8082
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7427094493350916,
+      "learning_rate": 3.108194046834081e-06,
+      "loss": 0.4359,
+      "step": 8083
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.7669299610378082,
+      "learning_rate": 3.107016859036357e-06,
+      "loss": 0.4728,
+      "step": 8084
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 3.8281721309768395,
+      "learning_rate": 3.1058397937121694e-06,
+      "loss": 0.5006,
+      "step": 8085
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4780911207277574,
+      "learning_rate": 3.1046628509376666e-06,
+      "loss": 0.4571,
+      "step": 8086
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8557642033012556,
+      "learning_rate": 3.1034860307889993e-06,
+      "loss": 0.471,
+      "step": 8087
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5114915537277782,
+      "learning_rate": 3.1023093333423036e-06,
+      "loss": 0.4278,
+      "step": 8088
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9119672267915637,
+      "learning_rate": 3.101132758673712e-06,
+      "loss": 0.4498,
+      "step": 8089
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5585712901812594,
+      "learning_rate": 3.0999563068593485e-06,
+      "loss": 0.4689,
+      "step": 8090
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4099783684387304,
+      "learning_rate": 3.0987799779753245e-06,
+      "loss": 0.4387,
+      "step": 8091
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.439110120624536,
+      "learning_rate": 3.0976037720977505e-06,
+      "loss": 0.4777,
+      "step": 8092
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8079205825377347,
+      "learning_rate": 3.0964276893027225e-06,
+      "loss": 0.4575,
+      "step": 8093
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.6049746453974447,
+      "learning_rate": 3.0952517296663332e-06,
+      "loss": 0.452,
+      "step": 8094
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4190559385822734,
+      "learning_rate": 3.0940758932646646e-06,
+      "loss": 0.4526,
+      "step": 8095
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.045835880420446,
+      "learning_rate": 3.0929001801737932e-06,
+      "loss": 0.4185,
+      "step": 8096
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.451319067717711,
+      "learning_rate": 3.0917245904697857e-06,
+      "loss": 0.4764,
+      "step": 8097
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5588377528101665,
+      "learning_rate": 3.0905491242287004e-06,
+      "loss": 0.4713,
+      "step": 8098
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.38529140347997,
+      "learning_rate": 3.0893737815265878e-06,
+      "loss": 0.4243,
+      "step": 8099
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.017633293929869,
+      "learning_rate": 3.088198562439493e-06,
+      "loss": 0.4533,
+      "step": 8100
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.114188137347059,
+      "learning_rate": 3.0870234670434506e-06,
+      "loss": 0.495,
+      "step": 8101
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8100860322323875,
+      "learning_rate": 3.0858484954144863e-06,
+      "loss": 0.4548,
+      "step": 8102
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.1960210478501754,
+      "learning_rate": 3.0846736476286222e-06,
+      "loss": 0.4426,
+      "step": 8103
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5392042988016463,
+      "learning_rate": 3.083498923761866e-06,
+      "loss": 0.4804,
+      "step": 8104
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5374065509737296,
+      "learning_rate": 3.0823243238902235e-06,
+      "loss": 0.493,
+      "step": 8105
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5958863729190773,
+      "learning_rate": 3.0811498480896872e-06,
+      "loss": 0.4168,
+      "step": 8106
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.9059708220234084,
+      "learning_rate": 3.079975496436247e-06,
+      "loss": 0.5173,
+      "step": 8107
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.557041729571183,
+      "learning_rate": 3.07880126900588e-06,
+      "loss": 0.4891,
+      "step": 8108
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.287258002531915,
+      "learning_rate": 3.0776271658745583e-06,
+      "loss": 0.4891,
+      "step": 8109
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5229634566716586,
+      "learning_rate": 3.0764531871182422e-06,
+      "loss": 0.4449,
+      "step": 8110
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4703024805919196,
+      "learning_rate": 3.0752793328128903e-06,
+      "loss": 0.4776,
+      "step": 8111
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8095430356990658,
+      "learning_rate": 3.0741056030344476e-06,
+      "loss": 0.4814,
+      "step": 8112
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8436207238147895,
+      "learning_rate": 3.0729319978588505e-06,
+      "loss": 0.4228,
+      "step": 8113
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5976924978678592,
+      "learning_rate": 3.0717585173620364e-06,
+      "loss": 0.4904,
+      "step": 8114
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5322658901538564,
+      "learning_rate": 3.07058516161992e-06,
+      "loss": 0.4538,
+      "step": 8115
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5642034487088647,
+      "learning_rate": 3.0694119307084224e-06,
+      "loss": 0.5059,
+      "step": 8116
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.545348273048814,
+      "learning_rate": 3.0682388247034455e-06,
+      "loss": 0.4789,
+      "step": 8117
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5304868884014815,
+      "learning_rate": 3.0670658436808918e-06,
+      "loss": 0.4751,
+      "step": 8118
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.771957334672217,
+      "learning_rate": 3.0658929877166464e-06,
+      "loss": 0.4728,
+      "step": 8119
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.089018682062842,
+      "learning_rate": 3.0647202568865973e-06,
+      "loss": 0.4896,
+      "step": 8120
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5361638208940827,
+      "learning_rate": 3.063547651266614e-06,
+      "loss": 0.4434,
+      "step": 8121
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.585971876334789,
+      "learning_rate": 3.0623751709325655e-06,
+      "loss": 0.4414,
+      "step": 8122
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5368608510986306,
+      "learning_rate": 3.061202815960308e-06,
+      "loss": 0.444,
+      "step": 8123
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5709463145910346,
+      "learning_rate": 3.0600305864256918e-06,
+      "loss": 0.4698,
+      "step": 8124
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.491771962829028,
+      "learning_rate": 3.0588584824045607e-06,
+      "loss": 0.4388,
+      "step": 8125
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9512206471581415,
+      "learning_rate": 3.0576865039727443e-06,
+      "loss": 0.5017,
+      "step": 8126
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.258454238825651,
+      "learning_rate": 3.056514651206072e-06,
+      "loss": 0.4689,
+      "step": 8127
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.777205834553369,
+      "learning_rate": 3.055342924180358e-06,
+      "loss": 0.4352,
+      "step": 8128
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.565830549088371,
+      "learning_rate": 3.054171322971414e-06,
+      "loss": 0.4808,
+      "step": 8129
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0270861352963903,
+      "learning_rate": 3.052999847655039e-06,
+      "loss": 0.4897,
+      "step": 8130
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.503145951232974,
+      "learning_rate": 3.051828498307028e-06,
+      "loss": 0.4092,
+      "step": 8131
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7852418683173914,
+      "learning_rate": 3.0506572750031637e-06,
+      "loss": 0.4986,
+      "step": 8132
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7238034247339618,
+      "learning_rate": 3.0494861778192244e-06,
+      "loss": 0.4939,
+      "step": 8133
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5546890509743488,
+      "learning_rate": 3.0483152068309757e-06,
+      "loss": 0.4875,
+      "step": 8134
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5299291633299098,
+      "learning_rate": 3.047144362114181e-06,
+      "loss": 0.4628,
+      "step": 8135
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7034223165638394,
+      "learning_rate": 3.0459736437445923e-06,
+      "loss": 0.4468,
+      "step": 8136
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6480248461493903,
+      "learning_rate": 3.0448030517979523e-06,
+      "loss": 0.4549,
+      "step": 8137
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.4255204878703114,
+      "learning_rate": 3.043632586349997e-06,
+      "loss": 0.4847,
+      "step": 8138
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.889181245140569,
+      "learning_rate": 3.0424622474764544e-06,
+      "loss": 0.4336,
+      "step": 8139
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8857978326130735,
+      "learning_rate": 3.041292035253044e-06,
+      "loss": 0.4556,
+      "step": 8140
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.933557231357053,
+      "learning_rate": 3.0401219497554757e-06,
+      "loss": 0.4665,
+      "step": 8141
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.194890223121409,
+      "learning_rate": 3.0389519910594546e-06,
+      "loss": 0.4635,
+      "step": 8142
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5286822373454189,
+      "learning_rate": 3.0377821592406727e-06,
+      "loss": 0.4749,
+      "step": 8143
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6087002311691982,
+      "learning_rate": 3.03661245437482e-06,
+      "loss": 0.4766,
+      "step": 8144
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5570596661068346,
+      "learning_rate": 3.035442876537571e-06,
+      "loss": 0.4906,
+      "step": 8145
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8020194337526472,
+      "learning_rate": 3.0342734258045987e-06,
+      "loss": 0.499,
+      "step": 8146
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7282402057997694,
+      "learning_rate": 3.033104102251565e-06,
+      "loss": 0.4557,
+      "step": 8147
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2994123144187069,
+      "learning_rate": 3.031934905954121e-06,
+      "loss": 0.4468,
+      "step": 8148
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 6.192736125801393,
+      "learning_rate": 3.0307658369879166e-06,
+      "loss": 0.4718,
+      "step": 8149
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.055779567573067,
+      "learning_rate": 3.0295968954285843e-06,
+      "loss": 0.5116,
+      "step": 8150
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4246759953451695,
+      "learning_rate": 3.0284280813517573e-06,
+      "loss": 0.4248,
+      "step": 8151
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7816028714276833,
+      "learning_rate": 3.027259394833052e-06,
+      "loss": 0.4964,
+      "step": 8152
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3141912048608082,
+      "learning_rate": 3.026090835948086e-06,
+      "loss": 0.4142,
+      "step": 8153
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5367527116014242,
+      "learning_rate": 3.024922404772458e-06,
+      "loss": 0.5168,
+      "step": 8154
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5601912569698757,
+      "learning_rate": 3.0237541013817696e-06,
+      "loss": 0.4537,
+      "step": 8155
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9212808418463638,
+      "learning_rate": 3.022585925851602e-06,
+      "loss": 0.5344,
+      "step": 8156
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9249692455907497,
+      "learning_rate": 3.0214178782575398e-06,
+      "loss": 0.4792,
+      "step": 8157
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5553376168459427,
+      "learning_rate": 3.020249958675153e-06,
+      "loss": 0.4919,
+      "step": 8158
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5754873056269578,
+      "learning_rate": 3.0190821671800036e-06,
+      "loss": 0.483,
+      "step": 8159
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9695789467971254,
+      "learning_rate": 3.0179145038476476e-06,
+      "loss": 0.4505,
+      "step": 8160
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7924532510880093,
+      "learning_rate": 3.016746968753629e-06,
+      "loss": 0.4163,
+      "step": 8161
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8949640970240083,
+      "learning_rate": 3.0155795619734885e-06,
+      "loss": 0.4256,
+      "step": 8162
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 8.498561765607665,
+      "learning_rate": 3.014412283582753e-06,
+      "loss": 0.45,
+      "step": 8163
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.840557414291357,
+      "learning_rate": 3.0132451336569456e-06,
+      "loss": 0.4796,
+      "step": 8164
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3982628845738012,
+      "learning_rate": 3.0120781122715792e-06,
+      "loss": 0.4468,
+      "step": 8165
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.411915879279902,
+      "learning_rate": 3.0109112195021597e-06,
+      "loss": 0.4517,
+      "step": 8166
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7663590898288921,
+      "learning_rate": 3.00974445542418e-06,
+      "loss": 0.4754,
+      "step": 8167
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 4.5685165903965315,
+      "learning_rate": 3.00857782011313e-06,
+      "loss": 0.4122,
+      "step": 8168
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8319429357519368,
+      "learning_rate": 3.0074113136444926e-06,
+      "loss": 0.3956,
+      "step": 8169
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7582723100033337,
+      "learning_rate": 3.006244936093734e-06,
+      "loss": 0.4117,
+      "step": 8170
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5314806916210415,
+      "learning_rate": 3.0050786875363223e-06,
+      "loss": 0.4934,
+      "step": 8171
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2160980175623233,
+      "learning_rate": 3.0039125680477083e-06,
+      "loss": 0.4287,
+      "step": 8172
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5365521514075609,
+      "learning_rate": 3.0027465777033405e-06,
+      "loss": 0.4623,
+      "step": 8173
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.4155931593043927,
+      "learning_rate": 3.0015807165786558e-06,
+      "loss": 0.4939,
+      "step": 8174
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4570468422069045,
+      "learning_rate": 3.000414984749086e-06,
+      "loss": 0.4386,
+      "step": 8175
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0362376372710047,
+      "learning_rate": 2.9992493822900493e-06,
+      "loss": 0.4244,
+      "step": 8176
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5268890266341643,
+      "learning_rate": 2.9980839092769617e-06,
+      "loss": 0.4952,
+      "step": 8177
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8424093816996632,
+      "learning_rate": 2.996918565785225e-06,
+      "loss": 0.4433,
+      "step": 8178
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5345731668737623,
+      "learning_rate": 2.9957533518902376e-06,
+      "loss": 0.3957,
+      "step": 8179
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7419438850059386,
+      "learning_rate": 2.9945882676673853e-06,
+      "loss": 0.4809,
+      "step": 8180
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.837067050587271,
+      "learning_rate": 2.9934233131920492e-06,
+      "loss": 0.4431,
+      "step": 8181
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.865787109757726,
+      "learning_rate": 2.992258488539601e-06,
+      "loss": 0.4587,
+      "step": 8182
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5083553290531755,
+      "learning_rate": 2.9910937937853997e-06,
+      "loss": 0.4388,
+      "step": 8183
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7808964105985536,
+      "learning_rate": 2.9899292290048044e-06,
+      "loss": 0.501,
+      "step": 8184
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.73766643203524,
+      "learning_rate": 2.9887647942731563e-06,
+      "loss": 0.4108,
+      "step": 8185
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.8654576279704345,
+      "learning_rate": 2.987600489665798e-06,
+      "loss": 0.4561,
+      "step": 8186
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.986497821393684,
+      "learning_rate": 2.9864363152580524e-06,
+      "loss": 0.4547,
+      "step": 8187
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9048572442369114,
+      "learning_rate": 2.9852722711252456e-06,
+      "loss": 0.466,
+      "step": 8188
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.575208243017937,
+      "learning_rate": 2.984108357342684e-06,
+      "loss": 0.4819,
+      "step": 8189
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.2347412976706735,
+      "learning_rate": 2.9829445739856788e-06,
+      "loss": 0.4964,
+      "step": 8190
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.598314507428206,
+      "learning_rate": 2.9817809211295174e-06,
+      "loss": 0.4724,
+      "step": 8191
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9845552463680445,
+      "learning_rate": 2.980617398849491e-06,
+      "loss": 0.4614,
+      "step": 8192
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7933392943722042,
+      "learning_rate": 2.979454007220879e-06,
+      "loss": 0.4422,
+      "step": 8193
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6753873903146643,
+      "learning_rate": 2.978290746318948e-06,
+      "loss": 0.4615,
+      "step": 8194
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9328252913577677,
+      "learning_rate": 2.9771276162189615e-06,
+      "loss": 0.4153,
+      "step": 8195
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.7414439884538382,
+      "learning_rate": 2.975964616996172e-06,
+      "loss": 0.4571,
+      "step": 8196
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5633142584691037,
+      "learning_rate": 2.974801748725824e-06,
+      "loss": 0.4559,
+      "step": 8197
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.641298584323138,
+      "learning_rate": 2.973639011483153e-06,
+      "loss": 0.5123,
+      "step": 8198
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4178594427642879,
+      "learning_rate": 2.9724764053433886e-06,
+      "loss": 0.4106,
+      "step": 8199
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.5018990867862554,
+      "learning_rate": 2.971313930381747e-06,
+      "loss": 0.4464,
+      "step": 8200
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.003077663117057,
+      "learning_rate": 2.970151586673441e-06,
+      "loss": 0.4763,
+      "step": 8201
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.0523952963003933,
+      "learning_rate": 2.9689893742936715e-06,
+      "loss": 0.4352,
+      "step": 8202
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9556635359587315,
+      "learning_rate": 2.967827293317631e-06,
+      "loss": 0.4427,
+      "step": 8203
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 7.1189133381954335,
+      "learning_rate": 2.9666653438205095e-06,
+      "loss": 0.4116,
+      "step": 8204
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.72891559671769,
+      "learning_rate": 2.9655035258774772e-06,
+      "loss": 0.4634,
+      "step": 8205
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.9393539305785878,
+      "learning_rate": 2.9643418395637076e-06,
+      "loss": 0.4805,
+      "step": 8206
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.8869886549698118,
+      "learning_rate": 2.9631802849543566e-06,
+      "loss": 0.4627,
+      "step": 8207
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5599968952526517,
+      "learning_rate": 2.9620188621245784e-06,
+      "loss": 0.4699,
+      "step": 8208
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6625822354179296,
+      "learning_rate": 2.9608575711495126e-06,
+      "loss": 0.4827,
+      "step": 8209
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6679442508616296,
+      "learning_rate": 2.9596964121042947e-06,
+      "loss": 0.4738,
+      "step": 8210
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.666576270733875,
+      "learning_rate": 2.9585353850640493e-06,
+      "loss": 0.4201,
+      "step": 8211
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.896265582095068,
+      "learning_rate": 2.9573744901038948e-06,
+      "loss": 0.4471,
+      "step": 8212
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5792897886569046,
+      "learning_rate": 2.956213727298938e-06,
+      "loss": 0.5022,
+      "step": 8213
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7002635747453678,
+      "learning_rate": 2.9550530967242792e-06,
+      "loss": 0.4753,
+      "step": 8214
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.500796774413423,
+      "learning_rate": 2.953892598455011e-06,
+      "loss": 0.4934,
+      "step": 8215
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.598761322657962,
+      "learning_rate": 2.9527322325662137e-06,
+      "loss": 0.4485,
+      "step": 8216
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9065096359666602,
+      "learning_rate": 2.9515719991329643e-06,
+      "loss": 0.4833,
+      "step": 8217
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3686941393225618,
+      "learning_rate": 2.9504118982303247e-06,
+      "loss": 0.413,
+      "step": 8218
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3139040235088357,
+      "learning_rate": 2.9492519299333573e-06,
+      "loss": 0.4613,
+      "step": 8219
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6044871145105861,
+      "learning_rate": 2.9480920943171043e-06,
+      "loss": 0.4376,
+      "step": 8220
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2663822308648935,
+      "learning_rate": 2.9469323914566113e-06,
+      "loss": 0.3854,
+      "step": 8221
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.520256989220859,
+      "learning_rate": 2.945772821426903e-06,
+      "loss": 0.4859,
+      "step": 8222
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5619964669443505,
+      "learning_rate": 2.94461338430301e-06,
+      "loss": 0.4116,
+      "step": 8223
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5446876132376467,
+      "learning_rate": 2.94345408015994e-06,
+      "loss": 0.4664,
+      "step": 8224
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5321443175789121,
+      "learning_rate": 2.9422949090727002e-06,
+      "loss": 0.4574,
+      "step": 8225
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5854540470994614,
+      "learning_rate": 2.9411358711162895e-06,
+      "loss": 0.4557,
+      "step": 8226
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5673594195810486,
+      "learning_rate": 2.9399769663656937e-06,
+      "loss": 0.4754,
+      "step": 8227
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6969962384126904,
+      "learning_rate": 2.938818194895894e-06,
+      "loss": 0.4986,
+      "step": 8228
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9260766414192725,
+      "learning_rate": 2.9376595567818598e-06,
+      "loss": 0.4723,
+      "step": 8229
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5517345577366306,
+      "learning_rate": 2.936501052098556e-06,
+      "loss": 0.459,
+      "step": 8230
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.796896044021153,
+      "learning_rate": 2.935342680920933e-06,
+      "loss": 0.4982,
+      "step": 8231
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9217778951487259,
+      "learning_rate": 2.934184443323939e-06,
+      "loss": 0.4424,
+      "step": 8232
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5820370672495068,
+      "learning_rate": 2.9330263393825083e-06,
+      "loss": 0.4529,
+      "step": 8233
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4939031886591057,
+      "learning_rate": 2.93186836917157e-06,
+      "loss": 0.4198,
+      "step": 8234
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6538945168280013,
+      "learning_rate": 2.9307105327660424e-06,
+      "loss": 0.4836,
+      "step": 8235
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6010270670961608,
+      "learning_rate": 2.9295528302408344e-06,
+      "loss": 0.4439,
+      "step": 8236
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.090542247128029,
+      "learning_rate": 2.9283952616708533e-06,
+      "loss": 0.467,
+      "step": 8237
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.649877439268315,
+      "learning_rate": 2.927237827130986e-06,
+      "loss": 0.4123,
+      "step": 8238
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6017103665985941,
+      "learning_rate": 2.9260805266961223e-06,
+      "loss": 0.4958,
+      "step": 8239
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4772572546893066,
+      "learning_rate": 2.9249233604411333e-06,
+      "loss": 0.4387,
+      "step": 8240
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7097143592457535,
+      "learning_rate": 2.9237663284408897e-06,
+      "loss": 0.4016,
+      "step": 8241
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8169986595543601,
+      "learning_rate": 2.9226094307702476e-06,
+      "loss": 0.4432,
+      "step": 8242
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.455766885992405,
+      "learning_rate": 2.921452667504059e-06,
+      "loss": 0.4199,
+      "step": 8243
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7458486229120573,
+      "learning_rate": 2.920296038717162e-06,
+      "loss": 0.4566,
+      "step": 8244
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5430650062282202,
+      "learning_rate": 2.9191395444843927e-06,
+      "loss": 0.4966,
+      "step": 8245
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8621342077114018,
+      "learning_rate": 2.917983184880571e-06,
+      "loss": 0.5137,
+      "step": 8246
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.5263195270427516,
+      "learning_rate": 2.916826959980514e-06,
+      "loss": 0.4929,
+      "step": 8247
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7143124195598418,
+      "learning_rate": 2.9156708698590273e-06,
+      "loss": 0.396,
+      "step": 8248
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6512225744344413,
+      "learning_rate": 2.914514914590909e-06,
+      "loss": 0.5167,
+      "step": 8249
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 4.046354989438265,
+      "learning_rate": 2.91335909425095e-06,
+      "loss": 0.5126,
+      "step": 8250
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8018917333452855,
+      "learning_rate": 2.912203408913925e-06,
+      "loss": 0.4447,
+      "step": 8251
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0472325603823958,
+      "learning_rate": 2.9110478586546086e-06,
+      "loss": 0.4253,
+      "step": 8252
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7327233528669361,
+      "learning_rate": 2.9098924435477636e-06,
+      "loss": 0.4946,
+      "step": 8253
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.980114705290418,
+      "learning_rate": 2.908737163668145e-06,
+      "loss": 0.5031,
+      "step": 8254
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5340968871623029,
+      "learning_rate": 2.9075820190904936e-06,
+      "loss": 0.5056,
+      "step": 8255
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.208119107141144,
+      "learning_rate": 2.9064270098895488e-06,
+      "loss": 0.4189,
+      "step": 8256
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.639106613865686,
+      "learning_rate": 2.9052721361400375e-06,
+      "loss": 0.4234,
+      "step": 8257
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7769743000105889,
+      "learning_rate": 2.9041173979166813e-06,
+      "loss": 0.4568,
+      "step": 8258
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 5.5079654871055235,
+      "learning_rate": 2.9029627952941854e-06,
+      "loss": 0.4062,
+      "step": 8259
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.628065600571878,
+      "learning_rate": 2.901808328347251e-06,
+      "loss": 0.4684,
+      "step": 8260
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5794870164140321,
+      "learning_rate": 2.900653997150578e-06,
+      "loss": 0.4794,
+      "step": 8261
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8121446360000792,
+      "learning_rate": 2.8994998017788417e-06,
+      "loss": 0.482,
+      "step": 8262
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4823476219949823,
+      "learning_rate": 2.898345742306723e-06,
+      "loss": 0.3933,
+      "step": 8263
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0361918507623584,
+      "learning_rate": 2.8971918188088815e-06,
+      "loss": 0.5002,
+      "step": 8264
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0558414802069973,
+      "learning_rate": 2.896038031359982e-06,
+      "loss": 0.466,
+      "step": 8265
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1774481971590314,
+      "learning_rate": 2.8948843800346687e-06,
+      "loss": 0.351,
+      "step": 8266
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5087912390487856,
+      "learning_rate": 2.8937308649075836e-06,
+      "loss": 0.4186,
+      "step": 8267
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5652916642586763,
+      "learning_rate": 2.8925774860533524e-06,
+      "loss": 0.4465,
+      "step": 8268
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5489859745171022,
+      "learning_rate": 2.8914242435466043e-06,
+      "loss": 0.4959,
+      "step": 8269
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5001235535866595,
+      "learning_rate": 2.890271137461948e-06,
+      "loss": 0.4462,
+      "step": 8270
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.343075215775687,
+      "learning_rate": 2.889118167873989e-06,
+      "loss": 0.4689,
+      "step": 8271
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9397795988072182,
+      "learning_rate": 2.8879653348573235e-06,
+      "loss": 0.4968,
+      "step": 8272
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6615440625796039,
+      "learning_rate": 2.886812638486537e-06,
+      "loss": 0.4679,
+      "step": 8273
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.518529275538549,
+      "learning_rate": 2.885660078836211e-06,
+      "loss": 0.455,
+      "step": 8274
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1638509373764467,
+      "learning_rate": 2.8845076559809103e-06,
+      "loss": 0.4403,
+      "step": 8275
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5621496897834805,
+      "learning_rate": 2.8833553699951968e-06,
+      "loss": 0.4857,
+      "step": 8276
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3736689648682092,
+      "learning_rate": 2.882203220953621e-06,
+      "loss": 0.4315,
+      "step": 8277
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6886079423347138,
+      "learning_rate": 2.881051208930729e-06,
+      "loss": 0.4527,
+      "step": 8278
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.3669679589894677,
+      "learning_rate": 2.87989933400105e-06,
+      "loss": 0.4652,
+      "step": 8279
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5709648877703546,
+      "learning_rate": 2.878747596239111e-06,
+      "loss": 0.4909,
+      "step": 8280
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.702603130119065,
+      "learning_rate": 2.8775959957194268e-06,
+      "loss": 0.421,
+      "step": 8281
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5842850027379044,
+      "learning_rate": 2.8764445325165057e-06,
+      "loss": 0.4882,
+      "step": 8282
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5246457164700038,
+      "learning_rate": 2.875293206704845e-06,
+      "loss": 0.4402,
+      "step": 8283
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8083703239667417,
+      "learning_rate": 2.8741420183589348e-06,
+      "loss": 0.415,
+      "step": 8284
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.676170385760368,
+      "learning_rate": 2.8729909675532563e-06,
+      "loss": 0.4529,
+      "step": 8285
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.747353601678995,
+      "learning_rate": 2.871840054362278e-06,
+      "loss": 0.4205,
+      "step": 8286
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7839235546008447,
+      "learning_rate": 2.8706892788604645e-06,
+      "loss": 0.453,
+      "step": 8287
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7023344674913623,
+      "learning_rate": 2.8695386411222693e-06,
+      "loss": 0.5045,
+      "step": 8288
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.2861785813172517,
+      "learning_rate": 2.8683881412221395e-06,
+      "loss": 0.4791,
+      "step": 8289
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.211967090717653,
+      "learning_rate": 2.867237779234505e-06,
+      "loss": 0.3838,
+      "step": 8290
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.865368609701087,
+      "learning_rate": 2.8660875552337974e-06,
+      "loss": 0.4767,
+      "step": 8291
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0595050006827917,
+      "learning_rate": 2.864937469294433e-06,
+      "loss": 0.4727,
+      "step": 8292
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8753089932077254,
+      "learning_rate": 2.8637875214908217e-06,
+      "loss": 0.4752,
+      "step": 8293
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.573684085456631,
+      "learning_rate": 2.8626377118973645e-06,
+      "loss": 0.4867,
+      "step": 8294
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8250185123723592,
+      "learning_rate": 2.861488040588448e-06,
+      "loss": 0.4569,
+      "step": 8295
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.0217553130530463,
+      "learning_rate": 2.8603385076384617e-06,
+      "loss": 0.4523,
+      "step": 8296
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5614057325078737,
+      "learning_rate": 2.859189113121772e-06,
+      "loss": 0.4893,
+      "step": 8297
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8505969155762023,
+      "learning_rate": 2.8580398571127486e-06,
+      "loss": 0.3796,
+      "step": 8298
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5173714784496488,
+      "learning_rate": 2.8568907396857415e-06,
+      "loss": 0.5161,
+      "step": 8299
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5475002076918463,
+      "learning_rate": 2.855741760915103e-06,
+      "loss": 0.501,
+      "step": 8300
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.1594465900801243,
+      "learning_rate": 2.8545929208751653e-06,
+      "loss": 0.4234,
+      "step": 8301
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8103977212711726,
+      "learning_rate": 2.8534442196402617e-06,
+      "loss": 0.4662,
+      "step": 8302
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.521399554962547,
+      "learning_rate": 2.8522956572847047e-06,
+      "loss": 0.4429,
+      "step": 8303
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4632441268805159,
+      "learning_rate": 2.851147233882812e-06,
+      "loss": 0.436,
+      "step": 8304
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7034870054871227,
+      "learning_rate": 2.849998949508883e-06,
+      "loss": 0.5054,
+      "step": 8305
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9543803211238833,
+      "learning_rate": 2.8488508042372076e-06,
+      "loss": 0.4428,
+      "step": 8306
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7372610364514034,
+      "learning_rate": 2.847702798142071e-06,
+      "loss": 0.3922,
+      "step": 8307
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.0050480757573963,
+      "learning_rate": 2.8465549312977487e-06,
+      "loss": 0.4517,
+      "step": 8308
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9251687330436067,
+      "learning_rate": 2.845407203778506e-06,
+      "loss": 0.4397,
+      "step": 8309
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.541528873678368,
+      "learning_rate": 2.844259615658598e-06,
+      "loss": 0.4849,
+      "step": 8310
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7542281992065132,
+      "learning_rate": 2.8431121670122724e-06,
+      "loss": 0.4707,
+      "step": 8311
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3925116782037286,
+      "learning_rate": 2.8419648579137677e-06,
+      "loss": 0.4252,
+      "step": 8312
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6915365684697257,
+      "learning_rate": 2.840817688437316e-06,
+      "loss": 0.3936,
+      "step": 8313
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7636968621207436,
+      "learning_rate": 2.839670658657133e-06,
+      "loss": 0.4472,
+      "step": 8314
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5575800331889267,
+      "learning_rate": 2.8385237686474303e-06,
+      "loss": 0.4699,
+      "step": 8315
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4642466297829557,
+      "learning_rate": 2.837377018482416e-06,
+      "loss": 0.4769,
+      "step": 8316
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5504902964577563,
+      "learning_rate": 2.836230408236278e-06,
+      "loss": 0.4717,
+      "step": 8317
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.4868786209281994,
+      "learning_rate": 2.835083937983202e-06,
+      "loss": 0.4724,
+      "step": 8318
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.466271145164999,
+      "learning_rate": 2.8339376077973628e-06,
+      "loss": 0.4861,
+      "step": 8319
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.598786724765163,
+      "learning_rate": 2.832791417752929e-06,
+      "loss": 0.4953,
+      "step": 8320
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.493471439789418,
+      "learning_rate": 2.8316453679240537e-06,
+      "loss": 0.4603,
+      "step": 8321
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.8763989217535593,
+      "learning_rate": 2.830499458384886e-06,
+      "loss": 0.4989,
+      "step": 8322
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2399243465219496,
+      "learning_rate": 2.8293536892095653e-06,
+      "loss": 0.4023,
+      "step": 8323
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 3.035756458421096,
+      "learning_rate": 2.828208060472224e-06,
+      "loss": 0.5114,
+      "step": 8324
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2155455974149905,
+      "learning_rate": 2.8270625722469776e-06,
+      "loss": 0.4782,
+      "step": 8325
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.628003121768669,
+      "learning_rate": 2.8259172246079402e-06,
+      "loss": 0.4042,
+      "step": 8326
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5870712803585254,
+      "learning_rate": 2.8247720176292147e-06,
+      "loss": 0.461,
+      "step": 8327
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3971352518875195,
+      "learning_rate": 2.8236269513848935e-06,
+      "loss": 0.4305,
+      "step": 8328
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.6430939273350604,
+      "learning_rate": 2.8224820259490636e-06,
+      "loss": 0.4626,
+      "step": 8329
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7146304547572568,
+      "learning_rate": 2.8213372413957947e-06,
+      "loss": 0.456,
+      "step": 8330
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9832411482587164,
+      "learning_rate": 2.8201925977991598e-06,
+      "loss": 0.477,
+      "step": 8331
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8731376214478976,
+      "learning_rate": 2.81904809523321e-06,
+      "loss": 0.4769,
+      "step": 8332
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 2.2825281635178962,
+      "learning_rate": 2.8179037337719977e-06,
+      "loss": 0.4094,
+      "step": 8333
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.8869194039794646,
+      "learning_rate": 2.8167595134895553e-06,
+      "loss": 0.485,
+      "step": 8334
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.9438768694801347,
+      "learning_rate": 2.8156154344599207e-06,
+      "loss": 0.4637,
+      "step": 8335
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7127709198925196,
+      "learning_rate": 2.8144714967571075e-06,
+      "loss": 0.3932,
+      "step": 8336
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4615523343789412,
+      "learning_rate": 2.813327700455131e-06,
+      "loss": 0.4314,
+      "step": 8337
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.565371734441393,
+      "learning_rate": 2.812184045627988e-06,
+      "loss": 0.4683,
+      "step": 8338
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7947520764056022,
+      "learning_rate": 2.811040532349678e-06,
+      "loss": 0.4466,
+      "step": 8339
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5716927810210861,
+      "learning_rate": 2.809897160694184e-06,
+      "loss": 0.4829,
+      "step": 8340
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.761223595511178,
+      "learning_rate": 2.8087539307354754e-06,
+      "loss": 0.4292,
+      "step": 8341
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2627338861524247,
+      "learning_rate": 2.8076108425475223e-06,
+      "loss": 0.444,
+      "step": 8342
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4287526818049328,
+      "learning_rate": 2.806467896204279e-06,
+      "loss": 0.4552,
+      "step": 8343
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.267878602824153,
+      "learning_rate": 2.8053250917796952e-06,
+      "loss": 0.4788,
+      "step": 8344
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8483628461080137,
+      "learning_rate": 2.8041824293477056e-06,
+      "loss": 0.4542,
+      "step": 8345
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6467845467099607,
+      "learning_rate": 2.8030399089822398e-06,
+      "loss": 0.4717,
+      "step": 8346
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6172119119763586,
+      "learning_rate": 2.801897530757218e-06,
+      "loss": 0.4363,
+      "step": 8347
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.404458643547706,
+      "learning_rate": 2.800755294746552e-06,
+      "loss": 0.4321,
+      "step": 8348
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9416211053333847,
+      "learning_rate": 2.7996132010241406e-06,
+      "loss": 0.4185,
+      "step": 8349
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.709488016779424,
+      "learning_rate": 2.7984712496638737e-06,
+      "loss": 0.3935,
+      "step": 8350
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.800367358551493,
+      "learning_rate": 2.797329440739641e-06,
+      "loss": 0.5001,
+      "step": 8351
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5586600921127364,
+      "learning_rate": 2.7961877743253105e-06,
+      "loss": 0.4761,
+      "step": 8352
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9689953269674476,
+      "learning_rate": 2.7950462504947483e-06,
+      "loss": 0.4774,
+      "step": 8353
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.546762584221291,
+      "learning_rate": 2.7939048693218084e-06,
+      "loss": 0.4389,
+      "step": 8354
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5685714991573594,
+      "learning_rate": 2.7927636308803406e-06,
+      "loss": 0.4797,
+      "step": 8355
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5265245492614934,
+      "learning_rate": 2.7916225352441757e-06,
+      "loss": 0.438,
+      "step": 8356
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5665146606972222,
+      "learning_rate": 2.790481582487144e-06,
+      "loss": 0.4888,
+      "step": 8357
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9738203611163148,
+      "learning_rate": 2.7893407726830633e-06,
+      "loss": 0.4465,
+      "step": 8358
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5877463799673889,
+      "learning_rate": 2.788200105905745e-06,
+      "loss": 0.5084,
+      "step": 8359
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.289425328467674,
+      "learning_rate": 2.787059582228984e-06,
+      "loss": 0.4557,
+      "step": 8360
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3600632331522529,
+      "learning_rate": 2.7859192017265734e-06,
+      "loss": 0.4491,
+      "step": 8361
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5846674638150436,
+      "learning_rate": 2.7847789644722945e-06,
+      "loss": 0.4691,
+      "step": 8362
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7295862358588767,
+      "learning_rate": 2.783638870539918e-06,
+      "loss": 0.4621,
+      "step": 8363
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1293941330170356,
+      "learning_rate": 2.7824989200032098e-06,
+      "loss": 0.4651,
+      "step": 8364
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.189229836462391,
+      "learning_rate": 2.7813591129359185e-06,
+      "loss": 0.4764,
+      "step": 8365
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8378040021790647,
+      "learning_rate": 2.78021944941179e-06,
+      "loss": 0.5054,
+      "step": 8366
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.141738031654644,
+      "learning_rate": 2.7790799295045595e-06,
+      "loss": 0.4571,
+      "step": 8367
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.002590834228274,
+      "learning_rate": 2.7779405532879545e-06,
+      "loss": 0.5377,
+      "step": 8368
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5624367821109841,
+      "learning_rate": 2.7768013208356835e-06,
+      "loss": 0.4628,
+      "step": 8369
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8460637358156968,
+      "learning_rate": 2.7756622322214646e-06,
+      "loss": 0.4131,
+      "step": 8370
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8655870718104048,
+      "learning_rate": 2.7745232875189866e-06,
+      "loss": 0.453,
+      "step": 8371
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6279504498558004,
+      "learning_rate": 2.7733844868019406e-06,
+      "loss": 0.4528,
+      "step": 8372
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 5.404658256272628,
+      "learning_rate": 2.7722458301440053e-06,
+      "loss": 0.4496,
+      "step": 8373
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.438696999281865,
+      "learning_rate": 2.771107317618851e-06,
+      "loss": 0.4156,
+      "step": 8374
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.284069264363936,
+      "learning_rate": 2.769968949300139e-06,
+      "loss": 0.4312,
+      "step": 8375
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.587769896459424,
+      "learning_rate": 2.7688307252615172e-06,
+      "loss": 0.476,
+      "step": 8376
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8013428202903021,
+      "learning_rate": 2.7676926455766284e-06,
+      "loss": 0.4594,
+      "step": 8377
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9031606263788614,
+      "learning_rate": 2.7665547103191054e-06,
+      "loss": 0.4725,
+      "step": 8378
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.141558528674537,
+      "learning_rate": 2.765416919562573e-06,
+      "loss": 0.4454,
+      "step": 8379
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.9625475439093094,
+      "learning_rate": 2.7642792733806405e-06,
+      "loss": 0.4441,
+      "step": 8380
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6386532775864326,
+      "learning_rate": 2.7631417718469147e-06,
+      "loss": 0.446,
+      "step": 8381
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5163095294031517,
+      "learning_rate": 2.7620044150349897e-06,
+      "loss": 0.4413,
+      "step": 8382
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7993799323997333,
+      "learning_rate": 2.760867203018451e-06,
+      "loss": 0.5195,
+      "step": 8383
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.399157036297125,
+      "learning_rate": 2.759730135870878e-06,
+      "loss": 0.4437,
+      "step": 8384
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6099059156589026,
+      "learning_rate": 2.75859321366583e-06,
+      "loss": 0.5127,
+      "step": 8385
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.072333514439562,
+      "learning_rate": 2.757456436476873e-06,
+      "loss": 0.462,
+      "step": 8386
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7351578955591256,
+      "learning_rate": 2.7563198043775493e-06,
+      "loss": 0.4621,
+      "step": 8387
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.688031589463551,
+      "learning_rate": 2.755183317441399e-06,
+      "loss": 0.4848,
+      "step": 8388
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.541015625171994,
+      "learning_rate": 2.754046975741951e-06,
+      "loss": 0.4711,
+      "step": 8389
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4927620728469397,
+      "learning_rate": 2.7529107793527284e-06,
+      "loss": 0.4529,
+      "step": 8390
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7690020372398174,
+      "learning_rate": 2.7517747283472364e-06,
+      "loss": 0.4486,
+      "step": 8391
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5273022297394172,
+      "learning_rate": 2.7506388227989778e-06,
+      "loss": 0.4845,
+      "step": 8392
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9213654365255597,
+      "learning_rate": 2.7495030627814455e-06,
+      "loss": 0.4256,
+      "step": 8393
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7624012089148893,
+      "learning_rate": 2.7483674483681213e-06,
+      "loss": 0.4227,
+      "step": 8394
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.472248459469368,
+      "learning_rate": 2.7472319796324793e-06,
+      "loss": 0.4334,
+      "step": 8395
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7581027745886375,
+      "learning_rate": 2.746096656647979e-06,
+      "loss": 0.4965,
+      "step": 8396
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6414922311337243,
+      "learning_rate": 2.744961479488077e-06,
+      "loss": 0.4498,
+      "step": 8397
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.3309071229241747,
+      "learning_rate": 2.743826448226217e-06,
+      "loss": 0.4042,
+      "step": 8398
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.824415640490874,
+      "learning_rate": 2.7426915629358364e-06,
+      "loss": 0.4779,
+      "step": 8399
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6230445499614004,
+      "learning_rate": 2.7415568236903567e-06,
+      "loss": 0.396,
+      "step": 8400
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3902487783454,
+      "learning_rate": 2.740422230563197e-06,
+      "loss": 0.4316,
+      "step": 8401
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0682433053236386,
+      "learning_rate": 2.739287783627762e-06,
+      "loss": 0.4449,
+      "step": 8402
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.075694457736074,
+      "learning_rate": 2.738153482957452e-06,
+      "loss": 0.4646,
+      "step": 8403
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8484991813968688,
+      "learning_rate": 2.7370193286256495e-06,
+      "loss": 0.4454,
+      "step": 8404
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.293613668368275,
+      "learning_rate": 2.7358853207057394e-06,
+      "loss": 0.4534,
+      "step": 8405
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.782463121768709,
+      "learning_rate": 2.7347514592710855e-06,
+      "loss": 0.4712,
+      "step": 8406
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5203310909192354,
+      "learning_rate": 2.733617744395049e-06,
+      "loss": 0.4353,
+      "step": 8407
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6304307686807669,
+      "learning_rate": 2.7324841761509793e-06,
+      "loss": 0.4466,
+      "step": 8408
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.084056826536372,
+      "learning_rate": 2.7313507546122177e-06,
+      "loss": 0.4692,
+      "step": 8409
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5836718673910857,
+      "learning_rate": 2.7302174798520958e-06,
+      "loss": 0.466,
+      "step": 8410
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5349119451230968,
+      "learning_rate": 2.7290843519439326e-06,
+      "loss": 0.4278,
+      "step": 8411
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.916947879716906,
+      "learning_rate": 2.72795137096104e-06,
+      "loss": 0.4467,
+      "step": 8412
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.2891707112800375,
+      "learning_rate": 2.726818536976722e-06,
+      "loss": 0.5197,
+      "step": 8413
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.438233219845548,
+      "learning_rate": 2.725685850064273e-06,
+      "loss": 0.4618,
+      "step": 8414
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7912730182133705,
+      "learning_rate": 2.7245533102969723e-06,
+      "loss": 0.4339,
+      "step": 8415
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7978089161464292,
+      "learning_rate": 2.7234209177480957e-06,
+      "loss": 0.4439,
+      "step": 8416
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8309355448519167,
+      "learning_rate": 2.7222886724909083e-06,
+      "loss": 0.4232,
+      "step": 8417
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4381159065529576,
+      "learning_rate": 2.7211565745986635e-06,
+      "loss": 0.4409,
+      "step": 8418
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4128080723197063,
+      "learning_rate": 2.7200246241446086e-06,
+      "loss": 0.442,
+      "step": 8419
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.298622242161887,
+      "learning_rate": 2.718892821201975e-06,
+      "loss": 0.4353,
+      "step": 8420
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.216261716874472,
+      "learning_rate": 2.717761165843995e-06,
+      "loss": 0.4954,
+      "step": 8421
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.674568964980296,
+      "learning_rate": 2.7166296581438806e-06,
+      "loss": 0.4312,
+      "step": 8422
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.825403042246879,
+      "learning_rate": 2.71549829817484e-06,
+      "loss": 0.4995,
+      "step": 8423
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8666790991962592,
+      "learning_rate": 2.7143670860100713e-06,
+      "loss": 0.4523,
+      "step": 8424
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1617083482127657,
+      "learning_rate": 2.7132360217227637e-06,
+      "loss": 0.4635,
+      "step": 8425
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 18.667210762189818,
+      "learning_rate": 2.7121051053860925e-06,
+      "loss": 0.4483,
+      "step": 8426
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.06093772514321,
+      "learning_rate": 2.7109743370732276e-06,
+      "loss": 0.4093,
+      "step": 8427
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3832167579246195,
+      "learning_rate": 2.7098437168573287e-06,
+      "loss": 0.4068,
+      "step": 8428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8066684483924873,
+      "learning_rate": 2.7087132448115463e-06,
+      "loss": 0.4502,
+      "step": 8429
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5061844776910662,
+      "learning_rate": 2.707582921009021e-06,
+      "loss": 0.4126,
+      "step": 8430
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5631107837904047,
+      "learning_rate": 2.7064527455228796e-06,
+      "loss": 0.4904,
+      "step": 8431
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6146969850143555,
+      "learning_rate": 2.7053227184262454e-06,
+      "loss": 0.4155,
+      "step": 8432
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6989812615926074,
+      "learning_rate": 2.704192839792229e-06,
+      "loss": 0.4233,
+      "step": 8433
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5768343822537383,
+      "learning_rate": 2.7030631096939355e-06,
+      "loss": 0.4276,
+      "step": 8434
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5914072490165013,
+      "learning_rate": 2.7019335282044513e-06,
+      "loss": 0.4897,
+      "step": 8435
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.9495717237674735,
+      "learning_rate": 2.700804095396862e-06,
+      "loss": 0.4889,
+      "step": 8436
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.159008380093327,
+      "learning_rate": 2.6996748113442397e-06,
+      "loss": 0.399,
+      "step": 8437
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4205268812198915,
+      "learning_rate": 2.6985456761196504e-06,
+      "loss": 0.4456,
+      "step": 8438
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5707560431560048,
+      "learning_rate": 2.6974166897961407e-06,
+      "loss": 0.4781,
+      "step": 8439
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5525412916814305,
+      "learning_rate": 2.696287852446761e-06,
+      "loss": 0.4906,
+      "step": 8440
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5591902389242657,
+      "learning_rate": 2.6951591641445463e-06,
+      "loss": 0.4899,
+      "step": 8441
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6514237805596739,
+      "learning_rate": 2.694030624962516e-06,
+      "loss": 0.4483,
+      "step": 8442
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5570701168885115,
+      "learning_rate": 2.6929022349736876e-06,
+      "loss": 0.4705,
+      "step": 8443
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7112678720265293,
+      "learning_rate": 2.6917739942510666e-06,
+      "loss": 0.4201,
+      "step": 8444
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 3.5183581164182973,
+      "learning_rate": 2.69064590286765e-06,
+      "loss": 0.4491,
+      "step": 8445
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.977618137465116,
+      "learning_rate": 2.6895179608964215e-06,
+      "loss": 0.4835,
+      "step": 8446
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.624648908913114,
+      "learning_rate": 2.6883901684103585e-06,
+      "loss": 0.4458,
+      "step": 8447
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7711426861807622,
+      "learning_rate": 2.6872625254824268e-06,
+      "loss": 0.4663,
+      "step": 8448
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3134556046696069,
+      "learning_rate": 2.686135032185587e-06,
+      "loss": 0.4006,
+      "step": 8449
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.340126581507843,
+      "learning_rate": 2.6850076885927812e-06,
+      "loss": 0.4624,
+      "step": 8450
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5775121088150093,
+      "learning_rate": 2.683880494776948e-06,
+      "loss": 0.4938,
+      "step": 8451
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3875635327016926,
+      "learning_rate": 2.6827534508110204e-06,
+      "loss": 0.4328,
+      "step": 8452
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8446132874692402,
+      "learning_rate": 2.6816265567679112e-06,
+      "loss": 0.476,
+      "step": 8453
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.967955552543585,
+      "learning_rate": 2.680499812720534e-06,
+      "loss": 0.5122,
+      "step": 8454
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5078882603539396,
+      "learning_rate": 2.67937321874178e-06,
+      "loss": 0.439,
+      "step": 8455
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8914479155023898,
+      "learning_rate": 2.678246774904546e-06,
+      "loss": 0.4425,
+      "step": 8456
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5940810631016251,
+      "learning_rate": 2.6771204812817077e-06,
+      "loss": 0.4842,
+      "step": 8457
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.534488177194589,
+      "learning_rate": 2.675994337946135e-06,
+      "loss": 0.4495,
+      "step": 8458
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.0514228141507864,
+      "learning_rate": 2.6748683449706887e-06,
+      "loss": 0.4997,
+      "step": 8459
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.3883575395705976,
+      "learning_rate": 2.6737425024282206e-06,
+      "loss": 0.3829,
+      "step": 8460
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.12968181554706,
+      "learning_rate": 2.6726168103915675e-06,
+      "loss": 0.4785,
+      "step": 8461
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4117946499171357,
+      "learning_rate": 2.6714912689335627e-06,
+      "loss": 0.4052,
+      "step": 8462
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8756036163836711,
+      "learning_rate": 2.6703658781270257e-06,
+      "loss": 0.4659,
+      "step": 8463
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.468828749774537,
+      "learning_rate": 2.66924063804477e-06,
+      "loss": 0.4377,
+      "step": 8464
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.309431145182377,
+      "learning_rate": 2.6681155487595967e-06,
+      "loss": 0.4578,
+      "step": 8465
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.7996191243156232,
+      "learning_rate": 2.6669906103442957e-06,
+      "loss": 0.4708,
+      "step": 8466
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.8765046758926833,
+      "learning_rate": 2.6658658228716503e-06,
+      "loss": 0.4406,
+      "step": 8467
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1124950900533443,
+      "learning_rate": 2.6647411864144322e-06,
+      "loss": 0.4353,
+      "step": 8468
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6868116547868814,
+      "learning_rate": 2.6636167010454074e-06,
+      "loss": 0.4134,
+      "step": 8469
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4886864699169808,
+      "learning_rate": 2.6624923668373225e-06,
+      "loss": 0.4429,
+      "step": 8470
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1118298259732167,
+      "learning_rate": 2.6613681838629245e-06,
+      "loss": 0.4653,
+      "step": 8471
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7302784910478501,
+      "learning_rate": 2.6602441521949455e-06,
+      "loss": 0.4425,
+      "step": 8472
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5630853160528666,
+      "learning_rate": 2.6591202719061093e-06,
+      "loss": 0.491,
+      "step": 8473
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1683949637166067,
+      "learning_rate": 2.657996543069129e-06,
+      "loss": 0.4348,
+      "step": 8474
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5872247644502637,
+      "learning_rate": 2.6568729657567083e-06,
+      "loss": 0.443,
+      "step": 8475
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.612841290188638,
+      "learning_rate": 2.6557495400415435e-06,
+      "loss": 0.4887,
+      "step": 8476
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.516457498165126,
+      "learning_rate": 2.654626265996315e-06,
+      "loss": 0.4446,
+      "step": 8477
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1979977524622476,
+      "learning_rate": 2.6535031436937e-06,
+      "loss": 0.3991,
+      "step": 8478
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5278825741034252,
+      "learning_rate": 2.6523801732063603e-06,
+      "loss": 0.4107,
+      "step": 8479
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5707282340576225,
+      "learning_rate": 2.6512573546069555e-06,
+      "loss": 0.4949,
+      "step": 8480
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8905851393611177,
+      "learning_rate": 2.6501346879681243e-06,
+      "loss": 0.4511,
+      "step": 8481
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5683133185116268,
+      "learning_rate": 2.6490121733625055e-06,
+      "loss": 0.5019,
+      "step": 8482
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6757818898066488,
+      "learning_rate": 2.647889810862724e-06,
+      "loss": 0.447,
+      "step": 8483
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5315035226698239,
+      "learning_rate": 2.6467676005413955e-06,
+      "loss": 0.4765,
+      "step": 8484
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.624385860488442,
+      "learning_rate": 2.6456455424711235e-06,
+      "loss": 0.4318,
+      "step": 8485
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5090856842110851,
+      "learning_rate": 2.6445236367245037e-06,
+      "loss": 0.4684,
+      "step": 8486
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8167839396921261,
+      "learning_rate": 2.6434018833741265e-06,
+      "loss": 0.4296,
+      "step": 8487
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5642067530405176,
+      "learning_rate": 2.642280282492563e-06,
+      "loss": 0.4662,
+      "step": 8488
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0329295704740686,
+      "learning_rate": 2.6411588341523827e-06,
+      "loss": 0.429,
+      "step": 8489
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8371561954660316,
+      "learning_rate": 2.6400375384261356e-06,
+      "loss": 0.4821,
+      "step": 8490
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.564146389919857,
+      "learning_rate": 2.638916395386377e-06,
+      "loss": 0.4751,
+      "step": 8491
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5552921235632774,
+      "learning_rate": 2.6377954051056374e-06,
+      "loss": 0.4775,
+      "step": 8492
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7284281295429405,
+      "learning_rate": 2.6366745676564445e-06,
+      "loss": 0.4726,
+      "step": 8493
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8370945594716481,
+      "learning_rate": 2.6355538831113153e-06,
+      "loss": 0.4031,
+      "step": 8494
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.040817039820376,
+      "learning_rate": 2.634433351542759e-06,
+      "loss": 0.4739,
+      "step": 8495
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5078401204528307,
+      "learning_rate": 2.633312973023268e-06,
+      "loss": 0.453,
+      "step": 8496
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.293889457053298,
+      "learning_rate": 2.632192747625332e-06,
+      "loss": 0.4403,
+      "step": 8497
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6745807258430505,
+      "learning_rate": 2.631072675421428e-06,
+      "loss": 0.471,
+      "step": 8498
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0634735259719865,
+      "learning_rate": 2.629952756484022e-06,
+      "loss": 0.4631,
+      "step": 8499
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8847559329080108,
+      "learning_rate": 2.628832990885575e-06,
+      "loss": 0.4264,
+      "step": 8500
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7825098481398827,
+      "learning_rate": 2.6277133786985297e-06,
+      "loss": 0.4513,
+      "step": 8501
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.567724441603882,
+      "learning_rate": 2.626593919995325e-06,
+      "loss": 0.4635,
+      "step": 8502
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6950739132535348,
+      "learning_rate": 2.6254746148483888e-06,
+      "loss": 0.4178,
+      "step": 8503
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.507370779644991,
+      "learning_rate": 2.6243554633301414e-06,
+      "loss": 0.4034,
+      "step": 8504
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9862938573743107,
+      "learning_rate": 2.623236465512985e-06,
+      "loss": 0.4807,
+      "step": 8505
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5831015312586768,
+      "learning_rate": 2.622117621469321e-06,
+      "loss": 0.4774,
+      "step": 8506
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4719524421445083,
+      "learning_rate": 2.6209989312715355e-06,
+      "loss": 0.3982,
+      "step": 8507
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6524371419219617,
+      "learning_rate": 2.6198803949920072e-06,
+      "loss": 0.4424,
+      "step": 8508
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5389395096091447,
+      "learning_rate": 2.618762012703104e-06,
+      "loss": 0.4764,
+      "step": 8509
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5818063829934148,
+      "learning_rate": 2.617643784477183e-06,
+      "loss": 0.4682,
+      "step": 8510
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8773826057078797,
+      "learning_rate": 2.616525710386595e-06,
+      "loss": 0.4931,
+      "step": 8511
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.6377089041080937,
+      "learning_rate": 2.6154077905036734e-06,
+      "loss": 0.4367,
+      "step": 8512
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4027818187605505,
+      "learning_rate": 2.6142900249007487e-06,
+      "loss": 0.4937,
+      "step": 8513
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.1403029345983993,
+      "learning_rate": 2.613172413650138e-06,
+      "loss": 0.4381,
+      "step": 8514
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5174338589195369,
+      "learning_rate": 2.612054956824152e-06,
+      "loss": 0.4698,
+      "step": 8515
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0666287030710415,
+      "learning_rate": 2.6109376544950847e-06,
+      "loss": 0.4822,
+      "step": 8516
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4342382356074892,
+      "learning_rate": 2.6098205067352255e-06,
+      "loss": 0.45,
+      "step": 8517
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5824856193519303,
+      "learning_rate": 2.6087035136168526e-06,
+      "loss": 0.4886,
+      "step": 8518
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1622282307878837,
+      "learning_rate": 2.6075866752122346e-06,
+      "loss": 0.4874,
+      "step": 8519
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3500344806688513,
+      "learning_rate": 2.606469991593631e-06,
+      "loss": 0.4618,
+      "step": 8520
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5917134376312063,
+      "learning_rate": 2.605353462833284e-06,
+      "loss": 0.4721,
+      "step": 8521
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.693608857713611,
+      "learning_rate": 2.6042370890034387e-06,
+      "loss": 0.5131,
+      "step": 8522
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6137690527316064,
+      "learning_rate": 2.603120870176318e-06,
+      "loss": 0.5005,
+      "step": 8523
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.1938158946481114,
+      "learning_rate": 2.602004806424144e-06,
+      "loss": 0.4397,
+      "step": 8524
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.152766065455655,
+      "learning_rate": 2.600888897819117e-06,
+      "loss": 0.4384,
+      "step": 8525
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4056162054945518,
+      "learning_rate": 2.599773144433445e-06,
+      "loss": 0.4124,
+      "step": 8526
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5641305285671968,
+      "learning_rate": 2.5986575463393084e-06,
+      "loss": 0.4952,
+      "step": 8527
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7002122904668833,
+      "learning_rate": 2.5975421036088867e-06,
+      "loss": 0.4981,
+      "step": 8528
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.69507720004764,
+      "learning_rate": 2.596426816314348e-06,
+      "loss": 0.4376,
+      "step": 8529
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.576468601082067,
+      "learning_rate": 2.59531168452785e-06,
+      "loss": 0.4381,
+      "step": 8530
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8429147412969076,
+      "learning_rate": 2.5941967083215434e-06,
+      "loss": 0.4373,
+      "step": 8531
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.965153882742663,
+      "learning_rate": 2.5930818877675595e-06,
+      "loss": 0.4456,
+      "step": 8532
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7499842478578287,
+      "learning_rate": 2.5919672229380287e-06,
+      "loss": 0.4288,
+      "step": 8533
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4858902344662497,
+      "learning_rate": 2.59085271390507e-06,
+      "loss": 0.4326,
+      "step": 8534
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.699835041205681,
+      "learning_rate": 2.5897383607407894e-06,
+      "loss": 0.4256,
+      "step": 8535
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.891612624963404,
+      "learning_rate": 2.5886241635172828e-06,
+      "loss": 0.4765,
+      "step": 8536
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4342743353297844,
+      "learning_rate": 2.587510122306638e-06,
+      "loss": 0.5153,
+      "step": 8537
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5554436417728643,
+      "learning_rate": 2.5863962371809327e-06,
+      "loss": 0.4616,
+      "step": 8538
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.3285952490353705,
+      "learning_rate": 2.5852825082122344e-06,
+      "loss": 0.5175,
+      "step": 8539
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3601572878960517,
+      "learning_rate": 2.584168935472598e-06,
+      "loss": 0.4244,
+      "step": 8540
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7248061866880613,
+      "learning_rate": 2.583055519034069e-06,
+      "loss": 0.4202,
+      "step": 8541
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5417979528186434,
+      "learning_rate": 2.5819422589686897e-06,
+      "loss": 0.4044,
+      "step": 8542
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.540173216546552,
+      "learning_rate": 2.5808291553484815e-06,
+      "loss": 0.496,
+      "step": 8543
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5293482959917508,
+      "learning_rate": 2.5797162082454625e-06,
+      "loss": 0.458,
+      "step": 8544
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.450903397994852,
+      "learning_rate": 2.5786034177316384e-06,
+      "loss": 0.47,
+      "step": 8545
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8479480616204647,
+      "learning_rate": 2.5774907838790076e-06,
+      "loss": 0.403,
+      "step": 8546
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6077915845405977,
+      "learning_rate": 2.5763783067595517e-06,
+      "loss": 0.4278,
+      "step": 8547
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.837164837608964,
+      "learning_rate": 2.575265986445249e-06,
+      "loss": 0.4479,
+      "step": 8548
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9004693878098304,
+      "learning_rate": 2.574153823008065e-06,
+      "loss": 0.4329,
+      "step": 8549
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6395205179966024,
+      "learning_rate": 2.5730418165199578e-06,
+      "loss": 0.5045,
+      "step": 8550
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2341567082617284,
+      "learning_rate": 2.5719299670528674e-06,
+      "loss": 0.4771,
+      "step": 8551
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.986118391247093,
+      "learning_rate": 2.5708182746787326e-06,
+      "loss": 0.4866,
+      "step": 8552
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.421030109128693,
+      "learning_rate": 2.5697067394694775e-06,
+      "loss": 0.453,
+      "step": 8553
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2025967267212385,
+      "learning_rate": 2.568595361497017e-06,
+      "loss": 0.4518,
+      "step": 8554
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.110569780574495,
+      "learning_rate": 2.5674841408332573e-06,
+      "loss": 0.4297,
+      "step": 8555
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5789818236593,
+      "learning_rate": 2.5663730775500884e-06,
+      "loss": 0.4786,
+      "step": 8556
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.850104729124118,
+      "learning_rate": 2.5652621717194015e-06,
+      "loss": 0.4091,
+      "step": 8557
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.731488864548277,
+      "learning_rate": 2.564151423413065e-06,
+      "loss": 0.4604,
+      "step": 8558
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.49003029567321,
+      "learning_rate": 2.5630408327029473e-06,
+      "loss": 0.3777,
+      "step": 8559
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0707340001976977,
+      "learning_rate": 2.5619303996608956e-06,
+      "loss": 0.4669,
+      "step": 8560
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3563625693508323,
+      "learning_rate": 2.5608201243587615e-06,
+      "loss": 0.4073,
+      "step": 8561
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5739350422634835,
+      "learning_rate": 2.559710006868372e-06,
+      "loss": 0.475,
+      "step": 8562
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6539876009477463,
+      "learning_rate": 2.558600047261556e-06,
+      "loss": 0.44,
+      "step": 8563
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.9642129818604936,
+      "learning_rate": 2.557490245610118e-06,
+      "loss": 0.5015,
+      "step": 8564
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5945089426742003,
+      "learning_rate": 2.5563806019858685e-06,
+      "loss": 0.4827,
+      "step": 8565
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.4074496052760956,
+      "learning_rate": 2.5552711164605993e-06,
+      "loss": 0.4507,
+      "step": 8566
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5297986086049986,
+      "learning_rate": 2.554161789106089e-06,
+      "loss": 0.4123,
+      "step": 8567
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9130238565323032,
+      "learning_rate": 2.553052619994111e-06,
+      "loss": 0.4755,
+      "step": 8568
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.9626426660853369,
+      "learning_rate": 2.551943609196428e-06,
+      "loss": 0.4997,
+      "step": 8569
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8235505115878488,
+      "learning_rate": 2.5508347567847918e-06,
+      "loss": 0.4507,
+      "step": 8570
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5009336228453374,
+      "learning_rate": 2.5497260628309417e-06,
+      "loss": 0.4378,
+      "step": 8571
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.299761076880709,
+      "learning_rate": 2.5486175274066103e-06,
+      "loss": 0.474,
+      "step": 8572
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.0266383355768753,
+      "learning_rate": 2.547509150583518e-06,
+      "loss": 0.4721,
+      "step": 8573
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5312536057926136,
+      "learning_rate": 2.546400932433377e-06,
+      "loss": 0.4383,
+      "step": 8574
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.8430456698327,
+      "learning_rate": 2.5452928730278838e-06,
+      "loss": 0.4483,
+      "step": 8575
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 4.088097365766597,
+      "learning_rate": 2.544184972438729e-06,
+      "loss": 0.4242,
+      "step": 8576
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7534001890111448,
+      "learning_rate": 2.543077230737598e-06,
+      "loss": 0.4751,
+      "step": 8577
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.376587771695501,
+      "learning_rate": 2.541969647996154e-06,
+      "loss": 0.4518,
+      "step": 8578
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7342260331926937,
+      "learning_rate": 2.540862224286058e-06,
+      "loss": 0.4812,
+      "step": 8579
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5734388931233861,
+      "learning_rate": 2.5397549596789596e-06,
+      "loss": 0.4952,
+      "step": 8580
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4332261127641648,
+      "learning_rate": 2.5386478542464983e-06,
+      "loss": 0.3881,
+      "step": 8581
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.2727493909285594,
+      "learning_rate": 2.537540908060299e-06,
+      "loss": 0.3958,
+      "step": 8582
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4639662977067223,
+      "learning_rate": 2.5364341211919817e-06,
+      "loss": 0.4752,
+      "step": 8583
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5761633576145906,
+      "learning_rate": 2.5353274937131545e-06,
+      "loss": 0.4901,
+      "step": 8584
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6340536612938814,
+      "learning_rate": 2.534221025695416e-06,
+      "loss": 0.4408,
+      "step": 8585
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5646026971952806,
+      "learning_rate": 2.53311471721035e-06,
+      "loss": 0.4779,
+      "step": 8586
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 3.5443148151215245,
+      "learning_rate": 2.5320085683295337e-06,
+      "loss": 0.4361,
+      "step": 8587
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3793015911878326,
+      "learning_rate": 2.530902579124535e-06,
+      "loss": 0.4311,
+      "step": 8588
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6410426161453069,
+      "learning_rate": 2.5297967496669096e-06,
+      "loss": 0.4645,
+      "step": 8589
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5528993356735784,
+      "learning_rate": 2.5286910800282048e-06,
+      "loss": 0.4958,
+      "step": 8590
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 2.475619101163341,
+      "learning_rate": 2.52758557027995e-06,
+      "loss": 0.4419,
+      "step": 8591
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.6119102276469228,
+      "learning_rate": 2.526480220493679e-06,
+      "loss": 0.4869,
+      "step": 8592
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.7778331900365023,
+      "learning_rate": 2.5253750307408996e-06,
+      "loss": 0.4321,
+      "step": 8593
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5411127459319414,
+      "learning_rate": 2.5242700010931207e-06,
+      "loss": 0.473,
+      "step": 8594
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6033165290274263,
+      "learning_rate": 2.52316513162183e-06,
+      "loss": 0.495,
+      "step": 8595
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5167209242897366,
+      "learning_rate": 2.522060422398519e-06,
+      "loss": 0.4819,
+      "step": 8596
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.009066943752502,
+      "learning_rate": 2.520955873494656e-06,
+      "loss": 0.4725,
+      "step": 8597
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5264861073931983,
+      "learning_rate": 2.519851484981704e-06,
+      "loss": 0.4712,
+      "step": 8598
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5077188082281688,
+      "learning_rate": 2.5187472569311167e-06,
+      "loss": 0.4536,
+      "step": 8599
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.019115630620905,
+      "learning_rate": 2.517643189414336e-06,
+      "loss": 0.4781,
+      "step": 8600
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6532095102760669,
+      "learning_rate": 2.5165392825027957e-06,
+      "loss": 0.4258,
+      "step": 8601
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8547903271265982,
+      "learning_rate": 2.515435536267913e-06,
+      "loss": 0.4306,
+      "step": 8602
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2216175528845805,
+      "learning_rate": 2.514331950781101e-06,
+      "loss": 0.4566,
+      "step": 8603
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9686083682659197,
+      "learning_rate": 2.51322852611376e-06,
+      "loss": 0.4442,
+      "step": 8604
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5932490974825848,
+      "learning_rate": 2.5121252623372826e-06,
+      "loss": 0.4123,
+      "step": 8605
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5724265732507499,
+      "learning_rate": 2.5110221595230435e-06,
+      "loss": 0.4122,
+      "step": 8606
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5805141920196887,
+      "learning_rate": 2.5099192177424155e-06,
+      "loss": 0.4808,
+      "step": 8607
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4585497866315096,
+      "learning_rate": 2.5088164370667558e-06,
+      "loss": 0.4799,
+      "step": 8608
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.169329087083359,
+      "learning_rate": 2.5077138175674144e-06,
+      "loss": 0.4654,
+      "step": 8609
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.40047169601641,
+      "learning_rate": 2.5066113593157303e-06,
+      "loss": 0.4812,
+      "step": 8610
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8158049722535092,
+      "learning_rate": 2.505509062383026e-06,
+      "loss": 0.4584,
+      "step": 8611
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.1137347449491664,
+      "learning_rate": 2.5044069268406256e-06,
+      "loss": 0.47,
+      "step": 8612
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7668796729463565,
+      "learning_rate": 2.503304952759831e-06,
+      "loss": 0.5049,
+      "step": 8613
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.493389780900634,
+      "learning_rate": 2.502203140211939e-06,
+      "loss": 0.4732,
+      "step": 8614
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7745690779243597,
+      "learning_rate": 2.5011014892682366e-06,
+      "loss": 0.4365,
+      "step": 8615
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.688350050063718,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.4749,
+      "step": 8616
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.5793764822172753,
+      "learning_rate": 2.498898672478493e-06,
+      "loss": 0.4334,
+      "step": 8617
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1798522161664264,
+      "learning_rate": 2.497797506774969e-06,
+      "loss": 0.4313,
+      "step": 8618
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5728479128864852,
+      "learning_rate": 2.4966965029606734e-06,
+      "loss": 0.4528,
+      "step": 8619
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1434383004298003,
+      "learning_rate": 2.495595661106839e-06,
+      "loss": 0.4011,
+      "step": 8620
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5424611765008539,
+      "learning_rate": 2.4944949812846915e-06,
+      "loss": 0.4689,
+      "step": 8621
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4775355701343522,
+      "learning_rate": 2.493394463565438e-06,
+      "loss": 0.4304,
+      "step": 8622
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6978010903989313,
+      "learning_rate": 2.4922941080202852e-06,
+      "loss": 0.4458,
+      "step": 8623
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4792508890475076,
+      "learning_rate": 2.4911939147204224e-06,
+      "loss": 0.4444,
+      "step": 8624
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5817313643310613,
+      "learning_rate": 2.490093883737033e-06,
+      "loss": 0.4179,
+      "step": 8625
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6923549140820773,
+      "learning_rate": 2.4889940151412824e-06,
+      "loss": 0.4164,
+      "step": 8626
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6833100477919725,
+      "learning_rate": 2.487894309004338e-06,
+      "loss": 0.3896,
+      "step": 8627
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9406500440740726,
+      "learning_rate": 2.4867947653973436e-06,
+      "loss": 0.4357,
+      "step": 8628
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4813109871955763,
+      "learning_rate": 2.4856953843914423e-06,
+      "loss": 0.5006,
+      "step": 8629
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.4881134140075156,
+      "learning_rate": 2.484596166057757e-06,
+      "loss": 0.4452,
+      "step": 8630
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9944873329607868,
+      "learning_rate": 2.483497110467412e-06,
+      "loss": 0.4985,
+      "step": 8631
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.022521918791755,
+      "learning_rate": 2.482398217691511e-06,
+      "loss": 0.4422,
+      "step": 8632
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6589273072944433,
+      "learning_rate": 2.481299487801151e-06,
+      "loss": 0.4147,
+      "step": 8633
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.6240499914796653,
+      "learning_rate": 2.48020092086742e-06,
+      "loss": 0.3867,
+      "step": 8634
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3407068340163475,
+      "learning_rate": 2.4791025169613925e-06,
+      "loss": 0.431,
+      "step": 8635
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4916918372388985,
+      "learning_rate": 2.478004276154137e-06,
+      "loss": 0.4317,
+      "step": 8636
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6608103333475932,
+      "learning_rate": 2.476906198516703e-06,
+      "loss": 0.4204,
+      "step": 8637
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5124608373167754,
+      "learning_rate": 2.475808284120138e-06,
+      "loss": 0.5029,
+      "step": 8638
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5698826708287567,
+      "learning_rate": 2.4747105330354748e-06,
+      "loss": 0.5198,
+      "step": 8639
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4527197190331629,
+      "learning_rate": 2.4736129453337384e-06,
+      "loss": 0.4326,
+      "step": 8640
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9973030042093867,
+      "learning_rate": 2.4725155210859373e-06,
+      "loss": 0.4569,
+      "step": 8641
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.7788717862448635,
+      "learning_rate": 2.4714182603630764e-06,
+      "loss": 0.4451,
+      "step": 8642
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5802608512960286,
+      "learning_rate": 2.4703211632361457e-06,
+      "loss": 0.4786,
+      "step": 8643
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6258557113890921,
+      "learning_rate": 2.4692242297761267e-06,
+      "loss": 0.4536,
+      "step": 8644
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6260850607556498,
+      "learning_rate": 2.4681274600539914e-06,
+      "loss": 0.4598,
+      "step": 8645
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7777149196640947,
+      "learning_rate": 2.4670308541406928e-06,
+      "loss": 0.4975,
+      "step": 8646
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3724422337579663,
+      "learning_rate": 2.465934412107189e-06,
+      "loss": 0.4833,
+      "step": 8647
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.200358260515967,
+      "learning_rate": 2.4648381340244116e-06,
+      "loss": 0.4854,
+      "step": 8648
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.669084606731499,
+      "learning_rate": 2.4637420199632906e-06,
+      "loss": 0.4839,
+      "step": 8649
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6659316772380972,
+      "learning_rate": 2.462646069994743e-06,
+      "loss": 0.4569,
+      "step": 8650
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5568474077030963,
+      "learning_rate": 2.4615502841896773e-06,
+      "loss": 0.4574,
+      "step": 8651
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3557823585281814,
+      "learning_rate": 2.460454662618985e-06,
+      "loss": 0.4195,
+      "step": 8652
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4003361552797249,
+      "learning_rate": 2.4593592053535548e-06,
+      "loss": 0.4713,
+      "step": 8653
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7484157385085242,
+      "learning_rate": 2.45826391246426e-06,
+      "loss": 0.4407,
+      "step": 8654
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.131802398963274,
+      "learning_rate": 2.457168784021965e-06,
+      "loss": 0.4606,
+      "step": 8655
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5690043519831829,
+      "learning_rate": 2.4560738200975252e-06,
+      "loss": 0.4774,
+      "step": 8656
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5434630962379705,
+      "learning_rate": 2.4549790207617794e-06,
+      "loss": 0.4787,
+      "step": 8657
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.564726948929696,
+      "learning_rate": 2.453884386085562e-06,
+      "loss": 0.4571,
+      "step": 8658
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9266085209091557,
+      "learning_rate": 2.452789916139694e-06,
+      "loss": 0.4381,
+      "step": 8659
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8626364164430838,
+      "learning_rate": 2.451695610994988e-06,
+      "loss": 0.4672,
+      "step": 8660
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.561103422228639,
+      "learning_rate": 2.4506014707222387e-06,
+      "loss": 0.4394,
+      "step": 8661
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.166281520429704,
+      "learning_rate": 2.4495074953922425e-06,
+      "loss": 0.4727,
+      "step": 8662
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4459813431808777,
+      "learning_rate": 2.448413685075774e-06,
+      "loss": 0.4384,
+      "step": 8663
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.747174031589353,
+      "learning_rate": 2.447320039843603e-06,
+      "loss": 0.3458,
+      "step": 8664
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4678498545576237,
+      "learning_rate": 2.446226559766483e-06,
+      "loss": 0.4436,
+      "step": 8665
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5563641565233948,
+      "learning_rate": 2.4451332449151653e-06,
+      "loss": 0.4762,
+      "step": 8666
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.638185921559047,
+      "learning_rate": 2.4440400953603864e-06,
+      "loss": 0.5084,
+      "step": 8667
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5263339387029313,
+      "learning_rate": 2.4429471111728683e-06,
+      "loss": 0.4799,
+      "step": 8668
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.854657421494241,
+      "learning_rate": 2.441854292423327e-06,
+      "loss": 0.4648,
+      "step": 8669
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5429165018187546,
+      "learning_rate": 2.4407616391824656e-06,
+      "loss": 0.4962,
+      "step": 8670
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5265043076079936,
+      "learning_rate": 2.4396691515209813e-06,
+      "loss": 0.4327,
+      "step": 8671
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5618279874013237,
+      "learning_rate": 2.438576829509551e-06,
+      "loss": 0.4737,
+      "step": 8672
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.1583303231816657,
+      "learning_rate": 2.437484673218849e-06,
+      "loss": 0.4506,
+      "step": 8673
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.586101024056693,
+      "learning_rate": 2.4363926827195367e-06,
+      "loss": 0.5102,
+      "step": 8674
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8869976814534242,
+      "learning_rate": 2.435300858082266e-06,
+      "loss": 0.4521,
+      "step": 8675
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.0961996825225055,
+      "learning_rate": 2.434209199377673e-06,
+      "loss": 0.4204,
+      "step": 8676
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6416289702852962,
+      "learning_rate": 2.433117706676386e-06,
+      "loss": 0.4125,
+      "step": 8677
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7160250390882927,
+      "learning_rate": 2.4320263800490302e-06,
+      "loss": 0.4856,
+      "step": 8678
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5549098930267665,
+      "learning_rate": 2.430935219566206e-06,
+      "loss": 0.5034,
+      "step": 8679
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5347285705408564,
+      "learning_rate": 2.4298442252985137e-06,
+      "loss": 0.4904,
+      "step": 8680
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.6905874609255225,
+      "learning_rate": 2.4287533973165343e-06,
+      "loss": 0.4594,
+      "step": 8681
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7937293985435099,
+      "learning_rate": 2.4276627356908503e-06,
+      "loss": 0.427,
+      "step": 8682
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9039307597113626,
+      "learning_rate": 2.4265722404920206e-06,
+      "loss": 0.4979,
+      "step": 8683
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6277690795983729,
+      "learning_rate": 2.4254819117906003e-06,
+      "loss": 0.4736,
+      "step": 8684
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7101031588502282,
+      "learning_rate": 2.4243917496571325e-06,
+      "loss": 0.4466,
+      "step": 8685
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9178211256721815,
+      "learning_rate": 2.4233017541621504e-06,
+      "loss": 0.4364,
+      "step": 8686
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.033452560934233,
+      "learning_rate": 2.4222119253761727e-06,
+      "loss": 0.5065,
+      "step": 8687
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4054579669995466,
+      "learning_rate": 2.4211222633697112e-06,
+      "loss": 0.4436,
+      "step": 8688
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5114033109457847,
+      "learning_rate": 2.4200327682132664e-06,
+      "loss": 0.4118,
+      "step": 8689
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5468618864721682,
+      "learning_rate": 2.418943439977325e-06,
+      "loss": 0.4454,
+      "step": 8690
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.450922365663008,
+      "learning_rate": 2.41785427873237e-06,
+      "loss": 0.4554,
+      "step": 8691
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.308682294931974,
+      "learning_rate": 2.4167652845488627e-06,
+      "loss": 0.4213,
+      "step": 8692
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.551988616539631,
+      "learning_rate": 2.4156764574972618e-06,
+      "loss": 0.4751,
+      "step": 8693
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.336032193924232,
+      "learning_rate": 2.4145877976480135e-06,
+      "loss": 0.4409,
+      "step": 8694
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0652399981037206,
+      "learning_rate": 2.413499305071555e-06,
+      "loss": 0.4625,
+      "step": 8695
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2609130787039096,
+      "learning_rate": 2.412410979838304e-06,
+      "loss": 0.3775,
+      "step": 8696
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3321685529579332,
+      "learning_rate": 2.4113228220186814e-06,
+      "loss": 0.4349,
+      "step": 8697
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8896809262748044,
+      "learning_rate": 2.410234831683084e-06,
+      "loss": 0.3922,
+      "step": 8698
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.9269334997883887,
+      "learning_rate": 2.4091470089019044e-06,
+      "loss": 0.3903,
+      "step": 8699
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.43302857964809,
+      "learning_rate": 2.4080593537455242e-06,
+      "loss": 0.468,
+      "step": 8700
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7937640036844844,
+      "learning_rate": 2.406971866284314e-06,
+      "loss": 0.4207,
+      "step": 8701
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.698001434002493,
+      "learning_rate": 2.4058845465886334e-06,
+      "loss": 0.4083,
+      "step": 8702
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7994236500960634,
+      "learning_rate": 2.404797394728827e-06,
+      "loss": 0.5186,
+      "step": 8703
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8225428990303667,
+      "learning_rate": 2.403710410775234e-06,
+      "loss": 0.4087,
+      "step": 8704
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8496258455235655,
+      "learning_rate": 2.4026235947981815e-06,
+      "loss": 0.439,
+      "step": 8705
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.910776108274756,
+      "learning_rate": 2.4015369468679866e-06,
+      "loss": 0.4466,
+      "step": 8706
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.6735818153322384,
+      "learning_rate": 2.4004504670549493e-06,
+      "loss": 0.4096,
+      "step": 8707
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.8797391118636217,
+      "learning_rate": 2.3993641554293673e-06,
+      "loss": 0.5002,
+      "step": 8708
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 4.978327153942626,
+      "learning_rate": 2.398278012061522e-06,
+      "loss": 0.4767,
+      "step": 8709
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2698675281431169,
+      "learning_rate": 2.3971920370216873e-06,
+      "loss": 0.4504,
+      "step": 8710
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2319502387286105,
+      "learning_rate": 2.3961062303801215e-06,
+      "loss": 0.4783,
+      "step": 8711
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.8360719569294925,
+      "learning_rate": 2.3950205922070742e-06,
+      "loss": 0.4039,
+      "step": 8712
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.0316487686099727,
+      "learning_rate": 2.39393512257279e-06,
+      "loss": 0.4442,
+      "step": 8713
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.801444738290949,
+      "learning_rate": 2.3928498215474933e-06,
+      "loss": 0.4964,
+      "step": 8714
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.73184748549202,
+      "learning_rate": 2.391764689201404e-06,
+      "loss": 0.4394,
+      "step": 8715
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.058526591733146,
+      "learning_rate": 2.390679725604724e-06,
+      "loss": 0.4617,
+      "step": 8716
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 9.03967655046963,
+      "learning_rate": 2.3895949308276558e-06,
+      "loss": 0.4795,
+      "step": 8717
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7912932690383438,
+      "learning_rate": 2.3885103049403783e-06,
+      "loss": 0.4372,
+      "step": 8718
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5605117839335046,
+      "learning_rate": 2.3874258480130684e-06,
+      "loss": 0.4656,
+      "step": 8719
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.556538726887634,
+      "learning_rate": 2.3863415601158884e-06,
+      "loss": 0.5036,
+      "step": 8720
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.2463965567200543,
+      "learning_rate": 2.385257441318992e-06,
+      "loss": 0.5037,
+      "step": 8721
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.813789513504855,
+      "learning_rate": 2.384173491692517e-06,
+      "loss": 0.4402,
+      "step": 8722
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.757445889349721,
+      "learning_rate": 2.3830897113065947e-06,
+      "loss": 0.4805,
+      "step": 8723
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5351100616852873,
+      "learning_rate": 2.3820061002313454e-06,
+      "loss": 0.4626,
+      "step": 8724
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5616114610000059,
+      "learning_rate": 2.380922658536876e-06,
+      "loss": 0.4854,
+      "step": 8725
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.431120166783938,
+      "learning_rate": 2.3798393862932873e-06,
+      "loss": 0.4711,
+      "step": 8726
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5083332342059331,
+      "learning_rate": 2.37875628357066e-06,
+      "loss": 0.4736,
+      "step": 8727
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6168500939661397,
+      "learning_rate": 2.377673350439072e-06,
+      "loss": 0.4297,
+      "step": 8728
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3929795471069135,
+      "learning_rate": 2.3765905869685883e-06,
+      "loss": 0.4428,
+      "step": 8729
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.553332261763214,
+      "learning_rate": 2.3755079932292635e-06,
+      "loss": 0.4802,
+      "step": 8730
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4712408387825002,
+      "learning_rate": 2.3744255692911345e-06,
+      "loss": 0.4004,
+      "step": 8731
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5426212009283794,
+      "learning_rate": 2.37334331522424e-06,
+      "loss": 0.4802,
+      "step": 8732
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5377603289534132,
+      "learning_rate": 2.3722612310985953e-06,
+      "loss": 0.4664,
+      "step": 8733
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7730832787599258,
+      "learning_rate": 2.3711793169842114e-06,
+      "loss": 0.4055,
+      "step": 8734
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6340121522007784,
+      "learning_rate": 2.3700975729510868e-06,
+      "loss": 0.4396,
+      "step": 8735
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5760830756576245,
+      "learning_rate": 2.3690159990692087e-06,
+      "loss": 0.4766,
+      "step": 8736
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8506566204935986,
+      "learning_rate": 2.367934595408556e-06,
+      "loss": 0.4954,
+      "step": 8737
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1439958788541276,
+      "learning_rate": 2.3668533620390895e-06,
+      "loss": 0.456,
+      "step": 8738
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8333177686380815,
+      "learning_rate": 2.3657722990307658e-06,
+      "loss": 0.4659,
+      "step": 8739
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.082158365443648,
+      "learning_rate": 2.3646914064535285e-06,
+      "loss": 0.4392,
+      "step": 8740
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.02661052503492,
+      "learning_rate": 2.363610684377312e-06,
+      "loss": 0.4633,
+      "step": 8741
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.160091996781083,
+      "learning_rate": 2.362530132872033e-06,
+      "loss": 0.4682,
+      "step": 8742
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5705629523476209,
+      "learning_rate": 2.3614497520076047e-06,
+      "loss": 0.4932,
+      "step": 8743
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.556365044463721,
+      "learning_rate": 2.360369541853925e-06,
+      "loss": 0.4571,
+      "step": 8744
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6269510513434418,
+      "learning_rate": 2.3592895024808837e-06,
+      "loss": 0.4977,
+      "step": 8745
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5624560286531658,
+      "learning_rate": 2.3582096339583587e-06,
+      "loss": 0.4796,
+      "step": 8746
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.10179051197027,
+      "learning_rate": 2.3571299363562104e-06,
+      "loss": 0.468,
+      "step": 8747
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2019275531017815,
+      "learning_rate": 2.356050409744302e-06,
+      "loss": 0.4014,
+      "step": 8748
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7148279912355804,
+      "learning_rate": 2.3549710541924714e-06,
+      "loss": 0.4015,
+      "step": 8749
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4593976164721274,
+      "learning_rate": 2.3538918697705553e-06,
+      "loss": 0.4381,
+      "step": 8750
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5434342728739264,
+      "learning_rate": 2.3528128565483698e-06,
+      "loss": 0.464,
+      "step": 8751
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.370781730100156,
+      "learning_rate": 2.351734014595734e-06,
+      "loss": 0.4182,
+      "step": 8752
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.917740908400957,
+      "learning_rate": 2.35065534398244e-06,
+      "loss": 0.4247,
+      "step": 8753
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.5026413348510474,
+      "learning_rate": 2.3495768447782808e-06,
+      "loss": 0.4439,
+      "step": 8754
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.764071283480364,
+      "learning_rate": 2.3484985170530315e-06,
+      "loss": 0.409,
+      "step": 8755
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.542035515467006,
+      "learning_rate": 2.34742036087646e-06,
+      "loss": 0.4857,
+      "step": 8756
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8717198331653573,
+      "learning_rate": 2.346342376318323e-06,
+      "loss": 0.4635,
+      "step": 8757
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0116756100216424,
+      "learning_rate": 2.3452645634483613e-06,
+      "loss": 0.4856,
+      "step": 8758
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4223974963036041,
+      "learning_rate": 2.344186922336309e-06,
+      "loss": 0.4801,
+      "step": 8759
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.368871237721846,
+      "learning_rate": 2.3431094530518888e-06,
+      "loss": 0.4299,
+      "step": 8760
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.520330667529179,
+      "learning_rate": 2.342032155664813e-06,
+      "loss": 0.4864,
+      "step": 8761
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5705362844642544,
+      "learning_rate": 2.340955030244778e-06,
+      "loss": 0.4903,
+      "step": 8762
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 7.216846974685011,
+      "learning_rate": 2.339878076861474e-06,
+      "loss": 0.4496,
+      "step": 8763
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.322440611483879,
+      "learning_rate": 2.3388012955845786e-06,
+      "loss": 0.3751,
+      "step": 8764
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4974417035979108,
+      "learning_rate": 2.33772468648376e-06,
+      "loss": 0.4752,
+      "step": 8765
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5098265348911137,
+      "learning_rate": 2.3366482496286672e-06,
+      "loss": 0.4416,
+      "step": 8766
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4855040288373205,
+      "learning_rate": 2.335571985088951e-06,
+      "loss": 0.4577,
+      "step": 8767
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5736189378000146,
+      "learning_rate": 2.3344958929342436e-06,
+      "loss": 0.4831,
+      "step": 8768
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7676328408650561,
+      "learning_rate": 2.3334199732341634e-06,
+      "loss": 0.4434,
+      "step": 8769
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.44241088476426,
+      "learning_rate": 2.3323442260583223e-06,
+      "loss": 0.4436,
+      "step": 8770
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.3992978635751228,
+      "learning_rate": 2.33126865147632e-06,
+      "loss": 0.4679,
+      "step": 8771
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.09087307772518,
+      "learning_rate": 2.3301932495577468e-06,
+      "loss": 0.4553,
+      "step": 8772
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.944289284012359,
+      "learning_rate": 2.3291180203721757e-06,
+      "loss": 0.4548,
+      "step": 8773
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.592144710149855,
+      "learning_rate": 2.328042963989175e-06,
+      "loss": 0.4152,
+      "step": 8774
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.289419589274149,
+      "learning_rate": 2.326968080478299e-06,
+      "loss": 0.447,
+      "step": 8775
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0442389268677466,
+      "learning_rate": 2.3258933699090934e-06,
+      "loss": 0.4645,
+      "step": 8776
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4767862079824787,
+      "learning_rate": 2.3248188323510866e-06,
+      "loss": 0.4949,
+      "step": 8777
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5575791928662546,
+      "learning_rate": 2.323744467873802e-06,
+      "loss": 0.4806,
+      "step": 8778
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.181235810894636,
+      "learning_rate": 2.3226702765467496e-06,
+      "loss": 0.5034,
+      "step": 8779
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2777678210077985,
+      "learning_rate": 2.3215962584394277e-06,
+      "loss": 0.5088,
+      "step": 8780
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 10.552099831855452,
+      "learning_rate": 2.3205224136213257e-06,
+      "loss": 0.444,
+      "step": 8781
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7754346767581746,
+      "learning_rate": 2.3194487421619143e-06,
+      "loss": 0.4058,
+      "step": 8782
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.344182909460469,
+      "learning_rate": 2.318375244130667e-06,
+      "loss": 0.5197,
+      "step": 8783
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3461524721266638,
+      "learning_rate": 2.3173019195970313e-06,
+      "loss": 0.4246,
+      "step": 8784
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.664310168159418,
+      "learning_rate": 2.316228768630454e-06,
+      "loss": 0.4196,
+      "step": 8785
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.003813943824683,
+      "learning_rate": 2.315155791300361e-06,
+      "loss": 0.4574,
+      "step": 8786
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5561281919265904,
+      "learning_rate": 2.3140829876761794e-06,
+      "loss": 0.4054,
+      "step": 8787
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8031096785030982,
+      "learning_rate": 2.3130103578273135e-06,
+      "loss": 0.5038,
+      "step": 8788
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8732727970377732,
+      "learning_rate": 2.311937901823162e-06,
+      "loss": 0.4708,
+      "step": 8789
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.369895655389187,
+      "learning_rate": 2.3108656197331118e-06,
+      "loss": 0.4523,
+      "step": 8790
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.2752789352993346,
+      "learning_rate": 2.309793511626538e-06,
+      "loss": 0.4474,
+      "step": 8791
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.608232228016393,
+      "learning_rate": 2.308721577572807e-06,
+      "loss": 0.4534,
+      "step": 8792
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5324486160938781,
+      "learning_rate": 2.307649817641268e-06,
+      "loss": 0.4597,
+      "step": 8793
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5424116957412364,
+      "learning_rate": 2.3065782319012625e-06,
+      "loss": 0.4805,
+      "step": 8794
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4370830468060798,
+      "learning_rate": 2.3055068204221226e-06,
+      "loss": 0.4478,
+      "step": 8795
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6971845653349664,
+      "learning_rate": 2.3044355832731685e-06,
+      "loss": 0.4706,
+      "step": 8796
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4142916033456723,
+      "learning_rate": 2.3033645205237043e-06,
+      "loss": 0.4383,
+      "step": 8797
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6465752809575275,
+      "learning_rate": 2.3022936322430275e-06,
+      "loss": 0.4822,
+      "step": 8798
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.1583248477051136,
+      "learning_rate": 2.3012229185004237e-06,
+      "loss": 0.4717,
+      "step": 8799
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5826932578766875,
+      "learning_rate": 2.3001523793651688e-06,
+      "loss": 0.5065,
+      "step": 8800
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5404024733051546,
+      "learning_rate": 2.299082014906519e-06,
+      "loss": 0.4674,
+      "step": 8801
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4957450252036628,
+      "learning_rate": 2.2980118251937315e-06,
+      "loss": 0.4216,
+      "step": 8802
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5752588296845698,
+      "learning_rate": 2.2969418102960467e-06,
+      "loss": 0.4983,
+      "step": 8803
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5223896494057233,
+      "learning_rate": 2.295871970282688e-06,
+      "loss": 0.4718,
+      "step": 8804
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9041600229989666,
+      "learning_rate": 2.294802305222876e-06,
+      "loss": 0.4312,
+      "step": 8805
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5169058452292408,
+      "learning_rate": 2.2937328151858156e-06,
+      "loss": 0.4399,
+      "step": 8806
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.493607344269368,
+      "learning_rate": 2.2926635002407046e-06,
+      "loss": 0.4618,
+      "step": 8807
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5539477899012804,
+      "learning_rate": 2.2915943604567203e-06,
+      "loss": 0.4676,
+      "step": 8808
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7997869972089129,
+      "learning_rate": 2.2905253959030386e-06,
+      "loss": 0.4128,
+      "step": 8809
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7885843396232923,
+      "learning_rate": 2.2894566066488194e-06,
+      "loss": 0.4351,
+      "step": 8810
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.51770918938979,
+      "learning_rate": 2.2883879927632136e-06,
+      "loss": 0.4564,
+      "step": 8811
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3128212868480458,
+      "learning_rate": 2.2873195543153555e-06,
+      "loss": 0.4388,
+      "step": 8812
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.149298372185613,
+      "learning_rate": 2.2862512913743734e-06,
+      "loss": 0.4808,
+      "step": 8813
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6532677180646076,
+      "learning_rate": 2.285183204009383e-06,
+      "loss": 0.3938,
+      "step": 8814
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6286394433991378,
+      "learning_rate": 2.284115292289488e-06,
+      "loss": 0.4213,
+      "step": 8815
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.89879466320086,
+      "learning_rate": 2.283047556283783e-06,
+      "loss": 0.3993,
+      "step": 8816
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.566456301592447,
+      "learning_rate": 2.281979996061342e-06,
+      "loss": 0.4567,
+      "step": 8817
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5120286926995476,
+      "learning_rate": 2.2809126116912443e-06,
+      "loss": 0.45,
+      "step": 8818
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5369697485171483,
+      "learning_rate": 2.2798454032425413e-06,
+      "loss": 0.4496,
+      "step": 8819
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0389247744033914,
+      "learning_rate": 2.278778370784284e-06,
+      "loss": 0.4151,
+      "step": 8820
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 8.838206905988423,
+      "learning_rate": 2.277711514385503e-06,
+      "loss": 0.4332,
+      "step": 8821
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4386275775480424,
+      "learning_rate": 2.2766448341152297e-06,
+      "loss": 0.4331,
+      "step": 8822
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.61286913567288,
+      "learning_rate": 2.275578330042471e-06,
+      "loss": 0.4725,
+      "step": 8823
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.8946472053284966,
+      "learning_rate": 2.2745120022362304e-06,
+      "loss": 0.4797,
+      "step": 8824
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6362844465428508,
+      "learning_rate": 2.2734458507654978e-06,
+      "loss": 0.4033,
+      "step": 8825
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6867924214973369,
+      "learning_rate": 2.2723798756992515e-06,
+      "loss": 0.4272,
+      "step": 8826
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9747343763163863,
+      "learning_rate": 2.2713140771064613e-06,
+      "loss": 0.4364,
+      "step": 8827
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.0082363570557376,
+      "learning_rate": 2.2702484550560787e-06,
+      "loss": 0.4699,
+      "step": 8828
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.10164129143919,
+      "learning_rate": 2.2691830096170503e-06,
+      "loss": 0.506,
+      "step": 8829
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.854927152118923,
+      "learning_rate": 2.268117740858309e-06,
+      "loss": 0.5502,
+      "step": 8830
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4937754557846141,
+      "learning_rate": 2.2670526488487774e-06,
+      "loss": 0.4466,
+      "step": 8831
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4937465814850117,
+      "learning_rate": 2.265987733657363e-06,
+      "loss": 0.4091,
+      "step": 8832
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4422124553944382,
+      "learning_rate": 2.2649229953529656e-06,
+      "loss": 0.4708,
+      "step": 8833
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3899277112183073,
+      "learning_rate": 2.263858434004472e-06,
+      "loss": 0.4503,
+      "step": 8834
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5844040864446882,
+      "learning_rate": 2.2627940496807588e-06,
+      "loss": 0.4265,
+      "step": 8835
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6941744003047006,
+      "learning_rate": 2.2617298424506918e-06,
+      "loss": 0.4782,
+      "step": 8836
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 5.717318726313983,
+      "learning_rate": 2.2606658123831175e-06,
+      "loss": 0.4421,
+      "step": 8837
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5847529787328378,
+      "learning_rate": 2.259601959546886e-06,
+      "loss": 0.4925,
+      "step": 8838
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5504788823467351,
+      "learning_rate": 2.2585382840108213e-06,
+      "loss": 0.4661,
+      "step": 8839
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.710090047101358,
+      "learning_rate": 2.2574747858437425e-06,
+      "loss": 0.4775,
+      "step": 8840
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7289786463566346,
+      "learning_rate": 2.256411465114458e-06,
+      "loss": 0.4498,
+      "step": 8841
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.107929901860206,
+      "learning_rate": 2.2553483218917644e-06,
+      "loss": 0.5099,
+      "step": 8842
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.9207998072396633,
+      "learning_rate": 2.254285356244442e-06,
+      "loss": 0.432,
+      "step": 8843
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.383567082357788,
+      "learning_rate": 2.2532225682412646e-06,
+      "loss": 0.4128,
+      "step": 8844
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4184029295374017,
+      "learning_rate": 2.2521599579509945e-06,
+      "loss": 0.449,
+      "step": 8845
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.9508461714939482,
+      "learning_rate": 2.25109752544238e-06,
+      "loss": 0.4423,
+      "step": 8846
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.008482865570436,
+      "learning_rate": 2.2500352707841622e-06,
+      "loss": 0.4669,
+      "step": 8847
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 2.776515179958293,
+      "learning_rate": 2.2489731940450625e-06,
+      "loss": 0.4838,
+      "step": 8848
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 3.317306725519793,
+      "learning_rate": 2.247911295293798e-06,
+      "loss": 0.4602,
+      "step": 8849
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5547771637607359,
+      "learning_rate": 2.2468495745990732e-06,
+      "loss": 0.4794,
+      "step": 8850
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.491712390905574,
+      "learning_rate": 2.245788032029581e-06,
+      "loss": 0.4536,
+      "step": 8851
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7838476396887042,
+      "learning_rate": 2.244726667653997e-06,
+      "loss": 0.4319,
+      "step": 8852
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.8499339875933396,
+      "learning_rate": 2.2436654815409964e-06,
+      "loss": 0.4707,
+      "step": 8853
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7661914415380568,
+      "learning_rate": 2.242604473759231e-06,
+      "loss": 0.4247,
+      "step": 8854
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6098408430758295,
+      "learning_rate": 2.241543644377352e-06,
+      "loss": 0.4587,
+      "step": 8855
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5159813986412654,
+      "learning_rate": 2.2404829934639865e-06,
+      "loss": 0.4422,
+      "step": 8856
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4452466883889428,
+      "learning_rate": 2.2394225210877653e-06,
+      "loss": 0.4382,
+      "step": 8857
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5300099823396506,
+      "learning_rate": 2.2383622273172934e-06,
+      "loss": 0.4384,
+      "step": 8858
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.410036168698587,
+      "learning_rate": 2.2373021122211728e-06,
+      "loss": 0.4705,
+      "step": 8859
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6855616743177466,
+      "learning_rate": 2.2362421758679916e-06,
+      "loss": 0.4123,
+      "step": 8860
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.601208286607932,
+      "learning_rate": 2.235182418326326e-06,
+      "loss": 0.4614,
+      "step": 8861
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.449006226234115,
+      "learning_rate": 2.2341228396647437e-06,
+      "loss": 0.4655,
+      "step": 8862
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1004065503937044,
+      "learning_rate": 2.2330634399517926e-06,
+      "loss": 0.4471,
+      "step": 8863
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7483681544224434,
+      "learning_rate": 2.2320042192560175e-06,
+      "loss": 0.41,
+      "step": 8864
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8212878501151792,
+      "learning_rate": 2.230945177645949e-06,
+      "loss": 0.4355,
+      "step": 8865
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.1134133405022233,
+      "learning_rate": 2.229886315190107e-06,
+      "loss": 0.4368,
+      "step": 8866
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.653425199290218,
+      "learning_rate": 2.228827631956994e-06,
+      "loss": 0.4838,
+      "step": 8867
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5187129633694805,
+      "learning_rate": 2.227769128015109e-06,
+      "loss": 0.4635,
+      "step": 8868
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1050924485044678,
+      "learning_rate": 2.2267108034329343e-06,
+      "loss": 0.4461,
+      "step": 8869
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7443921139464742,
+      "learning_rate": 2.225652658278943e-06,
+      "loss": 0.4587,
+      "step": 8870
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9735929362535298,
+      "learning_rate": 2.2245946926215965e-06,
+      "loss": 0.4552,
+      "step": 8871
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4806779958726386,
+      "learning_rate": 2.22353690652934e-06,
+      "loss": 0.4096,
+      "step": 8872
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6664587028003306,
+      "learning_rate": 2.2224793000706165e-06,
+      "loss": 0.4042,
+      "step": 8873
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8698759958732756,
+      "learning_rate": 2.221421873313848e-06,
+      "loss": 0.4567,
+      "step": 8874
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.2532583712821066,
+      "learning_rate": 2.2203646263274483e-06,
+      "loss": 0.437,
+      "step": 8875
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5752094964150016,
+      "learning_rate": 2.2193075591798214e-06,
+      "loss": 0.4742,
+      "step": 8876
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.124275472782954,
+      "learning_rate": 2.21825067193936e-06,
+      "loss": 0.4375,
+      "step": 8877
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6022341797398343,
+      "learning_rate": 2.217193964674439e-06,
+      "loss": 0.4521,
+      "step": 8878
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.669439483093396,
+      "learning_rate": 2.216137437453428e-06,
+      "loss": 0.4393,
+      "step": 8879
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5866524791355068,
+      "learning_rate": 2.2150810903446835e-06,
+      "loss": 0.4913,
+      "step": 8880
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6806622928456794,
+      "learning_rate": 2.214024923416549e-06,
+      "loss": 0.4437,
+      "step": 8881
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.452514326850183,
+      "learning_rate": 2.2129689367373593e-06,
+      "loss": 0.4496,
+      "step": 8882
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5622107964001402,
+      "learning_rate": 2.2119131303754316e-06,
+      "loss": 0.4505,
+      "step": 8883
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.68983625579445,
+      "learning_rate": 2.210857504399077e-06,
+      "loss": 0.535,
+      "step": 8884
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.281983929721882,
+      "learning_rate": 2.209802058876594e-06,
+      "loss": 0.4703,
+      "step": 8885
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3099412208279417,
+      "learning_rate": 2.2087467938762686e-06,
+      "loss": 0.4176,
+      "step": 8886
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6730334850973296,
+      "learning_rate": 2.207691709466371e-06,
+      "loss": 0.4114,
+      "step": 8887
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5680134646993322,
+      "learning_rate": 2.206636805715171e-06,
+      "loss": 0.4425,
+      "step": 8888
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5500234817171427,
+      "learning_rate": 2.205582082690913e-06,
+      "loss": 0.4506,
+      "step": 8889
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4492339526020959,
+      "learning_rate": 2.204527540461841e-06,
+      "loss": 0.4644,
+      "step": 8890
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.990111583427272,
+      "learning_rate": 2.203473179096176e-06,
+      "loss": 0.411,
+      "step": 8891
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5791222300962104,
+      "learning_rate": 2.2024189986621404e-06,
+      "loss": 0.4741,
+      "step": 8892
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6079402146861241,
+      "learning_rate": 2.2013649992279374e-06,
+      "loss": 0.4248,
+      "step": 8893
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.5184362187464617,
+      "learning_rate": 2.200311180861756e-06,
+      "loss": 0.4492,
+      "step": 8894
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.346843490522689,
+      "learning_rate": 2.199257543631779e-06,
+      "loss": 0.4288,
+      "step": 8895
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.719411981071385,
+      "learning_rate": 2.198204087606175e-06,
+      "loss": 0.4164,
+      "step": 8896
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8754511192056444,
+      "learning_rate": 2.1971508128531034e-06,
+      "loss": 0.4817,
+      "step": 8897
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2119384502315884,
+      "learning_rate": 2.1960977194407052e-06,
+      "loss": 0.4503,
+      "step": 8898
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.253599050783305,
+      "learning_rate": 2.195044807437117e-06,
+      "loss": 0.4391,
+      "step": 8899
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.054085900477556,
+      "learning_rate": 2.1939920769104606e-06,
+      "loss": 0.4627,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.163941639262285,
+      "learning_rate": 2.192939527928847e-06,
+      "loss": 0.4818,
+      "step": 8901
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.033932696319072,
+      "learning_rate": 2.191887160560373e-06,
+      "loss": 0.4583,
+      "step": 8902
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4270976456516458,
+      "learning_rate": 2.190834974873124e-06,
+      "loss": 0.452,
+      "step": 8903
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 11.860675687871824,
+      "learning_rate": 2.1897829709351816e-06,
+      "loss": 0.4249,
+      "step": 8904
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5612560136269833,
+      "learning_rate": 2.188731148814602e-06,
+      "loss": 0.449,
+      "step": 8905
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.862368711462543,
+      "learning_rate": 2.1876795085794417e-06,
+      "loss": 0.405,
+      "step": 8906
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.865863097226969,
+      "learning_rate": 2.1866280502977345e-06,
+      "loss": 0.459,
+      "step": 8907
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9782701601440664,
+      "learning_rate": 2.1855767740375156e-06,
+      "loss": 0.4475,
+      "step": 8908
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7652529203169653,
+      "learning_rate": 2.1845256798667956e-06,
+      "loss": 0.4677,
+      "step": 8909
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.0507245313661517,
+      "learning_rate": 2.1834747678535818e-06,
+      "loss": 0.3972,
+      "step": 8910
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.2667124990303824,
+      "learning_rate": 2.1824240380658655e-06,
+      "loss": 0.4434,
+      "step": 8911
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5420719700092652,
+      "learning_rate": 2.1813734905716305e-06,
+      "loss": 0.4724,
+      "step": 8912
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.568467584990093,
+      "learning_rate": 2.1803231254388406e-06,
+      "loss": 0.4626,
+      "step": 8913
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6134359736873944,
+      "learning_rate": 2.179272942735457e-06,
+      "loss": 0.4745,
+      "step": 8914
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.442892257311608,
+      "learning_rate": 2.1782229425294233e-06,
+      "loss": 0.4677,
+      "step": 8915
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.100262255585309,
+      "learning_rate": 2.177173124888674e-06,
+      "loss": 0.4557,
+      "step": 8916
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3427099857604725,
+      "learning_rate": 2.176123489881133e-06,
+      "loss": 0.4695,
+      "step": 8917
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7194150082813031,
+      "learning_rate": 2.175074037574706e-06,
+      "loss": 0.4675,
+      "step": 8918
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1314240298383442,
+      "learning_rate": 2.1740247680372927e-06,
+      "loss": 0.4601,
+      "step": 8919
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.855949414040905,
+      "learning_rate": 2.1729756813367807e-06,
+      "loss": 0.4619,
+      "step": 8920
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8873719564222389,
+      "learning_rate": 2.1719267775410456e-06,
+      "loss": 0.4642,
+      "step": 8921
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5828835973876518,
+      "learning_rate": 2.170878056717945e-06,
+      "loss": 0.4845,
+      "step": 8922
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7116217614610953,
+      "learning_rate": 2.169829518935337e-06,
+      "loss": 0.4399,
+      "step": 8923
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.063976523957411,
+      "learning_rate": 2.1687811642610545e-06,
+      "loss": 0.3979,
+      "step": 8924
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5441462325303021,
+      "learning_rate": 2.167732992762927e-06,
+      "loss": 0.5037,
+      "step": 8925
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9677746957089874,
+      "learning_rate": 2.16668500450877e-06,
+      "loss": 0.4443,
+      "step": 8926
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6562941220793181,
+      "learning_rate": 2.165637199566387e-06,
+      "loss": 0.5172,
+      "step": 8927
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5685035522870938,
+      "learning_rate": 2.1645895780035703e-06,
+      "loss": 0.4541,
+      "step": 8928
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8629978214457865,
+      "learning_rate": 2.1635421398880973e-06,
+      "loss": 0.4443,
+      "step": 8929
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5595506995331464,
+      "learning_rate": 2.1624948852877375e-06,
+      "loss": 0.4741,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.691843019428444,
+      "learning_rate": 2.1614478142702465e-06,
+      "loss": 0.4285,
+      "step": 8931
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 3.5176076962656144,
+      "learning_rate": 2.160400926903371e-06,
+      "loss": 0.4491,
+      "step": 8932
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.63210682468484,
+      "learning_rate": 2.159354223254839e-06,
+      "loss": 0.4373,
+      "step": 8933
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5803370079521174,
+      "learning_rate": 2.158307703392372e-06,
+      "loss": 0.4601,
+      "step": 8934
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8361520284800898,
+      "learning_rate": 2.1572613673836795e-06,
+      "loss": 0.4598,
+      "step": 8935
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.644538121924615,
+      "learning_rate": 2.15621521529646e-06,
+      "loss": 0.4267,
+      "step": 8936
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5643953606874178,
+      "learning_rate": 2.155169247198393e-06,
+      "loss": 0.4821,
+      "step": 8937
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3279099554743983,
+      "learning_rate": 2.1541234631571533e-06,
+      "loss": 0.4323,
+      "step": 8938
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.224607284366617,
+      "learning_rate": 2.1530778632404066e-06,
+      "loss": 0.5263,
+      "step": 8939
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9441309691186217,
+      "learning_rate": 2.1520324475157956e-06,
+      "loss": 0.4066,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6252829909269528,
+      "learning_rate": 2.1509872160509613e-06,
+      "loss": 0.3975,
+      "step": 8941
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6853830683461477,
+      "learning_rate": 2.1499421689135238e-06,
+      "loss": 0.5181,
+      "step": 8942
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5135213307313313,
+      "learning_rate": 2.148897306171103e-06,
+      "loss": 0.4361,
+      "step": 8943
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 4.457617931056689,
+      "learning_rate": 2.1478526278912955e-06,
+      "loss": 0.4784,
+      "step": 8944
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5260377366743276,
+      "learning_rate": 2.1468081341416912e-06,
+      "loss": 0.4389,
+      "step": 8945
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.9782107517266923,
+      "learning_rate": 2.145763824989868e-06,
+      "loss": 0.4756,
+      "step": 8946
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.007025125467285,
+      "learning_rate": 2.144719700503394e-06,
+      "loss": 0.4461,
+      "step": 8947
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5798570109214638,
+      "learning_rate": 2.143675760749818e-06,
+      "loss": 0.4313,
+      "step": 8948
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.011061969263494,
+      "learning_rate": 2.1426320057966844e-06,
+      "loss": 0.4546,
+      "step": 8949
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.301058737234699,
+      "learning_rate": 2.1415884357115213e-06,
+      "loss": 0.459,
+      "step": 8950
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.81892363766736,
+      "learning_rate": 2.1405450505618474e-06,
+      "loss": 0.4166,
+      "step": 8951
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6936392772729532,
+      "learning_rate": 2.1395018504151704e-06,
+      "loss": 0.4235,
+      "step": 8952
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5105893706325855,
+      "learning_rate": 2.1384588353389794e-06,
+      "loss": 0.3963,
+      "step": 8953
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.321971548827934,
+      "learning_rate": 2.137416005400758e-06,
+      "loss": 0.4713,
+      "step": 8954
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8070538892836439,
+      "learning_rate": 2.136373360667977e-06,
+      "loss": 0.4819,
+      "step": 8955
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8640226762022198,
+      "learning_rate": 2.135330901208095e-06,
+      "loss": 0.4304,
+      "step": 8956
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6654939189875235,
+      "learning_rate": 2.1342886270885526e-06,
+      "loss": 0.4595,
+      "step": 8957
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4511470481884279,
+      "learning_rate": 2.133246538376791e-06,
+      "loss": 0.4263,
+      "step": 8958
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8949903729220299,
+      "learning_rate": 2.1322046351402264e-06,
+      "loss": 0.5357,
+      "step": 8959
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8483101409038127,
+      "learning_rate": 2.13116291744627e-06,
+      "loss": 0.4539,
+      "step": 8960
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8035643891043378,
+      "learning_rate": 2.1301213853623206e-06,
+      "loss": 0.4152,
+      "step": 8961
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.956820849736373,
+      "learning_rate": 2.1290800389557626e-06,
+      "loss": 0.4353,
+      "step": 8962
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.702113032256644,
+      "learning_rate": 2.1280388782939725e-06,
+      "loss": 0.4402,
+      "step": 8963
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.754823921076016,
+      "learning_rate": 2.1269979034443076e-06,
+      "loss": 0.4655,
+      "step": 8964
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0251096046613744,
+      "learning_rate": 2.125957114474121e-06,
+      "loss": 0.4482,
+      "step": 8965
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.431460070861409,
+      "learning_rate": 2.124916511450749e-06,
+      "loss": 0.4246,
+      "step": 8966
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.4740637144662565,
+      "learning_rate": 2.1238760944415193e-06,
+      "loss": 0.4853,
+      "step": 8967
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5736244723557816,
+      "learning_rate": 2.122835863513742e-06,
+      "loss": 0.4817,
+      "step": 8968
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.7406323530372831,
+      "learning_rate": 2.1217958187347202e-06,
+      "loss": 0.4716,
+      "step": 8969
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 23.895272184965688,
+      "learning_rate": 2.120755960171745e-06,
+      "loss": 0.4219,
+      "step": 8970
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8415630562611935,
+      "learning_rate": 2.1197162878920914e-06,
+      "loss": 0.4629,
+      "step": 8971
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.246323001883261,
+      "learning_rate": 2.1186768019630283e-06,
+      "loss": 0.4528,
+      "step": 8972
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5775077323595279,
+      "learning_rate": 2.1176375024518037e-06,
+      "loss": 0.4975,
+      "step": 8973
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.0685712504700517,
+      "learning_rate": 2.1165983894256647e-06,
+      "loss": 0.4041,
+      "step": 8974
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8769084345059235,
+      "learning_rate": 2.1155594629518374e-06,
+      "loss": 0.4601,
+      "step": 8975
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5994686404379879,
+      "learning_rate": 2.1145207230975402e-06,
+      "loss": 0.4772,
+      "step": 8976
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.042815409577239,
+      "learning_rate": 2.1134821699299746e-06,
+      "loss": 0.4437,
+      "step": 8977
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7654912509264267,
+      "learning_rate": 2.11244380351634e-06,
+      "loss": 0.4473,
+      "step": 8978
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.423037569776539,
+      "learning_rate": 2.111405623923812e-06,
+      "loss": 0.4733,
+      "step": 8979
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.496522276719993,
+      "learning_rate": 2.1103676312195613e-06,
+      "loss": 0.4791,
+      "step": 8980
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5477178295596188,
+      "learning_rate": 2.109329825470745e-06,
+      "loss": 0.4504,
+      "step": 8981
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1280318578890047,
+      "learning_rate": 2.1082922067445073e-06,
+      "loss": 0.4473,
+      "step": 8982
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5299712522540068,
+      "learning_rate": 2.1072547751079832e-06,
+      "loss": 0.4549,
+      "step": 8983
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2376025277870144,
+      "learning_rate": 2.1062175306282885e-06,
+      "loss": 0.4675,
+      "step": 8984
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5206579654859457,
+      "learning_rate": 2.1051804733725344e-06,
+      "loss": 0.445,
+      "step": 8985
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.6246561210765194,
+      "learning_rate": 2.1041436034078165e-06,
+      "loss": 0.4884,
+      "step": 8986
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5689940381629038,
+      "learning_rate": 2.1031069208012215e-06,
+      "loss": 0.4835,
+      "step": 8987
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.887581087377493,
+      "learning_rate": 2.102070425619817e-06,
+      "loss": 0.4305,
+      "step": 8988
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5533238802699011,
+      "learning_rate": 2.1010341179306655e-06,
+      "loss": 0.4725,
+      "step": 8989
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5630686490331619,
+      "learning_rate": 2.0999979978008134e-06,
+      "loss": 0.4891,
+      "step": 8990
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5760721725173477,
+      "learning_rate": 2.0989620652972993e-06,
+      "loss": 0.4765,
+      "step": 8991
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8820932190718747,
+      "learning_rate": 2.097926320487141e-06,
+      "loss": 0.4796,
+      "step": 8992
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.417879228138292,
+      "learning_rate": 2.096890763437355e-06,
+      "loss": 0.4641,
+      "step": 8993
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7301338686617478,
+      "learning_rate": 2.0958553942149413e-06,
+      "loss": 0.4558,
+      "step": 8994
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5485477568464656,
+      "learning_rate": 2.0948202128868823e-06,
+      "loss": 0.4972,
+      "step": 8995
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5455553790606953,
+      "learning_rate": 2.0937852195201545e-06,
+      "loss": 0.434,
+      "step": 8996
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.089880353129698,
+      "learning_rate": 2.0927504141817216e-06,
+      "loss": 0.4297,
+      "step": 8997
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.806037263372854,
+      "learning_rate": 2.091715796938536e-06,
+      "loss": 0.4792,
+      "step": 8998
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1035219830148573,
+      "learning_rate": 2.0906813678575315e-06,
+      "loss": 0.4728,
+      "step": 8999
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.228726052321276,
+      "learning_rate": 2.089647127005637e-06,
+      "loss": 0.4432,
+      "step": 9000
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5839180102619852,
+      "learning_rate": 2.0886130744497664e-06,
+      "loss": 0.404,
+      "step": 9001
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.42224635361128,
+      "learning_rate": 2.087579210256822e-06,
+      "loss": 0.4393,
+      "step": 9002
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.5257432628567837,
+      "learning_rate": 2.086545534493692e-06,
+      "loss": 0.4647,
+      "step": 9003
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4674013468719789,
+      "learning_rate": 2.085512047227252e-06,
+      "loss": 0.449,
+      "step": 9004
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.060753583317764,
+      "learning_rate": 2.0844787485243745e-06,
+      "loss": 0.453,
+      "step": 9005
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7618055376179753,
+      "learning_rate": 2.083445638451906e-06,
+      "loss": 0.4613,
+      "step": 9006
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5422754477575824,
+      "learning_rate": 2.0824127170766904e-06,
+      "loss": 0.4661,
+      "step": 9007
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6175201461840238,
+      "learning_rate": 2.0813799844655525e-06,
+      "loss": 0.4802,
+      "step": 9008
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6608052641437412,
+      "learning_rate": 2.080347440685315e-06,
+      "loss": 0.4502,
+      "step": 9009
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6289195516378172,
+      "learning_rate": 2.079315085802776e-06,
+      "loss": 0.4424,
+      "step": 9010
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5768548825894335,
+      "learning_rate": 2.078282919884733e-06,
+      "loss": 0.487,
+      "step": 9011
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.738946655970322,
+      "learning_rate": 2.077250942997959e-06,
+      "loss": 0.4867,
+      "step": 9012
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.9968652139666454,
+      "learning_rate": 2.0762191552092297e-06,
+      "loss": 0.4363,
+      "step": 9013
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.413532780583224,
+      "learning_rate": 2.075187556585294e-06,
+      "loss": 0.4895,
+      "step": 9014
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5493171712768339,
+      "learning_rate": 2.074156147192897e-06,
+      "loss": 0.4795,
+      "step": 9015
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2347939832975197,
+      "learning_rate": 2.0731249270987704e-06,
+      "loss": 0.4646,
+      "step": 9016
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.638578408278056,
+      "learning_rate": 2.072093896369632e-06,
+      "loss": 0.4937,
+      "step": 9017
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6408582482948533,
+      "learning_rate": 2.07106305507219e-06,
+      "loss": 0.4195,
+      "step": 9018
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.742528398604224,
+      "learning_rate": 2.070032403273135e-06,
+      "loss": 0.4688,
+      "step": 9019
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.468767821780354,
+      "learning_rate": 2.0690019410391505e-06,
+      "loss": 0.4438,
+      "step": 9020
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3995615027725186,
+      "learning_rate": 2.067971668436907e-06,
+      "loss": 0.4166,
+      "step": 9021
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0048424312357143,
+      "learning_rate": 2.0669415855330616e-06,
+      "loss": 0.4124,
+      "step": 9022
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5710320791150603,
+      "learning_rate": 2.0659116923942575e-06,
+      "loss": 0.5199,
+      "step": 9023
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.594266709523323,
+      "learning_rate": 2.0648819890871285e-06,
+      "loss": 0.4721,
+      "step": 9024
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.558716562215718,
+      "learning_rate": 2.063852475678295e-06,
+      "loss": 0.4226,
+      "step": 9025
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4894564516670177,
+      "learning_rate": 2.0628231522343678e-06,
+      "loss": 0.4699,
+      "step": 9026
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7605007708350828,
+      "learning_rate": 2.061794018821936e-06,
+      "loss": 0.4051,
+      "step": 9027
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9585642006305908,
+      "learning_rate": 2.06076507550759e-06,
+      "loss": 0.4384,
+      "step": 9028
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.738196552594092,
+      "learning_rate": 2.0597363223578996e-06,
+      "loss": 0.4248,
+      "step": 9029
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5472938030587672,
+      "learning_rate": 2.058707759439422e-06,
+      "loss": 0.4787,
+      "step": 9030
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.8503939985995257,
+      "learning_rate": 2.057679386818704e-06,
+      "loss": 0.4982,
+      "step": 9031
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6700524863028299,
+      "learning_rate": 2.0566512045622805e-06,
+      "loss": 0.5455,
+      "step": 9032
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5123265766816918,
+      "learning_rate": 2.055623212736676e-06,
+      "loss": 0.4651,
+      "step": 9033
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5425342497021234,
+      "learning_rate": 2.054595411408396e-06,
+      "loss": 0.4521,
+      "step": 9034
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.729591550286699,
+      "learning_rate": 2.0535678006439396e-06,
+      "loss": 0.471,
+      "step": 9035
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8776881341316929,
+      "learning_rate": 2.052540380509792e-06,
+      "loss": 0.4177,
+      "step": 9036
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6643027711160012,
+      "learning_rate": 2.051513151072428e-06,
+      "loss": 0.4367,
+      "step": 9037
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9135374397307274,
+      "learning_rate": 2.0504861123983042e-06,
+      "loss": 0.4616,
+      "step": 9038
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5873912859784722,
+      "learning_rate": 2.0494592645538686e-06,
+      "loss": 0.4571,
+      "step": 9039
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.512977213640121,
+      "learning_rate": 2.048432607605563e-06,
+      "loss": 0.4658,
+      "step": 9040
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 6.335823981705311,
+      "learning_rate": 2.047406141619804e-06,
+      "loss": 0.423,
+      "step": 9041
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4187137457180805,
+      "learning_rate": 2.046379866663007e-06,
+      "loss": 0.451,
+      "step": 9042
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.03657010410487,
+      "learning_rate": 2.0453537828015653e-06,
+      "loss": 0.4411,
+      "step": 9043
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8738390984382853,
+      "learning_rate": 2.0443278901018725e-06,
+      "loss": 0.4555,
+      "step": 9044
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.698094442229802,
+      "learning_rate": 2.0433021886302966e-06,
+      "loss": 0.4815,
+      "step": 9045
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8242115093743156,
+      "learning_rate": 2.0422766784532035e-06,
+      "loss": 0.4607,
+      "step": 9046
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.1293595366669327,
+      "learning_rate": 2.041251359636936e-06,
+      "loss": 0.4592,
+      "step": 9047
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9513572425874517,
+      "learning_rate": 2.0402262322478387e-06,
+      "loss": 0.435,
+      "step": 9048
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 5.9822646478562636,
+      "learning_rate": 2.0392012963522314e-06,
+      "loss": 0.461,
+      "step": 9049
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.346509100721229,
+      "learning_rate": 2.038176552016426e-06,
+      "loss": 0.4662,
+      "step": 9050
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.809089129866771,
+      "learning_rate": 2.037151999306723e-06,
+      "loss": 0.4367,
+      "step": 9051
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1681372025468284,
+      "learning_rate": 2.0361276382894104e-06,
+      "loss": 0.4285,
+      "step": 9052
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8654387318564098,
+      "learning_rate": 2.035103469030764e-06,
+      "loss": 0.4467,
+      "step": 9053
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5548453404824296,
+      "learning_rate": 2.0340794915970426e-06,
+      "loss": 0.4682,
+      "step": 9054
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 4.623379998691161,
+      "learning_rate": 2.0330557060544982e-06,
+      "loss": 0.504,
+      "step": 9055
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3152380441153833,
+      "learning_rate": 2.0320321124693687e-06,
+      "loss": 0.4928,
+      "step": 9056
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.44446344924222,
+      "learning_rate": 2.03100871090788e-06,
+      "loss": 0.4374,
+      "step": 9057
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.9371770861031106,
+      "learning_rate": 2.029985501436243e-06,
+      "loss": 0.5088,
+      "step": 9058
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5161378238459426,
+      "learning_rate": 2.028962484120658e-06,
+      "loss": 0.4801,
+      "step": 9059
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.596385458370049,
+      "learning_rate": 2.027939659027314e-06,
+      "loss": 0.475,
+      "step": 9060
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8433846742063151,
+      "learning_rate": 2.0269170262223857e-06,
+      "loss": 0.4706,
+      "step": 9061
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3099279524651686,
+      "learning_rate": 2.0258945857720368e-06,
+      "loss": 0.4553,
+      "step": 9062
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5185901334993461,
+      "learning_rate": 2.024872337742418e-06,
+      "loss": 0.468,
+      "step": 9063
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.859923130356896,
+      "learning_rate": 2.0238502821996686e-06,
+      "loss": 0.408,
+      "step": 9064
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3521942265891753,
+      "learning_rate": 2.0228284192099106e-06,
+      "loss": 0.4353,
+      "step": 9065
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.3306694335385116,
+      "learning_rate": 2.021806748839259e-06,
+      "loss": 0.5052,
+      "step": 9066
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.783380075038979,
+      "learning_rate": 2.020785271153815e-06,
+      "loss": 0.4459,
+      "step": 9067
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0065249703452364,
+      "learning_rate": 2.019763986219669e-06,
+      "loss": 0.4692,
+      "step": 9068
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.84834275329012,
+      "learning_rate": 2.018742894102892e-06,
+      "loss": 0.4778,
+      "step": 9069
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.914032266461034,
+      "learning_rate": 2.0177219948695498e-06,
+      "loss": 0.4133,
+      "step": 9070
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6765420255273842,
+      "learning_rate": 2.0167012885856933e-06,
+      "loss": 0.4204,
+      "step": 9071
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.2937124151101687,
+      "learning_rate": 2.015680775317361e-06,
+      "loss": 0.4293,
+      "step": 9072
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.845473503153631,
+      "learning_rate": 2.0146604551305805e-06,
+      "loss": 0.4504,
+      "step": 9073
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5376945963040818,
+      "learning_rate": 2.0136403280913595e-06,
+      "loss": 0.4537,
+      "step": 9074
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.420506663673671,
+      "learning_rate": 2.012620394265706e-06,
+      "loss": 0.4476,
+      "step": 9075
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.46812424897621,
+      "learning_rate": 2.0116006537196033e-06,
+      "loss": 0.4879,
+      "step": 9076
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.1475708348329037,
+      "learning_rate": 2.0105811065190306e-06,
+      "loss": 0.4734,
+      "step": 9077
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7199294451489255,
+      "learning_rate": 2.0095617527299467e-06,
+      "loss": 0.4712,
+      "step": 9078
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8570547107638602,
+      "learning_rate": 2.0085425924183083e-06,
+      "loss": 0.4731,
+      "step": 9079
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.0896312839718028,
+      "learning_rate": 2.00752362565005e-06,
+      "loss": 0.4681,
+      "step": 9080
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 3.061011832455128,
+      "learning_rate": 2.006504852491099e-06,
+      "loss": 0.475,
+      "step": 9081
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7548912097951428,
+      "learning_rate": 2.0054862730073647e-06,
+      "loss": 0.4754,
+      "step": 9082
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7752719396753005,
+      "learning_rate": 2.0044678872647555e-06,
+      "loss": 0.4792,
+      "step": 9083
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.020631961341323,
+      "learning_rate": 2.0034496953291533e-06,
+      "loss": 0.4037,
+      "step": 9084
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5484566486418452,
+      "learning_rate": 2.002431697266436e-06,
+      "loss": 0.4601,
+      "step": 9085
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.7170834054576911,
+      "learning_rate": 2.0014138931424658e-06,
+      "loss": 0.4607,
+      "step": 9086
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.56762613332507,
+      "learning_rate": 2.000396283023095e-06,
+      "loss": 0.4715,
+      "step": 9087
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6321260865427072,
+      "learning_rate": 1.999378866974162e-06,
+      "loss": 0.4746,
+      "step": 9088
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5318285081404429,
+      "learning_rate": 1.9983616450614894e-06,
+      "loss": 0.4559,
+      "step": 9089
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.9360996999888145,
+      "learning_rate": 1.9973446173508913e-06,
+      "loss": 0.4115,
+      "step": 9090
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.920001063505966,
+      "learning_rate": 1.996327783908169e-06,
+      "loss": 0.5405,
+      "step": 9091
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.637046337926212,
+      "learning_rate": 1.995311144799111e-06,
+      "loss": 0.4646,
+      "step": 9092
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0892662190016984,
+      "learning_rate": 1.9942947000894903e-06,
+      "loss": 0.4765,
+      "step": 9093
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5470911614006504,
+      "learning_rate": 1.99327844984507e-06,
+      "loss": 0.4392,
+      "step": 9094
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.152385206313903,
+      "learning_rate": 1.992262394131601e-06,
+      "loss": 0.4473,
+      "step": 9095
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5429104070363848,
+      "learning_rate": 1.9912465330148208e-06,
+      "loss": 0.4847,
+      "step": 9096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.6103553685614254,
+      "learning_rate": 1.990230866560453e-06,
+      "loss": 0.4011,
+      "step": 9097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5783865783756478,
+      "learning_rate": 1.989215394834212e-06,
+      "loss": 0.4858,
+      "step": 9098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.0908368753540607,
+      "learning_rate": 1.988200117901797e-06,
+      "loss": 0.4849,
+      "step": 9099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8203995988067303,
+      "learning_rate": 1.9871850358288936e-06,
+      "loss": 0.4646,
+      "step": 9100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5586448525058509,
+      "learning_rate": 1.986170148681176e-06,
+      "loss": 0.4825,
+      "step": 9101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 2.901636238012747,
+      "learning_rate": 1.9851554565243086e-06,
+      "loss": 0.509,
+      "step": 9102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.552763840138365,
+      "learning_rate": 1.9841409594239402e-06,
+      "loss": 0.4677,
+      "step": 9103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4708208905118039,
+      "learning_rate": 1.9831266574457047e-06,
+      "loss": 0.4286,
+      "step": 9104
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9660444034794247,
+      "learning_rate": 1.982112550655228e-06,
+      "loss": 0.4539,
+      "step": 9105
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.553095877280426,
+      "learning_rate": 1.9810986391181215e-06,
+      "loss": 0.4674,
+      "step": 9106
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5755890513631077,
+      "learning_rate": 1.980084922899984e-06,
+      "loss": 0.4651,
+      "step": 9107
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4373639473080866,
+      "learning_rate": 1.979071402066403e-06,
+      "loss": 0.4511,
+      "step": 9108
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5649220923690779,
+      "learning_rate": 1.9780580766829484e-06,
+      "loss": 0.5041,
+      "step": 9109
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.539407448722253,
+      "learning_rate": 1.9770449468151833e-06,
+      "loss": 0.4831,
+      "step": 9110
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5224227635384933,
+      "learning_rate": 1.976032012528656e-06,
+      "loss": 0.4871,
+      "step": 9111
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.609974931255902,
+      "learning_rate": 1.9750192738889023e-06,
+      "loss": 0.4124,
+      "step": 9112
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4419942911751824,
+      "learning_rate": 1.974006730961442e-06,
+      "loss": 0.451,
+      "step": 9113
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.609049868820141,
+      "learning_rate": 1.9729943838117905e-06,
+      "loss": 0.4072,
+      "step": 9114
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6076993574642724,
+      "learning_rate": 1.971982232505441e-06,
+      "loss": 0.4505,
+      "step": 9115
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5596619700819974,
+      "learning_rate": 1.970970277107882e-06,
+      "loss": 0.4457,
+      "step": 9116
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5552951761548044,
+      "learning_rate": 1.9699585176845796e-06,
+      "loss": 0.4266,
+      "step": 9117
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6115211171873614,
+      "learning_rate": 1.9689469543009997e-06,
+      "loss": 0.4393,
+      "step": 9118
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6686097956781625,
+      "learning_rate": 1.967935587022588e-06,
+      "loss": 0.4445,
+      "step": 9119
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.722510681302889,
+      "learning_rate": 1.9669244159147756e-06,
+      "loss": 0.4319,
+      "step": 9120
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5196452251857038,
+      "learning_rate": 1.9659134410429857e-06,
+      "loss": 0.4391,
+      "step": 9121
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4980451849113845,
+      "learning_rate": 1.9649026624726276e-06,
+      "loss": 0.4359,
+      "step": 9122
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5861242357061968,
+      "learning_rate": 1.963892080269098e-06,
+      "loss": 0.4685,
+      "step": 9123
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4866437437874906,
+      "learning_rate": 1.9628816944977773e-06,
+      "loss": 0.4336,
+      "step": 9124
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.778217294257492,
+      "learning_rate": 1.961871505224038e-06,
+      "loss": 0.4535,
+      "step": 9125
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8422995642936062,
+      "learning_rate": 1.9608615125132387e-06,
+      "loss": 0.4909,
+      "step": 9126
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0333615509661427,
+      "learning_rate": 1.959851716430725e-06,
+      "loss": 0.4618,
+      "step": 9127
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.595272262520607,
+      "learning_rate": 1.9588421170418265e-06,
+      "loss": 0.4823,
+      "step": 9128
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5525596636398344,
+      "learning_rate": 1.957832714411863e-06,
+      "loss": 0.4619,
+      "step": 9129
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5089312348105903,
+      "learning_rate": 1.9568235086061467e-06,
+      "loss": 0.5012,
+      "step": 9130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5932644084783678,
+      "learning_rate": 1.9558144996899665e-06,
+      "loss": 0.4775,
+      "step": 9131
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4762967051107085,
+      "learning_rate": 1.9548056877286058e-06,
+      "loss": 0.4647,
+      "step": 9132
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0064854070553277,
+      "learning_rate": 1.953797072787334e-06,
+      "loss": 0.465,
+      "step": 9133
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6721437463299431,
+      "learning_rate": 1.952788654931408e-06,
+      "loss": 0.4281,
+      "step": 9134
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.4337388986184765,
+      "learning_rate": 1.951780434226068e-06,
+      "loss": 0.3888,
+      "step": 9135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.85172524650859,
+      "learning_rate": 1.9507724107365465e-06,
+      "loss": 0.469,
+      "step": 9136
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.901302928193326,
+      "learning_rate": 1.949764584528061e-06,
+      "loss": 0.5008,
+      "step": 9137
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6410772168762595,
+      "learning_rate": 1.948756955665819e-06,
+      "loss": 0.4112,
+      "step": 9138
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7054388433261165,
+      "learning_rate": 1.947749524215008e-06,
+      "loss": 0.4116,
+      "step": 9139
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0412121462996686,
+      "learning_rate": 1.94674229024081e-06,
+      "loss": 0.4587,
+      "step": 9140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7362147716078151,
+      "learning_rate": 1.945735253808392e-06,
+      "loss": 0.3564,
+      "step": 9141
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.617189113476337,
+      "learning_rate": 1.944728414982907e-06,
+      "loss": 0.4875,
+      "step": 9142
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.093336202199005,
+      "learning_rate": 1.9437217738294987e-06,
+      "loss": 0.4067,
+      "step": 9143
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.456723202227306,
+      "learning_rate": 1.9427153304132923e-06,
+      "loss": 0.4629,
+      "step": 9144
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3922613146136986,
+      "learning_rate": 1.941709084799404e-06,
+      "loss": 0.4023,
+      "step": 9145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9525647966324091,
+      "learning_rate": 1.9407030370529377e-06,
+      "loss": 0.4519,
+      "step": 9146
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5550076223116884,
+      "learning_rate": 1.9396971872389843e-06,
+      "loss": 0.4837,
+      "step": 9147
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.677272708607009,
+      "learning_rate": 1.938691535422616e-06,
+      "loss": 0.418,
+      "step": 9148
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7298142951579487,
+      "learning_rate": 1.9376860816689044e-06,
+      "loss": 0.4428,
+      "step": 9149
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5452945563889064,
+      "learning_rate": 1.9366808260428947e-06,
+      "loss": 0.4189,
+      "step": 9150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.767455687919906,
+      "learning_rate": 1.9356757686096297e-06,
+      "loss": 0.3797,
+      "step": 9151
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.549514929634445,
+      "learning_rate": 1.9346709094341325e-06,
+      "loss": 0.499,
+      "step": 9152
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5308333091790249,
+      "learning_rate": 1.933666248581418e-06,
+      "loss": 0.482,
+      "step": 9153
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5551199359990271,
+      "learning_rate": 1.9326617861164876e-06,
+      "loss": 0.4677,
+      "step": 9154
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.078679556563845,
+      "learning_rate": 1.9316575221043256e-06,
+      "loss": 0.4956,
+      "step": 9155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6984412788820709,
+      "learning_rate": 1.930653456609908e-06,
+      "loss": 0.4739,
+      "step": 9156
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3863778539989897,
+      "learning_rate": 1.929649589698196e-06,
+      "loss": 0.4696,
+      "step": 9157
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3208282700755762,
+      "learning_rate": 1.9286459214341413e-06,
+      "loss": 0.4139,
+      "step": 9158
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5560264677295821,
+      "learning_rate": 1.927642451882676e-06,
+      "loss": 0.4525,
+      "step": 9159
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.106273065112307,
+      "learning_rate": 1.9266391811087243e-06,
+      "loss": 0.518,
+      "step": 9160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4334736069047371,
+      "learning_rate": 1.9256361091771974e-06,
+      "loss": 0.4998,
+      "step": 9161
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7386219616256984,
+      "learning_rate": 1.9246332361529945e-06,
+      "loss": 0.4515,
+      "step": 9162
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5183076130130302,
+      "learning_rate": 1.9236305621009958e-06,
+      "loss": 0.4485,
+      "step": 9163
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5867848655026159,
+      "learning_rate": 1.922628087086074e-06,
+      "loss": 0.4857,
+      "step": 9164
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5093415147121736,
+      "learning_rate": 1.921625811173093e-06,
+      "loss": 0.4676,
+      "step": 9165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.874306764023957,
+      "learning_rate": 1.920623734426893e-06,
+      "loss": 0.4964,
+      "step": 9166
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.766747159618499,
+      "learning_rate": 1.9196218569123094e-06,
+      "loss": 0.4312,
+      "step": 9167
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5406331983404502,
+      "learning_rate": 1.918620178694161e-06,
+      "loss": 0.4471,
+      "step": 9168
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.260759109015618,
+      "learning_rate": 1.917618699837259e-06,
+      "loss": 0.4369,
+      "step": 9169
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.155857237969518,
+      "learning_rate": 1.9166174204063925e-06,
+      "loss": 0.4919,
+      "step": 9170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8222656194680276,
+      "learning_rate": 1.9156163404663453e-06,
+      "loss": 0.4673,
+      "step": 9171
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.747868198546036,
+      "learning_rate": 1.914615460081886e-06,
+      "loss": 0.4616,
+      "step": 9172
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2806562090282863,
+      "learning_rate": 1.913614779317772e-06,
+      "loss": 0.4271,
+      "step": 9173
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.8840310733404526,
+      "learning_rate": 1.912614298238743e-06,
+      "loss": 0.4259,
+      "step": 9174
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1425635769923828,
+      "learning_rate": 1.9116140169095292e-06,
+      "loss": 0.4592,
+      "step": 9175
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.554602568767729,
+      "learning_rate": 1.9106139353948493e-06,
+      "loss": 0.3769,
+      "step": 9176
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.224383722419359,
+      "learning_rate": 1.909614053759406e-06,
+      "loss": 0.469,
+      "step": 9177
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.878350654288025,
+      "learning_rate": 1.9086143720678925e-06,
+      "loss": 0.4799,
+      "step": 9178
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5199105857578176,
+      "learning_rate": 1.907614890384983e-06,
+      "loss": 0.4608,
+      "step": 9179
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.448375439959061,
+      "learning_rate": 1.9066156087753451e-06,
+      "loss": 0.4877,
+      "step": 9180
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4772809464921797,
+      "learning_rate": 1.9056165273036303e-06,
+      "loss": 0.4041,
+      "step": 9181
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5585764725422497,
+      "learning_rate": 1.9046176460344802e-06,
+      "loss": 0.4724,
+      "step": 9182
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5483104127030562,
+      "learning_rate": 1.903618965032516e-06,
+      "loss": 0.4465,
+      "step": 9183
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.759087172651646,
+      "learning_rate": 1.9026204843623569e-06,
+      "loss": 0.4474,
+      "step": 9184
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5382222947620958,
+      "learning_rate": 1.9016222040885994e-06,
+      "loss": 0.4849,
+      "step": 9185
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.2451304371676537,
+      "learning_rate": 1.9006241242758322e-06,
+      "loss": 0.4411,
+      "step": 9186
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7069006085466891,
+      "learning_rate": 1.8996262449886294e-06,
+      "loss": 0.4407,
+      "step": 9187
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.373472597296601,
+      "learning_rate": 1.8986285662915528e-06,
+      "loss": 0.4982,
+      "step": 9188
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7254849091100373,
+      "learning_rate": 1.8976310882491522e-06,
+      "loss": 0.4336,
+      "step": 9189
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0107084218907123,
+      "learning_rate": 1.8966338109259608e-06,
+      "loss": 0.4803,
+      "step": 9190
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7882591634732128,
+      "learning_rate": 1.8956367343865017e-06,
+      "loss": 0.4644,
+      "step": 9191
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8883673999002968,
+      "learning_rate": 1.8946398586952847e-06,
+      "loss": 0.4241,
+      "step": 9192
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5421223452292012,
+      "learning_rate": 1.8936431839168084e-06,
+      "loss": 0.4638,
+      "step": 9193
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3444914491058082,
+      "learning_rate": 1.8926467101155532e-06,
+      "loss": 0.5309,
+      "step": 9194
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5388673124207428,
+      "learning_rate": 1.8916504373559903e-06,
+      "loss": 0.465,
+      "step": 9195
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5987483196596364,
+      "learning_rate": 1.8906543657025784e-06,
+      "loss": 0.3882,
+      "step": 9196
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9093022392877852,
+      "learning_rate": 1.889658495219761e-06,
+      "loss": 0.4376,
+      "step": 9197
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 5.7973848421637735,
+      "learning_rate": 1.888662825971972e-06,
+      "loss": 0.4228,
+      "step": 9198
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4185810799590417,
+      "learning_rate": 1.887667358023625e-06,
+      "loss": 0.4673,
+      "step": 9199
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4879203975084994,
+      "learning_rate": 1.8866720914391313e-06,
+      "loss": 0.446,
+      "step": 9200
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.311890417718689,
+      "learning_rate": 1.8856770262828788e-06,
+      "loss": 0.4633,
+      "step": 9201
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1322965166696397,
+      "learning_rate": 1.884682162619249e-06,
+      "loss": 0.431,
+      "step": 9202
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0796013831000915,
+      "learning_rate": 1.8836875005126083e-06,
+      "loss": 0.5054,
+      "step": 9203
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.3121422649011376,
+      "learning_rate": 1.8826930400273108e-06,
+      "loss": 0.421,
+      "step": 9204
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5678717134258826,
+      "learning_rate": 1.8816987812276937e-06,
+      "loss": 0.4659,
+      "step": 9205
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5420242019248788,
+      "learning_rate": 1.8807047241780868e-06,
+      "loss": 0.4641,
+      "step": 9206
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.9214575866449635,
+      "learning_rate": 1.8797108689428034e-06,
+      "loss": 0.4601,
+      "step": 9207
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7336959115142554,
+      "learning_rate": 1.8787172155861449e-06,
+      "loss": 0.4275,
+      "step": 9208
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.8769876056339727,
+      "learning_rate": 1.8777237641724018e-06,
+      "loss": 0.466,
+      "step": 9209
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.426596782776659,
+      "learning_rate": 1.8767305147658443e-06,
+      "loss": 0.3977,
+      "step": 9210
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4462289258772492,
+      "learning_rate": 1.8757374674307365e-06,
+      "loss": 0.3972,
+      "step": 9211
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5932914619831127,
+      "learning_rate": 1.8747446222313282e-06,
+      "loss": 0.404,
+      "step": 9212
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5519372100243521,
+      "learning_rate": 1.873751979231856e-06,
+      "loss": 0.4905,
+      "step": 9213
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5142662442772574,
+      "learning_rate": 1.872759538496539e-06,
+      "loss": 0.4846,
+      "step": 9214
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.498870063757825,
+      "learning_rate": 1.8717673000895892e-06,
+      "loss": 0.4203,
+      "step": 9215
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.81282711386877,
+      "learning_rate": 1.870775264075203e-06,
+      "loss": 0.4317,
+      "step": 9216
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.9882558339664969,
+      "learning_rate": 1.869783430517565e-06,
+      "loss": 0.408,
+      "step": 9217
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.419733341748207,
+      "learning_rate": 1.8687917994808413e-06,
+      "loss": 0.4722,
+      "step": 9218
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3585326606217072,
+      "learning_rate": 1.867800371029193e-06,
+      "loss": 0.4435,
+      "step": 9219
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5414326528628461,
+      "learning_rate": 1.8668091452267656e-06,
+      "loss": 0.4861,
+      "step": 9220
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.84589213537367,
+      "learning_rate": 1.865818122137686e-06,
+      "loss": 0.4375,
+      "step": 9221
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5199006412952465,
+      "learning_rate": 1.8648273018260737e-06,
+      "loss": 0.3748,
+      "step": 9222
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5208804858357531,
+      "learning_rate": 1.8638366843560335e-06,
+      "loss": 0.4541,
+      "step": 9223
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.518205788056479,
+      "learning_rate": 1.8628462697916594e-06,
+      "loss": 0.454,
+      "step": 9224
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.6400893076547673,
+      "learning_rate": 1.8618560581970258e-06,
+      "loss": 0.484,
+      "step": 9225
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4480287237737612,
+      "learning_rate": 1.8608660496361997e-06,
+      "loss": 0.4083,
+      "step": 9226
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5329479549070404,
+      "learning_rate": 1.8598762441732343e-06,
+      "loss": 0.4236,
+      "step": 9227
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.1628207731848867,
+      "learning_rate": 1.8588866418721691e-06,
+      "loss": 0.4044,
+      "step": 9228
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.7215815356493622,
+      "learning_rate": 1.8578972427970281e-06,
+      "loss": 0.45,
+      "step": 9229
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.6611421374795294,
+      "learning_rate": 1.856908047011823e-06,
+      "loss": 0.494,
+      "step": 9230
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3915498632270553,
+      "learning_rate": 1.8559190545805589e-06,
+      "loss": 0.4175,
+      "step": 9231
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5541304575825532,
+      "learning_rate": 1.8549302655672175e-06,
+      "loss": 0.4754,
+      "step": 9232
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6711980062408236,
+      "learning_rate": 1.8539416800357752e-06,
+      "loss": 0.4579,
+      "step": 9233
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.63264840669654,
+      "learning_rate": 1.8529532980501874e-06,
+      "loss": 0.4018,
+      "step": 9234
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5190269963064156,
+      "learning_rate": 1.8519651196744076e-06,
+      "loss": 0.3929,
+      "step": 9235
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5045470009754347,
+      "learning_rate": 1.8509771449723646e-06,
+      "loss": 0.4788,
+      "step": 9236
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5421763058944935,
+      "learning_rate": 1.849989374007981e-06,
+      "loss": 0.4854,
+      "step": 9237
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9098538486400658,
+      "learning_rate": 1.8490018068451644e-06,
+      "loss": 0.4452,
+      "step": 9238
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3888801205920416,
+      "learning_rate": 1.8480144435478104e-06,
+      "loss": 0.4627,
+      "step": 9239
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6722898497132048,
+      "learning_rate": 1.847027284179797e-06,
+      "loss": 0.4271,
+      "step": 9240
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5033792356736404,
+      "learning_rate": 1.8460403288049934e-06,
+      "loss": 0.4277,
+      "step": 9241
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.391308248407966,
+      "learning_rate": 1.8450535774872546e-06,
+      "loss": 0.4019,
+      "step": 9242
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.140068029200228,
+      "learning_rate": 1.844067030290423e-06,
+      "loss": 0.4693,
+      "step": 9243
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9755847421866868,
+      "learning_rate": 1.8430806872783269e-06,
+      "loss": 0.4441,
+      "step": 9244
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5460951761563018,
+      "learning_rate": 1.842094548514779e-06,
+      "loss": 0.4513,
+      "step": 9245
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2005221529226167,
+      "learning_rate": 1.8411086140635831e-06,
+      "loss": 0.4939,
+      "step": 9246
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8903767635733937,
+      "learning_rate": 1.8401228839885271e-06,
+      "loss": 0.438,
+      "step": 9247
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0393010721104523,
+      "learning_rate": 1.839137358353389e-06,
+      "loss": 0.5114,
+      "step": 9248
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.946580249507811,
+      "learning_rate": 1.8381520372219265e-06,
+      "loss": 0.4953,
+      "step": 9249
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5556249492954387,
+      "learning_rate": 1.8371669206578914e-06,
+      "loss": 0.487,
+      "step": 9250
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9477080437888161,
+      "learning_rate": 1.8361820087250194e-06,
+      "loss": 0.4144,
+      "step": 9251
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.4589068885400316,
+      "learning_rate": 1.8351973014870339e-06,
+      "loss": 0.448,
+      "step": 9252
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5362232240013247,
+      "learning_rate": 1.834212799007639e-06,
+      "loss": 0.4662,
+      "step": 9253
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5367898453772608,
+      "learning_rate": 1.8332285013505374e-06,
+      "loss": 0.4171,
+      "step": 9254
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8864570404520737,
+      "learning_rate": 1.83224440857941e-06,
+      "loss": 0.4638,
+      "step": 9255
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.59273282789299,
+      "learning_rate": 1.8312605207579243e-06,
+      "loss": 0.4393,
+      "step": 9256
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5666707898029244,
+      "learning_rate": 1.8302768379497372e-06,
+      "loss": 0.5038,
+      "step": 9257
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.463357693049538,
+      "learning_rate": 1.8292933602184926e-06,
+      "loss": 0.4675,
+      "step": 9258
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9139614599752526,
+      "learning_rate": 1.8283100876278214e-06,
+      "loss": 0.4533,
+      "step": 9259
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3427453874249586,
+      "learning_rate": 1.8273270202413362e-06,
+      "loss": 0.4201,
+      "step": 9260
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5240255596541301,
+      "learning_rate": 1.8263441581226433e-06,
+      "loss": 0.4584,
+      "step": 9261
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.994949089005115,
+      "learning_rate": 1.825361501335331e-06,
+      "loss": 0.4578,
+      "step": 9262
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7777001466533116,
+      "learning_rate": 1.8243790499429787e-06,
+      "loss": 0.4535,
+      "step": 9263
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.813926574327787,
+      "learning_rate": 1.8233968040091455e-06,
+      "loss": 0.4297,
+      "step": 9264
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.038049685055486,
+      "learning_rate": 1.822414763597382e-06,
+      "loss": 0.4162,
+      "step": 9265
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5785531416329566,
+      "learning_rate": 1.8214329287712301e-06,
+      "loss": 0.4644,
+      "step": 9266
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.816192217168559,
+      "learning_rate": 1.820451299594208e-06,
+      "loss": 0.4696,
+      "step": 9267
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6655227772043057,
+      "learning_rate": 1.819469876129829e-06,
+      "loss": 0.4297,
+      "step": 9268
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.708322419406571,
+      "learning_rate": 1.8184886584415851e-06,
+      "loss": 0.4275,
+      "step": 9269
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6760379698829073,
+      "learning_rate": 1.8175076465929658e-06,
+      "loss": 0.4527,
+      "step": 9270
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.545372721176282,
+      "learning_rate": 1.816526840647438e-06,
+      "loss": 0.4836,
+      "step": 9271
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1280552135995947,
+      "learning_rate": 1.8155462406684598e-06,
+      "loss": 0.4463,
+      "step": 9272
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8441588535034579,
+      "learning_rate": 1.814565846719471e-06,
+      "loss": 0.4879,
+      "step": 9273
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9454659232218547,
+      "learning_rate": 1.8135856588639083e-06,
+      "loss": 0.5036,
+      "step": 9274
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.2368555782766117,
+      "learning_rate": 1.8126056771651835e-06,
+      "loss": 0.433,
+      "step": 9275
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0002766260060922,
+      "learning_rate": 1.8116259016867017e-06,
+      "loss": 0.4321,
+      "step": 9276
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5621497408715517,
+      "learning_rate": 1.8106463324918528e-06,
+      "loss": 0.4321,
+      "step": 9277
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4859241322226107,
+      "learning_rate": 1.809666969644014e-06,
+      "loss": 0.394,
+      "step": 9278
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.315578274882957,
+      "learning_rate": 1.808687813206551e-06,
+      "loss": 0.4011,
+      "step": 9279
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7546561153206832,
+      "learning_rate": 1.8077088632428098e-06,
+      "loss": 0.4231,
+      "step": 9280
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6751986167005537,
+      "learning_rate": 1.806730119816129e-06,
+      "loss": 0.4582,
+      "step": 9281
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3984430950892055,
+      "learning_rate": 1.805751582989832e-06,
+      "loss": 0.4586,
+      "step": 9282
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5345368492872156,
+      "learning_rate": 1.804773252827231e-06,
+      "loss": 0.4621,
+      "step": 9283
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8772328323655694,
+      "learning_rate": 1.803795129391619e-06,
+      "loss": 0.4708,
+      "step": 9284
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5103311658715841,
+      "learning_rate": 1.8028172127462806e-06,
+      "loss": 0.437,
+      "step": 9285
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8392821683791205,
+      "learning_rate": 1.801839502954486e-06,
+      "loss": 0.4159,
+      "step": 9286
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9755264681625297,
+      "learning_rate": 1.8008620000794923e-06,
+      "loss": 0.4176,
+      "step": 9287
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.744678326855679,
+      "learning_rate": 1.799884704184542e-06,
+      "loss": 0.4596,
+      "step": 9288
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3276883038910463,
+      "learning_rate": 1.798907615332865e-06,
+      "loss": 0.4224,
+      "step": 9289
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9183167014473168,
+      "learning_rate": 1.7979307335876795e-06,
+      "loss": 0.465,
+      "step": 9290
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7817855433985248,
+      "learning_rate": 1.7969540590121854e-06,
+      "loss": 0.4438,
+      "step": 9291
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5718021713930184,
+      "learning_rate": 1.7959775916695727e-06,
+      "loss": 0.4664,
+      "step": 9292
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5858898233712196,
+      "learning_rate": 1.7950013316230192e-06,
+      "loss": 0.4779,
+      "step": 9293
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.5968143485916064,
+      "learning_rate": 1.7940252789356882e-06,
+      "loss": 0.4526,
+      "step": 9294
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.3469827852523824,
+      "learning_rate": 1.793049433670726e-06,
+      "loss": 0.4858,
+      "step": 9295
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0758638365294835,
+      "learning_rate": 1.7920737958912704e-06,
+      "loss": 0.4653,
+      "step": 9296
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5430044710587435,
+      "learning_rate": 1.7910983656604436e-06,
+      "loss": 0.4693,
+      "step": 9297
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.7602142893216746,
+      "learning_rate": 1.7901231430413545e-06,
+      "loss": 0.4879,
+      "step": 9298
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3472476607751918,
+      "learning_rate": 1.7891481280971014e-06,
+      "loss": 0.4691,
+      "step": 9299
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.541496911455217,
+      "learning_rate": 1.7881733208907603e-06,
+      "loss": 0.4711,
+      "step": 9300
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4798230781164425,
+      "learning_rate": 1.787198721485407e-06,
+      "loss": 0.4784,
+      "step": 9301
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0178051855943138,
+      "learning_rate": 1.7862243299440917e-06,
+      "loss": 0.4781,
+      "step": 9302
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9868237610745132,
+      "learning_rate": 1.7852501463298605e-06,
+      "loss": 0.4474,
+      "step": 9303
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4928139370725808,
+      "learning_rate": 1.7842761707057355e-06,
+      "loss": 0.4428,
+      "step": 9304
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6491388189457152,
+      "learning_rate": 1.783302403134739e-06,
+      "loss": 0.4547,
+      "step": 9305
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5511368577955398,
+      "learning_rate": 1.7823288436798674e-06,
+      "loss": 0.4769,
+      "step": 9306
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5586850148950284,
+      "learning_rate": 1.7813554924041121e-06,
+      "loss": 0.4154,
+      "step": 9307
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.447723229103301,
+      "learning_rate": 1.7803823493704426e-06,
+      "loss": 0.485,
+      "step": 9308
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5967824889120015,
+      "learning_rate": 1.7794094146418266e-06,
+      "loss": 0.4526,
+      "step": 9309
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7050161123000305,
+      "learning_rate": 1.7784366882812066e-06,
+      "loss": 0.4809,
+      "step": 9310
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9088807439335018,
+      "learning_rate": 1.7774641703515189e-06,
+      "loss": 0.4201,
+      "step": 9311
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4174311003575504,
+      "learning_rate": 1.7764918609156835e-06,
+      "loss": 0.4769,
+      "step": 9312
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6198413680197977,
+      "learning_rate": 1.7755197600366076e-06,
+      "loss": 0.4949,
+      "step": 9313
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5362808728343472,
+      "learning_rate": 1.774547867777187e-06,
+      "loss": 0.4641,
+      "step": 9314
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3923999536205864,
+      "learning_rate": 1.7735761842002986e-06,
+      "loss": 0.4607,
+      "step": 9315
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9420880552116575,
+      "learning_rate": 1.7726047093688098e-06,
+      "loss": 0.4563,
+      "step": 9316
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6010606847688019,
+      "learning_rate": 1.7716334433455746e-06,
+      "loss": 0.4841,
+      "step": 9317
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5922081536444679,
+      "learning_rate": 1.7706623861934341e-06,
+      "loss": 0.4559,
+      "step": 9318
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.363557857965496,
+      "learning_rate": 1.7696915379752111e-06,
+      "loss": 0.4795,
+      "step": 9319
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.130864049268851,
+      "learning_rate": 1.7687208987537197e-06,
+      "loss": 0.4896,
+      "step": 9320
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6236941079567966,
+      "learning_rate": 1.767750468591759e-06,
+      "loss": 0.5046,
+      "step": 9321
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.571103114808913,
+      "learning_rate": 1.7667802475521145e-06,
+      "loss": 0.4693,
+      "step": 9322
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5359202050752966,
+      "learning_rate": 1.7658102356975588e-06,
+      "loss": 0.4683,
+      "step": 9323
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5456737112928702,
+      "learning_rate": 1.7648404330908496e-06,
+      "loss": 0.4762,
+      "step": 9324
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4335623476107258,
+      "learning_rate": 1.7638708397947345e-06,
+      "loss": 0.4111,
+      "step": 9325
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5906323543995217,
+      "learning_rate": 1.762901455871941e-06,
+      "loss": 0.4876,
+      "step": 9326
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5577277279246093,
+      "learning_rate": 1.761932281385188e-06,
+      "loss": 0.4681,
+      "step": 9327
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.804430177140995,
+      "learning_rate": 1.7609633163971806e-06,
+      "loss": 0.424,
+      "step": 9328
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6115224028505788,
+      "learning_rate": 1.7599945609706115e-06,
+      "loss": 0.4694,
+      "step": 9329
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1762393898851204,
+      "learning_rate": 1.7590260151681543e-06,
+      "loss": 0.3745,
+      "step": 9330
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.6722203559996953,
+      "learning_rate": 1.758057679052474e-06,
+      "loss": 0.4419,
+      "step": 9331
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.8229225259778827,
+      "learning_rate": 1.7570895526862202e-06,
+      "loss": 0.4483,
+      "step": 9332
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.07559915063841,
+      "learning_rate": 1.756121636132031e-06,
+      "loss": 0.4731,
+      "step": 9333
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4463621563456848,
+      "learning_rate": 1.75515392945253e-06,
+      "loss": 0.4036,
+      "step": 9334
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.681805116418943,
+      "learning_rate": 1.7541864327103208e-06,
+      "loss": 0.4379,
+      "step": 9335
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9468794944631487,
+      "learning_rate": 1.7532191459680076e-06,
+      "loss": 0.4798,
+      "step": 9336
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.58254993002841,
+      "learning_rate": 1.7522520692881661e-06,
+      "loss": 0.4886,
+      "step": 9337
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5990588669648145,
+      "learning_rate": 1.7512852027333694e-06,
+      "loss": 0.4393,
+      "step": 9338
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7985446580840798,
+      "learning_rate": 1.7503185463661666e-06,
+      "loss": 0.4427,
+      "step": 9339
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.61036880072945,
+      "learning_rate": 1.7493521002491054e-06,
+      "loss": 0.4524,
+      "step": 9340
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0426783606222543,
+      "learning_rate": 1.74838586444471e-06,
+      "loss": 0.41,
+      "step": 9341
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1366774435231237,
+      "learning_rate": 1.7474198390154974e-06,
+      "loss": 0.4091,
+      "step": 9342
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7556505387703467,
+      "learning_rate": 1.7464540240239626e-06,
+      "loss": 0.4505,
+      "step": 9343
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.6566058792548293,
+      "learning_rate": 1.7454884195325977e-06,
+      "loss": 0.4551,
+      "step": 9344
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.9662522560558913,
+      "learning_rate": 1.7445230256038764e-06,
+      "loss": 0.5178,
+      "step": 9345
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.509436944987248,
+      "learning_rate": 1.7435578423002553e-06,
+      "loss": 0.4226,
+      "step": 9346
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5255955807171936,
+      "learning_rate": 1.7425928696841815e-06,
+      "loss": 0.444,
+      "step": 9347
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0268476223997896,
+      "learning_rate": 1.741628107818088e-06,
+      "loss": 0.4717,
+      "step": 9348
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.0849529328624525,
+      "learning_rate": 1.7406635567643948e-06,
+      "loss": 0.4512,
+      "step": 9349
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1912060512591713,
+      "learning_rate": 1.7396992165855047e-06,
+      "loss": 0.5107,
+      "step": 9350
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7214610611756402,
+      "learning_rate": 1.7387350873438097e-06,
+      "loss": 0.4646,
+      "step": 9351
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5924600883896598,
+      "learning_rate": 1.7377711691016885e-06,
+      "loss": 0.5431,
+      "step": 9352
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.409935865581491,
+      "learning_rate": 1.736807461921507e-06,
+      "loss": 0.4661,
+      "step": 9353
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 3.547876913119003,
+      "learning_rate": 1.7358439658656119e-06,
+      "loss": 0.4549,
+      "step": 9354
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.966440331252713,
+      "learning_rate": 1.7348806809963404e-06,
+      "loss": 0.4637,
+      "step": 9355
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7112132617858995,
+      "learning_rate": 1.733917607376021e-06,
+      "loss": 0.4479,
+      "step": 9356
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4424168307166918,
+      "learning_rate": 1.7329547450669586e-06,
+      "loss": 0.4334,
+      "step": 9357
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4016292961403505,
+      "learning_rate": 1.73199209413145e-06,
+      "loss": 0.449,
+      "step": 9358
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4481316452250996,
+      "learning_rate": 1.7310296546317778e-06,
+      "loss": 0.4557,
+      "step": 9359
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5016184556563066,
+      "learning_rate": 1.730067426630212e-06,
+      "loss": 0.4622,
+      "step": 9360
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.2849569471795887,
+      "learning_rate": 1.7291054101890048e-06,
+      "loss": 0.4717,
+      "step": 9361
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7797826265308634,
+      "learning_rate": 1.7281436053703987e-06,
+      "loss": 0.416,
+      "step": 9362
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.8338266262279257,
+      "learning_rate": 1.7271820122366211e-06,
+      "loss": 0.4906,
+      "step": 9363
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.488351220973661,
+      "learning_rate": 1.7262206308498874e-06,
+      "loss": 0.4814,
+      "step": 9364
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5364365567445476,
+      "learning_rate": 1.725259461272395e-06,
+      "loss": 0.469,
+      "step": 9365
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5475691234460252,
+      "learning_rate": 1.7242985035663312e-06,
+      "loss": 0.4836,
+      "step": 9366
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.134605342680636,
+      "learning_rate": 1.7233377577938693e-06,
+      "loss": 0.4314,
+      "step": 9367
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9681300335814162,
+      "learning_rate": 1.7223772240171676e-06,
+      "loss": 0.4693,
+      "step": 9368
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.38245372498438,
+      "learning_rate": 1.721416902298374e-06,
+      "loss": 0.4229,
+      "step": 9369
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.2502208510489177,
+      "learning_rate": 1.7204567926996145e-06,
+      "loss": 0.4216,
+      "step": 9370
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5695482962660846,
+      "learning_rate": 1.7194968952830137e-06,
+      "loss": 0.4651,
+      "step": 9371
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.94518794761535,
+      "learning_rate": 1.7185372101106706e-06,
+      "loss": 0.4214,
+      "step": 9372
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1270999082332507,
+      "learning_rate": 1.717577737244679e-06,
+      "loss": 0.4356,
+      "step": 9373
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.803947583203103,
+      "learning_rate": 1.71661847674711e-06,
+      "loss": 0.4631,
+      "step": 9374
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5611308774126539,
+      "learning_rate": 1.7156594286800344e-06,
+      "loss": 0.4898,
+      "step": 9375
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7550734571660456,
+      "learning_rate": 1.7147005931054956e-06,
+      "loss": 0.4514,
+      "step": 9376
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.603361465568647,
+      "learning_rate": 1.7137419700855302e-06,
+      "loss": 0.4781,
+      "step": 9377
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.589085572953168,
+      "learning_rate": 1.7127835596821606e-06,
+      "loss": 0.438,
+      "step": 9378
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.339629372614089,
+      "learning_rate": 1.7118253619573944e-06,
+      "loss": 0.4657,
+      "step": 9379
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.815802526264609,
+      "learning_rate": 1.7108673769732275e-06,
+      "loss": 0.4439,
+      "step": 9380
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6688502173016868,
+      "learning_rate": 1.7099096047916375e-06,
+      "loss": 0.4582,
+      "step": 9381
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5982164421126173,
+      "learning_rate": 1.7089520454745912e-06,
+      "loss": 0.4237,
+      "step": 9382
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6970059074734893,
+      "learning_rate": 1.707994699084043e-06,
+      "loss": 0.4422,
+      "step": 9383
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4143705130229824,
+      "learning_rate": 1.7070375656819326e-06,
+      "loss": 0.4219,
+      "step": 9384
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5599999410261133,
+      "learning_rate": 1.706080645330182e-06,
+      "loss": 0.4577,
+      "step": 9385
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5586146254703207,
+      "learning_rate": 1.7051239380907054e-06,
+      "loss": 0.4774,
+      "step": 9386
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.249217766061195,
+      "learning_rate": 1.7041674440253991e-06,
+      "loss": 0.4788,
+      "step": 9387
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.395151535946371,
+      "learning_rate": 1.7032111631961502e-06,
+      "loss": 0.4569,
+      "step": 9388
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4964013857927392,
+      "learning_rate": 1.7022550956648237e-06,
+      "loss": 0.4444,
+      "step": 9389
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5279867411395442,
+      "learning_rate": 1.7012992414932773e-06,
+      "loss": 0.4287,
+      "step": 9390
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7686582251542629,
+      "learning_rate": 1.7003436007433583e-06,
+      "loss": 0.4121,
+      "step": 9391
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5200612202428085,
+      "learning_rate": 1.6993881734768897e-06,
+      "loss": 0.4541,
+      "step": 9392
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.841342268550445,
+      "learning_rate": 1.6984329597556886e-06,
+      "loss": 0.4745,
+      "step": 9393
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7093319680666754,
+      "learning_rate": 1.697477959641556e-06,
+      "loss": 0.4789,
+      "step": 9394
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5436758927180081,
+      "learning_rate": 1.6965231731962811e-06,
+      "loss": 0.4904,
+      "step": 9395
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8113144895817426,
+      "learning_rate": 1.6955686004816335e-06,
+      "loss": 0.4521,
+      "step": 9396
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5108512793777908,
+      "learning_rate": 1.6946142415593748e-06,
+      "loss": 0.4239,
+      "step": 9397
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.258795270221807,
+      "learning_rate": 1.6936600964912508e-06,
+      "loss": 0.5082,
+      "step": 9398
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7835922750894324,
+      "learning_rate": 1.6927061653389948e-06,
+      "loss": 0.414,
+      "step": 9399
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.558931009445231,
+      "learning_rate": 1.6917524481643216e-06,
+      "loss": 0.465,
+      "step": 9400
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5990602333267177,
+      "learning_rate": 1.6907989450289375e-06,
+      "loss": 0.4732,
+      "step": 9401
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.697666986131221,
+      "learning_rate": 1.6898456559945332e-06,
+      "loss": 0.3865,
+      "step": 9402
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7559005237630128,
+      "learning_rate": 1.6888925811227841e-06,
+      "loss": 0.4168,
+      "step": 9403
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1026836853962636,
+      "learning_rate": 1.687939720475355e-06,
+      "loss": 0.4133,
+      "step": 9404
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5645004944966443,
+      "learning_rate": 1.6869870741138906e-06,
+      "loss": 0.4667,
+      "step": 9405
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0396792270399624,
+      "learning_rate": 1.6860346421000311e-06,
+      "loss": 0.4497,
+      "step": 9406
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1391574834967404,
+      "learning_rate": 1.685082424495394e-06,
+      "loss": 0.4735,
+      "step": 9407
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2831335441920926,
+      "learning_rate": 1.6841304213615889e-06,
+      "loss": 0.437,
+      "step": 9408
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7141446056692833,
+      "learning_rate": 1.6831786327602045e-06,
+      "loss": 0.4728,
+      "step": 9409
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9154358509445302,
+      "learning_rate": 1.6822270587528273e-06,
+      "loss": 0.5231,
+      "step": 9410
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4284530672236355,
+      "learning_rate": 1.681275699401017e-06,
+      "loss": 0.3948,
+      "step": 9411
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.648456691912696,
+      "learning_rate": 1.6803245547663278e-06,
+      "loss": 0.4027,
+      "step": 9412
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0201403472041424,
+      "learning_rate": 1.6793736249102976e-06,
+      "loss": 0.4401,
+      "step": 9413
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.543950347878804,
+      "learning_rate": 1.6784229098944493e-06,
+      "loss": 0.4662,
+      "step": 9414
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7836923791801027,
+      "learning_rate": 1.6774724097802959e-06,
+      "loss": 0.4558,
+      "step": 9415
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5843778299191118,
+      "learning_rate": 1.676522124629329e-06,
+      "loss": 0.454,
+      "step": 9416
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1600457084528237,
+      "learning_rate": 1.675572054503033e-06,
+      "loss": 0.4858,
+      "step": 9417
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.575299610913128,
+      "learning_rate": 1.6746221994628764e-06,
+      "loss": 0.4033,
+      "step": 9418
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4931203483121034,
+      "learning_rate": 1.6736725595703145e-06,
+      "loss": 0.4551,
+      "step": 9419
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8352633826801457,
+      "learning_rate": 1.6727231348867856e-06,
+      "loss": 0.4587,
+      "step": 9420
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5924941548761598,
+      "learning_rate": 1.671773925473717e-06,
+      "loss": 0.4667,
+      "step": 9421
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5431038026124573,
+      "learning_rate": 1.6708249313925217e-06,
+      "loss": 0.3973,
+      "step": 9422
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.19925690089107,
+      "learning_rate": 1.6698761527045982e-06,
+      "loss": 0.4031,
+      "step": 9423
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6992884126470247,
+      "learning_rate": 1.6689275894713326e-06,
+      "loss": 0.4554,
+      "step": 9424
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5158162637385297,
+      "learning_rate": 1.6679792417540913e-06,
+      "loss": 0.4234,
+      "step": 9425
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5453318226443147,
+      "learning_rate": 1.6670311096142371e-06,
+      "loss": 0.4825,
+      "step": 9426
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3826283821909433,
+      "learning_rate": 1.666083193113109e-06,
+      "loss": 0.4079,
+      "step": 9427
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9418228459360969,
+      "learning_rate": 1.6651354923120367e-06,
+      "loss": 0.4107,
+      "step": 9428
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5313753955389569,
+      "learning_rate": 1.664188007272336e-06,
+      "loss": 0.4615,
+      "step": 9429
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 21.753136229551092,
+      "learning_rate": 1.6632407380553085e-06,
+      "loss": 0.4598,
+      "step": 9430
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.438014344140615,
+      "learning_rate": 1.6622936847222386e-06,
+      "loss": 0.4171,
+      "step": 9431
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.954942575705391,
+      "learning_rate": 1.6613468473344013e-06,
+      "loss": 0.4117,
+      "step": 9432
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.488685423206424,
+      "learning_rate": 1.6604002259530549e-06,
+      "loss": 0.4011,
+      "step": 9433
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6029407610868032,
+      "learning_rate": 1.659453820639446e-06,
+      "loss": 0.4283,
+      "step": 9434
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.794421523022401,
+      "learning_rate": 1.658507631454806e-06,
+      "loss": 0.464,
+      "step": 9435
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 10.459092209644698,
+      "learning_rate": 1.6575616584603493e-06,
+      "loss": 0.4383,
+      "step": 9436
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.729153750110327,
+      "learning_rate": 1.6566159017172812e-06,
+      "loss": 0.5503,
+      "step": 9437
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2554255729493307,
+      "learning_rate": 1.6556703612867904e-06,
+      "loss": 0.4288,
+      "step": 9438
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.413585252589411,
+      "learning_rate": 1.6547250372300538e-06,
+      "loss": 0.4772,
+      "step": 9439
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7488345497394422,
+      "learning_rate": 1.653779929608228e-06,
+      "loss": 0.4478,
+      "step": 9440
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7867792633302464,
+      "learning_rate": 1.6528350384824671e-06,
+      "loss": 0.4388,
+      "step": 9441
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5552530763629887,
+      "learning_rate": 1.6518903639138983e-06,
+      "loss": 0.4783,
+      "step": 9442
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.403988482679584,
+      "learning_rate": 1.6509459059636445e-06,
+      "loss": 0.4146,
+      "step": 9443
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5187400008910756,
+      "learning_rate": 1.6500016646928069e-06,
+      "loss": 0.4572,
+      "step": 9444
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5429090828856518,
+      "learning_rate": 1.6490576401624803e-06,
+      "loss": 0.4624,
+      "step": 9445
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.8945126541416097,
+      "learning_rate": 1.648113832433743e-06,
+      "loss": 0.4814,
+      "step": 9446
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5656748645440883,
+      "learning_rate": 1.6471702415676538e-06,
+      "loss": 0.4853,
+      "step": 9447
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.007872785565353,
+      "learning_rate": 1.646226867625264e-06,
+      "loss": 0.4346,
+      "step": 9448
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.958805964340988,
+      "learning_rate": 1.6452837106676089e-06,
+      "loss": 0.4651,
+      "step": 9449
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5760445978046895,
+      "learning_rate": 1.6443407707557103e-06,
+      "loss": 0.4191,
+      "step": 9450
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.208537671216683,
+      "learning_rate": 1.6433980479505728e-06,
+      "loss": 0.5142,
+      "step": 9451
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6558865705304067,
+      "learning_rate": 1.6424555423131905e-06,
+      "loss": 0.4232,
+      "step": 9452
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7879914755597162,
+      "learning_rate": 1.6415132539045424e-06,
+      "loss": 0.4403,
+      "step": 9453
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.735197036523958,
+      "learning_rate": 1.6405711827855952e-06,
+      "loss": 0.452,
+      "step": 9454
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7411948269339976,
+      "learning_rate": 1.639629329017296e-06,
+      "loss": 0.4381,
+      "step": 9455
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6955786518592089,
+      "learning_rate": 1.6386876926605816e-06,
+      "loss": 0.4735,
+      "step": 9456
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.247884104685819,
+      "learning_rate": 1.6377462737763794e-06,
+      "loss": 0.4553,
+      "step": 9457
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9240998028511493,
+      "learning_rate": 1.6368050724255935e-06,
+      "loss": 0.4466,
+      "step": 9458
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7293031957088278,
+      "learning_rate": 1.6358640886691213e-06,
+      "loss": 0.4111,
+      "step": 9459
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.559767266275453,
+      "learning_rate": 1.634923322567839e-06,
+      "loss": 0.4875,
+      "step": 9460
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.048214493890338,
+      "learning_rate": 1.6339827741826181e-06,
+      "loss": 0.3973,
+      "step": 9461
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1239997429149633,
+      "learning_rate": 1.6330424435743076e-06,
+      "loss": 0.4267,
+      "step": 9462
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 3.4019879411903484,
+      "learning_rate": 1.632102330803746e-06,
+      "loss": 0.4343,
+      "step": 9463
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9580660584644536,
+      "learning_rate": 1.631162435931758e-06,
+      "loss": 0.3897,
+      "step": 9464
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7203975426082199,
+      "learning_rate": 1.6302227590191543e-06,
+      "loss": 0.4667,
+      "step": 9465
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.560517183033107,
+      "learning_rate": 1.6292833001267288e-06,
+      "loss": 0.4751,
+      "step": 9466
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.865711654315479,
+      "learning_rate": 1.6283440593152644e-06,
+      "loss": 0.4865,
+      "step": 9467
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2372077869769353,
+      "learning_rate": 1.627405036645529e-06,
+      "loss": 0.4205,
+      "step": 9468
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7098413964338406,
+      "learning_rate": 1.6264662321782754e-06,
+      "loss": 0.4244,
+      "step": 9469
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.503438159074684,
+      "learning_rate": 1.6255276459742452e-06,
+      "loss": 0.497,
+      "step": 9470
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.606628312608506,
+      "learning_rate": 1.6245892780941601e-06,
+      "loss": 0.3892,
+      "step": 9471
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7582621109431151,
+      "learning_rate": 1.6236511285987333e-06,
+      "loss": 0.479,
+      "step": 9472
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.7475186005544148,
+      "learning_rate": 1.6227131975486616e-06,
+      "loss": 0.4826,
+      "step": 9473
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5311138999144461,
+      "learning_rate": 1.6217754850046297e-06,
+      "loss": 0.4862,
+      "step": 9474
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.850432386374592,
+      "learning_rate": 1.6208379910273014e-06,
+      "loss": 0.4534,
+      "step": 9475
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5669949954976595,
+      "learning_rate": 1.6199007156773378e-06,
+      "loss": 0.4969,
+      "step": 9476
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4869074465526095,
+      "learning_rate": 1.6189636590153746e-06,
+      "loss": 0.4501,
+      "step": 9477
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.491255001316474,
+      "learning_rate": 1.6180268211020412e-06,
+      "loss": 0.4272,
+      "step": 9478
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9957735508183045,
+      "learning_rate": 1.617090201997945e-06,
+      "loss": 0.4353,
+      "step": 9479
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 5.122290335082375,
+      "learning_rate": 1.616153801763689e-06,
+      "loss": 0.4475,
+      "step": 9480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 9.116308705620828,
+      "learning_rate": 1.615217620459857e-06,
+      "loss": 0.4787,
+      "step": 9481
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3082318900085268,
+      "learning_rate": 1.6142816581470156e-06,
+      "loss": 0.4574,
+      "step": 9482
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5697733745669809,
+      "learning_rate": 1.6133459148857216e-06,
+      "loss": 0.4371,
+      "step": 9483
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.0734756647600965,
+      "learning_rate": 1.6124103907365168e-06,
+      "loss": 0.4638,
+      "step": 9484
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9356342589704505,
+      "learning_rate": 1.6114750857599299e-06,
+      "loss": 0.4503,
+      "step": 9485
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.9925063712882152,
+      "learning_rate": 1.6105400000164707e-06,
+      "loss": 0.4001,
+      "step": 9486
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.529947549966917,
+      "learning_rate": 1.609605133566639e-06,
+      "loss": 0.4708,
+      "step": 9487
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6495789914128554,
+      "learning_rate": 1.60867048647092e-06,
+      "loss": 0.4685,
+      "step": 9488
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6722346052075252,
+      "learning_rate": 1.6077360587897867e-06,
+      "loss": 0.4613,
+      "step": 9489
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6700327105747919,
+      "learning_rate": 1.6068018505836901e-06,
+      "loss": 0.4214,
+      "step": 9490
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6803857038403462,
+      "learning_rate": 1.6058678619130735e-06,
+      "loss": 0.4462,
+      "step": 9491
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.682054659968237,
+      "learning_rate": 1.6049340928383694e-06,
+      "loss": 0.4963,
+      "step": 9492
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5366834013419136,
+      "learning_rate": 1.6040005434199869e-06,
+      "loss": 0.4837,
+      "step": 9493
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7070580308806913,
+      "learning_rate": 1.6030672137183283e-06,
+      "loss": 0.4479,
+      "step": 9494
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.917254791485374,
+      "learning_rate": 1.6021341037937739e-06,
+      "loss": 0.4863,
+      "step": 9495
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.808475486717067,
+      "learning_rate": 1.6012012137067013e-06,
+      "loss": 0.4548,
+      "step": 9496
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6753571088876864,
+      "learning_rate": 1.600268543517462e-06,
+      "loss": 0.4593,
+      "step": 9497
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4969076769005229,
+      "learning_rate": 1.5993360932864005e-06,
+      "loss": 0.4849,
+      "step": 9498
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9074840928268515,
+      "learning_rate": 1.5984038630738458e-06,
+      "loss": 0.4605,
+      "step": 9499
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6160669020640468,
+      "learning_rate": 1.5974718529401123e-06,
+      "loss": 0.4421,
+      "step": 9500
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 5.887074553743196,
+      "learning_rate": 1.5965400629454975e-06,
+      "loss": 0.4424,
+      "step": 9501
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.83158035012084,
+      "learning_rate": 1.595608493150288e-06,
+      "loss": 0.4464,
+      "step": 9502
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7320565965830719,
+      "learning_rate": 1.5946771436147561e-06,
+      "loss": 0.4332,
+      "step": 9503
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7855869176629924,
+      "learning_rate": 1.5937460143991579e-06,
+      "loss": 0.4244,
+      "step": 9504
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7249361960453422,
+      "learning_rate": 1.592815105563738e-06,
+      "loss": 0.4519,
+      "step": 9505
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.584797664218502,
+      "learning_rate": 1.5918844171687225e-06,
+      "loss": 0.4496,
+      "step": 9506
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.856571855287507,
+      "learning_rate": 1.590953949274327e-06,
+      "loss": 0.4652,
+      "step": 9507
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8055948117820615,
+      "learning_rate": 1.5900237019407511e-06,
+      "loss": 0.4248,
+      "step": 9508
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.4239989251735885,
+      "learning_rate": 1.5890936752281822e-06,
+      "loss": 0.4466,
+      "step": 9509
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4372588451993604,
+      "learning_rate": 1.5881638691967876e-06,
+      "loss": 0.4347,
+      "step": 9510
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9260101718843334,
+      "learning_rate": 1.5872342839067305e-06,
+      "loss": 0.5115,
+      "step": 9511
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7245122706338096,
+      "learning_rate": 1.586304919418149e-06,
+      "loss": 0.4313,
+      "step": 9512
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7412950487570953,
+      "learning_rate": 1.5853757757911737e-06,
+      "loss": 0.4374,
+      "step": 9513
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.78237475330738,
+      "learning_rate": 1.5844468530859193e-06,
+      "loss": 0.4631,
+      "step": 9514
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.137020579989209,
+      "learning_rate": 1.5835181513624848e-06,
+      "loss": 0.4487,
+      "step": 9515
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.6545369543014457,
+      "learning_rate": 1.5825896706809579e-06,
+      "loss": 0.4241,
+      "step": 9516
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8032011973221418,
+      "learning_rate": 1.5816614111014078e-06,
+      "loss": 0.4716,
+      "step": 9517
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5543070126425631,
+      "learning_rate": 1.5807333726838924e-06,
+      "loss": 0.4677,
+      "step": 9518
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5062740220644393,
+      "learning_rate": 1.5798055554884551e-06,
+      "loss": 0.4475,
+      "step": 9519
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.3760098652872506,
+      "learning_rate": 1.5788779595751252e-06,
+      "loss": 0.4429,
+      "step": 9520
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9306794140545294,
+      "learning_rate": 1.5779505850039152e-06,
+      "loss": 0.4637,
+      "step": 9521
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5703665620093,
+      "learning_rate": 1.577023431834825e-06,
+      "loss": 0.4788,
+      "step": 9522
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1967489297370695,
+      "learning_rate": 1.576096500127841e-06,
+      "loss": 0.4769,
+      "step": 9523
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7063140774317545,
+      "learning_rate": 1.5751697899429346e-06,
+      "loss": 0.4176,
+      "step": 9524
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5706641851239264,
+      "learning_rate": 1.5742433013400644e-06,
+      "loss": 0.4747,
+      "step": 9525
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8131725871098845,
+      "learning_rate": 1.5733170343791675e-06,
+      "loss": 0.4647,
+      "step": 9526
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5558082167399254,
+      "learning_rate": 1.5723909891201794e-06,
+      "loss": 0.3696,
+      "step": 9527
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4513936995032763,
+      "learning_rate": 1.5714651656230085e-06,
+      "loss": 0.3986,
+      "step": 9528
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8326119490423778,
+      "learning_rate": 1.5705395639475578e-06,
+      "loss": 0.4423,
+      "step": 9529
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.196816284472686,
+      "learning_rate": 1.5696141841537083e-06,
+      "loss": 0.4274,
+      "step": 9530
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7042805841630007,
+      "learning_rate": 1.5686890263013365e-06,
+      "loss": 0.4541,
+      "step": 9531
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8523043324055197,
+      "learning_rate": 1.5677640904502944e-06,
+      "loss": 0.4767,
+      "step": 9532
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.540166566621854,
+      "learning_rate": 1.5668393766604255e-06,
+      "loss": 0.4581,
+      "step": 9533
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3155390370897255,
+      "learning_rate": 1.5659148849915585e-06,
+      "loss": 0.3972,
+      "step": 9534
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.594892610223923,
+      "learning_rate": 1.564990615503507e-06,
+      "loss": 0.4966,
+      "step": 9535
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.750396699990423,
+      "learning_rate": 1.5640665682560678e-06,
+      "loss": 0.4343,
+      "step": 9536
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8670077921769563,
+      "learning_rate": 1.563142743309027e-06,
+      "loss": 0.4287,
+      "step": 9537
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5857018292928747,
+      "learning_rate": 1.5622191407221543e-06,
+      "loss": 0.4792,
+      "step": 9538
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.517864936443574,
+      "learning_rate": 1.5612957605552064e-06,
+      "loss": 0.442,
+      "step": 9539
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.690461840207118,
+      "learning_rate": 1.5603726028679255e-06,
+      "loss": 0.3945,
+      "step": 9540
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.520239876249671,
+      "learning_rate": 1.5594496677200366e-06,
+      "loss": 0.4275,
+      "step": 9541
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7890290913333489,
+      "learning_rate": 1.558526955171253e-06,
+      "loss": 0.4533,
+      "step": 9542
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.323409867952726,
+      "learning_rate": 1.557604465281274e-06,
+      "loss": 0.4962,
+      "step": 9543
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6111690265022294,
+      "learning_rate": 1.5566821981097836e-06,
+      "loss": 0.42,
+      "step": 9544
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.719315966510045,
+      "learning_rate": 1.5557601537164497e-06,
+      "loss": 0.4232,
+      "step": 9545
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.8168772331157212,
+      "learning_rate": 1.5548383321609272e-06,
+      "loss": 0.454,
+      "step": 9546
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 4.098846189225212,
+      "learning_rate": 1.5539167335028588e-06,
+      "loss": 0.492,
+      "step": 9547
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5404712007206263,
+      "learning_rate": 1.5529953578018691e-06,
+      "loss": 0.4703,
+      "step": 9548
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5318636608855462,
+      "learning_rate": 1.5520742051175708e-06,
+      "loss": 0.4772,
+      "step": 9549
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6133056320735004,
+      "learning_rate": 1.551153275509561e-06,
+      "loss": 0.4226,
+      "step": 9550
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5586398656457083,
+      "learning_rate": 1.5502325690374243e-06,
+      "loss": 0.4722,
+      "step": 9551
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7843552594030445,
+      "learning_rate": 1.5493120857607258e-06,
+      "loss": 0.4966,
+      "step": 9552
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3660796404631763,
+      "learning_rate": 1.5483918257390212e-06,
+      "loss": 0.4684,
+      "step": 9553
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4073057865492087,
+      "learning_rate": 1.5474717890318502e-06,
+      "loss": 0.4053,
+      "step": 9554
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5682954647072034,
+      "learning_rate": 1.5465519756987396e-06,
+      "loss": 0.4983,
+      "step": 9555
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.5165934129897196,
+      "learning_rate": 1.5456323857991967e-06,
+      "loss": 0.4705,
+      "step": 9556
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.702757020919363,
+      "learning_rate": 1.5447130193927202e-06,
+      "loss": 0.4301,
+      "step": 9557
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.690970463130095,
+      "learning_rate": 1.5437938765387906e-06,
+      "loss": 0.3771,
+      "step": 9558
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.878326313726984,
+      "learning_rate": 1.542874957296876e-06,
+      "loss": 0.46,
+      "step": 9559
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5634339958245722,
+      "learning_rate": 1.5419562617264312e-06,
+      "loss": 0.4798,
+      "step": 9560
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4606275065512824,
+      "learning_rate": 1.541037789886889e-06,
+      "loss": 0.4629,
+      "step": 9561
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7948825489499456,
+      "learning_rate": 1.5401195418376801e-06,
+      "loss": 0.3995,
+      "step": 9562
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8127300625938654,
+      "learning_rate": 1.5392015176382092e-06,
+      "loss": 0.476,
+      "step": 9563
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1496492507136784,
+      "learning_rate": 1.5382837173478748e-06,
+      "loss": 0.4718,
+      "step": 9564
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5008139438889498,
+      "learning_rate": 1.5373661410260515e-06,
+      "loss": 0.3987,
+      "step": 9565
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.767988731802174,
+      "learning_rate": 1.5364487887321128e-06,
+      "loss": 0.533,
+      "step": 9566
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.565293624225528,
+      "learning_rate": 1.535531660525405e-06,
+      "loss": 0.4886,
+      "step": 9567
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.180293049041982,
+      "learning_rate": 1.5346147564652664e-06,
+      "loss": 0.4694,
+      "step": 9568
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9830831666312374,
+      "learning_rate": 1.53369807661102e-06,
+      "loss": 0.4644,
+      "step": 9569
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5465052335030506,
+      "learning_rate": 1.5327816210219736e-06,
+      "loss": 0.4645,
+      "step": 9570
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.494685456068443,
+      "learning_rate": 1.531865389757422e-06,
+      "loss": 0.471,
+      "step": 9571
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5959141978547964,
+      "learning_rate": 1.5309493828766414e-06,
+      "loss": 0.4109,
+      "step": 9572
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.700318916623548,
+      "learning_rate": 1.5300336004388976e-06,
+      "loss": 0.4169,
+      "step": 9573
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6458351947150365,
+      "learning_rate": 1.5291180425034403e-06,
+      "loss": 0.4473,
+      "step": 9574
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5505808226334474,
+      "learning_rate": 1.5282027091295071e-06,
+      "loss": 0.4915,
+      "step": 9575
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6946074093814796,
+      "learning_rate": 1.5272876003763154e-06,
+      "loss": 0.4233,
+      "step": 9576
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.63728505550244,
+      "learning_rate": 1.5263727163030729e-06,
+      "loss": 0.4609,
+      "step": 9577
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5607493031343314,
+      "learning_rate": 1.5254580569689713e-06,
+      "loss": 0.4632,
+      "step": 9578
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.698320408923765,
+      "learning_rate": 1.52454362243319e-06,
+      "loss": 0.417,
+      "step": 9579
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1133433949544456,
+      "learning_rate": 1.5236294127548883e-06,
+      "loss": 0.4674,
+      "step": 9580
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4136501817371303,
+      "learning_rate": 1.5227154279932143e-06,
+      "loss": 0.485,
+      "step": 9581
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.4969586416456506,
+      "learning_rate": 1.5218016682073068e-06,
+      "loss": 0.4863,
+      "step": 9582
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4290515312334118,
+      "learning_rate": 1.5208881334562792e-06,
+      "loss": 0.5185,
+      "step": 9583
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.081993900914485,
+      "learning_rate": 1.5199748237992384e-06,
+      "loss": 0.4862,
+      "step": 9584
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7437198546062906,
+      "learning_rate": 1.519061739295274e-06,
+      "loss": 0.4606,
+      "step": 9585
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5912700496356162,
+      "learning_rate": 1.5181488800034627e-06,
+      "loss": 0.3727,
+      "step": 9586
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7711274858532204,
+      "learning_rate": 1.5172362459828627e-06,
+      "loss": 0.4261,
+      "step": 9587
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4229496297205668,
+      "learning_rate": 1.5163238372925215e-06,
+      "loss": 0.4726,
+      "step": 9588
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.83169196787833,
+      "learning_rate": 1.5154116539914709e-06,
+      "loss": 0.3916,
+      "step": 9589
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4578129682662675,
+      "learning_rate": 1.5144996961387292e-06,
+      "loss": 0.4679,
+      "step": 9590
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5427792402671615,
+      "learning_rate": 1.5135879637932966e-06,
+      "loss": 0.4881,
+      "step": 9591
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.693736627150592,
+      "learning_rate": 1.5126764570141617e-06,
+      "loss": 0.4274,
+      "step": 9592
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4386507718320383,
+      "learning_rate": 1.5117651758602975e-06,
+      "loss": 0.4608,
+      "step": 9593
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5710268488838196,
+      "learning_rate": 1.5108541203906635e-06,
+      "loss": 0.4796,
+      "step": 9594
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.1837446338123003,
+      "learning_rate": 1.5099432906642054e-06,
+      "loss": 0.3989,
+      "step": 9595
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.3467324347962193,
+      "learning_rate": 1.5090326867398476e-06,
+      "loss": 0.4476,
+      "step": 9596
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9092909157617979,
+      "learning_rate": 1.5081223086765113e-06,
+      "loss": 0.4181,
+      "step": 9597
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.6319775840444712,
+      "learning_rate": 1.5072121565330916e-06,
+      "loss": 0.3864,
+      "step": 9598
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9583816385388566,
+      "learning_rate": 1.5063022303684787e-06,
+      "loss": 0.4325,
+      "step": 9599
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 3.5529603790001194,
+      "learning_rate": 1.5053925302415374e-06,
+      "loss": 0.4394,
+      "step": 9600
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4241026037242772,
+      "learning_rate": 1.5044830562111311e-06,
+      "loss": 0.4198,
+      "step": 9601
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.8674276094398297,
+      "learning_rate": 1.5035738083360967e-06,
+      "loss": 0.5028,
+      "step": 9602
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7788418506101502,
+      "learning_rate": 1.5026647866752635e-06,
+      "loss": 0.3988,
+      "step": 9603
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5856901186244446,
+      "learning_rate": 1.501755991287443e-06,
+      "loss": 0.4846,
+      "step": 9604
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.9780219845521128,
+      "learning_rate": 1.500847422231434e-06,
+      "loss": 0.4242,
+      "step": 9605
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9590526702794837,
+      "learning_rate": 1.4999390795660207e-06,
+      "loss": 0.4731,
+      "step": 9606
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4575906782303634,
+      "learning_rate": 1.4990309633499683e-06,
+      "loss": 0.4557,
+      "step": 9607
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.658650490351291,
+      "learning_rate": 1.4981230736420332e-06,
+      "loss": 0.4157,
+      "step": 9608
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.635053793350725,
+      "learning_rate": 1.4972154105009546e-06,
+      "loss": 0.403,
+      "step": 9609
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7937839923334173,
+      "learning_rate": 1.4963079739854574e-06,
+      "loss": 0.4697,
+      "step": 9610
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.007751674592737,
+      "learning_rate": 1.4954007641542501e-06,
+      "loss": 0.4483,
+      "step": 9611
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.9161990258725385,
+      "learning_rate": 1.4944937810660282e-06,
+      "loss": 0.4666,
+      "step": 9612
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 2.2386863831883805,
+      "learning_rate": 1.493587024779473e-06,
+      "loss": 0.457,
+      "step": 9613
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5520118355003508,
+      "learning_rate": 1.492680495353252e-06,
+      "loss": 0.5036,
+      "step": 9614
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.1244214042995644,
+      "learning_rate": 1.4917741928460134e-06,
+      "loss": 0.433,
+      "step": 9615
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4019502683591856,
+      "learning_rate": 1.4908681173163931e-06,
+      "loss": 0.4509,
+      "step": 9616
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9891680347264522,
+      "learning_rate": 1.4899622688230186e-06,
+      "loss": 0.4819,
+      "step": 9617
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.159763312626244,
+      "learning_rate": 1.4890566474244923e-06,
+      "loss": 0.4488,
+      "step": 9618
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9339102527344705,
+      "learning_rate": 1.4881512531794074e-06,
+      "loss": 0.4697,
+      "step": 9619
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6953351534113665,
+      "learning_rate": 1.4872460861463428e-06,
+      "loss": 0.4052,
+      "step": 9620
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5247652228188053,
+      "learning_rate": 1.4863411463838622e-06,
+      "loss": 0.45,
+      "step": 9621
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5414550380181892,
+      "learning_rate": 1.4854364339505117e-06,
+      "loss": 0.4909,
+      "step": 9622
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5467590428689633,
+      "learning_rate": 1.4845319489048266e-06,
+      "loss": 0.4001,
+      "step": 9623
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3931780062135481,
+      "learning_rate": 1.4836276913053256e-06,
+      "loss": 0.4359,
+      "step": 9624
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.416778010420891,
+      "learning_rate": 1.482723661210514e-06,
+      "loss": 0.3635,
+      "step": 9625
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5649439013472531,
+      "learning_rate": 1.4818198586788795e-06,
+      "loss": 0.495,
+      "step": 9626
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6238112884185232,
+      "learning_rate": 1.4809162837688973e-06,
+      "loss": 0.4046,
+      "step": 9627
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.107925704751721,
+      "learning_rate": 1.4800129365390282e-06,
+      "loss": 0.4363,
+      "step": 9628
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.321423205510961,
+      "learning_rate": 1.4791098170477168e-06,
+      "loss": 0.4585,
+      "step": 9629
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.7286118436204627,
+      "learning_rate": 1.4782069253533965e-06,
+      "loss": 0.436,
+      "step": 9630
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8518329193865894,
+      "learning_rate": 1.4773042615144778e-06,
+      "loss": 0.392,
+      "step": 9631
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5473098367463618,
+      "learning_rate": 1.4764018255893686e-06,
+      "loss": 0.4798,
+      "step": 9632
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6071501785728153,
+      "learning_rate": 1.4754996176364494e-06,
+      "loss": 0.45,
+      "step": 9633
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4826179478004424,
+      "learning_rate": 1.4745976377140969e-06,
+      "loss": 0.4529,
+      "step": 9634
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7510696571586861,
+      "learning_rate": 1.4736958858806616e-06,
+      "loss": 0.4957,
+      "step": 9635
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.679493740338074,
+      "learning_rate": 1.472794362194493e-06,
+      "loss": 0.3896,
+      "step": 9636
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0704481842740536,
+      "learning_rate": 1.4718930667139131e-06,
+      "loss": 0.4859,
+      "step": 9637
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.13400537435942,
+      "learning_rate": 1.4709919994972366e-06,
+      "loss": 0.4419,
+      "step": 9638
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.828773668708369,
+      "learning_rate": 1.4700911606027612e-06,
+      "loss": 0.4428,
+      "step": 9639
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6926569199841242,
+      "learning_rate": 1.46919055008877e-06,
+      "loss": 0.478,
+      "step": 9640
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5257171656459595,
+      "learning_rate": 1.4682901680135332e-06,
+      "loss": 0.4702,
+      "step": 9641
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9173447501888918,
+      "learning_rate": 1.4673900144353003e-06,
+      "loss": 0.4863,
+      "step": 9642
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6745972367466786,
+      "learning_rate": 1.4664900894123123e-06,
+      "loss": 0.4137,
+      "step": 9643
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9858876001897323,
+      "learning_rate": 1.4655903930027937e-06,
+      "loss": 0.4207,
+      "step": 9644
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8356990604039494,
+      "learning_rate": 1.4646909252649544e-06,
+      "loss": 0.4483,
+      "step": 9645
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7294439850184085,
+      "learning_rate": 1.4637916862569855e-06,
+      "loss": 0.4204,
+      "step": 9646
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3642459233212287,
+      "learning_rate": 1.4628926760370688e-06,
+      "loss": 0.425,
+      "step": 9647
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 12.79114468182307,
+      "learning_rate": 1.461993894663369e-06,
+      "loss": 0.4828,
+      "step": 9648
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.885191909398489,
+      "learning_rate": 1.461095342194036e-06,
+      "loss": 0.479,
+      "step": 9649
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5446385798199103,
+      "learning_rate": 1.460197018687206e-06,
+      "loss": 0.4573,
+      "step": 9650
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4947546523808137,
+      "learning_rate": 1.4592989242009953e-06,
+      "loss": 0.4321,
+      "step": 9651
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0610985084441262,
+      "learning_rate": 1.458401058793516e-06,
+      "loss": 0.4104,
+      "step": 9652
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8461076201673023,
+      "learning_rate": 1.4575034225228528e-06,
+      "loss": 0.4946,
+      "step": 9653
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7472979095764258,
+      "learning_rate": 1.4566060154470846e-06,
+      "loss": 0.4457,
+      "step": 9654
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.242542189147063,
+      "learning_rate": 1.4557088376242718e-06,
+      "loss": 0.5229,
+      "step": 9655
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6033672748726733,
+      "learning_rate": 1.4548118891124623e-06,
+      "loss": 0.4779,
+      "step": 9656
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7161389469420267,
+      "learning_rate": 1.453915169969684e-06,
+      "loss": 0.408,
+      "step": 9657
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4131911299803783,
+      "learning_rate": 1.4530186802539558e-06,
+      "loss": 0.4149,
+      "step": 9658
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.83104449805615,
+      "learning_rate": 1.4521224200232786e-06,
+      "loss": 0.4386,
+      "step": 9659
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9539439458959489,
+      "learning_rate": 1.4512263893356392e-06,
+      "loss": 0.4481,
+      "step": 9660
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.296832807183728,
+      "learning_rate": 1.4503305882490126e-06,
+      "loss": 0.4384,
+      "step": 9661
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.542830261491004,
+      "learning_rate": 1.4494350168213511e-06,
+      "loss": 0.4623,
+      "step": 9662
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8351525186761743,
+      "learning_rate": 1.4485396751105996e-06,
+      "loss": 0.418,
+      "step": 9663
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.2847337403177157,
+      "learning_rate": 1.447644563174685e-06,
+      "loss": 0.4848,
+      "step": 9664
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.534628725416741,
+      "learning_rate": 1.446749681071522e-06,
+      "loss": 0.4817,
+      "step": 9665
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6961743451282096,
+      "learning_rate": 1.445855028859003e-06,
+      "loss": 0.452,
+      "step": 9666
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5279799934124707,
+      "learning_rate": 1.4449606065950173e-06,
+      "loss": 0.4812,
+      "step": 9667
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6628108141104814,
+      "learning_rate": 1.4440664143374283e-06,
+      "loss": 0.4062,
+      "step": 9668
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9440863062281821,
+      "learning_rate": 1.443172452144092e-06,
+      "loss": 0.4815,
+      "step": 9669
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5487369552085108,
+      "learning_rate": 1.4422787200728421e-06,
+      "loss": 0.4611,
+      "step": 9670
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.635404327655508,
+      "learning_rate": 1.441385218181507e-06,
+      "loss": 0.4288,
+      "step": 9671
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5539901173496475,
+      "learning_rate": 1.4404919465278938e-06,
+      "loss": 0.434,
+      "step": 9672
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7159622281753129,
+      "learning_rate": 1.439598905169794e-06,
+      "loss": 0.5101,
+      "step": 9673
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5960887156703283,
+      "learning_rate": 1.4387060941649878e-06,
+      "loss": 0.4447,
+      "step": 9674
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7102478509328658,
+      "learning_rate": 1.437813513571238e-06,
+      "loss": 0.4417,
+      "step": 9675
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.638983654801771,
+      "learning_rate": 1.4369211634462955e-06,
+      "loss": 0.4657,
+      "step": 9676
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7982156288296596,
+      "learning_rate": 1.4360290438478913e-06,
+      "loss": 0.3686,
+      "step": 9677
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5548789151374398,
+      "learning_rate": 1.4351371548337457e-06,
+      "loss": 0.408,
+      "step": 9678
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 5.045654916234841,
+      "learning_rate": 1.4342454964615627e-06,
+      "loss": 0.4873,
+      "step": 9679
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5091384744179397,
+      "learning_rate": 1.433354068789033e-06,
+      "loss": 0.3846,
+      "step": 9680
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0331724056136684,
+      "learning_rate": 1.4324628718738281e-06,
+      "loss": 0.4733,
+      "step": 9681
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8168027159557691,
+      "learning_rate": 1.4315719057736067e-06,
+      "loss": 0.3781,
+      "step": 9682
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8276691940837746,
+      "learning_rate": 1.4306811705460178e-06,
+      "loss": 0.4353,
+      "step": 9683
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7170013607744674,
+      "learning_rate": 1.4297906662486866e-06,
+      "loss": 0.4615,
+      "step": 9684
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.521713131086413,
+      "learning_rate": 1.4289003929392303e-06,
+      "loss": 0.4631,
+      "step": 9685
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.055493784451531,
+      "learning_rate": 1.4280103506752434e-06,
+      "loss": 0.4245,
+      "step": 9686
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.836280143244017,
+      "learning_rate": 1.427120539514318e-06,
+      "loss": 0.427,
+      "step": 9687
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.478532823067747,
+      "learning_rate": 1.4262309595140179e-06,
+      "loss": 0.4115,
+      "step": 9688
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.312490839099188,
+      "learning_rate": 1.4253416107318997e-06,
+      "loss": 0.4503,
+      "step": 9689
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3954164426067273,
+      "learning_rate": 1.4244524932255026e-06,
+      "loss": 0.4376,
+      "step": 9690
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.383651057757473,
+      "learning_rate": 1.4235636070523539e-06,
+      "loss": 0.4102,
+      "step": 9691
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7359183164421632,
+      "learning_rate": 1.4226749522699595e-06,
+      "loss": 0.4532,
+      "step": 9692
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.576892797630708,
+      "learning_rate": 1.4217865289358163e-06,
+      "loss": 0.4861,
+      "step": 9693
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9919035029096992,
+      "learning_rate": 1.4208983371074032e-06,
+      "loss": 0.4788,
+      "step": 9694
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5189714383760866,
+      "learning_rate": 1.4200103768421857e-06,
+      "loss": 0.4827,
+      "step": 9695
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8403890793835973,
+      "learning_rate": 1.4191226481976156e-06,
+      "loss": 0.4472,
+      "step": 9696
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6339010124317952,
+      "learning_rate": 1.4182351512311237e-06,
+      "loss": 0.434,
+      "step": 9697
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.086586402893464,
+      "learning_rate": 1.4173478860001328e-06,
+      "loss": 0.5368,
+      "step": 9698
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7552769623658644,
+      "learning_rate": 1.416460852562046e-06,
+      "loss": 0.5019,
+      "step": 9699
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.704840294793175,
+      "learning_rate": 1.4155740509742566e-06,
+      "loss": 0.4632,
+      "step": 9700
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4441749072943602,
+      "learning_rate": 1.4146874812941337e-06,
+      "loss": 0.4812,
+      "step": 9701
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6511908555842614,
+      "learning_rate": 1.4138011435790433e-06,
+      "loss": 0.4349,
+      "step": 9702
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5422171133102842,
+      "learning_rate": 1.412915037886326e-06,
+      "loss": 0.4836,
+      "step": 9703
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8992118184942701,
+      "learning_rate": 1.4120291642733152e-06,
+      "loss": 0.42,
+      "step": 9704
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5200262967941722,
+      "learning_rate": 1.4111435227973208e-06,
+      "loss": 0.4373,
+      "step": 9705
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3503431631095235,
+      "learning_rate": 1.4102581135156468e-06,
+      "loss": 0.3944,
+      "step": 9706
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6109928393556285,
+      "learning_rate": 1.4093729364855785e-06,
+      "loss": 0.4584,
+      "step": 9707
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.760680695785795,
+      "learning_rate": 1.408487991764383e-06,
+      "loss": 0.4235,
+      "step": 9708
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4555839654624207,
+      "learning_rate": 1.4076032794093158e-06,
+      "loss": 0.4189,
+      "step": 9709
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7517904030458151,
+      "learning_rate": 1.4067187994776166e-06,
+      "loss": 0.4695,
+      "step": 9710
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9138928488015565,
+      "learning_rate": 1.4058345520265127e-06,
+      "loss": 0.4391,
+      "step": 9711
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9197666283853205,
+      "learning_rate": 1.4049505371132095e-06,
+      "loss": 0.4925,
+      "step": 9712
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0472638830244523,
+      "learning_rate": 1.404066754794904e-06,
+      "loss": 0.4631,
+      "step": 9713
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 7.207484982972802,
+      "learning_rate": 1.4031832051287752e-06,
+      "loss": 0.4425,
+      "step": 9714
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7066049891329393,
+      "learning_rate": 1.4022998881719891e-06,
+      "loss": 0.458,
+      "step": 9715
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6129827361634426,
+      "learning_rate": 1.4014168039816929e-06,
+      "loss": 0.3964,
+      "step": 9716
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.110268117028117,
+      "learning_rate": 1.4005339526150197e-06,
+      "loss": 0.4511,
+      "step": 9717
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0035781261139505,
+      "learning_rate": 1.3996513341290946e-06,
+      "loss": 0.4731,
+      "step": 9718
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5289342487654921,
+      "learning_rate": 1.3987689485810168e-06,
+      "loss": 0.4722,
+      "step": 9719
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5322204423089738,
+      "learning_rate": 1.3978867960278786e-06,
+      "loss": 0.3959,
+      "step": 9720
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4515473684119469,
+      "learning_rate": 1.397004876526749e-06,
+      "loss": 0.4211,
+      "step": 9721
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8936942687901868,
+      "learning_rate": 1.3961231901346938e-06,
+      "loss": 0.4474,
+      "step": 9722
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 10.613215466898822,
+      "learning_rate": 1.3952417369087528e-06,
+      "loss": 0.4439,
+      "step": 9723
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 3.3998774352534946,
+      "learning_rate": 1.3943605169059554e-06,
+      "loss": 0.4603,
+      "step": 9724
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.622744009927646,
+      "learning_rate": 1.3934795301833154e-06,
+      "loss": 0.4416,
+      "step": 9725
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3747065419814466,
+      "learning_rate": 1.392598776797834e-06,
+      "loss": 0.4738,
+      "step": 9726
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8742286718314007,
+      "learning_rate": 1.3917182568064907e-06,
+      "loss": 0.4832,
+      "step": 9727
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5305018360697198,
+      "learning_rate": 1.390837970266256e-06,
+      "loss": 0.4251,
+      "step": 9728
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.694081055401206,
+      "learning_rate": 1.3899579172340839e-06,
+      "loss": 0.4881,
+      "step": 9729
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4563023400470618,
+      "learning_rate": 1.3890780977669117e-06,
+      "loss": 0.447,
+      "step": 9730
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5917020574858896,
+      "learning_rate": 1.388198511921664e-06,
+      "loss": 0.5025,
+      "step": 9731
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6261431732513043,
+      "learning_rate": 1.3873191597552465e-06,
+      "loss": 0.4911,
+      "step": 9732
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.878397864435762,
+      "learning_rate": 1.3864400413245532e-06,
+      "loss": 0.4863,
+      "step": 9733
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7808028072086688,
+      "learning_rate": 1.3855611566864624e-06,
+      "loss": 0.4719,
+      "step": 9734
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6844207196460712,
+      "learning_rate": 1.384682505897838e-06,
+      "loss": 0.4208,
+      "step": 9735
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.5627687040355727,
+      "learning_rate": 1.383804089015523e-06,
+      "loss": 0.4618,
+      "step": 9736
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6231723803329783,
+      "learning_rate": 1.3829259060963556e-06,
+      "loss": 0.4537,
+      "step": 9737
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5309707733894793,
+      "learning_rate": 1.3820479571971491e-06,
+      "loss": 0.4149,
+      "step": 9738
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.0083088880484525,
+      "learning_rate": 1.3811702423747065e-06,
+      "loss": 0.4542,
+      "step": 9739
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.424503448077591,
+      "learning_rate": 1.3802927616858157e-06,
+      "loss": 0.4607,
+      "step": 9740
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6868006566997644,
+      "learning_rate": 1.379415515187248e-06,
+      "loss": 0.5075,
+      "step": 9741
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.148107458961988,
+      "learning_rate": 1.3785385029357613e-06,
+      "loss": 0.4764,
+      "step": 9742
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6220890563820314,
+      "learning_rate": 1.3776617249880947e-06,
+      "loss": 0.4464,
+      "step": 9743
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5943503955041192,
+      "learning_rate": 1.3767851814009759e-06,
+      "loss": 0.4558,
+      "step": 9744
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.637870622727954,
+      "learning_rate": 1.375908872231116e-06,
+      "loss": 0.4537,
+      "step": 9745
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9365124975820953,
+      "learning_rate": 1.3750327975352128e-06,
+      "loss": 0.4461,
+      "step": 9746
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.814167621035966,
+      "learning_rate": 1.3741569573699437e-06,
+      "loss": 0.4428,
+      "step": 9747
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.437270261980394,
+      "learning_rate": 1.373281351791977e-06,
+      "loss": 0.4432,
+      "step": 9748
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.654862592197974,
+      "learning_rate": 1.3724059808579614e-06,
+      "loss": 0.4031,
+      "step": 9749
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.2013660152049366,
+      "learning_rate": 1.3715308446245341e-06,
+      "loss": 0.4495,
+      "step": 9750
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4853119548887193,
+      "learning_rate": 1.3706559431483163e-06,
+      "loss": 0.4481,
+      "step": 9751
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7335889532423319,
+      "learning_rate": 1.3697812764859075e-06,
+      "loss": 0.4155,
+      "step": 9752
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5775261972411954,
+      "learning_rate": 1.3689068446939046e-06,
+      "loss": 0.4711,
+      "step": 9753
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7033978075172276,
+      "learning_rate": 1.3680326478288768e-06,
+      "loss": 0.4691,
+      "step": 9754
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5869143012469478,
+      "learning_rate": 1.3671586859473879e-06,
+      "loss": 0.4545,
+      "step": 9755
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5480714256511907,
+      "learning_rate": 1.3662849591059756e-06,
+      "loss": 0.402,
+      "step": 9756
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.789185545663958,
+      "learning_rate": 1.3654114673611768e-06,
+      "loss": 0.3675,
+      "step": 9757
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.317689808810311,
+      "learning_rate": 1.3645382107694993e-06,
+      "loss": 0.3944,
+      "step": 9758
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4993788156783316,
+      "learning_rate": 1.363665189387444e-06,
+      "loss": 0.459,
+      "step": 9759
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4891979173073304,
+      "learning_rate": 1.362792403271494e-06,
+      "loss": 0.474,
+      "step": 9760
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8247490946867762,
+      "learning_rate": 1.361919852478119e-06,
+      "loss": 0.4309,
+      "step": 9761
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1624587517942544,
+      "learning_rate": 1.3610475370637694e-06,
+      "loss": 0.4307,
+      "step": 9762
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.967821326849219,
+      "learning_rate": 1.360175457084883e-06,
+      "loss": 0.3983,
+      "step": 9763
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4049153607276919,
+      "learning_rate": 1.359303612597883e-06,
+      "loss": 0.4611,
+      "step": 9764
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5869511785434793,
+      "learning_rate": 1.3584320036591769e-06,
+      "loss": 0.4449,
+      "step": 9765
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4878839656524625,
+      "learning_rate": 1.3575606303251582e-06,
+      "loss": 0.4765,
+      "step": 9766
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1391865289760585,
+      "learning_rate": 1.3566894926522e-06,
+      "loss": 0.4418,
+      "step": 9767
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 4.79582789396429,
+      "learning_rate": 1.3558185906966659e-06,
+      "loss": 0.4568,
+      "step": 9768
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5499720430556112,
+      "learning_rate": 1.354947924514901e-06,
+      "loss": 0.4457,
+      "step": 9769
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4413785096476557,
+      "learning_rate": 1.3540774941632395e-06,
+      "loss": 0.429,
+      "step": 9770
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.3919295298247376,
+      "learning_rate": 1.353207299697991e-06,
+      "loss": 0.4387,
+      "step": 9771
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5018319651659298,
+      "learning_rate": 1.352337341175463e-06,
+      "loss": 0.4359,
+      "step": 9772
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5264492331350257,
+      "learning_rate": 1.3514676186519354e-06,
+      "loss": 0.4789,
+      "step": 9773
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.327658411333996,
+      "learning_rate": 1.3505981321836804e-06,
+      "loss": 0.4332,
+      "step": 9774
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7856702861849734,
+      "learning_rate": 1.3497288818269515e-06,
+      "loss": 0.4272,
+      "step": 9775
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7233789048238943,
+      "learning_rate": 1.34885986763799e-06,
+      "loss": 0.4309,
+      "step": 9776
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6847997415998228,
+      "learning_rate": 1.3479910896730192e-06,
+      "loss": 0.4369,
+      "step": 9777
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7818468852197134,
+      "learning_rate": 1.3471225479882466e-06,
+      "loss": 0.4525,
+      "step": 9778
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5568295488108892,
+      "learning_rate": 1.3462542426398662e-06,
+      "loss": 0.4826,
+      "step": 9779
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5732143996625892,
+      "learning_rate": 1.3453861736840568e-06,
+      "loss": 0.4671,
+      "step": 9780
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5706864387830022,
+      "learning_rate": 1.3445183411769824e-06,
+      "loss": 0.4885,
+      "step": 9781
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.815658911544847,
+      "learning_rate": 1.3436507451747883e-06,
+      "loss": 0.4788,
+      "step": 9782
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6617200736913993,
+      "learning_rate": 1.3427833857336075e-06,
+      "loss": 0.4253,
+      "step": 9783
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4782593291817319,
+      "learning_rate": 1.3419162629095572e-06,
+      "loss": 0.4129,
+      "step": 9784
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8616842953290473,
+      "learning_rate": 1.34104937675874e-06,
+      "loss": 0.5035,
+      "step": 9785
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0308596685482283,
+      "learning_rate": 1.3401827273372426e-06,
+      "loss": 0.4239,
+      "step": 9786
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.62966683825278,
+      "learning_rate": 1.3393163147011323e-06,
+      "loss": 0.4814,
+      "step": 9787
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5726175159165826,
+      "learning_rate": 1.3384501389064709e-06,
+      "loss": 0.4621,
+      "step": 9788
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 11.473701068945438,
+      "learning_rate": 1.3375842000092936e-06,
+      "loss": 0.4214,
+      "step": 9789
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7708789622124228,
+      "learning_rate": 1.336718498065629e-06,
+      "loss": 0.4704,
+      "step": 9790
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7663202022689908,
+      "learning_rate": 1.3358530331314828e-06,
+      "loss": 0.4112,
+      "step": 9791
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5463211867620723,
+      "learning_rate": 1.3349878052628545e-06,
+      "loss": 0.4477,
+      "step": 9792
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5831405557254699,
+      "learning_rate": 1.3341228145157198e-06,
+      "loss": 0.5089,
+      "step": 9793
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6562798189496315,
+      "learning_rate": 1.3332580609460432e-06,
+      "loss": 0.4121,
+      "step": 9794
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.848636414408402,
+      "learning_rate": 1.3323935446097735e-06,
+      "loss": 0.4274,
+      "step": 9795
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5654602129231828,
+      "learning_rate": 1.3315292655628437e-06,
+      "loss": 0.4746,
+      "step": 9796
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5509125235374649,
+      "learning_rate": 1.3306652238611729e-06,
+      "loss": 0.456,
+      "step": 9797
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6180634667119023,
+      "learning_rate": 1.3298014195606601e-06,
+      "loss": 0.4136,
+      "step": 9798
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.683392706441607,
+      "learning_rate": 1.3289378527171948e-06,
+      "loss": 0.4956,
+      "step": 9799
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0790657076862082,
+      "learning_rate": 1.3280745233866477e-06,
+      "loss": 0.3967,
+      "step": 9800
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.349019600835016,
+      "learning_rate": 1.3272114316248774e-06,
+      "loss": 0.4709,
+      "step": 9801
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.6468145669126755,
+      "learning_rate": 1.326348577487721e-06,
+      "loss": 0.4098,
+      "step": 9802
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7165961360368753,
+      "learning_rate": 1.3254859610310055e-06,
+      "loss": 0.4703,
+      "step": 9803
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.425229221520358,
+      "learning_rate": 1.3246235823105424e-06,
+      "loss": 0.4426,
+      "step": 9804
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4234368658967118,
+      "learning_rate": 1.3237614413821266e-06,
+      "loss": 0.4654,
+      "step": 9805
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7920914462520112,
+      "learning_rate": 1.322899538301533e-06,
+      "loss": 0.4747,
+      "step": 9806
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.843531405480349,
+      "learning_rate": 1.3220378731245308e-06,
+      "loss": 0.4622,
+      "step": 9807
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5824400175471236,
+      "learning_rate": 1.3211764459068688e-06,
+      "loss": 0.4787,
+      "step": 9808
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6375655512626193,
+      "learning_rate": 1.3203152567042759e-06,
+      "loss": 0.4738,
+      "step": 9809
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9648093470506545,
+      "learning_rate": 1.319454305572473e-06,
+      "loss": 0.4488,
+      "step": 9810
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.541639658549343,
+      "learning_rate": 1.3185935925671612e-06,
+      "loss": 0.5024,
+      "step": 9811
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.2444307203724163,
+      "learning_rate": 1.31773311774403e-06,
+      "loss": 0.3797,
+      "step": 9812
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7301548652235517,
+      "learning_rate": 1.3168728811587471e-06,
+      "loss": 0.4219,
+      "step": 9813
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0353222835866855,
+      "learning_rate": 1.316012882866971e-06,
+      "loss": 0.4517,
+      "step": 9814
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8052040968909633,
+      "learning_rate": 1.3151531229243424e-06,
+      "loss": 0.5088,
+      "step": 9815
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3425719734328476,
+      "learning_rate": 1.3142936013864876e-06,
+      "loss": 0.4234,
+      "step": 9816
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6281791239865038,
+      "learning_rate": 1.3134343183090137e-06,
+      "loss": 0.4386,
+      "step": 9817
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5342242556977543,
+      "learning_rate": 1.3125752737475167e-06,
+      "loss": 0.4532,
+      "step": 9818
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8926627896884447,
+      "learning_rate": 1.311716467757576e-06,
+      "loss": 0.449,
+      "step": 9819
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.422622734935708,
+      "learning_rate": 1.3108579003947546e-06,
+      "loss": 0.3948,
+      "step": 9820
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5855818414106012,
+      "learning_rate": 1.309999571714603e-06,
+      "loss": 0.4644,
+      "step": 9821
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.488293214863266,
+      "learning_rate": 1.3091414817726483e-06,
+      "loss": 0.4581,
+      "step": 9822
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.205321705239975,
+      "learning_rate": 1.308283630624415e-06,
+      "loss": 0.4253,
+      "step": 9823
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.285906483377957,
+      "learning_rate": 1.3074260183254e-06,
+      "loss": 0.4629,
+      "step": 9824
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5537878799615814,
+      "learning_rate": 1.3065686449310922e-06,
+      "loss": 0.4511,
+      "step": 9825
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7795200422370296,
+      "learning_rate": 1.3057115104969586e-06,
+      "loss": 0.3775,
+      "step": 9826
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.870723241567607,
+      "learning_rate": 1.3048546150784609e-06,
+      "loss": 0.4295,
+      "step": 9827
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5407619162806406,
+      "learning_rate": 1.3039979587310343e-06,
+      "loss": 0.4587,
+      "step": 9828
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.869630397445088,
+      "learning_rate": 1.3031415415101039e-06,
+      "loss": 0.4071,
+      "step": 9829
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.965671893502372,
+      "learning_rate": 1.3022853634710803e-06,
+      "loss": 0.5198,
+      "step": 9830
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8517644058214042,
+      "learning_rate": 1.3014294246693565e-06,
+      "loss": 0.4298,
+      "step": 9831
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.931629945425655,
+      "learning_rate": 1.3005737251603117e-06,
+      "loss": 0.4668,
+      "step": 9832
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5241974859314537,
+      "learning_rate": 1.2997182649993068e-06,
+      "loss": 0.4772,
+      "step": 9833
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7567851186971823,
+      "learning_rate": 1.298863044241689e-06,
+      "loss": 0.4202,
+      "step": 9834
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5190000238083787,
+      "learning_rate": 1.2980080629427904e-06,
+      "loss": 0.4438,
+      "step": 9835
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.588945532635868,
+      "learning_rate": 1.2971533211579285e-06,
+      "loss": 0.4602,
+      "step": 9836
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.658398718537271,
+      "learning_rate": 1.296298818942402e-06,
+      "loss": 0.5749,
+      "step": 9837
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.326847104233507,
+      "learning_rate": 1.2954445563514966e-06,
+      "loss": 0.4386,
+      "step": 9838
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.4916592985201653,
+      "learning_rate": 1.2945905334404823e-06,
+      "loss": 0.4342,
+      "step": 9839
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.26873847420643,
+      "learning_rate": 1.2937367502646147e-06,
+      "loss": 0.4532,
+      "step": 9840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5683758370996783,
+      "learning_rate": 1.2928832068791275e-06,
+      "loss": 0.4822,
+      "step": 9841
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5455769787056626,
+      "learning_rate": 1.2920299033392492e-06,
+      "loss": 0.4319,
+      "step": 9842
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9132281596517118,
+      "learning_rate": 1.2911768397001873e-06,
+      "loss": 0.4885,
+      "step": 9843
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7972331351230315,
+      "learning_rate": 1.2903240160171304e-06,
+      "loss": 0.4187,
+      "step": 9844
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5648386324281731,
+      "learning_rate": 1.2894714323452568e-06,
+      "loss": 0.4717,
+      "step": 9845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1253718957580534,
+      "learning_rate": 1.288619088739727e-06,
+      "loss": 0.4887,
+      "step": 9846
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.3566837018351228,
+      "learning_rate": 1.287766985255689e-06,
+      "loss": 0.4555,
+      "step": 9847
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5524634094666402,
+      "learning_rate": 1.2869151219482695e-06,
+      "loss": 0.4535,
+      "step": 9848
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7896042531605219,
+      "learning_rate": 1.286063498872584e-06,
+      "loss": 0.4642,
+      "step": 9849
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.990259620710734,
+      "learning_rate": 1.2852121160837323e-06,
+      "loss": 0.4617,
+      "step": 9850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.172131941735793,
+      "learning_rate": 1.2843609736367978e-06,
+      "loss": 0.4756,
+      "step": 9851
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.04232448806431,
+      "learning_rate": 1.2835100715868465e-06,
+      "loss": 0.4368,
+      "step": 9852
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.0409907375930376,
+      "learning_rate": 1.2826594099889322e-06,
+      "loss": 0.4469,
+      "step": 9853
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1488936337963698,
+      "learning_rate": 1.2818089888980906e-06,
+      "loss": 0.4232,
+      "step": 9854
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5401719741359496,
+      "learning_rate": 1.2809588083693436e-06,
+      "loss": 0.4638,
+      "step": 9855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.1293472026315836,
+      "learning_rate": 1.2801088684576979e-06,
+      "loss": 0.4272,
+      "step": 9856
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6719735790227794,
+      "learning_rate": 1.2792591692181393e-06,
+      "loss": 0.4248,
+      "step": 9857
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.784254522689791,
+      "learning_rate": 1.278409710705648e-06,
+      "loss": 0.4354,
+      "step": 9858
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5275281126781837,
+      "learning_rate": 1.277560492975179e-06,
+      "loss": 0.4255,
+      "step": 9859
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5984211169223819,
+      "learning_rate": 1.276711516081678e-06,
+      "loss": 0.4526,
+      "step": 9860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.807004717862255,
+      "learning_rate": 1.2758627800800677e-06,
+      "loss": 0.491,
+      "step": 9861
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.9576853635297746,
+      "learning_rate": 1.2750142850252667e-06,
+      "loss": 0.5021,
+      "step": 9862
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8398768026562364,
+      "learning_rate": 1.2741660309721677e-06,
+      "loss": 0.4962,
+      "step": 9863
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 6.199763645472166,
+      "learning_rate": 1.2733180179756515e-06,
+      "loss": 0.4189,
+      "step": 9864
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5138608536669774,
+      "learning_rate": 1.2724702460905853e-06,
+      "loss": 0.449,
+      "step": 9865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5622340562567587,
+      "learning_rate": 1.271622715371818e-06,
+      "loss": 0.4845,
+      "step": 9866
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5342187282529401,
+      "learning_rate": 1.2707754258741845e-06,
+      "loss": 0.4659,
+      "step": 9867
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6251096065496584,
+      "learning_rate": 1.2699283776525017e-06,
+      "loss": 0.4797,
+      "step": 9868
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9495037112536284,
+      "learning_rate": 1.2690815707615727e-06,
+      "loss": 0.4678,
+      "step": 9869
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6086932940444034,
+      "learning_rate": 1.268235005256186e-06,
+      "loss": 0.4446,
+      "step": 9870
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7203885196696374,
+      "learning_rate": 1.267388681191114e-06,
+      "loss": 0.4484,
+      "step": 9871
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8644574076252163,
+      "learning_rate": 1.2665425986211094e-06,
+      "loss": 0.466,
+      "step": 9872
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7669623573024322,
+      "learning_rate": 1.2656967576009155e-06,
+      "loss": 0.422,
+      "step": 9873
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.028300996152387,
+      "learning_rate": 1.2648511581852557e-06,
+      "loss": 0.4337,
+      "step": 9874
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.842818952873279,
+      "learning_rate": 1.2640058004288402e-06,
+      "loss": 0.4038,
+      "step": 9875
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1706158749446196,
+      "learning_rate": 1.263160684386362e-06,
+      "loss": 0.4027,
+      "step": 9876
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5478624444558307,
+      "learning_rate": 1.2623158101124995e-06,
+      "loss": 0.4089,
+      "step": 9877
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4199987812309558,
+      "learning_rate": 1.2614711776619154e-06,
+      "loss": 0.4992,
+      "step": 9878
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7965907808414838,
+      "learning_rate": 1.2606267870892541e-06,
+      "loss": 0.4579,
+      "step": 9879
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.326134044212904,
+      "learning_rate": 1.259782638449148e-06,
+      "loss": 0.4601,
+      "step": 9880
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4429646459903784,
+      "learning_rate": 1.2589387317962126e-06,
+      "loss": 0.4423,
+      "step": 9881
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.849649173950172,
+      "learning_rate": 1.2580950671850488e-06,
+      "loss": 0.491,
+      "step": 9882
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.05358154404609,
+      "learning_rate": 1.2572516446702376e-06,
+      "loss": 0.4952,
+      "step": 9883
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9121105454089926,
+      "learning_rate": 1.2564084643063485e-06,
+      "loss": 0.4083,
+      "step": 9884
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7235992118250365,
+      "learning_rate": 1.255565526147935e-06,
+      "loss": 0.4742,
+      "step": 9885
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9650005750129307,
+      "learning_rate": 1.2547228302495334e-06,
+      "loss": 0.459,
+      "step": 9886
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5334263622553529,
+      "learning_rate": 1.253880376665667e-06,
+      "loss": 0.4773,
+      "step": 9887
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5196671081364501,
+      "learning_rate": 1.2530381654508377e-06,
+      "loss": 0.4672,
+      "step": 9888
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.599797577333411,
+      "learning_rate": 1.252196196659538e-06,
+      "loss": 0.471,
+      "step": 9889
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9724167519595683,
+      "learning_rate": 1.251354470346241e-06,
+      "loss": 0.4228,
+      "step": 9890
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8391791605091385,
+      "learning_rate": 1.250512986565408e-06,
+      "loss": 0.4731,
+      "step": 9891
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5489642498210031,
+      "learning_rate": 1.2496717453714769e-06,
+      "loss": 0.4806,
+      "step": 9892
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8260042414007853,
+      "learning_rate": 1.2488307468188805e-06,
+      "loss": 0.3926,
+      "step": 9893
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5828042894404579,
+      "learning_rate": 1.2479899909620258e-06,
+      "loss": 0.4676,
+      "step": 9894
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7757157675087814,
+      "learning_rate": 1.2471494778553128e-06,
+      "loss": 0.4054,
+      "step": 9895
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.29806407000764,
+      "learning_rate": 1.2463092075531158e-06,
+      "loss": 0.4547,
+      "step": 9896
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6058518900894856,
+      "learning_rate": 1.245469180109804e-06,
+      "loss": 0.4906,
+      "step": 9897
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1894937114798796,
+      "learning_rate": 1.2446293955797262e-06,
+      "loss": 0.5012,
+      "step": 9898
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7441777520870199,
+      "learning_rate": 1.2437898540172122e-06,
+      "loss": 0.5031,
+      "step": 9899
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5323359671712155,
+      "learning_rate": 1.2429505554765813e-06,
+      "loss": 0.4729,
+      "step": 9900
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4256854764996074,
+      "learning_rate": 1.2421115000121347e-06,
+      "loss": 0.462,
+      "step": 9901
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5698182656364748,
+      "learning_rate": 1.2412726876781594e-06,
+      "loss": 0.4525,
+      "step": 9902
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3352994204416662,
+      "learning_rate": 1.2404341185289226e-06,
+      "loss": 0.4005,
+      "step": 9903
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9110936550177373,
+      "learning_rate": 1.2395957926186802e-06,
+      "loss": 0.4255,
+      "step": 9904
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5794053515373636,
+      "learning_rate": 1.2387577100016706e-06,
+      "loss": 0.4853,
+      "step": 9905
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7277781653299185,
+      "learning_rate": 1.2379198707321189e-06,
+      "loss": 0.4378,
+      "step": 9906
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9080405587604494,
+      "learning_rate": 1.237082274864228e-06,
+      "loss": 0.4573,
+      "step": 9907
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9469390874965564,
+      "learning_rate": 1.2362449224521905e-06,
+      "loss": 0.5221,
+      "step": 9908
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6041981928925877,
+      "learning_rate": 1.2354078135501857e-06,
+      "loss": 0.398,
+      "step": 9909
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4548058323468214,
+      "learning_rate": 1.234570948212369e-06,
+      "loss": 0.4097,
+      "step": 9910
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6171399005921256,
+      "learning_rate": 1.2337343264928863e-06,
+      "loss": 0.4321,
+      "step": 9911
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.763854713977996,
+      "learning_rate": 1.232897948445866e-06,
+      "loss": 0.493,
+      "step": 9912
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5831070085581385,
+      "learning_rate": 1.2320618141254214e-06,
+      "loss": 0.4825,
+      "step": 9913
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.866635424288943,
+      "learning_rate": 1.2312259235856471e-06,
+      "loss": 0.4508,
+      "step": 9914
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.9752349673969443,
+      "learning_rate": 1.2303902768806252e-06,
+      "loss": 0.4699,
+      "step": 9915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.797462130394774,
+      "learning_rate": 1.2295548740644213e-06,
+      "loss": 0.5185,
+      "step": 9916
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 4.304825930292483,
+      "learning_rate": 1.2287197151910862e-06,
+      "loss": 0.4356,
+      "step": 9917
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5086793026636335,
+      "learning_rate": 1.227884800314651e-06,
+      "loss": 0.4641,
+      "step": 9918
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4596519792437321,
+      "learning_rate": 1.2270501294891341e-06,
+      "loss": 0.4481,
+      "step": 9919
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5448801912343616,
+      "learning_rate": 1.2262157027685384e-06,
+      "loss": 0.4972,
+      "step": 9920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2458879647467556,
+      "learning_rate": 1.2253815202068497e-06,
+      "loss": 0.4656,
+      "step": 9921
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.253077786647236,
+      "learning_rate": 1.2245475818580404e-06,
+      "loss": 0.4609,
+      "step": 9922
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5611500626285899,
+      "learning_rate": 1.2237138877760623e-06,
+      "loss": 0.4501,
+      "step": 9923
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5840354398186929,
+      "learning_rate": 1.2228804380148556e-06,
+      "loss": 0.4863,
+      "step": 9924
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1218853810262748,
+      "learning_rate": 1.222047232628344e-06,
+      "loss": 0.4536,
+      "step": 9925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.214266944519692,
+      "learning_rate": 1.221214271670435e-06,
+      "loss": 0.4082,
+      "step": 9926
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.474122939002315,
+      "learning_rate": 1.2203815551950165e-06,
+      "loss": 0.4345,
+      "step": 9927
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.104951180310517,
+      "learning_rate": 1.2195490832559704e-06,
+      "loss": 0.4321,
+      "step": 9928
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8340179677762505,
+      "learning_rate": 1.2187168559071517e-06,
+      "loss": 0.4338,
+      "step": 9929
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9407527022312259,
+      "learning_rate": 1.2178848732024069e-06,
+      "loss": 0.4398,
+      "step": 9930
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6665816190566394,
+      "learning_rate": 1.2170531351955605e-06,
+      "loss": 0.4085,
+      "step": 9931
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.4615416643228762,
+      "learning_rate": 1.2162216419404289e-06,
+      "loss": 0.438,
+      "step": 9932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7694842972383025,
+      "learning_rate": 1.2153903934908084e-06,
+      "loss": 0.4936,
+      "step": 9933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0637775037829873,
+      "learning_rate": 1.2145593899004777e-06,
+      "loss": 0.446,
+      "step": 9934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 3.450364050223724,
+      "learning_rate": 1.2137286312232022e-06,
+      "loss": 0.4466,
+      "step": 9935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2291485954652237,
+      "learning_rate": 1.2128981175127314e-06,
+      "loss": 0.4485,
+      "step": 9936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6169224279208168,
+      "learning_rate": 1.2120678488227994e-06,
+      "loss": 0.4509,
+      "step": 9937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2904789201098508,
+      "learning_rate": 1.211237825207121e-06,
+      "loss": 0.4667,
+      "step": 9938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6779455011223394,
+      "learning_rate": 1.2104080467193991e-06,
+      "loss": 0.4756,
+      "step": 9939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8871182208201456,
+      "learning_rate": 1.2095785134133187e-06,
+      "loss": 0.4948,
+      "step": 9940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7655107105732397,
+      "learning_rate": 1.208749225342552e-06,
+      "loss": 0.4383,
+      "step": 9941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2695138181530354,
+      "learning_rate": 1.207920182560749e-06,
+      "loss": 0.4496,
+      "step": 9942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8135839918449157,
+      "learning_rate": 1.2070913851215482e-06,
+      "loss": 0.424,
+      "step": 9943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0693347817800203,
+      "learning_rate": 1.2062628330785758e-06,
+      "loss": 0.4829,
+      "step": 9944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6251159898150476,
+      "learning_rate": 1.2054345264854344e-06,
+      "loss": 0.4375,
+      "step": 9945
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5535407938864633,
+      "learning_rate": 1.2046064653957147e-06,
+      "loss": 0.4467,
+      "step": 9946
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.384064629267377,
+      "learning_rate": 1.2037786498629916e-06,
+      "loss": 0.4314,
+      "step": 9947
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5791156704300235,
+      "learning_rate": 1.2029510799408256e-06,
+      "loss": 0.4431,
+      "step": 9948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6773274329880625,
+      "learning_rate": 1.2021237556827558e-06,
+      "loss": 0.4507,
+      "step": 9949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.431598623468071,
+      "learning_rate": 1.2012966771423112e-06,
+      "loss": 0.4252,
+      "step": 9950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6592970970066585,
+      "learning_rate": 1.2004698443730018e-06,
+      "loss": 0.4528,
+      "step": 9951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8403349594604146,
+      "learning_rate": 1.1996432574283245e-06,
+      "loss": 0.4828,
+      "step": 9952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5082034479910564,
+      "learning_rate": 1.1988169163617557e-06,
+      "loss": 0.4643,
+      "step": 9953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7542844335228094,
+      "learning_rate": 1.1979908212267599e-06,
+      "loss": 0.4257,
+      "step": 9954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5690702918212723,
+      "learning_rate": 1.1971649720767847e-06,
+      "loss": 0.4725,
+      "step": 9955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9817945666212584,
+      "learning_rate": 1.1963393689652603e-06,
+      "loss": 0.4124,
+      "step": 9956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8299977612464928,
+      "learning_rate": 1.1955140119456049e-06,
+      "loss": 0.48,
+      "step": 9957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6467384896579371,
+      "learning_rate": 1.1946889010712143e-06,
+      "loss": 0.3727,
+      "step": 9958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.3029892016173763,
+      "learning_rate": 1.193864036395474e-06,
+      "loss": 0.4493,
+      "step": 9959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9224826709130185,
+      "learning_rate": 1.1930394179717513e-06,
+      "loss": 0.4679,
+      "step": 9960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2186226574151795,
+      "learning_rate": 1.1922150458533993e-06,
+      "loss": 0.4339,
+      "step": 9961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.036283201440223,
+      "learning_rate": 1.19139092009375e-06,
+      "loss": 0.4187,
+      "step": 9962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4004004656923177,
+      "learning_rate": 1.1905670407461295e-06,
+      "loss": 0.4193,
+      "step": 9963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.1458357186531694,
+      "learning_rate": 1.189743407863836e-06,
+      "loss": 0.4948,
+      "step": 9964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6180800036755154,
+      "learning_rate": 1.1889200215001595e-06,
+      "loss": 0.387,
+      "step": 9965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5314298802656144,
+      "learning_rate": 1.188096881708372e-06,
+      "loss": 0.4666,
+      "step": 9966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9950293982609388,
+      "learning_rate": 1.18727398854173e-06,
+      "loss": 0.4727,
+      "step": 9967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7834056046502953,
+      "learning_rate": 1.1864513420534746e-06,
+      "loss": 0.4238,
+      "step": 9968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.468152471658757,
+      "learning_rate": 1.1856289422968271e-06,
+      "loss": 0.4463,
+      "step": 9969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9138845406716694,
+      "learning_rate": 1.184806789324997e-06,
+      "loss": 0.4564,
+      "step": 9970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8219396894224027,
+      "learning_rate": 1.1839848831911772e-06,
+      "loss": 0.4664,
+      "step": 9971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9908679180317457,
+      "learning_rate": 1.1831632239485446e-06,
+      "loss": 0.4371,
+      "step": 9972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4926536057576503,
+      "learning_rate": 1.1823418116502566e-06,
+      "loss": 0.4576,
+      "step": 9973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.951163140894335,
+      "learning_rate": 1.18152064634946e-06,
+      "loss": 0.4512,
+      "step": 9974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5585680077765179,
+      "learning_rate": 1.180699728099282e-06,
+      "loss": 0.4833,
+      "step": 9975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.2430595345526263,
+      "learning_rate": 1.1798790569528356e-06,
+      "loss": 0.4404,
+      "step": 9976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8148424587147551,
+      "learning_rate": 1.179058632963218e-06,
+      "loss": 0.4307,
+      "step": 9977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7420169557679468,
+      "learning_rate": 1.1782384561835052e-06,
+      "loss": 0.426,
+      "step": 9978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0582407115018397,
+      "learning_rate": 1.177418526666768e-06,
+      "loss": 0.3959,
+      "step": 9979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6864013496166899,
+      "learning_rate": 1.1765988444660508e-06,
+      "loss": 0.4894,
+      "step": 9980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.418173686730568,
+      "learning_rate": 1.1757794096343862e-06,
+      "loss": 0.4534,
+      "step": 9981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7759228502152764,
+      "learning_rate": 1.1749602222247908e-06,
+      "loss": 0.4532,
+      "step": 9982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4779370238483347,
+      "learning_rate": 1.1741412822902671e-06,
+      "loss": 0.4359,
+      "step": 9983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.6539670958047419,
+      "learning_rate": 1.1733225898837958e-06,
+      "loss": 0.4996,
+      "step": 9984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.7604816210236347,
+      "learning_rate": 1.172504145058347e-06,
+      "loss": 0.4669,
+      "step": 9985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.628253010602657,
+      "learning_rate": 1.1716859478668735e-06,
+      "loss": 0.4328,
+      "step": 9986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.576661623706323,
+      "learning_rate": 1.1708679983623122e-06,
+      "loss": 0.4275,
+      "step": 9987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.5528366063370465,
+      "learning_rate": 1.1700502965975808e-06,
+      "loss": 0.4376,
+      "step": 9988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5575971980865699,
+      "learning_rate": 1.169232842625585e-06,
+      "loss": 0.488,
+      "step": 9989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5605163162011814,
+      "learning_rate": 1.1684156364992133e-06,
+      "loss": 0.4654,
+      "step": 9990
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.014104133480851,
+      "learning_rate": 1.1675986782713372e-06,
+      "loss": 0.49,
+      "step": 9991
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.0143676496126663,
+      "learning_rate": 1.1667819679948145e-06,
+      "loss": 0.4435,
+      "step": 9992
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5501372127426185,
+      "learning_rate": 1.1659655057224834e-06,
+      "loss": 0.4581,
+      "step": 9993
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.8008272911397039,
+      "learning_rate": 1.1651492915071678e-06,
+      "loss": 0.4275,
+      "step": 9994
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5457069084323105,
+      "learning_rate": 1.1643333254016765e-06,
+      "loss": 0.4718,
+      "step": 9995
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.647085660820638,
+      "learning_rate": 1.1635176074588028e-06,
+      "loss": 0.5012,
+      "step": 9996
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.421620859981154,
+      "learning_rate": 1.1627021377313186e-06,
+      "loss": 0.4216,
+      "step": 9997
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.809655883939099,
+      "learning_rate": 1.1618869162719887e-06,
+      "loss": 0.5097,
+      "step": 9998
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.29139997838529,
+      "learning_rate": 1.1610719431335531e-06,
+      "loss": 0.4347,
+      "step": 9999
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5989818814749023,
+      "learning_rate": 1.1602572183687411e-06,
+      "loss": 0.5224,
+      "step": 10000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2937749957403455,
+      "learning_rate": 1.159442742030264e-06,
+      "loss": 0.4216,
+      "step": 10001
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4902885052366017,
+      "learning_rate": 1.1586285141708176e-06,
+      "loss": 0.5045,
+      "step": 10002
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7534818875848799,
+      "learning_rate": 1.157814534843082e-06,
+      "loss": 0.4205,
+      "step": 10003
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6569326609005501,
+      "learning_rate": 1.157000804099719e-06,
+      "loss": 0.4625,
+      "step": 10004
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.1153820032184525,
+      "learning_rate": 1.1561873219933762e-06,
+      "loss": 0.4792,
+      "step": 10005
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.313001344363856,
+      "learning_rate": 1.1553740885766857e-06,
+      "loss": 0.4271,
+      "step": 10006
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0491628184246715,
+      "learning_rate": 1.1545611039022637e-06,
+      "loss": 0.4403,
+      "step": 10007
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5193909216542221,
+      "learning_rate": 1.1537483680227058e-06,
+      "loss": 0.4806,
+      "step": 10008
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5499873585068724,
+      "learning_rate": 1.1529358809905971e-06,
+      "loss": 0.4809,
+      "step": 10009
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5125965011781745,
+      "learning_rate": 1.1521236428585047e-06,
+      "loss": 0.4527,
+      "step": 10010
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.473926970040196,
+      "learning_rate": 1.1513116536789792e-06,
+      "loss": 0.4537,
+      "step": 10011
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4358835663442993,
+      "learning_rate": 1.1504999135045558e-06,
+      "loss": 0.4445,
+      "step": 10012
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.646335808612089,
+      "learning_rate": 1.1496884223877498e-06,
+      "loss": 0.4642,
+      "step": 10013
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.568675840165188,
+      "learning_rate": 1.1488771803810684e-06,
+      "loss": 0.4282,
+      "step": 10014
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5611465754971974,
+      "learning_rate": 1.1480661875369947e-06,
+      "loss": 0.4305,
+      "step": 10015
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6794813328280251,
+      "learning_rate": 1.1472554439080007e-06,
+      "loss": 0.4677,
+      "step": 10016
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.440425831995939,
+      "learning_rate": 1.1464449495465368e-06,
+      "loss": 0.4176,
+      "step": 10017
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.797709889059092,
+      "learning_rate": 1.1456347045050463e-06,
+      "loss": 0.415,
+      "step": 10018
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.738777762315388,
+      "learning_rate": 1.144824708835947e-06,
+      "loss": 0.489,
+      "step": 10019
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7811877513908148,
+      "learning_rate": 1.1440149625916458e-06,
+      "loss": 0.4076,
+      "step": 10020
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5827299257063068,
+      "learning_rate": 1.1432054658245323e-06,
+      "loss": 0.4892,
+      "step": 10021
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8512464599301277,
+      "learning_rate": 1.1423962185869798e-06,
+      "loss": 0.44,
+      "step": 10022
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.537118812682858,
+      "learning_rate": 1.1415872209313466e-06,
+      "loss": 0.4744,
+      "step": 10023
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5649324672683547,
+      "learning_rate": 1.140778472909972e-06,
+      "loss": 0.4827,
+      "step": 10024
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6239141164720323,
+      "learning_rate": 1.1399699745751813e-06,
+      "loss": 0.444,
+      "step": 10025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6913938095130232,
+      "learning_rate": 1.1391617259792836e-06,
+      "loss": 0.5242,
+      "step": 10026
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5456945082691544,
+      "learning_rate": 1.1383537271745732e-06,
+      "loss": 0.4012,
+      "step": 10027
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6273633722324115,
+      "learning_rate": 1.1375459782133236e-06,
+      "loss": 0.4785,
+      "step": 10028
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.3777686672987817,
+      "learning_rate": 1.1367384791477964e-06,
+      "loss": 0.4541,
+      "step": 10029
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8233022556053948,
+      "learning_rate": 1.1359312300302361e-06,
+      "loss": 0.4733,
+      "step": 10030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.554051549842203,
+      "learning_rate": 1.1351242309128713e-06,
+      "loss": 0.4697,
+      "step": 10031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4047237450400798,
+      "learning_rate": 1.1343174818479103e-06,
+      "loss": 0.4407,
+      "step": 10032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5839521666223214,
+      "learning_rate": 1.133510982887553e-06,
+      "loss": 0.4754,
+      "step": 10033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.292550186325647,
+      "learning_rate": 1.132704734083978e-06,
+      "loss": 0.4588,
+      "step": 10034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5203647903507407,
+      "learning_rate": 1.1318987354893463e-06,
+      "loss": 0.4117,
+      "step": 10035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0278095392008386,
+      "learning_rate": 1.1310929871558068e-06,
+      "loss": 0.4154,
+      "step": 10036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.501440224635936,
+      "learning_rate": 1.1302874891354893e-06,
+      "loss": 0.4418,
+      "step": 10037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0775485183665126,
+      "learning_rate": 1.1294822414805106e-06,
+      "loss": 0.5021,
+      "step": 10038
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6244385112750148,
+      "learning_rate": 1.1286772442429667e-06,
+      "loss": 0.4278,
+      "step": 10039
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5100642143750571,
+      "learning_rate": 1.12787249747494e-06,
+      "loss": 0.4635,
+      "step": 10040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5164894743984914,
+      "learning_rate": 1.127068001228498e-06,
+      "loss": 0.4437,
+      "step": 10041
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8078510962786887,
+      "learning_rate": 1.1262637555556905e-06,
+      "loss": 0.4517,
+      "step": 10042
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5671661342163807,
+      "learning_rate": 1.12545976050855e-06,
+      "loss": 0.4318,
+      "step": 10043
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5310635909925292,
+      "learning_rate": 1.1246560161390925e-06,
+      "loss": 0.4857,
+      "step": 10044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8098884360179892,
+      "learning_rate": 1.1238525224993241e-06,
+      "loss": 0.4506,
+      "step": 10045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.029640331758974,
+      "learning_rate": 1.1230492796412258e-06,
+      "loss": 0.4725,
+      "step": 10046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5222749375809408,
+      "learning_rate": 1.1222462876167684e-06,
+      "loss": 0.4676,
+      "step": 10047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8984792153519288,
+      "learning_rate": 1.1214435464779006e-06,
+      "loss": 0.4544,
+      "step": 10048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5707293588161633,
+      "learning_rate": 1.1206410562765647e-06,
+      "loss": 0.4921,
+      "step": 10049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7187449274734654,
+      "learning_rate": 1.1198388170646758e-06,
+      "loss": 0.4653,
+      "step": 10050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.741097488078969,
+      "learning_rate": 1.119036828894141e-06,
+      "loss": 0.4321,
+      "step": 10051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.8068443278287956,
+      "learning_rate": 1.118235091816844e-06,
+      "loss": 0.4569,
+      "step": 10052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0535043314844557,
+      "learning_rate": 1.1174336058846608e-06,
+      "loss": 0.4297,
+      "step": 10053
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5286886783440068,
+      "learning_rate": 1.1166323711494438e-06,
+      "loss": 0.4638,
+      "step": 10054
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6937354309953243,
+      "learning_rate": 1.1158313876630311e-06,
+      "loss": 0.4444,
+      "step": 10055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8765475475668605,
+      "learning_rate": 1.1150306554772472e-06,
+      "loss": 0.4612,
+      "step": 10056
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.992382484442274,
+      "learning_rate": 1.1142301746438978e-06,
+      "loss": 0.4235,
+      "step": 10057
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9812451679340213,
+      "learning_rate": 1.113429945214774e-06,
+      "loss": 0.3851,
+      "step": 10058
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4420701126305544,
+      "learning_rate": 1.1126299672416474e-06,
+      "loss": 0.4877,
+      "step": 10059
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8827501000418154,
+      "learning_rate": 1.111830240776276e-06,
+      "loss": 0.4571,
+      "step": 10060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9510646900373567,
+      "learning_rate": 1.1110307658704023e-06,
+      "loss": 0.4637,
+      "step": 10061
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.343497489191553,
+      "learning_rate": 1.1102315425757514e-06,
+      "loss": 0.4265,
+      "step": 10062
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.461065753717907,
+      "learning_rate": 1.1094325709440306e-06,
+      "loss": 0.4333,
+      "step": 10063
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9664785611775422,
+      "learning_rate": 1.1086338510269324e-06,
+      "loss": 0.4431,
+      "step": 10064
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8027441426511144,
+      "learning_rate": 1.107835382876134e-06,
+      "loss": 0.4497,
+      "step": 10065
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9146061241076822,
+      "learning_rate": 1.1070371665432955e-06,
+      "loss": 0.4388,
+      "step": 10066
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 6.802076654544321,
+      "learning_rate": 1.1062392020800571e-06,
+      "loss": 0.4759,
+      "step": 10067
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.986980764309152,
+      "learning_rate": 1.1054414895380504e-06,
+      "loss": 0.4353,
+      "step": 10068
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.175705275420327,
+      "learning_rate": 1.1046440289688859e-06,
+      "loss": 0.4657,
+      "step": 10069
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 4.279797579439347,
+      "learning_rate": 1.1038468204241553e-06,
+      "loss": 0.4458,
+      "step": 10070
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5203828842706305,
+      "learning_rate": 1.1030498639554388e-06,
+      "loss": 0.4524,
+      "step": 10071
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9083944104413655,
+      "learning_rate": 1.1022531596142978e-06,
+      "loss": 0.4589,
+      "step": 10072
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4602612627777611,
+      "learning_rate": 1.1014567074522804e-06,
+      "loss": 0.407,
+      "step": 10073
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5200842627646898,
+      "learning_rate": 1.1006605075209127e-06,
+      "loss": 0.4651,
+      "step": 10074
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.527062461962958,
+      "learning_rate": 1.0998645598717088e-06,
+      "loss": 0.4612,
+      "step": 10075
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9121978114494584,
+      "learning_rate": 1.099068864556166e-06,
+      "loss": 0.4242,
+      "step": 10076
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.598748133599777,
+      "learning_rate": 1.0982734216257663e-06,
+      "loss": 0.4468,
+      "step": 10077
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.644147805144959,
+      "learning_rate": 1.0974782311319705e-06,
+      "loss": 0.4153,
+      "step": 10078
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6746698043604378,
+      "learning_rate": 1.0966832931262266e-06,
+      "loss": 0.4945,
+      "step": 10079
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6062556421542185,
+      "learning_rate": 1.09588860765997e-06,
+      "loss": 0.4372,
+      "step": 10080
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9156149015828148,
+      "learning_rate": 1.0950941747846123e-06,
+      "loss": 0.4627,
+      "step": 10081
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5239708024677558,
+      "learning_rate": 1.0942999945515542e-06,
+      "loss": 0.4788,
+      "step": 10082
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7085165876123345,
+      "learning_rate": 1.093506067012175e-06,
+      "loss": 0.435,
+      "step": 10083
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5268933168649017,
+      "learning_rate": 1.092712392217845e-06,
+      "loss": 0.4544,
+      "step": 10084
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.223037871441057,
+      "learning_rate": 1.0919189702199106e-06,
+      "loss": 0.4325,
+      "step": 10085
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5422378232099807,
+      "learning_rate": 1.0911258010697084e-06,
+      "loss": 0.4839,
+      "step": 10086
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3670030364387182,
+      "learning_rate": 1.0903328848185502e-06,
+      "loss": 0.4075,
+      "step": 10087
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4758455755496591,
+      "learning_rate": 1.0895402215177425e-06,
+      "loss": 0.4537,
+      "step": 10088
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2712702885966063,
+      "learning_rate": 1.0887478112185656e-06,
+      "loss": 0.3955,
+      "step": 10089
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5708667573810321,
+      "learning_rate": 1.0879556539722892e-06,
+      "loss": 0.4911,
+      "step": 10090
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8508735844326745,
+      "learning_rate": 1.0871637498301641e-06,
+      "loss": 0.4596,
+      "step": 10091
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5825759227327696,
+      "learning_rate": 1.0863720988434257e-06,
+      "loss": 0.4687,
+      "step": 10092
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.531806573215688,
+      "learning_rate": 1.0855807010632941e-06,
+      "loss": 0.45,
+      "step": 10093
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5500343390307103,
+      "learning_rate": 1.0847895565409694e-06,
+      "loss": 0.4556,
+      "step": 10094
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5167681809806384,
+      "learning_rate": 1.0839986653276385e-06,
+      "loss": 0.4665,
+      "step": 10095
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7299713853242789,
+      "learning_rate": 1.083208027474471e-06,
+      "loss": 0.4988,
+      "step": 10096
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5516558890094574,
+      "learning_rate": 1.0824176430326217e-06,
+      "loss": 0.4862,
+      "step": 10097
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.5045410156574004,
+      "learning_rate": 1.081627512053225e-06,
+      "loss": 0.4933,
+      "step": 10098
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9727140551739775,
+      "learning_rate": 1.0808376345874021e-06,
+      "loss": 0.4661,
+      "step": 10099
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6996984291407797,
+      "learning_rate": 1.0800480106862575e-06,
+      "loss": 0.4508,
+      "step": 10100
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.0836489190082763,
+      "learning_rate": 1.0792586404008788e-06,
+      "loss": 0.4154,
+      "step": 10101
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8713811642073703,
+      "learning_rate": 1.0784695237823368e-06,
+      "loss": 0.4478,
+      "step": 10102
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 5.34651316275778,
+      "learning_rate": 1.0776806608816863e-06,
+      "loss": 0.4779,
+      "step": 10103
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.468786105798487,
+      "learning_rate": 1.0768920517499681e-06,
+      "loss": 0.4599,
+      "step": 10104
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7699723849202884,
+      "learning_rate": 1.0761036964382004e-06,
+      "loss": 0.4194,
+      "step": 10105
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.704743010951251,
+      "learning_rate": 1.0753155949973903e-06,
+      "loss": 0.4728,
+      "step": 10106
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.9431877626586613,
+      "learning_rate": 1.074527747478527e-06,
+      "loss": 0.4507,
+      "step": 10107
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6056503841010683,
+      "learning_rate": 1.0737401539325848e-06,
+      "loss": 0.4602,
+      "step": 10108
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.379884067229351,
+      "learning_rate": 1.0729528144105171e-06,
+      "loss": 0.4448,
+      "step": 10109
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.164551580461938,
+      "learning_rate": 1.0721657289632654e-06,
+      "loss": 0.5092,
+      "step": 10110
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5196280155115981,
+      "learning_rate": 1.0713788976417522e-06,
+      "loss": 0.4763,
+      "step": 10111
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.569126974211081,
+      "learning_rate": 1.0705923204968855e-06,
+      "loss": 0.4598,
+      "step": 10112
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.3847303267708413,
+      "learning_rate": 1.0698059975795566e-06,
+      "loss": 0.4451,
+      "step": 10113
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.450786934867951,
+      "learning_rate": 1.0690199289406355e-06,
+      "loss": 0.4562,
+      "step": 10114
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5355269025012972,
+      "learning_rate": 1.0682341146309854e-06,
+      "loss": 0.4853,
+      "step": 10115
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4649698208335638,
+      "learning_rate": 1.0674485547014435e-06,
+      "loss": 0.4431,
+      "step": 10116
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8715216101613363,
+      "learning_rate": 1.0666632492028367e-06,
+      "loss": 0.4191,
+      "step": 10117
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5481972737754428,
+      "learning_rate": 1.0658781981859694e-06,
+      "loss": 0.4611,
+      "step": 10118
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5617909821057572,
+      "learning_rate": 1.0650934017016396e-06,
+      "loss": 0.4777,
+      "step": 10119
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 7.594774739197863,
+      "learning_rate": 1.0643088598006174e-06,
+      "loss": 0.4439,
+      "step": 10120
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.7465037877129606,
+      "learning_rate": 1.0635245725336647e-06,
+      "loss": 0.5007,
+      "step": 10121
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.8449610142401933,
+      "learning_rate": 1.06274053995152e-06,
+      "loss": 0.4208,
+      "step": 10122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7391144333403765,
+      "learning_rate": 1.061956762104913e-06,
+      "loss": 0.4399,
+      "step": 10123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.74180599198221,
+      "learning_rate": 1.0611732390445534e-06,
+      "loss": 0.4471,
+      "step": 10124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.545775071966047,
+      "learning_rate": 1.0603899708211312e-06,
+      "loss": 0.4709,
+      "step": 10125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5968866404031512,
+      "learning_rate": 1.0596069574853246e-06,
+      "loss": 0.4572,
+      "step": 10126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8624107754856385,
+      "learning_rate": 1.0588241990877924e-06,
+      "loss": 0.4588,
+      "step": 10127
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 9.650270285665199,
+      "learning_rate": 1.0580416956791805e-06,
+      "loss": 0.4794,
+      "step": 10128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.595881396622812,
+      "learning_rate": 1.0572594473101134e-06,
+      "loss": 0.4619,
+      "step": 10129
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5992030755187427,
+      "learning_rate": 1.0564774540312016e-06,
+      "loss": 0.3955,
+      "step": 10130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2050110777391208,
+      "learning_rate": 1.0556957158930397e-06,
+      "loss": 0.4798,
+      "step": 10131
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8822566815302626,
+      "learning_rate": 1.054914232946207e-06,
+      "loss": 0.4292,
+      "step": 10132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8888032067142406,
+      "learning_rate": 1.0541330052412612e-06,
+      "loss": 0.3733,
+      "step": 10133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.122169954676104,
+      "learning_rate": 1.0533520328287466e-06,
+      "loss": 0.4389,
+      "step": 10134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5848357495276764,
+      "learning_rate": 1.0525713157591955e-06,
+      "loss": 0.4672,
+      "step": 10135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.7891813849447913,
+      "learning_rate": 1.0517908540831146e-06,
+      "loss": 0.4424,
+      "step": 10136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6355553467586057,
+      "learning_rate": 1.0510106478510006e-06,
+      "loss": 0.4461,
+      "step": 10137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7127352407495429,
+      "learning_rate": 1.050230697113332e-06,
+      "loss": 0.4776,
+      "step": 10138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5268855595120363,
+      "learning_rate": 1.0494510019205716e-06,
+      "loss": 0.4742,
+      "step": 10139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4560115757899093,
+      "learning_rate": 1.0486715623231625e-06,
+      "loss": 0.4095,
+      "step": 10140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8363434091402573,
+      "learning_rate": 1.047892378371534e-06,
+      "loss": 0.4253,
+      "step": 10141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.599107681893172,
+      "learning_rate": 1.0471134501160983e-06,
+      "loss": 0.4325,
+      "step": 10142
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.280237265088869,
+      "learning_rate": 1.046334777607253e-06,
+      "loss": 0.4547,
+      "step": 10143
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6280701462092735,
+      "learning_rate": 1.0455563608953738e-06,
+      "loss": 0.4477,
+      "step": 10144
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 7.5957231272663535,
+      "learning_rate": 1.0447782000308255e-06,
+      "loss": 0.486,
+      "step": 10145
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5661373243191992,
+      "learning_rate": 1.0440002950639533e-06,
+      "loss": 0.4782,
+      "step": 10146
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.116309319991897,
+      "learning_rate": 1.0432226460450874e-06,
+      "loss": 0.4346,
+      "step": 10147
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.563210296679415,
+      "learning_rate": 1.042445253024541e-06,
+      "loss": 0.4381,
+      "step": 10148
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4462553222118308,
+      "learning_rate": 1.0416681160526072e-06,
+      "loss": 0.4943,
+      "step": 10149
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5557270660788208,
+      "learning_rate": 1.0408912351795707e-06,
+      "loss": 0.4693,
+      "step": 10150
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5732995197259659,
+      "learning_rate": 1.0401146104556909e-06,
+      "loss": 0.4755,
+      "step": 10151
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5629590446885313,
+      "learning_rate": 1.0393382419312164e-06,
+      "loss": 0.4815,
+      "step": 10152
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0746227246568245,
+      "learning_rate": 1.0385621296563741e-06,
+      "loss": 0.4695,
+      "step": 10153
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5774929791279447,
+      "learning_rate": 1.0377862736813826e-06,
+      "loss": 0.4666,
+      "step": 10154
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.250722500552263,
+      "learning_rate": 1.037010674056435e-06,
+      "loss": 0.4436,
+      "step": 10155
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5348842727797053,
+      "learning_rate": 1.0362353308317135e-06,
+      "loss": 0.4692,
+      "step": 10156
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5830933933544387,
+      "learning_rate": 1.035460244057378e-06,
+      "loss": 0.4945,
+      "step": 10157
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8278123751738586,
+      "learning_rate": 1.03468541378358e-06,
+      "loss": 0.3926,
+      "step": 10158
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.703471935626578,
+      "learning_rate": 1.0339108400604497e-06,
+      "loss": 0.4468,
+      "step": 10159
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.800743670222422,
+      "learning_rate": 1.0331365229380986e-06,
+      "loss": 0.4373,
+      "step": 10160
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5186875366033328,
+      "learning_rate": 1.0323624624666246e-06,
+      "loss": 0.4571,
+      "step": 10161
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0680486511731706,
+      "learning_rate": 1.0315886586961094e-06,
+      "loss": 0.4938,
+      "step": 10162
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5702533126976862,
+      "learning_rate": 1.030815111676618e-06,
+      "loss": 0.4813,
+      "step": 10163
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8372242516366455,
+      "learning_rate": 1.0300418214581953e-06,
+      "loss": 0.473,
+      "step": 10164
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.3402608279790478,
+      "learning_rate": 1.029268788090873e-06,
+      "loss": 0.469,
+      "step": 10165
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5423028570717174,
+      "learning_rate": 1.0284960116246663e-06,
+      "loss": 0.485,
+      "step": 10166
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8477678663288564,
+      "learning_rate": 1.027723492109573e-06,
+      "loss": 0.4616,
+      "step": 10167
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5457813744984346,
+      "learning_rate": 1.0269512295955725e-06,
+      "loss": 0.4729,
+      "step": 10168
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2194939557053823,
+      "learning_rate": 1.0261792241326285e-06,
+      "loss": 0.3903,
+      "step": 10169
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.984197579332957,
+      "learning_rate": 1.0254074757706927e-06,
+      "loss": 0.4394,
+      "step": 10170
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5415309075118928,
+      "learning_rate": 1.0246359845596927e-06,
+      "loss": 0.4477,
+      "step": 10171
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0898133796663094,
+      "learning_rate": 1.023864750549544e-06,
+      "loss": 0.4675,
+      "step": 10172
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9707821038854907,
+      "learning_rate": 1.0230937737901447e-06,
+      "loss": 0.445,
+      "step": 10173
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.356601087943694,
+      "learning_rate": 1.0223230543313772e-06,
+      "loss": 0.4358,
+      "step": 10174
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.1414530842857524,
+      "learning_rate": 1.0215525922231028e-06,
+      "loss": 0.4515,
+      "step": 10175
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.755855608755755,
+      "learning_rate": 1.0207823875151718e-06,
+      "loss": 0.4537,
+      "step": 10176
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.3748645557419277,
+      "learning_rate": 1.0200124402574146e-06,
+      "loss": 0.4621,
+      "step": 10177
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8587984269672981,
+      "learning_rate": 1.0192427504996471e-06,
+      "loss": 0.4121,
+      "step": 10178
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2371725035879613,
+      "learning_rate": 1.018473318291665e-06,
+      "loss": 0.4422,
+      "step": 10179
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5643178196292167,
+      "learning_rate": 1.0177041436832508e-06,
+      "loss": 0.4394,
+      "step": 10180
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.76954558248805,
+      "learning_rate": 1.0169352267241694e-06,
+      "loss": 0.3916,
+      "step": 10181
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.8456602351273816,
+      "learning_rate": 1.016166567464168e-06,
+      "loss": 0.4585,
+      "step": 10182
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9360083198211315,
+      "learning_rate": 1.0153981659529793e-06,
+      "loss": 0.4816,
+      "step": 10183
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9887583504601396,
+      "learning_rate": 1.0146300222403139e-06,
+      "loss": 0.4551,
+      "step": 10184
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6136260607012085,
+      "learning_rate": 1.0138621363758755e-06,
+      "loss": 0.4433,
+      "step": 10185
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.386772755302714,
+      "learning_rate": 1.0130945084093412e-06,
+      "loss": 0.4134,
+      "step": 10186
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.630934718784969,
+      "learning_rate": 1.0123271383903776e-06,
+      "loss": 0.4558,
+      "step": 10187
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4249852265575416,
+      "learning_rate": 1.0115600263686292e-06,
+      "loss": 0.4352,
+      "step": 10188
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 14.472719227417778,
+      "learning_rate": 1.0107931723937326e-06,
+      "loss": 0.4985,
+      "step": 10189
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.563726940326743,
+      "learning_rate": 1.0100265765152973e-06,
+      "loss": 0.4142,
+      "step": 10190
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7536737593022746,
+      "learning_rate": 1.0092602387829231e-06,
+      "loss": 0.456,
+      "step": 10191
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.914078101950237,
+      "learning_rate": 1.008494159246191e-06,
+      "loss": 0.4663,
+      "step": 10192
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.69250129495352,
+      "learning_rate": 1.0077283379546653e-06,
+      "loss": 0.4502,
+      "step": 10193
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5617546777973564,
+      "learning_rate": 1.0069627749578946e-06,
+      "loss": 0.4516,
+      "step": 10194
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.535590358477303,
+      "learning_rate": 1.0061974703054078e-06,
+      "loss": 0.4522,
+      "step": 10195
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.812779821157032,
+      "learning_rate": 1.00543242404672e-06,
+      "loss": 0.4317,
+      "step": 10196
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7600752864028086,
+      "learning_rate": 1.004667636231329e-06,
+      "loss": 0.3929,
+      "step": 10197
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.543665390703115,
+      "learning_rate": 1.003903106908717e-06,
+      "loss": 0.4597,
+      "step": 10198
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.467794311687954,
+      "learning_rate": 1.0031388361283446e-06,
+      "loss": 0.4686,
+      "step": 10199
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.761914508718542,
+      "learning_rate": 1.0023748239396608e-06,
+      "loss": 0.4778,
+      "step": 10200
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5866404601386759,
+      "learning_rate": 1.0016110703920966e-06,
+      "loss": 0.4683,
+      "step": 10201
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8504312362624236,
+      "learning_rate": 1.0008475755350655e-06,
+      "loss": 0.4326,
+      "step": 10202
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7673290957430992,
+      "learning_rate": 1.000084339417966e-06,
+      "loss": 0.43,
+      "step": 10203
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4637852832878877,
+      "learning_rate": 9.993213620901748e-07,
+      "loss": 0.4636,
+      "step": 10204
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8306487204951127,
+      "learning_rate": 9.985586436010602e-07,
+      "loss": 0.4476,
+      "step": 10205
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5479792378137343,
+      "learning_rate": 9.977961839999656e-07,
+      "loss": 0.4287,
+      "step": 10206
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2335205284329724,
+      "learning_rate": 9.970339833362224e-07,
+      "loss": 0.4882,
+      "step": 10207
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7333382107062971,
+      "learning_rate": 9.962720416591443e-07,
+      "loss": 0.3799,
+      "step": 10208
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0320695475648916,
+      "learning_rate": 9.955103590180287e-07,
+      "loss": 0.4507,
+      "step": 10209
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5718613082461896,
+      "learning_rate": 9.947489354621525e-07,
+      "loss": 0.4728,
+      "step": 10210
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8458296128320457,
+      "learning_rate": 9.939877710407814e-07,
+      "loss": 0.501,
+      "step": 10211
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.175564821698887,
+      "learning_rate": 9.932268658031607e-07,
+      "loss": 0.4355,
+      "step": 10212
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.131333332923354,
+      "learning_rate": 9.924662197985212e-07,
+      "loss": 0.4889,
+      "step": 10213
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6126079770849477,
+      "learning_rate": 9.917058330760742e-07,
+      "loss": 0.4724,
+      "step": 10214
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.8113405055326264,
+      "learning_rate": 9.909457056850159e-07,
+      "loss": 0.4848,
+      "step": 10215
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0699596809126817,
+      "learning_rate": 9.901858376745254e-07,
+      "loss": 0.4372,
+      "step": 10216
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.057509902737468,
+      "learning_rate": 9.894262290937667e-07,
+      "loss": 0.4431,
+      "step": 10217
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.174193223974668,
+      "learning_rate": 9.886668799918853e-07,
+      "loss": 0.4692,
+      "step": 10218
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.042545122643649,
+      "learning_rate": 9.879077904180067e-07,
+      "loss": 0.4941,
+      "step": 10219
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6030123872528108,
+      "learning_rate": 9.87148960421248e-07,
+      "loss": 0.4786,
+      "step": 10220
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8409863573274394,
+      "learning_rate": 9.863903900507011e-07,
+      "loss": 0.4517,
+      "step": 10221
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5449606679771221,
+      "learning_rate": 9.856320793554463e-07,
+      "loss": 0.4546,
+      "step": 10222
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9986242920848671,
+      "learning_rate": 9.848740283845427e-07,
+      "loss": 0.44,
+      "step": 10223
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.54743276200625,
+      "learning_rate": 9.841162371870388e-07,
+      "loss": 0.4556,
+      "step": 10224
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8319330720045848,
+      "learning_rate": 9.833587058119603e-07,
+      "loss": 0.4388,
+      "step": 10225
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.677950802104717,
+      "learning_rate": 9.82601434308319e-07,
+      "loss": 0.4177,
+      "step": 10226
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5409169777748895,
+      "learning_rate": 9.81844422725109e-07,
+      "loss": 0.4715,
+      "step": 10227
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2197829296552825,
+      "learning_rate": 9.810876711113087e-07,
+      "loss": 0.3991,
+      "step": 10228
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.7928168081855287,
+      "learning_rate": 9.803311795158804e-07,
+      "loss": 0.4411,
+      "step": 10229
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6313863757789209,
+      "learning_rate": 9.795749479877647e-07,
+      "loss": 0.47,
+      "step": 10230
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9420718221522324,
+      "learning_rate": 9.788189765758904e-07,
+      "loss": 0.4694,
+      "step": 10231
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6796659715781856,
+      "learning_rate": 9.780632653291676e-07,
+      "loss": 0.4478,
+      "step": 10232
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.2581204677731908,
+      "learning_rate": 9.773078142964926e-07,
+      "loss": 0.4327,
+      "step": 10233
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8283624635012161,
+      "learning_rate": 9.765526235267375e-07,
+      "loss": 0.465,
+      "step": 10234
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6861377692948445,
+      "learning_rate": 9.757976930687645e-07,
+      "loss": 0.4414,
+      "step": 10235
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8839549032528675,
+      "learning_rate": 9.750430229714163e-07,
+      "loss": 0.4785,
+      "step": 10236
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.8219899020139274,
+      "learning_rate": 9.742886132835195e-07,
+      "loss": 0.4568,
+      "step": 10237
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.5542221458246024,
+      "learning_rate": 9.735344640538842e-07,
+      "loss": 0.4604,
+      "step": 10238
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5291885483800136,
+      "learning_rate": 9.727805753312998e-07,
+      "loss": 0.4669,
+      "step": 10239
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3028155652579894,
+      "learning_rate": 9.720269471645455e-07,
+      "loss": 0.4357,
+      "step": 10240
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.55399613064397,
+      "learning_rate": 9.712735796023782e-07,
+      "loss": 0.4161,
+      "step": 10241
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5789348097584666,
+      "learning_rate": 9.705204726935391e-07,
+      "loss": 0.4795,
+      "step": 10242
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6750220213970957,
+      "learning_rate": 9.69767626486755e-07,
+      "loss": 0.4721,
+      "step": 10243
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.4960982561925205,
+      "learning_rate": 9.690150410307342e-07,
+      "loss": 0.4055,
+      "step": 10244
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.002021275009819,
+      "learning_rate": 9.682627163741653e-07,
+      "loss": 0.4747,
+      "step": 10245
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 3.452102472246556,
+      "learning_rate": 9.675106525657252e-07,
+      "loss": 0.454,
+      "step": 10246
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5461733158292336,
+      "learning_rate": 9.667588496540703e-07,
+      "loss": 0.5022,
+      "step": 10247
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.0046401364207433,
+      "learning_rate": 9.660073076878419e-07,
+      "loss": 0.4587,
+      "step": 10248
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.884001614234772,
+      "learning_rate": 9.652560267156647e-07,
+      "loss": 0.4558,
+      "step": 10249
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.2975346515012585,
+      "learning_rate": 9.645050067861433e-07,
+      "loss": 0.482,
+      "step": 10250
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4077653826448504,
+      "learning_rate": 9.63754247947869e-07,
+      "loss": 0.4467,
+      "step": 10251
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7385819549657786,
+      "learning_rate": 9.630037502494143e-07,
+      "loss": 0.5104,
+      "step": 10252
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8809381461273962,
+      "learning_rate": 9.62253513739338e-07,
+      "loss": 0.4465,
+      "step": 10253
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9198252244367415,
+      "learning_rate": 9.615035384661743e-07,
+      "loss": 0.4409,
+      "step": 10254
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.571633584458511,
+      "learning_rate": 9.60753824478451e-07,
+      "loss": 0.4721,
+      "step": 10255
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6279471394407021,
+      "learning_rate": 9.600043718246704e-07,
+      "loss": 0.4298,
+      "step": 10256
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5252276430729529,
+      "learning_rate": 9.59255180553323e-07,
+      "loss": 0.483,
+      "step": 10257
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7938880903349763,
+      "learning_rate": 9.585062507128767e-07,
+      "loss": 0.4777,
+      "step": 10258
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5651769452791178,
+      "learning_rate": 9.577575823517904e-07,
+      "loss": 0.454,
+      "step": 10259
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2160556900538735,
+      "learning_rate": 9.570091755185024e-07,
+      "loss": 0.4307,
+      "step": 10260
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5279663375566758,
+      "learning_rate": 9.5626103026143e-07,
+      "loss": 0.4615,
+      "step": 10261
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8164623997303837,
+      "learning_rate": 9.555131466289792e-07,
+      "loss": 0.4442,
+      "step": 10262
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0924895469259237,
+      "learning_rate": 9.54765524669537e-07,
+      "loss": 0.4184,
+      "step": 10263
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.130855503962928,
+      "learning_rate": 9.54018164431475e-07,
+      "loss": 0.4276,
+      "step": 10264
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3251020690627806,
+      "learning_rate": 9.532710659631434e-07,
+      "loss": 0.4154,
+      "step": 10265
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0217996353995815,
+      "learning_rate": 9.5252422931288e-07,
+      "loss": 0.4511,
+      "step": 10266
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1967683957335824,
+      "learning_rate": 9.517776545290041e-07,
+      "loss": 0.4048,
+      "step": 10267
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9186375248370817,
+      "learning_rate": 9.510313416598199e-07,
+      "loss": 0.4508,
+      "step": 10268
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.486553294585594,
+      "learning_rate": 9.502852907536098e-07,
+      "loss": 0.5231,
+      "step": 10269
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8377520954849014,
+      "learning_rate": 9.495395018586423e-07,
+      "loss": 0.4577,
+      "step": 10270
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8711266526113088,
+      "learning_rate": 9.487939750231728e-07,
+      "loss": 0.4018,
+      "step": 10271
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4857814936425724,
+      "learning_rate": 9.480487102954322e-07,
+      "loss": 0.4872,
+      "step": 10272
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.555337572606532,
+      "learning_rate": 9.473037077236413e-07,
+      "loss": 0.4755,
+      "step": 10273
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2174744059355738,
+      "learning_rate": 9.465589673559955e-07,
+      "loss": 0.424,
+      "step": 10274
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.118347666019394,
+      "learning_rate": 9.458144892406851e-07,
+      "loss": 0.4992,
+      "step": 10275
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5432466175226598,
+      "learning_rate": 9.450702734258721e-07,
+      "loss": 0.463,
+      "step": 10276
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.553520440288294,
+      "learning_rate": 9.44326319959708e-07,
+      "loss": 0.4794,
+      "step": 10277
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9463426116073292,
+      "learning_rate": 9.435826288903255e-07,
+      "loss": 0.4552,
+      "step": 10278
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.213634467337301,
+      "learning_rate": 9.42839200265842e-07,
+      "loss": 0.4496,
+      "step": 10279
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7451035748962336,
+      "learning_rate": 9.420960341343533e-07,
+      "loss": 0.5214,
+      "step": 10280
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 4.978678131634297,
+      "learning_rate": 9.413531305439428e-07,
+      "loss": 0.4649,
+      "step": 10281
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5653734423112623,
+      "learning_rate": 9.406104895426755e-07,
+      "loss": 0.4623,
+      "step": 10282
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.705521175647822,
+      "learning_rate": 9.39868111178599e-07,
+      "loss": 0.4061,
+      "step": 10283
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.249732587833357,
+      "learning_rate": 9.391259954997456e-07,
+      "loss": 0.4423,
+      "step": 10284
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7557518466595086,
+      "learning_rate": 9.383841425541268e-07,
+      "loss": 0.4089,
+      "step": 10285
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7884258835580211,
+      "learning_rate": 9.376425523897408e-07,
+      "loss": 0.469,
+      "step": 10286
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8110514647561224,
+      "learning_rate": 9.369012250545672e-07,
+      "loss": 0.475,
+      "step": 10287
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5295637057735154,
+      "learning_rate": 9.361601605965708e-07,
+      "loss": 0.4674,
+      "step": 10288
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5959763283997073,
+      "learning_rate": 9.354193590636945e-07,
+      "loss": 0.4581,
+      "step": 10289
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4084902483424653,
+      "learning_rate": 9.346788205038682e-07,
+      "loss": 0.46,
+      "step": 10290
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0642615119811527,
+      "learning_rate": 9.33938544965004e-07,
+      "loss": 0.4268,
+      "step": 10291
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9193524895874488,
+      "learning_rate": 9.331985324949988e-07,
+      "loss": 0.3976,
+      "step": 10292
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7494483074184157,
+      "learning_rate": 9.324587831417253e-07,
+      "loss": 0.4731,
+      "step": 10293
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0953088980565147,
+      "learning_rate": 9.317192969530492e-07,
+      "loss": 0.4876,
+      "step": 10294
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.59158419850571,
+      "learning_rate": 9.309800739768137e-07,
+      "loss": 0.4666,
+      "step": 10295
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3949582223503207,
+      "learning_rate": 9.302411142608431e-07,
+      "loss": 0.4396,
+      "step": 10296
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 6.514271748395274,
+      "learning_rate": 9.295024178529488e-07,
+      "loss": 0.406,
+      "step": 10297
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.1462051896753236,
+      "learning_rate": 9.287639848009228e-07,
+      "loss": 0.4275,
+      "step": 10298
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6638335972914338,
+      "learning_rate": 9.280258151525429e-07,
+      "loss": 0.4812,
+      "step": 10299
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5485263012621387,
+      "learning_rate": 9.272879089555642e-07,
+      "loss": 0.4746,
+      "step": 10300
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4822644987320721,
+      "learning_rate": 9.265502662577303e-07,
+      "loss": 0.4342,
+      "step": 10301
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.3107608682875815,
+      "learning_rate": 9.258128871067651e-07,
+      "loss": 0.4548,
+      "step": 10302
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0922364346486573,
+      "learning_rate": 9.250757715503783e-07,
+      "loss": 0.4197,
+      "step": 10303
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.4410525414933155,
+      "learning_rate": 9.243389196362568e-07,
+      "loss": 0.4901,
+      "step": 10304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7577238296997306,
+      "learning_rate": 9.23602331412074e-07,
+      "loss": 0.4418,
+      "step": 10305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.572376132149254,
+      "learning_rate": 9.228660069254908e-07,
+      "loss": 0.4062,
+      "step": 10306
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5418114903910072,
+      "learning_rate": 9.221299462241418e-07,
+      "loss": 0.4765,
+      "step": 10307
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5789479585277342,
+      "learning_rate": 9.213941493556522e-07,
+      "loss": 0.4395,
+      "step": 10308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.6820591129522606,
+      "learning_rate": 9.206586163676228e-07,
+      "loss": 0.4351,
+      "step": 10309
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.199853441489854,
+      "learning_rate": 9.199233473076474e-07,
+      "loss": 0.4287,
+      "step": 10310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1841568209031816,
+      "learning_rate": 9.191883422232923e-07,
+      "loss": 0.4498,
+      "step": 10311
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.32579340643818,
+      "learning_rate": 9.184536011621131e-07,
+      "loss": 0.4044,
+      "step": 10312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7108584612237394,
+      "learning_rate": 9.177191241716471e-07,
+      "loss": 0.4885,
+      "step": 10313
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5635000937771477,
+      "learning_rate": 9.169849112994145e-07,
+      "loss": 0.4314,
+      "step": 10314
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8802450056169515,
+      "learning_rate": 9.162509625929156e-07,
+      "loss": 0.483,
+      "step": 10315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5880258198498999,
+      "learning_rate": 9.155172780996369e-07,
+      "loss": 0.4836,
+      "step": 10316
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.942595668792461,
+      "learning_rate": 9.147838578670476e-07,
+      "loss": 0.4803,
+      "step": 10317
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9728958802542869,
+      "learning_rate": 9.140507019425981e-07,
+      "loss": 0.4667,
+      "step": 10318
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5643636428910844,
+      "learning_rate": 9.13317810373725e-07,
+      "loss": 0.4786,
+      "step": 10319
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5421649255762295,
+      "learning_rate": 9.12585183207842e-07,
+      "loss": 0.4662,
+      "step": 10320
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5524235680919116,
+      "learning_rate": 9.118528204923505e-07,
+      "loss": 0.4632,
+      "step": 10321
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7955650505880107,
+      "learning_rate": 9.111207222746332e-07,
+      "loss": 0.4266,
+      "step": 10322
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.7307945898360115,
+      "learning_rate": 9.103888886020579e-07,
+      "loss": 0.4036,
+      "step": 10323
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6035151461432573,
+      "learning_rate": 9.096573195219704e-07,
+      "loss": 0.4487,
+      "step": 10324
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.1929106941806102,
+      "learning_rate": 9.089260150817037e-07,
+      "loss": 0.4624,
+      "step": 10325
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5660181757846314,
+      "learning_rate": 9.081949753285718e-07,
+      "loss": 0.4302,
+      "step": 10326
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5867425608741932,
+      "learning_rate": 9.074642003098721e-07,
+      "loss": 0.438,
+      "step": 10327
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9704425104710601,
+      "learning_rate": 9.067336900728846e-07,
+      "loss": 0.4602,
+      "step": 10328
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5073233971025206,
+      "learning_rate": 9.060034446648735e-07,
+      "loss": 0.4338,
+      "step": 10329
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.084706068950284,
+      "learning_rate": 9.052734641330846e-07,
+      "loss": 0.4191,
+      "step": 10330
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.771881004230732,
+      "learning_rate": 9.045437485247449e-07,
+      "loss": 0.3819,
+      "step": 10331
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.622861835942293,
+      "learning_rate": 9.038142978870673e-07,
+      "loss": 0.4189,
+      "step": 10332
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5937708279263001,
+      "learning_rate": 9.030851122672457e-07,
+      "loss": 0.4857,
+      "step": 10333
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.5642762013169103,
+      "learning_rate": 9.023561917124596e-07,
+      "loss": 0.5107,
+      "step": 10334
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.87685458201007,
+      "learning_rate": 9.016275362698662e-07,
+      "loss": 0.4746,
+      "step": 10335
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0194265904069972,
+      "learning_rate": 9.008991459866096e-07,
+      "loss": 0.4218,
+      "step": 10336
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6916732363421423,
+      "learning_rate": 9.001710209098163e-07,
+      "loss": 0.4287,
+      "step": 10337
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5632858739406676,
+      "learning_rate": 8.994431610865945e-07,
+      "loss": 0.4771,
+      "step": 10338
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5283897132872917,
+      "learning_rate": 8.987155665640368e-07,
+      "loss": 0.4633,
+      "step": 10339
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2585277428787034,
+      "learning_rate": 8.979882373892152e-07,
+      "loss": 0.4379,
+      "step": 10340
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.59364825262942,
+      "learning_rate": 8.972611736091902e-07,
+      "loss": 0.4406,
+      "step": 10341
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.599927625327477,
+      "learning_rate": 8.965343752709993e-07,
+      "loss": 0.3911,
+      "step": 10342
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6899787358683087,
+      "learning_rate": 8.958078424216676e-07,
+      "loss": 0.4299,
+      "step": 10343
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9168804904398595,
+      "learning_rate": 8.95081575108197e-07,
+      "loss": 0.4807,
+      "step": 10344
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.069798021302125,
+      "learning_rate": 8.943555733775811e-07,
+      "loss": 0.425,
+      "step": 10345
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3943004014279132,
+      "learning_rate": 8.936298372767876e-07,
+      "loss": 0.3952,
+      "step": 10346
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9360949388802995,
+      "learning_rate": 8.929043668527715e-07,
+      "loss": 0.3859,
+      "step": 10347
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5787392059588699,
+      "learning_rate": 8.921791621524706e-07,
+      "loss": 0.465,
+      "step": 10348
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.2700749449816957,
+      "learning_rate": 8.914542232228041e-07,
+      "loss": 0.4674,
+      "step": 10349
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5973971838846036,
+      "learning_rate": 8.907295501106755e-07,
+      "loss": 0.4775,
+      "step": 10350
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.810281379395451,
+      "learning_rate": 8.900051428629686e-07,
+      "loss": 0.4809,
+      "step": 10351
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7596631884673937,
+      "learning_rate": 8.892810015265518e-07,
+      "loss": 0.4356,
+      "step": 10352
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.383423820983331,
+      "learning_rate": 8.885571261482773e-07,
+      "loss": 0.4504,
+      "step": 10353
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.5293575036812714,
+      "learning_rate": 8.878335167749797e-07,
+      "loss": 0.4547,
+      "step": 10354
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5329948584197477,
+      "learning_rate": 8.871101734534726e-07,
+      "loss": 0.4444,
+      "step": 10355
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.0164031748314137,
+      "learning_rate": 8.863870962305571e-07,
+      "loss": 0.4541,
+      "step": 10356
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.9878318176941,
+      "learning_rate": 8.856642851530145e-07,
+      "loss": 0.4141,
+      "step": 10357
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5456643252558189,
+      "learning_rate": 8.849417402676124e-07,
+      "loss": 0.4596,
+      "step": 10358
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.8861312861035753,
+      "learning_rate": 8.842194616210953e-07,
+      "loss": 0.4527,
+      "step": 10359
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5713905447653566,
+      "learning_rate": 8.834974492601933e-07,
+      "loss": 0.4632,
+      "step": 10360
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4134592203611467,
+      "learning_rate": 8.827757032316242e-07,
+      "loss": 0.3976,
+      "step": 10361
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.4968492898153474,
+      "learning_rate": 8.820542235820795e-07,
+      "loss": 0.4906,
+      "step": 10362
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 3.0828079652407085,
+      "learning_rate": 8.813330103582396e-07,
+      "loss": 0.463,
+      "step": 10363
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.705666856115678,
+      "learning_rate": 8.806120636067661e-07,
+      "loss": 0.4246,
+      "step": 10364
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.518197682824045,
+      "learning_rate": 8.798913833743045e-07,
+      "loss": 0.4563,
+      "step": 10365
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5171383302517183,
+      "learning_rate": 8.791709697074791e-07,
+      "loss": 0.4838,
+      "step": 10366
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.550740650647119,
+      "learning_rate": 8.784508226529015e-07,
+      "loss": 0.3847,
+      "step": 10367
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6963322196408366,
+      "learning_rate": 8.77730942257164e-07,
+      "loss": 0.4996,
+      "step": 10368
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7309523301521403,
+      "learning_rate": 8.770113285668436e-07,
+      "loss": 0.4089,
+      "step": 10369
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6586128791589392,
+      "learning_rate": 8.762919816284949e-07,
+      "loss": 0.37,
+      "step": 10370
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7656722472391493,
+      "learning_rate": 8.755729014886605e-07,
+      "loss": 0.5047,
+      "step": 10371
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5611340777647948,
+      "learning_rate": 8.748540881938644e-07,
+      "loss": 0.4674,
+      "step": 10372
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.969783835894969,
+      "learning_rate": 8.741355417906116e-07,
+      "loss": 0.4201,
+      "step": 10373
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.9939270554162345,
+      "learning_rate": 8.73417262325394e-07,
+      "loss": 0.4472,
+      "step": 10374
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.854829268638572,
+      "learning_rate": 8.726992498446785e-07,
+      "loss": 0.4519,
+      "step": 10375
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 2.349155445346147,
+      "learning_rate": 8.719815043949248e-07,
+      "loss": 0.4784,
+      "step": 10376
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 5.8306678559264045,
+      "learning_rate": 8.712640260225663e-07,
+      "loss": 0.4709,
+      "step": 10377
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.934961683327888,
+      "learning_rate": 8.70546814774026e-07,
+      "loss": 0.4467,
+      "step": 10378
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6903203542219105,
+      "learning_rate": 8.698298706957015e-07,
+      "loss": 0.4692,
+      "step": 10379
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7742561693359535,
+      "learning_rate": 8.691131938339842e-07,
+      "loss": 0.4619,
+      "step": 10380
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6538450090094314,
+      "learning_rate": 8.683967842352386e-07,
+      "loss": 0.5057,
+      "step": 10381
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.821025368120122,
+      "learning_rate": 8.676806419458156e-07,
+      "loss": 0.4166,
+      "step": 10382
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3437838864374085,
+      "learning_rate": 8.669647670120496e-07,
+      "loss": 0.436,
+      "step": 10383
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.3016530801643373,
+      "learning_rate": 8.662491594802563e-07,
+      "loss": 0.4759,
+      "step": 10384
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.343888290083774,
+      "learning_rate": 8.655338193967367e-07,
+      "loss": 0.4342,
+      "step": 10385
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5671307345206104,
+      "learning_rate": 8.648187468077684e-07,
+      "loss": 0.4856,
+      "step": 10386
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.809300376415277,
+      "learning_rate": 8.641039417596181e-07,
+      "loss": 0.4111,
+      "step": 10387
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7937873842626946,
+      "learning_rate": 8.63389404298532e-07,
+      "loss": 0.4076,
+      "step": 10388
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.536953704284807,
+      "learning_rate": 8.626751344707418e-07,
+      "loss": 0.4688,
+      "step": 10389
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1994912871828878,
+      "learning_rate": 8.619611323224563e-07,
+      "loss": 0.4205,
+      "step": 10390
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4989333246230003,
+      "learning_rate": 8.612473978998726e-07,
+      "loss": 0.4447,
+      "step": 10391
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.709582729769975,
+      "learning_rate": 8.605339312491679e-07,
+      "loss": 0.4478,
+      "step": 10392
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5257103068064729,
+      "learning_rate": 8.598207324165042e-07,
+      "loss": 0.4802,
+      "step": 10393
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.548270998571477,
+      "learning_rate": 8.59107801448022e-07,
+      "loss": 0.4655,
+      "step": 10394
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6897082800647134,
+      "learning_rate": 8.583951383898464e-07,
+      "loss": 0.4818,
+      "step": 10395
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8162914781800854,
+      "learning_rate": 8.576827432880902e-07,
+      "loss": 0.4079,
+      "step": 10396
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5219332710308813,
+      "learning_rate": 8.569706161888402e-07,
+      "loss": 0.4292,
+      "step": 10397
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.355615640721105,
+      "learning_rate": 8.56258757138172e-07,
+      "loss": 0.4518,
+      "step": 10398
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8559393422617085,
+      "learning_rate": 8.555471661821413e-07,
+      "loss": 0.4749,
+      "step": 10399
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.511759368637973,
+      "learning_rate": 8.548358433667886e-07,
+      "loss": 0.4598,
+      "step": 10400
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8516738310912404,
+      "learning_rate": 8.541247887381326e-07,
+      "loss": 0.4902,
+      "step": 10401
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7290323042681175,
+      "learning_rate": 8.5341400234218e-07,
+      "loss": 0.5068,
+      "step": 10402
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4562906546670134,
+      "learning_rate": 8.527034842249165e-07,
+      "loss": 0.426,
+      "step": 10403
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7327589930599883,
+      "learning_rate": 8.519932344323139e-07,
+      "loss": 0.4195,
+      "step": 10404
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4249916623598478,
+      "learning_rate": 8.51283253010321e-07,
+      "loss": 0.4659,
+      "step": 10405
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5381119532301166,
+      "learning_rate": 8.505735400048748e-07,
+      "loss": 0.4634,
+      "step": 10406
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7241813810346882,
+      "learning_rate": 8.498640954618926e-07,
+      "loss": 0.4035,
+      "step": 10407
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5900907041544682,
+      "learning_rate": 8.491549194272736e-07,
+      "loss": 0.4742,
+      "step": 10408
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5392694820345831,
+      "learning_rate": 8.484460119469035e-07,
+      "loss": 0.4233,
+      "step": 10409
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.798576735095686,
+      "learning_rate": 8.477373730666422e-07,
+      "loss": 0.4662,
+      "step": 10410
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.4661322394029295,
+      "learning_rate": 8.470290028323441e-07,
+      "loss": 0.439,
+      "step": 10411
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.05340044751119,
+      "learning_rate": 8.463209012898349e-07,
+      "loss": 0.5051,
+      "step": 10412
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.541423998263109,
+      "learning_rate": 8.456130684849317e-07,
+      "loss": 0.4881,
+      "step": 10413
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.449413169031657,
+      "learning_rate": 8.449055044634252e-07,
+      "loss": 0.4523,
+      "step": 10414
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.611654802954893,
+      "learning_rate": 8.441982092710998e-07,
+      "loss": 0.4483,
+      "step": 10415
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.5398940613810765,
+      "learning_rate": 8.434911829537129e-07,
+      "loss": 0.4558,
+      "step": 10416
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4587384967007637,
+      "learning_rate": 8.427844255570084e-07,
+      "loss": 0.4266,
+      "step": 10417
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.041512412770492,
+      "learning_rate": 8.420779371267134e-07,
+      "loss": 0.4424,
+      "step": 10418
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6879387679801476,
+      "learning_rate": 8.413717177085367e-07,
+      "loss": 0.4453,
+      "step": 10419
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0737336203807497,
+      "learning_rate": 8.406657673481711e-07,
+      "loss": 0.4025,
+      "step": 10420
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.6323107702586803,
+      "learning_rate": 8.399600860912882e-07,
+      "loss": 0.43,
+      "step": 10421
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 11.93673320629374,
+      "learning_rate": 8.39254673983545e-07,
+      "loss": 0.4602,
+      "step": 10422
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7602673856759,
+      "learning_rate": 8.385495310705821e-07,
+      "loss": 0.466,
+      "step": 10423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.199525647565027,
+      "learning_rate": 8.378446573980215e-07,
+      "loss": 0.4203,
+      "step": 10424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8172508039527577,
+      "learning_rate": 8.37140053011466e-07,
+      "loss": 0.4749,
+      "step": 10425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.555600705663999,
+      "learning_rate": 8.36435717956503e-07,
+      "loss": 0.4487,
+      "step": 10426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5600937912843675,
+      "learning_rate": 8.357316522787024e-07,
+      "loss": 0.4818,
+      "step": 10427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.564108326374525,
+      "learning_rate": 8.35027856023617e-07,
+      "loss": 0.4336,
+      "step": 10428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8942923099274132,
+      "learning_rate": 8.343243292367814e-07,
+      "loss": 0.5217,
+      "step": 10429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5478950951530912,
+      "learning_rate": 8.3362107196371e-07,
+      "loss": 0.4748,
+      "step": 10430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.7064674315709056,
+      "learning_rate": 8.329180842499074e-07,
+      "loss": 0.4475,
+      "step": 10431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.417799205265473,
+      "learning_rate": 8.322153661408522e-07,
+      "loss": 0.4613,
+      "step": 10432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5397847607225453,
+      "learning_rate": 8.315129176820108e-07,
+      "loss": 0.4745,
+      "step": 10433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5819000931362215,
+      "learning_rate": 8.308107389188303e-07,
+      "loss": 0.4834,
+      "step": 10434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.067798031180775,
+      "learning_rate": 8.301088298967425e-07,
+      "loss": 0.4475,
+      "step": 10435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.584377177634961,
+      "learning_rate": 8.294071906611573e-07,
+      "loss": 0.4604,
+      "step": 10436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 4.267941992448607,
+      "learning_rate": 8.287058212574712e-07,
+      "loss": 0.4429,
+      "step": 10437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.66659983254258,
+      "learning_rate": 8.28004721731061e-07,
+      "loss": 0.4247,
+      "step": 10438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5627511120482244,
+      "learning_rate": 8.273038921272897e-07,
+      "loss": 0.4869,
+      "step": 10439
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.8690265461064324,
+      "learning_rate": 8.266033324914962e-07,
+      "loss": 0.4462,
+      "step": 10440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6310837710931052,
+      "learning_rate": 8.259030428690079e-07,
+      "loss": 0.3962,
+      "step": 10441
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7576469164442905,
+      "learning_rate": 8.252030233051322e-07,
+      "loss": 0.4488,
+      "step": 10442
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0079198803829397,
+      "learning_rate": 8.245032738451598e-07,
+      "loss": 0.4568,
+      "step": 10443
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9395214021196525,
+      "learning_rate": 8.238037945343647e-07,
+      "loss": 0.4214,
+      "step": 10444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7831455162113885,
+      "learning_rate": 8.231045854179981e-07,
+      "loss": 0.4366,
+      "step": 10445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.565411501852786,
+      "learning_rate": 8.224056465413033e-07,
+      "loss": 0.4624,
+      "step": 10446
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1263394985690365,
+      "learning_rate": 8.217069779494968e-07,
+      "loss": 0.4622,
+      "step": 10447
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5564967220278327,
+      "learning_rate": 8.210085796877842e-07,
+      "loss": 0.454,
+      "step": 10448
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8633097240743695,
+      "learning_rate": 8.203104518013477e-07,
+      "loss": 0.5128,
+      "step": 10449
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6792719783856753,
+      "learning_rate": 8.196125943353594e-07,
+      "loss": 0.4169,
+      "step": 10450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.193644710044736,
+      "learning_rate": 8.189150073349661e-07,
+      "loss": 0.4513,
+      "step": 10451
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5672486236560986,
+      "learning_rate": 8.182176908453026e-07,
+      "loss": 0.4728,
+      "step": 10452
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4385328704938103,
+      "learning_rate": 8.175206449114841e-07,
+      "loss": 0.3883,
+      "step": 10453
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7665782052965864,
+      "learning_rate": 8.168238695786085e-07,
+      "loss": 0.4588,
+      "step": 10454
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1037436145980655,
+      "learning_rate": 8.161273648917573e-07,
+      "loss": 0.3922,
+      "step": 10455
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0230891308927696,
+      "learning_rate": 8.154311308959911e-07,
+      "loss": 0.4362,
+      "step": 10456
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.1669630473543564,
+      "learning_rate": 8.147351676363569e-07,
+      "loss": 0.4574,
+      "step": 10457
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7340495444242308,
+      "learning_rate": 8.140394751578818e-07,
+      "loss": 0.4766,
+      "step": 10458
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.541187111779539,
+      "learning_rate": 8.133440535055787e-07,
+      "loss": 0.467,
+      "step": 10459
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7945687677158546,
+      "learning_rate": 8.126489027244367e-07,
+      "loss": 0.4707,
+      "step": 10460
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5948763989509176,
+      "learning_rate": 8.119540228594331e-07,
+      "loss": 0.4514,
+      "step": 10461
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0623891789263,
+      "learning_rate": 8.112594139555258e-07,
+      "loss": 0.4628,
+      "step": 10462
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5384214814025952,
+      "learning_rate": 8.105650760576544e-07,
+      "loss": 0.4185,
+      "step": 10463
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0423502272740617,
+      "learning_rate": 8.098710092107436e-07,
+      "loss": 0.4325,
+      "step": 10464
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.702679491164323,
+      "learning_rate": 8.091772134596953e-07,
+      "loss": 0.416,
+      "step": 10465
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.625813032010719,
+      "learning_rate": 8.084836888494008e-07,
+      "loss": 0.4229,
+      "step": 10466
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9260244142622303,
+      "learning_rate": 8.077904354247274e-07,
+      "loss": 0.4503,
+      "step": 10467
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6132423590534624,
+      "learning_rate": 8.070974532305281e-07,
+      "loss": 0.4557,
+      "step": 10468
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 6.324202241150519,
+      "learning_rate": 8.064047423116395e-07,
+      "loss": 0.448,
+      "step": 10469
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.664319176533905,
+      "learning_rate": 8.057123027128788e-07,
+      "loss": 0.4788,
+      "step": 10470
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5482131992208511,
+      "learning_rate": 8.050201344790442e-07,
+      "loss": 0.4679,
+      "step": 10471
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.775362269388806,
+      "learning_rate": 8.043282376549194e-07,
+      "loss": 0.5079,
+      "step": 10472
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0416613657455334,
+      "learning_rate": 8.036366122852685e-07,
+      "loss": 0.4902,
+      "step": 10473
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5580741294411549,
+      "learning_rate": 8.029452584148395e-07,
+      "loss": 0.4381,
+      "step": 10474
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4791108179839572,
+      "learning_rate": 8.022541760883635e-07,
+      "loss": 0.4572,
+      "step": 10475
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9183146545131848,
+      "learning_rate": 8.015633653505494e-07,
+      "loss": 0.463,
+      "step": 10476
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9966746415375298,
+      "learning_rate": 8.00872826246093e-07,
+      "loss": 0.3996,
+      "step": 10477
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5423296937673391,
+      "learning_rate": 8.001825588196721e-07,
+      "loss": 0.4705,
+      "step": 10478
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.409019202988364,
+      "learning_rate": 7.994925631159467e-07,
+      "loss": 0.3938,
+      "step": 10479
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.658614317815432,
+      "learning_rate": 7.988028391795549e-07,
+      "loss": 0.4145,
+      "step": 10480
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.548083257395584,
+      "learning_rate": 7.981133870551261e-07,
+      "loss": 0.4779,
+      "step": 10481
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6787045973936443,
+      "learning_rate": 7.974242067872628e-07,
+      "loss": 0.4615,
+      "step": 10482
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6661808469186754,
+      "learning_rate": 7.967352984205573e-07,
+      "loss": 0.4011,
+      "step": 10483
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.441615375179931,
+      "learning_rate": 7.960466619995772e-07,
+      "loss": 0.4184,
+      "step": 10484
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.93345218590861,
+      "learning_rate": 7.953582975688795e-07,
+      "loss": 0.4368,
+      "step": 10485
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7915889610293307,
+      "learning_rate": 7.946702051730005e-07,
+      "loss": 0.4641,
+      "step": 10486
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4432525914788856,
+      "learning_rate": 7.93982384856457e-07,
+      "loss": 0.4187,
+      "step": 10487
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6832473353779829,
+      "learning_rate": 7.932948366637516e-07,
+      "loss": 0.421,
+      "step": 10488
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.840490557424005,
+      "learning_rate": 7.926075606393662e-07,
+      "loss": 0.5027,
+      "step": 10489
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5423022956824887,
+      "learning_rate": 7.9192055682777e-07,
+      "loss": 0.4685,
+      "step": 10490
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6758603850778215,
+      "learning_rate": 7.91233825273407e-07,
+      "loss": 0.5067,
+      "step": 10491
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.779663489257076,
+      "learning_rate": 7.905473660207097e-07,
+      "loss": 0.4588,
+      "step": 10492
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9660009401909029,
+      "learning_rate": 7.898611791140915e-07,
+      "loss": 0.4682,
+      "step": 10493
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.9830008667609105,
+      "learning_rate": 7.891752645979484e-07,
+      "loss": 0.4845,
+      "step": 10494
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.082095451632507,
+      "learning_rate": 7.884896225166561e-07,
+      "loss": 0.4598,
+      "step": 10495
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8006279290736278,
+      "learning_rate": 7.878042529145746e-07,
+      "loss": 0.417,
+      "step": 10496
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.941046344257525,
+      "learning_rate": 7.871191558360503e-07,
+      "loss": 0.5308,
+      "step": 10497
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6136279093325179,
+      "learning_rate": 7.864343313254041e-07,
+      "loss": 0.4389,
+      "step": 10498
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.523167780875091,
+      "learning_rate": 7.857497794269459e-07,
+      "loss": 0.4443,
+      "step": 10499
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8694191145655241,
+      "learning_rate": 7.850655001849616e-07,
+      "loss": 0.4757,
+      "step": 10500
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.0080366411927364,
+      "learning_rate": 7.843814936437277e-07,
+      "loss": 0.5011,
+      "step": 10501
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 5.063271031104307,
+      "learning_rate": 7.836977598474955e-07,
+      "loss": 0.4765,
+      "step": 10502
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.7384229019369908,
+      "learning_rate": 7.830142988405026e-07,
+      "loss": 0.4573,
+      "step": 10503
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5509632331126995,
+      "learning_rate": 7.823311106669685e-07,
+      "loss": 0.4835,
+      "step": 10504
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9171296323109506,
+      "learning_rate": 7.816481953710947e-07,
+      "loss": 0.4022,
+      "step": 10505
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3993772635970991,
+      "learning_rate": 7.809655529970633e-07,
+      "loss": 0.4226,
+      "step": 10506
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.410184111225007,
+      "learning_rate": 7.802831835890418e-07,
+      "loss": 0.4142,
+      "step": 10507
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5820284432433327,
+      "learning_rate": 7.79601087191178e-07,
+      "loss": 0.4665,
+      "step": 10508
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5314403847116487,
+      "learning_rate": 7.789192638476029e-07,
+      "loss": 0.4649,
+      "step": 10509
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5282810007839512,
+      "learning_rate": 7.78237713602431e-07,
+      "loss": 0.4705,
+      "step": 10510
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1412412354295554,
+      "learning_rate": 7.775564364997545e-07,
+      "loss": 0.4004,
+      "step": 10511
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0842889626919487,
+      "learning_rate": 7.768754325836531e-07,
+      "loss": 0.4309,
+      "step": 10512
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.091201168656958,
+      "learning_rate": 7.761947018981869e-07,
+      "loss": 0.4496,
+      "step": 10513
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6254376162901858,
+      "learning_rate": 7.755142444873992e-07,
+      "loss": 0.4252,
+      "step": 10514
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8434726523973886,
+      "learning_rate": 7.748340603953114e-07,
+      "loss": 0.4268,
+      "step": 10515
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5628724128854636,
+      "learning_rate": 7.741541496659344e-07,
+      "loss": 0.4491,
+      "step": 10516
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.178307691818149,
+      "learning_rate": 7.734745123432552e-07,
+      "loss": 0.4543,
+      "step": 10517
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.5852885678103217,
+      "learning_rate": 7.72795148471247e-07,
+      "loss": 0.4738,
+      "step": 10518
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0463769935032365,
+      "learning_rate": 7.721160580938603e-07,
+      "loss": 0.4684,
+      "step": 10519
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9538472192307197,
+      "learning_rate": 7.714372412550353e-07,
+      "loss": 0.4667,
+      "step": 10520
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.305163123853526,
+      "learning_rate": 7.707586979986903e-07,
+      "loss": 0.4575,
+      "step": 10521
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.0648033528671728,
+      "learning_rate": 7.700804283687241e-07,
+      "loss": 0.4717,
+      "step": 10522
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.209266467059784,
+      "learning_rate": 7.694024324090205e-07,
+      "loss": 0.4482,
+      "step": 10523
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7943535833943023,
+      "learning_rate": 7.687247101634449e-07,
+      "loss": 0.4341,
+      "step": 10524
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5405830788229496,
+      "learning_rate": 7.680472616758467e-07,
+      "loss": 0.4759,
+      "step": 10525
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3315461778543876,
+      "learning_rate": 7.673700869900536e-07,
+      "loss": 0.4454,
+      "step": 10526
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6853774075410262,
+      "learning_rate": 7.666931861498788e-07,
+      "loss": 0.4232,
+      "step": 10527
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5817766033926233,
+      "learning_rate": 7.66016559199117e-07,
+      "loss": 0.4506,
+      "step": 10528
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8565430079435985,
+      "learning_rate": 7.653402061815462e-07,
+      "loss": 0.4171,
+      "step": 10529
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9356066932565676,
+      "learning_rate": 7.646641271409233e-07,
+      "loss": 0.4531,
+      "step": 10530
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3723100509113055,
+      "learning_rate": 7.639883221209899e-07,
+      "loss": 0.4384,
+      "step": 10531
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.0359078845897915,
+      "learning_rate": 7.633127911654725e-07,
+      "loss": 0.415,
+      "step": 10532
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.678473949071921,
+      "learning_rate": 7.626375343180742e-07,
+      "loss": 0.5014,
+      "step": 10533
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6838617121346133,
+      "learning_rate": 7.619625516224854e-07,
+      "loss": 0.4165,
+      "step": 10534
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.398085281846617,
+      "learning_rate": 7.612878431223736e-07,
+      "loss": 0.4564,
+      "step": 10535
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6462042622435864,
+      "learning_rate": 7.606134088613954e-07,
+      "loss": 0.4807,
+      "step": 10536
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5108619678677597,
+      "learning_rate": 7.599392488831825e-07,
+      "loss": 0.4381,
+      "step": 10537
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.517588977747294,
+      "learning_rate": 7.59265363231354e-07,
+      "loss": 0.5155,
+      "step": 10538
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2031479033436567,
+      "learning_rate": 7.585917519495084e-07,
+      "loss": 0.4414,
+      "step": 10539
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7215911115858906,
+      "learning_rate": 7.579184150812297e-07,
+      "loss": 0.4741,
+      "step": 10540
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.114927707807161,
+      "learning_rate": 7.572453526700785e-07,
+      "loss": 0.4566,
+      "step": 10541
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.809764037563999,
+      "learning_rate": 7.565725647596028e-07,
+      "loss": 0.3765,
+      "step": 10542
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.888022302806769,
+      "learning_rate": 7.559000513933312e-07,
+      "loss": 0.4694,
+      "step": 10543
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 5.3171258366055145,
+      "learning_rate": 7.552278126147744e-07,
+      "loss": 0.4423,
+      "step": 10544
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5389018315423004,
+      "learning_rate": 7.545558484674264e-07,
+      "loss": 0.4687,
+      "step": 10545
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8164428281260794,
+      "learning_rate": 7.538841589947599e-07,
+      "loss": 0.4714,
+      "step": 10546
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5812016049619626,
+      "learning_rate": 7.532127442402337e-07,
+      "loss": 0.4561,
+      "step": 10547
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.7537437444967034,
+      "learning_rate": 7.525416042472877e-07,
+      "loss": 0.4706,
+      "step": 10548
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5567653317781628,
+      "learning_rate": 7.518707390593449e-07,
+      "loss": 0.4912,
+      "step": 10549
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6121919323594556,
+      "learning_rate": 7.512001487198051e-07,
+      "loss": 0.4809,
+      "step": 10550
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6031926595770527,
+      "learning_rate": 7.505298332720601e-07,
+      "loss": 0.4574,
+      "step": 10551
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4578192336840505,
+      "learning_rate": 7.498597927594748e-07,
+      "loss": 0.4564,
+      "step": 10552
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8163232364348196,
+      "learning_rate": 7.49190027225401e-07,
+      "loss": 0.42,
+      "step": 10553
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4241521212525727,
+      "learning_rate": 7.485205367131721e-07,
+      "loss": 0.4741,
+      "step": 10554
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.860552172848616,
+      "learning_rate": 7.478513212661021e-07,
+      "loss": 0.4345,
+      "step": 10555
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5689153823071813,
+      "learning_rate": 7.471823809274909e-07,
+      "loss": 0.4787,
+      "step": 10556
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.2615148959792037,
+      "learning_rate": 7.46513715740615e-07,
+      "loss": 0.4602,
+      "step": 10557
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.254356198053311,
+      "learning_rate": 7.458453257487369e-07,
+      "loss": 0.453,
+      "step": 10558
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5706240347495042,
+      "learning_rate": 7.451772109951016e-07,
+      "loss": 0.4214,
+      "step": 10559
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.302135577753881,
+      "learning_rate": 7.445093715229356e-07,
+      "loss": 0.5135,
+      "step": 10560
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 9.317015326389319,
+      "learning_rate": 7.438418073754456e-07,
+      "loss": 0.4705,
+      "step": 10561
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.55282029670015,
+      "learning_rate": 7.431745185958223e-07,
+      "loss": 0.4504,
+      "step": 10562
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.48795831693043,
+      "learning_rate": 7.425075052272396e-07,
+      "loss": 0.4232,
+      "step": 10563
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.096728701319735,
+      "learning_rate": 7.418407673128514e-07,
+      "loss": 0.4384,
+      "step": 10564
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5971121808522775,
+      "learning_rate": 7.411743048957965e-07,
+      "loss": 0.4235,
+      "step": 10565
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5255303463131702,
+      "learning_rate": 7.405081180191909e-07,
+      "loss": 0.4552,
+      "step": 10566
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.2773446879108823,
+      "learning_rate": 7.398422067261396e-07,
+      "loss": 0.4764,
+      "step": 10567
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5612833892498487,
+      "learning_rate": 7.391765710597237e-07,
+      "loss": 0.4762,
+      "step": 10568
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 7.219307298141768,
+      "learning_rate": 7.38511211063011e-07,
+      "loss": 0.4522,
+      "step": 10569
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5814998059988247,
+      "learning_rate": 7.37846126779046e-07,
+      "loss": 0.4762,
+      "step": 10570
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7357400480376641,
+      "learning_rate": 7.37181318250863e-07,
+      "loss": 0.4539,
+      "step": 10571
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.417097618210183,
+      "learning_rate": 7.365167855214711e-07,
+      "loss": 0.4158,
+      "step": 10572
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6116881681166537,
+      "learning_rate": 7.358525286338664e-07,
+      "loss": 0.4058,
+      "step": 10573
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6494477995560357,
+      "learning_rate": 7.351885476310244e-07,
+      "loss": 0.4724,
+      "step": 10574
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8962213527443266,
+      "learning_rate": 7.345248425559043e-07,
+      "loss": 0.4352,
+      "step": 10575
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6675876274353516,
+      "learning_rate": 7.338614134514482e-07,
+      "loss": 0.4667,
+      "step": 10576
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1430755958484577,
+      "learning_rate": 7.331982603605764e-07,
+      "loss": 0.4544,
+      "step": 10577
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3960839407345467,
+      "learning_rate": 7.325353833261956e-07,
+      "loss": 0.429,
+      "step": 10578
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.413680457817274,
+      "learning_rate": 7.31872782391193e-07,
+      "loss": 0.4515,
+      "step": 10579
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.6001233398623165,
+      "learning_rate": 7.312104575984397e-07,
+      "loss": 0.3942,
+      "step": 10580
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4972765224705613,
+      "learning_rate": 7.305484089907838e-07,
+      "loss": 0.4443,
+      "step": 10581
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9366631257451463,
+      "learning_rate": 7.298866366110607e-07,
+      "loss": 0.511,
+      "step": 10582
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.432471435266661,
+      "learning_rate": 7.292251405020862e-07,
+      "loss": 0.4247,
+      "step": 10583
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7552394220442418,
+      "learning_rate": 7.285639207066591e-07,
+      "loss": 0.4225,
+      "step": 10584
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2150558441291324,
+      "learning_rate": 7.279029772675572e-07,
+      "loss": 0.4029,
+      "step": 10585
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6752012027032102,
+      "learning_rate": 7.272423102275445e-07,
+      "loss": 0.4161,
+      "step": 10586
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8531403767023835,
+      "learning_rate": 7.26581919629366e-07,
+      "loss": 0.4319,
+      "step": 10587
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5229120767429257,
+      "learning_rate": 7.259218055157458e-07,
+      "loss": 0.4566,
+      "step": 10588
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8718696909609773,
+      "learning_rate": 7.252619679293937e-07,
+      "loss": 0.3962,
+      "step": 10589
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.22950862965234,
+      "learning_rate": 7.246024069130004e-07,
+      "loss": 0.3811,
+      "step": 10590
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7451374258481898,
+      "learning_rate": 7.239431225092397e-07,
+      "loss": 0.4488,
+      "step": 10591
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5238133579910506,
+      "learning_rate": 7.232841147607639e-07,
+      "loss": 0.4457,
+      "step": 10592
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1362877755698304,
+      "learning_rate": 7.226253837102109e-07,
+      "loss": 0.4429,
+      "step": 10593
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7402655114408039,
+      "learning_rate": 7.219669294002002e-07,
+      "loss": 0.3976,
+      "step": 10594
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5434919750140725,
+      "learning_rate": 7.213087518733341e-07,
+      "loss": 0.3986,
+      "step": 10595
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8967123775462449,
+      "learning_rate": 7.206508511721933e-07,
+      "loss": 0.4552,
+      "step": 10596
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.5326731976319565,
+      "learning_rate": 7.199932273393445e-07,
+      "loss": 0.4332,
+      "step": 10597
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6829190394143017,
+      "learning_rate": 7.19335880417335e-07,
+      "loss": 0.492,
+      "step": 10598
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2953998902493518,
+      "learning_rate": 7.186788104486936e-07,
+      "loss": 0.4273,
+      "step": 10599
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.858386332745713,
+      "learning_rate": 7.180220174759345e-07,
+      "loss": 0.4003,
+      "step": 10600
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.003565097261802,
+      "learning_rate": 7.173655015415465e-07,
+      "loss": 0.5118,
+      "step": 10601
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3693432797929233,
+      "learning_rate": 7.167092626880107e-07,
+      "loss": 0.4899,
+      "step": 10602
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6269481727048096,
+      "learning_rate": 7.160533009577808e-07,
+      "loss": 0.438,
+      "step": 10603
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.058163598176839,
+      "learning_rate": 7.153976163932996e-07,
+      "loss": 0.4285,
+      "step": 10604
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6441881616436609,
+      "learning_rate": 7.147422090369854e-07,
+      "loss": 0.4741,
+      "step": 10605
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5615128754202016,
+      "learning_rate": 7.140870789312471e-07,
+      "loss": 0.4669,
+      "step": 10606
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.3456162186333347,
+      "learning_rate": 7.134322261184662e-07,
+      "loss": 0.51,
+      "step": 10607
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7993210771523578,
+      "learning_rate": 7.127776506410134e-07,
+      "loss": 0.4509,
+      "step": 10608
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5409627791061821,
+      "learning_rate": 7.12123352541238e-07,
+      "loss": 0.4853,
+      "step": 10609
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.772569212832061,
+      "learning_rate": 7.114693318614723e-07,
+      "loss": 0.4592,
+      "step": 10610
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5598474010629473,
+      "learning_rate": 7.108155886440321e-07,
+      "loss": 0.4624,
+      "step": 10611
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7482030922763923,
+      "learning_rate": 7.101621229312111e-07,
+      "loss": 0.4464,
+      "step": 10612
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.039417855933371,
+      "learning_rate": 7.095089347652889e-07,
+      "loss": 0.462,
+      "step": 10613
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5682006385309659,
+      "learning_rate": 7.088560241885256e-07,
+      "loss": 0.457,
+      "step": 10614
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.9173123853531797,
+      "learning_rate": 7.082033912431658e-07,
+      "loss": 0.464,
+      "step": 10615
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.262541680649712,
+      "learning_rate": 7.075510359714305e-07,
+      "loss": 0.4394,
+      "step": 10616
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4127308117213195,
+      "learning_rate": 7.068989584155283e-07,
+      "loss": 0.4472,
+      "step": 10617
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.699450627630125,
+      "learning_rate": 7.062471586176473e-07,
+      "loss": 0.4832,
+      "step": 10618
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.1932431573612186,
+      "learning_rate": 7.055956366199595e-07,
+      "loss": 0.4566,
+      "step": 10619
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7379900483867532,
+      "learning_rate": 7.049443924646138e-07,
+      "loss": 0.4218,
+      "step": 10620
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2851025365906996,
+      "learning_rate": 7.042934261937484e-07,
+      "loss": 0.4281,
+      "step": 10621
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5554322575196475,
+      "learning_rate": 7.036427378494804e-07,
+      "loss": 0.4926,
+      "step": 10622
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5213528875887441,
+      "learning_rate": 7.029923274739053e-07,
+      "loss": 0.4723,
+      "step": 10623
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.199635282276161,
+      "learning_rate": 7.023421951091053e-07,
+      "loss": 0.4222,
+      "step": 10624
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.915413509761312,
+      "learning_rate": 7.016923407971432e-07,
+      "loss": 0.4813,
+      "step": 10625
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.09530607964085,
+      "learning_rate": 7.010427645800655e-07,
+      "loss": 0.4433,
+      "step": 10626
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 2.4282932174060985,
+      "learning_rate": 7.003934664998957e-07,
+      "loss": 0.4588,
+      "step": 10627
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.528510681393632,
+      "learning_rate": 6.997444465986442e-07,
+      "loss": 0.3815,
+      "step": 10628
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3746625250870323,
+      "learning_rate": 6.990957049183011e-07,
+      "loss": 0.4101,
+      "step": 10629
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.378387900466592,
+      "learning_rate": 6.984472415008408e-07,
+      "loss": 0.4524,
+      "step": 10630
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 3.1088558292974695,
+      "learning_rate": 6.977990563882165e-07,
+      "loss": 0.4007,
+      "step": 10631
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8216017181001483,
+      "learning_rate": 6.971511496223643e-07,
+      "loss": 0.5203,
+      "step": 10632
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.966337273078408,
+      "learning_rate": 6.965035212452043e-07,
+      "loss": 0.4615,
+      "step": 10633
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.7270422239114036,
+      "learning_rate": 6.958561712986367e-07,
+      "loss": 0.4433,
+      "step": 10634
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7750105060863808,
+      "learning_rate": 6.952090998245453e-07,
+      "loss": 0.5078,
+      "step": 10635
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6679588176040823,
+      "learning_rate": 6.945623068647922e-07,
+      "loss": 0.4258,
+      "step": 10636
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3607718985061528,
+      "learning_rate": 6.939157924612272e-07,
+      "loss": 0.3984,
+      "step": 10637
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5537740725478609,
+      "learning_rate": 6.932695566556763e-07,
+      "loss": 0.4549,
+      "step": 10638
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2090078497127363,
+      "learning_rate": 6.926235994899528e-07,
+      "loss": 0.4594,
+      "step": 10639
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2198465741682236,
+      "learning_rate": 6.919779210058447e-07,
+      "loss": 0.4492,
+      "step": 10640
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4685837115302212,
+      "learning_rate": 6.913325212451322e-07,
+      "loss": 0.4262,
+      "step": 10641
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6927511911117012,
+      "learning_rate": 6.906874002495678e-07,
+      "loss": 0.3991,
+      "step": 10642
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6964784704818188,
+      "learning_rate": 6.900425580608916e-07,
+      "loss": 0.4419,
+      "step": 10643
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5397056382885204,
+      "learning_rate": 6.893979947208235e-07,
+      "loss": 0.4807,
+      "step": 10644
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6845128980202997,
+      "learning_rate": 6.887537102710662e-07,
+      "loss": 0.4431,
+      "step": 10645
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.536268968551931,
+      "learning_rate": 6.881097047533047e-07,
+      "loss": 0.4872,
+      "step": 10646
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5194201580287918,
+      "learning_rate": 6.874659782092041e-07,
+      "loss": 0.4511,
+      "step": 10647
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3160002101378592,
+      "learning_rate": 6.868225306804132e-07,
+      "loss": 0.422,
+      "step": 10648
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7148668385569532,
+      "learning_rate": 6.861793622085622e-07,
+      "loss": 0.4259,
+      "step": 10649
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9040214725498845,
+      "learning_rate": 6.855364728352643e-07,
+      "loss": 0.408,
+      "step": 10650
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8621924258271785,
+      "learning_rate": 6.848938626021112e-07,
+      "loss": 0.461,
+      "step": 10651
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7124390789599107,
+      "learning_rate": 6.842515315506809e-07,
+      "loss": 0.4433,
+      "step": 10652
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.032863350494302,
+      "learning_rate": 6.836094797225306e-07,
+      "loss": 0.4329,
+      "step": 10653
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5382223817268618,
+      "learning_rate": 6.829677071592e-07,
+      "loss": 0.4832,
+      "step": 10654
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.312665022022054,
+      "learning_rate": 6.823262139022119e-07,
+      "loss": 0.464,
+      "step": 10655
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7930319769205327,
+      "learning_rate": 6.816849999930691e-07,
+      "loss": 0.4338,
+      "step": 10656
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7909627787217854,
+      "learning_rate": 6.810440654732592e-07,
+      "loss": 0.4075,
+      "step": 10657
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.54975043837616,
+      "learning_rate": 6.804034103842471e-07,
+      "loss": 0.4038,
+      "step": 10658
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.70822342377593,
+      "learning_rate": 6.797630347674833e-07,
+      "loss": 0.424,
+      "step": 10659
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0193018161484804,
+      "learning_rate": 6.791229386643999e-07,
+      "loss": 0.4893,
+      "step": 10660
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.700245250499907,
+      "learning_rate": 6.784831221164113e-07,
+      "loss": 0.4483,
+      "step": 10661
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6318722132310792,
+      "learning_rate": 6.778435851649101e-07,
+      "loss": 0.4449,
+      "step": 10662
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6731301835254497,
+      "learning_rate": 6.772043278512747e-07,
+      "loss": 0.5147,
+      "step": 10663
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5094104295199884,
+      "learning_rate": 6.765653502168646e-07,
+      "loss": 0.3971,
+      "step": 10664
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5054938064665913,
+      "learning_rate": 6.759266523030217e-07,
+      "loss": 0.461,
+      "step": 10665
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.5956615557668576,
+      "learning_rate": 6.752882341510663e-07,
+      "loss": 0.4563,
+      "step": 10666
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2219427086196077,
+      "learning_rate": 6.74650095802305e-07,
+      "loss": 0.4423,
+      "step": 10667
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0461035449146765,
+      "learning_rate": 6.74012237298024e-07,
+      "loss": 0.4971,
+      "step": 10668
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6654819515488573,
+      "learning_rate": 6.733746586794925e-07,
+      "loss": 0.4216,
+      "step": 10669
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8371940608724024,
+      "learning_rate": 6.727373599879617e-07,
+      "loss": 0.4323,
+      "step": 10670
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.756572135555636,
+      "learning_rate": 6.721003412646604e-07,
+      "loss": 0.4476,
+      "step": 10671
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.843198490948536,
+      "learning_rate": 6.714636025508081e-07,
+      "loss": 0.4297,
+      "step": 10672
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.059736616774512,
+      "learning_rate": 6.708271438875968e-07,
+      "loss": 0.4469,
+      "step": 10673
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.756322235707946,
+      "learning_rate": 6.701909653162076e-07,
+      "loss": 0.4584,
+      "step": 10674
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1092821559800905,
+      "learning_rate": 6.695550668777962e-07,
+      "loss": 0.4909,
+      "step": 10675
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4974114912072576,
+      "learning_rate": 6.689194486135092e-07,
+      "loss": 0.4561,
+      "step": 10676
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.4713085961367547,
+      "learning_rate": 6.682841105644672e-07,
+      "loss": 0.4235,
+      "step": 10677
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.003704943831131,
+      "learning_rate": 6.676490527717766e-07,
+      "loss": 0.4774,
+      "step": 10678
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.847168231471243,
+      "learning_rate": 6.670142752765246e-07,
+      "loss": 0.3983,
+      "step": 10679
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6862587570595589,
+      "learning_rate": 6.663797781197812e-07,
+      "loss": 0.4546,
+      "step": 10680
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7068724774272455,
+      "learning_rate": 6.657455613425979e-07,
+      "loss": 0.4418,
+      "step": 10681
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5351257022235056,
+      "learning_rate": 6.651116249860057e-07,
+      "loss": 0.4459,
+      "step": 10682
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.128996661846696,
+      "learning_rate": 6.644779690910208e-07,
+      "loss": 0.4545,
+      "step": 10683
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.626387138762449,
+      "learning_rate": 6.638445936986393e-07,
+      "loss": 0.3931,
+      "step": 10684
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.316828830127037,
+      "learning_rate": 6.632114988498417e-07,
+      "loss": 0.4442,
+      "step": 10685
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7934640524563779,
+      "learning_rate": 6.625786845855858e-07,
+      "loss": 0.4362,
+      "step": 10686
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.5812846553598687,
+      "learning_rate": 6.61946150946815e-07,
+      "loss": 0.4114,
+      "step": 10687
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7822974221690324,
+      "learning_rate": 6.613138979744532e-07,
+      "loss": 0.414,
+      "step": 10688
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.053307410662241,
+      "learning_rate": 6.606819257094066e-07,
+      "loss": 0.4487,
+      "step": 10689
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 4.820572103939041,
+      "learning_rate": 6.600502341925624e-07,
+      "loss": 0.457,
+      "step": 10690
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8400476219224644,
+      "learning_rate": 6.594188234647913e-07,
+      "loss": 0.4406,
+      "step": 10691
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8500534810364186,
+      "learning_rate": 6.587876935669446e-07,
+      "loss": 0.4305,
+      "step": 10692
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.781823387761016,
+      "learning_rate": 6.581568445398545e-07,
+      "loss": 0.4587,
+      "step": 10693
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.4031605541648275,
+      "learning_rate": 6.575262764243368e-07,
+      "loss": 0.4209,
+      "step": 10694
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0484288923151746,
+      "learning_rate": 6.568959892611882e-07,
+      "loss": 0.4873,
+      "step": 10695
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8766202190608015,
+      "learning_rate": 6.562659830911883e-07,
+      "loss": 0.4816,
+      "step": 10696
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7377699983800727,
+      "learning_rate": 6.556362579550962e-07,
+      "loss": 0.4995,
+      "step": 10697
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6917016214919727,
+      "learning_rate": 6.550068138936555e-07,
+      "loss": 0.4447,
+      "step": 10698
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5259794808055107,
+      "learning_rate": 6.543776509475897e-07,
+      "loss": 0.4073,
+      "step": 10699
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6962471054798165,
+      "learning_rate": 6.537487691576044e-07,
+      "loss": 0.4328,
+      "step": 10700
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5605011100392531,
+      "learning_rate": 6.531201685643901e-07,
+      "loss": 0.4673,
+      "step": 10701
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.81787194754758,
+      "learning_rate": 6.52491849208613e-07,
+      "loss": 0.3995,
+      "step": 10702
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9546298164334213,
+      "learning_rate": 6.518638111309261e-07,
+      "loss": 0.4361,
+      "step": 10703
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8255558778153074,
+      "learning_rate": 6.512360543719626e-07,
+      "loss": 0.4473,
+      "step": 10704
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.445829030253032,
+      "learning_rate": 6.506085789723382e-07,
+      "loss": 0.3999,
+      "step": 10705
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.191251097691553,
+      "learning_rate": 6.499813849726471e-07,
+      "loss": 0.443,
+      "step": 10706
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3895863329225655,
+      "learning_rate": 6.49354472413472e-07,
+      "loss": 0.4928,
+      "step": 10707
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.005762921725672,
+      "learning_rate": 6.487278413353703e-07,
+      "loss": 0.4381,
+      "step": 10708
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0280845369821043,
+      "learning_rate": 6.481014917788859e-07,
+      "loss": 0.4735,
+      "step": 10709
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9265798976973052,
+      "learning_rate": 6.474754237845399e-07,
+      "loss": 0.4614,
+      "step": 10710
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5695735178076162,
+      "learning_rate": 6.468496373928412e-07,
+      "loss": 0.4654,
+      "step": 10711
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.833141276809215,
+      "learning_rate": 6.462241326442775e-07,
+      "loss": 0.4018,
+      "step": 10712
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.203113768191496,
+      "learning_rate": 6.455989095793158e-07,
+      "loss": 0.4186,
+      "step": 10713
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2250349157697764,
+      "learning_rate": 6.449739682384082e-07,
+      "loss": 0.4521,
+      "step": 10714
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.6440796007398124,
+      "learning_rate": 6.443493086619884e-07,
+      "loss": 0.4536,
+      "step": 10715
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5371101995777372,
+      "learning_rate": 6.437249308904708e-07,
+      "loss": 0.482,
+      "step": 10716
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.05407619295554,
+      "learning_rate": 6.431008349642509e-07,
+      "loss": 0.4365,
+      "step": 10717
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.007079761125552,
+      "learning_rate": 6.424770209237069e-07,
+      "loss": 0.4205,
+      "step": 10718
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.205391552639151,
+      "learning_rate": 6.418534888091998e-07,
+      "loss": 0.4217,
+      "step": 10719
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5514636227149246,
+      "learning_rate": 6.412302386610714e-07,
+      "loss": 0.4642,
+      "step": 10720
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5290898880410322,
+      "learning_rate": 6.406072705196437e-07,
+      "loss": 0.4854,
+      "step": 10721
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9994688694439704,
+      "learning_rate": 6.399845844252218e-07,
+      "loss": 0.4705,
+      "step": 10722
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5213265426713367,
+      "learning_rate": 6.393621804180961e-07,
+      "loss": 0.455,
+      "step": 10723
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4853465900859177,
+      "learning_rate": 6.387400585385311e-07,
+      "loss": 0.4234,
+      "step": 10724
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.3982056414417987,
+      "learning_rate": 6.381182188267809e-07,
+      "loss": 0.4961,
+      "step": 10725
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.839140825583552,
+      "learning_rate": 6.374966613230732e-07,
+      "loss": 0.4942,
+      "step": 10726
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5467197117303432,
+      "learning_rate": 6.368753860676269e-07,
+      "loss": 0.4632,
+      "step": 10727
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7892157611266242,
+      "learning_rate": 6.362543931006343e-07,
+      "loss": 0.4098,
+      "step": 10728
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.2857626659349806,
+      "learning_rate": 6.356336824622744e-07,
+      "loss": 0.46,
+      "step": 10729
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5811190604182273,
+      "learning_rate": 6.350132541927057e-07,
+      "loss": 0.4885,
+      "step": 10730
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0408943141391895,
+      "learning_rate": 6.343931083320704e-07,
+      "loss": 0.4218,
+      "step": 10731
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4463160109204227,
+      "learning_rate": 6.337732449204886e-07,
+      "loss": 0.3877,
+      "step": 10732
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.69962972865419,
+      "learning_rate": 6.331536639980667e-07,
+      "loss": 0.417,
+      "step": 10733
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.7561165621652135,
+      "learning_rate": 6.325343656048899e-07,
+      "loss": 0.4256,
+      "step": 10734
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8768130782059609,
+      "learning_rate": 6.319153497810266e-07,
+      "loss": 0.4081,
+      "step": 10735
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.502886706519453,
+      "learning_rate": 6.312966165665263e-07,
+      "loss": 0.4612,
+      "step": 10736
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.171785716926124,
+      "learning_rate": 6.306781660014194e-07,
+      "loss": 0.4566,
+      "step": 10737
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 3.624382424834933,
+      "learning_rate": 6.300599981257193e-07,
+      "loss": 0.4844,
+      "step": 10738
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.547483583070166,
+      "learning_rate": 6.29442112979421e-07,
+      "loss": 0.4643,
+      "step": 10739
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.058531002574587,
+      "learning_rate": 6.288245106025015e-07,
+      "loss": 0.423,
+      "step": 10740
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0185327289834234,
+      "learning_rate": 6.282071910349152e-07,
+      "loss": 0.4103,
+      "step": 10741
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5220032886691861,
+      "learning_rate": 6.275901543166074e-07,
+      "loss": 0.4724,
+      "step": 10742
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6127179183379431,
+      "learning_rate": 6.269734004874956e-07,
+      "loss": 0.4629,
+      "step": 10743
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0900446525253376,
+      "learning_rate": 6.26356929587485e-07,
+      "loss": 0.4338,
+      "step": 10744
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.885178812776475,
+      "learning_rate": 6.257407416564576e-07,
+      "loss": 0.4254,
+      "step": 10745
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.53358432699037,
+      "learning_rate": 6.251248367342833e-07,
+      "loss": 0.4727,
+      "step": 10746
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.850430135838408,
+      "learning_rate": 6.2450921486081e-07,
+      "loss": 0.4494,
+      "step": 10747
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9969723715706393,
+      "learning_rate": 6.23893876075865e-07,
+      "loss": 0.4006,
+      "step": 10748
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.007078430430243,
+      "learning_rate": 6.232788204192624e-07,
+      "loss": 0.4395,
+      "step": 10749
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.389490615699665,
+      "learning_rate": 6.226640479307943e-07,
+      "loss": 0.4764,
+      "step": 10750
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.8297756184440575,
+      "learning_rate": 6.220495586502367e-07,
+      "loss": 0.4661,
+      "step": 10751
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5956634187715218,
+      "learning_rate": 6.21435352617345e-07,
+      "loss": 0.4813,
+      "step": 10752
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.0396832738084663,
+      "learning_rate": 6.208214298718579e-07,
+      "loss": 0.4326,
+      "step": 10753
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.034878129679466,
+      "learning_rate": 6.202077904534959e-07,
+      "loss": 0.4262,
+      "step": 10754
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.6042694590411406,
+      "learning_rate": 6.195944344019611e-07,
+      "loss": 0.4407,
+      "step": 10755
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.9815024060941804,
+      "learning_rate": 6.189813617569352e-07,
+      "loss": 0.4244,
+      "step": 10756
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.1980654936196093,
+      "learning_rate": 6.183685725580829e-07,
+      "loss": 0.4725,
+      "step": 10757
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5495542961519851,
+      "learning_rate": 6.177560668450539e-07,
+      "loss": 0.4988,
+      "step": 10758
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5828860896031169,
+      "learning_rate": 6.17143844657474e-07,
+      "loss": 0.4563,
+      "step": 10759
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6152607419167093,
+      "learning_rate": 6.165319060349551e-07,
+      "loss": 0.4755,
+      "step": 10760
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.8293710683555,
+      "learning_rate": 6.159202510170847e-07,
+      "loss": 0.4816,
+      "step": 10761
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.254212716978076,
+      "learning_rate": 6.15308879643442e-07,
+      "loss": 0.4216,
+      "step": 10762
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5356472859332781,
+      "learning_rate": 6.146977919535774e-07,
+      "loss": 0.3985,
+      "step": 10763
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7181886729506028,
+      "learning_rate": 6.140869879870287e-07,
+      "loss": 0.484,
+      "step": 10764
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5068125063763531,
+      "learning_rate": 6.134764677833149e-07,
+      "loss": 0.4658,
+      "step": 10765
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.676032320145416,
+      "learning_rate": 6.128662313819362e-07,
+      "loss": 0.4163,
+      "step": 10766
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9742997718596353,
+      "learning_rate": 6.122562788223724e-07,
+      "loss": 0.4986,
+      "step": 10767
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5058476103902625,
+      "learning_rate": 6.116466101440871e-07,
+      "loss": 0.4671,
+      "step": 10768
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1133596615408363,
+      "learning_rate": 6.110372253865255e-07,
+      "loss": 0.4354,
+      "step": 10769
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5683201992219515,
+      "learning_rate": 6.104281245891141e-07,
+      "loss": 0.4767,
+      "step": 10770
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8443396414802622,
+      "learning_rate": 6.098193077912618e-07,
+      "loss": 0.4378,
+      "step": 10771
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9546640569703932,
+      "learning_rate": 6.092107750323562e-07,
+      "loss": 0.4426,
+      "step": 10772
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.985797711230216,
+      "learning_rate": 6.086025263517692e-07,
+      "loss": 0.4264,
+      "step": 10773
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.163814367467736,
+      "learning_rate": 6.079945617888544e-07,
+      "loss": 0.4111,
+      "step": 10774
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5198349301332297,
+      "learning_rate": 6.07386881382947e-07,
+      "loss": 0.4133,
+      "step": 10775
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4770641914466405,
+      "learning_rate": 6.0677948517336e-07,
+      "loss": 0.465,
+      "step": 10776
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5733373793669696,
+      "learning_rate": 6.061723731993951e-07,
+      "loss": 0.4215,
+      "step": 10777
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5659410848480282,
+      "learning_rate": 6.055655455003289e-07,
+      "loss": 0.4572,
+      "step": 10778
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5392384193905682,
+      "learning_rate": 6.049590021154233e-07,
+      "loss": 0.4376,
+      "step": 10779
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4422965485470225,
+      "learning_rate": 6.043527430839208e-07,
+      "loss": 0.4066,
+      "step": 10780
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.01957811071372,
+      "learning_rate": 6.03746768445046e-07,
+      "loss": 0.4461,
+      "step": 10781
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.566111995609609,
+      "learning_rate": 6.031410782380049e-07,
+      "loss": 0.4185,
+      "step": 10782
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6194995263676935,
+      "learning_rate": 6.025356725019832e-07,
+      "loss": 0.4011,
+      "step": 10783
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8875681478744082,
+      "learning_rate": 6.019305512761508e-07,
+      "loss": 0.4145,
+      "step": 10784
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7007208466583958,
+      "learning_rate": 6.013257145996587e-07,
+      "loss": 0.4673,
+      "step": 10785
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5395393529176088,
+      "learning_rate": 6.007211625116394e-07,
+      "loss": 0.4513,
+      "step": 10786
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5478781205843473,
+      "learning_rate": 6.001168950512048e-07,
+      "loss": 0.4726,
+      "step": 10787
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5763235899281748,
+      "learning_rate": 5.995129122574517e-07,
+      "loss": 0.4755,
+      "step": 10788
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.990752247269342,
+      "learning_rate": 5.989092141694563e-07,
+      "loss": 0.488,
+      "step": 10789
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.565399298756159,
+      "learning_rate": 5.983058008262777e-07,
+      "loss": 0.4447,
+      "step": 10790
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.003246083554952,
+      "learning_rate": 5.97702672266956e-07,
+      "loss": 0.4448,
+      "step": 10791
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3993089887650623,
+      "learning_rate": 5.970998285305113e-07,
+      "loss": 0.4429,
+      "step": 10792
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3216908159238114,
+      "learning_rate": 5.964972696559496e-07,
+      "loss": 0.4244,
+      "step": 10793
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5344676289342227,
+      "learning_rate": 5.95894995682253e-07,
+      "loss": 0.3908,
+      "step": 10794
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.512207562490718,
+      "learning_rate": 5.952930066483897e-07,
+      "loss": 0.4978,
+      "step": 10795
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5225455933231761,
+      "learning_rate": 5.946913025933049e-07,
+      "loss": 0.4646,
+      "step": 10796
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.3586749670406686,
+      "learning_rate": 5.940898835559322e-07,
+      "loss": 0.4232,
+      "step": 10797
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5593505517596848,
+      "learning_rate": 5.934887495751796e-07,
+      "loss": 0.4223,
+      "step": 10798
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.8220238770516963,
+      "learning_rate": 5.928879006899396e-07,
+      "loss": 0.4473,
+      "step": 10799
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5536027658926895,
+      "learning_rate": 5.92287336939088e-07,
+      "loss": 0.475,
+      "step": 10800
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4541958896894158,
+      "learning_rate": 5.916870583614792e-07,
+      "loss": 0.399,
+      "step": 10801
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.637756580703952,
+      "learning_rate": 5.910870649959522e-07,
+      "loss": 0.4895,
+      "step": 10802
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5146636224458756,
+      "learning_rate": 5.904873568813236e-07,
+      "loss": 0.4372,
+      "step": 10803
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.013438246491489,
+      "learning_rate": 5.89887934056394e-07,
+      "loss": 0.437,
+      "step": 10804
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9922980137485204,
+      "learning_rate": 5.892887965599464e-07,
+      "loss": 0.3867,
+      "step": 10805
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.5359570081297016,
+      "learning_rate": 5.886899444307448e-07,
+      "loss": 0.4343,
+      "step": 10806
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4652502354706705,
+      "learning_rate": 5.88091377707532e-07,
+      "loss": 0.4436,
+      "step": 10807
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0566861749072802,
+      "learning_rate": 5.87493096429036e-07,
+      "loss": 0.5002,
+      "step": 10808
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.24133630827586,
+      "learning_rate": 5.868951006339635e-07,
+      "loss": 0.4898,
+      "step": 10809
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.30951765685275,
+      "learning_rate": 5.862973903610065e-07,
+      "loss": 0.493,
+      "step": 10810
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5417803423874918,
+      "learning_rate": 5.856999656488322e-07,
+      "loss": 0.4429,
+      "step": 10811
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.911492046118922,
+      "learning_rate": 5.85102826536097e-07,
+      "loss": 0.4498,
+      "step": 10812
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5400918949336364,
+      "learning_rate": 5.845059730614339e-07,
+      "loss": 0.4527,
+      "step": 10813
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5202719026276379,
+      "learning_rate": 5.839094052634575e-07,
+      "loss": 0.4696,
+      "step": 10814
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5429198491748485,
+      "learning_rate": 5.833131231807654e-07,
+      "loss": 0.4536,
+      "step": 10815
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3623802649636514,
+      "learning_rate": 5.827171268519365e-07,
+      "loss": 0.4241,
+      "step": 10816
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.160507762999765,
+      "learning_rate": 5.82121416315532e-07,
+      "loss": 0.4313,
+      "step": 10817
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5449607707765769,
+      "learning_rate": 5.815259916100918e-07,
+      "loss": 0.4767,
+      "step": 10818
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8159785381765345,
+      "learning_rate": 5.809308527741397e-07,
+      "loss": 0.367,
+      "step": 10819
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4829408520802603,
+      "learning_rate": 5.803359998461805e-07,
+      "loss": 0.459,
+      "step": 10820
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0951268489676216,
+      "learning_rate": 5.797414328647011e-07,
+      "loss": 0.437,
+      "step": 10821
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 9.422689292584558,
+      "learning_rate": 5.791471518681679e-07,
+      "loss": 0.4157,
+      "step": 10822
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0619657511312353,
+      "learning_rate": 5.785531568950309e-07,
+      "loss": 0.3925,
+      "step": 10823
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7916179152446619,
+      "learning_rate": 5.779594479837209e-07,
+      "loss": 0.4389,
+      "step": 10824
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.01182948138568,
+      "learning_rate": 5.773660251726493e-07,
+      "loss": 0.425,
+      "step": 10825
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9257717345465382,
+      "learning_rate": 5.767728885002116e-07,
+      "loss": 0.4323,
+      "step": 10826
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5709221319772775,
+      "learning_rate": 5.761800380047794e-07,
+      "loss": 0.4454,
+      "step": 10827
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5822537843938314,
+      "learning_rate": 5.755874737247141e-07,
+      "loss": 0.4192,
+      "step": 10828
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.442436894068763,
+      "learning_rate": 5.749951956983501e-07,
+      "loss": 0.4441,
+      "step": 10829
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5107128702364907,
+      "learning_rate": 5.744032039640096e-07,
+      "loss": 0.4749,
+      "step": 10830
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5734869784195678,
+      "learning_rate": 5.738114985599902e-07,
+      "loss": 0.4802,
+      "step": 10831
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.812160297709051,
+      "learning_rate": 5.732200795245785e-07,
+      "loss": 0.4362,
+      "step": 10832
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7317104284369116,
+      "learning_rate": 5.726289468960361e-07,
+      "loss": 0.4532,
+      "step": 10833
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.071701908294174,
+      "learning_rate": 5.720381007126092e-07,
+      "loss": 0.4298,
+      "step": 10834
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.503318717640116,
+      "learning_rate": 5.714475410125241e-07,
+      "loss": 0.3819,
+      "step": 10835
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.0042796740810926,
+      "learning_rate": 5.708572678339902e-07,
+      "loss": 0.4414,
+      "step": 10836
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.2880602654563855,
+      "learning_rate": 5.702672812151977e-07,
+      "loss": 0.3797,
+      "step": 10837
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6003581605275439,
+      "learning_rate": 5.696775811943167e-07,
+      "loss": 0.4342,
+      "step": 10838
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0254003841594748,
+      "learning_rate": 5.690881678095e-07,
+      "loss": 0.4486,
+      "step": 10839
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.6130870161573747,
+      "learning_rate": 5.684990410988833e-07,
+      "loss": 0.4116,
+      "step": 10840
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5333646815175068,
+      "learning_rate": 5.679102011005821e-07,
+      "loss": 0.4902,
+      "step": 10841
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6606103665435128,
+      "learning_rate": 5.673216478526916e-07,
+      "loss": 0.4655,
+      "step": 10842
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4302699390926856,
+      "learning_rate": 5.667333813932924e-07,
+      "loss": 0.3913,
+      "step": 10843
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.128687128433247,
+      "learning_rate": 5.66145401760444e-07,
+      "loss": 0.4219,
+      "step": 10844
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4151790613071067,
+      "learning_rate": 5.655577089921887e-07,
+      "loss": 0.4531,
+      "step": 10845
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.960544175415978,
+      "learning_rate": 5.649703031265463e-07,
+      "loss": 0.4136,
+      "step": 10846
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5500490480852088,
+      "learning_rate": 5.643831842015252e-07,
+      "loss": 0.4777,
+      "step": 10847
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5071035218244582,
+      "learning_rate": 5.637963522551099e-07,
+      "loss": 0.4581,
+      "step": 10848
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5424968768062859,
+      "learning_rate": 5.632098073252668e-07,
+      "loss": 0.4728,
+      "step": 10849
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9658832981977032,
+      "learning_rate": 5.626235494499449e-07,
+      "loss": 0.4336,
+      "step": 10850
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.806582633322733,
+      "learning_rate": 5.620375786670746e-07,
+      "loss": 0.4661,
+      "step": 10851
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.019587679592214,
+      "learning_rate": 5.614518950145687e-07,
+      "loss": 0.4553,
+      "step": 10852
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9067432855093487,
+      "learning_rate": 5.608664985303175e-07,
+      "loss": 0.3983,
+      "step": 10853
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6535079530856447,
+      "learning_rate": 5.602813892521963e-07,
+      "loss": 0.432,
+      "step": 10854
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3740540783042183,
+      "learning_rate": 5.596965672180621e-07,
+      "loss": 0.4499,
+      "step": 10855
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5738241806876595,
+      "learning_rate": 5.591120324657518e-07,
+      "loss": 0.4707,
+      "step": 10856
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 4.244159279122046,
+      "learning_rate": 5.585277850330828e-07,
+      "loss": 0.4346,
+      "step": 10857
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7453393786132343,
+      "learning_rate": 5.579438249578551e-07,
+      "loss": 0.4434,
+      "step": 10858
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6469922641319328,
+      "learning_rate": 5.573601522778527e-07,
+      "loss": 0.4563,
+      "step": 10859
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5708469778495701,
+      "learning_rate": 5.56776767030836e-07,
+      "loss": 0.4925,
+      "step": 10860
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7150207591033215,
+      "learning_rate": 5.561936692545511e-07,
+      "loss": 0.4137,
+      "step": 10861
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.554753261979144,
+      "learning_rate": 5.556108589867204e-07,
+      "loss": 0.463,
+      "step": 10862
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8912597914779397,
+      "learning_rate": 5.550283362650549e-07,
+      "loss": 0.4092,
+      "step": 10863
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9109225767576261,
+      "learning_rate": 5.544461011272406e-07,
+      "loss": 0.4596,
+      "step": 10864
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6054952096209463,
+      "learning_rate": 5.538641536109491e-07,
+      "loss": 0.473,
+      "step": 10865
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.460717511789836,
+      "learning_rate": 5.532824937538279e-07,
+      "loss": 0.4562,
+      "step": 10866
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8958464952457232,
+      "learning_rate": 5.527011215935152e-07,
+      "loss": 0.4841,
+      "step": 10867
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4102602154278276,
+      "learning_rate": 5.521200371676205e-07,
+      "loss": 0.4459,
+      "step": 10868
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.813225515684679,
+      "learning_rate": 5.51539240513741e-07,
+      "loss": 0.4529,
+      "step": 10869
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.1192440632913203,
+      "learning_rate": 5.509587316694537e-07,
+      "loss": 0.4115,
+      "step": 10870
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5054936648933233,
+      "learning_rate": 5.503785106723158e-07,
+      "loss": 0.4656,
+      "step": 10871
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.310831228962782,
+      "learning_rate": 5.497985775598691e-07,
+      "loss": 0.4901,
+      "step": 10872
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 21.034390259467692,
+      "learning_rate": 5.492189323696312e-07,
+      "loss": 0.4391,
+      "step": 10873
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.8294287366217235,
+      "learning_rate": 5.486395751391066e-07,
+      "loss": 0.4328,
+      "step": 10874
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9746278140422306,
+      "learning_rate": 5.48060505905778e-07,
+      "loss": 0.4982,
+      "step": 10875
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5766145404121513,
+      "learning_rate": 5.474817247071118e-07,
+      "loss": 0.4914,
+      "step": 10876
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4515331033336865,
+      "learning_rate": 5.469032315805522e-07,
+      "loss": 0.4779,
+      "step": 10877
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3005267914200878,
+      "learning_rate": 5.463250265635284e-07,
+      "loss": 0.4441,
+      "step": 10878
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.0966236538316863,
+      "learning_rate": 5.457471096934492e-07,
+      "loss": 0.4824,
+      "step": 10879
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.540918847395035,
+      "learning_rate": 5.451694810077052e-07,
+      "loss": 0.4803,
+      "step": 10880
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6410188473473428,
+      "learning_rate": 5.445921405436682e-07,
+      "loss": 0.426,
+      "step": 10881
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.3069103372556707,
+      "learning_rate": 5.440150883386913e-07,
+      "loss": 0.4174,
+      "step": 10882
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3113773610892514,
+      "learning_rate": 5.434383244301094e-07,
+      "loss": 0.3944,
+      "step": 10883
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.7556081142813822,
+      "learning_rate": 5.428618488552378e-07,
+      "loss": 0.4558,
+      "step": 10884
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6830569044444779,
+      "learning_rate": 5.422856616513733e-07,
+      "loss": 0.4389,
+      "step": 10885
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5261315779865384,
+      "learning_rate": 5.417097628557955e-07,
+      "loss": 0.4508,
+      "step": 10886
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5823560919071649,
+      "learning_rate": 5.411341525057645e-07,
+      "loss": 0.4723,
+      "step": 10887
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5149878131534045,
+      "learning_rate": 5.405588306385201e-07,
+      "loss": 0.418,
+      "step": 10888
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9139853125008583,
+      "learning_rate": 5.399837972912858e-07,
+      "loss": 0.4733,
+      "step": 10889
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3057611157787754,
+      "learning_rate": 5.394090525012652e-07,
+      "loss": 0.4586,
+      "step": 10890
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8697118132545676,
+      "learning_rate": 5.388345963056451e-07,
+      "loss": 0.41,
+      "step": 10891
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8412086396994791,
+      "learning_rate": 5.382604287415893e-07,
+      "loss": 0.3979,
+      "step": 10892
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5431152410329332,
+      "learning_rate": 5.376865498462463e-07,
+      "loss": 0.4659,
+      "step": 10893
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8782354999479223,
+      "learning_rate": 5.371129596567476e-07,
+      "loss": 0.4359,
+      "step": 10894
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4738863201221684,
+      "learning_rate": 5.365396582102017e-07,
+      "loss": 0.4061,
+      "step": 10895
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.865300320619555,
+      "learning_rate": 5.359666455437018e-07,
+      "loss": 0.4416,
+      "step": 10896
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9807918399546878,
+      "learning_rate": 5.353939216943183e-07,
+      "loss": 0.4199,
+      "step": 10897
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5496444305304409,
+      "learning_rate": 5.348214866991097e-07,
+      "loss": 0.4407,
+      "step": 10898
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.664488471224294,
+      "learning_rate": 5.342493405951088e-07,
+      "loss": 0.4543,
+      "step": 10899
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5374334593117986,
+      "learning_rate": 5.336774834193343e-07,
+      "loss": 0.4635,
+      "step": 10900
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.359286354499065,
+      "learning_rate": 5.33105915208782e-07,
+      "loss": 0.3914,
+      "step": 10901
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6388165609626366,
+      "learning_rate": 5.325346360004357e-07,
+      "loss": 0.4339,
+      "step": 10902
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5264002417774192,
+      "learning_rate": 5.319636458312532e-07,
+      "loss": 0.425,
+      "step": 10903
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0669631589083433,
+      "learning_rate": 5.313929447381777e-07,
+      "loss": 0.5052,
+      "step": 10904
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3205715580532176,
+      "learning_rate": 5.308225327581334e-07,
+      "loss": 0.4572,
+      "step": 10905
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8672978841287766,
+      "learning_rate": 5.302524099280243e-07,
+      "loss": 0.4327,
+      "step": 10906
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4016618676585857,
+      "learning_rate": 5.296825762847385e-07,
+      "loss": 0.4522,
+      "step": 10907
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5383318893556583,
+      "learning_rate": 5.29113031865141e-07,
+      "loss": 0.4663,
+      "step": 10908
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5331274848758256,
+      "learning_rate": 5.285437767060819e-07,
+      "loss": 0.473,
+      "step": 10909
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5353462402087394,
+      "learning_rate": 5.27974810844391e-07,
+      "loss": 0.4718,
+      "step": 10910
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.467879041583342,
+      "learning_rate": 5.274061343168807e-07,
+      "loss": 0.4464,
+      "step": 10911
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.811014807292322,
+      "learning_rate": 5.268377471603414e-07,
+      "loss": 0.4355,
+      "step": 10912
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9247970920665587,
+      "learning_rate": 5.262696494115483e-07,
+      "loss": 0.4841,
+      "step": 10913
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5240870374881936,
+      "learning_rate": 5.257018411072562e-07,
+      "loss": 0.4249,
+      "step": 10914
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6038187112985207,
+      "learning_rate": 5.25134322284202e-07,
+      "loss": 0.4438,
+      "step": 10915
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.0437388161851175,
+      "learning_rate": 5.245670929791036e-07,
+      "loss": 0.4498,
+      "step": 10916
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5039161400519319,
+      "learning_rate": 5.240001532286598e-07,
+      "loss": 0.4436,
+      "step": 10917
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6662791337627645,
+      "learning_rate": 5.234335030695514e-07,
+      "loss": 0.4816,
+      "step": 10918
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5946887567226384,
+      "learning_rate": 5.228671425384385e-07,
+      "loss": 0.4654,
+      "step": 10919
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9045963341688525,
+      "learning_rate": 5.223010716719645e-07,
+      "loss": 0.4131,
+      "step": 10920
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4916448129115687,
+      "learning_rate": 5.217352905067536e-07,
+      "loss": 0.4642,
+      "step": 10921
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0009105942146745,
+      "learning_rate": 5.211697990794118e-07,
+      "loss": 0.4309,
+      "step": 10922
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.776198953977961,
+      "learning_rate": 5.206045974265245e-07,
+      "loss": 0.4137,
+      "step": 10923
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.171524575346769,
+      "learning_rate": 5.200396855846596e-07,
+      "loss": 0.4797,
+      "step": 10924
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8214414534489494,
+      "learning_rate": 5.194750635903667e-07,
+      "loss": 0.4846,
+      "step": 10925
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5260738739702246,
+      "learning_rate": 5.189107314801756e-07,
+      "loss": 0.4783,
+      "step": 10926
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.514046117209759,
+      "learning_rate": 5.183466892905997e-07,
+      "loss": 0.4224,
+      "step": 10927
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.571039921632034,
+      "learning_rate": 5.177829370581277e-07,
+      "loss": 0.4983,
+      "step": 10928
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.657610666252025,
+      "learning_rate": 5.172194748192389e-07,
+      "loss": 0.4022,
+      "step": 10929
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5325803862687203,
+      "learning_rate": 5.166563026103844e-07,
+      "loss": 0.4814,
+      "step": 10930
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.101495184303436,
+      "learning_rate": 5.160934204680029e-07,
+      "loss": 0.438,
+      "step": 10931
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.245235616243393,
+      "learning_rate": 5.155308284285098e-07,
+      "loss": 0.4456,
+      "step": 10932
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8277363218902816,
+      "learning_rate": 5.149685265283078e-07,
+      "loss": 0.4615,
+      "step": 10933
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4798322389781593,
+      "learning_rate": 5.144065148037736e-07,
+      "loss": 0.4169,
+      "step": 10934
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5303499416254304,
+      "learning_rate": 5.13844793291271e-07,
+      "loss": 0.4855,
+      "step": 10935
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9138365050382489,
+      "learning_rate": 5.132833620271399e-07,
+      "loss": 0.4387,
+      "step": 10936
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3548282942326617,
+      "learning_rate": 5.127222210477067e-07,
+      "loss": 0.4752,
+      "step": 10937
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3340372347991212,
+      "learning_rate": 5.12161370389277e-07,
+      "loss": 0.444,
+      "step": 10938
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7390982665244552,
+      "learning_rate": 5.116008100881348e-07,
+      "loss": 0.431,
+      "step": 10939
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.729033023504952,
+      "learning_rate": 5.110405401805485e-07,
+      "loss": 0.4388,
+      "step": 10940
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5368454767663509,
+      "learning_rate": 5.104805607027668e-07,
+      "loss": 0.4783,
+      "step": 10941
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6008487244985252,
+      "learning_rate": 5.099208716910209e-07,
+      "loss": 0.475,
+      "step": 10942
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7334560289882073,
+      "learning_rate": 5.0936147318152e-07,
+      "loss": 0.3736,
+      "step": 10943
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5585106415293764,
+      "learning_rate": 5.088023652104568e-07,
+      "loss": 0.4335,
+      "step": 10944
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.775819680991112,
+      "learning_rate": 5.082435478140057e-07,
+      "loss": 0.4137,
+      "step": 10945
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9220030506697252,
+      "learning_rate": 5.076850210283224e-07,
+      "loss": 0.4376,
+      "step": 10946
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.2901663379713417,
+      "learning_rate": 5.071267848895401e-07,
+      "loss": 0.4601,
+      "step": 10947
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1680285529747505,
+      "learning_rate": 5.065688394337765e-07,
+      "loss": 0.446,
+      "step": 10948
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8301262076141867,
+      "learning_rate": 5.060111846971327e-07,
+      "loss": 0.4721,
+      "step": 10949
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7211615153631636,
+      "learning_rate": 5.054538207156856e-07,
+      "loss": 0.4458,
+      "step": 10950
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5182029070851061,
+      "learning_rate": 5.048967475254968e-07,
+      "loss": 0.4562,
+      "step": 10951
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6189059306816527,
+      "learning_rate": 5.04339965162608e-07,
+      "loss": 0.4001,
+      "step": 10952
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9055168091727923,
+      "learning_rate": 5.037834736630431e-07,
+      "loss": 0.4732,
+      "step": 10953
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1205733333738115,
+      "learning_rate": 5.032272730628052e-07,
+      "loss": 0.4328,
+      "step": 10954
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9872429198721748,
+      "learning_rate": 5.026713633978797e-07,
+      "loss": 0.4345,
+      "step": 10955
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 4.802664063243212,
+      "learning_rate": 5.021157447042341e-07,
+      "loss": 0.4033,
+      "step": 10956
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.398615388778024,
+      "learning_rate": 5.015604170178168e-07,
+      "loss": 0.4453,
+      "step": 10957
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5025946655643436,
+      "learning_rate": 5.010053803745546e-07,
+      "loss": 0.4612,
+      "step": 10958
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3875108706688568,
+      "learning_rate": 5.004506348103594e-07,
+      "loss": 0.3964,
+      "step": 10959
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2982524845645382,
+      "learning_rate": 4.998961803611213e-07,
+      "loss": 0.4712,
+      "step": 10960
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3033359280099868,
+      "learning_rate": 4.993420170627139e-07,
+      "loss": 0.4223,
+      "step": 10961
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7744552376192633,
+      "learning_rate": 4.987881449509913e-07,
+      "loss": 0.4591,
+      "step": 10962
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6700940168289402,
+      "learning_rate": 4.982345640617853e-07,
+      "loss": 0.4446,
+      "step": 10963
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6172525467242715,
+      "learning_rate": 4.976812744309156e-07,
+      "loss": 0.404,
+      "step": 10964
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5541547890582302,
+      "learning_rate": 4.97128276094177e-07,
+      "loss": 0.4487,
+      "step": 10965
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.795180220052461,
+      "learning_rate": 4.965755690873497e-07,
+      "loss": 0.4972,
+      "step": 10966
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.772764178635714,
+      "learning_rate": 4.960231534461896e-07,
+      "loss": 0.3899,
+      "step": 10967
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6916870512908477,
+      "learning_rate": 4.954710292064418e-07,
+      "loss": 0.4254,
+      "step": 10968
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6832493013628937,
+      "learning_rate": 4.949191964038241e-07,
+      "loss": 0.4546,
+      "step": 10969
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4603010368650682,
+      "learning_rate": 4.943676550740428e-07,
+      "loss": 0.4426,
+      "step": 10970
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.957370740015821,
+      "learning_rate": 4.938164052527778e-07,
+      "loss": 0.4446,
+      "step": 10971
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1051761391196138,
+      "learning_rate": 4.932654469756976e-07,
+      "loss": 0.4048,
+      "step": 10972
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4558718063756322,
+      "learning_rate": 4.927147802784482e-07,
+      "loss": 0.4611,
+      "step": 10973
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5121296279783231,
+      "learning_rate": 4.921644051966551e-07,
+      "loss": 0.4746,
+      "step": 10974
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7378633273797575,
+      "learning_rate": 4.916143217659286e-07,
+      "loss": 0.4456,
+      "step": 10975
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.5963217259718285,
+      "learning_rate": 4.91064530021857e-07,
+      "loss": 0.4507,
+      "step": 10976
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5914908812146715,
+      "learning_rate": 4.905150300000133e-07,
+      "loss": 0.4951,
+      "step": 10977
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5716469843884027,
+      "learning_rate": 4.899658217359471e-07,
+      "loss": 0.4683,
+      "step": 10978
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0361782927180196,
+      "learning_rate": 4.894169052651926e-07,
+      "loss": 0.4851,
+      "step": 10979
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7434227998681935,
+      "learning_rate": 4.888682806232631e-07,
+      "loss": 0.4674,
+      "step": 10980
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4416440587621024,
+      "learning_rate": 4.883199478456558e-07,
+      "loss": 0.4076,
+      "step": 10981
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.175439341363509,
+      "learning_rate": 4.877719069678449e-07,
+      "loss": 0.4031,
+      "step": 10982
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8180191666402474,
+      "learning_rate": 4.872241580252879e-07,
+      "loss": 0.4398,
+      "step": 10983
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0221037309435874,
+      "learning_rate": 4.866767010534268e-07,
+      "loss": 0.4792,
+      "step": 10984
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.0927173211045034,
+      "learning_rate": 4.861295360876772e-07,
+      "loss": 0.4248,
+      "step": 10985
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.521900667936592,
+      "learning_rate": 4.855826631634425e-07,
+      "loss": 0.4394,
+      "step": 10986
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.962458602319682,
+      "learning_rate": 4.850360823161032e-07,
+      "loss": 0.4409,
+      "step": 10987
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.8288190717926525,
+      "learning_rate": 4.844897935810245e-07,
+      "loss": 0.4034,
+      "step": 10988
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5754956543457765,
+      "learning_rate": 4.839437969935479e-07,
+      "loss": 0.3953,
+      "step": 10989
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4370702174886207,
+      "learning_rate": 4.833980925890003e-07,
+      "loss": 0.4565,
+      "step": 10990
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8381136063999433,
+      "learning_rate": 4.828526804026872e-07,
+      "loss": 0.4597,
+      "step": 10991
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.4472657339368933,
+      "learning_rate": 4.823075604698979e-07,
+      "loss": 0.3795,
+      "step": 10992
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9818725862183755,
+      "learning_rate": 4.817627328258983e-07,
+      "loss": 0.4473,
+      "step": 10993
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.50577307359984,
+      "learning_rate": 4.812181975059393e-07,
+      "loss": 0.4801,
+      "step": 10994
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5451923508688805,
+      "learning_rate": 4.806739545452516e-07,
+      "loss": 0.4783,
+      "step": 10995
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6751859823259365,
+      "learning_rate": 4.801300039790469e-07,
+      "loss": 0.4645,
+      "step": 10996
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5079735674860286,
+      "learning_rate": 4.795863458425193e-07,
+      "loss": 0.4326,
+      "step": 10997
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8596774659334627,
+      "learning_rate": 4.790429801708408e-07,
+      "loss": 0.4505,
+      "step": 10998
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6828437265802594,
+      "learning_rate": 4.784999069991675e-07,
+      "loss": 0.4265,
+      "step": 10999
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.852011303837463,
+      "learning_rate": 4.779571263626348e-07,
+      "loss": 0.4021,
+      "step": 11000
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9506692527695206,
+      "learning_rate": 4.774146382963618e-07,
+      "loss": 0.458,
+      "step": 11001
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8583617719503498,
+      "learning_rate": 4.768724428354432e-07,
+      "loss": 0.4738,
+      "step": 11002
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.193938393382482,
+      "learning_rate": 4.7633054001496313e-07,
+      "loss": 0.4351,
+      "step": 11003
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5042517900625831,
+      "learning_rate": 4.7578892986997804e-07,
+      "loss": 0.415,
+      "step": 11004
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.115526154518036,
+      "learning_rate": 4.752476124355304e-07,
+      "loss": 0.4728,
+      "step": 11005
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.587713683965019,
+      "learning_rate": 4.7470658774664334e-07,
+      "loss": 0.3938,
+      "step": 11006
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6697975863066161,
+      "learning_rate": 4.7416585583832054e-07,
+      "loss": 0.4207,
+      "step": 11007
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.9247078608840664,
+      "learning_rate": 4.736254167455473e-07,
+      "loss": 0.4493,
+      "step": 11008
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 3.6483100345600423,
+      "learning_rate": 4.730852705032868e-07,
+      "loss": 0.46,
+      "step": 11009
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7651148539912696,
+      "learning_rate": 4.725454171464883e-07,
+      "loss": 0.5046,
+      "step": 11010
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.3647511722311183,
+      "learning_rate": 4.720058567100777e-07,
+      "loss": 0.501,
+      "step": 11011
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.8047554035639717,
+      "learning_rate": 4.7146658922896607e-07,
+      "loss": 0.4499,
+      "step": 11012
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5396275407841262,
+      "learning_rate": 4.709276147380415e-07,
+      "loss": 0.4186,
+      "step": 11013
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.373214488335486,
+      "learning_rate": 4.7038893327217504e-07,
+      "loss": 0.4491,
+      "step": 11014
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5752977691237755,
+      "learning_rate": 4.698505448662194e-07,
+      "loss": 0.4068,
+      "step": 11015
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6235367153773126,
+      "learning_rate": 4.693124495550072e-07,
+      "loss": 0.4333,
+      "step": 11016
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9077259348158078,
+      "learning_rate": 4.6877464737335344e-07,
+      "loss": 0.4533,
+      "step": 11017
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0716259787283136,
+      "learning_rate": 4.682371383560508e-07,
+      "loss": 0.4502,
+      "step": 11018
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 8.23668520906571,
+      "learning_rate": 4.6769992253787865e-07,
+      "loss": 0.4499,
+      "step": 11019
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8833382083150905,
+      "learning_rate": 4.671629999535915e-07,
+      "loss": 0.4457,
+      "step": 11020
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9997230258626286,
+      "learning_rate": 4.666263706379287e-07,
+      "loss": 0.4033,
+      "step": 11021
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5658059775292958,
+      "learning_rate": 4.660900346256098e-07,
+      "loss": 0.462,
+      "step": 11022
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.964958672410041,
+      "learning_rate": 4.655539919513347e-07,
+      "loss": 0.519,
+      "step": 11023
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9088727839339725,
+      "learning_rate": 4.65018242649784e-07,
+      "loss": 0.4605,
+      "step": 11024
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1864388778310757,
+      "learning_rate": 4.6448278675562066e-07,
+      "loss": 0.4195,
+      "step": 11025
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6351581907831445,
+      "learning_rate": 4.639476243034874e-07,
+      "loss": 0.4386,
+      "step": 11026
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.464644086493101,
+      "learning_rate": 4.6341275532800936e-07,
+      "loss": 0.4323,
+      "step": 11027
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7431147858647225,
+      "learning_rate": 4.628781798637921e-07,
+      "loss": 0.4388,
+      "step": 11028
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7435532121922386,
+      "learning_rate": 4.623438979454203e-07,
+      "loss": 0.4644,
+      "step": 11029
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.531897520069842,
+      "learning_rate": 4.618099096074624e-07,
+      "loss": 0.4585,
+      "step": 11030
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.548102434405965,
+      "learning_rate": 4.612762148844668e-07,
+      "loss": 0.4373,
+      "step": 11031
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5863252569500517,
+      "learning_rate": 4.607428138109632e-07,
+      "loss": 0.4799,
+      "step": 11032
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5341649658252945,
+      "learning_rate": 4.602097064214611e-07,
+      "loss": 0.4108,
+      "step": 11033
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8485506589368792,
+      "learning_rate": 4.596768927504519e-07,
+      "loss": 0.4746,
+      "step": 11034
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.406488734472628,
+      "learning_rate": 4.5914437283240797e-07,
+      "loss": 0.4303,
+      "step": 11035
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1023013369516166,
+      "learning_rate": 4.5861214670178454e-07,
+      "loss": 0.4305,
+      "step": 11036
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.371459638917895,
+      "learning_rate": 4.5808021439301187e-07,
+      "loss": 0.4348,
+      "step": 11037
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.7219818569349385,
+      "learning_rate": 4.5754857594050905e-07,
+      "loss": 0.4844,
+      "step": 11038
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3413849910636684,
+      "learning_rate": 4.570172313786719e-07,
+      "loss": 0.4409,
+      "step": 11039
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7792495385356395,
+      "learning_rate": 4.5648618074187576e-07,
+      "loss": 0.4492,
+      "step": 11040
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.809878302252569,
+      "learning_rate": 4.5595542406448037e-07,
+      "loss": 0.4651,
+      "step": 11041
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5277547249318906,
+      "learning_rate": 4.5542496138082483e-07,
+      "loss": 0.4039,
+      "step": 11042
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2756840290280387,
+      "learning_rate": 4.5489479272522954e-07,
+      "loss": 0.4653,
+      "step": 11043
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.234351622996232,
+      "learning_rate": 4.5436491813199537e-07,
+      "loss": 0.4686,
+      "step": 11044
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4289956882369947,
+      "learning_rate": 4.538353376354038e-07,
+      "loss": 0.4557,
+      "step": 11045
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4691577427748748,
+      "learning_rate": 4.5330605126971906e-07,
+      "loss": 0.4013,
+      "step": 11046
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.75912939774734,
+      "learning_rate": 4.52777059069186e-07,
+      "loss": 0.5048,
+      "step": 11047
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.631659226972406,
+      "learning_rate": 4.5224836106802725e-07,
+      "loss": 0.4171,
+      "step": 11048
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5459743177321604,
+      "learning_rate": 4.5171995730045103e-07,
+      "loss": 0.4701,
+      "step": 11049
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8045602501803502,
+      "learning_rate": 4.5119184780064326e-07,
+      "loss": 0.439,
+      "step": 11050
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2593935877446603,
+      "learning_rate": 4.5066403260277226e-07,
+      "loss": 0.4548,
+      "step": 11051
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7272508472345955,
+      "learning_rate": 4.5013651174098837e-07,
+      "loss": 0.4533,
+      "step": 11052
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.5700512231630976,
+      "learning_rate": 4.496092852494183e-07,
+      "loss": 0.4275,
+      "step": 11053
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7779662086750219,
+      "learning_rate": 4.490823531621763e-07,
+      "loss": 0.5003,
+      "step": 11054
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.755303279194439,
+      "learning_rate": 4.485557155133524e-07,
+      "loss": 0.4891,
+      "step": 11055
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3269153614416394,
+      "learning_rate": 4.480293723370188e-07,
+      "loss": 0.3829,
+      "step": 11056
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5321619909982683,
+      "learning_rate": 4.4750332366723103e-07,
+      "loss": 0.463,
+      "step": 11057
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7386251950915632,
+      "learning_rate": 4.4697756953802296e-07,
+      "loss": 0.451,
+      "step": 11058
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1384207812147475,
+      "learning_rate": 4.464521099834096e-07,
+      "loss": 0.5124,
+      "step": 11059
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5470255194597082,
+      "learning_rate": 4.4592694503738767e-07,
+      "loss": 0.4803,
+      "step": 11060
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8946665068640247,
+      "learning_rate": 4.4540207473393495e-07,
+      "loss": 0.478,
+      "step": 11061
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9282180672194968,
+      "learning_rate": 4.4487749910700983e-07,
+      "loss": 0.4003,
+      "step": 11062
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 5.435205291567352,
+      "learning_rate": 4.44353218190553e-07,
+      "loss": 0.4354,
+      "step": 11063
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8126858841584292,
+      "learning_rate": 4.438292320184817e-07,
+      "loss": 0.3953,
+      "step": 11064
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.747794806915221,
+      "learning_rate": 4.4330554062469944e-07,
+      "loss": 0.5295,
+      "step": 11065
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9048302633146534,
+      "learning_rate": 4.427821440430879e-07,
+      "loss": 0.4653,
+      "step": 11066
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4801072815876237,
+      "learning_rate": 4.422590423075107e-07,
+      "loss": 0.4039,
+      "step": 11067
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.2195304546967476,
+      "learning_rate": 4.417362354518101e-07,
+      "loss": 0.4236,
+      "step": 11068
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.03114532722095,
+      "learning_rate": 4.4121372350981237e-07,
+      "loss": 0.4227,
+      "step": 11069
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5380196553927141,
+      "learning_rate": 4.406915065153233e-07,
+      "loss": 0.4577,
+      "step": 11070
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5299973016113624,
+      "learning_rate": 4.401695845021298e-07,
+      "loss": 0.4727,
+      "step": 11071
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1166077012706555,
+      "learning_rate": 4.3964795750399746e-07,
+      "loss": 0.4643,
+      "step": 11072
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9628230671405245,
+      "learning_rate": 4.3912662555467775e-07,
+      "loss": 0.4022,
+      "step": 11073
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.428174727125141,
+      "learning_rate": 4.386055886878998e-07,
+      "loss": 0.4568,
+      "step": 11074
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5599697611706638,
+      "learning_rate": 4.380848469373722e-07,
+      "loss": 0.4212,
+      "step": 11075
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3493566910356547,
+      "learning_rate": 4.375644003367874e-07,
+      "loss": 0.4529,
+      "step": 11076
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5665960389249406,
+      "learning_rate": 4.370442489198179e-07,
+      "loss": 0.4971,
+      "step": 11077
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.640238378215373,
+      "learning_rate": 4.365243927201168e-07,
+      "loss": 0.4231,
+      "step": 11078
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0784165762985,
+      "learning_rate": 4.360048317713167e-07,
+      "loss": 0.4206,
+      "step": 11079
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.783943435077842,
+      "learning_rate": 4.3548556610703394e-07,
+      "loss": 0.4323,
+      "step": 11080
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5721848985811098,
+      "learning_rate": 4.349665957608634e-07,
+      "loss": 0.4869,
+      "step": 11081
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7403417423485206,
+      "learning_rate": 4.3444792076638377e-07,
+      "loss": 0.4733,
+      "step": 11082
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5093830256977612,
+      "learning_rate": 4.339295411571498e-07,
+      "loss": 0.4959,
+      "step": 11083
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5818654274557407,
+      "learning_rate": 4.3341145696670086e-07,
+      "loss": 0.4678,
+      "step": 11084
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.467640757569211,
+      "learning_rate": 4.3289366822855837e-07,
+      "loss": 0.4348,
+      "step": 11085
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.107402953373491,
+      "learning_rate": 4.323761749762201e-07,
+      "loss": 0.4847,
+      "step": 11086
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0434247241650736,
+      "learning_rate": 4.318589772431686e-07,
+      "loss": 0.4694,
+      "step": 11087
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5864197595826266,
+      "learning_rate": 4.313420750628644e-07,
+      "loss": 0.4528,
+      "step": 11088
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5747658943338664,
+      "learning_rate": 4.308254684687524e-07,
+      "loss": 0.4639,
+      "step": 11089
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1738253868782342,
+      "learning_rate": 4.3030915749425527e-07,
+      "loss": 0.4832,
+      "step": 11090
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5485963096847947,
+      "learning_rate": 4.2979314217277746e-07,
+      "loss": 0.4454,
+      "step": 11091
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4513759898879213,
+      "learning_rate": 4.292774225377044e-07,
+      "loss": 0.4551,
+      "step": 11092
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.359294888208332,
+      "learning_rate": 4.2876199862240397e-07,
+      "loss": 0.4143,
+      "step": 11093
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.055505843926869,
+      "learning_rate": 4.282468704602216e-07,
+      "loss": 0.4714,
+      "step": 11094
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7351728595391682,
+      "learning_rate": 4.2773203808448573e-07,
+      "loss": 0.4229,
+      "step": 11095
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.842195055628895,
+      "learning_rate": 4.272175015285057e-07,
+      "loss": 0.4131,
+      "step": 11096
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7882111978495752,
+      "learning_rate": 4.2670326082557167e-07,
+      "loss": 0.4603,
+      "step": 11097
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7327045972385173,
+      "learning_rate": 4.261893160089553e-07,
+      "loss": 0.4475,
+      "step": 11098
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.650102187519729,
+      "learning_rate": 4.2567566711190554e-07,
+      "loss": 0.4435,
+      "step": 11099
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7589804294691203,
+      "learning_rate": 4.2516231416765583e-07,
+      "loss": 0.446,
+      "step": 11100
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8431037181242336,
+      "learning_rate": 4.246492572094202e-07,
+      "loss": 0.4329,
+      "step": 11101
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5659828880472241,
+      "learning_rate": 4.241364962703931e-07,
+      "loss": 0.4797,
+      "step": 11102
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8298914890583236,
+      "learning_rate": 4.236240313837475e-07,
+      "loss": 0.4847,
+      "step": 11103
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9504322843087247,
+      "learning_rate": 4.231118625826408e-07,
+      "loss": 0.4555,
+      "step": 11104
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.527842224653605,
+      "learning_rate": 4.2259998990020925e-07,
+      "loss": 0.4283,
+      "step": 11105
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9166044470507195,
+      "learning_rate": 4.2208841336957075e-07,
+      "loss": 0.4529,
+      "step": 11106
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.098630572928746,
+      "learning_rate": 4.2157713302382277e-07,
+      "loss": 0.422,
+      "step": 11107
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8797270150576728,
+      "learning_rate": 4.210661488960449e-07,
+      "loss": 0.4424,
+      "step": 11108
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4261053134837747,
+      "learning_rate": 4.20555461019298e-07,
+      "loss": 0.3732,
+      "step": 11109
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5512720004015147,
+      "learning_rate": 4.200450694266212e-07,
+      "loss": 0.4631,
+      "step": 11110
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5860907932607615,
+      "learning_rate": 4.195349741510374e-07,
+      "loss": 0.4602,
+      "step": 11111
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8385758329757602,
+      "learning_rate": 4.1902517522554866e-07,
+      "loss": 0.4846,
+      "step": 11112
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.5236980189463631,
+      "learning_rate": 4.185156726831391e-07,
+      "loss": 0.4259,
+      "step": 11113
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6843359058832088,
+      "learning_rate": 4.180064665567718e-07,
+      "loss": 0.4068,
+      "step": 11114
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5489712408136727,
+      "learning_rate": 4.174975568793915e-07,
+      "loss": 0.4696,
+      "step": 11115
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.714910879588225,
+      "learning_rate": 4.1698894368392527e-07,
+      "loss": 0.4368,
+      "step": 11116
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.065725231462918,
+      "learning_rate": 4.1648062700327953e-07,
+      "loss": 0.4302,
+      "step": 11117
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.6895355850815235,
+      "learning_rate": 4.1597260687034013e-07,
+      "loss": 0.4604,
+      "step": 11118
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.3833732998140564,
+      "learning_rate": 4.1546488331797587e-07,
+      "loss": 0.4258,
+      "step": 11119
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.4675202470914415,
+      "learning_rate": 4.149574563790382e-07,
+      "loss": 0.4398,
+      "step": 11120
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.9140230969136391,
+      "learning_rate": 4.144503260863536e-07,
+      "loss": 0.4241,
+      "step": 11121
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6376183652564233,
+      "learning_rate": 4.139434924727359e-07,
+      "loss": 0.4102,
+      "step": 11122
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5572895620921952,
+      "learning_rate": 4.1343695557097273e-07,
+      "loss": 0.4689,
+      "step": 11123
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7408529025844286,
+      "learning_rate": 4.129307154138401e-07,
+      "loss": 0.4504,
+      "step": 11124
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.098053940959106,
+      "learning_rate": 4.1242477203408904e-07,
+      "loss": 0.5142,
+      "step": 11125
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.561504258885627,
+      "learning_rate": 4.119191254644539e-07,
+      "loss": 0.474,
+      "step": 11126
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 3.8280724743599834,
+      "learning_rate": 4.1141377573764906e-07,
+      "loss": 0.4469,
+      "step": 11127
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7690310674729592,
+      "learning_rate": 4.1090872288637174e-07,
+      "loss": 0.4564,
+      "step": 11128
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5371923263062188,
+      "learning_rate": 4.1040396694329576e-07,
+      "loss": 0.4641,
+      "step": 11129
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.7144337657109012,
+      "learning_rate": 4.098995079410789e-07,
+      "loss": 0.4896,
+      "step": 11130
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5535176915037792,
+      "learning_rate": 4.093953459123595e-07,
+      "loss": 0.4849,
+      "step": 11131
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.6974414781112153,
+      "learning_rate": 4.088914808897554e-07,
+      "loss": 0.3875,
+      "step": 11132
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4327025747543147,
+      "learning_rate": 4.083879129058682e-07,
+      "loss": 0.4006,
+      "step": 11133
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.531642427612457,
+      "learning_rate": 4.078846419932753e-07,
+      "loss": 0.4116,
+      "step": 11134
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3440089981440855,
+      "learning_rate": 4.0738166818453883e-07,
+      "loss": 0.4336,
+      "step": 11135
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.1745102060302384,
+      "learning_rate": 4.0687899151220013e-07,
+      "loss": 0.487,
+      "step": 11136
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8795258286131635,
+      "learning_rate": 4.0637661200878363e-07,
+      "loss": 0.3881,
+      "step": 11137
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.880244844636084,
+      "learning_rate": 4.0587452970678953e-07,
+      "loss": 0.4214,
+      "step": 11138
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4795103430022716,
+      "learning_rate": 4.05372744638704e-07,
+      "loss": 0.4745,
+      "step": 11139
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 4.495783190228966,
+      "learning_rate": 4.048712568369911e-07,
+      "loss": 0.4568,
+      "step": 11140
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0233661883997396,
+      "learning_rate": 4.0437006633409715e-07,
+      "loss": 0.4748,
+      "step": 11141
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49765040198675514,
+      "learning_rate": 4.038691731624478e-07,
+      "loss": 0.4688,
+      "step": 11142
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3885208219486485,
+      "learning_rate": 4.0336857735445043e-07,
+      "loss": 0.4363,
+      "step": 11143
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.017277712770393,
+      "learning_rate": 4.028682789424937e-07,
+      "loss": 0.4493,
+      "step": 11144
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5870012594064757,
+      "learning_rate": 4.0236827795894497e-07,
+      "loss": 0.4356,
+      "step": 11145
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4991139979388177,
+      "learning_rate": 4.018685744361539e-07,
+      "loss": 0.4765,
+      "step": 11146
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0035029807323346,
+      "learning_rate": 4.0136916840645077e-07,
+      "loss": 0.4185,
+      "step": 11147
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8553420701018526,
+      "learning_rate": 4.0087005990214813e-07,
+      "loss": 0.4239,
+      "step": 11148
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6431159990692625,
+      "learning_rate": 4.00371248955535e-07,
+      "loss": 0.4547,
+      "step": 11149
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.006932616398925,
+      "learning_rate": 3.998727355988852e-07,
+      "loss": 0.4364,
+      "step": 11150
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6588544935478375,
+      "learning_rate": 3.9937451986445106e-07,
+      "loss": 0.4341,
+      "step": 11151
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.607829236175498,
+      "learning_rate": 3.9887660178446753e-07,
+      "loss": 0.4366,
+      "step": 11152
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6364493090416499,
+      "learning_rate": 3.983789813911498e-07,
+      "loss": 0.4478,
+      "step": 11153
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6423815707995204,
+      "learning_rate": 3.978816587166906e-07,
+      "loss": 0.4389,
+      "step": 11154
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5917284737552029,
+      "learning_rate": 3.9738463379326965e-07,
+      "loss": 0.4257,
+      "step": 11155
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8601407457175227,
+      "learning_rate": 3.968879066530407e-07,
+      "loss": 0.4395,
+      "step": 11156
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.71648421522421,
+      "learning_rate": 3.9639147732814356e-07,
+      "loss": 0.4904,
+      "step": 11157
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7223844515753464,
+      "learning_rate": 3.9589534585069366e-07,
+      "loss": 0.3652,
+      "step": 11158
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2905526601579997,
+      "learning_rate": 3.9539951225279374e-07,
+      "loss": 0.4261,
+      "step": 11159
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9678126799641222,
+      "learning_rate": 3.949039765665208e-07,
+      "loss": 0.4459,
+      "step": 11160
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7322365804689541,
+      "learning_rate": 3.944087388239376e-07,
+      "loss": 0.4157,
+      "step": 11161
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.429809090981671,
+      "learning_rate": 3.9391379905708237e-07,
+      "loss": 0.4199,
+      "step": 11162
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5127176248970764,
+      "learning_rate": 3.934191572979795e-07,
+      "loss": 0.4496,
+      "step": 11163
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5215306684772689,
+      "learning_rate": 3.9292481357863175e-07,
+      "loss": 0.445,
+      "step": 11164
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5355196048807527,
+      "learning_rate": 3.9243076793102077e-07,
+      "loss": 0.4505,
+      "step": 11165
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9755860829691727,
+      "learning_rate": 3.91937020387112e-07,
+      "loss": 0.4156,
+      "step": 11166
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.770825133660042,
+      "learning_rate": 3.9144357097885e-07,
+      "loss": 0.4661,
+      "step": 11167
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6167660160888846,
+      "learning_rate": 3.909504197381608e-07,
+      "loss": 0.4447,
+      "step": 11168
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.168457025153742,
+      "learning_rate": 3.90457566696949e-07,
+      "loss": 0.4483,
+      "step": 11169
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2577243769088193,
+      "learning_rate": 3.899650118871029e-07,
+      "loss": 0.428,
+      "step": 11170
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9777920249551537,
+      "learning_rate": 3.8947275534048977e-07,
+      "loss": 0.4255,
+      "step": 11171
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5827854011512574,
+      "learning_rate": 3.8898079708895917e-07,
+      "loss": 0.481,
+      "step": 11172
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6881681193543336,
+      "learning_rate": 3.884891371643379e-07,
+      "loss": 0.4776,
+      "step": 11173
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9202001014464545,
+      "learning_rate": 3.87997775598436e-07,
+      "loss": 0.4255,
+      "step": 11174
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.799322529494823,
+      "learning_rate": 3.8750671242304636e-07,
+      "loss": 0.4953,
+      "step": 11175
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5274787062340993,
+      "learning_rate": 3.870159476699381e-07,
+      "loss": 0.4343,
+      "step": 11176
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5083134779134512,
+      "learning_rate": 3.8652548137086297e-07,
+      "loss": 0.4482,
+      "step": 11177
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8203742842364898,
+      "learning_rate": 3.8603531355755454e-07,
+      "loss": 0.4405,
+      "step": 11178
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9579317628937916,
+      "learning_rate": 3.8554544426172566e-07,
+      "loss": 0.4467,
+      "step": 11179
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5503509841144032,
+      "learning_rate": 3.8505587351507003e-07,
+      "loss": 0.475,
+      "step": 11180
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1232782654738944,
+      "learning_rate": 3.845666013492616e-07,
+      "loss": 0.4173,
+      "step": 11181
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.210604519155103,
+      "learning_rate": 3.8407762779595623e-07,
+      "loss": 0.4128,
+      "step": 11182
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7613806451425842,
+      "learning_rate": 3.835889528867914e-07,
+      "loss": 0.4544,
+      "step": 11183
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.7304040712443616,
+      "learning_rate": 3.831005766533813e-07,
+      "loss": 0.4118,
+      "step": 11184
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.467988302086924,
+      "learning_rate": 3.826124991273239e-07,
+      "loss": 0.4517,
+      "step": 11185
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4506887612500328,
+      "learning_rate": 3.8212472034019787e-07,
+      "loss": 0.4643,
+      "step": 11186
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.19762559112552,
+      "learning_rate": 3.816372403235613e-07,
+      "loss": 0.4216,
+      "step": 11187
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5428591222245889,
+      "learning_rate": 3.8115005910895453e-07,
+      "loss": 0.4696,
+      "step": 11188
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6709501256682693,
+      "learning_rate": 3.806631767278951e-07,
+      "loss": 0.4457,
+      "step": 11189
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.347818018623209,
+      "learning_rate": 3.801765932118873e-07,
+      "loss": 0.3829,
+      "step": 11190
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.134665074654038,
+      "learning_rate": 3.7969030859240917e-07,
+      "loss": 0.4179,
+      "step": 11191
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9096282455278326,
+      "learning_rate": 3.792043229009246e-07,
+      "loss": 0.4159,
+      "step": 11192
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8413833469122418,
+      "learning_rate": 3.787186361688744e-07,
+      "loss": 0.4563,
+      "step": 11193
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9950115920133202,
+      "learning_rate": 3.7823324842768403e-07,
+      "loss": 0.5158,
+      "step": 11194
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.31672365372418,
+      "learning_rate": 3.7774815970875624e-07,
+      "loss": 0.4626,
+      "step": 11195
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.528051575456564,
+      "learning_rate": 3.772633700434769e-07,
+      "loss": 0.4833,
+      "step": 11196
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.2679686430784325,
+      "learning_rate": 3.7677887946320767e-07,
+      "loss": 0.4963,
+      "step": 11197
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0822057107203524,
+      "learning_rate": 3.7629468799929845e-07,
+      "loss": 0.4516,
+      "step": 11198
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2584542499398155,
+      "learning_rate": 3.758107956830753e-07,
+      "loss": 0.458,
+      "step": 11199
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.05002080538384,
+      "learning_rate": 3.753272025458432e-07,
+      "loss": 0.4493,
+      "step": 11200
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3085742794440849,
+      "learning_rate": 3.7484390861889153e-07,
+      "loss": 0.3982,
+      "step": 11201
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.4998934254205554,
+      "learning_rate": 3.743609139334886e-07,
+      "loss": 0.479,
+      "step": 11202
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.0320732209662187,
+      "learning_rate": 3.738782185208839e-07,
+      "loss": 0.4924,
+      "step": 11203
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8210167891000517,
+      "learning_rate": 3.7339582241230634e-07,
+      "loss": 0.4543,
+      "step": 11204
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3772200798027945,
+      "learning_rate": 3.7291372563896643e-07,
+      "loss": 0.4402,
+      "step": 11205
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4829788465856155,
+      "learning_rate": 3.7243192823205544e-07,
+      "loss": 0.4741,
+      "step": 11206
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1363770255691312,
+      "learning_rate": 3.719504302227461e-07,
+      "loss": 0.414,
+      "step": 11207
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5055676598897723,
+      "learning_rate": 3.714692316421886e-07,
+      "loss": 0.4652,
+      "step": 11208
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5650562037343111,
+      "learning_rate": 3.7098833252151633e-07,
+      "loss": 0.4449,
+      "step": 11209
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.053416737057283,
+      "learning_rate": 3.7050773289184494e-07,
+      "loss": 0.4279,
+      "step": 11210
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9053979519239272,
+      "learning_rate": 3.700274327842662e-07,
+      "loss": 0.4762,
+      "step": 11211
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.131708930989026,
+      "learning_rate": 3.6954743222985644e-07,
+      "loss": 0.4525,
+      "step": 11212
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.9492309929440115,
+      "learning_rate": 3.6906773125966963e-07,
+      "loss": 0.4571,
+      "step": 11213
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2333894648454606,
+      "learning_rate": 3.685883299047438e-07,
+      "loss": 0.4187,
+      "step": 11214
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.80297110438348,
+      "learning_rate": 3.681092281960935e-07,
+      "loss": 0.4388,
+      "step": 11215
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6756412323669294,
+      "learning_rate": 3.676304261647168e-07,
+      "loss": 0.4211,
+      "step": 11216
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1468896816334526,
+      "learning_rate": 3.671519238415916e-07,
+      "loss": 0.4483,
+      "step": 11217
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9665495799666641,
+      "learning_rate": 3.666737212576771e-07,
+      "loss": 0.4362,
+      "step": 11218
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5084502946220684,
+      "learning_rate": 3.661958184439113e-07,
+      "loss": 0.3862,
+      "step": 11219
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.247172816702885,
+      "learning_rate": 3.657182154312139e-07,
+      "loss": 0.5086,
+      "step": 11220
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4733995805753137,
+      "learning_rate": 3.6524091225048573e-07,
+      "loss": 0.4426,
+      "step": 11221
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6741091252571123,
+      "learning_rate": 3.647639089326072e-07,
+      "loss": 0.4142,
+      "step": 11222
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9150597790588353,
+      "learning_rate": 3.6428720550844123e-07,
+      "loss": 0.4377,
+      "step": 11223
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5591527812479657,
+      "learning_rate": 3.638108020088271e-07,
+      "loss": 0.4739,
+      "step": 11224
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5785559033991536,
+      "learning_rate": 3.633346984645908e-07,
+      "loss": 0.4709,
+      "step": 11225
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9393217086635213,
+      "learning_rate": 3.6285889490653317e-07,
+      "loss": 0.5065,
+      "step": 11226
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.0750015961136645,
+      "learning_rate": 3.623833913654401e-07,
+      "loss": 0.3872,
+      "step": 11227
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.846531809293167,
+      "learning_rate": 3.619081878720726e-07,
+      "loss": 0.4624,
+      "step": 11228
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 3.189120182490041,
+      "learning_rate": 3.6143328445718e-07,
+      "loss": 0.4669,
+      "step": 11229
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8147823092363453,
+      "learning_rate": 3.609586811514848e-07,
+      "loss": 0.4665,
+      "step": 11230
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8559804851063775,
+      "learning_rate": 3.6048437798569426e-07,
+      "loss": 0.3933,
+      "step": 11231
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6895121138224627,
+      "learning_rate": 3.600103749904954e-07,
+      "loss": 0.403,
+      "step": 11232
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8465626173969103,
+      "learning_rate": 3.595366721965554e-07,
+      "loss": 0.467,
+      "step": 11233
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5647456585426883,
+      "learning_rate": 3.590632696345231e-07,
+      "loss": 0.4701,
+      "step": 11234
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.009013782082817,
+      "learning_rate": 3.5859016733502504e-07,
+      "loss": 0.4634,
+      "step": 11235
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.586805668342278,
+      "learning_rate": 3.581173653286718e-07,
+      "loss": 0.4412,
+      "step": 11236
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.030110058531091,
+      "learning_rate": 3.5764486364605223e-07,
+      "loss": 0.4486,
+      "step": 11237
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.929175117218611,
+      "learning_rate": 3.571726623177385e-07,
+      "loss": 0.4102,
+      "step": 11238
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7564217298755755,
+      "learning_rate": 3.5670076137427847e-07,
+      "loss": 0.434,
+      "step": 11239
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8330279457739223,
+      "learning_rate": 3.5622916084620543e-07,
+      "loss": 0.4643,
+      "step": 11240
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.1385356145707313,
+      "learning_rate": 3.557578607640305e-07,
+      "loss": 0.4109,
+      "step": 11241
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6829586330125625,
+      "learning_rate": 3.552868611582466e-07,
+      "loss": 0.4019,
+      "step": 11242
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 6.521669728485595,
+      "learning_rate": 3.5481616205932767e-07,
+      "loss": 0.4217,
+      "step": 11243
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.627417484542981,
+      "learning_rate": 3.5434576349772485e-07,
+      "loss": 0.5044,
+      "step": 11244
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.8035499607896466,
+      "learning_rate": 3.5387566550387554e-07,
+      "loss": 0.4421,
+      "step": 11245
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.2312701092618523,
+      "learning_rate": 3.5340586810819144e-07,
+      "loss": 0.4549,
+      "step": 11246
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.8639923601783135,
+      "learning_rate": 3.5293637134106994e-07,
+      "loss": 0.4022,
+      "step": 11247
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3906026463714514,
+      "learning_rate": 3.524671752328857e-07,
+      "loss": 0.4425,
+      "step": 11248
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5507524293556348,
+      "learning_rate": 3.5199827981399594e-07,
+      "loss": 0.4714,
+      "step": 11249
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 4.4334582474343,
+      "learning_rate": 3.515296851147365e-07,
+      "loss": 0.4561,
+      "step": 11250
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4977533269852066,
+      "learning_rate": 3.5106139116542594e-07,
+      "loss": 0.4253,
+      "step": 11251
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7299628951329051,
+      "learning_rate": 3.5059339799636163e-07,
+      "loss": 0.4709,
+      "step": 11252
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6331679750197606,
+      "learning_rate": 3.501257056378221e-07,
+      "loss": 0.4863,
+      "step": 11253
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.233951295396871,
+      "learning_rate": 3.496583141200671e-07,
+      "loss": 0.4302,
+      "step": 11254
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5163712902913478,
+      "learning_rate": 3.4919122347333525e-07,
+      "loss": 0.4117,
+      "step": 11255
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.801016535145051,
+      "learning_rate": 3.4872443372784726e-07,
+      "loss": 0.4446,
+      "step": 11256
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9450585436580055,
+      "learning_rate": 3.4825794491380405e-07,
+      "loss": 0.4472,
+      "step": 11257
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.6690305201998483,
+      "learning_rate": 3.47791757061387e-07,
+      "loss": 0.4567,
+      "step": 11258
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.596938399897498,
+      "learning_rate": 3.4732587020075594e-07,
+      "loss": 0.4042,
+      "step": 11259
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.633037407709579,
+      "learning_rate": 3.468602843620561e-07,
+      "loss": 0.4794,
+      "step": 11260
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.7956112080581035,
+      "learning_rate": 3.4639499957540843e-07,
+      "loss": 0.4679,
+      "step": 11261
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.570303892062921,
+      "learning_rate": 3.4593001587091666e-07,
+      "loss": 0.4388,
+      "step": 11262
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5750040759752023,
+      "learning_rate": 3.454653332786634e-07,
+      "loss": 0.4611,
+      "step": 11263
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.9763764169020825,
+      "learning_rate": 3.4500095182871505e-07,
+      "loss": 0.4749,
+      "step": 11264
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5271912631586293,
+      "learning_rate": 3.445368715511166e-07,
+      "loss": 0.4241,
+      "step": 11265
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3161719424963043,
+      "learning_rate": 3.4407309247589114e-07,
+      "loss": 0.4764,
+      "step": 11266
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.3171954992292942,
+      "learning_rate": 3.4360961463304587e-07,
+      "loss": 0.4532,
+      "step": 11267
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.732159638137134,
+      "learning_rate": 3.4314643805256787e-07,
+      "loss": 0.4134,
+      "step": 11268
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8425792793175741,
+      "learning_rate": 3.4268356276442373e-07,
+      "loss": 0.4409,
+      "step": 11269
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.3244454559928642,
+      "learning_rate": 3.4222098879855946e-07,
+      "loss": 0.4513,
+      "step": 11270
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.84844128783743,
+      "learning_rate": 3.4175871618490455e-07,
+      "loss": 0.4285,
+      "step": 11271
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8364989667274885,
+      "learning_rate": 3.412967449533666e-07,
+      "loss": 0.4,
+      "step": 11272
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.720826969403055,
+      "learning_rate": 3.408350751338363e-07,
+      "loss": 0.4577,
+      "step": 11273
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4952967300242275,
+      "learning_rate": 3.403737067561802e-07,
+      "loss": 0.4244,
+      "step": 11274
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6792486451478024,
+      "learning_rate": 3.3991263985025e-07,
+      "loss": 0.4579,
+      "step": 11275
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.405480274448575,
+      "learning_rate": 3.394518744458758e-07,
+      "loss": 0.4045,
+      "step": 11276
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5755293694158577,
+      "learning_rate": 3.389914105728687e-07,
+      "loss": 0.4711,
+      "step": 11277
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.797505304131525,
+      "learning_rate": 3.385312482610209e-07,
+      "loss": 0.5049,
+      "step": 11278
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3801923791706727,
+      "learning_rate": 3.380713875401015e-07,
+      "loss": 0.4341,
+      "step": 11279
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7562848146804624,
+      "learning_rate": 3.3761182843986604e-07,
+      "loss": 0.4261,
+      "step": 11280
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2520444268079496,
+      "learning_rate": 3.3715257099004586e-07,
+      "loss": 0.4863,
+      "step": 11281
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6976790780636088,
+      "learning_rate": 3.3669361522035426e-07,
+      "loss": 0.4118,
+      "step": 11282
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4295970667448212,
+      "learning_rate": 3.3623496116048594e-07,
+      "loss": 0.4291,
+      "step": 11283
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5972504000799805,
+      "learning_rate": 3.357766088401149e-07,
+      "loss": 0.4549,
+      "step": 11284
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1297620285585066,
+      "learning_rate": 3.3531855828889517e-07,
+      "loss": 0.4567,
+      "step": 11285
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0201763946640727,
+      "learning_rate": 3.3486080953646196e-07,
+      "loss": 0.4452,
+      "step": 11286
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4356084084620493,
+      "learning_rate": 3.3440336261243213e-07,
+      "loss": 0.4678,
+      "step": 11287
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.957601814454286,
+      "learning_rate": 3.3394621754640146e-07,
+      "loss": 0.4485,
+      "step": 11288
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.204641086718789,
+      "learning_rate": 3.3348937436794683e-07,
+      "loss": 0.4006,
+      "step": 11289
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.063164485718259,
+      "learning_rate": 3.3303283310662516e-07,
+      "loss": 0.4201,
+      "step": 11290
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.918846002747808,
+      "learning_rate": 3.325765937919734e-07,
+      "loss": 0.4485,
+      "step": 11291
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.001033479431572,
+      "learning_rate": 3.321206564535101e-07,
+      "loss": 0.4627,
+      "step": 11292
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5881907082489715,
+      "learning_rate": 3.3166502112073507e-07,
+      "loss": 0.4645,
+      "step": 11293
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.463542343862739,
+      "learning_rate": 3.312096878231247e-07,
+      "loss": 0.4453,
+      "step": 11294
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5629842868900534,
+      "learning_rate": 3.307546565901415e-07,
+      "loss": 0.4811,
+      "step": 11295
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5926697933398304,
+      "learning_rate": 3.3029992745122307e-07,
+      "loss": 0.4473,
+      "step": 11296
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5122074229037237,
+      "learning_rate": 3.2984550043579145e-07,
+      "loss": 0.4471,
+      "step": 11297
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7315963044808953,
+      "learning_rate": 3.293913755732453e-07,
+      "loss": 0.4139,
+      "step": 11298
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6038755203354975,
+      "learning_rate": 3.2893755289296727e-07,
+      "loss": 0.4409,
+      "step": 11299
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6167336332645473,
+      "learning_rate": 3.2848403242432047e-07,
+      "loss": 0.4734,
+      "step": 11300
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.255135510269154,
+      "learning_rate": 3.2803081419664483e-07,
+      "loss": 0.4282,
+      "step": 11301
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5328681183211792,
+      "learning_rate": 3.2757789823926354e-07,
+      "loss": 0.4575,
+      "step": 11302
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9695160746563396,
+      "learning_rate": 3.2712528458147984e-07,
+      "loss": 0.4667,
+      "step": 11303
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5743806879613491,
+      "learning_rate": 3.2667297325257864e-07,
+      "loss": 0.4437,
+      "step": 11304
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.2580586402700162,
+      "learning_rate": 3.262209642818215e-07,
+      "loss": 0.4439,
+      "step": 11305
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.411325193736013,
+      "learning_rate": 3.2576925769845393e-07,
+      "loss": 0.4498,
+      "step": 11306
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5480874047937919,
+      "learning_rate": 3.253178535317003e-07,
+      "loss": 0.4622,
+      "step": 11307
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.868174691028942,
+      "learning_rate": 3.2486675181076786e-07,
+      "loss": 0.4257,
+      "step": 11308
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8368645604276665,
+      "learning_rate": 3.244159525648394e-07,
+      "loss": 0.4395,
+      "step": 11309
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.212415453784581,
+      "learning_rate": 3.23965455823082e-07,
+      "loss": 0.438,
+      "step": 11310
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.5916941555780544,
+      "learning_rate": 3.235152616146436e-07,
+      "loss": 0.4045,
+      "step": 11311
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6445125963401992,
+      "learning_rate": 3.230653699686498e-07,
+      "loss": 0.4447,
+      "step": 11312
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.662999763081008,
+      "learning_rate": 3.2261578091420886e-07,
+      "loss": 0.457,
+      "step": 11313
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.139957599849015,
+      "learning_rate": 3.2216649448040704e-07,
+      "loss": 0.4121,
+      "step": 11314
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.384133598054696,
+      "learning_rate": 3.2171751069631494e-07,
+      "loss": 0.441,
+      "step": 11315
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7865731249453882,
+      "learning_rate": 3.212688295909788e-07,
+      "loss": 0.4112,
+      "step": 11316
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8810960806389665,
+      "learning_rate": 3.208204511934293e-07,
+      "loss": 0.4665,
+      "step": 11317
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5744258616121598,
+      "learning_rate": 3.203723755326754e-07,
+      "loss": 0.4612,
+      "step": 11318
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5348738960636521,
+      "learning_rate": 3.1992460263770785e-07,
+      "loss": 0.4721,
+      "step": 11319
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7017684044963908,
+      "learning_rate": 3.194771325374951e-07,
+      "loss": 0.5283,
+      "step": 11320
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.2004786195128743,
+      "learning_rate": 3.1902996526098906e-07,
+      "loss": 0.4361,
+      "step": 11321
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0301826418746165,
+      "learning_rate": 3.18583100837121e-07,
+      "loss": 0.4674,
+      "step": 11322
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8679368170386168,
+      "learning_rate": 3.1813653929480215e-07,
+      "loss": 0.4034,
+      "step": 11323
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7842176739470226,
+      "learning_rate": 3.176902806629256e-07,
+      "loss": 0.4662,
+      "step": 11324
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8176489055260676,
+      "learning_rate": 3.172443249703616e-07,
+      "loss": 0.4688,
+      "step": 11325
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6398410562052315,
+      "learning_rate": 3.1679867224596426e-07,
+      "loss": 0.4201,
+      "step": 11326
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7633763032238536,
+      "learning_rate": 3.163533225185661e-07,
+      "loss": 0.4405,
+      "step": 11327
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1126162702991818,
+      "learning_rate": 3.159082758169818e-07,
+      "loss": 0.505,
+      "step": 11328
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.093901737775044,
+      "learning_rate": 3.1546353217000337e-07,
+      "loss": 0.4494,
+      "step": 11329
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4453150266778187,
+      "learning_rate": 3.150190916064072e-07,
+      "loss": 0.4399,
+      "step": 11330
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.901952173445501,
+      "learning_rate": 3.1457495415494643e-07,
+      "loss": 0.4039,
+      "step": 11331
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.275905669997325,
+      "learning_rate": 3.1413111984435696e-07,
+      "loss": 0.4455,
+      "step": 11332
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1209537237929874,
+      "learning_rate": 3.136875887033541e-07,
+      "loss": 0.4637,
+      "step": 11333
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5452202945300068,
+      "learning_rate": 3.132443607606339e-07,
+      "loss": 0.4814,
+      "step": 11334
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0593482072253075,
+      "learning_rate": 3.1280143604487333e-07,
+      "loss": 0.4407,
+      "step": 11335
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5469790320314923,
+      "learning_rate": 3.123588145847273e-07,
+      "loss": 0.4508,
+      "step": 11336
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.521174002232033,
+      "learning_rate": 3.1191649640883395e-07,
+      "loss": 0.4763,
+      "step": 11337
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.686434817031486,
+      "learning_rate": 3.114744815458104e-07,
+      "loss": 0.4415,
+      "step": 11338
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.547245960683708,
+      "learning_rate": 3.1103277002425494e-07,
+      "loss": 0.4387,
+      "step": 11339
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.005989508669002,
+      "learning_rate": 3.105913618727452e-07,
+      "loss": 0.4284,
+      "step": 11340
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5506089563682616,
+      "learning_rate": 3.101502571198395e-07,
+      "loss": 0.4575,
+      "step": 11341
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.850823503800598,
+      "learning_rate": 3.097094557940772e-07,
+      "loss": 0.3911,
+      "step": 11342
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.611517155265725,
+      "learning_rate": 3.0926895792397773e-07,
+      "loss": 0.4857,
+      "step": 11343
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7819940264159992,
+      "learning_rate": 3.0882876353803994e-07,
+      "loss": 0.4477,
+      "step": 11344
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 7.459706024797533,
+      "learning_rate": 3.0838887266474384e-07,
+      "loss": 0.4564,
+      "step": 11345
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9857827275160052,
+      "learning_rate": 3.0794928533255165e-07,
+      "loss": 0.4715,
+      "step": 11346
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.852616245136004,
+      "learning_rate": 3.0751000156990175e-07,
+      "loss": 0.4654,
+      "step": 11347
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7587980624301263,
+      "learning_rate": 3.070710214052169e-07,
+      "loss": 0.434,
+      "step": 11348
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5713148745825898,
+      "learning_rate": 3.0663234486689674e-07,
+      "loss": 0.4282,
+      "step": 11349
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5889383695196914,
+      "learning_rate": 3.0619397198332565e-07,
+      "loss": 0.4835,
+      "step": 11350
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.380683309233936,
+      "learning_rate": 3.057559027828633e-07,
+      "loss": 0.4882,
+      "step": 11351
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4736501459113696,
+      "learning_rate": 3.0531813729385294e-07,
+      "loss": 0.4491,
+      "step": 11352
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5551133746410761,
+      "learning_rate": 3.048806755446182e-07,
+      "loss": 0.4645,
+      "step": 11353
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8032008231829162,
+      "learning_rate": 3.0444351756346245e-07,
+      "loss": 0.4882,
+      "step": 11354
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.6726290999534124,
+      "learning_rate": 3.0400666337866756e-07,
+      "loss": 0.4682,
+      "step": 11355
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.491519928158693,
+      "learning_rate": 3.035701130184987e-07,
+      "loss": 0.4576,
+      "step": 11356
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.344926261831781,
+      "learning_rate": 3.0313386651119935e-07,
+      "loss": 0.4824,
+      "step": 11357
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5741899869775209,
+      "learning_rate": 3.026979238849953e-07,
+      "loss": 0.4982,
+      "step": 11358
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9130461305494726,
+      "learning_rate": 3.0226228516809066e-07,
+      "loss": 0.4759,
+      "step": 11359
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3645009846826872,
+      "learning_rate": 3.018269503886706e-07,
+      "loss": 0.3737,
+      "step": 11360
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8493573749084138,
+      "learning_rate": 3.013919195749004e-07,
+      "loss": 0.4389,
+      "step": 11361
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.4710799692358103,
+      "learning_rate": 3.0095719275492706e-07,
+      "loss": 0.5134,
+      "step": 11362
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.99393386916813,
+      "learning_rate": 3.0052276995687637e-07,
+      "loss": 0.4512,
+      "step": 11363
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9106719714730473,
+      "learning_rate": 3.0008865120885356e-07,
+      "loss": 0.415,
+      "step": 11364
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.705639712369887,
+      "learning_rate": 2.9965483653894846e-07,
+      "loss": 0.4809,
+      "step": 11365
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.448323036169266,
+      "learning_rate": 2.992213259752258e-07,
+      "loss": 0.4438,
+      "step": 11366
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.8341714831777236,
+      "learning_rate": 2.9878811954573374e-07,
+      "loss": 0.4113,
+      "step": 11367
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.563245717435171,
+      "learning_rate": 2.9835521727850035e-07,
+      "loss": 0.4082,
+      "step": 11368
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.8357519257263144,
+      "learning_rate": 2.979226192015339e-07,
+      "loss": 0.4238,
+      "step": 11369
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5867637843906715,
+      "learning_rate": 2.9749032534282405e-07,
+      "loss": 0.4019,
+      "step": 11370
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.320149599026252,
+      "learning_rate": 2.9705833573033747e-07,
+      "loss": 0.4417,
+      "step": 11371
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5305416131784837,
+      "learning_rate": 2.96626650392024e-07,
+      "loss": 0.4602,
+      "step": 11372
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5310197285914908,
+      "learning_rate": 2.96195269355814e-07,
+      "loss": 0.4519,
+      "step": 11373
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.206453211844582,
+      "learning_rate": 2.9576419264961684e-07,
+      "loss": 0.4195,
+      "step": 11374
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.627075121936036,
+      "learning_rate": 2.953334203013225e-07,
+      "loss": 0.4471,
+      "step": 11375
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0570038282300405,
+      "learning_rate": 2.9490295233880083e-07,
+      "loss": 0.4401,
+      "step": 11376
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9653415679385258,
+      "learning_rate": 2.944727887899035e-07,
+      "loss": 0.4807,
+      "step": 11377
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5174323364634673,
+      "learning_rate": 2.9404292968246095e-07,
+      "loss": 0.47,
+      "step": 11378
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.847390566248935,
+      "learning_rate": 2.9361337504428543e-07,
+      "loss": 0.5137,
+      "step": 11379
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6870357897874173,
+      "learning_rate": 2.9318412490316636e-07,
+      "loss": 0.4172,
+      "step": 11380
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6031952371289342,
+      "learning_rate": 2.927551792868782e-07,
+      "loss": 0.4591,
+      "step": 11381
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5360352078561933,
+      "learning_rate": 2.923265382231721e-07,
+      "loss": 0.4664,
+      "step": 11382
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.888519235564012,
+      "learning_rate": 2.918982017397809e-07,
+      "loss": 0.3777,
+      "step": 11383
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.22511278269066,
+      "learning_rate": 2.914701698644157e-07,
+      "loss": 0.4627,
+      "step": 11384
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.7058166092141644,
+      "learning_rate": 2.910424426247721e-07,
+      "loss": 0.4003,
+      "step": 11385
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.1476447299455805,
+      "learning_rate": 2.9061502004852194e-07,
+      "loss": 0.4712,
+      "step": 11386
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.548036668656327,
+      "learning_rate": 2.9018790216331917e-07,
+      "loss": 0.4813,
+      "step": 11387
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 2.0088872058147182,
+      "learning_rate": 2.897610889967983e-07,
+      "loss": 0.4794,
+      "step": 11388
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.820277018405758,
+      "learning_rate": 2.8933458057657295e-07,
+      "loss": 0.3888,
+      "step": 11389
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7648849737627514,
+      "learning_rate": 2.8890837693023863e-07,
+      "loss": 0.4696,
+      "step": 11390
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6697290973578673,
+      "learning_rate": 2.8848247808536847e-07,
+      "loss": 0.4708,
+      "step": 11391
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.7322855933915686,
+      "learning_rate": 2.8805688406951914e-07,
+      "loss": 0.4902,
+      "step": 11392
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6661093342394897,
+      "learning_rate": 2.8763159491022484e-07,
+      "loss": 0.4569,
+      "step": 11393
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.550928315296681,
+      "learning_rate": 2.87206610635003e-07,
+      "loss": 0.4722,
+      "step": 11394
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6933594117270645,
+      "learning_rate": 2.867819312713471e-07,
+      "loss": 0.4697,
+      "step": 11395
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.683783729663718,
+      "learning_rate": 2.863575568467353e-07,
+      "loss": 0.454,
+      "step": 11396
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.631256911272982,
+      "learning_rate": 2.859334873886227e-07,
+      "loss": 0.4355,
+      "step": 11397
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.62446836542238,
+      "learning_rate": 2.8550972292444744e-07,
+      "loss": 0.4442,
+      "step": 11398
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7297073422909943,
+      "learning_rate": 2.850862634816248e-07,
+      "loss": 0.3916,
+      "step": 11399
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6498571926319299,
+      "learning_rate": 2.8466310908755347e-07,
+      "loss": 0.4212,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2062971890786613,
+      "learning_rate": 2.842402597696109e-07,
+      "loss": 0.453,
+      "step": 11401
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9833679696466364,
+      "learning_rate": 2.8381771555515416e-07,
+      "loss": 0.5178,
+      "step": 11402
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.458659727497805,
+      "learning_rate": 2.8339547647152186e-07,
+      "loss": 0.4372,
+      "step": 11403
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6669920176993276,
+      "learning_rate": 2.8297354254603226e-07,
+      "loss": 0.3598,
+      "step": 11404
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6152094735551394,
+      "learning_rate": 2.82551913805984e-07,
+      "loss": 0.4501,
+      "step": 11405
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8700175843001834,
+      "learning_rate": 2.821305902786553e-07,
+      "loss": 0.4593,
+      "step": 11406
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1110618775428094,
+      "learning_rate": 2.8170957199130545e-07,
+      "loss": 0.4096,
+      "step": 11407
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.576036545255131,
+      "learning_rate": 2.812888589711743e-07,
+      "loss": 0.4446,
+      "step": 11408
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4702657997579975,
+      "learning_rate": 2.808684512454818e-07,
+      "loss": 0.424,
+      "step": 11409
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.212978056980427,
+      "learning_rate": 2.804483488414261e-07,
+      "loss": 0.4408,
+      "step": 11410
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5835566426546058,
+      "learning_rate": 2.800285517861884e-07,
+      "loss": 0.4556,
+      "step": 11411
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.628544965715317,
+      "learning_rate": 2.79609060106929e-07,
+      "loss": 0.4361,
+      "step": 11412
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8084291332964564,
+      "learning_rate": 2.7918987383078857e-07,
+      "loss": 0.4823,
+      "step": 11413
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5310464514391604,
+      "learning_rate": 2.787709929848881e-07,
+      "loss": 0.4619,
+      "step": 11414
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.3998772101075545,
+      "learning_rate": 2.78352417596327e-07,
+      "loss": 0.4576,
+      "step": 11415
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1718037059775135,
+      "learning_rate": 2.779341476921887e-07,
+      "loss": 0.4376,
+      "step": 11416
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5205833530637872,
+      "learning_rate": 2.7751618329953376e-07,
+      "loss": 0.4961,
+      "step": 11417
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.303675392030558,
+      "learning_rate": 2.770985244454044e-07,
+      "loss": 0.4473,
+      "step": 11418
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.542182332825767,
+      "learning_rate": 2.766811711568207e-07,
+      "loss": 0.4652,
+      "step": 11419
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.325617690032522,
+      "learning_rate": 2.7626412346078824e-07,
+      "loss": 0.424,
+      "step": 11420
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.70053174705483,
+      "learning_rate": 2.7584738138428604e-07,
+      "loss": 0.4394,
+      "step": 11421
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0024030876864947,
+      "learning_rate": 2.7543094495427913e-07,
+      "loss": 0.4314,
+      "step": 11422
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6713767100690384,
+      "learning_rate": 2.7501481419770884e-07,
+      "loss": 0.4287,
+      "step": 11423
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.766189884612762,
+      "learning_rate": 2.7459898914149963e-07,
+      "loss": 0.4385,
+      "step": 11424
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7557920078286193,
+      "learning_rate": 2.741834698125545e-07,
+      "loss": 0.4423,
+      "step": 11425
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5439389285858546,
+      "learning_rate": 2.737682562377564e-07,
+      "loss": 0.4793,
+      "step": 11426
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5388612115035586,
+      "learning_rate": 2.733533484439688e-07,
+      "loss": 0.462,
+      "step": 11427
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5729533662673044,
+      "learning_rate": 2.729387464580369e-07,
+      "loss": 0.475,
+      "step": 11428
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6717167305903013,
+      "learning_rate": 2.7252445030678475e-07,
+      "loss": 0.4029,
+      "step": 11429
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.357655547104963,
+      "learning_rate": 2.7211046001701604e-07,
+      "loss": 0.4295,
+      "step": 11430
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6012259188828346,
+      "learning_rate": 2.716967756155153e-07,
+      "loss": 0.457,
+      "step": 11431
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6790183752889436,
+      "learning_rate": 2.712833971290479e-07,
+      "loss": 0.4483,
+      "step": 11432
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9276594267081877,
+      "learning_rate": 2.7087032458435914e-07,
+      "loss": 0.4404,
+      "step": 11433
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6764322425125178,
+      "learning_rate": 2.7045755800817364e-07,
+      "loss": 0.4475,
+      "step": 11434
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2012675353302806,
+      "learning_rate": 2.700450974271962e-07,
+      "loss": 0.4434,
+      "step": 11435
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2039079946270856,
+      "learning_rate": 2.696329428681149e-07,
+      "loss": 0.44,
+      "step": 11436
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5352792845476458,
+      "learning_rate": 2.6922109435759346e-07,
+      "loss": 0.4878,
+      "step": 11437
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.687443549535484,
+      "learning_rate": 2.6880955192227885e-07,
+      "loss": 0.4485,
+      "step": 11438
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4417073023430471,
+      "learning_rate": 2.683983155887965e-07,
+      "loss": 0.4543,
+      "step": 11439
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4906272079818452,
+      "learning_rate": 2.6798738538375447e-07,
+      "loss": 0.437,
+      "step": 11440
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.243457288432882,
+      "learning_rate": 2.675767613337371e-07,
+      "loss": 0.4233,
+      "step": 11441
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.604024146459571,
+      "learning_rate": 2.671664434653132e-07,
+      "loss": 0.4342,
+      "step": 11442
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4625304856640504,
+      "learning_rate": 2.6675643180502865e-07,
+      "loss": 0.4449,
+      "step": 11443
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.662440333718254,
+      "learning_rate": 2.663467263794123e-07,
+      "loss": 0.474,
+      "step": 11444
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8502671932455674,
+      "learning_rate": 2.65937327214969e-07,
+      "loss": 0.3918,
+      "step": 11445
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7680351458876735,
+      "learning_rate": 2.655282343381882e-07,
+      "loss": 0.4038,
+      "step": 11446
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1142607570033145,
+      "learning_rate": 2.65119447775537e-07,
+      "loss": 0.4809,
+      "step": 11447
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6581699666275387,
+      "learning_rate": 2.6471096755346314e-07,
+      "loss": 0.4627,
+      "step": 11448
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.950991543973786,
+      "learning_rate": 2.6430279369839664e-07,
+      "loss": 0.4373,
+      "step": 11449
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5742220322790039,
+      "learning_rate": 2.638949262367424e-07,
+      "loss": 0.474,
+      "step": 11450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1846163652528445,
+      "learning_rate": 2.634873651948922e-07,
+      "loss": 0.4476,
+      "step": 11451
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.959303904278573,
+      "learning_rate": 2.6308011059921257e-07,
+      "loss": 0.4306,
+      "step": 11452
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.045963755850218,
+      "learning_rate": 2.626731624760542e-07,
+      "loss": 0.4186,
+      "step": 11453
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6883988659983487,
+      "learning_rate": 2.622665208517433e-07,
+      "loss": 0.4469,
+      "step": 11454
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1911554097917083,
+      "learning_rate": 2.618601857525921e-07,
+      "loss": 0.4113,
+      "step": 11455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7119376681944634,
+      "learning_rate": 2.614541572048884e-07,
+      "loss": 0.4641,
+      "step": 11456
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7382387058097901,
+      "learning_rate": 2.610484352349013e-07,
+      "loss": 0.4578,
+      "step": 11457
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.842338592108334,
+      "learning_rate": 2.606430198688814e-07,
+      "loss": 0.4156,
+      "step": 11458
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6047270523266723,
+      "learning_rate": 2.602379111330583e-07,
+      "loss": 0.4578,
+      "step": 11459
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5707814143339303,
+      "learning_rate": 2.5983310905364267e-07,
+      "loss": 0.4273,
+      "step": 11460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.014504903027408,
+      "learning_rate": 2.594286136568236e-07,
+      "loss": 0.3978,
+      "step": 11461
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1328381876816347,
+      "learning_rate": 2.5902442496877123e-07,
+      "loss": 0.4932,
+      "step": 11462
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.433364925832676,
+      "learning_rate": 2.586205430156369e-07,
+      "loss": 0.43,
+      "step": 11463
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.1491250236619726,
+      "learning_rate": 2.5821696782355134e-07,
+      "loss": 0.4287,
+      "step": 11464
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9174821088621523,
+      "learning_rate": 2.578136994186242e-07,
+      "loss": 0.4623,
+      "step": 11465
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9713032113604674,
+      "learning_rate": 2.5741073782694747e-07,
+      "loss": 0.5045,
+      "step": 11466
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5722450411662502,
+      "learning_rate": 2.570080830745919e-07,
+      "loss": 0.4682,
+      "step": 11467
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 5.043043298819018,
+      "learning_rate": 2.5660573518760833e-07,
+      "loss": 0.4239,
+      "step": 11468
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.4525646592560792,
+      "learning_rate": 2.562036941920298e-07,
+      "loss": 0.461,
+      "step": 11469
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5506322590572963,
+      "learning_rate": 2.5580196011386507e-07,
+      "loss": 0.4796,
+      "step": 11470
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5296074489700937,
+      "learning_rate": 2.554005329791087e-07,
+      "loss": 0.4776,
+      "step": 11471
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.62477518037893,
+      "learning_rate": 2.549994128137301e-07,
+      "loss": 0.4655,
+      "step": 11472
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5882824398765624,
+      "learning_rate": 2.5459859964368227e-07,
+      "loss": 0.4723,
+      "step": 11473
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2125354963534436,
+      "learning_rate": 2.541980934948979e-07,
+      "loss": 0.4498,
+      "step": 11474
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5654964210877123,
+      "learning_rate": 2.5379789439328893e-07,
+      "loss": 0.4927,
+      "step": 11475
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.70510662301389,
+      "learning_rate": 2.5339800236474645e-07,
+      "loss": 0.4355,
+      "step": 11476
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.6139649381691386,
+      "learning_rate": 2.529984174351441e-07,
+      "loss": 0.4594,
+      "step": 11477
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7511701146777976,
+      "learning_rate": 2.52599139630334e-07,
+      "loss": 0.4863,
+      "step": 11478
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7766756188876547,
+      "learning_rate": 2.5220016897614943e-07,
+      "loss": 0.468,
+      "step": 11479
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4947317481855442,
+      "learning_rate": 2.518015054984041e-07,
+      "loss": 0.4571,
+      "step": 11480
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.987195885614004,
+      "learning_rate": 2.5140314922288854e-07,
+      "loss": 0.4745,
+      "step": 11481
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9337409032654684,
+      "learning_rate": 2.5100510017537773e-07,
+      "loss": 0.4362,
+      "step": 11482
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8481453198409,
+      "learning_rate": 2.5060735838162433e-07,
+      "loss": 0.4275,
+      "step": 11483
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9586256030002387,
+      "learning_rate": 2.5020992386736274e-07,
+      "loss": 0.4326,
+      "step": 11484
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.300869195613282,
+      "learning_rate": 2.498127966583042e-07,
+      "loss": 0.4227,
+      "step": 11485
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6074281092426282,
+      "learning_rate": 2.4941597678014464e-07,
+      "loss": 0.4693,
+      "step": 11486
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4496022570779206,
+      "learning_rate": 2.490194642585564e-07,
+      "loss": 0.4339,
+      "step": 11487
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.0571224216844652,
+      "learning_rate": 2.486232591191945e-07,
+      "loss": 0.4335,
+      "step": 11488
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.359389310872638,
+      "learning_rate": 2.4822736138769064e-07,
+      "loss": 0.4979,
+      "step": 11489
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8411424742529219,
+      "learning_rate": 2.4783177108966107e-07,
+      "loss": 0.4602,
+      "step": 11490
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5397894798745554,
+      "learning_rate": 2.474364882507002e-07,
+      "loss": 0.4778,
+      "step": 11491
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7948989805836306,
+      "learning_rate": 2.4704151289638045e-07,
+      "loss": 0.4143,
+      "step": 11492
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9146951042539309,
+      "learning_rate": 2.466468450522574e-07,
+      "loss": 0.4293,
+      "step": 11493
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5695218852774204,
+      "learning_rate": 2.462524847438652e-07,
+      "loss": 0.3819,
+      "step": 11494
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.000545115941838,
+      "learning_rate": 2.458584319967194e-07,
+      "loss": 0.4175,
+      "step": 11495
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.574829380102967,
+      "learning_rate": 2.454646868363131e-07,
+      "loss": 0.4241,
+      "step": 11496
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6918417886550867,
+      "learning_rate": 2.4507124928812184e-07,
+      "loss": 0.4355,
+      "step": 11497
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9282118276250628,
+      "learning_rate": 2.4467811937760046e-07,
+      "loss": 0.3848,
+      "step": 11498
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 4.751103273554088,
+      "learning_rate": 2.442852971301846e-07,
+      "loss": 0.4671,
+      "step": 11499
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 16.979599586875334,
+      "learning_rate": 2.438927825712878e-07,
+      "loss": 0.4465,
+      "step": 11500
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6169477325319421,
+      "learning_rate": 2.4350057572630703e-07,
+      "loss": 0.4561,
+      "step": 11501
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5866350311536457,
+      "learning_rate": 2.431086766206159e-07,
+      "loss": 0.457,
+      "step": 11502
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.527767351703748,
+      "learning_rate": 2.427170852795713e-07,
+      "loss": 0.3784,
+      "step": 11503
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.884216163301964,
+      "learning_rate": 2.4232580172850793e-07,
+      "loss": 0.427,
+      "step": 11504
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5579895948039629,
+      "learning_rate": 2.4193482599274064e-07,
+      "loss": 0.4745,
+      "step": 11505
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.8752488804287286,
+      "learning_rate": 2.4154415809756695e-07,
+      "loss": 0.3918,
+      "step": 11506
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5353563110752431,
+      "learning_rate": 2.4115379806826045e-07,
+      "loss": 0.4613,
+      "step": 11507
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.7773073809189721,
+      "learning_rate": 2.4076374593007825e-07,
+      "loss": 0.4457,
+      "step": 11508
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6634825126636108,
+      "learning_rate": 2.4037400170825563e-07,
+      "loss": 0.4331,
+      "step": 11509
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.2898802316211877,
+      "learning_rate": 2.399845654280092e-07,
+      "loss": 0.4896,
+      "step": 11510
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9025798992848089,
+      "learning_rate": 2.395954371145343e-07,
+      "loss": 0.4425,
+      "step": 11511
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5200549513959155,
+      "learning_rate": 2.3920661679300685e-07,
+      "loss": 0.4664,
+      "step": 11512
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5994037983035863,
+      "learning_rate": 2.3881810448858345e-07,
+      "loss": 0.4922,
+      "step": 11513
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.627258752537254,
+      "learning_rate": 2.384299002264001e-07,
+      "loss": 0.4551,
+      "step": 11514
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 3.4129050262860194,
+      "learning_rate": 2.3804200403157453e-07,
+      "loss": 0.4703,
+      "step": 11515
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.5770364455989156,
+      "learning_rate": 2.37654415929201e-07,
+      "loss": 0.4302,
+      "step": 11516
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6457749771755348,
+      "learning_rate": 2.3726713594435623e-07,
+      "loss": 0.4548,
+      "step": 11517
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9965769024376148,
+      "learning_rate": 2.368801641020979e-07,
+      "loss": 0.4346,
+      "step": 11518
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.170950569971861,
+      "learning_rate": 2.3649350042746265e-07,
+      "loss": 0.4839,
+      "step": 11519
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.902604732530212,
+      "learning_rate": 2.3610714494546493e-07,
+      "loss": 0.4475,
+      "step": 11520
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5778077685182375,
+      "learning_rate": 2.3572109768110473e-07,
+      "loss": 0.4576,
+      "step": 11521
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5562866815272,
+      "learning_rate": 2.3533535865935597e-07,
+      "loss": 0.4886,
+      "step": 11522
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6254660906201128,
+      "learning_rate": 2.3494992790517757e-07,
+      "loss": 0.3944,
+      "step": 11523
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5323006559371795,
+      "learning_rate": 2.3456480544350345e-07,
+      "loss": 0.3926,
+      "step": 11524
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3175342823067306,
+      "learning_rate": 2.3417999129925374e-07,
+      "loss": 0.4315,
+      "step": 11525
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6252654989555713,
+      "learning_rate": 2.3379548549732456e-07,
+      "loss": 0.4237,
+      "step": 11526
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8537611675486931,
+      "learning_rate": 2.3341128806259162e-07,
+      "loss": 0.4407,
+      "step": 11527
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4698905978931054,
+      "learning_rate": 2.3302739901991277e-07,
+      "loss": 0.4562,
+      "step": 11528
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6425284597645118,
+      "learning_rate": 2.326438183941254e-07,
+      "loss": 0.4107,
+      "step": 11529
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5007639776151216,
+      "learning_rate": 2.3226054621004745e-07,
+      "loss": 0.464,
+      "step": 11530
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.056905218586717,
+      "learning_rate": 2.3187758249247406e-07,
+      "loss": 0.4718,
+      "step": 11531
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5152957575333061,
+      "learning_rate": 2.3149492726618373e-07,
+      "loss": 0.4407,
+      "step": 11532
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5526221538305488,
+      "learning_rate": 2.3111258055593387e-07,
+      "loss": 0.4593,
+      "step": 11533
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8977832804005321,
+      "learning_rate": 2.3073054238646197e-07,
+      "loss": 0.4654,
+      "step": 11534
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.3154200589164695,
+      "learning_rate": 2.3034881278248432e-07,
+      "loss": 0.4335,
+      "step": 11535
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7397881182992028,
+      "learning_rate": 2.2996739176869841e-07,
+      "loss": 0.405,
+      "step": 11536
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9948088429449033,
+      "learning_rate": 2.2958627936978394e-07,
+      "loss": 0.4391,
+      "step": 11537
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5491535799936376,
+      "learning_rate": 2.2920547561039563e-07,
+      "loss": 0.5129,
+      "step": 11538
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8423445602002524,
+      "learning_rate": 2.2882498051517266e-07,
+      "loss": 0.4742,
+      "step": 11539
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.456786406085758,
+      "learning_rate": 2.284447941087309e-07,
+      "loss": 0.4763,
+      "step": 11540
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5865854909279886,
+      "learning_rate": 2.2806491641567007e-07,
+      "loss": 0.4411,
+      "step": 11541
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.714004461077023,
+      "learning_rate": 2.2768534746056615e-07,
+      "loss": 0.4664,
+      "step": 11542
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8980090924975184,
+      "learning_rate": 2.273060872679772e-07,
+      "loss": 0.4344,
+      "step": 11543
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7582908799952768,
+      "learning_rate": 2.2692713586244086e-07,
+      "loss": 0.448,
+      "step": 11544
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7687432306654483,
+      "learning_rate": 2.2654849326847584e-07,
+      "loss": 0.4918,
+      "step": 11545
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.365381174851048,
+      "learning_rate": 2.2617015951057808e-07,
+      "loss": 0.4205,
+      "step": 11546
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1308940718252325,
+      "learning_rate": 2.257921346132258e-07,
+      "loss": 0.45,
+      "step": 11547
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1258958104859493,
+      "learning_rate": 2.254144186008772e-07,
+      "loss": 0.444,
+      "step": 11548
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0536017186212545,
+      "learning_rate": 2.2503701149796942e-07,
+      "loss": 0.4369,
+      "step": 11549
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.987336556551796,
+      "learning_rate": 2.246599133289218e-07,
+      "loss": 0.451,
+      "step": 11550
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.534520600854061,
+      "learning_rate": 2.242831241181298e-07,
+      "loss": 0.4563,
+      "step": 11551
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8006901115048144,
+      "learning_rate": 2.239066438899723e-07,
+      "loss": 0.4474,
+      "step": 11552
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5611517247848092,
+      "learning_rate": 2.23530472668807e-07,
+      "loss": 0.4742,
+      "step": 11553
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.43701236917575,
+      "learning_rate": 2.2315461047897225e-07,
+      "loss": 0.3991,
+      "step": 11554
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3178863987738128,
+      "learning_rate": 2.2277905734478466e-07,
+      "loss": 0.4664,
+      "step": 11555
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8549630883861659,
+      "learning_rate": 2.2240381329054317e-07,
+      "loss": 0.4583,
+      "step": 11556
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7553398528483806,
+      "learning_rate": 2.22028878340525e-07,
+      "loss": 0.4523,
+      "step": 11557
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9655068550644523,
+      "learning_rate": 2.2165425251898743e-07,
+      "loss": 0.4144,
+      "step": 11558
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.857684725555167,
+      "learning_rate": 2.212799358501694e-07,
+      "loss": 0.4714,
+      "step": 11559
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9652848317952962,
+      "learning_rate": 2.2090592835828817e-07,
+      "loss": 0.4937,
+      "step": 11560
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1636250670862816,
+      "learning_rate": 2.2053223006754166e-07,
+      "loss": 0.4448,
+      "step": 11561
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8285878564358173,
+      "learning_rate": 2.2015884100210717e-07,
+      "loss": 0.4099,
+      "step": 11562
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4770952271413116,
+      "learning_rate": 2.1978576118614315e-07,
+      "loss": 0.4794,
+      "step": 11563
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.642925565625324,
+      "learning_rate": 2.1941299064378697e-07,
+      "loss": 0.4624,
+      "step": 11564
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.1383780131702723,
+      "learning_rate": 2.1904052939915654e-07,
+      "loss": 0.4576,
+      "step": 11565
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3160913095182063,
+      "learning_rate": 2.1866837747634984e-07,
+      "loss": 0.527,
+      "step": 11566
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4668568066766676,
+      "learning_rate": 2.1829653489944368e-07,
+      "loss": 0.4025,
+      "step": 11567
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0666368858527995,
+      "learning_rate": 2.1792500169249665e-07,
+      "loss": 0.4298,
+      "step": 11568
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8285084925825668,
+      "learning_rate": 2.1755377787954724e-07,
+      "loss": 0.4079,
+      "step": 11569
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5435821511655918,
+      "learning_rate": 2.171828634846107e-07,
+      "loss": 0.4442,
+      "step": 11570
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.0623593439148813,
+      "learning_rate": 2.1681225853168619e-07,
+      "loss": 0.4361,
+      "step": 11571
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5450558251007471,
+      "learning_rate": 2.1644196304475285e-07,
+      "loss": 0.4657,
+      "step": 11572
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5219296057439682,
+      "learning_rate": 2.1607197704776593e-07,
+      "loss": 0.4352,
+      "step": 11573
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8415617436970722,
+      "learning_rate": 2.1570230056466412e-07,
+      "loss": 0.5017,
+      "step": 11574
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.515362573435689,
+      "learning_rate": 2.1533293361936381e-07,
+      "loss": 0.4751,
+      "step": 11575
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1913867957065527,
+      "learning_rate": 2.149638762357653e-07,
+      "loss": 0.4693,
+      "step": 11576
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8116784112370488,
+      "learning_rate": 2.1459512843774344e-07,
+      "loss": 0.4606,
+      "step": 11577
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.290734175068875,
+      "learning_rate": 2.1422669024915632e-07,
+      "loss": 0.431,
+      "step": 11578
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 4.157617377698731,
+      "learning_rate": 2.1385856169384212e-07,
+      "loss": 0.429,
+      "step": 11579
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5308058918915947,
+      "learning_rate": 2.1349074279561842e-07,
+      "loss": 0.456,
+      "step": 11580
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.46404386226592,
+      "learning_rate": 2.131232335782818e-07,
+      "loss": 0.4617,
+      "step": 11581
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6660260192758156,
+      "learning_rate": 2.1275603406560984e-07,
+      "loss": 0.4164,
+      "step": 11582
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5682696302155233,
+      "learning_rate": 2.123891442813597e-07,
+      "loss": 0.4748,
+      "step": 11583
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4616178722973794,
+      "learning_rate": 2.1202256424926904e-07,
+      "loss": 0.4912,
+      "step": 11584
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.263021123018609,
+      "learning_rate": 2.1165629399305553e-07,
+      "loss": 0.4166,
+      "step": 11585
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7393670491155402,
+      "learning_rate": 2.1129033353641527e-07,
+      "loss": 0.468,
+      "step": 11586
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6085495433722117,
+      "learning_rate": 2.1092468290302649e-07,
+      "loss": 0.4779,
+      "step": 11587
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6744697093702912,
+      "learning_rate": 2.1055934211654527e-07,
+      "loss": 0.4602,
+      "step": 11588
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.268146057217509,
+      "learning_rate": 2.1019431120060996e-07,
+      "loss": 0.456,
+      "step": 11589
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7486774258357767,
+      "learning_rate": 2.0982959017883554e-07,
+      "loss": 0.4097,
+      "step": 11590
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.320142242361531,
+      "learning_rate": 2.09465179074822e-07,
+      "loss": 0.4064,
+      "step": 11591
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.242958107989898,
+      "learning_rate": 2.0910107791214384e-07,
+      "loss": 0.4363,
+      "step": 11592
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.981069144351185,
+      "learning_rate": 2.0873728671435834e-07,
+      "loss": 0.4408,
+      "step": 11593
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.320356188281922,
+      "learning_rate": 2.0837380550500275e-07,
+      "loss": 0.435,
+      "step": 11594
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2015991794633907,
+      "learning_rate": 2.0801063430759384e-07,
+      "loss": 0.4901,
+      "step": 11595
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9990019182457437,
+      "learning_rate": 2.0764777314562834e-07,
+      "loss": 0.4306,
+      "step": 11596
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9163960061804437,
+      "learning_rate": 2.072852220425825e-07,
+      "loss": 0.4644,
+      "step": 11597
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.778514925806966,
+      "learning_rate": 2.069229810219131e-07,
+      "loss": 0.4941,
+      "step": 11598
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6211055040473192,
+      "learning_rate": 2.0656105010705696e-07,
+      "loss": 0.4179,
+      "step": 11599
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6831907000069037,
+      "learning_rate": 2.0619942932143034e-07,
+      "loss": 0.385,
+      "step": 11600
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8198307124341628,
+      "learning_rate": 2.0583811868842952e-07,
+      "loss": 0.4616,
+      "step": 11601
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5259222358063694,
+      "learning_rate": 2.0547711823143024e-07,
+      "loss": 0.4636,
+      "step": 11602
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4619442220726642,
+      "learning_rate": 2.0511642797378995e-07,
+      "loss": 0.458,
+      "step": 11603
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.066107231851946,
+      "learning_rate": 2.0475604793884385e-07,
+      "loss": 0.4493,
+      "step": 11604
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.846102312571656,
+      "learning_rate": 2.0439597814990943e-07,
+      "loss": 0.5673,
+      "step": 11605
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.682014747141216,
+      "learning_rate": 2.0403621863028022e-07,
+      "loss": 0.448,
+      "step": 11606
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7222181787300466,
+      "learning_rate": 2.0367676940323545e-07,
+      "loss": 0.405,
+      "step": 11607
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1382575977868075,
+      "learning_rate": 2.0331763049202868e-07,
+      "loss": 0.4307,
+      "step": 11608
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6968063808615348,
+      "learning_rate": 2.0295880191989637e-07,
+      "loss": 0.4625,
+      "step": 11609
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3394350511508932,
+      "learning_rate": 2.0260028371005324e-07,
+      "loss": 0.435,
+      "step": 11610
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.574160397364898,
+      "learning_rate": 2.0224207588569743e-07,
+      "loss": 0.4739,
+      "step": 11611
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4157438572398997,
+      "learning_rate": 2.0188417847000264e-07,
+      "loss": 0.4301,
+      "step": 11612
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.022310962266582,
+      "learning_rate": 2.015265914861242e-07,
+      "loss": 0.4337,
+      "step": 11613
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8392879442417496,
+      "learning_rate": 2.011693149571986e-07,
+      "loss": 0.5048,
+      "step": 11614
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5589643165917287,
+      "learning_rate": 2.0081234890634072e-07,
+      "loss": 0.4773,
+      "step": 11615
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.7195664601762457,
+      "learning_rate": 2.0045569335664595e-07,
+      "loss": 0.4187,
+      "step": 11616
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.308172576146993,
+      "learning_rate": 2.000993483311886e-07,
+      "loss": 0.4029,
+      "step": 11617
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5397308665185568,
+      "learning_rate": 1.9974331385302471e-07,
+      "loss": 0.4414,
+      "step": 11618
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6496242750131471,
+      "learning_rate": 1.9938758994518858e-07,
+      "loss": 0.4737,
+      "step": 11619
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5365566894805941,
+      "learning_rate": 1.990321766306963e-07,
+      "loss": 0.4651,
+      "step": 11620
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5735115217978257,
+      "learning_rate": 1.9867707393254055e-07,
+      "loss": 0.4782,
+      "step": 11621
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.2155802736907844,
+      "learning_rate": 1.9832228187369795e-07,
+      "loss": 0.4699,
+      "step": 11622
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4716095473700237,
+      "learning_rate": 1.9796780047712183e-07,
+      "loss": 0.4297,
+      "step": 11623
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.970630551248207,
+      "learning_rate": 1.976136297657477e-07,
+      "loss": 0.4208,
+      "step": 11624
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5832398212605996,
+      "learning_rate": 1.972597697624884e-07,
+      "loss": 0.4571,
+      "step": 11625
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4003509948693398,
+      "learning_rate": 1.9690622049024e-07,
+      "loss": 0.408,
+      "step": 11626
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.8614870074939063,
+      "learning_rate": 1.9655298197187646e-07,
+      "loss": 0.4654,
+      "step": 11627
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4231787220132968,
+      "learning_rate": 1.9620005423025056e-07,
+      "loss": 0.4819,
+      "step": 11628
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5170338117692749,
+      "learning_rate": 1.958474372881969e-07,
+      "loss": 0.4627,
+      "step": 11629
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6075157976621832,
+      "learning_rate": 1.9549513116852937e-07,
+      "loss": 0.4597,
+      "step": 11630
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3814228706855194,
+      "learning_rate": 1.9514313589404254e-07,
+      "loss": 0.4149,
+      "step": 11631
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5144638306441636,
+      "learning_rate": 1.947914514875088e-07,
+      "loss": 0.4471,
+      "step": 11632
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.827336960243377,
+      "learning_rate": 1.9444007797168208e-07,
+      "loss": 0.4891,
+      "step": 11633
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6389174844848375,
+      "learning_rate": 1.9408901536929592e-07,
+      "loss": 0.4981,
+      "step": 11634
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1337538953464175,
+      "learning_rate": 1.937382637030638e-07,
+      "loss": 0.4874,
+      "step": 11635
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3994266266616378,
+      "learning_rate": 1.9338782299567815e-07,
+      "loss": 0.4651,
+      "step": 11636
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.3322566128995708,
+      "learning_rate": 1.9303769326981193e-07,
+      "loss": 0.41,
+      "step": 11637
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.820720599621765,
+      "learning_rate": 1.9268787454812032e-07,
+      "loss": 0.3925,
+      "step": 11638
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5775772691937925,
+      "learning_rate": 1.9233836685323303e-07,
+      "loss": 0.4572,
+      "step": 11639
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.8084868230218274,
+      "learning_rate": 1.9198917020776532e-07,
+      "loss": 0.4723,
+      "step": 11640
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3103122734403823,
+      "learning_rate": 1.9164028463430738e-07,
+      "loss": 0.4105,
+      "step": 11641
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.1628537869307385,
+      "learning_rate": 1.91291710155434e-07,
+      "loss": 0.4057,
+      "step": 11642
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5206609882420195,
+      "learning_rate": 1.9094344679369602e-07,
+      "loss": 0.4058,
+      "step": 11643
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4943644369273077,
+      "learning_rate": 1.9059549457162652e-07,
+      "loss": 0.4503,
+      "step": 11644
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4623873662846762,
+      "learning_rate": 1.902478535117358e-07,
+      "loss": 0.4249,
+      "step": 11645
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.6449577851391033,
+      "learning_rate": 1.8990052363651812e-07,
+      "loss": 0.4197,
+      "step": 11646
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 3.5875981473689658,
+      "learning_rate": 1.8955350496844382e-07,
+      "loss": 0.4853,
+      "step": 11647
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.6546410222302237,
+      "learning_rate": 1.8920679752996441e-07,
+      "loss": 0.3984,
+      "step": 11648
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.680195054991957,
+      "learning_rate": 1.8886040134351248e-07,
+      "loss": 0.4421,
+      "step": 11649
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 2.143420574787109,
+      "learning_rate": 1.8851431643149843e-07,
+      "loss": 0.4128,
+      "step": 11650
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5329201202497394,
+      "learning_rate": 1.8816854281631435e-07,
+      "loss": 0.4747,
+      "step": 11651
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9878864414228619,
+      "learning_rate": 1.8782308052033072e-07,
+      "loss": 0.3681,
+      "step": 11652
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7494448824314872,
+      "learning_rate": 1.8747792956589793e-07,
+      "loss": 0.4904,
+      "step": 11653
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7084955679775087,
+      "learning_rate": 1.8713308997534819e-07,
+      "loss": 0.4709,
+      "step": 11654
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5248012759503514,
+      "learning_rate": 1.8678856177099193e-07,
+      "loss": 0.4722,
+      "step": 11655
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0269088247804787,
+      "learning_rate": 1.8644434497511855e-07,
+      "loss": 0.4441,
+      "step": 11656
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.018763957139318,
+      "learning_rate": 1.8610043960999968e-07,
+      "loss": 0.4347,
+      "step": 11657
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.9314586164153997,
+      "learning_rate": 1.8575684569788422e-07,
+      "loss": 0.461,
+      "step": 11658
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9558629127624854,
+      "learning_rate": 1.8541356326100436e-07,
+      "loss": 0.5043,
+      "step": 11659
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.549313648712087,
+      "learning_rate": 1.8507059232156677e-07,
+      "loss": 0.4707,
+      "step": 11660
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5672296151584785,
+      "learning_rate": 1.8472793290176428e-07,
+      "loss": 0.4638,
+      "step": 11661
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4357285619514586,
+      "learning_rate": 1.843855850237658e-07,
+      "loss": 0.4392,
+      "step": 11662
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9811760602233783,
+      "learning_rate": 1.8404354870971975e-07,
+      "loss": 0.452,
+      "step": 11663
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5425607736085312,
+      "learning_rate": 1.8370182398175563e-07,
+      "loss": 0.493,
+      "step": 11664
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.715600325251407,
+      "learning_rate": 1.8336041086198353e-07,
+      "loss": 0.4879,
+      "step": 11665
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7025627579313158,
+      "learning_rate": 1.8301930937249247e-07,
+      "loss": 0.3811,
+      "step": 11666
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1102612868056516,
+      "learning_rate": 1.8267851953534975e-07,
+      "loss": 0.432,
+      "step": 11667
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9142350793186051,
+      "learning_rate": 1.8233804137260502e-07,
+      "loss": 0.4279,
+      "step": 11668
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3686310315645536,
+      "learning_rate": 1.8199787490628672e-07,
+      "loss": 0.4218,
+      "step": 11669
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.094288253139774,
+      "learning_rate": 1.8165802015840394e-07,
+      "loss": 0.5162,
+      "step": 11670
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6267566053431828,
+      "learning_rate": 1.8131847715094354e-07,
+      "loss": 0.4153,
+      "step": 11671
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5263953086911533,
+      "learning_rate": 1.8097924590587345e-07,
+      "loss": 0.3842,
+      "step": 11672
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8573317461697971,
+      "learning_rate": 1.8064032644514284e-07,
+      "loss": 0.4834,
+      "step": 11673
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6050083062188751,
+      "learning_rate": 1.8030171879067858e-07,
+      "loss": 0.5432,
+      "step": 11674
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.85345786438777,
+      "learning_rate": 1.7996342296438817e-07,
+      "loss": 0.4472,
+      "step": 11675
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9006927653013674,
+      "learning_rate": 1.7962543898815798e-07,
+      "loss": 0.4083,
+      "step": 11676
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.43383714114517,
+      "learning_rate": 1.7928776688385774e-07,
+      "loss": 0.452,
+      "step": 11677
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5538166493161042,
+      "learning_rate": 1.7895040667333162e-07,
+      "loss": 0.4886,
+      "step": 11678
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5691896060784905,
+      "learning_rate": 1.7861335837840777e-07,
+      "loss": 0.4739,
+      "step": 11679
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.831635097430381,
+      "learning_rate": 1.7827662202089147e-07,
+      "loss": 0.449,
+      "step": 11680
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5493448542797732,
+      "learning_rate": 1.7794019762257143e-07,
+      "loss": 0.4695,
+      "step": 11681
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.595412397775315,
+      "learning_rate": 1.776040852052119e-07,
+      "loss": 0.3759,
+      "step": 11682
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.810619668862836,
+      "learning_rate": 1.772682847905599e-07,
+      "loss": 0.4705,
+      "step": 11683
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7690280015596287,
+      "learning_rate": 1.769327964003409e-07,
+      "loss": 0.4505,
+      "step": 11684
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9531642864754117,
+      "learning_rate": 1.7659762005626025e-07,
+      "loss": 0.4933,
+      "step": 11685
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3832544272593834,
+      "learning_rate": 1.762627557800045e-07,
+      "loss": 0.4736,
+      "step": 11686
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6428866634346775,
+      "learning_rate": 1.7592820359323748e-07,
+      "loss": 0.4457,
+      "step": 11687
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.459835124730534,
+      "learning_rate": 1.7559396351760516e-07,
+      "loss": 0.4679,
+      "step": 11688
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9869478851280151,
+      "learning_rate": 1.752600355747325e-07,
+      "loss": 0.4854,
+      "step": 11689
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8889244266455447,
+      "learning_rate": 1.7492641978622382e-07,
+      "loss": 0.4471,
+      "step": 11690
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.8149760252284493,
+      "learning_rate": 1.7459311617366415e-07,
+      "loss": 0.3922,
+      "step": 11691
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7654334978113753,
+      "learning_rate": 1.7426012475861676e-07,
+      "loss": 0.4416,
+      "step": 11692
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.546519247637802,
+      "learning_rate": 1.7392744556262665e-07,
+      "loss": 0.4635,
+      "step": 11693
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.213320408149971,
+      "learning_rate": 1.735950786072177e-07,
+      "loss": 0.4437,
+      "step": 11694
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4796898018508646,
+      "learning_rate": 1.7326302391389327e-07,
+      "loss": 0.4866,
+      "step": 11695
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7478029664733126,
+      "learning_rate": 1.7293128150413674e-07,
+      "loss": 0.4483,
+      "step": 11696
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6975966311324375,
+      "learning_rate": 1.725998513994126e-07,
+      "loss": 0.4207,
+      "step": 11697
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0983968243646123,
+      "learning_rate": 1.722687336211626e-07,
+      "loss": 0.4871,
+      "step": 11698
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5189287837012568,
+      "learning_rate": 1.7193792819081012e-07,
+      "loss": 0.4396,
+      "step": 11699
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7631721230813067,
+      "learning_rate": 1.7160743512975753e-07,
+      "loss": 0.4379,
+      "step": 11700
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.1318750828811144,
+      "learning_rate": 1.7127725445938827e-07,
+      "loss": 0.4334,
+      "step": 11701
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.2315353334409838,
+      "learning_rate": 1.70947386201063e-07,
+      "loss": 0.4221,
+      "step": 11702
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5154710352699057,
+      "learning_rate": 1.7061783037612468e-07,
+      "loss": 0.4226,
+      "step": 11703
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5780235033149828,
+      "learning_rate": 1.7028858700589567e-07,
+      "loss": 0.4612,
+      "step": 11704
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6857071600323215,
+      "learning_rate": 1.6995965611167676e-07,
+      "loss": 0.459,
+      "step": 11705
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8636272294791172,
+      "learning_rate": 1.6963103771474977e-07,
+      "loss": 0.4818,
+      "step": 11706
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7058722445479702,
+      "learning_rate": 1.6930273183637547e-07,
+      "loss": 0.444,
+      "step": 11707
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5762754065314423,
+      "learning_rate": 1.6897473849779522e-07,
+      "loss": 0.4674,
+      "step": 11708
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.896069031975638,
+      "learning_rate": 1.6864705772022927e-07,
+      "loss": 0.4441,
+      "step": 11709
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.388427535974961,
+      "learning_rate": 1.6831968952487897e-07,
+      "loss": 0.3808,
+      "step": 11710
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.209481080411894,
+      "learning_rate": 1.6799263393292352e-07,
+      "loss": 0.428,
+      "step": 11711
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5409480482351268,
+      "learning_rate": 1.6766589096552432e-07,
+      "loss": 0.4754,
+      "step": 11712
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6017806126677618,
+      "learning_rate": 1.6733946064382002e-07,
+      "loss": 0.4349,
+      "step": 11713
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.117835367983883,
+      "learning_rate": 1.6701334298893146e-07,
+      "loss": 0.4711,
+      "step": 11714
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5793243300401658,
+      "learning_rate": 1.6668753802195624e-07,
+      "loss": 0.4055,
+      "step": 11715
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7178818244728493,
+      "learning_rate": 1.6636204576397474e-07,
+      "loss": 0.4223,
+      "step": 11716
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.921887025210037,
+      "learning_rate": 1.660368662360462e-07,
+      "loss": 0.4347,
+      "step": 11717
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4684232702140438,
+      "learning_rate": 1.6571199945920824e-07,
+      "loss": 0.4019,
+      "step": 11718
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8215305646883584,
+      "learning_rate": 1.6538744545448016e-07,
+      "loss": 0.4282,
+      "step": 11719
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3457139713756714,
+      "learning_rate": 1.650632042428596e-07,
+      "loss": 0.4706,
+      "step": 11720
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8460228167536126,
+      "learning_rate": 1.647392758453259e-07,
+      "loss": 0.4424,
+      "step": 11721
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5670311731254835,
+      "learning_rate": 1.644156602828345e-07,
+      "loss": 0.4618,
+      "step": 11722
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5385163643714175,
+      "learning_rate": 1.640923575763248e-07,
+      "loss": 0.4623,
+      "step": 11723
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7504067138636776,
+      "learning_rate": 1.6376936774671282e-07,
+      "loss": 0.4806,
+      "step": 11724
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7419246664879728,
+      "learning_rate": 1.6344669081489684e-07,
+      "loss": 0.4719,
+      "step": 11725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.508917603475176,
+      "learning_rate": 1.6312432680175294e-07,
+      "loss": 0.5188,
+      "step": 11726
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5041972126193592,
+      "learning_rate": 1.6280227572813723e-07,
+      "loss": 0.4522,
+      "step": 11727
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8770011437191434,
+      "learning_rate": 1.6248053761488636e-07,
+      "loss": 0.4168,
+      "step": 11728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.065771088725907,
+      "learning_rate": 1.6215911248281646e-07,
+      "loss": 0.4602,
+      "step": 11729
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5790262580533339,
+      "learning_rate": 1.6183800035272312e-07,
+      "loss": 0.4678,
+      "step": 11730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.072923969698755,
+      "learning_rate": 1.615172012453825e-07,
+      "loss": 0.4187,
+      "step": 11731
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.072401735812162,
+      "learning_rate": 1.6119671518154966e-07,
+      "loss": 0.4527,
+      "step": 11732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.287261237985155,
+      "learning_rate": 1.6087654218195914e-07,
+      "loss": 0.4429,
+      "step": 11733
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.51499823437332,
+      "learning_rate": 1.6055668226732602e-07,
+      "loss": 0.4671,
+      "step": 11734
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5390546737187387,
+      "learning_rate": 1.602371354583443e-07,
+      "loss": 0.4722,
+      "step": 11735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5772369128030052,
+      "learning_rate": 1.5991790177568967e-07,
+      "loss": 0.4917,
+      "step": 11736
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6141040547021595,
+      "learning_rate": 1.595989812400145e-07,
+      "loss": 0.5119,
+      "step": 11737
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.1590915769767207,
+      "learning_rate": 1.5928037387195283e-07,
+      "loss": 0.4265,
+      "step": 11738
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7952018915691113,
+      "learning_rate": 1.5896207969211873e-07,
+      "loss": 0.4702,
+      "step": 11739
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3784287043932126,
+      "learning_rate": 1.586440987211052e-07,
+      "loss": 0.4555,
+      "step": 11740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8408478264027306,
+      "learning_rate": 1.583264309794863e-07,
+      "loss": 0.4219,
+      "step": 11741
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5522152831646425,
+      "learning_rate": 1.5800907648781226e-07,
+      "loss": 0.401,
+      "step": 11742
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5848418673087012,
+      "learning_rate": 1.5769203526661725e-07,
+      "loss": 0.4836,
+      "step": 11743
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6714697910516616,
+      "learning_rate": 1.5737530733641316e-07,
+      "loss": 0.4593,
+      "step": 11744
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.0432444829873164,
+      "learning_rate": 1.5705889271769193e-07,
+      "loss": 0.4546,
+      "step": 11745
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5415021101007814,
+      "learning_rate": 1.5674279143092386e-07,
+      "loss": 0.4756,
+      "step": 11746
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6342718162943994,
+      "learning_rate": 1.5642700349656315e-07,
+      "loss": 0.4146,
+      "step": 11747
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 4.007351409577253,
+      "learning_rate": 1.561115289350379e-07,
+      "loss": 0.3962,
+      "step": 11748
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 3.5274829211257144,
+      "learning_rate": 1.5579636776676065e-07,
+      "loss": 0.4788,
+      "step": 11749
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.068920914082165,
+      "learning_rate": 1.554815200121207e-07,
+      "loss": 0.4625,
+      "step": 11750
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9001688101723404,
+      "learning_rate": 1.5516698569148946e-07,
+      "loss": 0.448,
+      "step": 11751
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.072169792362131,
+      "learning_rate": 1.5485276482521683e-07,
+      "loss": 0.4276,
+      "step": 11752
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7416557488472462,
+      "learning_rate": 1.5453885743363152e-07,
+      "loss": 0.4124,
+      "step": 11753
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 7.585707352257947,
+      "learning_rate": 1.5422526353704337e-07,
+      "loss": 0.4148,
+      "step": 11754
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5493074466594068,
+      "learning_rate": 1.5391198315574175e-07,
+      "loss": 0.4154,
+      "step": 11755
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6782831014129347,
+      "learning_rate": 1.5359901630999596e-07,
+      "loss": 0.4625,
+      "step": 11756
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5275036593292459,
+      "learning_rate": 1.5328636302005317e-07,
+      "loss": 0.4483,
+      "step": 11757
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8550485810493866,
+      "learning_rate": 1.5297402330614276e-07,
+      "loss": 0.4682,
+      "step": 11758
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5052002794253732,
+      "learning_rate": 1.5266199718847185e-07,
+      "loss": 0.4214,
+      "step": 11759
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5472971086148162,
+      "learning_rate": 1.5235028468722934e-07,
+      "loss": 0.44,
+      "step": 11760
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8911524856423803,
+      "learning_rate": 1.5203888582258187e-07,
+      "loss": 0.4484,
+      "step": 11761
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7015379917537463,
+      "learning_rate": 1.5172780061467551e-07,
+      "loss": 0.4875,
+      "step": 11762
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.4568833628671496,
+      "learning_rate": 1.514170290836392e-07,
+      "loss": 0.4952,
+      "step": 11763
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5476256632750964,
+      "learning_rate": 1.511065712495774e-07,
+      "loss": 0.4603,
+      "step": 11764
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.304871431191685,
+      "learning_rate": 1.5079642713257792e-07,
+      "loss": 0.4606,
+      "step": 11765
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.8658886055179422,
+      "learning_rate": 1.5048659675270583e-07,
+      "loss": 0.4556,
+      "step": 11766
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7158709460536223,
+      "learning_rate": 1.5017708013000787e-07,
+      "loss": 0.4547,
+      "step": 11767
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5690187088992504,
+      "learning_rate": 1.4986787728450747e-07,
+      "loss": 0.4832,
+      "step": 11768
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.66863306240449,
+      "learning_rate": 1.4955898823621084e-07,
+      "loss": 0.476,
+      "step": 11769
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.6361697280701761,
+      "learning_rate": 1.49250413005102e-07,
+      "loss": 0.4889,
+      "step": 11770
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.6268884981300578,
+      "learning_rate": 1.4894215161114722e-07,
+      "loss": 0.4754,
+      "step": 11771
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.771910638155748,
+      "learning_rate": 1.4863420407428831e-07,
+      "loss": 0.4555,
+      "step": 11772
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.403239337633192,
+      "learning_rate": 1.4832657041444986e-07,
+      "loss": 0.468,
+      "step": 11773
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.128621604992297,
+      "learning_rate": 1.4801925065153544e-07,
+      "loss": 0.4175,
+      "step": 11774
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3295265146661326,
+      "learning_rate": 1.4771224480542857e-07,
+      "loss": 0.4661,
+      "step": 11775
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.513466561118754,
+      "learning_rate": 1.4740555289599278e-07,
+      "loss": 0.4737,
+      "step": 11776
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3361652599373923,
+      "learning_rate": 1.4709917494306836e-07,
+      "loss": 0.4472,
+      "step": 11777
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9373870785458802,
+      "learning_rate": 1.467931109664794e-07,
+      "loss": 0.4632,
+      "step": 11778
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.535779200380525,
+      "learning_rate": 1.4648736098602734e-07,
+      "loss": 0.4941,
+      "step": 11779
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5313879338354166,
+      "learning_rate": 1.4618192502149465e-07,
+      "loss": 0.3927,
+      "step": 11780
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.505408323235525,
+      "learning_rate": 1.4587680309264053e-07,
+      "loss": 0.4684,
+      "step": 11781
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5658704565250217,
+      "learning_rate": 1.4557199521920806e-07,
+      "loss": 0.4579,
+      "step": 11782
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.579745614303645,
+      "learning_rate": 1.452675014209165e-07,
+      "loss": 0.4513,
+      "step": 11783
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7317307861336095,
+      "learning_rate": 1.449633217174673e-07,
+      "loss": 0.4533,
+      "step": 11784
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5625232936981786,
+      "learning_rate": 1.4465945612853972e-07,
+      "loss": 0.3953,
+      "step": 11785
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8031530607266502,
+      "learning_rate": 1.4435590467379356e-07,
+      "loss": 0.3926,
+      "step": 11786
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.767036524024196,
+      "learning_rate": 1.4405266737286926e-07,
+      "loss": 0.4477,
+      "step": 11787
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6921663471674042,
+      "learning_rate": 1.4374974424538446e-07,
+      "loss": 0.4732,
+      "step": 11788
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5205772729405225,
+      "learning_rate": 1.4344713531093845e-07,
+      "loss": 0.4489,
+      "step": 11789
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.686222573396598,
+      "learning_rate": 1.4314484058910949e-07,
+      "loss": 0.4211,
+      "step": 11790
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5444070947995359,
+      "learning_rate": 1.4284286009945636e-07,
+      "loss": 0.4628,
+      "step": 11791
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.589692956958823,
+      "learning_rate": 1.4254119386151567e-07,
+      "loss": 0.4839,
+      "step": 11792
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8288403791287917,
+      "learning_rate": 1.4223984189480512e-07,
+      "loss": 0.4608,
+      "step": 11793
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7461559182713902,
+      "learning_rate": 1.4193880421882245e-07,
+      "loss": 0.426,
+      "step": 11794
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.814427882174773,
+      "learning_rate": 1.416380808530443e-07,
+      "loss": 0.4028,
+      "step": 11795
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8279203400288289,
+      "learning_rate": 1.4133767181692626e-07,
+      "loss": 0.4424,
+      "step": 11796
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5753394666998932,
+      "learning_rate": 1.410375771299044e-07,
+      "loss": 0.5023,
+      "step": 11797
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5094929337371277,
+      "learning_rate": 1.4073779681139655e-07,
+      "loss": 0.4065,
+      "step": 11798
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3697664726896503,
+      "learning_rate": 1.404383308807955e-07,
+      "loss": 0.4596,
+      "step": 11799
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9172992439108292,
+      "learning_rate": 1.4013917935747744e-07,
+      "loss": 0.4382,
+      "step": 11800
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.181812083023224,
+      "learning_rate": 1.398403422607969e-07,
+      "loss": 0.44,
+      "step": 11801
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6667817268286442,
+      "learning_rate": 1.3954181961008895e-07,
+      "loss": 0.4038,
+      "step": 11802
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.228499181348054,
+      "learning_rate": 1.3924361142466647e-07,
+      "loss": 0.4799,
+      "step": 11803
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5446933006211802,
+      "learning_rate": 1.389457177238235e-07,
+      "loss": 0.4672,
+      "step": 11804
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9628941708710212,
+      "learning_rate": 1.3864813852683346e-07,
+      "loss": 0.4206,
+      "step": 11805
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2225193719545264,
+      "learning_rate": 1.3835087385294988e-07,
+      "loss": 0.4079,
+      "step": 11806
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.748008663483608,
+      "learning_rate": 1.3805392372140514e-07,
+      "loss": 0.4172,
+      "step": 11807
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8159308094436208,
+      "learning_rate": 1.3775728815141055e-07,
+      "loss": 0.4658,
+      "step": 11808
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6111833964296154,
+      "learning_rate": 1.3746096716215906e-07,
+      "loss": 0.4822,
+      "step": 11809
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1832674905720406,
+      "learning_rate": 1.3716496077282205e-07,
+      "loss": 0.479,
+      "step": 11810
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7019671962116305,
+      "learning_rate": 1.3686926900255082e-07,
+      "loss": 0.4273,
+      "step": 11811
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5820665714242095,
+      "learning_rate": 1.3657389187047622e-07,
+      "loss": 0.4271,
+      "step": 11812
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5017301672721985,
+      "learning_rate": 1.3627882939570848e-07,
+      "loss": 0.4603,
+      "step": 11813
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5894047672238247,
+      "learning_rate": 1.359840815973379e-07,
+      "loss": 0.4129,
+      "step": 11814
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5779610539283278,
+      "learning_rate": 1.3568964849443421e-07,
+      "loss": 0.4441,
+      "step": 11815
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4367344687851928,
+      "learning_rate": 1.3539553010604666e-07,
+      "loss": 0.4902,
+      "step": 11816
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6874972787753546,
+      "learning_rate": 1.3510172645120556e-07,
+      "loss": 0.4714,
+      "step": 11817
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.867792170666657,
+      "learning_rate": 1.348082375489179e-07,
+      "loss": 0.4133,
+      "step": 11818
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.038960097807598,
+      "learning_rate": 1.34515063418173e-07,
+      "loss": 0.403,
+      "step": 11819
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5394822185844133,
+      "learning_rate": 1.342222040779384e-07,
+      "loss": 0.4609,
+      "step": 11820
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 7.48923574368581,
+      "learning_rate": 1.339296595471623e-07,
+      "loss": 0.4454,
+      "step": 11821
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.888465791767646,
+      "learning_rate": 1.3363742984477178e-07,
+      "loss": 0.4448,
+      "step": 11822
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4698326655386835,
+      "learning_rate": 1.3334551498967342e-07,
+      "loss": 0.4045,
+      "step": 11823
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.971319832167152,
+      "learning_rate": 1.3305391500075315e-07,
+      "loss": 0.4427,
+      "step": 11824
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7727151415737878,
+      "learning_rate": 1.327626298968787e-07,
+      "loss": 0.432,
+      "step": 11825
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2408672451898037,
+      "learning_rate": 1.32471659696895e-07,
+      "loss": 0.4308,
+      "step": 11826
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 10.576973474127147,
+      "learning_rate": 1.32181004419627e-07,
+      "loss": 0.4112,
+      "step": 11827
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8961084154561636,
+      "learning_rate": 1.3189066408387963e-07,
+      "loss": 0.4268,
+      "step": 11828
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.0636197551604933,
+      "learning_rate": 1.3160063870843842e-07,
+      "loss": 0.4425,
+      "step": 11829
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1016892919053314,
+      "learning_rate": 1.3131092831206727e-07,
+      "loss": 0.4146,
+      "step": 11830
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6057473556341042,
+      "learning_rate": 1.310215329135106e-07,
+      "loss": 0.4293,
+      "step": 11831
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5329633082483953,
+      "learning_rate": 1.3073245253149013e-07,
+      "loss": 0.4622,
+      "step": 11832
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5057841243858959,
+      "learning_rate": 1.3044368718471145e-07,
+      "loss": 0.4549,
+      "step": 11833
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.395646040385583,
+      "learning_rate": 1.3015523689185515e-07,
+      "loss": 0.416,
+      "step": 11834
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.01222230752039,
+      "learning_rate": 1.2986710167158466e-07,
+      "loss": 0.49,
+      "step": 11835
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 6.185042725076882,
+      "learning_rate": 1.2957928154254174e-07,
+      "loss": 0.4607,
+      "step": 11836
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.312650557878563,
+      "learning_rate": 1.2929177652334812e-07,
+      "loss": 0.4393,
+      "step": 11837
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.371576932473127,
+      "learning_rate": 1.2900458663260506e-07,
+      "loss": 0.4326,
+      "step": 11838
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.407646566140664,
+      "learning_rate": 1.287177118888927e-07,
+      "loss": 0.4396,
+      "step": 11839
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8307194252975234,
+      "learning_rate": 1.284311523107723e-07,
+      "loss": 0.4095,
+      "step": 11840
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.289413248782656,
+      "learning_rate": 1.281449079167829e-07,
+      "loss": 0.3817,
+      "step": 11841
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6516666861123825,
+      "learning_rate": 1.2785897872544585e-07,
+      "loss": 0.4696,
+      "step": 11842
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4494553662953884,
+      "learning_rate": 1.275733647552585e-07,
+      "loss": 0.412,
+      "step": 11843
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4287845930609309,
+      "learning_rate": 1.2728806602470112e-07,
+      "loss": 0.419,
+      "step": 11844
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5509627825945641,
+      "learning_rate": 1.2700308255223115e-07,
+      "loss": 0.4777,
+      "step": 11845
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.679124111023232,
+      "learning_rate": 1.2671841435628718e-07,
+      "loss": 0.4597,
+      "step": 11846
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7954668389820017,
+      "learning_rate": 1.2643406145528668e-07,
+      "loss": 0.4047,
+      "step": 11847
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2448436276847707,
+      "learning_rate": 1.2615002386762665e-07,
+      "loss": 0.4558,
+      "step": 11848
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.3496923961014495,
+      "learning_rate": 1.2586630161168456e-07,
+      "loss": 0.4378,
+      "step": 11849
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.1726860695772445,
+      "learning_rate": 1.2558289470581742e-07,
+      "loss": 0.5144,
+      "step": 11850
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6564585071185243,
+      "learning_rate": 1.2529980316835888e-07,
+      "loss": 0.4266,
+      "step": 11851
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5135238247412202,
+      "learning_rate": 1.2501702701762708e-07,
+      "loss": 0.4668,
+      "step": 11852
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5370747266075647,
+      "learning_rate": 1.2473456627191626e-07,
+      "loss": 0.4773,
+      "step": 11853
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5692125959339073,
+      "learning_rate": 1.2445242094950127e-07,
+      "loss": 0.4804,
+      "step": 11854
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5899320164988602,
+      "learning_rate": 1.2417059106863638e-07,
+      "loss": 0.4698,
+      "step": 11855
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5428892197021701,
+      "learning_rate": 1.2388907664755644e-07,
+      "loss": 0.4833,
+      "step": 11856
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1407326778783444,
+      "learning_rate": 1.236078777044747e-07,
+      "loss": 0.4428,
+      "step": 11857
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9358425855056336,
+      "learning_rate": 1.233269942575832e-07,
+      "loss": 0.4602,
+      "step": 11858
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0796795927982394,
+      "learning_rate": 1.2304642632505637e-07,
+      "loss": 0.4253,
+      "step": 11859
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2878413539009093,
+      "learning_rate": 1.227661739250452e-07,
+      "loss": 0.4794,
+      "step": 11860
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.724457566600411,
+      "learning_rate": 1.2248623707568352e-07,
+      "loss": 0.4119,
+      "step": 11861
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6095098691158997,
+      "learning_rate": 1.2220661579508075e-07,
+      "loss": 0.4726,
+      "step": 11862
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.341348735676004,
+      "learning_rate": 1.2192731010132853e-07,
+      "loss": 0.4496,
+      "step": 11863
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0959586589127315,
+      "learning_rate": 1.2164832001249903e-07,
+      "loss": 0.4609,
+      "step": 11864
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5378965620990586,
+      "learning_rate": 1.213696455466412e-07,
+      "loss": 0.4693,
+      "step": 11865
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5370492982726289,
+      "learning_rate": 1.2109128672178617e-07,
+      "loss": 0.455,
+      "step": 11866
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.881210373608352,
+      "learning_rate": 1.208132435559406e-07,
+      "loss": 0.4274,
+      "step": 11867
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8043808582421739,
+      "learning_rate": 1.205355160670968e-07,
+      "loss": 0.4434,
+      "step": 11868
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2422465420098794,
+      "learning_rate": 1.2025810427322149e-07,
+      "loss": 0.3805,
+      "step": 11869
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.0978954846853592,
+      "learning_rate": 1.1998100819226366e-07,
+      "loss": 0.4428,
+      "step": 11870
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 3.235537665464353,
+      "learning_rate": 1.1970422784215007e-07,
+      "loss": 0.4483,
+      "step": 11871
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9616152643456302,
+      "learning_rate": 1.1942776324078975e-07,
+      "loss": 0.4477,
+      "step": 11872
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.141307254675652,
+      "learning_rate": 1.1915161440606782e-07,
+      "loss": 0.4625,
+      "step": 11873
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5499304748956896,
+      "learning_rate": 1.188757813558511e-07,
+      "loss": 0.4539,
+      "step": 11874
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.906511863159606,
+      "learning_rate": 1.1860026410798641e-07,
+      "loss": 0.4581,
+      "step": 11875
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8112240483443718,
+      "learning_rate": 1.1832506268029898e-07,
+      "loss": 0.4797,
+      "step": 11876
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.688636235111136,
+      "learning_rate": 1.1805017709059452e-07,
+      "loss": 0.4368,
+      "step": 11877
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4453036420908236,
+      "learning_rate": 1.177756073566566e-07,
+      "loss": 0.4638,
+      "step": 11878
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9924358178709283,
+      "learning_rate": 1.1750135349625048e-07,
+      "loss": 0.4663,
+      "step": 11879
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9546958544048727,
+      "learning_rate": 1.1722741552711914e-07,
+      "loss": 0.4425,
+      "step": 11880
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8598562752448733,
+      "learning_rate": 1.1695379346698732e-07,
+      "loss": 0.3992,
+      "step": 11881
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.381059754564645,
+      "learning_rate": 1.1668048733355642e-07,
+      "loss": 0.4664,
+      "step": 11882
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.664146995013794,
+      "learning_rate": 1.1640749714451005e-07,
+      "loss": 0.4481,
+      "step": 11883
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 4.693232257969307,
+      "learning_rate": 1.1613482291751022e-07,
+      "loss": 0.4816,
+      "step": 11884
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7378869961476118,
+      "learning_rate": 1.1586246467019891e-07,
+      "loss": 0.4262,
+      "step": 11885
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5184969052000743,
+      "learning_rate": 1.1559042242019592e-07,
+      "loss": 0.3681,
+      "step": 11886
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5455151557725871,
+      "learning_rate": 1.1531869618510327e-07,
+      "loss": 0.4499,
+      "step": 11887
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.2787494235237395,
+      "learning_rate": 1.1504728598250136e-07,
+      "loss": 0.4623,
+      "step": 11888
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8588183388938833,
+      "learning_rate": 1.1477619182994948e-07,
+      "loss": 0.4085,
+      "step": 11889
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.379633366592021,
+      "learning_rate": 1.1450541374498747e-07,
+      "loss": 0.3935,
+      "step": 11890
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7485782022556036,
+      "learning_rate": 1.142349517451341e-07,
+      "loss": 0.4665,
+      "step": 11891
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.939857937332404,
+      "learning_rate": 1.1396480584788816e-07,
+      "loss": 0.4849,
+      "step": 11892
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8402968481618627,
+      "learning_rate": 1.1369497607072732e-07,
+      "loss": 0.4328,
+      "step": 11893
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.24751357273807,
+      "learning_rate": 1.134254624311093e-07,
+      "loss": 0.4644,
+      "step": 11894
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.503446791590182,
+      "learning_rate": 1.1315626494647126e-07,
+      "loss": 0.4626,
+      "step": 11895
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.612353147795168,
+      "learning_rate": 1.1288738363423091e-07,
+      "loss": 0.4853,
+      "step": 11896
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7931178423103957,
+      "learning_rate": 1.1261881851178325e-07,
+      "loss": 0.4546,
+      "step": 11897
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.317029515439848,
+      "learning_rate": 1.1235056959650381e-07,
+      "loss": 0.4648,
+      "step": 11898
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4037289430608917,
+      "learning_rate": 1.1208263690574983e-07,
+      "loss": 0.4456,
+      "step": 11899
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.48387276655709,
+      "learning_rate": 1.1181502045685411e-07,
+      "loss": 0.4614,
+      "step": 11900
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7392189250613297,
+      "learning_rate": 1.1154772026713278e-07,
+      "loss": 0.4368,
+      "step": 11901
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8668597737426498,
+      "learning_rate": 1.1128073635387815e-07,
+      "loss": 0.4094,
+      "step": 11902
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.7331114233805012,
+      "learning_rate": 1.1101406873436583e-07,
+      "loss": 0.4409,
+      "step": 11903
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5133502351552798,
+      "learning_rate": 1.1074771742584645e-07,
+      "loss": 0.4746,
+      "step": 11904
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5716708220319111,
+      "learning_rate": 1.1048168244555513e-07,
+      "loss": 0.4567,
+      "step": 11905
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4851554081146392,
+      "learning_rate": 1.1021596381070144e-07,
+      "loss": 0.4164,
+      "step": 11906
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4458411056171714,
+      "learning_rate": 1.0995056153847883e-07,
+      "loss": 0.4738,
+      "step": 11907
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.5420012444926368,
+      "learning_rate": 1.0968547564605747e-07,
+      "loss": 0.3982,
+      "step": 11908
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7359012604853308,
+      "learning_rate": 1.094207061505892e-07,
+      "loss": 0.4411,
+      "step": 11909
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9127195990424601,
+      "learning_rate": 1.0915625306920307e-07,
+      "loss": 0.4159,
+      "step": 11910
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5491286141939281,
+      "learning_rate": 1.0889211641900932e-07,
+      "loss": 0.4727,
+      "step": 11911
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.581806502731135,
+      "learning_rate": 1.0862829621709759e-07,
+      "loss": 0.4577,
+      "step": 11912
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5667499345099013,
+      "learning_rate": 1.0836479248053589e-07,
+      "loss": 0.4726,
+      "step": 11913
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.7898013645130377,
+      "learning_rate": 1.0810160522637336e-07,
+      "loss": 0.4765,
+      "step": 11914
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5324253851749927,
+      "learning_rate": 1.078387344716375e-07,
+      "loss": 0.4558,
+      "step": 11915
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.850392437971417,
+      "learning_rate": 1.0757618023333638e-07,
+      "loss": 0.4189,
+      "step": 11916
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6063968500130201,
+      "learning_rate": 1.0731394252845528e-07,
+      "loss": 0.4088,
+      "step": 11917
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.954495736238232,
+      "learning_rate": 1.0705202137396231e-07,
+      "loss": 0.4775,
+      "step": 11918
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.793191918679099,
+      "learning_rate": 1.0679041678680224e-07,
+      "loss": 0.4427,
+      "step": 11919
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5288792119075454,
+      "learning_rate": 1.0652912878390153e-07,
+      "loss": 0.4623,
+      "step": 11920
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5327368080932842,
+      "learning_rate": 1.0626815738216445e-07,
+      "loss": 0.47,
+      "step": 11921
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7568672842307334,
+      "learning_rate": 1.0600750259847581e-07,
+      "loss": 0.445,
+      "step": 11922
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6496598103123985,
+      "learning_rate": 1.057471644496999e-07,
+      "loss": 0.418,
+      "step": 11923
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8242774149216143,
+      "learning_rate": 1.0548714295267992e-07,
+      "loss": 0.4583,
+      "step": 11924
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.909606474818618,
+      "learning_rate": 1.0522743812423852e-07,
+      "loss": 0.4842,
+      "step": 11925
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5362836509408702,
+      "learning_rate": 1.0496804998117893e-07,
+      "loss": 0.4117,
+      "step": 11926
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.938663104269406,
+      "learning_rate": 1.0470897854028383e-07,
+      "loss": 0.4824,
+      "step": 11927
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7459478067827947,
+      "learning_rate": 1.0445022381831316e-07,
+      "loss": 0.4581,
+      "step": 11928
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.799667090012614,
+      "learning_rate": 1.0419178583200851e-07,
+      "loss": 0.4329,
+      "step": 11929
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.817532294050131,
+      "learning_rate": 1.0393366459809151e-07,
+      "loss": 0.4432,
+      "step": 11930
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5410605317043119,
+      "learning_rate": 1.0367586013326104e-07,
+      "loss": 0.4782,
+      "step": 11931
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 6.0164918633586515,
+      "learning_rate": 1.0341837245419761e-07,
+      "loss": 0.4298,
+      "step": 11932
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8023616736617374,
+      "learning_rate": 1.0316120157755904e-07,
+      "loss": 0.4301,
+      "step": 11933
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.08035251654865,
+      "learning_rate": 1.029043475199859e-07,
+      "loss": 0.4674,
+      "step": 11934
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5280545955846563,
+      "learning_rate": 1.026478102980949e-07,
+      "loss": 0.4237,
+      "step": 11935
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5355558222918961,
+      "learning_rate": 1.0239158992848442e-07,
+      "loss": 0.4775,
+      "step": 11936
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5548918877965017,
+      "learning_rate": 1.0213568642773064e-07,
+      "loss": 0.4738,
+      "step": 11937
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7010405354398268,
+      "learning_rate": 1.0188009981239089e-07,
+      "loss": 0.4567,
+      "step": 11938
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.971682598987677,
+      "learning_rate": 1.0162483009900137e-07,
+      "loss": 0.4605,
+      "step": 11939
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8486414200518737,
+      "learning_rate": 1.0136987730407776e-07,
+      "loss": 0.421,
+      "step": 11940
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.745743652754548,
+      "learning_rate": 1.0111524144411355e-07,
+      "loss": 0.4109,
+      "step": 11941
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6290193529352976,
+      "learning_rate": 1.0086092253558555e-07,
+      "loss": 0.4089,
+      "step": 11942
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.399057861432048,
+      "learning_rate": 1.0060692059494726e-07,
+      "loss": 0.4336,
+      "step": 11943
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1815565391110283,
+      "learning_rate": 1.0035323563863164e-07,
+      "loss": 0.4547,
+      "step": 11944
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7158134763297408,
+      "learning_rate": 1.0009986768305225e-07,
+      "loss": 0.4844,
+      "step": 11945
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9191455209149968,
+      "learning_rate": 9.98468167446015e-08,
+      "loss": 0.4335,
+      "step": 11946
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.612451584179278,
+      "learning_rate": 9.959408283965132e-08,
+      "loss": 0.3982,
+      "step": 11947
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.330184118253598,
+      "learning_rate": 9.934166598455364e-08,
+      "loss": 0.4807,
+      "step": 11948
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6963019474966408,
+      "learning_rate": 9.908956619563925e-08,
+      "loss": 0.4656,
+      "step": 11949
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5340791840168415,
+      "learning_rate": 9.883778348921846e-08,
+      "loss": 0.4865,
+      "step": 11950
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7673530539000595,
+      "learning_rate": 9.858631788158269e-08,
+      "loss": 0.4379,
+      "step": 11951
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.565390894706377,
+      "learning_rate": 9.833516938899889e-08,
+      "loss": 0.4651,
+      "step": 11952
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5371962289334098,
+      "learning_rate": 9.808433802771799e-08,
+      "loss": 0.4696,
+      "step": 11953
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.865953783534155,
+      "learning_rate": 9.783382381396811e-08,
+      "loss": 0.4583,
+      "step": 11954
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8496518662667734,
+      "learning_rate": 9.758362676395628e-08,
+      "loss": 0.4186,
+      "step": 11955
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5498701615126924,
+      "learning_rate": 9.73337468938712e-08,
+      "loss": 0.4842,
+      "step": 11956
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.653415683207238,
+      "learning_rate": 9.708418421987942e-08,
+      "loss": 0.4397,
+      "step": 11957
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.015837003194512,
+      "learning_rate": 9.683493875812688e-08,
+      "loss": 0.4137,
+      "step": 11958
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6087063513080675,
+      "learning_rate": 9.65860105247396e-08,
+      "loss": 0.4619,
+      "step": 11959
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.544591957260909,
+      "learning_rate": 9.6337399535823e-08,
+      "loss": 0.466,
+      "step": 11960
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4686581002285228,
+      "learning_rate": 9.608910580746145e-08,
+      "loss": 0.4207,
+      "step": 11961
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6442899317936135,
+      "learning_rate": 9.584112935572043e-08,
+      "loss": 0.4699,
+      "step": 11962
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.27312457993616,
+      "learning_rate": 9.559347019664267e-08,
+      "loss": 0.421,
+      "step": 11963
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.685791945300738,
+      "learning_rate": 9.53461283462509e-08,
+      "loss": 0.4614,
+      "step": 11964
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.9907961783823493,
+      "learning_rate": 9.509910382054899e-08,
+      "loss": 0.4612,
+      "step": 11965
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7820450876703018,
+      "learning_rate": 9.485239663551859e-08,
+      "loss": 0.4697,
+      "step": 11966
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.9810123048366193,
+      "learning_rate": 9.460600680712195e-08,
+      "loss": 0.4913,
+      "step": 11967
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.764547570391522,
+      "learning_rate": 9.435993435129853e-08,
+      "loss": 0.4745,
+      "step": 11968
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8924129971128811,
+      "learning_rate": 9.411417928397115e-08,
+      "loss": 0.4798,
+      "step": 11969
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.97297136998746,
+      "learning_rate": 9.386874162103821e-08,
+      "loss": 0.447,
+      "step": 11970
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5031608211588405,
+      "learning_rate": 9.362362137837978e-08,
+      "loss": 0.4246,
+      "step": 11971
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.1415973499809677,
+      "learning_rate": 9.337881857185426e-08,
+      "loss": 0.4136,
+      "step": 11972
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3171970365795675,
+      "learning_rate": 9.31343332173007e-08,
+      "loss": 0.4183,
+      "step": 11973
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.685742710756799,
+      "learning_rate": 9.289016533053696e-08,
+      "loss": 0.424,
+      "step": 11974
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8108158521644064,
+      "learning_rate": 9.264631492736043e-08,
+      "loss": 0.4484,
+      "step": 11975
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.927120095481386,
+      "learning_rate": 9.24027820235468e-08,
+      "loss": 0.4734,
+      "step": 11976
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.970899358122065,
+      "learning_rate": 9.215956663485348e-08,
+      "loss": 0.4696,
+      "step": 11977
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.792981198572327,
+      "learning_rate": 9.191666877701677e-08,
+      "loss": 0.4062,
+      "step": 11978
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.626086743518199,
+      "learning_rate": 9.167408846575022e-08,
+      "loss": 0.4593,
+      "step": 11979
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.737525146522079,
+      "learning_rate": 9.143182571674957e-08,
+      "loss": 0.4608,
+      "step": 11980
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 7.935269916152477,
+      "learning_rate": 9.118988054568844e-08,
+      "loss": 0.3818,
+      "step": 11981
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6439325173227501,
+      "learning_rate": 9.094825296822096e-08,
+      "loss": 0.4273,
+      "step": 11982
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7608430477874888,
+      "learning_rate": 9.070694299997906e-08,
+      "loss": 0.4532,
+      "step": 11983
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.442612307711923,
+      "learning_rate": 9.046595065657637e-08,
+      "loss": 0.4609,
+      "step": 11984
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7294814570900503,
+      "learning_rate": 9.022527595360375e-08,
+      "loss": 0.4221,
+      "step": 11985
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5299164766351682,
+      "learning_rate": 8.998491890663319e-08,
+      "loss": 0.4836,
+      "step": 11986
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4702638164515807,
+      "learning_rate": 8.974487953121502e-08,
+      "loss": 0.4399,
+      "step": 11987
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6158716646144935,
+      "learning_rate": 8.950515784287961e-08,
+      "loss": 0.4619,
+      "step": 11988
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.543920746919628,
+      "learning_rate": 8.926575385713731e-08,
+      "loss": 0.4503,
+      "step": 11989
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.005943766363169,
+      "learning_rate": 8.902666758947632e-08,
+      "loss": 0.4827,
+      "step": 11990
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 9.219600618481454,
+      "learning_rate": 8.878789905536533e-08,
+      "loss": 0.4136,
+      "step": 11991
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6054546682073494,
+      "learning_rate": 8.854944827025313e-08,
+      "loss": 0.4691,
+      "step": 11992
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7403299700870662,
+      "learning_rate": 8.831131524956682e-08,
+      "loss": 0.4097,
+      "step": 11993
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5500191021157173,
+      "learning_rate": 8.80735000087124e-08,
+      "loss": 0.4889,
+      "step": 11994
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8473662953143966,
+      "learning_rate": 8.783600256307701e-08,
+      "loss": 0.426,
+      "step": 11995
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.2492541375490505,
+      "learning_rate": 8.759882292802668e-08,
+      "loss": 0.4592,
+      "step": 11996
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5293817637611689,
+      "learning_rate": 8.736196111890638e-08,
+      "loss": 0.468,
+      "step": 11997
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7112564801047327,
+      "learning_rate": 8.71254171510405e-08,
+      "loss": 0.4274,
+      "step": 11998
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5482680416774683,
+      "learning_rate": 8.688919103973292e-08,
+      "loss": 0.4805,
+      "step": 11999
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2921730067327755,
+      "learning_rate": 8.665328280026808e-08,
+      "loss": 0.4523,
+      "step": 12000
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8922694503442175,
+      "learning_rate": 8.641769244790766e-08,
+      "loss": 0.4368,
+      "step": 12001
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.763895808284745,
+      "learning_rate": 8.618241999789556e-08,
+      "loss": 0.4167,
+      "step": 12002
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5721141942673393,
+      "learning_rate": 8.59474654654524e-08,
+      "loss": 0.427,
+      "step": 12003
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8860658456308081,
+      "learning_rate": 8.571282886577992e-08,
+      "loss": 0.4618,
+      "step": 12004
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5433129716077433,
+      "learning_rate": 8.547851021405873e-08,
+      "loss": 0.4994,
+      "step": 12005
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.754331412458918,
+      "learning_rate": 8.524450952544949e-08,
+      "loss": 0.4298,
+      "step": 12006
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6353948259040547,
+      "learning_rate": 8.50108268150901e-08,
+      "loss": 0.4357,
+      "step": 12007
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.384465789605593,
+      "learning_rate": 8.477746209810123e-08,
+      "loss": 0.4206,
+      "step": 12008
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5183290133880379,
+      "learning_rate": 8.454441538958025e-08,
+      "loss": 0.4745,
+      "step": 12009
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.0721333614224546,
+      "learning_rate": 8.431168670460566e-08,
+      "loss": 0.4232,
+      "step": 12010
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 4.984042666117495,
+      "learning_rate": 8.40792760582343e-08,
+      "loss": 0.4268,
+      "step": 12011
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5409935828601051,
+      "learning_rate": 8.384718346550302e-08,
+      "loss": 0.4686,
+      "step": 12012
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.443993223301895,
+      "learning_rate": 8.361540894142816e-08,
+      "loss": 0.4204,
+      "step": 12013
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.4094829799008024,
+      "learning_rate": 8.33839525010044e-08,
+      "loss": 0.3854,
+      "step": 12014
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.3902238271405243,
+      "learning_rate": 8.315281415920751e-08,
+      "loss": 0.4347,
+      "step": 12015
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8123749124699657,
+      "learning_rate": 8.29219939309911e-08,
+      "loss": 0.4353,
+      "step": 12016
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.148824752067686,
+      "learning_rate": 8.269149183128988e-08,
+      "loss": 0.4513,
+      "step": 12017
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 3.3997758670334317,
+      "learning_rate": 8.246130787501583e-08,
+      "loss": 0.456,
+      "step": 12018
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.482666262461286,
+      "learning_rate": 8.223144207706257e-08,
+      "loss": 0.4105,
+      "step": 12019
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5540410938636295,
+      "learning_rate": 8.200189445230212e-08,
+      "loss": 0.4704,
+      "step": 12020
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8120156422473894,
+      "learning_rate": 8.177266501558534e-08,
+      "loss": 0.4172,
+      "step": 12021
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8476028803184046,
+      "learning_rate": 8.154375378174317e-08,
+      "loss": 0.4135,
+      "step": 12022
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.705026539757959,
+      "learning_rate": 8.131516076558543e-08,
+      "loss": 0.3417,
+      "step": 12023
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8386646225885541,
+      "learning_rate": 8.108688598190306e-08,
+      "loss": 0.4464,
+      "step": 12024
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5201120201285709,
+      "learning_rate": 8.085892944546425e-08,
+      "loss": 0.4511,
+      "step": 12025
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7331608776490044,
+      "learning_rate": 8.06312911710172e-08,
+      "loss": 0.3949,
+      "step": 12026
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.7444903108343048,
+      "learning_rate": 8.040397117329068e-08,
+      "loss": 0.4283,
+      "step": 12027
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5353130995164832,
+      "learning_rate": 8.017696946699183e-08,
+      "loss": 0.4711,
+      "step": 12028
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.8191062973975802,
+      "learning_rate": 7.995028606680666e-08,
+      "loss": 0.4239,
+      "step": 12029
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 5.540892710039138,
+      "learning_rate": 7.972392098740234e-08,
+      "loss": 0.4686,
+      "step": 12030
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.613203055076282,
+      "learning_rate": 7.949787424342326e-08,
+      "loss": 0.4596,
+      "step": 12031
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5382350056105901,
+      "learning_rate": 7.927214584949549e-08,
+      "loss": 0.4601,
+      "step": 12032
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5409039780461194,
+      "learning_rate": 7.904673582022182e-08,
+      "loss": 0.4626,
+      "step": 12033
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8419304174987445,
+      "learning_rate": 7.882164417018779e-08,
+      "loss": 0.4346,
+      "step": 12034
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.046381171616162,
+      "learning_rate": 7.859687091395507e-08,
+      "loss": 0.4864,
+      "step": 12035
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7362520842688336,
+      "learning_rate": 7.837241606606704e-08,
+      "loss": 0.5111,
+      "step": 12036
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9985533586270405,
+      "learning_rate": 7.814827964104544e-08,
+      "loss": 0.441,
+      "step": 12037
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4875221980412505,
+      "learning_rate": 7.79244616533914e-08,
+      "loss": 0.46,
+      "step": 12038
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9080840248966062,
+      "learning_rate": 7.770096211758616e-08,
+      "loss": 0.449,
+      "step": 12039
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.096578013172229,
+      "learning_rate": 7.747778104808922e-08,
+      "loss": 0.4699,
+      "step": 12040
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5348432113625232,
+      "learning_rate": 7.725491845934075e-08,
+      "loss": 0.4683,
+      "step": 12041
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.390122300207347,
+      "learning_rate": 7.703237436575806e-08,
+      "loss": 0.4389,
+      "step": 12042
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7933012669045372,
+      "learning_rate": 7.681014878174187e-08,
+      "loss": 0.4556,
+      "step": 12043
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5535382962125575,
+      "learning_rate": 7.658824172166846e-08,
+      "loss": 0.4462,
+      "step": 12044
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4828021592565828,
+      "learning_rate": 7.636665319989522e-08,
+      "loss": 0.4121,
+      "step": 12045
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6677653342727699,
+      "learning_rate": 7.614538323075848e-08,
+      "loss": 0.4434,
+      "step": 12046
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.8003114051797455,
+      "learning_rate": 7.592443182857401e-08,
+      "loss": 0.3884,
+      "step": 12047
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.032165760348553,
+      "learning_rate": 7.570379900763813e-08,
+      "loss": 0.4645,
+      "step": 12048
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5871150212523286,
+      "learning_rate": 7.548348478222389e-08,
+      "loss": 0.4439,
+      "step": 12049
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7929947459670017,
+      "learning_rate": 7.526348916658654e-08,
+      "loss": 0.4837,
+      "step": 12050
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3993118595340133,
+      "learning_rate": 7.50438121749586e-08,
+      "loss": 0.4875,
+      "step": 12051
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9796587292852197,
+      "learning_rate": 7.482445382155368e-08,
+      "loss": 0.4285,
+      "step": 12052
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4114600189597066,
+      "learning_rate": 7.460541412056376e-08,
+      "loss": 0.4873,
+      "step": 12053
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9714049196770034,
+      "learning_rate": 7.438669308616031e-08,
+      "loss": 0.4226,
+      "step": 12054
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6486355661427448,
+      "learning_rate": 7.416829073249366e-08,
+      "loss": 0.3992,
+      "step": 12055
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6314097763733606,
+      "learning_rate": 7.395020707369527e-08,
+      "loss": 0.4039,
+      "step": 12056
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 4.670382789402163,
+      "learning_rate": 7.373244212387499e-08,
+      "loss": 0.4523,
+      "step": 12057
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4325088291008001,
+      "learning_rate": 7.351499589712042e-08,
+      "loss": 0.4255,
+      "step": 12058
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6009504742716,
+      "learning_rate": 7.329786840750142e-08,
+      "loss": 0.4536,
+      "step": 12059
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.743612576128621,
+      "learning_rate": 7.308105966906508e-08,
+      "loss": 0.4557,
+      "step": 12060
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7005295279212966,
+      "learning_rate": 7.286456969583911e-08,
+      "loss": 0.4287,
+      "step": 12061
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.1464286316114465,
+      "learning_rate": 7.264839850183003e-08,
+      "loss": 0.4371,
+      "step": 12062
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5683967303142914,
+      "learning_rate": 7.24325461010239e-08,
+      "loss": 0.4288,
+      "step": 12063
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.882147594621402,
+      "learning_rate": 7.221701250738622e-08,
+      "loss": 0.4865,
+      "step": 12064
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5881805689863375,
+      "learning_rate": 7.200179773486083e-08,
+      "loss": 0.4395,
+      "step": 12065
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1472904598078686,
+      "learning_rate": 7.17869017973727e-08,
+      "loss": 0.4904,
+      "step": 12066
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9338792417498625,
+      "learning_rate": 7.157232470882513e-08,
+      "loss": 0.4578,
+      "step": 12067
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5450994969963001,
+      "learning_rate": 7.13580664831015e-08,
+      "loss": 0.4691,
+      "step": 12068
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5002254441130725,
+      "learning_rate": 7.114412713406294e-08,
+      "loss": 0.423,
+      "step": 12069
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.356320751954262,
+      "learning_rate": 7.093050667555168e-08,
+      "loss": 0.4727,
+      "step": 12070
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2678645827548665,
+      "learning_rate": 7.071720512138836e-08,
+      "loss": 0.4195,
+      "step": 12071
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.41061581455058,
+      "learning_rate": 7.050422248537414e-08,
+      "loss": 0.3947,
+      "step": 12072
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5057553222935928,
+      "learning_rate": 7.029155878128746e-08,
+      "loss": 0.4665,
+      "step": 12073
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.605696499396263,
+      "learning_rate": 7.007921402288842e-08,
+      "loss": 0.436,
+      "step": 12074
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8168632830913305,
+      "learning_rate": 6.986718822391491e-08,
+      "loss": 0.428,
+      "step": 12075
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4194114817596504,
+      "learning_rate": 6.965548139808542e-08,
+      "loss": 0.4175,
+      "step": 12076
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5924585939362528,
+      "learning_rate": 6.944409355909565e-08,
+      "loss": 0.4687,
+      "step": 12077
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5749752855860285,
+      "learning_rate": 6.923302472062354e-08,
+      "loss": 0.4824,
+      "step": 12078
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5438675571520081,
+      "learning_rate": 6.902227489632485e-08,
+      "loss": 0.4815,
+      "step": 12079
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.4976537281507856,
+      "learning_rate": 6.881184409983421e-08,
+      "loss": 0.4177,
+      "step": 12080
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8124249486981636,
+      "learning_rate": 6.86017323447663e-08,
+      "loss": 0.4494,
+      "step": 12081
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6843678741358121,
+      "learning_rate": 6.839193964471525e-08,
+      "loss": 0.3948,
+      "step": 12082
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.527285501728673,
+      "learning_rate": 6.818246601325463e-08,
+      "loss": 0.4232,
+      "step": 12083
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.419924378425735,
+      "learning_rate": 6.797331146393638e-08,
+      "loss": 0.3952,
+      "step": 12084
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5542329978532223,
+      "learning_rate": 6.776447601029357e-08,
+      "loss": 0.4124,
+      "step": 12085
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.381132581234206,
+      "learning_rate": 6.75559596658365e-08,
+      "loss": 0.4306,
+      "step": 12086
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6266062903633327,
+      "learning_rate": 6.734776244405661e-08,
+      "loss": 0.411,
+      "step": 12087
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9210250492394096,
+      "learning_rate": 6.713988435842367e-08,
+      "loss": 0.4704,
+      "step": 12088
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9287624211710335,
+      "learning_rate": 6.693232542238692e-08,
+      "loss": 0.4052,
+      "step": 12089
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9268528997362258,
+      "learning_rate": 6.672508564937618e-08,
+      "loss": 0.4309,
+      "step": 12090
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9476670251623442,
+      "learning_rate": 6.651816505279796e-08,
+      "loss": 0.4564,
+      "step": 12091
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7563075051366521,
+      "learning_rate": 6.631156364604153e-08,
+      "loss": 0.4855,
+      "step": 12092
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9679931000451556,
+      "learning_rate": 6.61052814424723e-08,
+      "loss": 0.4603,
+      "step": 12093
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5478583157897122,
+      "learning_rate": 6.58993184554374e-08,
+      "loss": 0.4738,
+      "step": 12094
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7105812459440384,
+      "learning_rate": 6.569367469826171e-08,
+      "loss": 0.472,
+      "step": 12095
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5291457567537177,
+      "learning_rate": 6.548835018425015e-08,
+      "loss": 0.4716,
+      "step": 12096
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1333666078378886,
+      "learning_rate": 6.528334492668764e-08,
+      "loss": 0.5043,
+      "step": 12097
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.883826270473864,
+      "learning_rate": 6.507865893883692e-08,
+      "loss": 0.4708,
+      "step": 12098
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5617926602888514,
+      "learning_rate": 6.487429223394182e-08,
+      "loss": 0.4621,
+      "step": 12099
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5487135292953904,
+      "learning_rate": 6.467024482522344e-08,
+      "loss": 0.4556,
+      "step": 12100
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.080735979067863,
+      "learning_rate": 6.4466516725884e-08,
+      "loss": 0.4251,
+      "step": 12101
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5550096338531417,
+      "learning_rate": 6.426310794910461e-08,
+      "loss": 0.4718,
+      "step": 12102
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3219470163828018,
+      "learning_rate": 6.406001850804588e-08,
+      "loss": 0.4469,
+      "step": 12103
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0619843531829734,
+      "learning_rate": 6.385724841584672e-08,
+      "loss": 0.4093,
+      "step": 12104
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5224796500989117,
+      "learning_rate": 6.36547976856261e-08,
+      "loss": 0.4842,
+      "step": 12105
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.42412485217852,
+      "learning_rate": 6.345266633048241e-08,
+      "loss": 0.4935,
+      "step": 12106
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5558021060304932,
+      "learning_rate": 6.325085436349354e-08,
+      "loss": 0.4267,
+      "step": 12107
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.229519954525349,
+      "learning_rate": 6.304936179771571e-08,
+      "loss": 0.4617,
+      "step": 12108
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.776425275163013,
+      "learning_rate": 6.28481886461868e-08,
+      "loss": 0.4778,
+      "step": 12109
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.0293452167969104,
+      "learning_rate": 6.264733492192088e-08,
+      "loss": 0.4219,
+      "step": 12110
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8168790932509682,
+      "learning_rate": 6.244680063791419e-08,
+      "loss": 0.3921,
+      "step": 12111
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.568818975014232,
+      "learning_rate": 6.224658580713971e-08,
+      "loss": 0.5224,
+      "step": 12112
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3882406379384997,
+      "learning_rate": 6.204669044255151e-08,
+      "loss": 0.4598,
+      "step": 12113
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5767086610090396,
+      "learning_rate": 6.18471145570837e-08,
+      "loss": 0.4422,
+      "step": 12114
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9130092999954507,
+      "learning_rate": 6.164785816364704e-08,
+      "loss": 0.4806,
+      "step": 12115
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5033923641148645,
+      "learning_rate": 6.144892127513402e-08,
+      "loss": 0.4244,
+      "step": 12116
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.46121190796068,
+      "learning_rate": 6.125030390441489e-08,
+      "loss": 0.448,
+      "step": 12117
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.79412814643016,
+      "learning_rate": 6.105200606434102e-08,
+      "loss": 0.4732,
+      "step": 12118
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5249202068488625,
+      "learning_rate": 6.085402776774107e-08,
+      "loss": 0.4655,
+      "step": 12119
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.580471592105387,
+      "learning_rate": 6.065636902742423e-08,
+      "loss": 0.5075,
+      "step": 12120
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.504267498488612,
+      "learning_rate": 6.045902985617913e-08,
+      "loss": 0.4597,
+      "step": 12121
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.452123702608791,
+      "learning_rate": 6.026201026677281e-08,
+      "loss": 0.4219,
+      "step": 12122
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9315333289952938,
+      "learning_rate": 6.006531027195283e-08,
+      "loss": 0.4432,
+      "step": 12123
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.016069461751853,
+      "learning_rate": 5.986892988444404e-08,
+      "loss": 0.4479,
+      "step": 12124
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2834112262567636,
+      "learning_rate": 5.967286911695403e-08,
+      "loss": 0.4056,
+      "step": 12125
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1591761197743757,
+      "learning_rate": 5.9477127982165984e-08,
+      "loss": 0.4754,
+      "step": 12126
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.9542634257540719,
+      "learning_rate": 5.9281706492745337e-08,
+      "loss": 0.4493,
+      "step": 12127
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4679496009583155,
+      "learning_rate": 5.9086604661334756e-08,
+      "loss": 0.3981,
+      "step": 12128
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5567170839503993,
+      "learning_rate": 5.8891822500557474e-08,
+      "loss": 0.4796,
+      "step": 12129
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5534469475314527,
+      "learning_rate": 5.869736002301507e-08,
+      "loss": 0.4428,
+      "step": 12130
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 165.1323188151165,
+      "learning_rate": 5.8503217241289714e-08,
+      "loss": 0.4528,
+      "step": 12131
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.1583905077703123,
+      "learning_rate": 5.83093941679419e-08,
+      "loss": 0.4755,
+      "step": 12132
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5168212715548587,
+      "learning_rate": 5.811589081551161e-08,
+      "loss": 0.4414,
+      "step": 12133
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5492419677075632,
+      "learning_rate": 5.7922707196518266e-08,
+      "loss": 0.4515,
+      "step": 12134
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3464442182342977,
+      "learning_rate": 5.772984332346132e-08,
+      "loss": 0.3602,
+      "step": 12135
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.628395236424482,
+      "learning_rate": 5.753729920881745e-08,
+      "loss": 0.4281,
+      "step": 12136
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.095843470734586,
+      "learning_rate": 5.7345074865045035e-08,
+      "loss": 0.4279,
+      "step": 12137
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6520764461172746,
+      "learning_rate": 5.715317030458078e-08,
+      "loss": 0.4681,
+      "step": 12138
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7276730164990128,
+      "learning_rate": 5.696158553984032e-08,
+      "loss": 0.4216,
+      "step": 12139
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.624025671607213,
+      "learning_rate": 5.677032058321874e-08,
+      "loss": 0.4268,
+      "step": 12140
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2167050482682114,
+      "learning_rate": 5.6579375447090576e-08,
+      "loss": 0.4572,
+      "step": 12141
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7217232579254378,
+      "learning_rate": 5.638875014381096e-08,
+      "loss": 0.4298,
+      "step": 12142
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8711197038817466,
+      "learning_rate": 5.6198444685711116e-08,
+      "loss": 0.429,
+      "step": 12143
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8744199184200039,
+      "learning_rate": 5.6008459085105104e-08,
+      "loss": 0.4431,
+      "step": 12144
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 3.3754029644782606,
+      "learning_rate": 5.58187933542842e-08,
+      "loss": 0.48,
+      "step": 12145
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.603520604639879,
+      "learning_rate": 5.562944750551913e-08,
+      "loss": 0.4893,
+      "step": 12146
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.6475419867874217,
+      "learning_rate": 5.5440421551061216e-08,
+      "loss": 0.396,
+      "step": 12147
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5232736559847088,
+      "learning_rate": 5.525171550313957e-08,
+      "loss": 0.4579,
+      "step": 12148
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5367792868273121,
+      "learning_rate": 5.506332937396386e-08,
+      "loss": 0.4752,
+      "step": 12149
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.666094191354845,
+      "learning_rate": 5.487526317572156e-08,
+      "loss": 0.4597,
+      "step": 12150
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8068782758132576,
+      "learning_rate": 5.4687516920580716e-08,
+      "loss": 0.4569,
+      "step": 12151
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5510633400273337,
+      "learning_rate": 5.450009062068773e-08,
+      "loss": 0.4568,
+      "step": 12152
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.3912687435455138,
+      "learning_rate": 5.431298428817011e-08,
+      "loss": 0.3873,
+      "step": 12153
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7530046700622814,
+      "learning_rate": 5.412619793513263e-08,
+      "loss": 0.4362,
+      "step": 12154
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5405423546670909,
+      "learning_rate": 5.3939731573659504e-08,
+      "loss": 0.4569,
+      "step": 12155
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.2996063725361346,
+      "learning_rate": 5.3753585215816083e-08,
+      "loss": 0.4322,
+      "step": 12156
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.7489389309917094,
+      "learning_rate": 5.356775887364496e-08,
+      "loss": 0.4511,
+      "step": 12157
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8659760704169672,
+      "learning_rate": 5.338225255916929e-08,
+      "loss": 0.4514,
+      "step": 12158
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 2.5218827856497974,
+      "learning_rate": 5.319706628439003e-08,
+      "loss": 0.4333,
+      "step": 12159
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8867297634222016,
+      "learning_rate": 5.3012200061290374e-08,
+      "loss": 0.4457,
+      "step": 12160
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4000551009416529,
+      "learning_rate": 5.2827653901829646e-08,
+      "loss": 0.3902,
+      "step": 12161
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6200545568492413,
+      "learning_rate": 5.264342781794829e-08,
+      "loss": 0.3842,
+      "step": 12162
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.328773792435486,
+      "learning_rate": 5.2459521821563994e-08,
+      "loss": 0.4506,
+      "step": 12163
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.433551590036942,
+      "learning_rate": 5.22759359245778e-08,
+      "loss": 0.4113,
+      "step": 12164
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.2131692410590946,
+      "learning_rate": 5.209267013886521e-08,
+      "loss": 0.5002,
+      "step": 12165
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.9190334825657205,
+      "learning_rate": 5.1909724476284506e-08,
+      "loss": 0.455,
+      "step": 12166
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.797200642284039,
+      "learning_rate": 5.172709894867178e-08,
+      "loss": 0.4511,
+      "step": 12167
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6986365940501422,
+      "learning_rate": 5.154479356784259e-08,
+      "loss": 0.4261,
+      "step": 12168
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.283868373239577,
+      "learning_rate": 5.136280834559193e-08,
+      "loss": 0.483,
+      "step": 12169
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2296802254204233,
+      "learning_rate": 5.118114329369372e-08,
+      "loss": 0.4593,
+      "step": 12170
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6704065873590495,
+      "learning_rate": 5.09997984239019e-08,
+      "loss": 0.3777,
+      "step": 12171
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.613054292023631,
+      "learning_rate": 5.0818773747948744e-08,
+      "loss": 0.4532,
+      "step": 12172
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6671710088338947,
+      "learning_rate": 5.0638069277547106e-08,
+      "loss": 0.4454,
+      "step": 12173
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7875288782769443,
+      "learning_rate": 5.045768502438708e-08,
+      "loss": 0.5207,
+      "step": 12174
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1578022203620972,
+      "learning_rate": 5.027762100014044e-08,
+      "loss": 0.4654,
+      "step": 12175
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2157935910493225,
+      "learning_rate": 5.0097877216456755e-08,
+      "loss": 0.461,
+      "step": 12176
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7125320311153123,
+      "learning_rate": 4.9918453684965064e-08,
+      "loss": 0.4692,
+      "step": 12177
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7097222247374564,
+      "learning_rate": 4.9739350417273846e-08,
+      "loss": 0.4589,
+      "step": 12178
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.115027578580721,
+      "learning_rate": 4.956056742497106e-08,
+      "loss": 0.4324,
+      "step": 12179
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9496860954506348,
+      "learning_rate": 4.938210471962301e-08,
+      "loss": 0.4319,
+      "step": 12180
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6995035378925303,
+      "learning_rate": 4.920396231277713e-08,
+      "loss": 0.4313,
+      "step": 12181
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.791403557214716,
+      "learning_rate": 4.902614021595809e-08,
+      "loss": 0.4617,
+      "step": 12182
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8926969624183836,
+      "learning_rate": 4.8848638440671114e-08,
+      "loss": 0.4601,
+      "step": 12183
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7642352819950302,
+      "learning_rate": 4.8671456998400924e-08,
+      "loss": 0.4473,
+      "step": 12184
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.159602601577734,
+      "learning_rate": 4.8494595900609454e-08,
+      "loss": 0.4807,
+      "step": 12185
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.128217556712262,
+      "learning_rate": 4.831805515874089e-08,
+      "loss": 0.4223,
+      "step": 12186
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4839653266225885,
+      "learning_rate": 4.814183478421608e-08,
+      "loss": 0.4569,
+      "step": 12187
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6367361874834316,
+      "learning_rate": 4.796593478843703e-08,
+      "loss": 0.4188,
+      "step": 12188
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 6.874695916743228,
+      "learning_rate": 4.779035518278352e-08,
+      "loss": 0.4552,
+      "step": 12189
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.878999225241796,
+      "learning_rate": 4.7615095978616465e-08,
+      "loss": 0.4184,
+      "step": 12190
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8945250703808998,
+      "learning_rate": 4.744015718727346e-08,
+      "loss": 0.3877,
+      "step": 12191
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4281495448627544,
+      "learning_rate": 4.726553882007379e-08,
+      "loss": 0.3986,
+      "step": 12192
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8338645632583692,
+      "learning_rate": 4.709124088831507e-08,
+      "loss": 0.4245,
+      "step": 12193
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.9375192621050794,
+      "learning_rate": 4.691726340327274e-08,
+      "loss": 0.4672,
+      "step": 12194
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.559265579246486,
+      "learning_rate": 4.674360637620501e-08,
+      "loss": 0.4865,
+      "step": 12195
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9501957853212961,
+      "learning_rate": 4.657026981834623e-08,
+      "loss": 0.4692,
+      "step": 12196
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7653227726396006,
+      "learning_rate": 4.6397253740911306e-08,
+      "loss": 0.4655,
+      "step": 12197
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1589931507658484,
+      "learning_rate": 4.622455815509297e-08,
+      "loss": 0.4443,
+      "step": 12198
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7887900082419774,
+      "learning_rate": 4.605218307206616e-08,
+      "loss": 0.4196,
+      "step": 12199
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5796258830006363,
+      "learning_rate": 4.588012850298251e-08,
+      "loss": 0.468,
+      "step": 12200
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.020843040863877,
+      "learning_rate": 4.5708394458973684e-08,
+      "loss": 0.4111,
+      "step": 12201
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.191250185470242,
+      "learning_rate": 4.5536980951150224e-08,
+      "loss": 0.4359,
+      "step": 12202
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2164140324103676,
+      "learning_rate": 4.536588799060326e-08,
+      "loss": 0.4356,
+      "step": 12203
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6856964277341124,
+      "learning_rate": 4.5195115588402285e-08,
+      "loss": 0.4402,
+      "step": 12204
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6158079158037137,
+      "learning_rate": 4.502466375559511e-08,
+      "loss": 0.4386,
+      "step": 12205
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8127577270789712,
+      "learning_rate": 4.485453250321015e-08,
+      "loss": 0.4428,
+      "step": 12206
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.508409960259944,
+      "learning_rate": 4.4684721842254695e-08,
+      "loss": 0.4329,
+      "step": 12207
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5605080524883762,
+      "learning_rate": 4.4515231783716086e-08,
+      "loss": 0.4657,
+      "step": 12208
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6825498198534399,
+      "learning_rate": 4.434606233855887e-08,
+      "loss": 0.4074,
+      "step": 12209
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5272448982265743,
+      "learning_rate": 4.417721351772874e-08,
+      "loss": 0.4541,
+      "step": 12210
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4336231530695474,
+      "learning_rate": 4.4008685332149745e-08,
+      "loss": 0.4498,
+      "step": 12211
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.96325752225603,
+      "learning_rate": 4.384047779272538e-08,
+      "loss": 0.4657,
+      "step": 12212
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7679051193304456,
+      "learning_rate": 4.367259091033915e-08,
+      "loss": 0.4016,
+      "step": 12213
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1357838687223287,
+      "learning_rate": 4.3505024695851826e-08,
+      "loss": 0.4352,
+      "step": 12214
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1831472218454198,
+      "learning_rate": 4.33377791601064e-08,
+      "loss": 0.3924,
+      "step": 12215
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.585771024900894,
+      "learning_rate": 4.317085431392198e-08,
+      "loss": 0.4321,
+      "step": 12216
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6383268040798091,
+      "learning_rate": 4.3004250168098836e-08,
+      "loss": 0.4485,
+      "step": 12217
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8807358069181077,
+      "learning_rate": 4.283796673341667e-08,
+      "loss": 0.4061,
+      "step": 12218
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.582170223839725,
+      "learning_rate": 4.267200402063298e-08,
+      "loss": 0.4569,
+      "step": 12219
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.543225939011292,
+      "learning_rate": 4.250636204048586e-08,
+      "loss": 0.4511,
+      "step": 12220
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7590982331009064,
+      "learning_rate": 4.2341040803691166e-08,
+      "loss": 0.4552,
+      "step": 12221
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9528333234101705,
+      "learning_rate": 4.217604032094647e-08,
+      "loss": 0.4692,
+      "step": 12222
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.125541703692806,
+      "learning_rate": 4.2011360602925987e-08,
+      "loss": 0.4275,
+      "step": 12223
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3007772555223593,
+      "learning_rate": 4.184700166028455e-08,
+      "loss": 0.4378,
+      "step": 12224
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5897748730079242,
+      "learning_rate": 4.168296350365641e-08,
+      "loss": 0.4518,
+      "step": 12225
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4700859364213206,
+      "learning_rate": 4.151924614365366e-08,
+      "loss": 0.3961,
+      "step": 12226
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6105085606298144,
+      "learning_rate": 4.135584959086947e-08,
+      "loss": 0.4616,
+      "step": 12227
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7761485445290064,
+      "learning_rate": 4.119277385587539e-08,
+      "loss": 0.4722,
+      "step": 12228
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8239033176559623,
+      "learning_rate": 4.1030018949221315e-08,
+      "loss": 0.4868,
+      "step": 12229
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9100291850567426,
+      "learning_rate": 4.086758488143827e-08,
+      "loss": 0.4459,
+      "step": 12230
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.606513760791807,
+      "learning_rate": 4.0705471663035065e-08,
+      "loss": 0.4496,
+      "step": 12231
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6322041869369035,
+      "learning_rate": 4.054367930450054e-08,
+      "loss": 0.3929,
+      "step": 12232
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.921367039598112,
+      "learning_rate": 4.038220781630131e-08,
+      "loss": 0.4198,
+      "step": 12233
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6654984627644491,
+      "learning_rate": 4.022105720888625e-08,
+      "loss": 0.4631,
+      "step": 12234
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9114406106986481,
+      "learning_rate": 4.0060227492680346e-08,
+      "loss": 0.4321,
+      "step": 12235
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9159482544959077,
+      "learning_rate": 3.9899718678088615e-08,
+      "loss": 0.4723,
+      "step": 12236
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.6213108094001614,
+      "learning_rate": 3.973953077549719e-08,
+      "loss": 0.4484,
+      "step": 12237
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5540393688323295,
+      "learning_rate": 3.9579663795268894e-08,
+      "loss": 0.4497,
+      "step": 12238
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7377508136102546,
+      "learning_rate": 3.9420117747747675e-08,
+      "loss": 0.4299,
+      "step": 12239
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5321358645497081,
+      "learning_rate": 3.926089264325528e-08,
+      "loss": 0.4642,
+      "step": 12240
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.102315152921353,
+      "learning_rate": 3.9101988492093457e-08,
+      "loss": 0.4868,
+      "step": 12241
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7144132487324786,
+      "learning_rate": 3.894340530454399e-08,
+      "loss": 0.3879,
+      "step": 12242
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8071250167394868,
+      "learning_rate": 3.87851430908659e-08,
+      "loss": 0.4078,
+      "step": 12243
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.8470017989072924,
+      "learning_rate": 3.8627201861298756e-08,
+      "loss": 0.4481,
+      "step": 12244
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.978625695595467,
+      "learning_rate": 3.846958162606163e-08,
+      "loss": 0.4615,
+      "step": 12245
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.7093817018946185,
+      "learning_rate": 3.8312282395351897e-08,
+      "loss": 0.4345,
+      "step": 12246
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.0909561078046197,
+      "learning_rate": 3.815530417934754e-08,
+      "loss": 0.4328,
+      "step": 12247
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.491848422080435,
+      "learning_rate": 3.799864698820321e-08,
+      "loss": 0.4599,
+      "step": 12248
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.258644642161358,
+      "learning_rate": 3.784231083205525e-08,
+      "loss": 0.4154,
+      "step": 12249
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.2876207370542523,
+      "learning_rate": 3.76862957210189e-08,
+      "loss": 0.3973,
+      "step": 12250
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7184638176241906,
+      "learning_rate": 3.753060166518774e-08,
+      "loss": 0.4453,
+      "step": 12251
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.77533974378186,
+      "learning_rate": 3.737522867463483e-08,
+      "loss": 0.485,
+      "step": 12252
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5971882777627019,
+      "learning_rate": 3.72201767594127e-08,
+      "loss": 0.4669,
+      "step": 12253
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5874999510333814,
+      "learning_rate": 3.7065445929552744e-08,
+      "loss": 0.4568,
+      "step": 12254
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.8828794454335473,
+      "learning_rate": 3.691103619506642e-08,
+      "loss": 0.4453,
+      "step": 12255
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5288385871228325,
+      "learning_rate": 3.6756947565943504e-08,
+      "loss": 0.4698,
+      "step": 12256
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7319785221220998,
+      "learning_rate": 3.660318005215324e-08,
+      "loss": 0.4987,
+      "step": 12257
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.5563629644788217,
+      "learning_rate": 3.644973366364435e-08,
+      "loss": 0.4029,
+      "step": 12258
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4510350742571931,
+      "learning_rate": 3.6296608410344434e-08,
+      "loss": 0.4419,
+      "step": 12259
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7345225042378463,
+      "learning_rate": 3.614380430216058e-08,
+      "loss": 0.4484,
+      "step": 12260
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6197498834075845,
+      "learning_rate": 3.5991321348979316e-08,
+      "loss": 0.4375,
+      "step": 12261
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7244885332334154,
+      "learning_rate": 3.583915956066553e-08,
+      "loss": 0.4308,
+      "step": 12262
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.1979899734912385,
+      "learning_rate": 3.5687318947064676e-08,
+      "loss": 0.4077,
+      "step": 12263
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4689494533462557,
+      "learning_rate": 3.553579951799946e-08,
+      "loss": 0.4616,
+      "step": 12264
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7497343566923738,
+      "learning_rate": 3.53846012832737e-08,
+      "loss": 0.4211,
+      "step": 12265
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5627986779930028,
+      "learning_rate": 3.5233724252670134e-08,
+      "loss": 0.4668,
+      "step": 12266
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.544455862008603,
+      "learning_rate": 3.508316843594983e-08,
+      "loss": 0.4734,
+      "step": 12267
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7824636533899578,
+      "learning_rate": 3.4932933842853344e-08,
+      "loss": 0.4768,
+      "step": 12268
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5483257313524167,
+      "learning_rate": 3.4783020483101226e-08,
+      "loss": 0.4608,
+      "step": 12269
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.619005732447942,
+      "learning_rate": 3.463342836639239e-08,
+      "loss": 0.4387,
+      "step": 12270
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.175539456416044,
+      "learning_rate": 3.4484157502404636e-08,
+      "loss": 0.4443,
+      "step": 12271
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4769824548680204,
+      "learning_rate": 3.433520790079692e-08,
+      "loss": 0.4229,
+      "step": 12272
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.4195414604596017,
+      "learning_rate": 3.4186579571204856e-08,
+      "loss": 0.399,
+      "step": 12273
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9548132739694004,
+      "learning_rate": 3.403827252324521e-08,
+      "loss": 0.4315,
+      "step": 12274
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 5.277843897588021,
+      "learning_rate": 3.389028676651307e-08,
+      "loss": 0.4425,
+      "step": 12275
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4213875533276281,
+      "learning_rate": 3.374262231058245e-08,
+      "loss": 0.445,
+      "step": 12276
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 3.4152922269143065,
+      "learning_rate": 3.359527916500793e-08,
+      "loss": 0.4412,
+      "step": 12277
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7327429186035037,
+      "learning_rate": 3.3448257339322446e-08,
+      "loss": 0.4666,
+      "step": 12278
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.5239804060265414,
+      "learning_rate": 3.3301556843036705e-08,
+      "loss": 0.4433,
+      "step": 12279
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6389073043257443,
+      "learning_rate": 3.3155177685643136e-08,
+      "loss": 0.5153,
+      "step": 12280
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.9089056803210736,
+      "learning_rate": 3.3009119876612485e-08,
+      "loss": 0.4894,
+      "step": 12281
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7917171761536899,
+      "learning_rate": 3.286338342539386e-08,
+      "loss": 0.4897,
+      "step": 12282
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.750330503380991,
+      "learning_rate": 3.271796834141694e-08,
+      "loss": 0.5017,
+      "step": 12283
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.485750452860168,
+      "learning_rate": 3.257287463408865e-08,
+      "loss": 0.4362,
+      "step": 12284
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.71813865742657,
+      "learning_rate": 3.242810231279814e-08,
+      "loss": 0.444,
+      "step": 12285
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.508141695414921,
+      "learning_rate": 3.228365138691014e-08,
+      "loss": 0.4113,
+      "step": 12286
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.902796484319638,
+      "learning_rate": 3.213952186577163e-08,
+      "loss": 0.4562,
+      "step": 12287
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.550885577046666,
+      "learning_rate": 3.199571375870736e-08,
+      "loss": 0.4316,
+      "step": 12288
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5559343174319692,
+      "learning_rate": 3.185222707502156e-08,
+      "loss": 0.4691,
+      "step": 12289
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.201786523902043,
+      "learning_rate": 3.1709061823997356e-08,
+      "loss": 0.4446,
+      "step": 12290
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7683053461597373,
+      "learning_rate": 3.156621801489734e-08,
+      "loss": 0.3689,
+      "step": 12291
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2700657734605088,
+      "learning_rate": 3.1423695656964123e-08,
+      "loss": 0.4509,
+      "step": 12292
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.000739369004508,
+      "learning_rate": 3.128149475941755e-08,
+      "loss": 0.4454,
+      "step": 12293
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.793179808234318,
+      "learning_rate": 3.1139615331458596e-08,
+      "loss": 0.4245,
+      "step": 12294
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.038301359423851,
+      "learning_rate": 3.099805738226602e-08,
+      "loss": 0.4352,
+      "step": 12295
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5268989198223204,
+      "learning_rate": 3.085682092099918e-08,
+      "loss": 0.4585,
+      "step": 12296
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0993853307928116,
+      "learning_rate": 3.071590595679519e-08,
+      "loss": 0.4426,
+      "step": 12297
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0500907267790573,
+      "learning_rate": 3.057531249877233e-08,
+      "loss": 0.4231,
+      "step": 12298
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.787674634499823,
+      "learning_rate": 3.043504055602498e-08,
+      "loss": 0.453,
+      "step": 12299
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5125314362336555,
+      "learning_rate": 3.0295090137630324e-08,
+      "loss": 0.4401,
+      "step": 12300
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.743226770177226,
+      "learning_rate": 3.015546125264113e-08,
+      "loss": 0.387,
+      "step": 12301
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.832886323963274,
+      "learning_rate": 3.001615391009349e-08,
+      "loss": 0.4518,
+      "step": 12302
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8439139939251135,
+      "learning_rate": 2.987716811899799e-08,
+      "loss": 0.4536,
+      "step": 12303
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7431015270352062,
+      "learning_rate": 2.9738503888348534e-08,
+      "loss": 0.4455,
+      "step": 12304
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5146613005705547,
+      "learning_rate": 2.960016122711573e-08,
+      "loss": 0.3875,
+      "step": 12305
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5169523606476568,
+      "learning_rate": 2.946214014425075e-08,
+      "loss": 0.436,
+      "step": 12306
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.043498183237889,
+      "learning_rate": 2.932444064868256e-08,
+      "loss": 0.4254,
+      "step": 12307
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.592863184890767,
+      "learning_rate": 2.9187062749320704e-08,
+      "loss": 0.4328,
+      "step": 12308
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6593745597861853,
+      "learning_rate": 2.9050006455053625e-08,
+      "loss": 0.4651,
+      "step": 12309
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8018584495223355,
+      "learning_rate": 2.891327177474812e-08,
+      "loss": 0.4699,
+      "step": 12310
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9301548187060964,
+      "learning_rate": 2.8776858717250445e-08,
+      "loss": 0.4405,
+      "step": 12311
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5258197416621908,
+      "learning_rate": 2.8640767291387427e-08,
+      "loss": 0.4509,
+      "step": 12312
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.1931659833453594,
+      "learning_rate": 2.8504997505963695e-08,
+      "loss": 0.4655,
+      "step": 12313
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6087409343678611,
+      "learning_rate": 2.8369549369762218e-08,
+      "loss": 0.4666,
+      "step": 12314
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5624465476501812,
+      "learning_rate": 2.8234422891547654e-08,
+      "loss": 0.4252,
+      "step": 12315
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4663693670487221,
+      "learning_rate": 2.8099618080061896e-08,
+      "loss": 0.4408,
+      "step": 12316
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 19.06635210944333,
+      "learning_rate": 2.796513494402686e-08,
+      "loss": 0.4401,
+      "step": 12317
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 8.885232063756083,
+      "learning_rate": 2.7830973492143365e-08,
+      "loss": 0.4725,
+      "step": 12318
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.111824727592062,
+      "learning_rate": 2.7697133733091686e-08,
+      "loss": 0.4869,
+      "step": 12319
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7981570083496166,
+      "learning_rate": 2.7563615675530452e-08,
+      "loss": 0.3751,
+      "step": 12320
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4905198613775308,
+      "learning_rate": 2.7430419328098867e-08,
+      "loss": 0.3908,
+      "step": 12321
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.872076349735022,
+      "learning_rate": 2.7297544699413925e-08,
+      "loss": 0.3672,
+      "step": 12322
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 5.421741257911467,
+      "learning_rate": 2.716499179807264e-08,
+      "loss": 0.4444,
+      "step": 12323
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.624039835585337,
+      "learning_rate": 2.7032760632651477e-08,
+      "loss": 0.4546,
+      "step": 12324
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.13699220968566,
+      "learning_rate": 2.690085121170527e-08,
+      "loss": 0.426,
+      "step": 12325
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.371519041917397,
+      "learning_rate": 2.67692635437683e-08,
+      "loss": 0.3974,
+      "step": 12326
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.7528633945651872,
+      "learning_rate": 2.6637997637353752e-08,
+      "loss": 0.4801,
+      "step": 12327
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7920116230840426,
+      "learning_rate": 2.6507053500955393e-08,
+      "loss": 0.4531,
+      "step": 12328
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8057573905124469,
+      "learning_rate": 2.6376431143044223e-08,
+      "loss": 0.4239,
+      "step": 12329
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.1368041539187184,
+      "learning_rate": 2.6246130572071816e-08,
+      "loss": 0.4233,
+      "step": 12330
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5048181139696015,
+      "learning_rate": 2.6116151796467538e-08,
+      "loss": 0.4355,
+      "step": 12331
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6162267085902138,
+      "learning_rate": 2.598649482464244e-08,
+      "loss": 0.4565,
+      "step": 12332
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8058144886929828,
+      "learning_rate": 2.5857159664984254e-08,
+      "loss": 0.4149,
+      "step": 12333
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5896839669285732,
+      "learning_rate": 2.5728146325860183e-08,
+      "loss": 0.4573,
+      "step": 12334
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.627759581529985,
+      "learning_rate": 2.5599454815618542e-08,
+      "loss": 0.3881,
+      "step": 12335
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8290772097067245,
+      "learning_rate": 2.547108514258434e-08,
+      "loss": 0.4765,
+      "step": 12336
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5478055014297066,
+      "learning_rate": 2.534303731506371e-08,
+      "loss": 0.474,
+      "step": 12337
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.670776996613076,
+      "learning_rate": 2.5215311341340587e-08,
+      "loss": 0.5173,
+      "step": 12338
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.5531225872303787,
+      "learning_rate": 2.5087907229679465e-08,
+      "loss": 0.4399,
+      "step": 12339
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.125449561695311,
+      "learning_rate": 2.496082498832264e-08,
+      "loss": 0.4394,
+      "step": 12340
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.18680603653225,
+      "learning_rate": 2.4834064625492426e-08,
+      "loss": 0.4753,
+      "step": 12341
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9005813533696778,
+      "learning_rate": 2.4707626149389486e-08,
+      "loss": 0.4632,
+      "step": 12342
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8745101854580857,
+      "learning_rate": 2.4581509568194496e-08,
+      "loss": 0.4664,
+      "step": 12343
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.2581373497947888,
+      "learning_rate": 2.4455714890067593e-08,
+      "loss": 0.4004,
+      "step": 12344
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9772467133604987,
+      "learning_rate": 2.4330242123147273e-08,
+      "loss": 0.4486,
+      "step": 12345
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8955955870327175,
+      "learning_rate": 2.4205091275550928e-08,
+      "loss": 0.4303,
+      "step": 12346
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8156317972657583,
+      "learning_rate": 2.408026235537597e-08,
+      "loss": 0.4491,
+      "step": 12347
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5915565573319441,
+      "learning_rate": 2.395575537069872e-08,
+      "loss": 0.4276,
+      "step": 12348
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.545999848699551,
+      "learning_rate": 2.3831570329574948e-08,
+      "loss": 0.4236,
+      "step": 12349
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5578919054368918,
+      "learning_rate": 2.3707707240038237e-08,
+      "loss": 0.4018,
+      "step": 12350
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0503366308644626,
+      "learning_rate": 2.3584166110103835e-08,
+      "loss": 0.4351,
+      "step": 12351
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0182241847117792,
+      "learning_rate": 2.3460946947763684e-08,
+      "loss": 0.4586,
+      "step": 12352
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.723941289247087,
+      "learning_rate": 2.3338049760989745e-08,
+      "loss": 0.4359,
+      "step": 12353
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.231863281440146,
+      "learning_rate": 2.3215474557733985e-08,
+      "loss": 0.4563,
+      "step": 12354
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.476803359272614,
+      "learning_rate": 2.3093221345926732e-08,
+      "loss": 0.4407,
+      "step": 12355
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3673934526596825,
+      "learning_rate": 2.297129013347721e-08,
+      "loss": 0.4246,
+      "step": 12356
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.6722621444142756,
+      "learning_rate": 2.2849680928274663e-08,
+      "loss": 0.4734,
+      "step": 12357
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9019907675007741,
+      "learning_rate": 2.2728393738186695e-08,
+      "loss": 0.4344,
+      "step": 12358
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6298637796191917,
+      "learning_rate": 2.260742857106035e-08,
+      "loss": 0.4627,
+      "step": 12359
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8226549735745319,
+      "learning_rate": 2.2486785434722712e-08,
+      "loss": 0.4339,
+      "step": 12360
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.296583873858308,
+      "learning_rate": 2.2366464336978088e-08,
+      "loss": 0.4858,
+      "step": 12361
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.453059205472766,
+      "learning_rate": 2.224646528561192e-08,
+      "loss": 0.4317,
+      "step": 12362
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9994295560013642,
+      "learning_rate": 2.2126788288387434e-08,
+      "loss": 0.3928,
+      "step": 12363
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5494188192752956,
+      "learning_rate": 2.2007433353048447e-08,
+      "loss": 0.4853,
+      "step": 12364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 3.112445706361296,
+      "learning_rate": 2.1888400487315997e-08,
+      "loss": 0.3999,
+      "step": 12365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5452430482781085,
+      "learning_rate": 2.1769689698891705e-08,
+      "loss": 0.5108,
+      "step": 12366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8492918119962793,
+      "learning_rate": 2.1651300995456647e-08,
+      "loss": 0.4506,
+      "step": 12367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4972724682093894,
+      "learning_rate": 2.15332343846697e-08,
+      "loss": 0.432,
+      "step": 12368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.8818024684279302,
+      "learning_rate": 2.141548987416975e-08,
+      "loss": 0.4321,
+      "step": 12369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9899017050695995,
+      "learning_rate": 2.1298067471575146e-08,
+      "loss": 0.4494,
+      "step": 12370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7606900939875754,
+      "learning_rate": 2.1180967184482038e-08,
+      "loss": 0.4536,
+      "step": 12371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.968096899268923,
+      "learning_rate": 2.1064189020467694e-08,
+      "loss": 0.4077,
+      "step": 12372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5584148411790155,
+      "learning_rate": 2.0947732987086633e-08,
+      "loss": 0.4563,
+      "step": 12373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.4329334558067295,
+      "learning_rate": 2.0831599091873934e-08,
+      "loss": 0.42,
+      "step": 12374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6261095729089556,
+      "learning_rate": 2.0715787342343586e-08,
+      "loss": 0.453,
+      "step": 12375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.917747485383327,
+      "learning_rate": 2.060029774598793e-08,
+      "loss": 0.453,
+      "step": 12376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5866844751993328,
+      "learning_rate": 2.0485130310279322e-08,
+      "loss": 0.3979,
+      "step": 12377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.862112901942078,
+      "learning_rate": 2.037028504266847e-08,
+      "loss": 0.4562,
+      "step": 12378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5520047315398092,
+      "learning_rate": 2.0255761950586096e-08,
+      "loss": 0.4828,
+      "step": 12379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8075645153806543,
+      "learning_rate": 2.014156104144127e-08,
+      "loss": 0.4607,
+      "step": 12380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4334532620522746,
+      "learning_rate": 2.0027682322623087e-08,
+      "loss": 0.4378,
+      "step": 12381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5306064228313454,
+      "learning_rate": 1.991412580149954e-08,
+      "loss": 0.4261,
+      "step": 12382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9284743635933819,
+      "learning_rate": 1.9800891485416974e-08,
+      "loss": 0.4055,
+      "step": 12383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5043631965454617,
+      "learning_rate": 1.9687979381702306e-08,
+      "loss": 0.4403,
+      "step": 12384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5375569268696316,
+      "learning_rate": 1.9575389497659693e-08,
+      "loss": 0.468,
+      "step": 12385
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.08727103091835,
+      "learning_rate": 1.9463121840574416e-08,
+      "loss": 0.4142,
+      "step": 12386
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6574493790025686,
+      "learning_rate": 1.9351176417710115e-08,
+      "loss": 0.4573,
+      "step": 12387
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6610549354469388,
+      "learning_rate": 1.923955323630877e-08,
+      "loss": 0.4364,
+      "step": 12388
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6561434018668182,
+      "learning_rate": 1.9128252303592942e-08,
+      "loss": 0.3962,
+      "step": 12389
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.74990833792763,
+      "learning_rate": 1.9017273626762977e-08,
+      "loss": 0.4275,
+      "step": 12390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.0726889662546104,
+      "learning_rate": 1.89066172129998e-08,
+      "loss": 0.4517,
+      "step": 12391
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5685541489231947,
+      "learning_rate": 1.8796283069462683e-08,
+      "loss": 0.4848,
+      "step": 12392
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6556358915506932,
+      "learning_rate": 1.8686271203289254e-08,
+      "loss": 0.4567,
+      "step": 12393
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9056820492928797,
+      "learning_rate": 1.85765816215977e-08,
+      "loss": 0.4905,
+      "step": 12394
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.9497018895096259,
+      "learning_rate": 1.8467214331485128e-08,
+      "loss": 0.4273,
+      "step": 12395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7436125803114355,
+      "learning_rate": 1.835816934002699e-08,
+      "loss": 0.4769,
+      "step": 12396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6042549482850064,
+      "learning_rate": 1.8249446654278745e-08,
+      "loss": 0.3671,
+      "step": 12397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5176459774418807,
+      "learning_rate": 1.8141046281273667e-08,
+      "loss": 0.4364,
+      "step": 12398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4346893079265186,
+      "learning_rate": 1.8032968228026692e-08,
+      "loss": 0.4627,
+      "step": 12399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5640068258685682,
+      "learning_rate": 1.7925212501528898e-08,
+      "loss": 0.4771,
+      "step": 12400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5889853626905102,
+      "learning_rate": 1.7817779108752486e-08,
+      "loss": 0.4757,
+      "step": 12401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 4.709577586347337,
+      "learning_rate": 1.771066805664856e-08,
+      "loss": 0.4608,
+      "step": 12402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.95364414895742,
+      "learning_rate": 1.760387935214658e-08,
+      "loss": 0.4285,
+      "step": 12403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6627574704975139,
+      "learning_rate": 1.749741300215546e-08,
+      "loss": 0.4561,
+      "step": 12404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3749276363110394,
+      "learning_rate": 1.7391269013564137e-08,
+      "loss": 0.4683,
+      "step": 12405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.897443922160629,
+      "learning_rate": 1.7285447393239896e-08,
+      "loss": 0.4692,
+      "step": 12406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.8067648012740145,
+      "learning_rate": 1.717994814802837e-08,
+      "loss": 0.4647,
+      "step": 12407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3241150561047257,
+      "learning_rate": 1.707477128475632e-08,
+      "loss": 0.4062,
+      "step": 12408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5614959190544031,
+      "learning_rate": 1.696991681022775e-08,
+      "loss": 0.4802,
+      "step": 12409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7183736018550144,
+      "learning_rate": 1.686538473122723e-08,
+      "loss": 0.4653,
+      "step": 12410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.513748094700749,
+      "learning_rate": 1.6761175054517687e-08,
+      "loss": 0.356,
+      "step": 12411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5766378957803453,
+      "learning_rate": 1.6657287786840947e-08,
+      "loss": 0.4472,
+      "step": 12412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.001976903061876,
+      "learning_rate": 1.655372293491886e-08,
+      "loss": 0.4559,
+      "step": 12413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6760399318389718,
+      "learning_rate": 1.6450480505451614e-08,
+      "loss": 0.4358,
+      "step": 12414
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8678288470260627,
+      "learning_rate": 1.6347560505118877e-08,
+      "loss": 0.414,
+      "step": 12415
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7594823861541244,
+      "learning_rate": 1.6244962940579754e-08,
+      "loss": 0.4667,
+      "step": 12416
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6743331858330381,
+      "learning_rate": 1.6142687818471725e-08,
+      "loss": 0.4591,
+      "step": 12417
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5362704092320819,
+      "learning_rate": 1.6040735145411714e-08,
+      "loss": 0.4582,
+      "step": 12418
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1953725585209924,
+      "learning_rate": 1.5939104927997218e-08,
+      "loss": 0.4699,
+      "step": 12419
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7211307993693612,
+      "learning_rate": 1.5837797172801872e-08,
+      "loss": 0.4328,
+      "step": 12420
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5177193675839087,
+      "learning_rate": 1.5736811886381542e-08,
+      "loss": 0.4921,
+      "step": 12421
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6614835961182148,
+      "learning_rate": 1.563614907526878e-08,
+      "loss": 0.4426,
+      "step": 12422
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2722317561060597,
+      "learning_rate": 1.5535808745977264e-08,
+      "loss": 0.5116,
+      "step": 12423
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7264921149559496,
+      "learning_rate": 1.5435790904997916e-08,
+      "loss": 0.4587,
+      "step": 12424
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5321553756268829,
+      "learning_rate": 1.5336095558802776e-08,
+      "loss": 0.4614,
+      "step": 12425
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6128008886931653,
+      "learning_rate": 1.5236722713841136e-08,
+      "loss": 0.4895,
+      "step": 12426
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2370866148877075,
+      "learning_rate": 1.5137672376542843e-08,
+      "loss": 0.4655,
+      "step": 12427
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.647920554101873,
+      "learning_rate": 1.5038944553316115e-08,
+      "loss": 0.4008,
+      "step": 12428
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5034680432661958,
+      "learning_rate": 1.4940539250548058e-08,
+      "loss": 0.4441,
+      "step": 12429
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.362784014842779,
+      "learning_rate": 1.4842456474606359e-08,
+      "loss": 0.4948,
+      "step": 12430
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1619123305683585,
+      "learning_rate": 1.4744696231836497e-08,
+      "loss": 0.4318,
+      "step": 12431
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8835790837531212,
+      "learning_rate": 1.4647258528562858e-08,
+      "loss": 0.4293,
+      "step": 12432
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5230570462943147,
+      "learning_rate": 1.4550143371089841e-08,
+      "loss": 0.4644,
+      "step": 12433
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1458175927841308,
+      "learning_rate": 1.4453350765700757e-08,
+      "loss": 0.4646,
+      "step": 12434
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2452769847107312,
+      "learning_rate": 1.4356880718658373e-08,
+      "loss": 0.4045,
+      "step": 12435
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.684451015977942,
+      "learning_rate": 1.426073323620325e-08,
+      "loss": 0.5126,
+      "step": 12436
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5564080886814182,
+      "learning_rate": 1.416490832455708e-08,
+      "loss": 0.4564,
+      "step": 12437
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5360364455890111,
+      "learning_rate": 1.4069405989918795e-08,
+      "loss": 0.4856,
+      "step": 12438
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7875962319350327,
+      "learning_rate": 1.3974226238467337e-08,
+      "loss": 0.4406,
+      "step": 12439
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2361012055622402,
+      "learning_rate": 1.3879369076361115e-08,
+      "loss": 0.435,
+      "step": 12440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0111370813231217,
+      "learning_rate": 1.3784834509736888e-08,
+      "loss": 0.4896,
+      "step": 12441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0991391373320645,
+      "learning_rate": 1.3690622544711429e-08,
+      "loss": 0.409,
+      "step": 12442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5819138544642948,
+      "learning_rate": 1.3596733187379307e-08,
+      "loss": 0.4001,
+      "step": 12443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6426891537401123,
+      "learning_rate": 1.3503166443815663e-08,
+      "loss": 0.4505,
+      "step": 12444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.536805548788582,
+      "learning_rate": 1.3409922320074543e-08,
+      "loss": 0.4558,
+      "step": 12445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.036213115241153,
+      "learning_rate": 1.3317000822187232e-08,
+      "loss": 0.4884,
+      "step": 12446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5599256441556179,
+      "learning_rate": 1.3224401956167255e-08,
+      "loss": 0.471,
+      "step": 12447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7180533833538991,
+      "learning_rate": 1.3132125728004818e-08,
+      "loss": 0.4318,
+      "step": 12448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4900217684565824,
+      "learning_rate": 1.3040172143670148e-08,
+      "loss": 0.4495,
+      "step": 12449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9021945468911312,
+      "learning_rate": 1.2948541209112375e-08,
+      "loss": 0.4474,
+      "step": 12450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.0403115975809984,
+      "learning_rate": 1.2857232930260089e-08,
+      "loss": 0.4302,
+      "step": 12451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1025971055779142,
+      "learning_rate": 1.2766247313021341e-08,
+      "loss": 0.4879,
+      "step": 12452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9755778633438108,
+      "learning_rate": 1.2675584363281979e-08,
+      "loss": 0.475,
+      "step": 12453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6410505017855979,
+      "learning_rate": 1.2585244086907867e-08,
+      "loss": 0.4346,
+      "step": 12454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.320647738236473,
+      "learning_rate": 1.2495226489744328e-08,
+      "loss": 0.4043,
+      "step": 12455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6080114488495076,
+      "learning_rate": 1.2405531577615037e-08,
+      "loss": 0.4014,
+      "step": 12456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.077484636984275,
+      "learning_rate": 1.2316159356323132e-08,
+      "loss": 0.451,
+      "step": 12457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4593675162351005,
+      "learning_rate": 1.2227109831651763e-08,
+      "loss": 0.4389,
+      "step": 12458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6844183378142294,
+      "learning_rate": 1.2138383009360765e-08,
+      "loss": 0.3992,
+      "step": 12459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5246330768015395,
+      "learning_rate": 1.204997889519166e-08,
+      "loss": 0.4348,
+      "step": 12460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.085555009779293,
+      "learning_rate": 1.1961897494863761e-08,
+      "loss": 0.4301,
+      "step": 12461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9416403658433656,
+      "learning_rate": 1.1874138814075842e-08,
+      "loss": 0.4542,
+      "step": 12462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.5386038278228558,
+      "learning_rate": 1.1786702858506138e-08,
+      "loss": 0.4181,
+      "step": 12463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5406098473433208,
+      "learning_rate": 1.1699589633811236e-08,
+      "loss": 0.4708,
+      "step": 12464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3531787931654895,
+      "learning_rate": 1.161279914562774e-08,
+      "loss": 0.4606,
+      "step": 12465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3752613686105355,
+      "learning_rate": 1.1526331399570045e-08,
+      "loss": 0.4254,
+      "step": 12466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6261018432549799,
+      "learning_rate": 1.1440186401233123e-08,
+      "loss": 0.4538,
+      "step": 12467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7635071713066908,
+      "learning_rate": 1.1354364156189734e-08,
+      "loss": 0.438,
+      "step": 12468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6820209462723312,
+      "learning_rate": 1.1268864669993773e-08,
+      "loss": 0.4385,
+      "step": 12469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6472560419342481,
+      "learning_rate": 1.1183687948175814e-08,
+      "loss": 0.4733,
+      "step": 12470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.818097929327011,
+      "learning_rate": 1.109883399624756e-08,
+      "loss": 0.4247,
+      "step": 12471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.007440964239131,
+      "learning_rate": 1.1014302819697952e-08,
+      "loss": 0.4594,
+      "step": 12472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6816141890342104,
+      "learning_rate": 1.0930094423996508e-08,
+      "loss": 0.4831,
+      "step": 12473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.686752515404565,
+      "learning_rate": 1.0846208814591642e-08,
+      "loss": 0.4609,
+      "step": 12474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.1718206277850065,
+      "learning_rate": 1.0762645996910126e-08,
+      "loss": 0.4832,
+      "step": 12475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8219524276605636,
+      "learning_rate": 1.0679405976359303e-08,
+      "loss": 0.458,
+      "step": 12476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.070349208892602,
+      "learning_rate": 1.0596488758323198e-08,
+      "loss": 0.4414,
+      "step": 12477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6795091354404088,
+      "learning_rate": 1.0513894348168074e-08,
+      "loss": 0.4133,
+      "step": 12478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5278543022334349,
+      "learning_rate": 1.0431622751236326e-08,
+      "loss": 0.4756,
+      "step": 12479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6986593718091938,
+      "learning_rate": 1.0349673972852025e-08,
+      "loss": 0.4301,
+      "step": 12480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5694325962801887,
+      "learning_rate": 1.0268048018315935e-08,
+      "loss": 0.4633,
+      "step": 12481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.646592610833807,
+      "learning_rate": 1.0186744892909939e-08,
+      "loss": 0.4805,
+      "step": 12482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6287190815935564,
+      "learning_rate": 1.0105764601894274e-08,
+      "loss": 0.4353,
+      "step": 12483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5997022011670194,
+      "learning_rate": 1.0025107150508085e-08,
+      "loss": 0.439,
+      "step": 12484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8898280559415683,
+      "learning_rate": 9.944772543969417e-09,
+      "loss": 0.4542,
+      "step": 12485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9372268015740353,
+      "learning_rate": 9.864760787476336e-09,
+      "loss": 0.4609,
+      "step": 12486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5533978310676768,
+      "learning_rate": 9.785071886205256e-09,
+      "loss": 0.4693,
+      "step": 12487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.637118589972661,
+      "learning_rate": 9.705705845312052e-09,
+      "loss": 0.4497,
+      "step": 12488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1195306500732567,
+      "learning_rate": 9.626662669931507e-09,
+      "loss": 0.4441,
+      "step": 12489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5406200919018584,
+      "learning_rate": 9.547942365177865e-09,
+      "loss": 0.4616,
+      "step": 12490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.633634812555877,
+      "learning_rate": 9.469544936143715e-09,
+      "loss": 0.4279,
+      "step": 12491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6129967266922169,
+      "learning_rate": 9.391470387901669e-09,
+      "loss": 0.4574,
+      "step": 12492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7075401056109174,
+      "learning_rate": 9.31371872550324e-09,
+      "loss": 0.4309,
+      "step": 12493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5134047362786376,
+      "learning_rate": 9.23628995397774e-09,
+      "loss": 0.4877,
+      "step": 12494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.8002690524548104,
+      "learning_rate": 9.159184078336714e-09,
+      "loss": 0.434,
+      "step": 12495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.727504780408673,
+      "learning_rate": 9.082401103566729e-09,
+      "loss": 0.4242,
+      "step": 12496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.6428157749345744,
+      "learning_rate": 9.005941034636589e-09,
+      "loss": 0.4986,
+      "step": 12497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6031208444443603,
+      "learning_rate": 8.929803876493448e-09,
+      "loss": 0.4342,
+      "step": 12498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5116602925620355,
+      "learning_rate": 8.853989634062255e-09,
+      "loss": 0.4165,
+      "step": 12499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5207485320793253,
+      "learning_rate": 8.778498312249083e-09,
+      "loss": 0.4639,
+      "step": 12500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6557388787633334,
+      "learning_rate": 8.703329915937808e-09,
+      "loss": 0.4893,
+      "step": 12501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.555961993380804,
+      "learning_rate": 8.628484449991759e-09,
+      "loss": 0.4641,
+      "step": 12502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7873939258482971,
+      "learning_rate": 8.553961919252618e-09,
+      "loss": 0.4899,
+      "step": 12503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.176924203274366,
+      "learning_rate": 8.479762328543195e-09,
+      "loss": 0.4841,
+      "step": 12504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0594458894041816,
+      "learning_rate": 8.405885682662983e-09,
+      "loss": 0.4124,
+      "step": 12505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5285612407395224,
+      "learning_rate": 8.332331986392606e-09,
+      "loss": 0.4813,
+      "step": 12506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4827611726782732,
+      "learning_rate": 8.25910124448992e-09,
+      "loss": 0.4485,
+      "step": 12507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9188966963848202,
+      "learning_rate": 8.186193461693359e-09,
+      "loss": 0.4439,
+      "step": 12508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2877615313442303,
+      "learning_rate": 8.11360864272026e-09,
+      "loss": 0.4499,
+      "step": 12509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1060685943110413,
+      "learning_rate": 8.041346792266313e-09,
+      "loss": 0.4501,
+      "step": 12510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5991257398218082,
+      "learning_rate": 7.969407915007221e-09,
+      "loss": 0.4369,
+      "step": 12511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.3948486845115555,
+      "learning_rate": 7.897792015596484e-09,
+      "loss": 0.4424,
+      "step": 12512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.624702457056527,
+      "learning_rate": 7.826499098668728e-09,
+      "loss": 0.4198,
+      "step": 12513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.4245030949680832,
+      "learning_rate": 7.755529168835263e-09,
+      "loss": 0.4454,
+      "step": 12514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6885482230210522,
+      "learning_rate": 7.684882230689083e-09,
+      "loss": 0.4082,
+      "step": 12515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6320356095730846,
+      "learning_rate": 7.614558288799867e-09,
+      "loss": 0.4074,
+      "step": 12516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.812139307775009,
+      "learning_rate": 7.544557347717863e-09,
+      "loss": 0.4168,
+      "step": 12517
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5353127016172445,
+      "learning_rate": 7.47487941197167e-09,
+      "loss": 0.4709,
+      "step": 12518
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.2160211052553445,
+      "learning_rate": 7.405524486070459e-09,
+      "loss": 0.3932,
+      "step": 12519
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4763512548568691,
+      "learning_rate": 7.336492574500087e-09,
+      "loss": 0.4209,
+      "step": 12520
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5663983031717892,
+      "learning_rate": 7.267783681728091e-09,
+      "loss": 0.4826,
+      "step": 12521
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.200697643459193,
+      "learning_rate": 7.199397812198694e-09,
+      "loss": 0.4394,
+      "step": 12522
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.911818981207784,
+      "learning_rate": 7.13133497033669e-09,
+      "loss": 0.4324,
+      "step": 12523
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.59879226006257,
+      "learning_rate": 7.063595160546332e-09,
+      "loss": 0.4992,
+      "step": 12524
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.970817586930159,
+      "learning_rate": 6.9961783872091185e-09,
+      "loss": 0.47,
+      "step": 12525
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7914245449334705,
+      "learning_rate": 6.929084654688223e-09,
+      "loss": 0.4855,
+      "step": 12526
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.562617636154707,
+      "learning_rate": 6.862313967323508e-09,
+      "loss": 0.4592,
+      "step": 12527
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5172123400698967,
+      "learning_rate": 6.795866329434853e-09,
+      "loss": 0.4316,
+      "step": 12528
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.0669896121000284,
+      "learning_rate": 6.729741745322149e-09,
+      "loss": 0.4048,
+      "step": 12529
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 3.3857952260583914,
+      "learning_rate": 6.663940219263088e-09,
+      "loss": 0.4202,
+      "step": 12530
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.043134969167867,
+      "learning_rate": 6.598461755514818e-09,
+      "loss": 0.4681,
+      "step": 12531
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.604203662963431,
+      "learning_rate": 6.533306358313951e-09,
+      "loss": 0.4552,
+      "step": 12532
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.1187301365065365,
+      "learning_rate": 6.468474031876004e-09,
+      "loss": 0.4618,
+      "step": 12533
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8939579672034357,
+      "learning_rate": 6.4039647803953994e-09,
+      "loss": 0.4519,
+      "step": 12534
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.8553072704475941,
+      "learning_rate": 6.339778608046021e-09,
+      "loss": 0.4152,
+      "step": 12535
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3667794668592637,
+      "learning_rate": 6.275915518980102e-09,
+      "loss": 0.4094,
+      "step": 12536
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 4.8616680695991965,
+      "learning_rate": 6.212375517329894e-09,
+      "loss": 0.4965,
+      "step": 12537
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7325493851720626,
+      "learning_rate": 6.149158607206552e-09,
+      "loss": 0.4418,
+      "step": 12538
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.070073371111874,
+      "learning_rate": 6.086264792699581e-09,
+      "loss": 0.4642,
+      "step": 12539
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5267055785058776,
+      "learning_rate": 6.023694077878506e-09,
+      "loss": 0.4346,
+      "step": 12540
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.897340562724816,
+      "learning_rate": 5.961446466791199e-09,
+      "loss": 0.4368,
+      "step": 12541
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5575401288335324,
+      "learning_rate": 5.899521963465549e-09,
+      "loss": 0.4369,
+      "step": 12542
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.830500131131668,
+      "learning_rate": 5.837920571907796e-09,
+      "loss": 0.4485,
+      "step": 12543
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.737592730959085,
+      "learning_rate": 5.7766422961030855e-09,
+      "loss": 0.42,
+      "step": 12544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5180532914566667,
+      "learning_rate": 5.715687140016579e-09,
+      "loss": 0.4732,
+      "step": 12545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5399953821077093,
+      "learning_rate": 5.6550551075923445e-09,
+      "loss": 0.4859,
+      "step": 12546
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.835526287629584,
+      "learning_rate": 5.594746202751688e-09,
+      "loss": 0.4282,
+      "step": 12547
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9304038926712141,
+      "learning_rate": 5.534760429398156e-09,
+      "loss": 0.45,
+      "step": 12548
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8160448694699245,
+      "learning_rate": 5.475097791411421e-09,
+      "loss": 0.4387,
+      "step": 12549
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6161554019832003,
+      "learning_rate": 5.41575829265284e-09,
+      "loss": 0.4763,
+      "step": 12550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1088372761369976,
+      "learning_rate": 5.356741936960452e-09,
+      "loss": 0.4589,
+      "step": 12551
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0716707289278915,
+      "learning_rate": 5.298048728152872e-09,
+      "loss": 0.5001,
+      "step": 12552
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6048323588845608,
+      "learning_rate": 5.2396786700281695e-09,
+      "loss": 0.4213,
+      "step": 12553
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.185769982058536,
+      "learning_rate": 5.181631766362216e-09,
+      "loss": 0.4534,
+      "step": 12554
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8134513300830346,
+      "learning_rate": 5.12390802091034e-09,
+      "loss": 0.4304,
+      "step": 12555
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7380971816132587,
+      "learning_rate": 5.066507437407886e-09,
+      "loss": 0.457,
+      "step": 12556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5501853884556158,
+      "learning_rate": 5.009430019567996e-09,
+      "loss": 0.4683,
+      "step": 12557
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.783188806585626,
+      "learning_rate": 4.952675771083826e-09,
+      "loss": 0.4398,
+      "step": 12558
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.3761827483572775,
+      "learning_rate": 4.896244695626884e-09,
+      "loss": 0.4314,
+      "step": 12559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7419609398005242,
+      "learning_rate": 4.840136796848694e-09,
+      "loss": 0.4271,
+      "step": 12560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4873744251207768,
+      "learning_rate": 4.784352078379129e-09,
+      "loss": 0.3904,
+      "step": 12561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6377879071226575,
+      "learning_rate": 4.728890543827525e-09,
+      "loss": 0.4333,
+      "step": 12562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.74665235922418,
+      "learning_rate": 4.673752196782122e-09,
+      "loss": 0.4009,
+      "step": 12563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6837070454110004,
+      "learning_rate": 4.618937040810068e-09,
+      "loss": 0.4144,
+      "step": 12564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1052359316794678,
+      "learning_rate": 4.564445079457969e-09,
+      "loss": 0.461,
+      "step": 12565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5818245045843733,
+      "learning_rate": 4.510276316251339e-09,
+      "loss": 0.4229,
+      "step": 12566
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.779675202771309,
+      "learning_rate": 4.456430754695151e-09,
+      "loss": 0.4771,
+      "step": 12567
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.127673950283077,
+      "learning_rate": 4.4029083982727314e-09,
+      "loss": 0.5127,
+      "step": 12568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6734948827249356,
+      "learning_rate": 4.34970925044742e-09,
+      "loss": 0.4384,
+      "step": 12569
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8513899284300206,
+      "learning_rate": 4.296833314660354e-09,
+      "loss": 0.4326,
+      "step": 12570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8720860971901838,
+      "learning_rate": 4.244280594333239e-09,
+      "loss": 0.4191,
+      "step": 12571
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5922233995268702,
+      "learning_rate": 4.192051092866134e-09,
+      "loss": 0.4358,
+      "step": 12572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5047227126346328,
+      "learning_rate": 4.140144813637448e-09,
+      "loss": 0.4575,
+      "step": 12573
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5054844439859798,
+      "learning_rate": 4.08856176000616e-09,
+      "loss": 0.4882,
+      "step": 12574
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.543994646107952,
+      "learning_rate": 4.037301935309601e-09,
+      "loss": 0.3717,
+      "step": 12575
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.568550411337171,
+      "learning_rate": 3.986365342864007e-09,
+      "loss": 0.4686,
+      "step": 12576
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6765072465540531,
+      "learning_rate": 3.935751985965075e-09,
+      "loss": 0.4268,
+      "step": 12577
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5680153252172166,
+      "learning_rate": 3.885461867886852e-09,
+      "loss": 0.4592,
+      "step": 12578
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.743475879591402,
+      "learning_rate": 3.835494991883959e-09,
+      "loss": 0.4043,
+      "step": 12579
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.542144750645424,
+      "learning_rate": 3.785851361188808e-09,
+      "loss": 0.4415,
+      "step": 12580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5333926961892341,
+      "learning_rate": 3.736530979012721e-09,
+      "loss": 0.4617,
+      "step": 12581
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6365016367833727,
+      "learning_rate": 3.6875338485475866e-09,
+      "loss": 0.4253,
+      "step": 12582
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5217973327164985,
+      "learning_rate": 3.6388599729625383e-09,
+      "loss": 0.4607,
+      "step": 12583
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 16.75684485275876,
+      "learning_rate": 3.590509355407834e-09,
+      "loss": 0.4537,
+      "step": 12584
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7779382848013272,
+      "learning_rate": 3.5424819990104163e-09,
+      "loss": 0.4907,
+      "step": 12585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5397302420331073,
+      "learning_rate": 3.49477790687891e-09,
+      "loss": 0.4697,
+      "step": 12586
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.019218932599117,
+      "learning_rate": 3.4473970820986248e-09,
+      "loss": 0.437,
+      "step": 12587
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.2831027218224276,
+      "learning_rate": 3.4003395277354413e-09,
+      "loss": 0.4429,
+      "step": 12588
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.7428038821769722,
+      "learning_rate": 3.353605246833591e-09,
+      "loss": 0.4658,
+      "step": 12589
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8525318997912361,
+      "learning_rate": 3.307194242417322e-09,
+      "loss": 0.4467,
+      "step": 12590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7412943231851326,
+      "learning_rate": 3.261106517489232e-09,
+      "loss": 0.4505,
+      "step": 12591
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8502048825724473,
+      "learning_rate": 3.215342075030825e-09,
+      "loss": 0.4773,
+      "step": 12592
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.7600758568759587,
+      "learning_rate": 3.169900918003621e-09,
+      "loss": 0.4455,
+      "step": 12593
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8894457368547062,
+      "learning_rate": 3.1247830493469357e-09,
+      "loss": 0.4306,
+      "step": 12594
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7994146551691743,
+      "learning_rate": 3.0799884719795448e-09,
+      "loss": 0.4223,
+      "step": 12595
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6144149442660667,
+      "learning_rate": 3.035517188800796e-09,
+      "loss": 0.4332,
+      "step": 12596
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.6487461277474638,
+      "learning_rate": 2.9913692026867227e-09,
+      "loss": 0.494,
+      "step": 12597
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8387610012853146,
+      "learning_rate": 2.947544516494483e-09,
+      "loss": 0.461,
+      "step": 12598
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5622533905295612,
+      "learning_rate": 2.9040431330584762e-09,
+      "loss": 0.4856,
+      "step": 12599
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.730566070880151,
+      "learning_rate": 2.8608650551947835e-09,
+      "loss": 0.4485,
+      "step": 12600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6813296552044827,
+      "learning_rate": 2.818010285695616e-09,
+      "loss": 0.4401,
+      "step": 12601
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8829375974243217,
+      "learning_rate": 2.7754788273337553e-09,
+      "loss": 0.453,
+      "step": 12602
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4322393591207865,
+      "learning_rate": 2.733270682861444e-09,
+      "loss": 0.4748,
+      "step": 12603
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8279137379874328,
+      "learning_rate": 2.6913858550098314e-09,
+      "loss": 0.4555,
+      "step": 12604
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.620441580584949,
+      "learning_rate": 2.6498243464878614e-09,
+      "loss": 0.4354,
+      "step": 12605
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.723250227563526,
+      "learning_rate": 2.6085861599844943e-09,
+      "loss": 0.4761,
+      "step": 12606
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.653617630457076,
+      "learning_rate": 2.5676712981687058e-09,
+      "loss": 0.4205,
+      "step": 12607
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8081656612909613,
+      "learning_rate": 2.5270797636867127e-09,
+      "loss": 0.4807,
+      "step": 12608
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.5980962280382385,
+      "learning_rate": 2.486811559165303e-09,
+      "loss": 0.4742,
+      "step": 12609
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6159725732225896,
+      "learning_rate": 2.4468666872096148e-09,
+      "loss": 0.4149,
+      "step": 12610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.651409029062181,
+      "learning_rate": 2.407245150404247e-09,
+      "loss": 0.4721,
+      "step": 12611
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4651318218104874,
+      "learning_rate": 2.36794695131215e-09,
+      "loss": 0.435,
+      "step": 12612
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8897248070299966,
+      "learning_rate": 2.3289720924762893e-09,
+      "loss": 0.4097,
+      "step": 12613
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5459885522426507,
+      "learning_rate": 2.2903205764185364e-09,
+      "loss": 0.4722,
+      "step": 12614
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5094813596235412,
+      "learning_rate": 2.251992405638559e-09,
+      "loss": 0.4614,
+      "step": 12615
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5326465389649856,
+      "learning_rate": 2.21398758261715e-09,
+      "loss": 0.4607,
+      "step": 12616
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8250086540855375,
+      "learning_rate": 2.176306109812898e-09,
+      "loss": 0.41,
+      "step": 12617
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9695860736088986,
+      "learning_rate": 2.138947989663298e-09,
+      "loss": 0.4135,
+      "step": 12618
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 3.161567082916375,
+      "learning_rate": 2.101913224585861e-09,
+      "loss": 0.474,
+      "step": 12619
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2416319555802144,
+      "learning_rate": 2.065201816977003e-09,
+      "loss": 0.4995,
+      "step": 12620
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5850137362441887,
+      "learning_rate": 2.0288137692109357e-09,
+      "loss": 0.3967,
+      "step": 12621
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2629690739861874,
+      "learning_rate": 1.9927490836429975e-09,
+      "loss": 0.4719,
+      "step": 12622
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5208464892056612,
+      "learning_rate": 1.957007762605212e-09,
+      "loss": 0.4435,
+      "step": 12623
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7229118507059125,
+      "learning_rate": 1.9215898084112838e-09,
+      "loss": 0.4288,
+      "step": 12624
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.941189359487513,
+      "learning_rate": 1.8864952233521584e-09,
+      "loss": 0.4395,
+      "step": 12625
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6019764991918644,
+      "learning_rate": 1.8517240096982414e-09,
+      "loss": 0.4247,
+      "step": 12626
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8986391093999524,
+      "learning_rate": 1.8172761696994001e-09,
+      "loss": 0.489,
+      "step": 12627
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0394481341760646,
+      "learning_rate": 1.7831517055849623e-09,
+      "loss": 0.487,
+      "step": 12628
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2593047336199894,
+      "learning_rate": 1.7493506195614963e-09,
+      "loss": 0.4614,
+      "step": 12629
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5642282037206607,
+      "learning_rate": 1.7158729138166963e-09,
+      "loss": 0.4371,
+      "step": 12630
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7932297997838356,
+      "learning_rate": 1.6827185905160526e-09,
+      "loss": 0.4141,
+      "step": 12631
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0383482222754656,
+      "learning_rate": 1.6498876518050711e-09,
+      "loss": 0.4467,
+      "step": 12632
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8206322122454668,
+      "learning_rate": 1.6173800998076084e-09,
+      "loss": 0.489,
+      "step": 12633
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9885811727605542,
+      "learning_rate": 1.585195936626982e-09,
+      "loss": 0.4226,
+      "step": 12634
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5597764070803066,
+      "learning_rate": 1.5533351643459704e-09,
+      "loss": 0.4735,
+      "step": 12635
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.8587457064072965,
+      "learning_rate": 1.5217977850245925e-09,
+      "loss": 0.4089,
+      "step": 12636
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.0254855543485153,
+      "learning_rate": 1.4905838007045482e-09,
+      "loss": 0.4698,
+      "step": 12637
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.7780462741774332,
+      "learning_rate": 1.4596932134047782e-09,
+      "loss": 0.444,
+      "step": 12638
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5095873108317084,
+      "learning_rate": 1.4291260251236837e-09,
+      "loss": 0.4396,
+      "step": 12639
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.416634595873887,
+      "learning_rate": 1.3988822378396827e-09,
+      "loss": 0.3945,
+      "step": 12640
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.452242685705804,
+      "learning_rate": 1.3689618535084326e-09,
+      "loss": 0.4513,
+      "step": 12641
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5175640346460375,
+      "learning_rate": 1.3393648740667175e-09,
+      "loss": 0.444,
+      "step": 12642
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5284292319516303,
+      "learning_rate": 1.3100913014291172e-09,
+      "loss": 0.4663,
+      "step": 12643
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5361981828406261,
+      "learning_rate": 1.281141137489117e-09,
+      "loss": 0.4546,
+      "step": 12644
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.400674992174665,
+      "learning_rate": 1.2525143841202182e-09,
+      "loss": 0.3892,
+      "step": 12645
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.2224078853439946,
+      "learning_rate": 1.2242110431742727e-09,
+      "loss": 0.4413,
+      "step": 12646
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1059507453490065,
+      "learning_rate": 1.1962311164831485e-09,
+      "loss": 0.413,
+      "step": 12647
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.1326559083608303,
+      "learning_rate": 1.1685746058565095e-09,
+      "loss": 0.4486,
+      "step": 12648
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.012181921987285,
+      "learning_rate": 1.1412415130834797e-09,
+      "loss": 0.4484,
+      "step": 12649
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3939322267270886,
+      "learning_rate": 1.1142318399331997e-09,
+      "loss": 0.3886,
+      "step": 12650
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5624958257655308,
+      "learning_rate": 1.0875455881526055e-09,
+      "loss": 0.4612,
+      "step": 12651
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.9654969853414817,
+      "learning_rate": 1.0611827594680934e-09,
+      "loss": 0.483,
+      "step": 12652
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4940693652885892,
+      "learning_rate": 1.0351433555860769e-09,
+      "loss": 0.4191,
+      "step": 12653
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.9052627213490994,
+      "learning_rate": 1.0094273781907637e-09,
+      "loss": 0.4743,
+      "step": 12654
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6582757608629217,
+      "learning_rate": 9.840348289458234e-10,
+      "loss": 0.4644,
+      "step": 12655
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8414704778888966,
+      "learning_rate": 9.58965709494386e-10,
+      "loss": 0.4045,
+      "step": 12656
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5403204651039883,
+      "learning_rate": 9.342200214579323e-10,
+      "loss": 0.4109,
+      "step": 12657
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.516150066535453,
+      "learning_rate": 9.097977664385138e-10,
+      "loss": 0.4368,
+      "step": 12658
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.119426394438219,
+      "learning_rate": 8.856989460148679e-10,
+      "loss": 0.4289,
+      "step": 12659
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5717193044962845,
+      "learning_rate": 8.619235617474131e-10,
+      "loss": 0.4182,
+      "step": 12660
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5701016631999365,
+      "learning_rate": 8.38471615173253e-10,
+      "loss": 0.4469,
+      "step": 12661
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.4526411242547237,
+      "learning_rate": 8.153431078106177e-10,
+      "loss": 0.5089,
+      "step": 12662
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.8711326982969616,
+      "learning_rate": 7.925380411555327e-10,
+      "loss": 0.4796,
+      "step": 12663
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.82933748367458,
+      "learning_rate": 7.700564166834845e-10,
+      "loss": 0.4149,
+      "step": 12664
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.778140889547643,
+      "learning_rate": 7.478982358488651e-10,
+      "loss": 0.4111,
+      "step": 12665
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.551549633281428,
+      "learning_rate": 7.260635000855276e-10,
+      "loss": 0.4412,
+      "step": 12666
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6775553700036507,
+      "learning_rate": 7.045522108056757e-10,
+      "loss": 0.4213,
+      "step": 12667
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4542510969415225,
+      "learning_rate": 6.833643694015291e-10,
+      "loss": 0.4203,
+      "step": 12668
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4612066197198725,
+      "learning_rate": 6.62499977243658e-10,
+      "loss": 0.4268,
+      "step": 12669
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6071274144746788,
+      "learning_rate": 6.419590356826489e-10,
+      "loss": 0.4404,
+      "step": 12670
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.5635197407005026,
+      "learning_rate": 6.217415460463283e-10,
+      "loss": 0.4582,
+      "step": 12671
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.5138247265949745,
+      "learning_rate": 6.018475096436494e-10,
+      "loss": 0.4735,
+      "step": 12672
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5127880143163386,
+      "learning_rate": 5.822769277613604e-10,
+      "loss": 0.4465,
+      "step": 12673
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.926359490358433,
+      "learning_rate": 5.63029801665671e-10,
+      "loss": 0.4695,
+      "step": 12674
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.4810396036485027,
+      "learning_rate": 5.44106132601696e-10,
+      "loss": 0.4371,
+      "step": 12675
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7347582054933806,
+      "learning_rate": 5.255059217940117e-10,
+      "loss": 0.4459,
+      "step": 12676
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2379992398321513,
+      "learning_rate": 5.072291704460997e-10,
+      "loss": 0.3806,
+      "step": 12677
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.523233081733703,
+      "learning_rate": 4.892758797397923e-10,
+      "loss": 0.4726,
+      "step": 12678
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4533494827497468,
+      "learning_rate": 4.716460508380482e-10,
+      "loss": 0.4463,
+      "step": 12679
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5101038936748863,
+      "learning_rate": 4.543396848799564e-10,
+      "loss": 0.4486,
+      "step": 12680
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.730805620410329,
+      "learning_rate": 4.373567829857317e-10,
+      "loss": 0.4636,
+      "step": 12681
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4986104858271505,
+      "learning_rate": 4.20697346254495e-10,
+      "loss": 0.4653,
+      "step": 12682
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7828414255189737,
+      "learning_rate": 4.043613757642728e-10,
+      "loss": 0.509,
+      "step": 12683
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7409368545554322,
+      "learning_rate": 3.8834887257088726e-10,
+      "loss": 0.4569,
+      "step": 12684
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8107875358792018,
+      "learning_rate": 3.7265983771184176e-10,
+      "loss": 0.4634,
+      "step": 12685
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.1987832332917243,
+      "learning_rate": 3.5729427220076997e-10,
+      "loss": 0.3984,
+      "step": 12686
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5882555842386747,
+      "learning_rate": 3.422521770324316e-10,
+      "loss": 0.4267,
+      "step": 12687
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.13433649911325,
+      "learning_rate": 3.275335531799373e-10,
+      "loss": 0.4363,
+      "step": 12688
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.058063244847107,
+      "learning_rate": 3.131384015958583e-10,
+      "loss": 0.4152,
+      "step": 12689
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6775059766723466,
+      "learning_rate": 2.9906672321111664e-10,
+      "loss": 0.403,
+      "step": 12690
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2414625135016903,
+      "learning_rate": 2.853185189366503e-10,
+      "loss": 0.4378,
+      "step": 12691
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.194515766554616,
+      "learning_rate": 2.718937896611928e-10,
+      "loss": 0.4709,
+      "step": 12692
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.0364742613506723,
+      "learning_rate": 2.587925362540489e-10,
+      "loss": 0.4631,
+      "step": 12693
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7429766137277094,
+      "learning_rate": 2.460147595628737e-10,
+      "loss": 0.4512,
+      "step": 12694
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5571357964895418,
+      "learning_rate": 2.335604604131181e-10,
+      "loss": 0.4161,
+      "step": 12695
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.32499533573689,
+      "learning_rate": 2.2142963961246932e-10,
+      "loss": 0.4438,
+      "step": 12696
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4826414172354452,
+      "learning_rate": 2.0962229794418975e-10,
+      "loss": 0.3903,
+      "step": 12697
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9143893265601541,
+      "learning_rate": 1.9813843617266793e-10,
+      "loss": 0.4406,
+      "step": 12698
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.964557147942109,
+      "learning_rate": 1.8697805504175326e-10,
+      "loss": 0.4741,
+      "step": 12699
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5931841070559926,
+      "learning_rate": 1.761411552719805e-10,
+      "loss": 0.4752,
+      "step": 12700
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4855245468592573,
+      "learning_rate": 1.6562773756556573e-10,
+      "loss": 0.4715,
+      "step": 12701
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5683864777949463,
+      "learning_rate": 1.554378026025205e-10,
+      "loss": 0.4606,
+      "step": 12702
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5715258959469307,
+      "learning_rate": 1.4557135104231735e-10,
+      "loss": 0.4362,
+      "step": 12703
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.816903592909708,
+      "learning_rate": 1.3602838352277936e-10,
+      "loss": 0.4721,
+      "step": 12704
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6929083222985504,
+      "learning_rate": 1.2680890066119055e-10,
+      "loss": 0.439,
+      "step": 12705
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2301044291630188,
+      "learning_rate": 1.1791290305485093e-10,
+      "loss": 0.3951,
+      "step": 12706
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4302259131335344,
+      "learning_rate": 1.0934039127885598e-10,
+      "loss": 0.4174,
+      "step": 12707
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.782097524570738,
+      "learning_rate": 1.0109136588720703e-10,
+      "loss": 0.4739,
+      "step": 12708
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7439158114583095,
+      "learning_rate": 9.316582741503155e-11,
+      "loss": 0.4081,
+      "step": 12709
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.991228070544835,
+      "learning_rate": 8.556377637414237e-11,
+      "loss": 0.4222,
+      "step": 12710
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.55848693736396,
+      "learning_rate": 7.828521325636829e-11,
+      "loss": 0.4573,
+      "step": 12711
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7324051743234623,
+      "learning_rate": 7.133013853299897e-11,
+      "loss": 0.4124,
+      "step": 12712
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9375806151467723,
+      "learning_rate": 6.469855265367475e-11,
+      "loss": 0.4416,
+      "step": 12713
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5298100461225657,
+      "learning_rate": 5.839045604805194e-11,
+      "loss": 0.4134,
+      "step": 12714
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5515084004453629,
+      "learning_rate": 5.240584912358238e-11,
+      "loss": 0.4532,
+      "step": 12715
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5531722149703644,
+      "learning_rate": 4.674473226717879e-11,
+      "loss": 0.4737,
+      "step": 12716
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.8600555374023193,
+      "learning_rate": 4.1407105846325015e-11,
+      "loss": 0.4011,
+      "step": 12717
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.2951534367432265,
+      "learning_rate": 3.639297020519017e-11,
+      "loss": 0.4618,
+      "step": 12718
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5309432915281782,
+      "learning_rate": 3.170232566906961e-11,
+      "loss": 0.472,
+      "step": 12719
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.648632433604236,
+      "learning_rate": 2.7335172541054222e-11,
+      "loss": 0.472,
+      "step": 12720
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9891922851986015,
+      "learning_rate": 2.329151110314065e-11,
+      "loss": 0.4146,
+      "step": 12721
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4863681096782582,
+      "learning_rate": 1.9571341617896643e-11,
+      "loss": 0.4691,
+      "step": 12722
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5474883211736816,
+      "learning_rate": 1.6174664325685485e-11,
+      "loss": 0.4772,
+      "step": 12723
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7968771768996952,
+      "learning_rate": 1.3101479445776222e-11,
+      "loss": 0.432,
+      "step": 12724
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7850241950049794,
+      "learning_rate": 1.0351787177453886e-11,
+      "loss": 0.4187,
+      "step": 12725
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6177864248250622,
+      "learning_rate": 7.925587698909276e-12,
+      "loss": 0.4347,
+      "step": 12726
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9385634257563205,
+      "learning_rate": 5.8228811661287244e-12,
+      "loss": 0.4436,
+      "step": 12727
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.103776297698041,
+      "learning_rate": 4.043667716224775e-12,
+      "loss": 0.4528,
+      "step": 12728
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.541259578282624,
+      "learning_rate": 2.5879474641055114e-12,
+      "loss": 0.4331,
+      "step": 12729
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7721900415780467,
+      "learning_rate": 1.4557205035847787e-12,
+      "loss": 0.4684,
+      "step": 12730
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.6051586781797063,
+      "learning_rate": 6.46986907937297e-13,
+      "loss": 0.4647,
+      "step": 12731
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.6430257866029137,
+      "learning_rate": 1.6174672934354819e-13,
+      "loss": 0.4528,
+      "step": 12732
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.7634476721682628,
+      "learning_rate": 0.0,
+      "loss": 0.4617,
+      "step": 12733
+    },
+    {
+      "epoch": 1.0,
+      "step": 12733,
+      "total_flos": 3813892823678976.0,
+      "train_loss": 0.09614035478748979,
+      "train_runtime": 51328.1871,
+      "train_samples_per_second": 63.509,
+      "train_steps_per_second": 0.248
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12733,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 3813892823678976.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}