diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,90764 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 12962,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.614504080242568,
+      "learning_rate": 1.2853470437017994e-07,
+      "loss": 2.9582,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.570163404465948,
+      "learning_rate": 2.570694087403599e-07,
+      "loss": 2.8727,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.356797485109597,
+      "learning_rate": 3.8560411311053987e-07,
+      "loss": 2.8515,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.781751164059694,
+      "learning_rate": 5.141388174807198e-07,
+      "loss": 2.9528,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.94068714667129,
+      "learning_rate": 6.426735218508997e-07,
+      "loss": 2.838,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 14.207690351340352,
+      "learning_rate": 7.712082262210797e-07,
+      "loss": 2.8576,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.860289726894845,
+      "learning_rate": 8.997429305912597e-07,
+      "loss": 2.7774,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.35998549398397,
+      "learning_rate": 1.0282776349614395e-06,
+      "loss": 2.8728,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.849071113840285,
+      "learning_rate": 1.1568123393316196e-06,
+      "loss": 2.7707,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.076004211529174,
+      "learning_rate": 1.2853470437017995e-06,
+      "loss": 2.6408,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.240390429933518,
+      "learning_rate": 1.4138817480719794e-06,
+      "loss": 2.696,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.378123612137972,
+      "learning_rate": 1.5424164524421595e-06,
+      "loss": 2.7634,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.567312081792433,
+      "learning_rate": 1.6709511568123394e-06,
+      "loss": 2.5598,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.044425629898154,
+      "learning_rate": 1.7994858611825194e-06,
+      "loss": 2.5837,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.467778096569555,
+      "learning_rate": 1.928020565552699e-06,
+      "loss": 2.5774,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.218632565331351,
+      "learning_rate": 2.056555269922879e-06,
+      "loss": 2.6803,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.8545724267389976,
+      "learning_rate": 2.1850899742930593e-06,
+      "loss": 2.5352,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1826273541619954,
+      "learning_rate": 2.313624678663239e-06,
+      "loss": 2.4847,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.1569608664356097,
+      "learning_rate": 2.442159383033419e-06,
+      "loss": 2.464,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9987558130636356,
+      "learning_rate": 2.570694087403599e-06,
+      "loss": 2.6248,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.630066207289744,
+      "learning_rate": 2.699228791773779e-06,
+      "loss": 2.445,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.191065794331003,
+      "learning_rate": 2.8277634961439587e-06,
+      "loss": 2.4244,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.9851148257093616,
+      "learning_rate": 2.956298200514139e-06,
+      "loss": 2.4332,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.049303273666483,
+      "learning_rate": 3.084832904884319e-06,
+      "loss": 2.5698,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.6044393980477123,
+      "learning_rate": 3.213367609254499e-06,
+      "loss": 2.4001,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.6235482559708783,
+      "learning_rate": 3.3419023136246787e-06,
+      "loss": 2.4217,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2629172635241919,
+      "learning_rate": 3.470437017994859e-06,
+      "loss": 2.4073,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.300242551751265,
+      "learning_rate": 3.598971722365039e-06,
+      "loss": 2.5546,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.2481148376547564,
+      "learning_rate": 3.7275064267352188e-06,
+      "loss": 2.3982,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.177143906465669,
+      "learning_rate": 3.856041131105398e-06,
+      "loss": 2.3921,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1262339135292236,
+      "learning_rate": 3.984575835475578e-06,
+      "loss": 2.3848,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1201312761638555,
+      "learning_rate": 4.113110539845758e-06,
+      "loss": 2.3534,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.19831346765079,
+      "learning_rate": 4.241645244215939e-06,
+      "loss": 2.5288,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1262109382931946,
+      "learning_rate": 4.370179948586119e-06,
+      "loss": 2.358,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1677785952812396,
+      "learning_rate": 4.4987146529562985e-06,
+      "loss": 2.3458,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.1248316136252485,
+      "learning_rate": 4.627249357326478e-06,
+      "loss": 2.5331,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9438076309994886,
+      "learning_rate": 4.755784061696658e-06,
+      "loss": 2.3739,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8649453968522334,
+      "learning_rate": 4.884318766066838e-06,
+      "loss": 2.3573,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8424219839650771,
+      "learning_rate": 5.012853470437019e-06,
+      "loss": 2.3465,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8819647120903683,
+      "learning_rate": 5.141388174807198e-06,
+      "loss": 2.5345,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9075458149810234,
+      "learning_rate": 5.269922879177378e-06,
+      "loss": 2.3511,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0068870378933819,
+      "learning_rate": 5.398457583547558e-06,
+      "loss": 2.3529,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9514509413390595,
+      "learning_rate": 5.526992287917738e-06,
+      "loss": 2.326,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8556176001785393,
+      "learning_rate": 5.6555269922879175e-06,
+      "loss": 2.3198,
+      "step": 44
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9619745039800207,
+      "learning_rate": 5.784061696658098e-06,
+      "loss": 2.5048,
+      "step": 45
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.9876638609130287,
+      "learning_rate": 5.912596401028278e-06,
+      "loss": 2.3354,
+      "step": 46
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8688675959187675,
+      "learning_rate": 6.041131105398458e-06,
+      "loss": 2.3303,
+      "step": 47
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8025143826323919,
+      "learning_rate": 6.169665809768638e-06,
+      "loss": 2.4709,
+      "step": 48
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8680776063762177,
+      "learning_rate": 6.298200514138818e-06,
+      "loss": 2.2979,
+      "step": 49
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 1.0592149110299272,
+      "learning_rate": 6.426735218508998e-06,
+      "loss": 2.3169,
+      "step": 50
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8973191238306678,
+      "learning_rate": 6.5552699228791775e-06,
+      "loss": 2.3028,
+      "step": 51
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8449496542813714,
+      "learning_rate": 6.683804627249357e-06,
+      "loss": 2.4668,
+      "step": 52
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8771845820006734,
+      "learning_rate": 6.812339331619537e-06,
+      "loss": 2.3433,
+      "step": 53
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8790940901876545,
+      "learning_rate": 6.940874035989718e-06,
+      "loss": 2.298,
+      "step": 54
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.803435281615288,
+      "learning_rate": 7.069408740359898e-06,
+      "loss": 2.3524,
+      "step": 55
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7185911576146709,
+      "learning_rate": 7.197943444730078e-06,
+      "loss": 2.3055,
+      "step": 56
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8062159097676136,
+      "learning_rate": 7.326478149100258e-06,
+      "loss": 2.4896,
+      "step": 57
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7551608013791536,
+      "learning_rate": 7.4550128534704376e-06,
+      "loss": 2.3538,
+      "step": 58
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.743329166402104,
+      "learning_rate": 7.5835475578406175e-06,
+      "loss": 2.2654,
+      "step": 59
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.839762047748952,
+      "learning_rate": 7.712082262210796e-06,
+      "loss": 2.4416,
+      "step": 60
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7605774823430541,
+      "learning_rate": 7.840616966580976e-06,
+      "loss": 2.3077,
+      "step": 61
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8025973417994856,
+      "learning_rate": 7.969151670951156e-06,
+      "loss": 2.2836,
+      "step": 62
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.6864232970479806,
+      "learning_rate": 8.097686375321336e-06,
+      "loss": 2.2811,
+      "step": 63
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7940441918021376,
+      "learning_rate": 8.226221079691516e-06,
+      "loss": 2.2973,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7392450587141398,
+      "learning_rate": 8.354755784061698e-06,
+      "loss": 2.5009,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7406350762506708,
+      "learning_rate": 8.483290488431877e-06,
+      "loss": 2.3045,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7543156254603282,
+      "learning_rate": 8.611825192802057e-06,
+      "loss": 2.2644,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7848852496477787,
+      "learning_rate": 8.740359897172237e-06,
+      "loss": 2.4247,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7365416779921541,
+      "learning_rate": 8.868894601542417e-06,
+      "loss": 2.3137,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7350442586478796,
+      "learning_rate": 8.997429305912597e-06,
+      "loss": 2.2579,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7230483333064687,
+      "learning_rate": 9.125964010282777e-06,
+      "loss": 2.2739,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7269909307113563,
+      "learning_rate": 9.254498714652957e-06,
+      "loss": 2.4287,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7520259645621727,
+      "learning_rate": 9.383033419023137e-06,
+      "loss": 2.2747,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7960539911485558,
+      "learning_rate": 9.511568123393317e-06,
+      "loss": 2.2894,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7160407111609224,
+      "learning_rate": 9.640102827763496e-06,
+      "loss": 2.2722,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7409386960529175,
+      "learning_rate": 9.768637532133676e-06,
+      "loss": 2.2911,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7121957913508553,
+      "learning_rate": 9.897172236503858e-06,
+      "loss": 2.4432,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7342355334139097,
+      "learning_rate": 1.0025706940874038e-05,
+      "loss": 2.3059,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7503926601087365,
+      "learning_rate": 1.0154241645244216e-05,
+      "loss": 2.2978,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7431219726322306,
+      "learning_rate": 1.0282776349614396e-05,
+      "loss": 2.4642,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6912343863017675,
+      "learning_rate": 1.0411311053984576e-05,
+      "loss": 2.2826,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7535315971131915,
+      "learning_rate": 1.0539845758354756e-05,
+      "loss": 2.268,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7293404309596656,
+      "learning_rate": 1.0668380462724936e-05,
+      "loss": 2.2365,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7493574754176567,
+      "learning_rate": 1.0796915167095115e-05,
+      "loss": 2.4182,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7741535205635848,
+      "learning_rate": 1.0925449871465295e-05,
+      "loss": 2.2451,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7249225770927538,
+      "learning_rate": 1.1053984575835475e-05,
+      "loss": 2.2737,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.688407931585645,
+      "learning_rate": 1.1182519280205655e-05,
+      "loss": 2.2964,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7228442476379507,
+      "learning_rate": 1.1311053984575835e-05,
+      "loss": 2.252,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6985712773681425,
+      "learning_rate": 1.1439588688946017e-05,
+      "loss": 2.4318,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.743051302790943,
+      "learning_rate": 1.1568123393316196e-05,
+      "loss": 2.2714,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7352474699051454,
+      "learning_rate": 1.1696658097686376e-05,
+      "loss": 2.2312,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6857505375836221,
+      "learning_rate": 1.1825192802056556e-05,
+      "loss": 2.3752,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7357208435435285,
+      "learning_rate": 1.1953727506426736e-05,
+      "loss": 2.2498,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6828846057938996,
+      "learning_rate": 1.2082262210796916e-05,
+      "loss": 2.2675,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7717079623683487,
+      "learning_rate": 1.2210796915167096e-05,
+      "loss": 2.2534,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7853346252126335,
+      "learning_rate": 1.2339331619537276e-05,
+      "loss": 2.2224,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7346116207273206,
+      "learning_rate": 1.2467866323907456e-05,
+      "loss": 2.4562,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.741239462158256,
+      "learning_rate": 1.2596401028277636e-05,
+      "loss": 2.2545,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7603482137204725,
+      "learning_rate": 1.2724935732647817e-05,
+      "loss": 2.2573,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7028756516841389,
+      "learning_rate": 1.2853470437017995e-05,
+      "loss": 2.2877,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7381484539558425,
+      "learning_rate": 1.2982005141388177e-05,
+      "loss": 2.3774,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7947011145062155,
+      "learning_rate": 1.3110539845758355e-05,
+      "loss": 2.2394,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7327421330152774,
+      "learning_rate": 1.3239074550128535e-05,
+      "loss": 2.2408,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7945811841184374,
+      "learning_rate": 1.3367609254498715e-05,
+      "loss": 2.4034,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7489403340172621,
+      "learning_rate": 1.3496143958868895e-05,
+      "loss": 2.2526,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6899589256194841,
+      "learning_rate": 1.3624678663239075e-05,
+      "loss": 2.2861,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8663417148519842,
+      "learning_rate": 1.3753213367609254e-05,
+      "loss": 2.2177,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7342809824930605,
+      "learning_rate": 1.3881748071979436e-05,
+      "loss": 2.2449,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8160635567395097,
+      "learning_rate": 1.4010282776349614e-05,
+      "loss": 2.41,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8034197795553303,
+      "learning_rate": 1.4138817480719796e-05,
+      "loss": 2.2181,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7235235106141928,
+      "learning_rate": 1.4267352185089974e-05,
+      "loss": 2.2708,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.712692858685245,
+      "learning_rate": 1.4395886889460156e-05,
+      "loss": 2.2478,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7718128085655253,
+      "learning_rate": 1.4524421593830334e-05,
+      "loss": 2.3759,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7988679756765087,
+      "learning_rate": 1.4652956298200515e-05,
+      "loss": 2.2888,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7682800261550204,
+      "learning_rate": 1.4781491002570694e-05,
+      "loss": 2.2442,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8174401345904406,
+      "learning_rate": 1.4910025706940875e-05,
+      "loss": 2.2324,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7337342640199395,
+      "learning_rate": 1.5038560411311053e-05,
+      "loss": 2.373,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7233248531945753,
+      "learning_rate": 1.5167095115681235e-05,
+      "loss": 2.2296,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8951712559300193,
+      "learning_rate": 1.5295629820051416e-05,
+      "loss": 2.2088,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7515945129280316,
+      "learning_rate": 1.5424164524421593e-05,
+      "loss": 2.2358,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8472798169972598,
+      "learning_rate": 1.5552699228791776e-05,
+      "loss": 2.3821,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.815944237925127,
+      "learning_rate": 1.5681233933161953e-05,
+      "loss": 2.2392,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7637159644660926,
+      "learning_rate": 1.5809768637532136e-05,
+      "loss": 2.2272,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8367666448646426,
+      "learning_rate": 1.5938303341902313e-05,
+      "loss": 2.4248,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7114587818097278,
+      "learning_rate": 1.6066838046272496e-05,
+      "loss": 2.2506,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8019529997505153,
+      "learning_rate": 1.6195372750642672e-05,
+      "loss": 2.2143,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.832489195816639,
+      "learning_rate": 1.6323907455012856e-05,
+      "loss": 2.2487,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8340810174967511,
+      "learning_rate": 1.6452442159383032e-05,
+      "loss": 2.2134,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7424160283345604,
+      "learning_rate": 1.6580976863753215e-05,
+      "loss": 2.391,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8031178193982584,
+      "learning_rate": 1.6709511568123395e-05,
+      "loss": 2.2064,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7399645111812496,
+      "learning_rate": 1.6838046272493575e-05,
+      "loss": 2.2394,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8269335128138091,
+      "learning_rate": 1.6966580976863755e-05,
+      "loss": 2.2153,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7693441703791892,
+      "learning_rate": 1.7095115681233935e-05,
+      "loss": 2.3841,
+      "step": 133
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7676822031426157,
+      "learning_rate": 1.7223650385604115e-05,
+      "loss": 2.1814,
+      "step": 134
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7804318459497673,
+      "learning_rate": 1.7352185089974295e-05,
+      "loss": 2.2391,
+      "step": 135
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7438932382567364,
+      "learning_rate": 1.7480719794344475e-05,
+      "loss": 2.3838,
+      "step": 136
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.712586081466253,
+      "learning_rate": 1.7609254498714654e-05,
+      "loss": 2.2613,
+      "step": 137
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8600477181581904,
+      "learning_rate": 1.7737789203084834e-05,
+      "loss": 2.1858,
+      "step": 138
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7530114647052649,
+      "learning_rate": 1.7866323907455014e-05,
+      "loss": 2.2505,
+      "step": 139
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.803068386434486,
+      "learning_rate": 1.7994858611825194e-05,
+      "loss": 2.1718,
+      "step": 140
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8325994802276592,
+      "learning_rate": 1.8123393316195374e-05,
+      "loss": 2.4214,
+      "step": 141
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.749460914628141,
+      "learning_rate": 1.8251928020565554e-05,
+      "loss": 2.2305,
+      "step": 142
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7532191238420938,
+      "learning_rate": 1.8380462724935734e-05,
+      "loss": 2.23,
+      "step": 143
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9118671439579568,
+      "learning_rate": 1.8508997429305914e-05,
+      "loss": 2.2024,
+      "step": 144
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7432233383656482,
+      "learning_rate": 1.8637532133676093e-05,
+      "loss": 2.3729,
+      "step": 145
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8038797558993938,
+      "learning_rate": 1.8766066838046273e-05,
+      "loss": 2.2352,
+      "step": 146
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8068803892120634,
+      "learning_rate": 1.8894601542416453e-05,
+      "loss": 2.1927,
+      "step": 147
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7645221466645382,
+      "learning_rate": 1.9023136246786633e-05,
+      "loss": 2.2319,
+      "step": 148
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.725793203223981,
+      "learning_rate": 1.9151670951156813e-05,
+      "loss": 2.3541,
+      "step": 149
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8255924430645825,
+      "learning_rate": 1.9280205655526993e-05,
+      "loss": 2.1842,
+      "step": 150
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8265884765552166,
+      "learning_rate": 1.9408740359897173e-05,
+      "loss": 2.1864,
+      "step": 151
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8378283781456477,
+      "learning_rate": 1.9537275064267353e-05,
+      "loss": 2.2437,
+      "step": 152
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7598232159202175,
+      "learning_rate": 1.9665809768637533e-05,
+      "loss": 2.4008,
+      "step": 153
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8257843017410677,
+      "learning_rate": 1.9794344473007716e-05,
+      "loss": 2.2326,
+      "step": 154
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7753745188253554,
+      "learning_rate": 1.9922879177377892e-05,
+      "loss": 2.1824,
+      "step": 155
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7138011035699597,
+      "learning_rate": 2.0051413881748076e-05,
+      "loss": 2.2453,
+      "step": 156
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.774284079807111,
+      "learning_rate": 2.0179948586118252e-05,
+      "loss": 2.3797,
+      "step": 157
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8633054577213001,
+      "learning_rate": 2.0308483290488432e-05,
+      "loss": 2.1679,
+      "step": 158
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.805168376446495,
+      "learning_rate": 2.0437017994858612e-05,
+      "loss": 2.1878,
+      "step": 159
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7921651051962181,
+      "learning_rate": 2.0565552699228792e-05,
+      "loss": 2.1983,
+      "step": 160
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8062833144635174,
+      "learning_rate": 2.069408740359897e-05,
+      "loss": 2.358,
+      "step": 161
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.752102994062018,
+      "learning_rate": 2.082262210796915e-05,
+      "loss": 2.239,
+      "step": 162
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8531593906546486,
+      "learning_rate": 2.095115681233933e-05,
+      "loss": 2.2306,
+      "step": 163
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8449248806805474,
+      "learning_rate": 2.107969151670951e-05,
+      "loss": 2.1964,
+      "step": 164
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.842256756711545,
+      "learning_rate": 2.120822622107969e-05,
+      "loss": 2.413,
+      "step": 165
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8473662940224512,
+      "learning_rate": 2.133676092544987e-05,
+      "loss": 2.222,
+      "step": 166
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8042223052263614,
+      "learning_rate": 2.1465295629820054e-05,
+      "loss": 2.2017,
+      "step": 167
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7950694658117617,
+      "learning_rate": 2.159383033419023e-05,
+      "loss": 2.1971,
+      "step": 168
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7704212442597661,
+      "learning_rate": 2.1722365038560414e-05,
+      "loss": 2.3261,
+      "step": 169
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9416442899451295,
+      "learning_rate": 2.185089974293059e-05,
+      "loss": 2.1464,
+      "step": 170
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8068008414002802,
+      "learning_rate": 2.1979434447300774e-05,
+      "loss": 2.2209,
+      "step": 171
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8974292527858956,
+      "learning_rate": 2.210796915167095e-05,
+      "loss": 2.1655,
+      "step": 172
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8479575723898999,
+      "learning_rate": 2.2236503856041134e-05,
+      "loss": 2.3451,
+      "step": 173
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7986356237148048,
+      "learning_rate": 2.236503856041131e-05,
+      "loss": 2.2272,
+      "step": 174
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8286607794795989,
+      "learning_rate": 2.2493573264781493e-05,
+      "loss": 2.1686,
+      "step": 175
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.954741370859446,
+      "learning_rate": 2.262210796915167e-05,
+      "loss": 2.2007,
+      "step": 176
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8323671581646224,
+      "learning_rate": 2.2750642673521853e-05,
+      "loss": 2.3998,
+      "step": 177
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8806843936190892,
+      "learning_rate": 2.2879177377892033e-05,
+      "loss": 2.1764,
+      "step": 178
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.857802849990949,
+      "learning_rate": 2.3007712082262213e-05,
+      "loss": 2.183,
+      "step": 179
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8708433848297302,
+      "learning_rate": 2.3136246786632393e-05,
+      "loss": 2.1778,
+      "step": 180
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.840412034652977,
+      "learning_rate": 2.3264781491002573e-05,
+      "loss": 2.352,
+      "step": 181
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9344075211623188,
+      "learning_rate": 2.3393316195372753e-05,
+      "loss": 2.17,
+      "step": 182
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8812025888871806,
+      "learning_rate": 2.3521850899742933e-05,
+      "loss": 2.152,
+      "step": 183
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9509939865879609,
+      "learning_rate": 2.3650385604113112e-05,
+      "loss": 2.195,
+      "step": 184
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9945527159093788,
+      "learning_rate": 2.3778920308483292e-05,
+      "loss": 2.3279,
+      "step": 185
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.886769605585205,
+      "learning_rate": 2.3907455012853472e-05,
+      "loss": 2.1942,
+      "step": 186
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8959021518910648,
+      "learning_rate": 2.403598971722365e-05,
+      "loss": 2.2206,
+      "step": 187
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1859858907128553,
+      "learning_rate": 2.4164524421593832e-05,
+      "loss": 2.1445,
+      "step": 188
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0094984017630166,
+      "learning_rate": 2.4293059125964012e-05,
+      "loss": 2.3268,
+      "step": 189
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9238777016137953,
+      "learning_rate": 2.4421593830334192e-05,
+      "loss": 2.1705,
+      "step": 190
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0038666111133774,
+      "learning_rate": 2.455012853470437e-05,
+      "loss": 2.1728,
+      "step": 191
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9933981749687495,
+      "learning_rate": 2.467866323907455e-05,
+      "loss": 2.1753,
+      "step": 192
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8109137074634817,
+      "learning_rate": 2.480719794344473e-05,
+      "loss": 2.3185,
+      "step": 193
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0106594923453656,
+      "learning_rate": 2.493573264781491e-05,
+      "loss": 2.1354,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9243365227367555,
+      "learning_rate": 2.5064267352185088e-05,
+      "loss": 2.1799,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9321626469181074,
+      "learning_rate": 2.519280205655527e-05,
+      "loss": 2.167,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8646693304018523,
+      "learning_rate": 2.532133676092545e-05,
+      "loss": 2.347,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8372559532358477,
+      "learning_rate": 2.5449871465295634e-05,
+      "loss": 2.1813,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8396683369773447,
+      "learning_rate": 2.5578406169665807e-05,
+      "loss": 2.223,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8817676829226687,
+      "learning_rate": 2.570694087403599e-05,
+      "loss": 2.1483,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.776459798803132,
+      "learning_rate": 2.583547557840617e-05,
+      "loss": 2.3365,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8084753660969847,
+      "learning_rate": 2.5964010282776354e-05,
+      "loss": 2.1554,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8764255240490545,
+      "learning_rate": 2.6092544987146534e-05,
+      "loss": 2.1585,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8577833790854885,
+      "learning_rate": 2.622107969151671e-05,
+      "loss": 2.202,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8781993404197235,
+      "learning_rate": 2.634961439588689e-05,
+      "loss": 2.3809,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.889180704863445,
+      "learning_rate": 2.647814910025707e-05,
+      "loss": 2.1945,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9425559892998966,
+      "learning_rate": 2.6606683804627253e-05,
+      "loss": 2.1552,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.000696328058695,
+      "learning_rate": 2.673521850899743e-05,
+      "loss": 2.1882,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8298721511261294,
+      "learning_rate": 2.686375321336761e-05,
+      "loss": 2.3159,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8573486621291903,
+      "learning_rate": 2.699228791773779e-05,
+      "loss": 2.1083,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9666676202109203,
+      "learning_rate": 2.7120822622107973e-05,
+      "loss": 2.1577,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7871256442648927,
+      "learning_rate": 2.724935732647815e-05,
+      "loss": 2.2281,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8433162502641173,
+      "learning_rate": 2.737789203084833e-05,
+      "loss": 2.3337,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9560937954519017,
+      "learning_rate": 2.750642673521851e-05,
+      "loss": 2.1453,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9238370998287431,
+      "learning_rate": 2.7634961439588692e-05,
+      "loss": 2.1703,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9005431888321329,
+      "learning_rate": 2.7763496143958872e-05,
+      "loss": 2.1396,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9048485134303587,
+      "learning_rate": 2.789203084832905e-05,
+      "loss": 2.3822,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7455479070037024,
+      "learning_rate": 2.802056555269923e-05,
+      "loss": 2.1846,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.830115199097991,
+      "learning_rate": 2.8149100257069412e-05,
+      "loss": 2.1473,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8636395035972432,
+      "learning_rate": 2.827763496143959e-05,
+      "loss": 2.1965,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8054978869881467,
+      "learning_rate": 2.8406169665809768e-05,
+      "loss": 2.3366,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8681404882413228,
+      "learning_rate": 2.8534704370179948e-05,
+      "loss": 2.1337,
+      "step": 222
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8927113650861616,
+      "learning_rate": 2.866323907455013e-05,
+      "loss": 2.1411,
+      "step": 223
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.75671445839545,
+      "learning_rate": 2.879177377892031e-05,
+      "loss": 2.1998,
+      "step": 224
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7972525417440803,
+      "learning_rate": 2.892030848329049e-05,
+      "loss": 2.3657,
+      "step": 225
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8208457285454869,
+      "learning_rate": 2.9048843187660668e-05,
+      "loss": 2.1656,
+      "step": 226
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.815931009332894,
+      "learning_rate": 2.917737789203085e-05,
+      "loss": 2.1466,
+      "step": 227
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8193856937777181,
+      "learning_rate": 2.930591259640103e-05,
+      "loss": 2.1581,
+      "step": 228
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8665210116298263,
+      "learning_rate": 2.943444730077121e-05,
+      "loss": 2.3562,
+      "step": 229
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7522637518379005,
+      "learning_rate": 2.9562982005141387e-05,
+      "loss": 2.1879,
+      "step": 230
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8082362465992756,
+      "learning_rate": 2.969151670951157e-05,
+      "loss": 2.1543,
+      "step": 231
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8964399948187011,
+      "learning_rate": 2.982005141388175e-05,
+      "loss": 2.1378,
+      "step": 232
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8247864590624835,
+      "learning_rate": 2.994858611825193e-05,
+      "loss": 2.3513,
+      "step": 233
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8703640276954714,
+      "learning_rate": 3.0077120822622107e-05,
+      "loss": 2.1312,
+      "step": 234
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9344182165770607,
+      "learning_rate": 3.0205655526992287e-05,
+      "loss": 2.1212,
+      "step": 235
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8699682147643341,
+      "learning_rate": 3.033419023136247e-05,
+      "loss": 2.1938,
+      "step": 236
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9682396468502663,
+      "learning_rate": 3.046272493573265e-05,
+      "loss": 2.3009,
+      "step": 237
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9204904605625427,
+      "learning_rate": 3.059125964010283e-05,
+      "loss": 2.1922,
+      "step": 238
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9957959750973201,
+      "learning_rate": 3.0719794344473006e-05,
+      "loss": 2.1544,
+      "step": 239
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8798991594277322,
+      "learning_rate": 3.0848329048843186e-05,
+      "loss": 2.1357,
+      "step": 240
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7840988764470934,
+      "learning_rate": 3.097686375321337e-05,
+      "loss": 2.3482,
+      "step": 241
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.851310682314147,
+      "learning_rate": 3.110539845758355e-05,
+      "loss": 2.1685,
+      "step": 242
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8456942784305319,
+      "learning_rate": 3.1233933161953726e-05,
+      "loss": 2.2,
+      "step": 243
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8514480384706256,
+      "learning_rate": 3.1362467866323906e-05,
+      "loss": 2.1178,
+      "step": 244
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9099053986922037,
+      "learning_rate": 3.149100257069409e-05,
+      "loss": 2.2845,
+      "step": 245
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8279899922425604,
+      "learning_rate": 3.161953727506427e-05,
+      "loss": 2.1725,
+      "step": 246
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8313809058230442,
+      "learning_rate": 3.1748071979434445e-05,
+      "loss": 2.1451,
+      "step": 247
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9378075178520656,
+      "learning_rate": 3.1876606683804625e-05,
+      "loss": 2.1274,
+      "step": 248
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7622517964462712,
+      "learning_rate": 3.200514138817481e-05,
+      "loss": 2.1685,
+      "step": 249
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8909518195269749,
+      "learning_rate": 3.213367609254499e-05,
+      "loss": 2.3435,
+      "step": 250
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9650547266330559,
+      "learning_rate": 3.226221079691517e-05,
+      "loss": 2.1889,
+      "step": 251
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8733172155639845,
+      "learning_rate": 3.2390745501285345e-05,
+      "loss": 2.1284,
+      "step": 252
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0052534483850306,
+      "learning_rate": 3.251928020565553e-05,
+      "loss": 2.3263,
+      "step": 253
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8904747276526405,
+      "learning_rate": 3.264781491002571e-05,
+      "loss": 2.1521,
+      "step": 254
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9828398508170662,
+      "learning_rate": 3.277634961439589e-05,
+      "loss": 2.1533,
+      "step": 255
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9928896583152526,
+      "learning_rate": 3.2904884318766064e-05,
+      "loss": 2.1397,
+      "step": 256
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.94910024758784,
+      "learning_rate": 3.3033419023136244e-05,
+      "loss": 2.3181,
+      "step": 257
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9210982285355511,
+      "learning_rate": 3.316195372750643e-05,
+      "loss": 2.1382,
+      "step": 258
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8893051305631425,
+      "learning_rate": 3.329048843187661e-05,
+      "loss": 2.1345,
+      "step": 259
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0069840917191275,
+      "learning_rate": 3.341902313624679e-05,
+      "loss": 2.0934,
+      "step": 260
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8144084792935744,
+      "learning_rate": 3.3547557840616964e-05,
+      "loss": 2.183,
+      "step": 261
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8894335034020656,
+      "learning_rate": 3.367609254498715e-05,
+      "loss": 2.2968,
+      "step": 262
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9737452255198756,
+      "learning_rate": 3.380462724935733e-05,
+      "loss": 2.0878,
+      "step": 263
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9534263584781247,
+      "learning_rate": 3.393316195372751e-05,
+      "loss": 2.1539,
+      "step": 264
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8728334944976406,
+      "learning_rate": 3.406169665809768e-05,
+      "loss": 2.3197,
+      "step": 265
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8775640232734667,
+      "learning_rate": 3.419023136246787e-05,
+      "loss": 2.1279,
+      "step": 266
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0188555166573248,
+      "learning_rate": 3.431876606683805e-05,
+      "loss": 2.1845,
+      "step": 267
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9484249961824899,
+      "learning_rate": 3.444730077120823e-05,
+      "loss": 2.1497,
+      "step": 268
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8124571658973927,
+      "learning_rate": 3.45758354755784e-05,
+      "loss": 2.3162,
+      "step": 269
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0071685392965257,
+      "learning_rate": 3.470437017994859e-05,
+      "loss": 2.1484,
+      "step": 270
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8037777173426008,
+      "learning_rate": 3.483290488431877e-05,
+      "loss": 2.1098,
+      "step": 271
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9498689277672548,
+      "learning_rate": 3.496143958868895e-05,
+      "loss": 2.1618,
+      "step": 272
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8346628375739239,
+      "learning_rate": 3.508997429305913e-05,
+      "loss": 2.3119,
+      "step": 273
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.810258621972297,
+      "learning_rate": 3.521850899742931e-05,
+      "loss": 2.16,
+      "step": 274
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9832139792235742,
+      "learning_rate": 3.534704370179949e-05,
+      "loss": 2.0887,
+      "step": 275
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.817191678457258,
+      "learning_rate": 3.547557840616967e-05,
+      "loss": 2.1241,
+      "step": 276
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8660000347947862,
+      "learning_rate": 3.560411311053985e-05,
+      "loss": 2.3138,
+      "step": 277
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8149833299945981,
+      "learning_rate": 3.573264781491003e-05,
+      "loss": 2.1002,
+      "step": 278
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8700381938745791,
+      "learning_rate": 3.586118251928021e-05,
+      "loss": 2.1454,
+      "step": 279
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8608244430225308,
+      "learning_rate": 3.598971722365039e-05,
+      "loss": 2.1345,
+      "step": 280
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8092453261614396,
+      "learning_rate": 3.611825192802057e-05,
+      "loss": 2.1237,
+      "step": 281
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9037228073990298,
+      "learning_rate": 3.624678663239075e-05,
+      "loss": 2.3116,
+      "step": 282
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8225646743470585,
+      "learning_rate": 3.637532133676093e-05,
+      "loss": 2.1258,
+      "step": 283
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8619318488594805,
+      "learning_rate": 3.650385604113111e-05,
+      "loss": 2.1242,
+      "step": 284
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8155844239593858,
+      "learning_rate": 3.663239074550129e-05,
+      "loss": 2.2905,
+      "step": 285
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8589412595483565,
+      "learning_rate": 3.676092544987147e-05,
+      "loss": 2.1385,
+      "step": 286
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9899722294961518,
+      "learning_rate": 3.688946015424165e-05,
+      "loss": 2.0781,
+      "step": 287
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8877622740079876,
+      "learning_rate": 3.701799485861183e-05,
+      "loss": 2.1446,
+      "step": 288
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.05763635838708,
+      "learning_rate": 3.714652956298201e-05,
+      "loss": 2.3216,
+      "step": 289
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8500307408568465,
+      "learning_rate": 3.727506426735219e-05,
+      "loss": 2.0983,
+      "step": 290
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9452587122646533,
+      "learning_rate": 3.740359897172237e-05,
+      "loss": 2.165,
+      "step": 291
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8464230999175705,
+      "learning_rate": 3.753213367609255e-05,
+      "loss": 2.1721,
+      "step": 292
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8304922226407954,
+      "learning_rate": 3.766066838046273e-05,
+      "loss": 2.1029,
+      "step": 293
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9372525199525727,
+      "learning_rate": 3.7789203084832907e-05,
+      "loss": 2.2473,
+      "step": 294
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.827681917055771,
+      "learning_rate": 3.7917737789203086e-05,
+      "loss": 2.1189,
+      "step": 295
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9080064957599923,
+      "learning_rate": 3.8046272493573266e-05,
+      "loss": 2.1084,
+      "step": 296
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7999667851413097,
+      "learning_rate": 3.8174807197943446e-05,
+      "loss": 2.2356,
+      "step": 297
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.806392404946158,
+      "learning_rate": 3.8303341902313626e-05,
+      "loss": 2.1273,
+      "step": 298
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8292438743068946,
+      "learning_rate": 3.8431876606683806e-05,
+      "loss": 2.1474,
+      "step": 299
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9100527586708572,
+      "learning_rate": 3.8560411311053986e-05,
+      "loss": 2.097,
+      "step": 300
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8964113006938416,
+      "learning_rate": 3.8688946015424166e-05,
+      "loss": 2.2736,
+      "step": 301
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8617871886772082,
+      "learning_rate": 3.8817480719794346e-05,
+      "loss": 2.1159,
+      "step": 302
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8487198459931585,
+      "learning_rate": 3.8946015424164526e-05,
+      "loss": 2.1319,
+      "step": 303
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8677515989974598,
+      "learning_rate": 3.9074550128534705e-05,
+      "loss": 2.1331,
+      "step": 304
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7354985725581981,
+      "learning_rate": 3.9203084832904885e-05,
+      "loss": 2.1485,
+      "step": 305
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8500310764927246,
+      "learning_rate": 3.9331619537275065e-05,
+      "loss": 2.2693,
+      "step": 306
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8703580385080903,
+      "learning_rate": 3.9460154241645245e-05,
+      "loss": 2.1257,
+      "step": 307
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8518917215616878,
+      "learning_rate": 3.958868894601543e-05,
+      "loss": 2.097,
+      "step": 308
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8639792676320875,
+      "learning_rate": 3.9717223650385605e-05,
+      "loss": 2.2755,
+      "step": 309
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8303638114412453,
+      "learning_rate": 3.9845758354755785e-05,
+      "loss": 2.1282,
+      "step": 310
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7745614314235371,
+      "learning_rate": 3.9974293059125965e-05,
+      "loss": 2.0952,
+      "step": 311
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9118032615071021,
+      "learning_rate": 4.010282776349615e-05,
+      "loss": 2.1241,
+      "step": 312
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.847354847594381,
+      "learning_rate": 4.0231362467866324e-05,
+      "loss": 2.0603,
+      "step": 313
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8909382340071971,
+      "learning_rate": 4.0359897172236504e-05,
+      "loss": 2.2731,
+      "step": 314
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8614456618939683,
+      "learning_rate": 4.0488431876606684e-05,
+      "loss": 2.0796,
+      "step": 315
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8677412969107535,
+      "learning_rate": 4.0616966580976864e-05,
+      "loss": 2.1428,
+      "step": 316
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8263118413678917,
+      "learning_rate": 4.0745501285347044e-05,
+      "loss": 2.1178,
+      "step": 317
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7926581803625108,
+      "learning_rate": 4.0874035989717224e-05,
+      "loss": 2.3389,
+      "step": 318
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9553198751918666,
+      "learning_rate": 4.1002570694087404e-05,
+      "loss": 2.0861,
+      "step": 319
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.912762246066301,
+      "learning_rate": 4.1131105398457584e-05,
+      "loss": 2.1066,
+      "step": 320
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8088595278696781,
+      "learning_rate": 4.125964010282777e-05,
+      "loss": 2.2718,
+      "step": 321
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9162821547994565,
+      "learning_rate": 4.138817480719794e-05,
+      "loss": 2.1133,
+      "step": 322
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7552532762904902,
+      "learning_rate": 4.151670951156812e-05,
+      "loss": 2.1532,
+      "step": 323
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8563933631964012,
+      "learning_rate": 4.16452442159383e-05,
+      "loss": 2.1168,
+      "step": 324
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9445953598016068,
+      "learning_rate": 4.177377892030849e-05,
+      "loss": 2.0495,
+      "step": 325
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7547707013070538,
+      "learning_rate": 4.190231362467866e-05,
+      "loss": 2.2658,
+      "step": 326
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8198409608689639,
+      "learning_rate": 4.203084832904884e-05,
+      "loss": 2.068,
+      "step": 327
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9572399442192118,
+      "learning_rate": 4.215938303341902e-05,
+      "loss": 2.0412,
+      "step": 328
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8881850156329527,
+      "learning_rate": 4.228791773778921e-05,
+      "loss": 2.2808,
+      "step": 329
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7762157108582601,
+      "learning_rate": 4.241645244215938e-05,
+      "loss": 2.1584,
+      "step": 330
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8375757796062334,
+      "learning_rate": 4.254498714652956e-05,
+      "loss": 2.0602,
+      "step": 331
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9283286049859709,
+      "learning_rate": 4.267352185089974e-05,
+      "loss": 2.0707,
+      "step": 332
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7358818169110288,
+      "learning_rate": 4.280205655526993e-05,
+      "loss": 2.1139,
+      "step": 333
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9602188243305393,
+      "learning_rate": 4.293059125964011e-05,
+      "loss": 2.2886,
+      "step": 334
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7984682553531695,
+      "learning_rate": 4.305912596401028e-05,
+      "loss": 2.0655,
+      "step": 335
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8165703562307497,
+      "learning_rate": 4.318766066838046e-05,
+      "loss": 2.1523,
+      "step": 336
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8354262339886275,
+      "learning_rate": 4.331619537275065e-05,
+      "loss": 2.0885,
+      "step": 337
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8128764513509886,
+      "learning_rate": 4.344473007712083e-05,
+      "loss": 2.2437,
+      "step": 338
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.825527774423894,
+      "learning_rate": 4.3573264781491e-05,
+      "loss": 2.0837,
+      "step": 339
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8044921835942778,
+      "learning_rate": 4.370179948586118e-05,
+      "loss": 2.0612,
+      "step": 340
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8433194570271689,
+      "learning_rate": 4.383033419023137e-05,
+      "loss": 2.2951,
+      "step": 341
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7738317605120996,
+      "learning_rate": 4.395886889460155e-05,
+      "loss": 2.1568,
+      "step": 342
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8012292417236062,
+      "learning_rate": 4.408740359897173e-05,
+      "loss": 2.0523,
+      "step": 343
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8473619987616035,
+      "learning_rate": 4.42159383033419e-05,
+      "loss": 2.0864,
+      "step": 344
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8056024692378703,
+      "learning_rate": 4.434447300771208e-05,
+      "loss": 2.1009,
+      "step": 345
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.837167737608196,
+      "learning_rate": 4.447300771208227e-05,
+      "loss": 2.279,
+      "step": 346
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7570582597052302,
+      "learning_rate": 4.460154241645245e-05,
+      "loss": 2.0903,
+      "step": 347
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7519677905577871,
+      "learning_rate": 4.473007712082262e-05,
+      "loss": 2.1586,
+      "step": 348
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7740784841521778,
+      "learning_rate": 4.48586118251928e-05,
+      "loss": 2.0765,
+      "step": 349
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7368430313504458,
+      "learning_rate": 4.498714652956299e-05,
+      "loss": 2.2834,
+      "step": 350
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8693554172654819,
+      "learning_rate": 4.511568123393317e-05,
+      "loss": 2.0676,
+      "step": 351
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.799719553685666,
+      "learning_rate": 4.524421593830334e-05,
+      "loss": 2.09,
+      "step": 352
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9054201750990999,
+      "learning_rate": 4.537275064267352e-05,
+      "loss": 2.2791,
+      "step": 353
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9016969309095005,
+      "learning_rate": 4.5501285347043706e-05,
+      "loss": 2.1525,
+      "step": 354
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9191063043798773,
+      "learning_rate": 4.5629820051413886e-05,
+      "loss": 2.0865,
+      "step": 355
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.913842710979489,
+      "learning_rate": 4.5758354755784066e-05,
+      "loss": 2.1075,
+      "step": 356
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7883323826783198,
+      "learning_rate": 4.588688946015424e-05,
+      "loss": 2.0127,
+      "step": 357
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9350454557578266,
+      "learning_rate": 4.6015424164524426e-05,
+      "loss": 2.3214,
+      "step": 358
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8518201446433175,
+      "learning_rate": 4.6143958868894606e-05,
+      "loss": 2.0453,
+      "step": 359
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8523334988175477,
+      "learning_rate": 4.6272493573264786e-05,
+      "loss": 2.0779,
+      "step": 360
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7039829534606079,
+      "learning_rate": 4.640102827763496e-05,
+      "loss": 2.1315,
+      "step": 361
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8297513553222922,
+      "learning_rate": 4.6529562982005145e-05,
+      "loss": 2.2753,
+      "step": 362
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7983605852683096,
+      "learning_rate": 4.6658097686375325e-05,
+      "loss": 2.0706,
+      "step": 363
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8058756890740276,
+      "learning_rate": 4.6786632390745505e-05,
+      "loss": 2.1042,
+      "step": 364
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8100279961052802,
+      "learning_rate": 4.691516709511568e-05,
+      "loss": 2.1182,
+      "step": 365
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8765908325678339,
+      "learning_rate": 4.7043701799485865e-05,
+      "loss": 2.2436,
+      "step": 366
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7767884797247849,
+      "learning_rate": 4.7172236503856045e-05,
+      "loss": 2.1113,
+      "step": 367
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8669188181701808,
+      "learning_rate": 4.7300771208226225e-05,
+      "loss": 2.0812,
+      "step": 368
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8013753736440375,
+      "learning_rate": 4.7429305912596405e-05,
+      "loss": 2.1107,
+      "step": 369
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8184550380863294,
+      "learning_rate": 4.7557840616966585e-05,
+      "loss": 2.2659,
+      "step": 370
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8284297373516081,
+      "learning_rate": 4.7686375321336764e-05,
+      "loss": 2.0811,
+      "step": 371
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7532731279753423,
+      "learning_rate": 4.7814910025706944e-05,
+      "loss": 2.1254,
+      "step": 372
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8640461323037059,
+      "learning_rate": 4.7943444730077124e-05,
+      "loss": 2.1264,
+      "step": 373
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.73475651413898,
+      "learning_rate": 4.80719794344473e-05,
+      "loss": 2.2675,
+      "step": 374
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7682674945788471,
+      "learning_rate": 4.8200514138817484e-05,
+      "loss": 2.082,
+      "step": 375
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8181447462931017,
+      "learning_rate": 4.8329048843187664e-05,
+      "loss": 2.0638,
+      "step": 376
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9344356888550188,
+      "learning_rate": 4.8457583547557844e-05,
+      "loss": 2.0894,
+      "step": 377
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7988065826780911,
+      "learning_rate": 4.8586118251928024e-05,
+      "loss": 2.2891,
+      "step": 378
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8682307708465136,
+      "learning_rate": 4.8714652956298204e-05,
+      "loss": 2.1165,
+      "step": 379
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.892808523649001,
+      "learning_rate": 4.8843187660668383e-05,
+      "loss": 2.1021,
+      "step": 380
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8630464626378955,
+      "learning_rate": 4.897172236503856e-05,
+      "loss": 2.075,
+      "step": 381
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8721985941703809,
+      "learning_rate": 4.910025706940874e-05,
+      "loss": 2.2786,
+      "step": 382
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7583571634181621,
+      "learning_rate": 4.922879177377892e-05,
+      "loss": 2.0996,
+      "step": 383
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8640644130799212,
+      "learning_rate": 4.93573264781491e-05,
+      "loss": 2.0894,
+      "step": 384
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7433709916772382,
+      "learning_rate": 4.948586118251928e-05,
+      "loss": 2.1565,
+      "step": 385
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8026005748344281,
+      "learning_rate": 4.961439588688946e-05,
+      "loss": 2.2824,
+      "step": 386
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.879464100607665,
+      "learning_rate": 4.974293059125964e-05,
+      "loss": 2.0902,
+      "step": 387
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9045204292307262,
+      "learning_rate": 4.987146529562982e-05,
+      "loss": 2.0909,
+      "step": 388
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9658830197996352,
+      "learning_rate": 5e-05,
+      "loss": 2.0984,
+      "step": 389
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8232653956471361,
+      "learning_rate": 4.9999999219573654e-05,
+      "loss": 2.2913,
+      "step": 390
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8628948057280135,
+      "learning_rate": 4.999999687829465e-05,
+      "loss": 2.0863,
+      "step": 391
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8100726439528145,
+      "learning_rate": 4.9999992976163134e-05,
+      "loss": 2.1178,
+      "step": 392
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7836585604140691,
+      "learning_rate": 4.9999987513179356e-05,
+      "loss": 2.0862,
+      "step": 393
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0861402659570965,
+      "learning_rate": 4.9999980489343654e-05,
+      "loss": 2.2711,
+      "step": 394
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0821659517085256,
+      "learning_rate": 4.999997190465647e-05,
+      "loss": 2.1257,
+      "step": 395
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7729306412833703,
+      "learning_rate": 4.999996175911834e-05,
+      "loss": 2.0677,
+      "step": 396
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8229142493691448,
+      "learning_rate": 4.9999950052729894e-05,
+      "loss": 2.0697,
+      "step": 397
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8472627694967156,
+      "learning_rate": 4.9999936785491864e-05,
+      "loss": 2.2627,
+      "step": 398
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8070808701616811,
+      "learning_rate": 4.9999921957405074e-05,
+      "loss": 2.0669,
+      "step": 399
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9981962830207696,
+      "learning_rate": 4.999990556847045e-05,
+      "loss": 2.0874,
+      "step": 400
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9552103813766676,
+      "learning_rate": 4.9999887618689026e-05,
+      "loss": 2.0531,
+      "step": 401
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8118047843911405,
+      "learning_rate": 4.9999868108061924e-05,
+      "loss": 2.27,
+      "step": 402
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0649515536668732,
+      "learning_rate": 4.999984703659034e-05,
+      "loss": 2.0751,
+      "step": 403
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8397503691400678,
+      "learning_rate": 4.9999824404275606e-05,
+      "loss": 2.1119,
+      "step": 404
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7943271780521511,
+      "learning_rate": 4.999980021111914e-05,
+      "loss": 2.0389,
+      "step": 405
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9842156607524565,
+      "learning_rate": 4.999977445712244e-05,
+      "loss": 2.2834,
+      "step": 406
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0710314643105574,
+      "learning_rate": 4.999974714228712e-05,
+      "loss": 2.1163,
+      "step": 407
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7490094165284531,
+      "learning_rate": 4.999971826661488e-05,
+      "loss": 2.0819,
+      "step": 408
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8996838239970464,
+      "learning_rate": 4.9999687830107534e-05,
+      "loss": 2.1127,
+      "step": 409
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.825954463778069,
+      "learning_rate": 4.9999655832766966e-05,
+      "loss": 2.2581,
+      "step": 410
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8143067743139568,
+      "learning_rate": 4.9999622274595195e-05,
+      "loss": 2.0784,
+      "step": 411
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.114100938748624,
+      "learning_rate": 4.9999587155594293e-05,
+      "loss": 2.0502,
+      "step": 412
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0480702527765249,
+      "learning_rate": 4.999955047576648e-05,
+      "loss": 2.0777,
+      "step": 413
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7625278706498451,
+      "learning_rate": 4.999951223511401e-05,
+      "loss": 2.2702,
+      "step": 414
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9427066331418958,
+      "learning_rate": 4.999947243363931e-05,
+      "loss": 2.1061,
+      "step": 415
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8437621763333893,
+      "learning_rate": 4.999943107134483e-05,
+      "loss": 2.1052,
+      "step": 416
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8712966312607735,
+      "learning_rate": 4.999938814823317e-05,
+      "loss": 2.1043,
+      "step": 417
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.01293211230257,
+      "learning_rate": 4.999934366430702e-05,
+      "loss": 2.2984,
+      "step": 418
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0342316114886378,
+      "learning_rate": 4.999929761956913e-05,
+      "loss": 2.0889,
+      "step": 419
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8716933736049707,
+      "learning_rate": 4.9999250014022404e-05,
+      "loss": 2.0762,
+      "step": 420
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.1097916543460231,
+      "learning_rate": 4.9999200847669795e-05,
+      "loss": 2.0562,
+      "step": 421
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9730110208891356,
+      "learning_rate": 4.999915012051437e-05,
+      "loss": 2.2521,
+      "step": 422
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8441898111175365,
+      "learning_rate": 4.9999097832559315e-05,
+      "loss": 2.1052,
+      "step": 423
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9843318050039119,
+      "learning_rate": 4.999904398380788e-05,
+      "loss": 2.0956,
+      "step": 424
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8381330910956554,
+      "learning_rate": 4.999898857426343e-05,
+      "loss": 2.0962,
+      "step": 425
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8238884586298355,
+      "learning_rate": 4.999893160392942e-05,
+      "loss": 2.2767,
+      "step": 426
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8387452914407116,
+      "learning_rate": 4.999887307280941e-05,
+      "loss": 2.0725,
+      "step": 427
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8065959427415655,
+      "learning_rate": 4.9998812980907066e-05,
+      "loss": 2.0702,
+      "step": 428
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9162430394232872,
+      "learning_rate": 4.999875132822612e-05,
+      "loss": 2.1245,
+      "step": 429
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7315352407959287,
+      "learning_rate": 4.9998688114770444e-05,
+      "loss": 2.2965,
+      "step": 430
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7526227424480181,
+      "learning_rate": 4.999862334054396e-05,
+      "loss": 2.0827,
+      "step": 431
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7982820462086706,
+      "learning_rate": 4.999855700555073e-05,
+      "loss": 2.083,
+      "step": 432
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8254111799129937,
+      "learning_rate": 4.9998489109794886e-05,
+      "loss": 2.0499,
+      "step": 433
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.943682417353492,
+      "learning_rate": 4.999841965328068e-05,
+      "loss": 2.224,
+      "step": 434
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8211262943400262,
+      "learning_rate": 4.9998348636012425e-05,
+      "loss": 2.1633,
+      "step": 435
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.894248237844582,
+      "learning_rate": 4.999827605799458e-05,
+      "loss": 2.0915,
+      "step": 436
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8498766710321624,
+      "learning_rate": 4.999820191923166e-05,
+      "loss": 2.1312,
+      "step": 437
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7753817699423873,
+      "learning_rate": 4.99981262197283e-05,
+      "loss": 2.2832,
+      "step": 438
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8617140213936059,
+      "learning_rate": 4.999804895948922e-05,
+      "loss": 2.0921,
+      "step": 439
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9247910752335331,
+      "learning_rate": 4.999797013851926e-05,
+      "loss": 2.0767,
+      "step": 440
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7291350620687566,
+      "learning_rate": 4.9997889756823316e-05,
+      "loss": 2.1226,
+      "step": 441
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8024549436229903,
+      "learning_rate": 4.999780781440643e-05,
+      "loss": 2.2603,
+      "step": 442
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8068057277949261,
+      "learning_rate": 4.999772431127371e-05,
+      "loss": 2.0221,
+      "step": 443
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7133210792997057,
+      "learning_rate": 4.999763924743036e-05,
+      "loss": 2.0713,
+      "step": 444
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7195987481563844,
+      "learning_rate": 4.99975526228817e-05,
+      "loss": 2.1268,
+      "step": 445
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8302340811057716,
+      "learning_rate": 4.9997464437633144e-05,
+      "loss": 2.2817,
+      "step": 446
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7290367502371367,
+      "learning_rate": 4.9997374691690194e-05,
+      "loss": 2.1175,
+      "step": 447
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7687906770699118,
+      "learning_rate": 4.9997283385058444e-05,
+      "loss": 2.0849,
+      "step": 448
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8398500397240121,
+      "learning_rate": 4.99971905177436e-05,
+      "loss": 2.0612,
+      "step": 449
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8304813759530085,
+      "learning_rate": 4.999709608975146e-05,
+      "loss": 2.2927,
+      "step": 450
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8440126339882029,
+      "learning_rate": 4.999700010108792e-05,
+      "loss": 2.06,
+      "step": 451
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9082574090154942,
+      "learning_rate": 4.999690255175898e-05,
+      "loss": 2.1177,
+      "step": 452
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7700450496803274,
+      "learning_rate": 4.9996803441770715e-05,
+      "loss": 2.0007,
+      "step": 453
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8853145379477192,
+      "learning_rate": 4.9996702771129336e-05,
+      "loss": 2.3271,
+      "step": 454
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8751604240901071,
+      "learning_rate": 4.9996600539841096e-05,
+      "loss": 2.0703,
+      "step": 455
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7817455875133111,
+      "learning_rate": 4.9996496747912404e-05,
+      "loss": 2.055,
+      "step": 456
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8914872877842717,
+      "learning_rate": 4.999639139534973e-05,
+      "loss": 2.0885,
+      "step": 457
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8627105516470674,
+      "learning_rate": 4.999628448215966e-05,
+      "loss": 2.2488,
+      "step": 458
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.94988901800165,
+      "learning_rate": 4.999617600834886e-05,
+      "loss": 2.0616,
+      "step": 459
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7449114684799212,
+      "learning_rate": 4.99960659739241e-05,
+      "loss": 2.1143,
+      "step": 460
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8984075913582436,
+      "learning_rate": 4.999595437889226e-05,
+      "loss": 2.0737,
+      "step": 461
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9637318125513891,
+      "learning_rate": 4.99958412232603e-05,
+      "loss": 2.2363,
+      "step": 462
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.734514391471459,
+      "learning_rate": 4.999572650703528e-05,
+      "loss": 2.0785,
+      "step": 463
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9640707929274376,
+      "learning_rate": 4.9995610230224375e-05,
+      "loss": 2.047,
+      "step": 464
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.958605900611534,
+      "learning_rate": 4.9995492392834844e-05,
+      "loss": 2.1262,
+      "step": 465
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7942870212565192,
+      "learning_rate": 4.999537299487403e-05,
+      "loss": 2.155,
+      "step": 466
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0091026436455823,
+      "learning_rate": 4.9995252036349403e-05,
+      "loss": 2.2253,
+      "step": 467
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7574886362522087,
+      "learning_rate": 4.999512951726851e-05,
+      "loss": 2.0819,
+      "step": 468
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9080976371587154,
+      "learning_rate": 4.999500543763899e-05,
+      "loss": 2.0416,
+      "step": 469
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.842101937580771,
+      "learning_rate": 4.9994879797468605e-05,
+      "loss": 2.216,
+      "step": 470
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7974925990151202,
+      "learning_rate": 4.999475259676519e-05,
+      "loss": 2.0701,
+      "step": 471
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9395515818029764,
+      "learning_rate": 4.999462383553669e-05,
+      "loss": 2.0932,
+      "step": 472
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7662249798342431,
+      "learning_rate": 4.999449351379115e-05,
+      "loss": 2.0613,
+      "step": 473
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9843705290372647,
+      "learning_rate": 4.999436163153669e-05,
+      "loss": 2.2254,
+      "step": 474
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7764718211067133,
+      "learning_rate": 4.999422818878155e-05,
+      "loss": 2.0745,
+      "step": 475
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7954718879826685,
+      "learning_rate": 4.999409318553408e-05,
+      "loss": 2.0587,
+      "step": 476
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8859752029518048,
+      "learning_rate": 4.999395662180269e-05,
+      "loss": 2.0857,
+      "step": 477
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7225096838504672,
+      "learning_rate": 4.99938184975959e-05,
+      "loss": 2.2386,
+      "step": 478
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.827492685348671,
+      "learning_rate": 4.9993678812922356e-05,
+      "loss": 2.1525,
+      "step": 479
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7561700992463559,
+      "learning_rate": 4.999353756779076e-05,
+      "loss": 2.0558,
+      "step": 480
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.803135801979098,
+      "learning_rate": 4.999339476220994e-05,
+      "loss": 2.0807,
+      "step": 481
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8385660241106662,
+      "learning_rate": 4.999325039618881e-05,
+      "loss": 2.2535,
+      "step": 482
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7978445504286632,
+      "learning_rate": 4.999310446973638e-05,
+      "loss": 2.0686,
+      "step": 483
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8051040540770208,
+      "learning_rate": 4.999295698286177e-05,
+      "loss": 2.1121,
+      "step": 484
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8936295716752525,
+      "learning_rate": 4.999280793557418e-05,
+      "loss": 2.1334,
+      "step": 485
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7439477016178776,
+      "learning_rate": 4.999265732788292e-05,
+      "loss": 2.2998,
+      "step": 486
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7612729488072276,
+      "learning_rate": 4.999250515979739e-05,
+      "loss": 2.0642,
+      "step": 487
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.823381507772441,
+      "learning_rate": 4.999235143132708e-05,
+      "loss": 2.0587,
+      "step": 488
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8818117233228686,
+      "learning_rate": 4.999219614248161e-05,
+      "loss": 2.1089,
+      "step": 489
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7946174227550551,
+      "learning_rate": 4.9992039293270655e-05,
+      "loss": 2.179,
+      "step": 490
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.808321536214124,
+      "learning_rate": 4.999188088370402e-05,
+      "loss": 2.1266,
+      "step": 491
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0304105346551413,
+      "learning_rate": 4.9991720913791604e-05,
+      "loss": 2.0502,
+      "step": 492
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8010317810616022,
+      "learning_rate": 4.9991559383543375e-05,
+      "loss": 2.0472,
+      "step": 493
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1288761195489778,
+      "learning_rate": 4.999139629296942e-05,
+      "loss": 2.2365,
+      "step": 494
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0966092443581594,
+      "learning_rate": 4.999123164207993e-05,
+      "loss": 2.1008,
+      "step": 495
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.872445613599377,
+      "learning_rate": 4.999106543088519e-05,
+      "loss": 2.0911,
+      "step": 496
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8399811132152307,
+      "learning_rate": 4.999089765939556e-05,
+      "loss": 2.0984,
+      "step": 497
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8157452182265927,
+      "learning_rate": 4.999072832762153e-05,
+      "loss": 2.0636,
+      "step": 498
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7913392243514016,
+      "learning_rate": 4.999055743557366e-05,
+      "loss": 2.2203,
+      "step": 499
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9573702209199872,
+      "learning_rate": 4.999038498326263e-05,
+      "loss": 2.0432,
+      "step": 500
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8972903977893034,
+      "learning_rate": 4.9990210970699204e-05,
+      "loss": 2.1009,
+      "step": 501
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8210524161654924,
+      "learning_rate": 4.999003539789424e-05,
+      "loss": 2.2795,
+      "step": 502
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7701369965347726,
+      "learning_rate": 4.9989858264858704e-05,
+      "loss": 2.1121,
+      "step": 503
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8218291496648268,
+      "learning_rate": 4.998967957160365e-05,
+      "loss": 2.0224,
+      "step": 504
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7381038693476569,
+      "learning_rate": 4.998949931814025e-05,
+      "loss": 2.0335,
+      "step": 505
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0549697916733558,
+      "learning_rate": 4.998931750447975e-05,
+      "loss": 2.2512,
+      "step": 506
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.862611490633934,
+      "learning_rate": 4.99891341306335e-05,
+      "loss": 2.0708,
+      "step": 507
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8681370223790387,
+      "learning_rate": 4.998894919661294e-05,
+      "loss": 2.0249,
+      "step": 508
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8435041001435267,
+      "learning_rate": 4.998876270242963e-05,
+      "loss": 2.1056,
+      "step": 509
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7242803843850976,
+      "learning_rate": 4.998857464809521e-05,
+      "loss": 2.1529,
+      "step": 510
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8768372330042844,
+      "learning_rate": 4.998838503362141e-05,
+      "loss": 2.2559,
+      "step": 511
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9657078745831267,
+      "learning_rate": 4.998819385902008e-05,
+      "loss": 2.0714,
+      "step": 512
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.779099751623594,
+      "learning_rate": 4.998800112430316e-05,
+      "loss": 2.035,
+      "step": 513
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.835395994110178,
+      "learning_rate": 4.9987806829482674e-05,
+      "loss": 2.2612,
+      "step": 514
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8113856584369281,
+      "learning_rate": 4.9987610974570755e-05,
+      "loss": 2.0451,
+      "step": 515
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7372958763898732,
+      "learning_rate": 4.9987413559579636e-05,
+      "loss": 2.1364,
+      "step": 516
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9354273122997846,
+      "learning_rate": 4.998721458452163e-05,
+      "loss": 2.0553,
+      "step": 517
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.054942908460834,
+      "learning_rate": 4.9987014049409166e-05,
+      "loss": 2.2769,
+      "step": 518
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7067115783890964,
+      "learning_rate": 4.998681195425477e-05,
+      "loss": 2.1116,
+      "step": 519
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8711277437669996,
+      "learning_rate": 4.998660829907105e-05,
+      "loss": 2.0566,
+      "step": 520
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7411338331491617,
+      "learning_rate": 4.998640308387074e-05,
+      "loss": 2.0834,
+      "step": 521
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7683344720794699,
+      "learning_rate": 4.998619630866662e-05,
+      "loss": 2.0922,
+      "step": 522
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.910371811868455,
+      "learning_rate": 4.998598797347163e-05,
+      "loss": 2.2432,
+      "step": 523
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8449364689376,
+      "learning_rate": 4.998577807829876e-05,
+      "loss": 2.1126,
+      "step": 524
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8074557482892698,
+      "learning_rate": 4.998556662316113e-05,
+      "loss": 2.0238,
+      "step": 525
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8265954745821756,
+      "learning_rate": 4.9985353608071924e-05,
+      "loss": 2.2513,
+      "step": 526
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9321021662758424,
+      "learning_rate": 4.998513903304445e-05,
+      "loss": 2.0778,
+      "step": 527
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7223864797102112,
+      "learning_rate": 4.998492289809211e-05,
+      "loss": 2.097,
+      "step": 528
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8450822121312791,
+      "learning_rate": 4.998470520322839e-05,
+      "loss": 2.0525,
+      "step": 529
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8239669241930608,
+      "learning_rate": 4.9984485948466885e-05,
+      "loss": 2.0569,
+      "step": 530
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7194351685241752,
+      "learning_rate": 4.9984265133821285e-05,
+      "loss": 2.2723,
+      "step": 531
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7797691788690546,
+      "learning_rate": 4.9984042759305375e-05,
+      "loss": 2.081,
+      "step": 532
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7645428016155795,
+      "learning_rate": 4.9983818824933034e-05,
+      "loss": 2.0396,
+      "step": 533
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7905174111791649,
+      "learning_rate": 4.998359333071825e-05,
+      "loss": 2.0939,
+      "step": 534
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8112360346491397,
+      "learning_rate": 4.99833662766751e-05,
+      "loss": 2.2677,
+      "step": 535
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7480168307173948,
+      "learning_rate": 4.9983137662817756e-05,
+      "loss": 2.0835,
+      "step": 536
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7638784106124105,
+      "learning_rate": 4.9982907489160495e-05,
+      "loss": 2.0679,
+      "step": 537
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7446809569697774,
+      "learning_rate": 4.998267575571769e-05,
+      "loss": 2.2101,
+      "step": 538
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8256424001709195,
+      "learning_rate": 4.99824424625038e-05,
+      "loss": 2.0693,
+      "step": 539
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0535945192683556,
+      "learning_rate": 4.99822076095334e-05,
+      "loss": 2.0769,
+      "step": 540
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7751979279518619,
+      "learning_rate": 4.9981971196821154e-05,
+      "loss": 2.0937,
+      "step": 541
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8516367696336125,
+      "learning_rate": 4.998173322438181e-05,
+      "loss": 2.0542,
+      "step": 542
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8712691538460843,
+      "learning_rate": 4.998149369223024e-05,
+      "loss": 2.2249,
+      "step": 543
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.798473658618515,
+      "learning_rate": 4.9981252600381396e-05,
+      "loss": 2.1002,
+      "step": 544
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9008487150199367,
+      "learning_rate": 4.998100994885032e-05,
+      "loss": 2.0498,
+      "step": 545
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9204863140701205,
+      "learning_rate": 4.9980765737652166e-05,
+      "loss": 2.2054,
+      "step": 546
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.734580972708866,
+      "learning_rate": 4.9980519966802185e-05,
+      "loss": 2.0943,
+      "step": 547
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8603391453781783,
+      "learning_rate": 4.998027263631573e-05,
+      "loss": 2.0992,
+      "step": 548
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0484663923127682,
+      "learning_rate": 4.998002374620821e-05,
+      "loss": 2.0548,
+      "step": 549
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8023831852440654,
+      "learning_rate": 4.9979773296495215e-05,
+      "loss": 2.0565,
+      "step": 550
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1112580955384295,
+      "learning_rate": 4.997952128719233e-05,
+      "loss": 2.2747,
+      "step": 551
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.841365712502595,
+      "learning_rate": 4.997926771831533e-05,
+      "loss": 2.0796,
+      "step": 552
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0290392672159288,
+      "learning_rate": 4.9979012589880016e-05,
+      "loss": 2.1119,
+      "step": 553
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8556226930497965,
+      "learning_rate": 4.997875590190233e-05,
+      "loss": 2.032,
+      "step": 554
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8675387493299548,
+      "learning_rate": 4.997849765439831e-05,
+      "loss": 2.2422,
+      "step": 555
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8380065173917949,
+      "learning_rate": 4.9978237847384056e-05,
+      "loss": 2.0659,
+      "step": 556
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9092738694777557,
+      "learning_rate": 4.997797648087581e-05,
+      "loss": 2.0318,
+      "step": 557
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8886722847886525,
+      "learning_rate": 4.997771355488987e-05,
+      "loss": 2.2693,
+      "step": 558
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8860830104895143,
+      "learning_rate": 4.997744906944267e-05,
+      "loss": 2.1173,
+      "step": 559
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7234059020940159,
+      "learning_rate": 4.997718302455071e-05,
+      "loss": 2.0836,
+      "step": 560
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8871051934248314,
+      "learning_rate": 4.99769154202306e-05,
+      "loss": 2.0477,
+      "step": 561
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7697819244263002,
+      "learning_rate": 4.997664625649906e-05,
+      "loss": 2.0255,
+      "step": 562
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8989888184792155,
+      "learning_rate": 4.997637553337289e-05,
+      "loss": 2.2321,
+      "step": 563
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9919396328084688,
+      "learning_rate": 4.997610325086898e-05,
+      "loss": 2.0319,
+      "step": 564
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7548152415429998,
+      "learning_rate": 4.997582940900435e-05,
+      "loss": 2.1393,
+      "step": 565
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1222777399950583,
+      "learning_rate": 4.9975554007796075e-05,
+      "loss": 2.0713,
+      "step": 566
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7481217633436353,
+      "learning_rate": 4.997527704726137e-05,
+      "loss": 2.2721,
+      "step": 567
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0241147818502496,
+      "learning_rate": 4.9974998527417516e-05,
+      "loss": 2.0464,
+      "step": 568
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9176588570449181,
+      "learning_rate": 4.9974718448281905e-05,
+      "loss": 2.0913,
+      "step": 569
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9124820593264301,
+      "learning_rate": 4.997443680987202e-05,
+      "loss": 2.2672,
+      "step": 570
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0638159802716098,
+      "learning_rate": 4.9974153612205455e-05,
+      "loss": 2.0846,
+      "step": 571
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.746867800549365,
+      "learning_rate": 4.997386885529988e-05,
+      "loss": 2.0898,
+      "step": 572
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8165204294101939,
+      "learning_rate": 4.997358253917307e-05,
+      "loss": 2.0445,
+      "step": 573
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9051826744388111,
+      "learning_rate": 4.9973294663842904e-05,
+      "loss": 2.1151,
+      "step": 574
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7631780637342404,
+      "learning_rate": 4.997300522932737e-05,
+      "loss": 2.202,
+      "step": 575
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9389529809623927,
+      "learning_rate": 4.997271423564453e-05,
+      "loss": 2.092,
+      "step": 576
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7585799286700476,
+      "learning_rate": 4.9972421682812544e-05,
+      "loss": 2.0455,
+      "step": 577
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8323795123262308,
+      "learning_rate": 4.9972127570849685e-05,
+      "loss": 2.1076,
+      "step": 578
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7759156795048144,
+      "learning_rate": 4.9971831899774314e-05,
+      "loss": 2.2282,
+      "step": 579
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7526626135025627,
+      "learning_rate": 4.997153466960489e-05,
+      "loss": 2.0582,
+      "step": 580
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7387212144657123,
+      "learning_rate": 4.997123588035997e-05,
+      "loss": 2.0538,
+      "step": 581
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7232414088638426,
+      "learning_rate": 4.997093553205822e-05,
+      "loss": 2.0179,
+      "step": 582
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8764033237826514,
+      "learning_rate": 4.9970633624718365e-05,
+      "loss": 2.2522,
+      "step": 583
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7979401619614107,
+      "learning_rate": 4.997033015835928e-05,
+      "loss": 2.1133,
+      "step": 584
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8163702227581208,
+      "learning_rate": 4.99700251329999e-05,
+      "loss": 2.0159,
+      "step": 585
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0210158783867431,
+      "learning_rate": 4.996971854865927e-05,
+      "loss": 2.0582,
+      "step": 586
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9062636761679665,
+      "learning_rate": 4.996941040535653e-05,
+      "loss": 2.2816,
+      "step": 587
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7584401047861605,
+      "learning_rate": 4.9969100703110925e-05,
+      "loss": 2.0102,
+      "step": 588
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8224547912898316,
+      "learning_rate": 4.99687894419418e-05,
+      "loss": 2.0369,
+      "step": 589
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7323426756947437,
+      "learning_rate": 4.996847662186856e-05,
+      "loss": 2.1031,
+      "step": 590
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7679761335970001,
+      "learning_rate": 4.9968162242910757e-05,
+      "loss": 2.1985,
+      "step": 591
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8608716612778514,
+      "learning_rate": 4.9967846305088005e-05,
+      "loss": 2.0491,
+      "step": 592
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7686714367519798,
+      "learning_rate": 4.996752880842005e-05,
+      "loss": 2.0833,
+      "step": 593
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8174475967419659,
+      "learning_rate": 4.9967209752926695e-05,
+      "loss": 2.03,
+      "step": 594
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8003719052594509,
+      "learning_rate": 4.996688913862787e-05,
+      "loss": 2.2281,
+      "step": 595
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6971591958556124,
+      "learning_rate": 4.9966566965543594e-05,
+      "loss": 2.1018,
+      "step": 596
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8152650553138222,
+      "learning_rate": 4.9966243233693965e-05,
+      "loss": 2.0052,
+      "step": 597
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7612005329221397,
+      "learning_rate": 4.996591794309922e-05,
+      "loss": 2.0542,
+      "step": 598
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8592121253849749,
+      "learning_rate": 4.996559109377965e-05,
+      "loss": 2.2939,
+      "step": 599
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7527600388607134,
+      "learning_rate": 4.996526268575566e-05,
+      "loss": 2.0373,
+      "step": 600
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8181432522229931,
+      "learning_rate": 4.9964932719047774e-05,
+      "loss": 2.0315,
+      "step": 601
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.728095249099527,
+      "learning_rate": 4.996460119367657e-05,
+      "loss": 2.247,
+      "step": 602
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.823861225954546,
+      "learning_rate": 4.996426810966276e-05,
+      "loss": 2.1257,
+      "step": 603
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8566471527004964,
+      "learning_rate": 4.9963933467027135e-05,
+      "loss": 2.0245,
+      "step": 604
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9018949958093674,
+      "learning_rate": 4.996359726579059e-05,
+      "loss": 2.0317,
+      "step": 605
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6934804251111074,
+      "learning_rate": 4.996325950597411e-05,
+      "loss": 2.0841,
+      "step": 606
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0238246791964976,
+      "learning_rate": 4.9962920187598794e-05,
+      "loss": 2.2802,
+      "step": 607
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7658273860821063,
+      "learning_rate": 4.996257931068582e-05,
+      "loss": 1.9985,
+      "step": 608
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8999920583122112,
+      "learning_rate": 4.996223687525647e-05,
+      "loss": 2.1551,
+      "step": 609
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7115632169492504,
+      "learning_rate": 4.9961892881332115e-05,
+      "loss": 2.0671,
+      "step": 610
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.844916054757255,
+      "learning_rate": 4.9961547328934255e-05,
+      "loss": 2.2263,
+      "step": 611
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.979185090718177,
+      "learning_rate": 4.996120021808445e-05,
+      "loss": 2.0705,
+      "step": 612
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8409447018634192,
+      "learning_rate": 4.996085154880437e-05,
+      "loss": 2.0975,
+      "step": 613
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9493599629933573,
+      "learning_rate": 4.9960501321115774e-05,
+      "loss": 2.0209,
+      "step": 614
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0160571035520876,
+      "learning_rate": 4.9960149535040555e-05,
+      "loss": 2.2452,
+      "step": 615
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8700336432888692,
+      "learning_rate": 4.995979619060065e-05,
+      "loss": 2.0839,
+      "step": 616
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9711781661645156,
+      "learning_rate": 4.995944128781814e-05,
+      "loss": 2.017,
+      "step": 617
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7326138468653915,
+      "learning_rate": 4.995908482671517e-05,
+      "loss": 2.0629,
+      "step": 618
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9086365312372525,
+      "learning_rate": 4.9958726807314005e-05,
+      "loss": 2.2077,
+      "step": 619
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7792561192390194,
+      "learning_rate": 4.995836722963699e-05,
+      "loss": 2.0191,
+      "step": 620
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7027386571563832,
+      "learning_rate": 4.995800609370658e-05,
+      "loss": 2.0948,
+      "step": 621
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8441521648038912,
+      "learning_rate": 4.9957643399545314e-05,
+      "loss": 2.031,
+      "step": 622
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7996128938507705,
+      "learning_rate": 4.995727914717584e-05,
+      "loss": 2.2353,
+      "step": 623
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8796829820860176,
+      "learning_rate": 4.9956913336620914e-05,
+      "loss": 2.0402,
+      "step": 624
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7531181917265757,
+      "learning_rate": 4.995654596790335e-05,
+      "loss": 2.0885,
+      "step": 625
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7631970803856661,
+      "learning_rate": 4.9956177041046104e-05,
+      "loss": 2.0634,
+      "step": 626
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8219772969939948,
+      "learning_rate": 4.995580655607221e-05,
+      "loss": 2.2877,
+      "step": 627
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7625813947059649,
+      "learning_rate": 4.995543451300477e-05,
+      "loss": 2.0404,
+      "step": 628
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9709920769976625,
+      "learning_rate": 4.995506091186705e-05,
+      "loss": 2.0751,
+      "step": 629
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.749938883680589,
+      "learning_rate": 4.995468575268235e-05,
+      "loss": 2.0558,
+      "step": 630
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8616586664600657,
+      "learning_rate": 4.9954309035474114e-05,
+      "loss": 2.2479,
+      "step": 631
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7260712477242162,
+      "learning_rate": 4.9953930760265845e-05,
+      "loss": 2.0243,
+      "step": 632
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7800301390801145,
+      "learning_rate": 4.995355092708116e-05,
+      "loss": 2.0381,
+      "step": 633
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.736937019428342,
+      "learning_rate": 4.9953169535943783e-05,
+      "loss": 2.111,
+      "step": 634
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7440417604194494,
+      "learning_rate": 4.995278658687752e-05,
+      "loss": 2.2512,
+      "step": 635
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9139757044768461,
+      "learning_rate": 4.995240207990629e-05,
+      "loss": 2.0686,
+      "step": 636
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8458488429428564,
+      "learning_rate": 4.995201601505408e-05,
+      "loss": 2.0943,
+      "step": 637
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8822602846091177,
+      "learning_rate": 4.9951628392345005e-05,
+      "loss": 2.083,
+      "step": 638
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.751130550292683,
+      "learning_rate": 4.9951239211803276e-05,
+      "loss": 2.2268,
+      "step": 639
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9418203537671516,
+      "learning_rate": 4.995084847345317e-05,
+      "loss": 2.1336,
+      "step": 640
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8359962392838421,
+      "learning_rate": 4.9950456177319106e-05,
+      "loss": 2.0612,
+      "step": 641
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.803935196710273,
+      "learning_rate": 4.995006232342555e-05,
+      "loss": 2.039,
+      "step": 642
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7842009919222735,
+      "learning_rate": 4.994966691179711e-05,
+      "loss": 2.2302,
+      "step": 643
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7721257625274404,
+      "learning_rate": 4.9949269942458474e-05,
+      "loss": 2.1023,
+      "step": 644
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7282385906513733,
+      "learning_rate": 4.994887141543442e-05,
+      "loss": 2.0157,
+      "step": 645
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7311720520935104,
+      "learning_rate": 4.9948471330749824e-05,
+      "loss": 2.0833,
+      "step": 646
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6827996303665989,
+      "learning_rate": 4.994806968842968e-05,
+      "loss": 2.2676,
+      "step": 647
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7254043660775278,
+      "learning_rate": 4.994766648849904e-05,
+      "loss": 2.0642,
+      "step": 648
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7423240797752403,
+      "learning_rate": 4.994726173098312e-05,
+      "loss": 2.0143,
+      "step": 649
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7168180420701663,
+      "learning_rate": 4.994685541590715e-05,
+      "loss": 2.0438,
+      "step": 650
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7072408166290011,
+      "learning_rate": 4.99464475432965e-05,
+      "loss": 2.2326,
+      "step": 651
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7075539052966868,
+      "learning_rate": 4.994603811317667e-05,
+      "loss": 2.1136,
+      "step": 652
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8316250582793244,
+      "learning_rate": 4.994562712557319e-05,
+      "loss": 2.0761,
+      "step": 653
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7940173536652535,
+      "learning_rate": 4.994521458051172e-05,
+      "loss": 2.0333,
+      "step": 654
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8849979608600422,
+      "learning_rate": 4.994480047801804e-05,
+      "loss": 2.2608,
+      "step": 655
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6930111884757989,
+      "learning_rate": 4.9944384818117985e-05,
+      "loss": 2.0892,
+      "step": 656
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8911855013612321,
+      "learning_rate": 4.994396760083751e-05,
+      "loss": 2.0042,
+      "step": 657
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8295575785003704,
+      "learning_rate": 4.994354882620267e-05,
+      "loss": 1.9972,
+      "step": 658
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7464702859065359,
+      "learning_rate": 4.994312849423961e-05,
+      "loss": 2.2733,
+      "step": 659
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8064579456158069,
+      "learning_rate": 4.9942706604974555e-05,
+      "loss": 2.0438,
+      "step": 660
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7902967432062032,
+      "learning_rate": 4.9942283158433865e-05,
+      "loss": 2.0561,
+      "step": 661
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7873763760899675,
+      "learning_rate": 4.994185815464398e-05,
+      "loss": 2.0291,
+      "step": 662
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8795987679934034,
+      "learning_rate": 4.9941431593631414e-05,
+      "loss": 2.2468,
+      "step": 663
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7181087428112289,
+      "learning_rate": 4.994100347542282e-05,
+      "loss": 2.1,
+      "step": 664
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8611595095002097,
+      "learning_rate": 4.994057380004492e-05,
+      "loss": 2.1339,
+      "step": 665
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8267398110866929,
+      "learning_rate": 4.994014256752454e-05,
+      "loss": 2.0357,
+      "step": 666
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8075389777165766,
+      "learning_rate": 4.99397097778886e-05,
+      "loss": 2.2799,
+      "step": 667
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8544295929998521,
+      "learning_rate": 4.993927543116412e-05,
+      "loss": 2.0479,
+      "step": 668
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8784910813934775,
+      "learning_rate": 4.993883952737822e-05,
+      "loss": 2.0485,
+      "step": 669
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7081845553094777,
+      "learning_rate": 4.993840206655812e-05,
+      "loss": 2.0574,
+      "step": 670
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7016368079186255,
+      "learning_rate": 4.9937963048731134e-05,
+      "loss": 2.2637,
+      "step": 671
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8409725228445841,
+      "learning_rate": 4.993752247392467e-05,
+      "loss": 2.0161,
+      "step": 672
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8861996372853229,
+      "learning_rate": 4.993708034216622e-05,
+      "loss": 2.0386,
+      "step": 673
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7843481388070138,
+      "learning_rate": 4.993663665348341e-05,
+      "loss": 2.0615,
+      "step": 674
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8429688090966334,
+      "learning_rate": 4.993619140790393e-05,
+      "loss": 2.2218,
+      "step": 675
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.063902709714382,
+      "learning_rate": 4.993574460545557e-05,
+      "loss": 2.0353,
+      "step": 676
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8430582434241445,
+      "learning_rate": 4.9935296246166245e-05,
+      "loss": 2.1565,
+      "step": 677
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9687464770378915,
+      "learning_rate": 4.993484633006394e-05,
+      "loss": 2.0529,
+      "step": 678
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8519851547668436,
+      "learning_rate": 4.993439485717674e-05,
+      "loss": 2.2507,
+      "step": 679
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8263542105946974,
+      "learning_rate": 4.993394182753284e-05,
+      "loss": 2.0663,
+      "step": 680
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0702254450021844,
+      "learning_rate": 4.9933487241160516e-05,
+      "loss": 2.041,
+      "step": 681
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8437667150722616,
+      "learning_rate": 4.993303109808816e-05,
+      "loss": 2.044,
+      "step": 682
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.067139137316206,
+      "learning_rate": 4.9932573398344236e-05,
+      "loss": 2.0943,
+      "step": 683
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9999358260689343,
+      "learning_rate": 4.993211414195733e-05,
+      "loss": 2.2078,
+      "step": 684
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8929280356637327,
+      "learning_rate": 4.993165332895612e-05,
+      "loss": 2.0841,
+      "step": 685
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.953788517899573,
+      "learning_rate": 4.993119095936937e-05,
+      "loss": 2.066,
+      "step": 686
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8259038534549908,
+      "learning_rate": 4.9930727033225945e-05,
+      "loss": 2.2291,
+      "step": 687
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8592689725828306,
+      "learning_rate": 4.993026155055482e-05,
+      "loss": 2.0174,
+      "step": 688
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7815492604276704,
+      "learning_rate": 4.9929794511385044e-05,
+      "loss": 2.0364,
+      "step": 689
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7121539456110083,
+      "learning_rate": 4.992932591574578e-05,
+      "loss": 2.0971,
+      "step": 690
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7993185383495275,
+      "learning_rate": 4.99288557636663e-05,
+      "loss": 2.2007,
+      "step": 691
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7495261296115238,
+      "learning_rate": 4.9928384055175934e-05,
+      "loss": 2.041,
+      "step": 692
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8856085910383256,
+      "learning_rate": 4.992791079030414e-05,
+      "loss": 2.0365,
+      "step": 693
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7647786163909616,
+      "learning_rate": 4.9927435969080476e-05,
+      "loss": 2.0331,
+      "step": 694
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8602517470224418,
+      "learning_rate": 4.992695959153458e-05,
+      "loss": 2.2141,
+      "step": 695
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8208175579891402,
+      "learning_rate": 4.9926481657696186e-05,
+      "loss": 2.0653,
+      "step": 696
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8079899929003923,
+      "learning_rate": 4.9926002167595153e-05,
+      "loss": 2.0722,
+      "step": 697
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8167136440703118,
+      "learning_rate": 4.99255211212614e-05,
+      "loss": 2.0718,
+      "step": 698
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9042468225600693,
+      "learning_rate": 4.992503851872497e-05,
+      "loss": 2.2825,
+      "step": 699
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7273317571493322,
+      "learning_rate": 4.9924554360015986e-05,
+      "loss": 2.0564,
+      "step": 700
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8121517092205636,
+      "learning_rate": 4.992406864516468e-05,
+      "loss": 2.0045,
+      "step": 701
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8394669052353458,
+      "learning_rate": 4.992358137420138e-05,
+      "loss": 2.1349,
+      "step": 702
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7902429503949969,
+      "learning_rate": 4.992309254715651e-05,
+      "loss": 2.2078,
+      "step": 703
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9750147073260156,
+      "learning_rate": 4.9922602164060586e-05,
+      "loss": 2.0741,
+      "step": 704
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7831176577813663,
+      "learning_rate": 4.992211022494422e-05,
+      "loss": 2.0497,
+      "step": 705
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9215773769890955,
+      "learning_rate": 4.992161672983813e-05,
+      "loss": 2.0453,
+      "step": 706
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7117490398604537,
+      "learning_rate": 4.992112167877313e-05,
+      "loss": 2.2023,
+      "step": 707
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7623678902091698,
+      "learning_rate": 4.992062507178012e-05,
+      "loss": 2.0829,
+      "step": 708
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8021934458379629,
+      "learning_rate": 4.992012690889011e-05,
+      "loss": 2.0436,
+      "step": 709
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8240638728293711,
+      "learning_rate": 4.9919627190134214e-05,
+      "loss": 2.0802,
+      "step": 710
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8385189594782518,
+      "learning_rate": 4.9919125915543604e-05,
+      "loss": 2.2199,
+      "step": 711
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7448968165781115,
+      "learning_rate": 4.99186230851496e-05,
+      "loss": 2.0563,
+      "step": 712
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8913757722206899,
+      "learning_rate": 4.991811869898359e-05,
+      "loss": 2.0721,
+      "step": 713
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7106863283717292,
+      "learning_rate": 4.9917612757077056e-05,
+      "loss": 2.1059,
+      "step": 714
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7479462682033995,
+      "learning_rate": 4.9917105259461605e-05,
+      "loss": 1.9997,
+      "step": 715
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8834815894441659,
+      "learning_rate": 4.991659620616891e-05,
+      "loss": 2.2784,
+      "step": 716
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.714724318427169,
+      "learning_rate": 4.991608559723074e-05,
+      "loss": 2.0616,
+      "step": 717
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8042717531145045,
+      "learning_rate": 4.9915573432679005e-05,
+      "loss": 2.0689,
+      "step": 718
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7935815420813611,
+      "learning_rate": 4.991505971254566e-05,
+      "loss": 2.2461,
+      "step": 719
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7787098515282533,
+      "learning_rate": 4.991454443686278e-05,
+      "loss": 2.0594,
+      "step": 720
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9099736503795062,
+      "learning_rate": 4.991402760566254e-05,
+      "loss": 2.1082,
+      "step": 721
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7621661021654437,
+      "learning_rate": 4.991350921897721e-05,
+      "loss": 2.0533,
+      "step": 722
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9200315935884937,
+      "learning_rate": 4.991298927683915e-05,
+      "loss": 2.1915,
+      "step": 723
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7305474485812655,
+      "learning_rate": 4.9912467779280825e-05,
+      "loss": 2.0144,
+      "step": 724
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7781171457784287,
+      "learning_rate": 4.991194472633479e-05,
+      "loss": 2.0149,
+      "step": 725
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7233474849879434,
+      "learning_rate": 4.991142011803371e-05,
+      "loss": 2.0182,
+      "step": 726
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8977702470890788,
+      "learning_rate": 4.991089395441033e-05,
+      "loss": 2.0824,
+      "step": 727
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7223307154371491,
+      "learning_rate": 4.9910366235497503e-05,
+      "loss": 2.2124,
+      "step": 728
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7350429405381966,
+      "learning_rate": 4.9909836961328185e-05,
+      "loss": 2.0189,
+      "step": 729
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7981841479670778,
+      "learning_rate": 4.99093061319354e-05,
+      "loss": 2.0414,
+      "step": 730
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7145214242492883,
+      "learning_rate": 4.990877374735231e-05,
+      "loss": 2.2191,
+      "step": 731
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8365423771874068,
+      "learning_rate": 4.9908239807612153e-05,
+      "loss": 2.0124,
+      "step": 732
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6919689517950398,
+      "learning_rate": 4.9907704312748255e-05,
+      "loss": 2.1078,
+      "step": 733
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7618951027208049,
+      "learning_rate": 4.990716726279405e-05,
+      "loss": 2.0745,
+      "step": 734
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7305981607490506,
+      "learning_rate": 4.990662865778307e-05,
+      "loss": 2.0329,
+      "step": 735
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7463826793007735,
+      "learning_rate": 4.990608849774895e-05,
+      "loss": 2.2019,
+      "step": 736
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7064261747096207,
+      "learning_rate": 4.990554678272541e-05,
+      "loss": 2.0214,
+      "step": 737
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7913628349854176,
+      "learning_rate": 4.9905003512746264e-05,
+      "loss": 2.0662,
+      "step": 738
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.681483463675515,
+      "learning_rate": 4.990445868784544e-05,
+      "loss": 2.0323,
+      "step": 739
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6599138744460065,
+      "learning_rate": 4.9903912308056946e-05,
+      "loss": 2.2373,
+      "step": 740
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7496447704112126,
+      "learning_rate": 4.9903364373414904e-05,
+      "loss": 2.0558,
+      "step": 741
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6753875808215875,
+      "learning_rate": 4.990281488395352e-05,
+      "loss": 1.9622,
+      "step": 742
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.708285283325801,
+      "learning_rate": 4.990226383970709e-05,
+      "loss": 2.2127,
+      "step": 743
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6640098730573766,
+      "learning_rate": 4.9901711240710033e-05,
+      "loss": 2.044,
+      "step": 744
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6509362036911452,
+      "learning_rate": 4.990115708699685e-05,
+      "loss": 2.1253,
+      "step": 745
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6901334692916077,
+      "learning_rate": 4.990060137860212e-05,
+      "loss": 2.0085,
+      "step": 746
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7615038072105573,
+      "learning_rate": 4.990004411556056e-05,
+      "loss": 2.049,
+      "step": 747
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6931460809687433,
+      "learning_rate": 4.9899485297906954e-05,
+      "loss": 2.2397,
+      "step": 748
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6534189928436769,
+      "learning_rate": 4.989892492567619e-05,
+      "loss": 2.0762,
+      "step": 749
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7112982447357519,
+      "learning_rate": 4.989836299890326e-05,
+      "loss": 2.0737,
+      "step": 750
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7641410485605341,
+      "learning_rate": 4.989779951762324e-05,
+      "loss": 2.222,
+      "step": 751
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8505337638915853,
+      "learning_rate": 4.989723448187131e-05,
+      "loss": 2.0935,
+      "step": 752
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7147043281115546,
+      "learning_rate": 4.989666789168276e-05,
+      "loss": 2.0134,
+      "step": 753
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7304699287639821,
+      "learning_rate": 4.989609974709295e-05,
+      "loss": 2.0217,
+      "step": 754
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7256952314498918,
+      "learning_rate": 4.989553004813735e-05,
+      "loss": 2.2198,
+      "step": 755
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7552942679998607,
+      "learning_rate": 4.9894958794851544e-05,
+      "loss": 2.0898,
+      "step": 756
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7018188153634748,
+      "learning_rate": 4.9894385987271185e-05,
+      "loss": 2.0217,
+      "step": 757
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6743354492054912,
+      "learning_rate": 4.989381162543204e-05,
+      "loss": 2.0403,
+      "step": 758
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7935821089716525,
+      "learning_rate": 4.9893235709369976e-05,
+      "loss": 2.0584,
+      "step": 759
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.871835295142175,
+      "learning_rate": 4.9892658239120934e-05,
+      "loss": 2.1947,
+      "step": 760
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6787437045570935,
+      "learning_rate": 4.989207921472098e-05,
+      "loss": 2.006,
+      "step": 761
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0358893907234512,
+      "learning_rate": 4.9891498636206267e-05,
+      "loss": 2.0885,
+      "step": 762
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6983656004377002,
+      "learning_rate": 4.9890916503613025e-05,
+      "loss": 2.2546,
+      "step": 763
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7465843139766267,
+      "learning_rate": 4.989033281697762e-05,
+      "loss": 2.0957,
+      "step": 764
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7048198575000174,
+      "learning_rate": 4.988974757633649e-05,
+      "loss": 2.001,
+      "step": 765
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7771613760800516,
+      "learning_rate": 4.9889160781726166e-05,
+      "loss": 2.0773,
+      "step": 766
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6950254925834636,
+      "learning_rate": 4.988857243318328e-05,
+      "loss": 2.0548,
+      "step": 767
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8341697096984066,
+      "learning_rate": 4.988798253074457e-05,
+      "loss": 2.2359,
+      "step": 768
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7186223256349434,
+      "learning_rate": 4.988739107444688e-05,
+      "loss": 2.0475,
+      "step": 769
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7411821927818484,
+      "learning_rate": 4.988679806432712e-05,
+      "loss": 2.1333,
+      "step": 770
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8520646889766157,
+      "learning_rate": 4.9886203500422313e-05,
+      "loss": 2.071,
+      "step": 771
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.700077065116785,
+      "learning_rate": 4.9885607382769596e-05,
+      "loss": 2.2553,
+      "step": 772
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7404556462446583,
+      "learning_rate": 4.988500971140617e-05,
+      "loss": 2.0562,
+      "step": 773
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7289801149227552,
+      "learning_rate": 4.988441048636937e-05,
+      "loss": 2.0334,
+      "step": 774
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7153092758658268,
+      "learning_rate": 4.988380970769658e-05,
+      "loss": 2.2493,
+      "step": 775
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7462983965174421,
+      "learning_rate": 4.9883207375425334e-05,
+      "loss": 2.052,
+      "step": 776
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8049546590078663,
+      "learning_rate": 4.9882603489593226e-05,
+      "loss": 2.0558,
+      "step": 777
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7266973839399192,
+      "learning_rate": 4.988199805023796e-05,
+      "loss": 2.034,
+      "step": 778
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.782933669916175,
+      "learning_rate": 4.988139105739734e-05,
+      "loss": 1.9865,
+      "step": 779
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.704126248590267,
+      "learning_rate": 4.988078251110926e-05,
+      "loss": 2.2452,
+      "step": 780
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6504003730202843,
+      "learning_rate": 4.9880172411411716e-05,
+      "loss": 2.0553,
+      "step": 781
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7265142059382989,
+      "learning_rate": 4.9879560758342804e-05,
+      "loss": 2.0348,
+      "step": 782
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6737558360417277,
+      "learning_rate": 4.98789475519407e-05,
+      "loss": 2.1169,
+      "step": 783
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8120840395656,
+      "learning_rate": 4.9878332792243686e-05,
+      "loss": 2.2139,
+      "step": 784
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.711600918910579,
+      "learning_rate": 4.9877716479290174e-05,
+      "loss": 2.0253,
+      "step": 785
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7814369244521375,
+      "learning_rate": 4.98770986131186e-05,
+      "loss": 2.0456,
+      "step": 786
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6818013651486091,
+      "learning_rate": 4.987647919376758e-05,
+      "loss": 2.2463,
+      "step": 787
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.093077745814283,
+      "learning_rate": 4.9875858221275756e-05,
+      "loss": 2.0803,
+      "step": 788
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7075235454045999,
+      "learning_rate": 4.987523569568192e-05,
+      "loss": 2.0974,
+      "step": 789
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6675591461105268,
+      "learning_rate": 4.987461161702492e-05,
+      "loss": 2.0294,
+      "step": 790
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7661498172793076,
+      "learning_rate": 4.9873985985343734e-05,
+      "loss": 2.0526,
+      "step": 791
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7345125327153424,
+      "learning_rate": 4.987335880067742e-05,
+      "loss": 2.2249,
+      "step": 792
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7932124616173761,
+      "learning_rate": 4.987273006306513e-05,
+      "loss": 2.0285,
+      "step": 793
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6508623223374154,
+      "learning_rate": 4.987209977254612e-05,
+      "loss": 2.028,
+      "step": 794
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7867669477716656,
+      "learning_rate": 4.987146792915975e-05,
+      "loss": 2.1119,
+      "step": 795
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6763597433838995,
+      "learning_rate": 4.987083453294546e-05,
+      "loss": 2.2323,
+      "step": 796
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7683310136612457,
+      "learning_rate": 4.98701995839428e-05,
+      "loss": 2.0578,
+      "step": 797
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7761324967971456,
+      "learning_rate": 4.9869563082191415e-05,
+      "loss": 2.0284,
+      "step": 798
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.726820297459208,
+      "learning_rate": 4.986892502773103e-05,
+      "loss": 2.0347,
+      "step": 799
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7967007859485941,
+      "learning_rate": 4.98682854206015e-05,
+      "loss": 2.2171,
+      "step": 800
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7471516554376232,
+      "learning_rate": 4.986764426084275e-05,
+      "loss": 2.1015,
+      "step": 801
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7625045814073039,
+      "learning_rate": 4.98670015484948e-05,
+      "loss": 2.0452,
+      "step": 802
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8003251758240643,
+      "learning_rate": 4.9866357283597794e-05,
+      "loss": 2.0244,
+      "step": 803
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7053028460745614,
+      "learning_rate": 4.986571146619195e-05,
+      "loss": 2.2306,
+      "step": 804
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7142548230919835,
+      "learning_rate": 4.9865064096317586e-05,
+      "loss": 2.0227,
+      "step": 805
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7505401564644995,
+      "learning_rate": 4.9864415174015124e-05,
+      "loss": 2.0823,
+      "step": 806
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7597643084817601,
+      "learning_rate": 4.9863764699325075e-05,
+      "loss": 2.2234,
+      "step": 807
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.768779724659344,
+      "learning_rate": 4.986311267228805e-05,
+      "loss": 2.0879,
+      "step": 808
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8565394030667026,
+      "learning_rate": 4.986245909294477e-05,
+      "loss": 2.0535,
+      "step": 809
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6680538376376509,
+      "learning_rate": 4.986180396133603e-05,
+      "loss": 1.9681,
+      "step": 810
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8127246566167461,
+      "learning_rate": 4.986114727750273e-05,
+      "loss": 2.0534,
+      "step": 811
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7312139074271476,
+      "learning_rate": 4.986048904148587e-05,
+      "loss": 2.2078,
+      "step": 812
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8067891560152749,
+      "learning_rate": 4.9859829253326554e-05,
+      "loss": 2.046,
+      "step": 813
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6481194313573897,
+      "learning_rate": 4.9859167913065964e-05,
+      "loss": 2.1019,
+      "step": 814
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7939226992594048,
+      "learning_rate": 4.9858505020745406e-05,
+      "loss": 2.046,
+      "step": 815
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.82158712380577,
+      "learning_rate": 4.985784057640626e-05,
+      "loss": 2.2539,
+      "step": 816
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7419622301819795,
+      "learning_rate": 4.985717458009e-05,
+      "loss": 2.0159,
+      "step": 817
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7668192282059307,
+      "learning_rate": 4.985650703183822e-05,
+      "loss": 2.0512,
+      "step": 818
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7630692495225115,
+      "learning_rate": 4.985583793169258e-05,
+      "loss": 2.2402,
+      "step": 819
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6915707947014003,
+      "learning_rate": 4.985516727969489e-05,
+      "loss": 2.0336,
+      "step": 820
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.750965860634281,
+      "learning_rate": 4.9854495075886984e-05,
+      "loss": 2.0513,
+      "step": 821
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6914527944337195,
+      "learning_rate": 4.985382132031085e-05,
+      "loss": 2.0467,
+      "step": 822
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6926143551025613,
+      "learning_rate": 4.9853146013008554e-05,
+      "loss": 2.0265,
+      "step": 823
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7343547430760857,
+      "learning_rate": 4.985246915402224e-05,
+      "loss": 2.241,
+      "step": 824
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6710432336977166,
+      "learning_rate": 4.9851790743394186e-05,
+      "loss": 2.0356,
+      "step": 825
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7233493593153486,
+      "learning_rate": 4.9851110781166746e-05,
+      "loss": 2.1113,
+      "step": 826
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6718253909341829,
+      "learning_rate": 4.985042926738237e-05,
+      "loss": 2.012,
+      "step": 827
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8346908565575587,
+      "learning_rate": 4.98497462020836e-05,
+      "loss": 2.2423,
+      "step": 828
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.73973075154058,
+      "learning_rate": 4.984906158531309e-05,
+      "loss": 2.0238,
+      "step": 829
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.840117373218022,
+      "learning_rate": 4.984837541711359e-05,
+      "loss": 2.0475,
+      "step": 830
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6608448010903015,
+      "learning_rate": 4.984768769752794e-05,
+      "loss": 2.1229,
+      "step": 831
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9034436293011574,
+      "learning_rate": 4.984699842659905e-05,
+      "loss": 2.2663,
+      "step": 832
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6633732433207797,
+      "learning_rate": 4.9846307604369985e-05,
+      "loss": 2.0395,
+      "step": 833
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7343044334820176,
+      "learning_rate": 4.9845615230883865e-05,
+      "loss": 1.9828,
+      "step": 834
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7816831638217253,
+      "learning_rate": 4.984492130618391e-05,
+      "loss": 2.031,
+      "step": 835
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6939326101848108,
+      "learning_rate": 4.984422583031346e-05,
+      "loss": 2.1999,
+      "step": 836
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7522123002819346,
+      "learning_rate": 4.984352880331593e-05,
+      "loss": 2.0171,
+      "step": 837
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8171640916275302,
+      "learning_rate": 4.9842830225234835e-05,
+      "loss": 2.0524,
+      "step": 838
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6577343787602368,
+      "learning_rate": 4.9842130096113795e-05,
+      "loss": 2.1207,
+      "step": 839
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9935433694907367,
+      "learning_rate": 4.984142841599651e-05,
+      "loss": 2.2246,
+      "step": 840
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7223477368200224,
+      "learning_rate": 4.984072518492681e-05,
+      "loss": 2.0494,
+      "step": 841
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.849218800020168,
+      "learning_rate": 4.9840020402948574e-05,
+      "loss": 2.0295,
+      "step": 842
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7577179823050312,
+      "learning_rate": 4.983931407010583e-05,
+      "loss": 2.0514,
+      "step": 843
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7713770670130917,
+      "learning_rate": 4.983860618644266e-05,
+      "loss": 2.26,
+      "step": 844
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8329551518722105,
+      "learning_rate": 4.983789675200327e-05,
+      "loss": 2.107,
+      "step": 845
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8276628941021626,
+      "learning_rate": 4.983718576683195e-05,
+      "loss": 2.0514,
+      "step": 846
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.824191615076118,
+      "learning_rate": 4.983647323097309e-05,
+      "loss": 2.1109,
+      "step": 847
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9863282747024558,
+      "learning_rate": 4.9835759144471175e-05,
+      "loss": 2.2199,
+      "step": 848
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7489246358183148,
+      "learning_rate": 4.983504350737078e-05,
+      "loss": 2.0306,
+      "step": 849
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7440571818464188,
+      "learning_rate": 4.98343263197166e-05,
+      "loss": 2.0556,
+      "step": 850
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7850101165918115,
+      "learning_rate": 4.983360758155341e-05,
+      "loss": 2.1154,
+      "step": 851
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7008034623840368,
+      "learning_rate": 4.983288729292607e-05,
+      "loss": 2.2408,
+      "step": 852
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9255140032245944,
+      "learning_rate": 4.983216545387957e-05,
+      "loss": 2.0657,
+      "step": 853
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6977989356636721,
+      "learning_rate": 4.9831442064458964e-05,
+      "loss": 2.0126,
+      "step": 854
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.869289560394945,
+      "learning_rate": 4.983071712470942e-05,
+      "loss": 2.0777,
+      "step": 855
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8215997227558899,
+      "learning_rate": 4.982999063467619e-05,
+      "loss": 2.216,
+      "step": 856
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8457242573412425,
+      "learning_rate": 4.982926259440465e-05,
+      "loss": 2.133,
+      "step": 857
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7648923112512502,
+      "learning_rate": 4.982853300394025e-05,
+      "loss": 2.0303,
+      "step": 858
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8038686071630206,
+      "learning_rate": 4.9827801863328527e-05,
+      "loss": 2.0273,
+      "step": 859
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8461515344313622,
+      "learning_rate": 4.982706917261514e-05,
+      "loss": 2.247,
+      "step": 860
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6989260927839712,
+      "learning_rate": 4.9826334931845835e-05,
+      "loss": 2.0176,
+      "step": 861
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7934748715928471,
+      "learning_rate": 4.982559914106645e-05,
+      "loss": 2.0366,
+      "step": 862
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6917894664885689,
+      "learning_rate": 4.982486180032293e-05,
+      "loss": 2.1194,
+      "step": 863
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7250799543659454,
+      "learning_rate": 4.98241229096613e-05,
+      "loss": 2.2303,
+      "step": 864
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7269420689931193,
+      "learning_rate": 4.9823382469127686e-05,
+      "loss": 2.0437,
+      "step": 865
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6809825934741925,
+      "learning_rate": 4.982264047876835e-05,
+      "loss": 2.0315,
+      "step": 866
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7951886783491885,
+      "learning_rate": 4.982189693862957e-05,
+      "loss": 2.0548,
+      "step": 867
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9263725221802728,
+      "learning_rate": 4.982115184875781e-05,
+      "loss": 2.1664,
+      "step": 868
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6987917071579538,
+      "learning_rate": 4.982040520919957e-05,
+      "loss": 2.0739,
+      "step": 869
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0335954251331716,
+      "learning_rate": 4.9819657020001466e-05,
+      "loss": 2.1281,
+      "step": 870
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7497860936467335,
+      "learning_rate": 4.9818907281210206e-05,
+      "loss": 2.0432,
+      "step": 871
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8221237995850927,
+      "learning_rate": 4.9818155992872614e-05,
+      "loss": 2.1887,
+      "step": 872
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.925165863566801,
+      "learning_rate": 4.9817403155035595e-05,
+      "loss": 2.029,
+      "step": 873
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.79730469064726,
+      "learning_rate": 4.9816648767746135e-05,
+      "loss": 2.0166,
+      "step": 874
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8109329590212345,
+      "learning_rate": 4.9815892831051336e-05,
+      "loss": 2.0579,
+      "step": 875
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.811281504167831,
+      "learning_rate": 4.981513534499841e-05,
+      "loss": 2.2684,
+      "step": 876
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7600970178837713,
+      "learning_rate": 4.981437630963464e-05,
+      "loss": 2.0267,
+      "step": 877
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9612666325776142,
+      "learning_rate": 4.981361572500742e-05,
+      "loss": 2.0452,
+      "step": 878
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7421183489881553,
+      "learning_rate": 4.981285359116423e-05,
+      "loss": 2.0698,
+      "step": 879
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.826324154297863,
+      "learning_rate": 4.9812089908152656e-05,
+      "loss": 2.2681,
+      "step": 880
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.900948510584364,
+      "learning_rate": 4.981132467602038e-05,
+      "loss": 2.0657,
+      "step": 881
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7305268703880301,
+      "learning_rate": 4.9810557894815176e-05,
+      "loss": 2.0279,
+      "step": 882
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0192189923801236,
+      "learning_rate": 4.980978956458492e-05,
+      "loss": 2.0022,
+      "step": 883
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7921033842507202,
+      "learning_rate": 4.980901968537758e-05,
+      "loss": 2.251,
+      "step": 884
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9055087307965137,
+      "learning_rate": 4.9808248257241216e-05,
+      "loss": 2.021,
+      "step": 885
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.028227290666769,
+      "learning_rate": 4.980747528022401e-05,
+      "loss": 2.0538,
+      "step": 886
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6681327839597142,
+      "learning_rate": 4.98067007543742e-05,
+      "loss": 2.0145,
+      "step": 887
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9031349792497351,
+      "learning_rate": 4.980592467974016e-05,
+      "loss": 2.2398,
+      "step": 888
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.715313057331333,
+      "learning_rate": 4.980514705637034e-05,
+      "loss": 2.0258,
+      "step": 889
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7672590841322251,
+      "learning_rate": 4.980436788431328e-05,
+      "loss": 2.019,
+      "step": 890
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.785246739293584,
+      "learning_rate": 4.980358716361764e-05,
+      "loss": 2.0424,
+      "step": 891
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7162691780674956,
+      "learning_rate": 4.9802804894332156e-05,
+      "loss": 2.2233,
+      "step": 892
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8781188225279981,
+      "learning_rate": 4.980202107650567e-05,
+      "loss": 2.0545,
+      "step": 893
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6872812849863329,
+      "learning_rate": 4.980123571018711e-05,
+      "loss": 2.1034,
+      "step": 894
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8039306146356703,
+      "learning_rate": 4.980044879542552e-05,
+      "loss": 2.0484,
+      "step": 895
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7968291476576818,
+      "learning_rate": 4.979966033227004e-05,
+      "loss": 2.2398,
+      "step": 896
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7159424103456706,
+      "learning_rate": 4.9798870320769886e-05,
+      "loss": 2.0183,
+      "step": 897
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7672306920926802,
+      "learning_rate": 4.979807876097437e-05,
+      "loss": 2.0342,
+      "step": 898
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7310034897072122,
+      "learning_rate": 4.979728565293293e-05,
+      "loss": 1.9777,
+      "step": 899
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7945518271347839,
+      "learning_rate": 4.9796490996695086e-05,
+      "loss": 2.212,
+      "step": 900
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9091171486390567,
+      "learning_rate": 4.979569479231043e-05,
+      "loss": 2.097,
+      "step": 901
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6862795583698919,
+      "learning_rate": 4.9794897039828694e-05,
+      "loss": 2.0075,
+      "step": 902
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7937909251281441,
+      "learning_rate": 4.9794097739299664e-05,
+      "loss": 2.0301,
+      "step": 903
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.800197802644946,
+      "learning_rate": 4.979329689077327e-05,
+      "loss": 2.2486,
+      "step": 904
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8379376902739822,
+      "learning_rate": 4.979249449429948e-05,
+      "loss": 2.0303,
+      "step": 905
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6948687426230196,
+      "learning_rate": 4.979169054992842e-05,
+      "loss": 1.9992,
+      "step": 906
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8377682635008799,
+      "learning_rate": 4.979088505771027e-05,
+      "loss": 2.1031,
+      "step": 907
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6467559363841642,
+      "learning_rate": 4.979007801769533e-05,
+      "loss": 2.2709,
+      "step": 908
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7424639949571712,
+      "learning_rate": 4.9789269429933964e-05,
+      "loss": 2.0005,
+      "step": 909
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7432849247820381,
+      "learning_rate": 4.978845929447668e-05,
+      "loss": 2.0053,
+      "step": 910
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8244180404005984,
+      "learning_rate": 4.978764761137404e-05,
+      "loss": 1.9976,
+      "step": 911
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7312852236983929,
+      "learning_rate": 4.978683438067675e-05,
+      "loss": 2.2409,
+      "step": 912
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6755154803231432,
+      "learning_rate": 4.978601960243554e-05,
+      "loss": 2.1389,
+      "step": 913
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7008291060049643,
+      "learning_rate": 4.978520327670132e-05,
+      "loss": 2.0518,
+      "step": 914
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7482159958514782,
+      "learning_rate": 4.978438540352502e-05,
+      "loss": 1.9967,
+      "step": 915
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6732795052437436,
+      "learning_rate": 4.978356598295774e-05,
+      "loss": 2.2508,
+      "step": 916
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.735602100029352,
+      "learning_rate": 4.978274501505061e-05,
+      "loss": 2.0472,
+      "step": 917
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6458421689863666,
+      "learning_rate": 4.978192249985491e-05,
+      "loss": 2.0497,
+      "step": 918
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7434733474547877,
+      "learning_rate": 4.978109843742198e-05,
+      "loss": 2.0599,
+      "step": 919
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7670719285295212,
+      "learning_rate": 4.9780272827803265e-05,
+      "loss": 2.2025,
+      "step": 920
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8235117677091967,
+      "learning_rate": 4.9779445671050315e-05,
+      "loss": 2.0536,
+      "step": 921
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6662852030688867,
+      "learning_rate": 4.977861696721479e-05,
+      "loss": 2.0026,
+      "step": 922
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7213365024835765,
+      "learning_rate": 4.977778671634839e-05,
+      "loss": 2.0379,
+      "step": 923
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7001861314640812,
+      "learning_rate": 4.9776954918503e-05,
+      "loss": 2.2172,
+      "step": 924
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6729168726967532,
+      "learning_rate": 4.977612157373052e-05,
+      "loss": 2.104,
+      "step": 925
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6654466314413828,
+      "learning_rate": 4.9775286682082984e-05,
+      "loss": 2.0259,
+      "step": 926
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.733989863819936,
+      "learning_rate": 4.977445024361252e-05,
+      "loss": 2.0268,
+      "step": 927
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7558077891871695,
+      "learning_rate": 4.977361225837136e-05,
+      "loss": 2.2334,
+      "step": 928
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.676743415651291,
+      "learning_rate": 4.977277272641181e-05,
+      "loss": 2.0472,
+      "step": 929
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9118849471534284,
+      "learning_rate": 4.977193164778628e-05,
+      "loss": 2.0408,
+      "step": 930
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8876821118985034,
+      "learning_rate": 4.9771089022547304e-05,
+      "loss": 2.0464,
+      "step": 931
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6333579493885405,
+      "learning_rate": 4.977024485074747e-05,
+      "loss": 2.1092,
+      "step": 932
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7400077346057291,
+      "learning_rate": 4.97693991324395e-05,
+      "loss": 2.2154,
+      "step": 933
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0783608963601525,
+      "learning_rate": 4.9768551867676186e-05,
+      "loss": 2.0487,
+      "step": 934
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8185687668141374,
+      "learning_rate": 4.9767703056510416e-05,
+      "loss": 1.9996,
+      "step": 935
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.820888920686587,
+      "learning_rate": 4.9766852698995205e-05,
+      "loss": 2.2525,
+      "step": 936
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6874286826616396,
+      "learning_rate": 4.976600079518363e-05,
+      "loss": 2.021,
+      "step": 937
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7874940933643579,
+      "learning_rate": 4.9765147345128884e-05,
+      "loss": 2.0826,
+      "step": 938
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.71585370646607,
+      "learning_rate": 4.976429234888426e-05,
+      "loss": 2.0107,
+      "step": 939
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9736690348691248,
+      "learning_rate": 4.976343580650312e-05,
+      "loss": 2.241,
+      "step": 940
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7632309525353292,
+      "learning_rate": 4.976257771803896e-05,
+      "loss": 2.0125,
+      "step": 941
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8874391172412602,
+      "learning_rate": 4.9761718083545344e-05,
+      "loss": 2.0286,
+      "step": 942
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0125582744743706,
+      "learning_rate": 4.976085690307594e-05,
+      "loss": 2.0375,
+      "step": 943
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7545554717568546,
+      "learning_rate": 4.975999417668452e-05,
+      "loss": 2.1201,
+      "step": 944
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8204251486017243,
+      "learning_rate": 4.975912990442495e-05,
+      "loss": 2.2272,
+      "step": 945
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7925800742207649,
+      "learning_rate": 4.975826408635119e-05,
+      "loss": 2.0404,
+      "step": 946
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8528619441503231,
+      "learning_rate": 4.975739672251729e-05,
+      "loss": 1.9958,
+      "step": 947
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0799450757311806,
+      "learning_rate": 4.97565278129774e-05,
+      "loss": 2.2556,
+      "step": 948
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6630359748141954,
+      "learning_rate": 4.975565735778579e-05,
+      "loss": 2.04,
+      "step": 949
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0294361052356031,
+      "learning_rate": 4.9754785356996787e-05,
+      "loss": 2.0645,
+      "step": 950
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7712324828792639,
+      "learning_rate": 4.975391181066483e-05,
+      "loss": 1.9943,
+      "step": 951
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7424901430667604,
+      "learning_rate": 4.975303671884448e-05,
+      "loss": 2.0393,
+      "step": 952
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.843649540127684,
+      "learning_rate": 4.975216008159035e-05,
+      "loss": 2.1948,
+      "step": 953
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5891375173476674,
+      "learning_rate": 4.975128189895719e-05,
+      "loss": 1.9325,
+      "step": 954
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7569030117340539,
+      "learning_rate": 4.9750402170999814e-05,
+      "loss": 2.003,
+      "step": 955
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7571528370796616,
+      "learning_rate": 4.974952089777316e-05,
+      "loss": 2.2574,
+      "step": 956
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7062070482456363,
+      "learning_rate": 4.9748638079332236e-05,
+      "loss": 2.1115,
+      "step": 957
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9686026724538899,
+      "learning_rate": 4.974775371573217e-05,
+      "loss": 2.0151,
+      "step": 958
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6758375395265377,
+      "learning_rate": 4.974686780702817e-05,
+      "loss": 2.0299,
+      "step": 959
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0976813306232893,
+      "learning_rate": 4.974598035327555e-05,
+      "loss": 2.173,
+      "step": 960
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8152299831492347,
+      "learning_rate": 4.974509135452972e-05,
+      "loss": 2.0238,
+      "step": 961
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8807975321880066,
+      "learning_rate": 4.974420081084618e-05,
+      "loss": 2.0186,
+      "step": 962
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6950261951814034,
+      "learning_rate": 4.9743308722280536e-05,
+      "loss": 2.0838,
+      "step": 963
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8125022838282693,
+      "learning_rate": 4.974241508888847e-05,
+      "loss": 2.0443,
+      "step": 964
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6237207771715597,
+      "learning_rate": 4.9741519910725796e-05,
+      "loss": 2.2202,
+      "step": 965
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7897099917993983,
+      "learning_rate": 4.974062318784839e-05,
+      "loss": 2.0259,
+      "step": 966
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6978263002965192,
+      "learning_rate": 4.973972492031224e-05,
+      "loss": 2.0418,
+      "step": 967
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7031922639373333,
+      "learning_rate": 4.973882510817344e-05,
+      "loss": 2.2428,
+      "step": 968
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7226546147937737,
+      "learning_rate": 4.9737923751488146e-05,
+      "loss": 2.1011,
+      "step": 969
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6977130645612353,
+      "learning_rate": 4.973702085031264e-05,
+      "loss": 2.074,
+      "step": 970
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8065794558247578,
+      "learning_rate": 4.973611640470331e-05,
+      "loss": 2.0792,
+      "step": 971
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8448694034276166,
+      "learning_rate": 4.973521041471662e-05,
+      "loss": 2.2251,
+      "step": 972
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7159249637319535,
+      "learning_rate": 4.973430288040912e-05,
+      "loss": 2.0086,
+      "step": 973
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7876933856360586,
+      "learning_rate": 4.973339380183749e-05,
+      "loss": 1.9869,
+      "step": 974
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6559759951805296,
+      "learning_rate": 4.973248317905847e-05,
+      "loss": 2.1253,
+      "step": 975
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7135362861633066,
+      "learning_rate": 4.973157101212892e-05,
+      "loss": 1.9926,
+      "step": 976
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7323847052407415,
+      "learning_rate": 4.97306573011058e-05,
+      "loss": 2.245,
+      "step": 977
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6721835144845716,
+      "learning_rate": 4.972974204604614e-05,
+      "loss": 2.0218,
+      "step": 978
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7170384418801566,
+      "learning_rate": 4.972882524700709e-05,
+      "loss": 2.0825,
+      "step": 979
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7933813675003109,
+      "learning_rate": 4.972790690404591e-05,
+      "loss": 2.1797,
+      "step": 980
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6553433956107843,
+      "learning_rate": 4.972698701721989e-05,
+      "loss": 2.1613,
+      "step": 981
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6943969689560172,
+      "learning_rate": 4.9726065586586514e-05,
+      "loss": 1.9498,
+      "step": 982
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7959267362030705,
+      "learning_rate": 4.9725142612203265e-05,
+      "loss": 2.0123,
+      "step": 983
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6298556809262316,
+      "learning_rate": 4.972421809412781e-05,
+      "loss": 2.037,
+      "step": 984
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8254187216597044,
+      "learning_rate": 4.972329203241783e-05,
+      "loss": 2.2099,
+      "step": 985
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6321830946557987,
+      "learning_rate": 4.9722364427131175e-05,
+      "loss": 2.0229,
+      "step": 986
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7304206440517998,
+      "learning_rate": 4.972143527832575e-05,
+      "loss": 2.045,
+      "step": 987
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6197332727535526,
+      "learning_rate": 4.972050458605955e-05,
+      "loss": 2.0629,
+      "step": 988
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6639498186932975,
+      "learning_rate": 4.97195723503907e-05,
+      "loss": 2.2372,
+      "step": 989
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6580922492264214,
+      "learning_rate": 4.97186385713774e-05,
+      "loss": 2.0381,
+      "step": 990
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6623205429303454,
+      "learning_rate": 4.971770324907794e-05,
+      "loss": 2.0401,
+      "step": 991
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7211961374016485,
+      "learning_rate": 4.9716766383550734e-05,
+      "loss": 2.2275,
+      "step": 992
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7168788784949109,
+      "learning_rate": 4.9715827974854254e-05,
+      "loss": 2.0469,
+      "step": 993
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6104165856442945,
+      "learning_rate": 4.9714888023047104e-05,
+      "loss": 2.0707,
+      "step": 994
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7124639837637254,
+      "learning_rate": 4.971394652818796e-05,
+      "loss": 2.0214,
+      "step": 995
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7020182905430483,
+      "learning_rate": 4.9713003490335614e-05,
+      "loss": 2.0101,
+      "step": 996
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6275850417776819,
+      "learning_rate": 4.971205890954893e-05,
+      "loss": 2.2318,
+      "step": 997
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6901571679344544,
+      "learning_rate": 4.971111278588689e-05,
+      "loss": 1.9885,
+      "step": 998
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6518103261036434,
+      "learning_rate": 4.971016511940856e-05,
+      "loss": 2.0716,
+      "step": 999
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6234726733576571,
+      "learning_rate": 4.970921591017311e-05,
+      "loss": 2.0668,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.661947058249578,
+      "learning_rate": 4.9708265158239806e-05,
+      "loss": 2.1665,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6664500077721862,
+      "learning_rate": 4.970731286366801e-05,
+      "loss": 1.9984,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7111451604744019,
+      "learning_rate": 4.970635902651716e-05,
+      "loss": 2.0101,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7097839952634879,
+      "learning_rate": 4.970540364684683e-05,
+      "loss": 2.2351,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6495305803555842,
+      "learning_rate": 4.9704446724716643e-05,
+      "loss": 2.0391,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6183726723553741,
+      "learning_rate": 4.970348826018637e-05,
+      "loss": 2.1098,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7008105243982548,
+      "learning_rate": 4.9702528253315846e-05,
+      "loss": 2.0138,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6131618126007959,
+      "learning_rate": 4.970156670416499e-05,
+      "loss": 2.0179,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7841514087228121,
+      "learning_rate": 4.9700603612793864e-05,
+      "loss": 2.1816,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7992658807814428,
+      "learning_rate": 4.969963897926256e-05,
+      "loss": 2.0223,
+      "step": 1010
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9131768884047065,
+      "learning_rate": 4.969867280363134e-05,
+      "loss": 2.0633,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6761429244910706,
+      "learning_rate": 4.969770508596051e-05,
+      "loss": 2.1207,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9580250948896384,
+      "learning_rate": 4.969673582631049e-05,
+      "loss": 2.1782,
+      "step": 1013
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9106560560904544,
+      "learning_rate": 4.9695765024741796e-05,
+      "loss": 2.0303,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7384008520053995,
+      "learning_rate": 4.969479268131504e-05,
+      "loss": 2.0221,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8385369069628335,
+      "learning_rate": 4.9693818796090927e-05,
+      "loss": 2.0236,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7889825737723688,
+      "learning_rate": 4.969284336913026e-05,
+      "loss": 2.252,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6910643882549516,
+      "learning_rate": 4.969186640049394e-05,
+      "loss": 2.044,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7188617974176045,
+      "learning_rate": 4.9690887890242964e-05,
+      "loss": 2.1338,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7531724394565169,
+      "learning_rate": 4.9689907838438435e-05,
+      "loss": 2.0073,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6445354880030325,
+      "learning_rate": 4.968892624514152e-05,
+      "loss": 2.2523,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7243309486864397,
+      "learning_rate": 4.9687943110413516e-05,
+      "loss": 1.9924,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7009964450659534,
+      "learning_rate": 4.9686958434315804e-05,
+      "loss": 2.0295,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.707505304161887,
+      "learning_rate": 4.968597221690986e-05,
+      "loss": 2.218,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7355397671951441,
+      "learning_rate": 4.968498445825726e-05,
+      "loss": 2.1093,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6660555201393771,
+      "learning_rate": 4.968399515841967e-05,
+      "loss": 2.0032,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7212233879366329,
+      "learning_rate": 4.968300431745886e-05,
+      "loss": 1.991,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6601724424398042,
+      "learning_rate": 4.968201193543669e-05,
+      "loss": 2.0575,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7947924365624697,
+      "learning_rate": 4.9681018012415114e-05,
+      "loss": 2.1704,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6978287433527,
+      "learning_rate": 4.96800225484562e-05,
+      "loss": 2.0008,
+      "step": 1030
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.659031799836682,
+      "learning_rate": 4.9679025543622085e-05,
+      "loss": 2.095,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6936100193493179,
+      "learning_rate": 4.967802699797501e-05,
+      "loss": 2.0245,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6863481693580329,
+      "learning_rate": 4.967702691157734e-05,
+      "loss": 2.2254,
+      "step": 1033
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8047739314973198,
+      "learning_rate": 4.967602528449151e-05,
+      "loss": 2.0325,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.721649843661402,
+      "learning_rate": 4.967502211678005e-05,
+      "loss": 2.0201,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7442001868959572,
+      "learning_rate": 4.967401740850558e-05,
+      "loss": 2.2478,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6350403227026801,
+      "learning_rate": 4.9673011159730844e-05,
+      "loss": 2.0951,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6673287110831463,
+      "learning_rate": 4.967200337051867e-05,
+      "loss": 2.0471,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6992422278366218,
+      "learning_rate": 4.967099404093196e-05,
+      "loss": 1.9937,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6697359326786231,
+      "learning_rate": 4.9669983171033744e-05,
+      "loss": 2.0163,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7289081795903044,
+      "learning_rate": 4.9668970760887135e-05,
+      "loss": 2.2485,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6515271142029379,
+      "learning_rate": 4.966795681055534e-05,
+      "loss": 2.0277,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7645822991036333,
+      "learning_rate": 4.9666941320101655e-05,
+      "loss": 2.0732,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6673669356419604,
+      "learning_rate": 4.966592428958949e-05,
+      "loss": 2.0778,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8072304795674596,
+      "learning_rate": 4.966490571908235e-05,
+      "loss": 2.2673,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7547983630791196,
+      "learning_rate": 4.9663885608643815e-05,
+      "loss": 2.0216,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7585184589199094,
+      "learning_rate": 4.9662863958337577e-05,
+      "loss": 1.9981,
+      "step": 1047
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7120197062285855,
+      "learning_rate": 4.966184076822743e-05,
+      "loss": 2.067,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8317450870447444,
+      "learning_rate": 4.966081603837725e-05,
+      "loss": 2.2282,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7107590506437731,
+      "learning_rate": 4.965978976885102e-05,
+      "loss": 2.0561,
+      "step": 1050
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7189081973991424,
+      "learning_rate": 4.96587619597128e-05,
+      "loss": 2.0027,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.731351418285113,
+      "learning_rate": 4.965773261102678e-05,
+      "loss": 2.0292,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8858160599144653,
+      "learning_rate": 4.9656701722857203e-05,
+      "loss": 2.2305,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7457616477189996,
+      "learning_rate": 4.965566929526845e-05,
+      "loss": 2.0441,
+      "step": 1054
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8863291314227609,
+      "learning_rate": 4.965463532832498e-05,
+      "loss": 2.0729,
+      "step": 1055
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7711512030297716,
+      "learning_rate": 4.9653599822091336e-05,
+      "loss": 2.0952,
+      "step": 1056
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9152845673629448,
+      "learning_rate": 4.9652562776632186e-05,
+      "loss": 2.1901,
+      "step": 1057
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8398253263202407,
+      "learning_rate": 4.9651524192012255e-05,
+      "loss": 2.0426,
+      "step": 1058
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8054376367957267,
+      "learning_rate": 4.96504840682964e-05,
+      "loss": 2.0301,
+      "step": 1059
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.771070531991092,
+      "learning_rate": 4.9649442405549564e-05,
+      "loss": 2.0705,
+      "step": 1060
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.784323862701141,
+      "learning_rate": 4.964839920383677e-05,
+      "loss": 2.2204,
+      "step": 1061
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6701921618757497,
+      "learning_rate": 4.964735446322316e-05,
+      "loss": 2.1203,
+      "step": 1062
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8005268162701557,
+      "learning_rate": 4.964630818377396e-05,
+      "loss": 2.0721,
+      "step": 1063
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.736304255389527,
+      "learning_rate": 4.964526036555448e-05,
+      "loss": 2.0102,
+      "step": 1064
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7854911415831769,
+      "learning_rate": 4.964421100863016e-05,
+      "loss": 2.2211,
+      "step": 1065
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6912173675863236,
+      "learning_rate": 4.96431601130665e-05,
+      "loss": 2.03,
+      "step": 1066
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.872388200871447,
+      "learning_rate": 4.964210767892912e-05,
+      "loss": 1.9564,
+      "step": 1067
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8173542354284669,
+      "learning_rate": 4.9641053706283725e-05,
+      "loss": 2.0624,
+      "step": 1068
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7466383694390366,
+      "learning_rate": 4.9639998195196114e-05,
+      "loss": 2.1901,
+      "step": 1069
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9602262245815072,
+      "learning_rate": 4.96389411457322e-05,
+      "loss": 2.0349,
+      "step": 1070
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7223897110466765,
+      "learning_rate": 4.963788255795797e-05,
+      "loss": 2.0303,
+      "step": 1071
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7714234423989593,
+      "learning_rate": 4.963682243193951e-05,
+      "loss": 2.0497,
+      "step": 1072
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8033783682063672,
+      "learning_rate": 4.963576076774302e-05,
+      "loss": 2.2201,
+      "step": 1073
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6824882474199132,
+      "learning_rate": 4.9634697565434777e-05,
+      "loss": 2.0842,
+      "step": 1074
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.774489191820849,
+      "learning_rate": 4.9633632825081166e-05,
+      "loss": 2.0389,
+      "step": 1075
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7365024192543138,
+      "learning_rate": 4.963256654674866e-05,
+      "loss": 2.046,
+      "step": 1076
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.841699067956948,
+      "learning_rate": 4.963149873050383e-05,
+      "loss": 2.1969,
+      "step": 1077
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6731114895007412,
+      "learning_rate": 4.963042937641335e-05,
+      "loss": 2.0196,
+      "step": 1078
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7205027172424567,
+      "learning_rate": 4.962935848454397e-05,
+      "loss": 1.9993,
+      "step": 1079
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8527946200197376,
+      "learning_rate": 4.962828605496256e-05,
+      "loss": 2.0074,
+      "step": 1080
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7465398101084108,
+      "learning_rate": 4.9627212087736084e-05,
+      "loss": 2.2432,
+      "step": 1081
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.652738685562523,
+      "learning_rate": 4.962613658293158e-05,
+      "loss": 2.0646,
+      "step": 1082
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7861249464464168,
+      "learning_rate": 4.962505954061621e-05,
+      "loss": 2.0743,
+      "step": 1083
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.708890666243947,
+      "learning_rate": 4.962398096085721e-05,
+      "loss": 1.968,
+      "step": 1084
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7674057481884768,
+      "learning_rate": 4.962290084372191e-05,
+      "loss": 2.1829,
+      "step": 1085
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.748623140066446,
+      "learning_rate": 4.962181918927777e-05,
+      "loss": 1.9904,
+      "step": 1086
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6798828789939435,
+      "learning_rate": 4.962073599759231e-05,
+      "loss": 2.1426,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7301733815725464,
+      "learning_rate": 4.9619651268733144e-05,
+      "loss": 2.0079,
+      "step": 1088
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.68003568559764,
+      "learning_rate": 4.961856500276801e-05,
+      "loss": 2.2387,
+      "step": 1089
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6953394064931132,
+      "learning_rate": 4.961747719976474e-05,
+      "loss": 2.037,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7265080326878989,
+      "learning_rate": 4.961638785979123e-05,
+      "loss": 2.0535,
+      "step": 1091
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6410283530926739,
+      "learning_rate": 4.961529698291549e-05,
+      "loss": 2.0077,
+      "step": 1092
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6587762743896434,
+      "learning_rate": 4.961420456920566e-05,
+      "loss": 2.2499,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7541415627246982,
+      "learning_rate": 4.96131106187299e-05,
+      "loss": 1.9918,
+      "step": 1094
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7346643627920064,
+      "learning_rate": 4.961201513155654e-05,
+      "loss": 2.0532,
+      "step": 1095
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6976891395832233,
+      "learning_rate": 4.961091810775397e-05,
+      "loss": 2.0397,
+      "step": 1096
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8527623584074553,
+      "learning_rate": 4.960981954739067e-05,
+      "loss": 2.2478,
+      "step": 1097
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7338474575772996,
+      "learning_rate": 4.9608719450535236e-05,
+      "loss": 2.0291,
+      "step": 1098
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8544896969740515,
+      "learning_rate": 4.960761781725636e-05,
+      "loss": 2.1128,
+      "step": 1099
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6634281152760301,
+      "learning_rate": 4.96065146476228e-05,
+      "loss": 2.0321,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7655626578824397,
+      "learning_rate": 4.9605409941703464e-05,
+      "loss": 2.2023,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6733480948442727,
+      "learning_rate": 4.9604303699567286e-05,
+      "loss": 2.0059,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7740564112818662,
+      "learning_rate": 4.960319592128336e-05,
+      "loss": 1.9949,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6874472132497218,
+      "learning_rate": 4.960208660692084e-05,
+      "loss": 2.0361,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7380994477730938,
+      "learning_rate": 4.960097575654898e-05,
+      "loss": 2.2632,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.852312319985487,
+      "learning_rate": 4.9599863370237144e-05,
+      "loss": 2.0524,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6781156244779359,
+      "learning_rate": 4.959874944805478e-05,
+      "loss": 2.0221,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9680788472103373,
+      "learning_rate": 4.9597633990071426e-05,
+      "loss": 1.9844,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8531478478552657,
+      "learning_rate": 4.959651699635674e-05,
+      "loss": 2.1821,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.770146150294283,
+      "learning_rate": 4.959539846698045e-05,
+      "loss": 2.0408,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8851849038283203,
+      "learning_rate": 4.9594278402012395e-05,
+      "loss": 2.0737,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7026094322141042,
+      "learning_rate": 4.95931568015225e-05,
+      "loss": 2.0638,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.873459353307161,
+      "learning_rate": 4.95920336655808e-05,
+      "loss": 2.2284,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7931822077349445,
+      "learning_rate": 4.959090899425741e-05,
+      "loss": 2.0271,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.729894161943249,
+      "learning_rate": 4.958978278762255e-05,
+      "loss": 2.0379,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9047779535473272,
+      "learning_rate": 4.9588655045746534e-05,
+      "loss": 2.0224,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7682985554428163,
+      "learning_rate": 4.958752576869977e-05,
+      "loss": 2.2411,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8241863918368432,
+      "learning_rate": 4.958639495655276e-05,
+      "loss": 2.0876,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9637923812568879,
+      "learning_rate": 4.958526260937611e-05,
+      "loss": 1.9851,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7392565034796879,
+      "learning_rate": 4.958412872724052e-05,
+      "loss": 2.0205,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0138962511071723,
+      "learning_rate": 4.9582993310216774e-05,
+      "loss": 2.2164,
+      "step": 1121
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.667317663133804,
+      "learning_rate": 4.958185635837578e-05,
+      "loss": 2.023,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8238105349904448,
+      "learning_rate": 4.958071787178849e-05,
+      "loss": 2.0417,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7462304113404287,
+      "learning_rate": 4.9579577850526015e-05,
+      "loss": 2.0927,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7411967413119631,
+      "learning_rate": 4.957843629465951e-05,
+      "loss": 2.1775,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8957837918437386,
+      "learning_rate": 4.9577293204260265e-05,
+      "loss": 2.0178,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7530335422985333,
+      "learning_rate": 4.957614857939964e-05,
+      "loss": 2.0388,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6921811191967561,
+      "learning_rate": 4.9575002420149095e-05,
+      "loss": 2.059,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8618525567933475,
+      "learning_rate": 4.957385472658019e-05,
+      "loss": 2.2355,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6074498672199484,
+      "learning_rate": 4.957270549876459e-05,
+      "loss": 2.0668,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7810323804261484,
+      "learning_rate": 4.957155473677403e-05,
+      "loss": 2.0422,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6618596058067698,
+      "learning_rate": 4.957040244068038e-05,
+      "loss": 2.022,
+      "step": 1132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7021726401038475,
+      "learning_rate": 4.956924861055555e-05,
+      "loss": 2.2346,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7303958528888959,
+      "learning_rate": 4.9568093246471603e-05,
+      "loss": 2.0402,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6797179770565074,
+      "learning_rate": 4.956693634850067e-05,
+      "loss": 2.0024,
+      "step": 1135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7872261872073871,
+      "learning_rate": 4.956577791671497e-05,
+      "loss": 2.021,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9852649964706484,
+      "learning_rate": 4.956461795118684e-05,
+      "loss": 2.2566,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6945653714374064,
+      "learning_rate": 4.9563456451988694e-05,
+      "loss": 2.0682,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9116516976907111,
+      "learning_rate": 4.9562293419193054e-05,
+      "loss": 2.0412,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6666076139848338,
+      "learning_rate": 4.956112885287253e-05,
+      "loss": 2.0195,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9190447813968373,
+      "learning_rate": 4.9559962753099835e-05,
+      "loss": 2.1911,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8397680311380011,
+      "learning_rate": 4.955879511994778e-05,
+      "loss": 2.0148,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8182235791403191,
+      "learning_rate": 4.955762595348924e-05,
+      "loss": 2.0797,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9612082887683635,
+      "learning_rate": 4.955645525379723e-05,
+      "loss": 1.9672,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7293208653101693,
+      "learning_rate": 4.9555283020944844e-05,
+      "loss": 2.1946,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8979454429122249,
+      "learning_rate": 4.955410925500526e-05,
+      "loss": 2.0267,
+      "step": 1146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8792669306041917,
+      "learning_rate": 4.955293395605176e-05,
+      "loss": 2.0246,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7634640272201217,
+      "learning_rate": 4.955175712415773e-05,
+      "loss": 2.0883,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8582898375311494,
+      "learning_rate": 4.955057875939664e-05,
+      "loss": 2.0636,
+      "step": 1149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6254136814594533,
+      "learning_rate": 4.954939886184207e-05,
+      "loss": 2.2292,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.841958801591601,
+      "learning_rate": 4.9548217431567665e-05,
+      "loss": 2.0125,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7482969379492537,
+      "learning_rate": 4.9547034468647214e-05,
+      "loss": 2.0535,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8930921303952899,
+      "learning_rate": 4.9545849973154544e-05,
+      "loss": 2.1956,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.733474757353843,
+      "learning_rate": 4.954466394516364e-05,
+      "loss": 2.0234,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7105650425854,
+      "learning_rate": 4.954347638474852e-05,
+      "loss": 2.1365,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6912419971479322,
+      "learning_rate": 4.954228729198335e-05,
+      "loss": 2.0261,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6522808082938539,
+      "learning_rate": 4.954109666694235e-05,
+      "loss": 2.2071,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.807711966635246,
+      "learning_rate": 4.953990450969989e-05,
+      "loss": 2.0349,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.674344849250202,
+      "learning_rate": 4.953871082033036e-05,
+      "loss": 2.0277,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7286528576338047,
+      "learning_rate": 4.953751559890831e-05,
+      "loss": 2.0594,
+      "step": 1160
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7295389341607927,
+      "learning_rate": 4.953631884550836e-05,
+      "loss": 2.0911,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7102316954071105,
+      "learning_rate": 4.953512056020523e-05,
+      "loss": 2.2327,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6825915413046217,
+      "learning_rate": 4.9533920743073725e-05,
+      "loss": 1.9923,
+      "step": 1163
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.737601873845265,
+      "learning_rate": 4.9532719394188764e-05,
+      "loss": 1.9946,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7283855328717732,
+      "learning_rate": 4.9531516513625345e-05,
+      "loss": 2.2244,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7405415974333289,
+      "learning_rate": 4.953031210145858e-05,
+      "loss": 2.0497,
+      "step": 1166
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6350718675278094,
+      "learning_rate": 4.952910615776365e-05,
+      "loss": 2.0202,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6472411950872328,
+      "learning_rate": 4.9527898682615855e-05,
+      "loss": 2.0428,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7030429959244919,
+      "learning_rate": 4.952668967609058e-05,
+      "loss": 2.0084,
+      "step": 1169
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7266715028967139,
+      "learning_rate": 4.952547913826332e-05,
+      "loss": 2.1906,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6679616917197104,
+      "learning_rate": 4.952426706920963e-05,
+      "loss": 2.042,
+      "step": 1171
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7009517210809939,
+      "learning_rate": 4.952305346900521e-05,
+      "loss": 2.0376,
+      "step": 1172
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7035991577881616,
+      "learning_rate": 4.9521838337725814e-05,
+      "loss": 2.1955,
+      "step": 1173
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6990689106248329,
+      "learning_rate": 4.952062167544731e-05,
+      "loss": 2.127,
+      "step": 1174
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7225767453801756,
+      "learning_rate": 4.9519403482245666e-05,
+      "loss": 1.968,
+      "step": 1175
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6317048137352662,
+      "learning_rate": 4.9518183758196935e-05,
+      "loss": 2.0196,
+      "step": 1176
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7585815970200795,
+      "learning_rate": 4.951696250337727e-05,
+      "loss": 2.2158,
+      "step": 1177
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6414203807282861,
+      "learning_rate": 4.951573971786291e-05,
+      "loss": 2.0449,
+      "step": 1178
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6676496320229492,
+      "learning_rate": 4.951451540173021e-05,
+      "loss": 1.9986,
+      "step": 1179
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7573460421372907,
+      "learning_rate": 4.951328955505561e-05,
+      "loss": 2.082,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.648673490929942,
+      "learning_rate": 4.951206217791564e-05,
+      "loss": 2.0352,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8001062662861226,
+      "learning_rate": 4.951083327038693e-05,
+      "loss": 2.2163,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7891533039280032,
+      "learning_rate": 4.95096028325462e-05,
+      "loss": 2.0259,
+      "step": 1183
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6498202081550398,
+      "learning_rate": 4.950837086447028e-05,
+      "loss": 2.0162,
+      "step": 1184
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.851512622034079,
+      "learning_rate": 4.950713736623608e-05,
+      "loss": 2.2041,
+      "step": 1185
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6593544477671972,
+      "learning_rate": 4.950590233792062e-05,
+      "loss": 2.1312,
+      "step": 1186
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9568269752686577,
+      "learning_rate": 4.9504665779601e-05,
+      "loss": 2.0109,
+      "step": 1187
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8149241106213242,
+      "learning_rate": 4.950342769135443e-05,
+      "loss": 2.0578,
+      "step": 1188
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7001180957111567,
+      "learning_rate": 4.9502188073258215e-05,
+      "loss": 2.2101,
+      "step": 1189
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8304629760826161,
+      "learning_rate": 4.9500946925389734e-05,
+      "loss": 2.0463,
+      "step": 1190
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6827463911640193,
+      "learning_rate": 4.9499704247826486e-05,
+      "loss": 1.97,
+      "step": 1191
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7748230181632485,
+      "learning_rate": 4.949846004064605e-05,
+      "loss": 2.0596,
+      "step": 1192
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7744869918003887,
+      "learning_rate": 4.949721430392612e-05,
+      "loss": 1.974,
+      "step": 1193
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7183663680162962,
+      "learning_rate": 4.949596703774445e-05,
+      "loss": 2.1727,
+      "step": 1194
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8112647422572766,
+      "learning_rate": 4.949471824217894e-05,
+      "loss": 2.0321,
+      "step": 1195
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7032185809406998,
+      "learning_rate": 4.9493467917307537e-05,
+      "loss": 2.0183,
+      "step": 1196
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9315650264259715,
+      "learning_rate": 4.9492216063208306e-05,
+      "loss": 2.2519,
+      "step": 1197
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8683266485685761,
+      "learning_rate": 4.949096267995942e-05,
+      "loss": 2.0268,
+      "step": 1198
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.68553023361516,
+      "learning_rate": 4.948970776763911e-05,
+      "loss": 2.0655,
+      "step": 1199
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0074078112362301,
+      "learning_rate": 4.948845132632575e-05,
+      "loss": 1.9924,
+      "step": 1200
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6501335033968308,
+      "learning_rate": 4.9487193356097764e-05,
+      "loss": 2.0105,
+      "step": 1201
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8667444071589213,
+      "learning_rate": 4.9485933857033706e-05,
+      "loss": 2.2467,
+      "step": 1202
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9627874343059033,
+      "learning_rate": 4.948467282921221e-05,
+      "loss": 2.039,
+      "step": 1203
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6795434128030994,
+      "learning_rate": 4.9483410272712e-05,
+      "loss": 2.0446,
+      "step": 1204
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9108529482133553,
+      "learning_rate": 4.94821461876119e-05,
+      "loss": 2.0781,
+      "step": 1205
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7375020791839677,
+      "learning_rate": 4.948088057399084e-05,
+      "loss": 2.1952,
+      "step": 1206
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8337950335735858,
+      "learning_rate": 4.947961343192784e-05,
+      "loss": 2.0452,
+      "step": 1207
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7883459438546738,
+      "learning_rate": 4.947834476150201e-05,
+      "loss": 1.998,
+      "step": 1208
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7483379397419843,
+      "learning_rate": 4.947707456279256e-05,
+      "loss": 2.1965,
+      "step": 1209
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.891124485706932,
+      "learning_rate": 4.947580283587878e-05,
+      "loss": 2.0035,
+      "step": 1210
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5938230666014205,
+      "learning_rate": 4.9474529580840085e-05,
+      "loss": 2.0849,
+      "step": 1211
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7832320198452103,
+      "learning_rate": 4.9473254797755966e-05,
+      "loss": 2.0377,
+      "step": 1212
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7340048474777493,
+      "learning_rate": 4.9471978486706e-05,
+      "loss": 1.9824,
+      "step": 1213
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6411054456039458,
+      "learning_rate": 4.9470700647769904e-05,
+      "loss": 2.1894,
+      "step": 1214
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6750707950154565,
+      "learning_rate": 4.946942128102743e-05,
+      "loss": 1.9877,
+      "step": 1215
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7192942317818203,
+      "learning_rate": 4.946814038655846e-05,
+      "loss": 2.0281,
+      "step": 1216
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6375629204243052,
+      "learning_rate": 4.9466857964442964e-05,
+      "loss": 2.0929,
+      "step": 1217
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6631151567361007,
+      "learning_rate": 4.946557401476102e-05,
+      "loss": 2.17,
+      "step": 1218
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7004568972614962,
+      "learning_rate": 4.946428853759278e-05,
+      "loss": 2.0372,
+      "step": 1219
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7233063679299834,
+      "learning_rate": 4.946300153301851e-05,
+      "loss": 2.006,
+      "step": 1220
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.635386485366084,
+      "learning_rate": 4.946171300111855e-05,
+      "loss": 2.2403,
+      "step": 1221
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6471590731319148,
+      "learning_rate": 4.946042294197336e-05,
+      "loss": 1.993,
+      "step": 1222
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5973655684212206,
+      "learning_rate": 4.945913135566348e-05,
+      "loss": 2.0638,
+      "step": 1223
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6790031391159621,
+      "learning_rate": 4.9457838242269546e-05,
+      "loss": 2.021,
+      "step": 1224
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.623255845180241,
+      "learning_rate": 4.94565436018723e-05,
+      "loss": 2.0466,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6395915558908302,
+      "learning_rate": 4.9455247434552556e-05,
+      "loss": 2.2179,
+      "step": 1226
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6497210081429029,
+      "learning_rate": 4.945394974039126e-05,
+      "loss": 2.0188,
+      "step": 1227
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6588346421717401,
+      "learning_rate": 4.945265051946942e-05,
+      "loss": 2.0069,
+      "step": 1228
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7119112460293314,
+      "learning_rate": 4.945134977186816e-05,
+      "loss": 2.2778,
+      "step": 1229
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.57135677946212,
+      "learning_rate": 4.9450047497668676e-05,
+      "loss": 2.0541,
+      "step": 1230
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7191549598440876,
+      "learning_rate": 4.9448743696952286e-05,
+      "loss": 2.0439,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7157867781202208,
+      "learning_rate": 4.944743836980039e-05,
+      "loss": 2.035,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6393656820040531,
+      "learning_rate": 4.9446131516294485e-05,
+      "loss": 2.0361,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7531502674304072,
+      "learning_rate": 4.944482313651616e-05,
+      "loss": 2.2221,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6858199931976968,
+      "learning_rate": 4.944351323054711e-05,
+      "loss": 2.0587,
+      "step": 1235
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6473603601128464,
+      "learning_rate": 4.944220179846911e-05,
+      "loss": 2.1235,
+      "step": 1236
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6552420101560866,
+      "learning_rate": 4.944088884036404e-05,
+      "loss": 2.036,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6266792740480249,
+      "learning_rate": 4.9439574356313865e-05,
+      "loss": 2.2228,
+      "step": 1238
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7034071256095411,
+      "learning_rate": 4.943825834640068e-05,
+      "loss": 1.9987,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6827806777704185,
+      "learning_rate": 4.9436940810706615e-05,
+      "loss": 2.0333,
+      "step": 1240
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6347223248704427,
+      "learning_rate": 4.943562174931396e-05,
+      "loss": 2.1688,
+      "step": 1241
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.576430732108288,
+      "learning_rate": 4.943430116230504e-05,
+      "loss": 2.0933,
+      "step": 1242
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.689732601582731,
+      "learning_rate": 4.943297904976233e-05,
+      "loss": 2.0322,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6478854315264666,
+      "learning_rate": 4.9431655411768364e-05,
+      "loss": 2.0045,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6702488097472884,
+      "learning_rate": 4.9430330248405783e-05,
+      "loss": 2.0521,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6325607528456259,
+      "learning_rate": 4.942900355975732e-05,
+      "loss": 2.1882,
+      "step": 1246
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7121582656263283,
+      "learning_rate": 4.942767534590581e-05,
+      "loss": 2.0184,
+      "step": 1247
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7541101976557016,
+      "learning_rate": 4.9426345606934175e-05,
+      "loss": 2.0567,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6546008262859417,
+      "learning_rate": 4.9425014342925436e-05,
+      "loss": 2.0022,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6733489189131282,
+      "learning_rate": 4.942368155396271e-05,
+      "loss": 2.2083,
+      "step": 1250
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6980470194802821,
+      "learning_rate": 4.942234724012922e-05,
+      "loss": 1.9886,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6431429934939142,
+      "learning_rate": 4.942101140150826e-05,
+      "loss": 2.0319,
+      "step": 1252
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7025040199556309,
+      "learning_rate": 4.941967403818322e-05,
+      "loss": 2.1678,
+      "step": 1253
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6006639095021568,
+      "learning_rate": 4.941833515023763e-05,
+      "loss": 2.0953,
+      "step": 1254
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7116648130760442,
+      "learning_rate": 4.941699473775505e-05,
+      "loss": 2.0384,
+      "step": 1255
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6614144283823683,
+      "learning_rate": 4.9415652800819186e-05,
+      "loss": 2.0064,
+      "step": 1256
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7422982275451593,
+      "learning_rate": 4.941430933951381e-05,
+      "loss": 1.9888,
+      "step": 1257
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6830282249418417,
+      "learning_rate": 4.9412964353922814e-05,
+      "loss": 2.2137,
+      "step": 1258
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6865252342123821,
+      "learning_rate": 4.941161784413016e-05,
+      "loss": 2.0533,
+      "step": 1259
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6278576003474102,
+      "learning_rate": 4.941026981021992e-05,
+      "loss": 1.9389,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6387162366100033,
+      "learning_rate": 4.940892025227625e-05,
+      "loss": 2.0637,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.643538801418585,
+      "learning_rate": 4.9407569170383415e-05,
+      "loss": 2.2375,
+      "step": 1262
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6586212775070299,
+      "learning_rate": 4.940621656462577e-05,
+      "loss": 1.9945,
+      "step": 1263
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6856604374609993,
+      "learning_rate": 4.9404862435087765e-05,
+      "loss": 2.018,
+      "step": 1264
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6498519515362073,
+      "learning_rate": 4.940350678185394e-05,
+      "loss": 2.0121,
+      "step": 1265
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.78305577298325,
+      "learning_rate": 4.940214960500893e-05,
+      "loss": 2.2076,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8216115922378247,
+      "learning_rate": 4.940079090463747e-05,
+      "loss": 2.0733,
+      "step": 1267
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6509087305087776,
+      "learning_rate": 4.93994306808244e-05,
+      "loss": 2.0349,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7400337480556364,
+      "learning_rate": 4.939806893365464e-05,
+      "loss": 2.0341,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7059298938795996,
+      "learning_rate": 4.93967056632132e-05,
+      "loss": 2.2271,
+      "step": 1270
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7350343326413614,
+      "learning_rate": 4.939534086958521e-05,
+      "loss": 2.0629,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7837355257272042,
+      "learning_rate": 4.939397455285586e-05,
+      "loss": 2.0479,
+      "step": 1272
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6389854103590814,
+      "learning_rate": 4.939260671311047e-05,
+      "loss": 2.0584,
+      "step": 1273
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7361658984681914,
+      "learning_rate": 4.939123735043444e-05,
+      "loss": 2.18,
+      "step": 1274
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7460637094562477,
+      "learning_rate": 4.938986646491325e-05,
+      "loss": 2.0407,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6700816310591816,
+      "learning_rate": 4.938849405663251e-05,
+      "loss": 2.0328,
+      "step": 1276
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7239776831526815,
+      "learning_rate": 4.9387120125677885e-05,
+      "loss": 2.0594,
+      "step": 1277
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7791851613503944,
+      "learning_rate": 4.938574467213518e-05,
+      "loss": 2.2643,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6894871124071639,
+      "learning_rate": 4.938436769609025e-05,
+      "loss": 2.0878,
+      "step": 1279
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.900650097317208,
+      "learning_rate": 4.938298919762907e-05,
+      "loss": 2.0164,
+      "step": 1280
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6615225793496733,
+      "learning_rate": 4.9381609176837704e-05,
+      "loss": 1.9921,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.83935847146629,
+      "learning_rate": 4.938022763380232e-05,
+      "loss": 2.2264,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7050713243975505,
+      "learning_rate": 4.937884456860916e-05,
+      "loss": 1.9998,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6365375351547815,
+      "learning_rate": 4.93774599813446e-05,
+      "loss": 2.0232,
+      "step": 1284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7283724354987484,
+      "learning_rate": 4.937607387209505e-05,
+      "loss": 1.9966,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7911837829366256,
+      "learning_rate": 4.9374686240947075e-05,
+      "loss": 2.2074,
+      "step": 1286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.675441652626894,
+      "learning_rate": 4.9373297087987305e-05,
+      "loss": 1.9843,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8690720557947171,
+      "learning_rate": 4.937190641330247e-05,
+      "loss": 2.0517,
+      "step": 1288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.661902555249874,
+      "learning_rate": 4.9370514216979405e-05,
+      "loss": 1.991,
+      "step": 1289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8720815221336945,
+      "learning_rate": 4.9369120499105005e-05,
+      "loss": 2.1959,
+      "step": 1290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8070080440683559,
+      "learning_rate": 4.936772525976631e-05,
+      "loss": 2.0054,
+      "step": 1291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6766548776666609,
+      "learning_rate": 4.936632849905042e-05,
+      "loss": 2.0826,
+      "step": 1292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7497710861780327,
+      "learning_rate": 4.9364930217044537e-05,
+      "loss": 1.9755,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6591328786451597,
+      "learning_rate": 4.9363530413835976e-05,
+      "loss": 2.1417,
+      "step": 1294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8106457848170325,
+      "learning_rate": 4.9362129089512117e-05,
+      "loss": 2.0181,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6594006074696173,
+      "learning_rate": 4.936072624416046e-05,
+      "loss": 2.075,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7134499562293972,
+      "learning_rate": 4.935932187786858e-05,
+      "loss": 1.9969,
+      "step": 1297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7296888770916595,
+      "learning_rate": 4.935791599072418e-05,
+      "loss": 2.2605,
+      "step": 1298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6823858980290215,
+      "learning_rate": 4.935650858281501e-05,
+      "loss": 2.0379,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.969012912987145,
+      "learning_rate": 4.9355099654228945e-05,
+      "loss": 2.0309,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6737537170271464,
+      "learning_rate": 4.935368920505396e-05,
+      "loss": 1.9931,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.070565565169915,
+      "learning_rate": 4.935227723537811e-05,
+      "loss": 2.2284,
+      "step": 1302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.716436167594749,
+      "learning_rate": 4.935086374528955e-05,
+      "loss": 2.0107,
+      "step": 1303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7197695695049768,
+      "learning_rate": 4.934944873487654e-05,
+      "loss": 2.0889,
+      "step": 1304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8559310835590055,
+      "learning_rate": 4.9348032204227406e-05,
+      "loss": 2.0113,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6776819088013176,
+      "learning_rate": 4.9346614153430604e-05,
+      "loss": 2.1967,
+      "step": 1306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8774642127341507,
+      "learning_rate": 4.934519458257466e-05,
+      "loss": 2.056,
+      "step": 1307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6902292612230814,
+      "learning_rate": 4.93437734917482e-05,
+      "loss": 1.9993,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8858225906196834,
+      "learning_rate": 4.934235088103996e-05,
+      "loss": 2.029,
+      "step": 1309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8221413089279395,
+      "learning_rate": 4.934092675053875e-05,
+      "loss": 2.2356,
+      "step": 1310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7368355179078083,
+      "learning_rate": 4.9339501100333486e-05,
+      "loss": 2.0555,
+      "step": 1311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9112225608741429,
+      "learning_rate": 4.933807393051318e-05,
+      "loss": 1.9948,
+      "step": 1312
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6719035614778086,
+      "learning_rate": 4.933664524116694e-05,
+      "loss": 1.982,
+      "step": 1313
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.019917254334078,
+      "learning_rate": 4.933521503238395e-05,
+      "loss": 2.2106,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6554400597774033,
+      "learning_rate": 4.933378330425352e-05,
+      "loss": 1.9513,
+      "step": 1315
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7338500694121778,
+      "learning_rate": 4.933235005686504e-05,
+      "loss": 2.0033,
+      "step": 1316
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.725630760965212,
+      "learning_rate": 4.933091529030798e-05,
+      "loss": 2.0887,
+      "step": 1317
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6708245820323802,
+      "learning_rate": 4.932947900467193e-05,
+      "loss": 2.1988,
+      "step": 1318
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7575120674643232,
+      "learning_rate": 4.932804120004655e-05,
+      "loss": 2.0417,
+      "step": 1319
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6758547717404063,
+      "learning_rate": 4.932660187652162e-05,
+      "loss": 1.9296,
+      "step": 1320
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8203942084712488,
+      "learning_rate": 4.932516103418699e-05,
+      "loss": 2.1005,
+      "step": 1321
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7472642061522002,
+      "learning_rate": 4.932371867313264e-05,
+      "loss": 2.2003,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7215367806390354,
+      "learning_rate": 4.9322274793448595e-05,
+      "loss": 2.0837,
+      "step": 1323
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8281465675308213,
+      "learning_rate": 4.932082939522502e-05,
+      "loss": 2.0344,
+      "step": 1324
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6762818786916303,
+      "learning_rate": 4.931938247855216e-05,
+      "loss": 2.0468,
+      "step": 1325
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7774336220447284,
+      "learning_rate": 4.931793404352034e-05,
+      "loss": 2.2376,
+      "step": 1326
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6929016053717274,
+      "learning_rate": 4.9316484090220004e-05,
+      "loss": 1.9959,
+      "step": 1327
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6620327025676058,
+      "learning_rate": 4.931503261874166e-05,
+      "loss": 2.0216,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6691610602222117,
+      "learning_rate": 4.9313579629175955e-05,
+      "loss": 2.0597,
+      "step": 1329
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6837742184898068,
+      "learning_rate": 4.931212512161358e-05,
+      "loss": 2.1811,
+      "step": 1330
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6989880840977692,
+      "learning_rate": 4.931066909614536e-05,
+      "loss": 2.0448,
+      "step": 1331
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7094737200333332,
+      "learning_rate": 4.93092115528622e-05,
+      "loss": 2.051,
+      "step": 1332
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6346523142759848,
+      "learning_rate": 4.9307752491855096e-05,
+      "loss": 1.9916,
+      "step": 1333
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6398644637391256,
+      "learning_rate": 4.930629191321514e-05,
+      "loss": 2.2564,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.624015196136929,
+      "learning_rate": 4.9304829817033536e-05,
+      "loss": 2.0788,
+      "step": 1335
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7021210510783147,
+      "learning_rate": 4.930336620340156e-05,
+      "loss": 2.0178,
+      "step": 1336
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7179074690529313,
+      "learning_rate": 4.930190107241058e-05,
+      "loss": 2.036,
+      "step": 1337
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7221112935571207,
+      "learning_rate": 4.9300434424152093e-05,
+      "loss": 2.2437,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6790095362899412,
+      "learning_rate": 4.929896625871765e-05,
+      "loss": 2.0287,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6566081755875817,
+      "learning_rate": 4.929749657619893e-05,
+      "loss": 2.0673,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6720157735102602,
+      "learning_rate": 4.929602537668767e-05,
+      "loss": 2.1102,
+      "step": 1341
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6603505205475175,
+      "learning_rate": 4.929455266027574e-05,
+      "loss": 2.1874,
+      "step": 1342
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7339029255369434,
+      "learning_rate": 4.929307842705508e-05,
+      "loss": 1.9862,
+      "step": 1343
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6284914964558868,
+      "learning_rate": 4.9291602677117745e-05,
+      "loss": 1.9712,
+      "step": 1344
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6893421358684864,
+      "learning_rate": 4.929012541055585e-05,
+      "loss": 2.0037,
+      "step": 1345
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7258442383574379,
+      "learning_rate": 4.9288646627461645e-05,
+      "loss": 2.1871,
+      "step": 1346
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6574591573777764,
+      "learning_rate": 4.9287166327927455e-05,
+      "loss": 2.0176,
+      "step": 1347
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.646917874666992,
+      "learning_rate": 4.928568451204569e-05,
+      "loss": 2.0946,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6867314058364945,
+      "learning_rate": 4.928420117990887e-05,
+      "loss": 1.959,
+      "step": 1349
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7302301828126002,
+      "learning_rate": 4.928271633160962e-05,
+      "loss": 2.2413,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8222059838322406,
+      "learning_rate": 4.928122996724063e-05,
+      "loss": 2.0345,
+      "step": 1351
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.78961780159155,
+      "learning_rate": 4.92797420868947e-05,
+      "loss": 2.0132,
+      "step": 1352
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7620716258153973,
+      "learning_rate": 4.927825269066473e-05,
+      "loss": 2.0386,
+      "step": 1353
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.673236701377961,
+      "learning_rate": 4.92767617786437e-05,
+      "loss": 2.2611,
+      "step": 1354
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7473519180921998,
+      "learning_rate": 4.927526935092471e-05,
+      "loss": 2.0074,
+      "step": 1355
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7455857263136154,
+      "learning_rate": 4.927377540760092e-05,
+      "loss": 2.0332,
+      "step": 1356
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6684308072191367,
+      "learning_rate": 4.927227994876562e-05,
+      "loss": 2.0005,
+      "step": 1357
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7833241949470409,
+      "learning_rate": 4.927078297451217e-05,
+      "loss": 2.2257,
+      "step": 1358
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6724717742866224,
+      "learning_rate": 4.926928448493403e-05,
+      "loss": 2.0041,
+      "step": 1359
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7586771626220096,
+      "learning_rate": 4.9267784480124755e-05,
+      "loss": 2.0622,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7074053227017949,
+      "learning_rate": 4.9266282960178006e-05,
+      "loss": 2.01,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6660321763606343,
+      "learning_rate": 4.9264779925187524e-05,
+      "loss": 2.1879,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7068494611590002,
+      "learning_rate": 4.926327537524714e-05,
+      "loss": 2.0601,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6912167034183743,
+      "learning_rate": 4.926176931045081e-05,
+      "loss": 2.0017,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7897086876024525,
+      "learning_rate": 4.926026173089254e-05,
+      "loss": 2.0025,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8577721037195298,
+      "learning_rate": 4.925875263666648e-05,
+      "loss": 2.0945,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6607077861922018,
+      "learning_rate": 4.925724202786682e-05,
+      "loss": 2.1742,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8266398648704273,
+      "learning_rate": 4.92557299045879e-05,
+      "loss": 1.9941,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7324747896933725,
+      "learning_rate": 4.9254216266924114e-05,
+      "loss": 2.0076,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7938501904482388,
+      "learning_rate": 4.925270111496997e-05,
+      "loss": 2.2218,
+      "step": 1370
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6470950059920718,
+      "learning_rate": 4.925118444882006e-05,
+      "loss": 2.0211,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7670667226921369,
+      "learning_rate": 4.9249666268569086e-05,
+      "loss": 2.1107,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7477371405276937,
+      "learning_rate": 4.924814657431182e-05,
+      "loss": 2.0049,
+      "step": 1373
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6385979306079431,
+      "learning_rate": 4.924662536614314e-05,
+      "loss": 2.1831,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6470327373095323,
+      "learning_rate": 4.9245102644158046e-05,
+      "loss": 1.9855,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7842674063117663,
+      "learning_rate": 4.924357840845158e-05,
+      "loss": 2.018,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6581355206258155,
+      "learning_rate": 4.9242052659118935e-05,
+      "loss": 1.9806,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7782053478836232,
+      "learning_rate": 4.924052539625534e-05,
+      "loss": 2.2263,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7030568808399806,
+      "learning_rate": 4.923899661995617e-05,
+      "loss": 2.0524,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8329068407379203,
+      "learning_rate": 4.923746633031686e-05,
+      "loss": 2.0124,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7735910987759003,
+      "learning_rate": 4.9235934527432956e-05,
+      "loss": 2.0257,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7666188703619063,
+      "learning_rate": 4.923440121140009e-05,
+      "loss": 2.1792,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7339166525749653,
+      "learning_rate": 4.9232866382314014e-05,
+      "loss": 1.9829,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7588350992874737,
+      "learning_rate": 4.923133004027053e-05,
+      "loss": 2.0198,
+      "step": 1384
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6586022950087436,
+      "learning_rate": 4.922979218536557e-05,
+      "loss": 2.0854,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7254282207275699,
+      "learning_rate": 4.922825281769514e-05,
+      "loss": 2.0266,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8126760418600476,
+      "learning_rate": 4.922671193735536e-05,
+      "loss": 2.226,
+      "step": 1387
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.640145186003117,
+      "learning_rate": 4.922516954444243e-05,
+      "loss": 2.0387,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6627040198743919,
+      "learning_rate": 4.922362563905264e-05,
+      "loss": 2.0046,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7388868399071236,
+      "learning_rate": 4.9222080221282387e-05,
+      "loss": 2.1921,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7315726696665364,
+      "learning_rate": 4.922053329122817e-05,
+      "loss": 2.0809,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.645971740884652,
+      "learning_rate": 4.9218984848986544e-05,
+      "loss": 2.0262,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7486732423768055,
+      "learning_rate": 4.921743489465421e-05,
+      "loss": 2.0282,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6517187734383602,
+      "learning_rate": 4.921588342832792e-05,
+      "loss": 2.2206,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6837180450301867,
+      "learning_rate": 4.921433045010455e-05,
+      "loss": 2.0042,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7033723554480398,
+      "learning_rate": 4.921277596008106e-05,
+      "loss": 2.0309,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6239111702741212,
+      "learning_rate": 4.9211219958354496e-05,
+      "loss": 2.0923,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6479867167526784,
+      "learning_rate": 4.9209662445022003e-05,
+      "loss": 2.0287,
+      "step": 1398
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6640194757746232,
+      "learning_rate": 4.9208103420180826e-05,
+      "loss": 2.2056,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7296763663717897,
+      "learning_rate": 4.9206542883928305e-05,
+      "loss": 1.984,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6617176948865363,
+      "learning_rate": 4.920498083636188e-05,
+      "loss": 2.0014,
+      "step": 1401
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8683525948404506,
+      "learning_rate": 4.920341727757906e-05,
+      "loss": 2.19,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5884740506352705,
+      "learning_rate": 4.920185220767746e-05,
+      "loss": 2.0792,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7442585642204066,
+      "learning_rate": 4.920028562675481e-05,
+      "loss": 2.0436,
+      "step": 1404
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.622773421178493,
+      "learning_rate": 4.919871753490891e-05,
+      "loss": 1.9681,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6565300645157086,
+      "learning_rate": 4.9197147932237664e-05,
+      "loss": 2.204,
+      "step": 1406
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6632049598364192,
+      "learning_rate": 4.9195576818839064e-05,
+      "loss": 1.9777,
+      "step": 1407
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5882294410579186,
+      "learning_rate": 4.919400419481121e-05,
+      "loss": 2.0131,
+      "step": 1408
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6676869785017955,
+      "learning_rate": 4.919243006025228e-05,
+      "loss": 2.0167,
+      "step": 1409
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5893579848265511,
+      "learning_rate": 4.9190854415260556e-05,
+      "loss": 2.0163,
+      "step": 1410
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6268881006949795,
+      "learning_rate": 4.918927725993442e-05,
+      "loss": 2.2333,
+      "step": 1411
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6071670743541296,
+      "learning_rate": 4.918769859437232e-05,
+      "loss": 1.9968,
+      "step": 1412
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6715931927167169,
+      "learning_rate": 4.9186118418672844e-05,
+      "loss": 2.043,
+      "step": 1413
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.739076979411059,
+      "learning_rate": 4.918453673293463e-05,
+      "loss": 2.2027,
+      "step": 1414
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6128681262178067,
+      "learning_rate": 4.9182953537256435e-05,
+      "loss": 1.9508,
+      "step": 1415
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6802465883244089,
+      "learning_rate": 4.91813688317371e-05,
+      "loss": 2.0659,
+      "step": 1416
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.645269551867952,
+      "learning_rate": 4.917978261647558e-05,
+      "loss": 2.0554,
+      "step": 1417
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6956842109527737,
+      "learning_rate": 4.917819489157089e-05,
+      "loss": 2.0351,
+      "step": 1418
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6898499348151153,
+      "learning_rate": 4.917660565712216e-05,
+      "loss": 2.2164,
+      "step": 1419
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7366164056749555,
+      "learning_rate": 4.9175014913228634e-05,
+      "loss": 2.0809,
+      "step": 1420
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8131213082688976,
+      "learning_rate": 4.9173422659989607e-05,
+      "loss": 1.9771,
+      "step": 1421
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6265425338560084,
+      "learning_rate": 4.9171828897504495e-05,
+      "loss": 2.0688,
+      "step": 1422
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7843504494406331,
+      "learning_rate": 4.917023362587281e-05,
+      "loss": 2.2418,
+      "step": 1423
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6971086001904286,
+      "learning_rate": 4.916863684519414e-05,
+      "loss": 2.0333,
+      "step": 1424
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9228081100045981,
+      "learning_rate": 4.916703855556819e-05,
+      "loss": 2.0243,
+      "step": 1425
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8479742020759614,
+      "learning_rate": 4.916543875709473e-05,
+      "loss": 2.2056,
+      "step": 1426
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6598787123874069,
+      "learning_rate": 4.9163837449873665e-05,
+      "loss": 1.9996,
+      "step": 1427
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7174901121575203,
+      "learning_rate": 4.916223463400496e-05,
+      "loss": 2.0726,
+      "step": 1428
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6346016832985462,
+      "learning_rate": 4.9160630309588686e-05,
+      "loss": 2.0128,
+      "step": 1429
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6929303770388461,
+      "learning_rate": 4.9159024476725003e-05,
+      "loss": 1.9945,
+      "step": 1430
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6479037582618368,
+      "learning_rate": 4.9157417135514183e-05,
+      "loss": 2.2212,
+      "step": 1431
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7025884846329573,
+      "learning_rate": 4.915580828605656e-05,
+      "loss": 1.9632,
+      "step": 1432
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6762814430433215,
+      "learning_rate": 4.9154197928452596e-05,
+      "loss": 1.99,
+      "step": 1433
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.717427271152316,
+      "learning_rate": 4.915258606280283e-05,
+      "loss": 2.2395,
+      "step": 1434
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7645438601476247,
+      "learning_rate": 4.915097268920789e-05,
+      "loss": 2.0659,
+      "step": 1435
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7774831376577979,
+      "learning_rate": 4.914935780776851e-05,
+      "loss": 2.0418,
+      "step": 1436
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6832070153104234,
+      "learning_rate": 4.914774141858551e-05,
+      "loss": 2.041,
+      "step": 1437
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6548544928602852,
+      "learning_rate": 4.9146123521759824e-05,
+      "loss": 2.1845,
+      "step": 1438
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7154612609640302,
+      "learning_rate": 4.9144504117392445e-05,
+      "loss": 2.0099,
+      "step": 1439
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6506253952011838,
+      "learning_rate": 4.9142883205584485e-05,
+      "loss": 2.018,
+      "step": 1440
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6677169009202621,
+      "learning_rate": 4.9141260786437146e-05,
+      "loss": 2.0833,
+      "step": 1441
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6340480992538535,
+      "learning_rate": 4.913963686005172e-05,
+      "loss": 2.0133,
+      "step": 1442
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7369112995239134,
+      "learning_rate": 4.913801142652959e-05,
+      "loss": 2.2134,
+      "step": 1443
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7153467636897861,
+      "learning_rate": 4.913638448597227e-05,
+      "loss": 1.969,
+      "step": 1444
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8534291650550823,
+      "learning_rate": 4.913475603848129e-05,
+      "loss": 2.0239,
+      "step": 1445
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8527012168738066,
+      "learning_rate": 4.9133126084158344e-05,
+      "loss": 2.1673,
+      "step": 1446
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6865184574954233,
+      "learning_rate": 4.9131494623105204e-05,
+      "loss": 2.0749,
+      "step": 1447
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7413853463198135,
+      "learning_rate": 4.912986165542371e-05,
+      "loss": 2.0312,
+      "step": 1448
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8907502999290932,
+      "learning_rate": 4.912822718121583e-05,
+      "loss": 1.9743,
+      "step": 1449
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.62262153506801,
+      "learning_rate": 4.91265912005836e-05,
+      "loss": 1.9712,
+      "step": 1450
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.773709139725493,
+      "learning_rate": 4.9124953713629174e-05,
+      "loss": 2.2038,
+      "step": 1451
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7225878136381105,
+      "learning_rate": 4.9123314720454786e-05,
+      "loss": 2.0483,
+      "step": 1452
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6895708392708253,
+      "learning_rate": 4.912167422116275e-05,
+      "loss": 2.0664,
+      "step": 1453
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7912068401804979,
+      "learning_rate": 4.9120032215855496e-05,
+      "loss": 2.0234,
+      "step": 1454
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7033156723715096,
+      "learning_rate": 4.911838870463554e-05,
+      "loss": 2.2253,
+      "step": 1455
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7522270336473421,
+      "learning_rate": 4.911674368760551e-05,
+      "loss": 2.0282,
+      "step": 1456
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6380176066217332,
+      "learning_rate": 4.911509716486809e-05,
+      "loss": 1.9703,
+      "step": 1457
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8132472574000443,
+      "learning_rate": 4.911344913652609e-05,
+      "loss": 2.2366,
+      "step": 1458
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8299108877896236,
+      "learning_rate": 4.91117996026824e-05,
+      "loss": 2.0935,
+      "step": 1459
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7045345257515145,
+      "learning_rate": 4.9110148563439995e-05,
+      "loss": 2.0464,
+      "step": 1460
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8264982560829094,
+      "learning_rate": 4.9108496018901975e-05,
+      "loss": 2.0132,
+      "step": 1461
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6347887224176082,
+      "learning_rate": 4.9106841969171515e-05,
+      "loss": 1.9418,
+      "step": 1462
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8228208649537577,
+      "learning_rate": 4.9105186414351876e-05,
+      "loss": 2.1814,
+      "step": 1463
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6852758464567472,
+      "learning_rate": 4.910352935454642e-05,
+      "loss": 1.9875,
+      "step": 1464
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6862275519756383,
+      "learning_rate": 4.91018707898586e-05,
+      "loss": 1.991,
+      "step": 1465
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6894498481353553,
+      "learning_rate": 4.910021072039198e-05,
+      "loss": 2.0766,
+      "step": 1466
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6406473091023128,
+      "learning_rate": 4.90985491462502e-05,
+      "loss": 2.1514,
+      "step": 1467
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7072110393949574,
+      "learning_rate": 4.9096886067536996e-05,
+      "loss": 1.9942,
+      "step": 1468
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6918591279678531,
+      "learning_rate": 4.90952214843562e-05,
+      "loss": 1.993,
+      "step": 1469
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.663486934650943,
+      "learning_rate": 4.909355539681174e-05,
+      "loss": 2.032,
+      "step": 1470
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8722624576923758,
+      "learning_rate": 4.9091887805007646e-05,
+      "loss": 2.2325,
+      "step": 1471
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7249328263112711,
+      "learning_rate": 4.9090218709048016e-05,
+      "loss": 2.1094,
+      "step": 1472
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6907016675778929,
+      "learning_rate": 4.9088548109037075e-05,
+      "loss": 2.0002,
+      "step": 1473
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7851199550941093,
+      "learning_rate": 4.908687600507911e-05,
+      "loss": 2.0562,
+      "step": 1474
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6553147923177362,
+      "learning_rate": 4.908520239727853e-05,
+      "loss": 2.2066,
+      "step": 1475
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6799776154550105,
+      "learning_rate": 4.9083527285739815e-05,
+      "loss": 1.9599,
+      "step": 1476
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7863076043653596,
+      "learning_rate": 4.9081850670567556e-05,
+      "loss": 2.0393,
+      "step": 1477
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6203190189389729,
+      "learning_rate": 4.908017255186643e-05,
+      "loss": 2.0469,
+      "step": 1478
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7434557719813637,
+      "learning_rate": 4.907849292974121e-05,
+      "loss": 2.202,
+      "step": 1479
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.766558274567434,
+      "learning_rate": 4.9076811804296755e-05,
+      "loss": 2.0041,
+      "step": 1480
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.664942800646058,
+      "learning_rate": 4.907512917563802e-05,
+      "loss": 1.9843,
+      "step": 1481
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7293549543803105,
+      "learning_rate": 4.907344504387008e-05,
+      "loss": 2.0135,
+      "step": 1482
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.656950957100537,
+      "learning_rate": 4.907175940909807e-05,
+      "loss": 2.2146,
+      "step": 1483
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.643413637013205,
+      "learning_rate": 4.907007227142723e-05,
+      "loss": 2.0782,
+      "step": 1484
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6199355752174294,
+      "learning_rate": 4.906838363096289e-05,
+      "loss": 2.0068,
+      "step": 1485
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6248417641856743,
+      "learning_rate": 4.9066693487810486e-05,
+      "loss": 2.0399,
+      "step": 1486
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7214019985594473,
+      "learning_rate": 4.906500184207555e-05,
+      "loss": 2.2296,
+      "step": 1487
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.638379935770456,
+      "learning_rate": 4.906330869386367e-05,
+      "loss": 1.9561,
+      "step": 1488
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6697773568148789,
+      "learning_rate": 4.906161404328059e-05,
+      "loss": 1.9864,
+      "step": 1489
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5780295510364761,
+      "learning_rate": 4.905991789043209e-05,
+      "loss": 2.0631,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6523151894995519,
+      "learning_rate": 4.905822023542407e-05,
+      "loss": 2.2063,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6405427837066013,
+      "learning_rate": 4.905652107836254e-05,
+      "loss": 2.0093,
+      "step": 1492
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7534136238793245,
+      "learning_rate": 4.905482041935356e-05,
+      "loss": 2.0011,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.670397597019326,
+      "learning_rate": 4.905311825850333e-05,
+      "loss": 2.0301,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7016625641974874,
+      "learning_rate": 4.905141459591811e-05,
+      "loss": 2.1791,
+      "step": 1495
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7376418064013709,
+      "learning_rate": 4.904970943170427e-05,
+      "loss": 2.0212,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6303882548001669,
+      "learning_rate": 4.904800276596828e-05,
+      "loss": 2.0689,
+      "step": 1497
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.750003663635683,
+      "learning_rate": 4.904629459881668e-05,
+      "loss": 2.05,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7356302381575069,
+      "learning_rate": 4.904458493035612e-05,
+      "loss": 2.2429,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6686318271181354,
+      "learning_rate": 4.9042873760693345e-05,
+      "loss": 1.9609,
+      "step": 1500
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7809777811436641,
+      "learning_rate": 4.9041161089935194e-05,
+      "loss": 2.01,
+      "step": 1501
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6455630936021807,
+      "learning_rate": 4.903944691818859e-05,
+      "loss": 1.9877,
+      "step": 1502
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7342726700723904,
+      "learning_rate": 4.903773124556057e-05,
+      "loss": 2.222,
+      "step": 1503
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7438400129643592,
+      "learning_rate": 4.903601407215822e-05,
+      "loss": 2.0578,
+      "step": 1504
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6225040135293964,
+      "learning_rate": 4.903429539808878e-05,
+      "loss": 1.988,
+      "step": 1505
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7388586794179391,
+      "learning_rate": 4.903257522345954e-05,
+      "loss": 2.0275,
+      "step": 1506
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.676428974597801,
+      "learning_rate": 4.903085354837791e-05,
+      "loss": 2.1751,
+      "step": 1507
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6581480506518382,
+      "learning_rate": 4.902913037295136e-05,
+      "loss": 1.9707,
+      "step": 1508
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7250774251461055,
+      "learning_rate": 4.9027405697287495e-05,
+      "loss": 2.0678,
+      "step": 1509
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.756225713435815,
+      "learning_rate": 4.902567952149398e-05,
+      "loss": 1.9985,
+      "step": 1510
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6909473493638949,
+      "learning_rate": 4.902395184567859e-05,
+      "loss": 2.1961,
+      "step": 1511
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6870684254919689,
+      "learning_rate": 4.9022222669949197e-05,
+      "loss": 2.0012,
+      "step": 1512
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.675522815320918,
+      "learning_rate": 4.902049199441376e-05,
+      "loss": 1.9893,
+      "step": 1513
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.642581920324174,
+      "learning_rate": 4.901875981918033e-05,
+      "loss": 2.0157,
+      "step": 1514
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6328546304367614,
+      "learning_rate": 4.901702614435704e-05,
+      "loss": 2.2229,
+      "step": 1515
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6384657471304721,
+      "learning_rate": 4.9015290970052154e-05,
+      "loss": 2.05,
+      "step": 1516
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7481845841743291,
+      "learning_rate": 4.9013554296374e-05,
+      "loss": 2.019,
+      "step": 1517
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6071072945800561,
+      "learning_rate": 4.9011816123430984e-05,
+      "loss": 2.0117,
+      "step": 1518
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7733484827537251,
+      "learning_rate": 4.901007645133166e-05,
+      "loss": 2.2272,
+      "step": 1519
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7191250076478276,
+      "learning_rate": 4.900833528018463e-05,
+      "loss": 2.0274,
+      "step": 1520
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7835381137027306,
+      "learning_rate": 4.9006592610098585e-05,
+      "loss": 2.0886,
+      "step": 1521
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8706074794599856,
+      "learning_rate": 4.900484844118235e-05,
+      "loss": 2.013,
+      "step": 1522
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.728879837046103,
+      "learning_rate": 4.9003102773544806e-05,
+      "loss": 2.2453,
+      "step": 1523
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9895335942455465,
+      "learning_rate": 4.900135560729496e-05,
+      "loss": 2.0735,
+      "step": 1524
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6775610701308648,
+      "learning_rate": 4.8999606942541876e-05,
+      "loss": 2.0092,
+      "step": 1525
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8945486850427139,
+      "learning_rate": 4.899785677939474e-05,
+      "loss": 2.0308,
+      "step": 1526
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7183996339804891,
+      "learning_rate": 4.899610511796282e-05,
+      "loss": 2.1864,
+      "step": 1527
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6335110279820241,
+      "learning_rate": 4.899435195835548e-05,
+      "loss": 2.0696,
+      "step": 1528
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7512250537838179,
+      "learning_rate": 4.899259730068217e-05,
+      "loss": 2.025,
+      "step": 1529
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6574494636396642,
+      "learning_rate": 4.8990841145052445e-05,
+      "loss": 2.0099,
+      "step": 1530
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8200064744820464,
+      "learning_rate": 4.898908349157596e-05,
+      "loss": 2.1645,
+      "step": 1531
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8586679926042361,
+      "learning_rate": 4.898732434036244e-05,
+      "loss": 2.0106,
+      "step": 1532
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6652020727271293,
+      "learning_rate": 4.8985563691521716e-05,
+      "loss": 2.0548,
+      "step": 1533
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.02539381082077,
+      "learning_rate": 4.898380154516371e-05,
+      "loss": 2.0734,
+      "step": 1534
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8656481895243578,
+      "learning_rate": 4.898203790139846e-05,
+      "loss": 2.196,
+      "step": 1535
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8556766210930827,
+      "learning_rate": 4.8980272760336055e-05,
+      "loss": 2.0445,
+      "step": 1536
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9982375532611854,
+      "learning_rate": 4.897850612208671e-05,
+      "loss": 2.0179,
+      "step": 1537
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7142998302345301,
+      "learning_rate": 4.8976737986760725e-05,
+      "loss": 2.0623,
+      "step": 1538
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0613700319277675,
+      "learning_rate": 4.897496835446847e-05,
+      "loss": 2.2131,
+      "step": 1539
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6915873793323681,
+      "learning_rate": 4.8973197225320473e-05,
+      "loss": 2.1061,
+      "step": 1540
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9137395221800345,
+      "learning_rate": 4.897142459942728e-05,
+      "loss": 1.9875,
+      "step": 1541
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8664656178549999,
+      "learning_rate": 4.896965047689957e-05,
+      "loss": 1.9654,
+      "step": 1542
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8136781489277275,
+      "learning_rate": 4.896787485784811e-05,
+      "loss": 2.1726,
+      "step": 1543
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0280438076490326,
+      "learning_rate": 4.8966097742383765e-05,
+      "loss": 2.02,
+      "step": 1544
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7813941109639757,
+      "learning_rate": 4.8964319130617475e-05,
+      "loss": 1.9389,
+      "step": 1545
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.035533366829317,
+      "learning_rate": 4.8962539022660304e-05,
+      "loss": 2.0627,
+      "step": 1546
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7602538631759572,
+      "learning_rate": 4.896075741862338e-05,
+      "loss": 2.1579,
+      "step": 1547
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.789342306532993,
+      "learning_rate": 4.895897431861793e-05,
+      "loss": 1.9886,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9898172957004133,
+      "learning_rate": 4.8957189722755293e-05,
+      "loss": 1.9715,
+      "step": 1549
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7029873685542755,
+      "learning_rate": 4.895540363114688e-05,
+      "loss": 1.9931,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8891478307082431,
+      "learning_rate": 4.895361604390421e-05,
+      "loss": 2.2425,
+      "step": 1551
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7836904109212032,
+      "learning_rate": 4.895182696113888e-05,
+      "loss": 2.0798,
+      "step": 1552
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7116327861932176,
+      "learning_rate": 4.89500363829626e-05,
+      "loss": 1.9739,
+      "step": 1553
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9542128255001108,
+      "learning_rate": 4.894824430948716e-05,
+      "loss": 1.9707,
+      "step": 1554
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7535971870968055,
+      "learning_rate": 4.894645074082445e-05,
+      "loss": 2.1819,
+      "step": 1555
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8475963819539005,
+      "learning_rate": 4.8944655677086435e-05,
+      "loss": 2.028,
+      "step": 1556
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8359139022070011,
+      "learning_rate": 4.894285911838521e-05,
+      "loss": 2.032,
+      "step": 1557
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6385252085821237,
+      "learning_rate": 4.894106106483292e-05,
+      "loss": 2.017,
+      "step": 1558
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8793057892398624,
+      "learning_rate": 4.893926151654185e-05,
+      "loss": 2.229,
+      "step": 1559
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6070047003070554,
+      "learning_rate": 4.893746047362432e-05,
+      "loss": 2.0172,
+      "step": 1560
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8313779730718615,
+      "learning_rate": 4.893565793619281e-05,
+      "loss": 2.0286,
+      "step": 1561
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6392284320226778,
+      "learning_rate": 4.893385390435984e-05,
+      "loss": 1.9973,
+      "step": 1562
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7276386086022384,
+      "learning_rate": 4.8932048378238046e-05,
+      "loss": 2.1987,
+      "step": 1563
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7078267249440373,
+      "learning_rate": 4.893024135794015e-05,
+      "loss": 2.0001,
+      "step": 1564
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5761364664713644,
+      "learning_rate": 4.8928432843578996e-05,
+      "loss": 2.0929,
+      "step": 1565
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.694638786286016,
+      "learning_rate": 4.8926622835267463e-05,
+      "loss": 2.0052,
+      "step": 1566
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.602180231384444,
+      "learning_rate": 4.892481133311858e-05,
+      "loss": 2.1902,
+      "step": 1567
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6291354898008008,
+      "learning_rate": 4.892299833724544e-05,
+      "loss": 1.9592,
+      "step": 1568
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.666561263907537,
+      "learning_rate": 4.892118384776124e-05,
+      "loss": 1.9886,
+      "step": 1569
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7256364950114741,
+      "learning_rate": 4.891936786477925e-05,
+      "loss": 2.0393,
+      "step": 1570
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6332533624869824,
+      "learning_rate": 4.891755038841288e-05,
+      "loss": 2.2497,
+      "step": 1571
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7873463478785132,
+      "learning_rate": 4.891573141877557e-05,
+      "loss": 2.0263,
+      "step": 1572
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6103041586379244,
+      "learning_rate": 4.891391095598089e-05,
+      "loss": 2.0066,
+      "step": 1573
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.9432593701523323,
+      "learning_rate": 4.8912089000142524e-05,
+      "loss": 2.0334,
+      "step": 1574
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6640726176650178,
+      "learning_rate": 4.89102655513742e-05,
+      "loss": 2.2272,
+      "step": 1575
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6726939452587867,
+      "learning_rate": 4.8908440609789776e-05,
+      "loss": 2.045,
+      "step": 1576
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7856395934724953,
+      "learning_rate": 4.890661417550319e-05,
+      "loss": 2.0774,
+      "step": 1577
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.615826528286066,
+      "learning_rate": 4.8904786248628464e-05,
+      "loss": 2.0443,
+      "step": 1578
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8122669988640299,
+      "learning_rate": 4.8902956829279735e-05,
+      "loss": 2.2142,
+      "step": 1579
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7039585237719892,
+      "learning_rate": 4.890112591757121e-05,
+      "loss": 1.9933,
+      "step": 1580
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7762731046148803,
+      "learning_rate": 4.889929351361721e-05,
+      "loss": 2.0444,
+      "step": 1581
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7212878374222162,
+      "learning_rate": 4.889745961753213e-05,
+      "loss": 2.012,
+      "step": 1582
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.634371596967165,
+      "learning_rate": 4.889562422943047e-05,
+      "loss": 2.24,
+      "step": 1583
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6477497512795606,
+      "learning_rate": 4.8893787349426834e-05,
+      "loss": 2.0603,
+      "step": 1584
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6675986346829489,
+      "learning_rate": 4.889194897763589e-05,
+      "loss": 2.0018,
+      "step": 1585
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6876871842471952,
+      "learning_rate": 4.8890109114172424e-05,
+      "loss": 1.9922,
+      "step": 1586
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7691590454992162,
+      "learning_rate": 4.8888267759151294e-05,
+      "loss": 2.1962,
+      "step": 1587
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.65129547971366,
+      "learning_rate": 4.888642491268748e-05,
+      "loss": 2.0185,
+      "step": 1588
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6647064232407159,
+      "learning_rate": 4.8884580574896035e-05,
+      "loss": 2.0428,
+      "step": 1589
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7667339058106031,
+      "learning_rate": 4.888273474589209e-05,
+      "loss": 2.0748,
+      "step": 1590
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7962866151133833,
+      "learning_rate": 4.888088742579091e-05,
+      "loss": 2.179,
+      "step": 1591
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6636730443976754,
+      "learning_rate": 4.887903861470783e-05,
+      "loss": 2.0091,
+      "step": 1592
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8112790103102209,
+      "learning_rate": 4.8877188312758256e-05,
+      "loss": 2.0187,
+      "step": 1593
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.676995185450621,
+      "learning_rate": 4.887533652005774e-05,
+      "loss": 1.9847,
+      "step": 1594
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.964438644471062,
+      "learning_rate": 4.8873483236721876e-05,
+      "loss": 2.1875,
+      "step": 1595
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6755408188736353,
+      "learning_rate": 4.887162846286638e-05,
+      "loss": 2.0883,
+      "step": 1596
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8683921549347674,
+      "learning_rate": 4.8869772198607055e-05,
+      "loss": 2.0016,
+      "step": 1597
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7126116507965032,
+      "learning_rate": 4.886791444405979e-05,
+      "loss": 2.0219,
+      "step": 1598
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.678864731622033,
+      "learning_rate": 4.886605519934058e-05,
+      "loss": 2.2106,
+      "step": 1599
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6742139819262983,
+      "learning_rate": 4.886419446456549e-05,
+      "loss": 2.0287,
+      "step": 1600
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7170332726417835,
+      "learning_rate": 4.8862332239850703e-05,
+      "loss": 1.9693,
+      "step": 1601
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6758436084036683,
+      "learning_rate": 4.886046852531249e-05,
+      "loss": 2.0744,
+      "step": 1602
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6296118483492044,
+      "learning_rate": 4.885860332106721e-05,
+      "loss": 1.9917,
+      "step": 1603
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6878704827091733,
+      "learning_rate": 4.88567366272313e-05,
+      "loss": 2.1597,
+      "step": 1604
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.665578910103583,
+      "learning_rate": 4.885486844392132e-05,
+      "loss": 1.9791,
+      "step": 1605
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7055552918869917,
+      "learning_rate": 4.885299877125391e-05,
+      "loss": 2.0025,
+      "step": 1606
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6842684815151263,
+      "learning_rate": 4.885112760934579e-05,
+      "loss": 2.2387,
+      "step": 1607
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7917567223357809,
+      "learning_rate": 4.884925495831379e-05,
+      "loss": 2.0838,
+      "step": 1608
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6622367410573917,
+      "learning_rate": 4.884738081827484e-05,
+      "loss": 1.9954,
+      "step": 1609
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8085825632022866,
+      "learning_rate": 4.8845505189345934e-05,
+      "loss": 2.0259,
+      "step": 1610
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5928021443823616,
+      "learning_rate": 4.8843628071644165e-05,
+      "loss": 2.1899,
+      "step": 1611
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7038223703472594,
+      "learning_rate": 4.8841749465286756e-05,
+      "loss": 1.9877,
+      "step": 1612
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6738111926743082,
+      "learning_rate": 4.8839869370390975e-05,
+      "loss": 1.9798,
+      "step": 1613
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6367237971210636,
+      "learning_rate": 4.883798778707422e-05,
+      "loss": 2.0266,
+      "step": 1614
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6578258973002554,
+      "learning_rate": 4.8836104715453964e-05,
+      "loss": 2.0078,
+      "step": 1615
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6619893085531953,
+      "learning_rate": 4.883422015564776e-05,
+      "loss": 2.2095,
+      "step": 1616
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7054812614205657,
+      "learning_rate": 4.8832334107773284e-05,
+      "loss": 1.9746,
+      "step": 1617
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6519881658362222,
+      "learning_rate": 4.8830446571948276e-05,
+      "loss": 1.9801,
+      "step": 1618
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.646432943931197,
+      "learning_rate": 4.882855754829059e-05,
+      "loss": 2.2014,
+      "step": 1619
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6345553994015883,
+      "learning_rate": 4.882666703691817e-05,
+      "loss": 1.9976,
+      "step": 1620
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6086523847171076,
+      "learning_rate": 4.882477503794905e-05,
+      "loss": 2.1228,
+      "step": 1621
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7457897545748834,
+      "learning_rate": 4.882288155150134e-05,
+      "loss": 2.0059,
+      "step": 1622
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6650492792933501,
+      "learning_rate": 4.8820986577693276e-05,
+      "loss": 2.1606,
+      "step": 1623
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6978977152072615,
+      "learning_rate": 4.881909011664316e-05,
+      "loss": 2.0435,
+      "step": 1624
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6293032714526053,
+      "learning_rate": 4.8817192168469394e-05,
+      "loss": 2.0217,
+      "step": 1625
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6689911255570921,
+      "learning_rate": 4.881529273329048e-05,
+      "loss": 1.9957,
+      "step": 1626
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6303067500755346,
+      "learning_rate": 4.881339181122501e-05,
+      "loss": 2.1098,
+      "step": 1627
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6134512699079463,
+      "learning_rate": 4.881148940239165e-05,
+      "loss": 2.1788,
+      "step": 1628
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6117785189284667,
+      "learning_rate": 4.880958550690919e-05,
+      "loss": 1.9818,
+      "step": 1629
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6822191618286261,
+      "learning_rate": 4.8807680124896494e-05,
+      "loss": 2.0655,
+      "step": 1630
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.701812133039931,
+      "learning_rate": 4.880577325647252e-05,
+      "loss": 2.1826,
+      "step": 1631
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6547127302048279,
+      "learning_rate": 4.8803864901756334e-05,
+      "loss": 1.9343,
+      "step": 1632
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6370819513658245,
+      "learning_rate": 4.880195506086707e-05,
+      "loss": 2.0536,
+      "step": 1633
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6695398593518003,
+      "learning_rate": 4.8800043733923974e-05,
+      "loss": 1.954,
+      "step": 1634
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7313766525092967,
+      "learning_rate": 4.879813092104637e-05,
+      "loss": 2.0286,
+      "step": 1635
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7536167257279901,
+      "learning_rate": 4.8796216622353686e-05,
+      "loss": 2.2029,
+      "step": 1636
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.669188585636682,
+      "learning_rate": 4.879430083796545e-05,
+      "loss": 2.0162,
+      "step": 1637
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6985841310891843,
+      "learning_rate": 4.879238356800126e-05,
+      "loss": 1.981,
+      "step": 1638
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7026379047899727,
+      "learning_rate": 4.879046481258081e-05,
+      "loss": 2.0793,
+      "step": 1639
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7622717385476238,
+      "learning_rate": 4.8788544571823915e-05,
+      "loss": 2.1566,
+      "step": 1640
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6920474094895059,
+      "learning_rate": 4.878662284585046e-05,
+      "loss": 2.0048,
+      "step": 1641
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6975053090548534,
+      "learning_rate": 4.878469963478042e-05,
+      "loss": 1.9782,
+      "step": 1642
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7308433897087943,
+      "learning_rate": 4.878277493873388e-05,
+      "loss": 2.2721,
+      "step": 1643
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7687378647604103,
+      "learning_rate": 4.878084875783099e-05,
+      "loss": 2.0045,
+      "step": 1644
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.721376253300633,
+      "learning_rate": 4.8778921092192014e-05,
+      "loss": 2.0155,
+      "step": 1645
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7226996388170284,
+      "learning_rate": 4.877699194193731e-05,
+      "loss": 2.0904,
+      "step": 1646
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7450728313212015,
+      "learning_rate": 4.8775061307187333e-05,
+      "loss": 1.9734,
+      "step": 1647
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6431438065196093,
+      "learning_rate": 4.8773129188062593e-05,
+      "loss": 2.1727,
+      "step": 1648
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6648747749361291,
+      "learning_rate": 4.877119558468374e-05,
+      "loss": 2.0085,
+      "step": 1649
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7009929775636612,
+      "learning_rate": 4.8769260497171495e-05,
+      "loss": 2.0091,
+      "step": 1650
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7118196263174075,
+      "learning_rate": 4.876732392564667e-05,
+      "loss": 2.217,
+      "step": 1651
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8675359410523498,
+      "learning_rate": 4.8765385870230176e-05,
+      "loss": 2.0873,
+      "step": 1652
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6562636976045251,
+      "learning_rate": 4.876344633104301e-05,
+      "loss": 2.0165,
+      "step": 1653
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8688049381520782,
+      "learning_rate": 4.876150530820627e-05,
+      "loss": 2.0225,
+      "step": 1654
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6796720403257119,
+      "learning_rate": 4.875956280184113e-05,
+      "loss": 2.236,
+      "step": 1655
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6745944701673521,
+      "learning_rate": 4.875761881206888e-05,
+      "loss": 2.0252,
+      "step": 1656
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6894460831030285,
+      "learning_rate": 4.875567333901089e-05,
+      "loss": 2.0136,
+      "step": 1657
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6499894300623692,
+      "learning_rate": 4.875372638278862e-05,
+      "loss": 2.0469,
+      "step": 1658
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7034366928974665,
+      "learning_rate": 4.8751777943523634e-05,
+      "loss": 2.012,
+      "step": 1659
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6023193271663739,
+      "learning_rate": 4.8749828021337575e-05,
+      "loss": 2.1878,
+      "step": 1660
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7701695634234397,
+      "learning_rate": 4.8747876616352186e-05,
+      "loss": 1.9906,
+      "step": 1661
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7107925637929976,
+      "learning_rate": 4.8745923728689305e-05,
+      "loss": 2.0193,
+      "step": 1662
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7108363744203235,
+      "learning_rate": 4.874396935847085e-05,
+      "loss": 2.2194,
+      "step": 1663
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7780080001755454,
+      "learning_rate": 4.874201350581885e-05,
+      "loss": 2.1008,
+      "step": 1664
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7751381995421351,
+      "learning_rate": 4.87400561708554e-05,
+      "loss": 1.9841,
+      "step": 1665
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7150848145473261,
+      "learning_rate": 4.873809735370273e-05,
+      "loss": 2.039,
+      "step": 1666
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7543903194220497,
+      "learning_rate": 4.8736137054483114e-05,
+      "loss": 2.0149,
+      "step": 1667
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5988994073553893,
+      "learning_rate": 4.873417527331896e-05,
+      "loss": 2.1877,
+      "step": 1668
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6758694067084882,
+      "learning_rate": 4.873221201033273e-05,
+      "loss": 1.9786,
+      "step": 1669
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7166566400845973,
+      "learning_rate": 4.873024726564702e-05,
+      "loss": 2.0844,
+      "step": 1670
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.614621937838585,
+      "learning_rate": 4.8728281039384485e-05,
+      "loss": 1.9769,
+      "step": 1671
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7908840912065147,
+      "learning_rate": 4.8726313331667884e-05,
+      "loss": 2.1851,
+      "step": 1672
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8567382653011144,
+      "learning_rate": 4.872434414262008e-05,
+      "loss": 2.0293,
+      "step": 1673
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6801281216716422,
+      "learning_rate": 4.8722373472363995e-05,
+      "loss": 2.0525,
+      "step": 1674
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9845661601447437,
+      "learning_rate": 4.872040132102269e-05,
+      "loss": 2.2061,
+      "step": 1675
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8796047672036053,
+      "learning_rate": 4.871842768871928e-05,
+      "loss": 2.0246,
+      "step": 1676
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8034968704134742,
+      "learning_rate": 4.8716452575576996e-05,
+      "loss": 2.0719,
+      "step": 1677
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.917931829774112,
+      "learning_rate": 4.871447598171914e-05,
+      "loss": 2.0193,
+      "step": 1678
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6950799806501033,
+      "learning_rate": 4.8712497907269136e-05,
+      "loss": 1.9867,
+      "step": 1679
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7494405794121956,
+      "learning_rate": 4.871051835235047e-05,
+      "loss": 2.2111,
+      "step": 1680
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6827160001705807,
+      "learning_rate": 4.870853731708674e-05,
+      "loss": 1.9973,
+      "step": 1681
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6647011152862121,
+      "learning_rate": 4.870655480160162e-05,
+      "loss": 1.9859,
+      "step": 1682
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.617967958271628,
+      "learning_rate": 4.870457080601891e-05,
+      "loss": 2.0345,
+      "step": 1683
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6197415603830477,
+      "learning_rate": 4.870258533046245e-05,
+      "loss": 2.1756,
+      "step": 1684
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6347939948217158,
+      "learning_rate": 4.870059837505623e-05,
+      "loss": 1.977,
+      "step": 1685
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6514836045287381,
+      "learning_rate": 4.869860993992428e-05,
+      "loss": 2.0005,
+      "step": 1686
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6961361087451419,
+      "learning_rate": 4.8696620025190745e-05,
+      "loss": 2.043,
+      "step": 1687
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9934656610936521,
+      "learning_rate": 4.869462863097989e-05,
+      "loss": 2.1959,
+      "step": 1688
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6679691079294406,
+      "learning_rate": 4.869263575741602e-05,
+      "loss": 2.0317,
+      "step": 1689
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.897886914361417,
+      "learning_rate": 4.869064140462357e-05,
+      "loss": 2.044,
+      "step": 1690
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8105274619240014,
+      "learning_rate": 4.8688645572727056e-05,
+      "loss": 2.0175,
+      "step": 1691
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8083769139633945,
+      "learning_rate": 4.868664826185108e-05,
+      "loss": 2.211,
+      "step": 1692
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8407175834469127,
+      "learning_rate": 4.868464947212035e-05,
+      "loss": 1.962,
+      "step": 1693
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7955257249228689,
+      "learning_rate": 4.868264920365965e-05,
+      "loss": 2.0583,
+      "step": 1694
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7256681543945334,
+      "learning_rate": 4.8680647456593874e-05,
+      "loss": 2.0863,
+      "step": 1695
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7678117183049155,
+      "learning_rate": 4.867864423104799e-05,
+      "loss": 2.1936,
+      "step": 1696
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8129346087303382,
+      "learning_rate": 4.867663952714707e-05,
+      "loss": 2.0104,
+      "step": 1697
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.83111487245195,
+      "learning_rate": 4.8674633345016286e-05,
+      "loss": 1.978,
+      "step": 1698
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9700786480357096,
+      "learning_rate": 4.867262568478088e-05,
+      "loss": 2.0256,
+      "step": 1699
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7269961196035782,
+      "learning_rate": 4.867061654656621e-05,
+      "loss": 2.1374,
+      "step": 1700
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0480162515968983,
+      "learning_rate": 4.86686059304977e-05,
+      "loss": 2.0581,
+      "step": 1701
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6418276441783758,
+      "learning_rate": 4.8666593836700894e-05,
+      "loss": 2.0294,
+      "step": 1702
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8910414350926767,
+      "learning_rate": 4.86645802653014e-05,
+      "loss": 2.0612,
+      "step": 1703
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7781934119691012,
+      "learning_rate": 4.866256521642495e-05,
+      "loss": 2.1672,
+      "step": 1704
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7732012912474907,
+      "learning_rate": 4.866054869019735e-05,
+      "loss": 1.9943,
+      "step": 1705
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7562789775001519,
+      "learning_rate": 4.865853068674449e-05,
+      "loss": 1.9888,
+      "step": 1706
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8698717910725112,
+      "learning_rate": 4.865651120619237e-05,
+      "loss": 2.2336,
+      "step": 1707
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0561513301322838,
+      "learning_rate": 4.8654490248667066e-05,
+      "loss": 2.0689,
+      "step": 1708
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6765396813612639,
+      "learning_rate": 4.865246781429476e-05,
+      "loss": 1.9957,
+      "step": 1709
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8605045789616634,
+      "learning_rate": 4.865044390320173e-05,
+      "loss": 2.0038,
+      "step": 1710
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7478064347571007,
+      "learning_rate": 4.8648418515514324e-05,
+      "loss": 1.9915,
+      "step": 1711
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6858473212893377,
+      "learning_rate": 4.8646391651359e-05,
+      "loss": 2.1879,
+      "step": 1712
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8456511077570654,
+      "learning_rate": 4.864436331086231e-05,
+      "loss": 2.0012,
+      "step": 1713
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5993920168755866,
+      "learning_rate": 4.864233349415088e-05,
+      "loss": 2.0795,
+      "step": 1714
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6367367952583745,
+      "learning_rate": 4.864030220135144e-05,
+      "loss": 1.9592,
+      "step": 1715
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7510258089623232,
+      "learning_rate": 4.863826943259082e-05,
+      "loss": 2.2426,
+      "step": 1716
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6964838031100102,
+      "learning_rate": 4.8636235187995936e-05,
+      "loss": 1.9594,
+      "step": 1717
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7643991946731155,
+      "learning_rate": 4.863419946769379e-05,
+      "loss": 2.0051,
+      "step": 1718
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7387461030209671,
+      "learning_rate": 4.863216227181147e-05,
+      "loss": 1.9968,
+      "step": 1719
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7309874893534596,
+      "learning_rate": 4.863012360047618e-05,
+      "loss": 2.2375,
+      "step": 1720
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8503260656207527,
+      "learning_rate": 4.86280834538152e-05,
+      "loss": 2.0143,
+      "step": 1721
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6916383412403738,
+      "learning_rate": 4.862604183195591e-05,
+      "loss": 2.0233,
+      "step": 1722
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7300764838506615,
+      "learning_rate": 4.862399873502576e-05,
+      "loss": 2.0122,
+      "step": 1723
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6441696230647352,
+      "learning_rate": 4.862195416315232e-05,
+      "loss": 2.2013,
+      "step": 1724
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8475887569106031,
+      "learning_rate": 4.861990811646324e-05,
+      "loss": 1.9869,
+      "step": 1725
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6608040455635477,
+      "learning_rate": 4.861786059508626e-05,
+      "loss": 2.0757,
+      "step": 1726
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7768408457895387,
+      "learning_rate": 4.861581159914923e-05,
+      "loss": 2.0192,
+      "step": 1727
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7510877715204681,
+      "learning_rate": 4.8613761128780055e-05,
+      "loss": 2.2371,
+      "step": 1728
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7517153443662197,
+      "learning_rate": 4.861170918410678e-05,
+      "loss": 2.0471,
+      "step": 1729
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7452508890509288,
+      "learning_rate": 4.8609655765257486e-05,
+      "loss": 1.9749,
+      "step": 1730
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6394154916518292,
+      "learning_rate": 4.860760087236039e-05,
+      "loss": 1.9623,
+      "step": 1731
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.775191706083146,
+      "learning_rate": 4.8605544505543796e-05,
+      "loss": 2.2399,
+      "step": 1732
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.732192740280029,
+      "learning_rate": 4.860348666493609e-05,
+      "loss": 1.9739,
+      "step": 1733
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6941870679191552,
+      "learning_rate": 4.860142735066574e-05,
+      "loss": 1.9561,
+      "step": 1734
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7763156501705997,
+      "learning_rate": 4.859936656286132e-05,
+      "loss": 1.9537,
+      "step": 1735
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6504050006754993,
+      "learning_rate": 4.85973043016515e-05,
+      "loss": 2.1775,
+      "step": 1736
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7003261217402635,
+      "learning_rate": 4.8595240567165036e-05,
+      "loss": 1.9991,
+      "step": 1737
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6667079562988385,
+      "learning_rate": 4.859317535953077e-05,
+      "loss": 2.027,
+      "step": 1738
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6341486741370379,
+      "learning_rate": 4.859110867887764e-05,
+      "loss": 2.0825,
+      "step": 1739
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6467983622511467,
+      "learning_rate": 4.858904052533469e-05,
+      "loss": 2.1926,
+      "step": 1740
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.658715271192428,
+      "learning_rate": 4.858697089903102e-05,
+      "loss": 2.0488,
+      "step": 1741
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6425733424568277,
+      "learning_rate": 4.8584899800095864e-05,
+      "loss": 2.0168,
+      "step": 1742
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6061894869884389,
+      "learning_rate": 4.858282722865852e-05,
+      "loss": 2.0345,
+      "step": 1743
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.626872384955726,
+      "learning_rate": 4.85807531848484e-05,
+      "loss": 2.2092,
+      "step": 1744
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5975668826038398,
+      "learning_rate": 4.857867766879498e-05,
+      "loss": 2.1198,
+      "step": 1745
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6837035240959958,
+      "learning_rate": 4.8576600680627855e-05,
+      "loss": 1.9965,
+      "step": 1746
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6679771309077187,
+      "learning_rate": 4.8574522220476693e-05,
+      "loss": 1.9785,
+      "step": 1747
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6578942978104355,
+      "learning_rate": 4.857244228847126e-05,
+      "loss": 2.2314,
+      "step": 1748
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7001025761483601,
+      "learning_rate": 4.857036088474142e-05,
+      "loss": 1.9813,
+      "step": 1749
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6140900857659639,
+      "learning_rate": 4.856827800941711e-05,
+      "loss": 1.9905,
+      "step": 1750
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7319357294825499,
+      "learning_rate": 4.85661936626284e-05,
+      "loss": 2.07,
+      "step": 1751
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6979669462922744,
+      "learning_rate": 4.85641078445054e-05,
+      "loss": 2.1857,
+      "step": 1752
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0420516663690453,
+      "learning_rate": 4.856202055517834e-05,
+      "loss": 1.9879,
+      "step": 1753
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.766404546105132,
+      "learning_rate": 4.855993179477755e-05,
+      "loss": 1.9905,
+      "step": 1754
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5866602742975688,
+      "learning_rate": 4.855784156343342e-05,
+      "loss": 1.999,
+      "step": 1755
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6937368995058981,
+      "learning_rate": 4.8555749861276464e-05,
+      "loss": 2.2051,
+      "step": 1756
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.597893382237688,
+      "learning_rate": 4.8553656688437285e-05,
+      "loss": 2.0892,
+      "step": 1757
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6354028578838833,
+      "learning_rate": 4.855156204504655e-05,
+      "loss": 1.9973,
+      "step": 1758
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8012703333719722,
+      "learning_rate": 4.854946593123505e-05,
+      "loss": 2.0129,
+      "step": 1759
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6128584734467191,
+      "learning_rate": 4.854736834713364e-05,
+      "loss": 2.193,
+      "step": 1760
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6822113073547863,
+      "learning_rate": 4.8545269292873296e-05,
+      "loss": 2.0225,
+      "step": 1761
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5919420980363895,
+      "learning_rate": 4.854316876858506e-05,
+      "loss": 1.9967,
+      "step": 1762
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6405758031831892,
+      "learning_rate": 4.854106677440008e-05,
+      "loss": 1.9971,
+      "step": 1763
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7483749278589876,
+      "learning_rate": 4.853896331044959e-05,
+      "loss": 2.2328,
+      "step": 1764
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8002111886575328,
+      "learning_rate": 4.8536858376864926e-05,
+      "loss": 2.013,
+      "step": 1765
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.671024819792495,
+      "learning_rate": 4.8534751973777503e-05,
+      "loss": 2.0494,
+      "step": 1766
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6801693266773705,
+      "learning_rate": 4.8532644101318826e-05,
+      "loss": 2.0053,
+      "step": 1767
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7351967137811298,
+      "learning_rate": 4.853053475962051e-05,
+      "loss": 2.2053,
+      "step": 1768
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.802477307371425,
+      "learning_rate": 4.852842394881423e-05,
+      "loss": 2.0428,
+      "step": 1769
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6622435680034102,
+      "learning_rate": 4.85263116690318e-05,
+      "loss": 2.0499,
+      "step": 1770
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2915161924095737,
+      "learning_rate": 4.852419792040507e-05,
+      "loss": 2.012,
+      "step": 1771
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6785565182063865,
+      "learning_rate": 4.8522082703066035e-05,
+      "loss": 2.1959,
+      "step": 1772
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7987657957515331,
+      "learning_rate": 4.851996601714674e-05,
+      "loss": 1.9894,
+      "step": 1773
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7400324296134823,
+      "learning_rate": 4.851784786277935e-05,
+      "loss": 2.0346,
+      "step": 1774
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6903197819339421,
+      "learning_rate": 4.85157282400961e-05,
+      "loss": 1.9459,
+      "step": 1775
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7810173412521904,
+      "learning_rate": 4.851360714922933e-05,
+      "loss": 2.269,
+      "step": 1776
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6619569884971317,
+      "learning_rate": 4.851148459031147e-05,
+      "loss": 1.9881,
+      "step": 1777
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7205435777252497,
+      "learning_rate": 4.8509360563475045e-05,
+      "loss": 1.9773,
+      "step": 1778
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8441379042372801,
+      "learning_rate": 4.850723506885265e-05,
+      "loss": 2.01,
+      "step": 1779
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6344637065548948,
+      "learning_rate": 4.850510810657701e-05,
+      "loss": 2.1996,
+      "step": 1780
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.825882300894643,
+      "learning_rate": 4.85029796767809e-05,
+      "loss": 2.0101,
+      "step": 1781
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7931028711494718,
+      "learning_rate": 4.850084977959722e-05,
+      "loss": 2.0837,
+      "step": 1782
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7617546898901234,
+      "learning_rate": 4.849871841515895e-05,
+      "loss": 1.9774,
+      "step": 1783
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7939058964877811,
+      "learning_rate": 4.849658558359915e-05,
+      "loss": 2.2086,
+      "step": 1784
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.703272893238437,
+      "learning_rate": 4.849445128505099e-05,
+      "loss": 2.0059,
+      "step": 1785
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7225836065492971,
+      "learning_rate": 4.849231551964771e-05,
+      "loss": 1.9649,
+      "step": 1786
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8384191599681904,
+      "learning_rate": 4.849017828752267e-05,
+      "loss": 1.9603,
+      "step": 1787
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8981483033690908,
+      "learning_rate": 4.8488039588809296e-05,
+      "loss": 2.2443,
+      "step": 1788
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9480267823215893,
+      "learning_rate": 4.848589942364112e-05,
+      "loss": 1.9873,
+      "step": 1789
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7435715976088432,
+      "learning_rate": 4.8483757792151766e-05,
+      "loss": 1.9971,
+      "step": 1790
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9093753824461049,
+      "learning_rate": 4.848161469447493e-05,
+      "loss": 2.0132,
+      "step": 1791
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9136018486931248,
+      "learning_rate": 4.847947013074443e-05,
+      "loss": 2.1466,
+      "step": 1792
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7329625345457703,
+      "learning_rate": 4.8477324101094144e-05,
+      "loss": 2.0256,
+      "step": 1793
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9088924735466958,
+      "learning_rate": 4.847517660565808e-05,
+      "loss": 1.9981,
+      "step": 1794
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6709794582582006,
+      "learning_rate": 4.8473027644570296e-05,
+      "loss": 2.0677,
+      "step": 1795
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8333346130846855,
+      "learning_rate": 4.847087721796496e-05,
+      "loss": 2.2208,
+      "step": 1796
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7001607343194188,
+      "learning_rate": 4.846872532597635e-05,
+      "loss": 2.0311,
+      "step": 1797
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7770965623623268,
+      "learning_rate": 4.8466571968738804e-05,
+      "loss": 2.0218,
+      "step": 1798
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9596627643835487,
+      "learning_rate": 4.8464417146386764e-05,
+      "loss": 2.0097,
+      "step": 1799
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.906223162372791,
+      "learning_rate": 4.8462260859054766e-05,
+      "loss": 2.2109,
+      "step": 1800
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9874280350530131,
+      "learning_rate": 4.846010310687744e-05,
+      "loss": 2.0896,
+      "step": 1801
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0652570015294909,
+      "learning_rate": 4.84579438899895e-05,
+      "loss": 2.0179,
+      "step": 1802
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0234377552282155,
+      "learning_rate": 4.8455783208525754e-05,
+      "loss": 2.0279,
+      "step": 1803
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.177207769473578,
+      "learning_rate": 4.845362106262111e-05,
+      "loss": 2.2034,
+      "step": 1804
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8083954663005583,
+      "learning_rate": 4.845145745241054e-05,
+      "loss": 2.0515,
+      "step": 1805
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9474424782828786,
+      "learning_rate": 4.844929237802915e-05,
+      "loss": 2.0307,
+      "step": 1806
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0081894923458448,
+      "learning_rate": 4.8447125839612106e-05,
+      "loss": 2.0754,
+      "step": 1807
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.0398278686173632,
+      "learning_rate": 4.844495783729467e-05,
+      "loss": 2.1963,
+      "step": 1808
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8088859471188138,
+      "learning_rate": 4.84427883712122e-05,
+      "loss": 2.0053,
+      "step": 1809
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8088999841143488,
+      "learning_rate": 4.844061744150015e-05,
+      "loss": 2.0445,
+      "step": 1810
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8313782440265521,
+      "learning_rate": 4.843844504829405e-05,
+      "loss": 1.9776,
+      "step": 1811
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8764648906148655,
+      "learning_rate": 4.843627119172954e-05,
+      "loss": 2.2167,
+      "step": 1812
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7575454956956268,
+      "learning_rate": 4.843409587194234e-05,
+      "loss": 2.0908,
+      "step": 1813
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9162060400524485,
+      "learning_rate": 4.843191908906827e-05,
+      "loss": 2.0242,
+      "step": 1814
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7916566155701187,
+      "learning_rate": 4.842974084324323e-05,
+      "loss": 2.025,
+      "step": 1815
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7869885681845203,
+      "learning_rate": 4.8427561134603216e-05,
+      "loss": 2.2043,
+      "step": 1816
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7863558495998596,
+      "learning_rate": 4.8425379963284324e-05,
+      "loss": 1.9739,
+      "step": 1817
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7673110396455858,
+      "learning_rate": 4.842319732942272e-05,
+      "loss": 1.9612,
+      "step": 1818
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6860692483886814,
+      "learning_rate": 4.842101323315468e-05,
+      "loss": 2.1035,
+      "step": 1819
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9223342422367793,
+      "learning_rate": 4.841882767461657e-05,
+      "loss": 1.9804,
+      "step": 1820
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7213069056202914,
+      "learning_rate": 4.8416640653944844e-05,
+      "loss": 2.192,
+      "step": 1821
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.756393747374661,
+      "learning_rate": 4.8414452171276036e-05,
+      "loss": 2.0308,
+      "step": 1822
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8385305648117451,
+      "learning_rate": 4.8412262226746797e-05,
+      "loss": 1.9676,
+      "step": 1823
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7558679817858103,
+      "learning_rate": 4.8410070820493844e-05,
+      "loss": 2.207,
+      "step": 1824
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8893072318185858,
+      "learning_rate": 4.8407877952654003e-05,
+      "loss": 2.0141,
+      "step": 1825
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.789600043792327,
+      "learning_rate": 4.840568362336418e-05,
+      "loss": 2.0557,
+      "step": 1826
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7795883107531075,
+      "learning_rate": 4.840348783276137e-05,
+      "loss": 1.9814,
+      "step": 1827
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.809858936296519,
+      "learning_rate": 4.840129058098267e-05,
+      "loss": 2.1665,
+      "step": 1828
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.713842734851453,
+      "learning_rate": 4.839909186816527e-05,
+      "loss": 2.0311,
+      "step": 1829
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6990516633469149,
+      "learning_rate": 4.839689169444644e-05,
+      "loss": 1.998,
+      "step": 1830
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8448272917620883,
+      "learning_rate": 4.8394690059963534e-05,
+      "loss": 2.035,
+      "step": 1831
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6974921906357421,
+      "learning_rate": 4.8392486964854034e-05,
+      "loss": 2.08,
+      "step": 1832
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8253978659784993,
+      "learning_rate": 4.839028240925546e-05,
+      "loss": 2.2284,
+      "step": 1833
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6686619814700958,
+      "learning_rate": 4.838807639330548e-05,
+      "loss": 1.9985,
+      "step": 1834
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6646760242105587,
+      "learning_rate": 4.8385868917141796e-05,
+      "loss": 2.0572,
+      "step": 1835
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6952305784420879,
+      "learning_rate": 4.838365998090225e-05,
+      "loss": 2.155,
+      "step": 1836
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7589404232117765,
+      "learning_rate": 4.838144958472475e-05,
+      "loss": 2.0008,
+      "step": 1837
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.643111749401844,
+      "learning_rate": 4.8379237728747304e-05,
+      "loss": 2.0747,
+      "step": 1838
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6949470416109794,
+      "learning_rate": 4.837702441310799e-05,
+      "loss": 2.0142,
+      "step": 1839
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6166374309213687,
+      "learning_rate": 4.837480963794502e-05,
+      "loss": 2.1865,
+      "step": 1840
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7026553248019648,
+      "learning_rate": 4.837259340339665e-05,
+      "loss": 1.9819,
+      "step": 1841
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6210912148703593,
+      "learning_rate": 4.8370375709601254e-05,
+      "loss": 1.983,
+      "step": 1842
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8130754301565566,
+      "learning_rate": 4.8368156556697295e-05,
+      "loss": 1.9699,
+      "step": 1843
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5783749364887686,
+      "learning_rate": 4.8365935944823334e-05,
+      "loss": 2.0559,
+      "step": 1844
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7052941996369728,
+      "learning_rate": 4.8363713874117986e-05,
+      "loss": 2.2139,
+      "step": 1845
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7037952660026171,
+      "learning_rate": 4.8361490344720014e-05,
+      "loss": 1.9923,
+      "step": 1846
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6403449202149538,
+      "learning_rate": 4.8359265356768225e-05,
+      "loss": 2.0055,
+      "step": 1847
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6435531386604566,
+      "learning_rate": 4.835703891040154e-05,
+      "loss": 2.2097,
+      "step": 1848
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6201528865643201,
+      "learning_rate": 4.835481100575896e-05,
+      "loss": 2.0243,
+      "step": 1849
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6420688759533919,
+      "learning_rate": 4.8352581642979576e-05,
+      "loss": 2.0811,
+      "step": 1850
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6292286828867993,
+      "learning_rate": 4.83503508222026e-05,
+      "loss": 2.0127,
+      "step": 1851
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7858229513658695,
+      "learning_rate": 4.834811854356729e-05,
+      "loss": 1.9806,
+      "step": 1852
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6858820191676133,
+      "learning_rate": 4.8345884807213024e-05,
+      "loss": 2.2172,
+      "step": 1853
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5768020906808438,
+      "learning_rate": 4.834364961327926e-05,
+      "loss": 1.9764,
+      "step": 1854
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6775038752978104,
+      "learning_rate": 4.8341412961905554e-05,
+      "loss": 1.9987,
+      "step": 1855
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6548377793358822,
+      "learning_rate": 4.8339174853231536e-05,
+      "loss": 2.209,
+      "step": 1856
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6034509240509452,
+      "learning_rate": 4.833693528739697e-05,
+      "loss": 2.0124,
+      "step": 1857
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.849058493620179,
+      "learning_rate": 4.833469426454166e-05,
+      "loss": 2.0144,
+      "step": 1858
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6226748063919738,
+      "learning_rate": 4.833245178480552e-05,
+      "loss": 2.0212,
+      "step": 1859
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8929410495882787,
+      "learning_rate": 4.833020784832857e-05,
+      "loss": 2.1992,
+      "step": 1860
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9988434051563644,
+      "learning_rate": 4.832796245525089e-05,
+      "loss": 2.0101,
+      "step": 1861
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7577316817795313,
+      "learning_rate": 4.832571560571269e-05,
+      "loss": 2.0279,
+      "step": 1862
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8374854428542543,
+      "learning_rate": 4.832346729985423e-05,
+      "loss": 2.0645,
+      "step": 1863
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6492857480310614,
+      "learning_rate": 4.8321217537815907e-05,
+      "loss": 1.9531,
+      "step": 1864
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9223165128853924,
+      "learning_rate": 4.8318966319738156e-05,
+      "loss": 2.2079,
+      "step": 1865
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7060071403335778,
+      "learning_rate": 4.831671364576155e-05,
+      "loss": 1.9684,
+      "step": 1866
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8558848330627417,
+      "learning_rate": 4.831445951602671e-05,
+      "loss": 2.0156,
+      "step": 1867
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6866388768474617,
+      "learning_rate": 4.831220393067439e-05,
+      "loss": 2.222,
+      "step": 1868
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7799654960106693,
+      "learning_rate": 4.830994688984542e-05,
+      "loss": 2.0399,
+      "step": 1869
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6580485849579769,
+      "learning_rate": 4.830768839368069e-05,
+      "loss": 1.9877,
+      "step": 1870
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7177393830844447,
+      "learning_rate": 4.830542844232123e-05,
+      "loss": 1.9956,
+      "step": 1871
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6939986757423202,
+      "learning_rate": 4.830316703590814e-05,
+      "loss": 2.1717,
+      "step": 1872
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6069003436719608,
+      "learning_rate": 4.8300904174582585e-05,
+      "loss": 2.0164,
+      "step": 1873
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7233845452068256,
+      "learning_rate": 4.829863985848587e-05,
+      "loss": 2.0013,
+      "step": 1874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6915819562936643,
+      "learning_rate": 4.829637408775935e-05,
+      "loss": 2.0679,
+      "step": 1875
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6423483577736271,
+      "learning_rate": 4.829410686254449e-05,
+      "loss": 2.0362,
+      "step": 1876
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7685073532638866,
+      "learning_rate": 4.829183818298285e-05,
+      "loss": 2.1481,
+      "step": 1877
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6462386515353771,
+      "learning_rate": 4.828956804921607e-05,
+      "loss": 1.9984,
+      "step": 1878
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8045501896651835,
+      "learning_rate": 4.828729646138587e-05,
+      "loss": 1.9787,
+      "step": 1879
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.658713606850531,
+      "learning_rate": 4.828502341963409e-05,
+      "loss": 2.1825,
+      "step": 1880
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7051819751401018,
+      "learning_rate": 4.828274892410264e-05,
+      "loss": 2.0672,
+      "step": 1881
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7473361284914114,
+      "learning_rate": 4.828047297493353e-05,
+      "loss": 1.9811,
+      "step": 1882
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7964074112670938,
+      "learning_rate": 4.827819557226885e-05,
+      "loss": 2.0251,
+      "step": 1883
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6383579195253555,
+      "learning_rate": 4.8275916716250796e-05,
+      "loss": 2.0265,
+      "step": 1884
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.762900198290114,
+      "learning_rate": 4.8273636407021635e-05,
+      "loss": 2.1988,
+      "step": 1885
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6131933965567656,
+      "learning_rate": 4.827135464472375e-05,
+      "loss": 2.0072,
+      "step": 1886
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6246149938472124,
+      "learning_rate": 4.826907142949959e-05,
+      "loss": 1.9977,
+      "step": 1887
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0021715128205675,
+      "learning_rate": 4.826678676149171e-05,
+      "loss": 2.0794,
+      "step": 1888
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6416065211239265,
+      "learning_rate": 4.8264500640842744e-05,
+      "loss": 2.2062,
+      "step": 1889
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6671776828913605,
+      "learning_rate": 4.826221306769544e-05,
+      "loss": 1.9693,
+      "step": 1890
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.655776507763887,
+      "learning_rate": 4.82599240421926e-05,
+      "loss": 2.0237,
+      "step": 1891
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.776274065158058,
+      "learning_rate": 4.8257633564477156e-05,
+      "loss": 2.1839,
+      "step": 1892
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6304223728930962,
+      "learning_rate": 4.82553416346921e-05,
+      "loss": 1.9989,
+      "step": 1893
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6710501259125442,
+      "learning_rate": 4.825304825298053e-05,
+      "loss": 2.0553,
+      "step": 1894
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6477136174584734,
+      "learning_rate": 4.8250753419485634e-05,
+      "loss": 2.0014,
+      "step": 1895
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7097378178470494,
+      "learning_rate": 4.824845713435068e-05,
+      "loss": 1.9884,
+      "step": 1896
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6694838222329936,
+      "learning_rate": 4.824615939771905e-05,
+      "loss": 2.2176,
+      "step": 1897
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6585410827628839,
+      "learning_rate": 4.8243860209734185e-05,
+      "loss": 2.0357,
+      "step": 1898
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6307576693203919,
+      "learning_rate": 4.824155957053964e-05,
+      "loss": 1.9644,
+      "step": 1899
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.666368502614613,
+      "learning_rate": 4.823925748027905e-05,
+      "loss": 2.0461,
+      "step": 1900
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7578814712075588,
+      "learning_rate": 4.823695393909615e-05,
+      "loss": 2.1967,
+      "step": 1901
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.686373879384365,
+      "learning_rate": 4.823464894713475e-05,
+      "loss": 2.0403,
+      "step": 1902
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8093966061880892,
+      "learning_rate": 4.823234250453878e-05,
+      "loss": 2.0092,
+      "step": 1903
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6722061570107992,
+      "learning_rate": 4.823003461145221e-05,
+      "loss": 1.9932,
+      "step": 1904
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6468213522949562,
+      "learning_rate": 4.8227725268019155e-05,
+      "loss": 2.1959,
+      "step": 1905
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.71842789560096,
+      "learning_rate": 4.822541447438379e-05,
+      "loss": 2.1094,
+      "step": 1906
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6263837249534764,
+      "learning_rate": 4.822310223069039e-05,
+      "loss": 2.0399,
+      "step": 1907
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7170854163795776,
+      "learning_rate": 4.822078853708331e-05,
+      "loss": 2.0163,
+      "step": 1908
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6789374841806087,
+      "learning_rate": 4.8218473393707004e-05,
+      "loss": 2.1903,
+      "step": 1909
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6670422654854251,
+      "learning_rate": 4.8216156800706026e-05,
+      "loss": 1.9975,
+      "step": 1910
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7294112481300277,
+      "learning_rate": 4.8213838758225005e-05,
+      "loss": 1.9488,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6262879132211079,
+      "learning_rate": 4.821151926640866e-05,
+      "loss": 2.1806,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7180347551115842,
+      "learning_rate": 4.8209198325401815e-05,
+      "loss": 2.0577,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7097144834912172,
+      "learning_rate": 4.820687593534938e-05,
+      "loss": 1.987,
+      "step": 1914
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6546451942575606,
+      "learning_rate": 4.820455209639634e-05,
+      "loss": 1.9904,
+      "step": 1915
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7038046517450455,
+      "learning_rate": 4.820222680868778e-05,
+      "loss": 2.004,
+      "step": 1916
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.715691801539892,
+      "learning_rate": 4.819990007236889e-05,
+      "loss": 2.1634,
+      "step": 1917
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6782980425053148,
+      "learning_rate": 4.819757188758492e-05,
+      "loss": 2.0101,
+      "step": 1918
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6785400127678617,
+      "learning_rate": 4.8195242254481254e-05,
+      "loss": 2.0339,
+      "step": 1919
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6591438780865508,
+      "learning_rate": 4.819291117320332e-05,
+      "loss": 2.0253,
+      "step": 1920
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7028496178773267,
+      "learning_rate": 4.819057864389666e-05,
+      "loss": 2.1877,
+      "step": 1921
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6534432145073824,
+      "learning_rate": 4.8188244666706905e-05,
+      "loss": 2.0238,
+      "step": 1922
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7055207477321518,
+      "learning_rate": 4.818590924177978e-05,
+      "loss": 1.9773,
+      "step": 1923
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6952323058825431,
+      "learning_rate": 4.818357236926109e-05,
+      "loss": 2.2041,
+      "step": 1924
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.737407850953508,
+      "learning_rate": 4.818123404929673e-05,
+      "loss": 2.0453,
+      "step": 1925
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6750130506500975,
+      "learning_rate": 4.81788942820327e-05,
+      "loss": 2.0366,
+      "step": 1926
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7294570243800683,
+      "learning_rate": 4.817655306761508e-05,
+      "loss": 1.996,
+      "step": 1927
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6396067830300864,
+      "learning_rate": 4.8174210406190046e-05,
+      "loss": 1.9892,
+      "step": 1928
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6899997590369725,
+      "learning_rate": 4.817186629790385e-05,
+      "loss": 2.266,
+      "step": 1929
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.727265449128919,
+      "learning_rate": 4.816952074290285e-05,
+      "loss": 1.9957,
+      "step": 1930
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7189510403150956,
+      "learning_rate": 4.816717374133348e-05,
+      "loss": 2.1046,
+      "step": 1931
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6318497078622048,
+      "learning_rate": 4.8164825293342294e-05,
+      "loss": 1.9332,
+      "step": 1932
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7132285419760882,
+      "learning_rate": 4.816247539907589e-05,
+      "loss": 2.2157,
+      "step": 1933
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6304749875756387,
+      "learning_rate": 4.8160124058681e-05,
+      "loss": 1.9547,
+      "step": 1934
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6227453425372939,
+      "learning_rate": 4.815777127230442e-05,
+      "loss": 1.9644,
+      "step": 1935
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.819201227138268,
+      "learning_rate": 4.815541704009304e-05,
+      "loss": 1.9863,
+      "step": 1936
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6120868767218024,
+      "learning_rate": 4.8153061362193855e-05,
+      "loss": 2.1967,
+      "step": 1937
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7179887357398979,
+      "learning_rate": 4.8150704238753926e-05,
+      "loss": 1.9425,
+      "step": 1938
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5769099376860864,
+      "learning_rate": 4.814834566992043e-05,
+      "loss": 2.0275,
+      "step": 1939
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6524795357590113,
+      "learning_rate": 4.814598565584062e-05,
+      "loss": 1.9855,
+      "step": 1940
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6266808267090597,
+      "learning_rate": 4.814362419666184e-05,
+      "loss": 2.1876,
+      "step": 1941
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6636713796999617,
+      "learning_rate": 4.814126129253152e-05,
+      "loss": 1.9444,
+      "step": 1942
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6648586144715893,
+      "learning_rate": 4.8138896943597193e-05,
+      "loss": 2.025,
+      "step": 1943
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5964756895947774,
+      "learning_rate": 4.813653115000647e-05,
+      "loss": 2.0571,
+      "step": 1944
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6043784260492732,
+      "learning_rate": 4.813416391190707e-05,
+      "loss": 2.1955,
+      "step": 1945
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6485594120218846,
+      "learning_rate": 4.813179522944678e-05,
+      "loss": 1.9647,
+      "step": 1946
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.660139386834233,
+      "learning_rate": 4.812942510277347e-05,
+      "loss": 1.9782,
+      "step": 1947
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6625504557723374,
+      "learning_rate": 4.812705353203514e-05,
+      "loss": 2.0261,
+      "step": 1948
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6107333917391423,
+      "learning_rate": 4.8124680517379856e-05,
+      "loss": 2.221,
+      "step": 1949
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6368868631259151,
+      "learning_rate": 4.8122306058955765e-05,
+      "loss": 2.0587,
+      "step": 1950
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6312905399777267,
+      "learning_rate": 4.811993015691112e-05,
+      "loss": 2.0026,
+      "step": 1951
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6379599757539238,
+      "learning_rate": 4.811755281139425e-05,
+      "loss": 2.001,
+      "step": 1952
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6006845087547474,
+      "learning_rate": 4.8115174022553586e-05,
+      "loss": 2.198,
+      "step": 1953
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6085232232463885,
+      "learning_rate": 4.811279379053766e-05,
+      "loss": 2.0021,
+      "step": 1954
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6878227538795225,
+      "learning_rate": 4.811041211549506e-05,
+      "loss": 1.9862,
+      "step": 1955
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6502181177691235,
+      "learning_rate": 4.81080289975745e-05,
+      "loss": 2.0238,
+      "step": 1956
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7062331536371547,
+      "learning_rate": 4.810564443692475e-05,
+      "loss": 2.1624,
+      "step": 1957
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6823517708161424,
+      "learning_rate": 4.81032584336947e-05,
+      "loss": 1.9992,
+      "step": 1958
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7654489912713323,
+      "learning_rate": 4.810087098803332e-05,
+      "loss": 1.9859,
+      "step": 1959
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6343047554835318,
+      "learning_rate": 4.809848210008966e-05,
+      "loss": 1.9663,
+      "step": 1960
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7305276526388191,
+      "learning_rate": 4.8096091770012876e-05,
+      "loss": 2.2205,
+      "step": 1961
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5734896610164785,
+      "learning_rate": 4.809369999795219e-05,
+      "loss": 2.0651,
+      "step": 1962
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6876374327108107,
+      "learning_rate": 4.809130678405696e-05,
+      "loss": 2.0589,
+      "step": 1963
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6793109928680939,
+      "learning_rate": 4.808891212847657e-05,
+      "loss": 1.9602,
+      "step": 1964
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6450520195103199,
+      "learning_rate": 4.8086516031360546e-05,
+      "loss": 2.1662,
+      "step": 1965
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6866501038595605,
+      "learning_rate": 4.8084118492858486e-05,
+      "loss": 1.9506,
+      "step": 1966
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6866882009060182,
+      "learning_rate": 4.808171951312008e-05,
+      "loss": 2.0368,
+      "step": 1967
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5971420071909138,
+      "learning_rate": 4.80793190922951e-05,
+      "loss": 2.0533,
+      "step": 1968
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6137395946089833,
+      "learning_rate": 4.807691723053342e-05,
+      "loss": 2.1681,
+      "step": 1969
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6531782676571379,
+      "learning_rate": 4.807451392798499e-05,
+      "loss": 1.9953,
+      "step": 1970
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6139987833004621,
+      "learning_rate": 4.807210918479987e-05,
+      "loss": 2.0002,
+      "step": 1971
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6648496682012286,
+      "learning_rate": 4.806970300112819e-05,
+      "loss": 2.0317,
+      "step": 1972
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6266498790253385,
+      "learning_rate": 4.806729537712017e-05,
+      "loss": 2.1787,
+      "step": 1973
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6496864058024427,
+      "learning_rate": 4.806488631292614e-05,
+      "loss": 1.9903,
+      "step": 1974
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5938894547134859,
+      "learning_rate": 4.806247580869651e-05,
+      "loss": 2.0873,
+      "step": 1975
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.617821628766627,
+      "learning_rate": 4.8060063864581765e-05,
+      "loss": 2.0123,
+      "step": 1976
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6649397791322516,
+      "learning_rate": 4.805765048073251e-05,
+      "loss": 2.219,
+      "step": 1977
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6373434319364385,
+      "learning_rate": 4.80552356572994e-05,
+      "loss": 1.9543,
+      "step": 1978
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7608710489778561,
+      "learning_rate": 4.8052819394433225e-05,
+      "loss": 2.0333,
+      "step": 1979
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6116075345314019,
+      "learning_rate": 4.805040169228483e-05,
+      "loss": 2.0084,
+      "step": 1980
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7803187258145476,
+      "learning_rate": 4.804798255100515e-05,
+      "loss": 2.2264,
+      "step": 1981
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6517347828304965,
+      "learning_rate": 4.804556197074526e-05,
+      "loss": 2.0052,
+      "step": 1982
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7152816541509863,
+      "learning_rate": 4.804313995165625e-05,
+      "loss": 1.9957,
+      "step": 1983
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6822068536330941,
+      "learning_rate": 4.8040716493889346e-05,
+      "loss": 1.9954,
+      "step": 1984
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.651347863029358,
+      "learning_rate": 4.8038291597595865e-05,
+      "loss": 2.1961,
+      "step": 1985
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.673210263013792,
+      "learning_rate": 4.803586526292719e-05,
+      "loss": 2.0414,
+      "step": 1986
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6918784095515743,
+      "learning_rate": 4.803343749003482e-05,
+      "loss": 2.0696,
+      "step": 1987
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7075527213672979,
+      "learning_rate": 4.803100827907031e-05,
+      "loss": 2.0001,
+      "step": 1988
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5981893489644124,
+      "learning_rate": 4.802857763018535e-05,
+      "loss": 2.1629,
+      "step": 1989
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6081048844232899,
+      "learning_rate": 4.8026145543531685e-05,
+      "loss": 1.9654,
+      "step": 1990
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8983575774834642,
+      "learning_rate": 4.8023712019261155e-05,
+      "loss": 1.9739,
+      "step": 1991
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6052069133552328,
+      "learning_rate": 4.8021277057525705e-05,
+      "loss": 1.9801,
+      "step": 1992
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6536860243191052,
+      "learning_rate": 4.8018840658477346e-05,
+      "loss": 2.217,
+      "step": 1993
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.663562872718384,
+      "learning_rate": 4.801640282226821e-05,
+      "loss": 1.9379,
+      "step": 1994
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6527102329063471,
+      "learning_rate": 4.801396354905049e-05,
+      "loss": 1.9478,
+      "step": 1995
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6080631281335737,
+      "learning_rate": 4.8011522838976484e-05,
+      "loss": 2.0177,
+      "step": 1996
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7024265501649074,
+      "learning_rate": 4.800908069219857e-05,
+      "loss": 2.1758,
+      "step": 1997
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6714636259650334,
+      "learning_rate": 4.800663710886922e-05,
+      "loss": 1.9997,
+      "step": 1998
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5729278918161292,
+      "learning_rate": 4.800419208914101e-05,
+      "loss": 2.0836,
+      "step": 1999
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6197662764388889,
+      "learning_rate": 4.800174563316658e-05,
+      "loss": 2.0152,
+      "step": 2000
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6993932632955054,
+      "learning_rate": 4.799929774109868e-05,
+      "loss": 2.1788,
+      "step": 2001
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6362776724510202,
+      "learning_rate": 4.7996848413090134e-05,
+      "loss": 2.0305,
+      "step": 2002
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5515626975110235,
+      "learning_rate": 4.799439764929388e-05,
+      "loss": 1.9601,
+      "step": 2003
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6544932753332472,
+      "learning_rate": 4.79919454498629e-05,
+      "loss": 1.9933,
+      "step": 2004
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6665142569981537,
+      "learning_rate": 4.798949181495032e-05,
+      "loss": 2.2188,
+      "step": 2005
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6877752435755332,
+      "learning_rate": 4.7987036744709326e-05,
+      "loss": 2.0901,
+      "step": 2006
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7005530904857376,
+      "learning_rate": 4.7984580239293184e-05,
+      "loss": 2.0132,
+      "step": 2007
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6703711316980305,
+      "learning_rate": 4.798212229885528e-05,
+      "loss": 1.9825,
+      "step": 2008
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7980617066581983,
+      "learning_rate": 4.797966292354907e-05,
+      "loss": 2.2063,
+      "step": 2009
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7124983593212242,
+      "learning_rate": 4.797720211352811e-05,
+      "loss": 1.9624,
+      "step": 2010
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6286220111536401,
+      "learning_rate": 4.7974739868946015e-05,
+      "loss": 2.0244,
+      "step": 2011
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6824240014931058,
+      "learning_rate": 4.797227618995654e-05,
+      "loss": 2.0565,
+      "step": 2012
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6318707084812021,
+      "learning_rate": 4.7969811076713476e-05,
+      "loss": 2.1954,
+      "step": 2013
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8507820928260817,
+      "learning_rate": 4.7967344529370755e-05,
+      "loss": 2.0504,
+      "step": 2014
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.58717746675603,
+      "learning_rate": 4.796487654808236e-05,
+      "loss": 1.9898,
+      "step": 2015
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6394849373650409,
+      "learning_rate": 4.796240713300238e-05,
+      "loss": 1.9997,
+      "step": 2016
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6180948353544421,
+      "learning_rate": 4.795993628428499e-05,
+      "loss": 2.2295,
+      "step": 2017
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5926810606580063,
+      "learning_rate": 4.795746400208445e-05,
+      "loss": 2.0249,
+      "step": 2018
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6329189186096706,
+      "learning_rate": 4.795499028655513e-05,
+      "loss": 2.0,
+      "step": 2019
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6714982697090845,
+      "learning_rate": 4.795251513785146e-05,
+      "loss": 1.9714,
+      "step": 2020
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5925491323527572,
+      "learning_rate": 4.795003855612798e-05,
+      "loss": 2.2292,
+      "step": 2021
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6979715828390779,
+      "learning_rate": 4.794756054153932e-05,
+      "loss": 1.9905,
+      "step": 2022
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.61371908200805,
+      "learning_rate": 4.794508109424017e-05,
+      "loss": 1.9676,
+      "step": 2023
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8079443305970975,
+      "learning_rate": 4.794260021438535e-05,
+      "loss": 2.0677,
+      "step": 2024
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6251343873604478,
+      "learning_rate": 4.7940117902129756e-05,
+      "loss": 2.1648,
+      "step": 2025
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6899623895310409,
+      "learning_rate": 4.793763415762835e-05,
+      "loss": 1.9882,
+      "step": 2026
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.59601104007301,
+      "learning_rate": 4.793514898103623e-05,
+      "loss": 1.9357,
+      "step": 2027
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7354077466212326,
+      "learning_rate": 4.793266237250852e-05,
+      "loss": 1.9805,
+      "step": 2028
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.640648493276136,
+      "learning_rate": 4.7930174332200494e-05,
+      "loss": 2.2447,
+      "step": 2029
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7283083453591035,
+      "learning_rate": 4.792768486026749e-05,
+      "loss": 2.0533,
+      "step": 2030
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8283802324875262,
+      "learning_rate": 4.792519395686493e-05,
+      "loss": 2.0199,
+      "step": 2031
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6271099743607036,
+      "learning_rate": 4.792270162214833e-05,
+      "loss": 2.0296,
+      "step": 2032
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7249711768231938,
+      "learning_rate": 4.7920207856273306e-05,
+      "loss": 2.1937,
+      "step": 2033
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6344268820761548,
+      "learning_rate": 4.7917712659395535e-05,
+      "loss": 1.9857,
+      "step": 2034
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7262464653830112,
+      "learning_rate": 4.7915216031670815e-05,
+      "loss": 2.0272,
+      "step": 2035
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7068685898391689,
+      "learning_rate": 4.7912717973255025e-05,
+      "loss": 1.979,
+      "step": 2036
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6946991295264463,
+      "learning_rate": 4.791021848430413e-05,
+      "loss": 2.0553,
+      "step": 2037
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7196671744328093,
+      "learning_rate": 4.7907717564974176e-05,
+      "loss": 2.1877,
+      "step": 2038
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7350932805947377,
+      "learning_rate": 4.790521521542129e-05,
+      "loss": 1.9808,
+      "step": 2039
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7540355637161716,
+      "learning_rate": 4.790271143580174e-05,
+      "loss": 2.0044,
+      "step": 2040
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7341810212691784,
+      "learning_rate": 4.7900206226271823e-05,
+      "loss": 2.2063,
+      "step": 2041
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8960686007635085,
+      "learning_rate": 4.7897699586987955e-05,
+      "loss": 1.9756,
+      "step": 2042
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6861972935524167,
+      "learning_rate": 4.789519151810664e-05,
+      "loss": 2.0331,
+      "step": 2043
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0491275542035767,
+      "learning_rate": 4.7892682019784454e-05,
+      "loss": 2.0042,
+      "step": 2044
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0501591407251794,
+      "learning_rate": 4.789017109217809e-05,
+      "loss": 2.1573,
+      "step": 2045
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6687818144803892,
+      "learning_rate": 4.788765873544431e-05,
+      "loss": 2.0165,
+      "step": 2046
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1710635130302376,
+      "learning_rate": 4.788514494973996e-05,
+      "loss": 2.0076,
+      "step": 2047
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6839678899582097,
+      "learning_rate": 4.7882629735222004e-05,
+      "loss": 1.9858,
+      "step": 2048
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9295397337971999,
+      "learning_rate": 4.7880113092047475e-05,
+      "loss": 2.0672,
+      "step": 2049
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.1173896136462103,
+      "learning_rate": 4.787759502037348e-05,
+      "loss": 2.1929,
+      "step": 2050
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6115905052026146,
+      "learning_rate": 4.7875075520357246e-05,
+      "loss": 2.0094,
+      "step": 2051
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0712491627937821,
+      "learning_rate": 4.7872554592156085e-05,
+      "loss": 2.0632,
+      "step": 2052
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.900964613569214,
+      "learning_rate": 4.787003223592737e-05,
+      "loss": 2.1353,
+      "step": 2053
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6805119405914696,
+      "learning_rate": 4.786750845182859e-05,
+      "loss": 1.994,
+      "step": 2054
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9279147423168745,
+      "learning_rate": 4.786498324001732e-05,
+      "loss": 2.0383,
+      "step": 2055
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7407466233806347,
+      "learning_rate": 4.786245660065121e-05,
+      "loss": 2.006,
+      "step": 2056
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8411688707239905,
+      "learning_rate": 4.785992853388801e-05,
+      "loss": 2.226,
+      "step": 2057
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7817317712360857,
+      "learning_rate": 4.785739903988557e-05,
+      "loss": 2.0314,
+      "step": 2058
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.681575626284558,
+      "learning_rate": 4.7854868118801806e-05,
+      "loss": 1.9696,
+      "step": 2059
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8794660252609096,
+      "learning_rate": 4.785233577079473e-05,
+      "loss": 1.9634,
+      "step": 2060
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7363778253619299,
+      "learning_rate": 4.784980199602245e-05,
+      "loss": 2.1946,
+      "step": 2061
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7704576839771932,
+      "learning_rate": 4.7847266794643165e-05,
+      "loss": 2.0641,
+      "step": 2062
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6955109504115887,
+      "learning_rate": 4.784473016681516e-05,
+      "loss": 2.0238,
+      "step": 2063
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8911060735920282,
+      "learning_rate": 4.78421921126968e-05,
+      "loss": 2.0271,
+      "step": 2064
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.678539981222637,
+      "learning_rate": 4.783965263244654e-05,
+      "loss": 2.2187,
+      "step": 2065
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8116756337196732,
+      "learning_rate": 4.783711172622295e-05,
+      "loss": 1.9659,
+      "step": 2066
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6507957117608317,
+      "learning_rate": 4.783456939418465e-05,
+      "loss": 1.9952,
+      "step": 2067
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8306903190028612,
+      "learning_rate": 4.783202563649038e-05,
+      "loss": 2.1081,
+      "step": 2068
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6866248866498299,
+      "learning_rate": 4.782948045329895e-05,
+      "loss": 1.9792,
+      "step": 2069
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6740795505613211,
+      "learning_rate": 4.7826933844769273e-05,
+      "loss": 2.1661,
+      "step": 2070
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6695911551004502,
+      "learning_rate": 4.782438581106034e-05,
+      "loss": 1.9619,
+      "step": 2071
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6161563889870678,
+      "learning_rate": 4.782183635233124e-05,
+      "loss": 1.9638,
+      "step": 2072
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7081932149913969,
+      "learning_rate": 4.781928546874113e-05,
+      "loss": 2.1521,
+      "step": 2073
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5868916252570111,
+      "learning_rate": 4.7816733160449287e-05,
+      "loss": 2.0228,
+      "step": 2074
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6835951888062113,
+      "learning_rate": 4.7814179427615066e-05,
+      "loss": 2.041,
+      "step": 2075
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7369878413881155,
+      "learning_rate": 4.78116242703979e-05,
+      "loss": 2.0081,
+      "step": 2076
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7352898438406729,
+      "learning_rate": 4.780906768895731e-05,
+      "loss": 2.1729,
+      "step": 2077
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7692482600277273,
+      "learning_rate": 4.7806509683452937e-05,
+      "loss": 2.0021,
+      "step": 2078
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6717009874348432,
+      "learning_rate": 4.780395025404446e-05,
+      "loss": 1.961,
+      "step": 2079
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8596008915387896,
+      "learning_rate": 4.780138940089169e-05,
+      "loss": 2.0469,
+      "step": 2080
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6957045568805382,
+      "learning_rate": 4.779882712415451e-05,
+      "loss": 2.0324,
+      "step": 2081
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7573576574557406,
+      "learning_rate": 4.77962634239929e-05,
+      "loss": 2.2071,
+      "step": 2082
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8061973592927387,
+      "learning_rate": 4.779369830056691e-05,
+      "loss": 2.006,
+      "step": 2083
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.685884752514872,
+      "learning_rate": 4.7791131754036694e-05,
+      "loss": 2.0075,
+      "step": 2084
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7603544711573783,
+      "learning_rate": 4.778856378456249e-05,
+      "loss": 2.2021,
+      "step": 2085
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7646471324838224,
+      "learning_rate": 4.778599439230465e-05,
+      "loss": 2.0604,
+      "step": 2086
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8127998258039101,
+      "learning_rate": 4.778342357742355e-05,
+      "loss": 1.9673,
+      "step": 2087
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9142923762387998,
+      "learning_rate": 4.778085134007973e-05,
+      "loss": 1.9968,
+      "step": 2088
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7660587641226316,
+      "learning_rate": 4.777827768043377e-05,
+      "loss": 2.1904,
+      "step": 2089
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.898660551761882,
+      "learning_rate": 4.7775702598646366e-05,
+      "loss": 1.9953,
+      "step": 2090
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.924365474350514,
+      "learning_rate": 4.777312609487828e-05,
+      "loss": 1.9076,
+      "step": 2091
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8210222319038528,
+      "learning_rate": 4.7770548169290374e-05,
+      "loss": 2.0383,
+      "step": 2092
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9666485252711392,
+      "learning_rate": 4.7767968822043605e-05,
+      "loss": 2.0846,
+      "step": 2093
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7445174180160211,
+      "learning_rate": 4.7765388053299015e-05,
+      "loss": 2.21,
+      "step": 2094
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6692976444926391,
+      "learning_rate": 4.776280586321772e-05,
+      "loss": 1.9446,
+      "step": 2095
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6136148132588995,
+      "learning_rate": 4.7760222251960945e-05,
+      "loss": 1.9659,
+      "step": 2096
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7467297347061005,
+      "learning_rate": 4.7757637219689995e-05,
+      "loss": 2.1797,
+      "step": 2097
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7382853476037023,
+      "learning_rate": 4.775505076656626e-05,
+      "loss": 1.9875,
+      "step": 2098
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8627338129766725,
+      "learning_rate": 4.775246289275123e-05,
+      "loss": 2.0751,
+      "step": 2099
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7232280959031868,
+      "learning_rate": 4.7749873598406464e-05,
+      "loss": 2.013,
+      "step": 2100
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6658641609627705,
+      "learning_rate": 4.774728288369363e-05,
+      "loss": 1.969,
+      "step": 2101
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7188561573172418,
+      "learning_rate": 4.774469074877449e-05,
+      "loss": 2.1508,
+      "step": 2102
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6502199778916018,
+      "learning_rate": 4.774209719381086e-05,
+      "loss": 1.9486,
+      "step": 2103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6451419553568801,
+      "learning_rate": 4.773950221896467e-05,
+      "loss": 2.0225,
+      "step": 2104
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6527264398930251,
+      "learning_rate": 4.773690582439795e-05,
+      "loss": 2.0611,
+      "step": 2105
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7486942629641066,
+      "learning_rate": 4.7734308010272785e-05,
+      "loss": 2.1865,
+      "step": 2106
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6594084732036383,
+      "learning_rate": 4.773170877675139e-05,
+      "loss": 1.9565,
+      "step": 2107
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.682377732959749,
+      "learning_rate": 4.772910812399603e-05,
+      "loss": 2.0087,
+      "step": 2108
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6899658529279881,
+      "learning_rate": 4.7726506052169065e-05,
+      "loss": 2.2371,
+      "step": 2109
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8117764705299684,
+      "learning_rate": 4.7723902561432964e-05,
+      "loss": 1.9949,
+      "step": 2110
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6115773217696118,
+      "learning_rate": 4.772129765195028e-05,
+      "loss": 2.0599,
+      "step": 2111
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7327408694576583,
+      "learning_rate": 4.7718691323883644e-05,
+      "loss": 1.9613,
+      "step": 2112
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6346770945252507,
+      "learning_rate": 4.771608357739578e-05,
+      "loss": 1.9727,
+      "step": 2113
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7402626310764088,
+      "learning_rate": 4.771347441264949e-05,
+      "loss": 2.1908,
+      "step": 2114
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6964219120716855,
+      "learning_rate": 4.771086382980768e-05,
+      "loss": 2.0159,
+      "step": 2115
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6398498349708037,
+      "learning_rate": 4.770825182903335e-05,
+      "loss": 2.0072,
+      "step": 2116
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8105697010974076,
+      "learning_rate": 4.770563841048957e-05,
+      "loss": 2.0898,
+      "step": 2117
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5687941002014472,
+      "learning_rate": 4.770302357433951e-05,
+      "loss": 2.1923,
+      "step": 2118
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.759742122709048,
+      "learning_rate": 4.7700407320746414e-05,
+      "loss": 1.9789,
+      "step": 2119
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6839453939032593,
+      "learning_rate": 4.769778964987364e-05,
+      "loss": 2.0206,
+      "step": 2120
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6673818755837441,
+      "learning_rate": 4.769517056188461e-05,
+      "loss": 1.9856,
+      "step": 2121
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6906704042737747,
+      "learning_rate": 4.7692550056942855e-05,
+      "loss": 2.1954,
+      "step": 2122
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.65477996902443,
+      "learning_rate": 4.7689928135211966e-05,
+      "loss": 2.0563,
+      "step": 2123
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6225360719067285,
+      "learning_rate": 4.768730479685566e-05,
+      "loss": 2.0519,
+      "step": 2124
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5905329067382125,
+      "learning_rate": 4.76846800420377e-05,
+      "loss": 2.0434,
+      "step": 2125
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.626912742947821,
+      "learning_rate": 4.768205387092198e-05,
+      "loss": 2.1541,
+      "step": 2126
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6771285613181283,
+      "learning_rate": 4.7679426283672466e-05,
+      "loss": 2.0241,
+      "step": 2127
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.623145668525327,
+      "learning_rate": 4.7676797280453194e-05,
+      "loss": 2.0126,
+      "step": 2128
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6815742173356221,
+      "learning_rate": 4.767416686142831e-05,
+      "loss": 2.1446,
+      "step": 2129
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5941494331752277,
+      "learning_rate": 4.7671535026762035e-05,
+      "loss": 2.0679,
+      "step": 2130
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6488961613468942,
+      "learning_rate": 4.76689017766187e-05,
+      "loss": 1.9835,
+      "step": 2131
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5971346561956788,
+      "learning_rate": 4.76662671111627e-05,
+      "loss": 2.0281,
+      "step": 2132
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6922259929399077,
+      "learning_rate": 4.766363103055852e-05,
+      "loss": 1.9969,
+      "step": 2133
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7375878470013998,
+      "learning_rate": 4.7660993534970753e-05,
+      "loss": 2.2527,
+      "step": 2134
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7771057332345209,
+      "learning_rate": 4.765835462456407e-05,
+      "loss": 1.9622,
+      "step": 2135
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6883744502895583,
+      "learning_rate": 4.765571429950323e-05,
+      "loss": 2.0554,
+      "step": 2136
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6717032548924661,
+      "learning_rate": 4.765307255995307e-05,
+      "loss": 1.9963,
+      "step": 2137
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8194844959616974,
+      "learning_rate": 4.7650429406078525e-05,
+      "loss": 2.1399,
+      "step": 2138
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.682835427229905,
+      "learning_rate": 4.764778483804462e-05,
+      "loss": 2.0105,
+      "step": 2139
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9498380490845082,
+      "learning_rate": 4.764513885601647e-05,
+      "loss": 2.0031,
+      "step": 2140
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7035057578387215,
+      "learning_rate": 4.764249146015928e-05,
+      "loss": 2.1658,
+      "step": 2141
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8128048563623967,
+      "learning_rate": 4.763984265063832e-05,
+      "loss": 2.0525,
+      "step": 2142
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8574265438414361,
+      "learning_rate": 4.763719242761898e-05,
+      "loss": 1.9948,
+      "step": 2143
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6717391272248605,
+      "learning_rate": 4.7634540791266726e-05,
+      "loss": 2.0465,
+      "step": 2144
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7365831202557926,
+      "learning_rate": 4.76318877417471e-05,
+      "loss": 2.0318,
+      "step": 2145
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6025212183952929,
+      "learning_rate": 4.762923327922575e-05,
+      "loss": 2.2009,
+      "step": 2146
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6409536703728919,
+      "learning_rate": 4.7626577403868405e-05,
+      "loss": 1.9338,
+      "step": 2147
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7296461043848717,
+      "learning_rate": 4.762392011584088e-05,
+      "loss": 2.0974,
+      "step": 2148
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6625176358972278,
+      "learning_rate": 4.7621261415309074e-05,
+      "loss": 1.9974,
+      "step": 2149
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7108608134604655,
+      "learning_rate": 4.7618601302439e-05,
+      "loss": 2.143,
+      "step": 2150
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8277748439559212,
+      "learning_rate": 4.7615939777396715e-05,
+      "loss": 2.0087,
+      "step": 2151
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6338536422644437,
+      "learning_rate": 4.761327684034841e-05,
+      "loss": 1.9785,
+      "step": 2152
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7681194675428614,
+      "learning_rate": 4.761061249146033e-05,
+      "loss": 2.0017,
+      "step": 2153
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6113398171467649,
+      "learning_rate": 4.760794673089882e-05,
+      "loss": 2.2249,
+      "step": 2154
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7807389275853509,
+      "learning_rate": 4.760527955883033e-05,
+      "loss": 2.0717,
+      "step": 2155
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7329144150012221,
+      "learning_rate": 4.760261097542137e-05,
+      "loss": 1.9917,
+      "step": 2156
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6648096668284537,
+      "learning_rate": 4.7599940980838546e-05,
+      "loss": 1.9643,
+      "step": 2157
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6756730895888895,
+      "learning_rate": 4.759726957524857e-05,
+      "loss": 2.1812,
+      "step": 2158
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6185432429127606,
+      "learning_rate": 4.759459675881822e-05,
+      "loss": 2.0056,
+      "step": 2159
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7308715087862035,
+      "learning_rate": 4.759192253171437e-05,
+      "loss": 2.0104,
+      "step": 2160
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6355642578672368,
+      "learning_rate": 4.758924689410399e-05,
+      "loss": 2.1115,
+      "step": 2161
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6643475290261182,
+      "learning_rate": 4.7586569846154125e-05,
+      "loss": 2.2076,
+      "step": 2162
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6017302560997428,
+      "learning_rate": 4.7583891388031925e-05,
+      "loss": 1.9824,
+      "step": 2163
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6470610368288269,
+      "learning_rate": 4.75812115199046e-05,
+      "loss": 2.0207,
+      "step": 2164
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7109337064743986,
+      "learning_rate": 4.757853024193948e-05,
+      "loss": 2.0368,
+      "step": 2165
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7042583797246986,
+      "learning_rate": 4.757584755430396e-05,
+      "loss": 2.1905,
+      "step": 2166
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7838207228156526,
+      "learning_rate": 4.7573163457165534e-05,
+      "loss": 2.117,
+      "step": 2167
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7148682030422492,
+      "learning_rate": 4.757047795069178e-05,
+      "loss": 1.9718,
+      "step": 2168
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7888765709744396,
+      "learning_rate": 4.756779103505036e-05,
+      "loss": 1.9649,
+      "step": 2169
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8252544255032028,
+      "learning_rate": 4.7565102710409035e-05,
+      "loss": 2.1416,
+      "step": 2170
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6453524738858872,
+      "learning_rate": 4.756241297693566e-05,
+      "loss": 1.9725,
+      "step": 2171
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8951344953789735,
+      "learning_rate": 4.755972183479814e-05,
+      "loss": 1.9662,
+      "step": 2172
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6635546744142443,
+      "learning_rate": 4.755702928416452e-05,
+      "loss": 2.047,
+      "step": 2173
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6945403993087533,
+      "learning_rate": 4.755433532520289e-05,
+      "loss": 2.2012,
+      "step": 2174
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7612797178756655,
+      "learning_rate": 4.7551639958081454e-05,
+      "loss": 1.9838,
+      "step": 2175
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5700647447494938,
+      "learning_rate": 4.754894318296849e-05,
+      "loss": 1.9949,
+      "step": 2176
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7242183190116863,
+      "learning_rate": 4.754624500003236e-05,
+      "loss": 1.9739,
+      "step": 2177
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6305583716674078,
+      "learning_rate": 4.754354540944154e-05,
+      "loss": 2.1961,
+      "step": 2178
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6938293942764818,
+      "learning_rate": 4.754084441136457e-05,
+      "loss": 2.0865,
+      "step": 2179
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6773749508302627,
+      "learning_rate": 4.753814200597008e-05,
+      "loss": 2.0195,
+      "step": 2180
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8435730733284926,
+      "learning_rate": 4.753543819342679e-05,
+      "loss": 2.0346,
+      "step": 2181
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8816794609161358,
+      "learning_rate": 4.753273297390353e-05,
+      "loss": 2.1961,
+      "step": 2182
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.709258290640546,
+      "learning_rate": 4.753002634756918e-05,
+      "loss": 1.9799,
+      "step": 2183
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.669868895224133,
+      "learning_rate": 4.752731831459272e-05,
+      "loss": 2.0076,
+      "step": 2184
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6814754120180553,
+      "learning_rate": 4.752460887514324e-05,
+      "loss": 1.9609,
+      "step": 2185
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6959610837901892,
+      "learning_rate": 4.752189802938989e-05,
+      "loss": 2.2133,
+      "step": 2186
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.671086802904581,
+      "learning_rate": 4.7519185777501927e-05,
+      "loss": 1.9807,
+      "step": 2187
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6947207885745841,
+      "learning_rate": 4.751647211964869e-05,
+      "loss": 2.0377,
+      "step": 2188
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7706935463950707,
+      "learning_rate": 4.751375705599959e-05,
+      "loss": 2.0318,
+      "step": 2189
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6575969282435566,
+      "learning_rate": 4.751104058672415e-05,
+      "loss": 2.1668,
+      "step": 2190
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7027936953318382,
+      "learning_rate": 4.750832271199197e-05,
+      "loss": 1.9819,
+      "step": 2191
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.755792748557436,
+      "learning_rate": 4.750560343197274e-05,
+      "loss": 2.0773,
+      "step": 2192
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6651096194239591,
+      "learning_rate": 4.7502882746836234e-05,
+      "loss": 1.9705,
+      "step": 2193
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7122685757349444,
+      "learning_rate": 4.75001606567523e-05,
+      "loss": 2.132,
+      "step": 2194
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6180555298714466,
+      "learning_rate": 4.749743716189092e-05,
+      "loss": 2.0136,
+      "step": 2195
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7015893149285746,
+      "learning_rate": 4.74947122624221e-05,
+      "loss": 1.9645,
+      "step": 2196
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7161202047281838,
+      "learning_rate": 4.749198595851599e-05,
+      "loss": 2.0103,
+      "step": 2197
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9211103885903537,
+      "learning_rate": 4.748925825034281e-05,
+      "loss": 2.2655,
+      "step": 2198
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8479565679864598,
+      "learning_rate": 4.7486529138072834e-05,
+      "loss": 1.9188,
+      "step": 2199
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6857165073713344,
+      "learning_rate": 4.748379862187647e-05,
+      "loss": 2.0017,
+      "step": 2200
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7015212024503608,
+      "learning_rate": 4.748106670192419e-05,
+      "loss": 1.982,
+      "step": 2201
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.637524591985425,
+      "learning_rate": 4.7478333378386564e-05,
+      "loss": 2.1817,
+      "step": 2202
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7922389156854768,
+      "learning_rate": 4.747559865143425e-05,
+      "loss": 1.9866,
+      "step": 2203
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5721586071046244,
+      "learning_rate": 4.747286252123797e-05,
+      "loss": 2.0164,
+      "step": 2204
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8408036989563741,
+      "learning_rate": 4.747012498796856e-05,
+      "loss": 1.9922,
+      "step": 2205
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6903678311567681,
+      "learning_rate": 4.746738605179694e-05,
+      "loss": 2.1998,
+      "step": 2206
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7054701838058777,
+      "learning_rate": 4.7464645712894115e-05,
+      "loss": 1.9875,
+      "step": 2207
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7177184396523119,
+      "learning_rate": 4.746190397143116e-05,
+      "loss": 2.015,
+      "step": 2208
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7169249856911328,
+      "learning_rate": 4.745916082757928e-05,
+      "loss": 2.0019,
+      "step": 2209
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7588436353791614,
+      "learning_rate": 4.745641628150972e-05,
+      "loss": 2.2288,
+      "step": 2210
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7199095361220854,
+      "learning_rate": 4.745367033339383e-05,
+      "loss": 2.0128,
+      "step": 2211
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9067348744253737,
+      "learning_rate": 4.7450922983403067e-05,
+      "loss": 2.0238,
+      "step": 2212
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7484307023813518,
+      "learning_rate": 4.744817423170895e-05,
+      "loss": 1.9302,
+      "step": 2213
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8863818961107608,
+      "learning_rate": 4.744542407848309e-05,
+      "loss": 2.2086,
+      "step": 2214
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7061056169893236,
+      "learning_rate": 4.7442672523897205e-05,
+      "loss": 1.9792,
+      "step": 2215
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8597940885522649,
+      "learning_rate": 4.7439919568123075e-05,
+      "loss": 1.973,
+      "step": 2216
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6799079728415194,
+      "learning_rate": 4.743716521133258e-05,
+      "loss": 2.1038,
+      "step": 2217
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8826031338624678,
+      "learning_rate": 4.743440945369769e-05,
+      "loss": 2.2559,
+      "step": 2218
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6554724407852188,
+      "learning_rate": 4.743165229539045e-05,
+      "loss": 2.064,
+      "step": 2219
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8283170721138711,
+      "learning_rate": 4.742889373658301e-05,
+      "loss": 1.9921,
+      "step": 2220
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.769524174784727,
+      "learning_rate": 4.742613377744759e-05,
+      "loss": 2.0017,
+      "step": 2221
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9865132812343029,
+      "learning_rate": 4.742337241815651e-05,
+      "loss": 2.1628,
+      "step": 2222
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7770542092231372,
+      "learning_rate": 4.742060965888218e-05,
+      "loss": 2.0369,
+      "step": 2223
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9627244630515337,
+      "learning_rate": 4.741784549979707e-05,
+      "loss": 2.0156,
+      "step": 2224
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8230627296314685,
+      "learning_rate": 4.741507994107378e-05,
+      "loss": 1.9922,
+      "step": 2225
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0440074939971662,
+      "learning_rate": 4.741231298288496e-05,
+      "loss": 2.1958,
+      "step": 2226
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7999073981294432,
+      "learning_rate": 4.7409544625403376e-05,
+      "loss": 2.0093,
+      "step": 2227
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9101365691576917,
+      "learning_rate": 4.7406774868801854e-05,
+      "loss": 1.9779,
+      "step": 2228
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7328674918503169,
+      "learning_rate": 4.740400371325333e-05,
+      "loss": 2.1031,
+      "step": 2229
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8182139240926867,
+      "learning_rate": 4.740123115893081e-05,
+      "loss": 2.1808,
+      "step": 2230
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.770112800204449,
+      "learning_rate": 4.7398457206007404e-05,
+      "loss": 2.0,
+      "step": 2231
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7397900456538322,
+      "learning_rate": 4.7395681854656304e-05,
+      "loss": 2.046,
+      "step": 2232
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9795482557525457,
+      "learning_rate": 4.7392905105050786e-05,
+      "loss": 1.9678,
+      "step": 2233
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6687499942936698,
+      "learning_rate": 4.73901269573642e-05,
+      "loss": 2.1468,
+      "step": 2234
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9748739052543072,
+      "learning_rate": 4.738734741177002e-05,
+      "loss": 2.0643,
+      "step": 2235
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.611640619787887,
+      "learning_rate": 4.738456646844176e-05,
+      "loss": 1.9843,
+      "step": 2236
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.664586089769723,
+      "learning_rate": 4.738178412755306e-05,
+      "loss": 2.0164,
+      "step": 2237
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8897428639021979,
+      "learning_rate": 4.737900038927763e-05,
+      "loss": 2.2117,
+      "step": 2238
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6341164165295204,
+      "learning_rate": 4.737621525378927e-05,
+      "loss": 1.9878,
+      "step": 2239
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7995184631401555,
+      "learning_rate": 4.737342872126187e-05,
+      "loss": 1.9893,
+      "step": 2240
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7331363588765615,
+      "learning_rate": 4.7370640791869404e-05,
+      "loss": 2.0029,
+      "step": 2241
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7675685791988233,
+      "learning_rate": 4.736785146578593e-05,
+      "loss": 2.1929,
+      "step": 2242
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7447667360232648,
+      "learning_rate": 4.73650607431856e-05,
+      "loss": 1.9471,
+      "step": 2243
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6388670042930327,
+      "learning_rate": 4.736226862424265e-05,
+      "loss": 1.9826,
+      "step": 2244
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.831382957636824,
+      "learning_rate": 4.7359475109131404e-05,
+      "loss": 2.0248,
+      "step": 2245
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6354317487868717,
+      "learning_rate": 4.735668019802627e-05,
+      "loss": 2.147,
+      "step": 2246
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8316490916330804,
+      "learning_rate": 4.735388389110175e-05,
+      "loss": 1.9758,
+      "step": 2247
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8564514551536018,
+      "learning_rate": 4.735108618853242e-05,
+      "loss": 2.0641,
+      "step": 2248
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6832468664648573,
+      "learning_rate": 4.734828709049297e-05,
+      "loss": 2.017,
+      "step": 2249
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8471333581249486,
+      "learning_rate": 4.734548659715814e-05,
+      "loss": 2.2009,
+      "step": 2250
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6124243485508711,
+      "learning_rate": 4.734268470870279e-05,
+      "loss": 1.9736,
+      "step": 2251
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7743167500632855,
+      "learning_rate": 4.733988142530184e-05,
+      "loss": 1.9879,
+      "step": 2252
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6994242435417036,
+      "learning_rate": 4.733707674713032e-05,
+      "loss": 1.9677,
+      "step": 2253
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6255152331373967,
+      "learning_rate": 4.733427067436334e-05,
+      "loss": 2.0716,
+      "step": 2254
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8560530345274114,
+      "learning_rate": 4.7331463207176085e-05,
+      "loss": 2.1861,
+      "step": 2255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7110351004510782,
+      "learning_rate": 4.732865434574385e-05,
+      "loss": 1.9771,
+      "step": 2256
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7478884831751502,
+      "learning_rate": 4.732584409024199e-05,
+      "loss": 1.9959,
+      "step": 2257
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7481174060157333,
+      "learning_rate": 4.732303244084596e-05,
+      "loss": 2.22,
+      "step": 2258
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6823758471538995,
+      "learning_rate": 4.732021939773133e-05,
+      "loss": 2.0039,
+      "step": 2259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8592711529355944,
+      "learning_rate": 4.731740496107369e-05,
+      "loss": 2.0505,
+      "step": 2260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6770663520000079,
+      "learning_rate": 4.731458913104878e-05,
+      "loss": 2.0073,
+      "step": 2261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6564807024453408,
+      "learning_rate": 4.73117719078324e-05,
+      "loss": 2.148,
+      "step": 2262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8674542013229403,
+      "learning_rate": 4.7308953291600445e-05,
+      "loss": 1.9932,
+      "step": 2263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7098646804239714,
+      "learning_rate": 4.7306133282528884e-05,
+      "loss": 2.0072,
+      "step": 2264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7863901683057218,
+      "learning_rate": 4.7303311880793785e-05,
+      "loss": 1.97,
+      "step": 2265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8253347442705621,
+      "learning_rate": 4.730048908657131e-05,
+      "loss": 2.0451,
+      "step": 2266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8262551837465105,
+      "learning_rate": 4.729766490003768e-05,
+      "loss": 2.1899,
+      "step": 2267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7402038367310115,
+      "learning_rate": 4.729483932136923e-05,
+      "loss": 2.0274,
+      "step": 2268
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7775684593042144,
+      "learning_rate": 4.729201235074238e-05,
+      "loss": 1.9667,
+      "step": 2269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6828038902179513,
+      "learning_rate": 4.728918398833361e-05,
+      "loss": 2.2323,
+      "step": 2270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6458227013858772,
+      "learning_rate": 4.7286354234319524e-05,
+      "loss": 2.0015,
+      "step": 2271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6611350587631385,
+      "learning_rate": 4.728352308887679e-05,
+      "loss": 1.9831,
+      "step": 2272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6701283915036454,
+      "learning_rate": 4.728069055218216e-05,
+      "loss": 2.0867,
+      "step": 2273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5739344712462449,
+      "learning_rate": 4.727785662441249e-05,
+      "loss": 2.1736,
+      "step": 2274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6429531847119568,
+      "learning_rate": 4.7275021305744716e-05,
+      "loss": 2.0118,
+      "step": 2275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7812366594742745,
+      "learning_rate": 4.7272184596355844e-05,
+      "loss": 1.9718,
+      "step": 2276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7607882151070329,
+      "learning_rate": 4.7269346496422996e-05,
+      "loss": 2.0225,
+      "step": 2277
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6176206893234318,
+      "learning_rate": 4.726650700612336e-05,
+      "loss": 2.1607,
+      "step": 2278
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6333759925842067,
+      "learning_rate": 4.7263666125634224e-05,
+      "loss": 2.094,
+      "step": 2279
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6006265029852581,
+      "learning_rate": 4.726082385513295e-05,
+      "loss": 2.0022,
+      "step": 2280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.596499968652833,
+      "learning_rate": 4.725798019479699e-05,
+      "loss": 1.9532,
+      "step": 2281
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6467741825110729,
+      "learning_rate": 4.7255135144803884e-05,
+      "loss": 2.2438,
+      "step": 2282
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6138294467899366,
+      "learning_rate": 4.725228870533127e-05,
+      "loss": 2.0103,
+      "step": 2283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8036779673294767,
+      "learning_rate": 4.724944087655686e-05,
+      "loss": 1.964,
+      "step": 2284
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.82541946793859,
+      "learning_rate": 4.724659165865845e-05,
+      "loss": 2.0979,
+      "step": 2285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.626936905684608,
+      "learning_rate": 4.724374105181393e-05,
+      "loss": 1.9684,
+      "step": 2286
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6147490892790837,
+      "learning_rate": 4.724088905620127e-05,
+      "loss": 2.1902,
+      "step": 2287
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7228858432519913,
+      "learning_rate": 4.7238035671998555e-05,
+      "loss": 2.0171,
+      "step": 2288
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6408512904840232,
+      "learning_rate": 4.723518089938391e-05,
+      "loss": 2.028,
+      "step": 2289
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6334019702754733,
+      "learning_rate": 4.723232473853557e-05,
+      "loss": 2.1964,
+      "step": 2290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8022596608891602,
+      "learning_rate": 4.722946718963187e-05,
+      "loss": 2.0482,
+      "step": 2291
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.744178095420191,
+      "learning_rate": 4.7226608252851215e-05,
+      "loss": 1.994,
+      "step": 2292
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6400314233154624,
+      "learning_rate": 4.722374792837209e-05,
+      "loss": 2.0074,
+      "step": 2293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7664532159925955,
+      "learning_rate": 4.722088621637309e-05,
+      "loss": 2.214,
+      "step": 2294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6593368383900502,
+      "learning_rate": 4.721802311703287e-05,
+      "loss": 1.9766,
+      "step": 2295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9289243100644894,
+      "learning_rate": 4.72151586305302e-05,
+      "loss": 2.0148,
+      "step": 2296
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6120081628988943,
+      "learning_rate": 4.7212292757043916e-05,
+      "loss": 2.0284,
+      "step": 2297
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9375660098034472,
+      "learning_rate": 4.7209425496752945e-05,
+      "loss": 1.9266,
+      "step": 2298
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7405549037046271,
+      "learning_rate": 4.72065568498363e-05,
+      "loss": 2.163,
+      "step": 2299
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7500093703188777,
+      "learning_rate": 4.720368681647308e-05,
+      "loss": 1.9329,
+      "step": 2300
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.855537338743673,
+      "learning_rate": 4.720081539684248e-05,
+      "loss": 2.0065,
+      "step": 2301
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6596873923541644,
+      "learning_rate": 4.719794259112377e-05,
+      "loss": 2.1889,
+      "step": 2302
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.804295185037117,
+      "learning_rate": 4.719506839949631e-05,
+      "loss": 2.0001,
+      "step": 2303
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6570906406545723,
+      "learning_rate": 4.719219282213956e-05,
+      "loss": 2.1111,
+      "step": 2304
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7352542623935455,
+      "learning_rate": 4.718931585923304e-05,
+      "loss": 1.9935,
+      "step": 2305
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7498062613271765,
+      "learning_rate": 4.718643751095637e-05,
+      "loss": 2.1957,
+      "step": 2306
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7033133567511664,
+      "learning_rate": 4.718355777748927e-05,
+      "loss": 1.9835,
+      "step": 2307
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6600743093053663,
+      "learning_rate": 4.7180676659011516e-05,
+      "loss": 1.9515,
+      "step": 2308
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.683478817451781,
+      "learning_rate": 4.7177794155703004e-05,
+      "loss": 1.9768,
+      "step": 2309
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6630316729717854,
+      "learning_rate": 4.7174910267743695e-05,
+      "loss": 2.0506,
+      "step": 2310
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.64931722064103,
+      "learning_rate": 4.7172024995313634e-05,
+      "loss": 2.1634,
+      "step": 2311
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7905698167488725,
+      "learning_rate": 4.716913833859298e-05,
+      "loss": 1.9659,
+      "step": 2312
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6510276257360043,
+      "learning_rate": 4.7166250297761936e-05,
+      "loss": 1.9519,
+      "step": 2313
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7963777072486765,
+      "learning_rate": 4.716336087300083e-05,
+      "loss": 2.208,
+      "step": 2314
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6898933415254367,
+      "learning_rate": 4.716047006449005e-05,
+      "loss": 1.9863,
+      "step": 2315
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8683789186144247,
+      "learning_rate": 4.715757787241009e-05,
+      "loss": 2.0516,
+      "step": 2316
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8200129932086837,
+      "learning_rate": 4.715468429694152e-05,
+      "loss": 2.0297,
+      "step": 2317
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6914844773068379,
+      "learning_rate": 4.715178933826499e-05,
+      "loss": 1.9969,
+      "step": 2318
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8232841994859744,
+      "learning_rate": 4.7148892996561256e-05,
+      "loss": 2.2172,
+      "step": 2319
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7338345967617775,
+      "learning_rate": 4.714599527201113e-05,
+      "loss": 2.0036,
+      "step": 2320
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7050474363717165,
+      "learning_rate": 4.7143096164795554e-05,
+      "loss": 2.0321,
+      "step": 2321
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7349141590850311,
+      "learning_rate": 4.714019567509552e-05,
+      "loss": 2.0546,
+      "step": 2322
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6496724210225138,
+      "learning_rate": 4.713729380309211e-05,
+      "loss": 2.1735,
+      "step": 2323
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7487579177400563,
+      "learning_rate": 4.71343905489665e-05,
+      "loss": 1.9592,
+      "step": 2324
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6852632013917966,
+      "learning_rate": 4.7131485912899964e-05,
+      "loss": 1.985,
+      "step": 2325
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7545291562493922,
+      "learning_rate": 4.712857989507384e-05,
+      "loss": 2.1665,
+      "step": 2326
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.644067669464844,
+      "learning_rate": 4.712567249566958e-05,
+      "loss": 1.9815,
+      "step": 2327
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.902208787096129,
+      "learning_rate": 4.7122763714868676e-05,
+      "loss": 2.006,
+      "step": 2328
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7741701573641254,
+      "learning_rate": 4.7119853552852754e-05,
+      "loss": 1.9633,
+      "step": 2329
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8202262785045211,
+      "learning_rate": 4.711694200980351e-05,
+      "loss": 2.0201,
+      "step": 2330
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7125623033661771,
+      "learning_rate": 4.711402908590271e-05,
+      "loss": 2.1368,
+      "step": 2331
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8195414581391616,
+      "learning_rate": 4.7111114781332236e-05,
+      "loss": 1.9875,
+      "step": 2332
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6773615832984973,
+      "learning_rate": 4.710819909627403e-05,
+      "loss": 2.0147,
+      "step": 2333
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7920689320867647,
+      "learning_rate": 4.7105282030910125e-05,
+      "loss": 2.2293,
+      "step": 2334
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5946582109202839,
+      "learning_rate": 4.710236358542266e-05,
+      "loss": 2.0753,
+      "step": 2335
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.657712051303984,
+      "learning_rate": 4.7099443759993837e-05,
+      "loss": 1.983,
+      "step": 2336
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6048672789078698,
+      "learning_rate": 4.709652255480594e-05,
+      "loss": 1.9798,
+      "step": 2337
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7193073746587328,
+      "learning_rate": 4.709359997004138e-05,
+      "loss": 1.9517,
+      "step": 2338
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6935720562506166,
+      "learning_rate": 4.7090676005882604e-05,
+      "loss": 2.206,
+      "step": 2339
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7610782032850921,
+      "learning_rate": 4.7087750662512176e-05,
+      "loss": 1.9729,
+      "step": 2340
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8141273375887564,
+      "learning_rate": 4.708482394011273e-05,
+      "loss": 1.9995,
+      "step": 2341
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7121293148824727,
+      "learning_rate": 4.708189583886701e-05,
+      "loss": 1.9934,
+      "step": 2342
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7981987079489882,
+      "learning_rate": 4.707896635895781e-05,
+      "loss": 2.1746,
+      "step": 2343
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7721896421473234,
+      "learning_rate": 4.7076035500568036e-05,
+      "loss": 1.9816,
+      "step": 2344
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7399160644936708,
+      "learning_rate": 4.7073103263880684e-05,
+      "loss": 2.0032,
+      "step": 2345
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8728616936472825,
+      "learning_rate": 4.707016964907881e-05,
+      "loss": 2.1918,
+      "step": 2346
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7415057420742341,
+      "learning_rate": 4.7067234656345585e-05,
+      "loss": 2.0882,
+      "step": 2347
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8717150180252916,
+      "learning_rate": 4.706429828586424e-05,
+      "loss": 2.0265,
+      "step": 2348
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6964360105938643,
+      "learning_rate": 4.706136053781811e-05,
+      "loss": 1.982,
+      "step": 2349
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7922125513153762,
+      "learning_rate": 4.705842141239061e-05,
+      "loss": 1.9785,
+      "step": 2350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8620178246255268,
+      "learning_rate": 4.7055480909765246e-05,
+      "loss": 2.2115,
+      "step": 2351
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6431078819446007,
+      "learning_rate": 4.7052539030125605e-05,
+      "loss": 1.9849,
+      "step": 2352
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8656972101826599,
+      "learning_rate": 4.7049595773655354e-05,
+      "loss": 2.0261,
+      "step": 2353
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5748791145891335,
+      "learning_rate": 4.7046651140538254e-05,
+      "loss": 2.0088,
+      "step": 2354
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9484448275287364,
+      "learning_rate": 4.704370513095816e-05,
+      "loss": 2.1681,
+      "step": 2355
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6714215953690877,
+      "learning_rate": 4.7040757745099e-05,
+      "loss": 1.9677,
+      "step": 2356
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7568710651115563,
+      "learning_rate": 4.703780898314478e-05,
+      "loss": 2.0134,
+      "step": 2357
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7322763783823012,
+      "learning_rate": 4.7034858845279614e-05,
+      "loss": 2.1731,
+      "step": 2358
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6248436094859314,
+      "learning_rate": 4.7031907331687686e-05,
+      "loss": 2.0461,
+      "step": 2359
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7858325477692505,
+      "learning_rate": 4.7028954442553276e-05,
+      "loss": 1.9798,
+      "step": 2360
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6867747856137486,
+      "learning_rate": 4.7026000178060745e-05,
+      "loss": 1.9432,
+      "step": 2361
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7100661754213549,
+      "learning_rate": 4.702304453839453e-05,
+      "loss": 2.0057,
+      "step": 2362
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7705055658611255,
+      "learning_rate": 4.7020087523739186e-05,
+      "loss": 2.1714,
+      "step": 2363
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6361664936963715,
+      "learning_rate": 4.7017129134279307e-05,
+      "loss": 2.0617,
+      "step": 2364
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.62851620789004,
+      "learning_rate": 4.7014169370199615e-05,
+      "loss": 2.0086,
+      "step": 2365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7360203106700033,
+      "learning_rate": 4.701120823168489e-05,
+      "loss": 2.0948,
+      "step": 2366
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7045687881489059,
+      "learning_rate": 4.7008245718920005e-05,
+      "loss": 2.1657,
+      "step": 2367
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6437362584729972,
+      "learning_rate": 4.7005281832089934e-05,
+      "loss": 2.0071,
+      "step": 2368
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7407344383757167,
+      "learning_rate": 4.7002316571379715e-05,
+      "loss": 2.0272,
+      "step": 2369
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6528316540081227,
+      "learning_rate": 4.699934993697448e-05,
+      "loss": 1.9847,
+      "step": 2370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.707142817813905,
+      "learning_rate": 4.6996381929059466e-05,
+      "loss": 2.232,
+      "step": 2371
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7533255255702686,
+      "learning_rate": 4.699341254781995e-05,
+      "loss": 2.0115,
+      "step": 2372
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6602538492950344,
+      "learning_rate": 4.6990441793441344e-05,
+      "loss": 1.9749,
+      "step": 2373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7460006573247863,
+      "learning_rate": 4.698746966610912e-05,
+      "loss": 1.984,
+      "step": 2374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.644995270240468,
+      "learning_rate": 4.6984496166008835e-05,
+      "loss": 2.2163,
+      "step": 2375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8094113320180328,
+      "learning_rate": 4.6981521293326136e-05,
+      "loss": 1.994,
+      "step": 2376
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6921645976726828,
+      "learning_rate": 4.697854504824677e-05,
+      "loss": 1.9765,
+      "step": 2377
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6157273220243186,
+      "learning_rate": 4.697556743095654e-05,
+      "loss": 2.0594,
+      "step": 2378
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8109147459615079,
+      "learning_rate": 4.697258844164136e-05,
+      "loss": 2.2266,
+      "step": 2379
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.600012315143977,
+      "learning_rate": 4.696960808048721e-05,
+      "loss": 1.9498,
+      "step": 2380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7866080512635013,
+      "learning_rate": 4.696662634768019e-05,
+      "loss": 1.9794,
+      "step": 2381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6267702135950278,
+      "learning_rate": 4.6963643243406427e-05,
+      "loss": 1.9943,
+      "step": 2382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6558067975411246,
+      "learning_rate": 4.69606587678522e-05,
+      "loss": 2.2223,
+      "step": 2383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6278321107074327,
+      "learning_rate": 4.695767292120383e-05,
+      "loss": 2.0248,
+      "step": 2384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6757816358998711,
+      "learning_rate": 4.695468570364773e-05,
+      "loss": 2.0317,
+      "step": 2385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6539438843570307,
+      "learning_rate": 4.695169711537041e-05,
+      "loss": 1.9651,
+      "step": 2386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6967888616323279,
+      "learning_rate": 4.694870715655846e-05,
+      "loss": 2.1839,
+      "step": 2387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6999453531174823,
+      "learning_rate": 4.694571582739856e-05,
+      "loss": 1.99,
+      "step": 2388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6556003573154653,
+      "learning_rate": 4.6942723128077456e-05,
+      "loss": 1.9779,
+      "step": 2389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.784948135223806,
+      "learning_rate": 4.693972905878201e-05,
+      "loss": 2.2027,
+      "step": 2390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.63280086298786,
+      "learning_rate": 4.693673361969915e-05,
+      "loss": 2.0406,
+      "step": 2391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6852467634835429,
+      "learning_rate": 4.693373681101589e-05,
+      "loss": 2.0333,
+      "step": 2392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6550995047689948,
+      "learning_rate": 4.6930738632919334e-05,
+      "loss": 2.0042,
+      "step": 2393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6820817184019495,
+      "learning_rate": 4.6927739085596665e-05,
+      "loss": 2.0063,
+      "step": 2394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6858583753288315,
+      "learning_rate": 4.692473816923517e-05,
+      "loss": 2.1682,
+      "step": 2395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6994882500529027,
+      "learning_rate": 4.69217358840222e-05,
+      "loss": 1.9989,
+      "step": 2396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6813766758544803,
+      "learning_rate": 4.691873223014521e-05,
+      "loss": 2.0753,
+      "step": 2397
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.743267037534749,
+      "learning_rate": 4.691572720779171e-05,
+      "loss": 1.9845,
+      "step": 2398
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7464977535491373,
+      "learning_rate": 4.6912720817149344e-05,
+      "loss": 2.1735,
+      "step": 2399
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6988264981373272,
+      "learning_rate": 4.6909713058405786e-05,
+      "loss": 1.997,
+      "step": 2400
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7415633551152035,
+      "learning_rate": 4.6906703931748837e-05,
+      "loss": 1.9978,
+      "step": 2401
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6862739221398028,
+      "learning_rate": 4.690369343736636e-05,
+      "loss": 1.9716,
+      "step": 2402
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7525748910971672,
+      "learning_rate": 4.690068157544633e-05,
+      "loss": 2.2354,
+      "step": 2403
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7222412199060225,
+      "learning_rate": 4.689766834617677e-05,
+      "loss": 2.0065,
+      "step": 2404
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8914678611446539,
+      "learning_rate": 4.689465374974583e-05,
+      "loss": 1.9628,
+      "step": 2405
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6815979702097102,
+      "learning_rate": 4.68916377863417e-05,
+      "loss": 2.02,
+      "step": 2406
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9786463779104423,
+      "learning_rate": 4.6888620456152685e-05,
+      "loss": 2.2259,
+      "step": 2407
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6766761212809368,
+      "learning_rate": 4.6885601759367185e-05,
+      "loss": 2.0244,
+      "step": 2408
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9457390830997467,
+      "learning_rate": 4.6882581696173654e-05,
+      "loss": 2.0258,
+      "step": 2409
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6682738258013812,
+      "learning_rate": 4.6879560266760655e-05,
+      "loss": 2.0148,
+      "step": 2410
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9996729742622479,
+      "learning_rate": 4.6876537471316825e-05,
+      "loss": 2.1908,
+      "step": 2411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7582930961960058,
+      "learning_rate": 4.687351331003089e-05,
+      "loss": 1.9427,
+      "step": 2412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9209677685399305,
+      "learning_rate": 4.687048778309165e-05,
+      "loss": 2.0308,
+      "step": 2413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7093914039779269,
+      "learning_rate": 4.686746089068803e-05,
+      "loss": 1.9277,
+      "step": 2414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6020204047246959,
+      "learning_rate": 4.6864432633008984e-05,
+      "loss": 2.2177,
+      "step": 2415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6690675547998777,
+      "learning_rate": 4.686140301024358e-05,
+      "loss": 1.9599,
+      "step": 2416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6935338427150075,
+      "learning_rate": 4.6858372022580985e-05,
+      "loss": 1.9729,
+      "step": 2417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6817606223249436,
+      "learning_rate": 4.685533967021043e-05,
+      "loss": 1.995,
+      "step": 2418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6533889145307378,
+      "learning_rate": 4.685230595332123e-05,
+      "loss": 2.1781,
+      "step": 2419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6272304993663916,
+      "learning_rate": 4.6849270872102795e-05,
+      "loss": 1.9654,
+      "step": 2420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6208635882396977,
+      "learning_rate": 4.684623442674463e-05,
+      "loss": 1.9978,
+      "step": 2421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7090455630635236,
+      "learning_rate": 4.684319661743629e-05,
+      "loss": 2.0711,
+      "step": 2422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.785345367071294,
+      "learning_rate": 4.684015744436746e-05,
+      "loss": 2.1855,
+      "step": 2423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7682614166296461,
+      "learning_rate": 4.683711690772788e-05,
+      "loss": 1.996,
+      "step": 2424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.673925360401252,
+      "learning_rate": 4.683407500770739e-05,
+      "loss": 1.9595,
+      "step": 2425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7463416876586877,
+      "learning_rate": 4.6831031744495886e-05,
+      "loss": 1.9953,
+      "step": 2426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6868098204062206,
+      "learning_rate": 4.682798711828339e-05,
+      "loss": 2.1238,
+      "step": 2427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7755090925700292,
+      "learning_rate": 4.6824941129259984e-05,
+      "loss": 2.0348,
+      "step": 2428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6580598810441544,
+      "learning_rate": 4.6821893777615853e-05,
+      "loss": 1.9601,
+      "step": 2429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5903494072969236,
+      "learning_rate": 4.6818845063541237e-05,
+      "loss": 2.0145,
+      "step": 2430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6606038154073814,
+      "learning_rate": 4.68157949872265e-05,
+      "loss": 2.1669,
+      "step": 2431
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5882545759951369,
+      "learning_rate": 4.681274354886205e-05,
+      "loss": 2.011,
+      "step": 2432
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6009549583719351,
+      "learning_rate": 4.680969074863842e-05,
+      "loss": 1.991,
+      "step": 2433
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6440689349448566,
+      "learning_rate": 4.680663658674619e-05,
+      "loss": 2.0577,
+      "step": 2434
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.593953457515576,
+      "learning_rate": 4.680358106337607e-05,
+      "loss": 2.1275,
+      "step": 2435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6209700270235651,
+      "learning_rate": 4.6800524178718794e-05,
+      "loss": 1.9511,
+      "step": 2436
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.629379427169456,
+      "learning_rate": 4.6797465932965244e-05,
+      "loss": 1.9902,
+      "step": 2437
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6134594950209334,
+      "learning_rate": 4.6794406326306355e-05,
+      "loss": 1.972,
+      "step": 2438
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7002850129913979,
+      "learning_rate": 4.6791345358933136e-05,
+      "loss": 2.1939,
+      "step": 2439
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5788637489564847,
+      "learning_rate": 4.678828303103671e-05,
+      "loss": 2.0575,
+      "step": 2440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7493564311339685,
+      "learning_rate": 4.6785219342808267e-05,
+      "loss": 2.0173,
+      "step": 2441
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5873567821096899,
+      "learning_rate": 4.678215429443908e-05,
+      "loss": 1.9608,
+      "step": 2442
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6946516804751978,
+      "learning_rate": 4.677908788612052e-05,
+      "loss": 2.1833,
+      "step": 2443
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5749428253025071,
+      "learning_rate": 4.677602011804403e-05,
+      "loss": 1.9992,
+      "step": 2444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.667333628695,
+      "learning_rate": 4.677295099040115e-05,
+      "loss": 2.0027,
+      "step": 2445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.564125826582981,
+      "learning_rate": 4.6769880503383494e-05,
+      "loss": 2.0503,
+      "step": 2446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6558112280922334,
+      "learning_rate": 4.676680865718276e-05,
+      "loss": 2.1799,
+      "step": 2447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6205326355591049,
+      "learning_rate": 4.676373545199075e-05,
+      "loss": 1.9966,
+      "step": 2448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6013401005735152,
+      "learning_rate": 4.676066088799932e-05,
+      "loss": 1.993,
+      "step": 2449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7082665361921955,
+      "learning_rate": 4.675758496540044e-05,
+      "loss": 2.0307,
+      "step": 2450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6815068920718871,
+      "learning_rate": 4.675450768438615e-05,
+      "loss": 2.2017,
+      "step": 2451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7863016694375223,
+      "learning_rate": 4.6751429045148564e-05,
+      "loss": 1.9444,
+      "step": 2452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.641644814287374,
+      "learning_rate": 4.6748349047879915e-05,
+      "loss": 2.0592,
+      "step": 2453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6637420359780016,
+      "learning_rate": 4.674526769277249e-05,
+      "loss": 1.9671,
+      "step": 2454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8215598966312048,
+      "learning_rate": 4.674218498001867e-05,
+      "loss": 2.1768,
+      "step": 2455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6256409272116019,
+      "learning_rate": 4.6739100909810924e-05,
+      "loss": 1.9473,
+      "step": 2456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7411965677407634,
+      "learning_rate": 4.67360154823418e-05,
+      "loss": 2.0086,
+      "step": 2457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6874492364254969,
+      "learning_rate": 4.6732928697803934e-05,
+      "loss": 2.0286,
+      "step": 2458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7214778841287579,
+      "learning_rate": 4.6729840556390055e-05,
+      "loss": 2.2292,
+      "step": 2459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8902793631900654,
+      "learning_rate": 4.672675105829296e-05,
+      "loss": 2.0058,
+      "step": 2460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6500647663352871,
+      "learning_rate": 4.6723660203705534e-05,
+      "loss": 1.9868,
+      "step": 2461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8034596807430502,
+      "learning_rate": 4.6720567992820766e-05,
+      "loss": 1.9587,
+      "step": 2462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6284560820977981,
+      "learning_rate": 4.6717474425831696e-05,
+      "loss": 2.1853,
+      "step": 2463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7452628968630005,
+      "learning_rate": 4.6714379502931495e-05,
+      "loss": 1.9979,
+      "step": 2464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9020655853096545,
+      "learning_rate": 4.671128322431337e-05,
+      "loss": 2.1166,
+      "step": 2465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6905652344042049,
+      "learning_rate": 4.670818559017064e-05,
+      "loss": 1.9956,
+      "step": 2466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9088184554615418,
+      "learning_rate": 4.67050866006967e-05,
+      "loss": 2.1709,
+      "step": 2467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9334777492351023,
+      "learning_rate": 4.6701986256085046e-05,
+      "loss": 2.0273,
+      "step": 2468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7607925636820858,
+      "learning_rate": 4.669888455652923e-05,
+      "loss": 2.0003,
+      "step": 2469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.942315784948906,
+      "learning_rate": 4.669578150222291e-05,
+      "loss": 1.9725,
+      "step": 2470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6436585398899733,
+      "learning_rate": 4.669267709335983e-05,
+      "loss": 2.0215,
+      "step": 2471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9908122001421829,
+      "learning_rate": 4.66895713301338e-05,
+      "loss": 2.159,
+      "step": 2472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7194983822927975,
+      "learning_rate": 4.6686464212738724e-05,
+      "loss": 2.0396,
+      "step": 2473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.710400495671655,
+      "learning_rate": 4.6683355741368594e-05,
+      "loss": 2.0254,
+      "step": 2474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8116954626767107,
+      "learning_rate": 4.66802459162175e-05,
+      "loss": 2.1892,
+      "step": 2475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5833524321714841,
+      "learning_rate": 4.6677134737479576e-05,
+      "loss": 1.9489,
+      "step": 2476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.745788769867993,
+      "learning_rate": 4.6674022205349085e-05,
+      "loss": 2.0047,
+      "step": 2477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6438366094071706,
+      "learning_rate": 4.6670908320020346e-05,
+      "loss": 1.9714,
+      "step": 2478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7288017204382724,
+      "learning_rate": 4.666779308168778e-05,
+      "loss": 2.1895,
+      "step": 2479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8968327544160674,
+      "learning_rate": 4.666467649054587e-05,
+      "loss": 1.9725,
+      "step": 2480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6789129988395259,
+      "learning_rate": 4.666155854678921e-05,
+      "loss": 1.9716,
+      "step": 2481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8969814698472731,
+      "learning_rate": 4.6658439250612465e-05,
+      "loss": 1.966,
+      "step": 2482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6383086482161018,
+      "learning_rate": 4.665531860221038e-05,
+      "loss": 2.1536,
+      "step": 2483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.088937701805825,
+      "learning_rate": 4.665219660177779e-05,
+      "loss": 2.0763,
+      "step": 2484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7555931351443268,
+      "learning_rate": 4.664907324950962e-05,
+      "loss": 1.9906,
+      "step": 2485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.853980941632526,
+      "learning_rate": 4.664594854560087e-05,
+      "loss": 1.998,
+      "step": 2486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8620016157143996,
+      "learning_rate": 4.664282249024663e-05,
+      "loss": 2.1926,
+      "step": 2487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6415795421207837,
+      "learning_rate": 4.6639695083642064e-05,
+      "loss": 2.0103,
+      "step": 2488
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8729458814588819,
+      "learning_rate": 4.6636566325982446e-05,
+      "loss": 1.9789,
+      "step": 2489
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7013923046202719,
+      "learning_rate": 4.66334362174631e-05,
+      "loss": 2.0691,
+      "step": 2490
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9335828328259878,
+      "learning_rate": 4.6630304758279464e-05,
+      "loss": 2.1607,
+      "step": 2491
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8134259988417311,
+      "learning_rate": 4.6627171948627035e-05,
+      "loss": 2.0134,
+      "step": 2492
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6366219479972837,
+      "learning_rate": 4.662403778870142e-05,
+      "loss": 1.9299,
+      "step": 2493
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8955690564080252,
+      "learning_rate": 4.662090227869829e-05,
+      "loss": 2.0202,
+      "step": 2494
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6893279595012406,
+      "learning_rate": 4.661776541881341e-05,
+      "loss": 2.1555,
+      "step": 2495
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8340311214658663,
+      "learning_rate": 4.661462720924262e-05,
+      "loss": 2.0391,
+      "step": 2496
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6726823172223888,
+      "learning_rate": 4.661148765018187e-05,
+      "loss": 1.9803,
+      "step": 2497
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7181456520297506,
+      "learning_rate": 4.660834674182716e-05,
+      "loss": 1.9901,
+      "step": 2498
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8283961551635887,
+      "learning_rate": 4.6605204484374595e-05,
+      "loss": 2.1864,
+      "step": 2499
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5952219369838586,
+      "learning_rate": 4.6602060878020356e-05,
+      "loss": 1.9861,
+      "step": 2500
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9130134026602053,
+      "learning_rate": 4.659891592296071e-05,
+      "loss": 2.0087,
+      "step": 2501
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6755865748467891,
+      "learning_rate": 4.6595769619392017e-05,
+      "loss": 2.0502,
+      "step": 2502
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9135357547477468,
+      "learning_rate": 4.65926219675107e-05,
+      "loss": 1.9739,
+      "step": 2503
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.835995010263368,
+      "learning_rate": 4.65894729675133e-05,
+      "loss": 2.1934,
+      "step": 2504
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.702827364277536,
+      "learning_rate": 4.65863226195964e-05,
+      "loss": 1.9983,
+      "step": 2505
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8758870258136955,
+      "learning_rate": 4.658317092395671e-05,
+      "loss": 2.0134,
+      "step": 2506
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5975634318088306,
+      "learning_rate": 4.6580017880790996e-05,
+      "loss": 2.1986,
+      "step": 2507
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.721687439031074,
+      "learning_rate": 4.65768634902961e-05,
+      "loss": 2.0557,
+      "step": 2508
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6428451818146449,
+      "learning_rate": 4.657370775266898e-05,
+      "loss": 1.9956,
+      "step": 2509
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7195816553338311,
+      "learning_rate": 4.6570550668106664e-05,
+      "loss": 1.9929,
+      "step": 2510
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.888510653377431,
+      "learning_rate": 4.656739223680625e-05,
+      "loss": 2.2015,
+      "step": 2511
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6077243773500849,
+      "learning_rate": 4.656423245896494e-05,
+      "loss": 1.9667,
+      "step": 2512
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7652849593710155,
+      "learning_rate": 4.656107133478001e-05,
+      "loss": 2.002,
+      "step": 2513
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6359814216983952,
+      "learning_rate": 4.6557908864448824e-05,
+      "loss": 1.9514,
+      "step": 2514
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.703376085952886,
+      "learning_rate": 4.6554745048168825e-05,
+      "loss": 2.0369,
+      "step": 2515
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8130185953694179,
+      "learning_rate": 4.6551579886137544e-05,
+      "loss": 2.1346,
+      "step": 2516
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5785374215210156,
+      "learning_rate": 4.654841337855259e-05,
+      "loss": 2.0095,
+      "step": 2517
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7922265488236757,
+      "learning_rate": 4.6545245525611665e-05,
+      "loss": 1.9888,
+      "step": 2518
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6792421944658242,
+      "learning_rate": 4.654207632751255e-05,
+      "loss": 2.1704,
+      "step": 2519
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6529081425535241,
+      "learning_rate": 4.6538905784453126e-05,
+      "loss": 1.9981,
+      "step": 2520
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6233506140798122,
+      "learning_rate": 4.653573389663132e-05,
+      "loss": 2.0454,
+      "step": 2521
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6143485948861483,
+      "learning_rate": 4.6532560664245175e-05,
+      "loss": 2.0105,
+      "step": 2522
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5878004162490259,
+      "learning_rate": 4.652938608749281e-05,
+      "loss": 2.152,
+      "step": 2523
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6329684725306782,
+      "learning_rate": 4.6526210166572427e-05,
+      "loss": 1.9562,
+      "step": 2524
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6058239023321682,
+      "learning_rate": 4.652303290168231e-05,
+      "loss": 2.0145,
+      "step": 2525
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5722880139831528,
+      "learning_rate": 4.651985429302083e-05,
+      "loss": 2.0082,
+      "step": 2526
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5812835698838148,
+      "learning_rate": 4.651667434078645e-05,
+      "loss": 2.0681,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6398155508932993,
+      "learning_rate": 4.651349304517768e-05,
+      "loss": 2.1892,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6711898483730057,
+      "learning_rate": 4.6510310406393176e-05,
+      "loss": 2.0353,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6060037677450617,
+      "learning_rate": 4.650712642463162e-05,
+      "loss": 1.9769,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7229068100509315,
+      "learning_rate": 4.6503941100091816e-05,
+      "loss": 2.1607,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5770847722411103,
+      "learning_rate": 4.6500754432972615e-05,
+      "loss": 1.9753,
+      "step": 2532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6245591970449133,
+      "learning_rate": 4.6497566423473e-05,
+      "loss": 2.0326,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6484247102216382,
+      "learning_rate": 4.6494377071791996e-05,
+      "loss": 2.0061,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6980271774582169,
+      "learning_rate": 4.649118637812872e-05,
+      "loss": 1.9911,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6816791513504178,
+      "learning_rate": 4.648799434268241e-05,
+      "loss": 2.2098,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7105850829778596,
+      "learning_rate": 4.648480096565232e-05,
+      "loss": 1.9778,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6374545856159565,
+      "learning_rate": 4.6481606247237854e-05,
+      "loss": 1.9386,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7769387561562383,
+      "learning_rate": 4.647841018763846e-05,
+      "loss": 2.2127,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6266625671429538,
+      "learning_rate": 4.647521278705368e-05,
+      "loss": 2.0484,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6609623499769247,
+      "learning_rate": 4.647201404568314e-05,
+      "loss": 1.9669,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7607320207173331,
+      "learning_rate": 4.6468813963726556e-05,
+      "loss": 2.0151,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7523095930951635,
+      "learning_rate": 4.646561254138372e-05,
+      "loss": 2.1938,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6403148566506762,
+      "learning_rate": 4.646240977885451e-05,
+      "loss": 1.9833,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6661875918009426,
+      "learning_rate": 4.645920567633889e-05,
+      "loss": 2.0046,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6264816346894045,
+      "learning_rate": 4.645600023403691e-05,
+      "loss": 2.1022,
+      "step": 2546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5803142986930665,
+      "learning_rate": 4.645279345214868e-05,
+      "loss": 1.9611,
+      "step": 2547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.650348397766849,
+      "learning_rate": 4.644958533087443e-05,
+      "loss": 2.1951,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5958432771031661,
+      "learning_rate": 4.6446375870414446e-05,
+      "loss": 1.9965,
+      "step": 2549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6126603952489209,
+      "learning_rate": 4.644316507096912e-05,
+      "loss": 2.023,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6381507234176351,
+      "learning_rate": 4.64399529327389e-05,
+      "loss": 2.1905,
+      "step": 2551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5460717086865683,
+      "learning_rate": 4.643673945592435e-05,
+      "loss": 2.0771,
+      "step": 2552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6064345758734442,
+      "learning_rate": 4.643352464072608e-05,
+      "loss": 2.0046,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6930770852689816,
+      "learning_rate": 4.643030848734483e-05,
+      "loss": 1.9558,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5887981747283252,
+      "learning_rate": 4.642709099598137e-05,
+      "loss": 1.9727,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7391375565275787,
+      "learning_rate": 4.64238721668366e-05,
+      "loss": 2.1845,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6099042053901921,
+      "learning_rate": 4.642065200011147e-05,
+      "loss": 1.9381,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6410366767558648,
+      "learning_rate": 4.641743049600705e-05,
+      "loss": 2.0433,
+      "step": 2558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5850955014806185,
+      "learning_rate": 4.641420765472446e-05,
+      "loss": 1.968,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7167077705853634,
+      "learning_rate": 4.641098347646491e-05,
+      "loss": 2.1841,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6040635664927856,
+      "learning_rate": 4.6407757961429696e-05,
+      "loss": 2.0183,
+      "step": 2561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5906424527254666,
+      "learning_rate": 4.640453110982022e-05,
+      "loss": 1.9386,
+      "step": 2562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6744807983581735,
+      "learning_rate": 4.640130292183793e-05,
+      "loss": 2.1929,
+      "step": 2563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6552171141975681,
+      "learning_rate": 4.639807339768437e-05,
+      "loss": 2.1057,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5951574643730799,
+      "learning_rate": 4.63948425375612e-05,
+      "loss": 1.9701,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6844788004792994,
+      "learning_rate": 4.6391610341670105e-05,
+      "loss": 1.9789,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6408402412644457,
+      "learning_rate": 4.6388376810212905e-05,
+      "loss": 1.9848,
+      "step": 2567
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6167414151099627,
+      "learning_rate": 4.6385141943391466e-05,
+      "loss": 2.1842,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6755076403896096,
+      "learning_rate": 4.6381905741407774e-05,
+      "loss": 1.9806,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6814254611283641,
+      "learning_rate": 4.637866820446386e-05,
+      "loss": 1.9505,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6557317028419934,
+      "learning_rate": 4.637542933276186e-05,
+      "loss": 2.0537,
+      "step": 2571
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6007755011050209,
+      "learning_rate": 4.6372189126504004e-05,
+      "loss": 2.1509,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7141725935732299,
+      "learning_rate": 4.636894758589257e-05,
+      "loss": 2.0141,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6034986590006324,
+      "learning_rate": 4.6365704711129973e-05,
+      "loss": 1.9326,
+      "step": 2574
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6847327216872016,
+      "learning_rate": 4.636246050241864e-05,
+      "loss": 2.1582,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6466119695120638,
+      "learning_rate": 4.635921495996115e-05,
+      "loss": 2.0249,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5275950533765944,
+      "learning_rate": 4.635596808396011e-05,
+      "loss": 2.0512,
+      "step": 2577
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5945026369793002,
+      "learning_rate": 4.635271987461827e-05,
+      "loss": 1.9339,
+      "step": 2578
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6578289301259045,
+      "learning_rate": 4.634947033213839e-05,
+      "loss": 1.9694,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6301945375508965,
+      "learning_rate": 4.634621945672338e-05,
+      "loss": 2.1696,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6098334593422985,
+      "learning_rate": 4.6342967248576194e-05,
+      "loss": 2.0155,
+      "step": 2581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5744365610517691,
+      "learning_rate": 4.633971370789989e-05,
+      "loss": 2.0213,
+      "step": 2582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6221750693439922,
+      "learning_rate": 4.633645883489759e-05,
+      "loss": 2.0296,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6405853238989996,
+      "learning_rate": 4.633320262977251e-05,
+      "loss": 2.1845,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6666321309748923,
+      "learning_rate": 4.632994509272795e-05,
+      "loss": 2.0011,
+      "step": 2585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6275811478950984,
+      "learning_rate": 4.6326686223967295e-05,
+      "loss": 1.9741,
+      "step": 2586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6140932652479769,
+      "learning_rate": 4.632342602369401e-05,
+      "loss": 1.9649,
+      "step": 2587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5771464129573756,
+      "learning_rate": 4.632016449211163e-05,
+      "loss": 2.172,
+      "step": 2588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.64391990799126,
+      "learning_rate": 4.6316901629423806e-05,
+      "loss": 2.0208,
+      "step": 2589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6455979550112021,
+      "learning_rate": 4.6313637435834235e-05,
+      "loss": 1.9673,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7881033449711133,
+      "learning_rate": 4.631037191154672e-05,
+      "loss": 1.9972,
+      "step": 2591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6636328167441744,
+      "learning_rate": 4.630710505676514e-05,
+      "loss": 2.1885,
+      "step": 2592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.681236055433829,
+      "learning_rate": 4.6303836871693465e-05,
+      "loss": 1.965,
+      "step": 2593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6256080838698901,
+      "learning_rate": 4.630056735653574e-05,
+      "loss": 2.0288,
+      "step": 2594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.714541968053065,
+      "learning_rate": 4.629729651149608e-05,
+      "loss": 2.1176,
+      "step": 2595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7330178717854247,
+      "learning_rate": 4.629402433677871e-05,
+      "loss": 2.2151,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6947625806918035,
+      "learning_rate": 4.629075083258792e-05,
+      "loss": 1.9903,
+      "step": 2597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6977404663409702,
+      "learning_rate": 4.62874759991281e-05,
+      "loss": 1.9885,
+      "step": 2598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5868231309195191,
+      "learning_rate": 4.628419983660369e-05,
+      "loss": 1.9212,
+      "step": 2599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8153837171888013,
+      "learning_rate": 4.6280922345219255e-05,
+      "loss": 2.2106,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6169069233598656,
+      "learning_rate": 4.627764352517942e-05,
+      "loss": 2.0273,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8114057796826045,
+      "learning_rate": 4.627436337668888e-05,
+      "loss": 2.078,
+      "step": 2602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6880936366867727,
+      "learning_rate": 4.6271081899952435e-05,
+      "loss": 2.0006,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6814061456837147,
+      "learning_rate": 4.626779909517497e-05,
+      "loss": 2.2114,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6313515731882557,
+      "learning_rate": 4.626451496256143e-05,
+      "loss": 1.9582,
+      "step": 2605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7245878825509109,
+      "learning_rate": 4.6261229502316866e-05,
+      "loss": 1.9866,
+      "step": 2606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.648329670544724,
+      "learning_rate": 4.6257942714646405e-05,
+      "loss": 2.2024,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7756371981498712,
+      "learning_rate": 4.625465459975524e-05,
+      "loss": 2.056,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6740421069458721,
+      "learning_rate": 4.625136515784868e-05,
+      "loss": 1.9531,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6511230351410632,
+      "learning_rate": 4.62480743891321e-05,
+      "loss": 1.9944,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6263895442328945,
+      "learning_rate": 4.624478229381093e-05,
+      "loss": 1.9938,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6874119041578052,
+      "learning_rate": 4.624148887209073e-05,
+      "loss": 2.2105,
+      "step": 2612
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6195395378963254,
+      "learning_rate": 4.623819412417713e-05,
+      "loss": 2.0065,
+      "step": 2613
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6165988337872469,
+      "learning_rate": 4.6234898050275805e-05,
+      "loss": 2.0464,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6480559727054603,
+      "learning_rate": 4.623160065059257e-05,
+      "loss": 1.9912,
+      "step": 2615
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6252933703490435,
+      "learning_rate": 4.622830192533328e-05,
+      "loss": 2.1766,
+      "step": 2616
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6042034575515792,
+      "learning_rate": 4.622500187470389e-05,
+      "loss": 1.9972,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7110084457404399,
+      "learning_rate": 4.622170049891045e-05,
+      "loss": 1.9954,
+      "step": 2618
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6202881382255014,
+      "learning_rate": 4.6218397798159054e-05,
+      "loss": 1.9858,
+      "step": 2619
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.664962645004124,
+      "learning_rate": 4.621509377265592e-05,
+      "loss": 2.2143,
+      "step": 2620
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6597752179624428,
+      "learning_rate": 4.6211788422607334e-05,
+      "loss": 1.9589,
+      "step": 2621
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6344383572699662,
+      "learning_rate": 4.620848174821965e-05,
+      "loss": 1.9528,
+      "step": 2622
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7308068826075466,
+      "learning_rate": 4.620517374969933e-05,
+      "loss": 1.9578,
+      "step": 2623
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8427333752346938,
+      "learning_rate": 4.6201864427252896e-05,
+      "loss": 2.1922,
+      "step": 2624
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5974980874181669,
+      "learning_rate": 4.619855378108696e-05,
+      "loss": 1.9527,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.999716199534339,
+      "learning_rate": 4.619524181140823e-05,
+      "loss": 2.0393,
+      "step": 2626
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.650342992780213,
+      "learning_rate": 4.619192851842349e-05,
+      "loss": 1.973,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8062088109528851,
+      "learning_rate": 4.6188613902339587e-05,
+      "loss": 2.1825,
+      "step": 2628
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8563120718683732,
+      "learning_rate": 4.6185297963363475e-05,
+      "loss": 1.9669,
+      "step": 2629
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6138571827653525,
+      "learning_rate": 4.618198070170217e-05,
+      "loss": 1.9532,
+      "step": 2630
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7365087840322205,
+      "learning_rate": 4.61786621175628e-05,
+      "loss": 1.9943,
+      "step": 2631
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6883989549332556,
+      "learning_rate": 4.617534221115255e-05,
+      "loss": 2.1513,
+      "step": 2632
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6900936606222442,
+      "learning_rate": 4.61720209826787e-05,
+      "loss": 2.0413,
+      "step": 2633
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6495303036975949,
+      "learning_rate": 4.616869843234859e-05,
+      "loss": 1.9844,
+      "step": 2634
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7365622207438768,
+      "learning_rate": 4.616537456036969e-05,
+      "loss": 1.9726,
+      "step": 2635
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7641384791071775,
+      "learning_rate": 4.61620493669495e-05,
+      "loss": 2.1947,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6528772197007925,
+      "learning_rate": 4.6158722852295627e-05,
+      "loss": 1.9836,
+      "step": 2637
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7533304123828032,
+      "learning_rate": 4.6155395016615766e-05,
+      "loss": 1.9642,
+      "step": 2638
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7467104639484458,
+      "learning_rate": 4.615206586011769e-05,
+      "loss": 2.0856,
+      "step": 2639
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6372760177405661,
+      "learning_rate": 4.614873538300925e-05,
+      "loss": 2.1544,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6304150273704419,
+      "learning_rate": 4.614540358549837e-05,
+      "loss": 1.9661,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6541947253914094,
+      "learning_rate": 4.614207046779308e-05,
+      "loss": 1.9513,
+      "step": 2642
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5775906541814531,
+      "learning_rate": 4.6138736030101475e-05,
+      "loss": 1.9696,
+      "step": 2643
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6529792868366469,
+      "learning_rate": 4.613540027263175e-05,
+      "loss": 2.1803,
+      "step": 2644
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6877087040045944,
+      "learning_rate": 4.613206319559214e-05,
+      "loss": 2.0543,
+      "step": 2645
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6514369290234705,
+      "learning_rate": 4.612872479919103e-05,
+      "loss": 1.9421,
+      "step": 2646
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7543707099073174,
+      "learning_rate": 4.612538508363682e-05,
+      "loss": 1.9848,
+      "step": 2647
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6986106815333605,
+      "learning_rate": 4.6122044049138045e-05,
+      "loss": 2.1594,
+      "step": 2648
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6788654308420587,
+      "learning_rate": 4.611870169590328e-05,
+      "loss": 2.0143,
+      "step": 2649
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6031561172947046,
+      "learning_rate": 4.611535802414121e-05,
+      "loss": 2.009,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7350969260289836,
+      "learning_rate": 4.61120130340606e-05,
+      "loss": 2.0743,
+      "step": 2651
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7566004503572001,
+      "learning_rate": 4.6108666725870286e-05,
+      "loss": 2.2192,
+      "step": 2652
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6676839095512593,
+      "learning_rate": 4.6105319099779186e-05,
+      "loss": 2.0207,
+      "step": 2653
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6748837361455958,
+      "learning_rate": 4.610197015599632e-05,
+      "loss": 2.003,
+      "step": 2654
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7238881204449563,
+      "learning_rate": 4.6098619894730766e-05,
+      "loss": 1.9342,
+      "step": 2655
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6523191040277376,
+      "learning_rate": 4.6095268316191685e-05,
+      "loss": 2.1813,
+      "step": 2656
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.742330441610128,
+      "learning_rate": 4.6091915420588355e-05,
+      "loss": 2.1066,
+      "step": 2657
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.699416442856155,
+      "learning_rate": 4.6088561208130094e-05,
+      "loss": 1.9849,
+      "step": 2658
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6588847160306627,
+      "learning_rate": 4.6085205679026325e-05,
+      "loss": 1.955,
+      "step": 2659
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6638981429281201,
+      "learning_rate": 4.608184883348654e-05,
+      "loss": 2.1703,
+      "step": 2660
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6753161118399083,
+      "learning_rate": 4.6078490671720324e-05,
+      "loss": 1.9424,
+      "step": 2661
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7448796303868312,
+      "learning_rate": 4.607513119393735e-05,
+      "loss": 1.9945,
+      "step": 2662
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6639803543587954,
+      "learning_rate": 4.607177040034735e-05,
+      "loss": 1.9961,
+      "step": 2663
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5901072399585455,
+      "learning_rate": 4.6068408291160165e-05,
+      "loss": 2.2116,
+      "step": 2664
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7356833094454719,
+      "learning_rate": 4.6065044866585694e-05,
+      "loss": 2.0004,
+      "step": 2665
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7021501499989917,
+      "learning_rate": 4.606168012683394e-05,
+      "loss": 1.9792,
+      "step": 2666
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6148561520798782,
+      "learning_rate": 4.6058314072114966e-05,
+      "loss": 1.9557,
+      "step": 2667
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5675773032041115,
+      "learning_rate": 4.605494670263894e-05,
+      "loss": 2.1266,
+      "step": 2668
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6593238118243776,
+      "learning_rate": 4.6051578018616095e-05,
+      "loss": 1.9912,
+      "step": 2669
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7428701580648603,
+      "learning_rate": 4.604820802025675e-05,
+      "loss": 2.0746,
+      "step": 2670
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6066019308120973,
+      "learning_rate": 4.6044836707771314e-05,
+      "loss": 1.926,
+      "step": 2671
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8349110964232412,
+      "learning_rate": 4.604146408137026e-05,
+      "loss": 2.1856,
+      "step": 2672
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7760398526903683,
+      "learning_rate": 4.6038090141264165e-05,
+      "loss": 1.938,
+      "step": 2673
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.583652163062359,
+      "learning_rate": 4.603471488766368e-05,
+      "loss": 1.9962,
+      "step": 2674
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6732494966439296,
+      "learning_rate": 4.6031338320779534e-05,
+      "loss": 1.9694,
+      "step": 2675
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.592423970933599,
+      "learning_rate": 4.6027960440822533e-05,
+      "loss": 2.2178,
+      "step": 2676
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8374765612741423,
+      "learning_rate": 4.602458124800357e-05,
+      "loss": 1.9702,
+      "step": 2677
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6565341808899915,
+      "learning_rate": 4.6021200742533645e-05,
+      "loss": 2.0096,
+      "step": 2678
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6695837195252338,
+      "learning_rate": 4.601781892462379e-05,
+      "loss": 1.9901,
+      "step": 2679
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7057589195590709,
+      "learning_rate": 4.601443579448516e-05,
+      "loss": 2.1426,
+      "step": 2680
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6282253001598147,
+      "learning_rate": 4.601105135232897e-05,
+      "loss": 1.9709,
+      "step": 2681
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7592042474834655,
+      "learning_rate": 4.600766559836653e-05,
+      "loss": 2.0136,
+      "step": 2682
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6342157574431733,
+      "learning_rate": 4.600427853280922e-05,
+      "loss": 1.9526,
+      "step": 2683
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7069254077056205,
+      "learning_rate": 4.6000890155868524e-05,
+      "loss": 2.1806,
+      "step": 2684
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6246569782481525,
+      "learning_rate": 4.599750046775597e-05,
+      "loss": 1.9486,
+      "step": 2685
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7178039593665737,
+      "learning_rate": 4.599410946868321e-05,
+      "loss": 1.9738,
+      "step": 2686
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6769499236158081,
+      "learning_rate": 4.599071715886195e-05,
+      "loss": 1.9682,
+      "step": 2687
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7933150635822529,
+      "learning_rate": 4.5987323538503984e-05,
+      "loss": 2.1817,
+      "step": 2688
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7538366145808504,
+      "learning_rate": 4.598392860782119e-05,
+      "loss": 2.0383,
+      "step": 2689
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6930574343061014,
+      "learning_rate": 4.598053236702553e-05,
+      "loss": 2.0215,
+      "step": 2690
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7111614023256035,
+      "learning_rate": 4.597713481632905e-05,
+      "loss": 1.9819,
+      "step": 2691
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6052159770218407,
+      "learning_rate": 4.597373595594386e-05,
+      "loss": 2.1928,
+      "step": 2692
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7161440277678321,
+      "learning_rate": 4.5970335786082176e-05,
+      "loss": 1.9881,
+      "step": 2693
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5729690222187727,
+      "learning_rate": 4.596693430695628e-05,
+      "loss": 1.9705,
+      "step": 2694
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6668970882865378,
+      "learning_rate": 4.5963531518778535e-05,
+      "loss": 2.1186,
+      "step": 2695
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6602973681236253,
+      "learning_rate": 4.59601274217614e-05,
+      "loss": 2.179,
+      "step": 2696
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6645224252351648,
+      "learning_rate": 4.5956722016117404e-05,
+      "loss": 2.0442,
+      "step": 2697
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6187834337762769,
+      "learning_rate": 4.5953315302059166e-05,
+      "loss": 1.9885,
+      "step": 2698
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6341351296008236,
+      "learning_rate": 4.594990727979937e-05,
+      "loss": 2.0357,
+      "step": 2699
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6583401768576727,
+      "learning_rate": 4.594649794955079e-05,
+      "loss": 2.1957,
+      "step": 2700
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.698903175071379,
+      "learning_rate": 4.59430873115263e-05,
+      "loss": 2.0707,
+      "step": 2701
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6021695324554158,
+      "learning_rate": 4.593967536593884e-05,
+      "loss": 1.9307,
+      "step": 2702
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7372200389546351,
+      "learning_rate": 4.593626211300142e-05,
+      "loss": 2.0022,
+      "step": 2703
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7038707811265971,
+      "learning_rate": 4.593284755292715e-05,
+      "loss": 2.1875,
+      "step": 2704
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6931745239945363,
+      "learning_rate": 4.5929431685929204e-05,
+      "loss": 1.9846,
+      "step": 2705
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9749747679426978,
+      "learning_rate": 4.592601451222086e-05,
+      "loss": 1.9921,
+      "step": 2706
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6304867022249925,
+      "learning_rate": 4.592259603201546e-05,
+      "loss": 2.0321,
+      "step": 2707
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9620294751423234,
+      "learning_rate": 4.591917624552645e-05,
+      "loss": 2.149,
+      "step": 2708
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7179772686185427,
+      "learning_rate": 4.591575515296732e-05,
+      "loss": 1.9819,
+      "step": 2709
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7208279071564487,
+      "learning_rate": 4.591233275455168e-05,
+      "loss": 1.9119,
+      "step": 2710
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7848438247414634,
+      "learning_rate": 4.590890905049319e-05,
+      "loss": 1.9823,
+      "step": 2711
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7024802442464062,
+      "learning_rate": 4.590548404100561e-05,
+      "loss": 2.201,
+      "step": 2712
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6369270187313578,
+      "learning_rate": 4.5902057726302784e-05,
+      "loss": 2.0848,
+      "step": 2713
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6828478863561541,
+      "learning_rate": 4.5898630106598625e-05,
+      "loss": 1.993,
+      "step": 2714
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7181238008600065,
+      "learning_rate": 4.5895201182107134e-05,
+      "loss": 1.9576,
+      "step": 2715
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6776059441125085,
+      "learning_rate": 4.58917709530424e-05,
+      "loss": 2.118,
+      "step": 2716
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7677641259041942,
+      "learning_rate": 4.588833941961856e-05,
+      "loss": 2.0412,
+      "step": 2717
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7368767351520903,
+      "learning_rate": 4.58849065820499e-05,
+      "loss": 2.0143,
+      "step": 2718
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7477193378112147,
+      "learning_rate": 4.588147244055072e-05,
+      "loss": 1.9962,
+      "step": 2719
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6996894669502343,
+      "learning_rate": 4.587803699533543e-05,
+      "loss": 2.0909,
+      "step": 2720
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.696941883265145,
+      "learning_rate": 4.587460024661852e-05,
+      "loss": 2.2067,
+      "step": 2721
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6531248811442996,
+      "learning_rate": 4.587116219461456e-05,
+      "loss": 1.9893,
+      "step": 2722
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6570546490300734,
+      "learning_rate": 4.5867722839538205e-05,
+      "loss": 1.9928,
+      "step": 2723
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6718208182122217,
+      "learning_rate": 4.586428218160419e-05,
+      "loss": 2.1419,
+      "step": 2724
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6721359786965663,
+      "learning_rate": 4.586084022102732e-05,
+      "loss": 1.9408,
+      "step": 2725
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6473584802823314,
+      "learning_rate": 4.58573969580225e-05,
+      "loss": 2.0356,
+      "step": 2726
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6230859480633122,
+      "learning_rate": 4.58539523928047e-05,
+      "loss": 1.985,
+      "step": 2727
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8196654156442967,
+      "learning_rate": 4.5850506525588986e-05,
+      "loss": 2.1846,
+      "step": 2728
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6580276847375983,
+      "learning_rate": 4.5847059356590494e-05,
+      "loss": 2.0321,
+      "step": 2729
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7960055602568584,
+      "learning_rate": 4.5843610886024445e-05,
+      "loss": 2.0207,
+      "step": 2730
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6257598563119594,
+      "learning_rate": 4.584016111410614e-05,
+      "loss": 1.9627,
+      "step": 2731
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6786767927403474,
+      "learning_rate": 4.583671004105096e-05,
+      "loss": 2.065,
+      "step": 2732
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7355666708735067,
+      "learning_rate": 4.583325766707437e-05,
+      "loss": 2.1655,
+      "step": 2733
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8065959056684828,
+      "learning_rate": 4.582980399239193e-05,
+      "loss": 1.9871,
+      "step": 2734
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6523713117534081,
+      "learning_rate": 4.5826349017219244e-05,
+      "loss": 1.994,
+      "step": 2735
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7643096877117508,
+      "learning_rate": 4.582289274177204e-05,
+      "loss": 2.1942,
+      "step": 2736
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6550475104022917,
+      "learning_rate": 4.581943516626609e-05,
+      "loss": 2.02,
+      "step": 2737
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6967186393956918,
+      "learning_rate": 4.5815976290917285e-05,
+      "loss": 2.0852,
+      "step": 2738
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5945616636292965,
+      "learning_rate": 4.5812516115941564e-05,
+      "loss": 2.0147,
+      "step": 2739
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6756966445469972,
+      "learning_rate": 4.580905464155496e-05,
+      "loss": 2.1629,
+      "step": 2740
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6396973306899563,
+      "learning_rate": 4.580559186797358e-05,
+      "loss": 1.9821,
+      "step": 2741
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6490476811755255,
+      "learning_rate": 4.580212779541364e-05,
+      "loss": 2.0194,
+      "step": 2742
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6293092174399644,
+      "learning_rate": 4.57986624240914e-05,
+      "loss": 1.9912,
+      "step": 2743
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5853137351391694,
+      "learning_rate": 4.5795195754223227e-05,
+      "loss": 2.0212,
+      "step": 2744
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6222177440489787,
+      "learning_rate": 4.579172778602555e-05,
+      "loss": 2.1442,
+      "step": 2745
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6778188359882265,
+      "learning_rate": 4.578825851971489e-05,
+      "loss": 1.9919,
+      "step": 2746
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.587600823091895,
+      "learning_rate": 4.5784787955507856e-05,
+      "loss": 1.9981,
+      "step": 2747
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7124436175571408,
+      "learning_rate": 4.578131609362112e-05,
+      "loss": 2.1332,
+      "step": 2748
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.594628326436112,
+      "learning_rate": 4.5777842934271455e-05,
+      "loss": 1.9561,
+      "step": 2749
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6627225645232516,
+      "learning_rate": 4.577436847767569e-05,
+      "loss": 1.9689,
+      "step": 2750
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7631883599138035,
+      "learning_rate": 4.5770892724050754e-05,
+      "loss": 2.0177,
+      "step": 2751
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6207930598184283,
+      "learning_rate": 4.5767415673613664e-05,
+      "loss": 1.9675,
+      "step": 2752
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.789268479224515,
+      "learning_rate": 4.57639373265815e-05,
+      "loss": 2.2035,
+      "step": 2753
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6564910549515972,
+      "learning_rate": 4.576045768317143e-05,
+      "loss": 1.9582,
+      "step": 2754
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6885668117653446,
+      "learning_rate": 4.5756976743600694e-05,
+      "loss": 2.0111,
+      "step": 2755
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.726912666993224,
+      "learning_rate": 4.5753494508086634e-05,
+      "loss": 2.1499,
+      "step": 2756
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6295799172352778,
+      "learning_rate": 4.575001097684665e-05,
+      "loss": 2.0606,
+      "step": 2757
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6244854012366476,
+      "learning_rate": 4.5746526150098234e-05,
+      "loss": 1.9885,
+      "step": 2758
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7174359452624782,
+      "learning_rate": 4.574304002805897e-05,
+      "loss": 1.9835,
+      "step": 2759
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6227259967417186,
+      "learning_rate": 4.5739552610946495e-05,
+      "loss": 2.1767,
+      "step": 2760
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5984088042514014,
+      "learning_rate": 4.5736063898978553e-05,
+      "loss": 2.002,
+      "step": 2761
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6806538349505173,
+      "learning_rate": 4.573257389237296e-05,
+      "loss": 1.9991,
+      "step": 2762
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5726933289209153,
+      "learning_rate": 4.57290825913476e-05,
+      "loss": 2.0095,
+      "step": 2763
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7360269692392019,
+      "learning_rate": 4.572558999612047e-05,
+      "loss": 1.9736,
+      "step": 2764
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.645296711900151,
+      "learning_rate": 4.5722096106909595e-05,
+      "loss": 2.1521,
+      "step": 2765
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7201120969061646,
+      "learning_rate": 4.5718600923933144e-05,
+      "loss": 1.993,
+      "step": 2766
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6444093676344362,
+      "learning_rate": 4.571510444740932e-05,
+      "loss": 2.0214,
+      "step": 2767
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5893068827066283,
+      "learning_rate": 4.571160667755643e-05,
+      "loss": 2.1302,
+      "step": 2768
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6783447357780432,
+      "learning_rate": 4.5708107614592845e-05,
+      "loss": 2.0205,
+      "step": 2769
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6773233340650222,
+      "learning_rate": 4.570460725873703e-05,
+      "loss": 1.992,
+      "step": 2770
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6207222696489187,
+      "learning_rate": 4.570110561020753e-05,
+      "loss": 1.9541,
+      "step": 2771
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6739335409830205,
+      "learning_rate": 4.5697602669222964e-05,
+      "loss": 1.9661,
+      "step": 2772
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7606699792598559,
+      "learning_rate": 4.569409843600203e-05,
+      "loss": 2.1788,
+      "step": 2773
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6711149755631587,
+      "learning_rate": 4.569059291076352e-05,
+      "loss": 1.9947,
+      "step": 2774
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6465873488697467,
+      "learning_rate": 4.56870860937263e-05,
+      "loss": 2.0585,
+      "step": 2775
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6411975458999427,
+      "learning_rate": 4.568357798510931e-05,
+      "loss": 2.0042,
+      "step": 2776
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6888143430209183,
+      "learning_rate": 4.568006858513157e-05,
+      "loss": 2.1754,
+      "step": 2777
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8714153148177743,
+      "learning_rate": 4.56765578940122e-05,
+      "loss": 1.9843,
+      "step": 2778
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8656697630216207,
+      "learning_rate": 4.567304591197037e-05,
+      "loss": 1.9876,
+      "step": 2779
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7065480436484274,
+      "learning_rate": 4.566953263922537e-05,
+      "loss": 2.2021,
+      "step": 2780
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9803238992281242,
+      "learning_rate": 4.5666018075996516e-05,
+      "loss": 1.9878,
+      "step": 2781
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6178701951645434,
+      "learning_rate": 4.566250222250327e-05,
+      "loss": 1.9952,
+      "step": 2782
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8836848567588546,
+      "learning_rate": 4.5658985078965124e-05,
+      "loss": 1.9502,
+      "step": 2783
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8274683172290525,
+      "learning_rate": 4.5655466645601666e-05,
+      "loss": 2.013,
+      "step": 2784
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8645014730633772,
+      "learning_rate": 4.565194692263257e-05,
+      "loss": 2.1835,
+      "step": 2785
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2188305478409456,
+      "learning_rate": 4.5648425910277594e-05,
+      "loss": 2.0041,
+      "step": 2786
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6473009802391637,
+      "learning_rate": 4.5644903608756564e-05,
+      "loss": 1.999,
+      "step": 2787
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8297091501769635,
+      "learning_rate": 4.5641380018289384e-05,
+      "loss": 2.0303,
+      "step": 2788
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8988633567086005,
+      "learning_rate": 4.563785513909605e-05,
+      "loss": 2.2072,
+      "step": 2789
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7429093022412904,
+      "learning_rate": 4.563432897139664e-05,
+      "loss": 1.9594,
+      "step": 2790
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7786389766083004,
+      "learning_rate": 4.563080151541131e-05,
+      "loss": 1.9557,
+      "step": 2791
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6080534146007245,
+      "learning_rate": 4.5627272771360284e-05,
+      "loss": 2.1346,
+      "step": 2792
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7987251936834332,
+      "learning_rate": 4.562374273946388e-05,
+      "loss": 2.0202,
+      "step": 2793
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7283780239896972,
+      "learning_rate": 4.562021141994249e-05,
+      "loss": 2.0411,
+      "step": 2794
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6071314653849001,
+      "learning_rate": 4.561667881301659e-05,
+      "loss": 1.9707,
+      "step": 2795
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9252676566816939,
+      "learning_rate": 4.561314491890674e-05,
+      "loss": 1.9586,
+      "step": 2796
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6070199284302388,
+      "learning_rate": 4.560960973783357e-05,
+      "loss": 2.1933,
+      "step": 2797
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7924570906830115,
+      "learning_rate": 4.56060732700178e-05,
+      "loss": 2.0251,
+      "step": 2798
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6245592793203827,
+      "learning_rate": 4.560253551568022e-05,
+      "loss": 1.9828,
+      "step": 2799
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6175846472322825,
+      "learning_rate": 4.5598996475041716e-05,
+      "loss": 2.0201,
+      "step": 2800
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7242371757675056,
+      "learning_rate": 4.5595456148323235e-05,
+      "loss": 2.195,
+      "step": 2801
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6500032961745279,
+      "learning_rate": 4.559191453574582e-05,
+      "loss": 1.988,
+      "step": 2802
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6826942890717906,
+      "learning_rate": 4.558837163753059e-05,
+      "loss": 2.0091,
+      "step": 2803
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6229778172735722,
+      "learning_rate": 4.558482745389874e-05,
+      "loss": 1.9857,
+      "step": 2804
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.63313069533344,
+      "learning_rate": 4.558128198507153e-05,
+      "loss": 2.134,
+      "step": 2805
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5999104918410882,
+      "learning_rate": 4.557773523127036e-05,
+      "loss": 2.0455,
+      "step": 2806
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6294359560155657,
+      "learning_rate": 4.557418719271663e-05,
+      "loss": 1.9403,
+      "step": 2807
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5755913347329538,
+      "learning_rate": 4.5570637869631876e-05,
+      "loss": 1.9286,
+      "step": 2808
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.665712105095883,
+      "learning_rate": 4.55670872622377e-05,
+      "loss": 2.1782,
+      "step": 2809
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6305847707072353,
+      "learning_rate": 4.556353537075576e-05,
+      "loss": 1.9615,
+      "step": 2810
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6941507505291515,
+      "learning_rate": 4.555998219540783e-05,
+      "loss": 1.9509,
+      "step": 2811
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7167520786121081,
+      "learning_rate": 4.555642773641576e-05,
+      "loss": 2.2022,
+      "step": 2812
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8126009864774975,
+      "learning_rate": 4.5552871994001454e-05,
+      "loss": 2.0776,
+      "step": 2813
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6836392065689031,
+      "learning_rate": 4.554931496838692e-05,
+      "loss": 1.9893,
+      "step": 2814
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7756634706788672,
+      "learning_rate": 4.5545756659794234e-05,
+      "loss": 2.0069,
+      "step": 2815
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.60678672741813,
+      "learning_rate": 4.554219706844555e-05,
+      "loss": 2.016,
+      "step": 2816
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7056114940862528,
+      "learning_rate": 4.5538636194563126e-05,
+      "loss": 2.1671,
+      "step": 2817
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6152703262317227,
+      "learning_rate": 4.553507403836925e-05,
+      "loss": 2.0159,
+      "step": 2818
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6795958586602747,
+      "learning_rate": 4.5531510600086356e-05,
+      "loss": 1.9866,
+      "step": 2819
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6154295809613098,
+      "learning_rate": 4.552794587993691e-05,
+      "loss": 1.9898,
+      "step": 2820
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7639723736967522,
+      "learning_rate": 4.552437987814346e-05,
+      "loss": 2.202,
+      "step": 2821
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6465414274009734,
+      "learning_rate": 4.552081259492867e-05,
+      "loss": 2.0013,
+      "step": 2822
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6405697302521749,
+      "learning_rate": 4.5517244030515235e-05,
+      "loss": 1.9681,
+      "step": 2823
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7215042282520571,
+      "learning_rate": 4.5513674185125976e-05,
+      "loss": 2.1878,
+      "step": 2824
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6246500836520427,
+      "learning_rate": 4.551010305898377e-05,
+      "loss": 2.0427,
+      "step": 2825
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7177701784596208,
+      "learning_rate": 4.550653065231156e-05,
+      "loss": 1.9767,
+      "step": 2826
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6100182645441249,
+      "learning_rate": 4.550295696533241e-05,
+      "loss": 1.9884,
+      "step": 2827
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6253906657069692,
+      "learning_rate": 4.549938199826943e-05,
+      "loss": 1.9576,
+      "step": 2828
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6574391423232303,
+      "learning_rate": 4.549580575134581e-05,
+      "loss": 2.1227,
+      "step": 2829
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6410611009248435,
+      "learning_rate": 4.549222822478484e-05,
+      "loss": 1.9439,
+      "step": 2830
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6364492611243694,
+      "learning_rate": 4.548864941880988e-05,
+      "loss": 2.0551,
+      "step": 2831
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6567136142944537,
+      "learning_rate": 4.5485069333644365e-05,
+      "loss": 2.0015,
+      "step": 2832
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7503929174713984,
+      "learning_rate": 4.548148796951182e-05,
+      "loss": 2.2002,
+      "step": 2833
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5530971981044319,
+      "learning_rate": 4.5477905326635834e-05,
+      "loss": 1.9807,
+      "step": 2834
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6798052421349903,
+      "learning_rate": 4.54743214052401e-05,
+      "loss": 1.994,
+      "step": 2835
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6800876655630853,
+      "learning_rate": 4.547073620554837e-05,
+      "loss": 1.9732,
+      "step": 2836
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5855190065165854,
+      "learning_rate": 4.5467149727784475e-05,
+      "loss": 2.2371,
+      "step": 2837
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7001853668201681,
+      "learning_rate": 4.546356197217235e-05,
+      "loss": 2.005,
+      "step": 2838
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6267456078592311,
+      "learning_rate": 4.545997293893598e-05,
+      "loss": 1.9511,
+      "step": 2839
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6182001894960636,
+      "learning_rate": 4.545638262829945e-05,
+      "loss": 1.9601,
+      "step": 2840
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6826371254965823,
+      "learning_rate": 4.5452791040486916e-05,
+      "loss": 2.1518,
+      "step": 2841
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7307901848650923,
+      "learning_rate": 4.5449198175722616e-05,
+      "loss": 1.9406,
+      "step": 2842
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5820433448026973,
+      "learning_rate": 4.544560403423086e-05,
+      "loss": 2.0016,
+      "step": 2843
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6314021283020453,
+      "learning_rate": 4.544200861623606e-05,
+      "loss": 2.0596,
+      "step": 2844
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6428405994029349,
+      "learning_rate": 4.543841192196267e-05,
+      "loss": 2.1636,
+      "step": 2845
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6435160655798973,
+      "learning_rate": 4.543481395163528e-05,
+      "loss": 2.0009,
+      "step": 2846
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6594843761958827,
+      "learning_rate": 4.54312147054785e-05,
+      "loss": 1.9885,
+      "step": 2847
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5994132154448734,
+      "learning_rate": 4.542761418371705e-05,
+      "loss": 1.9679,
+      "step": 2848
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.80087867596022,
+      "learning_rate": 4.5424012386575724e-05,
+      "loss": 2.1821,
+      "step": 2849
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6222519603267266,
+      "learning_rate": 4.542040931427941e-05,
+      "loss": 2.0123,
+      "step": 2850
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8102587140396517,
+      "learning_rate": 4.541680496705304e-05,
+      "loss": 2.0074,
+      "step": 2851
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.642088701245558,
+      "learning_rate": 4.541319934512167e-05,
+      "loss": 1.956,
+      "step": 2852
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8197684316205066,
+      "learning_rate": 4.5409592448710416e-05,
+      "loss": 2.1598,
+      "step": 2853
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6182804461324081,
+      "learning_rate": 4.540598427804444e-05,
+      "loss": 1.9841,
+      "step": 2854
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7758159466236181,
+      "learning_rate": 4.540237483334905e-05,
+      "loss": 1.978,
+      "step": 2855
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7579463687232904,
+      "learning_rate": 4.539876411484958e-05,
+      "loss": 2.0344,
+      "step": 2856
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8260274656854518,
+      "learning_rate": 4.539515212277147e-05,
+      "loss": 2.1892,
+      "step": 2857
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5942964706685796,
+      "learning_rate": 4.539153885734022e-05,
+      "loss": 1.9459,
+      "step": 2858
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7282235040093462,
+      "learning_rate": 4.5387924318781424e-05,
+      "loss": 1.9631,
+      "step": 2859
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6374913565744571,
+      "learning_rate": 4.538430850732077e-05,
+      "loss": 1.9775,
+      "step": 2860
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7534812306001003,
+      "learning_rate": 4.538069142318398e-05,
+      "loss": 2.1807,
+      "step": 2861
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.590841649452214,
+      "learning_rate": 4.537707306659692e-05,
+      "loss": 2.0586,
+      "step": 2862
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8078345745360919,
+      "learning_rate": 4.537345343778546e-05,
+      "loss": 1.9146,
+      "step": 2863
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6232718261513431,
+      "learning_rate": 4.536983253697561e-05,
+      "loss": 2.0345,
+      "step": 2864
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7956778886829186,
+      "learning_rate": 4.536621036439344e-05,
+      "loss": 2.1656,
+      "step": 2865
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6220027138141474,
+      "learning_rate": 4.5362586920265085e-05,
+      "loss": 1.9612,
+      "step": 2866
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6011225600815533,
+      "learning_rate": 4.5358962204816775e-05,
+      "loss": 2.0283,
+      "step": 2867
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6305369247695293,
+      "learning_rate": 4.535533621827482e-05,
+      "loss": 2.0083,
+      "step": 2868
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.7610780113415268,
+      "learning_rate": 4.53517089608656e-05,
+      "loss": 2.2337,
+      "step": 2869
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6052291106605415,
+      "learning_rate": 4.5348080432815587e-05,
+      "loss": 1.9581,
+      "step": 2870
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 20.166794132613884,
+      "learning_rate": 4.534445063435132e-05,
+      "loss": 1.9625,
+      "step": 2871
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.43504401389042,
+      "learning_rate": 4.534081956569942e-05,
+      "loss": 2.7738,
+      "step": 2872
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 41.96701761382132,
+      "learning_rate": 4.53371872270866e-05,
+      "loss": 4.0904,
+      "step": 2873
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 43.03228087610891,
+      "learning_rate": 4.533355361873962e-05,
+      "loss": 3.6379,
+      "step": 2874
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 29.851978052510876,
+      "learning_rate": 4.532991874088537e-05,
+      "loss": 3.0166,
+      "step": 2875
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 30.87212281375283,
+      "learning_rate": 4.532628259375077e-05,
+      "loss": 3.7492,
+      "step": 2876
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 95.95200393937546,
+      "learning_rate": 4.532264517756284e-05,
+      "loss": 6.0122,
+      "step": 2877
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 12.859550550670065,
+      "learning_rate": 4.531900649254869e-05,
+      "loss": 3.0449,
+      "step": 2878
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 13.925404649856581,
+      "learning_rate": 4.531536653893549e-05,
+      "loss": 2.9083,
+      "step": 2879
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 22.55321120338719,
+      "learning_rate": 4.53117253169505e-05,
+      "loss": 2.8096,
+      "step": 2880
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 18.911858792571532,
+      "learning_rate": 4.530808282682106e-05,
+      "loss": 2.761,
+      "step": 2881
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 5.573116485632248,
+      "learning_rate": 4.5304439068774575e-05,
+      "loss": 2.4804,
+      "step": 2882
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.758761567269109,
+      "learning_rate": 4.5300794043038546e-05,
+      "loss": 2.3944,
+      "step": 2883
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.933964688501277,
+      "learning_rate": 4.529714774984055e-05,
+      "loss": 2.2645,
+      "step": 2884
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.7089508265074214,
+      "learning_rate": 4.529350018940824e-05,
+      "loss": 2.5618,
+      "step": 2885
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.1554988423120256,
+      "learning_rate": 4.528985136196934e-05,
+      "loss": 2.2297,
+      "step": 2886
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.155356489064996,
+      "learning_rate": 4.528620126775167e-05,
+      "loss": 2.3073,
+      "step": 2887
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 6.342922549592311,
+      "learning_rate": 4.528254990698311e-05,
+      "loss": 2.2671,
+      "step": 2888
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 4.250761253886918,
+      "learning_rate": 4.5278897279891646e-05,
+      "loss": 2.4565,
+      "step": 2889
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.230708315515155,
+      "learning_rate": 4.527524338670531e-05,
+      "loss": 2.2688,
+      "step": 2890
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9813439467396243,
+      "learning_rate": 4.5271588227652243e-05,
+      "loss": 2.2126,
+      "step": 2891
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.395299301826921,
+      "learning_rate": 4.5267931802960636e-05,
+      "loss": 2.1905,
+      "step": 2892
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4766864393796755,
+      "learning_rate": 4.5264274112858794e-05,
+      "loss": 2.3974,
+      "step": 2893
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.9217125126653207,
+      "learning_rate": 4.526061515757507e-05,
+      "loss": 2.1368,
+      "step": 2894
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0962335322072025,
+      "learning_rate": 4.5256954937337906e-05,
+      "loss": 2.1603,
+      "step": 2895
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9636543407468255,
+      "learning_rate": 4.5253293452375825e-05,
+      "loss": 2.1114,
+      "step": 2896
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.800738884958415,
+      "learning_rate": 4.524963070291744e-05,
+      "loss": 2.3236,
+      "step": 2897
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4290960109962267,
+      "learning_rate": 4.524596668919141e-05,
+      "loss": 2.1552,
+      "step": 2898
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1070494786359306,
+      "learning_rate": 4.5242301411426516e-05,
+      "loss": 2.0799,
+      "step": 2899
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6719305712665202,
+      "learning_rate": 4.5238634869851585e-05,
+      "loss": 2.1752,
+      "step": 2900
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7126071967751997,
+      "learning_rate": 4.523496706469554e-05,
+      "loss": 2.3153,
+      "step": 2901
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8126390229111553,
+      "learning_rate": 4.523129799618737e-05,
+      "loss": 2.1053,
+      "step": 2902
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6622483201105618,
+      "learning_rate": 4.522762766455615e-05,
+      "loss": 2.0884,
+      "step": 2903
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.1386163835490424,
+      "learning_rate": 4.522395607003105e-05,
+      "loss": 2.0542,
+      "step": 2904
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7274368793363496,
+      "learning_rate": 4.522028321284128e-05,
+      "loss": 2.2581,
+      "step": 2905
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.3954679378161785,
+      "learning_rate": 4.521660909321617e-05,
+      "loss": 2.1787,
+      "step": 2906
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6702922530941131,
+      "learning_rate": 4.521293371138509e-05,
+      "loss": 2.0398,
+      "step": 2907
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6784667388469936,
+      "learning_rate": 4.520925706757753e-05,
+      "loss": 2.0598,
+      "step": 2908
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6977581545666485,
+      "learning_rate": 4.5205579162023026e-05,
+      "loss": 2.2989,
+      "step": 2909
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.2077387236904,
+      "learning_rate": 4.5201899994951214e-05,
+      "loss": 2.085,
+      "step": 2910
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7423388495954832,
+      "learning_rate": 4.519821956659179e-05,
+      "loss": 2.1073,
+      "step": 2911
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5923475715394003,
+      "learning_rate": 4.519453787717455e-05,
+      "loss": 2.1546,
+      "step": 2912
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5899555901537639,
+      "learning_rate": 4.519085492692933e-05,
+      "loss": 2.2955,
+      "step": 2913
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5907632278075337,
+      "learning_rate": 4.518717071608611e-05,
+      "loss": 1.99,
+      "step": 2914
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5796207448440479,
+      "learning_rate": 4.518348524487488e-05,
+      "loss": 2.0275,
+      "step": 2915
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.653544279356977,
+      "learning_rate": 4.517979851352575e-05,
+      "loss": 2.0724,
+      "step": 2916
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9798715934058801,
+      "learning_rate": 4.517611052226891e-05,
+      "loss": 2.2267,
+      "step": 2917
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5531928906800047,
+      "learning_rate": 4.5172421271334596e-05,
+      "loss": 2.078,
+      "step": 2918
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6070879817857471,
+      "learning_rate": 4.516873076095315e-05,
+      "loss": 2.0652,
+      "step": 2919
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8722210199743303,
+      "learning_rate": 4.516503899135499e-05,
+      "loss": 2.0582,
+      "step": 2920
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6066966263477893,
+      "learning_rate": 4.516134596277061e-05,
+      "loss": 2.215,
+      "step": 2921
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5451661407756073,
+      "learning_rate": 4.5157651675430566e-05,
+      "loss": 2.0542,
+      "step": 2922
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.561401151559369,
+      "learning_rate": 4.515395612956552e-05,
+      "loss": 2.0164,
+      "step": 2923
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.095686439886581,
+      "learning_rate": 4.515025932540621e-05,
+      "loss": 2.0789,
+      "step": 2924
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5867301536369358,
+      "learning_rate": 4.5146561263183415e-05,
+      "loss": 2.2568,
+      "step": 2925
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5198274109784404,
+      "learning_rate": 4.5142861943128054e-05,
+      "loss": 2.0863,
+      "step": 2926
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6159994081806842,
+      "learning_rate": 4.5139161365471064e-05,
+      "loss": 2.0759,
+      "step": 2927
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5133054358121572,
+      "learning_rate": 4.513545953044349e-05,
+      "loss": 2.046,
+      "step": 2928
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5555267937876979,
+      "learning_rate": 4.513175643827647e-05,
+      "loss": 2.2429,
+      "step": 2929
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5531950471914892,
+      "learning_rate": 4.512805208920118e-05,
+      "loss": 2.0411,
+      "step": 2930
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5415491042252507,
+      "learning_rate": 4.5124346483448923e-05,
+      "loss": 2.0981,
+      "step": 2931
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5330258970161332,
+      "learning_rate": 4.512063962125104e-05,
+      "loss": 2.0328,
+      "step": 2932
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6063221214067308,
+      "learning_rate": 4.5116931502838954e-05,
+      "loss": 2.1844,
+      "step": 2933
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5297301841530939,
+      "learning_rate": 4.511322212844421e-05,
+      "loss": 2.0133,
+      "step": 2934
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6498937401538403,
+      "learning_rate": 4.5109511498298365e-05,
+      "loss": 2.0899,
+      "step": 2935
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5772001406626477,
+      "learning_rate": 4.5105799612633115e-05,
+      "loss": 2.0273,
+      "step": 2936
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6162029377259111,
+      "learning_rate": 4.51020864716802e-05,
+      "loss": 2.0866,
+      "step": 2937
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5330025861520211,
+      "learning_rate": 4.509837207567144e-05,
+      "loss": 2.2516,
+      "step": 2938
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5806761849143963,
+      "learning_rate": 4.5094656424838746e-05,
+      "loss": 2.0366,
+      "step": 2939
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5327825711569419,
+      "learning_rate": 4.509093951941411e-05,
+      "loss": 2.0163,
+      "step": 2940
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5273173302245768,
+      "learning_rate": 4.508722135962957e-05,
+      "loss": 2.2318,
+      "step": 2941
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6483279268246621,
+      "learning_rate": 4.5083501945717285e-05,
+      "loss": 2.0135,
+      "step": 2942
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.523400570331478,
+      "learning_rate": 4.507978127790947e-05,
+      "loss": 2.0888,
+      "step": 2943
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6060362394075197,
+      "learning_rate": 4.507605935643842e-05,
+      "loss": 2.0282,
+      "step": 2944
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6482317093725517,
+      "learning_rate": 4.507233618153651e-05,
+      "loss": 2.2234,
+      "step": 2945
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5557428825438474,
+      "learning_rate": 4.5068611753436194e-05,
+      "loss": 2.0408,
+      "step": 2946
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6543333379952416,
+      "learning_rate": 4.506488607236999e-05,
+      "loss": 1.9892,
+      "step": 2947
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5513633603021613,
+      "learning_rate": 4.506115913857054e-05,
+      "loss": 1.9946,
+      "step": 2948
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.570053182848793,
+      "learning_rate": 4.50574309522705e-05,
+      "loss": 2.0692,
+      "step": 2949
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5295210194236916,
+      "learning_rate": 4.505370151370265e-05,
+      "loss": 2.244,
+      "step": 2950
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5539428778005544,
+      "learning_rate": 4.504997082309983e-05,
+      "loss": 2.0081,
+      "step": 2951
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5551310281167554,
+      "learning_rate": 4.504623888069497e-05,
+      "loss": 2.0688,
+      "step": 2952
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5666713688531781,
+      "learning_rate": 4.504250568672106e-05,
+      "loss": 2.2231,
+      "step": 2953
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5435564900447342,
+      "learning_rate": 4.503877124141118e-05,
+      "loss": 2.0005,
+      "step": 2954
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9176725209323147,
+      "learning_rate": 4.5035035544998495e-05,
+      "loss": 2.0763,
+      "step": 2955
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5211079552103035,
+      "learning_rate": 4.503129859771623e-05,
+      "loss": 2.0328,
+      "step": 2956
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7508112094451146,
+      "learning_rate": 4.502756039979771e-05,
+      "loss": 2.2406,
+      "step": 2957
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5498865043891005,
+      "learning_rate": 4.502382095147631e-05,
+      "loss": 2.0041,
+      "step": 2958
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5929272534292733,
+      "learning_rate": 4.502008025298552e-05,
+      "loss": 2.0225,
+      "step": 2959
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.537894786026507,
+      "learning_rate": 4.501633830455887e-05,
+      "loss": 2.0411,
+      "step": 2960
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5770085207610334,
+      "learning_rate": 4.501259510642999e-05,
+      "loss": 2.2783,
+      "step": 2961
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.4964370893046266,
+      "learning_rate": 4.500885065883258e-05,
+      "loss": 2.1108,
+      "step": 2962
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5565558101560913,
+      "learning_rate": 4.5005104962000436e-05,
+      "loss": 1.9829,
+      "step": 2963
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5350522070567845,
+      "learning_rate": 4.50013580161674e-05,
+      "loss": 2.0104,
+      "step": 2964
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5002452148888681,
+      "learning_rate": 4.4997609821567405e-05,
+      "loss": 2.1977,
+      "step": 2965
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5190043063126969,
+      "learning_rate": 4.499386037843448e-05,
+      "loss": 2.036,
+      "step": 2966
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5389545859104026,
+      "learning_rate": 4.499010968700272e-05,
+      "loss": 2.0363,
+      "step": 2967
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.48623826020737193,
+      "learning_rate": 4.4986357747506294e-05,
+      "loss": 2.0875,
+      "step": 2968
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8390979814475955,
+      "learning_rate": 4.498260456017944e-05,
+      "loss": 1.9898,
+      "step": 2969
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5146167482196662,
+      "learning_rate": 4.497885012525651e-05,
+      "loss": 2.1814,
+      "step": 2970
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5450492026554314,
+      "learning_rate": 4.497509444297188e-05,
+      "loss": 2.0319,
+      "step": 2971
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5573433855941269,
+      "learning_rate": 4.4971337513560035e-05,
+      "loss": 2.005,
+      "step": 2972
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5267879523033264,
+      "learning_rate": 4.496757933725555e-05,
+      "loss": 2.2428,
+      "step": 2973
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5070948517884898,
+      "learning_rate": 4.4963819914293065e-05,
+      "loss": 2.0826,
+      "step": 2974
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5446417774020413,
+      "learning_rate": 4.496005924490728e-05,
+      "loss": 2.037,
+      "step": 2975
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5307354551005237,
+      "learning_rate": 4.495629732933301e-05,
+      "loss": 1.9974,
+      "step": 2976
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6865498864630941,
+      "learning_rate": 4.495253416780511e-05,
+      "loss": 2.1938,
+      "step": 2977
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.542953294256001,
+      "learning_rate": 4.494876976055853e-05,
+      "loss": 2.0188,
+      "step": 2978
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5551509845664487,
+      "learning_rate": 4.4945004107828306e-05,
+      "loss": 1.9786,
+      "step": 2979
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5064097033216102,
+      "learning_rate": 4.4941237209849543e-05,
+      "loss": 2.0564,
+      "step": 2980
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5435959713107112,
+      "learning_rate": 4.4937469066857405e-05,
+      "loss": 2.0109,
+      "step": 2981
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5859318797318583,
+      "learning_rate": 4.493369967908719e-05,
+      "loss": 2.1615,
+      "step": 2982
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6323865969125004,
+      "learning_rate": 4.49299290467742e-05,
+      "loss": 2.0481,
+      "step": 2983
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.551989575536368,
+      "learning_rate": 4.492615717015387e-05,
+      "loss": 1.9887,
+      "step": 2984
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.525058140906379,
+      "learning_rate": 4.492238404946169e-05,
+      "loss": 2.2161,
+      "step": 2985
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2797184377603426,
+      "learning_rate": 4.491860968493323e-05,
+      "loss": 2.0796,
+      "step": 2986
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6315811335961712,
+      "learning_rate": 4.4914834076804135e-05,
+      "loss": 2.0358,
+      "step": 2987
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5695245497296412,
+      "learning_rate": 4.491105722531014e-05,
+      "loss": 2.022,
+      "step": 2988
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5533580537009234,
+      "learning_rate": 4.4907279130687045e-05,
+      "loss": 1.9987,
+      "step": 2989
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6465884235430764,
+      "learning_rate": 4.490349979317073e-05,
+      "loss": 2.1776,
+      "step": 2990
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5298594952994982,
+      "learning_rate": 4.4899719212997167e-05,
+      "loss": 2.0113,
+      "step": 2991
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6758067072269289,
+      "learning_rate": 4.489593739040238e-05,
+      "loss": 2.0028,
+      "step": 2992
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.541197338964297,
+      "learning_rate": 4.489215432562248e-05,
+      "loss": 2.1124,
+      "step": 2993
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5292519909755773,
+      "learning_rate": 4.4888370018893666e-05,
+      "loss": 2.2642,
+      "step": 2994
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5542301378424841,
+      "learning_rate": 4.488458447045222e-05,
+      "loss": 2.0293,
+      "step": 2995
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.59741399901344,
+      "learning_rate": 4.488079768053447e-05,
+      "loss": 2.0269,
+      "step": 2996
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.549636190479805,
+      "learning_rate": 4.487700964937684e-05,
+      "loss": 2.1843,
+      "step": 2997
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5422848003070199,
+      "learning_rate": 4.487322037721586e-05,
+      "loss": 2.0089,
+      "step": 2998
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5463208575157817,
+      "learning_rate": 4.486942986428808e-05,
+      "loss": 2.0904,
+      "step": 2999
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6459195025201658,
+      "learning_rate": 4.4865638110830165e-05,
+      "loss": 2.0105,
+      "step": 3000
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5776394486813085,
+      "learning_rate": 4.486184511707886e-05,
+      "loss": 2.004,
+      "step": 3001
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4757109906581949,
+      "learning_rate": 4.485805088327096e-05,
+      "loss": 2.2294,
+      "step": 3002
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6572543101453379,
+      "learning_rate": 4.4854255409643374e-05,
+      "loss": 1.9792,
+      "step": 3003
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6001542724307072,
+      "learning_rate": 4.4850458696433056e-05,
+      "loss": 2.0043,
+      "step": 3004
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.538310238946539,
+      "learning_rate": 4.4846660743877056e-05,
+      "loss": 2.0936,
+      "step": 3005
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5412959976112217,
+      "learning_rate": 4.484286155221249e-05,
+      "loss": 2.1894,
+      "step": 3006
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6916963892320801,
+      "learning_rate": 4.483906112167656e-05,
+      "loss": 2.0347,
+      "step": 3007
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5378241418322235,
+      "learning_rate": 4.4835259452506544e-05,
+      "loss": 2.0322,
+      "step": 3008
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5293597395163336,
+      "learning_rate": 4.48314565449398e-05,
+      "loss": 2.2025,
+      "step": 3009
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5350689164969286,
+      "learning_rate": 4.482765239921375e-05,
+      "loss": 1.9953,
+      "step": 3010
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5125460775930459,
+      "learning_rate": 4.482384701556591e-05,
+      "loss": 2.0868,
+      "step": 3011
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5211971935313184,
+      "learning_rate": 4.482004039423385e-05,
+      "loss": 2.0176,
+      "step": 3012
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5327324633742367,
+      "learning_rate": 4.481623253545526e-05,
+      "loss": 2.0096,
+      "step": 3013
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5357325452896446,
+      "learning_rate": 4.481242343946787e-05,
+      "loss": 2.2573,
+      "step": 3014
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5107360098936532,
+      "learning_rate": 4.480861310650948e-05,
+      "loss": 2.0136,
+      "step": 3015
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5290318445763023,
+      "learning_rate": 4.4804801536818e-05,
+      "loss": 2.0329,
+      "step": 3016
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5453232158124682,
+      "learning_rate": 4.48009887306314e-05,
+      "loss": 2.1781,
+      "step": 3017
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5487487935389634,
+      "learning_rate": 4.4797174688187735e-05,
+      "loss": 2.045,
+      "step": 3018
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5505090687133408,
+      "learning_rate": 4.4793359409725115e-05,
+      "loss": 2.0087,
+      "step": 3019
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5448423394806262,
+      "learning_rate": 4.4789542895481765e-05,
+      "loss": 1.9594,
+      "step": 3020
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5858109388536379,
+      "learning_rate": 4.4785725145695944e-05,
+      "loss": 2.01,
+      "step": 3021
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6557275508743523,
+      "learning_rate": 4.478190616060603e-05,
+      "loss": 2.1915,
+      "step": 3022
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6174072734885235,
+      "learning_rate": 4.4778085940450446e-05,
+      "loss": 1.9886,
+      "step": 3023
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5644054020764692,
+      "learning_rate": 4.47742644854677e-05,
+      "loss": 2.07,
+      "step": 3024
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6221490422401692,
+      "learning_rate": 4.4770441795896396e-05,
+      "loss": 2.0201,
+      "step": 3025
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5690496479047262,
+      "learning_rate": 4.476661787197519e-05,
+      "loss": 2.2106,
+      "step": 3026
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5240771090098434,
+      "learning_rate": 4.4762792713942823e-05,
+      "loss": 2.0388,
+      "step": 3027
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6163574606302348,
+      "learning_rate": 4.475896632203812e-05,
+      "loss": 1.9703,
+      "step": 3028
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5223947186697642,
+      "learning_rate": 4.475513869649998e-05,
+      "loss": 2.1751,
+      "step": 3029
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5272071190591924,
+      "learning_rate": 4.4751309837567365e-05,
+      "loss": 2.0733,
+      "step": 3030
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5209451475983699,
+      "learning_rate": 4.4747479745479356e-05,
+      "loss": 1.9709,
+      "step": 3031
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5859910033416534,
+      "learning_rate": 4.474364842047505e-05,
+      "loss": 1.9599,
+      "step": 3032
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5208483879212257,
+      "learning_rate": 4.4739815862793665e-05,
+      "loss": 1.989,
+      "step": 3033
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5530816959492062,
+      "learning_rate": 4.4735982072674476e-05,
+      "loss": 2.22,
+      "step": 3034
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5502348586409187,
+      "learning_rate": 4.473214705035686e-05,
+      "loss": 2.0282,
+      "step": 3035
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.531787217889324,
+      "learning_rate": 4.472831079608024e-05,
+      "loss": 2.0597,
+      "step": 3036
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5415553944593385,
+      "learning_rate": 4.4724473310084127e-05,
+      "loss": 2.0026,
+      "step": 3037
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5673921069995485,
+      "learning_rate": 4.4720634592608116e-05,
+      "loss": 2.1918,
+      "step": 3038
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5315174309918465,
+      "learning_rate": 4.471679464389187e-05,
+      "loss": 2.0076,
+      "step": 3039
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5764970181306173,
+      "learning_rate": 4.471295346417515e-05,
+      "loss": 1.9987,
+      "step": 3040
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5904856550179022,
+      "learning_rate": 4.470911105369774e-05,
+      "loss": 2.1834,
+      "step": 3041
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5784748903700866,
+      "learning_rate": 4.4705267412699575e-05,
+      "loss": 2.094,
+      "step": 3042
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5652651017744937,
+      "learning_rate": 4.470142254142061e-05,
+      "loss": 1.9928,
+      "step": 3043
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5251313855767901,
+      "learning_rate": 4.46975764401009e-05,
+      "loss": 2.0166,
+      "step": 3044
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6055042450496926,
+      "learning_rate": 4.4693729108980575e-05,
+      "loss": 2.0269,
+      "step": 3045
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5659951318790962,
+      "learning_rate": 4.4689880548299835e-05,
+      "loss": 2.2015,
+      "step": 3046
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5883111019040398,
+      "learning_rate": 4.468603075829897e-05,
+      "loss": 2.0019,
+      "step": 3047
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.570511111654427,
+      "learning_rate": 4.4682179739218324e-05,
+      "loss": 1.99,
+      "step": 3048
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5144297455149546,
+      "learning_rate": 4.467832749129835e-05,
+      "loss": 2.0553,
+      "step": 3049
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5520219098859435,
+      "learning_rate": 4.467447401477954e-05,
+      "loss": 2.1901,
+      "step": 3050
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5605898048733032,
+      "learning_rate": 4.4670619309902505e-05,
+      "loss": 2.0287,
+      "step": 3051
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5462322543635677,
+      "learning_rate": 4.4666763376907885e-05,
+      "loss": 2.0158,
+      "step": 3052
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6783121051907095,
+      "learning_rate": 4.466290621603644e-05,
+      "loss": 2.041,
+      "step": 3053
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6015734271325199,
+      "learning_rate": 4.4659047827528986e-05,
+      "loss": 2.2325,
+      "step": 3054
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5001942558233803,
+      "learning_rate": 4.465518821162641e-05,
+      "loss": 2.0691,
+      "step": 3055
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5426410082398664,
+      "learning_rate": 4.465132736856969e-05,
+      "loss": 2.0224,
+      "step": 3056
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.553152788743633,
+      "learning_rate": 4.464746529859987e-05,
+      "loss": 1.9675,
+      "step": 3057
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5585410004654184,
+      "learning_rate": 4.464360200195808e-05,
+      "loss": 2.251,
+      "step": 3058
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6035344159380756,
+      "learning_rate": 4.463973747888553e-05,
+      "loss": 2.0561,
+      "step": 3059
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5716873282611552,
+      "learning_rate": 4.4635871729623465e-05,
+      "loss": 1.954,
+      "step": 3060
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5413273484563424,
+      "learning_rate": 4.463200475441327e-05,
+      "loss": 2.037,
+      "step": 3061
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5620685447740756,
+      "learning_rate": 4.4628136553496375e-05,
+      "loss": 2.2371,
+      "step": 3062
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5256083268662024,
+      "learning_rate": 4.462426712711428e-05,
+      "loss": 2.0082,
+      "step": 3063
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5839402228358669,
+      "learning_rate": 4.462039647550856e-05,
+      "loss": 1.975,
+      "step": 3064
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5243750347545947,
+      "learning_rate": 4.46165245989209e-05,
+      "loss": 1.9719,
+      "step": 3065
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5869131625149748,
+      "learning_rate": 4.461265149759301e-05,
+      "loss": 2.2012,
+      "step": 3066
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5523463400391795,
+      "learning_rate": 4.460877717176672e-05,
+      "loss": 2.0669,
+      "step": 3067
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5321862799436589,
+      "learning_rate": 4.460490162168392e-05,
+      "loss": 1.9933,
+      "step": 3068
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6087579949830706,
+      "learning_rate": 4.460102484758656e-05,
+      "loss": 2.0305,
+      "step": 3069
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5283733021628734,
+      "learning_rate": 4.4597146849716706e-05,
+      "loss": 2.2382,
+      "step": 3070
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5608277290882208,
+      "learning_rate": 4.459326762831647e-05,
+      "loss": 2.0233,
+      "step": 3071
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6122561098971167,
+      "learning_rate": 4.458938718362804e-05,
+      "loss": 1.9949,
+      "step": 3072
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5639313766544823,
+      "learning_rate": 4.458550551589369e-05,
+      "loss": 2.0509,
+      "step": 3073
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5873973947858502,
+      "learning_rate": 4.458162262535578e-05,
+      "loss": 2.2305,
+      "step": 3074
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5196865169561186,
+      "learning_rate": 4.4577738512256716e-05,
+      "loss": 1.9592,
+      "step": 3075
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5556544493886333,
+      "learning_rate": 4.457385317683902e-05,
+      "loss": 1.9434,
+      "step": 3076
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.538819078460489,
+      "learning_rate": 4.456996661934525e-05,
+      "loss": 2.0025,
+      "step": 3077
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.528421568111892,
+      "learning_rate": 4.456607884001808e-05,
+      "loss": 2.2043,
+      "step": 3078
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6105065073154947,
+      "learning_rate": 4.4562189839100214e-05,
+      "loss": 1.9861,
+      "step": 3079
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5140314215635273,
+      "learning_rate": 4.455829961683448e-05,
+      "loss": 2.0949,
+      "step": 3080
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5519385280790672,
+      "learning_rate": 4.455440817346375e-05,
+      "loss": 1.9669,
+      "step": 3081
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6072245415538848,
+      "learning_rate": 4.4550515509231e-05,
+      "loss": 2.1725,
+      "step": 3082
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5272914001221394,
+      "learning_rate": 4.454662162437924e-05,
+      "loss": 2.0246,
+      "step": 3083
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5878263895969793,
+      "learning_rate": 4.454272651915158e-05,
+      "loss": 1.9421,
+      "step": 3084
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5632942506680477,
+      "learning_rate": 4.4538830193791235e-05,
+      "loss": 2.0143,
+      "step": 3085
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6011929662413099,
+      "learning_rate": 4.453493264854145e-05,
+      "loss": 2.2458,
+      "step": 3086
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5667613697920171,
+      "learning_rate": 4.4531033883645575e-05,
+      "loss": 1.9677,
+      "step": 3087
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5986957592903076,
+      "learning_rate": 4.452713389934701e-05,
+      "loss": 1.9914,
+      "step": 3088
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5781240041222101,
+      "learning_rate": 4.4523232695889264e-05,
+      "loss": 1.959,
+      "step": 3089
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6047900673537016,
+      "learning_rate": 4.451933027351589e-05,
+      "loss": 2.1844,
+      "step": 3090
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.51749651286676,
+      "learning_rate": 4.451542663247054e-05,
+      "loss": 1.931,
+      "step": 3091
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.608999343968275,
+      "learning_rate": 4.451152177299694e-05,
+      "loss": 2.0843,
+      "step": 3092
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5613464348121842,
+      "learning_rate": 4.4507615695338875e-05,
+      "loss": 2.0248,
+      "step": 3093
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5398056805071721,
+      "learning_rate": 4.4503708399740226e-05,
+      "loss": 2.2453,
+      "step": 3094
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5497581670785875,
+      "learning_rate": 4.449979988644494e-05,
+      "loss": 1.964,
+      "step": 3095
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5752277970428307,
+      "learning_rate": 4.449589015569704e-05,
+      "loss": 2.0048,
+      "step": 3096
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5429129664166198,
+      "learning_rate": 4.449197920774062e-05,
+      "loss": 2.0177,
+      "step": 3097
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5791858818444132,
+      "learning_rate": 4.448806704281986e-05,
+      "loss": 2.2341,
+      "step": 3098
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5311848929243846,
+      "learning_rate": 4.4484153661179026e-05,
+      "loss": 1.9975,
+      "step": 3099
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5833882168181015,
+      "learning_rate": 4.448023906306243e-05,
+      "loss": 1.9617,
+      "step": 3100
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.636568355941607,
+      "learning_rate": 4.4476323248714483e-05,
+      "loss": 1.9783,
+      "step": 3101
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5190088671941866,
+      "learning_rate": 4.447240621837967e-05,
+      "loss": 2.1412,
+      "step": 3102
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6360666606981059,
+      "learning_rate": 4.446848797230254e-05,
+      "loss": 1.9726,
+      "step": 3103
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5417395305641296,
+      "learning_rate": 4.4464568510727725e-05,
+      "loss": 2.029,
+      "step": 3104
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5191390302221186,
+      "learning_rate": 4.446064783389994e-05,
+      "loss": 1.9548,
+      "step": 3105
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6407095709248257,
+      "learning_rate": 4.445672594206395e-05,
+      "loss": 2.1665,
+      "step": 3106
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5637518108239881,
+      "learning_rate": 4.445280283546465e-05,
+      "loss": 1.9612,
+      "step": 3107
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6665951520584106,
+      "learning_rate": 4.4448878514346934e-05,
+      "loss": 1.9967,
+      "step": 3108
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5235293209956146,
+      "learning_rate": 4.444495297895585e-05,
+      "loss": 2.0021,
+      "step": 3109
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5376388757881346,
+      "learning_rate": 4.4441026229536464e-05,
+      "loss": 2.1908,
+      "step": 3110
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5325745761634787,
+      "learning_rate": 4.443709826633394e-05,
+      "loss": 2.0606,
+      "step": 3111
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5406888155637817,
+      "learning_rate": 4.4433169089593536e-05,
+      "loss": 1.977,
+      "step": 3112
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.523972810914124,
+      "learning_rate": 4.4429238699560536e-05,
+      "loss": 2.0081,
+      "step": 3113
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5858073658422365,
+      "learning_rate": 4.442530709648036e-05,
+      "loss": 2.1435,
+      "step": 3114
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5872113277258418,
+      "learning_rate": 4.442137428059846e-05,
+      "loss": 2.0062,
+      "step": 3115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5492919489949836,
+      "learning_rate": 4.441744025216037e-05,
+      "loss": 2.0071,
+      "step": 3116
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.563643418318092,
+      "learning_rate": 4.4413505011411714e-05,
+      "loss": 2.0355,
+      "step": 3117
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6653275474714171,
+      "learning_rate": 4.44095685585982e-05,
+      "loss": 2.1831,
+      "step": 3118
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5606047937706002,
+      "learning_rate": 4.4405630893965575e-05,
+      "loss": 1.9636,
+      "step": 3119
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6809669035294279,
+      "learning_rate": 4.4401692017759696e-05,
+      "loss": 1.9992,
+      "step": 3120
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.601881053831601,
+      "learning_rate": 4.439775193022648e-05,
+      "loss": 1.9874,
+      "step": 3121
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6073000875923724,
+      "learning_rate": 4.439381063161192e-05,
+      "loss": 2.2055,
+      "step": 3122
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5848905662342224,
+      "learning_rate": 4.438986812216209e-05,
+      "loss": 2.0522,
+      "step": 3123
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.65239186504742,
+      "learning_rate": 4.438592440212315e-05,
+      "loss": 2.0192,
+      "step": 3124
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5781483351767635,
+      "learning_rate": 4.43819794717413e-05,
+      "loss": 1.9548,
+      "step": 3125
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5548028137848052,
+      "learning_rate": 4.437803333126285e-05,
+      "loss": 2.2145,
+      "step": 3126
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5811463187359873,
+      "learning_rate": 4.4374085980934163e-05,
+      "loss": 1.9876,
+      "step": 3127
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5499536181347944,
+      "learning_rate": 4.437013742100171e-05,
+      "loss": 1.9467,
+      "step": 3128
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5579082879306126,
+      "learning_rate": 4.4366187651712e-05,
+      "loss": 2.0134,
+      "step": 3129
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5679271814494812,
+      "learning_rate": 4.436223667331163e-05,
+      "loss": 2.2096,
+      "step": 3130
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5264303092148903,
+      "learning_rate": 4.435828448604729e-05,
+      "loss": 2.031,
+      "step": 3131
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5704734844603361,
+      "learning_rate": 4.4354331090165723e-05,
+      "loss": 2.0001,
+      "step": 3132
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6073458689126624,
+      "learning_rate": 4.435037648591376e-05,
+      "loss": 1.982,
+      "step": 3133
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.526037735751892,
+      "learning_rate": 4.434642067353829e-05,
+      "loss": 2.175,
+      "step": 3134
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6173599401846186,
+      "learning_rate": 4.434246365328631e-05,
+      "loss": 2.1007,
+      "step": 3135
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5558361519249272,
+      "learning_rate": 4.433850542540486e-05,
+      "loss": 1.9811,
+      "step": 3136
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6145086916061856,
+      "learning_rate": 4.433454599014107e-05,
+      "loss": 2.022,
+      "step": 3137
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5668464564979698,
+      "learning_rate": 4.4330585347742144e-05,
+      "loss": 2.2221,
+      "step": 3138
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5176113435286955,
+      "learning_rate": 4.432662349845537e-05,
+      "loss": 1.9587,
+      "step": 3139
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7541378393650352,
+      "learning_rate": 4.4322660442528086e-05,
+      "loss": 1.9759,
+      "step": 3140
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6357930037935415,
+      "learning_rate": 4.431869618020773e-05,
+      "loss": 1.9759,
+      "step": 3141
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5381859501403309,
+      "learning_rate": 4.4314730711741816e-05,
+      "loss": 2.2125,
+      "step": 3142
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6597049375354532,
+      "learning_rate": 4.4310764037377915e-05,
+      "loss": 1.9753,
+      "step": 3143
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5665729499487908,
+      "learning_rate": 4.430679615736368e-05,
+      "loss": 2.0038,
+      "step": 3144
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6275760667624822,
+      "learning_rate": 4.430282707194685e-05,
+      "loss": 1.9638,
+      "step": 3145
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5536298067015744,
+      "learning_rate": 4.429885678137523e-05,
+      "loss": 2.1439,
+      "step": 3146
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6063565792104014,
+      "learning_rate": 4.42948852858967e-05,
+      "loss": 2.0419,
+      "step": 3147
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6036209212881035,
+      "learning_rate": 4.429091258575922e-05,
+      "loss": 2.081,
+      "step": 3148
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5577181903040976,
+      "learning_rate": 4.4286938681210806e-05,
+      "loss": 2.0124,
+      "step": 3149
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6013356735735057,
+      "learning_rate": 4.428296357249958e-05,
+      "loss": 2.1882,
+      "step": 3150
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5494488809182315,
+      "learning_rate": 4.4278987259873725e-05,
+      "loss": 1.9539,
+      "step": 3151
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5816399600246024,
+      "learning_rate": 4.4275009743581496e-05,
+      "loss": 1.9172,
+      "step": 3152
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5485595113700477,
+      "learning_rate": 4.4271031023871226e-05,
+      "loss": 1.9945,
+      "step": 3153
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5072244229612652,
+      "learning_rate": 4.426705110099132e-05,
+      "loss": 2.0357,
+      "step": 3154
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5809621397385337,
+      "learning_rate": 4.426306997519026e-05,
+      "loss": 2.209,
+      "step": 3155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5407912006525201,
+      "learning_rate": 4.425908764671661e-05,
+      "loss": 2.0054,
+      "step": 3156
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6263865302044387,
+      "learning_rate": 4.4255104115819e-05,
+      "loss": 1.9729,
+      "step": 3157
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5433905362304362,
+      "learning_rate": 4.4251119382746134e-05,
+      "loss": 2.1682,
+      "step": 3158
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5713421487637378,
+      "learning_rate": 4.424713344774681e-05,
+      "loss": 1.9676,
+      "step": 3159
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5477496155262481,
+      "learning_rate": 4.4243146311069864e-05,
+      "loss": 2.0346,
+      "step": 3160
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6021761618875515,
+      "learning_rate": 4.423915797296425e-05,
+      "loss": 2.0064,
+      "step": 3161
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6531923646236424,
+      "learning_rate": 4.4235168433678964e-05,
+      "loss": 2.2051,
+      "step": 3162
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5666492547643132,
+      "learning_rate": 4.423117769346309e-05,
+      "loss": 1.9859,
+      "step": 3163
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6673958926570421,
+      "learning_rate": 4.422718575256579e-05,
+      "loss": 1.9826,
+      "step": 3164
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6099876724947764,
+      "learning_rate": 4.42231926112363e-05,
+      "loss": 1.9916,
+      "step": 3165
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.710935982016432,
+      "learning_rate": 4.421919826972391e-05,
+      "loss": 2.1488,
+      "step": 3166
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5346503417930785,
+      "learning_rate": 4.4215202728278036e-05,
+      "loss": 2.0467,
+      "step": 3167
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6288269047593603,
+      "learning_rate": 4.421120598714811e-05,
+      "loss": 1.9476,
+      "step": 3168
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5835183734857277,
+      "learning_rate": 4.4207208046583676e-05,
+      "loss": 1.9326,
+      "step": 3169
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.787401906139243,
+      "learning_rate": 4.420320890683434e-05,
+      "loss": 2.2128,
+      "step": 3170
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5571158610331691,
+      "learning_rate": 4.419920856814977e-05,
+      "loss": 1.944,
+      "step": 3171
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7772338409273627,
+      "learning_rate": 4.4195207030779753e-05,
+      "loss": 1.952,
+      "step": 3172
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6350613062603871,
+      "learning_rate": 4.419120429497409e-05,
+      "loss": 2.1112,
+      "step": 3173
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8228409984522378,
+      "learning_rate": 4.418720036098272e-05,
+      "loss": 2.1932,
+      "step": 3174
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7123757766598553,
+      "learning_rate": 4.41831952290556e-05,
+      "loss": 1.9612,
+      "step": 3175
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.711682191904777,
+      "learning_rate": 4.4179188899442785e-05,
+      "loss": 2.0422,
+      "step": 3176
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7173901475751948,
+      "learning_rate": 4.417518137239443e-05,
+      "loss": 2.0154,
+      "step": 3177
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5594013799117218,
+      "learning_rate": 4.417117264816072e-05,
+      "loss": 2.1322,
+      "step": 3178
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7223450837574492,
+      "learning_rate": 4.416716272699195e-05,
+      "loss": 2.0359,
+      "step": 3179
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5891216083427832,
+      "learning_rate": 4.4163151609138476e-05,
+      "loss": 1.9662,
+      "step": 3180
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6860727842026612,
+      "learning_rate": 4.4159139294850714e-05,
+      "loss": 1.9849,
+      "step": 3181
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6033572521887411,
+      "learning_rate": 4.415512578437918e-05,
+      "loss": 2.1296,
+      "step": 3182
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6047372855600465,
+      "learning_rate": 4.415111107797445e-05,
+      "loss": 1.9784,
+      "step": 3183
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.613395815152527,
+      "learning_rate": 4.414709517588719e-05,
+      "loss": 1.9669,
+      "step": 3184
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6426145319120751,
+      "learning_rate": 4.414307807836811e-05,
+      "loss": 2.0461,
+      "step": 3185
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5595279473275597,
+      "learning_rate": 4.4139059785668026e-05,
+      "loss": 1.9541,
+      "step": 3186
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6924084155046255,
+      "learning_rate": 4.413504029803782e-05,
+      "loss": 2.1733,
+      "step": 3187
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5854552082090089,
+      "learning_rate": 4.413101961572843e-05,
+      "loss": 2.0053,
+      "step": 3188
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5657365601591201,
+      "learning_rate": 4.4126997738990904e-05,
+      "loss": 1.997,
+      "step": 3189
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6407234745274227,
+      "learning_rate": 4.4122974668076325e-05,
+      "loss": 2.2066,
+      "step": 3190
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5916786023686055,
+      "learning_rate": 4.411895040323588e-05,
+      "loss": 2.0836,
+      "step": 3191
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6032633930374778,
+      "learning_rate": 4.4114924944720816e-05,
+      "loss": 2.0106,
+      "step": 3192
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6342046580559905,
+      "learning_rate": 4.411089829278247e-05,
+      "loss": 1.945,
+      "step": 3193
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6389317662153187,
+      "learning_rate": 4.410687044767223e-05,
+      "loss": 2.1984,
+      "step": 3194
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6542441894082088,
+      "learning_rate": 4.410284140964157e-05,
+      "loss": 1.9958,
+      "step": 3195
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5937683115309178,
+      "learning_rate": 4.409881117894205e-05,
+      "loss": 1.9803,
+      "step": 3196
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5925280636945824,
+      "learning_rate": 4.4094779755825286e-05,
+      "loss": 1.9499,
+      "step": 3197
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5232937824866151,
+      "learning_rate": 4.409074714054298e-05,
+      "loss": 2.0667,
+      "step": 3198
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5308756033358528,
+      "learning_rate": 4.4086713333346904e-05,
+      "loss": 2.1992,
+      "step": 3199
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.549670410488058,
+      "learning_rate": 4.40826783344889e-05,
+      "loss": 1.9567,
+      "step": 3200
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6109857083520281,
+      "learning_rate": 4.40786421442209e-05,
+      "loss": 1.9406,
+      "step": 3201
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5449108441190984,
+      "learning_rate": 4.407460476279489e-05,
+      "loss": 2.187,
+      "step": 3202
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5815709353056968,
+      "learning_rate": 4.4070566190462946e-05,
+      "loss": 2.0069,
+      "step": 3203
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5953099370600462,
+      "learning_rate": 4.406652642747721e-05,
+      "loss": 2.068,
+      "step": 3204
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5977740126693556,
+      "learning_rate": 4.40624854740899e-05,
+      "loss": 1.9676,
+      "step": 3205
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5857330017678563,
+      "learning_rate": 4.405844333055331e-05,
+      "loss": 2.011,
+      "step": 3206
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5383836634076384,
+      "learning_rate": 4.405439999711981e-05,
+      "loss": 2.2049,
+      "step": 3207
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5608673415896887,
+      "learning_rate": 4.405035547404185e-05,
+      "loss": 2.0398,
+      "step": 3208
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5436006016514275,
+      "learning_rate": 4.404630976157193e-05,
+      "loss": 1.9866,
+      "step": 3209
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5530611390368625,
+      "learning_rate": 4.404226285996265e-05,
+      "loss": 2.0972,
+      "step": 3210
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5285892806830715,
+      "learning_rate": 4.403821476946667e-05,
+      "loss": 2.1996,
+      "step": 3211
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.951716874295601,
+      "learning_rate": 4.4034165490336725e-05,
+      "loss": 1.9245,
+      "step": 3212
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.561144986898985,
+      "learning_rate": 4.403011502282564e-05,
+      "loss": 1.9978,
+      "step": 3213
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6404935613111418,
+      "learning_rate": 4.4026063367186296e-05,
+      "loss": 2.2001,
+      "step": 3214
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5516836754040394,
+      "learning_rate": 4.4022010523671655e-05,
+      "loss": 1.9813,
+      "step": 3215
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5183665819571563,
+      "learning_rate": 4.401795649253475e-05,
+      "loss": 2.0415,
+      "step": 3216
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6092366520886762,
+      "learning_rate": 4.40139012740287e-05,
+      "loss": 2.0294,
+      "step": 3217
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.564567496348813,
+      "learning_rate": 4.400984486840668e-05,
+      "loss": 1.9832,
+      "step": 3218
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5862631154947727,
+      "learning_rate": 4.400578727592195e-05,
+      "loss": 2.1957,
+      "step": 3219
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.557654457466685,
+      "learning_rate": 4.400172849682783e-05,
+      "loss": 1.9822,
+      "step": 3220
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6379741199535122,
+      "learning_rate": 4.399766853137776e-05,
+      "loss": 2.0048,
+      "step": 3221
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6000896596623431,
+      "learning_rate": 4.3993607379825184e-05,
+      "loss": 2.0727,
+      "step": 3222
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5509959785645787,
+      "learning_rate": 4.398954504242368e-05,
+      "loss": 2.1717,
+      "step": 3223
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.548364913786382,
+      "learning_rate": 4.3985481519426864e-05,
+      "loss": 1.9587,
+      "step": 3224
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5515049139128383,
+      "learning_rate": 4.398141681108844e-05,
+      "loss": 1.9693,
+      "step": 3225
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.605026834898499,
+      "learning_rate": 4.397735091766219e-05,
+      "loss": 2.1739,
+      "step": 3226
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5297361453332348,
+      "learning_rate": 4.397328383940196e-05,
+      "loss": 1.912,
+      "step": 3227
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5813205396252151,
+      "learning_rate": 4.396921557656168e-05,
+      "loss": 2.0003,
+      "step": 3228
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.571226219283037,
+      "learning_rate": 4.396514612939534e-05,
+      "loss": 2.0465,
+      "step": 3229
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5635189241268812,
+      "learning_rate": 4.3961075498157024e-05,
+      "loss": 1.9911,
+      "step": 3230
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.637102595991445,
+      "learning_rate": 4.395700368310086e-05,
+      "loss": 2.2116,
+      "step": 3231
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5920066194814708,
+      "learning_rate": 4.395293068448108e-05,
+      "loss": 1.9721,
+      "step": 3232
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.56295339516942,
+      "learning_rate": 4.394885650255198e-05,
+      "loss": 1.996,
+      "step": 3233
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6921409740498028,
+      "learning_rate": 4.394478113756792e-05,
+      "loss": 2.2119,
+      "step": 3234
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5468943033338183,
+      "learning_rate": 4.394070458978336e-05,
+      "loss": 2.0813,
+      "step": 3235
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5740285660313974,
+      "learning_rate": 4.393662685945279e-05,
+      "loss": 1.9983,
+      "step": 3236
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5914369146109548,
+      "learning_rate": 4.393254794683082e-05,
+      "loss": 1.9919,
+      "step": 3237
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6080906127468796,
+      "learning_rate": 4.3928467852172094e-05,
+      "loss": 2.0118,
+      "step": 3238
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.529411558711373,
+      "learning_rate": 4.392438657573137e-05,
+      "loss": 2.1488,
+      "step": 3239
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6052089419108222,
+      "learning_rate": 4.392030411776344e-05,
+      "loss": 1.9719,
+      "step": 3240
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5075801827915655,
+      "learning_rate": 4.39162204785232e-05,
+      "loss": 2.0834,
+      "step": 3241
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5661231696978897,
+      "learning_rate": 4.391213565826561e-05,
+      "loss": 1.9713,
+      "step": 3242
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5422298781509127,
+      "learning_rate": 4.390804965724569e-05,
+      "loss": 2.1621,
+      "step": 3243
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5721002124573331,
+      "learning_rate": 4.390396247571856e-05,
+      "loss": 2.0071,
+      "step": 3244
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5599377792621094,
+      "learning_rate": 4.3899874113939394e-05,
+      "loss": 1.9911,
+      "step": 3245
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5728331060929721,
+      "learning_rate": 4.389578457216345e-05,
+      "loss": 2.1599,
+      "step": 3246
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5723067273784446,
+      "learning_rate": 4.389169385064603e-05,
+      "loss": 2.0599,
+      "step": 3247
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5792381135786382,
+      "learning_rate": 4.3887601949642565e-05,
+      "loss": 1.9542,
+      "step": 3248
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5513730197014358,
+      "learning_rate": 4.3883508869408525e-05,
+      "loss": 2.0039,
+      "step": 3249
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5651637647062797,
+      "learning_rate": 4.387941461019944e-05,
+      "loss": 2.0264,
+      "step": 3250
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6029764896312595,
+      "learning_rate": 4.387531917227095e-05,
+      "loss": 2.1798,
+      "step": 3251
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5851202139593245,
+      "learning_rate": 4.3871222555878736e-05,
+      "loss": 1.9864,
+      "step": 3252
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5637226189453804,
+      "learning_rate": 4.386712476127858e-05,
+      "loss": 2.0785,
+      "step": 3253
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6260525313325669,
+      "learning_rate": 4.386302578872631e-05,
+      "loss": 1.9864,
+      "step": 3254
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6541163004093657,
+      "learning_rate": 4.385892563847785e-05,
+      "loss": 2.1882,
+      "step": 3255
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6197452157633231,
+      "learning_rate": 4.3854824310789193e-05,
+      "loss": 1.9975,
+      "step": 3256
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6661287455633321,
+      "learning_rate": 4.38507218059164e-05,
+      "loss": 1.9299,
+      "step": 3257
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6142717208272338,
+      "learning_rate": 4.384661812411559e-05,
+      "loss": 2.1479,
+      "step": 3258
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6815042646816536,
+      "learning_rate": 4.384251326564299e-05,
+      "loss": 1.9521,
+      "step": 3259
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6676586756612,
+      "learning_rate": 4.3838407230754885e-05,
+      "loss": 2.0306,
+      "step": 3260
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7664134930290358,
+      "learning_rate": 4.383430001970763e-05,
+      "loss": 1.9782,
+      "step": 3261
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5950400451396344,
+      "learning_rate": 4.383019163275764e-05,
+      "loss": 1.9454,
+      "step": 3262
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6731873750827059,
+      "learning_rate": 4.382608207016145e-05,
+      "loss": 2.1592,
+      "step": 3263
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7378730353702728,
+      "learning_rate": 4.38219713321756e-05,
+      "loss": 2.0057,
+      "step": 3264
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5912092684372481,
+      "learning_rate": 4.3817859419056764e-05,
+      "loss": 1.9883,
+      "step": 3265
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7862470092637582,
+      "learning_rate": 4.381374633106165e-05,
+      "loss": 2.0338,
+      "step": 3266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5452156022144231,
+      "learning_rate": 4.380963206844707e-05,
+      "loss": 2.1739,
+      "step": 3267
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7542629381111885,
+      "learning_rate": 4.380551663146989e-05,
+      "loss": 1.9914,
+      "step": 3268
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6568776710991511,
+      "learning_rate": 4.380140002038704e-05,
+      "loss": 1.9548,
+      "step": 3269
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0378808608981127,
+      "learning_rate": 4.379728223545557e-05,
+      "loss": 1.9983,
+      "step": 3270
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6024612506549143,
+      "learning_rate": 4.379316327693253e-05,
+      "loss": 2.1696,
+      "step": 3271
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6521915125566463,
+      "learning_rate": 4.3789043145075115e-05,
+      "loss": 2.0377,
+      "step": 3272
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8575947938912385,
+      "learning_rate": 4.378492184014054e-05,
+      "loss": 1.9609,
+      "step": 3273
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.631245056545762,
+      "learning_rate": 4.378079936238613e-05,
+      "loss": 1.986,
+      "step": 3274
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5619109139750522,
+      "learning_rate": 4.377667571206926e-05,
+      "loss": 2.203,
+      "step": 3275
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.717538494358374,
+      "learning_rate": 4.377255088944738e-05,
+      "loss": 1.9305,
+      "step": 3276
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6133662391586667,
+      "learning_rate": 4.3768424894778045e-05,
+      "loss": 1.9429,
+      "step": 3277
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5759920123206065,
+      "learning_rate": 4.376429772831883e-05,
+      "loss": 2.0151,
+      "step": 3278
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6415481196106938,
+      "learning_rate": 4.376016939032742e-05,
+      "loss": 2.1659,
+      "step": 3279
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5931806179380206,
+      "learning_rate": 4.3756039881061574e-05,
+      "loss": 2.0373,
+      "step": 3280
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.656448388706346,
+      "learning_rate": 4.3751909200779106e-05,
+      "loss": 1.987,
+      "step": 3281
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6091018394057396,
+      "learning_rate": 4.3747777349737903e-05,
+      "loss": 1.9788,
+      "step": 3282
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.574948755391725,
+      "learning_rate": 4.374364432819595e-05,
+      "loss": 2.2037,
+      "step": 3283
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5352410128152982,
+      "learning_rate": 4.3739510136411286e-05,
+      "loss": 2.0342,
+      "step": 3284
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6116577759711661,
+      "learning_rate": 4.373537477464201e-05,
+      "loss": 1.9694,
+      "step": 3285
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5453778945851593,
+      "learning_rate": 4.373123824314633e-05,
+      "loss": 1.9504,
+      "step": 3286
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6170036439003084,
+      "learning_rate": 4.3727100542182486e-05,
+      "loss": 2.2053,
+      "step": 3287
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6202737169495443,
+      "learning_rate": 4.372296167200883e-05,
+      "loss": 1.9792,
+      "step": 3288
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5979822188804368,
+      "learning_rate": 4.3718821632883756e-05,
+      "loss": 1.9862,
+      "step": 3289
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6348792791900096,
+      "learning_rate": 4.371468042506576e-05,
+      "loss": 1.9002,
+      "step": 3290
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6280045932829521,
+      "learning_rate": 4.371053804881337e-05,
+      "loss": 2.1955,
+      "step": 3291
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7290309835986681,
+      "learning_rate": 4.3706394504385226e-05,
+      "loss": 1.9748,
+      "step": 3292
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6870690638773312,
+      "learning_rate": 4.370224979204003e-05,
+      "loss": 1.9886,
+      "step": 3293
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6793562269941067,
+      "learning_rate": 4.369810391203655e-05,
+      "loss": 1.9683,
+      "step": 3294
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6613749313176212,
+      "learning_rate": 4.3693956864633626e-05,
+      "loss": 2.1846,
+      "step": 3295
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6103176616772819,
+      "learning_rate": 4.368980865009017e-05,
+      "loss": 1.9372,
+      "step": 3296
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6792652753475463,
+      "learning_rate": 4.368565926866519e-05,
+      "loss": 2.0678,
+      "step": 3297
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5682146921406908,
+      "learning_rate": 4.368150872061774e-05,
+      "loss": 1.9393,
+      "step": 3298
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7450505181751049,
+      "learning_rate": 4.3677357006206954e-05,
+      "loss": 2.1528,
+      "step": 3299
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6700431380599663,
+      "learning_rate": 4.367320412569204e-05,
+      "loss": 1.9847,
+      "step": 3300
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7540918979853474,
+      "learning_rate": 4.3669050079332274e-05,
+      "loss": 1.9376,
+      "step": 3301
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6213174934804948,
+      "learning_rate": 4.3664894867387027e-05,
+      "loss": 1.9505,
+      "step": 3302
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8059883836267961,
+      "learning_rate": 4.366073849011572e-05,
+      "loss": 2.2246,
+      "step": 3303
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5488375739823435,
+      "learning_rate": 4.3656580947777836e-05,
+      "loss": 2.0007,
+      "step": 3304
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6484767385118575,
+      "learning_rate": 4.365242224063297e-05,
+      "loss": 1.9556,
+      "step": 3305
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.662754931395236,
+      "learning_rate": 4.364826236894075e-05,
+      "loss": 1.9855,
+      "step": 3306
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6733678515166376,
+      "learning_rate": 4.364410133296091e-05,
+      "loss": 2.1629,
+      "step": 3307
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6135568690289923,
+      "learning_rate": 4.363993913295322e-05,
+      "loss": 1.9612,
+      "step": 3308
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7336750901568297,
+      "learning_rate": 4.3635775769177566e-05,
+      "loss": 2.1149,
+      "step": 3309
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5524292795243116,
+      "learning_rate": 4.3631611241893874e-05,
+      "loss": 1.9865,
+      "step": 3310
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6138231739158613,
+      "learning_rate": 4.362744555136214e-05,
+      "loss": 2.1715,
+      "step": 3311
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6072329554306661,
+      "learning_rate": 4.3623278697842465e-05,
+      "loss": 2.044,
+      "step": 3312
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6522704183978929,
+      "learning_rate": 4.3619110681594996e-05,
+      "loss": 1.9732,
+      "step": 3313
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6357362447883339,
+      "learning_rate": 4.361494150287996e-05,
+      "loss": 1.9921,
+      "step": 3314
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.319668401554552,
+      "learning_rate": 4.361077116195764e-05,
+      "loss": 2.1918,
+      "step": 3315
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6166332987090383,
+      "learning_rate": 4.360659965908843e-05,
+      "loss": 1.9881,
+      "step": 3316
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.765480390873279,
+      "learning_rate": 4.360242699453278e-05,
+      "loss": 2.0332,
+      "step": 3317
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6681156507324202,
+      "learning_rate": 4.3598253168551165e-05,
+      "loss": 2.0119,
+      "step": 3318
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7144967362057545,
+      "learning_rate": 4.359407818140422e-05,
+      "loss": 2.2072,
+      "step": 3319
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6340455968661336,
+      "learning_rate": 4.358990203335258e-05,
+      "loss": 1.9728,
+      "step": 3320
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3053541099892205,
+      "learning_rate": 4.3585724724656985e-05,
+      "loss": 1.9663,
+      "step": 3321
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7740309550279676,
+      "learning_rate": 4.358154625557825e-05,
+      "loss": 2.0592,
+      "step": 3322
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5487803186105508,
+      "learning_rate": 4.357736662637725e-05,
+      "loss": 2.1661,
+      "step": 3323
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7036017129551192,
+      "learning_rate": 4.357318583731492e-05,
+      "loss": 2.0091,
+      "step": 3324
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7003622111423636,
+      "learning_rate": 4.35690038886523e-05,
+      "loss": 2.0285,
+      "step": 3325
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6566763101795272,
+      "learning_rate": 4.3564820780650496e-05,
+      "loss": 1.9786,
+      "step": 3326
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.755847884163255,
+      "learning_rate": 4.356063651357066e-05,
+      "loss": 2.2053,
+      "step": 3327
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7177248356640873,
+      "learning_rate": 4.355645108767403e-05,
+      "loss": 2.0294,
+      "step": 3328
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 6.4502657034409685,
+      "learning_rate": 4.355226450322194e-05,
+      "loss": 1.9614,
+      "step": 3329
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6330405130890208,
+      "learning_rate": 4.354807676047575e-05,
+      "loss": 1.9972,
+      "step": 3330
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5886191931052352,
+      "learning_rate": 4.354388785969694e-05,
+      "loss": 2.2162,
+      "step": 3331
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6482163122056924,
+      "learning_rate": 4.353969780114703e-05,
+      "loss": 1.972,
+      "step": 3332
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6868287014165596,
+      "learning_rate": 4.3535506585087615e-05,
+      "loss": 1.9384,
+      "step": 3333
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.837657520789086,
+      "learning_rate": 4.353131421178038e-05,
+      "loss": 2.0328,
+      "step": 3334
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6424904278086672,
+      "learning_rate": 4.352712068148708e-05,
+      "loss": 2.1864,
+      "step": 3335
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.644194454034765,
+      "learning_rate": 4.352292599446951e-05,
+      "loss": 1.9619,
+      "step": 3336
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6396380955528167,
+      "learning_rate": 4.351873015098958e-05,
+      "loss": 2.0026,
+      "step": 3337
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5787244809705991,
+      "learning_rate": 4.351453315130926e-05,
+      "loss": 2.0064,
+      "step": 3338
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5886744309229474,
+      "learning_rate": 4.3510334995690566e-05,
+      "loss": 2.1857,
+      "step": 3339
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5810043770484355,
+      "learning_rate": 4.3506135684395624e-05,
+      "loss": 2.0112,
+      "step": 3340
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5734649583177113,
+      "learning_rate": 4.3501935217686593e-05,
+      "loss": 1.9991,
+      "step": 3341
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6017723494101405,
+      "learning_rate": 4.349773359582575e-05,
+      "loss": 1.9533,
+      "step": 3342
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6643558397714299,
+      "learning_rate": 4.34935308190754e-05,
+      "loss": 2.1964,
+      "step": 3343
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8339375034204594,
+      "learning_rate": 4.348932688769796e-05,
+      "loss": 1.9827,
+      "step": 3344
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7680698767254079,
+      "learning_rate": 4.348512180195587e-05,
+      "loss": 1.9871,
+      "step": 3345
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5630311519894491,
+      "learning_rate": 4.3480915562111704e-05,
+      "loss": 2.0061,
+      "step": 3346
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7274786665264129,
+      "learning_rate": 4.347670816842804e-05,
+      "loss": 2.2065,
+      "step": 3347
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5736075195281751,
+      "learning_rate": 4.347249962116759e-05,
+      "loss": 1.9347,
+      "step": 3348
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8256146733591994,
+      "learning_rate": 4.3468289920593105e-05,
+      "loss": 2.0248,
+      "step": 3349
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6201616414372438,
+      "learning_rate": 4.346407906696741e-05,
+      "loss": 2.0011,
+      "step": 3350
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5449729622105118,
+      "learning_rate": 4.3459867060553406e-05,
+      "loss": 2.1792,
+      "step": 3351
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5966042895886446,
+      "learning_rate": 4.345565390161406e-05,
+      "loss": 1.9734,
+      "step": 3352
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5739747873451795,
+      "learning_rate": 4.345143959041243e-05,
+      "loss": 2.0546,
+      "step": 3353
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5709630257652846,
+      "learning_rate": 4.3447224127211615e-05,
+      "loss": 1.9761,
+      "step": 3354
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5395760554946689,
+      "learning_rate": 4.344300751227483e-05,
+      "loss": 2.2084,
+      "step": 3355
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.649107722615574,
+      "learning_rate": 4.3438789745865316e-05,
+      "loss": 2.0233,
+      "step": 3356
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6035839514149086,
+      "learning_rate": 4.3434570828246404e-05,
+      "loss": 1.9721,
+      "step": 3357
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5679824546491449,
+      "learning_rate": 4.34303507596815e-05,
+      "loss": 1.9765,
+      "step": 3358
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6048435047577161,
+      "learning_rate": 4.34261295404341e-05,
+      "loss": 2.2147,
+      "step": 3359
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5830877140802417,
+      "learning_rate": 4.342190717076773e-05,
+      "loss": 1.9578,
+      "step": 3360
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5382855391966205,
+      "learning_rate": 4.341768365094601e-05,
+      "loss": 1.9491,
+      "step": 3361
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5322028188645407,
+      "learning_rate": 4.3413458981232645e-05,
+      "loss": 1.9556,
+      "step": 3362
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5714307426896813,
+      "learning_rate": 4.340923316189138e-05,
+      "loss": 2.1344,
+      "step": 3363
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.552061407346916,
+      "learning_rate": 4.3405006193186076e-05,
+      "loss": 1.976,
+      "step": 3364
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49554861170762793,
+      "learning_rate": 4.340077807538062e-05,
+      "loss": 2.0375,
+      "step": 3365
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5813390994549189,
+      "learning_rate": 4.3396548808738996e-05,
+      "loss": 1.9707,
+      "step": 3366
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5767326573369294,
+      "learning_rate": 4.339231839352526e-05,
+      "loss": 2.1926,
+      "step": 3367
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5883719675809757,
+      "learning_rate": 4.338808683000352e-05,
+      "loss": 1.9938,
+      "step": 3368
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6041851103058128,
+      "learning_rate": 4.3383854118437985e-05,
+      "loss": 1.9535,
+      "step": 3369
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5795335634206933,
+      "learning_rate": 4.337962025909291e-05,
+      "loss": 1.9767,
+      "step": 3370
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6099125154451848,
+      "learning_rate": 4.337538525223264e-05,
+      "loss": 2.0771,
+      "step": 3371
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6215558318063329,
+      "learning_rate": 4.337114909812158e-05,
+      "loss": 2.1499,
+      "step": 3372
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6072367112798203,
+      "learning_rate": 4.3366911797024213e-05,
+      "loss": 1.9993,
+      "step": 3373
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6916851379853962,
+      "learning_rate": 4.3362673349205084e-05,
+      "loss": 1.9861,
+      "step": 3374
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6713459140871444,
+      "learning_rate": 4.3358433754928826e-05,
+      "loss": 2.1917,
+      "step": 3375
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6458754280666313,
+      "learning_rate": 4.335419301446014e-05,
+      "loss": 1.9721,
+      "step": 3376
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7231586738055606,
+      "learning_rate": 4.334995112806377e-05,
+      "loss": 1.9626,
+      "step": 3377
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6093023144109259,
+      "learning_rate": 4.3345708096004574e-05,
+      "loss": 2.0173,
+      "step": 3378
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7243109850912278,
+      "learning_rate": 4.334146391854745e-05,
+      "loss": 2.2143,
+      "step": 3379
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6681658855781014,
+      "learning_rate": 4.3337218595957384e-05,
+      "loss": 1.9983,
+      "step": 3380
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6424942320027673,
+      "learning_rate": 4.333297212849944e-05,
+      "loss": 1.9515,
+      "step": 3381
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6980032375813264,
+      "learning_rate": 4.332872451643872e-05,
+      "loss": 2.0125,
+      "step": 3382
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7749696413298154,
+      "learning_rate": 4.332447576004044e-05,
+      "loss": 2.1838,
+      "step": 3383
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6050023657534769,
+      "learning_rate": 4.332022585956986e-05,
+      "loss": 1.9945,
+      "step": 3384
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5621619946895124,
+      "learning_rate": 4.331597481529232e-05,
+      "loss": 1.9254,
+      "step": 3385
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7756037528524355,
+      "learning_rate": 4.331172262747322e-05,
+      "loss": 1.9577,
+      "step": 3386
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6034669934600024,
+      "learning_rate": 4.3307469296378055e-05,
+      "loss": 2.1973,
+      "step": 3387
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6964275124768771,
+      "learning_rate": 4.330321482227237e-05,
+      "loss": 1.9351,
+      "step": 3388
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6379454919002772,
+      "learning_rate": 4.3298959205421795e-05,
+      "loss": 1.9611,
+      "step": 3389
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6556110713189239,
+      "learning_rate": 4.329470244609202e-05,
+      "loss": 2.0276,
+      "step": 3390
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5790557258477006,
+      "learning_rate": 4.329044454454882e-05,
+      "loss": 2.1435,
+      "step": 3391
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6192100637553798,
+      "learning_rate": 4.328618550105802e-05,
+      "loss": 1.966,
+      "step": 3392
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6438452810327356,
+      "learning_rate": 4.328192531588555e-05,
+      "loss": 1.9956,
+      "step": 3393
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6844017010091079,
+      "learning_rate": 4.327766398929737e-05,
+      "loss": 1.9829,
+      "step": 3394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6605226868598343,
+      "learning_rate": 4.327340152155954e-05,
+      "loss": 2.1895,
+      "step": 3395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6714904356371987,
+      "learning_rate": 4.3269137912938185e-05,
+      "loss": 2.0371,
+      "step": 3396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6537050452483303,
+      "learning_rate": 4.32648731636995e-05,
+      "loss": 1.9836,
+      "step": 3397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.596123098414866,
+      "learning_rate": 4.326060727410975e-05,
+      "loss": 1.9887,
+      "step": 3398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6259851481223893,
+      "learning_rate": 4.325634024443527e-05,
+      "loss": 2.1737,
+      "step": 3399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7210761280243484,
+      "learning_rate": 4.325207207494247e-05,
+      "loss": 1.9736,
+      "step": 3400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6334299517706657,
+      "learning_rate": 4.3247802765897825e-05,
+      "loss": 1.9371,
+      "step": 3401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5958427464469437,
+      "learning_rate": 4.324353231756789e-05,
+      "loss": 2.0299,
+      "step": 3402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5876429673431605,
+      "learning_rate": 4.3239260730219286e-05,
+      "loss": 1.9757,
+      "step": 3403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6369202401944646,
+      "learning_rate": 4.323498800411872e-05,
+      "loss": 2.1648,
+      "step": 3404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5352769254506956,
+      "learning_rate": 4.323071413953292e-05,
+      "loss": 1.9338,
+      "step": 3405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5900630856857613,
+      "learning_rate": 4.322643913672876e-05,
+      "loss": 1.9717,
+      "step": 3406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6130648586107842,
+      "learning_rate": 4.322216299597312e-05,
+      "loss": 2.1544,
+      "step": 3407
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5598692510465991,
+      "learning_rate": 4.321788571753299e-05,
+      "loss": 1.9428,
+      "step": 3408
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6758401584062906,
+      "learning_rate": 4.321360730167541e-05,
+      "loss": 2.0142,
+      "step": 3409
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6073161985836986,
+      "learning_rate": 4.3209327748667506e-05,
+      "loss": 1.9879,
+      "step": 3410
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6877559943121544,
+      "learning_rate": 4.320504705877646e-05,
+      "loss": 2.1397,
+      "step": 3411
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7371740557359265,
+      "learning_rate": 4.320076523226954e-05,
+      "loss": 2.015,
+      "step": 3412
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6645100746706,
+      "learning_rate": 4.319648226941408e-05,
+      "loss": 2.0135,
+      "step": 3413
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6395566917933644,
+      "learning_rate": 4.3192198170477474e-05,
+      "loss": 1.9693,
+      "step": 3414
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7190239991781591,
+      "learning_rate": 4.3187912935727204e-05,
+      "loss": 2.0273,
+      "step": 3415
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6344239998005745,
+      "learning_rate": 4.318362656543081e-05,
+      "loss": 2.1596,
+      "step": 3416
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5633102568156599,
+      "learning_rate": 4.3179339059855904e-05,
+      "loss": 1.9564,
+      "step": 3417
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.720819688413015,
+      "learning_rate": 4.3175050419270185e-05,
+      "loss": 2.0135,
+      "step": 3418
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.602534218540415,
+      "learning_rate": 4.3170760643941396e-05,
+      "loss": 2.1653,
+      "step": 3419
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6547663549626784,
+      "learning_rate": 4.316646973413738e-05,
+      "loss": 1.927,
+      "step": 3420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6419636206338514,
+      "learning_rate": 4.316217769012603e-05,
+      "loss": 2.0368,
+      "step": 3421
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5962335720010901,
+      "learning_rate": 4.315788451217531e-05,
+      "loss": 1.9468,
+      "step": 3422
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7040722502169042,
+      "learning_rate": 4.315359020055327e-05,
+      "loss": 1.9613,
+      "step": 3423
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6533006368460195,
+      "learning_rate": 4.314929475552801e-05,
+      "loss": 2.2073,
+      "step": 3424
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5992028038902805,
+      "learning_rate": 4.314499817736773e-05,
+      "loss": 2.0205,
+      "step": 3425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6436494245218551,
+      "learning_rate": 4.314070046634067e-05,
+      "loss": 1.9634,
+      "step": 3426
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6298600433972793,
+      "learning_rate": 4.313640162271514e-05,
+      "loss": 2.0242,
+      "step": 3427
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.622413120761477,
+      "learning_rate": 4.313210164675957e-05,
+      "loss": 2.1217,
+      "step": 3428
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6373264323441286,
+      "learning_rate": 4.312780053874239e-05,
+      "loss": 1.9176,
+      "step": 3429
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6206090643465045,
+      "learning_rate": 4.312349829893217e-05,
+      "loss": 1.9809,
+      "step": 3430
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5844543139264697,
+      "learning_rate": 4.3119194927597486e-05,
+      "loss": 2.1895,
+      "step": 3431
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5985566092734054,
+      "learning_rate": 4.3114890425007023e-05,
+      "loss": 1.9694,
+      "step": 3432
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5846860511863712,
+      "learning_rate": 4.311058479142954e-05,
+      "loss": 2.0556,
+      "step": 3433
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6473493468413358,
+      "learning_rate": 4.310627802713385e-05,
+      "loss": 1.9571,
+      "step": 3434
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5674972753856153,
+      "learning_rate": 4.310197013238884e-05,
+      "loss": 1.965,
+      "step": 3435
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.714437546061388,
+      "learning_rate": 4.309766110746346e-05,
+      "loss": 2.121,
+      "step": 3436
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5960384068287663,
+      "learning_rate": 4.309335095262676e-05,
+      "loss": 1.958,
+      "step": 3437
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5710089477893111,
+      "learning_rate": 4.308903966814782e-05,
+      "loss": 1.9827,
+      "step": 3438
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7151940356901383,
+      "learning_rate": 4.308472725429583e-05,
+      "loss": 2.1806,
+      "step": 3439
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.560446721333317,
+      "learning_rate": 4.308041371134003e-05,
+      "loss": 2.0418,
+      "step": 3440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.663684064300936,
+      "learning_rate": 4.307609903954971e-05,
+      "loss": 1.9948,
+      "step": 3441
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5690660750320651,
+      "learning_rate": 4.307178323919429e-05,
+      "loss": 2.0209,
+      "step": 3442
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7775104219250217,
+      "learning_rate": 4.3067466310543184e-05,
+      "loss": 2.151,
+      "step": 3443
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5791585319961575,
+      "learning_rate": 4.3063148253865934e-05,
+      "loss": 2.0003,
+      "step": 3444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6865458958024743,
+      "learning_rate": 4.305882906943214e-05,
+      "loss": 1.9677,
+      "step": 3445
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7435018116992348,
+      "learning_rate": 4.305450875751146e-05,
+      "loss": 2.0587,
+      "step": 3446
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.574147006628735,
+      "learning_rate": 4.3050187318373624e-05,
+      "loss": 2.0023,
+      "step": 3447
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8285860855126721,
+      "learning_rate": 4.3045864752288445e-05,
+      "loss": 2.172,
+      "step": 3448
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.712835756689722,
+      "learning_rate": 4.304154105952579e-05,
+      "loss": 1.9576,
+      "step": 3449
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6669105143428722,
+      "learning_rate": 4.303721624035561e-05,
+      "loss": 2.003,
+      "step": 3450
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6677012174665905,
+      "learning_rate": 4.303289029504793e-05,
+      "loss": 2.1459,
+      "step": 3451
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7140700264503651,
+      "learning_rate": 4.302856322387282e-05,
+      "loss": 2.0528,
+      "step": 3452
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5543060339324902,
+      "learning_rate": 4.302423502710044e-05,
+      "loss": 2.0025,
+      "step": 3453
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7029809762978481,
+      "learning_rate": 4.301990570500104e-05,
+      "loss": 1.9334,
+      "step": 3454
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6933270363547249,
+      "learning_rate": 4.3015575257844886e-05,
+      "loss": 1.9917,
+      "step": 3455
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6170275333723124,
+      "learning_rate": 4.301124368590236e-05,
+      "loss": 2.1338,
+      "step": 3456
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6934229534188497,
+      "learning_rate": 4.30069109894439e-05,
+      "loss": 1.9524,
+      "step": 3457
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7483831431055771,
+      "learning_rate": 4.300257716874001e-05,
+      "loss": 2.0487,
+      "step": 3458
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7167312180081552,
+      "learning_rate": 4.299824222406128e-05,
+      "loss": 2.0266,
+      "step": 3459
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6944895412136012,
+      "learning_rate": 4.2993906155678345e-05,
+      "loss": 2.1665,
+      "step": 3460
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.596947623670719,
+      "learning_rate": 4.298956896386192e-05,
+      "loss": 1.9717,
+      "step": 3461
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5938412641282599,
+      "learning_rate": 4.2985230648882815e-05,
+      "loss": 1.9447,
+      "step": 3462
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7036160327129625,
+      "learning_rate": 4.298089121101187e-05,
+      "loss": 2.1877,
+      "step": 3463
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5550288729252366,
+      "learning_rate": 4.297655065052001e-05,
+      "loss": 2.1042,
+      "step": 3464
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6381540682304744,
+      "learning_rate": 4.297220896767825e-05,
+      "loss": 1.969,
+      "step": 3465
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5916594263963448,
+      "learning_rate": 4.296786616275765e-05,
+      "loss": 1.9972,
+      "step": 3466
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5480182294382414,
+      "learning_rate": 4.296352223602936e-05,
+      "loss": 1.9891,
+      "step": 3467
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5908297581273191,
+      "learning_rate": 4.2959177187764576e-05,
+      "loss": 2.207,
+      "step": 3468
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.568584987974873,
+      "learning_rate": 4.2954831018234575e-05,
+      "loss": 1.9753,
+      "step": 3469
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5644334140005699,
+      "learning_rate": 4.295048372771072e-05,
+      "loss": 1.9906,
+      "step": 3470
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5875469908771835,
+      "learning_rate": 4.294613531646443e-05,
+      "loss": 2.0638,
+      "step": 3471
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5592134964810097,
+      "learning_rate": 4.294178578476718e-05,
+      "loss": 2.1786,
+      "step": 3472
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5542664728941481,
+      "learning_rate": 4.2937435132890534e-05,
+      "loss": 1.9597,
+      "step": 3473
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5675406162569849,
+      "learning_rate": 4.293308336110613e-05,
+      "loss": 1.9186,
+      "step": 3474
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.588528405367183,
+      "learning_rate": 4.2928730469685654e-05,
+      "loss": 2.1698,
+      "step": 3475
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5699638321728565,
+      "learning_rate": 4.292437645890089e-05,
+      "loss": 1.992,
+      "step": 3476
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5481057694382944,
+      "learning_rate": 4.292002132902366e-05,
+      "loss": 2.0218,
+      "step": 3477
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5445649490794684,
+      "learning_rate": 4.291566508032589e-05,
+      "loss": 2.0208,
+      "step": 3478
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5956980812701832,
+      "learning_rate": 4.291130771307955e-05,
+      "loss": 1.9786,
+      "step": 3479
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5826647802249104,
+      "learning_rate": 4.2906949227556684e-05,
+      "loss": 2.1842,
+      "step": 3480
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5498371002027868,
+      "learning_rate": 4.290258962402941e-05,
+      "loss": 1.9583,
+      "step": 3481
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5373103475212477,
+      "learning_rate": 4.289822890276992e-05,
+      "loss": 1.9537,
+      "step": 3482
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6016919352380754,
+      "learning_rate": 4.289386706405047e-05,
+      "loss": 2.051,
+      "step": 3483
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5436215769634162,
+      "learning_rate": 4.28895041081434e-05,
+      "loss": 2.1338,
+      "step": 3484
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5808837844726863,
+      "learning_rate": 4.2885140035321094e-05,
+      "loss": 2.016,
+      "step": 3485
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6209222730763857,
+      "learning_rate": 4.288077484585602e-05,
+      "loss": 1.9391,
+      "step": 3486
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5734919608423779,
+      "learning_rate": 4.2876408540020716e-05,
+      "loss": 1.9745,
+      "step": 3487
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.60775074838826,
+      "learning_rate": 4.287204111808778e-05,
+      "loss": 2.1612,
+      "step": 3488
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5629158467320676,
+      "learning_rate": 4.286767258032991e-05,
+      "loss": 2.0187,
+      "step": 3489
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5691637534575228,
+      "learning_rate": 4.286330292701983e-05,
+      "loss": 1.9552,
+      "step": 3490
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.611181244835064,
+      "learning_rate": 4.285893215843037e-05,
+      "loss": 1.9683,
+      "step": 3491
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.609905064067679,
+      "learning_rate": 4.285456027483441e-05,
+      "loss": 2.1753,
+      "step": 3492
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5841646607916439,
+      "learning_rate": 4.2850187276504896e-05,
+      "loss": 2.009,
+      "step": 3493
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.680033094241314,
+      "learning_rate": 4.2845813163714865e-05,
+      "loss": 1.9663,
+      "step": 3494
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6142553429540658,
+      "learning_rate": 4.28414379367374e-05,
+      "loss": 2.1743,
+      "step": 3495
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6085222728724563,
+      "learning_rate": 4.2837061595845676e-05,
+      "loss": 2.0248,
+      "step": 3496
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5852490288440427,
+      "learning_rate": 4.2832684141312916e-05,
+      "loss": 1.966,
+      "step": 3497
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5201748258194919,
+      "learning_rate": 4.282830557341243e-05,
+      "loss": 1.9415,
+      "step": 3498
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6259861356398713,
+      "learning_rate": 4.2823925892417586e-05,
+      "loss": 1.9744,
+      "step": 3499
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5707288307950139,
+      "learning_rate": 4.2819545098601833e-05,
+      "loss": 2.1729,
+      "step": 3500
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5469073015790462,
+      "learning_rate": 4.281516319223866e-05,
+      "loss": 1.9067,
+      "step": 3501
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5567885146327375,
+      "learning_rate": 4.281078017360167e-05,
+      "loss": 2.0487,
+      "step": 3502
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5760709394883137,
+      "learning_rate": 4.2806396042964506e-05,
+      "loss": 1.9843,
+      "step": 3503
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5365022875675651,
+      "learning_rate": 4.2802010800600887e-05,
+      "loss": 2.2496,
+      "step": 3504
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5724367302351668,
+      "learning_rate": 4.27976244467846e-05,
+      "loss": 1.9944,
+      "step": 3505
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5566146561870787,
+      "learning_rate": 4.2793236981789507e-05,
+      "loss": 1.9878,
+      "step": 3506
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5527466764045958,
+      "learning_rate": 4.278884840588953e-05,
+      "loss": 1.9375,
+      "step": 3507
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5495380399005712,
+      "learning_rate": 4.278445871935866e-05,
+      "loss": 2.1744,
+      "step": 3508
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5737131415803041,
+      "learning_rate": 4.278006792247099e-05,
+      "loss": 1.9496,
+      "step": 3509
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5389822741873364,
+      "learning_rate": 4.277567601550063e-05,
+      "loss": 1.994,
+      "step": 3510
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6161358148327485,
+      "learning_rate": 4.277128299872178e-05,
+      "loss": 1.9585,
+      "step": 3511
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6175374787626668,
+      "learning_rate": 4.276688887240874e-05,
+      "loss": 2.1893,
+      "step": 3512
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6075712968975086,
+      "learning_rate": 4.276249363683584e-05,
+      "loss": 1.9821,
+      "step": 3513
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5094120351756142,
+      "learning_rate": 4.2758097292277485e-05,
+      "loss": 2.0547,
+      "step": 3514
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6095360460770262,
+      "learning_rate": 4.2753699839008165e-05,
+      "loss": 1.928,
+      "step": 3515
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6031052188996743,
+      "learning_rate": 4.2749301277302436e-05,
+      "loss": 2.1757,
+      "step": 3516
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5632300347640559,
+      "learning_rate": 4.2744901607434907e-05,
+      "loss": 1.9888,
+      "step": 3517
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5516064651210182,
+      "learning_rate": 4.274050082968027e-05,
+      "loss": 1.9775,
+      "step": 3518
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6232206105397682,
+      "learning_rate": 4.27360989443133e-05,
+      "loss": 1.9351,
+      "step": 3519
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5510369828956985,
+      "learning_rate": 4.2731695951608796e-05,
+      "loss": 2.1984,
+      "step": 3520
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6898463706651472,
+      "learning_rate": 4.272729185184168e-05,
+      "loss": 1.9656,
+      "step": 3521
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5434207460363776,
+      "learning_rate": 4.2722886645286896e-05,
+      "loss": 1.9417,
+      "step": 3522
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6158325564207758,
+      "learning_rate": 4.27184803322195e-05,
+      "loss": 1.9336,
+      "step": 3523
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5631465895752555,
+      "learning_rate": 4.271407291291459e-05,
+      "loss": 2.1689,
+      "step": 3524
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6401951007254659,
+      "learning_rate": 4.270966438764733e-05,
+      "loss": 1.9601,
+      "step": 3525
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5830976587396529,
+      "learning_rate": 4.270525475669297e-05,
+      "loss": 1.9174,
+      "step": 3526
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6569767585344545,
+      "learning_rate": 4.270084402032682e-05,
+      "loss": 2.0499,
+      "step": 3527
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6160872431867402,
+      "learning_rate": 4.269643217882427e-05,
+      "loss": 2.1516,
+      "step": 3528
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5621347350661055,
+      "learning_rate": 4.2692019232460755e-05,
+      "loss": 1.9803,
+      "step": 3529
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6838445083236101,
+      "learning_rate": 4.2687605181511796e-05,
+      "loss": 1.9519,
+      "step": 3530
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6362231150004192,
+      "learning_rate": 4.2683190026252984e-05,
+      "loss": 1.9883,
+      "step": 3531
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6638870352443634,
+      "learning_rate": 4.267877376695998e-05,
+      "loss": 2.1864,
+      "step": 3532
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5835125272321948,
+      "learning_rate": 4.26743564039085e-05,
+      "loss": 2.0174,
+      "step": 3533
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5749277383190746,
+      "learning_rate": 4.266993793737434e-05,
+      "loss": 1.9989,
+      "step": 3534
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6846675944468018,
+      "learning_rate": 4.266551836763336e-05,
+      "loss": 1.9736,
+      "step": 3535
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6025763259762551,
+      "learning_rate": 4.2661097694961506e-05,
+      "loss": 2.1971,
+      "step": 3536
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6117460003735535,
+      "learning_rate": 4.2656675919634756e-05,
+      "loss": 2.0025,
+      "step": 3537
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6263141921681835,
+      "learning_rate": 4.2652253041929205e-05,
+      "loss": 1.9023,
+      "step": 3538
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6386828240778866,
+      "learning_rate": 4.264782906212098e-05,
+      "loss": 2.0203,
+      "step": 3539
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5962488546407848,
+      "learning_rate": 4.264340398048628e-05,
+      "loss": 2.177,
+      "step": 3540
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6400321250285554,
+      "learning_rate": 4.263897779730139e-05,
+      "loss": 1.9227,
+      "step": 3541
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.73282255923245,
+      "learning_rate": 4.2634550512842654e-05,
+      "loss": 1.9957,
+      "step": 3542
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5829278847102797,
+      "learning_rate": 4.263012212738649e-05,
+      "loss": 1.9299,
+      "step": 3543
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7623627785148214,
+      "learning_rate": 4.262569264120937e-05,
+      "loss": 2.1304,
+      "step": 3544
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6822532274608936,
+      "learning_rate": 4.262126205458785e-05,
+      "loss": 2.0581,
+      "step": 3545
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5622786746964618,
+      "learning_rate": 4.2616830367798555e-05,
+      "loss": 1.9939,
+      "step": 3546
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8214571913001184,
+      "learning_rate": 4.2612397581118165e-05,
+      "loss": 1.9589,
+      "step": 3547
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5774588481513999,
+      "learning_rate": 4.260796369482344e-05,
+      "loss": 2.1912,
+      "step": 3548
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.692268545416117,
+      "learning_rate": 4.2603528709191206e-05,
+      "loss": 1.9431,
+      "step": 3549
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6791923928694871,
+      "learning_rate": 4.259909262449836e-05,
+      "loss": 1.9993,
+      "step": 3550
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5751758330601876,
+      "learning_rate": 4.259465544102186e-05,
+      "loss": 2.0241,
+      "step": 3551
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.631682776599926,
+      "learning_rate": 4.259021715903874e-05,
+      "loss": 2.1648,
+      "step": 3552
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6333248117430617,
+      "learning_rate": 4.2585777778826106e-05,
+      "loss": 2.0004,
+      "step": 3553
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5575869903289661,
+      "learning_rate": 4.258133730066112e-05,
+      "loss": 2.0006,
+      "step": 3554
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6207432318493338,
+      "learning_rate": 4.257689572482101e-05,
+      "loss": 1.9789,
+      "step": 3555
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6236564842883453,
+      "learning_rate": 4.25724530515831e-05,
+      "loss": 2.1376,
+      "step": 3556
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6018490087744807,
+      "learning_rate": 4.256800928122475e-05,
+      "loss": 1.9989,
+      "step": 3557
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5837737169990872,
+      "learning_rate": 4.2563564414023425e-05,
+      "loss": 2.0121,
+      "step": 3558
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5678897506246137,
+      "learning_rate": 4.2559118450256606e-05,
+      "loss": 1.9457,
+      "step": 3559
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5795636050288442,
+      "learning_rate": 4.25546713902019e-05,
+      "loss": 2.145,
+      "step": 3560
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5758600301418108,
+      "learning_rate": 4.255022323413693e-05,
+      "loss": 1.9447,
+      "step": 3561
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6114102311399662,
+      "learning_rate": 4.2545773982339435e-05,
+      "loss": 1.9254,
+      "step": 3562
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5894951209657461,
+      "learning_rate": 4.254132363508718e-05,
+      "loss": 1.9254,
+      "step": 3563
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6579455900537443,
+      "learning_rate": 4.2536872192658036e-05,
+      "loss": 2.2953,
+      "step": 3564
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5968628123988545,
+      "learning_rate": 4.2532419655329914e-05,
+      "loss": 1.9795,
+      "step": 3565
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6507131490752657,
+      "learning_rate": 4.252796602338081e-05,
+      "loss": 1.9623,
+      "step": 3566
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6137508647944913,
+      "learning_rate": 4.252351129708878e-05,
+      "loss": 2.0012,
+      "step": 3567
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6348735388023643,
+      "learning_rate": 4.251905547673195e-05,
+      "loss": 2.1883,
+      "step": 3568
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5625714378638849,
+      "learning_rate": 4.2514598562588525e-05,
+      "loss": 1.9412,
+      "step": 3569
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6659271808445483,
+      "learning_rate": 4.251014055493675e-05,
+      "loss": 2.0677,
+      "step": 3570
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5666380485461262,
+      "learning_rate": 4.2505681454054976e-05,
+      "loss": 1.9743,
+      "step": 3571
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6660798478226465,
+      "learning_rate": 4.250122126022158e-05,
+      "loss": 2.189,
+      "step": 3572
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6894931664937108,
+      "learning_rate": 4.2496759973715057e-05,
+      "loss": 1.9843,
+      "step": 3573
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6268832561340679,
+      "learning_rate": 4.249229759481392e-05,
+      "loss": 1.9398,
+      "step": 3574
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.69379858918675,
+      "learning_rate": 4.2487834123796796e-05,
+      "loss": 1.967,
+      "step": 3575
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.652530298300391,
+      "learning_rate": 4.2483369560942334e-05,
+      "loss": 2.1919,
+      "step": 3576
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6466408698572437,
+      "learning_rate": 4.247890390652929e-05,
+      "loss": 1.9517,
+      "step": 3577
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7099083335149067,
+      "learning_rate": 4.2474437160836475e-05,
+      "loss": 2.0187,
+      "step": 3578
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6085220034334953,
+      "learning_rate": 4.246996932414276e-05,
+      "loss": 1.9295,
+      "step": 3579
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6055128721805431,
+      "learning_rate": 4.246550039672709e-05,
+      "loss": 2.1425,
+      "step": 3580
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6548863555268384,
+      "learning_rate": 4.246103037886848e-05,
+      "loss": 1.9823,
+      "step": 3581
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5452235542876803,
+      "learning_rate": 4.2456559270846006e-05,
+      "loss": 2.0255,
+      "step": 3582
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.60667387122125,
+      "learning_rate": 4.245208707293883e-05,
+      "loss": 1.9773,
+      "step": 3583
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5532194008719898,
+      "learning_rate": 4.2447613785426166e-05,
+      "loss": 2.158,
+      "step": 3584
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6363204850272001,
+      "learning_rate": 4.2443139408587295e-05,
+      "loss": 1.98,
+      "step": 3585
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6124182363098932,
+      "learning_rate": 4.243866394270157e-05,
+      "loss": 1.9029,
+      "step": 3586
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6270547089256402,
+      "learning_rate": 4.243418738804842e-05,
+      "loss": 1.9367,
+      "step": 3587
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5550182301349645,
+      "learning_rate": 4.242970974490732e-05,
+      "loss": 2.1668,
+      "step": 3588
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.524093186701037,
+      "learning_rate": 4.242523101355784e-05,
+      "loss": 2.0382,
+      "step": 3589
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.58839779694442,
+      "learning_rate": 4.242075119427961e-05,
+      "loss": 1.9615,
+      "step": 3590
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5847928684804756,
+      "learning_rate": 4.241627028735231e-05,
+      "loss": 1.9393,
+      "step": 3591
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5915792754609337,
+      "learning_rate": 4.2411788293055714e-05,
+      "loss": 2.1656,
+      "step": 3592
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5692197619522463,
+      "learning_rate": 4.240730521166964e-05,
+      "loss": 1.9471,
+      "step": 3593
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5302539583684146,
+      "learning_rate": 4.2402821043474e-05,
+      "loss": 1.9468,
+      "step": 3594
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5800767614250936,
+      "learning_rate": 4.239833578874874e-05,
+      "loss": 2.0516,
+      "step": 3595
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.610295258296427,
+      "learning_rate": 4.2393849447773906e-05,
+      "loss": 2.169,
+      "step": 3596
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5541249686121217,
+      "learning_rate": 4.238936202082959e-05,
+      "loss": 1.978,
+      "step": 3597
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6451316626778092,
+      "learning_rate": 4.238487350819597e-05,
+      "loss": 1.922,
+      "step": 3598
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.594026999119571,
+      "learning_rate": 4.238038391015328e-05,
+      "loss": 1.9515,
+      "step": 3599
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6913281622954055,
+      "learning_rate": 4.237589322698182e-05,
+      "loss": 2.1972,
+      "step": 3600
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5723853604082905,
+      "learning_rate": 4.2371401458961965e-05,
+      "loss": 2.0434,
+      "step": 3601
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5638424577995461,
+      "learning_rate": 4.2366908606374154e-05,
+      "loss": 1.9402,
+      "step": 3602
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5650151943904794,
+      "learning_rate": 4.23624146694989e-05,
+      "loss": 1.9579,
+      "step": 3603
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6580841009359153,
+      "learning_rate": 4.2357919648616755e-05,
+      "loss": 2.1871,
+      "step": 3604
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6327577715920094,
+      "learning_rate": 4.235342354400839e-05,
+      "loss": 1.9753,
+      "step": 3605
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5814117013969183,
+      "learning_rate": 4.23489263559545e-05,
+      "loss": 1.9585,
+      "step": 3606
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6753239856025098,
+      "learning_rate": 4.2344428084735865e-05,
+      "loss": 2.0341,
+      "step": 3607
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5747846217294922,
+      "learning_rate": 4.2339928730633335e-05,
+      "loss": 1.9766,
+      "step": 3608
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6735753855103506,
+      "learning_rate": 4.2335428293927823e-05,
+      "loss": 2.1485,
+      "step": 3609
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.617259619016361,
+      "learning_rate": 4.23309267749003e-05,
+      "loss": 1.9316,
+      "step": 3610
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5617756648780682,
+      "learning_rate": 4.232642417383182e-05,
+      "loss": 1.9694,
+      "step": 3611
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6015126110128293,
+      "learning_rate": 4.2321920491003505e-05,
+      "loss": 2.1618,
+      "step": 3612
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5470485102142324,
+      "learning_rate": 4.231741572669654e-05,
+      "loss": 1.9627,
+      "step": 3613
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6729148117988755,
+      "learning_rate": 4.2312909881192164e-05,
+      "loss": 1.9969,
+      "step": 3614
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6340268730218388,
+      "learning_rate": 4.23084029547717e-05,
+      "loss": 1.9647,
+      "step": 3615
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6350587568855608,
+      "learning_rate": 4.2303894947716525e-05,
+      "loss": 2.1933,
+      "step": 3616
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5810210642753171,
+      "learning_rate": 4.229938586030812e-05,
+      "loss": 1.9459,
+      "step": 3617
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6103930794581048,
+      "learning_rate": 4.2294875692827976e-05,
+      "loss": 1.9919,
+      "step": 3618
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5695218794060828,
+      "learning_rate": 4.229036444555769e-05,
+      "loss": 1.9077,
+      "step": 3619
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.538446693709765,
+      "learning_rate": 4.228585211877893e-05,
+      "loss": 2.0638,
+      "step": 3620
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5967420469338331,
+      "learning_rate": 4.2281338712773416e-05,
+      "loss": 2.1515,
+      "step": 3621
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.566508999026493,
+      "learning_rate": 4.227682422782293e-05,
+      "loss": 1.9358,
+      "step": 3622
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5647954476459947,
+      "learning_rate": 4.227230866420932e-05,
+      "loss": 1.9863,
+      "step": 3623
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.621558945295215,
+      "learning_rate": 4.226779202221453e-05,
+      "loss": 2.146,
+      "step": 3624
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5398533495844497,
+      "learning_rate": 4.2263274302120546e-05,
+      "loss": 1.9566,
+      "step": 3625
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.577457045530722,
+      "learning_rate": 4.225875550420943e-05,
+      "loss": 2.07,
+      "step": 3626
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5599499195681623,
+      "learning_rate": 4.2254235628763306e-05,
+      "loss": 1.9593,
+      "step": 3627
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5528940413054003,
+      "learning_rate": 4.224971467606437e-05,
+      "loss": 2.1501,
+      "step": 3628
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5897714662077573,
+      "learning_rate": 4.224519264639488e-05,
+      "loss": 1.9869,
+      "step": 3629
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.58070952960925,
+      "learning_rate": 4.2240669540037184e-05,
+      "loss": 1.955,
+      "step": 3630
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.599571654178837,
+      "learning_rate": 4.223614535727364e-05,
+      "loss": 1.9775,
+      "step": 3631
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6161370519779155,
+      "learning_rate": 4.2231620098386745e-05,
+      "loss": 2.0454,
+      "step": 3632
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5641901942704897,
+      "learning_rate": 4.2227093763659015e-05,
+      "loss": 2.18,
+      "step": 3633
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6682596416905594,
+      "learning_rate": 4.222256635337305e-05,
+      "loss": 1.9559,
+      "step": 3634
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.62565096363727,
+      "learning_rate": 4.221803786781152e-05,
+      "loss": 1.9523,
+      "step": 3635
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5632185529378687,
+      "learning_rate": 4.221350830725714e-05,
+      "loss": 2.2169,
+      "step": 3636
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6509183960603653,
+      "learning_rate": 4.220897767199274e-05,
+      "loss": 1.965,
+      "step": 3637
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6192762582637117,
+      "learning_rate": 4.2204445962301157e-05,
+      "loss": 2.068,
+      "step": 3638
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.680031407651879,
+      "learning_rate": 4.219991317846534e-05,
+      "loss": 1.9664,
+      "step": 3639
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6869830079391385,
+      "learning_rate": 4.219537932076828e-05,
+      "loss": 1.9401,
+      "step": 3640
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6236129672761939,
+      "learning_rate": 4.219084438949305e-05,
+      "loss": 2.1334,
+      "step": 3641
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7697485014305365,
+      "learning_rate": 4.218630838492278e-05,
+      "loss": 1.9315,
+      "step": 3642
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6271973277901026,
+      "learning_rate": 4.218177130734068e-05,
+      "loss": 1.9801,
+      "step": 3643
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7424544855148989,
+      "learning_rate": 4.2177233157030005e-05,
+      "loss": 2.1955,
+      "step": 3644
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7151312443573993,
+      "learning_rate": 4.21726939342741e-05,
+      "loss": 2.0733,
+      "step": 3645
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5976545533176514,
+      "learning_rate": 4.216815363935637e-05,
+      "loss": 1.9701,
+      "step": 3646
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9199907789832565,
+      "learning_rate": 4.2163612272560285e-05,
+      "loss": 1.9659,
+      "step": 3647
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7140632782975725,
+      "learning_rate": 4.2159069834169366e-05,
+      "loss": 2.1708,
+      "step": 3648
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8030525450295289,
+      "learning_rate": 4.215452632446723e-05,
+      "loss": 1.9654,
+      "step": 3649
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7162447764218811,
+      "learning_rate": 4.214998174373754e-05,
+      "loss": 2.0132,
+      "step": 3650
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.717803541402079,
+      "learning_rate": 4.2145436092264036e-05,
+      "loss": 1.9878,
+      "step": 3651
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7983918925216665,
+      "learning_rate": 4.214088937033053e-05,
+      "loss": 1.9996,
+      "step": 3652
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6256055976217183,
+      "learning_rate": 4.213634157822087e-05,
+      "loss": 2.1418,
+      "step": 3653
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6976242666758139,
+      "learning_rate": 4.213179271621901e-05,
+      "loss": 1.945,
+      "step": 3654
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7891417967130395,
+      "learning_rate": 4.2127242784608964e-05,
+      "loss": 1.9337,
+      "step": 3655
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7066755555934711,
+      "learning_rate": 4.2122691783674786e-05,
+      "loss": 2.1451,
+      "step": 3656
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6921345097396429,
+      "learning_rate": 4.211813971370061e-05,
+      "loss": 2.0568,
+      "step": 3657
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.66107110219017,
+      "learning_rate": 4.211358657497065e-05,
+      "loss": 1.9833,
+      "step": 3658
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6506189651012769,
+      "learning_rate": 4.2109032367769184e-05,
+      "loss": 1.9113,
+      "step": 3659
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7299732300683766,
+      "learning_rate": 4.210447709238054e-05,
+      "loss": 2.1513,
+      "step": 3660
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6536139581473334,
+      "learning_rate": 4.209992074908912e-05,
+      "loss": 2.0024,
+      "step": 3661
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7630059739097375,
+      "learning_rate": 4.20953633381794e-05,
+      "loss": 1.9434,
+      "step": 3662
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.627320424637434,
+      "learning_rate": 4.209080485993592e-05,
+      "loss": 2.0309,
+      "step": 3663
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6609549788504543,
+      "learning_rate": 4.208624531464328e-05,
+      "loss": 1.9348,
+      "step": 3664
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8144441512133345,
+      "learning_rate": 4.2081684702586146e-05,
+      "loss": 2.1636,
+      "step": 3665
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6368279950705021,
+      "learning_rate": 4.207712302404927e-05,
+      "loss": 1.9436,
+      "step": 3666
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.67509957430749,
+      "learning_rate": 4.207256027931745e-05,
+      "loss": 1.9727,
+      "step": 3667
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7246471972744529,
+      "learning_rate": 4.2067996468675554e-05,
+      "loss": 2.1569,
+      "step": 3668
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.667741795146805,
+      "learning_rate": 4.206343159240852e-05,
+      "loss": 2.0284,
+      "step": 3669
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7179018458694636,
+      "learning_rate": 4.205886565080135e-05,
+      "loss": 1.9626,
+      "step": 3670
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7388418184937322,
+      "learning_rate": 4.205429864413912e-05,
+      "loss": 2.0461,
+      "step": 3671
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7171662474538194,
+      "learning_rate": 4.2049730572706975e-05,
+      "loss": 1.9874,
+      "step": 3672
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5848568328763255,
+      "learning_rate": 4.2045161436790105e-05,
+      "loss": 2.1068,
+      "step": 3673
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6474219331114105,
+      "learning_rate": 4.204059123667378e-05,
+      "loss": 1.9379,
+      "step": 3674
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7147258856222993,
+      "learning_rate": 4.203601997264333e-05,
+      "loss": 1.9313,
+      "step": 3675
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.534804032294127,
+      "learning_rate": 4.203144764498418e-05,
+      "loss": 1.9753,
+      "step": 3676
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7070121154045783,
+      "learning_rate": 4.2026874253981784e-05,
+      "loss": 2.1294,
+      "step": 3677
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6047364049344168,
+      "learning_rate": 4.2022299799921676e-05,
+      "loss": 1.9219,
+      "step": 3678
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6540707221114954,
+      "learning_rate": 4.2017724283089465e-05,
+      "loss": 1.9848,
+      "step": 3679
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6470687965261076,
+      "learning_rate": 4.201314770377082e-05,
+      "loss": 2.1433,
+      "step": 3680
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6776147701039897,
+      "learning_rate": 4.200857006225146e-05,
+      "loss": 1.9983,
+      "step": 3681
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6336883861353526,
+      "learning_rate": 4.200399135881721e-05,
+      "loss": 2.0358,
+      "step": 3682
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6296053071156608,
+      "learning_rate": 4.199941159375392e-05,
+      "loss": 1.9927,
+      "step": 3683
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6452340141289513,
+      "learning_rate": 4.199483076734754e-05,
+      "loss": 1.9312,
+      "step": 3684
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8134644464980094,
+      "learning_rate": 4.1990248879884044e-05,
+      "loss": 2.1655,
+      "step": 3685
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.706150096135686,
+      "learning_rate": 4.198566593164952e-05,
+      "loss": 1.9844,
+      "step": 3686
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7440402121958606,
+      "learning_rate": 4.1981081922930086e-05,
+      "loss": 1.9808,
+      "step": 3687
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8651733446142279,
+      "learning_rate": 4.197649685401195e-05,
+      "loss": 2.0224,
+      "step": 3688
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.716079448776735,
+      "learning_rate": 4.197191072518139e-05,
+      "loss": 2.1476,
+      "step": 3689
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7528872485841568,
+      "learning_rate": 4.19673235367247e-05,
+      "loss": 1.98,
+      "step": 3690
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5623067613899365,
+      "learning_rate": 4.1962735288928305e-05,
+      "loss": 1.9248,
+      "step": 3691
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8775328457312157,
+      "learning_rate": 4.195814598207866e-05,
+      "loss": 2.1676,
+      "step": 3692
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6086395262624,
+      "learning_rate": 4.1953555616462306e-05,
+      "loss": 1.9684,
+      "step": 3693
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7308533708234124,
+      "learning_rate": 4.194896419236581e-05,
+      "loss": 2.0282,
+      "step": 3694
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6431569238318813,
+      "learning_rate": 4.194437171007587e-05,
+      "loss": 1.9371,
+      "step": 3695
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.651955892258961,
+      "learning_rate": 4.193977816987919e-05,
+      "loss": 1.9868,
+      "step": 3696
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6106225390211322,
+      "learning_rate": 4.193518357206256e-05,
+      "loss": 2.1603,
+      "step": 3697
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7169168433234424,
+      "learning_rate": 4.193058791691285e-05,
+      "loss": 1.9885,
+      "step": 3698
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6363548164402616,
+      "learning_rate": 4.192599120471699e-05,
+      "loss": 2.0001,
+      "step": 3699
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.590361218538044,
+      "learning_rate": 4.192139343576196e-05,
+      "loss": 2.0251,
+      "step": 3700
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7181527413578267,
+      "learning_rate": 4.1916794610334816e-05,
+      "loss": 2.1661,
+      "step": 3701
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6217415248038604,
+      "learning_rate": 4.191219472872271e-05,
+      "loss": 1.9796,
+      "step": 3702
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6439466922081448,
+      "learning_rate": 4.19075937912128e-05,
+      "loss": 1.9892,
+      "step": 3703
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6608511847562606,
+      "learning_rate": 4.1902991798092336e-05,
+      "loss": 1.936,
+      "step": 3704
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6023502963879678,
+      "learning_rate": 4.1898388749648674e-05,
+      "loss": 2.1612,
+      "step": 3705
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5909319257948227,
+      "learning_rate": 4.189378464616918e-05,
+      "loss": 1.9095,
+      "step": 3706
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5664454631982433,
+      "learning_rate": 4.18891794879413e-05,
+      "loss": 2.0211,
+      "step": 3707
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.592439128006405,
+      "learning_rate": 4.188457327525257e-05,
+      "loss": 1.9933,
+      "step": 3708
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6306148919285279,
+      "learning_rate": 4.187996600839056e-05,
+      "loss": 2.1649,
+      "step": 3709
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5813971131414821,
+      "learning_rate": 4.187535768764293e-05,
+      "loss": 1.9943,
+      "step": 3710
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6002910987017142,
+      "learning_rate": 4.18707483132974e-05,
+      "loss": 1.9844,
+      "step": 3711
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6474852591967267,
+      "learning_rate": 4.1866137885641736e-05,
+      "loss": 2.1512,
+      "step": 3712
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5539846145113527,
+      "learning_rate": 4.18615264049638e-05,
+      "loss": 2.019,
+      "step": 3713
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5927722425702392,
+      "learning_rate": 4.1856913871551506e-05,
+      "loss": 1.9364,
+      "step": 3714
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6342850121416378,
+      "learning_rate": 4.185230028569283e-05,
+      "loss": 1.9444,
+      "step": 3715
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6604777794372962,
+      "learning_rate": 4.18476856476758e-05,
+      "loss": 1.9796,
+      "step": 3716
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5853262536874004,
+      "learning_rate": 4.1843069957788564e-05,
+      "loss": 2.1418,
+      "step": 3717
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6001695196600496,
+      "learning_rate": 4.183845321631926e-05,
+      "loss": 1.9306,
+      "step": 3718
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.647535376377552,
+      "learning_rate": 4.1833835423556164e-05,
+      "loss": 2.0501,
+      "step": 3719
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6275327676420367,
+      "learning_rate": 4.1829216579787554e-05,
+      "loss": 1.9902,
+      "step": 3720
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5887762892279188,
+      "learning_rate": 4.1824596685301827e-05,
+      "loss": 2.1678,
+      "step": 3721
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6468948178157242,
+      "learning_rate": 4.181997574038741e-05,
+      "loss": 1.9457,
+      "step": 3722
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5916795784286839,
+      "learning_rate": 4.18153537453328e-05,
+      "loss": 1.9751,
+      "step": 3723
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6727172830218018,
+      "learning_rate": 4.1810730700426585e-05,
+      "loss": 1.995,
+      "step": 3724
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5879229392431792,
+      "learning_rate": 4.1806106605957385e-05,
+      "loss": 2.2419,
+      "step": 3725
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5803117247129603,
+      "learning_rate": 4.1801481462213923e-05,
+      "loss": 2.0243,
+      "step": 3726
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6433096028801361,
+      "learning_rate": 4.179685526948494e-05,
+      "loss": 1.9618,
+      "step": 3727
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6557278814775276,
+      "learning_rate": 4.179222802805928e-05,
+      "loss": 1.9889,
+      "step": 3728
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6609669072280904,
+      "learning_rate": 4.178759973822584e-05,
+      "loss": 2.1725,
+      "step": 3729
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6070272964311362,
+      "learning_rate": 4.178297040027358e-05,
+      "loss": 1.9638,
+      "step": 3730
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6422282982085109,
+      "learning_rate": 4.1778340014491537e-05,
+      "loss": 2.0354,
+      "step": 3731
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7263148650498896,
+      "learning_rate": 4.1773708581168805e-05,
+      "loss": 1.967,
+      "step": 3732
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6768927745311344,
+      "learning_rate": 4.176907610059453e-05,
+      "loss": 2.1521,
+      "step": 3733
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6760268932536927,
+      "learning_rate": 4.1764442573057936e-05,
+      "loss": 1.9839,
+      "step": 3734
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8601180800964421,
+      "learning_rate": 4.1759807998848335e-05,
+      "loss": 1.9589,
+      "step": 3735
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6381109812213627,
+      "learning_rate": 4.175517237825507e-05,
+      "loss": 1.9288,
+      "step": 3736
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8315830592472763,
+      "learning_rate": 4.175053571156756e-05,
+      "loss": 2.2191,
+      "step": 3737
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6895210048175766,
+      "learning_rate": 4.174589799907529e-05,
+      "loss": 2.0246,
+      "step": 3738
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6992436444783046,
+      "learning_rate": 4.1741259241067814e-05,
+      "loss": 1.9902,
+      "step": 3739
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.674783121252672,
+      "learning_rate": 4.1736619437834754e-05,
+      "loss": 1.955,
+      "step": 3740
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7897336829256808,
+      "learning_rate": 4.1731978589665774e-05,
+      "loss": 2.1828,
+      "step": 3741
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.691425194650292,
+      "learning_rate": 4.172733669685065e-05,
+      "loss": 1.9785,
+      "step": 3742
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6987882457763207,
+      "learning_rate": 4.1722693759679165e-05,
+      "loss": 1.9203,
+      "step": 3743
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5805207964582163,
+      "learning_rate": 4.171804977844122e-05,
+      "loss": 2.0378,
+      "step": 3744
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7442671714135737,
+      "learning_rate": 4.1713404753426744e-05,
+      "loss": 2.1423,
+      "step": 3745
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6770209912668741,
+      "learning_rate": 4.170875868492575e-05,
+      "loss": 2.0107,
+      "step": 3746
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6920145614139671,
+      "learning_rate": 4.170411157322831e-05,
+      "loss": 1.9467,
+      "step": 3747
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7737534641922775,
+      "learning_rate": 4.169946341862456e-05,
+      "loss": 1.9685,
+      "step": 3748
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5401961910033874,
+      "learning_rate": 4.169481422140471e-05,
+      "loss": 2.1523,
+      "step": 3749
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7122973873269187,
+      "learning_rate": 4.169016398185902e-05,
+      "loss": 2.0053,
+      "step": 3750
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.625318261047493,
+      "learning_rate": 4.168551270027783e-05,
+      "loss": 2.0251,
+      "step": 3751
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5696754666203367,
+      "learning_rate": 4.168086037695153e-05,
+      "loss": 1.9869,
+      "step": 3752
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6690044794565485,
+      "learning_rate": 4.16762070121706e-05,
+      "loss": 2.1924,
+      "step": 3753
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5967770646292277,
+      "learning_rate": 4.167155260622555e-05,
+      "loss": 2.0024,
+      "step": 3754
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.598320225419967,
+      "learning_rate": 4.1666897159406984e-05,
+      "loss": 1.9937,
+      "step": 3755
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.589541358190962,
+      "learning_rate": 4.166224067200556e-05,
+      "loss": 2.0132,
+      "step": 3756
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6334616475154086,
+      "learning_rate": 4.1657583144312004e-05,
+      "loss": 2.1651,
+      "step": 3757
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6560129408766884,
+      "learning_rate": 4.1652924576617104e-05,
+      "loss": 1.9557,
+      "step": 3758
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6146363391234785,
+      "learning_rate": 4.1648264969211704e-05,
+      "loss": 1.9219,
+      "step": 3759
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5892504839759339,
+      "learning_rate": 4.164360432238672e-05,
+      "loss": 1.9663,
+      "step": 3760
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6571287875756067,
+      "learning_rate": 4.163894263643317e-05,
+      "loss": 2.1751,
+      "step": 3761
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6322659039374557,
+      "learning_rate": 4.163427991164206e-05,
+      "loss": 2.0274,
+      "step": 3762
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6154836965409817,
+      "learning_rate": 4.162961614830452e-05,
+      "loss": 1.9442,
+      "step": 3763
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6225135413794903,
+      "learning_rate": 4.162495134671173e-05,
+      "loss": 1.9632,
+      "step": 3764
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.795049432926198,
+      "learning_rate": 4.162028550715493e-05,
+      "loss": 2.1672,
+      "step": 3765
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7213558375102108,
+      "learning_rate": 4.1615618629925424e-05,
+      "loss": 1.9773,
+      "step": 3766
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6433100509610542,
+      "learning_rate": 4.161095071531459e-05,
+      "loss": 1.9275,
+      "step": 3767
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6187964614148259,
+      "learning_rate": 4.1606281763613866e-05,
+      "loss": 1.9209,
+      "step": 3768
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6156113816969525,
+      "learning_rate": 4.160161177511475e-05,
+      "loss": 2.1876,
+      "step": 3769
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6513655067227441,
+      "learning_rate": 4.1596940750108815e-05,
+      "loss": 1.9644,
+      "step": 3770
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5770425557614786,
+      "learning_rate": 4.159226868888768e-05,
+      "loss": 1.9649,
+      "step": 3771
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.704248245161816,
+      "learning_rate": 4.1587595591743046e-05,
+      "loss": 1.9884,
+      "step": 3772
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7243638607066908,
+      "learning_rate": 4.158292145896668e-05,
+      "loss": 2.1617,
+      "step": 3773
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6268560745072803,
+      "learning_rate": 4.1578246290850406e-05,
+      "loss": 1.958,
+      "step": 3774
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7851414752569611,
+      "learning_rate": 4.157357008768611e-05,
+      "loss": 2.0203,
+      "step": 3775
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6003588521340534,
+      "learning_rate": 4.156889284976575e-05,
+      "loss": 1.9558,
+      "step": 3776
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7828218098407062,
+      "learning_rate": 4.156421457738133e-05,
+      "loss": 2.1569,
+      "step": 3777
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.640437062290225,
+      "learning_rate": 4.155953527082496e-05,
+      "loss": 1.9903,
+      "step": 3778
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7795827255519555,
+      "learning_rate": 4.155485493038877e-05,
+      "loss": 1.9329,
+      "step": 3779
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6161625806171004,
+      "learning_rate": 4.155017355636497e-05,
+      "loss": 2.0008,
+      "step": 3780
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8121995485210262,
+      "learning_rate": 4.1545491149045855e-05,
+      "loss": 2.2146,
+      "step": 3781
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5827979963584551,
+      "learning_rate": 4.154080770872375e-05,
+      "loss": 1.9621,
+      "step": 3782
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.745533884517018,
+      "learning_rate": 4.153612323569108e-05,
+      "loss": 1.9825,
+      "step": 3783
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6806726306196428,
+      "learning_rate": 4.15314377302403e-05,
+      "loss": 1.9997,
+      "step": 3784
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6785866854606835,
+      "learning_rate": 4.1526751192663935e-05,
+      "loss": 2.1872,
+      "step": 3785
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6685545687231972,
+      "learning_rate": 4.1522063623254614e-05,
+      "loss": 1.9986,
+      "step": 3786
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6158387432072003,
+      "learning_rate": 4.151737502230498e-05,
+      "loss": 2.0551,
+      "step": 3787
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5783198797559468,
+      "learning_rate": 4.151268539010777e-05,
+      "loss": 1.9408,
+      "step": 3788
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7293824441419369,
+      "learning_rate": 4.150799472695578e-05,
+      "loss": 2.1578,
+      "step": 3789
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.614454177983162,
+      "learning_rate": 4.150330303314186e-05,
+      "loss": 1.9386,
+      "step": 3790
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6715579437601324,
+      "learning_rate": 4.1498610308958944e-05,
+      "loss": 1.9401,
+      "step": 3791
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6604314832863485,
+      "learning_rate": 4.14939165547e-05,
+      "loss": 1.9834,
+      "step": 3792
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6463366341505608,
+      "learning_rate": 4.148922177065808e-05,
+      "loss": 2.1602,
+      "step": 3793
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6125310322242448,
+      "learning_rate": 4.1484525957126306e-05,
+      "loss": 1.9544,
+      "step": 3794
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6675931853812922,
+      "learning_rate": 4.147982911439786e-05,
+      "loss": 1.9501,
+      "step": 3795
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5795349176630711,
+      "learning_rate": 4.147513124276599e-05,
+      "loss": 1.9548,
+      "step": 3796
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6957731277957201,
+      "learning_rate": 4.1470432342523986e-05,
+      "loss": 2.2193,
+      "step": 3797
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5775759814995692,
+      "learning_rate": 4.146573241396523e-05,
+      "loss": 1.9176,
+      "step": 3798
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7166914214504236,
+      "learning_rate": 4.1461031457383155e-05,
+      "loss": 1.9598,
+      "step": 3799
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6249005085959003,
+      "learning_rate": 4.1456329473071266e-05,
+      "loss": 2.0379,
+      "step": 3800
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6984090408310287,
+      "learning_rate": 4.1451626461323124e-05,
+      "loss": 2.2095,
+      "step": 3801
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6744236103759379,
+      "learning_rate": 4.144692242243235e-05,
+      "loss": 1.9289,
+      "step": 3802
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6968875299821077,
+      "learning_rate": 4.144221735669265e-05,
+      "loss": 1.9086,
+      "step": 3803
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5967936196153527,
+      "learning_rate": 4.143751126439777e-05,
+      "loss": 2.0279,
+      "step": 3804
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6990424801649224,
+      "learning_rate": 4.143280414584154e-05,
+      "loss": 2.2224,
+      "step": 3805
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5412543314057722,
+      "learning_rate": 4.1428096001317834e-05,
+      "loss": 2.0582,
+      "step": 3806
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5916577335081531,
+      "learning_rate": 4.142338683112061e-05,
+      "loss": 1.9157,
+      "step": 3807
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6903530340385335,
+      "learning_rate": 4.1418676635543875e-05,
+      "loss": 1.9892,
+      "step": 3808
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6203696010839237,
+      "learning_rate": 4.141396541488171e-05,
+      "loss": 2.1407,
+      "step": 3809
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5971054918061296,
+      "learning_rate": 4.1409253169428255e-05,
+      "loss": 1.9674,
+      "step": 3810
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5558782665163996,
+      "learning_rate": 4.140453989947771e-05,
+      "loss": 1.9476,
+      "step": 3811
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5782427070370348,
+      "learning_rate": 4.139982560532435e-05,
+      "loss": 2.0582,
+      "step": 3812
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.585325603865789,
+      "learning_rate": 4.13951102872625e-05,
+      "loss": 2.1974,
+      "step": 3813
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6051801542283487,
+      "learning_rate": 4.1390393945586576e-05,
+      "loss": 1.9737,
+      "step": 3814
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5680878182741933,
+      "learning_rate": 4.1385676580591016e-05,
+      "loss": 2.0094,
+      "step": 3815
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6363873734387459,
+      "learning_rate": 4.1380958192570346e-05,
+      "loss": 1.9964,
+      "step": 3816
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6007227104821521,
+      "learning_rate": 4.137623878181917e-05,
+      "loss": 2.1475,
+      "step": 3817
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6815804682887711,
+      "learning_rate": 4.137151834863213e-05,
+      "loss": 2.0552,
+      "step": 3818
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6222419737891229,
+      "learning_rate": 4.136679689330394e-05,
+      "loss": 1.9573,
+      "step": 3819
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6444219139377292,
+      "learning_rate": 4.1362074416129395e-05,
+      "loss": 1.9884,
+      "step": 3820
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6472578944575412,
+      "learning_rate": 4.1357350917403314e-05,
+      "loss": 2.133,
+      "step": 3821
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5571814318918725,
+      "learning_rate": 4.135262639742062e-05,
+      "loss": 1.8615,
+      "step": 3822
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6234326717590418,
+      "learning_rate": 4.1347900856476285e-05,
+      "loss": 1.9736,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.522757702886885,
+      "learning_rate": 4.134317429486534e-05,
+      "loss": 1.9592,
+      "step": 3824
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.587477050938539,
+      "learning_rate": 4.133844671288288e-05,
+      "loss": 2.0882,
+      "step": 3825
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6210542377236401,
+      "learning_rate": 4.133371811082408e-05,
+      "loss": 2.1116,
+      "step": 3826
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5511949637425722,
+      "learning_rate": 4.1328988488984156e-05,
+      "loss": 1.9628,
+      "step": 3827
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5773843327705133,
+      "learning_rate": 4.1324257847658385e-05,
+      "loss": 1.9791,
+      "step": 3828
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7238198762020395,
+      "learning_rate": 4.131952618714215e-05,
+      "loss": 2.1866,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6251000848677782,
+      "learning_rate": 4.1314793507730855e-05,
+      "loss": 1.9598,
+      "step": 3830
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7334419789332116,
+      "learning_rate": 4.131005980971997e-05,
+      "loss": 2.0115,
+      "step": 3831
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.61654808122787,
+      "learning_rate": 4.130532509340505e-05,
+      "loss": 1.9669,
+      "step": 3832
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7098605691584817,
+      "learning_rate": 4.13005893590817e-05,
+      "loss": 2.0955,
+      "step": 3833
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6233435367167875,
+      "learning_rate": 4.129585260704559e-05,
+      "loss": 1.9566,
+      "step": 3834
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5885074043276831,
+      "learning_rate": 4.129111483759247e-05,
+      "loss": 1.967,
+      "step": 3835
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6161076841351686,
+      "learning_rate": 4.128637605101811e-05,
+      "loss": 1.9802,
+      "step": 3836
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.524378497797327,
+      "learning_rate": 4.1281636247618386e-05,
+      "loss": 2.0529,
+      "step": 3837
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.667027443925809,
+      "learning_rate": 4.127689542768923e-05,
+      "loss": 2.1559,
+      "step": 3838
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6157032631702479,
+      "learning_rate": 4.127215359152663e-05,
+      "loss": 1.9427,
+      "step": 3839
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5900975255754316,
+      "learning_rate": 4.1267410739426625e-05,
+      "loss": 1.9955,
+      "step": 3840
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8209839775849878,
+      "learning_rate": 4.126266687168535e-05,
+      "loss": 2.1679,
+      "step": 3841
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6071237933102744,
+      "learning_rate": 4.125792198859896e-05,
+      "loss": 1.9518,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7098167601851108,
+      "learning_rate": 4.1253176090463716e-05,
+      "loss": 2.0249,
+      "step": 3843
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.627390929245382,
+      "learning_rate": 4.124842917757593e-05,
+      "loss": 1.9704,
+      "step": 3844
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6204529250606994,
+      "learning_rate": 4.1243681250231945e-05,
+      "loss": 2.1617,
+      "step": 3845
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6361972559365728,
+      "learning_rate": 4.123893230872822e-05,
+      "loss": 1.9475,
+      "step": 3846
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5938836620228498,
+      "learning_rate": 4.123418235336123e-05,
+      "loss": 1.9606,
+      "step": 3847
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6716685132204049,
+      "learning_rate": 4.122943138442755e-05,
+      "loss": 1.9911,
+      "step": 3848
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6664801702246396,
+      "learning_rate": 4.12246794022238e-05,
+      "loss": 2.0585,
+      "step": 3849
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6241313133185189,
+      "learning_rate": 4.121992640704665e-05,
+      "loss": 2.1804,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6469842430902533,
+      "learning_rate": 4.121517239919287e-05,
+      "loss": 1.9379,
+      "step": 3851
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6268787637599117,
+      "learning_rate": 4.121041737895926e-05,
+      "loss": 1.9418,
+      "step": 3852
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6673374740659214,
+      "learning_rate": 4.12056613466427e-05,
+      "loss": 2.1749,
+      "step": 3853
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6010038114486702,
+      "learning_rate": 4.1200904302540136e-05,
+      "loss": 1.9673,
+      "step": 3854
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5980945872837211,
+      "learning_rate": 4.1196146246948555e-05,
+      "loss": 1.9698,
+      "step": 3855
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6505349547823434,
+      "learning_rate": 4.119138718016502e-05,
+      "loss": 2.0493,
+      "step": 3856
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6289551684898544,
+      "learning_rate": 4.1186627102486674e-05,
+      "loss": 1.9395,
+      "step": 3857
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6917665537546885,
+      "learning_rate": 4.1181866014210705e-05,
+      "loss": 2.1503,
+      "step": 3858
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6641802093286319,
+      "learning_rate": 4.117710391563436e-05,
+      "loss": 1.9582,
+      "step": 3859
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5819769923215931,
+      "learning_rate": 4.1172340807054966e-05,
+      "loss": 2.0016,
+      "step": 3860
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7389922370062213,
+      "learning_rate": 4.116757668876989e-05,
+      "loss": 2.1497,
+      "step": 3861
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5921953994301721,
+      "learning_rate": 4.1162811561076584e-05,
+      "loss": 2.0115,
+      "step": 3862
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7361337257409508,
+      "learning_rate": 4.115804542427256e-05,
+      "loss": 1.9585,
+      "step": 3863
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6390181972698236,
+      "learning_rate": 4.1153278278655386e-05,
+      "loss": 1.9763,
+      "step": 3864
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6448865750421973,
+      "learning_rate": 4.1148510124522674e-05,
+      "loss": 2.1723,
+      "step": 3865
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6366256531070978,
+      "learning_rate": 4.114374096217214e-05,
+      "loss": 1.979,
+      "step": 3866
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6361417826220834,
+      "learning_rate": 4.113897079190154e-05,
+      "loss": 1.9386,
+      "step": 3867
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7078568579808728,
+      "learning_rate": 4.1134199614008695e-05,
+      "loss": 2.0403,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.663271888653804,
+      "learning_rate": 4.112942742879149e-05,
+      "loss": 1.9351,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6333105124393905,
+      "learning_rate": 4.112465423654786e-05,
+      "loss": 2.1861,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.590322939627224,
+      "learning_rate": 4.111988003757583e-05,
+      "loss": 2.0093,
+      "step": 3871
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6361424907464709,
+      "learning_rate": 4.111510483217347e-05,
+      "loss": 1.9733,
+      "step": 3872
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.601702979866309,
+      "learning_rate": 4.1110328620638904e-05,
+      "loss": 2.1791,
+      "step": 3873
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.61119646632048,
+      "learning_rate": 4.110555140327035e-05,
+      "loss": 2.0144,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6710618336666347,
+      "learning_rate": 4.110077318036605e-05,
+      "loss": 1.8737,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7984900059817356,
+      "learning_rate": 4.1095993952224344e-05,
+      "loss": 1.9586,
+      "step": 3876
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6716178927325301,
+      "learning_rate": 4.109121371914361e-05,
+      "loss": 2.1351,
+      "step": 3877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7196671657757124,
+      "learning_rate": 4.108643248142229e-05,
+      "loss": 1.9578,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6514675716239146,
+      "learning_rate": 4.108165023935891e-05,
+      "loss": 1.9881,
+      "step": 3879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7576338022037286,
+      "learning_rate": 4.1076866993252044e-05,
+      "loss": 2.0097,
+      "step": 3880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6239123027027155,
+      "learning_rate": 4.107208274340032e-05,
+      "loss": 1.9381,
+      "step": 3881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8501749336640424,
+      "learning_rate": 4.106729749010245e-05,
+      "loss": 2.1388,
+      "step": 3882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8044566692530245,
+      "learning_rate": 4.106251123365719e-05,
+      "loss": 2.0012,
+      "step": 3883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7300007290830302,
+      "learning_rate": 4.105772397436337e-05,
+      "loss": 1.9595,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8729567628153265,
+      "learning_rate": 4.105293571251988e-05,
+      "loss": 2.1429,
+      "step": 3885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6933568471502953,
+      "learning_rate": 4.1048146448425656e-05,
+      "loss": 1.9622,
+      "step": 3886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6015375889673975,
+      "learning_rate": 4.104335618237972e-05,
+      "loss": 2.0398,
+      "step": 3887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6893158007409019,
+      "learning_rate": 4.103856491468116e-05,
+      "loss": 2.0003,
+      "step": 3888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7030049488384783,
+      "learning_rate": 4.1033772645629095e-05,
+      "loss": 1.9561,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8214531954570513,
+      "learning_rate": 4.1028979375522745e-05,
+      "loss": 2.1591,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7519576420218346,
+      "learning_rate": 4.1024185104661354e-05,
+      "loss": 1.9539,
+      "step": 3891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7359641436046099,
+      "learning_rate": 4.101938983334427e-05,
+      "loss": 1.9944,
+      "step": 3892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7713006825437768,
+      "learning_rate": 4.1014593561870855e-05,
+      "loss": 2.0327,
+      "step": 3893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8861839918944877,
+      "learning_rate": 4.1009796290540586e-05,
+      "loss": 2.1134,
+      "step": 3894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6414954375270823,
+      "learning_rate": 4.1004998019652966e-05,
+      "loss": 1.9314,
+      "step": 3895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9701965403462911,
+      "learning_rate": 4.1000198749507555e-05,
+      "loss": 1.9541,
+      "step": 3896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6865817510739247,
+      "learning_rate": 4.099539848040401e-05,
+      "loss": 2.1611,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8992472177473495,
+      "learning_rate": 4.0990597212642036e-05,
+      "loss": 1.9594,
+      "step": 3898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8409069075456934,
+      "learning_rate": 4.0985794946521385e-05,
+      "loss": 2.0357,
+      "step": 3899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6275592996063134,
+      "learning_rate": 4.098099168234187e-05,
+      "loss": 1.9735,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.874378122345896,
+      "learning_rate": 4.097618742040341e-05,
+      "loss": 1.9749,
+      "step": 3901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8336322301344887,
+      "learning_rate": 4.0971382161005936e-05,
+      "loss": 2.16,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.756447500049861,
+      "learning_rate": 4.0966575904449455e-05,
+      "loss": 1.9693,
+      "step": 3903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7280308604770998,
+      "learning_rate": 4.096176865103405e-05,
+      "loss": 1.9693,
+      "step": 3904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7692940030010743,
+      "learning_rate": 4.095696040105986e-05,
+      "loss": 1.9629,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7008616252657456,
+      "learning_rate": 4.095215115482707e-05,
+      "loss": 2.1898,
+      "step": 3906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7994155487578607,
+      "learning_rate": 4.094734091263596e-05,
+      "loss": 1.9899,
+      "step": 3907
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6206056063284756,
+      "learning_rate": 4.0942529674786835e-05,
+      "loss": 1.9675,
+      "step": 3908
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7773799855828458,
+      "learning_rate": 4.0937717441580094e-05,
+      "loss": 2.1524,
+      "step": 3909
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6650349460778588,
+      "learning_rate": 4.093290421331618e-05,
+      "loss": 1.9509,
+      "step": 3910
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5244867637826205,
+      "learning_rate": 4.0928089990295596e-05,
+      "loss": 2.0021,
+      "step": 3911
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8067763805729119,
+      "learning_rate": 4.0923274772818926e-05,
+      "loss": 1.9896,
+      "step": 3912
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.709736541042762,
+      "learning_rate": 4.091845856118679e-05,
+      "loss": 1.9559,
+      "step": 3913
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7596323234363336,
+      "learning_rate": 4.0913641355699905e-05,
+      "loss": 2.1884,
+      "step": 3914
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7109707777121279,
+      "learning_rate": 4.0908823156659e-05,
+      "loss": 1.9795,
+      "step": 3915
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7377098262679318,
+      "learning_rate": 4.090400396436492e-05,
+      "loss": 2.0191,
+      "step": 3916
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6128321289481717,
+      "learning_rate": 4.089918377911853e-05,
+      "loss": 2.1359,
+      "step": 3917
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6487596788145268,
+      "learning_rate": 4.089436260122079e-05,
+      "loss": 2.0608,
+      "step": 3918
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5528278559792273,
+      "learning_rate": 4.088954043097269e-05,
+      "loss": 1.9681,
+      "step": 3919
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6877384704484403,
+      "learning_rate": 4.088471726867531e-05,
+      "loss": 1.9483,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5882308502786506,
+      "learning_rate": 4.087989311462978e-05,
+      "loss": 1.9969,
+      "step": 3921
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6646453401542289,
+      "learning_rate": 4.0875067969137274e-05,
+      "loss": 2.1848,
+      "step": 3922
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5977845743029465,
+      "learning_rate": 4.087024183249906e-05,
+      "loss": 1.9554,
+      "step": 3923
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7494689848784223,
+      "learning_rate": 4.086541470501646e-05,
+      "loss": 2.0282,
+      "step": 3924
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5366623985619712,
+      "learning_rate": 4.086058658699083e-05,
+      "loss": 1.9795,
+      "step": 3925
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6488769816750848,
+      "learning_rate": 4.085575747872363e-05,
+      "loss": 2.1629,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5618492324406718,
+      "learning_rate": 4.085092738051636e-05,
+      "loss": 1.9689,
+      "step": 3927
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.528711944747505,
+      "learning_rate": 4.084609629267057e-05,
+      "loss": 1.9608,
+      "step": 3928
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6366696275030724,
+      "learning_rate": 4.08412642154879e-05,
+      "loss": 2.2045,
+      "step": 3929
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5767927321291471,
+      "learning_rate": 4.083643114927001e-05,
+      "loss": 2.0507,
+      "step": 3930
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6262810275058607,
+      "learning_rate": 4.083159709431867e-05,
+      "loss": 1.9761,
+      "step": 3931
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.603558159394882,
+      "learning_rate": 4.08267620509357e-05,
+      "loss": 1.9745,
+      "step": 3932
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5701235201637165,
+      "learning_rate": 4.0821926019422954e-05,
+      "loss": 1.9826,
+      "step": 3933
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5999710044862449,
+      "learning_rate": 4.0817089000082354e-05,
+      "loss": 2.1976,
+      "step": 3934
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6436869824710597,
+      "learning_rate": 4.081225099321592e-05,
+      "loss": 2.0037,
+      "step": 3935
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5418557418995505,
+      "learning_rate": 4.08074119991257e-05,
+      "loss": 2.0913,
+      "step": 3936
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.652760146539893,
+      "learning_rate": 4.0802572018113804e-05,
+      "loss": 1.9648,
+      "step": 3937
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.615856227564425,
+      "learning_rate": 4.0797731050482414e-05,
+      "loss": 2.1223,
+      "step": 3938
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5337244278544372,
+      "learning_rate": 4.079288909653378e-05,
+      "loss": 1.9399,
+      "step": 3939
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6072894525284867,
+      "learning_rate": 4.078804615657021e-05,
+      "loss": 2.0075,
+      "step": 3940
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6463664617478929,
+      "learning_rate": 4.078320223089406e-05,
+      "loss": 1.9656,
+      "step": 3941
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6053858812193568,
+      "learning_rate": 4.077835731980774e-05,
+      "loss": 2.1589,
+      "step": 3942
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6018016711849219,
+      "learning_rate": 4.077351142361376e-05,
+      "loss": 1.977,
+      "step": 3943
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5989532290503011,
+      "learning_rate": 4.076866454261466e-05,
+      "loss": 1.9794,
+      "step": 3944
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6046932673990401,
+      "learning_rate": 4.0763816677113064e-05,
+      "loss": 1.993,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6115075893355694,
+      "learning_rate": 4.075896782741163e-05,
+      "loss": 2.1672,
+      "step": 3946
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7013460309228474,
+      "learning_rate": 4.075411799381309e-05,
+      "loss": 1.9498,
+      "step": 3947
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5536393500853389,
+      "learning_rate": 4.074926717662025e-05,
+      "loss": 1.9795,
+      "step": 3948
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5944213003659885,
+      "learning_rate": 4.074441537613596e-05,
+      "loss": 2.0185,
+      "step": 3949
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6489011507801296,
+      "learning_rate": 4.0739562592663134e-05,
+      "loss": 2.1803,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5330079028742769,
+      "learning_rate": 4.0734708826504755e-05,
+      "loss": 1.9274,
+      "step": 3951
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.611252372527336,
+      "learning_rate": 4.0729854077963866e-05,
+      "loss": 1.965,
+      "step": 3952
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6750862461291328,
+      "learning_rate": 4.072499834734357e-05,
+      "loss": 1.9219,
+      "step": 3953
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.655206426453125,
+      "learning_rate": 4.072014163494702e-05,
+      "loss": 2.1192,
+      "step": 3954
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7161222534966286,
+      "learning_rate": 4.071528394107746e-05,
+      "loss": 1.9455,
+      "step": 3955
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6284146936861615,
+      "learning_rate": 4.0710425266038145e-05,
+      "loss": 1.9684,
+      "step": 3956
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6844628586151592,
+      "learning_rate": 4.070556561013246e-05,
+      "loss": 2.0016,
+      "step": 3957
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7755004546749505,
+      "learning_rate": 4.070070497366378e-05,
+      "loss": 2.196,
+      "step": 3958
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6326489269355683,
+      "learning_rate": 4.069584335693559e-05,
+      "loss": 1.9699,
+      "step": 3959
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8097240980216724,
+      "learning_rate": 4.0690980760251426e-05,
+      "loss": 1.9669,
+      "step": 3960
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6677787931024276,
+      "learning_rate": 4.068611718391487e-05,
+      "loss": 2.0097,
+      "step": 3961
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.688784473718283,
+      "learning_rate": 4.068125262822958e-05,
+      "loss": 2.1466,
+      "step": 3962
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7263695693723677,
+      "learning_rate": 4.067638709349927e-05,
+      "loss": 1.9492,
+      "step": 3963
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6962450382958119,
+      "learning_rate": 4.06715205800277e-05,
+      "loss": 1.9657,
+      "step": 3964
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7149673577783155,
+      "learning_rate": 4.0666653088118734e-05,
+      "loss": 1.9378,
+      "step": 3965
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8102051381467807,
+      "learning_rate": 4.066178461807625e-05,
+      "loss": 2.1599,
+      "step": 3966
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6126262700724647,
+      "learning_rate": 4.065691517020421e-05,
+      "loss": 2.0774,
+      "step": 3967
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6809055049876596,
+      "learning_rate": 4.065204474480664e-05,
+      "loss": 1.9269,
+      "step": 3968
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6687990448358586,
+      "learning_rate": 4.0647173342187615e-05,
+      "loss": 1.9146,
+      "step": 3969
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.707960760128314,
+      "learning_rate": 4.064230096265128e-05,
+      "loss": 2.161,
+      "step": 3970
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6992339231666369,
+      "learning_rate": 4.0637427606501833e-05,
+      "loss": 1.9466,
+      "step": 3971
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6548835889887717,
+      "learning_rate": 4.063255327404354e-05,
+      "loss": 1.9599,
+      "step": 3972
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6445787414692051,
+      "learning_rate": 4.062767796558073e-05,
+      "loss": 1.9736,
+      "step": 3973
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6183177348038901,
+      "learning_rate": 4.062280168141778e-05,
+      "loss": 2.1858,
+      "step": 3974
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5799925654320878,
+      "learning_rate": 4.0617924421859144e-05,
+      "loss": 1.9523,
+      "step": 3975
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5812799647488383,
+      "learning_rate": 4.061304618720932e-05,
+      "loss": 1.9651,
+      "step": 3976
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6254982255938857,
+      "learning_rate": 4.060816697777289e-05,
+      "loss": 1.8825,
+      "step": 3977
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6802300448778695,
+      "learning_rate": 4.060328679385448e-05,
+      "loss": 2.1185,
+      "step": 3978
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6273730577655596,
+      "learning_rate": 4.059840563575876e-05,
+      "loss": 1.9549,
+      "step": 3979
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5928283508554224,
+      "learning_rate": 4.059352350379051e-05,
+      "loss": 2.005,
+      "step": 3980
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6396703254040137,
+      "learning_rate": 4.058864039825452e-05,
+      "loss": 1.9481,
+      "step": 3981
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.633288209867391,
+      "learning_rate": 4.058375631945568e-05,
+      "loss": 2.1884,
+      "step": 3982
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5722078112696001,
+      "learning_rate": 4.05788712676989e-05,
+      "loss": 1.9676,
+      "step": 3983
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.673606589168583,
+      "learning_rate": 4.0573985243289195e-05,
+      "loss": 1.9897,
+      "step": 3984
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6106146849179724,
+      "learning_rate": 4.0569098246531615e-05,
+      "loss": 1.9488,
+      "step": 3985
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6830378494852114,
+      "learning_rate": 4.056421027773126e-05,
+      "loss": 2.1669,
+      "step": 3986
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5848723800476755,
+      "learning_rate": 4.055932133719333e-05,
+      "loss": 2.0048,
+      "step": 3987
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.56868799365382,
+      "learning_rate": 4.0554431425223046e-05,
+      "loss": 1.9433,
+      "step": 3988
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6174129913622555,
+      "learning_rate": 4.054954054212571e-05,
+      "loss": 2.0008,
+      "step": 3989
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6788571167867056,
+      "learning_rate": 4.054464868820667e-05,
+      "loss": 2.1301,
+      "step": 3990
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5954529197305675,
+      "learning_rate": 4.0539755863771367e-05,
+      "loss": 1.9696,
+      "step": 3991
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.747791661504132,
+      "learning_rate": 4.0534862069125254e-05,
+      "loss": 2.0345,
+      "step": 3992
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6381212657735547,
+      "learning_rate": 4.052996730457389e-05,
+      "loss": 1.9531,
+      "step": 3993
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7038711363901147,
+      "learning_rate": 4.052507157042287e-05,
+      "loss": 2.1784,
+      "step": 3994
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6146548233511422,
+      "learning_rate": 4.052017486697784e-05,
+      "loss": 1.9686,
+      "step": 3995
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6803219620205003,
+      "learning_rate": 4.0515277194544554e-05,
+      "loss": 1.9474,
+      "step": 3996
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6469660966486128,
+      "learning_rate": 4.051037855342876e-05,
+      "loss": 1.9724,
+      "step": 3997
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7130486245223812,
+      "learning_rate": 4.050547894393633e-05,
+      "loss": 2.1833,
+      "step": 3998
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6113273627151158,
+      "learning_rate": 4.0500578366373134e-05,
+      "loss": 1.9974,
+      "step": 3999
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7685298234374351,
+      "learning_rate": 4.049567682104516e-05,
+      "loss": 1.9503,
+      "step": 4000
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6208536724757183,
+      "learning_rate": 4.049077430825843e-05,
+      "loss": 2.0512,
+      "step": 4001
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8181386935703927,
+      "learning_rate": 4.048587082831901e-05,
+      "loss": 2.1764,
+      "step": 4002
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6101616068591584,
+      "learning_rate": 4.048096638153306e-05,
+      "loss": 1.9338,
+      "step": 4003
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6578592184641787,
+      "learning_rate": 4.047606096820679e-05,
+      "loss": 1.9506,
+      "step": 4004
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7672901417031481,
+      "learning_rate": 4.0471154588646455e-05,
+      "loss": 2.0373,
+      "step": 4005
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6933477230872233,
+      "learning_rate": 4.0466247243158376e-05,
+      "loss": 2.1908,
+      "step": 4006
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8540479356400489,
+      "learning_rate": 4.046133893204894e-05,
+      "loss": 1.9528,
+      "step": 4007
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6878251296685576,
+      "learning_rate": 4.04564296556246e-05,
+      "loss": 1.9557,
+      "step": 4008
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7105579766739942,
+      "learning_rate": 4.045151941419188e-05,
+      "loss": 1.9516,
+      "step": 4009
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8189345288667786,
+      "learning_rate": 4.044660820805731e-05,
+      "loss": 2.149,
+      "step": 4010
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5344511898656646,
+      "learning_rate": 4.044169603752753e-05,
+      "loss": 2.0342,
+      "step": 4011
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6863763187420581,
+      "learning_rate": 4.0436782902909234e-05,
+      "loss": 1.9208,
+      "step": 4012
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.725220391489797,
+      "learning_rate": 4.043186880450916e-05,
+      "loss": 1.9492,
+      "step": 4013
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6705762837053068,
+      "learning_rate": 4.042695374263413e-05,
+      "loss": 2.2205,
+      "step": 4014
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8170922709142047,
+      "learning_rate": 4.0422037717591e-05,
+      "loss": 1.9958,
+      "step": 4015
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6541893670330549,
+      "learning_rate": 4.0417120729686694e-05,
+      "loss": 1.9424,
+      "step": 4016
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8970390296533451,
+      "learning_rate": 4.041220277922822e-05,
+      "loss": 2.0265,
+      "step": 4017
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7807026077532898,
+      "learning_rate": 4.040728386652261e-05,
+      "loss": 2.1844,
+      "step": 4018
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7573740902493367,
+      "learning_rate": 4.040236399187696e-05,
+      "loss": 2.013,
+      "step": 4019
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.833450128425402,
+      "learning_rate": 4.039744315559846e-05,
+      "loss": 1.9287,
+      "step": 4020
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6706700585675034,
+      "learning_rate": 4.0392521357994315e-05,
+      "loss": 1.9206,
+      "step": 4021
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6587641498120164,
+      "learning_rate": 4.0387598599371844e-05,
+      "loss": 2.1812,
+      "step": 4022
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7913950435021273,
+      "learning_rate": 4.0382674880038377e-05,
+      "loss": 2.0243,
+      "step": 4023
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7083470022022188,
+      "learning_rate": 4.0377750200301325e-05,
+      "loss": 1.9828,
+      "step": 4024
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7614277294632309,
+      "learning_rate": 4.037282456046815e-05,
+      "loss": 1.9319,
+      "step": 4025
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7367759670715945,
+      "learning_rate": 4.036789796084638e-05,
+      "loss": 2.1518,
+      "step": 4026
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8088455260098338,
+      "learning_rate": 4.036297040174362e-05,
+      "loss": 1.9458,
+      "step": 4027
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6464907246155637,
+      "learning_rate": 4.035804188346749e-05,
+      "loss": 1.9363,
+      "step": 4028
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7437880248343192,
+      "learning_rate": 4.035311240632572e-05,
+      "loss": 2.0191,
+      "step": 4029
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7913735731501759,
+      "learning_rate": 4.0348181970626075e-05,
+      "loss": 2.1226,
+      "step": 4030
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6486981255868638,
+      "learning_rate": 4.0343250576676375e-05,
+      "loss": 1.9796,
+      "step": 4031
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6713058533336963,
+      "learning_rate": 4.03383182247845e-05,
+      "loss": 1.9512,
+      "step": 4032
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6447257924957461,
+      "learning_rate": 4.033338491525842e-05,
+      "loss": 1.9843,
+      "step": 4033
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6190412919800601,
+      "learning_rate": 4.032845064840612e-05,
+      "loss": 2.1511,
+      "step": 4034
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5563706820721747,
+      "learning_rate": 4.032351542453569e-05,
+      "loss": 1.9204,
+      "step": 4035
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6621936152872246,
+      "learning_rate": 4.0318579243955226e-05,
+      "loss": 2.0138,
+      "step": 4036
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5951986835296329,
+      "learning_rate": 4.0313642106972924e-05,
+      "loss": 1.8928,
+      "step": 4037
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7612010858730531,
+      "learning_rate": 4.0308704013897046e-05,
+      "loss": 2.1399,
+      "step": 4038
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6335425540729737,
+      "learning_rate": 4.030376496503589e-05,
+      "loss": 1.9681,
+      "step": 4039
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6512516302522636,
+      "learning_rate": 4.029882496069781e-05,
+      "loss": 1.8656,
+      "step": 4040
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6668516782976716,
+      "learning_rate": 4.029388400119124e-05,
+      "loss": 1.9722,
+      "step": 4041
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6543969273058352,
+      "learning_rate": 4.0288942086824666e-05,
+      "loss": 2.0603,
+      "step": 4042
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7095261547541843,
+      "learning_rate": 4.028399921790663e-05,
+      "loss": 2.111,
+      "step": 4043
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7275530170920769,
+      "learning_rate": 4.0279055394745735e-05,
+      "loss": 1.955,
+      "step": 4044
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6040600540830103,
+      "learning_rate": 4.027411061765064e-05,
+      "loss": 1.9669,
+      "step": 4045
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8112563656138503,
+      "learning_rate": 4.0269164886930075e-05,
+      "loss": 2.1463,
+      "step": 4046
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6998928681235164,
+      "learning_rate": 4.026421820289281e-05,
+      "loss": 1.9192,
+      "step": 4047
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.765060021458267,
+      "learning_rate": 4.025927056584771e-05,
+      "loss": 2.0376,
+      "step": 4048
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6231500733024301,
+      "learning_rate": 4.025432197610365e-05,
+      "loss": 1.9835,
+      "step": 4049
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6408508241256927,
+      "learning_rate": 4.0249372433969604e-05,
+      "loss": 2.1321,
+      "step": 4050
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6238985287983541,
+      "learning_rate": 4.0244421939754596e-05,
+      "loss": 1.9963,
+      "step": 4051
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6009776238444348,
+      "learning_rate": 4.0239470493767704e-05,
+      "loss": 1.9228,
+      "step": 4052
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6494780339165497,
+      "learning_rate": 4.023451809631805e-05,
+      "loss": 1.9753,
+      "step": 4053
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5178308636558642,
+      "learning_rate": 4.022956474771486e-05,
+      "loss": 2.0482,
+      "step": 4054
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7089063311765201,
+      "learning_rate": 4.0224610448267374e-05,
+      "loss": 2.1219,
+      "step": 4055
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6881296990431022,
+      "learning_rate": 4.021965519828491e-05,
+      "loss": 1.9831,
+      "step": 4056
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5829662410557068,
+      "learning_rate": 4.021469899807685e-05,
+      "loss": 1.9113,
+      "step": 4057
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7235164656320588,
+      "learning_rate": 4.020974184795262e-05,
+      "loss": 2.1464,
+      "step": 4058
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.613069299045964,
+      "learning_rate": 4.020478374822174e-05,
+      "loss": 1.9108,
+      "step": 4059
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.566183753466578,
+      "learning_rate": 4.019982469919374e-05,
+      "loss": 2.0192,
+      "step": 4060
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6621362930446608,
+      "learning_rate": 4.0194864701178236e-05,
+      "loss": 1.9858,
+      "step": 4061
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5832710915149316,
+      "learning_rate": 4.0189903754484916e-05,
+      "loss": 2.134,
+      "step": 4062
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5300774396538228,
+      "learning_rate": 4.0184941859423495e-05,
+      "loss": 1.9766,
+      "step": 4063
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.594071852916594,
+      "learning_rate": 4.017997901630378e-05,
+      "loss": 1.9055,
+      "step": 4064
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6664524059594977,
+      "learning_rate": 4.0175015225435605e-05,
+      "loss": 1.9479,
+      "step": 4065
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5964200763707886,
+      "learning_rate": 4.01700504871289e-05,
+      "loss": 2.1945,
+      "step": 4066
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6767975939423628,
+      "learning_rate": 4.016508480169361e-05,
+      "loss": 2.0171,
+      "step": 4067
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6941159025302919,
+      "learning_rate": 4.0160118169439784e-05,
+      "loss": 1.9541,
+      "step": 4068
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6975722077131571,
+      "learning_rate": 4.0155150590677505e-05,
+      "loss": 1.931,
+      "step": 4069
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6624950585239348,
+      "learning_rate": 4.015018206571691e-05,
+      "loss": 2.149,
+      "step": 4070
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6328763987781628,
+      "learning_rate": 4.0145212594868206e-05,
+      "loss": 1.9379,
+      "step": 4071
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6509446988740296,
+      "learning_rate": 4.014024217844167e-05,
+      "loss": 1.9456,
+      "step": 4072
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.701774793259767,
+      "learning_rate": 4.0135270816747616e-05,
+      "loss": 1.9844,
+      "step": 4073
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5660466799537208,
+      "learning_rate": 4.013029851009642e-05,
+      "loss": 1.9132,
+      "step": 4074
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9094343887196557,
+      "learning_rate": 4.012532525879854e-05,
+      "loss": 2.2041,
+      "step": 4075
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6423610864906237,
+      "learning_rate": 4.012035106316446e-05,
+      "loss": 1.9683,
+      "step": 4076
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7225926440714548,
+      "learning_rate": 4.011537592350475e-05,
+      "loss": 1.9301,
+      "step": 4077
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5885874847862682,
+      "learning_rate": 4.0110399840130034e-05,
+      "loss": 2.1384,
+      "step": 4078
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.591280796075023,
+      "learning_rate": 4.010542281335097e-05,
+      "loss": 2.0321,
+      "step": 4079
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6572687791010098,
+      "learning_rate": 4.0100444843478316e-05,
+      "loss": 1.9494,
+      "step": 4080
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7116668405227692,
+      "learning_rate": 4.009546593082284e-05,
+      "loss": 1.9259,
+      "step": 4081
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6647111284321758,
+      "learning_rate": 4.0090486075695425e-05,
+      "loss": 2.181,
+      "step": 4082
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5488426282994372,
+      "learning_rate": 4.0085505278406965e-05,
+      "loss": 1.9043,
+      "step": 4083
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6767314701789492,
+      "learning_rate": 4.0080523539268445e-05,
+      "loss": 1.9771,
+      "step": 4084
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5424138911592522,
+      "learning_rate": 4.0075540858590883e-05,
+      "loss": 2.0103,
+      "step": 4085
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6156755671808553,
+      "learning_rate": 4.007055723668538e-05,
+      "loss": 1.9421,
+      "step": 4086
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5763372860269943,
+      "learning_rate": 4.006557267386306e-05,
+      "loss": 2.1497,
+      "step": 4087
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5724485197989426,
+      "learning_rate": 4.006058717043516e-05,
+      "loss": 1.9518,
+      "step": 4088
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6776327466516002,
+      "learning_rate": 4.005560072671293e-05,
+      "loss": 1.9684,
+      "step": 4089
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5891820338573822,
+      "learning_rate": 4.00506133430077e-05,
+      "loss": 2.1816,
+      "step": 4090
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5833905310001206,
+      "learning_rate": 4.004562501963085e-05,
+      "loss": 2.0274,
+      "step": 4091
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5584142273382847,
+      "learning_rate": 4.0040635756893815e-05,
+      "loss": 1.9575,
+      "step": 4092
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6308651916216067,
+      "learning_rate": 4.00356455551081e-05,
+      "loss": 1.9465,
+      "step": 4093
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7290515560126466,
+      "learning_rate": 4.0030654414585274e-05,
+      "loss": 2.1995,
+      "step": 4094
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5546197178194088,
+      "learning_rate": 4.0025662335636935e-05,
+      "loss": 1.9544,
+      "step": 4095
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7673565137874488,
+      "learning_rate": 4.002066931857478e-05,
+      "loss": 1.9614,
+      "step": 4096
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6278296844606506,
+      "learning_rate": 4.001567536371053e-05,
+      "loss": 1.976,
+      "step": 4097
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.726820248686488,
+      "learning_rate": 4.001068047135598e-05,
+      "loss": 2.0045,
+      "step": 4098
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7171819765104391,
+      "learning_rate": 4.000568464182298e-05,
+      "loss": 2.174,
+      "step": 4099
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6500659507829263,
+      "learning_rate": 4.0000687875423434e-05,
+      "loss": 1.9581,
+      "step": 4100
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6972264706245118,
+      "learning_rate": 3.999569017246934e-05,
+      "loss": 1.9549,
+      "step": 4101
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5898157219807859,
+      "learning_rate": 3.99906915332727e-05,
+      "loss": 2.1611,
+      "step": 4102
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6506849130612727,
+      "learning_rate": 3.998569195814559e-05,
+      "loss": 1.9515,
+      "step": 4103
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6667674551883965,
+      "learning_rate": 3.998069144740018e-05,
+      "loss": 2.0535,
+      "step": 4104
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6384655034348754,
+      "learning_rate": 3.9975690001348656e-05,
+      "loss": 1.9372,
+      "step": 4105
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7788984270879734,
+      "learning_rate": 3.997068762030328e-05,
+      "loss": 2.0143,
+      "step": 4106
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.563719882087236,
+      "learning_rate": 3.996568430457639e-05,
+      "loss": 2.1671,
+      "step": 4107
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7170814891339737,
+      "learning_rate": 3.9960680054480336e-05,
+      "loss": 1.9732,
+      "step": 4108
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6033591562490503,
+      "learning_rate": 3.995567487032756e-05,
+      "loss": 1.9629,
+      "step": 4109
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5668281568158399,
+      "learning_rate": 3.9950668752430584e-05,
+      "loss": 1.9932,
+      "step": 4110
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7483063348298942,
+      "learning_rate": 3.994566170110192e-05,
+      "loss": 2.1169,
+      "step": 4111
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6168689571475807,
+      "learning_rate": 3.994065371665421e-05,
+      "loss": 1.9824,
+      "step": 4112
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7382031205135126,
+      "learning_rate": 3.99356447994001e-05,
+      "loss": 1.9817,
+      "step": 4113
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8086618970629478,
+      "learning_rate": 3.993063494965234e-05,
+      "loss": 2.1664,
+      "step": 4114
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.639274337963226,
+      "learning_rate": 3.9925624167723686e-05,
+      "loss": 2.0025,
+      "step": 4115
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7961330145395729,
+      "learning_rate": 3.9920612453927014e-05,
+      "loss": 2.0132,
+      "step": 4116
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6386178915926113,
+      "learning_rate": 3.9915599808575204e-05,
+      "loss": 1.9651,
+      "step": 4117
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.716198739890325,
+      "learning_rate": 3.991058623198123e-05,
+      "loss": 1.9417,
+      "step": 4118
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8806786422332284,
+      "learning_rate": 3.99055717244581e-05,
+      "loss": 2.2179,
+      "step": 4119
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6563777969517502,
+      "learning_rate": 3.990055628631889e-05,
+      "loss": 1.9123,
+      "step": 4120
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8316082380364341,
+      "learning_rate": 3.9895539917876734e-05,
+      "loss": 2.0081,
+      "step": 4121
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6531517471364706,
+      "learning_rate": 3.989052261944484e-05,
+      "loss": 2.1482,
+      "step": 4122
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9491221114104248,
+      "learning_rate": 3.9885504391336446e-05,
+      "loss": 2.0655,
+      "step": 4123
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5969032576001848,
+      "learning_rate": 3.988048523386485e-05,
+      "loss": 1.9306,
+      "step": 4124
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6930020906973378,
+      "learning_rate": 3.987546514734345e-05,
+      "loss": 1.8648,
+      "step": 4125
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7359572301489702,
+      "learning_rate": 3.9870444132085624e-05,
+      "loss": 2.1703,
+      "step": 4126
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6229532862650323,
+      "learning_rate": 3.98654221884049e-05,
+      "loss": 1.9508,
+      "step": 4127
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6771936719024844,
+      "learning_rate": 3.9860399316614796e-05,
+      "loss": 1.9391,
+      "step": 4128
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6030715902413544,
+      "learning_rate": 3.985537551702892e-05,
+      "loss": 2.0164,
+      "step": 4129
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5462154005934228,
+      "learning_rate": 3.9850350789960915e-05,
+      "loss": 1.9326,
+      "step": 4130
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.618767830556047,
+      "learning_rate": 3.984532513572451e-05,
+      "loss": 2.1289,
+      "step": 4131
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.568389126318579,
+      "learning_rate": 3.984029855463347e-05,
+      "loss": 1.9101,
+      "step": 4132
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6222226575696356,
+      "learning_rate": 3.983527104700162e-05,
+      "loss": 1.968,
+      "step": 4133
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6016777030481472,
+      "learning_rate": 3.9830242613142856e-05,
+      "loss": 2.1575,
+      "step": 4134
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6511502054969768,
+      "learning_rate": 3.982521325337112e-05,
+      "loss": 2.0416,
+      "step": 4135
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6534986331541173,
+      "learning_rate": 3.982018296800042e-05,
+      "loss": 1.9462,
+      "step": 4136
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6222479396519502,
+      "learning_rate": 3.981515175734481e-05,
+      "loss": 1.9251,
+      "step": 4137
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6610825183974631,
+      "learning_rate": 3.981011962171842e-05,
+      "loss": 1.9168,
+      "step": 4138
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6508214780388522,
+      "learning_rate": 3.9805086561435424e-05,
+      "loss": 2.1497,
+      "step": 4139
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6701706073600894,
+      "learning_rate": 3.9800052576810044e-05,
+      "loss": 1.9367,
+      "step": 4140
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6906092533384679,
+      "learning_rate": 3.9795017668156584e-05,
+      "loss": 2.0669,
+      "step": 4141
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6498540114859206,
+      "learning_rate": 3.978998183578939e-05,
+      "loss": 2.0183,
+      "step": 4142
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6682406019479503,
+      "learning_rate": 3.978494508002287e-05,
+      "loss": 2.1736,
+      "step": 4143
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5991606630291433,
+      "learning_rate": 3.97799074011715e-05,
+      "loss": 1.9409,
+      "step": 4144
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.679494942795775,
+      "learning_rate": 3.977486879954979e-05,
+      "loss": 1.9034,
+      "step": 4145
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6458545376487609,
+      "learning_rate": 3.976982927547232e-05,
+      "loss": 2.1579,
+      "step": 4146
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5785166783159018,
+      "learning_rate": 3.976478882925373e-05,
+      "loss": 2.0051,
+      "step": 4147
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6818663488915552,
+      "learning_rate": 3.975974746120872e-05,
+      "loss": 1.9113,
+      "step": 4148
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6200890306013992,
+      "learning_rate": 3.975470517165205e-05,
+      "loss": 1.9861,
+      "step": 4149
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6172356514311024,
+      "learning_rate": 3.974966196089851e-05,
+      "loss": 2.0036,
+      "step": 4150
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6667376797902171,
+      "learning_rate": 3.974461782926299e-05,
+      "loss": 2.1371,
+      "step": 4151
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5861910524732865,
+      "learning_rate": 3.97395727770604e-05,
+      "loss": 1.9552,
+      "step": 4152
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5958368244158986,
+      "learning_rate": 3.973452680460574e-05,
+      "loss": 1.9543,
+      "step": 4153
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5202759880150478,
+      "learning_rate": 3.972947991221403e-05,
+      "loss": 2.0163,
+      "step": 4154
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7141005077994359,
+      "learning_rate": 3.972443210020038e-05,
+      "loss": 2.1162,
+      "step": 4155
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6307918856221736,
+      "learning_rate": 3.9719383368879946e-05,
+      "loss": 1.924,
+      "step": 4156
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6467399845723995,
+      "learning_rate": 3.971433371856794e-05,
+      "loss": 1.9376,
+      "step": 4157
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6559244853490612,
+      "learning_rate": 3.970928314957963e-05,
+      "loss": 1.9626,
+      "step": 4158
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6787886862581384,
+      "learning_rate": 3.970423166223035e-05,
+      "loss": 2.1393,
+      "step": 4159
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6743291378837947,
+      "learning_rate": 3.9699179256835484e-05,
+      "loss": 2.0393,
+      "step": 4160
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6788849442685472,
+      "learning_rate": 3.9694125933710464e-05,
+      "loss": 1.9498,
+      "step": 4161
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.689141391814546,
+      "learning_rate": 3.96890716931708e-05,
+      "loss": 1.9654,
+      "step": 4162
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8747882560307242,
+      "learning_rate": 3.968401653553204e-05,
+      "loss": 2.1416,
+      "step": 4163
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5828133104793877,
+      "learning_rate": 3.9678960461109816e-05,
+      "loss": 1.9815,
+      "step": 4164
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6578941699912043,
+      "learning_rate": 3.967390347021978e-05,
+      "loss": 1.9491,
+      "step": 4165
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6690431374311795,
+      "learning_rate": 3.966884556317767e-05,
+      "loss": 2.0695,
+      "step": 4166
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5929844733012585,
+      "learning_rate": 3.966378674029927e-05,
+      "loss": 2.1407,
+      "step": 4167
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.736529336430372,
+      "learning_rate": 3.965872700190042e-05,
+      "loss": 1.9762,
+      "step": 4168
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6092250471496666,
+      "learning_rate": 3.9653666348297024e-05,
+      "loss": 1.8881,
+      "step": 4169
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6962698082383253,
+      "learning_rate": 3.964860477980504e-05,
+      "loss": 2.0023,
+      "step": 4170
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.574335469008907,
+      "learning_rate": 3.9643542296740486e-05,
+      "loss": 2.142,
+      "step": 4171
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.685396584835887,
+      "learning_rate": 3.963847889941943e-05,
+      "loss": 2.0115,
+      "step": 4172
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5942806629623203,
+      "learning_rate": 3.9633414588158e-05,
+      "loss": 1.9602,
+      "step": 4173
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6221320088094293,
+      "learning_rate": 3.9628349363272375e-05,
+      "loss": 1.9347,
+      "step": 4174
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6190268661794806,
+      "learning_rate": 3.962328322507881e-05,
+      "loss": 2.1584,
+      "step": 4175
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5854184979955704,
+      "learning_rate": 3.96182161738936e-05,
+      "loss": 1.9456,
+      "step": 4176
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6761812111916473,
+      "learning_rate": 3.961314821003309e-05,
+      "loss": 1.9894,
+      "step": 4177
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6677913233325974,
+      "learning_rate": 3.960807933381372e-05,
+      "loss": 2.0588,
+      "step": 4178
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6302217176351721,
+      "learning_rate": 3.960300954555194e-05,
+      "loss": 2.1687,
+      "step": 4179
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6817323947918734,
+      "learning_rate": 3.959793884556428e-05,
+      "loss": 2.017,
+      "step": 4180
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5683292709927948,
+      "learning_rate": 3.959286723416733e-05,
+      "loss": 1.92,
+      "step": 4181
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7522464788947631,
+      "learning_rate": 3.958779471167773e-05,
+      "loss": 1.9146,
+      "step": 4182
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5944227507167528,
+      "learning_rate": 3.958272127841218e-05,
+      "loss": 2.1575,
+      "step": 4183
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.611117292103314,
+      "learning_rate": 3.957764693468743e-05,
+      "loss": 1.9312,
+      "step": 4184
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5486262008145932,
+      "learning_rate": 3.95725716808203e-05,
+      "loss": 2.055,
+      "step": 4185
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5441369815693254,
+      "learning_rate": 3.956749551712765e-05,
+      "loss": 1.9596,
+      "step": 4186
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7913050312491605,
+      "learning_rate": 3.9562418443926405e-05,
+      "loss": 2.1651,
+      "step": 4187
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5654104549632992,
+      "learning_rate": 3.9557340461533566e-05,
+      "loss": 1.9722,
+      "step": 4188
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6621842290953107,
+      "learning_rate": 3.955226157026615e-05,
+      "loss": 1.9271,
+      "step": 4189
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5553873429162748,
+      "learning_rate": 3.954718177044126e-05,
+      "loss": 1.975,
+      "step": 4190
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5891143605088437,
+      "learning_rate": 3.954210106237606e-05,
+      "loss": 2.21,
+      "step": 4191
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5878468753216854,
+      "learning_rate": 3.953701944638775e-05,
+      "loss": 1.959,
+      "step": 4192
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5670275086268481,
+      "learning_rate": 3.953193692279359e-05,
+      "loss": 1.9134,
+      "step": 4193
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6460113726730338,
+      "learning_rate": 3.95268534919109e-05,
+      "loss": 1.9846,
+      "step": 4194
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.616892812174001,
+      "learning_rate": 3.9521769154057085e-05,
+      "loss": 2.0986,
+      "step": 4195
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5844693900956399,
+      "learning_rate": 3.9516683909549556e-05,
+      "loss": 1.9758,
+      "step": 4196
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7695225825371955,
+      "learning_rate": 3.9511597758705823e-05,
+      "loss": 2.0918,
+      "step": 4197
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6108845788741807,
+      "learning_rate": 3.9506510701843415e-05,
+      "loss": 1.9384,
+      "step": 4198
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7358984466156699,
+      "learning_rate": 3.9501422739279956e-05,
+      "loss": 2.1465,
+      "step": 4199
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5858954601419409,
+      "learning_rate": 3.94963338713331e-05,
+      "loss": 1.9532,
+      "step": 4200
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6010029902198067,
+      "learning_rate": 3.9491244098320575e-05,
+      "loss": 1.9341,
+      "step": 4201
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7187548294193934,
+      "learning_rate": 3.948615342056014e-05,
+      "loss": 2.0105,
+      "step": 4202
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7410447610254437,
+      "learning_rate": 3.948106183836964e-05,
+      "loss": 2.237,
+      "step": 4203
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.773160094178751,
+      "learning_rate": 3.947596935206697e-05,
+      "loss": 1.9106,
+      "step": 4204
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7089167731289175,
+      "learning_rate": 3.9470875961970046e-05,
+      "loss": 1.9779,
+      "step": 4205
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6271017327688269,
+      "learning_rate": 3.94657816683969e-05,
+      "loss": 1.9877,
+      "step": 4206
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7965188562601738,
+      "learning_rate": 3.9460686471665575e-05,
+      "loss": 2.1462,
+      "step": 4207
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5556564766072845,
+      "learning_rate": 3.945559037209419e-05,
+      "loss": 1.9445,
+      "step": 4208
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6317411375553889,
+      "learning_rate": 3.945049337000091e-05,
+      "loss": 2.0155,
+      "step": 4209
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6331097891280774,
+      "learning_rate": 3.944539546570396e-05,
+      "loss": 1.9554,
+      "step": 4210
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5788739374701954,
+      "learning_rate": 3.944029665952163e-05,
+      "loss": 2.1392,
+      "step": 4211
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5807134202646161,
+      "learning_rate": 3.943519695177226e-05,
+      "loss": 1.9856,
+      "step": 4212
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5754240782281697,
+      "learning_rate": 3.943009634277425e-05,
+      "loss": 1.9365,
+      "step": 4213
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6038530614824176,
+      "learning_rate": 3.942499483284603e-05,
+      "loss": 1.9384,
+      "step": 4214
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6271145163137049,
+      "learning_rate": 3.9419892422306135e-05,
+      "loss": 2.1352,
+      "step": 4215
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5607835222965799,
+      "learning_rate": 3.941478911147312e-05,
+      "loss": 2.0255,
+      "step": 4216
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6467699712595358,
+      "learning_rate": 3.940968490066559e-05,
+      "loss": 1.9653,
+      "step": 4217
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5785419738591893,
+      "learning_rate": 3.940457979020224e-05,
+      "loss": 1.9432,
+      "step": 4218
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6162449185767824,
+      "learning_rate": 3.93994737804018e-05,
+      "loss": 2.2019,
+      "step": 4219
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.693963652144824,
+      "learning_rate": 3.9394366871583056e-05,
+      "loss": 1.9089,
+      "step": 4220
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6798087583273534,
+      "learning_rate": 3.938925906406486e-05,
+      "loss": 1.9198,
+      "step": 4221
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5918439343952753,
+      "learning_rate": 3.9384150358166104e-05,
+      "loss": 2.0332,
+      "step": 4222
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8535466206240492,
+      "learning_rate": 3.937904075420575e-05,
+      "loss": 2.1347,
+      "step": 4223
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7366642374484893,
+      "learning_rate": 3.937393025250281e-05,
+      "loss": 2.0028,
+      "step": 4224
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7072022369107425,
+      "learning_rate": 3.936881885337636e-05,
+      "loss": 1.9384,
+      "step": 4225
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7737388429356021,
+      "learning_rate": 3.9363706557145517e-05,
+      "loss": 1.9621,
+      "step": 4226
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6076327274112083,
+      "learning_rate": 3.935859336412946e-05,
+      "loss": 2.1192,
+      "step": 4227
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6815344495596531,
+      "learning_rate": 3.935347927464744e-05,
+      "loss": 2.0016,
+      "step": 4228
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6615853294729421,
+      "learning_rate": 3.934836428901875e-05,
+      "loss": 1.9588,
+      "step": 4229
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6235311926772648,
+      "learning_rate": 3.9343248407562714e-05,
+      "loss": 1.9616,
+      "step": 4230
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6510289824220373,
+      "learning_rate": 3.933813163059877e-05,
+      "loss": 2.1426,
+      "step": 4231
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6026470396838804,
+      "learning_rate": 3.933301395844636e-05,
+      "loss": 1.9037,
+      "step": 4232
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6478324745573237,
+      "learning_rate": 3.932789539142501e-05,
+      "loss": 1.9229,
+      "step": 4233
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6088147320383938,
+      "learning_rate": 3.932277592985429e-05,
+      "loss": 2.0331,
+      "step": 4234
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5955612231736747,
+      "learning_rate": 3.9317655574053834e-05,
+      "loss": 2.1625,
+      "step": 4235
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7205104638984847,
+      "learning_rate": 3.9312534324343315e-05,
+      "loss": 1.9806,
+      "step": 4236
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5683236566141067,
+      "learning_rate": 3.930741218104248e-05,
+      "loss": 1.8869,
+      "step": 4237
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6544598333093086,
+      "learning_rate": 3.9302289144471133e-05,
+      "loss": 1.9651,
+      "step": 4238
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6059811547721828,
+      "learning_rate": 3.9297165214949116e-05,
+      "loss": 2.1661,
+      "step": 4239
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6112877250749854,
+      "learning_rate": 3.929204039279634e-05,
+      "loss": 2.0016,
+      "step": 4240
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6280721444993135,
+      "learning_rate": 3.9286914678332766e-05,
+      "loss": 1.9413,
+      "step": 4241
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6559420218860966,
+      "learning_rate": 3.9281788071878424e-05,
+      "loss": 1.9214,
+      "step": 4242
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7951705025332012,
+      "learning_rate": 3.927666057375338e-05,
+      "loss": 2.1691,
+      "step": 4243
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6606755004220511,
+      "learning_rate": 3.9271532184277756e-05,
+      "loss": 1.9617,
+      "step": 4244
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6690951439912823,
+      "learning_rate": 3.926640290377176e-05,
+      "loss": 1.907,
+      "step": 4245
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5447368601992475,
+      "learning_rate": 3.926127273255562e-05,
+      "loss": 1.8932,
+      "step": 4246
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6504429514132948,
+      "learning_rate": 3.925614167094963e-05,
+      "loss": 2.1912,
+      "step": 4247
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6345062244363715,
+      "learning_rate": 3.925100971927415e-05,
+      "loss": 1.9291,
+      "step": 4248
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5416326667971353,
+      "learning_rate": 3.92458768778496e-05,
+      "loss": 1.9463,
+      "step": 4249
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6624254902130771,
+      "learning_rate": 3.9240743146996425e-05,
+      "loss": 1.9344,
+      "step": 4250
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.654982927006778,
+      "learning_rate": 3.9235608527035165e-05,
+      "loss": 2.1271,
+      "step": 4251
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6756192245296007,
+      "learning_rate": 3.9230473018286373e-05,
+      "loss": 1.9666,
+      "step": 4252
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6637202496845914,
+      "learning_rate": 3.922533662107068e-05,
+      "loss": 2.025,
+      "step": 4253
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7659217657022728,
+      "learning_rate": 3.92201993357088e-05,
+      "loss": 1.9497,
+      "step": 4254
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6193462346349753,
+      "learning_rate": 3.921506116252145e-05,
+      "loss": 2.1023,
+      "step": 4255
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7100720856542366,
+      "learning_rate": 3.920992210182944e-05,
+      "loss": 1.9934,
+      "step": 4256
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5583371440301395,
+      "learning_rate": 3.920478215395361e-05,
+      "loss": 1.9546,
+      "step": 4257
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7031087404109865,
+      "learning_rate": 3.919964131921487e-05,
+      "loss": 1.9732,
+      "step": 4258
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6378499759337778,
+      "learning_rate": 3.919449959793421e-05,
+      "loss": 2.0445,
+      "step": 4259
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6315181914154472,
+      "learning_rate": 3.918935699043261e-05,
+      "loss": 2.1002,
+      "step": 4260
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6398218696219451,
+      "learning_rate": 3.918421349703117e-05,
+      "loss": 1.9075,
+      "step": 4261
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6279144776807133,
+      "learning_rate": 3.9179069118051e-05,
+      "loss": 1.9694,
+      "step": 4262
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6009627840596174,
+      "learning_rate": 3.917392385381331e-05,
+      "loss": 2.1633,
+      "step": 4263
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6551057753180778,
+      "learning_rate": 3.9168777704639316e-05,
+      "loss": 1.9732,
+      "step": 4264
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.662148499968243,
+      "learning_rate": 3.9163630670850326e-05,
+      "loss": 2.0494,
+      "step": 4265
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6359425411127517,
+      "learning_rate": 3.9158482752767686e-05,
+      "loss": 1.9126,
+      "step": 4266
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7941442442133257,
+      "learning_rate": 3.915333395071281e-05,
+      "loss": 2.1518,
+      "step": 4267
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6103810564615236,
+      "learning_rate": 3.914818426500715e-05,
+      "loss": 1.9139,
+      "step": 4268
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8147103784645705,
+      "learning_rate": 3.914303369597222e-05,
+      "loss": 1.9763,
+      "step": 4269
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6973755384124137,
+      "learning_rate": 3.913788224392959e-05,
+      "loss": 1.973,
+      "step": 4270
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7215870957036106,
+      "learning_rate": 3.91327299092009e-05,
+      "loss": 2.1512,
+      "step": 4271
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6969747208710462,
+      "learning_rate": 3.9127576692107825e-05,
+      "loss": 2.0517,
+      "step": 4272
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7328462801748226,
+      "learning_rate": 3.91224225929721e-05,
+      "loss": 1.9495,
+      "step": 4273
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7118597534176863,
+      "learning_rate": 3.911726761211551e-05,
+      "loss": 1.9704,
+      "step": 4274
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6432313818551401,
+      "learning_rate": 3.91121117498599e-05,
+      "loss": 2.1426,
+      "step": 4275
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6427143902515152,
+      "learning_rate": 3.9106955006527196e-05,
+      "loss": 1.8722,
+      "step": 4276
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7940138342593733,
+      "learning_rate": 3.910179738243933e-05,
+      "loss": 1.9619,
+      "step": 4277
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6804206834817563,
+      "learning_rate": 3.909663887791832e-05,
+      "loss": 2.0195,
+      "step": 4278
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8703396583957184,
+      "learning_rate": 3.9091479493286245e-05,
+      "loss": 2.191,
+      "step": 4279
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7819235342186928,
+      "learning_rate": 3.908631922886521e-05,
+      "loss": 1.9707,
+      "step": 4280
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7830730519810698,
+      "learning_rate": 3.90811580849774e-05,
+      "loss": 1.9649,
+      "step": 4281
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8258504872351909,
+      "learning_rate": 3.907599606194503e-05,
+      "loss": 1.9669,
+      "step": 4282
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8381687185617055,
+      "learning_rate": 3.9070833160090415e-05,
+      "loss": 2.1841,
+      "step": 4283
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9962400620273721,
+      "learning_rate": 3.906566937973588e-05,
+      "loss": 1.9938,
+      "step": 4284
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7057815614552411,
+      "learning_rate": 3.906050472120382e-05,
+      "loss": 1.9663,
+      "step": 4285
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9838644196236582,
+      "learning_rate": 3.9055339184816686e-05,
+      "loss": 1.9667,
+      "step": 4286
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6803952165310518,
+      "learning_rate": 3.905017277089698e-05,
+      "loss": 2.1447,
+      "step": 4287
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.8107295805123587,
+      "learning_rate": 3.9045005479767285e-05,
+      "loss": 1.9513,
+      "step": 4288
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6636508748800759,
+      "learning_rate": 3.9039837311750185e-05,
+      "loss": 1.9765,
+      "step": 4289
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6921984582453629,
+      "learning_rate": 3.9034668267168374e-05,
+      "loss": 2.0056,
+      "step": 4290
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6569365539144029,
+      "learning_rate": 3.902949834634456e-05,
+      "loss": 1.9962,
+      "step": 4291
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5861950974921066,
+      "learning_rate": 3.902432754960153e-05,
+      "loss": 2.1495,
+      "step": 4292
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.770440891264291,
+      "learning_rate": 3.901915587726212e-05,
+      "loss": 1.9043,
+      "step": 4293
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5990832208443667,
+      "learning_rate": 3.901398332964922e-05,
+      "loss": 1.9989,
+      "step": 4294
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6385444299617757,
+      "learning_rate": 3.900880990708576e-05,
+      "loss": 2.1644,
+      "step": 4295
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6277662409168571,
+      "learning_rate": 3.9003635609894753e-05,
+      "loss": 2.0426,
+      "step": 4296
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6105182656034237,
+      "learning_rate": 3.8998460438399256e-05,
+      "loss": 1.9699,
+      "step": 4297
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7437269659624429,
+      "learning_rate": 3.8993284392922356e-05,
+      "loss": 1.9393,
+      "step": 4298
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6414177949379191,
+      "learning_rate": 3.898810747378723e-05,
+      "loss": 2.1556,
+      "step": 4299
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6308444532499686,
+      "learning_rate": 3.8982929681317085e-05,
+      "loss": 1.9512,
+      "step": 4300
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6251334020427104,
+      "learning_rate": 3.8977751015835204e-05,
+      "loss": 1.9447,
+      "step": 4301
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6893460225896615,
+      "learning_rate": 3.89725714776649e-05,
+      "loss": 1.9364,
+      "step": 4302
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6024077758931757,
+      "learning_rate": 3.8967391067129554e-05,
+      "loss": 1.9991,
+      "step": 4303
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.653707150620456,
+      "learning_rate": 3.8962209784552615e-05,
+      "loss": 2.1409,
+      "step": 4304
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6336298885563304,
+      "learning_rate": 3.8957027630257545e-05,
+      "loss": 1.9564,
+      "step": 4305
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6230794956562666,
+      "learning_rate": 3.895184460456792e-05,
+      "loss": 1.9522,
+      "step": 4306
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6579539485605063,
+      "learning_rate": 3.8946660707807316e-05,
+      "loss": 2.1485,
+      "step": 4307
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5840641715645104,
+      "learning_rate": 3.894147594029939e-05,
+      "loss": 2.0345,
+      "step": 4308
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6744405857109206,
+      "learning_rate": 3.893629030236784e-05,
+      "loss": 2.0084,
+      "step": 4309
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6286397086298595,
+      "learning_rate": 3.8931103794336445e-05,
+      "loss": 1.9664,
+      "step": 4310
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7740672559762488,
+      "learning_rate": 3.8925916416529e-05,
+      "loss": 2.1496,
+      "step": 4311
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6060055357519761,
+      "learning_rate": 3.8920728169269395e-05,
+      "loss": 1.9364,
+      "step": 4312
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.619143471942503,
+      "learning_rate": 3.891553905288153e-05,
+      "loss": 1.9894,
+      "step": 4313
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6153381781409927,
+      "learning_rate": 3.89103490676894e-05,
+      "loss": 1.9647,
+      "step": 4314
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6093023878888556,
+      "learning_rate": 3.8905158214017036e-05,
+      "loss": 2.0394,
+      "step": 4315
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6673039096948462,
+      "learning_rate": 3.889996649218852e-05,
+      "loss": 2.1217,
+      "step": 4316
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5923850916158117,
+      "learning_rate": 3.889477390252799e-05,
+      "loss": 1.9286,
+      "step": 4317
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.66482560773539,
+      "learning_rate": 3.8889580445359644e-05,
+      "loss": 1.9539,
+      "step": 4318
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7097007569225484,
+      "learning_rate": 3.8884386121007735e-05,
+      "loss": 2.1569,
+      "step": 4319
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.648878996987364,
+      "learning_rate": 3.887919092979656e-05,
+      "loss": 1.9512,
+      "step": 4320
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7942225475166459,
+      "learning_rate": 3.8873994872050476e-05,
+      "loss": 2.0265,
+      "step": 4321
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6908363231555014,
+      "learning_rate": 3.8868797948093907e-05,
+      "loss": 1.9484,
+      "step": 4322
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7984410250597462,
+      "learning_rate": 3.88636001582513e-05,
+      "loss": 1.9443,
+      "step": 4323
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7372637636973678,
+      "learning_rate": 3.8858401502847184e-05,
+      "loss": 2.1734,
+      "step": 4324
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7212128789613863,
+      "learning_rate": 3.885320198220613e-05,
+      "loss": 1.9871,
+      "step": 4325
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5613039960028418,
+      "learning_rate": 3.884800159665276e-05,
+      "loss": 1.9985,
+      "step": 4326
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5810105846741793,
+      "learning_rate": 3.884280034651177e-05,
+      "loss": 2.0023,
+      "step": 4327
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.603984859952067,
+      "learning_rate": 3.8837598232107875e-05,
+      "loss": 2.1717,
+      "step": 4328
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5867783536452,
+      "learning_rate": 3.8832395253765885e-05,
+      "loss": 1.9714,
+      "step": 4329
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6227137080868063,
+      "learning_rate": 3.882719141181063e-05,
+      "loss": 1.9408,
+      "step": 4330
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5776335378307313,
+      "learning_rate": 3.882198670656702e-05,
+      "loss": 2.1308,
+      "step": 4331
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5850417625810728,
+      "learning_rate": 3.881678113835999e-05,
+      "loss": 1.9955,
+      "step": 4332
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.575204070898516,
+      "learning_rate": 3.8811574707514554e-05,
+      "loss": 1.8894,
+      "step": 4333
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5405457335076372,
+      "learning_rate": 3.880636741435577e-05,
+      "loss": 2.0448,
+      "step": 4334
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6162480331095656,
+      "learning_rate": 3.880115925920875e-05,
+      "loss": 1.9569,
+      "step": 4335
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6264051050997514,
+      "learning_rate": 3.879595024239866e-05,
+      "loss": 2.1405,
+      "step": 4336
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5916383912221922,
+      "learning_rate": 3.879074036425073e-05,
+      "loss": 1.9364,
+      "step": 4337
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6654279517878134,
+      "learning_rate": 3.878552962509022e-05,
+      "loss": 1.9617,
+      "step": 4338
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5542521841408227,
+      "learning_rate": 3.878031802524246e-05,
+      "loss": 2.1424,
+      "step": 4339
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5934101125279453,
+      "learning_rate": 3.877510556503283e-05,
+      "loss": 2.049,
+      "step": 4340
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6888907264278066,
+      "learning_rate": 3.8769892244786777e-05,
+      "loss": 1.9537,
+      "step": 4341
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5756450583300693,
+      "learning_rate": 3.8764678064829786e-05,
+      "loss": 1.9567,
+      "step": 4342
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6200812269677566,
+      "learning_rate": 3.875946302548738e-05,
+      "loss": 1.9485,
+      "step": 4343
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6677011459949054,
+      "learning_rate": 3.8754247127085196e-05,
+      "loss": 2.1762,
+      "step": 4344
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.626758298631475,
+      "learning_rate": 3.874903036994885e-05,
+      "loss": 1.9647,
+      "step": 4345
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8052057796280683,
+      "learning_rate": 3.874381275440405e-05,
+      "loss": 2.0362,
+      "step": 4346
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5778446515164742,
+      "learning_rate": 3.873859428077656e-05,
+      "loss": 1.9055,
+      "step": 4347
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0083069561266784,
+      "learning_rate": 3.8733374949392196e-05,
+      "loss": 2.1773,
+      "step": 4348
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.604158060705805,
+      "learning_rate": 3.8728154760576817e-05,
+      "loss": 1.9015,
+      "step": 4349
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9759068361012028,
+      "learning_rate": 3.8722933714656326e-05,
+      "loss": 1.9355,
+      "step": 4350
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6302412004092347,
+      "learning_rate": 3.8717711811956724e-05,
+      "loss": 2.1948,
+      "step": 4351
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9263154794949575,
+      "learning_rate": 3.8712489052804016e-05,
+      "loss": 2.0147,
+      "step": 4352
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.712102964740718,
+      "learning_rate": 3.8707265437524276e-05,
+      "loss": 1.9625,
+      "step": 4353
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7703831464616566,
+      "learning_rate": 3.8702040966443647e-05,
+      "loss": 1.9142,
+      "step": 4354
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6927806315144396,
+      "learning_rate": 3.8696815639888316e-05,
+      "loss": 1.9254,
+      "step": 4355
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7080955706200333,
+      "learning_rate": 3.869158945818451e-05,
+      "loss": 2.1294,
+      "step": 4356
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6949109716807168,
+      "learning_rate": 3.8686362421658536e-05,
+      "loss": 1.9396,
+      "step": 4357
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6792063113000733,
+      "learning_rate": 3.868113453063672e-05,
+      "loss": 2.0811,
+      "step": 4358
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5793790023500945,
+      "learning_rate": 3.867590578544548e-05,
+      "loss": 1.918,
+      "step": 4359
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5617554401257666,
+      "learning_rate": 3.867067618641126e-05,
+      "loss": 2.0772,
+      "step": 4360
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5864697959019031,
+      "learning_rate": 3.8665445733860564e-05,
+      "loss": 1.9212,
+      "step": 4361
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6403485845629397,
+      "learning_rate": 3.866021442811995e-05,
+      "loss": 1.9769,
+      "step": 4362
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6047298692041521,
+      "learning_rate": 3.865498226951603e-05,
+      "loss": 2.1612,
+      "step": 4363
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6036323065027595,
+      "learning_rate": 3.864974925837547e-05,
+      "loss": 1.9416,
+      "step": 4364
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6426389247713166,
+      "learning_rate": 3.8644515395024996e-05,
+      "loss": 2.0031,
+      "step": 4365
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6451309771276795,
+      "learning_rate": 3.863928067979137e-05,
+      "loss": 1.9549,
+      "step": 4366
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.657265998403602,
+      "learning_rate": 3.863404511300142e-05,
+      "loss": 1.9616,
+      "step": 4367
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6926548820296801,
+      "learning_rate": 3.862880869498202e-05,
+      "loss": 2.1542,
+      "step": 4368
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6532338635759335,
+      "learning_rate": 3.862357142606011e-05,
+      "loss": 1.9475,
+      "step": 4369
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6792288448474958,
+      "learning_rate": 3.8618333306562656e-05,
+      "loss": 1.9408,
+      "step": 4370
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6737249830675784,
+      "learning_rate": 3.8613094336816724e-05,
+      "loss": 2.0663,
+      "step": 4371
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5585360617454067,
+      "learning_rate": 3.860785451714939e-05,
+      "loss": 2.1716,
+      "step": 4372
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6505144601677421,
+      "learning_rate": 3.8602613847887784e-05,
+      "loss": 1.9449,
+      "step": 4373
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5817532365077317,
+      "learning_rate": 3.859737232935911e-05,
+      "loss": 1.9681,
+      "step": 4374
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5983945038749495,
+      "learning_rate": 3.859212996189064e-05,
+      "loss": 1.9459,
+      "step": 4375
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7314114556193463,
+      "learning_rate": 3.858688674580965e-05,
+      "loss": 2.1313,
+      "step": 4376
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5909548692520957,
+      "learning_rate": 3.85816426814435e-05,
+      "loss": 2.0277,
+      "step": 4377
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6623723731456705,
+      "learning_rate": 3.857639776911961e-05,
+      "loss": 2.0025,
+      "step": 4378
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6125545316274319,
+      "learning_rate": 3.857115200916543e-05,
+      "loss": 1.9303,
+      "step": 4379
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7056132544504672,
+      "learning_rate": 3.8565905401908483e-05,
+      "loss": 2.082,
+      "step": 4380
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5864366618594531,
+      "learning_rate": 3.8560657947676325e-05,
+      "loss": 1.9583,
+      "step": 4381
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.625493065204582,
+      "learning_rate": 3.855540964679658e-05,
+      "loss": 1.9232,
+      "step": 4382
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5816248450736996,
+      "learning_rate": 3.855016049959693e-05,
+      "loss": 2.0525,
+      "step": 4383
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5736163535612869,
+      "learning_rate": 3.85449105064051e-05,
+      "loss": 2.1753,
+      "step": 4384
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6021596190690466,
+      "learning_rate": 3.8539659667548855e-05,
+      "loss": 1.9268,
+      "step": 4385
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6033668431664059,
+      "learning_rate": 3.8534407983356036e-05,
+      "loss": 1.9663,
+      "step": 4386
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5648118830651282,
+      "learning_rate": 3.852915545415452e-05,
+      "loss": 1.9762,
+      "step": 4387
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6291279868575248,
+      "learning_rate": 3.852390208027226e-05,
+      "loss": 2.0589,
+      "step": 4388
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5625199705409151,
+      "learning_rate": 3.851864786203724e-05,
+      "loss": 2.0122,
+      "step": 4389
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5596898088384515,
+      "learning_rate": 3.851339279977748e-05,
+      "loss": 1.9733,
+      "step": 4390
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6822178071031308,
+      "learning_rate": 3.85081368938211e-05,
+      "loss": 1.9744,
+      "step": 4391
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7498263891852862,
+      "learning_rate": 3.850288014449625e-05,
+      "loss": 2.1548,
+      "step": 4392
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5981234279327199,
+      "learning_rate": 3.849762255213112e-05,
+      "loss": 1.9223,
+      "step": 4393
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7302829030366942,
+      "learning_rate": 3.849236411705395e-05,
+      "loss": 1.9801,
+      "step": 4394
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6763328139377589,
+      "learning_rate": 3.8487104839593074e-05,
+      "loss": 2.1811,
+      "step": 4395
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6171388269099434,
+      "learning_rate": 3.848184472007683e-05,
+      "loss": 2.0034,
+      "step": 4396
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6999880754595095,
+      "learning_rate": 3.847658375883364e-05,
+      "loss": 1.9193,
+      "step": 4397
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5972728499854988,
+      "learning_rate": 3.847132195619195e-05,
+      "loss": 1.9404,
+      "step": 4398
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6070771283482037,
+      "learning_rate": 3.846605931248031e-05,
+      "loss": 1.979,
+      "step": 4399
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6099604479022251,
+      "learning_rate": 3.846079582802725e-05,
+      "loss": 2.182,
+      "step": 4400
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5608375736396208,
+      "learning_rate": 3.845553150316141e-05,
+      "loss": 1.9627,
+      "step": 4401
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5672638693170281,
+      "learning_rate": 3.8450266338211474e-05,
+      "loss": 2.0279,
+      "step": 4402
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6660033917176815,
+      "learning_rate": 3.844500033350614e-05,
+      "loss": 1.9438,
+      "step": 4403
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6282443245464764,
+      "learning_rate": 3.8439733489374215e-05,
+      "loss": 2.1681,
+      "step": 4404
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5622702359925827,
+      "learning_rate": 3.84344658061445e-05,
+      "loss": 1.9212,
+      "step": 4405
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7458993525389047,
+      "learning_rate": 3.8429197284145916e-05,
+      "loss": 1.9342,
+      "step": 4406
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6235234656751504,
+      "learning_rate": 3.8423927923707356e-05,
+      "loss": 1.9417,
+      "step": 4407
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7945183695090019,
+      "learning_rate": 3.841865772515785e-05,
+      "loss": 2.1982,
+      "step": 4408
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5578727356276851,
+      "learning_rate": 3.8413386688826406e-05,
+      "loss": 1.9745,
+      "step": 4409
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7824336407270559,
+      "learning_rate": 3.8408114815042126e-05,
+      "loss": 1.975,
+      "step": 4410
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6216351827846404,
+      "learning_rate": 3.840284210413415e-05,
+      "loss": 1.976,
+      "step": 4411
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8872750739774029,
+      "learning_rate": 3.839756855643169e-05,
+      "loss": 2.2,
+      "step": 4412
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6151798659465543,
+      "learning_rate": 3.839229417226399e-05,
+      "loss": 1.9122,
+      "step": 4413
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7198660855755996,
+      "learning_rate": 3.838701895196034e-05,
+      "loss": 2.0136,
+      "step": 4414
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6931496767608801,
+      "learning_rate": 3.8381742895850106e-05,
+      "loss": 1.9569,
+      "step": 4415
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6448028692130285,
+      "learning_rate": 3.8376466004262684e-05,
+      "loss": 2.1357,
+      "step": 4416
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6644199868732823,
+      "learning_rate": 3.837118827752755e-05,
+      "loss": 1.9638,
+      "step": 4417
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7433498353314635,
+      "learning_rate": 3.836590971597419e-05,
+      "loss": 1.9767,
+      "step": 4418
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7621119758579099,
+      "learning_rate": 3.836063031993217e-05,
+      "loss": 1.9419,
+      "step": 4419
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.623431164086012,
+      "learning_rate": 3.835535008973113e-05,
+      "loss": 2.1778,
+      "step": 4420
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6519281478657304,
+      "learning_rate": 3.8350069025700714e-05,
+      "loss": 1.9413,
+      "step": 4421
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6741287649515262,
+      "learning_rate": 3.8344787128170644e-05,
+      "loss": 1.8843,
+      "step": 4422
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6543883916637756,
+      "learning_rate": 3.833950439747069e-05,
+      "loss": 1.9404,
+      "step": 4423
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.639618998694041,
+      "learning_rate": 3.833422083393068e-05,
+      "loss": 2.1588,
+      "step": 4424
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7792888971563574,
+      "learning_rate": 3.832893643788048e-05,
+      "loss": 1.9739,
+      "step": 4425
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5972069956774114,
+      "learning_rate": 3.8323651209650026e-05,
+      "loss": 1.9193,
+      "step": 4426
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.63954927210489,
+      "learning_rate": 3.831836514956929e-05,
+      "loss": 2.0351,
+      "step": 4427
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6868894707069756,
+      "learning_rate": 3.83130782579683e-05,
+      "loss": 2.104,
+      "step": 4428
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7123481801193416,
+      "learning_rate": 3.8307790535177146e-05,
+      "loss": 1.9262,
+      "step": 4429
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6752594360746733,
+      "learning_rate": 3.830250198152596e-05,
+      "loss": 1.9485,
+      "step": 4430
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.68835547310715,
+      "learning_rate": 3.829721259734494e-05,
+      "loss": 1.9427,
+      "step": 4431
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6021798034399275,
+      "learning_rate": 3.829192238296428e-05,
+      "loss": 2.1595,
+      "step": 4432
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6177533265273177,
+      "learning_rate": 3.828663133871433e-05,
+      "loss": 2.0128,
+      "step": 4433
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5531690626980392,
+      "learning_rate": 3.828133946492539e-05,
+      "loss": 1.9513,
+      "step": 4434
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6151763873065207,
+      "learning_rate": 3.827604676192787e-05,
+      "loss": 1.9621,
+      "step": 4435
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7075990693408382,
+      "learning_rate": 3.82707532300522e-05,
+      "loss": 2.1406,
+      "step": 4436
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5348308837089606,
+      "learning_rate": 3.82654588696289e-05,
+      "loss": 1.9271,
+      "step": 4437
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7167115773271293,
+      "learning_rate": 3.826016368098851e-05,
+      "loss": 1.9938,
+      "step": 4438
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.596114671235856,
+      "learning_rate": 3.8254867664461624e-05,
+      "loss": 2.0721,
+      "step": 4439
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6546141997728795,
+      "learning_rate": 3.82495708203789e-05,
+      "loss": 2.1227,
+      "step": 4440
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.575612789577337,
+      "learning_rate": 3.824427314907103e-05,
+      "loss": 1.9296,
+      "step": 4441
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.686612674473643,
+      "learning_rate": 3.823897465086878e-05,
+      "loss": 1.9645,
+      "step": 4442
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5830525850696061,
+      "learning_rate": 3.8233675326102966e-05,
+      "loss": 1.9667,
+      "step": 4443
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7455931264622753,
+      "learning_rate": 3.8228375175104424e-05,
+      "loss": 2.1471,
+      "step": 4444
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.544897455350434,
+      "learning_rate": 3.8223074198204075e-05,
+      "loss": 2.0501,
+      "step": 4445
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5614974197327285,
+      "learning_rate": 3.82177723957329e-05,
+      "loss": 1.9458,
+      "step": 4446
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7265878008804447,
+      "learning_rate": 3.821246976802187e-05,
+      "loss": 1.9887,
+      "step": 4447
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6900083680557307,
+      "learning_rate": 3.820716631540209e-05,
+      "loss": 2.0828,
+      "step": 4448
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5796403133224415,
+      "learning_rate": 3.820186203820466e-05,
+      "loss": 1.8965,
+      "step": 4449
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7623232512879177,
+      "learning_rate": 3.8196556936760746e-05,
+      "loss": 1.9752,
+      "step": 4450
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.579542656646331,
+      "learning_rate": 3.8191251011401566e-05,
+      "loss": 1.9393,
+      "step": 4451
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7949012346198973,
+      "learning_rate": 3.81859442624584e-05,
+      "loss": 2.1719,
+      "step": 4452
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6544462388500241,
+      "learning_rate": 3.818063669026256e-05,
+      "loss": 1.9141,
+      "step": 4453
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7140605103685778,
+      "learning_rate": 3.817532829514543e-05,
+      "loss": 1.9451,
+      "step": 4454
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6435867400521149,
+      "learning_rate": 3.817001907743842e-05,
+      "loss": 1.9955,
+      "step": 4455
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8514679103256637,
+      "learning_rate": 3.8164709037473024e-05,
+      "loss": 2.1574,
+      "step": 4456
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8132548458648815,
+      "learning_rate": 3.815939817558076e-05,
+      "loss": 1.9691,
+      "step": 4457
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6757275666489964,
+      "learning_rate": 3.81540864920932e-05,
+      "loss": 2.0172,
+      "step": 4458
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7751533094119201,
+      "learning_rate": 3.814877398734199e-05,
+      "loss": 1.9146,
+      "step": 4459
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7394760283740757,
+      "learning_rate": 3.8143460661658806e-05,
+      "loss": 2.1714,
+      "step": 4460
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6619551291268945,
+      "learning_rate": 3.8138146515375366e-05,
+      "loss": 1.9629,
+      "step": 4461
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6562099669408633,
+      "learning_rate": 3.8132831548823465e-05,
+      "loss": 1.9479,
+      "step": 4462
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.67942495489139,
+      "learning_rate": 3.8127515762334956e-05,
+      "loss": 1.9477,
+      "step": 4463
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6520331737264866,
+      "learning_rate": 3.81221991562417e-05,
+      "loss": 2.1779,
+      "step": 4464
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7283209839781335,
+      "learning_rate": 3.811688173087564e-05,
+      "loss": 1.951,
+      "step": 4465
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8320847976965166,
+      "learning_rate": 3.811156348656877e-05,
+      "loss": 1.8864,
+      "step": 4466
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7255569647509523,
+      "learning_rate": 3.810624442365313e-05,
+      "loss": 1.9596,
+      "step": 4467
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8860217960053602,
+      "learning_rate": 3.810092454246081e-05,
+      "loss": 2.1661,
+      "step": 4468
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6937212016566178,
+      "learning_rate": 3.809560384332395e-05,
+      "loss": 1.912,
+      "step": 4469
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8612660362157865,
+      "learning_rate": 3.809028232657474e-05,
+      "loss": 2.044,
+      "step": 4470
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6364416117114474,
+      "learning_rate": 3.8084959992545435e-05,
+      "loss": 1.9291,
+      "step": 4471
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8559921493649146,
+      "learning_rate": 3.807963684156831e-05,
+      "loss": 2.1473,
+      "step": 4472
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6460329994162132,
+      "learning_rate": 3.807431287397574e-05,
+      "loss": 1.9031,
+      "step": 4473
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7874675989436741,
+      "learning_rate": 3.80689880901001e-05,
+      "loss": 1.99,
+      "step": 4474
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6465324431444173,
+      "learning_rate": 3.8063662490273846e-05,
+      "loss": 2.0194,
+      "step": 4475
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9237668974303416,
+      "learning_rate": 3.805833607482947e-05,
+      "loss": 2.0214,
+      "step": 4476
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6687627969254356,
+      "learning_rate": 3.8053008844099535e-05,
+      "loss": 2.1645,
+      "step": 4477
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.711514186846853,
+      "learning_rate": 3.8047680798416634e-05,
+      "loss": 1.9223,
+      "step": 4478
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6976499322967237,
+      "learning_rate": 3.804235193811341e-05,
+      "loss": 1.9665,
+      "step": 4479
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6610651322310258,
+      "learning_rate": 3.8037022263522584e-05,
+      "loss": 2.1418,
+      "step": 4480
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6567772680014503,
+      "learning_rate": 3.8031691774976904e-05,
+      "loss": 1.9634,
+      "step": 4481
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6559507738504657,
+      "learning_rate": 3.802636047280917e-05,
+      "loss": 1.9563,
+      "step": 4482
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.653243600597226,
+      "learning_rate": 3.802102835735223e-05,
+      "loss": 2.0213,
+      "step": 4483
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5940082702885116,
+      "learning_rate": 3.8015695428939005e-05,
+      "loss": 2.1738,
+      "step": 4484
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.78616982316246,
+      "learning_rate": 3.801036168790244e-05,
+      "loss": 1.9112,
+      "step": 4485
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5933655010084593,
+      "learning_rate": 3.800502713457556e-05,
+      "loss": 1.9269,
+      "step": 4486
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6409105469604371,
+      "learning_rate": 3.799969176929139e-05,
+      "loss": 1.9534,
+      "step": 4487
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7339980246127552,
+      "learning_rate": 3.799435559238306e-05,
+      "loss": 2.1626,
+      "step": 4488
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5962549545325753,
+      "learning_rate": 3.7989018604183744e-05,
+      "loss": 2.0306,
+      "step": 4489
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6200094656196918,
+      "learning_rate": 3.798368080502663e-05,
+      "loss": 1.9883,
+      "step": 4490
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6756244266265415,
+      "learning_rate": 3.797834219524498e-05,
+      "loss": 1.9493,
+      "step": 4491
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5571328057955283,
+      "learning_rate": 3.797300277517212e-05,
+      "loss": 2.1108,
+      "step": 4492
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.619284757357046,
+      "learning_rate": 3.7967662545141394e-05,
+      "loss": 1.9247,
+      "step": 4493
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6761823760877507,
+      "learning_rate": 3.796232150548622e-05,
+      "loss": 1.9609,
+      "step": 4494
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5857258520826704,
+      "learning_rate": 3.795697965654007e-05,
+      "loss": 2.0241,
+      "step": 4495
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5817072957315393,
+      "learning_rate": 3.795163699863645e-05,
+      "loss": 2.1535,
+      "step": 4496
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7111934921047798,
+      "learning_rate": 3.7946293532108924e-05,
+      "loss": 1.9019,
+      "step": 4497
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5520885716905221,
+      "learning_rate": 3.794094925729111e-05,
+      "loss": 1.9475,
+      "step": 4498
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.648726651824061,
+      "learning_rate": 3.793560417451667e-05,
+      "loss": 1.908,
+      "step": 4499
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6470228555717719,
+      "learning_rate": 3.793025828411932e-05,
+      "loss": 2.1746,
+      "step": 4500
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7383317037246777,
+      "learning_rate": 3.792491158643283e-05,
+      "loss": 2.0167,
+      "step": 4501
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7120164336355987,
+      "learning_rate": 3.7919564081791016e-05,
+      "loss": 1.9274,
+      "step": 4502
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6461439735147896,
+      "learning_rate": 3.791421577052774e-05,
+      "loss": 1.9233,
+      "step": 4503
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7043594748429448,
+      "learning_rate": 3.790886665297691e-05,
+      "loss": 2.1682,
+      "step": 4504
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.748685788347361,
+      "learning_rate": 3.790351672947251e-05,
+      "loss": 1.9622,
+      "step": 4505
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.599421878969426,
+      "learning_rate": 3.7898166000348553e-05,
+      "loss": 1.9282,
+      "step": 4506
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7748510910912709,
+      "learning_rate": 3.789281446593911e-05,
+      "loss": 2.0331,
+      "step": 4507
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6130002666553089,
+      "learning_rate": 3.788746212657829e-05,
+      "loss": 1.9256,
+      "step": 4508
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7125775114491252,
+      "learning_rate": 3.788210898260026e-05,
+      "loss": 2.1376,
+      "step": 4509
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6629979972676115,
+      "learning_rate": 3.787675503433926e-05,
+      "loss": 1.9068,
+      "step": 4510
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7042523414491536,
+      "learning_rate": 3.7871400282129535e-05,
+      "loss": 1.9617,
+      "step": 4511
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7484435187424696,
+      "learning_rate": 3.786604472630541e-05,
+      "loss": 2.1678,
+      "step": 4512
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7139644599465554,
+      "learning_rate": 3.786068836720126e-05,
+      "loss": 1.9437,
+      "step": 4513
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7328390801583808,
+      "learning_rate": 3.78553312051515e-05,
+      "loss": 2.0425,
+      "step": 4514
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.802550608482326,
+      "learning_rate": 3.784997324049061e-05,
+      "loss": 1.9291,
+      "step": 4515
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5774978766688895,
+      "learning_rate": 3.78446144735531e-05,
+      "loss": 2.1619,
+      "step": 4516
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7348814858362367,
+      "learning_rate": 3.7839254904673535e-05,
+      "loss": 1.9604,
+      "step": 4517
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6635614605971497,
+      "learning_rate": 3.7833894534186547e-05,
+      "loss": 1.9451,
+      "step": 4518
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7963735056630429,
+      "learning_rate": 3.78285333624268e-05,
+      "loss": 1.9446,
+      "step": 4519
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5528505259637595,
+      "learning_rate": 3.7823171389729015e-05,
+      "loss": 2.0135,
+      "step": 4520
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9342537638908034,
+      "learning_rate": 3.7817808616427954e-05,
+      "loss": 2.1162,
+      "step": 4521
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5836072343727278,
+      "learning_rate": 3.7812445042858445e-05,
+      "loss": 1.9477,
+      "step": 4522
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6562720672801294,
+      "learning_rate": 3.7807080669355364e-05,
+      "loss": 1.9449,
+      "step": 4523
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7239260240502563,
+      "learning_rate": 3.780171549625362e-05,
+      "loss": 2.1631,
+      "step": 4524
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6211668040384145,
+      "learning_rate": 3.7796349523888194e-05,
+      "loss": 1.9366,
+      "step": 4525
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.601126692496266,
+      "learning_rate": 3.779098275259409e-05,
+      "loss": 2.0233,
+      "step": 4526
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6654757451046609,
+      "learning_rate": 3.778561518270639e-05,
+      "loss": 1.9472,
+      "step": 4527
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6809630006499146,
+      "learning_rate": 3.778024681456021e-05,
+      "loss": 2.145,
+      "step": 4528
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.611702543825706,
+      "learning_rate": 3.7774877648490715e-05,
+      "loss": 1.967,
+      "step": 4529
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.629286604144254,
+      "learning_rate": 3.776950768483313e-05,
+      "loss": 1.9239,
+      "step": 4530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6652129692701337,
+      "learning_rate": 3.7764136923922714e-05,
+      "loss": 1.9758,
+      "step": 4531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6611273692524587,
+      "learning_rate": 3.775876536609481e-05,
+      "loss": 2.0146,
+      "step": 4532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6568179251180883,
+      "learning_rate": 3.7753393011684756e-05,
+      "loss": 2.1347,
+      "step": 4533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7095948969757396,
+      "learning_rate": 3.774801986102798e-05,
+      "loss": 1.9302,
+      "step": 4534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5734242437712674,
+      "learning_rate": 3.774264591445997e-05,
+      "loss": 1.9522,
+      "step": 4535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5928707338214555,
+      "learning_rate": 3.7737271172316225e-05,
+      "loss": 2.1217,
+      "step": 4536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6125327935601353,
+      "learning_rate": 3.7731895634932316e-05,
+      "loss": 1.9343,
+      "step": 4537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5503117636574768,
+      "learning_rate": 3.7726519302643846e-05,
+      "loss": 2.0371,
+      "step": 4538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6093649456146506,
+      "learning_rate": 3.77211421757865e-05,
+      "loss": 1.95,
+      "step": 4539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5441527425897305,
+      "learning_rate": 3.771576425469599e-05,
+      "loss": 1.9412,
+      "step": 4540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.626570346505215,
+      "learning_rate": 3.771038553970808e-05,
+      "loss": 2.1658,
+      "step": 4541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6631777972993519,
+      "learning_rate": 3.7705006031158583e-05,
+      "loss": 1.9206,
+      "step": 4542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5529961738904918,
+      "learning_rate": 3.7699625729383367e-05,
+      "loss": 1.9418,
+      "step": 4543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.596120332339497,
+      "learning_rate": 3.769424463471835e-05,
+      "loss": 2.1932,
+      "step": 4544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5402253345057711,
+      "learning_rate": 3.7688862747499475e-05,
+      "loss": 2.0261,
+      "step": 4545
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6280176015203585,
+      "learning_rate": 3.768348006806279e-05,
+      "loss": 1.9731,
+      "step": 4546
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5648992552483801,
+      "learning_rate": 3.767809659674433e-05,
+      "loss": 1.9255,
+      "step": 4547
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6884928959955109,
+      "learning_rate": 3.767271233388022e-05,
+      "loss": 2.0994,
+      "step": 4548
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5801285625143808,
+      "learning_rate": 3.766732727980661e-05,
+      "loss": 1.9387,
+      "step": 4549
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.665814298024065,
+      "learning_rate": 3.766194143485974e-05,
+      "loss": 1.9447,
+      "step": 4550
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.625749741598418,
+      "learning_rate": 3.7656554799375826e-05,
+      "loss": 2.0227,
+      "step": 4551
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7114293565033241,
+      "learning_rate": 3.765116737369121e-05,
+      "loss": 1.9427,
+      "step": 4552
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7288519728761113,
+      "learning_rate": 3.7645779158142244e-05,
+      "loss": 2.1523,
+      "step": 4553
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6360551637515837,
+      "learning_rate": 3.7640390153065344e-05,
+      "loss": 1.9417,
+      "step": 4554
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6212972976738201,
+      "learning_rate": 3.7635000358796944e-05,
+      "loss": 1.9311,
+      "step": 4555
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8032232522016756,
+      "learning_rate": 3.762960977567357e-05,
+      "loss": 2.1672,
+      "step": 4556
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5694755003946796,
+      "learning_rate": 3.7624218404031776e-05,
+      "loss": 2.1032,
+      "step": 4557
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7995504593559443,
+      "learning_rate": 3.761882624420817e-05,
+      "loss": 1.9575,
+      "step": 4558
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6626166008222993,
+      "learning_rate": 3.761343329653939e-05,
+      "loss": 1.9275,
+      "step": 4559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7426541296423854,
+      "learning_rate": 3.7608039561362165e-05,
+      "loss": 1.9405,
+      "step": 4560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.692703020133761,
+      "learning_rate": 3.760264503901323e-05,
+      "loss": 2.147,
+      "step": 4561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.632365593826782,
+      "learning_rate": 3.7597249729829385e-05,
+      "loss": 1.9448,
+      "step": 4562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5947682356225279,
+      "learning_rate": 3.759185363414749e-05,
+      "loss": 1.9847,
+      "step": 4563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5834192104801217,
+      "learning_rate": 3.758645675230446e-05,
+      "loss": 1.9446,
+      "step": 4564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.593489115597403,
+      "learning_rate": 3.758105908463721e-05,
+      "loss": 2.1533,
+      "step": 4565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5725828287935691,
+      "learning_rate": 3.7575660631482765e-05,
+      "loss": 1.9059,
+      "step": 4566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5894493672114854,
+      "learning_rate": 3.757026139317816e-05,
+      "loss": 1.9845,
+      "step": 4567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6519983219193053,
+      "learning_rate": 3.7564861370060506e-05,
+      "loss": 2.1509,
+      "step": 4568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6498794438670903,
+      "learning_rate": 3.755946056246693e-05,
+      "loss": 1.9526,
+      "step": 4569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6132216474169538,
+      "learning_rate": 3.755405897073464e-05,
+      "loss": 1.9738,
+      "step": 4570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7414255346001511,
+      "learning_rate": 3.754865659520087e-05,
+      "loss": 2.0065,
+      "step": 4571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6065536364009866,
+      "learning_rate": 3.754325343620293e-05,
+      "loss": 2.0018,
+      "step": 4572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7188275435021811,
+      "learning_rate": 3.753784949407814e-05,
+      "loss": 2.0866,
+      "step": 4573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6628994937750039,
+      "learning_rate": 3.7532444769163894e-05,
+      "loss": 1.913,
+      "step": 4574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6052328597426868,
+      "learning_rate": 3.7527039261797645e-05,
+      "loss": 1.9235,
+      "step": 4575
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6140602934212733,
+      "learning_rate": 3.752163297231687e-05,
+      "loss": 2.0424,
+      "step": 4576
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6206060788932011,
+      "learning_rate": 3.751622590105911e-05,
+      "loss": 2.0892,
+      "step": 4577
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6504112075589991,
+      "learning_rate": 3.751081804836194e-05,
+      "loss": 2.0051,
+      "step": 4578
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5746055505232299,
+      "learning_rate": 3.7505409414563017e-05,
+      "loss": 1.9563,
+      "step": 4579
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6647188108881102,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 2.1705,
+      "step": 4580
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5775354576732673,
+      "learning_rate": 3.7494589805010635e-05,
+      "loss": 1.9662,
+      "step": 4581
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5691829658577531,
+      "learning_rate": 3.74891788299327e-05,
+      "loss": 2.0184,
+      "step": 4582
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5773646747269596,
+      "learning_rate": 3.748376707510403e-05,
+      "loss": 1.9053,
+      "step": 4583
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6010039732758989,
+      "learning_rate": 3.747835454086249e-05,
+      "loss": 1.9363,
+      "step": 4584
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6254087318681725,
+      "learning_rate": 3.747294122754601e-05,
+      "loss": 2.1394,
+      "step": 4585
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6252779143695171,
+      "learning_rate": 3.7467527135492583e-05,
+      "loss": 1.9304,
+      "step": 4586
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6472007465791715,
+      "learning_rate": 3.746211226504021e-05,
+      "loss": 2.0061,
+      "step": 4587
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5835858692826364,
+      "learning_rate": 3.745669661652696e-05,
+      "loss": 2.0139,
+      "step": 4588
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7380918619140692,
+      "learning_rate": 3.745128019029099e-05,
+      "loss": 2.1064,
+      "step": 4589
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5414109110382487,
+      "learning_rate": 3.744586298667043e-05,
+      "loss": 1.8944,
+      "step": 4590
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.545539508588627,
+      "learning_rate": 3.7440445006003524e-05,
+      "loss": 1.8998,
+      "step": 4591
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6462009997921957,
+      "learning_rate": 3.7435026248628525e-05,
+      "loss": 1.9209,
+      "step": 4592
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5849379071625298,
+      "learning_rate": 3.742960671488376e-05,
+      "loss": 2.2009,
+      "step": 4593
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6122243423458665,
+      "learning_rate": 3.7424186405107576e-05,
+      "loss": 2.0145,
+      "step": 4594
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6869390300064838,
+      "learning_rate": 3.74187653196384e-05,
+      "loss": 1.9925,
+      "step": 4595
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.562011134177658,
+      "learning_rate": 3.741334345881469e-05,
+      "loss": 1.945,
+      "step": 4596
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.70982922155756,
+      "learning_rate": 3.7407920822974955e-05,
+      "loss": 2.1168,
+      "step": 4597
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6793283496652244,
+      "learning_rate": 3.740249741245773e-05,
+      "loss": 1.9613,
+      "step": 4598
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6071544331153595,
+      "learning_rate": 3.739707322760166e-05,
+      "loss": 1.9349,
+      "step": 4599
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7589780363811607,
+      "learning_rate": 3.7391648268745374e-05,
+      "loss": 2.1522,
+      "step": 4600
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5527873324192897,
+      "learning_rate": 3.738622253622758e-05,
+      "loss": 2.0132,
+      "step": 4601
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6481257788788469,
+      "learning_rate": 3.738079603038703e-05,
+      "loss": 1.8904,
+      "step": 4602
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7039334567917342,
+      "learning_rate": 3.737536875156252e-05,
+      "loss": 1.9387,
+      "step": 4603
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.581156301829832,
+      "learning_rate": 3.736994070009291e-05,
+      "loss": 1.9444,
+      "step": 4604
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6517879382266721,
+      "learning_rate": 3.736451187631708e-05,
+      "loss": 2.1677,
+      "step": 4605
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5979746394869523,
+      "learning_rate": 3.735908228057397e-05,
+      "loss": 1.9477,
+      "step": 4606
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5357640041017192,
+      "learning_rate": 3.735365191320259e-05,
+      "loss": 2.0413,
+      "step": 4607
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5755159086779102,
+      "learning_rate": 3.734822077454197e-05,
+      "loss": 1.9118,
+      "step": 4608
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6934114478928152,
+      "learning_rate": 3.734278886493119e-05,
+      "loss": 2.0967,
+      "step": 4609
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6273077499067635,
+      "learning_rate": 3.7337356184709404e-05,
+      "loss": 1.9517,
+      "step": 4610
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6147067829217797,
+      "learning_rate": 3.733192273421579e-05,
+      "loss": 1.9286,
+      "step": 4611
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6378051186075927,
+      "learning_rate": 3.7326488513789575e-05,
+      "loss": 2.1279,
+      "step": 4612
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.584584887172597,
+      "learning_rate": 3.732105352377004e-05,
+      "loss": 2.013,
+      "step": 4613
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5904111437185489,
+      "learning_rate": 3.731561776449653e-05,
+      "loss": 1.9558,
+      "step": 4614
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5988825537706705,
+      "learning_rate": 3.73101812363084e-05,
+      "loss": 1.9887,
+      "step": 4615
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6920621311075653,
+      "learning_rate": 3.7304743939545086e-05,
+      "loss": 1.9542,
+      "step": 4616
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.598883419367774,
+      "learning_rate": 3.729930587454605e-05,
+      "loss": 2.1527,
+      "step": 4617
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7763987532804442,
+      "learning_rate": 3.729386704165084e-05,
+      "loss": 1.9328,
+      "step": 4618
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5822567768376813,
+      "learning_rate": 3.728842744119899e-05,
+      "loss": 2.018,
+      "step": 4619
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7015109472716764,
+      "learning_rate": 3.728298707353014e-05,
+      "loss": 1.9737,
+      "step": 4620
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5499912200871575,
+      "learning_rate": 3.7277545938983944e-05,
+      "loss": 2.1166,
+      "step": 4621
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6120910517543294,
+      "learning_rate": 3.7272104037900126e-05,
+      "loss": 1.9298,
+      "step": 4622
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5852681695519044,
+      "learning_rate": 3.7266661370618426e-05,
+      "loss": 1.979,
+      "step": 4623
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6191672159876855,
+      "learning_rate": 3.7261217937478666e-05,
+      "loss": 1.9266,
+      "step": 4624
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5847564464928743,
+      "learning_rate": 3.725577373882071e-05,
+      "loss": 2.174,
+      "step": 4625
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5490442193106174,
+      "learning_rate": 3.725032877498444e-05,
+      "loss": 1.9365,
+      "step": 4626
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6520073043825677,
+      "learning_rate": 3.7244883046309825e-05,
+      "loss": 1.9205,
+      "step": 4627
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6419321881823317,
+      "learning_rate": 3.7239436553136855e-05,
+      "loss": 1.9641,
+      "step": 4628
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5850553093256736,
+      "learning_rate": 3.723398929580558e-05,
+      "loss": 2.1423,
+      "step": 4629
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6727839364118863,
+      "learning_rate": 3.7228541274656095e-05,
+      "loss": 1.9831,
+      "step": 4630
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6275247077144798,
+      "learning_rate": 3.722309249002854e-05,
+      "loss": 1.9254,
+      "step": 4631
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6077404667060448,
+      "learning_rate": 3.7217642942263116e-05,
+      "loss": 2.0486,
+      "step": 4632
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8651727763107769,
+      "learning_rate": 3.721219263170005e-05,
+      "loss": 2.144,
+      "step": 4633
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6093622984162658,
+      "learning_rate": 3.720674155867962e-05,
+      "loss": 1.9273,
+      "step": 4634
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6736119062326317,
+      "learning_rate": 3.720128972354217e-05,
+      "loss": 1.9216,
+      "step": 4635
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6630065011592754,
+      "learning_rate": 3.719583712662809e-05,
+      "loss": 1.9821,
+      "step": 4636
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6053358140119253,
+      "learning_rate": 3.719038376827778e-05,
+      "loss": 2.1618,
+      "step": 4637
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5799597435925008,
+      "learning_rate": 3.718492964883175e-05,
+      "loss": 1.9346,
+      "step": 4638
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6228775409115401,
+      "learning_rate": 3.717947476863049e-05,
+      "loss": 2.0254,
+      "step": 4639
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6759821630854005,
+      "learning_rate": 3.7174019128014595e-05,
+      "loss": 1.9418,
+      "step": 4640
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6459516164249385,
+      "learning_rate": 3.7168562727324665e-05,
+      "loss": 2.0994,
+      "step": 4641
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8044837298201137,
+      "learning_rate": 3.716310556690138e-05,
+      "loss": 1.983,
+      "step": 4642
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.617742525089041,
+      "learning_rate": 3.715764764708544e-05,
+      "loss": 1.9656,
+      "step": 4643
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6550254102178965,
+      "learning_rate": 3.7152188968217626e-05,
+      "loss": 2.0012,
+      "step": 4644
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6018839108485253,
+      "learning_rate": 3.714672953063872e-05,
+      "loss": 2.1283,
+      "step": 4645
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5707417904779045,
+      "learning_rate": 3.714126933468959e-05,
+      "loss": 1.9697,
+      "step": 4646
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6809655619899423,
+      "learning_rate": 3.713580838071115e-05,
+      "loss": 1.9183,
+      "step": 4647
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5941669650791962,
+      "learning_rate": 3.7130346669044334e-05,
+      "loss": 1.9262,
+      "step": 4648
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7518815762349834,
+      "learning_rate": 3.712488420003013e-05,
+      "loss": 2.0998,
+      "step": 4649
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5725059687229308,
+      "learning_rate": 3.711942097400961e-05,
+      "loss": 2.0428,
+      "step": 4650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6596717414011584,
+      "learning_rate": 3.711395699132384e-05,
+      "loss": 1.9658,
+      "step": 4651
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6030059234909771,
+      "learning_rate": 3.7108492252313975e-05,
+      "loss": 1.9322,
+      "step": 4652
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5484816064757371,
+      "learning_rate": 3.710302675732119e-05,
+      "loss": 2.1407,
+      "step": 4653
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.647664935855046,
+      "learning_rate": 3.709756050668673e-05,
+      "loss": 2.0041,
+      "step": 4654
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.579449708215444,
+      "learning_rate": 3.709209350075187e-05,
+      "loss": 1.9277,
+      "step": 4655
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7497199718254093,
+      "learning_rate": 3.7086625739857936e-05,
+      "loss": 2.0176,
+      "step": 4656
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.59496679369951,
+      "learning_rate": 3.70811572243463e-05,
+      "loss": 2.1822,
+      "step": 4657
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6484314983319903,
+      "learning_rate": 3.707568795455839e-05,
+      "loss": 1.9863,
+      "step": 4658
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5690762120048712,
+      "learning_rate": 3.707021793083568e-05,
+      "loss": 1.9237,
+      "step": 4659
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6705043008743519,
+      "learning_rate": 3.7064747153519666e-05,
+      "loss": 1.9436,
+      "step": 4660
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6605600977110833,
+      "learning_rate": 3.705927562295193e-05,
+      "loss": 2.1577,
+      "step": 4661
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5789292077746689,
+      "learning_rate": 3.705380333947408e-05,
+      "loss": 1.974,
+      "step": 4662
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7034311252114132,
+      "learning_rate": 3.704833030342777e-05,
+      "loss": 2.0697,
+      "step": 4663
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5831696916825349,
+      "learning_rate": 3.7042856515154695e-05,
+      "loss": 1.9434,
+      "step": 4664
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9460172960487814,
+      "learning_rate": 3.703738197499663e-05,
+      "loss": 2.1367,
+      "step": 4665
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5654026776816257,
+      "learning_rate": 3.7031906683295336e-05,
+      "loss": 1.9181,
+      "step": 4666
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.846738244729651,
+      "learning_rate": 3.702643064039269e-05,
+      "loss": 1.9382,
+      "step": 4667
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6118843728036611,
+      "learning_rate": 3.7020953846630574e-05,
+      "loss": 1.9653,
+      "step": 4668
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7572111987931353,
+      "learning_rate": 3.7015476302350924e-05,
+      "loss": 2.1739,
+      "step": 4669
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5949346599788876,
+      "learning_rate": 3.700999800789574e-05,
+      "loss": 1.9535,
+      "step": 4670
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7023064675007107,
+      "learning_rate": 3.700451896360702e-05,
+      "loss": 1.9342,
+      "step": 4671
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7893816354472377,
+      "learning_rate": 3.699903916982688e-05,
+      "loss": 1.9422,
+      "step": 4672
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5871888810771583,
+      "learning_rate": 3.699355862689743e-05,
+      "loss": 2.1476,
+      "step": 4673
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7243676654051104,
+      "learning_rate": 3.6988077335160844e-05,
+      "loss": 1.9757,
+      "step": 4674
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6273895543975264,
+      "learning_rate": 3.698259529495933e-05,
+      "loss": 1.9991,
+      "step": 4675
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6656282908365744,
+      "learning_rate": 3.697711250663518e-05,
+      "loss": 1.944,
+      "step": 4676
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7190901801235662,
+      "learning_rate": 3.697162897053069e-05,
+      "loss": 2.1093,
+      "step": 4677
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.761173362190466,
+      "learning_rate": 3.696614468698822e-05,
+      "loss": 1.984,
+      "step": 4678
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6424122385025459,
+      "learning_rate": 3.6960659656350186e-05,
+      "loss": 1.9448,
+      "step": 4679
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.890952911421364,
+      "learning_rate": 3.695517387895902e-05,
+      "loss": 1.9384,
+      "step": 4680
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6989737145639305,
+      "learning_rate": 3.694968735515725e-05,
+      "loss": 2.1309,
+      "step": 4681
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7220639939720056,
+      "learning_rate": 3.694420008528739e-05,
+      "loss": 1.9886,
+      "step": 4682
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7970861329113466,
+      "learning_rate": 3.6938712069692074e-05,
+      "loss": 1.9776,
+      "step": 4683
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6411988496901369,
+      "learning_rate": 3.693322330871391e-05,
+      "loss": 1.9829,
+      "step": 4684
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7227587877892967,
+      "learning_rate": 3.692773380269558e-05,
+      "loss": 2.1231,
+      "step": 4685
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6833635231869992,
+      "learning_rate": 3.692224355197984e-05,
+      "loss": 1.94,
+      "step": 4686
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6645386764457574,
+      "learning_rate": 3.691675255690946e-05,
+      "loss": 1.9642,
+      "step": 4687
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6965385015727485,
+      "learning_rate": 3.691126081782725e-05,
+      "loss": 1.9576,
+      "step": 4688
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7688105334265255,
+      "learning_rate": 3.69057683350761e-05,
+      "loss": 2.1369,
+      "step": 4689
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.621924953292858,
+      "learning_rate": 3.6900275108998925e-05,
+      "loss": 1.9282,
+      "step": 4690
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7901117835822349,
+      "learning_rate": 3.689478113993869e-05,
+      "loss": 1.9855,
+      "step": 4691
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6908094168079423,
+      "learning_rate": 3.688928642823839e-05,
+      "loss": 1.9461,
+      "step": 4692
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7312677029226718,
+      "learning_rate": 3.688379097424111e-05,
+      "loss": 2.1617,
+      "step": 4693
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6717444361560017,
+      "learning_rate": 3.6878294778289934e-05,
+      "loss": 1.9665,
+      "step": 4694
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8909326024452142,
+      "learning_rate": 3.687279784072802e-05,
+      "loss": 1.9488,
+      "step": 4695
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7271402416990989,
+      "learning_rate": 3.686730016189856e-05,
+      "loss": 1.9845,
+      "step": 4696
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6966872470924167,
+      "learning_rate": 3.68618017421448e-05,
+      "loss": 2.1745,
+      "step": 4697
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6524514981011487,
+      "learning_rate": 3.685630258181003e-05,
+      "loss": 1.9557,
+      "step": 4698
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.737429228420455,
+      "learning_rate": 3.6850802681237586e-05,
+      "loss": 1.9708,
+      "step": 4699
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6159700820543748,
+      "learning_rate": 3.684530204077083e-05,
+      "loss": 2.0117,
+      "step": 4700
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6066024934144761,
+      "learning_rate": 3.6839800660753224e-05,
+      "loss": 2.1746,
+      "step": 4701
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6359433901257521,
+      "learning_rate": 3.683429854152821e-05,
+      "loss": 1.9269,
+      "step": 4702
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6289758389005274,
+      "learning_rate": 3.682879568343933e-05,
+      "loss": 1.9624,
+      "step": 4703
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6541772419167793,
+      "learning_rate": 3.682329208683014e-05,
+      "loss": 1.9802,
+      "step": 4704
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6026982855266556,
+      "learning_rate": 3.681778775204425e-05,
+      "loss": 2.1602,
+      "step": 4705
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7724504028348298,
+      "learning_rate": 3.681228267942533e-05,
+      "loss": 2.0269,
+      "step": 4706
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6211520140337724,
+      "learning_rate": 3.680677686931707e-05,
+      "loss": 1.9657,
+      "step": 4707
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7451407836550549,
+      "learning_rate": 3.680127032206323e-05,
+      "loss": 2.0122,
+      "step": 4708
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8191650691320711,
+      "learning_rate": 3.67957630380076e-05,
+      "loss": 2.1867,
+      "step": 4709
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6745953566886871,
+      "learning_rate": 3.6790255017494025e-05,
+      "loss": 1.9312,
+      "step": 4710
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7616853766009727,
+      "learning_rate": 3.6784746260866394e-05,
+      "loss": 1.96,
+      "step": 4711
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7295046801112041,
+      "learning_rate": 3.677923676846864e-05,
+      "loss": 2.0049,
+      "step": 4712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.694161046260058,
+      "learning_rate": 3.6773726540644746e-05,
+      "loss": 2.1122,
+      "step": 4713
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6638928477942693,
+      "learning_rate": 3.676821557773874e-05,
+      "loss": 1.9574,
+      "step": 4714
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6624925543645899,
+      "learning_rate": 3.676270388009469e-05,
+      "loss": 1.9721,
+      "step": 4715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6660978387182667,
+      "learning_rate": 3.675719144805671e-05,
+      "loss": 1.9273,
+      "step": 4716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.9183644923551257,
+      "learning_rate": 3.675167828196896e-05,
+      "loss": 2.1165,
+      "step": 4717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6559180353931214,
+      "learning_rate": 3.6746164382175674e-05,
+      "loss": 1.9967,
+      "step": 4718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7120517330747328,
+      "learning_rate": 3.6740649749021084e-05,
+      "loss": 1.9402,
+      "step": 4719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6409065073598506,
+      "learning_rate": 3.67351343828495e-05,
+      "loss": 1.9214,
+      "step": 4720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6731334120056713,
+      "learning_rate": 3.6729618284005266e-05,
+      "loss": 2.1697,
+      "step": 4721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6367337670013409,
+      "learning_rate": 3.672410145283278e-05,
+      "loss": 1.9172,
+      "step": 4722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.650152979979224,
+      "learning_rate": 3.671858388967647e-05,
+      "loss": 1.9346,
+      "step": 4723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6358511097484627,
+      "learning_rate": 3.671306559488083e-05,
+      "loss": 1.8852,
+      "step": 4724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7829484360152792,
+      "learning_rate": 3.670754656879038e-05,
+      "loss": 1.9841,
+      "step": 4725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5437294739990459,
+      "learning_rate": 3.6702026811749715e-05,
+      "loss": 2.1451,
+      "step": 4726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8352595488512609,
+      "learning_rate": 3.669650632410343e-05,
+      "loss": 1.9545,
+      "step": 4727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6711262599207909,
+      "learning_rate": 3.669098510619621e-05,
+      "loss": 1.9366,
+      "step": 4728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.645527468457305,
+      "learning_rate": 3.6685463158372764e-05,
+      "loss": 2.08,
+      "step": 4729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7214608629543685,
+      "learning_rate": 3.667994048097785e-05,
+      "loss": 1.9247,
+      "step": 4730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6107787139542818,
+      "learning_rate": 3.667441707435626e-05,
+      "loss": 2.0255,
+      "step": 4731
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5908272047238038,
+      "learning_rate": 3.666889293885286e-05,
+      "loss": 1.9859,
+      "step": 4732
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7067818259224394,
+      "learning_rate": 3.666336807481253e-05,
+      "loss": 2.1366,
+      "step": 4733
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6059287932626503,
+      "learning_rate": 3.6657842482580225e-05,
+      "loss": 1.9511,
+      "step": 4734
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6573364868243076,
+      "learning_rate": 3.6652316162500925e-05,
+      "loss": 1.9758,
+      "step": 4735
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6362458790995237,
+      "learning_rate": 3.664678911491964e-05,
+      "loss": 1.9272,
+      "step": 4736
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.722959643300314,
+      "learning_rate": 3.664126134018149e-05,
+      "loss": 2.0347,
+      "step": 4737
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6504901755942433,
+      "learning_rate": 3.663573283863155e-05,
+      "loss": 2.1483,
+      "step": 4738
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7863910147687538,
+      "learning_rate": 3.663020361061501e-05,
+      "loss": 1.9469,
+      "step": 4739
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6254541932283408,
+      "learning_rate": 3.6624673656477085e-05,
+      "loss": 1.913,
+      "step": 4740
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6893589688892566,
+      "learning_rate": 3.661914297656304e-05,
+      "loss": 2.1198,
+      "step": 4741
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5943159851954114,
+      "learning_rate": 3.661361157121816e-05,
+      "loss": 1.9524,
+      "step": 4742
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6369550146610724,
+      "learning_rate": 3.660807944078778e-05,
+      "loss": 2.0145,
+      "step": 4743
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.598288693577857,
+      "learning_rate": 3.6602546585617336e-05,
+      "loss": 1.9399,
+      "step": 4744
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6413492942640383,
+      "learning_rate": 3.659701300605224e-05,
+      "loss": 2.1052,
+      "step": 4745
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5991417708818674,
+      "learning_rate": 3.6591478702437976e-05,
+      "loss": 1.9681,
+      "step": 4746
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6241898250773444,
+      "learning_rate": 3.658594367512008e-05,
+      "loss": 1.9846,
+      "step": 4747
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5775838621061189,
+      "learning_rate": 3.6580407924444114e-05,
+      "loss": 1.9257,
+      "step": 4748
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5942154168384388,
+      "learning_rate": 3.657487145075572e-05,
+      "loss": 2.1569,
+      "step": 4749
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5953021834184642,
+      "learning_rate": 3.656933425440055e-05,
+      "loss": 2.0365,
+      "step": 4750
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6199424827545824,
+      "learning_rate": 3.656379633572431e-05,
+      "loss": 1.9542,
+      "step": 4751
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6194022955660908,
+      "learning_rate": 3.6558257695072764e-05,
+      "loss": 1.9068,
+      "step": 4752
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6733998969564405,
+      "learning_rate": 3.6552718332791705e-05,
+      "loss": 2.1807,
+      "step": 4753
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7006791335968228,
+      "learning_rate": 3.654717824922698e-05,
+      "loss": 1.9273,
+      "step": 4754
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5792439530125585,
+      "learning_rate": 3.654163744472448e-05,
+      "loss": 1.952,
+      "step": 4755
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6828276972817997,
+      "learning_rate": 3.653609591963014e-05,
+      "loss": 2.0,
+      "step": 4756
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.625322314376544,
+      "learning_rate": 3.653055367428993e-05,
+      "loss": 1.9745,
+      "step": 4757
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7885891218744532,
+      "learning_rate": 3.65250107090499e-05,
+      "loss": 2.1744,
+      "step": 4758
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6003271068487426,
+      "learning_rate": 3.65194670242561e-05,
+      "loss": 1.8843,
+      "step": 4759
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8609184344399764,
+      "learning_rate": 3.651392262025465e-05,
+      "loss": 1.9825,
+      "step": 4760
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6101162523801026,
+      "learning_rate": 3.6508377497391714e-05,
+      "loss": 2.1402,
+      "step": 4761
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8534718123886675,
+      "learning_rate": 3.6502831656013496e-05,
+      "loss": 2.0237,
+      "step": 4762
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.696970966415562,
+      "learning_rate": 3.649728509646624e-05,
+      "loss": 1.9946,
+      "step": 4763
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9040466391560001,
+      "learning_rate": 3.649173781909624e-05,
+      "loss": 1.9549,
+      "step": 4764
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7925418800453099,
+      "learning_rate": 3.648618982424984e-05,
+      "loss": 2.1358,
+      "step": 4765
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6176131791561329,
+      "learning_rate": 3.648064111227343e-05,
+      "loss": 1.9197,
+      "step": 4766
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.9971629430327391,
+      "learning_rate": 3.647509168351342e-05,
+      "loss": 1.9576,
+      "step": 4767
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6735379764414092,
+      "learning_rate": 3.6469541538316305e-05,
+      "loss": 1.9922,
+      "step": 4768
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7999531466813391,
+      "learning_rate": 3.6463990677028584e-05,
+      "loss": 1.9518,
+      "step": 4769
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7424020831917181,
+      "learning_rate": 3.645843909999684e-05,
+      "loss": 2.1188,
+      "step": 4770
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8578322560891996,
+      "learning_rate": 3.645288680756766e-05,
+      "loss": 1.9651,
+      "step": 4771
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7497432528668888,
+      "learning_rate": 3.6447333800087714e-05,
+      "loss": 1.9299,
+      "step": 4772
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6746771859867038,
+      "learning_rate": 3.644178007790369e-05,
+      "loss": 2.1816,
+      "step": 4773
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7167248304321332,
+      "learning_rate": 3.643622564136233e-05,
+      "loss": 2.0064,
+      "step": 4774
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5722167492806013,
+      "learning_rate": 3.6430670490810424e-05,
+      "loss": 1.8921,
+      "step": 4775
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6989817001031443,
+      "learning_rate": 3.64251146265948e-05,
+      "loss": 1.9661,
+      "step": 4776
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7228226005619175,
+      "learning_rate": 3.641955804906234e-05,
+      "loss": 1.988,
+      "step": 4777
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6480983627475008,
+      "learning_rate": 3.641400075855995e-05,
+      "loss": 2.1424,
+      "step": 4778
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7215222448976328,
+      "learning_rate": 3.640844275543461e-05,
+      "loss": 1.9284,
+      "step": 4779
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6382733868481769,
+      "learning_rate": 3.6402884040033325e-05,
+      "loss": 2.0043,
+      "step": 4780
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.72673542802697,
+      "learning_rate": 3.639732461270313e-05,
+      "loss": 2.0087,
+      "step": 4781
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7352246283865421,
+      "learning_rate": 3.6391764473791146e-05,
+      "loss": 2.1643,
+      "step": 4782
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.730022278407708,
+      "learning_rate": 3.638620362364451e-05,
+      "loss": 1.9394,
+      "step": 4783
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7040886428413377,
+      "learning_rate": 3.638064206261041e-05,
+      "loss": 1.9217,
+      "step": 4784
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.64922167465325,
+      "learning_rate": 3.6375079791036066e-05,
+      "loss": 2.135,
+      "step": 4785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6802041321356217,
+      "learning_rate": 3.636951680926876e-05,
+      "loss": 1.9149,
+      "step": 4786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6351113225951748,
+      "learning_rate": 3.6363953117655815e-05,
+      "loss": 1.9879,
+      "step": 4787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6329126638139047,
+      "learning_rate": 3.635838871654459e-05,
+      "loss": 1.9912,
+      "step": 4788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6524280305072329,
+      "learning_rate": 3.63528236062825e-05,
+      "loss": 1.934,
+      "step": 4789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.59834984059157,
+      "learning_rate": 3.6347257787216984e-05,
+      "loss": 2.1532,
+      "step": 4790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.819354849115079,
+      "learning_rate": 3.634169125969556e-05,
+      "loss": 1.9114,
+      "step": 4791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5574501477083332,
+      "learning_rate": 3.6336124024065756e-05,
+      "loss": 1.9329,
+      "step": 4792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8849853051718746,
+      "learning_rate": 3.633055608067515e-05,
+      "loss": 2.0182,
+      "step": 4793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.734568881521357,
+      "learning_rate": 3.6324987429871396e-05,
+      "loss": 2.1256,
+      "step": 4794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7362110709940252,
+      "learning_rate": 3.631941807200214e-05,
+      "loss": 1.948,
+      "step": 4795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7688663947144935,
+      "learning_rate": 3.631384800741512e-05,
+      "loss": 1.9248,
+      "step": 4796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8674906297711769,
+      "learning_rate": 3.6308277236458087e-05,
+      "loss": 2.1146,
+      "step": 4797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6068105948524357,
+      "learning_rate": 3.630270575947885e-05,
+      "loss": 1.9392,
+      "step": 4798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6256214382059277,
+      "learning_rate": 3.629713357682526e-05,
+      "loss": 1.999,
+      "step": 4799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6179384383323173,
+      "learning_rate": 3.629156068884521e-05,
+      "loss": 1.9301,
+      "step": 4800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6741669033673735,
+      "learning_rate": 3.628598709588664e-05,
+      "loss": 1.9412,
+      "step": 4801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6883006040068876,
+      "learning_rate": 3.6280412798297536e-05,
+      "loss": 2.1709,
+      "step": 4802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6397816357689469,
+      "learning_rate": 3.6274837796425915e-05,
+      "loss": 1.9127,
+      "step": 4803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6680436694491237,
+      "learning_rate": 3.626926209061985e-05,
+      "loss": 1.916,
+      "step": 4804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6352155594469302,
+      "learning_rate": 3.626368568122746e-05,
+      "loss": 1.9939,
+      "step": 4805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6595464689303211,
+      "learning_rate": 3.625810856859691e-05,
+      "loss": 2.1554,
+      "step": 4806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6368539874875424,
+      "learning_rate": 3.625253075307638e-05,
+      "loss": 1.9122,
+      "step": 4807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.576583132290356,
+      "learning_rate": 3.624695223501413e-05,
+      "loss": 1.9027,
+      "step": 4808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8398802022237688,
+      "learning_rate": 3.624137301475846e-05,
+      "loss": 1.9326,
+      "step": 4809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6850803838526457,
+      "learning_rate": 3.6235793092657677e-05,
+      "loss": 2.1494,
+      "step": 4810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8661114941413354,
+      "learning_rate": 3.623021246906018e-05,
+      "loss": 1.9252,
+      "step": 4811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7045008245636176,
+      "learning_rate": 3.622463114431439e-05,
+      "loss": 2.0383,
+      "step": 4812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7955888263780125,
+      "learning_rate": 3.621904911876876e-05,
+      "loss": 1.9266,
+      "step": 4813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7030212117884018,
+      "learning_rate": 3.62134663927718e-05,
+      "loss": 2.1516,
+      "step": 4814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6507131457896277,
+      "learning_rate": 3.620788296667206e-05,
+      "loss": 1.9343,
+      "step": 4815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6526919725551548,
+      "learning_rate": 3.620229884081816e-05,
+      "loss": 1.9412,
+      "step": 4816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8476586273965105,
+      "learning_rate": 3.619671401555872e-05,
+      "loss": 2.1126,
+      "step": 4817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6220980261084174,
+      "learning_rate": 3.619112849124242e-05,
+      "loss": 2.0459,
+      "step": 4818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6651082744578745,
+      "learning_rate": 3.6185542268218e-05,
+      "loss": 1.9323,
+      "step": 4819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6903831526174564,
+      "learning_rate": 3.617995534683422e-05,
+      "loss": 1.9248,
+      "step": 4820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7014190415757046,
+      "learning_rate": 3.61743677274399e-05,
+      "loss": 1.9466,
+      "step": 4821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6976815471371333,
+      "learning_rate": 3.61687794103839e-05,
+      "loss": 2.0984,
+      "step": 4822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7382070190135815,
+      "learning_rate": 3.6163190396015115e-05,
+      "loss": 1.9494,
+      "step": 4823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7367671758006374,
+      "learning_rate": 3.61576006846825e-05,
+      "loss": 1.9747,
+      "step": 4824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7143127810736432,
+      "learning_rate": 3.615201027673504e-05,
+      "loss": 1.9418,
+      "step": 4825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6618584439855055,
+      "learning_rate": 3.6146419172521755e-05,
+      "loss": 2.1213,
+      "step": 4826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.628905297473289,
+      "learning_rate": 3.614082737239174e-05,
+      "loss": 1.9553,
+      "step": 4827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.745367877986337,
+      "learning_rate": 3.613523487669409e-05,
+      "loss": 1.9467,
+      "step": 4828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7563125224947433,
+      "learning_rate": 3.612964168577799e-05,
+      "loss": 2.159,
+      "step": 4829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6378071852890695,
+      "learning_rate": 3.6124047799992644e-05,
+      "loss": 2.0224,
+      "step": 4830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7233810739824218,
+      "learning_rate": 3.611845321968729e-05,
+      "loss": 1.9317,
+      "step": 4831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6379880379420682,
+      "learning_rate": 3.611285794521123e-05,
+      "loss": 1.9459,
+      "step": 4832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6098284416219479,
+      "learning_rate": 3.610726197691379e-05,
+      "loss": 1.9352,
+      "step": 4833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6395096838876897,
+      "learning_rate": 3.610166531514436e-05,
+      "loss": 2.1485,
+      "step": 4834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7531053626722699,
+      "learning_rate": 3.609606796025235e-05,
+      "loss": 1.935,
+      "step": 4835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5613887591457554,
+      "learning_rate": 3.609046991258724e-05,
+      "loss": 2.0268,
+      "step": 4836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6253388748590373,
+      "learning_rate": 3.608487117249853e-05,
+      "loss": 1.8833,
+      "step": 4837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7218240222397928,
+      "learning_rate": 3.607927174033577e-05,
+      "loss": 2.1792,
+      "step": 4838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5902828546062213,
+      "learning_rate": 3.607367161644857e-05,
+      "loss": 1.9617,
+      "step": 4839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7533100967655497,
+      "learning_rate": 3.6068070801186556e-05,
+      "loss": 1.9455,
+      "step": 4840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.558009128469453,
+      "learning_rate": 3.606246929489941e-05,
+      "loss": 1.9555,
+      "step": 4841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6486365976513113,
+      "learning_rate": 3.605686709793686e-05,
+      "loss": 2.1386,
+      "step": 4842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6321387886310299,
+      "learning_rate": 3.605126421064868e-05,
+      "loss": 1.9676,
+      "step": 4843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.65173096131481,
+      "learning_rate": 3.604566063338467e-05,
+      "loss": 1.9478,
+      "step": 4844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6228662913301347,
+      "learning_rate": 3.604005636649469e-05,
+      "loss": 1.9486,
+      "step": 4845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.624371585335125,
+      "learning_rate": 3.603445141032864e-05,
+      "loss": 2.115,
+      "step": 4846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5968152884125457,
+      "learning_rate": 3.602884576523646e-05,
+      "loss": 1.9123,
+      "step": 4847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.747768162418612,
+      "learning_rate": 3.602323943156813e-05,
+      "loss": 1.9538,
+      "step": 4848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6437661254966658,
+      "learning_rate": 3.601763240967367e-05,
+      "loss": 2.0026,
+      "step": 4849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6028479013512997,
+      "learning_rate": 3.601202469990317e-05,
+      "loss": 2.1292,
+      "step": 4850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.701882364962353,
+      "learning_rate": 3.600641630260671e-05,
+      "loss": 1.9085,
+      "step": 4851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6543341446933661,
+      "learning_rate": 3.600080721813449e-05,
+      "loss": 1.9492,
+      "step": 4852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6241082689790098,
+      "learning_rate": 3.599519744683667e-05,
+      "loss": 1.9316,
+      "step": 4853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7367186568189865,
+      "learning_rate": 3.59895869890635e-05,
+      "loss": 2.1732,
+      "step": 4854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5782439832713605,
+      "learning_rate": 3.598397584516527e-05,
+      "loss": 2.0535,
+      "step": 4855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7170534204782196,
+      "learning_rate": 3.5978364015492316e-05,
+      "loss": 1.9535,
+      "step": 4856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5957469533969957,
+      "learning_rate": 3.597275150039499e-05,
+      "loss": 1.8993,
+      "step": 4857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7500520501754648,
+      "learning_rate": 3.596713830022369e-05,
+      "loss": 2.1402,
+      "step": 4858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6488243955121235,
+      "learning_rate": 3.596152441532892e-05,
+      "loss": 1.8859,
+      "step": 4859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6621914445148278,
+      "learning_rate": 3.595590984606114e-05,
+      "loss": 1.9178,
+      "step": 4860
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6487321180927275,
+      "learning_rate": 3.59502945927709e-05,
+      "loss": 1.9983,
+      "step": 4861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6293516161470537,
+      "learning_rate": 3.5944678655808776e-05,
+      "loss": 2.1425,
+      "step": 4862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.676160178749284,
+      "learning_rate": 3.5939062035525415e-05,
+      "loss": 1.9776,
+      "step": 4863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7545774529736037,
+      "learning_rate": 3.5933444732271466e-05,
+      "loss": 1.9758,
+      "step": 4864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7370220867433243,
+      "learning_rate": 3.592782674639765e-05,
+      "loss": 1.915,
+      "step": 4865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7241577712820746,
+      "learning_rate": 3.592220807825471e-05,
+      "loss": 2.1472,
+      "step": 4866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6535015884361915,
+      "learning_rate": 3.5916588728193454e-05,
+      "loss": 2.004,
+      "step": 4867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8188842802311205,
+      "learning_rate": 3.5910968696564715e-05,
+      "loss": 1.9369,
+      "step": 4868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6576870158359984,
+      "learning_rate": 3.5905347983719384e-05,
+      "loss": 1.9254,
+      "step": 4869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.862925927629241,
+      "learning_rate": 3.589972659000838e-05,
+      "loss": 2.1162,
+      "step": 4870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6625663915038383,
+      "learning_rate": 3.589410451578266e-05,
+      "loss": 1.9184,
+      "step": 4871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7044216793811316,
+      "learning_rate": 3.5888481761393236e-05,
+      "loss": 1.9468,
+      "step": 4872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5931061083974939,
+      "learning_rate": 3.588285832719117e-05,
+      "loss": 1.9318,
+      "step": 4873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7799332121425692,
+      "learning_rate": 3.5877234213527555e-05,
+      "loss": 2.1624,
+      "step": 4874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5893027717145287,
+      "learning_rate": 3.5871609420753515e-05,
+      "loss": 1.9149,
+      "step": 4875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.625378115703623,
+      "learning_rate": 3.586598394922025e-05,
+      "loss": 1.9441,
+      "step": 4876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.625346058798477,
+      "learning_rate": 3.586035779927896e-05,
+      "loss": 1.9113,
+      "step": 4877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7466397160493984,
+      "learning_rate": 3.585473097128092e-05,
+      "loss": 2.1596,
+      "step": 4878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6286924315052096,
+      "learning_rate": 3.584910346557743e-05,
+      "loss": 1.942,
+      "step": 4879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7182046538500895,
+      "learning_rate": 3.584347528251983e-05,
+      "loss": 1.9596,
+      "step": 4880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6564135419418355,
+      "learning_rate": 3.583784642245954e-05,
+      "loss": 1.9692,
+      "step": 4881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6719896381860888,
+      "learning_rate": 3.583221688574798e-05,
+      "loss": 2.1145,
+      "step": 4882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6626820275549926,
+      "learning_rate": 3.582658667273661e-05,
+      "loss": 1.943,
+      "step": 4883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6305779435852528,
+      "learning_rate": 3.582095578377695e-05,
+      "loss": 1.9804,
+      "step": 4884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7769676042755611,
+      "learning_rate": 3.581532421922058e-05,
+      "loss": 1.9346,
+      "step": 4885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6636510673320748,
+      "learning_rate": 3.580969197941907e-05,
+      "loss": 2.1317,
+      "step": 4886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5964077681184267,
+      "learning_rate": 3.58040590647241e-05,
+      "loss": 1.9183,
+      "step": 4887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7905805499799053,
+      "learning_rate": 3.5798425475487326e-05,
+      "loss": 1.8928,
+      "step": 4888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6668423946944634,
+      "learning_rate": 3.57927912120605e-05,
+      "loss": 1.9675,
+      "step": 4889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.693228153826767,
+      "learning_rate": 3.578715627479537e-05,
+      "loss": 2.1948,
+      "step": 4890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6261805740600579,
+      "learning_rate": 3.578152066404377e-05,
+      "loss": 1.9651,
+      "step": 4891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5750266366947432,
+      "learning_rate": 3.5775884380157535e-05,
+      "loss": 1.9995,
+      "step": 4892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5897497716502431,
+      "learning_rate": 3.5770247423488576e-05,
+      "loss": 1.9676,
+      "step": 4893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6938595282712247,
+      "learning_rate": 3.5764609794388824e-05,
+      "loss": 2.1646,
+      "step": 4894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5818011666412489,
+      "learning_rate": 3.575897149321026e-05,
+      "loss": 1.9607,
+      "step": 4895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.781899539806489,
+      "learning_rate": 3.57533325203049e-05,
+      "loss": 1.9604,
+      "step": 4896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6104675117794495,
+      "learning_rate": 3.574769287602482e-05,
+      "loss": 1.9061,
+      "step": 4897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7949453918841302,
+      "learning_rate": 3.574205256072212e-05,
+      "loss": 2.1601,
+      "step": 4898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6233673003276599,
+      "learning_rate": 3.573641157474896e-05,
+      "loss": 1.9039,
+      "step": 4899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6875097083360802,
+      "learning_rate": 3.5730769918457505e-05,
+      "loss": 1.918,
+      "step": 4900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6298407881285911,
+      "learning_rate": 3.5725127592200004e-05,
+      "loss": 1.9486,
+      "step": 4901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6879005699387595,
+      "learning_rate": 3.571948459632874e-05,
+      "loss": 2.201,
+      "step": 4902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.585224647237516,
+      "learning_rate": 3.5713840931196e-05,
+      "loss": 1.9238,
+      "step": 4903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6634765641228637,
+      "learning_rate": 3.5708196597154156e-05,
+      "loss": 1.9315,
+      "step": 4904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6282448634397899,
+      "learning_rate": 3.570255159455561e-05,
+      "loss": 1.9945,
+      "step": 4905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6665441768835734,
+      "learning_rate": 3.569690592375281e-05,
+      "loss": 2.188,
+      "step": 4906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6435155328731993,
+      "learning_rate": 3.569125958509822e-05,
+      "loss": 1.9922,
+      "step": 4907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.841636529822681,
+      "learning_rate": 3.5685612578944386e-05,
+      "loss": 1.9607,
+      "step": 4908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.579221368705233,
+      "learning_rate": 3.5679964905643845e-05,
+      "loss": 1.962,
+      "step": 4909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.0878884409838723,
+      "learning_rate": 3.567431656554923e-05,
+      "loss": 2.1543,
+      "step": 4910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6524587137712553,
+      "learning_rate": 3.566866755901318e-05,
+      "loss": 2.0048,
+      "step": 4911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8985314317031181,
+      "learning_rate": 3.566301788638838e-05,
+      "loss": 1.9171,
+      "step": 4912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7219355575710548,
+      "learning_rate": 3.565736754802758e-05,
+      "loss": 1.959,
+      "step": 4913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9435077329563867,
+      "learning_rate": 3.565171654428353e-05,
+      "loss": 2.141,
+      "step": 4914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8014999943373552,
+      "learning_rate": 3.564606487550907e-05,
+      "loss": 1.9465,
+      "step": 4915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8448964834738598,
+      "learning_rate": 3.564041254205703e-05,
+      "loss": 1.8788,
+      "step": 4916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7853180201926094,
+      "learning_rate": 3.563475954428034e-05,
+      "loss": 2.0332,
+      "step": 4917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8271357735683744,
+      "learning_rate": 3.562910588253191e-05,
+      "loss": 2.1353,
+      "step": 4918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7044526796355662,
+      "learning_rate": 3.562345155716474e-05,
+      "loss": 1.8886,
+      "step": 4919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7620216103355353,
+      "learning_rate": 3.561779656853184e-05,
+      "loss": 1.9581,
+      "step": 4920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7557933361188971,
+      "learning_rate": 3.5612140916986295e-05,
+      "loss": 1.9606,
+      "step": 4921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7488438260002206,
+      "learning_rate": 3.5606484602881184e-05,
+      "loss": 2.1036,
+      "step": 4922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8269938362513124,
+      "learning_rate": 3.560082762656967e-05,
+      "loss": 2.0296,
+      "step": 4923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6384114549163041,
+      "learning_rate": 3.5595169988404945e-05,
+      "loss": 1.9036,
+      "step": 4924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.796767598397149,
+      "learning_rate": 3.558951168874022e-05,
+      "loss": 1.9511,
+      "step": 4925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6574925523142856,
+      "learning_rate": 3.5583852727928784e-05,
+      "loss": 2.1584,
+      "step": 4926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7779394647116428,
+      "learning_rate": 3.5578193106323945e-05,
+      "loss": 1.8908,
+      "step": 4927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7288697736838304,
+      "learning_rate": 3.557253282427905e-05,
+      "loss": 1.9244,
+      "step": 4928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8030971869065199,
+      "learning_rate": 3.55668718821475e-05,
+      "loss": 1.9508,
+      "step": 4929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5713921305334442,
+      "learning_rate": 3.556121028028272e-05,
+      "loss": 2.1512,
+      "step": 4930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8450327594662159,
+      "learning_rate": 3.555554801903821e-05,
+      "loss": 1.9129,
+      "step": 4931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6902317094232462,
+      "learning_rate": 3.5549885098767466e-05,
+      "loss": 1.9306,
+      "step": 4932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.767841052157392,
+      "learning_rate": 3.5544221519824054e-05,
+      "loss": 1.9266,
+      "step": 4933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8576160523484292,
+      "learning_rate": 3.553855728256158e-05,
+      "loss": 2.1496,
+      "step": 4934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7284099207394579,
+      "learning_rate": 3.5532892387333684e-05,
+      "loss": 1.9734,
+      "step": 4935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8124906419782457,
+      "learning_rate": 3.552722683449404e-05,
+      "loss": 1.9928,
+      "step": 4936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6548622394625544,
+      "learning_rate": 3.5521560624396374e-05,
+      "loss": 1.9522,
+      "step": 4937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7241905203745482,
+      "learning_rate": 3.551589375739446e-05,
+      "loss": 2.1221,
+      "step": 4938
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7048180412665803,
+      "learning_rate": 3.55102262338421e-05,
+      "loss": 1.8886,
+      "step": 4939
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.72704645228362,
+      "learning_rate": 3.550455805409314e-05,
+      "loss": 1.9417,
+      "step": 4940
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.739575565672194,
+      "learning_rate": 3.549888921850146e-05,
+      "loss": 1.9104,
+      "step": 4941
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8694837718308387,
+      "learning_rate": 3.5493219727421005e-05,
+      "loss": 2.0068,
+      "step": 4942
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6557171092454571,
+      "learning_rate": 3.548754958120573e-05,
+      "loss": 2.1225,
+      "step": 4943
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7203125294237179,
+      "learning_rate": 3.548187878020965e-05,
+      "loss": 1.9071,
+      "step": 4944
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.733140397295901,
+      "learning_rate": 3.547620732478681e-05,
+      "loss": 1.9407,
+      "step": 4945
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7584475131848014,
+      "learning_rate": 3.547053521529132e-05,
+      "loss": 2.1462,
+      "step": 4946
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6869873449389399,
+      "learning_rate": 3.54648624520773e-05,
+      "loss": 1.8969,
+      "step": 4947
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7851867765378411,
+      "learning_rate": 3.545918903549893e-05,
+      "loss": 1.9889,
+      "step": 4948
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6615166145106793,
+      "learning_rate": 3.545351496591042e-05,
+      "loss": 1.9647,
+      "step": 4949
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9990315158721332,
+      "learning_rate": 3.544784024366602e-05,
+      "loss": 2.1321,
+      "step": 4950
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6313974984637188,
+      "learning_rate": 3.5442164869120035e-05,
+      "loss": 1.9503,
+      "step": 4951
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9676895609822459,
+      "learning_rate": 3.5436488842626796e-05,
+      "loss": 1.9408,
+      "step": 4952
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.78490869894415,
+      "learning_rate": 3.5430812164540696e-05,
+      "loss": 1.9296,
+      "step": 4953
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7165856847143085,
+      "learning_rate": 3.542513483521613e-05,
+      "loss": 2.0675,
+      "step": 4954
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8895762981929953,
+      "learning_rate": 3.541945685500758e-05,
+      "loss": 2.1712,
+      "step": 4955
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7316392203747876,
+      "learning_rate": 3.5413778224269526e-05,
+      "loss": 1.9718,
+      "step": 4956
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.672448181299301,
+      "learning_rate": 3.540809894335652e-05,
+      "loss": 1.9351,
+      "step": 4957
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9657108158480306,
+      "learning_rate": 3.540241901262314e-05,
+      "loss": 2.2049,
+      "step": 4958
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7977634046555407,
+      "learning_rate": 3.5396738432424004e-05,
+      "loss": 2.0132,
+      "step": 4959
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6908695068692342,
+      "learning_rate": 3.5391057203113774e-05,
+      "loss": 1.9712,
+      "step": 4960
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7776545020907336,
+      "learning_rate": 3.5385375325047166e-05,
+      "loss": 1.9858,
+      "step": 4961
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6761425136956788,
+      "learning_rate": 3.53796927985789e-05,
+      "loss": 2.1517,
+      "step": 4962
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6420998580007115,
+      "learning_rate": 3.5374009624063784e-05,
+      "loss": 1.9702,
+      "step": 4963
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6368251509227237,
+      "learning_rate": 3.5368325801856614e-05,
+      "loss": 1.9241,
+      "step": 4964
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6608187805130326,
+      "learning_rate": 3.536264133231229e-05,
+      "loss": 1.9057,
+      "step": 4965
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6747763875665331,
+      "learning_rate": 3.5356956215785675e-05,
+      "loss": 2.1867,
+      "step": 4966
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5750928136821136,
+      "learning_rate": 3.535127045263176e-05,
+      "loss": 2.0105,
+      "step": 4967
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6220107590320654,
+      "learning_rate": 3.534558404320549e-05,
+      "loss": 1.9061,
+      "step": 4968
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6439262243647773,
+      "learning_rate": 3.5339896987861905e-05,
+      "loss": 1.991,
+      "step": 4969
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6117646048809102,
+      "learning_rate": 3.5334209286956076e-05,
+      "loss": 2.1053,
+      "step": 4970
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5962279058014149,
+      "learning_rate": 3.532852094084312e-05,
+      "loss": 1.9544,
+      "step": 4971
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.541364344839125,
+      "learning_rate": 3.532283194987816e-05,
+      "loss": 1.9493,
+      "step": 4972
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5976789882406672,
+      "learning_rate": 3.531714231441639e-05,
+      "loss": 2.0178,
+      "step": 4973
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6112561014419314,
+      "learning_rate": 3.531145203481305e-05,
+      "loss": 1.9598,
+      "step": 4974
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5996308575886273,
+      "learning_rate": 3.53057611114234e-05,
+      "loss": 2.1614,
+      "step": 4975
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6022651087823346,
+      "learning_rate": 3.530006954460274e-05,
+      "loss": 1.929,
+      "step": 4976
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6955935245841902,
+      "learning_rate": 3.529437733470643e-05,
+      "loss": 1.9489,
+      "step": 4977
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6997916345993581,
+      "learning_rate": 3.528868448208985e-05,
+      "loss": 2.1126,
+      "step": 4978
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6566727541303923,
+      "learning_rate": 3.5282990987108426e-05,
+      "loss": 1.9735,
+      "step": 4979
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6589771113389836,
+      "learning_rate": 3.5277296850117635e-05,
+      "loss": 1.9152,
+      "step": 4980
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8259057297426624,
+      "learning_rate": 3.527160207147299e-05,
+      "loss": 1.9651,
+      "step": 4981
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6598443558399713,
+      "learning_rate": 3.526590665153002e-05,
+      "loss": 2.1058,
+      "step": 4982
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8279044215666387,
+      "learning_rate": 3.5260210590644327e-05,
+      "loss": 1.9355,
+      "step": 4983
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7152544510800812,
+      "learning_rate": 3.525451388917154e-05,
+      "loss": 1.9371,
+      "step": 4984
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7716289501710197,
+      "learning_rate": 3.5248816547467315e-05,
+      "loss": 2.011,
+      "step": 4985
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7228448912774911,
+      "learning_rate": 3.524311856588738e-05,
+      "loss": 1.9735,
+      "step": 4986
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7886004156693251,
+      "learning_rate": 3.523741994478747e-05,
+      "loss": 2.1254,
+      "step": 4987
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6564041856943515,
+      "learning_rate": 3.523172068452337e-05,
+      "loss": 1.9346,
+      "step": 4988
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7224110735626381,
+      "learning_rate": 3.5226020785450934e-05,
+      "loss": 1.9472,
+      "step": 4989
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7593298512664858,
+      "learning_rate": 3.522032024792599e-05,
+      "loss": 2.1758,
+      "step": 4990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5950790309987041,
+      "learning_rate": 3.521461907230447e-05,
+      "loss": 1.9615,
+      "step": 4991
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6328394128244219,
+      "learning_rate": 3.520891725894233e-05,
+      "loss": 2.048,
+      "step": 4992
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6572697820220891,
+      "learning_rate": 3.520321480819554e-05,
+      "loss": 1.9425,
+      "step": 4993
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6348769675057717,
+      "learning_rate": 3.519751172042013e-05,
+      "loss": 1.8967,
+      "step": 4994
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7013175129305821,
+      "learning_rate": 3.519180799597217e-05,
+      "loss": 2.1684,
+      "step": 4995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6169724105799803,
+      "learning_rate": 3.518610363520778e-05,
+      "loss": 1.951,
+      "step": 4996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6601227095969193,
+      "learning_rate": 3.518039863848309e-05,
+      "loss": 1.9641,
+      "step": 4997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7020753131930161,
+      "learning_rate": 3.517469300615428e-05,
+      "loss": 2.0315,
+      "step": 4998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7163562740692608,
+      "learning_rate": 3.51689867385776e-05,
+      "loss": 2.1527,
+      "step": 4999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7087977709407456,
+      "learning_rate": 3.516327983610929e-05,
+      "loss": 1.9428,
+      "step": 5000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6083215029152462,
+      "learning_rate": 3.515757229910568e-05,
+      "loss": 1.9484,
+      "step": 5001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9528040944593796,
+      "learning_rate": 3.5151864127923096e-05,
+      "loss": 2.1106,
+      "step": 5002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5690130556597205,
+      "learning_rate": 3.514615532291793e-05,
+      "loss": 1.9163,
+      "step": 5003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9381438920999967,
+      "learning_rate": 3.5140445884446606e-05,
+      "loss": 2.0292,
+      "step": 5004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.616909352266048,
+      "learning_rate": 3.5134735812865586e-05,
+      "loss": 1.9625,
+      "step": 5005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8110078976931545,
+      "learning_rate": 3.512902510853138e-05,
+      "loss": 1.9324,
+      "step": 5006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6071756179973448,
+      "learning_rate": 3.512331377180052e-05,
+      "loss": 2.1176,
+      "step": 5007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7455602813805682,
+      "learning_rate": 3.51176018030296e-05,
+      "loss": 1.949,
+      "step": 5008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6383701987621012,
+      "learning_rate": 3.511188920257523e-05,
+      "loss": 1.8885,
+      "step": 5009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6369806590233833,
+      "learning_rate": 3.5106175970794074e-05,
+      "loss": 1.9903,
+      "step": 5010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.669527485607258,
+      "learning_rate": 3.5100462108042844e-05,
+      "loss": 2.0959,
+      "step": 5011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6052509258200691,
+      "learning_rate": 3.509474761467826e-05,
+      "loss": 1.9715,
+      "step": 5012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6352368860994843,
+      "learning_rate": 3.508903249105712e-05,
+      "loss": 1.9829,
+      "step": 5013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.648413771381859,
+      "learning_rate": 3.508331673753624e-05,
+      "loss": 2.1369,
+      "step": 5014
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6130999575956402,
+      "learning_rate": 3.507760035447246e-05,
+      "loss": 1.9496,
+      "step": 5015
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5758889217406541,
+      "learning_rate": 3.5071883342222695e-05,
+      "loss": 2.0148,
+      "step": 5016
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5820698654532159,
+      "learning_rate": 3.506616570114387e-05,
+      "loss": 1.9013,
+      "step": 5017
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6273154194628972,
+      "learning_rate": 3.506044743159298e-05,
+      "loss": 1.927,
+      "step": 5018
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5980274872029098,
+      "learning_rate": 3.505472853392703e-05,
+      "loss": 2.1856,
+      "step": 5019
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6521227812636816,
+      "learning_rate": 3.504900900850305e-05,
+      "loss": 1.9512,
+      "step": 5020
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6477494473361861,
+      "learning_rate": 3.504328885567818e-05,
+      "loss": 1.9951,
+      "step": 5021
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6135790221560959,
+      "learning_rate": 3.5037568075809514e-05,
+      "loss": 2.1131,
+      "step": 5022
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7305116539084799,
+      "learning_rate": 3.503184666925424e-05,
+      "loss": 2.0182,
+      "step": 5023
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6656652082082719,
+      "learning_rate": 3.5026124636369564e-05,
+      "loss": 1.9354,
+      "step": 5024
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.9870328842349357,
+      "learning_rate": 3.5020401977512745e-05,
+      "loss": 1.9769,
+      "step": 5025
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6597119962730804,
+      "learning_rate": 3.5014678693041065e-05,
+      "loss": 1.9542,
+      "step": 5026
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8940749741245584,
+      "learning_rate": 3.500895478331185e-05,
+      "loss": 2.2081,
+      "step": 5027
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6404446824598085,
+      "learning_rate": 3.500323024868247e-05,
+      "loss": 1.9265,
+      "step": 5028
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.811943312396918,
+      "learning_rate": 3.499750508951033e-05,
+      "loss": 1.994,
+      "step": 5029
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5744118806779627,
+      "learning_rate": 3.499177930615288e-05,
+      "loss": 1.951,
+      "step": 5030
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8107830469888612,
+      "learning_rate": 3.4986052898967596e-05,
+      "loss": 2.1361,
+      "step": 5031
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6383419898317009,
+      "learning_rate": 3.498032586831202e-05,
+      "loss": 1.9297,
+      "step": 5032
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5795718250406815,
+      "learning_rate": 3.497459821454368e-05,
+      "loss": 1.9623,
+      "step": 5033
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8100169774083417,
+      "learning_rate": 3.49688699380202e-05,
+      "loss": 2.1598,
+      "step": 5034
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.638604078315698,
+      "learning_rate": 3.496314103909922e-05,
+      "loss": 2.0557,
+      "step": 5035
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5789338985855809,
+      "learning_rate": 3.4957411518138424e-05,
+      "loss": 1.933,
+      "step": 5036
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7036222176596998,
+      "learning_rate": 3.495168137549551e-05,
+      "loss": 1.9634,
+      "step": 5037
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6122702391052178,
+      "learning_rate": 3.494595061152825e-05,
+      "loss": 1.9126,
+      "step": 5038
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6107893314926427,
+      "learning_rate": 3.494021922659444e-05,
+      "loss": 2.0844,
+      "step": 5039
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7109800160386094,
+      "learning_rate": 3.4934487221051896e-05,
+      "loss": 1.9057,
+      "step": 5040
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.632276211162601,
+      "learning_rate": 3.4928754595258507e-05,
+      "loss": 2.0348,
+      "step": 5041
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.684807517385371,
+      "learning_rate": 3.492302134957218e-05,
+      "loss": 1.9132,
+      "step": 5042
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8238146461518109,
+      "learning_rate": 3.491728748435087e-05,
+      "loss": 2.1217,
+      "step": 5043
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.704284460585149,
+      "learning_rate": 3.491155299995256e-05,
+      "loss": 1.9214,
+      "step": 5044
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6484063574546234,
+      "learning_rate": 3.490581789673527e-05,
+      "loss": 1.8958,
+      "step": 5045
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6476115655836854,
+      "learning_rate": 3.4900082175057084e-05,
+      "loss": 2.1198,
+      "step": 5046
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.588357662251807,
+      "learning_rate": 3.48943458352761e-05,
+      "loss": 1.9271,
+      "step": 5047
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6138398041989258,
+      "learning_rate": 3.488860887775045e-05,
+      "loss": 1.9617,
+      "step": 5048
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7224529356434634,
+      "learning_rate": 3.4882871302838326e-05,
+      "loss": 1.9337,
+      "step": 5049
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6085381517050298,
+      "learning_rate": 3.4877133110897953e-05,
+      "loss": 1.9781,
+      "step": 5050
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.689891322227274,
+      "learning_rate": 3.4871394302287576e-05,
+      "loss": 2.1299,
+      "step": 5051
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5673033528489645,
+      "learning_rate": 3.4865654877365503e-05,
+      "loss": 1.9211,
+      "step": 5052
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7063793365332953,
+      "learning_rate": 3.4859914836490063e-05,
+      "loss": 1.9259,
+      "step": 5053
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5602466605579784,
+      "learning_rate": 3.485417418001964e-05,
+      "loss": 1.9988,
+      "step": 5054
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6672311114395776,
+      "learning_rate": 3.484843290831264e-05,
+      "loss": 2.0885,
+      "step": 5055
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6463298415058347,
+      "learning_rate": 3.484269102172751e-05,
+      "loss": 1.9214,
+      "step": 5056
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6085751309892113,
+      "learning_rate": 3.483694852062276e-05,
+      "loss": 1.9387,
+      "step": 5057
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6513838993862322,
+      "learning_rate": 3.483120540535688e-05,
+      "loss": 1.9219,
+      "step": 5058
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5525022947518117,
+      "learning_rate": 3.482546167628847e-05,
+      "loss": 2.1275,
+      "step": 5059
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6651188752857388,
+      "learning_rate": 3.4819717333776126e-05,
+      "loss": 2.0334,
+      "step": 5060
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6299755797869692,
+      "learning_rate": 3.481397237817849e-05,
+      "loss": 1.9483,
+      "step": 5061
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6649929131221995,
+      "learning_rate": 3.480822680985424e-05,
+      "loss": 1.9433,
+      "step": 5062
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6515585989302406,
+      "learning_rate": 3.480248062916209e-05,
+      "loss": 2.117,
+      "step": 5063
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6126332814546808,
+      "learning_rate": 3.479673383646081e-05,
+      "loss": 1.9619,
+      "step": 5064
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6522764331628698,
+      "learning_rate": 3.47909864321092e-05,
+      "loss": 1.9519,
+      "step": 5065
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.572194786763794,
+      "learning_rate": 3.4785238416466074e-05,
+      "loss": 2.0115,
+      "step": 5066
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5890212179732006,
+      "learning_rate": 3.477948978989031e-05,
+      "loss": 2.1467,
+      "step": 5067
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6135908098063265,
+      "learning_rate": 3.477374055274083e-05,
+      "loss": 1.9341,
+      "step": 5068
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5592208126466649,
+      "learning_rate": 3.4767990705376576e-05,
+      "loss": 1.9307,
+      "step": 5069
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.605346290090158,
+      "learning_rate": 3.476224024815653e-05,
+      "loss": 1.9213,
+      "step": 5070
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6277069168830549,
+      "learning_rate": 3.475648918143973e-05,
+      "loss": 2.1826,
+      "step": 5071
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6185838947817488,
+      "learning_rate": 3.475073750558522e-05,
+      "loss": 2.0301,
+      "step": 5072
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.598390697223414,
+      "learning_rate": 3.4744985220952115e-05,
+      "loss": 1.9124,
+      "step": 5073
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5998682517890915,
+      "learning_rate": 3.4739232327899554e-05,
+      "loss": 1.9641,
+      "step": 5074
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6185740550141803,
+      "learning_rate": 3.47334788267867e-05,
+      "loss": 2.1101,
+      "step": 5075
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6503722389680315,
+      "learning_rate": 3.4727724717972775e-05,
+      "loss": 1.9473,
+      "step": 5076
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6252433398013737,
+      "learning_rate": 3.4721970001817043e-05,
+      "loss": 1.8973,
+      "step": 5077
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8114455057663799,
+      "learning_rate": 3.471621467867878e-05,
+      "loss": 1.897,
+      "step": 5078
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6976009962830167,
+      "learning_rate": 3.471045874891732e-05,
+      "loss": 2.1843,
+      "step": 5079
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6506971509349028,
+      "learning_rate": 3.4704702212892026e-05,
+      "loss": 1.9501,
+      "step": 5080
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.735880400545291,
+      "learning_rate": 3.469894507096231e-05,
+      "loss": 1.9218,
+      "step": 5081
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6150772405365771,
+      "learning_rate": 3.469318732348761e-05,
+      "loss": 1.9252,
+      "step": 5082
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7313290514504924,
+      "learning_rate": 3.468742897082741e-05,
+      "loss": 2.1071,
+      "step": 5083
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6328257819271537,
+      "learning_rate": 3.4681670013341204e-05,
+      "loss": 1.9276,
+      "step": 5084
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.650721268479889,
+      "learning_rate": 3.467591045138857e-05,
+      "loss": 1.9722,
+      "step": 5085
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6296090873765905,
+      "learning_rate": 3.467015028532911e-05,
+      "loss": 1.9098,
+      "step": 5086
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6848619633990509,
+      "learning_rate": 3.4664389515522435e-05,
+      "loss": 2.1327,
+      "step": 5087
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6510473511446323,
+      "learning_rate": 3.465862814232822e-05,
+      "loss": 1.9506,
+      "step": 5088
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7600050779940416,
+      "learning_rate": 3.465286616610617e-05,
+      "loss": 1.928,
+      "step": 5089
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6204448387553597,
+      "learning_rate": 3.464710358721604e-05,
+      "loss": 1.9436,
+      "step": 5090
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7748582401322675,
+      "learning_rate": 3.4641340406017585e-05,
+      "loss": 2.1407,
+      "step": 5091
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6826420282782385,
+      "learning_rate": 3.463557662287065e-05,
+      "loss": 1.9402,
+      "step": 5092
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6613574219917621,
+      "learning_rate": 3.4629812238135085e-05,
+      "loss": 1.9415,
+      "step": 5093
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6959311685514882,
+      "learning_rate": 3.4624047252170776e-05,
+      "loss": 1.933,
+      "step": 5094
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.678041244402601,
+      "learning_rate": 3.461828166533766e-05,
+      "loss": 2.1241,
+      "step": 5095
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5685807298551447,
+      "learning_rate": 3.461251547799571e-05,
+      "loss": 1.9364,
+      "step": 5096
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6884281035315439,
+      "learning_rate": 3.460674869050493e-05,
+      "loss": 1.9722,
+      "step": 5097
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6516877836476774,
+      "learning_rate": 3.460098130322536e-05,
+      "loss": 1.9021,
+      "step": 5098
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6246031735400638,
+      "learning_rate": 3.459521331651709e-05,
+      "loss": 2.1431,
+      "step": 5099
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6126417966221773,
+      "learning_rate": 3.458944473074023e-05,
+      "loss": 1.922,
+      "step": 5100
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6750844793239923,
+      "learning_rate": 3.4583675546254947e-05,
+      "loss": 1.9017,
+      "step": 5101
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6670977954542251,
+      "learning_rate": 3.4577905763421424e-05,
+      "loss": 1.9748,
+      "step": 5102
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7198755473431313,
+      "learning_rate": 3.457213538259989e-05,
+      "loss": 2.1711,
+      "step": 5103
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5750742742821333,
+      "learning_rate": 3.456636440415064e-05,
+      "loss": 1.931,
+      "step": 5104
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7123374391740395,
+      "learning_rate": 3.456059282843395e-05,
+      "loss": 1.9418,
+      "step": 5105
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7983166650273611,
+      "learning_rate": 3.455482065581017e-05,
+      "loss": 1.9267,
+      "step": 5106
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.763056605363453,
+      "learning_rate": 3.4549047886639685e-05,
+      "loss": 2.1487,
+      "step": 5107
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7288017896855449,
+      "learning_rate": 3.454327452128292e-05,
+      "loss": 1.893,
+      "step": 5108
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6259020509081491,
+      "learning_rate": 3.453750056010031e-05,
+      "loss": 1.9561,
+      "step": 5109
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6858846928339414,
+      "learning_rate": 3.453172600345237e-05,
+      "loss": 2.0127,
+      "step": 5110
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7253630144634484,
+      "learning_rate": 3.4525950851699615e-05,
+      "loss": 2.1008,
+      "step": 5111
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6431262069070509,
+      "learning_rate": 3.452017510520262e-05,
+      "loss": 1.9834,
+      "step": 5112
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6694494870553823,
+      "learning_rate": 3.451439876432197e-05,
+      "loss": 1.9657,
+      "step": 5113
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6672835466280991,
+      "learning_rate": 3.450862182941832e-05,
+      "loss": 1.9153,
+      "step": 5114
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7988169812524438,
+      "learning_rate": 3.4502844300852365e-05,
+      "loss": 2.1356,
+      "step": 5115
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5528837048793839,
+      "learning_rate": 3.449706617898479e-05,
+      "loss": 2.01,
+      "step": 5116
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8164440584757596,
+      "learning_rate": 3.4491287464176356e-05,
+      "loss": 1.9029,
+      "step": 5117
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6419419950076354,
+      "learning_rate": 3.448550815678786e-05,
+      "loss": 1.8905,
+      "step": 5118
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7065539452675651,
+      "learning_rate": 3.447972825718012e-05,
+      "loss": 2.1337,
+      "step": 5119
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6587030263711571,
+      "learning_rate": 3.4473947765714e-05,
+      "loss": 1.9352,
+      "step": 5120
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6222426182698918,
+      "learning_rate": 3.4468166682750396e-05,
+      "loss": 1.8794,
+      "step": 5121
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.598407123406919,
+      "learning_rate": 3.4462385008650256e-05,
+      "loss": 2.0528,
+      "step": 5122
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6774662075057788,
+      "learning_rate": 3.4456602743774546e-05,
+      "loss": 2.1225,
+      "step": 5123
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6068316567688256,
+      "learning_rate": 3.4450819888484284e-05,
+      "loss": 1.9259,
+      "step": 5124
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6501330345693576,
+      "learning_rate": 3.44450364431405e-05,
+      "loss": 1.9687,
+      "step": 5125
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8185496278360419,
+      "learning_rate": 3.443925240810429e-05,
+      "loss": 1.9759,
+      "step": 5126
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6778249108819614,
+      "learning_rate": 3.443346778373678e-05,
+      "loss": 2.1438,
+      "step": 5127
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5906612399808495,
+      "learning_rate": 3.442768257039913e-05,
+      "loss": 2.0352,
+      "step": 5128
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6928571371964668,
+      "learning_rate": 3.442189676845252e-05,
+      "loss": 1.9655,
+      "step": 5129
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.664110489034663,
+      "learning_rate": 3.441611037825819e-05,
+      "loss": 1.9252,
+      "step": 5130
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5979560810888491,
+      "learning_rate": 3.4410323400177403e-05,
+      "loss": 2.1254,
+      "step": 5131
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6088116302105208,
+      "learning_rate": 3.4404535834571475e-05,
+      "loss": 1.937,
+      "step": 5132
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6322525961600001,
+      "learning_rate": 3.439874768180174e-05,
+      "loss": 1.9341,
+      "step": 5133
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5610481539786276,
+      "learning_rate": 3.439295894222957e-05,
+      "loss": 2.0191,
+      "step": 5134
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.695969413194961,
+      "learning_rate": 3.438716961621638e-05,
+      "loss": 2.1643,
+      "step": 5135
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6649232098148886,
+      "learning_rate": 3.438137970412364e-05,
+      "loss": 1.9539,
+      "step": 5136
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.776742545986537,
+      "learning_rate": 3.437558920631282e-05,
+      "loss": 1.9219,
+      "step": 5137
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7022998487518577,
+      "learning_rate": 3.436979812314545e-05,
+      "loss": 1.9381,
+      "step": 5138
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.754871912085761,
+      "learning_rate": 3.4364006454983085e-05,
+      "loss": 2.163,
+      "step": 5139
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6805855494567521,
+      "learning_rate": 3.435821420218734e-05,
+      "loss": 1.9095,
+      "step": 5140
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6674972179146449,
+      "learning_rate": 3.435242136511984e-05,
+      "loss": 1.9671,
+      "step": 5141
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7204555658970736,
+      "learning_rate": 3.434662794414224e-05,
+      "loss": 1.9212,
+      "step": 5142
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6916334257099047,
+      "learning_rate": 3.434083393961627e-05,
+      "loss": 2.1897,
+      "step": 5143
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7097736671270447,
+      "learning_rate": 3.4335039351903655e-05,
+      "loss": 1.9501,
+      "step": 5144
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.78387467883872,
+      "learning_rate": 3.43292441813662e-05,
+      "loss": 1.9351,
+      "step": 5145
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7159987457329327,
+      "learning_rate": 3.4323448428365685e-05,
+      "loss": 1.8985,
+      "step": 5146
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8751289202502383,
+      "learning_rate": 3.431765209326399e-05,
+      "loss": 2.189,
+      "step": 5147
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8509625811467332,
+      "learning_rate": 3.4311855176423e-05,
+      "loss": 1.8786,
+      "step": 5148
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7757564645598389,
+      "learning_rate": 3.430605767820464e-05,
+      "loss": 1.9526,
+      "step": 5149
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1200283631676975,
+      "learning_rate": 3.430025959897086e-05,
+      "loss": 1.9862,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.895042336920376,
+      "learning_rate": 3.429446093908368e-05,
+      "loss": 2.1759,
+      "step": 5151
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1542142256196717,
+      "learning_rate": 3.428866169890511e-05,
+      "loss": 1.9336,
+      "step": 5152
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1158703541461679,
+      "learning_rate": 3.4282861878797235e-05,
+      "loss": 1.971,
+      "step": 5153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7370615823645804,
+      "learning_rate": 3.4277061479122165e-05,
+      "loss": 1.9838,
+      "step": 5154
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.22579558972302,
+      "learning_rate": 3.427126050024203e-05,
+      "loss": 2.1637,
+      "step": 5155
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6998609977686412,
+      "learning_rate": 3.426545894251901e-05,
+      "loss": 1.9077,
+      "step": 5156
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0131026859083376,
+      "learning_rate": 3.425965680631533e-05,
+      "loss": 1.9276,
+      "step": 5157
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7119424614406688,
+      "learning_rate": 3.4253854091993235e-05,
+      "loss": 1.9032,
+      "step": 5158
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.776952531289085,
+      "learning_rate": 3.424805079991501e-05,
+      "loss": 2.013,
+      "step": 5159
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7172813218013243,
+      "learning_rate": 3.424224693044298e-05,
+      "loss": 2.1131,
+      "step": 5160
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.728952270225371,
+      "learning_rate": 3.423644248393952e-05,
+      "loss": 1.9089,
+      "step": 5161
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6588193576705801,
+      "learning_rate": 3.4230637460767e-05,
+      "loss": 1.9099,
+      "step": 5162
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7497192611985066,
+      "learning_rate": 3.4224831861287864e-05,
+      "loss": 2.127,
+      "step": 5163
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6744664704845145,
+      "learning_rate": 3.4219025685864574e-05,
+      "loss": 1.9533,
+      "step": 5164
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7467798192513476,
+      "learning_rate": 3.421321893485965e-05,
+      "loss": 1.9963,
+      "step": 5165
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7099871825638486,
+      "learning_rate": 3.420741160863561e-05,
+      "loss": 1.9176,
+      "step": 5166
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8651618133622729,
+      "learning_rate": 3.4201603707555037e-05,
+      "loss": 2.0902,
+      "step": 5167
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7327223380270227,
+      "learning_rate": 3.419579523198054e-05,
+      "loss": 1.976,
+      "step": 5168
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7989076192145107,
+      "learning_rate": 3.418998618227478e-05,
+      "loss": 1.945,
+      "step": 5169
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.680794191298754,
+      "learning_rate": 3.418417655880043e-05,
+      "loss": 1.9292,
+      "step": 5170
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6296391274873299,
+      "learning_rate": 3.41783663619202e-05,
+      "loss": 2.1723,
+      "step": 5171
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.714581826324136,
+      "learning_rate": 3.417255559199686e-05,
+      "loss": 1.995,
+      "step": 5172
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6517959202400113,
+      "learning_rate": 3.416674424939318e-05,
+      "loss": 1.9151,
+      "step": 5173
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7430243647891767,
+      "learning_rate": 3.416093233447201e-05,
+      "loss": 1.9562,
+      "step": 5174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7116917518055694,
+      "learning_rate": 3.41551198475962e-05,
+      "loss": 2.1601,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8043435232940346,
+      "learning_rate": 3.414930678912864e-05,
+      "loss": 1.9281,
+      "step": 5176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7233351272705736,
+      "learning_rate": 3.414349315943227e-05,
+      "loss": 1.9291,
+      "step": 5177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6889890429916535,
+      "learning_rate": 3.4137678958870065e-05,
+      "loss": 2.0542,
+      "step": 5178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7207627617406904,
+      "learning_rate": 3.413186418780503e-05,
+      "loss": 2.1352,
+      "step": 5179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6497740913649946,
+      "learning_rate": 3.41260488466002e-05,
+      "loss": 1.9394,
+      "step": 5180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8390346520117312,
+      "learning_rate": 3.412023293561864e-05,
+      "loss": 1.9551,
+      "step": 5181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6449023928647714,
+      "learning_rate": 3.411441645522347e-05,
+      "loss": 1.938,
+      "step": 5182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8190888596988902,
+      "learning_rate": 3.410859940577784e-05,
+      "loss": 2.1311,
+      "step": 5183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6824027525943179,
+      "learning_rate": 3.4102781787644936e-05,
+      "loss": 1.9881,
+      "step": 5184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6869264416125215,
+      "learning_rate": 3.4096963601187964e-05,
+      "loss": 1.9124,
+      "step": 5185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7333373906617172,
+      "learning_rate": 3.409114484677019e-05,
+      "loss": 1.9365,
+      "step": 5186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7163243037554896,
+      "learning_rate": 3.408532552475489e-05,
+      "loss": 2.1024,
+      "step": 5187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6718368610810902,
+      "learning_rate": 3.407950563550539e-05,
+      "loss": 1.8952,
+      "step": 5188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.849432910916129,
+      "learning_rate": 3.4073685179385055e-05,
+      "loss": 1.9269,
+      "step": 5189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.625871996430383,
+      "learning_rate": 3.4067864156757285e-05,
+      "loss": 2.0522,
+      "step": 5190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7717397180791657,
+      "learning_rate": 3.40620425679855e-05,
+      "loss": 1.9598,
+      "step": 5191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7919723249946629,
+      "learning_rate": 3.405622041343317e-05,
+      "loss": 2.1885,
+      "step": 5192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6797896803009621,
+      "learning_rate": 3.40503976934638e-05,
+      "loss": 1.9569,
+      "step": 5193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8431752009887734,
+      "learning_rate": 3.404457440844092e-05,
+      "loss": 1.9272,
+      "step": 5194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6138715832117367,
+      "learning_rate": 3.403875055872809e-05,
+      "loss": 2.141,
+      "step": 5195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7532240803932055,
+      "learning_rate": 3.403292614468895e-05,
+      "loss": 1.9997,
+      "step": 5196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8005897661975706,
+      "learning_rate": 3.402710116668711e-05,
+      "loss": 1.9514,
+      "step": 5197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.561932043487301,
+      "learning_rate": 3.402127562508628e-05,
+      "loss": 1.9238,
+      "step": 5198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8603888360652088,
+      "learning_rate": 3.4015449520250123e-05,
+      "loss": 2.1168,
+      "step": 5199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6083117502512289,
+      "learning_rate": 3.4009622852542436e-05,
+      "loss": 1.9282,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5974118334823821,
+      "learning_rate": 3.400379562232698e-05,
+      "loss": 1.9231,
+      "step": 5201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8409479283076279,
+      "learning_rate": 3.399796782996757e-05,
+      "loss": 1.989,
+      "step": 5202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6229204406977977,
+      "learning_rate": 3.399213947582805e-05,
+      "loss": 1.969,
+      "step": 5203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8096436491396863,
+      "learning_rate": 3.3986310560272346e-05,
+      "loss": 2.1007,
+      "step": 5204
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7074908868530467,
+      "learning_rate": 3.398048108366435e-05,
+      "loss": 1.9152,
+      "step": 5205
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7490728766885801,
+      "learning_rate": 3.3974651046368025e-05,
+      "loss": 1.9479,
+      "step": 5206
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7154405943947776,
+      "learning_rate": 3.396882044874736e-05,
+      "loss": 2.0789,
+      "step": 5207
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6013850840004539,
+      "learning_rate": 3.39629892911664e-05,
+      "loss": 1.9299,
+      "step": 5208
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7772636156467689,
+      "learning_rate": 3.3957157573989196e-05,
+      "loss": 2.0059,
+      "step": 5209
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5949626288378417,
+      "learning_rate": 3.395132529757984e-05,
+      "loss": 1.9148,
+      "step": 5210
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7835083503038963,
+      "learning_rate": 3.394549246230248e-05,
+      "loss": 1.9476,
+      "step": 5211
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6232077671416494,
+      "learning_rate": 3.3939659068521277e-05,
+      "loss": 2.1236,
+      "step": 5212
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5492288599285367,
+      "learning_rate": 3.3933825116600426e-05,
+      "loss": 1.952,
+      "step": 5213
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7358976361594033,
+      "learning_rate": 3.392799060690418e-05,
+      "loss": 1.932,
+      "step": 5214
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7269091727364939,
+      "learning_rate": 3.392215553979679e-05,
+      "loss": 2.0332,
+      "step": 5215
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7253663669870873,
+      "learning_rate": 3.3916319915642593e-05,
+      "loss": 2.1421,
+      "step": 5216
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6163733515459247,
+      "learning_rate": 3.3910483734805907e-05,
+      "loss": 1.9196,
+      "step": 5217
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6822118512181622,
+      "learning_rate": 3.390464699765112e-05,
+      "loss": 1.8877,
+      "step": 5218
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7501477255554793,
+      "learning_rate": 3.3898809704542645e-05,
+      "loss": 2.1221,
+      "step": 5219
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6202622808850948,
+      "learning_rate": 3.3892971855844905e-05,
+      "loss": 1.9652,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8934645349024312,
+      "learning_rate": 3.388713345192241e-05,
+      "loss": 2.0123,
+      "step": 5221
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.661446574841369,
+      "learning_rate": 3.388129449313966e-05,
+      "loss": 1.9432,
+      "step": 5222
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7651857552611896,
+      "learning_rate": 3.3875454979861224e-05,
+      "loss": 1.885,
+      "step": 5223
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.728772148889215,
+      "learning_rate": 3.386961491245165e-05,
+      "loss": 2.1677,
+      "step": 5224
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.9152544393392451,
+      "learning_rate": 3.386377429127559e-05,
+      "loss": 1.9461,
+      "step": 5225
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7011376699196756,
+      "learning_rate": 3.385793311669769e-05,
+      "loss": 1.9182,
+      "step": 5226
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6548572417212039,
+      "learning_rate": 3.385209138908264e-05,
+      "loss": 2.1628,
+      "step": 5227
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.755820390360719,
+      "learning_rate": 3.384624910879515e-05,
+      "loss": 2.0142,
+      "step": 5228
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7237642728698935,
+      "learning_rate": 3.384040627619999e-05,
+      "loss": 1.9402,
+      "step": 5229
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5743953572042363,
+      "learning_rate": 3.3834562891661953e-05,
+      "loss": 1.9626,
+      "step": 5230
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.692353865690534,
+      "learning_rate": 3.3828718955545856e-05,
+      "loss": 2.0868,
+      "step": 5231
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.536562168032147,
+      "learning_rate": 3.382287446821657e-05,
+      "loss": 1.9185,
+      "step": 5232
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6804395895438056,
+      "learning_rate": 3.3817029430038984e-05,
+      "loss": 1.9131,
+      "step": 5233
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.545639237880857,
+      "learning_rate": 3.3811183841378026e-05,
+      "loss": 2.0617,
+      "step": 5234
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6906458231078563,
+      "learning_rate": 3.3805337702598665e-05,
+      "loss": 1.9701,
+      "step": 5235
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7034130215905531,
+      "learning_rate": 3.37994910140659e-05,
+      "loss": 2.132,
+      "step": 5236
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6340922560260684,
+      "learning_rate": 3.379364377614476e-05,
+      "loss": 1.8999,
+      "step": 5237
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.8457953857030698,
+      "learning_rate": 3.378779598920031e-05,
+      "loss": 1.9566,
+      "step": 5238
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6584616535164477,
+      "learning_rate": 3.3781947653597664e-05,
+      "loss": 2.1106,
+      "step": 5239
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6286440522181891,
+      "learning_rate": 3.377609876970194e-05,
+      "loss": 2.0084,
+      "step": 5240
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6934126197897016,
+      "learning_rate": 3.3770249337878324e-05,
+      "loss": 1.9316,
+      "step": 5241
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6051795487898327,
+      "learning_rate": 3.376439935849201e-05,
+      "loss": 1.9447,
+      "step": 5242
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7181415837208338,
+      "learning_rate": 3.3758548831908236e-05,
+      "loss": 1.9312,
+      "step": 5243
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6367579874078206,
+      "learning_rate": 3.3752697758492286e-05,
+      "loss": 2.1441,
+      "step": 5244
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6806886415859817,
+      "learning_rate": 3.374684613860945e-05,
+      "loss": 1.943,
+      "step": 5245
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6978920071413222,
+      "learning_rate": 3.374099397262508e-05,
+      "loss": 2.0205,
+      "step": 5246
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6608756208767823,
+      "learning_rate": 3.3735141260904546e-05,
+      "loss": 1.9107,
+      "step": 5247
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6651245321275171,
+      "learning_rate": 3.372928800381326e-05,
+      "loss": 2.0967,
+      "step": 5248
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6118157271258999,
+      "learning_rate": 3.3723434201716664e-05,
+      "loss": 1.9745,
+      "step": 5249
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6605765584694455,
+      "learning_rate": 3.3717579854980223e-05,
+      "loss": 1.9419,
+      "step": 5250
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6217061660600536,
+      "learning_rate": 3.371172496396947e-05,
+      "loss": 2.1354,
+      "step": 5251
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6241051748013673,
+      "learning_rate": 3.3705869529049935e-05,
+      "loss": 2.0294,
+      "step": 5252
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6555008498835914,
+      "learning_rate": 3.3700013550587206e-05,
+      "loss": 2.0088,
+      "step": 5253
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5542913075640968,
+      "learning_rate": 3.369415702894688e-05,
+      "loss": 1.9328,
+      "step": 5254
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6435369289148044,
+      "learning_rate": 3.368829996449463e-05,
+      "loss": 1.9698,
+      "step": 5255
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6417280768547451,
+      "learning_rate": 3.368244235759611e-05,
+      "loss": 2.1303,
+      "step": 5256
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6886953462472054,
+      "learning_rate": 3.3676584208617044e-05,
+      "loss": 1.9449,
+      "step": 5257
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6210181981287859,
+      "learning_rate": 3.3670725517923186e-05,
+      "loss": 1.9421,
+      "step": 5258
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6393147429425503,
+      "learning_rate": 3.366486628588031e-05,
+      "loss": 1.9926,
+      "step": 5259
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6190023792123103,
+      "learning_rate": 3.365900651285424e-05,
+      "loss": 2.1097,
+      "step": 5260
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5924726801705815,
+      "learning_rate": 3.3653146199210816e-05,
+      "loss": 2.0024,
+      "step": 5261
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6666711527369532,
+      "learning_rate": 3.364728534531593e-05,
+      "loss": 1.9541,
+      "step": 5262
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7360866032981873,
+      "learning_rate": 3.36414239515355e-05,
+      "loss": 2.153,
+      "step": 5263
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6590162771064586,
+      "learning_rate": 3.3635562018235464e-05,
+      "loss": 1.9536,
+      "step": 5264
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6593605614296099,
+      "learning_rate": 3.362969954578182e-05,
+      "loss": 1.9707,
+      "step": 5265
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6401941888386696,
+      "learning_rate": 3.3623836534540585e-05,
+      "loss": 1.9432,
+      "step": 5266
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6374019819026574,
+      "learning_rate": 3.36179729848778e-05,
+      "loss": 1.9842,
+      "step": 5267
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5891797848415179,
+      "learning_rate": 3.361210889715957e-05,
+      "loss": 2.1383,
+      "step": 5268
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6791328842955693,
+      "learning_rate": 3.3606244271751985e-05,
+      "loss": 1.9178,
+      "step": 5269
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7093524321758472,
+      "learning_rate": 3.360037910902123e-05,
+      "loss": 1.9443,
+      "step": 5270
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6167685012106178,
+      "learning_rate": 3.359451340933347e-05,
+      "loss": 2.0262,
+      "step": 5271
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7361821044169229,
+      "learning_rate": 3.3588647173054924e-05,
+      "loss": 2.1283,
+      "step": 5272
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7803715391666181,
+      "learning_rate": 3.3582780400551864e-05,
+      "loss": 1.9659,
+      "step": 5273
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.612198908835539,
+      "learning_rate": 3.357691309219057e-05,
+      "loss": 1.9545,
+      "step": 5274
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6046219214048734,
+      "learning_rate": 3.357104524833733e-05,
+      "loss": 1.947,
+      "step": 5275
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6937068032221769,
+      "learning_rate": 3.356517686935855e-05,
+      "loss": 2.1869,
+      "step": 5276
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6028598148032348,
+      "learning_rate": 3.355930795562058e-05,
+      "loss": 2.0009,
+      "step": 5277
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.652012611738902,
+      "learning_rate": 3.3553438507489857e-05,
+      "loss": 1.9558,
+      "step": 5278
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6308769959970407,
+      "learning_rate": 3.354756852533282e-05,
+      "loss": 1.961,
+      "step": 5279
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.553347355467075,
+      "learning_rate": 3.3541698009515974e-05,
+      "loss": 2.1434,
+      "step": 5280
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6119953478046416,
+      "learning_rate": 3.353582696040583e-05,
+      "loss": 1.8929,
+      "step": 5281
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6226083767859955,
+      "learning_rate": 3.352995537836894e-05,
+      "loss": 1.9324,
+      "step": 5282
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5923486480085616,
+      "learning_rate": 3.35240832637719e-05,
+      "loss": 2.0477,
+      "step": 5283
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.753970142294538,
+      "learning_rate": 3.3518210616981314e-05,
+      "loss": 2.1125,
+      "step": 5284
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.614150341462612,
+      "learning_rate": 3.351233743836386e-05,
+      "loss": 1.9301,
+      "step": 5285
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8223604775807146,
+      "learning_rate": 3.35064637282862e-05,
+      "loss": 1.9172,
+      "step": 5286
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7853024617602508,
+      "learning_rate": 3.350058948711506e-05,
+      "loss": 1.9313,
+      "step": 5287
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7462653367780909,
+      "learning_rate": 3.34947147152172e-05,
+      "loss": 2.1277,
+      "step": 5288
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6979593356434883,
+      "learning_rate": 3.3488839412959404e-05,
+      "loss": 1.8937,
+      "step": 5289
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.591860886886428,
+      "learning_rate": 3.34829635807085e-05,
+      "loss": 2.0153,
+      "step": 5290
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.677212997866265,
+      "learning_rate": 3.3477087218831324e-05,
+      "loss": 1.9012,
+      "step": 5291
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7726955864109494,
+      "learning_rate": 3.347121032769476e-05,
+      "loss": 2.1036,
+      "step": 5292
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.684098049955425,
+      "learning_rate": 3.346533290766574e-05,
+      "loss": 1.9699,
+      "step": 5293
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6406872766735322,
+      "learning_rate": 3.3459454959111216e-05,
+      "loss": 1.932,
+      "step": 5294
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7139993032799766,
+      "learning_rate": 3.345357648239816e-05,
+      "loss": 1.9265,
+      "step": 5295
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7681864756959326,
+      "learning_rate": 3.3447697477893604e-05,
+      "loss": 2.1703,
+      "step": 5296
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.718380306151459,
+      "learning_rate": 3.344181794596458e-05,
+      "loss": 1.9471,
+      "step": 5297
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6512072857555886,
+      "learning_rate": 3.343593788697819e-05,
+      "loss": 1.9173,
+      "step": 5298
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6574842632797722,
+      "learning_rate": 3.3430057301301545e-05,
+      "loss": 1.9076,
+      "step": 5299
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.633373791204063,
+      "learning_rate": 3.342417618930178e-05,
+      "loss": 2.1679,
+      "step": 5300
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5966169302392149,
+      "learning_rate": 3.341829455134609e-05,
+      "loss": 1.931,
+      "step": 5301
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6409462902292787,
+      "learning_rate": 3.3412412387801694e-05,
+      "loss": 2.0083,
+      "step": 5302
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5929602334546218,
+      "learning_rate": 3.3406529699035826e-05,
+      "loss": 1.9623,
+      "step": 5303
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5766598760441138,
+      "learning_rate": 3.3400646485415784e-05,
+      "loss": 2.1611,
+      "step": 5304
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6439478219571193,
+      "learning_rate": 3.339476274730886e-05,
+      "loss": 1.8813,
+      "step": 5305
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.69001488862924,
+      "learning_rate": 3.338887848508242e-05,
+      "loss": 1.9284,
+      "step": 5306
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6379634097320331,
+      "learning_rate": 3.338299369910383e-05,
+      "loss": 1.9562,
+      "step": 5307
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7348691268440591,
+      "learning_rate": 3.337710838974051e-05,
+      "loss": 2.1931,
+      "step": 5308
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.586891011009019,
+      "learning_rate": 3.3371222557359894e-05,
+      "loss": 1.932,
+      "step": 5309
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7085664646717423,
+      "learning_rate": 3.336533620232946e-05,
+      "loss": 1.9048,
+      "step": 5310
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6181471965254565,
+      "learning_rate": 3.3359449325016725e-05,
+      "loss": 1.911,
+      "step": 5311
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5879129201872632,
+      "learning_rate": 3.335356192578923e-05,
+      "loss": 2.1318,
+      "step": 5312
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6472406361103712,
+      "learning_rate": 3.334767400501455e-05,
+      "loss": 1.9394,
+      "step": 5313
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5692944377047722,
+      "learning_rate": 3.334178556306028e-05,
+      "loss": 1.9983,
+      "step": 5314
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5861367237346459,
+      "learning_rate": 3.333589660029407e-05,
+      "loss": 1.8992,
+      "step": 5315
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5696021726610209,
+      "learning_rate": 3.3330007117083596e-05,
+      "loss": 2.1327,
+      "step": 5316
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5841289705934051,
+      "learning_rate": 3.3324117113796554e-05,
+      "loss": 1.9826,
+      "step": 5317
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5924441354951783,
+      "learning_rate": 3.3318226590800676e-05,
+      "loss": 1.911,
+      "step": 5318
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5634746192330765,
+      "learning_rate": 3.331233554846375e-05,
+      "loss": 1.9629,
+      "step": 5319
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6060996788835779,
+      "learning_rate": 3.330644398715357e-05,
+      "loss": 2.0848,
+      "step": 5320
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6252072286373167,
+      "learning_rate": 3.3300551907237966e-05,
+      "loss": 2.0015,
+      "step": 5321
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5882399546778966,
+      "learning_rate": 3.3294659309084795e-05,
+      "loss": 1.8883,
+      "step": 5322
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6282342767627428,
+      "learning_rate": 3.328876619306199e-05,
+      "loss": 1.9294,
+      "step": 5323
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7159374445704546,
+      "learning_rate": 3.3282872559537445e-05,
+      "loss": 2.1121,
+      "step": 5324
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6241206143967585,
+      "learning_rate": 3.327697840887914e-05,
+      "loss": 1.9856,
+      "step": 5325
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6719809961015538,
+      "learning_rate": 3.327108374145507e-05,
+      "loss": 1.9072,
+      "step": 5326
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6331125481645862,
+      "learning_rate": 3.326518855763327e-05,
+      "loss": 2.0484,
+      "step": 5327
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7214303477512397,
+      "learning_rate": 3.325929285778179e-05,
+      "loss": 2.1567,
+      "step": 5328
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5861930785042947,
+      "learning_rate": 3.325339664226873e-05,
+      "loss": 1.9386,
+      "step": 5329
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5961781830149099,
+      "learning_rate": 3.3247499911462204e-05,
+      "loss": 1.9516,
+      "step": 5330
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7115045127246735,
+      "learning_rate": 3.324160266573038e-05,
+      "loss": 1.9597,
+      "step": 5331
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7211409242964176,
+      "learning_rate": 3.323570490544145e-05,
+      "loss": 2.1304,
+      "step": 5332
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8215052802825855,
+      "learning_rate": 3.3229806630963626e-05,
+      "loss": 1.9903,
+      "step": 5333
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6843410490338113,
+      "learning_rate": 3.322390784266517e-05,
+      "loss": 1.969,
+      "step": 5334
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6323225278612354,
+      "learning_rate": 3.321800854091435e-05,
+      "loss": 1.9766,
+      "step": 5335
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.710342787216553,
+      "learning_rate": 3.3212108726079504e-05,
+      "loss": 2.1685,
+      "step": 5336
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6274321409641549,
+      "learning_rate": 3.320620839852897e-05,
+      "loss": 1.9584,
+      "step": 5337
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6616181259084215,
+      "learning_rate": 3.3200307558631145e-05,
+      "loss": 1.8952,
+      "step": 5338
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6943513058564507,
+      "learning_rate": 3.319440620675442e-05,
+      "loss": 2.011,
+      "step": 5339
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6387000761587426,
+      "learning_rate": 3.318850434326726e-05,
+      "loss": 2.0719,
+      "step": 5340
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7940324939661112,
+      "learning_rate": 3.318260196853813e-05,
+      "loss": 1.9866,
+      "step": 5341
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7355360939558231,
+      "learning_rate": 3.3176699082935545e-05,
+      "loss": 1.9287,
+      "step": 5342
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5906442451607258,
+      "learning_rate": 3.317079568682804e-05,
+      "loss": 1.8854,
+      "step": 5343
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9712313172517418,
+      "learning_rate": 3.31648917805842e-05,
+      "loss": 2.1572,
+      "step": 5344
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6233534474131831,
+      "learning_rate": 3.315898736457263e-05,
+      "loss": 2.0169,
+      "step": 5345
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8243955919054732,
+      "learning_rate": 3.3153082439161955e-05,
+      "loss": 1.9047,
+      "step": 5346
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7424397036822097,
+      "learning_rate": 3.314717700472085e-05,
+      "loss": 1.9045,
+      "step": 5347
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.762248712150822,
+      "learning_rate": 3.314127106161801e-05,
+      "loss": 2.1389,
+      "step": 5348
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.763049404309697,
+      "learning_rate": 3.3135364610222176e-05,
+      "loss": 1.8824,
+      "step": 5349
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7094588973745759,
+      "learning_rate": 3.312945765090211e-05,
+      "loss": 1.9473,
+      "step": 5350
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6796285764005191,
+      "learning_rate": 3.31235501840266e-05,
+      "loss": 1.9082,
+      "step": 5351
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8510634566621629,
+      "learning_rate": 3.311764220996449e-05,
+      "loss": 2.2012,
+      "step": 5352
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7778469676119967,
+      "learning_rate": 3.311173372908462e-05,
+      "loss": 1.9048,
+      "step": 5353
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.614625148193122,
+      "learning_rate": 3.310582474175589e-05,
+      "loss": 1.97,
+      "step": 5354
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7622313466563077,
+      "learning_rate": 3.309991524834723e-05,
+      "loss": 1.9458,
+      "step": 5355
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6917900793142051,
+      "learning_rate": 3.309400524922758e-05,
+      "loss": 2.143,
+      "step": 5356
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7178232395697096,
+      "learning_rate": 3.308809474476593e-05,
+      "loss": 1.9195,
+      "step": 5357
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7657897095043855,
+      "learning_rate": 3.30821837353313e-05,
+      "loss": 2.0424,
+      "step": 5358
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7310202310255814,
+      "learning_rate": 3.3076272221292735e-05,
+      "loss": 1.9099,
+      "step": 5359
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.647379420288956,
+      "learning_rate": 3.307036020301932e-05,
+      "loss": 2.0979,
+      "step": 5360
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8142841419592612,
+      "learning_rate": 3.306444768088016e-05,
+      "loss": 1.9309,
+      "step": 5361
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7231203214133937,
+      "learning_rate": 3.305853465524441e-05,
+      "loss": 1.9071,
+      "step": 5362
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7493323650010434,
+      "learning_rate": 3.305262112648123e-05,
+      "loss": 1.9176,
+      "step": 5363
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6321117970619141,
+      "learning_rate": 3.3046707094959836e-05,
+      "loss": 2.143,
+      "step": 5364
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8904915285736467,
+      "learning_rate": 3.304079256104945e-05,
+      "loss": 1.8658,
+      "step": 5365
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5726260305969468,
+      "learning_rate": 3.3034877525119365e-05,
+      "loss": 1.9868,
+      "step": 5366
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8061666037883968,
+      "learning_rate": 3.3028961987538866e-05,
+      "loss": 1.9327,
+      "step": 5367
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6501762678587258,
+      "learning_rate": 3.3023045948677276e-05,
+      "loss": 2.1338,
+      "step": 5368
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6640013442665536,
+      "learning_rate": 3.3017129408903975e-05,
+      "loss": 1.9243,
+      "step": 5369
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6466435086930428,
+      "learning_rate": 3.301121236858835e-05,
+      "loss": 2.025,
+      "step": 5370
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6449506710824027,
+      "learning_rate": 3.3005294828099834e-05,
+      "loss": 1.9383,
+      "step": 5371
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8232486897759911,
+      "learning_rate": 3.2999376787807864e-05,
+      "loss": 2.1042,
+      "step": 5372
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7443728813879235,
+      "learning_rate": 3.2993458248081944e-05,
+      "loss": 1.936,
+      "step": 5373
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7590241250146753,
+      "learning_rate": 3.2987539209291584e-05,
+      "loss": 1.9107,
+      "step": 5374
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5787622383389469,
+      "learning_rate": 3.2981619671806335e-05,
+      "loss": 1.8928,
+      "step": 5375
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.825712362311262,
+      "learning_rate": 3.297569963599579e-05,
+      "loss": 2.1247,
+      "step": 5376
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5490339581906496,
+      "learning_rate": 3.2969779102229545e-05,
+      "loss": 2.0606,
+      "step": 5377
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6229286280893256,
+      "learning_rate": 3.2963858070877254e-05,
+      "loss": 1.8879,
+      "step": 5378
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6643637550041719,
+      "learning_rate": 3.295793654230858e-05,
+      "loss": 1.9659,
+      "step": 5379
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6463529286718409,
+      "learning_rate": 3.2952014516893234e-05,
+      "loss": 2.1603,
+      "step": 5380
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6484300848430331,
+      "learning_rate": 3.294609199500097e-05,
+      "loss": 1.9299,
+      "step": 5381
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6379223253281268,
+      "learning_rate": 3.2940168977001526e-05,
+      "loss": 1.9676,
+      "step": 5382
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5928337322562393,
+      "learning_rate": 3.293424546326472e-05,
+      "loss": 2.0145,
+      "step": 5383
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6034318649683867,
+      "learning_rate": 3.292832145416037e-05,
+      "loss": 2.1486,
+      "step": 5384
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6452086210617555,
+      "learning_rate": 3.292239695005835e-05,
+      "loss": 1.9154,
+      "step": 5385
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.565199029063573,
+      "learning_rate": 3.291647195132853e-05,
+      "loss": 1.8916,
+      "step": 5386
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.611928703412391,
+      "learning_rate": 3.291054645834085e-05,
+      "loss": 1.9355,
+      "step": 5387
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.641325278371438,
+      "learning_rate": 3.2904620471465265e-05,
+      "loss": 2.0621,
+      "step": 5388
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5642565882519102,
+      "learning_rate": 3.2898693991071745e-05,
+      "loss": 2.0146,
+      "step": 5389
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6294307780830736,
+      "learning_rate": 3.28927670175303e-05,
+      "loss": 1.8861,
+      "step": 5390
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6456823197143003,
+      "learning_rate": 3.2886839551211e-05,
+      "loss": 2.0031,
+      "step": 5391
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6659878202889287,
+      "learning_rate": 3.28809115924839e-05,
+      "loss": 2.162,
+      "step": 5392
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6324399690362816,
+      "learning_rate": 3.2874983141719115e-05,
+      "loss": 1.9593,
+      "step": 5393
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.602348209580063,
+      "learning_rate": 3.286905419928678e-05,
+      "loss": 1.9467,
+      "step": 5394
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5756152802113174,
+      "learning_rate": 3.286312476555707e-05,
+      "loss": 1.9929,
+      "step": 5395
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6515698233982119,
+      "learning_rate": 3.2857194840900176e-05,
+      "loss": 2.1354,
+      "step": 5396
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6215240250757839,
+      "learning_rate": 3.285126442568633e-05,
+      "loss": 1.9331,
+      "step": 5397
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6643247006869244,
+      "learning_rate": 3.284533352028579e-05,
+      "loss": 1.8621,
+      "step": 5398
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6937257105518946,
+      "learning_rate": 3.2839402125068855e-05,
+      "loss": 1.9668,
+      "step": 5399
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6590564962345531,
+      "learning_rate": 3.283347024040583e-05,
+      "loss": 2.1266,
+      "step": 5400
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6415555727897525,
+      "learning_rate": 3.282753786666708e-05,
+      "loss": 1.9967,
+      "step": 5401
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7811069309996574,
+      "learning_rate": 3.2821605004222986e-05,
+      "loss": 1.9463,
+      "step": 5402
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6789272133585802,
+      "learning_rate": 3.2815671653443966e-05,
+      "loss": 1.9341,
+      "step": 5403
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6393927864589408,
+      "learning_rate": 3.2809737814700446e-05,
+      "loss": 2.1274,
+      "step": 5404
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.720660913358279,
+      "learning_rate": 3.280380348836292e-05,
+      "loss": 1.9389,
+      "step": 5405
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5962042049683706,
+      "learning_rate": 3.279786867480189e-05,
+      "loss": 1.9441,
+      "step": 5406
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6429573779968364,
+      "learning_rate": 3.279193337438786e-05,
+      "loss": 1.9165,
+      "step": 5407
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5831571987918922,
+      "learning_rate": 3.278599758749143e-05,
+      "loss": 2.0326,
+      "step": 5408
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6558730896963416,
+      "learning_rate": 3.278006131448319e-05,
+      "loss": 2.1279,
+      "step": 5409
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7726335095758782,
+      "learning_rate": 3.2774124555733764e-05,
+      "loss": 1.9557,
+      "step": 5410
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.634100231303346,
+      "learning_rate": 3.27681873116138e-05,
+      "loss": 1.9333,
+      "step": 5411
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8239728322209793,
+      "learning_rate": 3.2762249582493984e-05,
+      "loss": 2.1315,
+      "step": 5412
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6687361561821789,
+      "learning_rate": 3.2756311368745053e-05,
+      "loss": 1.92,
+      "step": 5413
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6801857455707073,
+      "learning_rate": 3.275037267073773e-05,
+      "loss": 1.9852,
+      "step": 5414
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8131654143789754,
+      "learning_rate": 3.274443348884281e-05,
+      "loss": 1.9139,
+      "step": 5415
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7387883901254066,
+      "learning_rate": 3.273849382343107e-05,
+      "loss": 2.1647,
+      "step": 5416
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8309319043134368,
+      "learning_rate": 3.2732553674873396e-05,
+      "loss": 1.9756,
+      "step": 5417
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6997358030108618,
+      "learning_rate": 3.272661304354062e-05,
+      "loss": 1.9328,
+      "step": 5418
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6219879378199415,
+      "learning_rate": 3.2720671929803645e-05,
+      "loss": 1.9756,
+      "step": 5419
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7561806188237168,
+      "learning_rate": 3.271473033403341e-05,
+      "loss": 2.031,
+      "step": 5420
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6677278711034889,
+      "learning_rate": 3.270878825660087e-05,
+      "loss": 2.1608,
+      "step": 5421
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6945289703547172,
+      "learning_rate": 3.2702845697877e-05,
+      "loss": 1.9434,
+      "step": 5422
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7416432792052098,
+      "learning_rate": 3.2696902658232835e-05,
+      "loss": 1.9324,
+      "step": 5423
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7182562747949965,
+      "learning_rate": 3.269095913803942e-05,
+      "loss": 2.1002,
+      "step": 5424
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6315693837351214,
+      "learning_rate": 3.268501513766783e-05,
+      "loss": 1.8572,
+      "step": 5425
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8830270748114272,
+      "learning_rate": 3.267907065748917e-05,
+      "loss": 2.009,
+      "step": 5426
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5447074759827956,
+      "learning_rate": 3.2673125697874585e-05,
+      "loss": 1.9355,
+      "step": 5427
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8267359268744935,
+      "learning_rate": 3.266718025919524e-05,
+      "loss": 1.9089,
+      "step": 5428
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.739055047598812,
+      "learning_rate": 3.266123434182233e-05,
+      "loss": 2.1481,
+      "step": 5429
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6565457511060875,
+      "learning_rate": 3.265528794612709e-05,
+      "loss": 1.8831,
+      "step": 5430
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8211801035284677,
+      "learning_rate": 3.264934107248079e-05,
+      "loss": 1.9369,
+      "step": 5431
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6170832083440241,
+      "learning_rate": 3.264339372125468e-05,
+      "loss": 2.0434,
+      "step": 5432
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8385473880863283,
+      "learning_rate": 3.263744589282011e-05,
+      "loss": 2.1591,
+      "step": 5433
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6369368375090488,
+      "learning_rate": 3.2631497587548415e-05,
+      "loss": 1.912,
+      "step": 5434
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6332721939866883,
+      "learning_rate": 3.262554880581098e-05,
+      "loss": 1.9064,
+      "step": 5435
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8214737652989269,
+      "learning_rate": 3.26195995479792e-05,
+      "loss": 2.0765,
+      "step": 5436
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6195597384186993,
+      "learning_rate": 3.261364981442452e-05,
+      "loss": 1.9174,
+      "step": 5437
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7362173295151322,
+      "learning_rate": 3.26076996055184e-05,
+      "loss": 1.9226,
+      "step": 5438
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6311151273567933,
+      "learning_rate": 3.260174892163235e-05,
+      "loss": 2.0216,
+      "step": 5439
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.554020995852219,
+      "learning_rate": 3.259579776313789e-05,
+      "loss": 1.9296,
+      "step": 5440
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.881589836792152,
+      "learning_rate": 3.258984613040655e-05,
+      "loss": 2.1549,
+      "step": 5441
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5622356008906851,
+      "learning_rate": 3.2583894023809956e-05,
+      "loss": 1.9503,
+      "step": 5442
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7343325030107479,
+      "learning_rate": 3.25779414437197e-05,
+      "loss": 1.9232,
+      "step": 5443
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6680789443046955,
+      "learning_rate": 3.2571988390507416e-05,
+      "loss": 2.1192,
+      "step": 5444
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6236030928499064,
+      "learning_rate": 3.2566034864544804e-05,
+      "loss": 2.0281,
+      "step": 5445
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6963852352563481,
+      "learning_rate": 3.2560080866203545e-05,
+      "loss": 1.9275,
+      "step": 5446
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5962980734923388,
+      "learning_rate": 3.2554126395855385e-05,
+      "loss": 1.9583,
+      "step": 5447
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.618694015992922,
+      "learning_rate": 3.2548171453872076e-05,
+      "loss": 2.1159,
+      "step": 5448
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6097435989779305,
+      "learning_rate": 3.254221604062542e-05,
+      "loss": 1.969,
+      "step": 5449
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5839369802511799,
+      "learning_rate": 3.253626015648723e-05,
+      "loss": 1.976,
+      "step": 5450
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5962427501335629,
+      "learning_rate": 3.253030380182936e-05,
+      "loss": 1.99,
+      "step": 5451
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6226046656599126,
+      "learning_rate": 3.252434697702369e-05,
+      "loss": 2.0126,
+      "step": 5452
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6503967928360542,
+      "learning_rate": 3.2518389682442126e-05,
+      "loss": 2.1454,
+      "step": 5453
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6229639323652619,
+      "learning_rate": 3.25124319184566e-05,
+      "loss": 1.957,
+      "step": 5454
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6585957902523308,
+      "learning_rate": 3.2506473685439096e-05,
+      "loss": 1.9081,
+      "step": 5455
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7721535816273783,
+      "learning_rate": 3.25005149837616e-05,
+      "loss": 2.1421,
+      "step": 5456
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6653778234987971,
+      "learning_rate": 3.249455581379614e-05,
+      "loss": 2.0075,
+      "step": 5457
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6623979170580239,
+      "learning_rate": 3.248859617591477e-05,
+      "loss": 1.963,
+      "step": 5458
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6740516845005219,
+      "learning_rate": 3.248263607048958e-05,
+      "loss": 1.9438,
+      "step": 5459
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6493063185098801,
+      "learning_rate": 3.247667549789268e-05,
+      "loss": 1.9434,
+      "step": 5460
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7555680808154528,
+      "learning_rate": 3.247071445849622e-05,
+      "loss": 2.1238,
+      "step": 5461
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6736019075933498,
+      "learning_rate": 3.246475295267236e-05,
+      "loss": 1.9137,
+      "step": 5462
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7857940747407072,
+      "learning_rate": 3.245879098079331e-05,
+      "loss": 2.0145,
+      "step": 5463
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5621384193787339,
+      "learning_rate": 3.245282854323129e-05,
+      "loss": 1.942,
+      "step": 5464
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7935732787791099,
+      "learning_rate": 3.244686564035857e-05,
+      "loss": 2.1737,
+      "step": 5465
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5655421507494806,
+      "learning_rate": 3.2440902272547435e-05,
+      "loss": 1.9185,
+      "step": 5466
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7009622857266311,
+      "learning_rate": 3.243493844017021e-05,
+      "loss": 1.9592,
+      "step": 5467
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5869157228127593,
+      "learning_rate": 3.242897414359923e-05,
+      "loss": 2.1205,
+      "step": 5468
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7839826632691027,
+      "learning_rate": 3.2423009383206876e-05,
+      "loss": 1.9139,
+      "step": 5469
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6880840664426252,
+      "learning_rate": 3.241704415936555e-05,
+      "loss": 2.0146,
+      "step": 5470
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6882937678105371,
+      "learning_rate": 3.241107847244769e-05,
+      "loss": 1.9057,
+      "step": 5471
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6652235381871626,
+      "learning_rate": 3.240511232282575e-05,
+      "loss": 1.9161,
+      "step": 5472
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6758711507376635,
+      "learning_rate": 3.239914571087223e-05,
+      "loss": 2.1404,
+      "step": 5473
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6828744622650211,
+      "learning_rate": 3.239317863695965e-05,
+      "loss": 1.9411,
+      "step": 5474
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7109241498687872,
+      "learning_rate": 3.238721110146055e-05,
+      "loss": 1.946,
+      "step": 5475
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6202686161398212,
+      "learning_rate": 3.2381243104747515e-05,
+      "loss": 2.005,
+      "step": 5476
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6073099042464479,
+      "learning_rate": 3.237527464719315e-05,
+      "loss": 2.1776,
+      "step": 5477
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6354792856647093,
+      "learning_rate": 3.2369305729170104e-05,
+      "loss": 1.9666,
+      "step": 5478
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6266403012475132,
+      "learning_rate": 3.236333635105101e-05,
+      "loss": 1.9383,
+      "step": 5479
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7119905599063874,
+      "learning_rate": 3.235736651320858e-05,
+      "loss": 2.1329,
+      "step": 5480
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5632344650770692,
+      "learning_rate": 3.235139621601554e-05,
+      "loss": 1.8772,
+      "step": 5481
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6895694931197734,
+      "learning_rate": 3.234542545984464e-05,
+      "loss": 1.9847,
+      "step": 5482
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.57348800410612,
+      "learning_rate": 3.233945424506864e-05,
+      "loss": 1.9015,
+      "step": 5483
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7073989494594237,
+      "learning_rate": 3.233348257206036e-05,
+      "loss": 1.9312,
+      "step": 5484
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6053516722515452,
+      "learning_rate": 3.232751044119264e-05,
+      "loss": 2.1203,
+      "step": 5485
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7736522186800224,
+      "learning_rate": 3.232153785283835e-05,
+      "loss": 1.965,
+      "step": 5486
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.66586357634712,
+      "learning_rate": 3.2315564807370366e-05,
+      "loss": 1.8798,
+      "step": 5487
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.654245001755994,
+      "learning_rate": 3.230959130516161e-05,
+      "loss": 1.9973,
+      "step": 5488
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6061864268640438,
+      "learning_rate": 3.230361734658505e-05,
+      "loss": 2.0671,
+      "step": 5489
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7680848019920755,
+      "learning_rate": 3.229764293201366e-05,
+      "loss": 1.9135,
+      "step": 5490
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6036311509983753,
+      "learning_rate": 3.2291668061820436e-05,
+      "loss": 1.884,
+      "step": 5491
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7897024667505609,
+      "learning_rate": 3.228569273637841e-05,
+      "loss": 1.9369,
+      "step": 5492
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6258449289895553,
+      "learning_rate": 3.227971695606067e-05,
+      "loss": 2.0891,
+      "step": 5493
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6916593594434459,
+      "learning_rate": 3.227374072124029e-05,
+      "loss": 2.0673,
+      "step": 5494
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6518117188330713,
+      "learning_rate": 3.22677640322904e-05,
+      "loss": 1.9512,
+      "step": 5495
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6092611170364233,
+      "learning_rate": 3.226178688958415e-05,
+      "loss": 1.9958,
+      "step": 5496
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6112249442078188,
+      "learning_rate": 3.225580929349469e-05,
+      "loss": 2.1535,
+      "step": 5497
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6476680973428768,
+      "learning_rate": 3.224983124439526e-05,
+      "loss": 1.905,
+      "step": 5498
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6653773667041585,
+      "learning_rate": 3.224385274265909e-05,
+      "loss": 1.8951,
+      "step": 5499
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6582542361566793,
+      "learning_rate": 3.2237873788659426e-05,
+      "loss": 2.0665,
+      "step": 5500
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6227587498986435,
+      "learning_rate": 3.223189438276957e-05,
+      "loss": 2.061,
+      "step": 5501
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.676275070605795,
+      "learning_rate": 3.222591452536283e-05,
+      "loss": 1.8989,
+      "step": 5502
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5952723171790377,
+      "learning_rate": 3.2219934216812577e-05,
+      "loss": 1.9088,
+      "step": 5503
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6425753209429459,
+      "learning_rate": 3.221395345749216e-05,
+      "loss": 1.9099,
+      "step": 5504
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6438114003272419,
+      "learning_rate": 3.220797224777499e-05,
+      "loss": 2.1085,
+      "step": 5505
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5728218918483337,
+      "learning_rate": 3.220199058803451e-05,
+      "loss": 1.9183,
+      "step": 5506
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6184371594927183,
+      "learning_rate": 3.219600847864417e-05,
+      "loss": 2.0645,
+      "step": 5507
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.627973492224833,
+      "learning_rate": 3.219002591997746e-05,
+      "loss": 1.9195,
+      "step": 5508
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7046136339300492,
+      "learning_rate": 3.218404291240789e-05,
+      "loss": 2.1546,
+      "step": 5509
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7030012737852505,
+      "learning_rate": 3.217805945630901e-05,
+      "loss": 1.9139,
+      "step": 5510
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6097944281340764,
+      "learning_rate": 3.2172075552054396e-05,
+      "loss": 1.9455,
+      "step": 5511
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6250208566547525,
+      "learning_rate": 3.2166091200017634e-05,
+      "loss": 1.9487,
+      "step": 5512
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6892774558588758,
+      "learning_rate": 3.216010640057236e-05,
+      "loss": 2.1717,
+      "step": 5513
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6820070281238775,
+      "learning_rate": 3.215412115409223e-05,
+      "loss": 1.932,
+      "step": 5514
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.719605332206247,
+      "learning_rate": 3.2148135460950926e-05,
+      "loss": 1.9452,
+      "step": 5515
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7167754269098408,
+      "learning_rate": 3.214214932152216e-05,
+      "loss": 1.9655,
+      "step": 5516
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7925441305797717,
+      "learning_rate": 3.213616273617967e-05,
+      "loss": 2.1331,
+      "step": 5517
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9049723788698983,
+      "learning_rate": 3.213017570529723e-05,
+      "loss": 1.9337,
+      "step": 5518
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7241100713603708,
+      "learning_rate": 3.212418822924862e-05,
+      "loss": 1.9803,
+      "step": 5519
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8024369542836135,
+      "learning_rate": 3.211820030840768e-05,
+      "loss": 1.8979,
+      "step": 5520
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8232868303700673,
+      "learning_rate": 3.211221194314825e-05,
+      "loss": 2.1254,
+      "step": 5521
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8988761093127632,
+      "learning_rate": 3.2106223133844207e-05,
+      "loss": 1.9475,
+      "step": 5522
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7382648375615933,
+      "learning_rate": 3.2100233880869466e-05,
+      "loss": 1.9,
+      "step": 5523
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7229839387085425,
+      "learning_rate": 3.209424418459795e-05,
+      "loss": 1.9229,
+      "step": 5524
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8075262864280117,
+      "learning_rate": 3.208825404540363e-05,
+      "loss": 2.1694,
+      "step": 5525
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0498941409031897,
+      "learning_rate": 3.2082263463660484e-05,
+      "loss": 1.9127,
+      "step": 5526
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7824956633924119,
+      "learning_rate": 3.207627243974254e-05,
+      "loss": 1.9317,
+      "step": 5527
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.695756948294812,
+      "learning_rate": 3.207028097402385e-05,
+      "loss": 1.9239,
+      "step": 5528
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.869502937624567,
+      "learning_rate": 3.2064289066878456e-05,
+      "loss": 2.1836,
+      "step": 5529
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6652202611194128,
+      "learning_rate": 3.205829671868048e-05,
+      "loss": 1.911,
+      "step": 5530
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7013150867735364,
+      "learning_rate": 3.205230392980404e-05,
+      "loss": 1.9228,
+      "step": 5531
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5496800131590572,
+      "learning_rate": 3.20463107006233e-05,
+      "loss": 2.0032,
+      "step": 5532
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7706156566641672,
+      "learning_rate": 3.2040317031512436e-05,
+      "loss": 2.1507,
+      "step": 5533
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6176484443005774,
+      "learning_rate": 3.2034322922845655e-05,
+      "loss": 1.9458,
+      "step": 5534
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5469379133064656,
+      "learning_rate": 3.2028328374997196e-05,
+      "loss": 1.9403,
+      "step": 5535
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.639909366424169,
+      "learning_rate": 3.202233338834133e-05,
+      "loss": 1.9298,
+      "step": 5536
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6565769827278872,
+      "learning_rate": 3.201633796325233e-05,
+      "loss": 2.1173,
+      "step": 5537
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5550613749009332,
+      "learning_rate": 3.201034210010453e-05,
+      "loss": 2.0534,
+      "step": 5538
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6131837862882319,
+      "learning_rate": 3.200434579927228e-05,
+      "loss": 1.9521,
+      "step": 5539
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.653824083945356,
+      "learning_rate": 3.1998349061129936e-05,
+      "loss": 1.883,
+      "step": 5540
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6178267137072395,
+      "learning_rate": 3.199235188605192e-05,
+      "loss": 2.1442,
+      "step": 5541
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7666114406094763,
+      "learning_rate": 3.198635427441266e-05,
+      "loss": 1.9135,
+      "step": 5542
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6891744295974432,
+      "learning_rate": 3.198035622658658e-05,
+      "loss": 1.9241,
+      "step": 5543
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8053479588396891,
+      "learning_rate": 3.1974357742948196e-05,
+      "loss": 2.0171,
+      "step": 5544
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.645338650666459,
+      "learning_rate": 3.1968358823872006e-05,
+      "loss": 2.1128,
+      "step": 5545
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5481985921662712,
+      "learning_rate": 3.196235946973255e-05,
+      "loss": 1.9602,
+      "step": 5546
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6978525522805725,
+      "learning_rate": 3.195635968090439e-05,
+      "loss": 1.92,
+      "step": 5547
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6134188586923429,
+      "learning_rate": 3.195035945776212e-05,
+      "loss": 1.9142,
+      "step": 5548
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7686567179918987,
+      "learning_rate": 3.194435880068034e-05,
+      "loss": 2.1173,
+      "step": 5549
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6178725405164761,
+      "learning_rate": 3.193835771003373e-05,
+      "loss": 1.9653,
+      "step": 5550
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6683409509862176,
+      "learning_rate": 3.193235618619694e-05,
+      "loss": 1.9064,
+      "step": 5551
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6730131551392984,
+      "learning_rate": 3.192635422954467e-05,
+      "loss": 1.898,
+      "step": 5552
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6284940050193553,
+      "learning_rate": 3.1920351840451657e-05,
+      "loss": 2.1349,
+      "step": 5553
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6939775343718775,
+      "learning_rate": 3.191434901929265e-05,
+      "loss": 1.9622,
+      "step": 5554
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6332149772626218,
+      "learning_rate": 3.190834576644242e-05,
+      "loss": 1.9276,
+      "step": 5555
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6489657017125712,
+      "learning_rate": 3.190234208227578e-05,
+      "loss": 1.9118,
+      "step": 5556
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6571878936188528,
+      "learning_rate": 3.1896337967167586e-05,
+      "loss": 2.1403,
+      "step": 5557
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5664121402087847,
+      "learning_rate": 3.1890333421492674e-05,
+      "loss": 1.9243,
+      "step": 5558
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7547122291313327,
+      "learning_rate": 3.1884328445625937e-05,
+      "loss": 1.8924,
+      "step": 5559
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6396239459256722,
+      "learning_rate": 3.18783230399423e-05,
+      "loss": 1.9602,
+      "step": 5560
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.635847358947385,
+      "learning_rate": 3.187231720481669e-05,
+      "loss": 2.1302,
+      "step": 5561
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6876565785077545,
+      "learning_rate": 3.186631094062409e-05,
+      "loss": 1.9288,
+      "step": 5562
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6089386780474755,
+      "learning_rate": 3.186030424773949e-05,
+      "loss": 2.0301,
+      "step": 5563
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5876104511747701,
+      "learning_rate": 3.1854297126537916e-05,
+      "loss": 1.9272,
+      "step": 5564
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6949961849253385,
+      "learning_rate": 3.184828957739441e-05,
+      "loss": 2.1834,
+      "step": 5565
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5832294479535828,
+      "learning_rate": 3.184228160068405e-05,
+      "loss": 1.9594,
+      "step": 5566
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.71915185201773,
+      "learning_rate": 3.183627319678195e-05,
+      "loss": 1.9801,
+      "step": 5567
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5781935776393736,
+      "learning_rate": 3.1830264366063226e-05,
+      "loss": 1.9394,
+      "step": 5568
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6775409416722471,
+      "learning_rate": 3.182425510890303e-05,
+      "loss": 2.1399,
+      "step": 5569
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6175411048432007,
+      "learning_rate": 3.181824542567656e-05,
+      "loss": 1.916,
+      "step": 5570
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6560404504115196,
+      "learning_rate": 3.181223531675901e-05,
+      "loss": 1.9763,
+      "step": 5571
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7631719080507129,
+      "learning_rate": 3.180622478252564e-05,
+      "loss": 1.9363,
+      "step": 5572
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7248499098020429,
+      "learning_rate": 3.1800213823351685e-05,
+      "loss": 2.125,
+      "step": 5573
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7869909938454049,
+      "learning_rate": 3.1794202439612444e-05,
+      "loss": 1.9454,
+      "step": 5574
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6510848517505108,
+      "learning_rate": 3.1788190631683244e-05,
+      "loss": 1.9797,
+      "step": 5575
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6339881798816766,
+      "learning_rate": 3.178217839993941e-05,
+      "loss": 1.9007,
+      "step": 5576
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7958293017951832,
+      "learning_rate": 3.1776165744756314e-05,
+      "loss": 2.1362,
+      "step": 5577
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6717389265772505,
+      "learning_rate": 3.1770152666509353e-05,
+      "loss": 1.9518,
+      "step": 5578
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6488673600725035,
+      "learning_rate": 3.176413916557395e-05,
+      "loss": 1.9405,
+      "step": 5579
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6598453826940345,
+      "learning_rate": 3.1758125242325554e-05,
+      "loss": 1.9191,
+      "step": 5580
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6914219150642945,
+      "learning_rate": 3.175211089713963e-05,
+      "loss": 2.1558,
+      "step": 5581
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6103646244784562,
+      "learning_rate": 3.174609613039169e-05,
+      "loss": 1.8826,
+      "step": 5582
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6979368666189041,
+      "learning_rate": 3.174008094245725e-05,
+      "loss": 1.9413,
+      "step": 5583
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6121559807410863,
+      "learning_rate": 3.173406533371187e-05,
+      "loss": 1.8936,
+      "step": 5584
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.710151507960139,
+      "learning_rate": 3.172804930453113e-05,
+      "loss": 2.101,
+      "step": 5585
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6376532148603582,
+      "learning_rate": 3.172203285529063e-05,
+      "loss": 1.9711,
+      "step": 5586
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5740796587668582,
+      "learning_rate": 3.1716015986366005e-05,
+      "loss": 1.9534,
+      "step": 5587
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5661990887156229,
+      "learning_rate": 3.170999869813291e-05,
+      "loss": 2.0021,
+      "step": 5588
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.657153919988527,
+      "learning_rate": 3.170398099096703e-05,
+      "loss": 2.1396,
+      "step": 5589
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6297288853270356,
+      "learning_rate": 3.169796286524408e-05,
+      "loss": 1.989,
+      "step": 5590
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5844081551829533,
+      "learning_rate": 3.169194432133979e-05,
+      "loss": 1.9137,
+      "step": 5591
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6411655285201882,
+      "learning_rate": 3.1685925359629924e-05,
+      "loss": 1.9242,
+      "step": 5592
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5957149856691657,
+      "learning_rate": 3.1679905980490274e-05,
+      "loss": 2.1384,
+      "step": 5593
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6402308920806045,
+      "learning_rate": 3.167388618429665e-05,
+      "loss": 1.9841,
+      "step": 5594
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5446746250188561,
+      "learning_rate": 3.166786597142489e-05,
+      "loss": 1.874,
+      "step": 5595
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.597463264678572,
+      "learning_rate": 3.166184534225087e-05,
+      "loss": 1.8861,
+      "step": 5596
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6503667002407351,
+      "learning_rate": 3.165582429715049e-05,
+      "loss": 2.1238,
+      "step": 5597
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5549774539797778,
+      "learning_rate": 3.164980283649964e-05,
+      "loss": 1.9113,
+      "step": 5598
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6981562770849379,
+      "learning_rate": 3.164378096067428e-05,
+      "loss": 1.9204,
+      "step": 5599
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6651272135530678,
+      "learning_rate": 3.1637758670050396e-05,
+      "loss": 1.9922,
+      "step": 5600
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6938228074042723,
+      "learning_rate": 3.1631735965003955e-05,
+      "loss": 2.1331,
+      "step": 5601
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6700507001131234,
+      "learning_rate": 3.1625712845911e-05,
+      "loss": 1.9222,
+      "step": 5602
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5687402131253495,
+      "learning_rate": 3.161968931314757e-05,
+      "loss": 1.9802,
+      "step": 5603
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7408564606199575,
+      "learning_rate": 3.1613665367089744e-05,
+      "loss": 1.9038,
+      "step": 5604
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6017354416684018,
+      "learning_rate": 3.160764100811362e-05,
+      "loss": 2.0945,
+      "step": 5605
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6633917462520157,
+      "learning_rate": 3.160161623659532e-05,
+      "loss": 1.9678,
+      "step": 5606
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5746071633726025,
+      "learning_rate": 3.1595591052911e-05,
+      "loss": 1.9158,
+      "step": 5607
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6561268664491572,
+      "learning_rate": 3.1589565457436834e-05,
+      "loss": 1.9476,
+      "step": 5608
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6480544738128139,
+      "learning_rate": 3.1583539450549036e-05,
+      "loss": 2.1389,
+      "step": 5609
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.649009919750529,
+      "learning_rate": 3.157751303262381e-05,
+      "loss": 1.9035,
+      "step": 5610
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6935282012022059,
+      "learning_rate": 3.1571486204037426e-05,
+      "loss": 1.9505,
+      "step": 5611
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6754449665646393,
+      "learning_rate": 3.1565458965166166e-05,
+      "loss": 2.0261,
+      "step": 5612
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8167251778993289,
+      "learning_rate": 3.1559431316386335e-05,
+      "loss": 2.091,
+      "step": 5613
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6611146503877938,
+      "learning_rate": 3.155340325807425e-05,
+      "loss": 1.9362,
+      "step": 5614
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7208759102287223,
+      "learning_rate": 3.1547374790606293e-05,
+      "loss": 1.8937,
+      "step": 5615
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6500901377578676,
+      "learning_rate": 3.154134591435882e-05,
+      "loss": 1.8853,
+      "step": 5616
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7823771440414354,
+      "learning_rate": 3.153531662970825e-05,
+      "loss": 2.1279,
+      "step": 5617
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8283768467376464,
+      "learning_rate": 3.1529286937031016e-05,
+      "loss": 1.9031,
+      "step": 5618
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.835448366300111,
+      "learning_rate": 3.1523256836703573e-05,
+      "loss": 1.9898,
+      "step": 5619
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7975719300266685,
+      "learning_rate": 3.151722632910241e-05,
+      "loss": 1.8939,
+      "step": 5620
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9297861943058162,
+      "learning_rate": 3.151119541460403e-05,
+      "loss": 2.1439,
+      "step": 5621
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6940642597277984,
+      "learning_rate": 3.150516409358498e-05,
+      "loss": 1.9605,
+      "step": 5622
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8434133156426439,
+      "learning_rate": 3.14991323664218e-05,
+      "loss": 1.9468,
+      "step": 5623
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7307312587276985,
+      "learning_rate": 3.149310023349109e-05,
+      "loss": 1.9061,
+      "step": 5624
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6250806731301657,
+      "learning_rate": 3.148706769516947e-05,
+      "loss": 1.9862,
+      "step": 5625
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7314703992850408,
+      "learning_rate": 3.148103475183356e-05,
+      "loss": 2.1846,
+      "step": 5626
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7413703977005917,
+      "learning_rate": 3.147500140386002e-05,
+      "loss": 1.9285,
+      "step": 5627
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6007459767992146,
+      "learning_rate": 3.146896765162554e-05,
+      "loss": 1.9477,
+      "step": 5628
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7063260244334643,
+      "learning_rate": 3.1462933495506844e-05,
+      "loss": 2.1046,
+      "step": 5629
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6583464691097333,
+      "learning_rate": 3.1456898935880654e-05,
+      "loss": 1.9515,
+      "step": 5630
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.595900724631551,
+      "learning_rate": 3.1450863973123735e-05,
+      "loss": 2.0259,
+      "step": 5631
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7036715175892024,
+      "learning_rate": 3.1444828607612875e-05,
+      "loss": 1.9199,
+      "step": 5632
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7564329608678626,
+      "learning_rate": 3.1438792839724895e-05,
+      "loss": 2.1449,
+      "step": 5633
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5711565792424932,
+      "learning_rate": 3.1432756669836624e-05,
+      "loss": 1.9007,
+      "step": 5634
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8805444805309195,
+      "learning_rate": 3.1426720098324936e-05,
+      "loss": 1.95,
+      "step": 5635
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6766885487840181,
+      "learning_rate": 3.14206831255667e-05,
+      "loss": 1.9246,
+      "step": 5636
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8552016492847282,
+      "learning_rate": 3.141464575193884e-05,
+      "loss": 1.9995,
+      "step": 5637
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7930754126393524,
+      "learning_rate": 3.1408607977818306e-05,
+      "loss": 2.126,
+      "step": 5638
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6392465697169919,
+      "learning_rate": 3.140256980358204e-05,
+      "loss": 1.9739,
+      "step": 5639
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6505359465621804,
+      "learning_rate": 3.139653122960704e-05,
+      "loss": 1.9368,
+      "step": 5640
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9011323354788731,
+      "learning_rate": 3.139049225627031e-05,
+      "loss": 2.1172,
+      "step": 5641
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6137511859891005,
+      "learning_rate": 3.1384452883948906e-05,
+      "loss": 1.9576,
+      "step": 5642
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8351084893802556,
+      "learning_rate": 3.137841311301988e-05,
+      "loss": 2.0007,
+      "step": 5643
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7214810994985507,
+      "learning_rate": 3.137237294386032e-05,
+      "loss": 1.9329,
+      "step": 5644
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6237961656532114,
+      "learning_rate": 3.136633237684733e-05,
+      "loss": 1.9463,
+      "step": 5645
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7963826803090053,
+      "learning_rate": 3.1360291412358055e-05,
+      "loss": 2.1485,
+      "step": 5646
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7336081837304146,
+      "learning_rate": 3.135425005076967e-05,
+      "loss": 1.9525,
+      "step": 5647
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6102364683341208,
+      "learning_rate": 3.134820829245935e-05,
+      "loss": 1.9593,
+      "step": 5648
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8611686594721094,
+      "learning_rate": 3.1342166137804305e-05,
+      "loss": 2.0992,
+      "step": 5649
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7283125798332221,
+      "learning_rate": 3.133612358718177e-05,
+      "loss": 2.0374,
+      "step": 5650
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7680529055810161,
+      "learning_rate": 3.1330080640969013e-05,
+      "loss": 1.9495,
+      "step": 5651
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8126057726636566,
+      "learning_rate": 3.132403729954331e-05,
+      "loss": 1.9785,
+      "step": 5652
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.907884186185073,
+      "learning_rate": 3.131799356328199e-05,
+      "loss": 2.1239,
+      "step": 5653
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8293905053910474,
+      "learning_rate": 3.1311949432562374e-05,
+      "loss": 1.9556,
+      "step": 5654
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9725845588753346,
+      "learning_rate": 3.130590490776182e-05,
+      "loss": 1.9439,
+      "step": 5655
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5958940545997812,
+      "learning_rate": 3.129985998925772e-05,
+      "loss": 1.9927,
+      "step": 5656
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9049355484599579,
+      "learning_rate": 3.1293814677427475e-05,
+      "loss": 1.9656,
+      "step": 5657
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7549136803220631,
+      "learning_rate": 3.128776897264853e-05,
+      "loss": 2.1347,
+      "step": 5658
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6720022788834428,
+      "learning_rate": 3.128172287529834e-05,
+      "loss": 1.9347,
+      "step": 5659
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6583619839814041,
+      "learning_rate": 3.127567638575438e-05,
+      "loss": 1.9452,
+      "step": 5660
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8349044684190652,
+      "learning_rate": 3.126962950439416e-05,
+      "loss": 2.1666,
+      "step": 5661
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6399870177308284,
+      "learning_rate": 3.126358223159522e-05,
+      "loss": 2.0164,
+      "step": 5662
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6637986645778988,
+      "learning_rate": 3.1257534567735105e-05,
+      "loss": 1.9026,
+      "step": 5663
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7877965929691404,
+      "learning_rate": 3.12514865131914e-05,
+      "loss": 1.9328,
+      "step": 5664
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6548425843584832,
+      "learning_rate": 3.1245438068341716e-05,
+      "loss": 2.118,
+      "step": 5665
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8428785557154008,
+      "learning_rate": 3.123938923356367e-05,
+      "loss": 1.909,
+      "step": 5666
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7076931364083444,
+      "learning_rate": 3.1233340009234916e-05,
+      "loss": 1.948,
+      "step": 5667
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9209240343831496,
+      "learning_rate": 3.122729039573315e-05,
+      "loss": 2.0336,
+      "step": 5668
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6994883790430878,
+      "learning_rate": 3.1221240393436065e-05,
+      "loss": 1.9338,
+      "step": 5669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7258248339480239,
+      "learning_rate": 3.121519000272138e-05,
+      "loss": 2.0986,
+      "step": 5670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7379881905749857,
+      "learning_rate": 3.120913922396684e-05,
+      "loss": 1.9479,
+      "step": 5671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7116379554679477,
+      "learning_rate": 3.1203088057550244e-05,
+      "loss": 1.9304,
+      "step": 5672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6419883003008122,
+      "learning_rate": 3.1197036503849375e-05,
+      "loss": 2.1337,
+      "step": 5673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5931873374662571,
+      "learning_rate": 3.119098456324206e-05,
+      "loss": 1.9675,
+      "step": 5674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6365796046810436,
+      "learning_rate": 3.118493223610613e-05,
+      "loss": 1.9293,
+      "step": 5675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6763007361020857,
+      "learning_rate": 3.1178879522819496e-05,
+      "loss": 1.8999,
+      "step": 5676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5798559912838727,
+      "learning_rate": 3.117282642376002e-05,
+      "loss": 1.9353,
+      "step": 5677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6138782324790876,
+      "learning_rate": 3.1166772939305624e-05,
+      "loss": 2.1953,
+      "step": 5678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6053371296245269,
+      "learning_rate": 3.116071906983427e-05,
+      "loss": 1.9841,
+      "step": 5679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5811291216702393,
+      "learning_rate": 3.115466481572391e-05,
+      "loss": 1.9546,
+      "step": 5680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5583008752776615,
+      "learning_rate": 3.114861017735255e-05,
+      "loss": 2.0254,
+      "step": 5681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6959782999116002,
+      "learning_rate": 3.114255515509818e-05,
+      "loss": 2.1192,
+      "step": 5682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6288505739894407,
+      "learning_rate": 3.1136499749338884e-05,
+      "loss": 1.9543,
+      "step": 5683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6022301968296992,
+      "learning_rate": 3.113044396045268e-05,
+      "loss": 1.9214,
+      "step": 5684
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6258144349345304,
+      "learning_rate": 3.1124387788817675e-05,
+      "loss": 2.0662,
+      "step": 5685
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6814655204425702,
+      "learning_rate": 3.111833123481198e-05,
+      "loss": 1.9214,
+      "step": 5686
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5806089413387597,
+      "learning_rate": 3.111227429881374e-05,
+      "loss": 2.0211,
+      "step": 5687
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6995239871323705,
+      "learning_rate": 3.11062169812011e-05,
+      "loss": 1.9629,
+      "step": 5688
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6256702290667359,
+      "learning_rate": 3.1100159282352245e-05,
+      "loss": 1.8834,
+      "step": 5689
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7182005483318241,
+      "learning_rate": 3.1094101202645396e-05,
+      "loss": 2.0998,
+      "step": 5690
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5897869539217989,
+      "learning_rate": 3.1088042742458774e-05,
+      "loss": 1.9761,
+      "step": 5691
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.712672048422269,
+      "learning_rate": 3.108198390217062e-05,
+      "loss": 1.9743,
+      "step": 5692
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5958747449883048,
+      "learning_rate": 3.1075924682159235e-05,
+      "loss": 1.9697,
+      "step": 5693
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7223463824161566,
+      "learning_rate": 3.1069865082802915e-05,
+      "loss": 2.1399,
+      "step": 5694
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6468503222814457,
+      "learning_rate": 3.106380510447998e-05,
+      "loss": 1.9359,
+      "step": 5695
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6973626502674143,
+      "learning_rate": 3.105774474756878e-05,
+      "loss": 1.8797,
+      "step": 5696
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5766560756694544,
+      "learning_rate": 3.105168401244769e-05,
+      "loss": 2.1195,
+      "step": 5697
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6723336036288439,
+      "learning_rate": 3.1045622899495105e-05,
+      "loss": 1.9232,
+      "step": 5698
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5929080382003645,
+      "learning_rate": 3.103956140908944e-05,
+      "loss": 1.9183,
+      "step": 5699
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6355364240260755,
+      "learning_rate": 3.1033499541609154e-05,
+      "loss": 1.9449,
+      "step": 5700
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6227281413567548,
+      "learning_rate": 3.102743729743271e-05,
+      "loss": 1.9086,
+      "step": 5701
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6266973476625657,
+      "learning_rate": 3.1021374676938584e-05,
+      "loss": 2.1226,
+      "step": 5702
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6539028996800659,
+      "learning_rate": 3.10153116805053e-05,
+      "loss": 1.9272,
+      "step": 5703
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6681228223165171,
+      "learning_rate": 3.1009248308511394e-05,
+      "loss": 1.9208,
+      "step": 5704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6159281972341079,
+      "learning_rate": 3.100318456133544e-05,
+      "loss": 2.1283,
+      "step": 5705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7052663774126838,
+      "learning_rate": 3.0997120439356006e-05,
+      "loss": 1.9883,
+      "step": 5706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6501867927821594,
+      "learning_rate": 3.099105594295171e-05,
+      "loss": 1.9268,
+      "step": 5707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6176458681958992,
+      "learning_rate": 3.098499107250118e-05,
+      "loss": 1.9281,
+      "step": 5708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6047726259960889,
+      "learning_rate": 3.097892582838307e-05,
+      "loss": 1.8908,
+      "step": 5709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.671476239877798,
+      "learning_rate": 3.0972860210976046e-05,
+      "loss": 2.0854,
+      "step": 5710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6699469831719684,
+      "learning_rate": 3.096679422065883e-05,
+      "loss": 1.8663,
+      "step": 5711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6341770898239449,
+      "learning_rate": 3.096072785781015e-05,
+      "loss": 1.9885,
+      "step": 5712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6226054819923315,
+      "learning_rate": 3.095466112280873e-05,
+      "loss": 1.9509,
+      "step": 5713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6880827740637214,
+      "learning_rate": 3.094859401603336e-05,
+      "loss": 2.0943,
+      "step": 5714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.662426860200876,
+      "learning_rate": 3.0942526537862827e-05,
+      "loss": 1.9288,
+      "step": 5715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6754217997764672,
+      "learning_rate": 3.0936458688675946e-05,
+      "loss": 1.9322,
+      "step": 5716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8095193802523499,
+      "learning_rate": 3.093039046885156e-05,
+      "loss": 2.1083,
+      "step": 5717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5588495657601058,
+      "learning_rate": 3.092432187876853e-05,
+      "loss": 1.9885,
+      "step": 5718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6877874677781738,
+      "learning_rate": 3.091825291880576e-05,
+      "loss": 1.9208,
+      "step": 5719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7145755284049263,
+      "learning_rate": 3.0912183589342145e-05,
+      "loss": 1.9058,
+      "step": 5720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6521407847622906,
+      "learning_rate": 3.090611389075661e-05,
+      "loss": 1.9402,
+      "step": 5721
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6378500896549859,
+      "learning_rate": 3.0900043823428135e-05,
+      "loss": 2.112,
+      "step": 5722
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6985318167672901,
+      "learning_rate": 3.0893973387735687e-05,
+      "loss": 1.9081,
+      "step": 5723
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7396519116591681,
+      "learning_rate": 3.088790258405826e-05,
+      "loss": 1.9504,
+      "step": 5724
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6966317296586871,
+      "learning_rate": 3.088183141277489e-05,
+      "loss": 1.9531,
+      "step": 5725
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8077000601469055,
+      "learning_rate": 3.087575987426462e-05,
+      "loss": 2.0959,
+      "step": 5726
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6151521071501472,
+      "learning_rate": 3.086968796890652e-05,
+      "loss": 1.9102,
+      "step": 5727
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6104241636886986,
+      "learning_rate": 3.0863615697079684e-05,
+      "loss": 1.9434,
+      "step": 5728
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.730877139267078,
+      "learning_rate": 3.0857543059163244e-05,
+      "loss": 1.8747,
+      "step": 5729
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6452296799580808,
+      "learning_rate": 3.085147005553632e-05,
+      "loss": 2.1487,
+      "step": 5730
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6712202380413377,
+      "learning_rate": 3.084539668657808e-05,
+      "loss": 1.9207,
+      "step": 5731
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8546406976221929,
+      "learning_rate": 3.083932295266771e-05,
+      "loss": 1.9288,
+      "step": 5732
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.842712235570716,
+      "learning_rate": 3.0833248854184426e-05,
+      "loss": 1.9128,
+      "step": 5733
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6239950412938232,
+      "learning_rate": 3.0827174391507455e-05,
+      "loss": 2.1123,
+      "step": 5734
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.0118108361250793,
+      "learning_rate": 3.082109956501604e-05,
+      "loss": 1.9422,
+      "step": 5735
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6014669940825281,
+      "learning_rate": 3.081502437508946e-05,
+      "loss": 1.9319,
+      "step": 5736
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7780192406599721,
+      "learning_rate": 3.080894882210704e-05,
+      "loss": 2.0019,
+      "step": 5737
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8025051982257406,
+      "learning_rate": 3.0802872906448066e-05,
+      "loss": 2.0809,
+      "step": 5738
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6156759401258579,
+      "learning_rate": 3.07967966284919e-05,
+      "loss": 1.939,
+      "step": 5739
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6413774824302707,
+      "learning_rate": 3.079071998861791e-05,
+      "loss": 1.926,
+      "step": 5740
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7518592741742948,
+      "learning_rate": 3.078464298720548e-05,
+      "loss": 1.9426,
+      "step": 5741
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7987791983415258,
+      "learning_rate": 3.077856562463402e-05,
+      "loss": 2.1456,
+      "step": 5742
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.737582029992249,
+      "learning_rate": 3.0772487901282977e-05,
+      "loss": 1.9901,
+      "step": 5743
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7857942931527094,
+      "learning_rate": 3.0766409817531795e-05,
+      "loss": 1.9798,
+      "step": 5744
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.70463175540639,
+      "learning_rate": 3.076033137375996e-05,
+      "loss": 1.9092,
+      "step": 5745
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8344312086943486,
+      "learning_rate": 3.075425257034697e-05,
+      "loss": 2.0679,
+      "step": 5746
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.760871066110913,
+      "learning_rate": 3.074817340767236e-05,
+      "loss": 1.9413,
+      "step": 5747
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6373982971954872,
+      "learning_rate": 3.074209388611567e-05,
+      "loss": 1.9049,
+      "step": 5748
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7423452997813833,
+      "learning_rate": 3.0736014006056465e-05,
+      "loss": 2.0043,
+      "step": 5749
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6956013611059232,
+      "learning_rate": 3.072993376787434e-05,
+      "loss": 2.1138,
+      "step": 5750
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.608469913008348,
+      "learning_rate": 3.072385317194891e-05,
+      "loss": 1.8915,
+      "step": 5751
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.756110784860381,
+      "learning_rate": 3.071777221865982e-05,
+      "loss": 1.8639,
+      "step": 5752
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6833615027991287,
+      "learning_rate": 3.071169090838672e-05,
+      "loss": 1.9444,
+      "step": 5753
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9189515067290148,
+      "learning_rate": 3.0705609241509284e-05,
+      "loss": 2.0931,
+      "step": 5754
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7567967621122482,
+      "learning_rate": 3.0699527218407236e-05,
+      "loss": 2.023,
+      "step": 5755
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7913196701736857,
+      "learning_rate": 3.0693444839460276e-05,
+      "loss": 1.9515,
+      "step": 5756
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7879972756291038,
+      "learning_rate": 3.0687362105048175e-05,
+      "loss": 1.9332,
+      "step": 5757
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8393335126977188,
+      "learning_rate": 3.068127901555069e-05,
+      "loss": 2.1426,
+      "step": 5758
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.827387334529188,
+      "learning_rate": 3.067519557134762e-05,
+      "loss": 1.9112,
+      "step": 5759
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6661379443057521,
+      "learning_rate": 3.0669111772818784e-05,
+      "loss": 1.9109,
+      "step": 5760
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6901456791948904,
+      "learning_rate": 3.066302762034399e-05,
+      "loss": 2.0122,
+      "step": 5761
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7603237042872206,
+      "learning_rate": 3.065694311430313e-05,
+      "loss": 2.1706,
+      "step": 5762
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7077419352404956,
+      "learning_rate": 3.065085825507607e-05,
+      "loss": 1.8644,
+      "step": 5763
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6731703485827933,
+      "learning_rate": 3.064477304304272e-05,
+      "loss": 1.9301,
+      "step": 5764
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7114193835034094,
+      "learning_rate": 3.0638687478582986e-05,
+      "loss": 1.9856,
+      "step": 5765
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6114941728036395,
+      "learning_rate": 3.063260156207684e-05,
+      "loss": 2.1226,
+      "step": 5766
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6161022771549636,
+      "learning_rate": 3.062651529390424e-05,
+      "loss": 1.9145,
+      "step": 5767
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6684632530489678,
+      "learning_rate": 3.0620428674445165e-05,
+      "loss": 2.0433,
+      "step": 5768
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6118152255175296,
+      "learning_rate": 3.061434170407965e-05,
+      "loss": 1.9451,
+      "step": 5769
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6348987746710177,
+      "learning_rate": 3.0608254383187714e-05,
+      "loss": 2.0961,
+      "step": 5770
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.594512903309019,
+      "learning_rate": 3.060216671214942e-05,
+      "loss": 1.9386,
+      "step": 5771
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7134842303316508,
+      "learning_rate": 3.059607869134484e-05,
+      "loss": 1.918,
+      "step": 5772
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6122237054063439,
+      "learning_rate": 3.0589990321154076e-05,
+      "loss": 1.9393,
+      "step": 5773
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6732249884679624,
+      "learning_rate": 3.058390160195725e-05,
+      "loss": 2.1934,
+      "step": 5774
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5879932233373174,
+      "learning_rate": 3.057781253413452e-05,
+      "loss": 1.9816,
+      "step": 5775
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6194843408405601,
+      "learning_rate": 3.057172311806603e-05,
+      "loss": 1.9238,
+      "step": 5776
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6374707411736114,
+      "learning_rate": 3.056563335413198e-05,
+      "loss": 1.8724,
+      "step": 5777
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6014176864905327,
+      "learning_rate": 3.055954324271258e-05,
+      "loss": 2.1753,
+      "step": 5778
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6591498976409459,
+      "learning_rate": 3.055345278418804e-05,
+      "loss": 1.9615,
+      "step": 5779
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6067070575975457,
+      "learning_rate": 3.054736197893865e-05,
+      "loss": 2.0082,
+      "step": 5780
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7425535479890805,
+      "learning_rate": 3.0541270827344645e-05,
+      "loss": 1.902,
+      "step": 5781
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6934969637879375,
+      "learning_rate": 3.0535179329786336e-05,
+      "loss": 2.1057,
+      "step": 5782
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6481807528759657,
+      "learning_rate": 3.052908748664406e-05,
+      "loss": 1.9545,
+      "step": 5783
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7054520982920129,
+      "learning_rate": 3.052299529829813e-05,
+      "loss": 1.9623,
+      "step": 5784
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7810095589884813,
+      "learning_rate": 3.051690276512891e-05,
+      "loss": 1.9144,
+      "step": 5785
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6483734637919202,
+      "learning_rate": 3.051080988751679e-05,
+      "loss": 2.1813,
+      "step": 5786
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6941977951243173,
+      "learning_rate": 3.0504716665842174e-05,
+      "loss": 1.8944,
+      "step": 5787
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7810229032536363,
+      "learning_rate": 3.049862310048548e-05,
+      "loss": 1.9613,
+      "step": 5788
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5888749814878694,
+      "learning_rate": 3.049252919182715e-05,
+      "loss": 1.8577,
+      "step": 5789
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.9656947102507605,
+      "learning_rate": 3.048643494024766e-05,
+      "loss": 2.1447,
+      "step": 5790
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6301585562103391,
+      "learning_rate": 3.0480340346127505e-05,
+      "loss": 1.9057,
+      "step": 5791
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7956605755982032,
+      "learning_rate": 3.0474245409847185e-05,
+      "loss": 1.9545,
+      "step": 5792
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7283144757418191,
+      "learning_rate": 3.0468150131787238e-05,
+      "loss": 1.8993,
+      "step": 5793
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7746248303744437,
+      "learning_rate": 3.046205451232821e-05,
+      "loss": 2.0672,
+      "step": 5794
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6175403243772434,
+      "learning_rate": 3.0455958551850682e-05,
+      "loss": 1.9436,
+      "step": 5795
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5948147446023655,
+      "learning_rate": 3.044986225073525e-05,
+      "loss": 1.9149,
+      "step": 5796
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6912182738719852,
+      "learning_rate": 3.0443765609362528e-05,
+      "loss": 1.9246,
+      "step": 5797
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5967504650693743,
+      "learning_rate": 3.0437668628113153e-05,
+      "loss": 2.1317,
+      "step": 5798
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5895091483558293,
+      "learning_rate": 3.043157130736779e-05,
+      "loss": 1.976,
+      "step": 5799
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5451492981864545,
+      "learning_rate": 3.0425473647507115e-05,
+      "loss": 1.8784,
+      "step": 5800
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5733678371711695,
+      "learning_rate": 3.041937564891183e-05,
+      "loss": 1.9525,
+      "step": 5801
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6781993896953907,
+      "learning_rate": 3.0413277311962668e-05,
+      "loss": 2.0332,
+      "step": 5802
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6823059792144617,
+      "learning_rate": 3.040717863704035e-05,
+      "loss": 1.9345,
+      "step": 5803
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6834360640782295,
+      "learning_rate": 3.0401079624525668e-05,
+      "loss": 1.9425,
+      "step": 5804
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6150847125185505,
+      "learning_rate": 3.03949802747994e-05,
+      "loss": 1.9875,
+      "step": 5805
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6982209554060324,
+      "learning_rate": 3.0388880588242347e-05,
+      "loss": 2.1652,
+      "step": 5806
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6245965626235614,
+      "learning_rate": 3.038278056523533e-05,
+      "loss": 1.9285,
+      "step": 5807
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6570708627418822,
+      "learning_rate": 3.037668020615922e-05,
+      "loss": 1.917,
+      "step": 5808
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5709824748440416,
+      "learning_rate": 3.037057951139488e-05,
+      "loss": 1.8992,
+      "step": 5809
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5846108487519303,
+      "learning_rate": 3.036447848132319e-05,
+      "loss": 2.1099,
+      "step": 5810
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7109491030643428,
+      "learning_rate": 3.0358377116325075e-05,
+      "loss": 2.0347,
+      "step": 5811
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6270243396177597,
+      "learning_rate": 3.0352275416781463e-05,
+      "loss": 1.907,
+      "step": 5812
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6364134201729412,
+      "learning_rate": 3.0346173383073307e-05,
+      "loss": 1.9384,
+      "step": 5813
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6674900450779712,
+      "learning_rate": 3.0340071015581583e-05,
+      "loss": 2.1365,
+      "step": 5814
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5837690441920116,
+      "learning_rate": 3.0333968314687296e-05,
+      "loss": 1.9339,
+      "step": 5815
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6997362233780504,
+      "learning_rate": 3.0327865280771445e-05,
+      "loss": 1.9428,
+      "step": 5816
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6275056328771808,
+      "learning_rate": 3.032176191421508e-05,
+      "loss": 1.9665,
+      "step": 5817
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6967948643742953,
+      "learning_rate": 3.0315658215399256e-05,
+      "loss": 2.1072,
+      "step": 5818
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6505052680559663,
+      "learning_rate": 3.0309554184705058e-05,
+      "loss": 1.9309,
+      "step": 5819
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6099629371334845,
+      "learning_rate": 3.0303449822513575e-05,
+      "loss": 1.9285,
+      "step": 5820
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6247597694038475,
+      "learning_rate": 3.0297345129205933e-05,
+      "loss": 1.9295,
+      "step": 5821
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6442285496729565,
+      "learning_rate": 3.0291240105163272e-05,
+      "loss": 2.1128,
+      "step": 5822
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6061204908232027,
+      "learning_rate": 3.0285134750766754e-05,
+      "loss": 2.0129,
+      "step": 5823
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6482506134397642,
+      "learning_rate": 3.0279029066397563e-05,
+      "loss": 1.9383,
+      "step": 5824
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.690503483661787,
+      "learning_rate": 3.02729230524369e-05,
+      "loss": 1.9615,
+      "step": 5825
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.627034202065041,
+      "learning_rate": 3.026681670926599e-05,
+      "loss": 2.0785,
+      "step": 5826
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7078633677456196,
+      "learning_rate": 3.0260710037266082e-05,
+      "loss": 1.9748,
+      "step": 5827
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6251306460929781,
+      "learning_rate": 3.0254603036818423e-05,
+      "loss": 1.8976,
+      "step": 5828
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.644219769805675,
+      "learning_rate": 3.0248495708304315e-05,
+      "loss": 2.007,
+      "step": 5829
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6517518456469886,
+      "learning_rate": 3.024238805210506e-05,
+      "loss": 2.1587,
+      "step": 5830
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5818419248020729,
+      "learning_rate": 3.0236280068601987e-05,
+      "loss": 1.9044,
+      "step": 5831
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7287355598340334,
+      "learning_rate": 3.0230171758176428e-05,
+      "loss": 1.9495,
+      "step": 5832
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6768832661481076,
+      "learning_rate": 3.022406312120976e-05,
+      "loss": 1.9114,
+      "step": 5833
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6207307031416446,
+      "learning_rate": 3.0217954158083384e-05,
+      "loss": 2.1368,
+      "step": 5834
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7258980759954399,
+      "learning_rate": 3.0211844869178686e-05,
+      "loss": 1.9388,
+      "step": 5835
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6388601802372186,
+      "learning_rate": 3.02057352548771e-05,
+      "loss": 1.9966,
+      "step": 5836
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6227692425556223,
+      "learning_rate": 3.0199625315560078e-05,
+      "loss": 1.8945,
+      "step": 5837
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5706924334742495,
+      "learning_rate": 3.0193515051609083e-05,
+      "loss": 2.1304,
+      "step": 5838
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6382479775592301,
+      "learning_rate": 3.018740446340561e-05,
+      "loss": 1.9298,
+      "step": 5839
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7066930751598656,
+      "learning_rate": 3.018129355133117e-05,
+      "loss": 1.9715,
+      "step": 5840
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5358642174619391,
+      "learning_rate": 3.0175182315767285e-05,
+      "loss": 1.9365,
+      "step": 5841
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6866363591956317,
+      "learning_rate": 3.0169070757095507e-05,
+      "loss": 2.0069,
+      "step": 5842
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.600059190155287,
+      "learning_rate": 3.016295887569741e-05,
+      "loss": 2.0635,
+      "step": 5843
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.568068985388958,
+      "learning_rate": 3.0156846671954576e-05,
+      "loss": 1.9238,
+      "step": 5844
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7167827825017467,
+      "learning_rate": 3.0150734146248626e-05,
+      "loss": 1.9442,
+      "step": 5845
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5374978304369556,
+      "learning_rate": 3.0144621298961173e-05,
+      "loss": 2.1267,
+      "step": 5846
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6757165958268606,
+      "learning_rate": 3.0138508130473887e-05,
+      "loss": 1.9142,
+      "step": 5847
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6948448174678183,
+      "learning_rate": 3.0132394641168427e-05,
+      "loss": 2.007,
+      "step": 5848
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6077437555357943,
+      "learning_rate": 3.012628083142649e-05,
+      "loss": 1.9272,
+      "step": 5849
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7672750313597165,
+      "learning_rate": 3.012016670162977e-05,
+      "loss": 2.1119,
+      "step": 5850
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5958027950611892,
+      "learning_rate": 3.0114052252160013e-05,
+      "loss": 1.8916,
+      "step": 5851
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6362691561284282,
+      "learning_rate": 3.0107937483398967e-05,
+      "loss": 1.9124,
+      "step": 5852
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6498122597783662,
+      "learning_rate": 3.0101822395728403e-05,
+      "loss": 1.921,
+      "step": 5853
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.604066288166943,
+      "learning_rate": 3.0095706989530096e-05,
+      "loss": 2.1652,
+      "step": 5854
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6256895403644306,
+      "learning_rate": 3.008959126518588e-05,
+      "loss": 1.9921,
+      "step": 5855
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6248806026458444,
+      "learning_rate": 3.0083475223077574e-05,
+      "loss": 1.893,
+      "step": 5856
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6109976970539495,
+      "learning_rate": 3.007735886358702e-05,
+      "loss": 1.9482,
+      "step": 5857
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7434508351032425,
+      "learning_rate": 3.0071242187096093e-05,
+      "loss": 2.0882,
+      "step": 5858
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5989737490732091,
+      "learning_rate": 3.0065125193986682e-05,
+      "loss": 1.9545,
+      "step": 5859
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7457862906738794,
+      "learning_rate": 3.00590078846407e-05,
+      "loss": 1.9459,
+      "step": 5860
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6021411980242811,
+      "learning_rate": 3.0052890259440074e-05,
+      "loss": 2.0195,
+      "step": 5861
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7722521010901666,
+      "learning_rate": 3.0046772318766753e-05,
+      "loss": 1.9465,
+      "step": 5862
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6555835462201973,
+      "learning_rate": 3.00406540630027e-05,
+      "loss": 2.1223,
+      "step": 5863
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6467735934076191,
+      "learning_rate": 3.0034535492529904e-05,
+      "loss": 1.9069,
+      "step": 5864
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6481897514592124,
+      "learning_rate": 3.002841660773038e-05,
+      "loss": 1.8667,
+      "step": 5865
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6425457021568652,
+      "learning_rate": 3.0022297408986146e-05,
+      "loss": 2.1423,
+      "step": 5866
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6027530357585592,
+      "learning_rate": 3.0016177896679255e-05,
+      "loss": 2.0292,
+      "step": 5867
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6240375855683122,
+      "learning_rate": 3.0010058071191772e-05,
+      "loss": 1.9629,
+      "step": 5868
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5726931814811277,
+      "learning_rate": 3.0003937932905778e-05,
+      "loss": 1.8965,
+      "step": 5869
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6322633159460269,
+      "learning_rate": 2.9997817482203388e-05,
+      "loss": 2.1334,
+      "step": 5870
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5850224271326997,
+      "learning_rate": 2.999169671946671e-05,
+      "loss": 1.9914,
+      "step": 5871
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5361120137306,
+      "learning_rate": 2.9985575645077912e-05,
+      "loss": 1.8992,
+      "step": 5872
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5719195961820016,
+      "learning_rate": 2.9979454259419138e-05,
+      "loss": 2.0251,
+      "step": 5873
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5979140306424473,
+      "learning_rate": 2.9973332562872585e-05,
+      "loss": 1.8826,
+      "step": 5874
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6108812639731959,
+      "learning_rate": 2.9967210555820446e-05,
+      "loss": 2.0957,
+      "step": 5875
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5824445855687179,
+      "learning_rate": 2.9961088238644937e-05,
+      "loss": 1.9262,
+      "step": 5876
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5955769642508199,
+      "learning_rate": 2.9954965611728324e-05,
+      "loss": 1.9625,
+      "step": 5877
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6965183239099628,
+      "learning_rate": 2.9948842675452843e-05,
+      "loss": 2.154,
+      "step": 5878
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5728572728103319,
+      "learning_rate": 2.9942719430200782e-05,
+      "loss": 2.0247,
+      "step": 5879
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5873205140366959,
+      "learning_rate": 2.993659587635444e-05,
+      "loss": 1.8742,
+      "step": 5880
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5757376840800539,
+      "learning_rate": 2.9930472014296152e-05,
+      "loss": 1.915,
+      "step": 5881
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7817920063761639,
+      "learning_rate": 2.992434784440823e-05,
+      "loss": 2.1434,
+      "step": 5882
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6871254658795236,
+      "learning_rate": 2.9918223367073038e-05,
+      "loss": 1.9094,
+      "step": 5883
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6902483786040653,
+      "learning_rate": 2.9912098582672964e-05,
+      "loss": 1.941,
+      "step": 5884
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6770423390800582,
+      "learning_rate": 2.99059734915904e-05,
+      "loss": 1.9193,
+      "step": 5885
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.654342515134921,
+      "learning_rate": 2.989984809420775e-05,
+      "loss": 2.0164,
+      "step": 5886
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6264815406638509,
+      "learning_rate": 2.989372239090745e-05,
+      "loss": 2.114,
+      "step": 5887
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7499529035550805,
+      "learning_rate": 2.9887596382071965e-05,
+      "loss": 1.9465,
+      "step": 5888
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6541832866045406,
+      "learning_rate": 2.9881470068083763e-05,
+      "loss": 1.907,
+      "step": 5889
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6916484585900159,
+      "learning_rate": 2.9875343449325323e-05,
+      "loss": 2.1104,
+      "step": 5890
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6748633955122447,
+      "learning_rate": 2.986921652617916e-05,
+      "loss": 1.9339,
+      "step": 5891
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6295381116869506,
+      "learning_rate": 2.986308929902782e-05,
+      "loss": 2.0131,
+      "step": 5892
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6349348436545615,
+      "learning_rate": 2.985696176825382e-05,
+      "loss": 1.9599,
+      "step": 5893
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6700301491722512,
+      "learning_rate": 2.9850833934239752e-05,
+      "loss": 1.9179,
+      "step": 5894
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5780490373822696,
+      "learning_rate": 2.9844705797368193e-05,
+      "loss": 2.0867,
+      "step": 5895
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5788800115791749,
+      "learning_rate": 2.9838577358021758e-05,
+      "loss": 1.9781,
+      "step": 5896
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5914250052963899,
+      "learning_rate": 2.9832448616583042e-05,
+      "loss": 1.9192,
+      "step": 5897
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5997608681927281,
+      "learning_rate": 2.982631957343472e-05,
+      "loss": 1.9879,
+      "step": 5898
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5849522910100994,
+      "learning_rate": 2.9820190228959445e-05,
+      "loss": 2.1126,
+      "step": 5899
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6182279116715895,
+      "learning_rate": 2.9814060583539883e-05,
+      "loss": 1.9309,
+      "step": 5900
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.567416295909268,
+      "learning_rate": 2.9807930637558733e-05,
+      "loss": 1.9174,
+      "step": 5901
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6423126065949616,
+      "learning_rate": 2.980180039139874e-05,
+      "loss": 2.1392,
+      "step": 5902
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6030654245451491,
+      "learning_rate": 2.9795669845442614e-05,
+      "loss": 1.9263,
+      "step": 5903
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.547532943309135,
+      "learning_rate": 2.978953900007312e-05,
+      "loss": 1.9269,
+      "step": 5904
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5745689368935785,
+      "learning_rate": 2.9783407855673025e-05,
+      "loss": 1.9176,
+      "step": 5905
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6950814474030836,
+      "learning_rate": 2.9777276412625132e-05,
+      "loss": 1.9407,
+      "step": 5906
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6198642487117558,
+      "learning_rate": 2.977114467131225e-05,
+      "loss": 2.1529,
+      "step": 5907
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5575607576923353,
+      "learning_rate": 2.9765012632117202e-05,
+      "loss": 1.9215,
+      "step": 5908
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6393011271732744,
+      "learning_rate": 2.9758880295422835e-05,
+      "loss": 1.9153,
+      "step": 5909
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5837295561767646,
+      "learning_rate": 2.975274766161203e-05,
+      "loss": 1.9391,
+      "step": 5910
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7341705152820133,
+      "learning_rate": 2.9746614731067658e-05,
+      "loss": 2.1301,
+      "step": 5911
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5781172239487322,
+      "learning_rate": 2.9740481504172634e-05,
+      "loss": 1.9259,
+      "step": 5912
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6829223166055113,
+      "learning_rate": 2.9734347981309868e-05,
+      "loss": 1.934,
+      "step": 5913
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6853101534161736,
+      "learning_rate": 2.9728214162862317e-05,
+      "loss": 2.1407,
+      "step": 5914
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6815798800857179,
+      "learning_rate": 2.9722080049212926e-05,
+      "loss": 1.9295,
+      "step": 5915
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6201937169440519,
+      "learning_rate": 2.971594564074468e-05,
+      "loss": 1.9333,
+      "step": 5916
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6875413870756768,
+      "learning_rate": 2.9709810937840577e-05,
+      "loss": 2.0487,
+      "step": 5917
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6372626365975891,
+      "learning_rate": 2.9703675940883625e-05,
+      "loss": 1.9225,
+      "step": 5918
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6004064747032164,
+      "learning_rate": 2.9697540650256854e-05,
+      "loss": 2.0802,
+      "step": 5919
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5515527150902807,
+      "learning_rate": 2.9691405066343336e-05,
+      "loss": 1.8321,
+      "step": 5920
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6598469618246547,
+      "learning_rate": 2.9685269189526123e-05,
+      "loss": 1.9312,
+      "step": 5921
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6479674919511998,
+      "learning_rate": 2.9679133020188297e-05,
+      "loss": 2.1647,
+      "step": 5922
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5835508491915485,
+      "learning_rate": 2.9672996558712986e-05,
+      "loss": 1.9966,
+      "step": 5923
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6440159767565734,
+      "learning_rate": 2.9666859805483305e-05,
+      "loss": 1.9207,
+      "step": 5924
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6500028754809394,
+      "learning_rate": 2.9660722760882387e-05,
+      "loss": 1.9209,
+      "step": 5925
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5388042773105715,
+      "learning_rate": 2.96545854252934e-05,
+      "loss": 1.9555,
+      "step": 5926
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5819980533110257,
+      "learning_rate": 2.964844779909952e-05,
+      "loss": 2.1586,
+      "step": 5927
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5728317851910332,
+      "learning_rate": 2.9642309882683955e-05,
+      "loss": 1.9181,
+      "step": 5928
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5808699988925805,
+      "learning_rate": 2.9636171676429914e-05,
+      "loss": 2.0103,
+      "step": 5929
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5857061366846728,
+      "learning_rate": 2.9630033180720625e-05,
+      "loss": 1.922,
+      "step": 5930
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5757872076384501,
+      "learning_rate": 2.9623894395939344e-05,
+      "loss": 2.1355,
+      "step": 5931
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6384088026438381,
+      "learning_rate": 2.9617755322469336e-05,
+      "loss": 1.8799,
+      "step": 5932
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5739669432667489,
+      "learning_rate": 2.96116159606939e-05,
+      "loss": 1.9056,
+      "step": 5933
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7187417204353276,
+      "learning_rate": 2.9605476310996333e-05,
+      "loss": 2.1319,
+      "step": 5934
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5733529554445508,
+      "learning_rate": 2.9599336373759957e-05,
+      "loss": 1.9944,
+      "step": 5935
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6047752585470587,
+      "learning_rate": 2.959319614936812e-05,
+      "loss": 1.9309,
+      "step": 5936
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.62405650142704,
+      "learning_rate": 2.958705563820418e-05,
+      "loss": 1.8992,
+      "step": 5937
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6292321132456158,
+      "learning_rate": 2.9580914840651508e-05,
+      "loss": 1.9078,
+      "step": 5938
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6208311396776585,
+      "learning_rate": 2.9574773757093505e-05,
+      "loss": 2.06,
+      "step": 5939
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5850421044653606,
+      "learning_rate": 2.9568632387913586e-05,
+      "loss": 1.9316,
+      "step": 5940
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6455775185099398,
+      "learning_rate": 2.9562490733495173e-05,
+      "loss": 2.0399,
+      "step": 5941
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5914731793845333,
+      "learning_rate": 2.955634879422173e-05,
+      "loss": 1.9121,
+      "step": 5942
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6141170356408592,
+      "learning_rate": 2.95502065704767e-05,
+      "loss": 2.1214,
+      "step": 5943
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6032996933972882,
+      "learning_rate": 2.954406406264359e-05,
+      "loss": 1.9002,
+      "step": 5944
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6409087770806413,
+      "learning_rate": 2.953792127110589e-05,
+      "loss": 1.9393,
+      "step": 5945
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6407734269741273,
+      "learning_rate": 2.953177819624713e-05,
+      "loss": 1.8988,
+      "step": 5946
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6535392139358569,
+      "learning_rate": 2.9525634838450834e-05,
+      "loss": 2.0778,
+      "step": 5947
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6638810545480212,
+      "learning_rate": 2.951949119810056e-05,
+      "loss": 2.0145,
+      "step": 5948
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6015073651731506,
+      "learning_rate": 2.951334727557989e-05,
+      "loss": 1.9383,
+      "step": 5949
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6427577626067742,
+      "learning_rate": 2.9507203071272404e-05,
+      "loss": 1.9349,
+      "step": 5950
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7472902852878983,
+      "learning_rate": 2.9501058585561713e-05,
+      "loss": 2.143,
+      "step": 5951
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6198206890509291,
+      "learning_rate": 2.9494913818831443e-05,
+      "loss": 1.9639,
+      "step": 5952
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5979936800180621,
+      "learning_rate": 2.9488768771465246e-05,
+      "loss": 1.9325,
+      "step": 5953
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5777256783062885,
+      "learning_rate": 2.948262344384677e-05,
+      "loss": 2.0319,
+      "step": 5954
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7053294183986611,
+      "learning_rate": 2.9476477836359694e-05,
+      "loss": 2.1467,
+      "step": 5955
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.563736640740669,
+      "learning_rate": 2.947033194938772e-05,
+      "loss": 1.9303,
+      "step": 5956
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.755739956087039,
+      "learning_rate": 2.946418578331456e-05,
+      "loss": 1.9339,
+      "step": 5957
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5795689857617631,
+      "learning_rate": 2.9458039338523936e-05,
+      "loss": 1.8778,
+      "step": 5958
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6715298216653058,
+      "learning_rate": 2.9451892615399607e-05,
+      "loss": 2.1671,
+      "step": 5959
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6365313182095643,
+      "learning_rate": 2.9445745614325337e-05,
+      "loss": 1.9715,
+      "step": 5960
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7217220329131973,
+      "learning_rate": 2.9439598335684893e-05,
+      "loss": 1.918,
+      "step": 5961
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6536711268682291,
+      "learning_rate": 2.943345077986209e-05,
+      "loss": 1.9244,
+      "step": 5962
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5971157805791927,
+      "learning_rate": 2.9427302947240747e-05,
+      "loss": 2.1523,
+      "step": 5963
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6721388121043124,
+      "learning_rate": 2.9421154838204702e-05,
+      "loss": 1.8868,
+      "step": 5964
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5901097327456989,
+      "learning_rate": 2.9415006453137777e-05,
+      "loss": 1.9206,
+      "step": 5965
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6013916989419127,
+      "learning_rate": 2.940885779242387e-05,
+      "loss": 1.9623,
+      "step": 5966
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6500430553437768,
+      "learning_rate": 2.940270885644687e-05,
+      "loss": 2.109,
+      "step": 5967
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5305325770273547,
+      "learning_rate": 2.939655964559066e-05,
+      "loss": 1.9062,
+      "step": 5968
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5686425637501226,
+      "learning_rate": 2.9390410160239162e-05,
+      "loss": 1.9069,
+      "step": 5969
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6181288397718674,
+      "learning_rate": 2.9384260400776332e-05,
+      "loss": 1.9082,
+      "step": 5970
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6010878185833171,
+      "learning_rate": 2.9378110367586115e-05,
+      "loss": 2.0718,
+      "step": 5971
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6404898518074024,
+      "learning_rate": 2.9371960061052478e-05,
+      "loss": 2.0232,
+      "step": 5972
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6122741082486938,
+      "learning_rate": 2.936580948155941e-05,
+      "loss": 1.9552,
+      "step": 5973
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7802081123553569,
+      "learning_rate": 2.9359658629490927e-05,
+      "loss": 1.913,
+      "step": 5974
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6671633665411347,
+      "learning_rate": 2.935350750523105e-05,
+      "loss": 2.187,
+      "step": 5975
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6844170526407003,
+      "learning_rate": 2.9347356109163803e-05,
+      "loss": 1.9088,
+      "step": 5976
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7637585308538776,
+      "learning_rate": 2.9341204441673266e-05,
+      "loss": 1.9218,
+      "step": 5977
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6329181075836443,
+      "learning_rate": 2.9335052503143488e-05,
+      "loss": 1.9727,
+      "step": 5978
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6628045370020772,
+      "learning_rate": 2.9328900293958583e-05,
+      "loss": 2.134,
+      "step": 5979
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7207841108382089,
+      "learning_rate": 2.932274781450265e-05,
+      "loss": 1.9223,
+      "step": 5980
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6168220054113233,
+      "learning_rate": 2.9316595065159815e-05,
+      "loss": 1.9768,
+      "step": 5981
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7358260198982544,
+      "learning_rate": 2.9310442046314208e-05,
+      "loss": 1.9494,
+      "step": 5982
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.717560500509033,
+      "learning_rate": 2.930428875835e-05,
+      "loss": 2.1116,
+      "step": 5983
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7831643477036163,
+      "learning_rate": 2.929813520165136e-05,
+      "loss": 1.9086,
+      "step": 5984
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5854716702326174,
+      "learning_rate": 2.929198137660249e-05,
+      "loss": 1.9832,
+      "step": 5985
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7115282403627274,
+      "learning_rate": 2.9285827283587587e-05,
+      "loss": 1.8967,
+      "step": 5986
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7733198993500927,
+      "learning_rate": 2.9279672922990874e-05,
+      "loss": 2.1253,
+      "step": 5987
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6566746307875865,
+      "learning_rate": 2.9273518295196606e-05,
+      "loss": 1.9572,
+      "step": 5988
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7421258753400743,
+      "learning_rate": 2.9267363400589036e-05,
+      "loss": 1.8925,
+      "step": 5989
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6622406955733426,
+      "learning_rate": 2.926120823955244e-05,
+      "loss": 1.9287,
+      "step": 5990
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7180158241893673,
+      "learning_rate": 2.9255052812471095e-05,
+      "loss": 2.1562,
+      "step": 5991
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7136305695771915,
+      "learning_rate": 2.9248897119729336e-05,
+      "loss": 1.9286,
+      "step": 5992
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6507400524003629,
+      "learning_rate": 2.9242741161711467e-05,
+      "loss": 1.9081,
+      "step": 5993
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6761297807425809,
+      "learning_rate": 2.9236584938801842e-05,
+      "loss": 1.9281,
+      "step": 5994
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7256114366956248,
+      "learning_rate": 2.9230428451384807e-05,
+      "loss": 2.1501,
+      "step": 5995
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8544743399555044,
+      "learning_rate": 2.922427169984476e-05,
+      "loss": 1.9347,
+      "step": 5996
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.796932107963117,
+      "learning_rate": 2.9218114684566067e-05,
+      "loss": 1.9943,
+      "step": 5997
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8625829152119489,
+      "learning_rate": 2.9211957405933144e-05,
+      "loss": 1.8956,
+      "step": 5998
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8907416044556657,
+      "learning_rate": 2.9205799864330425e-05,
+      "loss": 2.1307,
+      "step": 5999
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8481354418486149,
+      "learning_rate": 2.9199642060142333e-05,
+      "loss": 1.9001,
+      "step": 6000
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8110458884367083,
+      "learning_rate": 2.919348399375334e-05,
+      "loss": 1.9431,
+      "step": 6001
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6983623568008385,
+      "learning_rate": 2.9187325665547915e-05,
+      "loss": 1.9364,
+      "step": 6002
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.885705006631116,
+      "learning_rate": 2.918116707591054e-05,
+      "loss": 2.1654,
+      "step": 6003
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6231988290505693,
+      "learning_rate": 2.9175008225225735e-05,
+      "loss": 1.9289,
+      "step": 6004
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9352950724195798,
+      "learning_rate": 2.9168849113878015e-05,
+      "loss": 1.9102,
+      "step": 6005
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6339089231534399,
+      "learning_rate": 2.9162689742251913e-05,
+      "loss": 1.8939,
+      "step": 6006
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9594279193490054,
+      "learning_rate": 2.9156530110732e-05,
+      "loss": 2.1353,
+      "step": 6007
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6788463141369327,
+      "learning_rate": 2.9150370219702826e-05,
+      "loss": 1.9246,
+      "step": 6008
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8518331399669477,
+      "learning_rate": 2.9144210069548993e-05,
+      "loss": 1.9682,
+      "step": 6009
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7246511535194453,
+      "learning_rate": 2.91380496606551e-05,
+      "loss": 1.9815,
+      "step": 6010
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9307261416658241,
+      "learning_rate": 2.9131888993405772e-05,
+      "loss": 2.1287,
+      "step": 6011
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6475655844128312,
+      "learning_rate": 2.9125728068185632e-05,
+      "loss": 1.8807,
+      "step": 6012
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7629216445526122,
+      "learning_rate": 2.911956688537934e-05,
+      "loss": 1.9203,
+      "step": 6013
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5200944476500788,
+      "learning_rate": 2.9113405445371572e-05,
+      "loss": 1.879,
+      "step": 6014
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7816722641709961,
+      "learning_rate": 2.9107243748547002e-05,
+      "loss": 2.1537,
+      "step": 6015
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5489949333037935,
+      "learning_rate": 2.910108179529032e-05,
+      "loss": 1.9887,
+      "step": 6016
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.741292539163876,
+      "learning_rate": 2.9094919585986263e-05,
+      "loss": 1.9236,
+      "step": 6017
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5803594920771034,
+      "learning_rate": 2.9088757121019562e-05,
+      "loss": 1.9517,
+      "step": 6018
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7821026421642108,
+      "learning_rate": 2.908259440077495e-05,
+      "loss": 2.1103,
+      "step": 6019
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7419957883203638,
+      "learning_rate": 2.9076431425637186e-05,
+      "loss": 1.921,
+      "step": 6020
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7552171151522863,
+      "learning_rate": 2.907026819599108e-05,
+      "loss": 1.9162,
+      "step": 6021
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6109273457977054,
+      "learning_rate": 2.9064104712221402e-05,
+      "loss": 1.9777,
+      "step": 6022
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8284574135245422,
+      "learning_rate": 2.9057940974712967e-05,
+      "loss": 2.123,
+      "step": 6023
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5625931250820133,
+      "learning_rate": 2.9051776983850608e-05,
+      "loss": 1.9472,
+      "step": 6024
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7130654200441631,
+      "learning_rate": 2.9045612740019172e-05,
+      "loss": 1.9037,
+      "step": 6025
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6123249707788289,
+      "learning_rate": 2.9039448243603508e-05,
+      "loss": 1.9549,
+      "step": 6026
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6920475267902129,
+      "learning_rate": 2.9033283494988496e-05,
+      "loss": 2.126,
+      "step": 6027
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6273737386071347,
+      "learning_rate": 2.902711849455903e-05,
+      "loss": 2.0022,
+      "step": 6028
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7092443818611979,
+      "learning_rate": 2.9020953242700012e-05,
+      "loss": 1.9157,
+      "step": 6029
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6526950394158906,
+      "learning_rate": 2.9014787739796357e-05,
+      "loss": 1.8765,
+      "step": 6030
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6886205787344196,
+      "learning_rate": 2.900862198623302e-05,
+      "loss": 2.0765,
+      "step": 6031
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.608315282366492,
+      "learning_rate": 2.9002455982394944e-05,
+      "loss": 1.9495,
+      "step": 6032
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6204321830864946,
+      "learning_rate": 2.899628972866709e-05,
+      "loss": 1.9036,
+      "step": 6033
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7325716062979188,
+      "learning_rate": 2.8990123225434456e-05,
+      "loss": 1.9479,
+      "step": 6034
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7898018552032495,
+      "learning_rate": 2.8983956473082035e-05,
+      "loss": 2.1685,
+      "step": 6035
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6740754423569145,
+      "learning_rate": 2.8977789471994855e-05,
+      "loss": 1.9594,
+      "step": 6036
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5931584178554936,
+      "learning_rate": 2.897162222255792e-05,
+      "loss": 1.8588,
+      "step": 6037
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6105781461985351,
+      "learning_rate": 2.8965454725156304e-05,
+      "loss": 1.964,
+      "step": 6038
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7733784573568997,
+      "learning_rate": 2.8959286980175065e-05,
+      "loss": 2.1584,
+      "step": 6039
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6629231910999642,
+      "learning_rate": 2.895311898799927e-05,
+      "loss": 1.8734,
+      "step": 6040
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6336630662818065,
+      "learning_rate": 2.8946950749014012e-05,
+      "loss": 1.9989,
+      "step": 6041
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6976598515104996,
+      "learning_rate": 2.89407822636044e-05,
+      "loss": 1.9579,
+      "step": 6042
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6588442803929545,
+      "learning_rate": 2.8934613532155576e-05,
+      "loss": 2.1298,
+      "step": 6043
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5984901173273851,
+      "learning_rate": 2.892844455505266e-05,
+      "loss": 1.9541,
+      "step": 6044
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6124076658784989,
+      "learning_rate": 2.8922275332680804e-05,
+      "loss": 1.88,
+      "step": 6045
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.693190037613699,
+      "learning_rate": 2.8916105865425193e-05,
+      "loss": 1.9359,
+      "step": 6046
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6854303850161515,
+      "learning_rate": 2.890993615367101e-05,
+      "loss": 2.1658,
+      "step": 6047
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.620317473937023,
+      "learning_rate": 2.890376619780344e-05,
+      "loss": 1.9665,
+      "step": 6048
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6017633292389617,
+      "learning_rate": 2.8897595998207717e-05,
+      "loss": 1.9135,
+      "step": 6049
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6695049851816347,
+      "learning_rate": 2.8891425555269058e-05,
+      "loss": 1.9482,
+      "step": 6050
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6579856773407556,
+      "learning_rate": 2.888525486937272e-05,
+      "loss": 2.1544,
+      "step": 6051
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6943533170253708,
+      "learning_rate": 2.8879083940903957e-05,
+      "loss": 1.9584,
+      "step": 6052
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7728571549998186,
+      "learning_rate": 2.8872912770248044e-05,
+      "loss": 1.9971,
+      "step": 6053
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.790100486381364,
+      "learning_rate": 2.8866741357790283e-05,
+      "loss": 1.921,
+      "step": 6054
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8250976546382595,
+      "learning_rate": 2.8860569703915975e-05,
+      "loss": 2.1303,
+      "step": 6055
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5617133295986723,
+      "learning_rate": 2.8854397809010436e-05,
+      "loss": 1.8921,
+      "step": 6056
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7769489026042259,
+      "learning_rate": 2.884822567345901e-05,
+      "loss": 1.9517,
+      "step": 6057
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6641212926741072,
+      "learning_rate": 2.8842053297647047e-05,
+      "loss": 1.9326,
+      "step": 6058
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6963086897952515,
+      "learning_rate": 2.883588068195991e-05,
+      "loss": 1.9416,
+      "step": 6059
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7279431345902708,
+      "learning_rate": 2.882970782678298e-05,
+      "loss": 2.107,
+      "step": 6060
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6968588048479214,
+      "learning_rate": 2.882353473250167e-05,
+      "loss": 1.9444,
+      "step": 6061
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7109336254401915,
+      "learning_rate": 2.8817361399501375e-05,
+      "loss": 1.9478,
+      "step": 6062
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7157506190781077,
+      "learning_rate": 2.881118782816752e-05,
+      "loss": 2.1161,
+      "step": 6063
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6585982358935669,
+      "learning_rate": 2.8805014018885556e-05,
+      "loss": 1.9232,
+      "step": 6064
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6466991773353137,
+      "learning_rate": 2.8798839972040935e-05,
+      "loss": 1.8945,
+      "step": 6065
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6788620826862066,
+      "learning_rate": 2.8792665688019127e-05,
+      "loss": 1.9795,
+      "step": 6066
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7929808849009231,
+      "learning_rate": 2.8786491167205613e-05,
+      "loss": 2.1469,
+      "step": 6067
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.765492373526748,
+      "learning_rate": 2.8780316409985915e-05,
+      "loss": 1.8901,
+      "step": 6068
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6430955345410521,
+      "learning_rate": 2.8774141416745524e-05,
+      "loss": 1.8766,
+      "step": 6069
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7970949401462726,
+      "learning_rate": 2.8767966187869983e-05,
+      "loss": 1.9247,
+      "step": 6070
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7224962921434076,
+      "learning_rate": 2.8761790723744834e-05,
+      "loss": 2.1156,
+      "step": 6071
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7898177335206944,
+      "learning_rate": 2.8755615024755633e-05,
+      "loss": 1.9961,
+      "step": 6072
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8446886317861725,
+      "learning_rate": 2.8749439091287962e-05,
+      "loss": 1.9131,
+      "step": 6073
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7758007328045595,
+      "learning_rate": 2.87432629237274e-05,
+      "loss": 1.9052,
+      "step": 6074
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.951844290078726,
+      "learning_rate": 2.8737086522459563e-05,
+      "loss": 2.1448,
+      "step": 6075
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5842599623911655,
+      "learning_rate": 2.8730909887870057e-05,
+      "loss": 1.9293,
+      "step": 6076
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9858072108986572,
+      "learning_rate": 2.872473302034452e-05,
+      "loss": 1.9186,
+      "step": 6077
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6881601169750788,
+      "learning_rate": 2.87185559202686e-05,
+      "loss": 2.0033,
+      "step": 6078
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7856721072346264,
+      "learning_rate": 2.871237858802796e-05,
+      "loss": 1.9302,
+      "step": 6079
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7997111429447697,
+      "learning_rate": 2.8706201024008267e-05,
+      "loss": 2.1093,
+      "step": 6080
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.626295539419116,
+      "learning_rate": 2.8700023228595225e-05,
+      "loss": 1.8793,
+      "step": 6081
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7778128002423056,
+      "learning_rate": 2.869384520217453e-05,
+      "loss": 1.9124,
+      "step": 6082
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7314199482865292,
+      "learning_rate": 2.868766694513191e-05,
+      "loss": 2.0705,
+      "step": 6083
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5738717192730649,
+      "learning_rate": 2.868148845785308e-05,
+      "loss": 2.0203,
+      "step": 6084
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.792030696512876,
+      "learning_rate": 2.867530974072381e-05,
+      "loss": 1.9414,
+      "step": 6085
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7172438226032153,
+      "learning_rate": 2.8669130794129863e-05,
+      "loss": 1.9345,
+      "step": 6086
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8222602144114246,
+      "learning_rate": 2.8662951618456996e-05,
+      "loss": 2.1253,
+      "step": 6087
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7206897007433584,
+      "learning_rate": 2.865677221409101e-05,
+      "loss": 1.9173,
+      "step": 6088
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6571108072797999,
+      "learning_rate": 2.8650592581417718e-05,
+      "loss": 1.8818,
+      "step": 6089
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.911692898390632,
+      "learning_rate": 2.8644412720822933e-05,
+      "loss": 1.9718,
+      "step": 6090
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5496703802805056,
+      "learning_rate": 2.863823263269249e-05,
+      "loss": 1.8786,
+      "step": 6091
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7499032976311122,
+      "learning_rate": 2.863205231741224e-05,
+      "loss": 2.1219,
+      "step": 6092
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6175999930661693,
+      "learning_rate": 2.862587177536804e-05,
+      "loss": 1.8987,
+      "step": 6093
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6292828661928492,
+      "learning_rate": 2.861969100694577e-05,
+      "loss": 1.9428,
+      "step": 6094
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7337999951298932,
+      "learning_rate": 2.861351001253132e-05,
+      "loss": 2.1104,
+      "step": 6095
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.675859228106822,
+      "learning_rate": 2.8607328792510595e-05,
+      "loss": 1.898,
+      "step": 6096
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.661865080266223,
+      "learning_rate": 2.8601147347269514e-05,
+      "loss": 1.9985,
+      "step": 6097
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5937158924458813,
+      "learning_rate": 2.8594965677194013e-05,
+      "loss": 1.9324,
+      "step": 6098
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6074448734327125,
+      "learning_rate": 2.858878378267003e-05,
+      "loss": 2.1222,
+      "step": 6099
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6270234800502213,
+      "learning_rate": 2.858260166408354e-05,
+      "loss": 1.892,
+      "step": 6100
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6092753354246233,
+      "learning_rate": 2.8576419321820507e-05,
+      "loss": 1.8842,
+      "step": 6101
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6812256014787198,
+      "learning_rate": 2.857023675626692e-05,
+      "loss": 1.9156,
+      "step": 6102
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6009945673494914,
+      "learning_rate": 2.856405396780879e-05,
+      "loss": 2.0136,
+      "step": 6103
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7282707919402174,
+      "learning_rate": 2.8557870956832132e-05,
+      "loss": 2.151,
+      "step": 6104
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5514084935556735,
+      "learning_rate": 2.855168772372297e-05,
+      "loss": 1.9077,
+      "step": 6105
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6554203992770687,
+      "learning_rate": 2.854550426886735e-05,
+      "loss": 1.9441,
+      "step": 6106
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7310541932742173,
+      "learning_rate": 2.8539320592651336e-05,
+      "loss": 2.1253,
+      "step": 6107
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.594755774719508,
+      "learning_rate": 2.8533136695461004e-05,
+      "loss": 1.8911,
+      "step": 6108
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.618015005561634,
+      "learning_rate": 2.8526952577682425e-05,
+      "loss": 1.9736,
+      "step": 6109
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5911892757457342,
+      "learning_rate": 2.8520768239701707e-05,
+      "loss": 1.9123,
+      "step": 6110
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7290930809107585,
+      "learning_rate": 2.8514583681904972e-05,
+      "loss": 1.9127,
+      "step": 6111
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5634256326604189,
+      "learning_rate": 2.8508398904678334e-05,
+      "loss": 2.1181,
+      "step": 6112
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6367236849331929,
+      "learning_rate": 2.8502213908407937e-05,
+      "loss": 1.9486,
+      "step": 6113
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6176580245284353,
+      "learning_rate": 2.8496028693479938e-05,
+      "loss": 1.9372,
+      "step": 6114
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7009382742760467,
+      "learning_rate": 2.8489843260280513e-05,
+      "loss": 2.0241,
+      "step": 6115
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6130760663920718,
+      "learning_rate": 2.8483657609195835e-05,
+      "loss": 2.1075,
+      "step": 6116
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6104278131448414,
+      "learning_rate": 2.8477471740612094e-05,
+      "loss": 1.896,
+      "step": 6117
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5945735973185974,
+      "learning_rate": 2.8471285654915512e-05,
+      "loss": 1.9348,
+      "step": 6118
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6541514904034934,
+      "learning_rate": 2.8465099352492307e-05,
+      "loss": 2.0572,
+      "step": 6119
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5787861155529801,
+      "learning_rate": 2.845891283372871e-05,
+      "loss": 1.9138,
+      "step": 6120
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5971763521927723,
+      "learning_rate": 2.845272609901098e-05,
+      "loss": 1.9641,
+      "step": 6121
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6436377618779282,
+      "learning_rate": 2.844653914872537e-05,
+      "loss": 1.9603,
+      "step": 6122
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6244497706521172,
+      "learning_rate": 2.844035198325817e-05,
+      "loss": 1.8929,
+      "step": 6123
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6290549793758208,
+      "learning_rate": 2.8434164602995655e-05,
+      "loss": 2.129,
+      "step": 6124
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6793031626704319,
+      "learning_rate": 2.8427977008324142e-05,
+      "loss": 1.8963,
+      "step": 6125
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6042645089643837,
+      "learning_rate": 2.8421789199629944e-05,
+      "loss": 1.9367,
+      "step": 6126
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7570077202776087,
+      "learning_rate": 2.8415601177299377e-05,
+      "loss": 2.1181,
+      "step": 6127
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5524239611151556,
+      "learning_rate": 2.8409412941718805e-05,
+      "loss": 2.0019,
+      "step": 6128
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6257208772204805,
+      "learning_rate": 2.840322449327458e-05,
+      "loss": 1.9449,
+      "step": 6129
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6241737322143389,
+      "learning_rate": 2.8397035832353065e-05,
+      "loss": 1.9102,
+      "step": 6130
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5951944577986961,
+      "learning_rate": 2.8390846959340638e-05,
+      "loss": 2.1369,
+      "step": 6131
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5806862807364876,
+      "learning_rate": 2.8384657874623716e-05,
+      "loss": 1.9321,
+      "step": 6132
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.626804204217459,
+      "learning_rate": 2.8378468578588703e-05,
+      "loss": 1.9192,
+      "step": 6133
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5864251812693121,
+      "learning_rate": 2.837227907162201e-05,
+      "loss": 2.0457,
+      "step": 6134
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6549455162764742,
+      "learning_rate": 2.836608935411008e-05,
+      "loss": 1.9088,
+      "step": 6135
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5614088505271921,
+      "learning_rate": 2.8359899426439374e-05,
+      "loss": 2.0919,
+      "step": 6136
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7003340149978751,
+      "learning_rate": 2.8353709288996333e-05,
+      "loss": 1.9978,
+      "step": 6137
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5766927691662888,
+      "learning_rate": 2.8347518942167444e-05,
+      "loss": 1.9665,
+      "step": 6138
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.548015639669285,
+      "learning_rate": 2.83413283863392e-05,
+      "loss": 2.1375,
+      "step": 6139
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5188959315707593,
+      "learning_rate": 2.8335137621898095e-05,
+      "loss": 1.8921,
+      "step": 6140
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5845271040985408,
+      "learning_rate": 2.8328946649230647e-05,
+      "loss": 1.9146,
+      "step": 6141
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5985565151098161,
+      "learning_rate": 2.8322755468723387e-05,
+      "loss": 1.9291,
+      "step": 6142
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5583602207056978,
+      "learning_rate": 2.8316564080762854e-05,
+      "loss": 1.9573,
+      "step": 6143
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5808160069754263,
+      "learning_rate": 2.8310372485735597e-05,
+      "loss": 2.1284,
+      "step": 6144
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6245984797766763,
+      "learning_rate": 2.8304180684028187e-05,
+      "loss": 1.9659,
+      "step": 6145
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5463832253494476,
+      "learning_rate": 2.8297988676027203e-05,
+      "loss": 1.9819,
+      "step": 6146
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6290164631193843,
+      "learning_rate": 2.8291796462119246e-05,
+      "loss": 1.9101,
+      "step": 6147
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5937066422120306,
+      "learning_rate": 2.8285604042690894e-05,
+      "loss": 2.1315,
+      "step": 6148
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6532670149783002,
+      "learning_rate": 2.8279411418128798e-05,
+      "loss": 1.9147,
+      "step": 6149
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6641452628773036,
+      "learning_rate": 2.8273218588819573e-05,
+      "loss": 1.9303,
+      "step": 6150
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6492564426112483,
+      "learning_rate": 2.8267025555149867e-05,
+      "loss": 2.1218,
+      "step": 6151
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6654462429076632,
+      "learning_rate": 2.8260832317506326e-05,
+      "loss": 1.9879,
+      "step": 6152
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6035637213693915,
+      "learning_rate": 2.825463887627563e-05,
+      "loss": 1.8904,
+      "step": 6153
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5842202944278496,
+      "learning_rate": 2.8248445231844466e-05,
+      "loss": 1.9454,
+      "step": 6154
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6810767396249412,
+      "learning_rate": 2.8242251384599516e-05,
+      "loss": 1.9095,
+      "step": 6155
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6713845080647166,
+      "learning_rate": 2.8236057334927494e-05,
+      "loss": 2.1393,
+      "step": 6156
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5653520002476087,
+      "learning_rate": 2.8229863083215113e-05,
+      "loss": 1.9318,
+      "step": 6157
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7087058206744752,
+      "learning_rate": 2.822366862984912e-05,
+      "loss": 1.8836,
+      "step": 6158
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6229565829493346,
+      "learning_rate": 2.8217473975216247e-05,
+      "loss": 1.954,
+      "step": 6159
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7815621087382343,
+      "learning_rate": 2.8211279119703253e-05,
+      "loss": 2.0912,
+      "step": 6160
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5396882479747217,
+      "learning_rate": 2.820508406369692e-05,
+      "loss": 1.9726,
+      "step": 6161
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7976339825171728,
+      "learning_rate": 2.8198888807584013e-05,
+      "loss": 1.9525,
+      "step": 6162
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6038848610897559,
+      "learning_rate": 2.8192693351751342e-05,
+      "loss": 1.858,
+      "step": 6163
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.703904694513044,
+      "learning_rate": 2.8186497696585707e-05,
+      "loss": 2.1853,
+      "step": 6164
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5839196280763157,
+      "learning_rate": 2.8180301842473926e-05,
+      "loss": 1.9819,
+      "step": 6165
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7046973964425847,
+      "learning_rate": 2.817410578980284e-05,
+      "loss": 1.9532,
+      "step": 6166
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5848964288431363,
+      "learning_rate": 2.8167909538959292e-05,
+      "loss": 1.9219,
+      "step": 6167
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.754947494831866,
+      "learning_rate": 2.8161713090330132e-05,
+      "loss": 2.1235,
+      "step": 6168
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6366837932948054,
+      "learning_rate": 2.8155516444302237e-05,
+      "loss": 1.8788,
+      "step": 6169
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7108136727958364,
+      "learning_rate": 2.8149319601262492e-05,
+      "loss": 1.8844,
+      "step": 6170
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6030695718504766,
+      "learning_rate": 2.8143122561597784e-05,
+      "loss": 2.0339,
+      "step": 6171
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6403831619276319,
+      "learning_rate": 2.8136925325695017e-05,
+      "loss": 2.1471,
+      "step": 6172
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6154977051718946,
+      "learning_rate": 2.813072789394113e-05,
+      "loss": 1.8562,
+      "step": 6173
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7061323377689693,
+      "learning_rate": 2.812453026672302e-05,
+      "loss": 1.9135,
+      "step": 6174
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6152239140420555,
+      "learning_rate": 2.811833244442766e-05,
+      "loss": 1.9313,
+      "step": 6175
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7293025277019005,
+      "learning_rate": 2.8112134427441998e-05,
+      "loss": 2.1367,
+      "step": 6176
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.576462525545831,
+      "learning_rate": 2.8105936216152996e-05,
+      "loss": 2.0389,
+      "step": 6177
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6069390237623686,
+      "learning_rate": 2.809973781094763e-05,
+      "loss": 1.9373,
+      "step": 6178
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6314378157304235,
+      "learning_rate": 2.8093539212212905e-05,
+      "loss": 1.9404,
+      "step": 6179
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7083517174409519,
+      "learning_rate": 2.8087340420335818e-05,
+      "loss": 2.1407,
+      "step": 6180
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6085739442812378,
+      "learning_rate": 2.8081141435703384e-05,
+      "loss": 1.8795,
+      "step": 6181
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.686347668277774,
+      "learning_rate": 2.8074942258702625e-05,
+      "loss": 1.9214,
+      "step": 6182
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7562017784454768,
+      "learning_rate": 2.8068742889720602e-05,
+      "loss": 2.1409,
+      "step": 6183
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6131328833498226,
+      "learning_rate": 2.8062543329144347e-05,
+      "loss": 2.0185,
+      "step": 6184
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7561551937959166,
+      "learning_rate": 2.8056343577360932e-05,
+      "loss": 1.9472,
+      "step": 6185
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7064216136558119,
+      "learning_rate": 2.805014363475743e-05,
+      "loss": 1.9215,
+      "step": 6186
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5945729312441711,
+      "learning_rate": 2.804394350172093e-05,
+      "loss": 1.8748,
+      "step": 6187
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8767651977002929,
+      "learning_rate": 2.8037743178638536e-05,
+      "loss": 2.1211,
+      "step": 6188
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7120981152521955,
+      "learning_rate": 2.8031542665897353e-05,
+      "loss": 1.9144,
+      "step": 6189
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6513010227419045,
+      "learning_rate": 2.802534196388451e-05,
+      "loss": 1.9783,
+      "step": 6190
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6527548377306654,
+      "learning_rate": 2.8019141072987138e-05,
+      "loss": 1.8964,
+      "step": 6191
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7180908197047047,
+      "learning_rate": 2.8012939993592386e-05,
+      "loss": 2.0873,
+      "step": 6192
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5644956398775843,
+      "learning_rate": 2.800673872608741e-05,
+      "loss": 1.9113,
+      "step": 6193
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6723416665426786,
+      "learning_rate": 2.8000537270859393e-05,
+      "loss": 1.8887,
+      "step": 6194
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6503820382184902,
+      "learning_rate": 2.7994335628295497e-05,
+      "loss": 1.9419,
+      "step": 6195
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6320788178361572,
+      "learning_rate": 2.798813379878293e-05,
+      "loss": 2.1741,
+      "step": 6196
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5783328004963862,
+      "learning_rate": 2.798193178270889e-05,
+      "loss": 1.9284,
+      "step": 6197
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6536369794030019,
+      "learning_rate": 2.7975729580460613e-05,
+      "loss": 1.9154,
+      "step": 6198
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.599713590673247,
+      "learning_rate": 2.79695271924253e-05,
+      "loss": 1.9504,
+      "step": 6199
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6992959079936097,
+      "learning_rate": 2.796332461899021e-05,
+      "loss": 2.1409,
+      "step": 6200
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5718521089859425,
+      "learning_rate": 2.7957121860542595e-05,
+      "loss": 1.9513,
+      "step": 6201
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6516463898451217,
+      "learning_rate": 2.795091891746971e-05,
+      "loss": 1.9985,
+      "step": 6202
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6394872787664603,
+      "learning_rate": 2.7944715790158826e-05,
+      "loss": 1.9417,
+      "step": 6203
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6762012398491848,
+      "learning_rate": 2.7938512478997247e-05,
+      "loss": 2.1098,
+      "step": 6204
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5778828320413963,
+      "learning_rate": 2.7932308984372264e-05,
+      "loss": 1.8797,
+      "step": 6205
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5898964487828565,
+      "learning_rate": 2.792610530667118e-05,
+      "loss": 1.9294,
+      "step": 6206
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7816938189025686,
+      "learning_rate": 2.7919901446281327e-05,
+      "loss": 1.9356,
+      "step": 6207
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5807735827522142,
+      "learning_rate": 2.7913697403590022e-05,
+      "loss": 2.2003,
+      "step": 6208
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7416073177594824,
+      "learning_rate": 2.790749317898463e-05,
+      "loss": 1.9687,
+      "step": 6209
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6570188210974146,
+      "learning_rate": 2.7901288772852486e-05,
+      "loss": 1.9539,
+      "step": 6210
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6492437513193234,
+      "learning_rate": 2.7895084185580968e-05,
+      "loss": 1.9314,
+      "step": 6211
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7676241786619878,
+      "learning_rate": 2.7888879417557447e-05,
+      "loss": 2.0943,
+      "step": 6212
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6137155261653936,
+      "learning_rate": 2.788267446916932e-05,
+      "loss": 1.9101,
+      "step": 6213
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7993431988100764,
+      "learning_rate": 2.787646934080398e-05,
+      "loss": 1.9186,
+      "step": 6214
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6681505255680475,
+      "learning_rate": 2.7870264032848847e-05,
+      "loss": 2.001,
+      "step": 6215
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6817111604437344,
+      "learning_rate": 2.7864058545691335e-05,
+      "loss": 2.0707,
+      "step": 6216
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7807959088349454,
+      "learning_rate": 2.7857852879718883e-05,
+      "loss": 1.9475,
+      "step": 6217
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6483716676870414,
+      "learning_rate": 2.785164703531894e-05,
+      "loss": 1.8711,
+      "step": 6218
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7819843106273833,
+      "learning_rate": 2.7845441012878958e-05,
+      "loss": 1.9566,
+      "step": 6219
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8153074197466349,
+      "learning_rate": 2.7839234812786398e-05,
+      "loss": 2.1171,
+      "step": 6220
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7899667362837469,
+      "learning_rate": 2.7833028435428744e-05,
+      "loss": 2.0442,
+      "step": 6221
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7751969810908127,
+      "learning_rate": 2.7826821881193488e-05,
+      "loss": 1.9313,
+      "step": 6222
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7277906156214009,
+      "learning_rate": 2.782061515046814e-05,
+      "loss": 1.934,
+      "step": 6223
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6810980516631513,
+      "learning_rate": 2.781440824364019e-05,
+      "loss": 2.1377,
+      "step": 6224
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.683502361825806,
+      "learning_rate": 2.7808201161097164e-05,
+      "loss": 1.9203,
+      "step": 6225
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7533439572775853,
+      "learning_rate": 2.780199390322662e-05,
+      "loss": 1.9185,
+      "step": 6226
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6924290575414431,
+      "learning_rate": 2.7795786470416075e-05,
+      "loss": 2.053,
+      "step": 6227
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6600270829700075,
+      "learning_rate": 2.77895788630531e-05,
+      "loss": 2.1536,
+      "step": 6228
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6329372795595144,
+      "learning_rate": 2.778337108152525e-05,
+      "loss": 1.9058,
+      "step": 6229
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.639861170322704,
+      "learning_rate": 2.7777163126220118e-05,
+      "loss": 1.9785,
+      "step": 6230
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7040172819721391,
+      "learning_rate": 2.7770954997525277e-05,
+      "loss": 1.9785,
+      "step": 6231
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7186256311206687,
+      "learning_rate": 2.7764746695828338e-05,
+      "loss": 2.1167,
+      "step": 6232
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5552706841250771,
+      "learning_rate": 2.77585382215169e-05,
+      "loss": 2.0279,
+      "step": 6233
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6550710334390859,
+      "learning_rate": 2.7752329574978587e-05,
+      "loss": 1.9226,
+      "step": 6234
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7397534607079066,
+      "learning_rate": 2.7746120756601036e-05,
+      "loss": 1.8638,
+      "step": 6235
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6768705864741504,
+      "learning_rate": 2.773991176677188e-05,
+      "loss": 2.086,
+      "step": 6236
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6123262735101047,
+      "learning_rate": 2.773370260587878e-05,
+      "loss": 1.8896,
+      "step": 6237
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6817026461177041,
+      "learning_rate": 2.7727493274309397e-05,
+      "loss": 1.9399,
+      "step": 6238
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6704821989407344,
+      "learning_rate": 2.77212837724514e-05,
+      "loss": 2.013,
+      "step": 6239
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6769753792453249,
+      "learning_rate": 2.7715074100692485e-05,
+      "loss": 2.1333,
+      "step": 6240
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6488463278760194,
+      "learning_rate": 2.7708864259420343e-05,
+      "loss": 1.9872,
+      "step": 6241
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7098015590597466,
+      "learning_rate": 2.7702654249022658e-05,
+      "loss": 1.9316,
+      "step": 6242
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6843690213052005,
+      "learning_rate": 2.769644406988718e-05,
+      "loss": 1.9908,
+      "step": 6243
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8762668527890835,
+      "learning_rate": 2.7690233722401622e-05,
+      "loss": 2.0975,
+      "step": 6244
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5770720704679605,
+      "learning_rate": 2.7684023206953718e-05,
+      "loss": 1.939,
+      "step": 6245
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7628804012746702,
+      "learning_rate": 2.7677812523931214e-05,
+      "loss": 2.0027,
+      "step": 6246
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6034161507441359,
+      "learning_rate": 2.7671601673721875e-05,
+      "loss": 1.9441,
+      "step": 6247
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7497617206672372,
+      "learning_rate": 2.7665390656713474e-05,
+      "loss": 2.1,
+      "step": 6248
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6197471813617871,
+      "learning_rate": 2.7659179473293785e-05,
+      "loss": 1.9101,
+      "step": 6249
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6742153165713498,
+      "learning_rate": 2.7652968123850586e-05,
+      "loss": 1.9398,
+      "step": 6250
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6490613315645393,
+      "learning_rate": 2.7646756608771707e-05,
+      "loss": 1.9596,
+      "step": 6251
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6230335999776684,
+      "learning_rate": 2.7640544928444927e-05,
+      "loss": 2.1408,
+      "step": 6252
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5873823734704713,
+      "learning_rate": 2.7634333083258078e-05,
+      "loss": 1.8557,
+      "step": 6253
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7394570898366446,
+      "learning_rate": 2.7628121073598996e-05,
+      "loss": 1.9595,
+      "step": 6254
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.603952903734535,
+      "learning_rate": 2.762190889985552e-05,
+      "loss": 1.9438,
+      "step": 6255
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7066202772028929,
+      "learning_rate": 2.76156965624155e-05,
+      "loss": 2.0827,
+      "step": 6256
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.619922854523307,
+      "learning_rate": 2.7609484061666797e-05,
+      "loss": 1.9264,
+      "step": 6257
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6373186179076702,
+      "learning_rate": 2.7603271397997283e-05,
+      "loss": 2.0163,
+      "step": 6258
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7085674482236836,
+      "learning_rate": 2.759705857179484e-05,
+      "loss": 1.9401,
+      "step": 6259
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8460768376705932,
+      "learning_rate": 2.7590845583447366e-05,
+      "loss": 2.1166,
+      "step": 6260
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6407379021518438,
+      "learning_rate": 2.7584632433342755e-05,
+      "loss": 1.8905,
+      "step": 6261
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.822180832229733,
+      "learning_rate": 2.7578419121868927e-05,
+      "loss": 1.8983,
+      "step": 6262
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6852591998882038,
+      "learning_rate": 2.7572205649413795e-05,
+      "loss": 1.9379,
+      "step": 6263
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9455577536761596,
+      "learning_rate": 2.7565992016365305e-05,
+      "loss": 2.2034,
+      "step": 6264
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6990673944455624,
+      "learning_rate": 2.755977822311139e-05,
+      "loss": 1.9272,
+      "step": 6265
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8673850285092596,
+      "learning_rate": 2.7553564270040004e-05,
+      "loss": 1.91,
+      "step": 6266
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7243019347774309,
+      "learning_rate": 2.7547350157539102e-05,
+      "loss": 1.9443,
+      "step": 6267
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7808072509537783,
+      "learning_rate": 2.754113588599667e-05,
+      "loss": 2.0962,
+      "step": 6268
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6098559360605493,
+      "learning_rate": 2.7534921455800688e-05,
+      "loss": 1.9306,
+      "step": 6269
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7130255314547961,
+      "learning_rate": 2.7528706867339155e-05,
+      "loss": 2.0059,
+      "step": 6270
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6684206521914795,
+      "learning_rate": 2.7522492121000053e-05,
+      "loss": 1.9323,
+      "step": 6271
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8614901461539285,
+      "learning_rate": 2.75162772171714e-05,
+      "loss": 2.1314,
+      "step": 6272
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6344629885555926,
+      "learning_rate": 2.7510062156241233e-05,
+      "loss": 1.9489,
+      "step": 6273
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7758436503945126,
+      "learning_rate": 2.7503846938597573e-05,
+      "loss": 1.8288,
+      "step": 6274
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.605220214384712,
+      "learning_rate": 2.7497631564628463e-05,
+      "loss": 1.9024,
+      "step": 6275
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7868727511479419,
+      "learning_rate": 2.7491416034721947e-05,
+      "loss": 2.1032,
+      "step": 6276
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.631998244984183,
+      "learning_rate": 2.74852003492661e-05,
+      "loss": 1.9976,
+      "step": 6277
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.677611523979507,
+      "learning_rate": 2.7478984508648982e-05,
+      "loss": 1.9082,
+      "step": 6278
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6037355082206972,
+      "learning_rate": 2.7472768513258674e-05,
+      "loss": 1.9162,
+      "step": 6279
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7325722620758016,
+      "learning_rate": 2.746655236348328e-05,
+      "loss": 2.1248,
+      "step": 6280
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6374940967422656,
+      "learning_rate": 2.746033605971088e-05,
+      "loss": 1.9327,
+      "step": 6281
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6198196505856042,
+      "learning_rate": 2.745411960232959e-05,
+      "loss": 1.9322,
+      "step": 6282
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.8806410171992154,
+      "learning_rate": 2.7447902991727538e-05,
+      "loss": 2.0063,
+      "step": 6283
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6195516100498301,
+      "learning_rate": 2.7441686228292845e-05,
+      "loss": 2.1015,
+      "step": 6284
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6724804372020051,
+      "learning_rate": 2.7435469312413646e-05,
+      "loss": 1.929,
+      "step": 6285
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7303353372836486,
+      "learning_rate": 2.74292522444781e-05,
+      "loss": 1.9625,
+      "step": 6286
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6388992246806826,
+      "learning_rate": 2.742303502487435e-05,
+      "loss": 1.9379,
+      "step": 6287
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6474355819923994,
+      "learning_rate": 2.741681765399058e-05,
+      "loss": 2.1119,
+      "step": 6288
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.706857478088858,
+      "learning_rate": 2.7410600132214943e-05,
+      "loss": 2.0147,
+      "step": 6289
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6004460174570148,
+      "learning_rate": 2.740438245993564e-05,
+      "loss": 1.9207,
+      "step": 6290
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6111838454979798,
+      "learning_rate": 2.739816463754087e-05,
+      "loss": 1.9263,
+      "step": 6291
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7399164053496282,
+      "learning_rate": 2.739194666541882e-05,
+      "loss": 2.0721,
+      "step": 6292
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6003525359252971,
+      "learning_rate": 2.7385728543957712e-05,
+      "loss": 1.881,
+      "step": 6293
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6059831079086222,
+      "learning_rate": 2.737951027354578e-05,
+      "loss": 1.9259,
+      "step": 6294
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8741102411667938,
+      "learning_rate": 2.7373291854571242e-05,
+      "loss": 1.9761,
+      "step": 6295
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6247820810764958,
+      "learning_rate": 2.736707328742234e-05,
+      "loss": 1.9127,
+      "step": 6296
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9251347795626271,
+      "learning_rate": 2.7360854572487328e-05,
+      "loss": 2.0991,
+      "step": 6297
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6272725172961547,
+      "learning_rate": 2.7354635710154476e-05,
+      "loss": 1.9306,
+      "step": 6298
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9061320015216355,
+      "learning_rate": 2.734841670081204e-05,
+      "loss": 1.9352,
+      "step": 6299
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5739559382400902,
+      "learning_rate": 2.7342197544848298e-05,
+      "loss": 2.1169,
+      "step": 6300
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8159598425787489,
+      "learning_rate": 2.7335978242651538e-05,
+      "loss": 1.988,
+      "step": 6301
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5958532971495627,
+      "learning_rate": 2.7329758794610067e-05,
+      "loss": 1.8957,
+      "step": 6302
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6864914249263036,
+      "learning_rate": 2.732353920111218e-05,
+      "loss": 1.9273,
+      "step": 6303
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7457351298560015,
+      "learning_rate": 2.73173194625462e-05,
+      "loss": 2.1102,
+      "step": 6304
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6434126360288767,
+      "learning_rate": 2.7311099579300443e-05,
+      "loss": 1.9104,
+      "step": 6305
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7744130679943296,
+      "learning_rate": 2.730487955176324e-05,
+      "loss": 1.9333,
+      "step": 6306
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6963486857136643,
+      "learning_rate": 2.7298659380322945e-05,
+      "loss": 1.9047,
+      "step": 6307
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6944298369570352,
+      "learning_rate": 2.7292439065367897e-05,
+      "loss": 1.961,
+      "step": 6308
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7594025378885407,
+      "learning_rate": 2.7286218607286474e-05,
+      "loss": 2.0973,
+      "step": 6309
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6702155753060279,
+      "learning_rate": 2.7279998006467013e-05,
+      "loss": 1.9062,
+      "step": 6310
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7141238335824098,
+      "learning_rate": 2.7273777263297918e-05,
+      "loss": 1.9002,
+      "step": 6311
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7850918367434662,
+      "learning_rate": 2.7267556378167562e-05,
+      "loss": 2.0622,
+      "step": 6312
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7853737021525058,
+      "learning_rate": 2.7261335351464356e-05,
+      "loss": 1.9067,
+      "step": 6313
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7180192248553696,
+      "learning_rate": 2.7255114183576686e-05,
+      "loss": 1.9566,
+      "step": 6314
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7293039767112508,
+      "learning_rate": 2.7248892874892973e-05,
+      "loss": 1.9476,
+      "step": 6315
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.0638097239181599,
+      "learning_rate": 2.7242671425801653e-05,
+      "loss": 2.1024,
+      "step": 6316
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6103435582410526,
+      "learning_rate": 2.723644983669113e-05,
+      "loss": 1.9338,
+      "step": 6317
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9525562407998518,
+      "learning_rate": 2.7230228107949852e-05,
+      "loss": 1.9767,
+      "step": 6318
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7404284053967155,
+      "learning_rate": 2.7224006239966276e-05,
+      "loss": 1.9288,
+      "step": 6319
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7664604999074345,
+      "learning_rate": 2.7217784233128863e-05,
+      "loss": 1.9619,
+      "step": 6320
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8083402247020052,
+      "learning_rate": 2.721156208782606e-05,
+      "loss": 2.1128,
+      "step": 6321
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8589773230302058,
+      "learning_rate": 2.7205339804446352e-05,
+      "loss": 1.9453,
+      "step": 6322
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8037461703300274,
+      "learning_rate": 2.719911738337822e-05,
+      "loss": 1.9369,
+      "step": 6323
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.902684100404686,
+      "learning_rate": 2.7192894825010157e-05,
+      "loss": 2.036,
+      "step": 6324
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7397437051689738,
+      "learning_rate": 2.718667212973066e-05,
+      "loss": 1.9031,
+      "step": 6325
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8863541555251446,
+      "learning_rate": 2.718044929792824e-05,
+      "loss": 2.0063,
+      "step": 6326
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6734382630234651,
+      "learning_rate": 2.7174226329991415e-05,
+      "loss": 1.9337,
+      "step": 6327
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7439477573656978,
+      "learning_rate": 2.7168003226308704e-05,
+      "loss": 1.9413,
+      "step": 6328
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8405441101287547,
+      "learning_rate": 2.7161779987268644e-05,
+      "loss": 2.1336,
+      "step": 6329
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7845992489489523,
+      "learning_rate": 2.715555661325978e-05,
+      "loss": 1.9426,
+      "step": 6330
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.684220392349569,
+      "learning_rate": 2.7149333104670665e-05,
+      "loss": 1.9015,
+      "step": 6331
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7445128853872242,
+      "learning_rate": 2.7143109461889854e-05,
+      "loss": 2.0821,
+      "step": 6332
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6254908015871811,
+      "learning_rate": 2.7136885685305914e-05,
+      "loss": 2.0031,
+      "step": 6333
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7128510934743082,
+      "learning_rate": 2.7130661775307425e-05,
+      "loss": 1.9537,
+      "step": 6334
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5891874934560332,
+      "learning_rate": 2.712443773228296e-05,
+      "loss": 1.9551,
+      "step": 6335
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7148586621068198,
+      "learning_rate": 2.7118213556621124e-05,
+      "loss": 2.0711,
+      "step": 6336
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5901565348664094,
+      "learning_rate": 2.7111989248710517e-05,
+      "loss": 1.899,
+      "step": 6337
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6258220849754685,
+      "learning_rate": 2.710576480893975e-05,
+      "loss": 1.9008,
+      "step": 6338
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6659096296748049,
+      "learning_rate": 2.7099540237697425e-05,
+      "loss": 1.985,
+      "step": 6339
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6057466423014145,
+      "learning_rate": 2.7093315535372177e-05,
+      "loss": 1.9626,
+      "step": 6340
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6249653946376338,
+      "learning_rate": 2.7087090702352645e-05,
+      "loss": 2.1248,
+      "step": 6341
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.671220399695298,
+      "learning_rate": 2.7080865739027466e-05,
+      "loss": 1.9481,
+      "step": 6342
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7340574327747496,
+      "learning_rate": 2.7074640645785294e-05,
+      "loss": 1.9409,
+      "step": 6343
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6511500081887206,
+      "learning_rate": 2.706841542301477e-05,
+      "loss": 2.0593,
+      "step": 6344
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6696786609030081,
+      "learning_rate": 2.7062190071104593e-05,
+      "loss": 1.9721,
+      "step": 6345
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5917858614202364,
+      "learning_rate": 2.7055964590443406e-05,
+      "loss": 1.9453,
+      "step": 6346
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6694715221395168,
+      "learning_rate": 2.704973898141991e-05,
+      "loss": 1.9565,
+      "step": 6347
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6180434968984945,
+      "learning_rate": 2.7043513244422787e-05,
+      "loss": 2.1066,
+      "step": 6348
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7467616967784585,
+      "learning_rate": 2.703728737984073e-05,
+      "loss": 1.9194,
+      "step": 6349
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6380841685937406,
+      "learning_rate": 2.7031061388062462e-05,
+      "loss": 1.9194,
+      "step": 6350
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6718007430123296,
+      "learning_rate": 2.7024835269476683e-05,
+      "loss": 1.9237,
+      "step": 6351
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6836913112075511,
+      "learning_rate": 2.701860902447212e-05,
+      "loss": 1.9175,
+      "step": 6352
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7741449222510344,
+      "learning_rate": 2.7012382653437502e-05,
+      "loss": 2.1073,
+      "step": 6353
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.674702675501228,
+      "learning_rate": 2.700615615676157e-05,
+      "loss": 1.8957,
+      "step": 6354
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.692738056319925,
+      "learning_rate": 2.6999929534833067e-05,
+      "loss": 1.9649,
+      "step": 6355
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7722980354248258,
+      "learning_rate": 2.6993702788040752e-05,
+      "loss": 2.1176,
+      "step": 6356
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6283095319800152,
+      "learning_rate": 2.6987475916773375e-05,
+      "loss": 1.9743,
+      "step": 6357
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7239859593458725,
+      "learning_rate": 2.698124892141971e-05,
+      "loss": 1.9184,
+      "step": 6358
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7990575234945749,
+      "learning_rate": 2.6975021802368543e-05,
+      "loss": 1.9814,
+      "step": 6359
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7371279858209258,
+      "learning_rate": 2.6968794560008652e-05,
+      "loss": 1.9591,
+      "step": 6360
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.57694478099228,
+      "learning_rate": 2.6962567194728817e-05,
+      "loss": 2.1411,
+      "step": 6361
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9142407508577615,
+      "learning_rate": 2.695633970691786e-05,
+      "loss": 1.9122,
+      "step": 6362
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7185005034699699,
+      "learning_rate": 2.6950112096964582e-05,
+      "loss": 1.9562,
+      "step": 6363
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7208891590353855,
+      "learning_rate": 2.694388436525779e-05,
+      "loss": 1.9924,
+      "step": 6364
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7494004827651258,
+      "learning_rate": 2.69376565121863e-05,
+      "loss": 2.1343,
+      "step": 6365
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7569180915849669,
+      "learning_rate": 2.6931428538138965e-05,
+      "loss": 1.9115,
+      "step": 6366
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5876711456789118,
+      "learning_rate": 2.6925200443504612e-05,
+      "loss": 1.9731,
+      "step": 6367
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9403923503396039,
+      "learning_rate": 2.6918972228672083e-05,
+      "loss": 2.131,
+      "step": 6368
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6975016124287862,
+      "learning_rate": 2.691274389403023e-05,
+      "loss": 1.9383,
+      "step": 6369
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6691601448721995,
+      "learning_rate": 2.690651543996793e-05,
+      "loss": 1.9907,
+      "step": 6370
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7262561516712985,
+      "learning_rate": 2.690028686687403e-05,
+      "loss": 1.8903,
+      "step": 6371
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7972537822963063,
+      "learning_rate": 2.689405817513742e-05,
+      "loss": 1.9515,
+      "step": 6372
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6614371799723311,
+      "learning_rate": 2.6887829365146972e-05,
+      "loss": 2.1152,
+      "step": 6373
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7331056166193822,
+      "learning_rate": 2.688160043729158e-05,
+      "loss": 1.9515,
+      "step": 6374
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.640742495327223,
+      "learning_rate": 2.6875371391960154e-05,
+      "loss": 1.856,
+      "step": 6375
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6652746394054031,
+      "learning_rate": 2.6869142229541576e-05,
+      "loss": 1.9937,
+      "step": 6376
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.623689740819728,
+      "learning_rate": 2.6862912950424773e-05,
+      "loss": 2.1254,
+      "step": 6377
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6182803784742932,
+      "learning_rate": 2.6856683554998664e-05,
+      "loss": 1.9425,
+      "step": 6378
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6217050867780594,
+      "learning_rate": 2.6850454043652172e-05,
+      "loss": 1.9413,
+      "step": 6379
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7833390775011795,
+      "learning_rate": 2.6844224416774232e-05,
+      "loss": 1.9165,
+      "step": 6380
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.655663190329253,
+      "learning_rate": 2.683799467475379e-05,
+      "loss": 2.0983,
+      "step": 6381
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8302883528141125,
+      "learning_rate": 2.6831764817979776e-05,
+      "loss": 1.999,
+      "step": 6382
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7625945604520008,
+      "learning_rate": 2.682553484684117e-05,
+      "loss": 1.8862,
+      "step": 6383
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8751855810560286,
+      "learning_rate": 2.681930476172692e-05,
+      "loss": 1.9799,
+      "step": 6384
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7121713981357994,
+      "learning_rate": 2.6813074563026007e-05,
+      "loss": 2.1168,
+      "step": 6385
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.617725812316514,
+      "learning_rate": 2.6806844251127396e-05,
+      "loss": 1.9398,
+      "step": 6386
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8606776568208174,
+      "learning_rate": 2.6800613826420073e-05,
+      "loss": 1.8627,
+      "step": 6387
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6623085162935066,
+      "learning_rate": 2.679438328929304e-05,
+      "loss": 1.9927,
+      "step": 6388
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7345463378154149,
+      "learning_rate": 2.6788152640135284e-05,
+      "loss": 2.1387,
+      "step": 6389
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7439848691363004,
+      "learning_rate": 2.678192187933581e-05,
+      "loss": 1.9252,
+      "step": 6390
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6534642994032598,
+      "learning_rate": 2.6775691007283627e-05,
+      "loss": 1.9147,
+      "step": 6391
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6808793804983158,
+      "learning_rate": 2.676946002436777e-05,
+      "loss": 1.8683,
+      "step": 6392
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6173372498357584,
+      "learning_rate": 2.6763228930977248e-05,
+      "loss": 2.1185,
+      "step": 6393
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.694527177978532,
+      "learning_rate": 2.6756997727501104e-05,
+      "loss": 1.9406,
+      "step": 6394
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7075817026082758,
+      "learning_rate": 2.675076641432837e-05,
+      "loss": 1.9601,
+      "step": 6395
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6227519438643089,
+      "learning_rate": 2.6744534991848098e-05,
+      "loss": 1.8889,
+      "step": 6396
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6963333923752751,
+      "learning_rate": 2.6738303460449342e-05,
+      "loss": 2.122,
+      "step": 6397
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5646072270557386,
+      "learning_rate": 2.673207182052116e-05,
+      "loss": 1.9139,
+      "step": 6398
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6538531963622448,
+      "learning_rate": 2.6725840072452613e-05,
+      "loss": 1.9323,
+      "step": 6399
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5916119778924133,
+      "learning_rate": 2.6719608216632785e-05,
+      "loss": 2.1269,
+      "step": 6400
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5195173999519469,
+      "learning_rate": 2.6713376253450752e-05,
+      "loss": 1.9843,
+      "step": 6401
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6435930025104557,
+      "learning_rate": 2.6707144183295596e-05,
+      "loss": 1.9354,
+      "step": 6402
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5893239821802132,
+      "learning_rate": 2.6700912006556428e-05,
+      "loss": 1.9228,
+      "step": 6403
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6378697042324742,
+      "learning_rate": 2.6694679723622317e-05,
+      "loss": 1.9492,
+      "step": 6404
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6668632156797163,
+      "learning_rate": 2.6688447334882406e-05,
+      "loss": 2.0899,
+      "step": 6405
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5982915319177238,
+      "learning_rate": 2.668221484072579e-05,
+      "loss": 1.9246,
+      "step": 6406
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6175635190154075,
+      "learning_rate": 2.667598224154158e-05,
+      "loss": 1.9874,
+      "step": 6407
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7548753674621945,
+      "learning_rate": 2.6669749537718914e-05,
+      "loss": 1.9697,
+      "step": 6408
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6075771625951945,
+      "learning_rate": 2.666351672964693e-05,
+      "loss": 2.0868,
+      "step": 6409
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7002282873478666,
+      "learning_rate": 2.665728381771477e-05,
+      "loss": 1.8903,
+      "step": 6410
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7007065430102314,
+      "learning_rate": 2.6651050802311562e-05,
+      "loss": 1.8867,
+      "step": 6411
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5632550025659581,
+      "learning_rate": 2.664481768382647e-05,
+      "loss": 1.9242,
+      "step": 6412
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8547363471462214,
+      "learning_rate": 2.6638584462648663e-05,
+      "loss": 2.1747,
+      "step": 6413
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6100679688798435,
+      "learning_rate": 2.6632351139167288e-05,
+      "loss": 1.9371,
+      "step": 6414
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6077719150285603,
+      "learning_rate": 2.6626117713771526e-05,
+      "loss": 1.9165,
+      "step": 6415
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8199742785346006,
+      "learning_rate": 2.661988418685055e-05,
+      "loss": 1.9009,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6128586632987055,
+      "learning_rate": 2.6613650558793563e-05,
+      "loss": 2.0733,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6736447465802785,
+      "learning_rate": 2.6607416829989735e-05,
+      "loss": 1.916,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7715410248714414,
+      "learning_rate": 2.660118300082827e-05,
+      "loss": 1.9957,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5569462191645618,
+      "learning_rate": 2.659494907169837e-05,
+      "loss": 1.9038,
+      "step": 6420
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7890419025935528,
+      "learning_rate": 2.6588715042989253e-05,
+      "loss": 2.0887,
+      "step": 6421
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6375991649496321,
+      "learning_rate": 2.658248091509013e-05,
+      "loss": 1.9304,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6158677615368133,
+      "learning_rate": 2.6576246688390216e-05,
+      "loss": 1.8878,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.608750926854217,
+      "learning_rate": 2.657001236327875e-05,
+      "loss": 1.8975,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7927305185666081,
+      "learning_rate": 2.6563777940144963e-05,
+      "loss": 2.074,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5685464491802472,
+      "learning_rate": 2.6557543419378096e-05,
+      "loss": 2.0006,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7335274472471456,
+      "learning_rate": 2.6551308801367385e-05,
+      "loss": 1.926,
+      "step": 6427
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6472048511415299,
+      "learning_rate": 2.6545074086502108e-05,
+      "loss": 1.8998,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6226244305351178,
+      "learning_rate": 2.6538839275171495e-05,
+      "loss": 2.1072,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6405350237664789,
+      "learning_rate": 2.653260436776483e-05,
+      "loss": 1.9084,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.722162397830998,
+      "learning_rate": 2.6526369364671383e-05,
+      "loss": 1.9867,
+      "step": 6431
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5839888910470223,
+      "learning_rate": 2.6520134266280423e-05,
+      "loss": 1.9984,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9008688104355426,
+      "learning_rate": 2.651389907298123e-05,
+      "loss": 2.1423,
+      "step": 6433
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6555502658788382,
+      "learning_rate": 2.6507663785163102e-05,
+      "loss": 1.8524,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7458811807973122,
+      "learning_rate": 2.6501428403215333e-05,
+      "loss": 1.8861,
+      "step": 6435
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6126057195498361,
+      "learning_rate": 2.6495192927527218e-05,
+      "loss": 1.9225,
+      "step": 6436
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8195722288029695,
+      "learning_rate": 2.648895735848807e-05,
+      "loss": 2.0759,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5526570915836037,
+      "learning_rate": 2.6482721696487194e-05,
+      "loss": 2.0409,
+      "step": 6438
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8074294850016852,
+      "learning_rate": 2.6476485941913907e-05,
+      "loss": 1.9042,
+      "step": 6439
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6485487909687631,
+      "learning_rate": 2.6470250095157544e-05,
+      "loss": 1.952,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8235302213354181,
+      "learning_rate": 2.6464014156607425e-05,
+      "loss": 2.1099,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7040433830711543,
+      "learning_rate": 2.645777812665289e-05,
+      "loss": 1.9142,
+      "step": 6442
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6830980778597556,
+      "learning_rate": 2.6451542005683273e-05,
+      "loss": 1.8876,
+      "step": 6443
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6751410545835068,
+      "learning_rate": 2.6445305794087933e-05,
+      "loss": 1.9794,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8226084228673222,
+      "learning_rate": 2.6439069492256207e-05,
+      "loss": 2.1037,
+      "step": 6445
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7110663717682013,
+      "learning_rate": 2.6432833100577464e-05,
+      "loss": 1.9225,
+      "step": 6446
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7825733965495835,
+      "learning_rate": 2.6426596619441064e-05,
+      "loss": 1.9231,
+      "step": 6447
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6180548781077644,
+      "learning_rate": 2.6420360049236374e-05,
+      "loss": 1.9135,
+      "step": 6448
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6477929110806399,
+      "learning_rate": 2.6414123390352774e-05,
+      "loss": 2.1098,
+      "step": 6449
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6782042156932895,
+      "learning_rate": 2.6407886643179642e-05,
+      "loss": 1.9674,
+      "step": 6450
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6313235860653843,
+      "learning_rate": 2.6401649808106365e-05,
+      "loss": 1.9113,
+      "step": 6451
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.590025222204397,
+      "learning_rate": 2.6395412885522332e-05,
+      "loss": 1.9045,
+      "step": 6452
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7673810216325135,
+      "learning_rate": 2.638917587581694e-05,
+      "loss": 2.1172,
+      "step": 6453
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6403630723685011,
+      "learning_rate": 2.638293877937959e-05,
+      "loss": 1.9157,
+      "step": 6454
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7478088267732901,
+      "learning_rate": 2.6376701596599685e-05,
+      "loss": 1.908,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7785362892239078,
+      "learning_rate": 2.6370464327866646e-05,
+      "loss": 1.8698,
+      "step": 6456
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6430051336674103,
+      "learning_rate": 2.6364226973569893e-05,
+      "loss": 2.1497,
+      "step": 6457
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6146206002096767,
+      "learning_rate": 2.6357989534098848e-05,
+      "loss": 1.9205,
+      "step": 6458
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5705869803621944,
+      "learning_rate": 2.635175200984293e-05,
+      "loss": 1.8827,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7840429222527192,
+      "learning_rate": 2.634551440119159e-05,
+      "loss": 1.9066,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.580990543547921,
+      "learning_rate": 2.633927670853425e-05,
+      "loss": 2.1212,
+      "step": 6461
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9785381031087194,
+      "learning_rate": 2.6333038932260366e-05,
+      "loss": 1.9464,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5972155522336412,
+      "learning_rate": 2.6326801072759388e-05,
+      "loss": 1.9938,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6739338406529702,
+      "learning_rate": 2.6320563130420762e-05,
+      "loss": 1.9367,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7606005480781356,
+      "learning_rate": 2.6314325105633958e-05,
+      "loss": 2.058,
+      "step": 6465
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5444003072458795,
+      "learning_rate": 2.630808699878844e-05,
+      "loss": 1.9059,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.603631708142459,
+      "learning_rate": 2.6301848810273676e-05,
+      "loss": 1.9152,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.682464709504073,
+      "learning_rate": 2.6295610540479144e-05,
+      "loss": 1.9084,
+      "step": 6468
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6001165683683274,
+      "learning_rate": 2.628937218979432e-05,
+      "loss": 2.1897,
+      "step": 6469
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6119697145205543,
+      "learning_rate": 2.6283133758608697e-05,
+      "loss": 1.9305,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6257823626186746,
+      "learning_rate": 2.6276895247311762e-05,
+      "loss": 1.9475,
+      "step": 6471
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.585516300421782,
+      "learning_rate": 2.627065665629301e-05,
+      "loss": 1.9104,
+      "step": 6472
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6904933322458482,
+      "learning_rate": 2.6264417985941934e-05,
+      "loss": 2.1072,
+      "step": 6473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5985068987064216,
+      "learning_rate": 2.625817923664806e-05,
+      "loss": 1.8894,
+      "step": 6474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7306728873087909,
+      "learning_rate": 2.625194040880089e-05,
+      "loss": 1.9422,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6368274552658867,
+      "learning_rate": 2.6245701502789926e-05,
+      "loss": 1.8867,
+      "step": 6476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7380934310298212,
+      "learning_rate": 2.6239462519004708e-05,
+      "loss": 2.1191,
+      "step": 6477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5841557164929541,
+      "learning_rate": 2.6233223457834753e-05,
+      "loss": 1.8899,
+      "step": 6478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6219095969145549,
+      "learning_rate": 2.6226984319669588e-05,
+      "loss": 1.9118,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5612442367264727,
+      "learning_rate": 2.6220745104898747e-05,
+      "loss": 1.9137,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6326602545118899,
+      "learning_rate": 2.6214505813911782e-05,
+      "loss": 2.0018,
+      "step": 6481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5925353628429101,
+      "learning_rate": 2.6208266447098235e-05,
+      "loss": 2.0961,
+      "step": 6482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6412055795370248,
+      "learning_rate": 2.620202700484764e-05,
+      "loss": 1.9081,
+      "step": 6483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6140144795272754,
+      "learning_rate": 2.6195787487549554e-05,
+      "loss": 1.9592,
+      "step": 6484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6572373121609558,
+      "learning_rate": 2.6189547895593562e-05,
+      "loss": 2.1652,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5594062648790008,
+      "learning_rate": 2.6183308229369196e-05,
+      "loss": 1.9392,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6535715238490272,
+      "learning_rate": 2.6177068489266037e-05,
+      "loss": 1.8974,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5662751974351232,
+      "learning_rate": 2.617082867567366e-05,
+      "loss": 1.9877,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6078156268936326,
+      "learning_rate": 2.6164588788981636e-05,
+      "loss": 2.0873,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6188123406915048,
+      "learning_rate": 2.615834882957955e-05,
+      "loss": 1.9034,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6312126786743251,
+      "learning_rate": 2.6152108797856988e-05,
+      "loss": 1.9074,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6345859343140237,
+      "learning_rate": 2.614586869420354e-05,
+      "loss": 1.9321,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7595484134926646,
+      "learning_rate": 2.6139628519008806e-05,
+      "loss": 2.1029,
+      "step": 6493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8093661214668302,
+      "learning_rate": 2.613338827266238e-05,
+      "loss": 1.9802,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6021368062009931,
+      "learning_rate": 2.6127147955553864e-05,
+      "loss": 1.8901,
+      "step": 6495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7603385245092367,
+      "learning_rate": 2.6120907568072883e-05,
+      "loss": 1.9472,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7216827404475619,
+      "learning_rate": 2.611466711060902e-05,
+      "loss": 2.1154,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5904495242187202,
+      "learning_rate": 2.610842658355192e-05,
+      "loss": 1.8947,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6135900713172538,
+      "learning_rate": 2.6102185987291193e-05,
+      "loss": 1.922,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7206665211888598,
+      "learning_rate": 2.6095945322216475e-05,
+      "loss": 1.9861,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6640372432298424,
+      "learning_rate": 2.6089704588717375e-05,
+      "loss": 2.1029,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5726830092544704,
+      "learning_rate": 2.6083463787183545e-05,
+      "loss": 1.9174,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6900662103599454,
+      "learning_rate": 2.607722291800463e-05,
+      "loss": 1.8905,
+      "step": 6503
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7248149977913002,
+      "learning_rate": 2.6070981981570253e-05,
+      "loss": 1.9053,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6124037329195494,
+      "learning_rate": 2.6064740978270068e-05,
+      "loss": 2.0932,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6154079137857783,
+      "learning_rate": 2.6058499908493734e-05,
+      "loss": 1.95,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6410602571612857,
+      "learning_rate": 2.6052258772630904e-05,
+      "loss": 1.9309,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7614619568929409,
+      "learning_rate": 2.6046017571071233e-05,
+      "loss": 1.914,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6217193523042852,
+      "learning_rate": 2.6039776304204394e-05,
+      "loss": 2.0801,
+      "step": 6509
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7819015384910641,
+      "learning_rate": 2.603353497242004e-05,
+      "loss": 1.9156,
+      "step": 6510
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6537187431996537,
+      "learning_rate": 2.6027293576107858e-05,
+      "loss": 1.9082,
+      "step": 6511
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.673756057435738,
+      "learning_rate": 2.6021052115657513e-05,
+      "loss": 1.9758,
+      "step": 6512
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6239100848813969,
+      "learning_rate": 2.6014810591458693e-05,
+      "loss": 1.969,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.706847355813064,
+      "learning_rate": 2.600856900390108e-05,
+      "loss": 2.1096,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5976473344697637,
+      "learning_rate": 2.600232735337436e-05,
+      "loss": 1.9286,
+      "step": 6515
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5737124371633711,
+      "learning_rate": 2.5996085640268226e-05,
+      "loss": 1.9315,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.674491351773786,
+      "learning_rate": 2.5989843864972367e-05,
+      "loss": 2.1327,
+      "step": 6517
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6371537861890869,
+      "learning_rate": 2.5983602027876497e-05,
+      "loss": 1.9665,
+      "step": 6518
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5739627246603651,
+      "learning_rate": 2.597736012937031e-05,
+      "loss": 2.0252,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5989448315254908,
+      "learning_rate": 2.597111816984351e-05,
+      "loss": 1.917,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7805313688530142,
+      "learning_rate": 2.596487614968583e-05,
+      "loss": 2.1378,
+      "step": 6521
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.556115292494162,
+      "learning_rate": 2.595863406928695e-05,
+      "loss": 1.9427,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6606468995493125,
+      "learning_rate": 2.5952391929036607e-05,
+      "loss": 1.9071,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7276127053848916,
+      "learning_rate": 2.5946149729324532e-05,
+      "loss": 1.9441,
+      "step": 6524
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.647926321420105,
+      "learning_rate": 2.593990747054044e-05,
+      "loss": 2.0367,
+      "step": 6525
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6235778308925776,
+      "learning_rate": 2.5933665153074066e-05,
+      "loss": 2.0931,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6940199105293988,
+      "learning_rate": 2.592742277731513e-05,
+      "loss": 1.877,
+      "step": 6527
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6561189324662845,
+      "learning_rate": 2.5921180343653397e-05,
+      "loss": 1.8998,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7855837172417871,
+      "learning_rate": 2.5914937852478582e-05,
+      "loss": 2.1151,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6618184785480731,
+      "learning_rate": 2.5908695304180436e-05,
+      "loss": 1.9466,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6232953111634054,
+      "learning_rate": 2.5902452699148706e-05,
+      "loss": 1.9658,
+      "step": 6531
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7454729600205101,
+      "learning_rate": 2.5896210037773155e-05,
+      "loss": 1.9026,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7197768050331184,
+      "learning_rate": 2.5889967320443525e-05,
+      "loss": 2.0729,
+      "step": 6533
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6301825914804862,
+      "learning_rate": 2.588372454754958e-05,
+      "loss": 1.9216,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7410699171847606,
+      "learning_rate": 2.5877481719481084e-05,
+      "loss": 1.961,
+      "step": 6535
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7172321039022076,
+      "learning_rate": 2.58712388366278e-05,
+      "loss": 1.9479,
+      "step": 6536
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6191187844892279,
+      "learning_rate": 2.5864995899379492e-05,
+      "loss": 1.9541,
+      "step": 6537
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8812027701116729,
+      "learning_rate": 2.5858752908125933e-05,
+      "loss": 2.143,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6271000850811852,
+      "learning_rate": 2.5852509863256906e-05,
+      "loss": 1.9163,
+      "step": 6539
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7495511646401828,
+      "learning_rate": 2.5846266765162185e-05,
+      "loss": 1.9088,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7451757222132465,
+      "learning_rate": 2.5840023614231555e-05,
+      "loss": 2.1284,
+      "step": 6541
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6579809278641219,
+      "learning_rate": 2.5833780410854803e-05,
+      "loss": 1.8942,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7037405966466191,
+      "learning_rate": 2.5827537155421717e-05,
+      "loss": 1.9404,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6230426557387939,
+      "learning_rate": 2.5821293848322074e-05,
+      "loss": 2.0121,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6133644587188964,
+      "learning_rate": 2.5815050489945693e-05,
+      "loss": 1.8939,
+      "step": 6545
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6627209813478885,
+      "learning_rate": 2.5808807080682357e-05,
+      "loss": 2.1186,
+      "step": 6546
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.670813289271015,
+      "learning_rate": 2.5802563620921877e-05,
+      "loss": 1.8857,
+      "step": 6547
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6815405987066258,
+      "learning_rate": 2.5796320111054046e-05,
+      "loss": 1.93,
+      "step": 6548
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6400182904428378,
+      "learning_rate": 2.5790076551468678e-05,
+      "loss": 2.113,
+      "step": 6549
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.696414317097647,
+      "learning_rate": 2.5783832942555597e-05,
+      "loss": 2.0038,
+      "step": 6550
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6315532102106972,
+      "learning_rate": 2.57775892847046e-05,
+      "loss": 1.8825,
+      "step": 6551
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6095253865456688,
+      "learning_rate": 2.5771345578305505e-05,
+      "loss": 1.8986,
+      "step": 6552
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.754273026621178,
+      "learning_rate": 2.5765101823748138e-05,
+      "loss": 2.0933,
+      "step": 6553
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6101022920436412,
+      "learning_rate": 2.5758858021422327e-05,
+      "loss": 1.9192,
+      "step": 6554
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6922359524671611,
+      "learning_rate": 2.5752614171717888e-05,
+      "loss": 1.9294,
+      "step": 6555
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7249867819013628,
+      "learning_rate": 2.574637027502465e-05,
+      "loss": 1.988,
+      "step": 6556
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6196852222361716,
+      "learning_rate": 2.5740126331732457e-05,
+      "loss": 1.9272,
+      "step": 6557
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7469761063330268,
+      "learning_rate": 2.573388234223113e-05,
+      "loss": 2.0865,
+      "step": 6558
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5995948496574428,
+      "learning_rate": 2.5727638306910518e-05,
+      "loss": 1.9225,
+      "step": 6559
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6704497077651614,
+      "learning_rate": 2.5721394226160454e-05,
+      "loss": 1.8708,
+      "step": 6560
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6901500235670213,
+      "learning_rate": 2.571515010037079e-05,
+      "loss": 2.1018,
+      "step": 6561
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7004204854742906,
+      "learning_rate": 2.5708905929931365e-05,
+      "loss": 1.9743,
+      "step": 6562
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6347871873770429,
+      "learning_rate": 2.570266171523203e-05,
+      "loss": 1.9534,
+      "step": 6563
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6473977143340702,
+      "learning_rate": 2.5696417456662635e-05,
+      "loss": 1.936,
+      "step": 6564
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6020022940889188,
+      "learning_rate": 2.569017315461304e-05,
+      "loss": 2.1082,
+      "step": 6565
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6402247237128365,
+      "learning_rate": 2.5683928809473097e-05,
+      "loss": 1.9557,
+      "step": 6566
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5779581137798265,
+      "learning_rate": 2.567768442163267e-05,
+      "loss": 1.9866,
+      "step": 6567
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5404170883335784,
+      "learning_rate": 2.5671439991481626e-05,
+      "loss": 2.0246,
+      "step": 6568
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6021764084011956,
+      "learning_rate": 2.5665195519409825e-05,
+      "loss": 1.9269,
+      "step": 6569
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.572015219090514,
+      "learning_rate": 2.5658951005807124e-05,
+      "loss": 2.134,
+      "step": 6570
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5833732534412206,
+      "learning_rate": 2.5652706451063414e-05,
+      "loss": 1.8916,
+      "step": 6571
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6407664972173585,
+      "learning_rate": 2.564646185556856e-05,
+      "loss": 1.8819,
+      "step": 6572
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5724012674693831,
+      "learning_rate": 2.564021721971243e-05,
+      "loss": 2.0751,
+      "step": 6573
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6070858081393523,
+      "learning_rate": 2.5633972543884903e-05,
+      "loss": 1.9252,
+      "step": 6574
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5840496408222126,
+      "learning_rate": 2.5627727828475884e-05,
+      "loss": 1.9422,
+      "step": 6575
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5945039173567086,
+      "learning_rate": 2.562148307387523e-05,
+      "loss": 1.8583,
+      "step": 6576
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7302556369481313,
+      "learning_rate": 2.561523828047283e-05,
+      "loss": 1.9206,
+      "step": 6577
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7939308568202998,
+      "learning_rate": 2.5608993448658576e-05,
+      "loss": 2.1211,
+      "step": 6578
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6155024660609106,
+      "learning_rate": 2.5602748578822366e-05,
+      "loss": 1.9103,
+      "step": 6579
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7240653221384331,
+      "learning_rate": 2.5596503671354082e-05,
+      "loss": 1.9608,
+      "step": 6580
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5702597143346358,
+      "learning_rate": 2.5590258726643622e-05,
+      "loss": 1.9685,
+      "step": 6581
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5627715759863336,
+      "learning_rate": 2.5584013745080887e-05,
+      "loss": 2.1114,
+      "step": 6582
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6022562582895445,
+      "learning_rate": 2.557776872705578e-05,
+      "loss": 1.9566,
+      "step": 6583
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6007677309871938,
+      "learning_rate": 2.5571523672958187e-05,
+      "loss": 1.9237,
+      "step": 6584
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.612852549201217,
+      "learning_rate": 2.556527858317803e-05,
+      "loss": 2.1438,
+      "step": 6585
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.578082825537637,
+      "learning_rate": 2.5559033458105214e-05,
+      "loss": 1.8709,
+      "step": 6586
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6058746397140969,
+      "learning_rate": 2.555278829812963e-05,
+      "loss": 2.0134,
+      "step": 6587
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6107176893918208,
+      "learning_rate": 2.5546543103641208e-05,
+      "loss": 1.9255,
+      "step": 6588
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5478592703328194,
+      "learning_rate": 2.5540297875029855e-05,
+      "loss": 1.8468,
+      "step": 6589
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6146878771812208,
+      "learning_rate": 2.553405261268549e-05,
+      "loss": 2.117,
+      "step": 6590
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5628835116745478,
+      "learning_rate": 2.5527807316998014e-05,
+      "loss": 1.95,
+      "step": 6591
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6261919377483969,
+      "learning_rate": 2.552156198835737e-05,
+      "loss": 1.9386,
+      "step": 6592
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5894541127077549,
+      "learning_rate": 2.5515316627153467e-05,
+      "loss": 1.9644,
+      "step": 6593
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5803564882349125,
+      "learning_rate": 2.550907123377623e-05,
+      "loss": 2.1085,
+      "step": 6594
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5893129725344164,
+      "learning_rate": 2.5502825808615576e-05,
+      "loss": 1.9014,
+      "step": 6595
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5996772034919101,
+      "learning_rate": 2.549658035206145e-05,
+      "loss": 1.9233,
+      "step": 6596
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5835587734110113,
+      "learning_rate": 2.549033486450378e-05,
+      "loss": 1.9002,
+      "step": 6597
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.622206582894593,
+      "learning_rate": 2.548408934633248e-05,
+      "loss": 2.1586,
+      "step": 6598
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.630864665239744,
+      "learning_rate": 2.5477843797937484e-05,
+      "loss": 1.9711,
+      "step": 6599
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6289054898510056,
+      "learning_rate": 2.5471598219708752e-05,
+      "loss": 1.9334,
+      "step": 6600
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7711055284334014,
+      "learning_rate": 2.5465352612036203e-05,
+      "loss": 1.9544,
+      "step": 6601
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5941108726458566,
+      "learning_rate": 2.5459106975309776e-05,
+      "loss": 2.0669,
+      "step": 6602
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7933195619145668,
+      "learning_rate": 2.5452861309919413e-05,
+      "loss": 1.9245,
+      "step": 6603
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6616488977692362,
+      "learning_rate": 2.544661561625506e-05,
+      "loss": 1.9095,
+      "step": 6604
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6973954469667004,
+      "learning_rate": 2.544036989470666e-05,
+      "loss": 2.1797,
+      "step": 6605
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6452638391147235,
+      "learning_rate": 2.5434124145664158e-05,
+      "loss": 1.9915,
+      "step": 6606
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6125737213034402,
+      "learning_rate": 2.54278783695175e-05,
+      "loss": 1.928,
+      "step": 6607
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7839168365955907,
+      "learning_rate": 2.542163256665664e-05,
+      "loss": 1.925,
+      "step": 6608
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6268008301437035,
+      "learning_rate": 2.5415386737471525e-05,
+      "loss": 1.8918,
+      "step": 6609
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6244913844101304,
+      "learning_rate": 2.540914088235211e-05,
+      "loss": 2.1603,
+      "step": 6610
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7305445559108353,
+      "learning_rate": 2.5402895001688347e-05,
+      "loss": 1.8815,
+      "step": 6611
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6097771757402122,
+      "learning_rate": 2.5396649095870202e-05,
+      "loss": 2.0172,
+      "step": 6612
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7104598836472195,
+      "learning_rate": 2.539040316528762e-05,
+      "loss": 1.9346,
+      "step": 6613
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8201450661432044,
+      "learning_rate": 2.5384157210330562e-05,
+      "loss": 2.1178,
+      "step": 6614
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5908416191008313,
+      "learning_rate": 2.5377911231388994e-05,
+      "loss": 1.8939,
+      "step": 6615
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7295479850326737,
+      "learning_rate": 2.537166522885287e-05,
+      "loss": 1.9063,
+      "step": 6616
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7544398311537737,
+      "learning_rate": 2.5365419203112163e-05,
+      "loss": 2.0982,
+      "step": 6617
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6296044960054024,
+      "learning_rate": 2.5359173154556837e-05,
+      "loss": 1.9909,
+      "step": 6618
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7362914480592476,
+      "learning_rate": 2.5352927083576856e-05,
+      "loss": 1.88,
+      "step": 6619
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6252893834786406,
+      "learning_rate": 2.5346680990562183e-05,
+      "loss": 1.9465,
+      "step": 6620
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6992760315702686,
+      "learning_rate": 2.5340434875902786e-05,
+      "loss": 1.8827,
+      "step": 6621
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7032932056196257,
+      "learning_rate": 2.533418873998865e-05,
+      "loss": 2.0916,
+      "step": 6622
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5785093266873672,
+      "learning_rate": 2.5327942583209737e-05,
+      "loss": 1.9074,
+      "step": 6623
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7826355115178765,
+      "learning_rate": 2.5321696405956018e-05,
+      "loss": 1.905,
+      "step": 6624
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5951128670746993,
+      "learning_rate": 2.5315450208617474e-05,
+      "loss": 1.9331,
+      "step": 6625
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6259712750432452,
+      "learning_rate": 2.5309203991584073e-05,
+      "loss": 2.111,
+      "step": 6626
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6766416194774147,
+      "learning_rate": 2.5302957755245805e-05,
+      "loss": 1.9084,
+      "step": 6627
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6037285168651996,
+      "learning_rate": 2.529671149999263e-05,
+      "loss": 1.8978,
+      "step": 6628
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6127325765893574,
+      "learning_rate": 2.5290465226214543e-05,
+      "loss": 1.9681,
+      "step": 6629
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5841792177438028,
+      "learning_rate": 2.528421893430152e-05,
+      "loss": 2.1269,
+      "step": 6630
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.617397095474975,
+      "learning_rate": 2.5277972624643537e-05,
+      "loss": 1.908,
+      "step": 6631
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5633587189688742,
+      "learning_rate": 2.527172629763058e-05,
+      "loss": 1.9798,
+      "step": 6632
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.589941329595494,
+      "learning_rate": 2.5265479953652637e-05,
+      "loss": 1.921,
+      "step": 6633
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6327104751330486,
+      "learning_rate": 2.525923359309969e-05,
+      "loss": 2.1278,
+      "step": 6634
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6186814560941952,
+      "learning_rate": 2.5252987216361723e-05,
+      "loss": 1.8606,
+      "step": 6635
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6145968434655722,
+      "learning_rate": 2.524674082382873e-05,
+      "loss": 1.925,
+      "step": 6636
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6630123417458124,
+      "learning_rate": 2.524049441589069e-05,
+      "loss": 1.9979,
+      "step": 6637
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6087926754554818,
+      "learning_rate": 2.5234247992937593e-05,
+      "loss": 2.1213,
+      "step": 6638
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5742631285193094,
+      "learning_rate": 2.522800155535943e-05,
+      "loss": 1.9055,
+      "step": 6639
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6937187167504592,
+      "learning_rate": 2.5221755103546207e-05,
+      "loss": 1.8946,
+      "step": 6640
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6016097139590426,
+      "learning_rate": 2.521550863788789e-05,
+      "loss": 1.965,
+      "step": 6641
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6677489183552128,
+      "learning_rate": 2.520926215877448e-05,
+      "loss": 2.0933,
+      "step": 6642
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5908826423794707,
+      "learning_rate": 2.520301566659598e-05,
+      "loss": 2.0006,
+      "step": 6643
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.737560666984083,
+      "learning_rate": 2.5196769161742386e-05,
+      "loss": 1.9149,
+      "step": 6644
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5887517383948331,
+      "learning_rate": 2.5190522644603676e-05,
+      "loss": 1.9481,
+      "step": 6645
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7762551666231263,
+      "learning_rate": 2.5184276115569843e-05,
+      "loss": 2.1162,
+      "step": 6646
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6401774345065644,
+      "learning_rate": 2.5178029575030916e-05,
+      "loss": 1.902,
+      "step": 6647
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8045769911011456,
+      "learning_rate": 2.517178302337686e-05,
+      "loss": 1.888,
+      "step": 6648
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5811641553877283,
+      "learning_rate": 2.5165536460997685e-05,
+      "loss": 1.9934,
+      "step": 6649
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6901640646059801,
+      "learning_rate": 2.5159289888283392e-05,
+      "loss": 2.0762,
+      "step": 6650
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6641364706304923,
+      "learning_rate": 2.5153043305623974e-05,
+      "loss": 1.9632,
+      "step": 6651
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6830171689028038,
+      "learning_rate": 2.5146796713409438e-05,
+      "loss": 1.9316,
+      "step": 6652
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6258924788664454,
+      "learning_rate": 2.5140550112029775e-05,
+      "loss": 1.8769,
+      "step": 6653
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6910472998453835,
+      "learning_rate": 2.513430350187499e-05,
+      "loss": 2.1065,
+      "step": 6654
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6023365616298708,
+      "learning_rate": 2.512805688333509e-05,
+      "loss": 1.9598,
+      "step": 6655
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.651133693258318,
+      "learning_rate": 2.5121810256800072e-05,
+      "loss": 1.9025,
+      "step": 6656
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6834977825180181,
+      "learning_rate": 2.5115563622659936e-05,
+      "loss": 1.9227,
+      "step": 6657
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5868590545569323,
+      "learning_rate": 2.51093169813047e-05,
+      "loss": 2.0791,
+      "step": 6658
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6011948248089714,
+      "learning_rate": 2.510307033312434e-05,
+      "loss": 1.9041,
+      "step": 6659
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.799201478210623,
+      "learning_rate": 2.509682367850888e-05,
+      "loss": 1.8954,
+      "step": 6660
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6106814700613309,
+      "learning_rate": 2.5090577017848323e-05,
+      "loss": 1.9026,
+      "step": 6661
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.9408401486680976,
+      "learning_rate": 2.5084330351532675e-05,
+      "loss": 2.1758,
+      "step": 6662
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6554253349014847,
+      "learning_rate": 2.5078083679951924e-05,
+      "loss": 1.9211,
+      "step": 6663
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7121920533366128,
+      "learning_rate": 2.5071837003496095e-05,
+      "loss": 1.9488,
+      "step": 6664
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7799823614125067,
+      "learning_rate": 2.5065590322555193e-05,
+      "loss": 1.9208,
+      "step": 6665
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6308743232210926,
+      "learning_rate": 2.5059343637519212e-05,
+      "loss": 2.1013,
+      "step": 6666
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7979534779466928,
+      "learning_rate": 2.505309694877816e-05,
+      "loss": 1.8896,
+      "step": 6667
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6173264794306369,
+      "learning_rate": 2.5046850256722048e-05,
+      "loss": 2.0149,
+      "step": 6668
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7864081946453376,
+      "learning_rate": 2.5040603561740894e-05,
+      "loss": 1.899,
+      "step": 6669
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7364723312186022,
+      "learning_rate": 2.5034356864224685e-05,
+      "loss": 2.0625,
+      "step": 6670
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8247596131665619,
+      "learning_rate": 2.5028110164563434e-05,
+      "loss": 1.9049,
+      "step": 6671
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7132997313923205,
+      "learning_rate": 2.502186346314715e-05,
+      "loss": 1.9216,
+      "step": 6672
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7463588765872892,
+      "learning_rate": 2.501561676036584e-05,
+      "loss": 1.9391,
+      "step": 6673
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6538608084078793,
+      "learning_rate": 2.5009370056609512e-05,
+      "loss": 2.1831,
+      "step": 6674
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6643066015080465,
+      "learning_rate": 2.5003123352268176e-05,
+      "loss": 1.885,
+      "step": 6675
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.720157166336245,
+      "learning_rate": 2.4996876647731833e-05,
+      "loss": 1.9004,
+      "step": 6676
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.612813983232992,
+      "learning_rate": 2.4990629943390497e-05,
+      "loss": 1.9212,
+      "step": 6677
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5979299110423234,
+      "learning_rate": 2.4984383239634167e-05,
+      "loss": 2.103,
+      "step": 6678
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6832136148110365,
+      "learning_rate": 2.4978136536852856e-05,
+      "loss": 1.8912,
+      "step": 6679
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5999895238823745,
+      "learning_rate": 2.497188983543658e-05,
+      "loss": 1.9868,
+      "step": 6680
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7620609205589449,
+      "learning_rate": 2.4965643135775324e-05,
+      "loss": 1.9181,
+      "step": 6681
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6423405015893627,
+      "learning_rate": 2.4959396438259115e-05,
+      "loss": 2.0946,
+      "step": 6682
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7936071310518991,
+      "learning_rate": 2.4953149743277958e-05,
+      "loss": 1.894,
+      "step": 6683
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7077760201526929,
+      "learning_rate": 2.4946903051221842e-05,
+      "loss": 1.8507,
+      "step": 6684
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5887798889227074,
+      "learning_rate": 2.49406563624808e-05,
+      "loss": 1.8935,
+      "step": 6685
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8497011918958953,
+      "learning_rate": 2.4934409677444813e-05,
+      "loss": 2.1291,
+      "step": 6686
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6640881353257697,
+      "learning_rate": 2.4928162996503904e-05,
+      "loss": 1.9164,
+      "step": 6687
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8178929498874834,
+      "learning_rate": 2.4921916320048082e-05,
+      "loss": 1.9089,
+      "step": 6688
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6573935696544478,
+      "learning_rate": 2.491566964846733e-05,
+      "loss": 1.8432,
+      "step": 6689
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.0749758311940638,
+      "learning_rate": 2.4909422982151683e-05,
+      "loss": 2.1485,
+      "step": 6690
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.633142134461561,
+      "learning_rate": 2.4903176321491125e-05,
+      "loss": 1.9045,
+      "step": 6691
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8586981657256701,
+      "learning_rate": 2.489692966687566e-05,
+      "loss": 1.8607,
+      "step": 6692
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6127538538250511,
+      "learning_rate": 2.4890683018695317e-05,
+      "loss": 1.9529,
+      "step": 6693
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8780725795540756,
+      "learning_rate": 2.488443637734007e-05,
+      "loss": 2.0996,
+      "step": 6694
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5397061852810898,
+      "learning_rate": 2.4878189743199927e-05,
+      "loss": 1.8956,
+      "step": 6695
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7186891442921187,
+      "learning_rate": 2.4871943116664917e-05,
+      "loss": 1.9034,
+      "step": 6696
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6146453114973145,
+      "learning_rate": 2.4865696498125013e-05,
+      "loss": 1.9031,
+      "step": 6697
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6217876282561405,
+      "learning_rate": 2.4859449887970238e-05,
+      "loss": 2.1317,
+      "step": 6698
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7026887031648654,
+      "learning_rate": 2.485320328659057e-05,
+      "loss": 1.9936,
+      "step": 6699
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5953854447743928,
+      "learning_rate": 2.4846956694376025e-05,
+      "loss": 1.9078,
+      "step": 6700
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6630793745386788,
+      "learning_rate": 2.4840710111716614e-05,
+      "loss": 1.9119,
+      "step": 6701
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7620197279551758,
+      "learning_rate": 2.4834463539002317e-05,
+      "loss": 2.1157,
+      "step": 6702
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6243289771808985,
+      "learning_rate": 2.482821697662315e-05,
+      "loss": 1.9319,
+      "step": 6703
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6316703398832386,
+      "learning_rate": 2.482197042496909e-05,
+      "loss": 1.972,
+      "step": 6704
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.678801402402233,
+      "learning_rate": 2.4815723884430153e-05,
+      "loss": 1.9919,
+      "step": 6705
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5825680057129189,
+      "learning_rate": 2.4809477355396336e-05,
+      "loss": 2.1045,
+      "step": 6706
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7844047551011293,
+      "learning_rate": 2.480323083825762e-05,
+      "loss": 1.9317,
+      "step": 6707
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6614241627814321,
+      "learning_rate": 2.479698433340402e-05,
+      "loss": 1.9111,
+      "step": 6708
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6151569102807789,
+      "learning_rate": 2.4790737841225524e-05,
+      "loss": 1.9354,
+      "step": 6709
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6900031948324293,
+      "learning_rate": 2.4784491362112113e-05,
+      "loss": 2.1168,
+      "step": 6710
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6581494566532976,
+      "learning_rate": 2.4778244896453803e-05,
+      "loss": 1.9439,
+      "step": 6711
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6657633475316107,
+      "learning_rate": 2.477199844464057e-05,
+      "loss": 1.9274,
+      "step": 6712
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7144507489745362,
+      "learning_rate": 2.4765752007062406e-05,
+      "loss": 1.9003,
+      "step": 6713
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7623985367728991,
+      "learning_rate": 2.475950558410932e-05,
+      "loss": 2.1121,
+      "step": 6714
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6632496185060474,
+      "learning_rate": 2.475325917617128e-05,
+      "loss": 1.907,
+      "step": 6715
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.740477192041616,
+      "learning_rate": 2.4747012783638272e-05,
+      "loss": 1.9347,
+      "step": 6716
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5760196357184919,
+      "learning_rate": 2.4740766406900315e-05,
+      "loss": 1.9901,
+      "step": 6717
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7993674826330882,
+      "learning_rate": 2.4734520046347362e-05,
+      "loss": 2.097,
+      "step": 6718
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6563414546606893,
+      "learning_rate": 2.472827370236942e-05,
+      "loss": 1.9282,
+      "step": 6719
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7020165258798277,
+      "learning_rate": 2.472202737535647e-05,
+      "loss": 1.9044,
+      "step": 6720
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5795772846152659,
+      "learning_rate": 2.471578106569848e-05,
+      "loss": 1.888,
+      "step": 6721
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7765358359830016,
+      "learning_rate": 2.4709534773785462e-05,
+      "loss": 2.1495,
+      "step": 6722
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5982578700613664,
+      "learning_rate": 2.4703288500007372e-05,
+      "loss": 1.8962,
+      "step": 6723
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7118588956260865,
+      "learning_rate": 2.4697042244754208e-05,
+      "loss": 1.9662,
+      "step": 6724
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.607385491131942,
+      "learning_rate": 2.469079600841593e-05,
+      "loss": 1.9547,
+      "step": 6725
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7996836038041695,
+      "learning_rate": 2.468454979138253e-05,
+      "loss": 2.0931,
+      "step": 6726
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.593839911260235,
+      "learning_rate": 2.467830359404399e-05,
+      "loss": 1.8912,
+      "step": 6727
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7344143099607419,
+      "learning_rate": 2.4672057416790266e-05,
+      "loss": 1.8788,
+      "step": 6728
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6128024914626605,
+      "learning_rate": 2.4665811260011353e-05,
+      "loss": 1.8832,
+      "step": 6729
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5686391140724573,
+      "learning_rate": 2.465956512409722e-05,
+      "loss": 1.956,
+      "step": 6730
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6762861090534112,
+      "learning_rate": 2.4653319009437823e-05,
+      "loss": 2.108,
+      "step": 6731
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.680451203074685,
+      "learning_rate": 2.4647072916423157e-05,
+      "loss": 1.9085,
+      "step": 6732
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.713221164139241,
+      "learning_rate": 2.4640826845443165e-05,
+      "loss": 1.9304,
+      "step": 6733
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7218507434423468,
+      "learning_rate": 2.463458079688784e-05,
+      "loss": 2.1379,
+      "step": 6734
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6526674507286963,
+      "learning_rate": 2.4628334771147136e-05,
+      "loss": 1.933,
+      "step": 6735
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6464778420717437,
+      "learning_rate": 2.4622088768611008e-05,
+      "loss": 1.9781,
+      "step": 6736
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5949194859757655,
+      "learning_rate": 2.4615842789669447e-05,
+      "loss": 1.8958,
+      "step": 6737
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5796289542634331,
+      "learning_rate": 2.460959683471239e-05,
+      "loss": 2.1069,
+      "step": 6738
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6106510587974017,
+      "learning_rate": 2.46033509041298e-05,
+      "loss": 1.9695,
+      "step": 6739
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.624575271276694,
+      "learning_rate": 2.4597104998311656e-05,
+      "loss": 1.8893,
+      "step": 6740
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5724035341380189,
+      "learning_rate": 2.4590859117647892e-05,
+      "loss": 1.9249,
+      "step": 6741
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6213111489395294,
+      "learning_rate": 2.4584613262528474e-05,
+      "loss": 1.9701,
+      "step": 6742
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.697271578999779,
+      "learning_rate": 2.4578367433343365e-05,
+      "loss": 2.1436,
+      "step": 6743
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6243794315928739,
+      "learning_rate": 2.4572121630482505e-05,
+      "loss": 1.9727,
+      "step": 6744
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5818349899542608,
+      "learning_rate": 2.4565875854335855e-05,
+      "loss": 1.9395,
+      "step": 6745
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6990770382030523,
+      "learning_rate": 2.4559630105293348e-05,
+      "loss": 2.1298,
+      "step": 6746
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6459763055133859,
+      "learning_rate": 2.4553384383744947e-05,
+      "loss": 1.9348,
+      "step": 6747
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5520473712682026,
+      "learning_rate": 2.45471386900806e-05,
+      "loss": 1.9527,
+      "step": 6748
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6405644222285571,
+      "learning_rate": 2.4540893024690233e-05,
+      "loss": 1.8977,
+      "step": 6749
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6213141383741682,
+      "learning_rate": 2.453464738796381e-05,
+      "loss": 2.1142,
+      "step": 6750
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5806709769443219,
+      "learning_rate": 2.452840178029125e-05,
+      "loss": 1.8977,
+      "step": 6751
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7474948406269726,
+      "learning_rate": 2.4522156202062515e-05,
+      "loss": 1.9109,
+      "step": 6752
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.677271788432471,
+      "learning_rate": 2.4515910653667534e-05,
+      "loss": 1.9066,
+      "step": 6753
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5992337099234126,
+      "learning_rate": 2.450966513549623e-05,
+      "loss": 2.1152,
+      "step": 6754
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5830395139034389,
+      "learning_rate": 2.450341964793855e-05,
+      "loss": 1.9954,
+      "step": 6755
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5840318406664341,
+      "learning_rate": 2.4497174191384427e-05,
+      "loss": 1.9175,
+      "step": 6756
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5804305668201363,
+      "learning_rate": 2.4490928766223774e-05,
+      "loss": 1.9423,
+      "step": 6757
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7197838169407231,
+      "learning_rate": 2.4484683372846542e-05,
+      "loss": 2.0937,
+      "step": 6758
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6172439288259548,
+      "learning_rate": 2.4478438011642637e-05,
+      "loss": 1.8963,
+      "step": 6759
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6235788729125963,
+      "learning_rate": 2.4472192683001985e-05,
+      "loss": 1.9157,
+      "step": 6760
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5959534785565576,
+      "learning_rate": 2.4465947387314524e-05,
+      "loss": 1.9853,
+      "step": 6761
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5975694025758582,
+      "learning_rate": 2.445970212497015e-05,
+      "loss": 1.9189,
+      "step": 6762
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5889693467279564,
+      "learning_rate": 2.445345689635879e-05,
+      "loss": 2.0962,
+      "step": 6763
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.638321035293485,
+      "learning_rate": 2.4447211701870377e-05,
+      "loss": 1.9231,
+      "step": 6764
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5869589860641043,
+      "learning_rate": 2.4440966541894792e-05,
+      "loss": 1.8795,
+      "step": 6765
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5699530902625876,
+      "learning_rate": 2.4434721416821975e-05,
+      "loss": 2.1158,
+      "step": 6766
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6050436652707758,
+      "learning_rate": 2.4428476327041816e-05,
+      "loss": 2.0187,
+      "step": 6767
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6708975022240512,
+      "learning_rate": 2.4422231272944224e-05,
+      "loss": 1.9343,
+      "step": 6768
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5464048787942786,
+      "learning_rate": 2.4415986254919115e-05,
+      "loss": 1.9244,
+      "step": 6769
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6976577689042746,
+      "learning_rate": 2.440974127335638e-05,
+      "loss": 2.1208,
+      "step": 6770
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5730240615629292,
+      "learning_rate": 2.4403496328645927e-05,
+      "loss": 1.9276,
+      "step": 6771
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7571714232426864,
+      "learning_rate": 2.439725142117764e-05,
+      "loss": 1.9251,
+      "step": 6772
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6625868087940432,
+      "learning_rate": 2.4391006551341427e-05,
+      "loss": 1.9773,
+      "step": 6773
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.63218111458179,
+      "learning_rate": 2.438476171952718e-05,
+      "loss": 1.8817,
+      "step": 6774
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8624729973375317,
+      "learning_rate": 2.437851692612478e-05,
+      "loss": 2.1295,
+      "step": 6775
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6057129145136635,
+      "learning_rate": 2.437227217152412e-05,
+      "loss": 1.8881,
+      "step": 6776
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6593773451083194,
+      "learning_rate": 2.43660274561151e-05,
+      "loss": 1.9353,
+      "step": 6777
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7473076565520184,
+      "learning_rate": 2.4359782780287575e-05,
+      "loss": 2.0973,
+      "step": 6778
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6276330959639643,
+      "learning_rate": 2.4353538144431454e-05,
+      "loss": 1.9284,
+      "step": 6779
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7225193005370535,
+      "learning_rate": 2.4347293548936595e-05,
+      "loss": 1.9298,
+      "step": 6780
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6884143929539044,
+      "learning_rate": 2.4341048994192878e-05,
+      "loss": 1.9273,
+      "step": 6781
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.669166626157464,
+      "learning_rate": 2.4334804480590188e-05,
+      "loss": 2.066,
+      "step": 6782
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7284814876054923,
+      "learning_rate": 2.4328560008518377e-05,
+      "loss": 1.9085,
+      "step": 6783
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6239209916842718,
+      "learning_rate": 2.4322315578367334e-05,
+      "loss": 1.9425,
+      "step": 6784
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7803236080098342,
+      "learning_rate": 2.4316071190526905e-05,
+      "loss": 1.8679,
+      "step": 6785
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6369172920398191,
+      "learning_rate": 2.430982684538696e-05,
+      "loss": 1.9699,
+      "step": 6786
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7015151308620332,
+      "learning_rate": 2.430358254333737e-05,
+      "loss": 2.1318,
+      "step": 6787
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6460169474328119,
+      "learning_rate": 2.429733828476798e-05,
+      "loss": 1.8822,
+      "step": 6788
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5749543660504861,
+      "learning_rate": 2.4291094070068638e-05,
+      "loss": 1.8478,
+      "step": 6789
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6489338329207572,
+      "learning_rate": 2.4284849899629217e-05,
+      "loss": 2.0708,
+      "step": 6790
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6296951755367514,
+      "learning_rate": 2.4278605773839548e-05,
+      "loss": 1.9237,
+      "step": 6791
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.563245463226585,
+      "learning_rate": 2.4272361693089495e-05,
+      "loss": 1.9968,
+      "step": 6792
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7218330509470475,
+      "learning_rate": 2.4266117657768874e-05,
+      "loss": 1.9514,
+      "step": 6793
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6303823179613774,
+      "learning_rate": 2.4259873668267553e-05,
+      "loss": 1.8751,
+      "step": 6794
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7076920535276987,
+      "learning_rate": 2.425362972497536e-05,
+      "loss": 2.0869,
+      "step": 6795
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.656480926864414,
+      "learning_rate": 2.424738582828212e-05,
+      "loss": 1.963,
+      "step": 6796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7363724995566108,
+      "learning_rate": 2.4241141978577675e-05,
+      "loss": 1.9566,
+      "step": 6797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7277395544747648,
+      "learning_rate": 2.4234898176251868e-05,
+      "loss": 1.9801,
+      "step": 6798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8379841473739562,
+      "learning_rate": 2.42286544216945e-05,
+      "loss": 2.0948,
+      "step": 6799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6659577058511705,
+      "learning_rate": 2.422241071529541e-05,
+      "loss": 1.9077,
+      "step": 6800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7714648438478618,
+      "learning_rate": 2.421616705744441e-05,
+      "loss": 1.9729,
+      "step": 6801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6628822201225683,
+      "learning_rate": 2.4209923448531318e-05,
+      "loss": 2.0946,
+      "step": 6802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5952311084057027,
+      "learning_rate": 2.4203679888945963e-05,
+      "loss": 1.8796,
+      "step": 6803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6836393055619138,
+      "learning_rate": 2.419743637907813e-05,
+      "loss": 2.0108,
+      "step": 6804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6903539165802698,
+      "learning_rate": 2.4191192919317653e-05,
+      "loss": 1.9067,
+      "step": 6805
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.708099156768583,
+      "learning_rate": 2.4184949510054313e-05,
+      "loss": 1.9122,
+      "step": 6806
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6796728385968701,
+      "learning_rate": 2.4178706151677925e-05,
+      "loss": 2.0893,
+      "step": 6807
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6327854860876525,
+      "learning_rate": 2.4172462844578295e-05,
+      "loss": 1.9177,
+      "step": 6808
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7414510937443434,
+      "learning_rate": 2.4166219589145202e-05,
+      "loss": 1.9004,
+      "step": 6809
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7781712396548345,
+      "learning_rate": 2.4159976385768444e-05,
+      "loss": 2.1063,
+      "step": 6810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.621843997687271,
+      "learning_rate": 2.4153733234837817e-05,
+      "loss": 1.9824,
+      "step": 6811
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7435361891021265,
+      "learning_rate": 2.4147490136743096e-05,
+      "loss": 1.8615,
+      "step": 6812
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7920600782553857,
+      "learning_rate": 2.414124709187408e-05,
+      "loss": 1.9069,
+      "step": 6813
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6051091967503096,
+      "learning_rate": 2.4135004100620517e-05,
+      "loss": 1.9507,
+      "step": 6814
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6672953505824236,
+      "learning_rate": 2.4128761163372203e-05,
+      "loss": 2.1116,
+      "step": 6815
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6394488405411185,
+      "learning_rate": 2.4122518280518922e-05,
+      "loss": 1.9096,
+      "step": 6816
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7029695181896743,
+      "learning_rate": 2.4116275452450424e-05,
+      "loss": 1.9282,
+      "step": 6817
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6386608439110977,
+      "learning_rate": 2.4110032679556484e-05,
+      "loss": 1.916,
+      "step": 6818
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7052065251058428,
+      "learning_rate": 2.410378996222685e-05,
+      "loss": 2.089,
+      "step": 6819
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6749873781333043,
+      "learning_rate": 2.4097547300851296e-05,
+      "loss": 1.8826,
+      "step": 6820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7167518747102501,
+      "learning_rate": 2.4091304695819574e-05,
+      "loss": 1.9472,
+      "step": 6821
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7432231883065988,
+      "learning_rate": 2.4085062147521427e-05,
+      "loss": 2.1274,
+      "step": 6822
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6918260628988899,
+      "learning_rate": 2.407881965634661e-05,
+      "loss": 1.9907,
+      "step": 6823
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6266301248228702,
+      "learning_rate": 2.407257722268487e-05,
+      "loss": 1.8851,
+      "step": 6824
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7717103865401205,
+      "learning_rate": 2.406633484692594e-05,
+      "loss": 1.9052,
+      "step": 6825
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6047759068097547,
+      "learning_rate": 2.406009252945957e-05,
+      "loss": 1.9096,
+      "step": 6826
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8291367883720913,
+      "learning_rate": 2.4053850270675474e-05,
+      "loss": 2.1058,
+      "step": 6827
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5644723587796809,
+      "learning_rate": 2.404760807096339e-05,
+      "loss": 1.9254,
+      "step": 6828
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5899525154037856,
+      "learning_rate": 2.4041365930713058e-05,
+      "loss": 1.9809,
+      "step": 6829
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6335833926549232,
+      "learning_rate": 2.403512385031418e-05,
+      "loss": 1.896,
+      "step": 6830
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.651372802921379,
+      "learning_rate": 2.4028881830156484e-05,
+      "loss": 2.1591,
+      "step": 6831
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.588525335985015,
+      "learning_rate": 2.4022639870629695e-05,
+      "loss": 1.8998,
+      "step": 6832
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6097101569219551,
+      "learning_rate": 2.4016397972123502e-05,
+      "loss": 1.9111,
+      "step": 6833
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6704457411346033,
+      "learning_rate": 2.4010156135027635e-05,
+      "loss": 2.0914,
+      "step": 6834
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6255031233854543,
+      "learning_rate": 2.4003914359731783e-05,
+      "loss": 1.9935,
+      "step": 6835
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6501068350777695,
+      "learning_rate": 2.3997672646625642e-05,
+      "loss": 1.9106,
+      "step": 6836
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5436276998753237,
+      "learning_rate": 2.3991430996098926e-05,
+      "loss": 1.9294,
+      "step": 6837
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6567220748704277,
+      "learning_rate": 2.3985189408541313e-05,
+      "loss": 1.9128,
+      "step": 6838
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7579279803622189,
+      "learning_rate": 2.3978947884342496e-05,
+      "loss": 2.0822,
+      "step": 6839
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6192621446964727,
+      "learning_rate": 2.397270642389215e-05,
+      "loss": 1.9266,
+      "step": 6840
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6473291489568552,
+      "learning_rate": 2.3966465027579965e-05,
+      "loss": 1.8916,
+      "step": 6841
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.680980680998492,
+      "learning_rate": 2.396022369579562e-05,
+      "loss": 1.9915,
+      "step": 6842
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5652075683804121,
+      "learning_rate": 2.395398242892877e-05,
+      "loss": 2.0981,
+      "step": 6843
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6204969797342692,
+      "learning_rate": 2.39477412273691e-05,
+      "loss": 1.8843,
+      "step": 6844
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.624807351666636,
+      "learning_rate": 2.3941500091506275e-05,
+      "loss": 1.9323,
+      "step": 6845
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6369086123194133,
+      "learning_rate": 2.3935259021729935e-05,
+      "loss": 1.9658,
+      "step": 6846
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6217138948400651,
+      "learning_rate": 2.392901801842976e-05,
+      "loss": 2.0679,
+      "step": 6847
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5476398496472505,
+      "learning_rate": 2.392277708199538e-05,
+      "loss": 1.9317,
+      "step": 6848
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6066945188728551,
+      "learning_rate": 2.3916536212816454e-05,
+      "loss": 1.9177,
+      "step": 6849
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5873595244939671,
+      "learning_rate": 2.391029541128263e-05,
+      "loss": 1.9272,
+      "step": 6850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7177781426609902,
+      "learning_rate": 2.3904054677783534e-05,
+      "loss": 2.1173,
+      "step": 6851
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5705803263904187,
+      "learning_rate": 2.3897814012708813e-05,
+      "loss": 1.9098,
+      "step": 6852
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6393458847388561,
+      "learning_rate": 2.3891573416448085e-05,
+      "loss": 1.9201,
+      "step": 6853
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6531558168703689,
+      "learning_rate": 2.388533288939098e-05,
+      "loss": 1.9883,
+      "step": 6854
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6441874429943146,
+      "learning_rate": 2.387909243192713e-05,
+      "loss": 2.0952,
+      "step": 6855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.609066281396595,
+      "learning_rate": 2.3872852044446138e-05,
+      "loss": 1.9278,
+      "step": 6856
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6546182544626906,
+      "learning_rate": 2.386661172733762e-05,
+      "loss": 1.8682,
+      "step": 6857
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.675446999757002,
+      "learning_rate": 2.38603714809912e-05,
+      "loss": 1.8778,
+      "step": 6858
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7444320340379087,
+      "learning_rate": 2.3854131305796462e-05,
+      "loss": 2.1021,
+      "step": 6859
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6819584812800328,
+      "learning_rate": 2.384789120214302e-05,
+      "loss": 2.0283,
+      "step": 6860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5771843788204984,
+      "learning_rate": 2.3841651170420452e-05,
+      "loss": 1.8946,
+      "step": 6861
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6172188110496265,
+      "learning_rate": 2.3835411211018363e-05,
+      "loss": 1.9105,
+      "step": 6862
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7448204919284072,
+      "learning_rate": 2.382917132432635e-05,
+      "loss": 2.0971,
+      "step": 6863
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5821705947384628,
+      "learning_rate": 2.3822931510733965e-05,
+      "loss": 1.9144,
+      "step": 6864
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6005340323705847,
+      "learning_rate": 2.3816691770630803e-05,
+      "loss": 1.8606,
+      "step": 6865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6484878491223819,
+      "learning_rate": 2.3810452104406444e-05,
+      "loss": 2.0203,
+      "step": 6866
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6096836462339968,
+      "learning_rate": 2.380421251245044e-05,
+      "loss": 2.1376,
+      "step": 6867
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7009150752552232,
+      "learning_rate": 2.379797299515237e-05,
+      "loss": 1.9604,
+      "step": 6868
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6553904931408051,
+      "learning_rate": 2.3791733552901774e-05,
+      "loss": 1.9431,
+      "step": 6869
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6193090366345624,
+      "learning_rate": 2.3785494186088217e-05,
+      "loss": 1.9141,
+      "step": 6870
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7736903595860248,
+      "learning_rate": 2.3779254895101256e-05,
+      "loss": 2.0722,
+      "step": 6871
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6561150774563063,
+      "learning_rate": 2.3773015680330414e-05,
+      "loss": 1.8629,
+      "step": 6872
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5182705139781527,
+      "learning_rate": 2.3766776542165256e-05,
+      "loss": 1.904,
+      "step": 6873
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5810549487202489,
+      "learning_rate": 2.3760537480995298e-05,
+      "loss": 1.9271,
+      "step": 6874
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8188873168414694,
+      "learning_rate": 2.3754298497210073e-05,
+      "loss": 2.063,
+      "step": 6875
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.702023092199348,
+      "learning_rate": 2.374805959119912e-05,
+      "loss": 1.8776,
+      "step": 6876
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5803541173771632,
+      "learning_rate": 2.3741820763351945e-05,
+      "loss": 1.9236,
+      "step": 6877
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6470537838943383,
+      "learning_rate": 2.373558201405806e-05,
+      "loss": 1.8677,
+      "step": 6878
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6631028917874696,
+      "learning_rate": 2.3729343343707e-05,
+      "loss": 2.1301,
+      "step": 6879
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.623893806627512,
+      "learning_rate": 2.372310475268824e-05,
+      "loss": 1.9193,
+      "step": 6880
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6154738471162933,
+      "learning_rate": 2.3716866241391312e-05,
+      "loss": 1.9054,
+      "step": 6881
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6468615388599988,
+      "learning_rate": 2.3710627810205686e-05,
+      "loss": 1.9558,
+      "step": 6882
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6929877096400356,
+      "learning_rate": 2.3704389459520855e-05,
+      "loss": 2.1175,
+      "step": 6883
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5970163384843282,
+      "learning_rate": 2.369815118972633e-05,
+      "loss": 1.9196,
+      "step": 6884
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6720238952797764,
+      "learning_rate": 2.3691913001211567e-05,
+      "loss": 1.9358,
+      "step": 6885
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6089547532220582,
+      "learning_rate": 2.368567489436605e-05,
+      "loss": 1.8918,
+      "step": 6886
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6229818684048509,
+      "learning_rate": 2.367943686957924e-05,
+      "loss": 2.0991,
+      "step": 6887
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.570459583142368,
+      "learning_rate": 2.3673198927240618e-05,
+      "loss": 1.9391,
+      "step": 6888
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5588454685117317,
+      "learning_rate": 2.3666961067739646e-05,
+      "loss": 1.9468,
+      "step": 6889
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6496758721614757,
+      "learning_rate": 2.3660723291465757e-05,
+      "loss": 1.8884,
+      "step": 6890
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5668348741268975,
+      "learning_rate": 2.365448559880842e-05,
+      "loss": 2.1318,
+      "step": 6891
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6481880115062101,
+      "learning_rate": 2.3648247990157077e-05,
+      "loss": 1.9129,
+      "step": 6892
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5671130049193563,
+      "learning_rate": 2.3642010465901158e-05,
+      "loss": 1.9233,
+      "step": 6893
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.565254955019476,
+      "learning_rate": 2.3635773026430112e-05,
+      "loss": 1.8941,
+      "step": 6894
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5617922894977659,
+      "learning_rate": 2.362953567213336e-05,
+      "loss": 2.0552,
+      "step": 6895
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5900120968178542,
+      "learning_rate": 2.3623298403400318e-05,
+      "loss": 1.9113,
+      "step": 6896
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.59758674281974,
+      "learning_rate": 2.3617061220620422e-05,
+      "loss": 1.9938,
+      "step": 6897
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6684784597662042,
+      "learning_rate": 2.3610824124183063e-05,
+      "loss": 1.929,
+      "step": 6898
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6054468772221763,
+      "learning_rate": 2.3604587114477677e-05,
+      "loss": 2.0654,
+      "step": 6899
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5964270755876917,
+      "learning_rate": 2.359835019189364e-05,
+      "loss": 1.884,
+      "step": 6900
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5766920747884321,
+      "learning_rate": 2.3592113356820357e-05,
+      "loss": 1.9003,
+      "step": 6901
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6625545172774548,
+      "learning_rate": 2.358587660964723e-05,
+      "loss": 1.8528,
+      "step": 6902
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5957701778591085,
+      "learning_rate": 2.3579639950763628e-05,
+      "loss": 2.1228,
+      "step": 6903
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5332229546551666,
+      "learning_rate": 2.3573403380558935e-05,
+      "loss": 2.0034,
+      "step": 6904
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5887189021077798,
+      "learning_rate": 2.3567166899422542e-05,
+      "loss": 1.9305,
+      "step": 6905
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.688491201664447,
+      "learning_rate": 2.35609305077438e-05,
+      "loss": 1.8917,
+      "step": 6906
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6404379052876008,
+      "learning_rate": 2.355469420591208e-05,
+      "loss": 2.0586,
+      "step": 6907
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6316202744321762,
+      "learning_rate": 2.354845799431673e-05,
+      "loss": 1.94,
+      "step": 6908
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6118239124633812,
+      "learning_rate": 2.3542221873347117e-05,
+      "loss": 1.9468,
+      "step": 6909
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.604040834599595,
+      "learning_rate": 2.3535985843392585e-05,
+      "loss": 1.9769,
+      "step": 6910
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7044546934138206,
+      "learning_rate": 2.352974990484246e-05,
+      "loss": 2.1114,
+      "step": 6911
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.589162926996751,
+      "learning_rate": 2.3523514058086092e-05,
+      "loss": 1.9311,
+      "step": 6912
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7182601365819111,
+      "learning_rate": 2.3517278303512812e-05,
+      "loss": 1.8462,
+      "step": 6913
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6005610685268739,
+      "learning_rate": 2.3511042641511936e-05,
+      "loss": 1.9243,
+      "step": 6914
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7061502814093771,
+      "learning_rate": 2.350480707247279e-05,
+      "loss": 2.1309,
+      "step": 6915
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5499280755367931,
+      "learning_rate": 2.3498571596784672e-05,
+      "loss": 1.9845,
+      "step": 6916
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8461499592735158,
+      "learning_rate": 2.34923362148369e-05,
+      "loss": 1.9292,
+      "step": 6917
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.630147152218096,
+      "learning_rate": 2.348610092701878e-05,
+      "loss": 1.9001,
+      "step": 6918
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7115403702532992,
+      "learning_rate": 2.3479865733719583e-05,
+      "loss": 2.1299,
+      "step": 6919
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6065194402136416,
+      "learning_rate": 2.3473630635328626e-05,
+      "loss": 1.8717,
+      "step": 6920
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6420221356455456,
+      "learning_rate": 2.3467395632235175e-05,
+      "loss": 1.8935,
+      "step": 6921
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6782267090405564,
+      "learning_rate": 2.3461160724828504e-05,
+      "loss": 2.0023,
+      "step": 6922
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6184186246641804,
+      "learning_rate": 2.34549259134979e-05,
+      "loss": 2.1346,
+      "step": 6923
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5995745473603085,
+      "learning_rate": 2.3448691198632618e-05,
+      "loss": 1.9167,
+      "step": 6924
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5724627745614244,
+      "learning_rate": 2.3442456580621906e-05,
+      "loss": 1.903,
+      "step": 6925
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6059457914412485,
+      "learning_rate": 2.3436222059855046e-05,
+      "loss": 1.9007,
+      "step": 6926
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7008904390475549,
+      "learning_rate": 2.3429987636721255e-05,
+      "loss": 2.094,
+      "step": 6927
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5477791583777565,
+      "learning_rate": 2.3423753311609793e-05,
+      "loss": 1.945,
+      "step": 6928
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5938890089036918,
+      "learning_rate": 2.341751908490988e-05,
+      "loss": 1.9185,
+      "step": 6929
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6770447220216039,
+      "learning_rate": 2.3411284957010746e-05,
+      "loss": 1.91,
+      "step": 6930
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.70596531483098,
+      "learning_rate": 2.3405050928301632e-05,
+      "loss": 2.1294,
+      "step": 6931
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5728492471548148,
+      "learning_rate": 2.3398816999171734e-05,
+      "loss": 1.8909,
+      "step": 6932
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6840785627932982,
+      "learning_rate": 2.3392583170010277e-05,
+      "loss": 1.8891,
+      "step": 6933
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6181292946772861,
+      "learning_rate": 2.3386349441206442e-05,
+      "loss": 1.939,
+      "step": 6934
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7004103747028102,
+      "learning_rate": 2.3380115813149452e-05,
+      "loss": 2.1441,
+      "step": 6935
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.623852175591801,
+      "learning_rate": 2.3373882286228487e-05,
+      "loss": 1.9123,
+      "step": 6936
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6033455416480228,
+      "learning_rate": 2.336764886083272e-05,
+      "loss": 1.9239,
+      "step": 6937
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6598874445654936,
+      "learning_rate": 2.3361415537351343e-05,
+      "loss": 1.9185,
+      "step": 6938
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7243329797556521,
+      "learning_rate": 2.3355182316173534e-05,
+      "loss": 2.172,
+      "step": 6939
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6220763789740399,
+      "learning_rate": 2.334894919768844e-05,
+      "loss": 1.9007,
+      "step": 6940
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6732834833874951,
+      "learning_rate": 2.3342716182285244e-05,
+      "loss": 1.9679,
+      "step": 6941
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6290743538021611,
+      "learning_rate": 2.3336483270353075e-05,
+      "loss": 1.896,
+      "step": 6942
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7647600318389801,
+      "learning_rate": 2.3330250462281085e-05,
+      "loss": 2.1093,
+      "step": 6943
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6296311758101999,
+      "learning_rate": 2.332401775845843e-05,
+      "loss": 1.9479,
+      "step": 6944
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6859822628177811,
+      "learning_rate": 2.331778515927422e-05,
+      "loss": 1.9131,
+      "step": 6945
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5937622532003612,
+      "learning_rate": 2.3311552665117596e-05,
+      "loss": 1.8953,
+      "step": 6946
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.62087894693453,
+      "learning_rate": 2.3305320276377686e-05,
+      "loss": 1.9931,
+      "step": 6947
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6519940319026273,
+      "learning_rate": 2.3299087993443578e-05,
+      "loss": 2.0583,
+      "step": 6948
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5664911436855777,
+      "learning_rate": 2.3292855816704406e-05,
+      "loss": 1.8998,
+      "step": 6949
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5829259214555926,
+      "learning_rate": 2.3286623746549254e-05,
+      "loss": 1.9355,
+      "step": 6950
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6418516789200658,
+      "learning_rate": 2.3280391783367214e-05,
+      "loss": 2.1351,
+      "step": 6951
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7138007940635156,
+      "learning_rate": 2.3274159927547392e-05,
+      "loss": 1.9246,
+      "step": 6952
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6026687616839503,
+      "learning_rate": 2.326792817947885e-05,
+      "loss": 2.0096,
+      "step": 6953
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6200187077054666,
+      "learning_rate": 2.326169653955067e-05,
+      "loss": 1.8937,
+      "step": 6954
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6451318910676073,
+      "learning_rate": 2.3255465008151905e-05,
+      "loss": 2.1026,
+      "step": 6955
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6039961510056435,
+      "learning_rate": 2.3249233585671636e-05,
+      "loss": 1.9181,
+      "step": 6956
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6301588727294065,
+      "learning_rate": 2.3243002272498908e-05,
+      "loss": 1.8861,
+      "step": 6957
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6535534668817495,
+      "learning_rate": 2.3236771069022755e-05,
+      "loss": 1.9266,
+      "step": 6958
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6153450633121307,
+      "learning_rate": 2.3230539975632238e-05,
+      "loss": 2.0735,
+      "step": 6959
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5817948886325971,
+      "learning_rate": 2.3224308992716382e-05,
+      "loss": 1.9895,
+      "step": 6960
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6647028274507037,
+      "learning_rate": 2.3218078120664197e-05,
+      "loss": 1.8726,
+      "step": 6961
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6404021436884167,
+      "learning_rate": 2.321184735986473e-05,
+      "loss": 1.9051,
+      "step": 6962
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6104217841963283,
+      "learning_rate": 2.3205616710706966e-05,
+      "loss": 2.0636,
+      "step": 6963
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.562373572926197,
+      "learning_rate": 2.3199386173579926e-05,
+      "loss": 1.8685,
+      "step": 6964
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6511361838470012,
+      "learning_rate": 2.319315574887261e-05,
+      "loss": 1.8733,
+      "step": 6965
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5410075843839882,
+      "learning_rate": 2.3186925436973996e-05,
+      "loss": 1.9435,
+      "step": 6966
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6898610281706871,
+      "learning_rate": 2.318069523827308e-05,
+      "loss": 2.0948,
+      "step": 6967
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6079141679002845,
+      "learning_rate": 2.3174465153158835e-05,
+      "loss": 1.8928,
+      "step": 6968
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6755281245097672,
+      "learning_rate": 2.316823518202022e-05,
+      "loss": 1.8831,
+      "step": 6969
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6964386660556257,
+      "learning_rate": 2.316200532524622e-05,
+      "loss": 1.9361,
+      "step": 6970
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7251950405762728,
+      "learning_rate": 2.3155775583225774e-05,
+      "loss": 2.1257,
+      "step": 6971
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6855173412700594,
+      "learning_rate": 2.314954595634783e-05,
+      "loss": 2.0141,
+      "step": 6972
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6617032673510361,
+      "learning_rate": 2.3143316445001342e-05,
+      "loss": 1.9213,
+      "step": 6973
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7249613973160681,
+      "learning_rate": 2.3137087049575233e-05,
+      "loss": 1.9375,
+      "step": 6974
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8622200660367813,
+      "learning_rate": 2.3130857770458436e-05,
+      "loss": 2.0947,
+      "step": 6975
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5616372763328307,
+      "learning_rate": 2.3124628608039856e-05,
+      "loss": 1.9241,
+      "step": 6976
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.813247333180486,
+      "learning_rate": 2.3118399562708415e-05,
+      "loss": 1.8971,
+      "step": 6977
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7622497027040896,
+      "learning_rate": 2.3112170634853034e-05,
+      "loss": 1.9777,
+      "step": 6978
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6114024234879807,
+      "learning_rate": 2.3105941824862586e-05,
+      "loss": 1.8715,
+      "step": 6979
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7345018359457328,
+      "learning_rate": 2.3099713133125967e-05,
+      "loss": 2.1165,
+      "step": 6980
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6750225725094887,
+      "learning_rate": 2.3093484560032074e-05,
+      "loss": 1.8935,
+      "step": 6981
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6722410139024572,
+      "learning_rate": 2.308725610596977e-05,
+      "loss": 1.9261,
+      "step": 6982
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6002796526862085,
+      "learning_rate": 2.308102777132793e-05,
+      "loss": 2.0829,
+      "step": 6983
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6908741450115808,
+      "learning_rate": 2.3074799556495393e-05,
+      "loss": 2.0061,
+      "step": 6984
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6850103079609436,
+      "learning_rate": 2.3068571461861037e-05,
+      "loss": 1.8672,
+      "step": 6985
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.591922588698335,
+      "learning_rate": 2.3062343487813707e-05,
+      "loss": 1.9733,
+      "step": 6986
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6652737791434807,
+      "learning_rate": 2.305611563474222e-05,
+      "loss": 2.1129,
+      "step": 6987
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5924844956271506,
+      "learning_rate": 2.3049887903035434e-05,
+      "loss": 1.9452,
+      "step": 6988
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5684996174977273,
+      "learning_rate": 2.3043660293082146e-05,
+      "loss": 1.8864,
+      "step": 6989
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.617460985069477,
+      "learning_rate": 2.3037432805271182e-05,
+      "loss": 1.916,
+      "step": 6990
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5281311389438791,
+      "learning_rate": 2.303120543999136e-05,
+      "loss": 2.0102,
+      "step": 6991
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6979070499430653,
+      "learning_rate": 2.3024978197631462e-05,
+      "loss": 2.1233,
+      "step": 6992
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5409055517401112,
+      "learning_rate": 2.3018751078580287e-05,
+      "loss": 1.9143,
+      "step": 6993
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6920508027160223,
+      "learning_rate": 2.3012524083226634e-05,
+      "loss": 1.8603,
+      "step": 6994
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5625641840138551,
+      "learning_rate": 2.3006297211959253e-05,
+      "loss": 2.1046,
+      "step": 6995
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6173929899220647,
+      "learning_rate": 2.300007046516694e-05,
+      "loss": 1.9481,
+      "step": 6996
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5926224062531321,
+      "learning_rate": 2.2993843843238434e-05,
+      "loss": 1.9576,
+      "step": 6997
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6107784547323151,
+      "learning_rate": 2.2987617346562497e-05,
+      "loss": 1.9127,
+      "step": 6998
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.579781604217849,
+      "learning_rate": 2.2981390975527885e-05,
+      "loss": 2.1035,
+      "step": 6999
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5959741326920936,
+      "learning_rate": 2.2975164730523323e-05,
+      "loss": 1.9177,
+      "step": 7000
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.593494398983973,
+      "learning_rate": 2.296893861193755e-05,
+      "loss": 1.8908,
+      "step": 7001
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6607170372594442,
+      "learning_rate": 2.296271262015927e-05,
+      "loss": 1.8771,
+      "step": 7002
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5588231769927112,
+      "learning_rate": 2.2956486755577222e-05,
+      "loss": 1.9588,
+      "step": 7003
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.648435541585525,
+      "learning_rate": 2.29502610185801e-05,
+      "loss": 2.0974,
+      "step": 7004
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6408984872079808,
+      "learning_rate": 2.2944035409556596e-05,
+      "loss": 1.8749,
+      "step": 7005
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6234841430035625,
+      "learning_rate": 2.2937809928895413e-05,
+      "loss": 1.8781,
+      "step": 7006
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7036612319436399,
+      "learning_rate": 2.2931584576985232e-05,
+      "loss": 2.1122,
+      "step": 7007
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5788380251171683,
+      "learning_rate": 2.292535935421471e-05,
+      "loss": 1.9241,
+      "step": 7008
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6810023589386787,
+      "learning_rate": 2.291913426097254e-05,
+      "loss": 1.9958,
+      "step": 7009
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6200881174158608,
+      "learning_rate": 2.2912909297647357e-05,
+      "loss": 1.9272,
+      "step": 7010
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6523238988010632,
+      "learning_rate": 2.2906684464627826e-05,
+      "loss": 1.868,
+      "step": 7011
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6205149522966484,
+      "learning_rate": 2.2900459762302588e-05,
+      "loss": 2.1478,
+      "step": 7012
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5813464064364479,
+      "learning_rate": 2.289423519106026e-05,
+      "loss": 1.9218,
+      "step": 7013
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5636650836602358,
+      "learning_rate": 2.2888010751289485e-05,
+      "loss": 1.8626,
+      "step": 7014
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6417591896562348,
+      "learning_rate": 2.288178644337888e-05,
+      "loss": 1.9584,
+      "step": 7015
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6052769187449286,
+      "learning_rate": 2.287556226771704e-05,
+      "loss": 2.0555,
+      "step": 7016
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.591835350630733,
+      "learning_rate": 2.2869338224692584e-05,
+      "loss": 1.9158,
+      "step": 7017
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6124122539614082,
+      "learning_rate": 2.286311431469409e-05,
+      "loss": 1.9145,
+      "step": 7018
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6667524110642084,
+      "learning_rate": 2.285689053811015e-05,
+      "loss": 2.1128,
+      "step": 7019
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6081311829630692,
+      "learning_rate": 2.285066689532934e-05,
+      "loss": 1.853,
+      "step": 7020
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.687556115632128,
+      "learning_rate": 2.2844443386740223e-05,
+      "loss": 1.8619,
+      "step": 7021
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6382971623618422,
+      "learning_rate": 2.2838220012731365e-05,
+      "loss": 1.9921,
+      "step": 7022
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6057546403634597,
+      "learning_rate": 2.2831996773691302e-05,
+      "loss": 1.9165,
+      "step": 7023
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6067936404980967,
+      "learning_rate": 2.2825773670008594e-05,
+      "loss": 2.0945,
+      "step": 7024
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.663462780835987,
+      "learning_rate": 2.281955070207177e-05,
+      "loss": 1.8995,
+      "step": 7025
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6777704425548444,
+      "learning_rate": 2.2813327870269342e-05,
+      "loss": 1.9194,
+      "step": 7026
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7056129903038216,
+      "learning_rate": 2.2807105174989842e-05,
+      "loss": 2.0746,
+      "step": 7027
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7074882555796836,
+      "learning_rate": 2.2800882616621782e-05,
+      "loss": 1.9873,
+      "step": 7028
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7463059703195273,
+      "learning_rate": 2.279466019555365e-05,
+      "loss": 1.8292,
+      "step": 7029
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5826551625192009,
+      "learning_rate": 2.2788437912173953e-05,
+      "loss": 1.9047,
+      "step": 7030
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.751733655989344,
+      "learning_rate": 2.2782215766871143e-05,
+      "loss": 1.9323,
+      "step": 7031
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6846844569600771,
+      "learning_rate": 2.2775993760033723e-05,
+      "loss": 2.0615,
+      "step": 7032
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6789548700333394,
+      "learning_rate": 2.2769771892050154e-05,
+      "loss": 1.9029,
+      "step": 7033
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5713224070040431,
+      "learning_rate": 2.2763550163308877e-05,
+      "loss": 1.9606,
+      "step": 7034
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6912832095467348,
+      "learning_rate": 2.2757328574198363e-05,
+      "loss": 1.9541,
+      "step": 7035
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6331018500034579,
+      "learning_rate": 2.275110712510703e-05,
+      "loss": 2.1113,
+      "step": 7036
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7698986496254807,
+      "learning_rate": 2.2744885816423313e-05,
+      "loss": 1.9137,
+      "step": 7037
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6602287729423949,
+      "learning_rate": 2.2738664648535653e-05,
+      "loss": 1.9441,
+      "step": 7038
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8384485578352099,
+      "learning_rate": 2.273244362183244e-05,
+      "loss": 2.1119,
+      "step": 7039
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6644004986678486,
+      "learning_rate": 2.2726222736702084e-05,
+      "loss": 1.9533,
+      "step": 7040
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6269602940696372,
+      "learning_rate": 2.2720001993532993e-05,
+      "loss": 1.8806,
+      "step": 7041
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9113031923531637,
+      "learning_rate": 2.2713781392713535e-05,
+      "loss": 1.9083,
+      "step": 7042
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.674585077090671,
+      "learning_rate": 2.2707560934632105e-05,
+      "loss": 1.8596,
+      "step": 7043
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8963966075763483,
+      "learning_rate": 2.270134061967706e-05,
+      "loss": 2.1252,
+      "step": 7044
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5903137312976403,
+      "learning_rate": 2.2695120448236758e-05,
+      "loss": 1.9081,
+      "step": 7045
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8005989408765137,
+      "learning_rate": 2.2688900420699566e-05,
+      "loss": 1.9593,
+      "step": 7046
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6172194273735723,
+      "learning_rate": 2.2682680537453807e-05,
+      "loss": 1.8885,
+      "step": 7047
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.650572484361134,
+      "learning_rate": 2.2676460798887828e-05,
+      "loss": 2.1261,
+      "step": 7048
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5911149005308879,
+      "learning_rate": 2.2670241205389936e-05,
+      "loss": 1.9152,
+      "step": 7049
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6391682859057229,
+      "learning_rate": 2.2664021757348465e-05,
+      "loss": 1.9219,
+      "step": 7050
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6184611343089382,
+      "learning_rate": 2.2657802455151715e-05,
+      "loss": 2.0442,
+      "step": 7051
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.617959307286655,
+      "learning_rate": 2.2651583299187967e-05,
+      "loss": 1.9262,
+      "step": 7052
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5855954349373058,
+      "learning_rate": 2.2645364289845527e-05,
+      "loss": 1.9751,
+      "step": 7053
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6068063082177816,
+      "learning_rate": 2.2639145427512675e-05,
+      "loss": 1.9319,
+      "step": 7054
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6119161880553763,
+      "learning_rate": 2.2632926712577664e-05,
+      "loss": 1.8622,
+      "step": 7055
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7401214494617729,
+      "learning_rate": 2.2626708145428767e-05,
+      "loss": 2.0986,
+      "step": 7056
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.650526575413362,
+      "learning_rate": 2.262048972645423e-05,
+      "loss": 1.8896,
+      "step": 7057
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6153992105687388,
+      "learning_rate": 2.2614271456042287e-05,
+      "loss": 1.867,
+      "step": 7058
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6958404556936367,
+      "learning_rate": 2.2608053334581188e-05,
+      "loss": 1.9781,
+      "step": 7059
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8525084546608769,
+      "learning_rate": 2.2601835362459137e-05,
+      "loss": 2.1342,
+      "step": 7060
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5961400193711438,
+      "learning_rate": 2.2595617540064362e-05,
+      "loss": 1.8963,
+      "step": 7061
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7133749822418753,
+      "learning_rate": 2.2589399867785066e-05,
+      "loss": 1.8691,
+      "step": 7062
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.769794224711012,
+      "learning_rate": 2.2583182346009427e-05,
+      "loss": 1.9175,
+      "step": 7063
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6764473170183589,
+      "learning_rate": 2.2576964975125653e-05,
+      "loss": 2.136,
+      "step": 7064
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6971862049428335,
+      "learning_rate": 2.2570747755521906e-05,
+      "loss": 1.9702,
+      "step": 7065
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6229210677628433,
+      "learning_rate": 2.2564530687586353e-05,
+      "loss": 1.892,
+      "step": 7066
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7336707328230961,
+      "learning_rate": 2.2558313771707164e-05,
+      "loss": 1.9791,
+      "step": 7067
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6234561192939817,
+      "learning_rate": 2.2552097008272465e-05,
+      "loss": 2.1177,
+      "step": 7068
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6286443785210944,
+      "learning_rate": 2.2545880397670418e-05,
+      "loss": 1.9123,
+      "step": 7069
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6793349183753359,
+      "learning_rate": 2.2539663940289126e-05,
+      "loss": 1.912,
+      "step": 7070
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6109221225200698,
+      "learning_rate": 2.253344763651673e-05,
+      "loss": 1.9851,
+      "step": 7071
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6411202618269222,
+      "learning_rate": 2.2527231486741335e-05,
+      "loss": 2.0817,
+      "step": 7072
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6277596865889165,
+      "learning_rate": 2.2521015491351024e-05,
+      "loss": 1.8819,
+      "step": 7073
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6490474024323507,
+      "learning_rate": 2.2514799650733903e-05,
+      "loss": 1.8837,
+      "step": 7074
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6498254397721837,
+      "learning_rate": 2.2508583965278056e-05,
+      "loss": 1.959,
+      "step": 7075
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6953074094947508,
+      "learning_rate": 2.2502368435371546e-05,
+      "loss": 2.1674,
+      "step": 7076
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7777982722019334,
+      "learning_rate": 2.249615306140244e-05,
+      "loss": 1.9823,
+      "step": 7077
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6517462064972002,
+      "learning_rate": 2.248993784375877e-05,
+      "loss": 1.8857,
+      "step": 7078
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6744181231832219,
+      "learning_rate": 2.2483722782828602e-05,
+      "loss": 1.9429,
+      "step": 7079
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6113470241886698,
+      "learning_rate": 2.247750787899996e-05,
+      "loss": 2.0814,
+      "step": 7080
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6370281683862717,
+      "learning_rate": 2.2471293132660854e-05,
+      "loss": 1.9052,
+      "step": 7081
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6864563819048298,
+      "learning_rate": 2.2465078544199315e-05,
+      "loss": 1.8725,
+      "step": 7082
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5688018938728608,
+      "learning_rate": 2.245886411400333e-05,
+      "loss": 2.0922,
+      "step": 7083
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6339467111976518,
+      "learning_rate": 2.2452649842460897e-05,
+      "loss": 1.985,
+      "step": 7084
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5978812465602381,
+      "learning_rate": 2.2446435729960005e-05,
+      "loss": 1.9389,
+      "step": 7085
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6214510825501665,
+      "learning_rate": 2.244022177688862e-05,
+      "loss": 1.8599,
+      "step": 7086
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6063567140804703,
+      "learning_rate": 2.2434007983634698e-05,
+      "loss": 1.9341,
+      "step": 7087
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6603026354214003,
+      "learning_rate": 2.2427794350586208e-05,
+      "loss": 2.0587,
+      "step": 7088
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6498869179808687,
+      "learning_rate": 2.242158087813108e-05,
+      "loss": 1.9256,
+      "step": 7089
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5645758018308584,
+      "learning_rate": 2.241536756665725e-05,
+      "loss": 1.9558,
+      "step": 7090
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5449768337855257,
+      "learning_rate": 2.240915441655264e-05,
+      "loss": 1.906,
+      "step": 7091
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6862921970938655,
+      "learning_rate": 2.2402941428205158e-05,
+      "loss": 2.1415,
+      "step": 7092
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6004315119994686,
+      "learning_rate": 2.239672860200272e-05,
+      "loss": 1.8968,
+      "step": 7093
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6422995537792966,
+      "learning_rate": 2.239051593833321e-05,
+      "loss": 1.9705,
+      "step": 7094
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6016649101271928,
+      "learning_rate": 2.23843034375845e-05,
+      "loss": 1.9008,
+      "step": 7095
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7433319391126177,
+      "learning_rate": 2.2378091100144486e-05,
+      "loss": 2.165,
+      "step": 7096
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6302796150504841,
+      "learning_rate": 2.2371878926401007e-05,
+      "loss": 1.9497,
+      "step": 7097
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7388779265309346,
+      "learning_rate": 2.236566691674193e-05,
+      "loss": 1.8547,
+      "step": 7098
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6020923971267467,
+      "learning_rate": 2.235945507155508e-05,
+      "loss": 1.9102,
+      "step": 7099
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5904257903188902,
+      "learning_rate": 2.2353243391228302e-05,
+      "loss": 2.085,
+      "step": 7100
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6017370117735489,
+      "learning_rate": 2.2347031876149416e-05,
+      "loss": 1.9043,
+      "step": 7101
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6253550887842904,
+      "learning_rate": 2.234082052670622e-05,
+      "loss": 1.9643,
+      "step": 7102
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5765322305345614,
+      "learning_rate": 2.2334609343286532e-05,
+      "loss": 1.9271,
+      "step": 7103
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5997474369010031,
+      "learning_rate": 2.232839832627813e-05,
+      "loss": 2.1051,
+      "step": 7104
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.61023107429698,
+      "learning_rate": 2.232218747606879e-05,
+      "loss": 1.9001,
+      "step": 7105
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7220776415909888,
+      "learning_rate": 2.231597679304629e-05,
+      "loss": 1.9247,
+      "step": 7106
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.602710895891604,
+      "learning_rate": 2.230976627759838e-05,
+      "loss": 1.9171,
+      "step": 7107
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6274813470842318,
+      "learning_rate": 2.230355593011282e-05,
+      "loss": 2.0654,
+      "step": 7108
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6293276020049982,
+      "learning_rate": 2.2297345750977345e-05,
+      "loss": 1.8803,
+      "step": 7109
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6902353520763248,
+      "learning_rate": 2.2291135740579666e-05,
+      "loss": 1.9062,
+      "step": 7110
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5897435616821807,
+      "learning_rate": 2.2284925899307524e-05,
+      "loss": 1.8983,
+      "step": 7111
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6894477058140261,
+      "learning_rate": 2.2278716227548602e-05,
+      "loss": 2.0887,
+      "step": 7112
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5631959590502968,
+      "learning_rate": 2.2272506725690602e-05,
+      "loss": 1.8773,
+      "step": 7113
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5674763541325322,
+      "learning_rate": 2.2266297394121224e-05,
+      "loss": 1.8867,
+      "step": 7114
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.644300278402013,
+      "learning_rate": 2.2260088233228124e-05,
+      "loss": 1.9708,
+      "step": 7115
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6722651746994859,
+      "learning_rate": 2.2253879243398976e-05,
+      "loss": 2.1445,
+      "step": 7116
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5909110145659097,
+      "learning_rate": 2.2247670425021416e-05,
+      "loss": 1.8911,
+      "step": 7117
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7002382137309457,
+      "learning_rate": 2.2241461778483107e-05,
+      "loss": 1.9433,
+      "step": 7118
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6592361345742014,
+      "learning_rate": 2.2235253304171674e-05,
+      "loss": 1.951,
+      "step": 7119
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5902326106815577,
+      "learning_rate": 2.222904500247473e-05,
+      "loss": 2.0921,
+      "step": 7120
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.657334138711458,
+      "learning_rate": 2.2222836873779888e-05,
+      "loss": 1.9693,
+      "step": 7121
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5950747318325049,
+      "learning_rate": 2.221662891847476e-05,
+      "loss": 1.9305,
+      "step": 7122
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6781177497196382,
+      "learning_rate": 2.2210421136946905e-05,
+      "loss": 1.8639,
+      "step": 7123
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7076692536163869,
+      "learning_rate": 2.2204213529583934e-05,
+      "loss": 2.0714,
+      "step": 7124
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7220402174776631,
+      "learning_rate": 2.2198006096773385e-05,
+      "loss": 1.9185,
+      "step": 7125
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6844509331292953,
+      "learning_rate": 2.2191798838902835e-05,
+      "loss": 1.9145,
+      "step": 7126
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7607255365029794,
+      "learning_rate": 2.218559175635982e-05,
+      "loss": 2.0228,
+      "step": 7127
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8821401844558856,
+      "learning_rate": 2.2179384849531867e-05,
+      "loss": 2.092,
+      "step": 7128
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6025159822689634,
+      "learning_rate": 2.217317811880651e-05,
+      "loss": 1.9039,
+      "step": 7129
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6164067586434357,
+      "learning_rate": 2.216697156457126e-05,
+      "loss": 1.9117,
+      "step": 7130
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7827460749410914,
+      "learning_rate": 2.2160765187213604e-05,
+      "loss": 1.8779,
+      "step": 7131
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6121111607789798,
+      "learning_rate": 2.2154558987121055e-05,
+      "loss": 2.0825,
+      "step": 7132
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6589871554842689,
+      "learning_rate": 2.214835296468107e-05,
+      "loss": 1.9043,
+      "step": 7133
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6041585164555597,
+      "learning_rate": 2.2142147120281116e-05,
+      "loss": 1.8571,
+      "step": 7134
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7757840823701326,
+      "learning_rate": 2.213594145430867e-05,
+      "loss": 1.9162,
+      "step": 7135
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5754137911446694,
+      "learning_rate": 2.2129735967151162e-05,
+      "loss": 2.118,
+      "step": 7136
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6791259800983376,
+      "learning_rate": 2.212353065919603e-05,
+      "loss": 1.8735,
+      "step": 7137
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.706139065615846,
+      "learning_rate": 2.2117325530830687e-05,
+      "loss": 1.8688,
+      "step": 7138
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7306988954336419,
+      "learning_rate": 2.2111120582442556e-05,
+      "loss": 1.914,
+      "step": 7139
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7065324494053647,
+      "learning_rate": 2.210491581441904e-05,
+      "loss": 2.1362,
+      "step": 7140
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6015279587928598,
+      "learning_rate": 2.2098711227147524e-05,
+      "loss": 1.8962,
+      "step": 7141
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6616721901679053,
+      "learning_rate": 2.2092506821015374e-05,
+      "loss": 1.8964,
+      "step": 7142
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6736976791115087,
+      "learning_rate": 2.208630259640998e-05,
+      "loss": 1.9148,
+      "step": 7143
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6437729616502058,
+      "learning_rate": 2.2080098553718682e-05,
+      "loss": 2.0852,
+      "step": 7144
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5836531457632554,
+      "learning_rate": 2.207389469332883e-05,
+      "loss": 1.9099,
+      "step": 7145
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.780522278964798,
+      "learning_rate": 2.206769101562774e-05,
+      "loss": 2.0065,
+      "step": 7146
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6294901526396731,
+      "learning_rate": 2.2061487521002756e-05,
+      "loss": 1.8784,
+      "step": 7147
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.659262610955476,
+      "learning_rate": 2.205528420984118e-05,
+      "loss": 2.1164,
+      "step": 7148
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7506478021350258,
+      "learning_rate": 2.2049081082530296e-05,
+      "loss": 1.9208,
+      "step": 7149
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6470726917128564,
+      "learning_rate": 2.2042878139457414e-05,
+      "loss": 1.9499,
+      "step": 7150
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6630278570646492,
+      "learning_rate": 2.2036675381009796e-05,
+      "loss": 1.8826,
+      "step": 7151
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5836915036791764,
+      "learning_rate": 2.2030472807574702e-05,
+      "loss": 2.1425,
+      "step": 7152
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6778376505702007,
+      "learning_rate": 2.2024270419539396e-05,
+      "loss": 1.9369,
+      "step": 7153
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6020272307999652,
+      "learning_rate": 2.201806821729111e-05,
+      "loss": 1.897,
+      "step": 7154
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6378113685334109,
+      "learning_rate": 2.201186620121707e-05,
+      "loss": 1.9331,
+      "step": 7155
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7196204078253812,
+      "learning_rate": 2.200566437170451e-05,
+      "loss": 2.0464,
+      "step": 7156
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6870507394380193,
+      "learning_rate": 2.199946272914061e-05,
+      "loss": 1.9842,
+      "step": 7157
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5903679297446236,
+      "learning_rate": 2.1993261273912595e-05,
+      "loss": 2.0015,
+      "step": 7158
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7061143685041945,
+      "learning_rate": 2.198706000640762e-05,
+      "loss": 1.9472,
+      "step": 7159
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7848484033832144,
+      "learning_rate": 2.198085892701286e-05,
+      "loss": 2.096,
+      "step": 7160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6541871937377336,
+      "learning_rate": 2.1974658036115496e-05,
+      "loss": 1.9383,
+      "step": 7161
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7265036105007652,
+      "learning_rate": 2.1968457334102653e-05,
+      "loss": 1.9324,
+      "step": 7162
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6588526125122433,
+      "learning_rate": 2.1962256821361474e-05,
+      "loss": 1.8729,
+      "step": 7163
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.698886202931842,
+      "learning_rate": 2.1956056498279072e-05,
+      "loss": 1.9966,
+      "step": 7164
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6022777693830814,
+      "learning_rate": 2.1949856365242573e-05,
+      "loss": 2.0972,
+      "step": 7165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6355794948438966,
+      "learning_rate": 2.194365642263908e-05,
+      "loss": 1.9053,
+      "step": 7166
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6804858176317747,
+      "learning_rate": 2.193745667085566e-05,
+      "loss": 1.8692,
+      "step": 7167
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8027507694284006,
+      "learning_rate": 2.1931257110279404e-05,
+      "loss": 2.0691,
+      "step": 7168
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5709697890511846,
+      "learning_rate": 2.192505774129738e-05,
+      "loss": 1.892,
+      "step": 7169
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8380485138601247,
+      "learning_rate": 2.1918858564296625e-05,
+      "loss": 1.9127,
+      "step": 7170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6667429478176476,
+      "learning_rate": 2.1912659579664194e-05,
+      "loss": 2.0148,
+      "step": 7171
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8201416312174684,
+      "learning_rate": 2.19064607877871e-05,
+      "loss": 2.1194,
+      "step": 7172
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6685640113804465,
+      "learning_rate": 2.1900262189052374e-05,
+      "loss": 1.8882,
+      "step": 7173
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.734331434556597,
+      "learning_rate": 2.1894063783847017e-05,
+      "loss": 1.9626,
+      "step": 7174
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7149495157399715,
+      "learning_rate": 2.1887865572558008e-05,
+      "loss": 1.8565,
+      "step": 7175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6527159256163847,
+      "learning_rate": 2.188166755557234e-05,
+      "loss": 2.0872,
+      "step": 7176
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6156317127751327,
+      "learning_rate": 2.1875469733276984e-05,
+      "loss": 1.9913,
+      "step": 7177
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6557012985822566,
+      "learning_rate": 2.1869272106058877e-05,
+      "loss": 1.9005,
+      "step": 7178
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6481861921031615,
+      "learning_rate": 2.1863074674304985e-05,
+      "loss": 1.8875,
+      "step": 7179
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6915144718678985,
+      "learning_rate": 2.1856877438402225e-05,
+      "loss": 2.0598,
+      "step": 7180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5659897229314858,
+      "learning_rate": 2.1850680398737507e-05,
+      "loss": 1.9198,
+      "step": 7181
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7180510336848404,
+      "learning_rate": 2.1844483555697765e-05,
+      "loss": 1.9171,
+      "step": 7182
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6970089792392445,
+      "learning_rate": 2.183828690966987e-05,
+      "loss": 2.0298,
+      "step": 7183
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7449044811721637,
+      "learning_rate": 2.183209046104072e-05,
+      "loss": 2.0964,
+      "step": 7184
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6001297285281055,
+      "learning_rate": 2.1825894210197166e-05,
+      "loss": 1.9219,
+      "step": 7185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7751054949107328,
+      "learning_rate": 2.1819698157526077e-05,
+      "loss": 1.9525,
+      "step": 7186
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6261356355558343,
+      "learning_rate": 2.1813502303414306e-05,
+      "loss": 1.8893,
+      "step": 7187
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5529366505647058,
+      "learning_rate": 2.1807306648248667e-05,
+      "loss": 2.0767,
+      "step": 7188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5735127884941142,
+      "learning_rate": 2.180111119241599e-05,
+      "loss": 1.9527,
+      "step": 7189
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6297245764528177,
+      "learning_rate": 2.179491593630309e-05,
+      "loss": 1.8792,
+      "step": 7190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6635401985904764,
+      "learning_rate": 2.178872088029675e-05,
+      "loss": 1.9294,
+      "step": 7191
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5366730749779657,
+      "learning_rate": 2.1782526024783765e-05,
+      "loss": 2.1127,
+      "step": 7192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.637139723089521,
+      "learning_rate": 2.1776331370150886e-05,
+      "loss": 1.8955,
+      "step": 7193
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6160504497415435,
+      "learning_rate": 2.1770136916784886e-05,
+      "loss": 1.9309,
+      "step": 7194
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.595431003120135,
+      "learning_rate": 2.1763942665072515e-05,
+      "loss": 1.9768,
+      "step": 7195
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6425601155605549,
+      "learning_rate": 2.1757748615400486e-05,
+      "loss": 1.9279,
+      "step": 7196
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6066156650627091,
+      "learning_rate": 2.175155476815554e-05,
+      "loss": 2.1865,
+      "step": 7197
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6224862680018973,
+      "learning_rate": 2.174536112372437e-05,
+      "loss": 1.9068,
+      "step": 7198
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6522896334917498,
+      "learning_rate": 2.1739167682493673e-05,
+      "loss": 1.8614,
+      "step": 7199
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6530008537812411,
+      "learning_rate": 2.1732974444850143e-05,
+      "loss": 2.0685,
+      "step": 7200
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6220273372286494,
+      "learning_rate": 2.1726781411180436e-05,
+      "loss": 1.8989,
+      "step": 7201
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6875934935557652,
+      "learning_rate": 2.1720588581871205e-05,
+      "loss": 2.0173,
+      "step": 7202
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.647425471230818,
+      "learning_rate": 2.1714395957309108e-05,
+      "loss": 1.895,
+      "step": 7203
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6975663778019863,
+      "learning_rate": 2.170820353788076e-05,
+      "loss": 2.132,
+      "step": 7204
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.652413674375587,
+      "learning_rate": 2.1702011323972806e-05,
+      "loss": 1.9458,
+      "step": 7205
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5861768745372263,
+      "learning_rate": 2.169581931597182e-05,
+      "loss": 1.8725,
+      "step": 7206
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6773070762895714,
+      "learning_rate": 2.1689627514264406e-05,
+      "loss": 1.9253,
+      "step": 7207
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6701305382543101,
+      "learning_rate": 2.168343591923715e-05,
+      "loss": 1.9629,
+      "step": 7208
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6551773956243744,
+      "learning_rate": 2.1677244531276615e-05,
+      "loss": 2.1104,
+      "step": 7209
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7680767086947965,
+      "learning_rate": 2.167105335076935e-05,
+      "loss": 1.8678,
+      "step": 7210
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6623728301108752,
+      "learning_rate": 2.1664862378101907e-05,
+      "loss": 1.9475,
+      "step": 7211
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6544914406511996,
+      "learning_rate": 2.1658671613660805e-05,
+      "loss": 2.0887,
+      "step": 7212
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7121214269180476,
+      "learning_rate": 2.1652481057832565e-05,
+      "loss": 1.9314,
+      "step": 7213
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5707876974294038,
+      "learning_rate": 2.1646290711003673e-05,
+      "loss": 1.9861,
+      "step": 7214
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6926351362731875,
+      "learning_rate": 2.1640100573560635e-05,
+      "loss": 1.8641,
+      "step": 7215
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6122877260474989,
+      "learning_rate": 2.1633910645889928e-05,
+      "loss": 1.8801,
+      "step": 7216
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6669921394037489,
+      "learning_rate": 2.162772092837799e-05,
+      "loss": 2.1207,
+      "step": 7217
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5918065698270532,
+      "learning_rate": 2.1621531421411306e-05,
+      "loss": 1.9429,
+      "step": 7218
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6268004447010787,
+      "learning_rate": 2.1615342125376287e-05,
+      "loss": 1.8946,
+      "step": 7219
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5598774952385015,
+      "learning_rate": 2.1609153040659358e-05,
+      "loss": 1.9713,
+      "step": 7220
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7073847708333127,
+      "learning_rate": 2.1602964167646947e-05,
+      "loss": 2.1206,
+      "step": 7221
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5838708180880664,
+      "learning_rate": 2.1596775506725424e-05,
+      "loss": 1.8763,
+      "step": 7222
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5800284778255044,
+      "learning_rate": 2.1590587058281197e-05,
+      "loss": 1.8516,
+      "step": 7223
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.657618647389656,
+      "learning_rate": 2.158439882270063e-05,
+      "loss": 2.1243,
+      "step": 7224
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5562270119868195,
+      "learning_rate": 2.1578210800370065e-05,
+      "loss": 1.9324,
+      "step": 7225
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5721847601824016,
+      "learning_rate": 2.1572022991675867e-05,
+      "loss": 1.993,
+      "step": 7226
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5690747419460479,
+      "learning_rate": 2.156583539700435e-05,
+      "loss": 1.911,
+      "step": 7227
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5881490877444991,
+      "learning_rate": 2.155964801674183e-05,
+      "loss": 1.9269,
+      "step": 7228
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6207478585431586,
+      "learning_rate": 2.155346085127463e-05,
+      "loss": 2.1148,
+      "step": 7229
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6411459013281953,
+      "learning_rate": 2.1547273900989026e-05,
+      "loss": 1.9268,
+      "step": 7230
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.558632660448574,
+      "learning_rate": 2.1541087166271298e-05,
+      "loss": 1.8909,
+      "step": 7231
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6412867247519702,
+      "learning_rate": 2.15349006475077e-05,
+      "loss": 2.1073,
+      "step": 7232
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5938810372108251,
+      "learning_rate": 2.152871434508449e-05,
+      "loss": 1.9594,
+      "step": 7233
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6025254743018162,
+      "learning_rate": 2.1522528259387915e-05,
+      "loss": 1.8702,
+      "step": 7234
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5538153892434978,
+      "learning_rate": 2.1516342390804174e-05,
+      "loss": 1.9244,
+      "step": 7235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6292929905538829,
+      "learning_rate": 2.1510156739719486e-05,
+      "loss": 2.0605,
+      "step": 7236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.589755722392195,
+      "learning_rate": 2.1503971306520068e-05,
+      "loss": 1.9073,
+      "step": 7237
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6042969964154412,
+      "learning_rate": 2.1497786091592066e-05,
+      "loss": 1.9172,
+      "step": 7238
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5792345466784315,
+      "learning_rate": 2.1491601095321678e-05,
+      "loss": 1.9593,
+      "step": 7239
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6446262371470731,
+      "learning_rate": 2.1485416318095034e-05,
+      "loss": 1.8965,
+      "step": 7240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6075509528148927,
+      "learning_rate": 2.1479231760298292e-05,
+      "loss": 2.127,
+      "step": 7241
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5905849354335069,
+      "learning_rate": 2.147304742231758e-05,
+      "loss": 1.8929,
+      "step": 7242
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6174299116019957,
+      "learning_rate": 2.1466863304539002e-05,
+      "loss": 1.9116,
+      "step": 7243
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6828452486405352,
+      "learning_rate": 2.1460679407348663e-05,
+      "loss": 2.0654,
+      "step": 7244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.585932819302075,
+      "learning_rate": 2.1454495731132656e-05,
+      "loss": 1.9564,
+      "step": 7245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6512186412691896,
+      "learning_rate": 2.1448312276277033e-05,
+      "loss": 1.9333,
+      "step": 7246
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7471593192695719,
+      "learning_rate": 2.1442129043167874e-05,
+      "loss": 1.866,
+      "step": 7247
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6217386041095689,
+      "learning_rate": 2.1435946032191214e-05,
+      "loss": 1.9202,
+      "step": 7248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.966798505952979,
+      "learning_rate": 2.1429763243733078e-05,
+      "loss": 2.1007,
+      "step": 7249
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5942478061829854,
+      "learning_rate": 2.14235806781795e-05,
+      "loss": 1.8993,
+      "step": 7250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8772466218251117,
+      "learning_rate": 2.1417398335916466e-05,
+      "loss": 2.0149,
+      "step": 7251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.674575726489538,
+      "learning_rate": 2.1411216217329975e-05,
+      "loss": 1.8765,
+      "step": 7252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6232146997120385,
+      "learning_rate": 2.1405034322805993e-05,
+      "loss": 2.0617,
+      "step": 7253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.780735005891494,
+      "learning_rate": 2.1398852652730485e-05,
+      "loss": 1.9695,
+      "step": 7254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8718858315235172,
+      "learning_rate": 2.139267120748941e-05,
+      "loss": 1.9176,
+      "step": 7255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6427575541782409,
+      "learning_rate": 2.1386489987468685e-05,
+      "loss": 2.1129,
+      "step": 7256
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6988038591974072,
+      "learning_rate": 2.1380308993054232e-05,
+      "loss": 1.9629,
+      "step": 7257
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8126967783396629,
+      "learning_rate": 2.1374128224631968e-05,
+      "loss": 1.9121,
+      "step": 7258
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6480828606258465,
+      "learning_rate": 2.1367947682587766e-05,
+      "loss": 1.8874,
+      "step": 7259
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7559179858354168,
+      "learning_rate": 2.136176736730752e-05,
+      "loss": 1.8846,
+      "step": 7260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6264785205289588,
+      "learning_rate": 2.135558727917707e-05,
+      "loss": 2.0998,
+      "step": 7261
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6855266999770363,
+      "learning_rate": 2.1349407418582285e-05,
+      "loss": 1.9076,
+      "step": 7262
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6841639230184681,
+      "learning_rate": 2.1343227785908996e-05,
+      "loss": 1.8807,
+      "step": 7263
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5186418502883205,
+      "learning_rate": 2.133704838154301e-05,
+      "loss": 1.924,
+      "step": 7264
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6156556500052351,
+      "learning_rate": 2.1330869205870153e-05,
+      "loss": 2.0754,
+      "step": 7265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6744965324455826,
+      "learning_rate": 2.1324690259276194e-05,
+      "loss": 1.8892,
+      "step": 7266
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5900687565550549,
+      "learning_rate": 2.1318511542146917e-05,
+      "loss": 1.8578,
+      "step": 7267
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5964819836294853,
+      "learning_rate": 2.1312333054868104e-05,
+      "loss": 2.0788,
+      "step": 7268
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6430949106129895,
+      "learning_rate": 2.130615479782548e-05,
+      "loss": 1.9343,
+      "step": 7269
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6288165514492883,
+      "learning_rate": 2.1299976771404777e-05,
+      "loss": 1.9859,
+      "step": 7270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6304949588771951,
+      "learning_rate": 2.129379897599174e-05,
+      "loss": 1.9052,
+      "step": 7271
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6029355302321225,
+      "learning_rate": 2.1287621411972045e-05,
+      "loss": 1.9103,
+      "step": 7272
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8158719594910049,
+      "learning_rate": 2.1281444079731407e-05,
+      "loss": 2.0746,
+      "step": 7273
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6241475591582576,
+      "learning_rate": 2.1275266979655487e-05,
+      "loss": 1.9031,
+      "step": 7274
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5825824557244752,
+      "learning_rate": 2.1269090112129942e-05,
+      "loss": 1.8996,
+      "step": 7275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7931492913138694,
+      "learning_rate": 2.1262913477540443e-05,
+      "loss": 1.9908,
+      "step": 7276
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8127104084335548,
+      "learning_rate": 2.1256737076272604e-05,
+      "loss": 2.1054,
+      "step": 7277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7282542847005522,
+      "learning_rate": 2.125056090871204e-05,
+      "loss": 1.9168,
+      "step": 7278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8202587071699139,
+      "learning_rate": 2.1244384975244373e-05,
+      "loss": 1.9363,
+      "step": 7279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7050566243680747,
+      "learning_rate": 2.123820927625517e-05,
+      "loss": 1.8982,
+      "step": 7280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6849295607571945,
+      "learning_rate": 2.1232033812130026e-05,
+      "loss": 2.1236,
+      "step": 7281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6767800671926392,
+      "learning_rate": 2.1225858583254478e-05,
+      "loss": 1.9975,
+      "step": 7282
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6911295310121451,
+      "learning_rate": 2.1219683590014087e-05,
+      "loss": 1.914,
+      "step": 7283
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7651820409333433,
+      "learning_rate": 2.121350883279439e-05,
+      "loss": 1.9307,
+      "step": 7284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7892108656387842,
+      "learning_rate": 2.120733431198088e-05,
+      "loss": 2.0932,
+      "step": 7285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7156816524560146,
+      "learning_rate": 2.1201160027959077e-05,
+      "loss": 1.9326,
+      "step": 7286
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6845370047602501,
+      "learning_rate": 2.119498598111445e-05,
+      "loss": 1.8781,
+      "step": 7287
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7145397295165097,
+      "learning_rate": 2.1188812171832484e-05,
+      "loss": 2.0986,
+      "step": 7288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6967629017218039,
+      "learning_rate": 2.1182638600498638e-05,
+      "loss": 1.9753,
+      "step": 7289
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6563082609900799,
+      "learning_rate": 2.1176465267498335e-05,
+      "loss": 1.9112,
+      "step": 7290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.765324743082738,
+      "learning_rate": 2.1170292173217016e-05,
+      "loss": 1.8701,
+      "step": 7291
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6646597247079286,
+      "learning_rate": 2.1164119318040097e-05,
+      "loss": 1.9354,
+      "step": 7292
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5549407964088721,
+      "learning_rate": 2.1157946702352956e-05,
+      "loss": 2.1084,
+      "step": 7293
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6800305775096903,
+      "learning_rate": 2.1151774326540995e-05,
+      "loss": 1.8631,
+      "step": 7294
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5770964736927102,
+      "learning_rate": 2.1145602190989573e-05,
+      "loss": 1.9447,
+      "step": 7295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6505654133223174,
+      "learning_rate": 2.1139430296084028e-05,
+      "loss": 1.8349,
+      "step": 7296
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5746499518406967,
+      "learning_rate": 2.1133258642209723e-05,
+      "loss": 2.0849,
+      "step": 7297
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7566038324024239,
+      "learning_rate": 2.112708722975196e-05,
+      "loss": 1.904,
+      "step": 7298
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6691683081121365,
+      "learning_rate": 2.112091605909606e-05,
+      "loss": 1.8989,
+      "step": 7299
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5890623129883249,
+      "learning_rate": 2.111474513062729e-05,
+      "loss": 2.0621,
+      "step": 7300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6424871177697925,
+      "learning_rate": 2.110857444473095e-05,
+      "loss": 1.9735,
+      "step": 7301
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5936359825885794,
+      "learning_rate": 2.11024040017923e-05,
+      "loss": 1.9183,
+      "step": 7302
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6188357063057587,
+      "learning_rate": 2.1096233802196566e-05,
+      "loss": 1.9112,
+      "step": 7303
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5953951359081591,
+      "learning_rate": 2.1090063846328994e-05,
+      "loss": 1.8927,
+      "step": 7304
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6504911132892564,
+      "learning_rate": 2.108389413457481e-05,
+      "loss": 2.0904,
+      "step": 7305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6652314757090627,
+      "learning_rate": 2.10777246673192e-05,
+      "loss": 1.9195,
+      "step": 7306
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5849517295895404,
+      "learning_rate": 2.1071555444947356e-05,
+      "loss": 1.9119,
+      "step": 7307
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6236488856542042,
+      "learning_rate": 2.106538646784443e-05,
+      "loss": 1.8592,
+      "step": 7308
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6425707000101294,
+      "learning_rate": 2.10592177363956e-05,
+      "loss": 2.1318,
+      "step": 7309
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.737830631510237,
+      "learning_rate": 2.1053049250986e-05,
+      "loss": 1.9342,
+      "step": 7310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6455412232761667,
+      "learning_rate": 2.1046881012000737e-05,
+      "loss": 1.9663,
+      "step": 7311
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6239959017085986,
+      "learning_rate": 2.1040713019824947e-05,
+      "loss": 1.8808,
+      "step": 7312
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7998685955768088,
+      "learning_rate": 2.1034545274843702e-05,
+      "loss": 2.1561,
+      "step": 7313
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6866229055262413,
+      "learning_rate": 2.1028377777442076e-05,
+      "loss": 1.9229,
+      "step": 7314
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6133971319030368,
+      "learning_rate": 2.1022210528005158e-05,
+      "loss": 1.8722,
+      "step": 7315
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6879325007277484,
+      "learning_rate": 2.101604352691797e-05,
+      "loss": 1.9324,
+      "step": 7316
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.703810009496628,
+      "learning_rate": 2.1009876774565546e-05,
+      "loss": 2.1263,
+      "step": 7317
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6151415270233368,
+      "learning_rate": 2.1003710271332914e-05,
+      "loss": 1.8897,
+      "step": 7318
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6151800921032786,
+      "learning_rate": 2.0997544017605062e-05,
+      "loss": 1.933,
+      "step": 7319
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6247689092229443,
+      "learning_rate": 2.099137801376699e-05,
+      "loss": 1.9846,
+      "step": 7320
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5750809373967433,
+      "learning_rate": 2.0985212260203646e-05,
+      "loss": 2.0955,
+      "step": 7321
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6239106670710417,
+      "learning_rate": 2.097904675729999e-05,
+      "loss": 1.8829,
+      "step": 7322
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6638413466005451,
+      "learning_rate": 2.0972881505440977e-05,
+      "loss": 1.8935,
+      "step": 7323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7277011628816623,
+      "learning_rate": 2.0966716505011507e-05,
+      "loss": 1.8811,
+      "step": 7324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7630823199090656,
+      "learning_rate": 2.096055175639649e-05,
+      "loss": 2.1231,
+      "step": 7325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6393340549565498,
+      "learning_rate": 2.0954387259980834e-05,
+      "loss": 1.9858,
+      "step": 7326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.655572543583511,
+      "learning_rate": 2.0948223016149395e-05,
+      "loss": 1.9043,
+      "step": 7327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7111844582569002,
+      "learning_rate": 2.0942059025287042e-05,
+      "loss": 1.921,
+      "step": 7328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.762136831360782,
+      "learning_rate": 2.0935895287778607e-05,
+      "loss": 2.1168,
+      "step": 7329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5797093263866309,
+      "learning_rate": 2.0929731804008927e-05,
+      "loss": 1.8717,
+      "step": 7330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.766374980836551,
+      "learning_rate": 2.0923568574362816e-05,
+      "loss": 1.8769,
+      "step": 7331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6227355717302747,
+      "learning_rate": 2.091740559922506e-05,
+      "loss": 1.9866,
+      "step": 7332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6500367381723878,
+      "learning_rate": 2.091124287898045e-05,
+      "loss": 2.1388,
+      "step": 7333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6039645270186607,
+      "learning_rate": 2.090508041401374e-05,
+      "loss": 1.9131,
+      "step": 7334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5906587379808558,
+      "learning_rate": 2.089891820470968e-05,
+      "loss": 1.8895,
+      "step": 7335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6914270717833539,
+      "learning_rate": 2.089275625145301e-05,
+      "loss": 1.9106,
+      "step": 7336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6782870065893911,
+      "learning_rate": 2.088659455462843e-05,
+      "loss": 2.1223,
+      "step": 7337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6588272938888124,
+      "learning_rate": 2.0880433114620655e-05,
+      "loss": 1.9942,
+      "step": 7338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.702487022460331,
+      "learning_rate": 2.0874271931814374e-05,
+      "loss": 1.8877,
+      "step": 7339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6768333364061985,
+      "learning_rate": 2.086811100659423e-05,
+      "loss": 1.8878,
+      "step": 7340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6666201058640474,
+      "learning_rate": 2.0861950339344903e-05,
+      "loss": 2.1104,
+      "step": 7341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6435467248136331,
+      "learning_rate": 2.0855789930451013e-05,
+      "loss": 1.8925,
+      "step": 7342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6320909845695809,
+      "learning_rate": 2.0849629780297177e-05,
+      "loss": 1.9657,
+      "step": 7343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6593159093040133,
+      "learning_rate": 2.0843469889268012e-05,
+      "loss": 1.9709,
+      "step": 7344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6994436826118214,
+      "learning_rate": 2.0837310257748093e-05,
+      "loss": 2.1234,
+      "step": 7345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5976903538861065,
+      "learning_rate": 2.0831150886121998e-05,
+      "loss": 1.8707,
+      "step": 7346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7060931059646764,
+      "learning_rate": 2.0824991774774267e-05,
+      "loss": 1.9359,
+      "step": 7347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6027539226784565,
+      "learning_rate": 2.081883292408946e-05,
+      "loss": 1.9441,
+      "step": 7348
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.705751071262903,
+      "learning_rate": 2.0812674334452098e-05,
+      "loss": 2.0955,
+      "step": 7349
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6501754017414836,
+      "learning_rate": 2.0806516006246667e-05,
+      "loss": 1.8891,
+      "step": 7350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6324623236416029,
+      "learning_rate": 2.0800357939857666e-05,
+      "loss": 1.9692,
+      "step": 7351
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6050688502804717,
+      "learning_rate": 2.0794200135669584e-05,
+      "loss": 1.8858,
+      "step": 7352
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.729193483662727,
+      "learning_rate": 2.078804259406686e-05,
+      "loss": 2.1256,
+      "step": 7353
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5865860927438855,
+      "learning_rate": 2.0781885315433945e-05,
+      "loss": 1.953,
+      "step": 7354
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.609515615648004,
+      "learning_rate": 2.0775728300155247e-05,
+      "loss": 1.9525,
+      "step": 7355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6442694883220836,
+      "learning_rate": 2.0769571548615192e-05,
+      "loss": 1.8235,
+      "step": 7356
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6364531234374721,
+      "learning_rate": 2.0763415061198167e-05,
+      "loss": 2.1417,
+      "step": 7357
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5782392300388424,
+      "learning_rate": 2.0757258838288535e-05,
+      "loss": 1.9585,
+      "step": 7358
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5868560016539307,
+      "learning_rate": 2.075110288027067e-05,
+      "loss": 1.9465,
+      "step": 7359
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6356635225885338,
+      "learning_rate": 2.074494718752891e-05,
+      "loss": 1.898,
+      "step": 7360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6293176831216851,
+      "learning_rate": 2.0738791760447566e-05,
+      "loss": 2.1025,
+      "step": 7361
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6016449991054256,
+      "learning_rate": 2.0732636599410973e-05,
+      "loss": 1.9183,
+      "step": 7362
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6386926365839832,
+      "learning_rate": 2.07264817048034e-05,
+      "loss": 1.9901,
+      "step": 7363
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.653190534971601,
+      "learning_rate": 2.0720327077009125e-05,
+      "loss": 1.9277,
+      "step": 7364
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6671414947956004,
+      "learning_rate": 2.0714172716412422e-05,
+      "loss": 2.1002,
+      "step": 7365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6823965286627397,
+      "learning_rate": 2.0708018623397518e-05,
+      "loss": 1.9273,
+      "step": 7366
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6225618021359707,
+      "learning_rate": 2.070186479834865e-05,
+      "loss": 1.8743,
+      "step": 7367
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6371447632782691,
+      "learning_rate": 2.0695711241650005e-05,
+      "loss": 1.895,
+      "step": 7368
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6596598709758402,
+      "learning_rate": 2.068955795368579e-05,
+      "loss": 2.11,
+      "step": 7369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6276795787017508,
+      "learning_rate": 2.0683404934840194e-05,
+      "loss": 1.9647,
+      "step": 7370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.645061532308815,
+      "learning_rate": 2.0677252185497357e-05,
+      "loss": 1.8867,
+      "step": 7371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6170649555892768,
+      "learning_rate": 2.0671099706041413e-05,
+      "loss": 1.8827,
+      "step": 7372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6017387813683415,
+      "learning_rate": 2.0664947496856514e-05,
+      "loss": 2.0895,
+      "step": 7373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5851078981143658,
+      "learning_rate": 2.0658795558326743e-05,
+      "loss": 1.8712,
+      "step": 7374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5973031696518187,
+      "learning_rate": 2.0652643890836206e-05,
+      "loss": 1.9821,
+      "step": 7375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6775773923427596,
+      "learning_rate": 2.064649249476896e-05,
+      "loss": 1.9546,
+      "step": 7376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6113860286992369,
+      "learning_rate": 2.0640341370509076e-05,
+      "loss": 2.1126,
+      "step": 7377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6929904803168362,
+      "learning_rate": 2.0634190518440595e-05,
+      "loss": 1.8947,
+      "step": 7378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7734925789046544,
+      "learning_rate": 2.0628039938947528e-05,
+      "loss": 1.9135,
+      "step": 7379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6616484667557089,
+      "learning_rate": 2.0621889632413897e-05,
+      "loss": 1.9011,
+      "step": 7380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7183585584146585,
+      "learning_rate": 2.0615739599223677e-05,
+      "loss": 2.1266,
+      "step": 7381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7285490928099024,
+      "learning_rate": 2.0609589839760837e-05,
+      "loss": 1.9792,
+      "step": 7382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6570023923710698,
+      "learning_rate": 2.0603440354409352e-05,
+      "loss": 1.8497,
+      "step": 7383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7165867499995137,
+      "learning_rate": 2.0597291143553137e-05,
+      "loss": 1.8766,
+      "step": 7384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7159072975781767,
+      "learning_rate": 2.059114220757613e-05,
+      "loss": 2.1021,
+      "step": 7385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5531413868131627,
+      "learning_rate": 2.0584993546862226e-05,
+      "loss": 1.8672,
+      "step": 7386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6555195095108425,
+      "learning_rate": 2.0578845161795307e-05,
+      "loss": 1.9261,
+      "step": 7387
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6276633997778613,
+      "learning_rate": 2.057269705275926e-05,
+      "loss": 2.019,
+      "step": 7388
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6387465327063436,
+      "learning_rate": 2.056654922013791e-05,
+      "loss": 2.0728,
+      "step": 7389
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6036005097645605,
+      "learning_rate": 2.0560401664315103e-05,
+      "loss": 1.9428,
+      "step": 7390
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6996799062165104,
+      "learning_rate": 2.0554254385674672e-05,
+      "loss": 1.8858,
+      "step": 7391
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6748253409045181,
+      "learning_rate": 2.05481073846004e-05,
+      "loss": 1.9227,
+      "step": 7392
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7628063743483974,
+      "learning_rate": 2.0541960661476063e-05,
+      "loss": 2.1167,
+      "step": 7393
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.658668073202702,
+      "learning_rate": 2.0535814216685447e-05,
+      "loss": 1.987,
+      "step": 7394
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7283869058421752,
+      "learning_rate": 2.0529668050612285e-05,
+      "loss": 1.9042,
+      "step": 7395
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6069732163069816,
+      "learning_rate": 2.0523522163640316e-05,
+      "loss": 1.8773,
+      "step": 7396
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6449391014440827,
+      "learning_rate": 2.0517376556153236e-05,
+      "loss": 2.1408,
+      "step": 7397
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6505226632544117,
+      "learning_rate": 2.051123122853476e-05,
+      "loss": 1.8859,
+      "step": 7398
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5534992712997799,
+      "learning_rate": 2.0505086181168563e-05,
+      "loss": 1.8782,
+      "step": 7399
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6117614993161784,
+      "learning_rate": 2.049894141443829e-05,
+      "loss": 1.9413,
+      "step": 7400
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6527196665692702,
+      "learning_rate": 2.049279692872761e-05,
+      "loss": 2.0682,
+      "step": 7401
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.640524576464648,
+      "learning_rate": 2.048665272442012e-05,
+      "loss": 1.8479,
+      "step": 7402
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.658504031410833,
+      "learning_rate": 2.0480508801899444e-05,
+      "loss": 1.8991,
+      "step": 7403
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7941770572559292,
+      "learning_rate": 2.047436516154918e-05,
+      "loss": 1.9206,
+      "step": 7404
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6878974126968476,
+      "learning_rate": 2.046822180375288e-05,
+      "loss": 2.0717,
+      "step": 7405
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.71425617941636,
+      "learning_rate": 2.0462078728894113e-05,
+      "loss": 2.0106,
+      "step": 7406
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8648724224857081,
+      "learning_rate": 2.0455935937356418e-05,
+      "loss": 1.9283,
+      "step": 7407
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.626782430205911,
+      "learning_rate": 2.0449793429523302e-05,
+      "loss": 1.8618,
+      "step": 7408
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7414114821446328,
+      "learning_rate": 2.0443651205778284e-05,
+      "loss": 2.1174,
+      "step": 7409
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6968132671299696,
+      "learning_rate": 2.0437509266504832e-05,
+      "loss": 1.9201,
+      "step": 7410
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7281002042507275,
+      "learning_rate": 2.0431367612086416e-05,
+      "loss": 1.8651,
+      "step": 7411
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6356516589441528,
+      "learning_rate": 2.04252262429065e-05,
+      "loss": 1.8927,
+      "step": 7412
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7431854250079011,
+      "learning_rate": 2.0419085159348498e-05,
+      "loss": 1.9704,
+      "step": 7413
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8332549538721065,
+      "learning_rate": 2.041294436179583e-05,
+      "loss": 2.1103,
+      "step": 7414
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5797416300580368,
+      "learning_rate": 2.0406803850631883e-05,
+      "loss": 1.8642,
+      "step": 7415
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7284414722257874,
+      "learning_rate": 2.040066362624004e-05,
+      "loss": 1.8877,
+      "step": 7416
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7216333573857389,
+      "learning_rate": 2.0394523689003673e-05,
+      "loss": 2.0827,
+      "step": 7417
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6284776663562003,
+      "learning_rate": 2.0388384039306106e-05,
+      "loss": 1.8757,
+      "step": 7418
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5951179781957419,
+      "learning_rate": 2.038224467753066e-05,
+      "loss": 1.9637,
+      "step": 7419
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7521373977327651,
+      "learning_rate": 2.0376105604060662e-05,
+      "loss": 1.9383,
+      "step": 7420
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6787394134085954,
+      "learning_rate": 2.0369966819279384e-05,
+      "loss": 2.0367,
+      "step": 7421
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6373840400038623,
+      "learning_rate": 2.03638283235701e-05,
+      "loss": 1.8479,
+      "step": 7422
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.737724974414012,
+      "learning_rate": 2.035769011731605e-05,
+      "loss": 1.8992,
+      "step": 7423
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6356524877603701,
+      "learning_rate": 2.0351552200900483e-05,
+      "loss": 1.9707,
+      "step": 7424
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6993567715220683,
+      "learning_rate": 2.034541457470661e-05,
+      "loss": 1.9438,
+      "step": 7425
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6278721841817009,
+      "learning_rate": 2.0339277239117622e-05,
+      "loss": 2.0964,
+      "step": 7426
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7216983972602918,
+      "learning_rate": 2.0333140194516705e-05,
+      "loss": 1.9102,
+      "step": 7427
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5843360294576814,
+      "learning_rate": 2.032700344128702e-05,
+      "loss": 1.9312,
+      "step": 7428
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7878950908674687,
+      "learning_rate": 2.0320866979811702e-05,
+      "loss": 2.055,
+      "step": 7429
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6416528585355762,
+      "learning_rate": 2.0314730810473886e-05,
+      "loss": 1.8602,
+      "step": 7430
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.645743870292766,
+      "learning_rate": 2.0308594933656673e-05,
+      "loss": 2.0019,
+      "step": 7431
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6327059867524288,
+      "learning_rate": 2.030245934974314e-05,
+      "loss": 1.895,
+      "step": 7432
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5957065992942077,
+      "learning_rate": 2.0296324059116384e-05,
+      "loss": 1.8854,
+      "step": 7433
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6853454415237912,
+      "learning_rate": 2.0290189062159426e-05,
+      "loss": 2.1029,
+      "step": 7434
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5830829976964061,
+      "learning_rate": 2.028405435925533e-05,
+      "loss": 1.8808,
+      "step": 7435
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.661245305855937,
+      "learning_rate": 2.027791995078708e-05,
+      "loss": 1.8743,
+      "step": 7436
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7477937381320924,
+      "learning_rate": 2.0271785837137685e-05,
+      "loss": 2.0629,
+      "step": 7437
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.706015409677576,
+      "learning_rate": 2.0265652018690135e-05,
+      "loss": 1.9585,
+      "step": 7438
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7595538683054777,
+      "learning_rate": 2.0259518495827372e-05,
+      "loss": 1.9137,
+      "step": 7439
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5607071566507457,
+      "learning_rate": 2.025338526893234e-05,
+      "loss": 1.8659,
+      "step": 7440
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8435936549757441,
+      "learning_rate": 2.0247252338387976e-05,
+      "loss": 2.1034,
+      "step": 7441
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6633180763715603,
+      "learning_rate": 2.0241119704577168e-05,
+      "loss": 1.8802,
+      "step": 7442
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5874910635053093,
+      "learning_rate": 2.023498736788281e-05,
+      "loss": 1.8765,
+      "step": 7443
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7818123573771695,
+      "learning_rate": 2.0228855328687756e-05,
+      "loss": 1.9539,
+      "step": 7444
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6425650006863312,
+      "learning_rate": 2.022272358737487e-05,
+      "loss": 1.8853,
+      "step": 7445
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7359990422258524,
+      "learning_rate": 2.021659214432698e-05,
+      "loss": 2.118,
+      "step": 7446
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6259407194982133,
+      "learning_rate": 2.0210460999926887e-05,
+      "loss": 1.8953,
+      "step": 7447
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.627919178173089,
+      "learning_rate": 2.0204330154557395e-05,
+      "loss": 1.8928,
+      "step": 7448
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5863685512161994,
+      "learning_rate": 2.0198199608601263e-05,
+      "loss": 2.0774,
+      "step": 7449
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6990582930259674,
+      "learning_rate": 2.0192069362441262e-05,
+      "loss": 2.0039,
+      "step": 7450
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6186505863705892,
+      "learning_rate": 2.0185939416460133e-05,
+      "loss": 1.8586,
+      "step": 7451
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5804258153121453,
+      "learning_rate": 2.0179809771040564e-05,
+      "loss": 1.8572,
+      "step": 7452
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6708033748391804,
+      "learning_rate": 2.0173680426565282e-05,
+      "loss": 2.0746,
+      "step": 7453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5670344301200633,
+      "learning_rate": 2.016755138341696e-05,
+      "loss": 1.8525,
+      "step": 7454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6115158828790319,
+      "learning_rate": 2.016142264197825e-05,
+      "loss": 1.9054,
+      "step": 7455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5784293196161365,
+      "learning_rate": 2.015529420263181e-05,
+      "loss": 1.9921,
+      "step": 7456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5936630638482707,
+      "learning_rate": 2.0149166065760254e-05,
+      "loss": 1.8874,
+      "step": 7457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6060218442028465,
+      "learning_rate": 2.0143038231746182e-05,
+      "loss": 2.0903,
+      "step": 7458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5762059045394486,
+      "learning_rate": 2.0136910700972193e-05,
+      "loss": 1.9866,
+      "step": 7459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5996693658244541,
+      "learning_rate": 2.0130783473820843e-05,
+      "loss": 1.9174,
+      "step": 7460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6404792004372338,
+      "learning_rate": 2.0124656550674693e-05,
+      "loss": 2.1065,
+      "step": 7461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6121254749829028,
+      "learning_rate": 2.0118529931916247e-05,
+      "loss": 1.9494,
+      "step": 7462
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6516623998966036,
+      "learning_rate": 2.0112403617928037e-05,
+      "loss": 1.8624,
+      "step": 7463
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5898528230755222,
+      "learning_rate": 2.0106277609092557e-05,
+      "loss": 1.9155,
+      "step": 7464
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6476386149013306,
+      "learning_rate": 2.0100151905792257e-05,
+      "loss": 1.8746,
+      "step": 7465
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6260923744569447,
+      "learning_rate": 2.00940265084096e-05,
+      "loss": 2.1048,
+      "step": 7466
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5857347870856231,
+      "learning_rate": 2.008790141732704e-05,
+      "loss": 1.8918,
+      "step": 7467
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5651491519892858,
+      "learning_rate": 2.0081776632926965e-05,
+      "loss": 1.8136,
+      "step": 7468
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5522517109868166,
+      "learning_rate": 2.007565215559178e-05,
+      "loss": 1.9117,
+      "step": 7469
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6346710075024526,
+      "learning_rate": 2.0069527985703857e-05,
+      "loss": 2.1084,
+      "step": 7470
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6665164053294654,
+      "learning_rate": 2.0063404123645556e-05,
+      "loss": 1.8704,
+      "step": 7471
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5901807610479669,
+      "learning_rate": 2.0057280569799223e-05,
+      "loss": 1.9049,
+      "step": 7472
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.635968917172567,
+      "learning_rate": 2.0051157324547163e-05,
+      "loss": 2.1212,
+      "step": 7473
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.605951827647163,
+      "learning_rate": 2.0045034388271682e-05,
+      "loss": 1.9667,
+      "step": 7474
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6223305054312995,
+      "learning_rate": 2.0038911761355066e-05,
+      "loss": 1.9933,
+      "step": 7475
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5882975045259876,
+      "learning_rate": 2.003278944417956e-05,
+      "loss": 1.8758,
+      "step": 7476
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6558002044253558,
+      "learning_rate": 2.0026667437127425e-05,
+      "loss": 1.9152,
+      "step": 7477
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6593655389980609,
+      "learning_rate": 2.0020545740580868e-05,
+      "loss": 2.0409,
+      "step": 7478
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5719161447590927,
+      "learning_rate": 2.001442435492209e-05,
+      "loss": 1.9346,
+      "step": 7479
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6163830647250391,
+      "learning_rate": 2.000830328053329e-05,
+      "loss": 1.8942,
+      "step": 7480
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6298884309303193,
+      "learning_rate": 2.0002182517796618e-05,
+      "loss": 1.9555,
+      "step": 7481
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6152251923132696,
+      "learning_rate": 1.9996062067094228e-05,
+      "loss": 2.0791,
+      "step": 7482
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6171229527220804,
+      "learning_rate": 1.9989941928808237e-05,
+      "loss": 1.8887,
+      "step": 7483
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7590752427832077,
+      "learning_rate": 1.9983822103320747e-05,
+      "loss": 1.879,
+      "step": 7484
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6410255358750863,
+      "learning_rate": 1.9977702591013857e-05,
+      "loss": 2.147,
+      "step": 7485
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6097421190168215,
+      "learning_rate": 1.997158339226963e-05,
+      "loss": 1.7997,
+      "step": 7486
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6546721909584039,
+      "learning_rate": 1.9965464507470095e-05,
+      "loss": 1.9504,
+      "step": 7487
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.664879367497345,
+      "learning_rate": 1.9959345936997306e-05,
+      "loss": 1.9098,
+      "step": 7488
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6574980666547309,
+      "learning_rate": 1.9953227681233256e-05,
+      "loss": 1.9149,
+      "step": 7489
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6670381344607366,
+      "learning_rate": 1.9947109740559936e-05,
+      "loss": 2.1465,
+      "step": 7490
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6490529542813748,
+      "learning_rate": 1.9940992115359303e-05,
+      "loss": 1.8828,
+      "step": 7491
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7091719813146908,
+      "learning_rate": 1.993487480601332e-05,
+      "loss": 1.8648,
+      "step": 7492
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.663230339097832,
+      "learning_rate": 1.992875781290392e-05,
+      "loss": 1.9842,
+      "step": 7493
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6451242447074287,
+      "learning_rate": 1.992264113641299e-05,
+      "loss": 2.0838,
+      "step": 7494
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6761461984001201,
+      "learning_rate": 1.991652477692244e-05,
+      "loss": 1.8725,
+      "step": 7495
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6764888629671318,
+      "learning_rate": 1.991040873481413e-05,
+      "loss": 1.8958,
+      "step": 7496
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6348643339141381,
+      "learning_rate": 1.9904293010469906e-05,
+      "loss": 1.8808,
+      "step": 7497
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6372306706892445,
+      "learning_rate": 1.989817760427161e-05,
+      "loss": 2.0818,
+      "step": 7498
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5985030957992815,
+      "learning_rate": 1.9892062516601036e-05,
+      "loss": 1.8452,
+      "step": 7499
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6256361061427789,
+      "learning_rate": 1.988594774783999e-05,
+      "loss": 1.9576,
+      "step": 7500
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6360511425134711,
+      "learning_rate": 1.9879833298370238e-05,
+      "loss": 1.8892,
+      "step": 7501
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6999622105388574,
+      "learning_rate": 1.9873719168573517e-05,
+      "loss": 2.0999,
+      "step": 7502
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5395675035008263,
+      "learning_rate": 1.9867605358831583e-05,
+      "loss": 1.8828,
+      "step": 7503
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6807422017356013,
+      "learning_rate": 1.986149186952612e-05,
+      "loss": 1.9035,
+      "step": 7504
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6111838818270434,
+      "learning_rate": 1.9855378701038822e-05,
+      "loss": 2.1125,
+      "step": 7505
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5613667252474608,
+      "learning_rate": 1.9849265853751383e-05,
+      "loss": 1.9703,
+      "step": 7506
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6696172171386587,
+      "learning_rate": 1.9843153328045427e-05,
+      "loss": 1.9168,
+      "step": 7507
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.599626743049355,
+      "learning_rate": 1.9837041124302592e-05,
+      "loss": 1.8893,
+      "step": 7508
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6155077201274783,
+      "learning_rate": 1.98309292429045e-05,
+      "loss": 1.874,
+      "step": 7509
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6281653756446095,
+      "learning_rate": 1.982481768423272e-05,
+      "loss": 2.1106,
+      "step": 7510
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6422081517210234,
+      "learning_rate": 1.9818706448668843e-05,
+      "loss": 1.8745,
+      "step": 7511
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5335942756138671,
+      "learning_rate": 1.9812595536594393e-05,
+      "loss": 1.9267,
+      "step": 7512
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5873523027594504,
+      "learning_rate": 1.9806484948390916e-05,
+      "loss": 1.8589,
+      "step": 7513
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6550029628677968,
+      "learning_rate": 1.980037468443993e-05,
+      "loss": 2.0874,
+      "step": 7514
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5955135580517517,
+      "learning_rate": 1.979426474512291e-05,
+      "loss": 1.8733,
+      "step": 7515
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6351558349762674,
+      "learning_rate": 1.978815513082133e-05,
+      "loss": 1.8892,
+      "step": 7516
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5896240603083429,
+      "learning_rate": 1.9782045841916625e-05,
+      "loss": 2.1121,
+      "step": 7517
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6049194885746961,
+      "learning_rate": 1.977593687879024e-05,
+      "loss": 1.993,
+      "step": 7518
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6476357499497972,
+      "learning_rate": 1.976982824182358e-05,
+      "loss": 1.8646,
+      "step": 7519
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6076586817350905,
+      "learning_rate": 1.9763719931398022e-05,
+      "loss": 1.9183,
+      "step": 7520
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5797685028477424,
+      "learning_rate": 1.9757611947894943e-05,
+      "loss": 1.925,
+      "step": 7521
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6096386985363572,
+      "learning_rate": 1.9751504291695694e-05,
+      "loss": 2.1126,
+      "step": 7522
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5943040080437846,
+      "learning_rate": 1.974539696318158e-05,
+      "loss": 1.88,
+      "step": 7523
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5497812217171836,
+      "learning_rate": 1.973928996273393e-05,
+      "loss": 1.9296,
+      "step": 7524
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5525565185515982,
+      "learning_rate": 1.9733183290734016e-05,
+      "loss": 1.8942,
+      "step": 7525
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.577473909009619,
+      "learning_rate": 1.97270769475631e-05,
+      "loss": 2.0955,
+      "step": 7526
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5569508128697931,
+      "learning_rate": 1.9720970933602443e-05,
+      "loss": 1.8713,
+      "step": 7527
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5597587193983119,
+      "learning_rate": 1.971486524923325e-05,
+      "loss": 1.7974,
+      "step": 7528
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5770106661095573,
+      "learning_rate": 1.970875989483674e-05,
+      "loss": 1.859,
+      "step": 7529
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6099842965258105,
+      "learning_rate": 1.9702654870794072e-05,
+      "loss": 2.101,
+      "step": 7530
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5573426705048805,
+      "learning_rate": 1.9696550177486427e-05,
+      "loss": 1.9299,
+      "step": 7531
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5369077858623763,
+      "learning_rate": 1.969044581529495e-05,
+      "loss": 1.9068,
+      "step": 7532
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6089210354151908,
+      "learning_rate": 1.9684341784600747e-05,
+      "loss": 1.888,
+      "step": 7533
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7114880549646531,
+      "learning_rate": 1.967823808578492e-05,
+      "loss": 2.1132,
+      "step": 7534
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5641173395370542,
+      "learning_rate": 1.967213471922856e-05,
+      "loss": 1.8494,
+      "step": 7535
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6504060120101758,
+      "learning_rate": 1.9666031685312713e-05,
+      "loss": 1.8698,
+      "step": 7536
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7751881732319554,
+      "learning_rate": 1.9659928984418423e-05,
+      "loss": 2.0236,
+      "step": 7537
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.65248424949771,
+      "learning_rate": 1.96538266169267e-05,
+      "loss": 2.1419,
+      "step": 7538
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7119397484158105,
+      "learning_rate": 1.9647724583218543e-05,
+      "loss": 1.9427,
+      "step": 7539
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6102276425834192,
+      "learning_rate": 1.9641622883674937e-05,
+      "loss": 1.9154,
+      "step": 7540
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6075383596120245,
+      "learning_rate": 1.9635521518676813e-05,
+      "loss": 1.8703,
+      "step": 7541
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6724357469914515,
+      "learning_rate": 1.9629420488605128e-05,
+      "loss": 2.0996,
+      "step": 7542
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5967935028428883,
+      "learning_rate": 1.9623319793840787e-05,
+      "loss": 1.9402,
+      "step": 7543
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6798387192989632,
+      "learning_rate": 1.961721943476467e-05,
+      "loss": 1.8666,
+      "step": 7544
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5955447451059178,
+      "learning_rate": 1.9611119411757665e-05,
+      "loss": 1.8419,
+      "step": 7545
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6602241962033706,
+      "learning_rate": 1.9605019725200605e-05,
+      "loss": 2.0853,
+      "step": 7546
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6837053047327782,
+      "learning_rate": 1.959892037547433e-05,
+      "loss": 1.9084,
+      "step": 7547
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6317991216773129,
+      "learning_rate": 1.9592821362959653e-05,
+      "loss": 1.9026,
+      "step": 7548
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6683632362655592,
+      "learning_rate": 1.9586722688037338e-05,
+      "loss": 1.9967,
+      "step": 7549
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6633348025790817,
+      "learning_rate": 1.9580624351088176e-05,
+      "loss": 2.0673,
+      "step": 7550
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.602838920944642,
+      "learning_rate": 1.957452635249289e-05,
+      "loss": 1.8658,
+      "step": 7551
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.664692394167682,
+      "learning_rate": 1.956842869263221e-05,
+      "loss": 1.8804,
+      "step": 7552
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7348265005897375,
+      "learning_rate": 1.9562331371886853e-05,
+      "loss": 1.8952,
+      "step": 7553
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6979152235019084,
+      "learning_rate": 1.9556234390637478e-05,
+      "loss": 2.1225,
+      "step": 7554
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6594635941213265,
+      "learning_rate": 1.955013774926475e-05,
+      "loss": 1.9475,
+      "step": 7555
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7153893886509993,
+      "learning_rate": 1.9544041448149323e-05,
+      "loss": 1.8636,
+      "step": 7556
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6232601181222743,
+      "learning_rate": 1.9537945487671794e-05,
+      "loss": 1.8631,
+      "step": 7557
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6831352141143314,
+      "learning_rate": 1.9531849868212775e-05,
+      "loss": 2.0824,
+      "step": 7558
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5646218633526511,
+      "learning_rate": 1.9525754590152818e-05,
+      "loss": 1.8719,
+      "step": 7559
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.636855344582006,
+      "learning_rate": 1.9519659653872498e-05,
+      "loss": 1.8465,
+      "step": 7560
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.669076372448756,
+      "learning_rate": 1.9513565059752343e-05,
+      "loss": 1.8569,
+      "step": 7561
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5967328342704521,
+      "learning_rate": 1.9507470808172855e-05,
+      "loss": 2.1491,
+      "step": 7562
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7086050301968272,
+      "learning_rate": 1.9501376899514533e-05,
+      "loss": 1.8971,
+      "step": 7563
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6896803251796363,
+      "learning_rate": 1.949528333415783e-05,
+      "loss": 1.9368,
+      "step": 7564
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6199854808112897,
+      "learning_rate": 1.948919011248321e-05,
+      "loss": 1.841,
+      "step": 7565
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.687056408491387,
+      "learning_rate": 1.9483097234871094e-05,
+      "loss": 2.071,
+      "step": 7566
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7042535060235954,
+      "learning_rate": 1.9477004701701874e-05,
+      "loss": 1.9079,
+      "step": 7567
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5570597359259216,
+      "learning_rate": 1.9470912513355942e-05,
+      "loss": 1.9658,
+      "step": 7568
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6629395922030592,
+      "learning_rate": 1.9464820670213663e-05,
+      "loss": 1.8729,
+      "step": 7569
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7239165032245175,
+      "learning_rate": 1.9458729172655357e-05,
+      "loss": 2.0527,
+      "step": 7570
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6368581978630224,
+      "learning_rate": 1.9452638021061364e-05,
+      "loss": 1.8782,
+      "step": 7571
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5908619781751578,
+      "learning_rate": 1.944654721581196e-05,
+      "loss": 1.8752,
+      "step": 7572
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7667053191603881,
+      "learning_rate": 1.944045675728743e-05,
+      "loss": 1.8701,
+      "step": 7573
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6085696768060196,
+      "learning_rate": 1.9434366645868028e-05,
+      "loss": 2.1137,
+      "step": 7574
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.753390430733746,
+      "learning_rate": 1.9428276881933977e-05,
+      "loss": 1.8631,
+      "step": 7575
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.63005162168378,
+      "learning_rate": 1.9422187465865495e-05,
+      "loss": 1.9281,
+      "step": 7576
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.578794465516103,
+      "learning_rate": 1.941609839804275e-05,
+      "loss": 1.8995,
+      "step": 7577
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6501152993703927,
+      "learning_rate": 1.9410009678845923e-05,
+      "loss": 2.1268,
+      "step": 7578
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6252823273451281,
+      "learning_rate": 1.9403921308655174e-05,
+      "loss": 1.9138,
+      "step": 7579
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5666028147532545,
+      "learning_rate": 1.939783328785059e-05,
+      "loss": 1.9616,
+      "step": 7580
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5504643615582026,
+      "learning_rate": 1.9391745616812285e-05,
+      "loss": 1.8821,
+      "step": 7581
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6995082849356324,
+      "learning_rate": 1.9385658295920357e-05,
+      "loss": 2.0964,
+      "step": 7582
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6160606732619602,
+      "learning_rate": 1.9379571325554834e-05,
+      "loss": 1.8769,
+      "step": 7583
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5878447444420979,
+      "learning_rate": 1.9373484706095775e-05,
+      "loss": 1.8947,
+      "step": 7584
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7063162727936425,
+      "learning_rate": 1.936739843792316e-05,
+      "loss": 1.8695,
+      "step": 7585
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5468114631380889,
+      "learning_rate": 1.9361312521417013e-05,
+      "loss": 2.1268,
+      "step": 7586
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6016263376687625,
+      "learning_rate": 1.9355226956957293e-05,
+      "loss": 1.8518,
+      "step": 7587
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5718826876833379,
+      "learning_rate": 1.934914174492393e-05,
+      "loss": 1.8744,
+      "step": 7588
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6938886081879279,
+      "learning_rate": 1.934305688569687e-05,
+      "loss": 1.8735,
+      "step": 7589
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.611434917581038,
+      "learning_rate": 1.9336972379656015e-05,
+      "loss": 2.0847,
+      "step": 7590
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6697880557438126,
+      "learning_rate": 1.9330888227181225e-05,
+      "loss": 1.8967,
+      "step": 7591
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.644376929742968,
+      "learning_rate": 1.9324804428652386e-05,
+      "loss": 1.874,
+      "step": 7592
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6863416530407801,
+      "learning_rate": 1.931872098444931e-05,
+      "loss": 1.9604,
+      "step": 7593
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6221475829163491,
+      "learning_rate": 1.9312637894951824e-05,
+      "loss": 2.0881,
+      "step": 7594
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.658100105169115,
+      "learning_rate": 1.9306555160539726e-05,
+      "loss": 1.8559,
+      "step": 7595
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6787245946516081,
+      "learning_rate": 1.9300472781592767e-05,
+      "loss": 1.8732,
+      "step": 7596
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5583594449455466,
+      "learning_rate": 1.9294390758490715e-05,
+      "loss": 1.8852,
+      "step": 7597
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.633676691118613,
+      "learning_rate": 1.928830909161329e-05,
+      "loss": 2.1102,
+      "step": 7598
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6366774170746637,
+      "learning_rate": 1.928222778134018e-05,
+      "loss": 1.9181,
+      "step": 7599
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5583888740415418,
+      "learning_rate": 1.927614682805109e-05,
+      "loss": 1.8672,
+      "step": 7600
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6867549589236431,
+      "learning_rate": 1.9270066232125663e-05,
+      "loss": 1.9151,
+      "step": 7601
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6007781558950073,
+      "learning_rate": 1.9263985993943537e-05,
+      "loss": 2.1635,
+      "step": 7602
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5531207100722589,
+      "learning_rate": 1.925790611388434e-05,
+      "loss": 1.879,
+      "step": 7603
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.611486952598035,
+      "learning_rate": 1.925182659232764e-05,
+      "loss": 1.9385,
+      "step": 7604
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6065691808500147,
+      "learning_rate": 1.9245747429653037e-05,
+      "loss": 1.9616,
+      "step": 7605
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6324305611315473,
+      "learning_rate": 1.9239668626240044e-05,
+      "loss": 2.1589,
+      "step": 7606
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6595277076650461,
+      "learning_rate": 1.923359018246821e-05,
+      "loss": 1.8229,
+      "step": 7607
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6920637110195723,
+      "learning_rate": 1.9227512098717036e-05,
+      "loss": 1.8776,
+      "step": 7608
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6464730782806439,
+      "learning_rate": 1.9221434375365983e-05,
+      "loss": 1.9045,
+      "step": 7609
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.651696193506599,
+      "learning_rate": 1.9215357012794532e-05,
+      "loss": 2.1166,
+      "step": 7610
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5421532025738192,
+      "learning_rate": 1.92092800113821e-05,
+      "loss": 2.0113,
+      "step": 7611
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5955992891345671,
+      "learning_rate": 1.9203203371508106e-05,
+      "loss": 1.8837,
+      "step": 7612
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6563145396795834,
+      "learning_rate": 1.9197127093551946e-05,
+      "loss": 1.863,
+      "step": 7613
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7051864180143707,
+      "learning_rate": 1.9191051177892968e-05,
+      "loss": 2.0835,
+      "step": 7614
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5824176436838076,
+      "learning_rate": 1.9184975624910532e-05,
+      "loss": 1.9472,
+      "step": 7615
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6618173573657026,
+      "learning_rate": 1.917890043498397e-05,
+      "loss": 1.8721,
+      "step": 7616
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7768219192627006,
+      "learning_rate": 1.917282560849255e-05,
+      "loss": 1.9269,
+      "step": 7617
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6594454665436558,
+      "learning_rate": 1.916675114581558e-05,
+      "loss": 2.1636,
+      "step": 7618
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7607844642340414,
+      "learning_rate": 1.9160677047332294e-05,
+      "loss": 1.8994,
+      "step": 7619
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5839653610549591,
+      "learning_rate": 1.915460331342192e-05,
+      "loss": 1.9218,
+      "step": 7620
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6375672953904478,
+      "learning_rate": 1.914852994446369e-05,
+      "loss": 1.8676,
+      "step": 7621
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6728387487782017,
+      "learning_rate": 1.914245694083677e-05,
+      "loss": 2.0477,
+      "step": 7622
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5714255795539871,
+      "learning_rate": 1.9136384302920312e-05,
+      "loss": 1.8571,
+      "step": 7623
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.706055256844909,
+      "learning_rate": 1.913031203109349e-05,
+      "loss": 2.0368,
+      "step": 7624
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.603293087212047,
+      "learning_rate": 1.9124240125735392e-05,
+      "loss": 1.8949,
+      "step": 7625
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7114611120341588,
+      "learning_rate": 1.9118168587225125e-05,
+      "loss": 2.0529,
+      "step": 7626
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5596024286181308,
+      "learning_rate": 1.9112097415941747e-05,
+      "loss": 1.8863,
+      "step": 7627
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7051319325984238,
+      "learning_rate": 1.9106026612264316e-05,
+      "loss": 1.8772,
+      "step": 7628
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7242464445597571,
+      "learning_rate": 1.9099956176571868e-05,
+      "loss": 1.9038,
+      "step": 7629
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5763240749728643,
+      "learning_rate": 1.9093886109243386e-05,
+      "loss": 1.9214,
+      "step": 7630
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.8007476587952197,
+      "learning_rate": 1.9087816410657867e-05,
+      "loss": 2.0672,
+      "step": 7631
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6473781147589445,
+      "learning_rate": 1.9081747081194243e-05,
+      "loss": 1.8783,
+      "step": 7632
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6372584365790553,
+      "learning_rate": 1.9075678121231466e-05,
+      "loss": 1.9152,
+      "step": 7633
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5798538309930465,
+      "learning_rate": 1.906960953114845e-05,
+      "loss": 2.0876,
+      "step": 7634
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6779708767780566,
+      "learning_rate": 1.9063541311324063e-05,
+      "loss": 1.9284,
+      "step": 7635
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5876085668122086,
+      "learning_rate": 1.9057473462137183e-05,
+      "loss": 2.0232,
+      "step": 7636
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6045640124363295,
+      "learning_rate": 1.9051405983966654e-05,
+      "loss": 1.9048,
+      "step": 7637
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6803396659912817,
+      "learning_rate": 1.9045338877191273e-05,
+      "loss": 2.0784,
+      "step": 7638
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6129913206714195,
+      "learning_rate": 1.903927214218986e-05,
+      "loss": 1.8991,
+      "step": 7639
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6158775224672377,
+      "learning_rate": 1.9033205779341174e-05,
+      "loss": 1.8946,
+      "step": 7640
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6578830794031483,
+      "learning_rate": 1.902713978902395e-05,
+      "loss": 1.8633,
+      "step": 7641
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6716137483931451,
+      "learning_rate": 1.9021074171616944e-05,
+      "loss": 1.9555,
+      "step": 7642
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6002736563887744,
+      "learning_rate": 1.9015008927498826e-05,
+      "loss": 2.0767,
+      "step": 7643
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.637040197673345,
+      "learning_rate": 1.9008944057048303e-05,
+      "loss": 1.9329,
+      "step": 7644
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5944808095142273,
+      "learning_rate": 1.9002879560644e-05,
+      "loss": 1.9367,
+      "step": 7645
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6928997993598467,
+      "learning_rate": 1.8996815438664562e-05,
+      "loss": 2.1515,
+      "step": 7646
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5642027716309448,
+      "learning_rate": 1.8990751691488605e-05,
+      "loss": 1.8794,
+      "step": 7647
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5426371725645626,
+      "learning_rate": 1.8984688319494707e-05,
+      "loss": 1.8746,
+      "step": 7648
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5755880918336511,
+      "learning_rate": 1.8978625323061422e-05,
+      "loss": 2.025,
+      "step": 7649
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5959507840409963,
+      "learning_rate": 1.8972562702567305e-05,
+      "loss": 1.8835,
+      "step": 7650
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5611507187399206,
+      "learning_rate": 1.896650045839085e-05,
+      "loss": 2.1309,
+      "step": 7651
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5395980553674166,
+      "learning_rate": 1.8960438590910567e-05,
+      "loss": 1.8881,
+      "step": 7652
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5809652481476495,
+      "learning_rate": 1.8954377100504904e-05,
+      "loss": 1.8932,
+      "step": 7653
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.595966721291081,
+      "learning_rate": 1.894831598755232e-05,
+      "loss": 2.1252,
+      "step": 7654
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5735137991161089,
+      "learning_rate": 1.894225525243123e-05,
+      "loss": 1.9248,
+      "step": 7655
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5451566076934459,
+      "learning_rate": 1.8936194895520027e-05,
+      "loss": 1.8523,
+      "step": 7656
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5920669946347059,
+      "learning_rate": 1.893013491719709e-05,
+      "loss": 1.8835,
+      "step": 7657
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5737393704937246,
+      "learning_rate": 1.892407531784077e-05,
+      "loss": 2.1175,
+      "step": 7658
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.650125666799056,
+      "learning_rate": 1.8918016097829376e-05,
+      "loss": 1.864,
+      "step": 7659
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5931904915324301,
+      "learning_rate": 1.8911957257541235e-05,
+      "loss": 1.905,
+      "step": 7660
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5930247996829384,
+      "learning_rate": 1.8905898797354603e-05,
+      "loss": 1.9136,
+      "step": 7661
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5787902606841083,
+      "learning_rate": 1.8899840717647747e-05,
+      "loss": 1.8665,
+      "step": 7662
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5587285463762074,
+      "learning_rate": 1.8893783018798906e-05,
+      "loss": 2.0741,
+      "step": 7663
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6440621057514645,
+      "learning_rate": 1.888772570118626e-05,
+      "loss": 1.913,
+      "step": 7664
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6042225792984329,
+      "learning_rate": 1.8881668765188024e-05,
+      "loss": 1.8945,
+      "step": 7665
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6284955078525634,
+      "learning_rate": 1.887561221118233e-05,
+      "loss": 2.1106,
+      "step": 7666
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5557327841831808,
+      "learning_rate": 1.8869556039547327e-05,
+      "loss": 2.0176,
+      "step": 7667
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5555472631674698,
+      "learning_rate": 1.8863500250661132e-05,
+      "loss": 1.9302,
+      "step": 7668
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6076933169783432,
+      "learning_rate": 1.8857444844901816e-05,
+      "loss": 1.9055,
+      "step": 7669
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6086386264331047,
+      "learning_rate": 1.8851389822647452e-05,
+      "loss": 2.0957,
+      "step": 7670
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.589318693426939,
+      "learning_rate": 1.8845335184276093e-05,
+      "loss": 1.8835,
+      "step": 7671
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6233262412149501,
+      "learning_rate": 1.8839280930165734e-05,
+      "loss": 1.8331,
+      "step": 7672
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.578148119549877,
+      "learning_rate": 1.883322706069438e-05,
+      "loss": 2.0009,
+      "step": 7673
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6096090856143407,
+      "learning_rate": 1.8827173576239988e-05,
+      "loss": 1.9099,
+      "step": 7674
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6330269737034272,
+      "learning_rate": 1.8821120477180503e-05,
+      "loss": 2.0796,
+      "step": 7675
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6019862947484418,
+      "learning_rate": 1.881506776389387e-05,
+      "loss": 1.947,
+      "step": 7676
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6449003333759745,
+      "learning_rate": 1.880901543675795e-05,
+      "loss": 1.9347,
+      "step": 7677
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7011496248020106,
+      "learning_rate": 1.8802963496150635e-05,
+      "loss": 2.0848,
+      "step": 7678
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6035376871469559,
+      "learning_rate": 1.8796911942449758e-05,
+      "loss": 1.9058,
+      "step": 7679
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6177581435147302,
+      "learning_rate": 1.879086077603316e-05,
+      "loss": 1.979,
+      "step": 7680
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.631416139822634,
+      "learning_rate": 1.8784809997278633e-05,
+      "loss": 1.906,
+      "step": 7681
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5876923499469475,
+      "learning_rate": 1.877875960656394e-05,
+      "loss": 1.8442,
+      "step": 7682
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5814283022024014,
+      "learning_rate": 1.877270960426685e-05,
+      "loss": 2.0722,
+      "step": 7683
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6296999651721421,
+      "learning_rate": 1.8766659990765083e-05,
+      "loss": 1.8841,
+      "step": 7684
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6039020759272172,
+      "learning_rate": 1.8760610766436335e-05,
+      "loss": 1.8647,
+      "step": 7685
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.591855462939525,
+      "learning_rate": 1.8754561931658296e-05,
+      "loss": 1.98,
+      "step": 7686
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5790357429639996,
+      "learning_rate": 1.874851348680861e-05,
+      "loss": 2.1164,
+      "step": 7687
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6119618387567786,
+      "learning_rate": 1.87424654322649e-05,
+      "loss": 1.8756,
+      "step": 7688
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6679472949345356,
+      "learning_rate": 1.8736417768404792e-05,
+      "loss": 1.9204,
+      "step": 7689
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5625197921548905,
+      "learning_rate": 1.873037049560585e-05,
+      "loss": 2.0564,
+      "step": 7690
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7021173541041683,
+      "learning_rate": 1.8724323614245626e-05,
+      "loss": 1.8542,
+      "step": 7691
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6189479197917626,
+      "learning_rate": 1.871827712470167e-05,
+      "loss": 1.9473,
+      "step": 7692
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6044790573714799,
+      "learning_rate": 1.8712231027351475e-05,
+      "loss": 1.8695,
+      "step": 7693
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6365205537043823,
+      "learning_rate": 1.870618532257253e-05,
+      "loss": 1.956,
+      "step": 7694
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7221952991466091,
+      "learning_rate": 1.870014001074229e-05,
+      "loss": 2.0779,
+      "step": 7695
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6370865901904805,
+      "learning_rate": 1.8694095092238186e-05,
+      "loss": 1.8588,
+      "step": 7696
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7765337784512784,
+      "learning_rate": 1.868805056743764e-05,
+      "loss": 1.8748,
+      "step": 7697
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5785412881356533,
+      "learning_rate": 1.868200643671802e-05,
+      "loss": 1.9235,
+      "step": 7698
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7833795339311975,
+      "learning_rate": 1.8675962700456696e-05,
+      "loss": 2.0906,
+      "step": 7699
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7302896496016024,
+      "learning_rate": 1.8669919359030996e-05,
+      "loss": 1.924,
+      "step": 7700
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6321616853204284,
+      "learning_rate": 1.8663876412818238e-05,
+      "loss": 1.9304,
+      "step": 7701
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6107288873436914,
+      "learning_rate": 1.8657833862195708e-05,
+      "loss": 2.1407,
+      "step": 7702
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6918899882460737,
+      "learning_rate": 1.8651791707540656e-05,
+      "loss": 1.9031,
+      "step": 7703
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5983769026275159,
+      "learning_rate": 1.864574994923033e-05,
+      "loss": 1.9669,
+      "step": 7704
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6237951129478277,
+      "learning_rate": 1.8639708587641947e-05,
+      "loss": 1.9017,
+      "step": 7705
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6324504575700634,
+      "learning_rate": 1.8633667623152673e-05,
+      "loss": 1.8928,
+      "step": 7706
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6281415625933285,
+      "learning_rate": 1.8627627056139692e-05,
+      "loss": 2.1019,
+      "step": 7707
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6614321870183782,
+      "learning_rate": 1.862158688698013e-05,
+      "loss": 1.8833,
+      "step": 7708
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6280023889953259,
+      "learning_rate": 1.8615547116051097e-05,
+      "loss": 1.8863,
+      "step": 7709
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6248936056974818,
+      "learning_rate": 1.8609507743729694e-05,
+      "loss": 2.083,
+      "step": 7710
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6450804657683494,
+      "learning_rate": 1.8603468770392964e-05,
+      "loss": 1.9482,
+      "step": 7711
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6406340177667768,
+      "learning_rate": 1.859743019641797e-05,
+      "loss": 1.897,
+      "step": 7712
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6726984593177419,
+      "learning_rate": 1.8591392022181703e-05,
+      "loss": 1.8997,
+      "step": 7713
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6072362926928292,
+      "learning_rate": 1.8585354248061155e-05,
+      "loss": 1.8714,
+      "step": 7714
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5853056436290325,
+      "learning_rate": 1.8579316874433305e-05,
+      "loss": 2.0732,
+      "step": 7715
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5740760803316709,
+      "learning_rate": 1.8573279901675073e-05,
+      "loss": 1.866,
+      "step": 7716
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5924569794105321,
+      "learning_rate": 1.856724333016337e-05,
+      "loss": 1.9407,
+      "step": 7717
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5658751626210109,
+      "learning_rate": 1.8561207160275107e-05,
+      "loss": 1.9026,
+      "step": 7718
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6292890968869956,
+      "learning_rate": 1.8555171392387127e-05,
+      "loss": 2.0437,
+      "step": 7719
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5884278253519223,
+      "learning_rate": 1.8549136026876278e-05,
+      "loss": 1.8897,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5941294603155431,
+      "learning_rate": 1.8543101064119355e-05,
+      "loss": 1.9087,
+      "step": 7721
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6409276571564958,
+      "learning_rate": 1.853706650449317e-05,
+      "loss": 2.0555,
+      "step": 7722
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6044054904189281,
+      "learning_rate": 1.853103234837447e-05,
+      "loss": 1.9716,
+      "step": 7723
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6079388183732213,
+      "learning_rate": 1.8524998596139987e-05,
+      "loss": 1.914,
+      "step": 7724
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6633073225337965,
+      "learning_rate": 1.8518965248166455e-05,
+      "loss": 1.8744,
+      "step": 7725
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6743036103359565,
+      "learning_rate": 1.8512932304830537e-05,
+      "loss": 1.9096,
+      "step": 7726
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5814667760419157,
+      "learning_rate": 1.8506899766508908e-05,
+      "loss": 2.1345,
+      "step": 7727
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5985511854595066,
+      "learning_rate": 1.8500867633578206e-05,
+      "loss": 1.8968,
+      "step": 7728
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6194039839198576,
+      "learning_rate": 1.8494835906415026e-05,
+      "loss": 1.9508,
+      "step": 7729
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.630587953349486,
+      "learning_rate": 1.848880458539597e-05,
+      "loss": 1.8826,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.666797202881205,
+      "learning_rate": 1.84827736708976e-05,
+      "loss": 2.0822,
+      "step": 7731
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5535212223626834,
+      "learning_rate": 1.847674316329643e-05,
+      "loss": 1.8889,
+      "step": 7732
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5835434616088623,
+      "learning_rate": 1.8470713062968993e-05,
+      "loss": 1.8848,
+      "step": 7733
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6145188838056859,
+      "learning_rate": 1.846468337029176e-05,
+      "loss": 2.0532,
+      "step": 7734
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5934392537100782,
+      "learning_rate": 1.8458654085641185e-05,
+      "loss": 1.9976,
+      "step": 7735
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.617151614941236,
+      "learning_rate": 1.8452625209393716e-05,
+      "loss": 1.8462,
+      "step": 7736
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6344854068682579,
+      "learning_rate": 1.844659674192575e-05,
+      "loss": 1.9037,
+      "step": 7737
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6264861498899883,
+      "learning_rate": 1.8440568683613668e-05,
+      "loss": 1.8669,
+      "step": 7738
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5723406710105847,
+      "learning_rate": 1.8434541034833837e-05,
+      "loss": 2.1105,
+      "step": 7739
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5758895671529257,
+      "learning_rate": 1.8428513795962577e-05,
+      "loss": 1.9048,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5647157800545446,
+      "learning_rate": 1.84224869673762e-05,
+      "loss": 1.8653,
+      "step": 7741
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5416794166380693,
+      "learning_rate": 1.8416460549450977e-05,
+      "loss": 2.0007,
+      "step": 7742
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5715399729988719,
+      "learning_rate": 1.8410434542563162e-05,
+      "loss": 2.0936,
+      "step": 7743
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6038103633140979,
+      "learning_rate": 1.8404408947089004e-05,
+      "loss": 1.8245,
+      "step": 7744
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6865303338332547,
+      "learning_rate": 1.8398383763404683e-05,
+      "loss": 1.8822,
+      "step": 7745
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5997130829296533,
+      "learning_rate": 1.839235899188639e-05,
+      "loss": 1.8675,
+      "step": 7746
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.595432684154704,
+      "learning_rate": 1.838633463291026e-05,
+      "loss": 2.0342,
+      "step": 7747
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.544027924728637,
+      "learning_rate": 1.8380310686852435e-05,
+      "loss": 1.9624,
+      "step": 7748
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6739226841857188,
+      "learning_rate": 1.837428715408901e-05,
+      "loss": 1.9288,
+      "step": 7749
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6483018252734095,
+      "learning_rate": 1.8368264034996047e-05,
+      "loss": 1.8997,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.598280320278809,
+      "learning_rate": 1.836224132994961e-05,
+      "loss": 2.1073,
+      "step": 7751
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6217780171334572,
+      "learning_rate": 1.835621903932572e-05,
+      "loss": 1.8639,
+      "step": 7752
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6310027601285159,
+      "learning_rate": 1.8350197163500362e-05,
+      "loss": 1.8839,
+      "step": 7753
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7017178980785458,
+      "learning_rate": 1.8344175702849523e-05,
+      "loss": 1.9401,
+      "step": 7754
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7024779219891703,
+      "learning_rate": 1.8338154657749128e-05,
+      "loss": 2.0857,
+      "step": 7755
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6629654364110027,
+      "learning_rate": 1.8332134028575104e-05,
+      "loss": 1.8908,
+      "step": 7756
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6335287202610022,
+      "learning_rate": 1.8326113815703356e-05,
+      "loss": 1.8859,
+      "step": 7757
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7035097575842879,
+      "learning_rate": 1.8320094019509725e-05,
+      "loss": 1.9079,
+      "step": 7758
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6674914708528802,
+      "learning_rate": 1.8314074640370078e-05,
+      "loss": 2.1142,
+      "step": 7759
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7532304095738104,
+      "learning_rate": 1.8308055678660214e-05,
+      "loss": 1.9745,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5922029710105496,
+      "learning_rate": 1.830203713475592e-05,
+      "loss": 1.9778,
+      "step": 7761
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7977118274274759,
+      "learning_rate": 1.829601900903297e-05,
+      "loss": 1.9227,
+      "step": 7762
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6873866370585509,
+      "learning_rate": 1.8290001301867095e-05,
+      "loss": 2.0672,
+      "step": 7763
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6362474355501763,
+      "learning_rate": 1.8283984013633994e-05,
+      "loss": 1.8277,
+      "step": 7764
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7595118133481907,
+      "learning_rate": 1.8277967144709375e-05,
+      "loss": 1.8891,
+      "step": 7765
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6614081217776412,
+      "learning_rate": 1.8271950695468877e-05,
+      "loss": 1.8994,
+      "step": 7766
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8510539012794145,
+      "learning_rate": 1.8265934666288138e-05,
+      "loss": 2.1158,
+      "step": 7767
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6470941999315772,
+      "learning_rate": 1.8259919057542753e-05,
+      "loss": 1.8989,
+      "step": 7768
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7667160417145628,
+      "learning_rate": 1.8253903869608313e-05,
+      "loss": 1.9311,
+      "step": 7769
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.61583410037006,
+      "learning_rate": 1.8247889102860376e-05,
+      "loss": 1.8668,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6115283995877181,
+      "learning_rate": 1.8241874757674452e-05,
+      "loss": 2.0583,
+      "step": 7771
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6218731402090237,
+      "learning_rate": 1.8235860834426054e-05,
+      "loss": 1.8667,
+      "step": 7772
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6225194025522878,
+      "learning_rate": 1.8229847333490656e-05,
+      "loss": 1.9771,
+      "step": 7773
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.585509736846809,
+      "learning_rate": 1.8223834255243695e-05,
+      "loss": 1.8606,
+      "step": 7774
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6233976512524234,
+      "learning_rate": 1.8217821600060604e-05,
+      "loss": 2.1005,
+      "step": 7775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6346375244118908,
+      "learning_rate": 1.821180936831676e-05,
+      "loss": 1.883,
+      "step": 7776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5644265855740093,
+      "learning_rate": 1.8205797560387555e-05,
+      "loss": 1.8575,
+      "step": 7777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6023407771751983,
+      "learning_rate": 1.819978617664832e-05,
+      "loss": 1.9059,
+      "step": 7778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5831614747697381,
+      "learning_rate": 1.819377521747436e-05,
+      "loss": 2.1664,
+      "step": 7779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6005857545495962,
+      "learning_rate": 1.8187764683240987e-05,
+      "loss": 1.8854,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6008182808219541,
+      "learning_rate": 1.8181754574323446e-05,
+      "loss": 1.9133,
+      "step": 7781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6701638954489487,
+      "learning_rate": 1.817574489109697e-05,
+      "loss": 1.9047,
+      "step": 7782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7797914057030707,
+      "learning_rate": 1.8169735633936784e-05,
+      "loss": 2.0497,
+      "step": 7783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5536464968490604,
+      "learning_rate": 1.8163726803218062e-05,
+      "loss": 1.9119,
+      "step": 7784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6704797841778731,
+      "learning_rate": 1.815771839931595e-05,
+      "loss": 1.9655,
+      "step": 7785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6139415734454531,
+      "learning_rate": 1.8151710422605596e-05,
+      "loss": 1.8866,
+      "step": 7786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7212104729721152,
+      "learning_rate": 1.8145702873462093e-05,
+      "loss": 2.0949,
+      "step": 7787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6511792471052809,
+      "learning_rate": 1.8139695752260523e-05,
+      "loss": 1.8368,
+      "step": 7788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.590368059501848,
+      "learning_rate": 1.8133689059375914e-05,
+      "loss": 1.8924,
+      "step": 7789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7139687289831533,
+      "learning_rate": 1.812768279518331e-05,
+      "loss": 1.9371,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7018265757162682,
+      "learning_rate": 1.8121676960057707e-05,
+      "loss": 2.1112,
+      "step": 7791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7482878230500701,
+      "learning_rate": 1.8115671554374066e-05,
+      "loss": 1.9234,
+      "step": 7792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6287800205060552,
+      "learning_rate": 1.8109666578507336e-05,
+      "loss": 1.9281,
+      "step": 7793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7918995443383934,
+      "learning_rate": 1.810366203283242e-05,
+      "loss": 1.8884,
+      "step": 7794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6771759262240806,
+      "learning_rate": 1.8097657917724213e-05,
+      "loss": 2.0591,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6795670841674857,
+      "learning_rate": 1.8091654233557588e-05,
+      "loss": 1.8965,
+      "step": 7796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6368717744136054,
+      "learning_rate": 1.808565098070736e-05,
+      "loss": 1.9103,
+      "step": 7797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6647974750713925,
+      "learning_rate": 1.8079648159548346e-05,
+      "loss": 2.0173,
+      "step": 7798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6999442391893519,
+      "learning_rate": 1.807364577045534e-05,
+      "loss": 2.1013,
+      "step": 7799
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6073664548570264,
+      "learning_rate": 1.8067643813803066e-05,
+      "loss": 1.8972,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7299174686241575,
+      "learning_rate": 1.806164228996628e-05,
+      "loss": 1.9237,
+      "step": 7801
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6031002959850923,
+      "learning_rate": 1.8055641199319662e-05,
+      "loss": 1.8913,
+      "step": 7802
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.600973150473757,
+      "learning_rate": 1.804964054223789e-05,
+      "loss": 2.0941,
+      "step": 7803
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5593916640226199,
+      "learning_rate": 1.8043640319095618e-05,
+      "loss": 2.0043,
+      "step": 7804
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6171216055780611,
+      "learning_rate": 1.8037640530267458e-05,
+      "loss": 1.8854,
+      "step": 7805
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5691476928332655,
+      "learning_rate": 1.8031641176127993e-05,
+      "loss": 1.9638,
+      "step": 7806
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5861069403314976,
+      "learning_rate": 1.802564225705181e-05,
+      "loss": 2.091,
+      "step": 7807
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5431777712881763,
+      "learning_rate": 1.8019643773413418e-05,
+      "loss": 1.8843,
+      "step": 7808
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6019298883685359,
+      "learning_rate": 1.801364572558735e-05,
+      "loss": 1.8988,
+      "step": 7809
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5790242965063109,
+      "learning_rate": 1.800764811394808e-05,
+      "loss": 1.9872,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6369057263097551,
+      "learning_rate": 1.8001650938870056e-05,
+      "loss": 2.0637,
+      "step": 7811
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5517602621617457,
+      "learning_rate": 1.7995654200727725e-05,
+      "loss": 1.8532,
+      "step": 7812
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5596762157021434,
+      "learning_rate": 1.7989657899895475e-05,
+      "loss": 1.9205,
+      "step": 7813
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6250233531507094,
+      "learning_rate": 1.7983662036747682e-05,
+      "loss": 1.9066,
+      "step": 7814
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5848115747426713,
+      "learning_rate": 1.7977666611658682e-05,
+      "loss": 2.0706,
+      "step": 7815
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5774253218411696,
+      "learning_rate": 1.7971671625002813e-05,
+      "loss": 1.9222,
+      "step": 7816
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.680425238664995,
+      "learning_rate": 1.7965677077154358e-05,
+      "loss": 1.8846,
+      "step": 7817
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6590543752562206,
+      "learning_rate": 1.795968296848757e-05,
+      "loss": 1.921,
+      "step": 7818
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5760212269972022,
+      "learning_rate": 1.79536892993767e-05,
+      "loss": 2.1341,
+      "step": 7819
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6014094710801818,
+      "learning_rate": 1.794769607019597e-05,
+      "loss": 1.8809,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6076167046700441,
+      "learning_rate": 1.7941703281319526e-05,
+      "loss": 1.9012,
+      "step": 7821
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.565255626712296,
+      "learning_rate": 1.7935710933121553e-05,
+      "loss": 1.9993,
+      "step": 7822
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6269915996758877,
+      "learning_rate": 1.792971902597616e-05,
+      "loss": 2.1131,
+      "step": 7823
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5768757771115035,
+      "learning_rate": 1.7923727560257457e-05,
+      "loss": 1.9345,
+      "step": 7824
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6170705832434309,
+      "learning_rate": 1.791773653633952e-05,
+      "loss": 1.8921,
+      "step": 7825
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5969539826377342,
+      "learning_rate": 1.7911745954596373e-05,
+      "loss": 1.8605,
+      "step": 7826
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6066256125022826,
+      "learning_rate": 1.7905755815402058e-05,
+      "loss": 2.0766,
+      "step": 7827
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5944986377650323,
+      "learning_rate": 1.7899766119130547e-05,
+      "loss": 1.8571,
+      "step": 7828
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5858110077330112,
+      "learning_rate": 1.7893776866155796e-05,
+      "loss": 1.9526,
+      "step": 7829
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5677919074652281,
+      "learning_rate": 1.7887788056851763e-05,
+      "loss": 1.8803,
+      "step": 7830
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6300631390689946,
+      "learning_rate": 1.788179969159233e-05,
+      "loss": 2.0733,
+      "step": 7831
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5922216597867523,
+      "learning_rate": 1.787581177075138e-05,
+      "loss": 1.9638,
+      "step": 7832
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5736583336007127,
+      "learning_rate": 1.786982429470278e-05,
+      "loss": 1.8957,
+      "step": 7833
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6553455345654143,
+      "learning_rate": 1.786383726382034e-05,
+      "loss": 1.9086,
+      "step": 7834
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5445063604073471,
+      "learning_rate": 1.7857850678477853e-05,
+      "loss": 2.1732,
+      "step": 7835
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5949209908124793,
+      "learning_rate": 1.785186453904908e-05,
+      "loss": 1.891,
+      "step": 7836
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.601220615707868,
+      "learning_rate": 1.7845878845907777e-05,
+      "loss": 1.947,
+      "step": 7837
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5717188789287955,
+      "learning_rate": 1.7839893599427655e-05,
+      "loss": 1.919,
+      "step": 7838
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6157986394718005,
+      "learning_rate": 1.7833908799982375e-05,
+      "loss": 2.0781,
+      "step": 7839
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5897577683343103,
+      "learning_rate": 1.7827924447945607e-05,
+      "loss": 1.8707,
+      "step": 7840
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.544929613959994,
+      "learning_rate": 1.782194054369099e-05,
+      "loss": 1.9656,
+      "step": 7841
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6379077735656136,
+      "learning_rate": 1.781595708759211e-05,
+      "loss": 1.8916,
+      "step": 7842
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6000366617506121,
+      "learning_rate": 1.780997408002255e-05,
+      "loss": 2.1222,
+      "step": 7843
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6179728577898299,
+      "learning_rate": 1.7803991521355833e-05,
+      "loss": 1.9332,
+      "step": 7844
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6523102133110338,
+      "learning_rate": 1.7798009411965488e-05,
+      "loss": 1.8804,
+      "step": 7845
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7151860094893757,
+      "learning_rate": 1.779202775222501e-05,
+      "loss": 1.8776,
+      "step": 7846
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6217671189010079,
+      "learning_rate": 1.7786046542507843e-05,
+      "loss": 1.9642,
+      "step": 7847
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6190687440613986,
+      "learning_rate": 1.7780065783187433e-05,
+      "loss": 2.0869,
+      "step": 7848
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5916595747307961,
+      "learning_rate": 1.7774085474637174e-05,
+      "loss": 1.8928,
+      "step": 7849
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6289275709568086,
+      "learning_rate": 1.7768105617230436e-05,
+      "loss": 1.8397,
+      "step": 7850
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6002033890970713,
+      "learning_rate": 1.7762126211340584e-05,
+      "loss": 2.1026,
+      "step": 7851
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6448147130258309,
+      "learning_rate": 1.775614725734092e-05,
+      "loss": 1.8848,
+      "step": 7852
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.555557086988092,
+      "learning_rate": 1.775016875560474e-05,
+      "loss": 1.9457,
+      "step": 7853
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6107464537399255,
+      "learning_rate": 1.7744190706505314e-05,
+      "loss": 1.8869,
+      "step": 7854
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5869432236206674,
+      "learning_rate": 1.7738213110415862e-05,
+      "loss": 2.0733,
+      "step": 7855
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6157146054263752,
+      "learning_rate": 1.773223596770961e-05,
+      "loss": 1.9056,
+      "step": 7856
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6584867691868652,
+      "learning_rate": 1.7726259278759716e-05,
+      "loss": 1.9017,
+      "step": 7857
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6264213813329534,
+      "learning_rate": 1.772028304393933e-05,
+      "loss": 1.8453,
+      "step": 7858
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7118522151676001,
+      "learning_rate": 1.771430726362159e-05,
+      "loss": 2.1332,
+      "step": 7859
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5368481785471216,
+      "learning_rate": 1.7708331938179573e-05,
+      "loss": 2.0009,
+      "step": 7860
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5710401867510809,
+      "learning_rate": 1.770235706798635e-05,
+      "loss": 1.8649,
+      "step": 7861
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6109026683273162,
+      "learning_rate": 1.7696382653414954e-05,
+      "loss": 1.9261,
+      "step": 7862
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5869656608413821,
+      "learning_rate": 1.769040869483839e-05,
+      "loss": 2.1461,
+      "step": 7863
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6068625502702995,
+      "learning_rate": 1.7684435192629646e-05,
+      "loss": 1.9137,
+      "step": 7864
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5809425731552649,
+      "learning_rate": 1.767846214716166e-05,
+      "loss": 1.8826,
+      "step": 7865
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5677613606869302,
+      "learning_rate": 1.7672489558807358e-05,
+      "loss": 1.9327,
+      "step": 7866
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6166180087456454,
+      "learning_rate": 1.7666517427939647e-05,
+      "loss": 1.8603,
+      "step": 7867
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5764775943284338,
+      "learning_rate": 1.7660545754931366e-05,
+      "loss": 2.0908,
+      "step": 7868
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5927651333936975,
+      "learning_rate": 1.7654574540155375e-05,
+      "loss": 1.8477,
+      "step": 7869
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6216857035616014,
+      "learning_rate": 1.7648603783984467e-05,
+      "loss": 1.9295,
+      "step": 7870
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.59652069168757,
+      "learning_rate": 1.7642633486791417e-05,
+      "loss": 2.1204,
+      "step": 7871
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5288864603747875,
+      "learning_rate": 1.7636663648949e-05,
+      "loss": 2.0155,
+      "step": 7872
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5794646324459539,
+      "learning_rate": 1.7630694270829902e-05,
+      "loss": 1.9097,
+      "step": 7873
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.607118808374081,
+      "learning_rate": 1.762472535280685e-05,
+      "loss": 1.8734,
+      "step": 7874
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.639673190950008,
+      "learning_rate": 1.7618756895252488e-05,
+      "loss": 2.0686,
+      "step": 7875
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5895451515529958,
+      "learning_rate": 1.7612788898539448e-05,
+      "loss": 1.8973,
+      "step": 7876
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6220253777305389,
+      "learning_rate": 1.7606821363040354e-05,
+      "loss": 1.9109,
+      "step": 7877
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7342770980332243,
+      "learning_rate": 1.760085428912777e-05,
+      "loss": 1.9813,
+      "step": 7878
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5549089337712442,
+      "learning_rate": 1.759488767717425e-05,
+      "loss": 1.8884,
+      "step": 7879
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7152730103099663,
+      "learning_rate": 1.7588921527552317e-05,
+      "loss": 2.0773,
+      "step": 7880
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6943605257457981,
+      "learning_rate": 1.7582955840634458e-05,
+      "loss": 1.852,
+      "step": 7881
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6611591251877922,
+      "learning_rate": 1.7576990616793137e-05,
+      "loss": 1.8504,
+      "step": 7882
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5935913874892953,
+      "learning_rate": 1.7571025856400776e-05,
+      "loss": 2.0245,
+      "step": 7883
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.684131538217943,
+      "learning_rate": 1.7565061559829796e-05,
+      "loss": 1.9444,
+      "step": 7884
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6948814435617148,
+      "learning_rate": 1.7559097727452574e-05,
+      "loss": 1.8492,
+      "step": 7885
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6375461746237897,
+      "learning_rate": 1.7553134359641434e-05,
+      "loss": 1.8901,
+      "step": 7886
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7377319021951324,
+      "learning_rate": 1.754717145676871e-05,
+      "loss": 2.1231,
+      "step": 7887
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6110164086481045,
+      "learning_rate": 1.75412090192067e-05,
+      "loss": 1.8822,
+      "step": 7888
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5990784653626101,
+      "learning_rate": 1.7535247047327647e-05,
+      "loss": 1.8701,
+      "step": 7889
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6213652946377782,
+      "learning_rate": 1.7529285541503793e-05,
+      "loss": 1.9267,
+      "step": 7890
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6712567975216626,
+      "learning_rate": 1.7523324502107322e-05,
+      "loss": 1.9811,
+      "step": 7891
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6564395912470025,
+      "learning_rate": 1.751736392951042e-05,
+      "loss": 2.0606,
+      "step": 7892
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6159426270957444,
+      "learning_rate": 1.7511403824085237e-05,
+      "loss": 1.8315,
+      "step": 7893
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.78930003480808,
+      "learning_rate": 1.7505444186203863e-05,
+      "loss": 1.8509,
+      "step": 7894
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6570447162742777,
+      "learning_rate": 1.749948501623841e-05,
+      "loss": 2.063,
+      "step": 7895
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5827395437317916,
+      "learning_rate": 1.7493526314560914e-05,
+      "loss": 1.8527,
+      "step": 7896
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6478877776408313,
+      "learning_rate": 1.7487568081543402e-05,
+      "loss": 1.9804,
+      "step": 7897
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6965102124717057,
+      "learning_rate": 1.7481610317557886e-05,
+      "loss": 1.907,
+      "step": 7898
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5703523088545954,
+      "learning_rate": 1.7475653022976322e-05,
+      "loss": 1.8926,
+      "step": 7899
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7261274965392124,
+      "learning_rate": 1.7469696198170645e-05,
+      "loss": 2.0773,
+      "step": 7900
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5916676651136727,
+      "learning_rate": 1.7463739843512777e-05,
+      "loss": 1.9314,
+      "step": 7901
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6047730800406996,
+      "learning_rate": 1.7457783959374586e-05,
+      "loss": 1.9088,
+      "step": 7902
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6273730038263792,
+      "learning_rate": 1.7451828546127933e-05,
+      "loss": 2.0144,
+      "step": 7903
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6265284706347284,
+      "learning_rate": 1.7445873604144618e-05,
+      "loss": 2.0741,
+      "step": 7904
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5870418695440645,
+      "learning_rate": 1.7439919133796454e-05,
+      "loss": 1.8845,
+      "step": 7905
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5911542426237453,
+      "learning_rate": 1.7433965135455205e-05,
+      "loss": 1.8938,
+      "step": 7906
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7172121942309946,
+      "learning_rate": 1.7428011609492583e-05,
+      "loss": 2.0289,
+      "step": 7907
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6178258872050997,
+      "learning_rate": 1.742205855628031e-05,
+      "loss": 1.9205,
+      "step": 7908
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5437219516372771,
+      "learning_rate": 1.741610597619005e-05,
+      "loss": 1.9571,
+      "step": 7909
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5808736437971359,
+      "learning_rate": 1.741015386959345e-05,
+      "loss": 1.8994,
+      "step": 7910
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6386257751831419,
+      "learning_rate": 1.7404202236862126e-05,
+      "loss": 1.908,
+      "step": 7911
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5856127589847484,
+      "learning_rate": 1.7398251078367653e-05,
+      "loss": 2.1146,
+      "step": 7912
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5849799831314236,
+      "learning_rate": 1.7392300394481597e-05,
+      "loss": 1.8754,
+      "step": 7913
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.577959089254045,
+      "learning_rate": 1.738635018557549e-05,
+      "loss": 1.8759,
+      "step": 7914
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7037860917785987,
+      "learning_rate": 1.73804004520208e-05,
+      "loss": 2.1306,
+      "step": 7915
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5369241546013384,
+      "learning_rate": 1.737445119418903e-05,
+      "loss": 1.9128,
+      "step": 7916
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5684730169652269,
+      "learning_rate": 1.736850241245159e-05,
+      "loss": 1.8928,
+      "step": 7917
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6438222618279876,
+      "learning_rate": 1.736255410717989e-05,
+      "loss": 1.9077,
+      "step": 7918
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.589757951522892,
+      "learning_rate": 1.7356606278745325e-05,
+      "loss": 2.0778,
+      "step": 7919
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6387325451317934,
+      "learning_rate": 1.735065892751922e-05,
+      "loss": 1.9,
+      "step": 7920
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6508457076639893,
+      "learning_rate": 1.7344712053872904e-05,
+      "loss": 1.8971,
+      "step": 7921
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6313086783591261,
+      "learning_rate": 1.7338765658177673e-05,
+      "loss": 1.9701,
+      "step": 7922
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6351958365253333,
+      "learning_rate": 1.7332819740804763e-05,
+      "loss": 1.909,
+      "step": 7923
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5854426547907485,
+      "learning_rate": 1.7326874302125425e-05,
+      "loss": 2.088,
+      "step": 7924
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6417122455130803,
+      "learning_rate": 1.7320929342510835e-05,
+      "loss": 1.9338,
+      "step": 7925
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6150854322462879,
+      "learning_rate": 1.7314984862332174e-05,
+      "loss": 1.8561,
+      "step": 7926
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6194812651014531,
+      "learning_rate": 1.7309040861960588e-05,
+      "loss": 2.1163,
+      "step": 7927
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6255036604490646,
+      "learning_rate": 1.7303097341767168e-05,
+      "loss": 1.9524,
+      "step": 7928
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5913152906707704,
+      "learning_rate": 1.729715430212301e-05,
+      "loss": 1.8699,
+      "step": 7929
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.630252163603967,
+      "learning_rate": 1.7291211743399142e-05,
+      "loss": 1.8665,
+      "step": 7930
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6743948522811635,
+      "learning_rate": 1.7285269665966597e-05,
+      "loss": 1.8937,
+      "step": 7931
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5741099734215458,
+      "learning_rate": 1.7279328070196367e-05,
+      "loss": 2.0774,
+      "step": 7932
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5663571055188146,
+      "learning_rate": 1.7273386956459387e-05,
+      "loss": 1.8711,
+      "step": 7933
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5469398317891959,
+      "learning_rate": 1.7267446325126613e-05,
+      "loss": 1.984,
+      "step": 7934
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.590180910990213,
+      "learning_rate": 1.726150617656893e-05,
+      "loss": 1.9194,
+      "step": 7935
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6495951347388106,
+      "learning_rate": 1.7255566511157202e-05,
+      "loss": 2.1191,
+      "step": 7936
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5711526199658411,
+      "learning_rate": 1.724962732926228e-05,
+      "loss": 1.9093,
+      "step": 7937
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6701154678327149,
+      "learning_rate": 1.724368863125495e-05,
+      "loss": 1.8442,
+      "step": 7938
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6142836841447056,
+      "learning_rate": 1.723775041750601e-05,
+      "loss": 2.0674,
+      "step": 7939
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6181215189326719,
+      "learning_rate": 1.7231812688386207e-05,
+      "loss": 1.9722,
+      "step": 7940
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6913465182502596,
+      "learning_rate": 1.722587544426624e-05,
+      "loss": 1.9164,
+      "step": 7941
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5774474714587979,
+      "learning_rate": 1.7219938685516813e-05,
+      "loss": 1.8898,
+      "step": 7942
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6105863409077543,
+      "learning_rate": 1.721400241250857e-05,
+      "loss": 1.8607,
+      "step": 7943
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6597433325417092,
+      "learning_rate": 1.7208066625612137e-05,
+      "loss": 2.0587,
+      "step": 7944
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5939433443558272,
+      "learning_rate": 1.7202131325198128e-05,
+      "loss": 1.9571,
+      "step": 7945
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6551387871101169,
+      "learning_rate": 1.7196196511637084e-05,
+      "loss": 1.8808,
+      "step": 7946
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6203557296536555,
+      "learning_rate": 1.7190262185299553e-05,
+      "loss": 1.9544,
+      "step": 7947
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6123601245854662,
+      "learning_rate": 1.7184328346556043e-05,
+      "loss": 2.0858,
+      "step": 7948
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6121778463590674,
+      "learning_rate": 1.7178394995777017e-05,
+      "loss": 1.886,
+      "step": 7949
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6997189840662202,
+      "learning_rate": 1.7172462133332928e-05,
+      "loss": 1.8644,
+      "step": 7950
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6597529635170581,
+      "learning_rate": 1.7166529759594175e-05,
+      "loss": 1.9119,
+      "step": 7951
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6709672561238749,
+      "learning_rate": 1.716059787493115e-05,
+      "loss": 2.1265,
+      "step": 7952
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6120303691723041,
+      "learning_rate": 1.7154666479714217e-05,
+      "loss": 1.9662,
+      "step": 7953
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7134738882650948,
+      "learning_rate": 1.7148735574313678e-05,
+      "loss": 1.8967,
+      "step": 7954
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5841904401426483,
+      "learning_rate": 1.7142805159099823e-05,
+      "loss": 1.9011,
+      "step": 7955
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5456764169555162,
+      "learning_rate": 1.7136875234442935e-05,
+      "loss": 2.0628,
+      "step": 7956
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6893908449369878,
+      "learning_rate": 1.713094580071322e-05,
+      "loss": 1.8831,
+      "step": 7957
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5742596989805362,
+      "learning_rate": 1.712501685828089e-05,
+      "loss": 1.9134,
+      "step": 7958
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6545005024332411,
+      "learning_rate": 1.71190884075161e-05,
+      "loss": 1.9738,
+      "step": 7959
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6106907723159131,
+      "learning_rate": 1.7113160448789004e-05,
+      "loss": 2.0977,
+      "step": 7960
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6451038191553446,
+      "learning_rate": 1.7107232982469703e-05,
+      "loss": 1.9291,
+      "step": 7961
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7239972334341865,
+      "learning_rate": 1.7101306008928264e-05,
+      "loss": 1.9091,
+      "step": 7962
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.662887261971421,
+      "learning_rate": 1.7095379528534747e-05,
+      "loss": 1.8839,
+      "step": 7963
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6312263534688255,
+      "learning_rate": 1.7089453541659156e-05,
+      "loss": 2.0567,
+      "step": 7964
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7483024201679156,
+      "learning_rate": 1.7083528048671472e-05,
+      "loss": 1.945,
+      "step": 7965
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.635652201073088,
+      "learning_rate": 1.707760304994166e-05,
+      "loss": 1.8519,
+      "step": 7966
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.607762180081153,
+      "learning_rate": 1.7071678545839637e-05,
+      "loss": 1.9039,
+      "step": 7967
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7066561170774155,
+      "learning_rate": 1.706575453673528e-05,
+      "loss": 2.0635,
+      "step": 7968
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5958641381948798,
+      "learning_rate": 1.705983102299848e-05,
+      "loss": 1.8844,
+      "step": 7969
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5972990343197131,
+      "learning_rate": 1.7053908004999034e-05,
+      "loss": 1.9133,
+      "step": 7970
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6463145125872612,
+      "learning_rate": 1.7047985483106765e-05,
+      "loss": 1.946,
+      "step": 7971
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.588798646411515,
+      "learning_rate": 1.7042063457691427e-05,
+      "loss": 2.0574,
+      "step": 7972
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5766981243692957,
+      "learning_rate": 1.7036141929122752e-05,
+      "loss": 1.8695,
+      "step": 7973
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5941651588530273,
+      "learning_rate": 1.703022089777046e-05,
+      "loss": 1.9097,
+      "step": 7974
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5730527789485307,
+      "learning_rate": 1.702430036400422e-05,
+      "loss": 1.9164,
+      "step": 7975
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.631103891784438,
+      "learning_rate": 1.7018380328193674e-05,
+      "loss": 2.0435,
+      "step": 7976
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5534488734333493,
+      "learning_rate": 1.7012460790708425e-05,
+      "loss": 1.7841,
+      "step": 7977
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5734930696584916,
+      "learning_rate": 1.7006541751918065e-05,
+      "loss": 1.9679,
+      "step": 7978
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5548599008500483,
+      "learning_rate": 1.7000623212192145e-05,
+      "loss": 1.8742,
+      "step": 7979
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6245666297705111,
+      "learning_rate": 1.699470517190018e-05,
+      "loss": 2.0611,
+      "step": 7980
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.606544668125844,
+      "learning_rate": 1.698878763141165e-05,
+      "loss": 1.9305,
+      "step": 7981
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6158267457037985,
+      "learning_rate": 1.698287059109603e-05,
+      "loss": 1.9005,
+      "step": 7982
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.560299826039455,
+      "learning_rate": 1.6976954051322723e-05,
+      "loss": 1.84,
+      "step": 7983
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6148135511510494,
+      "learning_rate": 1.697103801246115e-05,
+      "loss": 2.1081,
+      "step": 7984
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5975577107629523,
+      "learning_rate": 1.696512247488064e-05,
+      "loss": 1.8782,
+      "step": 7985
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6683331777920812,
+      "learning_rate": 1.695920743895055e-05,
+      "loss": 1.9223,
+      "step": 7986
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6126261745148434,
+      "learning_rate": 1.6953292905040176e-05,
+      "loss": 1.8976,
+      "step": 7987
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6679495069078509,
+      "learning_rate": 1.6947378873518773e-05,
+      "loss": 2.0959,
+      "step": 7988
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6938483756843288,
+      "learning_rate": 1.69414653447556e-05,
+      "loss": 1.9325,
+      "step": 7989
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5716343155119477,
+      "learning_rate": 1.6935552319119842e-05,
+      "loss": 1.9796,
+      "step": 7990
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6483450979338703,
+      "learning_rate": 1.692963979698068e-05,
+      "loss": 1.9206,
+      "step": 7991
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5561805460373539,
+      "learning_rate": 1.692372777870727e-05,
+      "loss": 2.1099,
+      "step": 7992
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6956486869538951,
+      "learning_rate": 1.6917816264668705e-05,
+      "loss": 1.9271,
+      "step": 7993
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.602162545252122,
+      "learning_rate": 1.6911905255234073e-05,
+      "loss": 1.8839,
+      "step": 7994
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5901686064514177,
+      "learning_rate": 1.6905994750772427e-05,
+      "loss": 1.8695,
+      "step": 7995
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.569357300453312,
+      "learning_rate": 1.690008475165278e-05,
+      "loss": 2.127,
+      "step": 7996
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6560344160546161,
+      "learning_rate": 1.6894175258244116e-05,
+      "loss": 1.8329,
+      "step": 7997
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7627662379042516,
+      "learning_rate": 1.6888266270915386e-05,
+      "loss": 1.8988,
+      "step": 7998
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.597755388550354,
+      "learning_rate": 1.688235779003552e-05,
+      "loss": 1.9014,
+      "step": 7999
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6881618806615041,
+      "learning_rate": 1.687644981597341e-05,
+      "loss": 2.0975,
+      "step": 8000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7977816978091596,
+      "learning_rate": 1.6870542349097896e-05,
+      "loss": 1.8794,
+      "step": 8001
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6734383211323276,
+      "learning_rate": 1.6864635389777823e-05,
+      "loss": 1.9372,
+      "step": 8002
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7203124991084491,
+      "learning_rate": 1.6858728938381996e-05,
+      "loss": 1.8298,
+      "step": 8003
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6864009574616214,
+      "learning_rate": 1.6852822995279162e-05,
+      "loss": 2.1231,
+      "step": 8004
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6852291387443736,
+      "learning_rate": 1.684691756083806e-05,
+      "loss": 1.8848,
+      "step": 8005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6096937971114873,
+      "learning_rate": 1.684101263542738e-05,
+      "loss": 1.9214,
+      "step": 8006
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7029510562460212,
+      "learning_rate": 1.68351082194158e-05,
+      "loss": 1.8697,
+      "step": 8007
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7127286186459997,
+      "learning_rate": 1.6829204313171966e-05,
+      "loss": 2.123,
+      "step": 8008
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6243969119573524,
+      "learning_rate": 1.682330091706446e-05,
+      "loss": 1.9176,
+      "step": 8009
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7802181979540191,
+      "learning_rate": 1.6817398031461877e-05,
+      "loss": 1.8532,
+      "step": 8010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5591925316772424,
+      "learning_rate": 1.681149565673275e-05,
+      "loss": 1.836,
+      "step": 8011
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9567767408081158,
+      "learning_rate": 1.680559379324558e-05,
+      "loss": 2.0832,
+      "step": 8012
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5533084232426075,
+      "learning_rate": 1.6799692441368864e-05,
+      "loss": 1.8367,
+      "step": 8013
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7830512871304877,
+      "learning_rate": 1.6793791601471032e-05,
+      "loss": 1.8869,
+      "step": 8014
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.616983149810398,
+      "learning_rate": 1.67878912739205e-05,
+      "loss": 1.9416,
+      "step": 8015
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6995988982109247,
+      "learning_rate": 1.6781991459085655e-05,
+      "loss": 2.0572,
+      "step": 8016
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.75433762192975,
+      "learning_rate": 1.6776092157334837e-05,
+      "loss": 1.933,
+      "step": 8017
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6567393243415063,
+      "learning_rate": 1.6770193369036387e-05,
+      "loss": 1.8639,
+      "step": 8018
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7404922005921006,
+      "learning_rate": 1.6764295094558558e-05,
+      "loss": 1.8564,
+      "step": 8019
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6152472413509229,
+      "learning_rate": 1.6758397334269617e-05,
+      "loss": 2.0461,
+      "step": 8020
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6866267694510039,
+      "learning_rate": 1.6752500088537802e-05,
+      "loss": 1.9558,
+      "step": 8021
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6343017888169888,
+      "learning_rate": 1.674660335773128e-05,
+      "loss": 1.8898,
+      "step": 8022
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6819270641416761,
+      "learning_rate": 1.674070714221822e-05,
+      "loss": 1.917,
+      "step": 8023
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7417602449024082,
+      "learning_rate": 1.673481144236674e-05,
+      "loss": 2.0914,
+      "step": 8024
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5716904858969567,
+      "learning_rate": 1.6728916258544934e-05,
+      "loss": 1.9304,
+      "step": 8025
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.9571997724407122,
+      "learning_rate": 1.6723021591120868e-05,
+      "loss": 1.914,
+      "step": 8026
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6170397039361923,
+      "learning_rate": 1.671712744046256e-05,
+      "loss": 1.972,
+      "step": 8027
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7542187494579101,
+      "learning_rate": 1.671123380693802e-05,
+      "loss": 2.0995,
+      "step": 8028
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6364200626318585,
+      "learning_rate": 1.6705340690915204e-05,
+      "loss": 1.8761,
+      "step": 8029
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6798029863896007,
+      "learning_rate": 1.669944809276204e-05,
+      "loss": 1.9271,
+      "step": 8030
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6346467238584688,
+      "learning_rate": 1.669355601284644e-05,
+      "loss": 1.8897,
+      "step": 8031
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6258482406132627,
+      "learning_rate": 1.6687664451536257e-05,
+      "loss": 2.1138,
+      "step": 8032
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6937016290866295,
+      "learning_rate": 1.668177340919932e-05,
+      "loss": 1.9456,
+      "step": 8033
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6942954855108763,
+      "learning_rate": 1.6675882886203455e-05,
+      "loss": 1.9401,
+      "step": 8034
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6378264902439612,
+      "learning_rate": 1.6669992882916407e-05,
+      "loss": 1.9037,
+      "step": 8035
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7453561900100464,
+      "learning_rate": 1.666410339970593e-05,
+      "loss": 2.1065,
+      "step": 8036
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6114200967986942,
+      "learning_rate": 1.6658214436939727e-05,
+      "loss": 1.8579,
+      "step": 8037
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7159341784228856,
+      "learning_rate": 1.6652325994985457e-05,
+      "loss": 1.8801,
+      "step": 8038
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6965514550871307,
+      "learning_rate": 1.6646438074210773e-05,
+      "loss": 1.8748,
+      "step": 8039
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5231232297080328,
+      "learning_rate": 1.664055067498328e-05,
+      "loss": 2.1366,
+      "step": 8040
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7130533992530428,
+      "learning_rate": 1.663466379767054e-05,
+      "loss": 1.9333,
+      "step": 8041
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7032800803893571,
+      "learning_rate": 1.6628777442640116e-05,
+      "loss": 1.9024,
+      "step": 8042
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6120943397108941,
+      "learning_rate": 1.66228916102595e-05,
+      "loss": 1.8988,
+      "step": 8043
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6785806725400173,
+      "learning_rate": 1.6617006300896183e-05,
+      "loss": 2.08,
+      "step": 8044
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5878644076970218,
+      "learning_rate": 1.6611121514917585e-05,
+      "loss": 1.8978,
+      "step": 8045
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5831659900656351,
+      "learning_rate": 1.660523725269114e-05,
+      "loss": 1.9658,
+      "step": 8046
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5602775535804109,
+      "learning_rate": 1.659935351458423e-05,
+      "loss": 1.824,
+      "step": 8047
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6002489386191445,
+      "learning_rate": 1.6593470300964177e-05,
+      "loss": 2.0646,
+      "step": 8048
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5564676984264488,
+      "learning_rate": 1.658758761219831e-05,
+      "loss": 1.8593,
+      "step": 8049
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6264873869681522,
+      "learning_rate": 1.6581705448653916e-05,
+      "loss": 1.8952,
+      "step": 8050
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6058678630819602,
+      "learning_rate": 1.6575823810698227e-05,
+      "loss": 1.8987,
+      "step": 8051
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5782693188947625,
+      "learning_rate": 1.656994269869847e-05,
+      "loss": 2.1094,
+      "step": 8052
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6357996098006783,
+      "learning_rate": 1.6564062113021814e-05,
+      "loss": 1.8721,
+      "step": 8053
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.515303778245165,
+      "learning_rate": 1.655818205403542e-05,
+      "loss": 1.8825,
+      "step": 8054
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7160412712075651,
+      "learning_rate": 1.6552302522106405e-05,
+      "loss": 1.8827,
+      "step": 8055
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.0602593173735444,
+      "learning_rate": 1.6546423517601838e-05,
+      "loss": 2.1012,
+      "step": 8056
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6502543486562072,
+      "learning_rate": 1.654054504088879e-05,
+      "loss": 1.8866,
+      "step": 8057
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.593784157115339,
+      "learning_rate": 1.653466709233426e-05,
+      "loss": 2.0046,
+      "step": 8058
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6077538852885653,
+      "learning_rate": 1.652878967230524e-05,
+      "loss": 1.891,
+      "step": 8059
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5954354868094053,
+      "learning_rate": 1.6522912781168685e-05,
+      "loss": 2.1012,
+      "step": 8060
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5827105956302798,
+      "learning_rate": 1.651703641929151e-05,
+      "loss": 1.9216,
+      "step": 8061
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5644056136574299,
+      "learning_rate": 1.651116058704059e-05,
+      "loss": 1.9216,
+      "step": 8062
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6259181331048347,
+      "learning_rate": 1.6505285284782802e-05,
+      "loss": 1.8646,
+      "step": 8063
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5904613980319957,
+      "learning_rate": 1.6499410512884945e-05,
+      "loss": 2.1182,
+      "step": 8064
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5345021678364518,
+      "learning_rate": 1.6493536271713816e-05,
+      "loss": 1.9739,
+      "step": 8065
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6547393274403956,
+      "learning_rate": 1.6487662561636154e-05,
+      "loss": 1.9231,
+      "step": 8066
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.55041248391652,
+      "learning_rate": 1.6481789383018682e-05,
+      "loss": 1.8416,
+      "step": 8067
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5587007572077188,
+      "learning_rate": 1.647591673622811e-05,
+      "loss": 2.0942,
+      "step": 8068
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6008870407817143,
+      "learning_rate": 1.6470044621631063e-05,
+      "loss": 1.8771,
+      "step": 8069
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5771818797147513,
+      "learning_rate": 1.6464173039594172e-05,
+      "loss": 1.8802,
+      "step": 8070
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5496455565021258,
+      "learning_rate": 1.6458301990484032e-05,
+      "loss": 1.9363,
+      "step": 8071
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5569270637929251,
+      "learning_rate": 1.645243147466718e-05,
+      "loss": 2.1303,
+      "step": 8072
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6290749331081951,
+      "learning_rate": 1.6446561492510156e-05,
+      "loss": 1.8487,
+      "step": 8073
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6659922722590701,
+      "learning_rate": 1.6440692044379425e-05,
+      "loss": 1.947,
+      "step": 8074
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6061690885773195,
+      "learning_rate": 1.643482313064146e-05,
+      "loss": 1.9335,
+      "step": 8075
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5302501742388732,
+      "learning_rate": 1.6428954751662674e-05,
+      "loss": 2.0805,
+      "step": 8076
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5643789972439974,
+      "learning_rate": 1.6423086907809442e-05,
+      "loss": 1.9611,
+      "step": 8077
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5978386274008384,
+      "learning_rate": 1.6417219599448142e-05,
+      "loss": 1.8761,
+      "step": 8078
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5626958470384708,
+      "learning_rate": 1.641135282694508e-05,
+      "loss": 1.8671,
+      "step": 8079
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6167363174344708,
+      "learning_rate": 1.6405486590666532e-05,
+      "loss": 2.0731,
+      "step": 8080
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5752784107257081,
+      "learning_rate": 1.6399620890978778e-05,
+      "loss": 1.9475,
+      "step": 8081
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5718197703786133,
+      "learning_rate": 1.6393755728248017e-05,
+      "loss": 1.8973,
+      "step": 8082
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6568533469105671,
+      "learning_rate": 1.6387891102840437e-05,
+      "loss": 1.9846,
+      "step": 8083
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5827023216585865,
+      "learning_rate": 1.6382027015122202e-05,
+      "loss": 1.8643,
+      "step": 8084
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.581859477546589,
+      "learning_rate": 1.6376163465459417e-05,
+      "loss": 2.087,
+      "step": 8085
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6712733176972534,
+      "learning_rate": 1.6370300454218184e-05,
+      "loss": 1.8773,
+      "step": 8086
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5836764467520185,
+      "learning_rate": 1.6364437981764542e-05,
+      "loss": 1.8948,
+      "step": 8087
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6397438923478695,
+      "learning_rate": 1.6358576048464502e-05,
+      "loss": 2.0889,
+      "step": 8088
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5826641247811815,
+      "learning_rate": 1.6352714654684072e-05,
+      "loss": 1.9598,
+      "step": 8089
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5377883252599942,
+      "learning_rate": 1.6346853800789186e-05,
+      "loss": 1.8654,
+      "step": 8090
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.627100768298841,
+      "learning_rate": 1.6340993487145774e-05,
+      "loss": 1.8855,
+      "step": 8091
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5961259797891036,
+      "learning_rate": 1.6335133714119694e-05,
+      "loss": 2.1214,
+      "step": 8092
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6048038036339286,
+      "learning_rate": 1.6329274482076817e-05,
+      "loss": 1.8785,
+      "step": 8093
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5718464391050702,
+      "learning_rate": 1.6323415791382965e-05,
+      "loss": 1.8645,
+      "step": 8094
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5973612566044231,
+      "learning_rate": 1.6317557642403896e-05,
+      "loss": 1.8553,
+      "step": 8095
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6668321467019985,
+      "learning_rate": 1.631170003550538e-05,
+      "loss": 1.9516,
+      "step": 8096
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6406095896078066,
+      "learning_rate": 1.6305842971053123e-05,
+      "loss": 2.0735,
+      "step": 8097
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5694496974267224,
+      "learning_rate": 1.6299986449412797e-05,
+      "loss": 1.8583,
+      "step": 8098
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6471903012742594,
+      "learning_rate": 1.629413047095007e-05,
+      "loss": 1.8983,
+      "step": 8099
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6770508875523102,
+      "learning_rate": 1.628827503603053e-05,
+      "loss": 2.0704,
+      "step": 8100
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6027164083039429,
+      "learning_rate": 1.628242014501978e-05,
+      "loss": 1.912,
+      "step": 8101
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6107296977646794,
+      "learning_rate": 1.627656579828335e-05,
+      "loss": 1.974,
+      "step": 8102
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6033916630560902,
+      "learning_rate": 1.6270711996186745e-05,
+      "loss": 1.9159,
+      "step": 8103
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6333899100718187,
+      "learning_rate": 1.6264858739095457e-05,
+      "loss": 2.0812,
+      "step": 8104
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6231071912438237,
+      "learning_rate": 1.625900602737493e-05,
+      "loss": 1.8665,
+      "step": 8105
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.604826120684237,
+      "learning_rate": 1.6253153861390552e-05,
+      "loss": 1.83,
+      "step": 8106
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6174874650203618,
+      "learning_rate": 1.6247302241507723e-05,
+      "loss": 1.9156,
+      "step": 8107
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5875313358786617,
+      "learning_rate": 1.6241451168091767e-05,
+      "loss": 1.9797,
+      "step": 8108
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6266089327591563,
+      "learning_rate": 1.6235600641507993e-05,
+      "loss": 2.0616,
+      "step": 8109
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5832433585138151,
+      "learning_rate": 1.6229750662121682e-05,
+      "loss": 1.8802,
+      "step": 8110
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5934576083999693,
+      "learning_rate": 1.6223901230298062e-05,
+      "loss": 1.8223,
+      "step": 8111
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6013801694484762,
+      "learning_rate": 1.621805234640235e-05,
+      "loss": 2.0854,
+      "step": 8112
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.588291008313574,
+      "learning_rate": 1.6212204010799692e-05,
+      "loss": 1.9156,
+      "step": 8113
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5721952548075647,
+      "learning_rate": 1.6206356223855247e-05,
+      "loss": 1.9885,
+      "step": 8114
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5699337598355331,
+      "learning_rate": 1.6200508985934113e-05,
+      "loss": 1.9044,
+      "step": 8115
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6409316333204232,
+      "learning_rate": 1.619466229740134e-05,
+      "loss": 1.8509,
+      "step": 8116
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6857036804747035,
+      "learning_rate": 1.6188816158621977e-05,
+      "loss": 2.0861,
+      "step": 8117
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5625723024370958,
+      "learning_rate": 1.6182970569961025e-05,
+      "loss": 1.9028,
+      "step": 8118
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6138287574014092,
+      "learning_rate": 1.617712553178344e-05,
+      "loss": 1.9157,
+      "step": 8119
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6368201486613727,
+      "learning_rate": 1.6171281044454153e-05,
+      "loss": 1.9289,
+      "step": 8120
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5930858523858696,
+      "learning_rate": 1.6165437108338052e-05,
+      "loss": 2.0664,
+      "step": 8121
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6615515391843885,
+      "learning_rate": 1.6159593723800012e-05,
+      "loss": 1.8469,
+      "step": 8122
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.601377849741503,
+      "learning_rate": 1.615375089120486e-05,
+      "loss": 1.8893,
+      "step": 8123
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.587868546604295,
+      "learning_rate": 1.6147908610917367e-05,
+      "loss": 2.0896,
+      "step": 8124
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6414396838159648,
+      "learning_rate": 1.6142066883302315e-05,
+      "loss": 1.8464,
+      "step": 8125
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6636118403606509,
+      "learning_rate": 1.613622570872441e-05,
+      "loss": 1.8806,
+      "step": 8126
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5315243631804367,
+      "learning_rate": 1.613038508754835e-05,
+      "loss": 1.8916,
+      "step": 8127
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5443498168180257,
+      "learning_rate": 1.6124545020138792e-05,
+      "loss": 1.8791,
+      "step": 8128
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6572416383230438,
+      "learning_rate": 1.6118705506860344e-05,
+      "loss": 2.0854,
+      "step": 8129
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6465249023691553,
+      "learning_rate": 1.611286654807759e-05,
+      "loss": 1.9097,
+      "step": 8130
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5706125345136264,
+      "learning_rate": 1.61070281441551e-05,
+      "loss": 1.8534,
+      "step": 8131
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.65104707713467,
+      "learning_rate": 1.6101190295457364e-05,
+      "loss": 2.069,
+      "step": 8132
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6095387153789739,
+      "learning_rate": 1.6095353002348883e-05,
+      "loss": 1.9897,
+      "step": 8133
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6541979343828871,
+      "learning_rate": 1.6089516265194096e-05,
+      "loss": 1.8967,
+      "step": 8134
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7041542623288523,
+      "learning_rate": 1.6083680084357406e-05,
+      "loss": 1.9204,
+      "step": 8135
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6013131482303682,
+      "learning_rate": 1.6077844460203206e-05,
+      "loss": 2.0851,
+      "step": 8136
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5713963058728061,
+      "learning_rate": 1.6072009393095827e-05,
+      "loss": 1.8316,
+      "step": 8137
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7461458228105966,
+      "learning_rate": 1.6066174883399583e-05,
+      "loss": 1.8683,
+      "step": 8138
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5712805059533478,
+      "learning_rate": 1.606034093147873e-05,
+      "loss": 1.9879,
+      "step": 8139
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5743408072585097,
+      "learning_rate": 1.6054507537697524e-05,
+      "loss": 1.8603,
+      "step": 8140
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.62309224352662,
+      "learning_rate": 1.6048674702420167e-05,
+      "loss": 2.0376,
+      "step": 8141
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6157498767052207,
+      "learning_rate": 1.6042842426010813e-05,
+      "loss": 1.9029,
+      "step": 8142
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.581655074193218,
+      "learning_rate": 1.6037010708833606e-05,
+      "loss": 1.8947,
+      "step": 8143
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5974668525946777,
+      "learning_rate": 1.6031179551252646e-05,
+      "loss": 2.032,
+      "step": 8144
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5023700833854972,
+      "learning_rate": 1.602534895363198e-05,
+      "loss": 1.9623,
+      "step": 8145
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6143190152339365,
+      "learning_rate": 1.6019518916335662e-05,
+      "loss": 1.8728,
+      "step": 8146
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5465392302896779,
+      "learning_rate": 1.6013689439727663e-05,
+      "loss": 1.906,
+      "step": 8147
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6719044804724982,
+      "learning_rate": 1.600786052417194e-05,
+      "loss": 1.8757,
+      "step": 8148
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.595666847333441,
+      "learning_rate": 1.6002032170032443e-05,
+      "loss": 2.1048,
+      "step": 8149
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7392415314386237,
+      "learning_rate": 1.599620437767303e-05,
+      "loss": 1.8648,
+      "step": 8150
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6580973388601695,
+      "learning_rate": 1.599037714745757e-05,
+      "loss": 1.9703,
+      "step": 8151
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6301201200346209,
+      "learning_rate": 1.5984550479749882e-05,
+      "loss": 1.8936,
+      "step": 8152
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7824068914292536,
+      "learning_rate": 1.5978724374913732e-05,
+      "loss": 2.0885,
+      "step": 8153
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5938557897911837,
+      "learning_rate": 1.5972898833312893e-05,
+      "loss": 1.9188,
+      "step": 8154
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5867135984751461,
+      "learning_rate": 1.5967073855311056e-05,
+      "loss": 1.8744,
+      "step": 8155
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5764174067017281,
+      "learning_rate": 1.5961249441271903e-05,
+      "loss": 2.098,
+      "step": 8156
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5878453083439222,
+      "learning_rate": 1.595542559155909e-05,
+      "loss": 1.8809,
+      "step": 8157
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6221777970606243,
+      "learning_rate": 1.594960230653621e-05,
+      "loss": 1.9561,
+      "step": 8158
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5714815620707111,
+      "learning_rate": 1.594377958656684e-05,
+      "loss": 1.8767,
+      "step": 8159
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6778263028093088,
+      "learning_rate": 1.5937957432014506e-05,
+      "loss": 1.9047,
+      "step": 8160
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5769145023959021,
+      "learning_rate": 1.593213584324272e-05,
+      "loss": 2.1239,
+      "step": 8161
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6434559364530071,
+      "learning_rate": 1.592631482061495e-05,
+      "loss": 1.8986,
+      "step": 8162
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6005564674170942,
+      "learning_rate": 1.5920494364494616e-05,
+      "loss": 1.883,
+      "step": 8163
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.561925540620594,
+      "learning_rate": 1.5914674475245115e-05,
+      "loss": 1.9474,
+      "step": 8164
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6616754726054338,
+      "learning_rate": 1.590885515322982e-05,
+      "loss": 2.1058,
+      "step": 8165
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5455132285433694,
+      "learning_rate": 1.590303639881204e-05,
+      "loss": 1.8727,
+      "step": 8166
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6495259356825086,
+      "learning_rate": 1.5897218212355077e-05,
+      "loss": 1.8749,
+      "step": 8167
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6147887527625137,
+      "learning_rate": 1.5891400594222163e-05,
+      "loss": 1.8735,
+      "step": 8168
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6270307980884934,
+      "learning_rate": 1.5885583544776534e-05,
+      "loss": 2.118,
+      "step": 8169
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6925350570389632,
+      "learning_rate": 1.5879767064381375e-05,
+      "loss": 1.9491,
+      "step": 8170
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5710676017424163,
+      "learning_rate": 1.5873951153399808e-05,
+      "loss": 1.9204,
+      "step": 8171
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6574959194793875,
+      "learning_rate": 1.5868135812194977e-05,
+      "loss": 1.9109,
+      "step": 8172
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6029470086877811,
+      "learning_rate": 1.5862321041129934e-05,
+      "loss": 2.0745,
+      "step": 8173
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6605546681612915,
+      "learning_rate": 1.5856506840567724e-05,
+      "loss": 1.926,
+      "step": 8174
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.631577726888011,
+      "learning_rate": 1.5850693210871364e-05,
+      "loss": 1.8794,
+      "step": 8175
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5567390926673188,
+      "learning_rate": 1.584488015240381e-05,
+      "loss": 2.0051,
+      "step": 8176
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6831978573502983,
+      "learning_rate": 1.583906766552799e-05,
+      "loss": 2.0304,
+      "step": 8177
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.625262748015255,
+      "learning_rate": 1.5833255750606825e-05,
+      "loss": 1.8433,
+      "step": 8178
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7187073260242481,
+      "learning_rate": 1.582744440800315e-05,
+      "loss": 1.8865,
+      "step": 8179
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6115921162894893,
+      "learning_rate": 1.582163363807981e-05,
+      "loss": 1.8639,
+      "step": 8180
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.746344698450898,
+      "learning_rate": 1.581582344119958e-05,
+      "loss": 2.1107,
+      "step": 8181
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6236604105758277,
+      "learning_rate": 1.5810013817725216e-05,
+      "loss": 1.9527,
+      "step": 8182
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6120799183592102,
+      "learning_rate": 1.580420476801946e-05,
+      "loss": 1.8962,
+      "step": 8183
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6785819149712855,
+      "learning_rate": 1.5798396292444966e-05,
+      "loss": 1.8985,
+      "step": 8184
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5700043150546789,
+      "learning_rate": 1.5792588391364393e-05,
+      "loss": 2.1238,
+      "step": 8185
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6551128092140062,
+      "learning_rate": 1.578678106514036e-05,
+      "loss": 1.8622,
+      "step": 8186
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5569663536734482,
+      "learning_rate": 1.578097431413543e-05,
+      "loss": 1.9006,
+      "step": 8187
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6273827823125815,
+      "learning_rate": 1.577516813871215e-05,
+      "loss": 2.1056,
+      "step": 8188
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5433925996740084,
+      "learning_rate": 1.5769362539233007e-05,
+      "loss": 1.9165,
+      "step": 8189
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5739637496134635,
+      "learning_rate": 1.576355751606049e-05,
+      "loss": 1.8591,
+      "step": 8190
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6371749102403651,
+      "learning_rate": 1.5757753069557022e-05,
+      "loss": 1.8659,
+      "step": 8191
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5614759520207058,
+      "learning_rate": 1.5751949200084993e-05,
+      "loss": 1.9139,
+      "step": 8192
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.791254540914331,
+      "learning_rate": 1.5746145908006778e-05,
+      "loss": 2.0515,
+      "step": 8193
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5432546090652003,
+      "learning_rate": 1.574034319368468e-05,
+      "loss": 1.8471,
+      "step": 8194
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6580852910200371,
+      "learning_rate": 1.5734541057480993e-05,
+      "loss": 1.9622,
+      "step": 8195
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6109275097657062,
+      "learning_rate": 1.5728739499757984e-05,
+      "loss": 1.876,
+      "step": 8196
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7098363063708722,
+      "learning_rate": 1.5722938520877838e-05,
+      "loss": 2.0967,
+      "step": 8197
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7277659192718275,
+      "learning_rate": 1.5717138121202764e-05,
+      "loss": 1.9058,
+      "step": 8198
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5582410277991453,
+      "learning_rate": 1.571133830109489e-05,
+      "loss": 1.8916,
+      "step": 8199
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6894891811058498,
+      "learning_rate": 1.5705539060916323e-05,
+      "loss": 1.8803,
+      "step": 8200
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6723858026494186,
+      "learning_rate": 1.569974040102914e-05,
+      "loss": 2.0869,
+      "step": 8201
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.65493513551692,
+      "learning_rate": 1.5693942321795366e-05,
+      "loss": 1.8294,
+      "step": 8202
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6075682452729689,
+      "learning_rate": 1.5688144823576996e-05,
+      "loss": 1.8741,
+      "step": 8203
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6680175207035346,
+      "learning_rate": 1.5682347906736013e-05,
+      "loss": 1.8766,
+      "step": 8204
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7719439429786313,
+      "learning_rate": 1.5676551571634317e-05,
+      "loss": 2.0715,
+      "step": 8205
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5527198752998035,
+      "learning_rate": 1.567075581863382e-05,
+      "loss": 1.8949,
+      "step": 8206
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7039805020857247,
+      "learning_rate": 1.566496064809635e-05,
+      "loss": 1.9128,
+      "step": 8207
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.678894358340965,
+      "learning_rate": 1.565916606038374e-05,
+      "loss": 1.8901,
+      "step": 8208
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6226494381252005,
+      "learning_rate": 1.5653372055857772e-05,
+      "loss": 2.052,
+      "step": 8209
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6068881925983989,
+      "learning_rate": 1.564757863488017e-05,
+      "loss": 1.8918,
+      "step": 8210
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.710772123518608,
+      "learning_rate": 1.5641785797812665e-05,
+      "loss": 1.8985,
+      "step": 8211
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5776936970601333,
+      "learning_rate": 1.563599354501692e-05,
+      "loss": 1.8813,
+      "step": 8212
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5847462085535985,
+      "learning_rate": 1.5630201876854556e-05,
+      "loss": 2.1456,
+      "step": 8213
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5717775815285082,
+      "learning_rate": 1.562441079368719e-05,
+      "loss": 1.8814,
+      "step": 8214
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6150031356894797,
+      "learning_rate": 1.5618620295876365e-05,
+      "loss": 1.8829,
+      "step": 8215
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5620292866264495,
+      "learning_rate": 1.561283038378362e-05,
+      "loss": 1.863,
+      "step": 8216
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5698428993535279,
+      "learning_rate": 1.5607041057770443e-05,
+      "loss": 2.1106,
+      "step": 8217
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6124337157572259,
+      "learning_rate": 1.560125231819827e-05,
+      "loss": 1.8827,
+      "step": 8218
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6624098066420331,
+      "learning_rate": 1.559546416542853e-05,
+      "loss": 1.8848,
+      "step": 8219
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5558559020972711,
+      "learning_rate": 1.55896765998226e-05,
+      "loss": 1.9914,
+      "step": 8220
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6771763274958177,
+      "learning_rate": 1.558388962174181e-05,
+      "loss": 2.1127,
+      "step": 8221
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7440466134590237,
+      "learning_rate": 1.5578103231547485e-05,
+      "loss": 1.8577,
+      "step": 8222
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5614016490479711,
+      "learning_rate": 1.5572317429600874e-05,
+      "loss": 1.8803,
+      "step": 8223
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5485893364419799,
+      "learning_rate": 1.5566532216263214e-05,
+      "loss": 1.8685,
+      "step": 8224
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6493762376982991,
+      "learning_rate": 1.5560747591895707e-05,
+      "loss": 2.0743,
+      "step": 8225
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5732961619871015,
+      "learning_rate": 1.5554963556859502e-05,
+      "loss": 1.9832,
+      "step": 8226
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5707537925842697,
+      "learning_rate": 1.5549180111515732e-05,
+      "loss": 1.8635,
+      "step": 8227
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7161740029301322,
+      "learning_rate": 1.5543397256225457e-05,
+      "loss": 1.9338,
+      "step": 8228
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5728558791292477,
+      "learning_rate": 1.553761499134974e-05,
+      "loss": 2.0471,
+      "step": 8229
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6120882533885048,
+      "learning_rate": 1.5531833317249607e-05,
+      "loss": 1.9354,
+      "step": 8230
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6077008716899439,
+      "learning_rate": 1.552605223428601e-05,
+      "loss": 1.8686,
+      "step": 8231
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5953917133863135,
+      "learning_rate": 1.5520271742819883e-05,
+      "loss": 1.9322,
+      "step": 8232
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7260728902014262,
+      "learning_rate": 1.551449184321215e-05,
+      "loss": 2.1019,
+      "step": 8233
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5898128424381098,
+      "learning_rate": 1.550871253582365e-05,
+      "loss": 1.8632,
+      "step": 8234
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.635470300023864,
+      "learning_rate": 1.550293382101522e-05,
+      "loss": 1.9043,
+      "step": 8235
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6179457716260498,
+      "learning_rate": 1.549715569914764e-05,
+      "loss": 1.8717,
+      "step": 8236
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.786198922971832,
+      "learning_rate": 1.5491378170581673e-05,
+      "loss": 2.0932,
+      "step": 8237
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5624426271090147,
+      "learning_rate": 1.5485601235678036e-05,
+      "loss": 1.9573,
+      "step": 8238
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6588177540776137,
+      "learning_rate": 1.547982489479739e-05,
+      "loss": 1.8455,
+      "step": 8239
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7217626135836811,
+      "learning_rate": 1.547404914830039e-05,
+      "loss": 1.9021,
+      "step": 8240
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6935998256831328,
+      "learning_rate": 1.5468273996547634e-05,
+      "loss": 2.092,
+      "step": 8241
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.698241910157609,
+      "learning_rate": 1.5462499439899687e-05,
+      "loss": 1.8916,
+      "step": 8242
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6767138739625068,
+      "learning_rate": 1.5456725478717088e-05,
+      "loss": 1.8829,
+      "step": 8243
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7023304067465366,
+      "learning_rate": 1.5450952113360317e-05,
+      "loss": 1.8764,
+      "step": 8244
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6698611768945949,
+      "learning_rate": 1.544517934418983e-05,
+      "loss": 2.1403,
+      "step": 8245
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6979222954784412,
+      "learning_rate": 1.543940717156606e-05,
+      "loss": 1.8792,
+      "step": 8246
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6839476217949796,
+      "learning_rate": 1.5433635595849362e-05,
+      "loss": 1.846,
+      "step": 8247
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6525108767904783,
+      "learning_rate": 1.5427864617400106e-05,
+      "loss": 1.8906,
+      "step": 8248
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6955640084463295,
+      "learning_rate": 1.5422094236578582e-05,
+      "loss": 2.1157,
+      "step": 8249
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.664098642127394,
+      "learning_rate": 1.5416324453745056e-05,
+      "loss": 1.8921,
+      "step": 8250
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7276079778300725,
+      "learning_rate": 1.5410555269259775e-05,
+      "loss": 1.9502,
+      "step": 8251
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.696146237692106,
+      "learning_rate": 1.5404786683482917e-05,
+      "loss": 1.906,
+      "step": 8252
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7047891673446766,
+      "learning_rate": 1.539901869677464e-05,
+      "loss": 2.0576,
+      "step": 8253
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.712452434085597,
+      "learning_rate": 1.5393251309495078e-05,
+      "loss": 1.8319,
+      "step": 8254
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6657790011896529,
+      "learning_rate": 1.53874845220043e-05,
+      "loss": 1.868,
+      "step": 8255
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.632458669051041,
+      "learning_rate": 1.538171833466235e-05,
+      "loss": 1.8939,
+      "step": 8256
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.875252718223018,
+      "learning_rate": 1.537595274782923e-05,
+      "loss": 2.1076,
+      "step": 8257
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.668075048234358,
+      "learning_rate": 1.5370187761864925e-05,
+      "loss": 1.8784,
+      "step": 8258
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6593903000086498,
+      "learning_rate": 1.5364423377129362e-05,
+      "loss": 1.8844,
+      "step": 8259
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6841854400582955,
+      "learning_rate": 1.535865959398242e-05,
+      "loss": 1.8852,
+      "step": 8260
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.8715785511613727,
+      "learning_rate": 1.535289641278398e-05,
+      "loss": 2.0431,
+      "step": 8261
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6125051896033447,
+      "learning_rate": 1.5347133833893834e-05,
+      "loss": 1.8947,
+      "step": 8262
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6869835391280711,
+      "learning_rate": 1.5341371857671782e-05,
+      "loss": 1.9404,
+      "step": 8263
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7505560144207256,
+      "learning_rate": 1.5335610484477574e-05,
+      "loss": 1.8743,
+      "step": 8264
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5855600688663578,
+      "learning_rate": 1.5329849714670894e-05,
+      "loss": 2.1,
+      "step": 8265
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5937593652213607,
+      "learning_rate": 1.5324089548611427e-05,
+      "loss": 1.8643,
+      "step": 8266
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6755362418864037,
+      "learning_rate": 1.53183299866588e-05,
+      "loss": 1.855,
+      "step": 8267
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6361405546204497,
+      "learning_rate": 1.53125710291726e-05,
+      "loss": 1.858,
+      "step": 8268
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6494852668769009,
+      "learning_rate": 1.5306812676512398e-05,
+      "loss": 2.0751,
+      "step": 8269
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6303017898280341,
+      "learning_rate": 1.5301054929037696e-05,
+      "loss": 1.8737,
+      "step": 8270
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6548025075988254,
+      "learning_rate": 1.529529778710797e-05,
+      "loss": 1.8957,
+      "step": 8271
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6117351221638625,
+      "learning_rate": 1.5289541251082685e-05,
+      "loss": 1.9119,
+      "step": 8272
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7217486094577534,
+      "learning_rate": 1.5283785321321227e-05,
+      "loss": 2.0763,
+      "step": 8273
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6044672171650762,
+      "learning_rate": 1.527802999818297e-05,
+      "loss": 1.9282,
+      "step": 8274
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6058762495678016,
+      "learning_rate": 1.5272275282027227e-05,
+      "loss": 1.8802,
+      "step": 8275
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6435101311129808,
+      "learning_rate": 1.5266521173213306e-05,
+      "loss": 1.9759,
+      "step": 8276
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5998722890330975,
+      "learning_rate": 1.526076767210046e-05,
+      "loss": 2.1124,
+      "step": 8277
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6601584134614529,
+      "learning_rate": 1.5255014779047888e-05,
+      "loss": 1.8473,
+      "step": 8278
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6360509053621417,
+      "learning_rate": 1.5249262494414779e-05,
+      "loss": 1.8821,
+      "step": 8279
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5754572570256475,
+      "learning_rate": 1.5243510818560278e-05,
+      "loss": 1.9129,
+      "step": 8280
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7073060019844518,
+      "learning_rate": 1.5237759751843472e-05,
+      "loss": 2.1283,
+      "step": 8281
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5409692528019727,
+      "learning_rate": 1.5232009294623433e-05,
+      "loss": 1.9735,
+      "step": 8282
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5741485297007035,
+      "learning_rate": 1.5226259447259173e-05,
+      "loss": 1.8682,
+      "step": 8283
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.559161722635525,
+      "learning_rate": 1.5220510210109692e-05,
+      "loss": 1.8501,
+      "step": 8284
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5986846022673626,
+      "learning_rate": 1.521476158353394e-05,
+      "loss": 2.1265,
+      "step": 8285
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6098624342768384,
+      "learning_rate": 1.5209013567890809e-05,
+      "loss": 1.8849,
+      "step": 8286
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5847258215495189,
+      "learning_rate": 1.5203266163539195e-05,
+      "loss": 1.8308,
+      "step": 8287
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5826882362576981,
+      "learning_rate": 1.5197519370837914e-05,
+      "loss": 1.9887,
+      "step": 8288
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6131710668401387,
+      "learning_rate": 1.5191773190145765e-05,
+      "loss": 2.0917,
+      "step": 8289
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6737468307936287,
+      "learning_rate": 1.5186027621821519e-05,
+      "loss": 1.9153,
+      "step": 8290
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5850376527198209,
+      "learning_rate": 1.5180282666223876e-05,
+      "loss": 1.8875,
+      "step": 8291
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6631375221770343,
+      "learning_rate": 1.5174538323711526e-05,
+      "loss": 1.9183,
+      "step": 8292
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5964725218158218,
+      "learning_rate": 1.516879459464312e-05,
+      "loss": 2.1034,
+      "step": 8293
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5944332287387949,
+      "learning_rate": 1.516305147937725e-05,
+      "loss": 1.9352,
+      "step": 8294
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6335496563320524,
+      "learning_rate": 1.5157308978272491e-05,
+      "loss": 1.8337,
+      "step": 8295
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6109860443396689,
+      "learning_rate": 1.5151567091687366e-05,
+      "loss": 1.8912,
+      "step": 8296
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6440179883257581,
+      "learning_rate": 1.5145825819980358e-05,
+      "loss": 2.1178,
+      "step": 8297
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6459999153058461,
+      "learning_rate": 1.5140085163509937e-05,
+      "loss": 1.8803,
+      "step": 8298
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.721322398752142,
+      "learning_rate": 1.5134345122634502e-05,
+      "loss": 1.8711,
+      "step": 8299
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7338500907295465,
+      "learning_rate": 1.5128605697712423e-05,
+      "loss": 1.9667,
+      "step": 8300
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.58298275120827,
+      "learning_rate": 1.5122866889102056e-05,
+      "loss": 1.892,
+      "step": 8301
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6829896328795524,
+      "learning_rate": 1.5117128697161675e-05,
+      "loss": 2.1107,
+      "step": 8302
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6022251087855807,
+      "learning_rate": 1.5111391122249558e-05,
+      "loss": 1.8814,
+      "step": 8303
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6012433719777383,
+      "learning_rate": 1.5105654164723906e-05,
+      "loss": 1.9031,
+      "step": 8304
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7934396214992626,
+      "learning_rate": 1.5099917824942917e-05,
+      "loss": 2.1218,
+      "step": 8305
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5628738128644897,
+      "learning_rate": 1.5094182103264734e-05,
+      "loss": 1.9238,
+      "step": 8306
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6208050780137758,
+      "learning_rate": 1.5088447000047445e-05,
+      "loss": 1.9961,
+      "step": 8307
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.678462383842667,
+      "learning_rate": 1.5082712515649138e-05,
+      "loss": 1.883,
+      "step": 8308
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5734550053619724,
+      "learning_rate": 1.5076978650427826e-05,
+      "loss": 2.0707,
+      "step": 8309
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6609625791081309,
+      "learning_rate": 1.5071245404741496e-05,
+      "loss": 1.8878,
+      "step": 8310
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5957946995340899,
+      "learning_rate": 1.5065512778948113e-05,
+      "loss": 1.931,
+      "step": 8311
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5922935311235172,
+      "learning_rate": 1.5059780773405569e-05,
+      "loss": 1.9025,
+      "step": 8312
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6260024107634815,
+      "learning_rate": 1.5054049388471752e-05,
+      "loss": 1.924,
+      "step": 8313
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6567235781613207,
+      "learning_rate": 1.5048318624504499e-05,
+      "loss": 2.103,
+      "step": 8314
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6422142457679305,
+      "learning_rate": 1.504258848186158e-05,
+      "loss": 1.9157,
+      "step": 8315
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6090879090397989,
+      "learning_rate": 1.5036858960900784e-05,
+      "loss": 1.9059,
+      "step": 8316
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5860529407594672,
+      "learning_rate": 1.5031130061979804e-05,
+      "loss": 2.0723,
+      "step": 8317
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6070556915618565,
+      "learning_rate": 1.5025401785456323e-05,
+      "loss": 1.892,
+      "step": 8318
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5812505114581853,
+      "learning_rate": 1.5019674131687997e-05,
+      "loss": 1.9216,
+      "step": 8319
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5483005522117286,
+      "learning_rate": 1.501394710103241e-05,
+      "loss": 1.9201,
+      "step": 8320
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6217673454563182,
+      "learning_rate": 1.5008220693847133e-05,
+      "loss": 2.065,
+      "step": 8321
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5785405452553793,
+      "learning_rate": 1.5002494910489675e-05,
+      "loss": 1.9242,
+      "step": 8322
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5759303158367849,
+      "learning_rate": 1.4996769751317535e-05,
+      "loss": 1.888,
+      "step": 8323
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6392581200989591,
+      "learning_rate": 1.4991045216688162e-05,
+      "loss": 1.9223,
+      "step": 8324
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5625454460311049,
+      "learning_rate": 1.4985321306958944e-05,
+      "loss": 2.0063,
+      "step": 8325
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6748058489025044,
+      "learning_rate": 1.4979598022487254e-05,
+      "loss": 2.11,
+      "step": 8326
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5538524199096411,
+      "learning_rate": 1.4973875363630435e-05,
+      "loss": 1.8799,
+      "step": 8327
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6342147857005462,
+      "learning_rate": 1.4968153330745766e-05,
+      "loss": 1.854,
+      "step": 8328
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6144464139240059,
+      "learning_rate": 1.4962431924190495e-05,
+      "loss": 2.0567,
+      "step": 8329
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6177040436793608,
+      "learning_rate": 1.4956711144321828e-05,
+      "loss": 1.8772,
+      "step": 8330
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5697342071879058,
+      "learning_rate": 1.4950990991496943e-05,
+      "loss": 1.9566,
+      "step": 8331
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5800006235252373,
+      "learning_rate": 1.4945271466072986e-05,
+      "loss": 1.9476,
+      "step": 8332
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6809219626952162,
+      "learning_rate": 1.4939552568407023e-05,
+      "loss": 1.8311,
+      "step": 8333
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6479902340806296,
+      "learning_rate": 1.4933834298856125e-05,
+      "loss": 2.125,
+      "step": 8334
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6766866714776264,
+      "learning_rate": 1.4928116657777314e-05,
+      "loss": 1.8615,
+      "step": 8335
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.624355159845173,
+      "learning_rate": 1.4922399645527541e-05,
+      "loss": 1.8756,
+      "step": 8336
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6115643295686684,
+      "learning_rate": 1.4916683262463771e-05,
+      "loss": 2.0561,
+      "step": 8337
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6002869454987185,
+      "learning_rate": 1.4910967508942885e-05,
+      "loss": 1.9247,
+      "step": 8338
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5905011575551776,
+      "learning_rate": 1.4905252385321738e-05,
+      "loss": 1.8617,
+      "step": 8339
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.665978707803345,
+      "learning_rate": 1.4899537891957166e-05,
+      "loss": 1.8919,
+      "step": 8340
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5565799305527607,
+      "learning_rate": 1.4893824029205927e-05,
+      "loss": 2.064,
+      "step": 8341
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5470586564642844,
+      "learning_rate": 1.4888110797424782e-05,
+      "loss": 1.8414,
+      "step": 8342
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5678370175685654,
+      "learning_rate": 1.488239819697041e-05,
+      "loss": 1.927,
+      "step": 8343
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.548526596206015,
+      "learning_rate": 1.4876686228199476e-05,
+      "loss": 1.9146,
+      "step": 8344
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5649908976600285,
+      "learning_rate": 1.4870974891468625e-05,
+      "loss": 1.9099,
+      "step": 8345
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6362717066922087,
+      "learning_rate": 1.4865264187134418e-05,
+      "loss": 2.0419,
+      "step": 8346
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5799718748809531,
+      "learning_rate": 1.4859554115553392e-05,
+      "loss": 1.9097,
+      "step": 8347
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7240228413988585,
+      "learning_rate": 1.4853844677082077e-05,
+      "loss": 1.9195,
+      "step": 8348
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6464349700166891,
+      "learning_rate": 1.4848135872076913e-05,
+      "loss": 2.0589,
+      "step": 8349
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6094590967829089,
+      "learning_rate": 1.4842427700894335e-05,
+      "loss": 1.9454,
+      "step": 8350
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.692654192896129,
+      "learning_rate": 1.4836720163890714e-05,
+      "loss": 1.8715,
+      "step": 8351
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5742143711666354,
+      "learning_rate": 1.483101326142241e-05,
+      "loss": 1.9249,
+      "step": 8352
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7213744787539169,
+      "learning_rate": 1.4825306993845727e-05,
+      "loss": 2.0971,
+      "step": 8353
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5890199882972156,
+      "learning_rate": 1.4819601361516922e-05,
+      "loss": 1.8812,
+      "step": 8354
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5900389314356714,
+      "learning_rate": 1.4813896364792231e-05,
+      "loss": 1.861,
+      "step": 8355
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6403901725483335,
+      "learning_rate": 1.4808192004027833e-05,
+      "loss": 1.959,
+      "step": 8356
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5522444832337519,
+      "learning_rate": 1.4802488279579873e-05,
+      "loss": 1.9121,
+      "step": 8357
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6040347173582511,
+      "learning_rate": 1.479678519180447e-05,
+      "loss": 2.09,
+      "step": 8358
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5918151870480092,
+      "learning_rate": 1.4791082741057674e-05,
+      "loss": 1.9118,
+      "step": 8359
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6178034941637252,
+      "learning_rate": 1.4785380927695525e-05,
+      "loss": 1.8849,
+      "step": 8360
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5923204492687153,
+      "learning_rate": 1.4779679752074016e-05,
+      "loss": 2.1176,
+      "step": 8361
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5937402765545233,
+      "learning_rate": 1.4773979214549075e-05,
+      "loss": 1.9154,
+      "step": 8362
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5496306359528972,
+      "learning_rate": 1.476827931547663e-05,
+      "loss": 1.8407,
+      "step": 8363
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6835988362926855,
+      "learning_rate": 1.4762580055212535e-05,
+      "loss": 1.8992,
+      "step": 8364
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6618641400451147,
+      "learning_rate": 1.4756881434112621e-05,
+      "loss": 1.876,
+      "step": 8365
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.558850417452291,
+      "learning_rate": 1.4751183452532685e-05,
+      "loss": 2.123,
+      "step": 8366
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6686298556803639,
+      "learning_rate": 1.4745486110828469e-05,
+      "loss": 1.9513,
+      "step": 8367
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6551899067909253,
+      "learning_rate": 1.4739789409355676e-05,
+      "loss": 1.9246,
+      "step": 8368
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.632266763181478,
+      "learning_rate": 1.4734093348469986e-05,
+      "loss": 1.9525,
+      "step": 8369
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6531190688063936,
+      "learning_rate": 1.4728397928527022e-05,
+      "loss": 2.1023,
+      "step": 8370
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5759786680048378,
+      "learning_rate": 1.4722703149882372e-05,
+      "loss": 1.9022,
+      "step": 8371
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6418757468144191,
+      "learning_rate": 1.4717009012891575e-05,
+      "loss": 1.8912,
+      "step": 8372
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6208691702859511,
+      "learning_rate": 1.4711315517910158e-05,
+      "loss": 2.1196,
+      "step": 8373
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6593373464025568,
+      "learning_rate": 1.4705622665293584e-05,
+      "loss": 1.8869,
+      "step": 8374
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6248539678183344,
+      "learning_rate": 1.4699930455397266e-05,
+      "loss": 1.9512,
+      "step": 8375
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6262350862361253,
+      "learning_rate": 1.4694238888576611e-05,
+      "loss": 1.9118,
+      "step": 8376
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7035650364266526,
+      "learning_rate": 1.4688547965186956e-05,
+      "loss": 1.8821,
+      "step": 8377
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6402756077840743,
+      "learning_rate": 1.4682857685583613e-05,
+      "loss": 2.133,
+      "step": 8378
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.757344491442185,
+      "learning_rate": 1.4677168050121853e-05,
+      "loss": 1.8662,
+      "step": 8379
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6612969382278916,
+      "learning_rate": 1.467147905915689e-05,
+      "loss": 1.8899,
+      "step": 8380
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5714825887140408,
+      "learning_rate": 1.4665790713043923e-05,
+      "loss": 1.975,
+      "step": 8381
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.788005066422033,
+      "learning_rate": 1.4660103012138101e-05,
+      "loss": 2.1076,
+      "step": 8382
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5671699072782598,
+      "learning_rate": 1.4654415956794518e-05,
+      "loss": 1.9042,
+      "step": 8383
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7448664976130364,
+      "learning_rate": 1.4648729547368256e-05,
+      "loss": 1.8882,
+      "step": 8384
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5978554749946987,
+      "learning_rate": 1.4643043784214327e-05,
+      "loss": 1.9024,
+      "step": 8385
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6162028055246319,
+      "learning_rate": 1.4637358667687717e-05,
+      "loss": 2.1219,
+      "step": 8386
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6358085826389558,
+      "learning_rate": 1.4631674198143386e-05,
+      "loss": 1.9215,
+      "step": 8387
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6253095673043599,
+      "learning_rate": 1.4625990375936227e-05,
+      "loss": 1.8879,
+      "step": 8388
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5891575975168217,
+      "learning_rate": 1.4620307201421108e-05,
+      "loss": 1.8759,
+      "step": 8389
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5844824307884063,
+      "learning_rate": 1.4614624674952842e-05,
+      "loss": 2.0822,
+      "step": 8390
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6076718295782593,
+      "learning_rate": 1.4608942796886222e-05,
+      "loss": 1.8833,
+      "step": 8391
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.610044044050712,
+      "learning_rate": 1.4603261567576005e-05,
+      "loss": 1.9266,
+      "step": 8392
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6041904689968519,
+      "learning_rate": 1.4597580987376867e-05,
+      "loss": 2.0367,
+      "step": 8393
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5486691567567659,
+      "learning_rate": 1.4591901056643481e-05,
+      "loss": 1.9878,
+      "step": 8394
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.60939536967728,
+      "learning_rate": 1.4586221775730483e-05,
+      "loss": 1.8734,
+      "step": 8395
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7025366799675828,
+      "learning_rate": 1.458054314499242e-05,
+      "loss": 1.9182,
+      "step": 8396
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5760048945283269,
+      "learning_rate": 1.4574865164783874e-05,
+      "loss": 1.9016,
+      "step": 8397
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6066221588001964,
+      "learning_rate": 1.4569187835459313e-05,
+      "loss": 2.093,
+      "step": 8398
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6190251384747406,
+      "learning_rate": 1.4563511157373194e-05,
+      "loss": 1.8641,
+      "step": 8399
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5231950066328885,
+      "learning_rate": 1.455783513087997e-05,
+      "loss": 1.9563,
+      "step": 8400
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6522367159585565,
+      "learning_rate": 1.4552159756333988e-05,
+      "loss": 1.8752,
+      "step": 8401
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6378421827321957,
+      "learning_rate": 1.4546485034089582e-05,
+      "loss": 2.0896,
+      "step": 8402
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5467172471794276,
+      "learning_rate": 1.4540810964501078e-05,
+      "loss": 1.9024,
+      "step": 8403
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6147354972623675,
+      "learning_rate": 1.4535137547922705e-05,
+      "loss": 1.8624,
+      "step": 8404
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6367802605389595,
+      "learning_rate": 1.4529464784708685e-05,
+      "loss": 2.1177,
+      "step": 8405
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6344875033579794,
+      "learning_rate": 1.4523792675213191e-05,
+      "loss": 1.9227,
+      "step": 8406
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5676284131238892,
+      "learning_rate": 1.4518121219790349e-05,
+      "loss": 1.8619,
+      "step": 8407
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6291751101252852,
+      "learning_rate": 1.4512450418794279e-05,
+      "loss": 1.8821,
+      "step": 8408
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6382279217446245,
+      "learning_rate": 1.4506780272579004e-05,
+      "loss": 1.9027,
+      "step": 8409
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5791729965552005,
+      "learning_rate": 1.4501110781498545e-05,
+      "loss": 2.0835,
+      "step": 8410
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5808640192402451,
+      "learning_rate": 1.4495441945906868e-05,
+      "loss": 1.9074,
+      "step": 8411
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7144172097834264,
+      "learning_rate": 1.4489773766157905e-05,
+      "loss": 1.9677,
+      "step": 8412
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5637800667798738,
+      "learning_rate": 1.4484106242605544e-05,
+      "loss": 1.8858,
+      "step": 8413
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5957657132017352,
+      "learning_rate": 1.4478439375603628e-05,
+      "loss": 2.0952,
+      "step": 8414
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5997853878901057,
+      "learning_rate": 1.4472773165505965e-05,
+      "loss": 1.9064,
+      "step": 8415
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5553566637781663,
+      "learning_rate": 1.4467107612666325e-05,
+      "loss": 1.8395,
+      "step": 8416
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5707236780590035,
+      "learning_rate": 1.4461442717438423e-05,
+      "loss": 1.8524,
+      "step": 8417
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5481949855272942,
+      "learning_rate": 1.4455778480175947e-05,
+      "loss": 2.1056,
+      "step": 8418
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6026943304665449,
+      "learning_rate": 1.4450114901232536e-05,
+      "loss": 1.933,
+      "step": 8419
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6367322090836317,
+      "learning_rate": 1.4444451980961793e-05,
+      "loss": 1.9229,
+      "step": 8420
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5822269286977031,
+      "learning_rate": 1.4438789719717278e-05,
+      "loss": 1.8309,
+      "step": 8421
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5598154256625976,
+      "learning_rate": 1.4433128117852504e-05,
+      "loss": 2.0662,
+      "step": 8422
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6089580740826536,
+      "learning_rate": 1.4427467175720963e-05,
+      "loss": 1.9101,
+      "step": 8423
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6353665177947495,
+      "learning_rate": 1.442180689367606e-05,
+      "loss": 1.8802,
+      "step": 8424
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5852755869134891,
+      "learning_rate": 1.4416147272071217e-05,
+      "loss": 1.9708,
+      "step": 8425
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5779986415234073,
+      "learning_rate": 1.4410488311259779e-05,
+      "loss": 2.046,
+      "step": 8426
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6119835022295036,
+      "learning_rate": 1.4404830011595061e-05,
+      "loss": 1.8668,
+      "step": 8427
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6115490455169098,
+      "learning_rate": 1.439917237343033e-05,
+      "loss": 1.9312,
+      "step": 8428
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5523500629186943,
+      "learning_rate": 1.4393515397118817e-05,
+      "loss": 1.8471,
+      "step": 8429
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6853963367796468,
+      "learning_rate": 1.438785908301371e-05,
+      "loss": 2.0839,
+      "step": 8430
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5360391308820578,
+      "learning_rate": 1.4382203431468166e-05,
+      "loss": 1.946,
+      "step": 8431
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5828613091107977,
+      "learning_rate": 1.4376548442835264e-05,
+      "loss": 1.8952,
+      "step": 8432
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6538969990510641,
+      "learning_rate": 1.4370894117468093e-05,
+      "loss": 1.8855,
+      "step": 8433
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6040078346928618,
+      "learning_rate": 1.4365240455719676e-05,
+      "loss": 2.0953,
+      "step": 8434
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5997250562285082,
+      "learning_rate": 1.4359587457942966e-05,
+      "loss": 1.9147,
+      "step": 8435
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6396231405765763,
+      "learning_rate": 1.4353935124490944e-05,
+      "loss": 1.8607,
+      "step": 8436
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5480410615934718,
+      "learning_rate": 1.4348283455716477e-05,
+      "loss": 1.9447,
+      "step": 8437
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6175805536951733,
+      "learning_rate": 1.4342632451972423e-05,
+      "loss": 2.0425,
+      "step": 8438
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7063913237575837,
+      "learning_rate": 1.4336982113611627e-05,
+      "loss": 1.8255,
+      "step": 8439
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5839859235407079,
+      "learning_rate": 1.4331332440986833e-05,
+      "loss": 1.8658,
+      "step": 8440
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6021326410331537,
+      "learning_rate": 1.432568343445077e-05,
+      "loss": 1.8743,
+      "step": 8441
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6228495152851654,
+      "learning_rate": 1.4320035094356163e-05,
+      "loss": 2.0522,
+      "step": 8442
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5945449277991676,
+      "learning_rate": 1.431438742105562e-05,
+      "loss": 1.9605,
+      "step": 8443
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6207654692319132,
+      "learning_rate": 1.4308740414901786e-05,
+      "loss": 1.9179,
+      "step": 8444
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6675638651978442,
+      "learning_rate": 1.4303094076247198e-05,
+      "loss": 1.8708,
+      "step": 8445
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.579769209260611,
+      "learning_rate": 1.4297448405444386e-05,
+      "loss": 2.0948,
+      "step": 8446
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5662214284765524,
+      "learning_rate": 1.4291803402845852e-05,
+      "loss": 1.9296,
+      "step": 8447
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.600883709921342,
+      "learning_rate": 1.4286159068804012e-05,
+      "loss": 1.8854,
+      "step": 8448
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5407549030319233,
+      "learning_rate": 1.4280515403671269e-05,
+      "loss": 1.9608,
+      "step": 8449
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6162919457358811,
+      "learning_rate": 1.4274872407800002e-05,
+      "loss": 2.0603,
+      "step": 8450
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6100484827663457,
+      "learning_rate": 1.4269230081542504e-05,
+      "loss": 1.9084,
+      "step": 8451
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.636818154258723,
+      "learning_rate": 1.426358842525105e-05,
+      "loss": 1.8274,
+      "step": 8452
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6708886457156216,
+      "learning_rate": 1.4257947439277883e-05,
+      "loss": 1.8816,
+      "step": 8453
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5963036956356728,
+      "learning_rate": 1.4252307123975184e-05,
+      "loss": 2.0885,
+      "step": 8454
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5860545459372719,
+      "learning_rate": 1.4246667479695102e-05,
+      "loss": 1.8928,
+      "step": 8455
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5687690006596076,
+      "learning_rate": 1.424102850678975e-05,
+      "loss": 1.9581,
+      "step": 8456
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5661182447874715,
+      "learning_rate": 1.4235390205611182e-05,
+      "loss": 1.9138,
+      "step": 8457
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6227025414832638,
+      "learning_rate": 1.422975257651143e-05,
+      "loss": 2.0783,
+      "step": 8458
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6077340263005613,
+      "learning_rate": 1.4224115619842466e-05,
+      "loss": 1.9016,
+      "step": 8459
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5448702481143165,
+      "learning_rate": 1.4218479335956237e-05,
+      "loss": 1.7924,
+      "step": 8460
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6442842715583527,
+      "learning_rate": 1.4212843725204627e-05,
+      "loss": 1.8729,
+      "step": 8461
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.610442592279379,
+      "learning_rate": 1.4207208787939504e-05,
+      "loss": 2.0597,
+      "step": 8462
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.571406911976173,
+      "learning_rate": 1.4201574524512673e-05,
+      "loss": 1.9445,
+      "step": 8463
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5648312902225111,
+      "learning_rate": 1.4195940935275906e-05,
+      "loss": 1.8947,
+      "step": 8464
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7253143853235747,
+      "learning_rate": 1.4190308020580931e-05,
+      "loss": 1.8727,
+      "step": 8465
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5946337252723355,
+      "learning_rate": 1.4184675780779433e-05,
+      "loss": 2.0862,
+      "step": 8466
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.582638211781166,
+      "learning_rate": 1.4179044216223054e-05,
+      "loss": 1.8253,
+      "step": 8467
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6323046328427233,
+      "learning_rate": 1.4173413327263401e-05,
+      "loss": 1.9609,
+      "step": 8468
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5870801734128277,
+      "learning_rate": 1.4167783114252031e-05,
+      "loss": 1.9358,
+      "step": 8469
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5997595708730505,
+      "learning_rate": 1.4162153577540472e-05,
+      "loss": 2.0762,
+      "step": 8470
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6321850180025512,
+      "learning_rate": 1.4156524717480162e-05,
+      "loss": 1.9419,
+      "step": 8471
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5563100483904052,
+      "learning_rate": 1.4150896534422574e-05,
+      "loss": 1.8244,
+      "step": 8472
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5696775635313168,
+      "learning_rate": 1.4145269028719096e-05,
+      "loss": 1.8365,
+      "step": 8473
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6878602264916693,
+      "learning_rate": 1.4139642200721043e-05,
+      "loss": 2.0902,
+      "step": 8474
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6334518736809681,
+      "learning_rate": 1.4134016050779755e-05,
+      "loss": 1.9119,
+      "step": 8475
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5501471026815904,
+      "learning_rate": 1.412839057924648e-05,
+      "loss": 1.9088,
+      "step": 8476
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7509255308771137,
+      "learning_rate": 1.4122765786472447e-05,
+      "loss": 1.8323,
+      "step": 8477
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6452526786712618,
+      "learning_rate": 1.4117141672808837e-05,
+      "loss": 2.0613,
+      "step": 8478
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5759912342349364,
+      "learning_rate": 1.4111518238606763e-05,
+      "loss": 1.856,
+      "step": 8479
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.551535028670185,
+      "learning_rate": 1.4105895484217346e-05,
+      "loss": 1.9951,
+      "step": 8480
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6794228304001027,
+      "learning_rate": 1.4100273409991638e-05,
+      "loss": 1.8793,
+      "step": 8481
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.658221075901024,
+      "learning_rate": 1.4094652016280618e-05,
+      "loss": 2.0915,
+      "step": 8482
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5688625301369679,
+      "learning_rate": 1.4089031303435279e-05,
+      "loss": 1.8485,
+      "step": 8483
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6129076960848894,
+      "learning_rate": 1.4083411271806554e-05,
+      "loss": 1.8873,
+      "step": 8484
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5701859911290824,
+      "learning_rate": 1.4077791921745286e-05,
+      "loss": 1.9065,
+      "step": 8485
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6010933927380878,
+      "learning_rate": 1.4072173253602364e-05,
+      "loss": 2.0676,
+      "step": 8486
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5426578050004103,
+      "learning_rate": 1.4066555267728543e-05,
+      "loss": 1.9963,
+      "step": 8487
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5553893945729792,
+      "learning_rate": 1.4060937964474585e-05,
+      "loss": 1.8465,
+      "step": 8488
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6730931781520945,
+      "learning_rate": 1.4055321344191228e-05,
+      "loss": 1.9021,
+      "step": 8489
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5653963508369981,
+      "learning_rate": 1.4049705407229103e-05,
+      "loss": 2.0911,
+      "step": 8490
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5862701075061838,
+      "learning_rate": 1.4044090153938871e-05,
+      "loss": 1.875,
+      "step": 8491
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5737780929865324,
+      "learning_rate": 1.403847558467109e-05,
+      "loss": 1.8891,
+      "step": 8492
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5486844852116715,
+      "learning_rate": 1.4032861699776297e-05,
+      "loss": 1.955,
+      "step": 8493
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6260768524180835,
+      "learning_rate": 1.4027248499605028e-05,
+      "loss": 2.0892,
+      "step": 8494
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5527390406160034,
+      "learning_rate": 1.4021635984507698e-05,
+      "loss": 1.9001,
+      "step": 8495
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5770027029984508,
+      "learning_rate": 1.4016024154834722e-05,
+      "loss": 1.8301,
+      "step": 8496
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5907552349465512,
+      "learning_rate": 1.4010413010936508e-05,
+      "loss": 1.9299,
+      "step": 8497
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6062324672381054,
+      "learning_rate": 1.400480255316334e-05,
+      "loss": 2.0519,
+      "step": 8498
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5978566604214869,
+      "learning_rate": 1.399919278186552e-05,
+      "loss": 1.9683,
+      "step": 8499
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5792716538486619,
+      "learning_rate": 1.3993583697393287e-05,
+      "loss": 1.8334,
+      "step": 8500
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6486085030997955,
+      "learning_rate": 1.398797530009684e-05,
+      "loss": 1.9165,
+      "step": 8501
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6364920988361558,
+      "learning_rate": 1.3982367590326334e-05,
+      "loss": 2.1785,
+      "step": 8502
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6034087084768031,
+      "learning_rate": 1.397676056843188e-05,
+      "loss": 1.8647,
+      "step": 8503
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6287188234446461,
+      "learning_rate": 1.3971154234763545e-05,
+      "loss": 1.8712,
+      "step": 8504
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7453187348247245,
+      "learning_rate": 1.3965548589671363e-05,
+      "loss": 1.9466,
+      "step": 8505
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7166162908838116,
+      "learning_rate": 1.3959943633505312e-05,
+      "loss": 2.0211,
+      "step": 8506
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6406683667796296,
+      "learning_rate": 1.3954339366615334e-05,
+      "loss": 1.8351,
+      "step": 8507
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7466193511423979,
+      "learning_rate": 1.3948735789351325e-05,
+      "loss": 1.8836,
+      "step": 8508
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.621780792721436,
+      "learning_rate": 1.3943132902063144e-05,
+      "loss": 1.917,
+      "step": 8509
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.57014835824952,
+      "learning_rate": 1.3937530705100593e-05,
+      "loss": 2.0921,
+      "step": 8510
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6341438096239674,
+      "learning_rate": 1.3931929198813451e-05,
+      "loss": 1.9834,
+      "step": 8511
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6290707977599788,
+      "learning_rate": 1.3926328383551434e-05,
+      "loss": 1.8971,
+      "step": 8512
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6191908593992131,
+      "learning_rate": 1.3920728259664232e-05,
+      "loss": 1.8975,
+      "step": 8513
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5833253584950423,
+      "learning_rate": 1.3915128827501475e-05,
+      "loss": 2.0609,
+      "step": 8514
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6005955911894293,
+      "learning_rate": 1.3909530087412765e-05,
+      "loss": 1.8793,
+      "step": 8515
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5971804806446672,
+      "learning_rate": 1.3903932039747653e-05,
+      "loss": 1.8696,
+      "step": 8516
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.58994638363175,
+      "learning_rate": 1.3898334684855647e-05,
+      "loss": 1.8766,
+      "step": 8517
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5751458873973857,
+      "learning_rate": 1.3892738023086216e-05,
+      "loss": 1.9172,
+      "step": 8518
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5993312938921292,
+      "learning_rate": 1.3887142054788777e-05,
+      "loss": 2.0507,
+      "step": 8519
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5672412881572056,
+      "learning_rate": 1.3881546780312721e-05,
+      "loss": 1.9102,
+      "step": 8520
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5807934501430776,
+      "learning_rate": 1.3875952200007358e-05,
+      "loss": 1.9018,
+      "step": 8521
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6413226685929957,
+      "learning_rate": 1.3870358314222007e-05,
+      "loss": 2.0798,
+      "step": 8522
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.580456197932248,
+      "learning_rate": 1.3864765123305906e-05,
+      "loss": 1.8941,
+      "step": 8523
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5822159749078407,
+      "learning_rate": 1.3859172627608266e-05,
+      "loss": 1.9366,
+      "step": 8524
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6010838648521095,
+      "learning_rate": 1.3853580827478257e-05,
+      "loss": 1.8688,
+      "step": 8525
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5884866549209309,
+      "learning_rate": 1.3847989723264964e-05,
+      "loss": 2.0797,
+      "step": 8526
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.545073034576549,
+      "learning_rate": 1.38423993153175e-05,
+      "loss": 1.8967,
+      "step": 8527
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5609372254453845,
+      "learning_rate": 1.383680960398489e-05,
+      "loss": 1.8685,
+      "step": 8528
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5645828410811794,
+      "learning_rate": 1.3831220589616101e-05,
+      "loss": 1.8775,
+      "step": 8529
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5511053468525992,
+      "learning_rate": 1.3825632272560096e-05,
+      "loss": 2.0025,
+      "step": 8530
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6517839308943691,
+      "learning_rate": 1.382004465316579e-05,
+      "loss": 2.0883,
+      "step": 8531
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.563880873236184,
+      "learning_rate": 1.3814457731782003e-05,
+      "loss": 1.9002,
+      "step": 8532
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5884613126007006,
+      "learning_rate": 1.3808871508757587e-05,
+      "loss": 1.913,
+      "step": 8533
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6063064397574519,
+      "learning_rate": 1.3803285984441294e-05,
+      "loss": 2.0526,
+      "step": 8534
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5800230364768042,
+      "learning_rate": 1.3797701159181839e-05,
+      "loss": 1.8493,
+      "step": 8535
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5499297325574661,
+      "learning_rate": 1.3792117033327942e-05,
+      "loss": 1.9589,
+      "step": 8536
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6478588615736024,
+      "learning_rate": 1.3786533607228214e-05,
+      "loss": 1.8968,
+      "step": 8537
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6329771958707212,
+      "learning_rate": 1.3780950881231256e-05,
+      "loss": 2.0763,
+      "step": 8538
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5358917360353697,
+      "learning_rate": 1.3775368855685625e-05,
+      "loss": 1.8916,
+      "step": 8539
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.637109025169848,
+      "learning_rate": 1.3769787530939817e-05,
+      "loss": 1.891,
+      "step": 8540
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5673965260128847,
+      "learning_rate": 1.3764206907342331e-05,
+      "loss": 1.8718,
+      "step": 8541
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6521987901552707,
+      "learning_rate": 1.3758626985241552e-05,
+      "loss": 2.0852,
+      "step": 8542
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5571003005752742,
+      "learning_rate": 1.3753047764985864e-05,
+      "loss": 1.9525,
+      "step": 8543
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6230516333206275,
+      "learning_rate": 1.3747469246923628e-05,
+      "loss": 1.9026,
+      "step": 8544
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6340404145684204,
+      "learning_rate": 1.3741891431403098e-05,
+      "loss": 1.8577,
+      "step": 8545
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6600575389834339,
+      "learning_rate": 1.3736314318772541e-05,
+      "loss": 2.0429,
+      "step": 8546
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6498873619680668,
+      "learning_rate": 1.3730737909380154e-05,
+      "loss": 1.8626,
+      "step": 8547
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5623245105571738,
+      "learning_rate": 1.372516220357409e-05,
+      "loss": 1.8865,
+      "step": 8548
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6607093895623688,
+      "learning_rate": 1.3719587201702471e-05,
+      "loss": 1.9498,
+      "step": 8549
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6563606008018121,
+      "learning_rate": 1.3714012904113363e-05,
+      "loss": 1.8501,
+      "step": 8550
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6198558532331665,
+      "learning_rate": 1.3708439311154794e-05,
+      "loss": 2.0818,
+      "step": 8551
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5700943533081028,
+      "learning_rate": 1.3702866423174746e-05,
+      "loss": 1.8936,
+      "step": 8552
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6431288884653616,
+      "learning_rate": 1.3697294240521155e-05,
+      "loss": 1.8353,
+      "step": 8553
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7200407733227885,
+      "learning_rate": 1.369172276354192e-05,
+      "loss": 2.0534,
+      "step": 8554
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5515226836936954,
+      "learning_rate": 1.3686151992584887e-05,
+      "loss": 1.9655,
+      "step": 8555
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5671820526737776,
+      "learning_rate": 1.3680581927997866e-05,
+      "loss": 1.8547,
+      "step": 8556
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6704553659143992,
+      "learning_rate": 1.3675012570128612e-05,
+      "loss": 1.8765,
+      "step": 8557
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.592979961376165,
+      "learning_rate": 1.3669443919324849e-05,
+      "loss": 2.0761,
+      "step": 8558
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6612134705520717,
+      "learning_rate": 1.3663875975934248e-05,
+      "loss": 1.9118,
+      "step": 8559
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6290439276435041,
+      "learning_rate": 1.3658308740304442e-05,
+      "loss": 1.9336,
+      "step": 8560
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6302113149443741,
+      "learning_rate": 1.3652742212783015e-05,
+      "loss": 1.9294,
+      "step": 8561
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.659622096518625,
+      "learning_rate": 1.3647176393717509e-05,
+      "loss": 1.9164,
+      "step": 8562
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5893036326325897,
+      "learning_rate": 1.3641611283455414e-05,
+      "loss": 2.0492,
+      "step": 8563
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6128457363763159,
+      "learning_rate": 1.3636046882344191e-05,
+      "loss": 1.8784,
+      "step": 8564
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5718976069132464,
+      "learning_rate": 1.3630483190731246e-05,
+      "loss": 1.9224,
+      "step": 8565
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6117637899435415,
+      "learning_rate": 1.3624920208963943e-05,
+      "loss": 2.0845,
+      "step": 8566
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5645666454087888,
+      "learning_rate": 1.3619357937389607e-05,
+      "loss": 1.8944,
+      "step": 8567
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6152548072642815,
+      "learning_rate": 1.3613796376355493e-05,
+      "loss": 1.8912,
+      "step": 8568
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5636969463493041,
+      "learning_rate": 1.3608235526208851e-05,
+      "loss": 1.8919,
+      "step": 8569
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.552762268172415,
+      "learning_rate": 1.3602675387296878e-05,
+      "loss": 2.0277,
+      "step": 8570
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5999415492243528,
+      "learning_rate": 1.3597115959966683e-05,
+      "loss": 1.9362,
+      "step": 8571
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6282172049431161,
+      "learning_rate": 1.3591557244565401e-05,
+      "loss": 1.9132,
+      "step": 8572
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5534342352584115,
+      "learning_rate": 1.3585999241440046e-05,
+      "loss": 1.8664,
+      "step": 8573
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6279400443349271,
+      "learning_rate": 1.358044195093766e-05,
+      "loss": 1.9564,
+      "step": 8574
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6930726390298751,
+      "learning_rate": 1.3574885373405205e-05,
+      "loss": 2.0806,
+      "step": 8575
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6196094716214805,
+      "learning_rate": 1.3569329509189572e-05,
+      "loss": 1.8736,
+      "step": 8576
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5951292243552486,
+      "learning_rate": 1.3563774358637666e-05,
+      "loss": 1.8785,
+      "step": 8577
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7286490270615972,
+      "learning_rate": 1.3558219922096318e-05,
+      "loss": 2.0865,
+      "step": 8578
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5723622479482635,
+      "learning_rate": 1.3552666199912285e-05,
+      "loss": 1.8561,
+      "step": 8579
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6699243728236776,
+      "learning_rate": 1.3547113192432348e-05,
+      "loss": 1.9449,
+      "step": 8580
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.604466441751994,
+      "learning_rate": 1.3541560900003172e-05,
+      "loss": 1.8642,
+      "step": 8581
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6409522116104243,
+      "learning_rate": 1.3536009322971417e-05,
+      "loss": 1.824,
+      "step": 8582
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6572523274642812,
+      "learning_rate": 1.353045846168371e-05,
+      "loss": 2.0556,
+      "step": 8583
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.610483919632084,
+      "learning_rate": 1.352490831648659e-05,
+      "loss": 1.8716,
+      "step": 8584
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6766008775200689,
+      "learning_rate": 1.3519358887726585e-05,
+      "loss": 1.9247,
+      "step": 8585
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6005071354693734,
+      "learning_rate": 1.3513810175750167e-05,
+      "loss": 1.991,
+      "step": 8586
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6092829767934,
+      "learning_rate": 1.3508262180903757e-05,
+      "loss": 2.0896,
+      "step": 8587
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7091040707658487,
+      "learning_rate": 1.350271490353377e-05,
+      "loss": 1.8663,
+      "step": 8588
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6521786289256927,
+      "learning_rate": 1.349716834398651e-05,
+      "loss": 1.8663,
+      "step": 8589
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.581375004306449,
+      "learning_rate": 1.3491622502608278e-05,
+      "loss": 2.0701,
+      "step": 8590
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6356977326703559,
+      "learning_rate": 1.3486077379745351e-05,
+      "loss": 1.9062,
+      "step": 8591
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5360480351020158,
+      "learning_rate": 1.3480532975743903e-05,
+      "loss": 1.9742,
+      "step": 8592
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6339020989294368,
+      "learning_rate": 1.3474989290950105e-05,
+      "loss": 1.9318,
+      "step": 8593
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6414836588003963,
+      "learning_rate": 1.346944632571007e-05,
+      "loss": 1.818,
+      "step": 8594
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.599580117894308,
+      "learning_rate": 1.3463904080369872e-05,
+      "loss": 2.0789,
+      "step": 8595
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5615634126070519,
+      "learning_rate": 1.3458362555275531e-05,
+      "loss": 1.8855,
+      "step": 8596
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6112730891410497,
+      "learning_rate": 1.3452821750773032e-05,
+      "loss": 1.8439,
+      "step": 8597
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5481926855027783,
+      "learning_rate": 1.3447281667208305e-05,
+      "loss": 1.9753,
+      "step": 8598
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5859228881668126,
+      "learning_rate": 1.3441742304927247e-05,
+      "loss": 2.0449,
+      "step": 8599
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6293356603405511,
+      "learning_rate": 1.3436203664275699e-05,
+      "loss": 1.8994,
+      "step": 8600
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5443950816119576,
+      "learning_rate": 1.3430665745599458e-05,
+      "loss": 1.907,
+      "step": 8601
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5837096071741881,
+      "learning_rate": 1.3425128549244284e-05,
+      "loss": 1.9029,
+      "step": 8602
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5931245888165687,
+      "learning_rate": 1.3419592075555887e-05,
+      "loss": 2.1035,
+      "step": 8603
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5616027434009185,
+      "learning_rate": 1.3414056324879931e-05,
+      "loss": 1.8675,
+      "step": 8604
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5342495136644386,
+      "learning_rate": 1.3408521297562032e-05,
+      "loss": 1.9561,
+      "step": 8605
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5621810607453066,
+      "learning_rate": 1.340298699394777e-05,
+      "loss": 1.9151,
+      "step": 8606
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5983527875689366,
+      "learning_rate": 1.3397453414382668e-05,
+      "loss": 2.0649,
+      "step": 8607
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6036558960586824,
+      "learning_rate": 1.3391920559212218e-05,
+      "loss": 1.8935,
+      "step": 8608
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6834322158330604,
+      "learning_rate": 1.3386388428781854e-05,
+      "loss": 1.8728,
+      "step": 8609
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.556994328282094,
+      "learning_rate": 1.3380857023436966e-05,
+      "loss": 2.1294,
+      "step": 8610
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5596047244802386,
+      "learning_rate": 1.3375326343522912e-05,
+      "loss": 1.9589,
+      "step": 8611
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.7035321085298167,
+      "learning_rate": 1.3369796389384987e-05,
+      "loss": 1.8387,
+      "step": 8612
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6056846529259186,
+      "learning_rate": 1.3364267161368452e-05,
+      "loss": 1.8656,
+      "step": 8613
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6670725191531433,
+      "learning_rate": 1.3358738659818529e-05,
+      "loss": 1.8667,
+      "step": 8614
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6668562559978036,
+      "learning_rate": 1.3353210885080356e-05,
+      "loss": 2.0726,
+      "step": 8615
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6006221780662695,
+      "learning_rate": 1.3347683837499081e-05,
+      "loss": 1.8713,
+      "step": 8616
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5437169844469553,
+      "learning_rate": 1.3342157517419784e-05,
+      "loss": 1.9375,
+      "step": 8617
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6638882216412134,
+      "learning_rate": 1.3336631925187464e-05,
+      "loss": 1.8647,
+      "step": 8618
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6079776967948477,
+      "learning_rate": 1.3331107061147152e-05,
+      "loss": 2.0908,
+      "step": 8619
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5663355289423392,
+      "learning_rate": 1.3325582925643736e-05,
+      "loss": 1.8549,
+      "step": 8620
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5505350455027813,
+      "learning_rate": 1.3320059519022155e-05,
+      "loss": 1.8134,
+      "step": 8621
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6372308119260377,
+      "learning_rate": 1.3314536841627245e-05,
+      "loss": 2.0911,
+      "step": 8622
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6125913963207185,
+      "learning_rate": 1.330901489380379e-05,
+      "loss": 1.9626,
+      "step": 8623
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5903497701163174,
+      "learning_rate": 1.3303493675896567e-05,
+      "loss": 1.8914,
+      "step": 8624
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5583309349050712,
+      "learning_rate": 1.3297973188250298e-05,
+      "loss": 1.8922,
+      "step": 8625
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6648625876564936,
+      "learning_rate": 1.3292453431209616e-05,
+      "loss": 1.9234,
+      "step": 8626
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6481893625143103,
+      "learning_rate": 1.3286934405119183e-05,
+      "loss": 2.1129,
+      "step": 8627
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6127304809058359,
+      "learning_rate": 1.3281416110323541e-05,
+      "loss": 1.8729,
+      "step": 8628
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5871048174818186,
+      "learning_rate": 1.3275898547167223e-05,
+      "loss": 1.97,
+      "step": 8629
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6042252402753703,
+      "learning_rate": 1.3270381715994742e-05,
+      "loss": 1.898,
+      "step": 8630
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6207772287126526,
+      "learning_rate": 1.3264865617150508e-05,
+      "loss": 2.0631,
+      "step": 8631
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5761240972595963,
+      "learning_rate": 1.3259350250978914e-05,
+      "loss": 1.9032,
+      "step": 8632
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6168258906492459,
+      "learning_rate": 1.3253835617824333e-05,
+      "loss": 1.9284,
+      "step": 8633
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6313108853701159,
+      "learning_rate": 1.324832171803104e-05,
+      "loss": 1.8823,
+      "step": 8634
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5543843764108831,
+      "learning_rate": 1.3242808551943297e-05,
+      "loss": 2.0722,
+      "step": 8635
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6124466219280645,
+      "learning_rate": 1.323729611990532e-05,
+      "loss": 1.992,
+      "step": 8636
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5751113074120218,
+      "learning_rate": 1.3231784422261256e-05,
+      "loss": 1.8129,
+      "step": 8637
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5787250491079584,
+      "learning_rate": 1.322627345935526e-05,
+      "loss": 1.8522,
+      "step": 8638
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.61873852970918,
+      "learning_rate": 1.3220763231531363e-05,
+      "loss": 2.0986,
+      "step": 8639
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6139043570440373,
+      "learning_rate": 1.321525373913361e-05,
+      "loss": 1.8932,
+      "step": 8640
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.647349524027658,
+      "learning_rate": 1.3209744982505981e-05,
+      "loss": 1.8338,
+      "step": 8641
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5659427456241436,
+      "learning_rate": 1.3204236961992406e-05,
+      "loss": 1.9393,
+      "step": 8642
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5873982729163277,
+      "learning_rate": 1.3198729677936777e-05,
+      "loss": 2.1229,
+      "step": 8643
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6567380965448689,
+      "learning_rate": 1.3193223130682936e-05,
+      "loss": 1.8528,
+      "step": 8644
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6325024617637471,
+      "learning_rate": 1.3187717320574678e-05,
+      "loss": 1.9045,
+      "step": 8645
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5699967685561543,
+      "learning_rate": 1.318221224795575e-05,
+      "loss": 1.8783,
+      "step": 8646
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7058278910740858,
+      "learning_rate": 1.3176707913169864e-05,
+      "loss": 2.0709,
+      "step": 8647
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6344523613551085,
+      "learning_rate": 1.3171204316560672e-05,
+      "loss": 1.9545,
+      "step": 8648
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5531069157011015,
+      "learning_rate": 1.3165701458471791e-05,
+      "loss": 1.9011,
+      "step": 8649
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6963226978869259,
+      "learning_rate": 1.3160199339246782e-05,
+      "loss": 1.8903,
+      "step": 8650
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6847067252507361,
+      "learning_rate": 1.3154697959229171e-05,
+      "loss": 2.0682,
+      "step": 8651
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5702210634528158,
+      "learning_rate": 1.3149197318762425e-05,
+      "loss": 1.8759,
+      "step": 8652
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6337334650472904,
+      "learning_rate": 1.3143697418189976e-05,
+      "loss": 1.8783,
+      "step": 8653
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6429446647410276,
+      "learning_rate": 1.31381982578552e-05,
+      "loss": 1.9204,
+      "step": 8654
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5727540017137633,
+      "learning_rate": 1.3132699838101442e-05,
+      "loss": 2.0938,
+      "step": 8655
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6290632682576639,
+      "learning_rate": 1.3127202159271983e-05,
+      "loss": 1.8858,
+      "step": 8656
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6711671010643725,
+      "learning_rate": 1.3121705221710068e-05,
+      "loss": 1.9096,
+      "step": 8657
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6011141916907088,
+      "learning_rate": 1.3116209025758891e-05,
+      "loss": 1.8669,
+      "step": 8658
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6049575083717088,
+      "learning_rate": 1.3110713571761607e-05,
+      "loss": 2.053,
+      "step": 8659
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5716973986454544,
+      "learning_rate": 1.3105218860061316e-05,
+      "loss": 2.0254,
+      "step": 8660
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6318554013884782,
+      "learning_rate": 1.3099724891001086e-05,
+      "loss": 1.8212,
+      "step": 8661
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5822776419222844,
+      "learning_rate": 1.30942316649239e-05,
+      "loss": 1.8674,
+      "step": 8662
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6539473255754054,
+      "learning_rate": 1.3088739182172754e-05,
+      "loss": 2.0807,
+      "step": 8663
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5726907059048509,
+      "learning_rate": 1.3083247443090557e-05,
+      "loss": 1.8876,
+      "step": 8664
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6364942106172853,
+      "learning_rate": 1.3077756448020163e-05,
+      "loss": 1.9013,
+      "step": 8665
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6030113906748293,
+      "learning_rate": 1.307226619730442e-05,
+      "loss": 1.8653,
+      "step": 8666
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.60789121040305,
+      "learning_rate": 1.3066776691286108e-05,
+      "loss": 2.1202,
+      "step": 8667
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6401168588582484,
+      "learning_rate": 1.306128793030793e-05,
+      "loss": 1.8666,
+      "step": 8668
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6399983815398669,
+      "learning_rate": 1.3055799914712613e-05,
+      "loss": 1.8938,
+      "step": 8669
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5508051395557823,
+      "learning_rate": 1.3050312644842755e-05,
+      "loss": 1.8825,
+      "step": 8670
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6947899353881971,
+      "learning_rate": 1.3044826121040976e-05,
+      "loss": 2.0546,
+      "step": 8671
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6564700720229625,
+      "learning_rate": 1.303934034364983e-05,
+      "loss": 1.8261,
+      "step": 8672
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5809735156918099,
+      "learning_rate": 1.3033855313011779e-05,
+      "loss": 1.9565,
+      "step": 8673
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5619105179484444,
+      "learning_rate": 1.3028371029469321e-05,
+      "loss": 1.8757,
+      "step": 8674
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6853825321241442,
+      "learning_rate": 1.3022887493364828e-05,
+      "loss": 2.0622,
+      "step": 8675
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6197992129906217,
+      "learning_rate": 1.3017404705040665e-05,
+      "loss": 1.882,
+      "step": 8676
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6033869209812119,
+      "learning_rate": 1.301192266483917e-05,
+      "loss": 1.8982,
+      "step": 8677
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6420913792748507,
+      "learning_rate": 1.3006441373102582e-05,
+      "loss": 1.8986,
+      "step": 8678
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6021570211050509,
+      "learning_rate": 1.300096083017312e-05,
+      "loss": 2.1299,
+      "step": 8679
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6610012231118445,
+      "learning_rate": 1.2995481036392987e-05,
+      "loss": 1.8846,
+      "step": 8680
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5652170866260483,
+      "learning_rate": 1.2990001992104275e-05,
+      "loss": 1.8632,
+      "step": 8681
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.579067354333605,
+      "learning_rate": 1.2984523697649082e-05,
+      "loss": 1.9132,
+      "step": 8682
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6874822570013811,
+      "learning_rate": 1.2979046153369431e-05,
+      "loss": 2.0483,
+      "step": 8683
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5956231186543932,
+      "learning_rate": 1.2973569359607307e-05,
+      "loss": 1.8835,
+      "step": 8684
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5349684656914901,
+      "learning_rate": 1.2968093316704668e-05,
+      "loss": 1.9981,
+      "step": 8685
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6591905216131088,
+      "learning_rate": 1.2962618025003387e-05,
+      "loss": 1.9206,
+      "step": 8686
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5736513586558483,
+      "learning_rate": 1.2957143484845307e-05,
+      "loss": 2.0825,
+      "step": 8687
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.581533273416157,
+      "learning_rate": 1.2951669696572239e-05,
+      "loss": 1.8792,
+      "step": 8688
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5878595518519131,
+      "learning_rate": 1.2946196660525925e-05,
+      "loss": 1.8829,
+      "step": 8689
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5631029998026801,
+      "learning_rate": 1.2940724377048072e-05,
+      "loss": 1.8717,
+      "step": 8690
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6711591498302996,
+      "learning_rate": 1.2935252846480336e-05,
+      "loss": 2.0881,
+      "step": 8691
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6632857706655103,
+      "learning_rate": 1.2929782069164328e-05,
+      "loss": 1.8821,
+      "step": 8692
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5705579918390798,
+      "learning_rate": 1.292431204544161e-05,
+      "loss": 1.8638,
+      "step": 8693
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6644491969516353,
+      "learning_rate": 1.2918842775653703e-05,
+      "loss": 1.8954,
+      "step": 8694
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5713225296310331,
+      "learning_rate": 1.291337426014207e-05,
+      "loss": 2.0361,
+      "step": 8695
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6832915837978382,
+      "learning_rate": 1.2907906499248135e-05,
+      "loss": 1.8719,
+      "step": 8696
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5582669651517246,
+      "learning_rate": 1.2902439493313273e-05,
+      "loss": 1.8815,
+      "step": 8697
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7521352281468984,
+      "learning_rate": 1.2896973242678812e-05,
+      "loss": 1.9321,
+      "step": 8698
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6377691421672715,
+      "learning_rate": 1.2891507747686033e-05,
+      "loss": 1.9999,
+      "step": 8699
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5736105985285261,
+      "learning_rate": 1.2886043008676175e-05,
+      "loss": 1.8454,
+      "step": 8700
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7559359865908138,
+      "learning_rate": 1.28805790259904e-05,
+      "loss": 1.8744,
+      "step": 8701
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6358727315505697,
+      "learning_rate": 1.2875115799969873e-05,
+      "loss": 1.8908,
+      "step": 8702
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5709064312530099,
+      "learning_rate": 1.2869653330955677e-05,
+      "loss": 2.1159,
+      "step": 8703
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6272993111009666,
+      "learning_rate": 1.2864191619288857e-05,
+      "loss": 1.9256,
+      "step": 8704
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6534564457869096,
+      "learning_rate": 1.2858730665310409e-05,
+      "loss": 1.8781,
+      "step": 8705
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6502841609183282,
+      "learning_rate": 1.285327046936128e-05,
+      "loss": 1.8261,
+      "step": 8706
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5600801028258066,
+      "learning_rate": 1.2847811031782381e-05,
+      "loss": 2.0815,
+      "step": 8707
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5605817157645283,
+      "learning_rate": 1.2842352352914566e-05,
+      "loss": 1.8538,
+      "step": 8708
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5952696877176576,
+      "learning_rate": 1.2836894433098623e-05,
+      "loss": 1.8875,
+      "step": 8709
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6160494490497608,
+      "learning_rate": 1.2831437272675337e-05,
+      "loss": 1.9535,
+      "step": 8710
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5622216973261807,
+      "learning_rate": 1.282598087198542e-05,
+      "loss": 2.0664,
+      "step": 8711
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6090350722916298,
+      "learning_rate": 1.2820525231369513e-05,
+      "loss": 1.92,
+      "step": 8712
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6111858315437384,
+      "learning_rate": 1.2815070351168257e-05,
+      "loss": 1.7962,
+      "step": 8713
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6091675126430611,
+      "learning_rate": 1.2809616231722226e-05,
+      "loss": 1.86,
+      "step": 8714
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6199471517289616,
+      "learning_rate": 1.2804162873371917e-05,
+      "loss": 2.073,
+      "step": 8715
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7489872968338708,
+      "learning_rate": 1.2798710276457838e-05,
+      "loss": 1.9163,
+      "step": 8716
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5888493704435825,
+      "learning_rate": 1.279325844132038e-05,
+      "loss": 1.8946,
+      "step": 8717
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5820104097078809,
+      "learning_rate": 1.2787807368299953e-05,
+      "loss": 1.8968,
+      "step": 8718
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6798128080974845,
+      "learning_rate": 1.2782357057736893e-05,
+      "loss": 2.0939,
+      "step": 8719
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6774942971158335,
+      "learning_rate": 1.2776907509971453e-05,
+      "loss": 1.8877,
+      "step": 8720
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6069237605141844,
+      "learning_rate": 1.2771458725343913e-05,
+      "loss": 1.8867,
+      "step": 8721
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6015872260863675,
+      "learning_rate": 1.2766010704194429e-05,
+      "loss": 1.877,
+      "step": 8722
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5476016472505173,
+      "learning_rate": 1.2760563446863144e-05,
+      "loss": 2.1346,
+      "step": 8723
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5648279566715566,
+      "learning_rate": 1.2755116953690185e-05,
+      "loss": 1.8936,
+      "step": 8724
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6231963609137519,
+      "learning_rate": 1.274967122501557e-05,
+      "loss": 1.9462,
+      "step": 8725
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6387268435225342,
+      "learning_rate": 1.2744226261179292e-05,
+      "loss": 1.9723,
+      "step": 8726
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5736787226257911,
+      "learning_rate": 1.2738782062521338e-05,
+      "loss": 2.0329,
+      "step": 8727
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6097582913622503,
+      "learning_rate": 1.273333862938158e-05,
+      "loss": 1.8371,
+      "step": 8728
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.617895908236208,
+      "learning_rate": 1.2727895962099886e-05,
+      "loss": 1.9564,
+      "step": 8729
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5677837355447011,
+      "learning_rate": 1.2722454061016059e-05,
+      "loss": 1.8741,
+      "step": 8730
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6623722375513699,
+      "learning_rate": 1.2717012926469867e-05,
+      "loss": 2.0816,
+      "step": 8731
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6722682742782295,
+      "learning_rate": 1.2711572558801016e-05,
+      "loss": 1.9458,
+      "step": 8732
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6579497860217047,
+      "learning_rate": 1.2706132958349171e-05,
+      "loss": 1.8893,
+      "step": 8733
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6600464651354379,
+      "learning_rate": 1.2700694125453949e-05,
+      "loss": 1.8756,
+      "step": 8734
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6848976771683926,
+      "learning_rate": 1.2695256060454922e-05,
+      "loss": 1.9648,
+      "step": 8735
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7453968710889805,
+      "learning_rate": 1.2689818763691608e-05,
+      "loss": 2.0642,
+      "step": 8736
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5756189041668505,
+      "learning_rate": 1.2684382235503478e-05,
+      "loss": 1.9392,
+      "step": 8737
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6617490335438091,
+      "learning_rate": 1.2678946476229959e-05,
+      "loss": 1.8536,
+      "step": 8738
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7506384916913145,
+      "learning_rate": 1.2673511486210429e-05,
+      "loss": 2.1051,
+      "step": 8739
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.581873285162329,
+      "learning_rate": 1.2668077265784214e-05,
+      "loss": 1.8506,
+      "step": 8740
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.595695703777978,
+      "learning_rate": 1.2662643815290598e-05,
+      "loss": 1.9281,
+      "step": 8741
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6947511778094969,
+      "learning_rate": 1.265721113506881e-05,
+      "loss": 1.8439,
+      "step": 8742
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6449173664146374,
+      "learning_rate": 1.2651779225458038e-05,
+      "loss": 1.998,
+      "step": 8743
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6224350317698285,
+      "learning_rate": 1.2646348086797417e-05,
+      "loss": 1.8989,
+      "step": 8744
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6790410130954976,
+      "learning_rate": 1.2640917719426032e-05,
+      "loss": 1.9553,
+      "step": 8745
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.583889671221694,
+      "learning_rate": 1.263548812368293e-05,
+      "loss": 1.9089,
+      "step": 8746
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5969628962423894,
+      "learning_rate": 1.2630059299907096e-05,
+      "loss": 1.9486,
+      "step": 8747
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5840452399107695,
+      "learning_rate": 1.262463124843748e-05,
+      "loss": 2.0745,
+      "step": 8748
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6437609757353934,
+      "learning_rate": 1.261920396961297e-05,
+      "loss": 1.871,
+      "step": 8749
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5717221585252856,
+      "learning_rate": 1.2613777463772431e-05,
+      "loss": 1.8757,
+      "step": 8750
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5723605872703034,
+      "learning_rate": 1.2608351731254628e-05,
+      "loss": 2.1028,
+      "step": 8751
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5586435057115842,
+      "learning_rate": 1.2602926772398344e-05,
+      "loss": 1.8964,
+      "step": 8752
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5954265559492661,
+      "learning_rate": 1.2597502587542267e-05,
+      "loss": 1.8493,
+      "step": 8753
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5942395390469012,
+      "learning_rate": 1.2592079177025056e-05,
+      "loss": 1.9371,
+      "step": 8754
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6281826293343765,
+      "learning_rate": 1.2586656541185327e-05,
+      "loss": 2.0527,
+      "step": 8755
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6432272039943556,
+      "learning_rate": 1.2581234680361604e-05,
+      "loss": 1.9081,
+      "step": 8756
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6475550466941161,
+      "learning_rate": 1.2575813594892425e-05,
+      "loss": 1.9237,
+      "step": 8757
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5851996478684167,
+      "learning_rate": 1.2570393285116255e-05,
+      "loss": 1.8926,
+      "step": 8758
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6510044105894766,
+      "learning_rate": 1.2564973751371472e-05,
+      "loss": 2.0843,
+      "step": 8759
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7124363025477015,
+      "learning_rate": 1.2559554993996475e-05,
+      "loss": 1.9566,
+      "step": 8760
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5451627289675968,
+      "learning_rate": 1.2554137013329575e-05,
+      "loss": 1.8964,
+      "step": 8761
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5661923208633487,
+      "learning_rate": 1.2548719809709014e-05,
+      "loss": 1.8784,
+      "step": 8762
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6699562863447603,
+      "learning_rate": 1.2543303383473042e-05,
+      "loss": 2.1363,
+      "step": 8763
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.714096483959045,
+      "learning_rate": 1.2537887734959802e-05,
+      "loss": 1.9116,
+      "step": 8764
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5793116362609599,
+      "learning_rate": 1.2532472864507422e-05,
+      "loss": 1.9082,
+      "step": 8765
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6544742867909544,
+      "learning_rate": 1.2527058772453994e-05,
+      "loss": 1.9758,
+      "step": 8766
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7800509719635721,
+      "learning_rate": 1.252164545913751e-05,
+      "loss": 1.9075,
+      "step": 8767
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5501263086274368,
+      "learning_rate": 1.2516232924895981e-05,
+      "loss": 2.0744,
+      "step": 8768
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.644350824962762,
+      "learning_rate": 1.2510821170067305e-05,
+      "loss": 1.8681,
+      "step": 8769
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7992225186303318,
+      "learning_rate": 1.2505410194989359e-05,
+      "loss": 1.8871,
+      "step": 8770
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6096261619417377,
+      "learning_rate": 1.2500000000000006e-05,
+      "loss": 2.0819,
+      "step": 8771
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6788025443794176,
+      "learning_rate": 1.2494590585436994e-05,
+      "loss": 1.9351,
+      "step": 8772
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6759960931689781,
+      "learning_rate": 1.2489181951638054e-05,
+      "loss": 1.8647,
+      "step": 8773
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5961771150630039,
+      "learning_rate": 1.24837740989409e-05,
+      "loss": 1.8634,
+      "step": 8774
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.628692511645392,
+      "learning_rate": 1.2478367027683138e-05,
+      "loss": 2.0496,
+      "step": 8775
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7063952800830876,
+      "learning_rate": 1.2472960738202364e-05,
+      "loss": 1.8908,
+      "step": 8776
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7021218590966916,
+      "learning_rate": 1.2467555230836112e-05,
+      "loss": 1.8638,
+      "step": 8777
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5610603436504511,
+      "learning_rate": 1.2462150505921872e-05,
+      "loss": 1.9725,
+      "step": 8778
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6714642554802152,
+      "learning_rate": 1.2456746563797081e-05,
+      "loss": 1.8773,
+      "step": 8779
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6677076745605204,
+      "learning_rate": 1.2451343404799132e-05,
+      "loss": 2.073,
+      "step": 8780
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5861338936795244,
+      "learning_rate": 1.2445941029265357e-05,
+      "loss": 1.8537,
+      "step": 8781
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.619341397626377,
+      "learning_rate": 1.2440539437533075e-05,
+      "loss": 1.8861,
+      "step": 8782
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6662572628293926,
+      "learning_rate": 1.24351386299395e-05,
+      "loss": 2.0641,
+      "step": 8783
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.705633029341517,
+      "learning_rate": 1.2429738606821842e-05,
+      "loss": 1.8344,
+      "step": 8784
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6145125928571025,
+      "learning_rate": 1.2424339368517241e-05,
+      "loss": 1.9325,
+      "step": 8785
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5865299997432878,
+      "learning_rate": 1.2418940915362792e-05,
+      "loss": 1.8549,
+      "step": 8786
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7310860049384014,
+      "learning_rate": 1.241354324769555e-05,
+      "loss": 2.0772,
+      "step": 8787
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5932857037420229,
+      "learning_rate": 1.240814636585251e-05,
+      "loss": 1.8949,
+      "step": 8788
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6091244034045606,
+      "learning_rate": 1.2402750270170619e-05,
+      "loss": 1.8901,
+      "step": 8789
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7422317422346834,
+      "learning_rate": 1.2397354960986778e-05,
+      "loss": 1.8867,
+      "step": 8790
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5548355465916321,
+      "learning_rate": 1.2391960438637843e-05,
+      "loss": 1.9537,
+      "step": 8791
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6548998747751967,
+      "learning_rate": 1.2386566703460612e-05,
+      "loss": 2.0772,
+      "step": 8792
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6379166524627451,
+      "learning_rate": 1.238117375579184e-05,
+      "loss": 1.9479,
+      "step": 8793
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6107371196293422,
+      "learning_rate": 1.2375781595968229e-05,
+      "loss": 1.8995,
+      "step": 8794
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6392314146193777,
+      "learning_rate": 1.2370390224326434e-05,
+      "loss": 2.0415,
+      "step": 8795
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.583465814577676,
+      "learning_rate": 1.236499964120306e-05,
+      "loss": 1.9118,
+      "step": 8796
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6644074057218952,
+      "learning_rate": 1.2359609846934675e-05,
+      "loss": 1.967,
+      "step": 8797
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5946191306395001,
+      "learning_rate": 1.2354220841857755e-05,
+      "loss": 1.8817,
+      "step": 8798
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6077100075031053,
+      "learning_rate": 1.234883262630879e-05,
+      "loss": 1.9032,
+      "step": 8799
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6666430370992059,
+      "learning_rate": 1.2343445200624174e-05,
+      "loss": 2.0219,
+      "step": 8800
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5626193383892211,
+      "learning_rate": 1.233805856514027e-05,
+      "loss": 1.7948,
+      "step": 8801
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6549132647330811,
+      "learning_rate": 1.2332672720193395e-05,
+      "loss": 1.8796,
+      "step": 8802
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6518883757999413,
+      "learning_rate": 1.2327287666119784e-05,
+      "loss": 1.9805,
+      "step": 8803
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6077235601742319,
+      "learning_rate": 1.2321903403255672e-05,
+      "loss": 2.0926,
+      "step": 8804
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6216077700134023,
+      "learning_rate": 1.2316519931937223e-05,
+      "loss": 1.8612,
+      "step": 8805
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6743369199991857,
+      "learning_rate": 1.2311137252500521e-05,
+      "loss": 1.9098,
+      "step": 8806
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6188633363446723,
+      "learning_rate": 1.2305755365281658e-05,
+      "loss": 2.0393,
+      "step": 8807
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5770180604480277,
+      "learning_rate": 1.2300374270616646e-05,
+      "loss": 1.8961,
+      "step": 8808
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5520806635559293,
+      "learning_rate": 1.229499396884142e-05,
+      "loss": 1.9026,
+      "step": 8809
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6135668284606721,
+      "learning_rate": 1.2289614460291932e-05,
+      "loss": 1.8928,
+      "step": 8810
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6145064538993978,
+      "learning_rate": 1.228423574530402e-05,
+      "loss": 1.8568,
+      "step": 8811
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.585015988062962,
+      "learning_rate": 1.22788578242135e-05,
+      "loss": 2.0865,
+      "step": 8812
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5957807166786357,
+      "learning_rate": 1.2273480697356163e-05,
+      "loss": 1.836,
+      "step": 8813
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5567437321043897,
+      "learning_rate": 1.2268104365067702e-05,
+      "loss": 1.9116,
+      "step": 8814
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6368441550066674,
+      "learning_rate": 1.2262728827683776e-05,
+      "loss": 2.1079,
+      "step": 8815
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5674927031967735,
+      "learning_rate": 1.2257354085540034e-05,
+      "loss": 1.9611,
+      "step": 8816
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5725891608370609,
+      "learning_rate": 1.2251980138972009e-05,
+      "loss": 1.8493,
+      "step": 8817
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6423943796850572,
+      "learning_rate": 1.2246606988315251e-05,
+      "loss": 1.8758,
+      "step": 8818
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6585748741224486,
+      "learning_rate": 1.2241234633905202e-05,
+      "loss": 1.8859,
+      "step": 8819
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6650097545311923,
+      "learning_rate": 1.2235863076077279e-05,
+      "loss": 2.0411,
+      "step": 8820
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.593182295754304,
+      "learning_rate": 1.2230492315166881e-05,
+      "loss": 1.8899,
+      "step": 8821
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6889164621736298,
+      "learning_rate": 1.2225122351509294e-05,
+      "loss": 1.9523,
+      "step": 8822
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5929854939059906,
+      "learning_rate": 1.2219753185439801e-05,
+      "loss": 1.8882,
+      "step": 8823
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6254372045136838,
+      "learning_rate": 1.2214384817293618e-05,
+      "loss": 2.0553,
+      "step": 8824
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6058496520986417,
+      "learning_rate": 1.2209017247405916e-05,
+      "loss": 1.8983,
+      "step": 8825
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6168582899922456,
+      "learning_rate": 1.2203650476111814e-05,
+      "loss": 1.9247,
+      "step": 8826
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6046463546199636,
+      "learning_rate": 1.2198284503746382e-05,
+      "loss": 2.087,
+      "step": 8827
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6461352854307241,
+      "learning_rate": 1.2192919330644639e-05,
+      "loss": 1.925,
+      "step": 8828
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6035699882418979,
+      "learning_rate": 1.2187554957141556e-05,
+      "loss": 1.9059,
+      "step": 8829
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5421313133058074,
+      "learning_rate": 1.218219138357205e-05,
+      "loss": 1.8411,
+      "step": 8830
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5412003136050683,
+      "learning_rate": 1.2176828610270994e-05,
+      "loss": 1.9012,
+      "step": 8831
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5596224861619358,
+      "learning_rate": 1.2171466637573206e-05,
+      "loss": 2.0319,
+      "step": 8832
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6028913675159295,
+      "learning_rate": 1.2166105465813458e-05,
+      "loss": 1.9113,
+      "step": 8833
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5302855995389067,
+      "learning_rate": 1.2160745095326467e-05,
+      "loss": 1.9256,
+      "step": 8834
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5375513492834442,
+      "learning_rate": 1.2155385526446906e-05,
+      "loss": 1.8875,
+      "step": 8835
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5805135815755255,
+      "learning_rate": 1.2150026759509392e-05,
+      "loss": 2.1003,
+      "step": 8836
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6524744472052294,
+      "learning_rate": 1.21446687948485e-05,
+      "loss": 1.9039,
+      "step": 8837
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6662964546437532,
+      "learning_rate": 1.2139311632798741e-05,
+      "loss": 1.8916,
+      "step": 8838
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5951191752042527,
+      "learning_rate": 1.2133955273694595e-05,
+      "loss": 2.0872,
+      "step": 8839
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5958990018643873,
+      "learning_rate": 1.2128599717870474e-05,
+      "loss": 1.9315,
+      "step": 8840
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6241624007750445,
+      "learning_rate": 1.2123244965660748e-05,
+      "loss": 1.8489,
+      "step": 8841
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6068461162751405,
+      "learning_rate": 1.211789101739974e-05,
+      "loss": 1.9049,
+      "step": 8842
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5675627637799707,
+      "learning_rate": 1.2112537873421714e-05,
+      "loss": 1.8964,
+      "step": 8843
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5508374290162419,
+      "learning_rate": 1.2107185534060906e-05,
+      "loss": 2.0937,
+      "step": 8844
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5976408528316398,
+      "learning_rate": 1.2101833999651446e-05,
+      "loss": 1.8768,
+      "step": 8845
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5834468549355947,
+      "learning_rate": 1.2096483270527489e-05,
+      "loss": 1.8995,
+      "step": 8846
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5437978974703385,
+      "learning_rate": 1.2091133347023099e-05,
+      "loss": 1.9792,
+      "step": 8847
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5937528487611289,
+      "learning_rate": 1.2085784229472266e-05,
+      "loss": 2.0644,
+      "step": 8848
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5584538869829736,
+      "learning_rate": 1.2080435918208997e-05,
+      "loss": 1.9118,
+      "step": 8849
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5983310220381675,
+      "learning_rate": 1.2075088413567168e-05,
+      "loss": 1.8529,
+      "step": 8850
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5942272224507877,
+      "learning_rate": 1.2069741715880676e-05,
+      "loss": 1.8811,
+      "step": 8851
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6019786597771181,
+      "learning_rate": 1.2064395825483335e-05,
+      "loss": 2.0718,
+      "step": 8852
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6501016498188937,
+      "learning_rate": 1.2059050742708888e-05,
+      "loss": 1.999,
+      "step": 8853
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5902441354163089,
+      "learning_rate": 1.2053706467891074e-05,
+      "loss": 1.8607,
+      "step": 8854
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5682853050172568,
+      "learning_rate": 1.2048363001363557e-05,
+      "loss": 1.8497,
+      "step": 8855
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6331533186245029,
+      "learning_rate": 1.204302034345993e-05,
+      "loss": 2.0867,
+      "step": 8856
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5568823901478561,
+      "learning_rate": 1.2037678494513787e-05,
+      "loss": 1.9061,
+      "step": 8857
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5942284791509133,
+      "learning_rate": 1.203233745485862e-05,
+      "loss": 1.8537,
+      "step": 8858
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5681410685758121,
+      "learning_rate": 1.2026997224827884e-05,
+      "loss": 1.9808,
+      "step": 8859
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5806763038192028,
+      "learning_rate": 1.2021657804755027e-05,
+      "loss": 2.0807,
+      "step": 8860
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5715662261749004,
+      "learning_rate": 1.201631919497338e-05,
+      "loss": 1.8743,
+      "step": 8861
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6674112982937261,
+      "learning_rate": 1.2010981395816257e-05,
+      "loss": 1.8547,
+      "step": 8862
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.7059145218294421,
+      "learning_rate": 1.2005644407616943e-05,
+      "loss": 1.9202,
+      "step": 8863
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6036741212008151,
+      "learning_rate": 1.200030823070861e-05,
+      "loss": 2.1309,
+      "step": 8864
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6185832540199722,
+      "learning_rate": 1.1994972865424458e-05,
+      "loss": 1.9757,
+      "step": 8865
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6150503598869486,
+      "learning_rate": 1.1989638312097567e-05,
+      "loss": 1.8713,
+      "step": 8866
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6263990557912182,
+      "learning_rate": 1.1984304571060994e-05,
+      "loss": 1.882,
+      "step": 8867
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6550664095643863,
+      "learning_rate": 1.1978971642647777e-05,
+      "loss": 2.1173,
+      "step": 8868
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.663525514048317,
+      "learning_rate": 1.1973639527190842e-05,
+      "loss": 1.8936,
+      "step": 8869
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.609334133523607,
+      "learning_rate": 1.1968308225023104e-05,
+      "loss": 1.9084,
+      "step": 8870
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6579913967017702,
+      "learning_rate": 1.1962977736477418e-05,
+      "loss": 2.0755,
+      "step": 8871
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5412014913664036,
+      "learning_rate": 1.1957648061886592e-05,
+      "loss": 1.9265,
+      "step": 8872
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6529214344120083,
+      "learning_rate": 1.1952319201583375e-05,
+      "loss": 1.8839,
+      "step": 8873
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.690778401363497,
+      "learning_rate": 1.1946991155900473e-05,
+      "loss": 1.908,
+      "step": 8874
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5600634155290309,
+      "learning_rate": 1.1941663925170536e-05,
+      "loss": 1.876,
+      "step": 8875
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.574977154053681,
+      "learning_rate": 1.1936337509726165e-05,
+      "loss": 2.1155,
+      "step": 8876
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6845992120299016,
+      "learning_rate": 1.1931011909899909e-05,
+      "loss": 1.8366,
+      "step": 8877
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6181392346632707,
+      "learning_rate": 1.1925687126024268e-05,
+      "loss": 1.9286,
+      "step": 8878
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5679508684837916,
+      "learning_rate": 1.1920363158431691e-05,
+      "loss": 1.806,
+      "step": 8879
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5953247742092924,
+      "learning_rate": 1.1915040007454576e-05,
+      "loss": 2.0837,
+      "step": 8880
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6235760091207185,
+      "learning_rate": 1.1909717673425266e-05,
+      "loss": 1.8956,
+      "step": 8881
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5604918472928406,
+      "learning_rate": 1.1904396156676058e-05,
+      "loss": 1.8441,
+      "step": 8882
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6777452150674497,
+      "learning_rate": 1.1899075457539196e-05,
+      "loss": 1.8254,
+      "step": 8883
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6023650247270822,
+      "learning_rate": 1.1893755576346874e-05,
+      "loss": 2.1022,
+      "step": 8884
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6027617035413784,
+      "learning_rate": 1.1888436513431233e-05,
+      "loss": 1.8963,
+      "step": 8885
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6539234529159501,
+      "learning_rate": 1.1883118269124364e-05,
+      "loss": 1.8679,
+      "step": 8886
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6230029363436617,
+      "learning_rate": 1.1877800843758308e-05,
+      "loss": 1.8864,
+      "step": 8887
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6428341291666918,
+      "learning_rate": 1.187248423766505e-05,
+      "loss": 2.0338,
+      "step": 8888
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.613766123947066,
+      "learning_rate": 1.186716845117653e-05,
+      "loss": 1.8405,
+      "step": 8889
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5532933969053196,
+      "learning_rate": 1.1861853484624638e-05,
+      "loss": 1.9795,
+      "step": 8890
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6169487386176906,
+      "learning_rate": 1.1856539338341213e-05,
+      "loss": 1.868,
+      "step": 8891
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5960371645660122,
+      "learning_rate": 1.1851226012658015e-05,
+      "loss": 2.087,
+      "step": 8892
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5721951551728145,
+      "learning_rate": 1.18459135079068e-05,
+      "loss": 1.8576,
+      "step": 8893
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6220000348786602,
+      "learning_rate": 1.1840601824419254e-05,
+      "loss": 1.9147,
+      "step": 8894
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5900780264400767,
+      "learning_rate": 1.1835290962526977e-05,
+      "loss": 1.8963,
+      "step": 8895
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5439461672235679,
+      "learning_rate": 1.1829980922561578e-05,
+      "loss": 2.0645,
+      "step": 8896
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6130964610757279,
+      "learning_rate": 1.1824671704854574e-05,
+      "loss": 1.8606,
+      "step": 8897
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6065454504427,
+      "learning_rate": 1.181936330973744e-05,
+      "loss": 1.8275,
+      "step": 8898
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6110003175225011,
+      "learning_rate": 1.181405573754161e-05,
+      "loss": 1.9116,
+      "step": 8899
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6051730352268181,
+      "learning_rate": 1.1808748988598432e-05,
+      "loss": 2.1061,
+      "step": 8900
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5542959880384529,
+      "learning_rate": 1.1803443063239253e-05,
+      "loss": 1.9033,
+      "step": 8901
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5991864687821332,
+      "learning_rate": 1.179813796179535e-05,
+      "loss": 1.8917,
+      "step": 8902
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5835249598383039,
+      "learning_rate": 1.1792833684597907e-05,
+      "loss": 1.9983,
+      "step": 8903
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.618029452236078,
+      "learning_rate": 1.1787530231978134e-05,
+      "loss": 2.0739,
+      "step": 8904
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6509764650270039,
+      "learning_rate": 1.1782227604267115e-05,
+      "loss": 1.8991,
+      "step": 8905
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5639464677399753,
+      "learning_rate": 1.1776925801795918e-05,
+      "loss": 1.8663,
+      "step": 8906
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6631683405041704,
+      "learning_rate": 1.1771624824895584e-05,
+      "loss": 1.8441,
+      "step": 8907
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7244930170314408,
+      "learning_rate": 1.1766324673897048e-05,
+      "loss": 2.044,
+      "step": 8908
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.536179444579546,
+      "learning_rate": 1.1761025349131216e-05,
+      "loss": 1.9369,
+      "step": 8909
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.698844605755115,
+      "learning_rate": 1.1755726850928978e-05,
+      "loss": 1.8842,
+      "step": 8910
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.586336241273513,
+      "learning_rate": 1.1750429179621113e-05,
+      "loss": 1.9146,
+      "step": 8911
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6575251979932237,
+      "learning_rate": 1.1745132335538383e-05,
+      "loss": 2.1108,
+      "step": 8912
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7304544296409403,
+      "learning_rate": 1.1739836319011496e-05,
+      "loss": 1.9049,
+      "step": 8913
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5706283641550365,
+      "learning_rate": 1.1734541130371091e-05,
+      "loss": 1.8179,
+      "step": 8914
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7423358573463116,
+      "learning_rate": 1.17292467699478e-05,
+      "loss": 1.9911,
+      "step": 8915
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6387765270577971,
+      "learning_rate": 1.1723953238072138e-05,
+      "loss": 2.0679,
+      "step": 8916
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6428468345137285,
+      "learning_rate": 1.1718660535074617e-05,
+      "loss": 1.8761,
+      "step": 8917
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6521253366768809,
+      "learning_rate": 1.1713368661285679e-05,
+      "loss": 1.8638,
+      "step": 8918
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.552787582184753,
+      "learning_rate": 1.170807761703572e-05,
+      "loss": 1.9127,
+      "step": 8919
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6269348775012485,
+      "learning_rate": 1.1702787402655077e-05,
+      "loss": 2.0874,
+      "step": 8920
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6747665622647425,
+      "learning_rate": 1.1697498018474045e-05,
+      "loss": 1.987,
+      "step": 8921
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6037883675324819,
+      "learning_rate": 1.169220946482286e-05,
+      "loss": 1.8746,
+      "step": 8922
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7517759081716553,
+      "learning_rate": 1.1686921742031706e-05,
+      "loss": 1.9458,
+      "step": 8923
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.689681542275298,
+      "learning_rate": 1.168163485043072e-05,
+      "loss": 2.0796,
+      "step": 8924
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6042231150947777,
+      "learning_rate": 1.1676348790349984e-05,
+      "loss": 1.8489,
+      "step": 8925
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7444022091593935,
+      "learning_rate": 1.1671063562119527e-05,
+      "loss": 1.9177,
+      "step": 8926
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5372101566769004,
+      "learning_rate": 1.1665779166069329e-05,
+      "loss": 1.9518,
+      "step": 8927
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7321614206163566,
+      "learning_rate": 1.1660495602529317e-05,
+      "loss": 2.0587,
+      "step": 8928
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5558316182831381,
+      "learning_rate": 1.1655212871829363e-05,
+      "loss": 1.8393,
+      "step": 8929
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5861100510809774,
+      "learning_rate": 1.1649930974299292e-05,
+      "loss": 1.8132,
+      "step": 8930
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6681001288460169,
+      "learning_rate": 1.1644649910268874e-05,
+      "loss": 1.8697,
+      "step": 8931
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.713585293462639,
+      "learning_rate": 1.1639369680067826e-05,
+      "loss": 2.0735,
+      "step": 8932
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6395908342134343,
+      "learning_rate": 1.1634090284025818e-05,
+      "loss": 1.8464,
+      "step": 8933
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6980486669324184,
+      "learning_rate": 1.1628811722472461e-05,
+      "loss": 1.9269,
+      "step": 8934
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6742857271931811,
+      "learning_rate": 1.1623533995737318e-05,
+      "loss": 1.8838,
+      "step": 8935
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6776423470723552,
+      "learning_rate": 1.1618257104149898e-05,
+      "loss": 2.0972,
+      "step": 8936
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6642118329494529,
+      "learning_rate": 1.1612981048039662e-05,
+      "loss": 1.8567,
+      "step": 8937
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5561285820399503,
+      "learning_rate": 1.1607705827736023e-05,
+      "loss": 1.8689,
+      "step": 8938
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7644962606635218,
+      "learning_rate": 1.1602431443568309e-05,
+      "loss": 1.9054,
+      "step": 8939
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5814468708664705,
+      "learning_rate": 1.1597157895865849e-05,
+      "loss": 2.1014,
+      "step": 8940
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6463987066225427,
+      "learning_rate": 1.1591885184957888e-05,
+      "loss": 1.8453,
+      "step": 8941
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7092471512929696,
+      "learning_rate": 1.15866133111736e-05,
+      "loss": 1.8881,
+      "step": 8942
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6306521969659294,
+      "learning_rate": 1.1581342274842156e-05,
+      "loss": 1.9119,
+      "step": 8943
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8056851128874467,
+      "learning_rate": 1.1576072076292646e-05,
+      "loss": 2.0935,
+      "step": 8944
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5987624892222396,
+      "learning_rate": 1.1570802715854088e-05,
+      "loss": 1.8671,
+      "step": 8945
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7241315503274826,
+      "learning_rate": 1.1565534193855502e-05,
+      "loss": 1.9342,
+      "step": 8946
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7160714612811266,
+      "learning_rate": 1.1560266510625786e-05,
+      "loss": 1.9409,
+      "step": 8947
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5935638016157081,
+      "learning_rate": 1.1554999666493857e-05,
+      "loss": 2.0735,
+      "step": 8948
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.7389213777941352,
+      "learning_rate": 1.154973366178854e-05,
+      "loss": 1.8752,
+      "step": 8949
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6656055625503415,
+      "learning_rate": 1.1544468496838585e-05,
+      "loss": 1.8694,
+      "step": 8950
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6598248342308699,
+      "learning_rate": 1.1539204171972756e-05,
+      "loss": 1.8352,
+      "step": 8951
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6899235881956752,
+      "learning_rate": 1.1533940687519704e-05,
+      "loss": 1.9687,
+      "step": 8952
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6734762562583627,
+      "learning_rate": 1.1528678043808041e-05,
+      "loss": 2.0654,
+      "step": 8953
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.588025099494492,
+      "learning_rate": 1.1523416241166371e-05,
+      "loss": 1.8717,
+      "step": 8954
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6447800920734613,
+      "learning_rate": 1.1518155279923179e-05,
+      "loss": 1.904,
+      "step": 8955
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5886658190258836,
+      "learning_rate": 1.1512895160406928e-05,
+      "loss": 2.1072,
+      "step": 8956
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6128230081493294,
+      "learning_rate": 1.1507635882946055e-05,
+      "loss": 1.8783,
+      "step": 8957
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5646989950639825,
+      "learning_rate": 1.1502377447868895e-05,
+      "loss": 1.977,
+      "step": 8958
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6610830664979006,
+      "learning_rate": 1.149711985550376e-05,
+      "loss": 1.8535,
+      "step": 8959
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6162754333595482,
+      "learning_rate": 1.1491863106178905e-05,
+      "loss": 2.0493,
+      "step": 8960
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5623413457056937,
+      "learning_rate": 1.1486607200222516e-05,
+      "loss": 1.904,
+      "step": 8961
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.630117947374411,
+      "learning_rate": 1.1481352137962775e-05,
+      "loss": 1.8877,
+      "step": 8962
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6497054129635882,
+      "learning_rate": 1.1476097919727746e-05,
+      "loss": 1.9027,
+      "step": 8963
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5903965738586977,
+      "learning_rate": 1.1470844545845483e-05,
+      "loss": 2.0835,
+      "step": 8964
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6425310664912935,
+      "learning_rate": 1.1465592016643973e-05,
+      "loss": 1.987,
+      "step": 8965
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6062504648587747,
+      "learning_rate": 1.1460340332451152e-05,
+      "loss": 1.8975,
+      "step": 8966
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6036239617901162,
+      "learning_rate": 1.145508949359491e-05,
+      "loss": 1.8289,
+      "step": 8967
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6590383456361064,
+      "learning_rate": 1.1449839500403072e-05,
+      "loss": 2.0469,
+      "step": 8968
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5849750715025981,
+      "learning_rate": 1.1444590353203421e-05,
+      "loss": 1.8668,
+      "step": 8969
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5630310422033301,
+      "learning_rate": 1.1439342052323682e-05,
+      "loss": 1.8713,
+      "step": 8970
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6500602837798692,
+      "learning_rate": 1.1434094598091527e-05,
+      "loss": 1.9178,
+      "step": 8971
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6798654638692546,
+      "learning_rate": 1.1428847990834576e-05,
+      "loss": 2.0882,
+      "step": 8972
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5567403147370228,
+      "learning_rate": 1.1423602230880396e-05,
+      "loss": 1.8623,
+      "step": 8973
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.638065412208186,
+      "learning_rate": 1.1418357318556503e-05,
+      "loss": 1.8666,
+      "step": 8974
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6264968581154093,
+      "learning_rate": 1.1413113254190355e-05,
+      "loss": 1.8236,
+      "step": 8975
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6431754052019392,
+      "learning_rate": 1.1407870038109367e-05,
+      "loss": 2.0461,
+      "step": 8976
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6445769863296433,
+      "learning_rate": 1.1402627670640886e-05,
+      "loss": 1.9494,
+      "step": 8977
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5650827138130116,
+      "learning_rate": 1.1397386152112222e-05,
+      "loss": 1.8509,
+      "step": 8978
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6146498328787695,
+      "learning_rate": 1.1392145482850619e-05,
+      "loss": 1.8565,
+      "step": 8979
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.581297197723896,
+      "learning_rate": 1.1386905663183279e-05,
+      "loss": 2.0822,
+      "step": 8980
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5799198228726854,
+      "learning_rate": 1.138166669343734e-05,
+      "loss": 1.8947,
+      "step": 8981
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6046031559924472,
+      "learning_rate": 1.1376428573939896e-05,
+      "loss": 1.8561,
+      "step": 8982
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5858783894629405,
+      "learning_rate": 1.1371191305017981e-05,
+      "loss": 1.9879,
+      "step": 8983
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6035202749223898,
+      "learning_rate": 1.1365954886998584e-05,
+      "loss": 1.9014,
+      "step": 8984
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6150137928215254,
+      "learning_rate": 1.1360719320208644e-05,
+      "loss": 2.0604,
+      "step": 8985
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6308692671391954,
+      "learning_rate": 1.1355484604975007e-05,
+      "loss": 1.8866,
+      "step": 8986
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5441074458398664,
+      "learning_rate": 1.1350250741624527e-05,
+      "loss": 1.8655,
+      "step": 8987
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5849942020288769,
+      "learning_rate": 1.134501773048398e-05,
+      "loss": 2.0749,
+      "step": 8988
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5441499006926636,
+      "learning_rate": 1.133978557188005e-05,
+      "loss": 1.9374,
+      "step": 8989
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5771236807412552,
+      "learning_rate": 1.1334554266139438e-05,
+      "loss": 1.8723,
+      "step": 8990
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5852552216350425,
+      "learning_rate": 1.1329323813588752e-05,
+      "loss": 1.8802,
+      "step": 8991
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5829212308814022,
+      "learning_rate": 1.132409421455452e-05,
+      "loss": 2.09,
+      "step": 8992
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5872424583607305,
+      "learning_rate": 1.1318865469363288e-05,
+      "loss": 1.9085,
+      "step": 8993
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6049008394115704,
+      "learning_rate": 1.1313637578341468e-05,
+      "loss": 1.9283,
+      "step": 8994
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5677177169557862,
+      "learning_rate": 1.130841054181549e-05,
+      "loss": 1.863,
+      "step": 8995
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5592029557494277,
+      "learning_rate": 1.1303184360111695e-05,
+      "loss": 1.9552,
+      "step": 8996
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6133598281410187,
+      "learning_rate": 1.1297959033556352e-05,
+      "loss": 2.0903,
+      "step": 8997
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6049364312685719,
+      "learning_rate": 1.1292734562475731e-05,
+      "loss": 1.8911,
+      "step": 8998
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5942166249025516,
+      "learning_rate": 1.1287510947195998e-05,
+      "loss": 1.8813,
+      "step": 8999
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5876166467493542,
+      "learning_rate": 1.1282288188043277e-05,
+      "loss": 2.0959,
+      "step": 9000
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5487513118352851,
+      "learning_rate": 1.1277066285343676e-05,
+      "loss": 1.8581,
+      "step": 9001
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5524661667346662,
+      "learning_rate": 1.1271845239423196e-05,
+      "loss": 1.8847,
+      "step": 9002
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5491182847563656,
+      "learning_rate": 1.1266625050607801e-05,
+      "loss": 1.9084,
+      "step": 9003
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5738646314008986,
+      "learning_rate": 1.1261405719223445e-05,
+      "loss": 2.0226,
+      "step": 9004
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5421923009558371,
+      "learning_rate": 1.1256187245595956e-05,
+      "loss": 1.8445,
+      "step": 9005
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5859893055896866,
+      "learning_rate": 1.1250969630051161e-05,
+      "loss": 1.902,
+      "step": 9006
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6179165487566488,
+      "learning_rate": 1.1245752872914813e-05,
+      "loss": 1.8717,
+      "step": 9007
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5920237269196194,
+      "learning_rate": 1.1240536974512616e-05,
+      "loss": 1.9672,
+      "step": 9008
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5996679037578515,
+      "learning_rate": 1.1235321935170223e-05,
+      "loss": 2.0506,
+      "step": 9009
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6058979153581787,
+      "learning_rate": 1.123010775521323e-05,
+      "loss": 1.8233,
+      "step": 9010
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5778100973649726,
+      "learning_rate": 1.1224894434967165e-05,
+      "loss": 1.8551,
+      "step": 9011
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.613146258795698,
+      "learning_rate": 1.121968197475755e-05,
+      "loss": 2.0915,
+      "step": 9012
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5547189804107404,
+      "learning_rate": 1.1214470374909792e-05,
+      "loss": 1.88,
+      "step": 9013
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5669628106820677,
+      "learning_rate": 1.120925963574928e-05,
+      "loss": 1.912,
+      "step": 9014
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6637462011967322,
+      "learning_rate": 1.1204049757601342e-05,
+      "loss": 1.8702,
+      "step": 9015
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6329998871514068,
+      "learning_rate": 1.1198840740791253e-05,
+      "loss": 1.885,
+      "step": 9016
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6415384885773406,
+      "learning_rate": 1.1193632585644234e-05,
+      "loss": 2.0906,
+      "step": 9017
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5885668861348707,
+      "learning_rate": 1.1188425292485452e-05,
+      "loss": 1.8545,
+      "step": 9018
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6120703274370523,
+      "learning_rate": 1.1183218861640016e-05,
+      "loss": 1.8903,
+      "step": 9019
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6447567062713911,
+      "learning_rate": 1.1178013293432987e-05,
+      "loss": 2.0459,
+      "step": 9020
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6119559065528996,
+      "learning_rate": 1.1172808588189371e-05,
+      "loss": 2.0141,
+      "step": 9021
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5624525125415928,
+      "learning_rate": 1.116760474623412e-05,
+      "loss": 1.9349,
+      "step": 9022
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6079882131605245,
+      "learning_rate": 1.1162401767892128e-05,
+      "loss": 1.8844,
+      "step": 9023
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6587785908513255,
+      "learning_rate": 1.115719965348824e-05,
+      "loss": 2.0327,
+      "step": 9024
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6628911569822925,
+      "learning_rate": 1.1151998403347244e-05,
+      "loss": 1.878,
+      "step": 9025
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6310908490151304,
+      "learning_rate": 1.1146798017793878e-05,
+      "loss": 1.8653,
+      "step": 9026
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6924362140089038,
+      "learning_rate": 1.1141598497152822e-05,
+      "loss": 1.9822,
+      "step": 9027
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5751115490065275,
+      "learning_rate": 1.1136399841748705e-05,
+      "loss": 1.9184,
+      "step": 9028
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.709002621491511,
+      "learning_rate": 1.1131202051906098e-05,
+      "loss": 2.0618,
+      "step": 9029
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6492433261792121,
+      "learning_rate": 1.1126005127949521e-05,
+      "loss": 1.8877,
+      "step": 9030
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5857699762376762,
+      "learning_rate": 1.1120809070203439e-05,
+      "loss": 1.7941,
+      "step": 9031
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6875346506182057,
+      "learning_rate": 1.1115613878992273e-05,
+      "loss": 2.1179,
+      "step": 9032
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5931447089468332,
+      "learning_rate": 1.1110419554640353e-05,
+      "loss": 1.9593,
+      "step": 9033
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5659188177705613,
+      "learning_rate": 1.1105226097472011e-05,
+      "loss": 1.841,
+      "step": 9034
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6412129882068893,
+      "learning_rate": 1.110003350781149e-05,
+      "loss": 1.8918,
+      "step": 9035
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6367599403419106,
+      "learning_rate": 1.1094841785982964e-05,
+      "loss": 1.8299,
+      "step": 9036
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7019631410934414,
+      "learning_rate": 1.10896509323106e-05,
+      "loss": 2.0714,
+      "step": 9037
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5648713714214756,
+      "learning_rate": 1.108446094711848e-05,
+      "loss": 1.8836,
+      "step": 9038
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6315288337701673,
+      "learning_rate": 1.1079271830730612e-05,
+      "loss": 1.8922,
+      "step": 9039
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7319047211924279,
+      "learning_rate": 1.1074083583471009e-05,
+      "loss": 1.8929,
+      "step": 9040
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.786482489658743,
+      "learning_rate": 1.106889620566357e-05,
+      "loss": 2.0979,
+      "step": 9041
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6220242033234383,
+      "learning_rate": 1.106370969763216e-05,
+      "loss": 1.9183,
+      "step": 9042
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7118737215799696,
+      "learning_rate": 1.1058524059700625e-05,
+      "loss": 1.8291,
+      "step": 9043
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7103684660373127,
+      "learning_rate": 1.1053339292192685e-05,
+      "loss": 2.0744,
+      "step": 9044
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5451329559212765,
+      "learning_rate": 1.104815539543208e-05,
+      "loss": 1.9221,
+      "step": 9045
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7052156067961092,
+      "learning_rate": 1.1042972369742457e-05,
+      "loss": 1.8948,
+      "step": 9046
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7082693283465217,
+      "learning_rate": 1.1037790215447388e-05,
+      "loss": 1.8731,
+      "step": 9047
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6234874898157641,
+      "learning_rate": 1.103260893287045e-05,
+      "loss": 1.8789,
+      "step": 9048
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6569024854737763,
+      "learning_rate": 1.1027428522335109e-05,
+      "loss": 2.0847,
+      "step": 9049
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7109161951346876,
+      "learning_rate": 1.1022248984164795e-05,
+      "loss": 1.8649,
+      "step": 9050
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6294946277085567,
+      "learning_rate": 1.101707031868292e-05,
+      "loss": 1.8409,
+      "step": 9051
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5459657403953199,
+      "learning_rate": 1.1011892526212778e-05,
+      "loss": 1.9045,
+      "step": 9052
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7548898115060142,
+      "learning_rate": 1.100671560707765e-05,
+      "loss": 2.0695,
+      "step": 9053
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6757110874202107,
+      "learning_rate": 1.1001539561600752e-05,
+      "loss": 1.8898,
+      "step": 9054
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6284426952565466,
+      "learning_rate": 1.0996364390105249e-05,
+      "loss": 1.8846,
+      "step": 9055
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7151443084353183,
+      "learning_rate": 1.099119009291424e-05,
+      "loss": 2.07,
+      "step": 9056
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6037357861414693,
+      "learning_rate": 1.0986016670350789e-05,
+      "loss": 1.8184,
+      "step": 9057
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5663437382965097,
+      "learning_rate": 1.0980844122737876e-05,
+      "loss": 1.9734,
+      "step": 9058
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.636895098695771,
+      "learning_rate": 1.0975672450398474e-05,
+      "loss": 1.8723,
+      "step": 9059
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.585696111484951,
+      "learning_rate": 1.0970501653655446e-05,
+      "loss": 1.8623,
+      "step": 9060
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6018070926836733,
+      "learning_rate": 1.0965331732831635e-05,
+      "loss": 2.0335,
+      "step": 9061
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.589457054084724,
+      "learning_rate": 1.096016268824982e-05,
+      "loss": 1.879,
+      "step": 9062
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5881471392466731,
+      "learning_rate": 1.0954994520232725e-05,
+      "loss": 1.9167,
+      "step": 9063
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5892091476514738,
+      "learning_rate": 1.094982722910302e-05,
+      "loss": 1.9291,
+      "step": 9064
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5824712548333607,
+      "learning_rate": 1.0944660815183322e-05,
+      "loss": 2.066,
+      "step": 9065
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6014375458918635,
+      "learning_rate": 1.0939495278796189e-05,
+      "loss": 1.881,
+      "step": 9066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.659247263436494,
+      "learning_rate": 1.093433062026413e-05,
+      "loss": 1.8505,
+      "step": 9067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5729738108592379,
+      "learning_rate": 1.092916683990959e-05,
+      "loss": 1.8158,
+      "step": 9068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6485817235210111,
+      "learning_rate": 1.092400393805497e-05,
+      "loss": 2.1079,
+      "step": 9069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5696438041200594,
+      "learning_rate": 1.091884191502261e-05,
+      "loss": 1.9341,
+      "step": 9070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5811623664936997,
+      "learning_rate": 1.0913680771134796e-05,
+      "loss": 1.9206,
+      "step": 9071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6172993000228082,
+      "learning_rate": 1.090852050671376e-05,
+      "loss": 1.8677,
+      "step": 9072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6330724661300089,
+      "learning_rate": 1.090336112208168e-05,
+      "loss": 2.0588,
+      "step": 9073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5584832196605616,
+      "learning_rate": 1.0898202617560679e-05,
+      "loss": 1.8711,
+      "step": 9074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6649929924127639,
+      "learning_rate": 1.0893044993472807e-05,
+      "loss": 1.8367,
+      "step": 9075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5651047852612145,
+      "learning_rate": 1.0887888250140097e-05,
+      "loss": 1.9192,
+      "step": 9076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5994202001465511,
+      "learning_rate": 1.0882732387884495e-05,
+      "loss": 2.0713,
+      "step": 9077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5774354138388746,
+      "learning_rate": 1.0877577407027908e-05,
+      "loss": 1.8924,
+      "step": 9078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5672691327615841,
+      "learning_rate": 1.087242330789218e-05,
+      "loss": 1.8676,
+      "step": 9079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5945602608888274,
+      "learning_rate": 1.0867270090799101e-05,
+      "loss": 1.867,
+      "step": 9080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.56695498511441,
+      "learning_rate": 1.0862117756070408e-05,
+      "loss": 2.0643,
+      "step": 9081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6187372765823543,
+      "learning_rate": 1.0856966304027794e-05,
+      "loss": 1.8731,
+      "step": 9082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5944595247670489,
+      "learning_rate": 1.0851815734992856e-05,
+      "loss": 1.9807,
+      "step": 9083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5949849583666381,
+      "learning_rate": 1.0846666049287193e-05,
+      "loss": 1.8762,
+      "step": 9084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.553331177626958,
+      "learning_rate": 1.0841517247232322e-05,
+      "loss": 2.0812,
+      "step": 9085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5816110901336285,
+      "learning_rate": 1.0836369329149673e-05,
+      "loss": 1.883,
+      "step": 9086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6027924349682429,
+      "learning_rate": 1.0831222295360693e-05,
+      "loss": 1.8307,
+      "step": 9087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5891867453439753,
+      "learning_rate": 1.0826076146186703e-05,
+      "loss": 2.0519,
+      "step": 9088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5764090670972408,
+      "learning_rate": 1.0820930881948996e-05,
+      "loss": 1.9517,
+      "step": 9089
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.639198495741725,
+      "learning_rate": 1.0815786502968844e-05,
+      "loss": 1.8725,
+      "step": 9090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6119584558755411,
+      "learning_rate": 1.081064300956739e-05,
+      "loss": 1.8848,
+      "step": 9091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6049137738412469,
+      "learning_rate": 1.0805500402065793e-05,
+      "loss": 1.8525,
+      "step": 9092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.594575451201536,
+      "learning_rate": 1.0800358680785131e-05,
+      "loss": 2.064,
+      "step": 9093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7778320771048078,
+      "learning_rate": 1.079521784604639e-05,
+      "loss": 1.8795,
+      "step": 9094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5881063539509765,
+      "learning_rate": 1.0790077898170573e-05,
+      "loss": 1.9414,
+      "step": 9095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6233175357656116,
+      "learning_rate": 1.0784938837478556e-05,
+      "loss": 1.8184,
+      "step": 9096
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6708820210209048,
+      "learning_rate": 1.0779800664291198e-05,
+      "loss": 2.0366,
+      "step": 9097
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6427088609917424,
+      "learning_rate": 1.077466337892932e-05,
+      "loss": 1.888,
+      "step": 9098
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6058515373365156,
+      "learning_rate": 1.0769526981713637e-05,
+      "loss": 1.8744,
+      "step": 9099
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6066250554972017,
+      "learning_rate": 1.0764391472964846e-05,
+      "loss": 1.8707,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.541145392620769,
+      "learning_rate": 1.0759256853003578e-05,
+      "loss": 2.1121,
+      "step": 9101
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6215361040592173,
+      "learning_rate": 1.0754123122150403e-05,
+      "loss": 1.9273,
+      "step": 9102
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6156539468634563,
+      "learning_rate": 1.0748990280725846e-05,
+      "loss": 1.8976,
+      "step": 9103
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.597990725334442,
+      "learning_rate": 1.0743858329050371e-05,
+      "loss": 1.8982,
+      "step": 9104
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6452914495627337,
+      "learning_rate": 1.0738727267444387e-05,
+      "loss": 2.0279,
+      "step": 9105
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6054249485480094,
+      "learning_rate": 1.0733597096228249e-05,
+      "loss": 1.835,
+      "step": 9106
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5516434133446267,
+      "learning_rate": 1.0728467815722248e-05,
+      "loss": 1.916,
+      "step": 9107
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.570161196032374,
+      "learning_rate": 1.0723339426246631e-05,
+      "loss": 1.8933,
+      "step": 9108
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5996528106237009,
+      "learning_rate": 1.0718211928121585e-05,
+      "loss": 2.0236,
+      "step": 9109
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.585697169691358,
+      "learning_rate": 1.0713085321667238e-05,
+      "loss": 1.8724,
+      "step": 9110
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6188694330078404,
+      "learning_rate": 1.0707959607203665e-05,
+      "loss": 1.8512,
+      "step": 9111
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5919157344421371,
+      "learning_rate": 1.0702834785050891e-05,
+      "loss": 1.8169,
+      "step": 9112
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6611276542969717,
+      "learning_rate": 1.0697710855528872e-05,
+      "loss": 2.0389,
+      "step": 9113
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5837143797542886,
+      "learning_rate": 1.0692587818957523e-05,
+      "loss": 1.9498,
+      "step": 9114
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.592156555293683,
+      "learning_rate": 1.0687465675656688e-05,
+      "loss": 1.844,
+      "step": 9115
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5896782269159244,
+      "learning_rate": 1.0682344425946172e-05,
+      "loss": 1.8374,
+      "step": 9116
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5969475499696985,
+      "learning_rate": 1.0677224070145711e-05,
+      "loss": 2.0708,
+      "step": 9117
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6530626391598137,
+      "learning_rate": 1.067210460857499e-05,
+      "loss": 1.9103,
+      "step": 9118
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6318855688340977,
+      "learning_rate": 1.0666986041553639e-05,
+      "loss": 1.8565,
+      "step": 9119
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5507586475693985,
+      "learning_rate": 1.0661868369401232e-05,
+      "loss": 1.9463,
+      "step": 9120
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6459514645073638,
+      "learning_rate": 1.0656751592437295e-05,
+      "loss": 2.0995,
+      "step": 9121
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.618714926262875,
+      "learning_rate": 1.0651635710981259e-05,
+      "loss": 1.8599,
+      "step": 9122
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.640309620779375,
+      "learning_rate": 1.064652072535256e-05,
+      "loss": 1.8614,
+      "step": 9123
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6005298326894235,
+      "learning_rate": 1.0641406635870548e-05,
+      "loss": 1.8736,
+      "step": 9124
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5912396905761028,
+      "learning_rate": 1.0636293442854486e-05,
+      "loss": 2.0848,
+      "step": 9125
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5368890569683601,
+      "learning_rate": 1.0631181146623645e-05,
+      "loss": 1.9706,
+      "step": 9126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5780489721092229,
+      "learning_rate": 1.0626069747497189e-05,
+      "loss": 1.8754,
+      "step": 9127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7092488767632632,
+      "learning_rate": 1.0620959245794252e-05,
+      "loss": 1.843,
+      "step": 9128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6404823415655542,
+      "learning_rate": 1.0615849641833905e-05,
+      "loss": 2.0839,
+      "step": 9129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6090130922027726,
+      "learning_rate": 1.061074093593514e-05,
+      "loss": 1.8885,
+      "step": 9130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6531765974865161,
+      "learning_rate": 1.0605633128416941e-05,
+      "loss": 1.9048,
+      "step": 9131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6355725339599639,
+      "learning_rate": 1.0600526219598208e-05,
+      "loss": 1.9389,
+      "step": 9132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6311116807264902,
+      "learning_rate": 1.059542020979776e-05,
+      "loss": 2.0494,
+      "step": 9133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5977921596929435,
+      "learning_rate": 1.0590315099334419e-05,
+      "loss": 1.874,
+      "step": 9134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6165528990222782,
+      "learning_rate": 1.0585210888526898e-05,
+      "loss": 1.8753,
+      "step": 9135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7361342408858902,
+      "learning_rate": 1.0580107577693868e-05,
+      "loss": 1.8894,
+      "step": 9136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5651574776338373,
+      "learning_rate": 1.0575005167153976e-05,
+      "loss": 2.071,
+      "step": 9137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.53760120567633,
+      "learning_rate": 1.0569903657225766e-05,
+      "loss": 2.0183,
+      "step": 9138
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7385791444295767,
+      "learning_rate": 1.056480304822774e-05,
+      "loss": 1.8683,
+      "step": 9139
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6451145320076157,
+      "learning_rate": 1.0559703340478378e-05,
+      "loss": 1.8854,
+      "step": 9140
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5810387693752033,
+      "learning_rate": 1.055460453429604e-05,
+      "loss": 2.1023,
+      "step": 9141
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5908092875667219,
+      "learning_rate": 1.0549506629999104e-05,
+      "loss": 1.8354,
+      "step": 9142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5982518529546936,
+      "learning_rate": 1.0544409627905823e-05,
+      "loss": 1.8512,
+      "step": 9143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6154504164744401,
+      "learning_rate": 1.0539313528334427e-05,
+      "loss": 1.8631,
+      "step": 9144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6007975283152956,
+      "learning_rate": 1.053421833160311e-05,
+      "loss": 2.0937,
+      "step": 9145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6324812558371196,
+      "learning_rate": 1.0529124038029958e-05,
+      "loss": 1.8819,
+      "step": 9146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6269389856174107,
+      "learning_rate": 1.0524030647933045e-05,
+      "loss": 1.8783,
+      "step": 9147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5913707038973363,
+      "learning_rate": 1.0518938161630365e-05,
+      "loss": 1.831,
+      "step": 9148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5759501360372297,
+      "learning_rate": 1.0513846579439866e-05,
+      "loss": 2.0864,
+      "step": 9149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5372606769174609,
+      "learning_rate": 1.0508755901679434e-05,
+      "loss": 1.8412,
+      "step": 9150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5898067814951259,
+      "learning_rate": 1.0503666128666903e-05,
+      "loss": 1.9408,
+      "step": 9151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.647693619439521,
+      "learning_rate": 1.049857726072005e-05,
+      "loss": 1.8602,
+      "step": 9152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5743667830549007,
+      "learning_rate": 1.0493489298156589e-05,
+      "loss": 2.0675,
+      "step": 9153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5847235096846988,
+      "learning_rate": 1.0488402241294187e-05,
+      "loss": 1.8692,
+      "step": 9154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6210592134530121,
+      "learning_rate": 1.0483316090450448e-05,
+      "loss": 1.8803,
+      "step": 9155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.637971455570303,
+      "learning_rate": 1.0478230845942922e-05,
+      "loss": 1.8875,
+      "step": 9156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.598139127248657,
+      "learning_rate": 1.0473146508089102e-05,
+      "loss": 2.1185,
+      "step": 9157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5761171302761096,
+      "learning_rate": 1.0468063077206422e-05,
+      "loss": 1.855,
+      "step": 9158
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7166434478134002,
+      "learning_rate": 1.0462980553612262e-05,
+      "loss": 1.9408,
+      "step": 9159
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6374484654212716,
+      "learning_rate": 1.0457898937623947e-05,
+      "loss": 1.8762,
+      "step": 9160
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6478868611453413,
+      "learning_rate": 1.0452818229558741e-05,
+      "loss": 2.113,
+      "step": 9161
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6128435431946185,
+      "learning_rate": 1.0447738429733855e-05,
+      "loss": 1.8489,
+      "step": 9162
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6614622207371631,
+      "learning_rate": 1.044265953846644e-05,
+      "loss": 1.9346,
+      "step": 9163
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6086541404354413,
+      "learning_rate": 1.0437581556073594e-05,
+      "loss": 1.8777,
+      "step": 9164
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6600978241140333,
+      "learning_rate": 1.0432504482872355e-05,
+      "loss": 2.0631,
+      "step": 9165
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5955817409664877,
+      "learning_rate": 1.0427428319179705e-05,
+      "loss": 1.9142,
+      "step": 9166
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6612831621137155,
+      "learning_rate": 1.0422353065312573e-05,
+      "loss": 1.9185,
+      "step": 9167
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6342337894144602,
+      "learning_rate": 1.0417278721587834e-05,
+      "loss": 1.8357,
+      "step": 9168
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6358924008997958,
+      "learning_rate": 1.0412205288322272e-05,
+      "loss": 2.1075,
+      "step": 9169
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6253993767833744,
+      "learning_rate": 1.0407132765832672e-05,
+      "loss": 1.9541,
+      "step": 9170
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6434223852720643,
+      "learning_rate": 1.040206115443573e-05,
+      "loss": 1.8634,
+      "step": 9171
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7152873966179553,
+      "learning_rate": 1.0396990454448064e-05,
+      "loss": 1.8401,
+      "step": 9172
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6087864617375922,
+      "learning_rate": 1.0391920666186283e-05,
+      "loss": 2.0705,
+      "step": 9173
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.617145200602466,
+      "learning_rate": 1.0386851789966906e-05,
+      "loss": 1.8484,
+      "step": 9174
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6190056093802709,
+      "learning_rate": 1.0381783826106407e-05,
+      "loss": 1.8507,
+      "step": 9175
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6541125680774376,
+      "learning_rate": 1.03767167749212e-05,
+      "loss": 1.9178,
+      "step": 9176
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5851915526654308,
+      "learning_rate": 1.0371650636727625e-05,
+      "loss": 2.0858,
+      "step": 9177
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5937121348079563,
+      "learning_rate": 1.0366585411842004e-05,
+      "loss": 1.8649,
+      "step": 9178
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5598166854965481,
+      "learning_rate": 1.036152110058058e-05,
+      "loss": 1.854,
+      "step": 9179
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6081817359937839,
+      "learning_rate": 1.0356457703259512e-05,
+      "loss": 1.885,
+      "step": 9180
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6222803257389463,
+      "learning_rate": 1.0351395220194961e-05,
+      "loss": 2.0485,
+      "step": 9181
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5549624694026287,
+      "learning_rate": 1.034633365170298e-05,
+      "loss": 1.9528,
+      "step": 9182
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.562846511501338,
+      "learning_rate": 1.034127299809958e-05,
+      "loss": 1.8949,
+      "step": 9183
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.562859662000185,
+      "learning_rate": 1.0336213259700742e-05,
+      "loss": 1.9456,
+      "step": 9184
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5694966323681612,
+      "learning_rate": 1.033115443682234e-05,
+      "loss": 2.0711,
+      "step": 9185
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6002793016046768,
+      "learning_rate": 1.0326096529780221e-05,
+      "loss": 1.9055,
+      "step": 9186
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5856210081111061,
+      "learning_rate": 1.0321039538890196e-05,
+      "loss": 1.8865,
+      "step": 9187
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5369310040700239,
+      "learning_rate": 1.0315983464467965e-05,
+      "loss": 1.9726,
+      "step": 9188
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5365621363147519,
+      "learning_rate": 1.031092830682921e-05,
+      "loss": 2.0697,
+      "step": 9189
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5416529600726808,
+      "learning_rate": 1.0305874066289545e-05,
+      "loss": 1.8485,
+      "step": 9190
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5773556099348008,
+      "learning_rate": 1.0300820743164517e-05,
+      "loss": 1.8534,
+      "step": 9191
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6056971620359966,
+      "learning_rate": 1.0295768337769654e-05,
+      "loss": 1.8932,
+      "step": 9192
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.632840341869504,
+      "learning_rate": 1.0290716850420373e-05,
+      "loss": 2.1244,
+      "step": 9193
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5715013093662434,
+      "learning_rate": 1.0285666281432054e-05,
+      "loss": 1.995,
+      "step": 9194
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5758681065116839,
+      "learning_rate": 1.0280616631120058e-05,
+      "loss": 1.8977,
+      "step": 9195
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6121426899550583,
+      "learning_rate": 1.0275567899799627e-05,
+      "loss": 1.8658,
+      "step": 9196
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7007971085436007,
+      "learning_rate": 1.0270520087785979e-05,
+      "loss": 2.0747,
+      "step": 9197
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6246483726393967,
+      "learning_rate": 1.0265473195394271e-05,
+      "loss": 1.8726,
+      "step": 9198
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6167679172073429,
+      "learning_rate": 1.0260427222939603e-05,
+      "loss": 1.8508,
+      "step": 9199
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6576087787338054,
+      "learning_rate": 1.0255382170737017e-05,
+      "loss": 1.8802,
+      "step": 9200
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6744936214540866,
+      "learning_rate": 1.0250338039101492e-05,
+      "loss": 1.9684,
+      "step": 9201
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6939741958046834,
+      "learning_rate": 1.0245294828347956e-05,
+      "loss": 2.1056,
+      "step": 9202
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5684010768455341,
+      "learning_rate": 1.024025253879128e-05,
+      "loss": 1.8547,
+      "step": 9203
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.7084248380870669,
+      "learning_rate": 1.0235211170746272e-05,
+      "loss": 1.8738,
+      "step": 9204
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6754908617304664,
+      "learning_rate": 1.0230170724527687e-05,
+      "loss": 2.1127,
+      "step": 9205
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5801160443704926,
+      "learning_rate": 1.022513120045022e-05,
+      "loss": 1.8622,
+      "step": 9206
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6979906434215114,
+      "learning_rate": 1.0220092598828506e-05,
+      "loss": 1.9711,
+      "step": 9207
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5714614698029041,
+      "learning_rate": 1.021505491997713e-05,
+      "loss": 1.8611,
+      "step": 9208
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.573599780203101,
+      "learning_rate": 1.0210018164210613e-05,
+      "loss": 2.0945,
+      "step": 9209
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5933984367883695,
+      "learning_rate": 1.0204982331843421e-05,
+      "loss": 1.8765,
+      "step": 9210
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6089982838662868,
+      "learning_rate": 1.019994742318996e-05,
+      "loss": 1.8725,
+      "step": 9211
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6617475841757581,
+      "learning_rate": 1.0194913438564585e-05,
+      "loss": 1.9014,
+      "step": 9212
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5486714680552005,
+      "learning_rate": 1.0189880378281583e-05,
+      "loss": 1.9564,
+      "step": 9213
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5624858566544474,
+      "learning_rate": 1.018484824265519e-05,
+      "loss": 2.0566,
+      "step": 9214
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6607205287791696,
+      "learning_rate": 1.0179817031999592e-05,
+      "loss": 1.9036,
+      "step": 9215
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.581678025841982,
+      "learning_rate": 1.0174786746628882e-05,
+      "loss": 1.9132,
+      "step": 9216
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5755037036557926,
+      "learning_rate": 1.0169757386857148e-05,
+      "loss": 2.0799,
+      "step": 9217
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6000446840684024,
+      "learning_rate": 1.0164728952998393e-05,
+      "loss": 1.8344,
+      "step": 9218
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5289722188984569,
+      "learning_rate": 1.0159701445366537e-05,
+      "loss": 1.8892,
+      "step": 9219
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6116870511369378,
+      "learning_rate": 1.0154674864275492e-05,
+      "loss": 1.8453,
+      "step": 9220
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6018517427676207,
+      "learning_rate": 1.0149649210039095e-05,
+      "loss": 2.0773,
+      "step": 9221
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5771253800150048,
+      "learning_rate": 1.0144624482971082e-05,
+      "loss": 1.8452,
+      "step": 9222
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5581315325428106,
+      "learning_rate": 1.0139600683385211e-05,
+      "loss": 1.8911,
+      "step": 9223
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6220632438330608,
+      "learning_rate": 1.0134577811595098e-05,
+      "loss": 1.8631,
+      "step": 9224
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.528929299601843,
+      "learning_rate": 1.012955586791437e-05,
+      "loss": 1.9446,
+      "step": 9225
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5723426600599879,
+      "learning_rate": 1.0124534852656569e-05,
+      "loss": 2.0402,
+      "step": 9226
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5786359998373249,
+      "learning_rate": 1.0119514766135146e-05,
+      "loss": 1.862,
+      "step": 9227
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6125392689275904,
+      "learning_rate": 1.011449560866356e-05,
+      "loss": 1.842,
+      "step": 9228
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5790911191526162,
+      "learning_rate": 1.0109477380555172e-05,
+      "loss": 2.0697,
+      "step": 9229
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.554725571518991,
+      "learning_rate": 1.0104460082123263e-05,
+      "loss": 1.8748,
+      "step": 9230
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5634733682984511,
+      "learning_rate": 1.009944371368112e-05,
+      "loss": 1.8902,
+      "step": 9231
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6302671616588267,
+      "learning_rate": 1.0094428275541914e-05,
+      "loss": 1.9676,
+      "step": 9232
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6394008497227877,
+      "learning_rate": 1.0089413768018774e-05,
+      "loss": 1.8946,
+      "step": 9233
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5990141180389505,
+      "learning_rate": 1.0084400191424804e-05,
+      "loss": 2.1013,
+      "step": 9234
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6270817832424838,
+      "learning_rate": 1.0079387546072996e-05,
+      "loss": 1.8989,
+      "step": 9235
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6068966137492929,
+      "learning_rate": 1.007437583227632e-05,
+      "loss": 1.8663,
+      "step": 9236
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5884188737728199,
+      "learning_rate": 1.0069365050347674e-05,
+      "loss": 2.0516,
+      "step": 9237
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6081604267299447,
+      "learning_rate": 1.0064355200599897e-05,
+      "loss": 1.9799,
+      "step": 9238
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6268185048734554,
+      "learning_rate": 1.0059346283345802e-05,
+      "loss": 1.9341,
+      "step": 9239
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5728982552830466,
+      "learning_rate": 1.0054338298898086e-05,
+      "loss": 1.8625,
+      "step": 9240
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5748449065888819,
+      "learning_rate": 1.0049331247569418e-05,
+      "loss": 2.0726,
+      "step": 9241
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6534999963919573,
+      "learning_rate": 1.0044325129672438e-05,
+      "loss": 1.9365,
+      "step": 9242
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6369739411115005,
+      "learning_rate": 1.0039319945519671e-05,
+      "loss": 1.8588,
+      "step": 9243
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6036211428324699,
+      "learning_rate": 1.0034315695423619e-05,
+      "loss": 1.9864,
+      "step": 9244
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5631335037004745,
+      "learning_rate": 1.002931237969672e-05,
+      "loss": 1.9031,
+      "step": 9245
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6229290213338599,
+      "learning_rate": 1.002430999865135e-05,
+      "loss": 2.0904,
+      "step": 9246
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.614150747591734,
+      "learning_rate": 1.0019308552599827e-05,
+      "loss": 1.9015,
+      "step": 9247
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6020129951521515,
+      "learning_rate": 1.0014308041854414e-05,
+      "loss": 1.8864,
+      "step": 9248
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5790584168232801,
+      "learning_rate": 1.0009308466727313e-05,
+      "loss": 2.0549,
+      "step": 9249
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5634899495336889,
+      "learning_rate": 1.0004309827530667e-05,
+      "loss": 1.9986,
+      "step": 9250
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5895714685956942,
+      "learning_rate": 9.999312124576563e-06,
+      "loss": 1.848,
+      "step": 9251
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6765046884733992,
+      "learning_rate": 9.994315358177028e-06,
+      "loss": 1.831,
+      "step": 9252
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5876158533732383,
+      "learning_rate": 9.989319528644028e-06,
+      "loss": 1.8814,
+      "step": 9253
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6043639542678267,
+      "learning_rate": 9.984324636289477e-06,
+      "loss": 2.0544,
+      "step": 9254
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5817899931944838,
+      "learning_rate": 9.979330681425226e-06,
+      "loss": 1.9131,
+      "step": 9255
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6290646883876165,
+      "learning_rate": 9.974337664363065e-06,
+      "loss": 1.9108,
+      "step": 9256
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5638846314317384,
+      "learning_rate": 9.969345585414732e-06,
+      "loss": 1.8902,
+      "step": 9257
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6045638657858438,
+      "learning_rate": 9.964354444891902e-06,
+      "loss": 2.0607,
+      "step": 9258
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5782976219339114,
+      "learning_rate": 9.959364243106189e-06,
+      "loss": 1.8826,
+      "step": 9259
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5915224915822787,
+      "learning_rate": 9.954374980369158e-06,
+      "loss": 1.8847,
+      "step": 9260
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6635595631159843,
+      "learning_rate": 9.949386656992305e-06,
+      "loss": 2.0862,
+      "step": 9261
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.627402783443919,
+      "learning_rate": 9.944399273287081e-06,
+      "loss": 1.8954,
+      "step": 9262
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5471039164639492,
+      "learning_rate": 9.93941282956484e-06,
+      "loss": 1.9705,
+      "step": 9263
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6161588034091146,
+      "learning_rate": 9.93442732613694e-06,
+      "loss": 1.8726,
+      "step": 9264
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5791925210441221,
+      "learning_rate": 9.929442763314638e-06,
+      "loss": 1.8929,
+      "step": 9265
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6022714609027946,
+      "learning_rate": 9.92445914140912e-06,
+      "loss": 2.0459,
+      "step": 9266
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5699188505675418,
+      "learning_rate": 9.919476460731559e-06,
+      "loss": 1.8858,
+      "step": 9267
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5560944783914413,
+      "learning_rate": 9.914494721593043e-06,
+      "loss": 1.8523,
+      "step": 9268
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6095869623516343,
+      "learning_rate": 9.909513924304576e-06,
+      "loss": 1.9441,
+      "step": 9269
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5579724162565891,
+      "learning_rate": 9.904534069177165e-06,
+      "loss": 2.0803,
+      "step": 9270
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6019718343321604,
+      "learning_rate": 9.89955515652169e-06,
+      "loss": 1.8929,
+      "step": 9271
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6244384296453995,
+      "learning_rate": 9.894577186649028e-06,
+      "loss": 1.887,
+      "step": 9272
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5687379927810733,
+      "learning_rate": 9.889600159869977e-06,
+      "loss": 2.0805,
+      "step": 9273
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5901804755337678,
+      "learning_rate": 9.884624076495245e-06,
+      "loss": 1.8696,
+      "step": 9274
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6001413139691167,
+      "learning_rate": 9.879648936835537e-06,
+      "loss": 1.9881,
+      "step": 9275
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5814126159887417,
+      "learning_rate": 9.87467474120147e-06,
+      "loss": 1.8028,
+      "step": 9276
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6070093577798656,
+      "learning_rate": 9.869701489903576e-06,
+      "loss": 1.8732,
+      "step": 9277
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5828542656286376,
+      "learning_rate": 9.864729183252395e-06,
+      "loss": 2.0749,
+      "step": 9278
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6883931659187985,
+      "learning_rate": 9.859757821558337e-06,
+      "loss": 1.8866,
+      "step": 9279
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6404242987489672,
+      "learning_rate": 9.85478740513179e-06,
+      "loss": 1.8502,
+      "step": 9280
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5586243460154057,
+      "learning_rate": 9.8498179342831e-06,
+      "loss": 1.9476,
+      "step": 9281
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6985995473612314,
+      "learning_rate": 9.844849409322507e-06,
+      "loss": 2.0613,
+      "step": 9282
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6077835055473094,
+      "learning_rate": 9.839881830560222e-06,
+      "loss": 1.8633,
+      "step": 9283
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6164790779805088,
+      "learning_rate": 9.834915198306394e-06,
+      "loss": 1.8772,
+      "step": 9284
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6740379464332183,
+      "learning_rate": 9.829949512871112e-06,
+      "loss": 1.8693,
+      "step": 9285
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5611387469339576,
+      "learning_rate": 9.8249847745644e-06,
+      "loss": 2.0391,
+      "step": 9286
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6192591806087414,
+      "learning_rate": 9.820020983696228e-06,
+      "loss": 1.902,
+      "step": 9287
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6901033779313758,
+      "learning_rate": 9.815058140576502e-06,
+      "loss": 1.9258,
+      "step": 9288
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5461962916869743,
+      "learning_rate": 9.810096245515094e-06,
+      "loss": 1.812,
+      "step": 9289
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6273547982635934,
+      "learning_rate": 9.805135298821768e-06,
+      "loss": 1.9908,
+      "step": 9290
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6162977639484502,
+      "learning_rate": 9.80017530080627e-06,
+      "loss": 1.9147,
+      "step": 9291
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5834847988762599,
+      "learning_rate": 9.795216251778266e-06,
+      "loss": 1.871,
+      "step": 9292
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5876223912801993,
+      "learning_rate": 9.790258152047376e-06,
+      "loss": 2.0444,
+      "step": 9293
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5290928720043104,
+      "learning_rate": 9.785301001923155e-06,
+      "loss": 1.9755,
+      "step": 9294
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5791252143136313,
+      "learning_rate": 9.780344801715096e-06,
+      "loss": 1.8599,
+      "step": 9295
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.588392931823372,
+      "learning_rate": 9.775389551732633e-06,
+      "loss": 1.9003,
+      "step": 9296
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5529629571744131,
+      "learning_rate": 9.770435252285146e-06,
+      "loss": 1.8638,
+      "step": 9297
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.537131013553097,
+      "learning_rate": 9.76548190368195e-06,
+      "loss": 2.1159,
+      "step": 9298
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5839084429182194,
+      "learning_rate": 9.760529506232306e-06,
+      "loss": 1.842,
+      "step": 9299
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5553734722973852,
+      "learning_rate": 9.755578060245406e-06,
+      "loss": 1.9605,
+      "step": 9300
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5456225138734266,
+      "learning_rate": 9.750627566030395e-06,
+      "loss": 1.8193,
+      "step": 9301
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5394570531306575,
+      "learning_rate": 9.745678023896353e-06,
+      "loss": 2.0634,
+      "step": 9302
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.598553817223424,
+      "learning_rate": 9.7407294341523e-06,
+      "loss": 1.8828,
+      "step": 9303
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5753502339997031,
+      "learning_rate": 9.735781797107188e-06,
+      "loss": 1.9376,
+      "step": 9304
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.557206453832556,
+      "learning_rate": 9.730835113069931e-06,
+      "loss": 2.0949,
+      "step": 9305
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5718145664194376,
+      "learning_rate": 9.725889382349362e-06,
+      "loss": 1.9781,
+      "step": 9306
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6252340794307238,
+      "learning_rate": 9.72094460525427e-06,
+      "loss": 1.8936,
+      "step": 9307
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6068262860839396,
+      "learning_rate": 9.71600078209337e-06,
+      "loss": 1.7976,
+      "step": 9308
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5593851232236426,
+      "learning_rate": 9.711057913175334e-06,
+      "loss": 1.8861,
+      "step": 9309
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6343592871161907,
+      "learning_rate": 9.706115998808757e-06,
+      "loss": 2.0901,
+      "step": 9310
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.603726929348084,
+      "learning_rate": 9.701175039302191e-06,
+      "loss": 1.9028,
+      "step": 9311
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.540844101733581,
+      "learning_rate": 9.696235034964122e-06,
+      "loss": 1.9085,
+      "step": 9312
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5818997421729912,
+      "learning_rate": 9.691295986102953e-06,
+      "loss": 1.889,
+      "step": 9313
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5774282749784887,
+      "learning_rate": 9.686357893027073e-06,
+      "loss": 2.0891,
+      "step": 9314
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5455396012599567,
+      "learning_rate": 9.68142075604479e-06,
+      "loss": 1.8386,
+      "step": 9315
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5560231317072828,
+      "learning_rate": 9.67648457546432e-06,
+      "loss": 1.9105,
+      "step": 9316
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5911131358901061,
+      "learning_rate": 9.671549351593889e-06,
+      "loss": 1.8223,
+      "step": 9317
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5950767093686201,
+      "learning_rate": 9.666615084741593e-06,
+      "loss": 2.1064,
+      "step": 9318
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6629213462850306,
+      "learning_rate": 9.661681775215498e-06,
+      "loss": 1.8686,
+      "step": 9319
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5162172326512674,
+      "learning_rate": 9.656749423323639e-06,
+      "loss": 1.8834,
+      "step": 9320
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5952177313652275,
+      "learning_rate": 9.65181802937393e-06,
+      "loss": 1.8938,
+      "step": 9321
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6794197089375936,
+      "learning_rate": 9.646887593674278e-06,
+      "loss": 2.0959,
+      "step": 9322
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6171538800800993,
+      "learning_rate": 9.641958116532518e-06,
+      "loss": 1.8379,
+      "step": 9323
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6181360025815352,
+      "learning_rate": 9.637029598256386e-06,
+      "loss": 1.9082,
+      "step": 9324
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.600046441319299,
+      "learning_rate": 9.632102039153628e-06,
+      "loss": 1.9493,
+      "step": 9325
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5735399393542862,
+      "learning_rate": 9.627175439531862e-06,
+      "loss": 2.069,
+      "step": 9326
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6123741758597079,
+      "learning_rate": 9.62224979969868e-06,
+      "loss": 1.86,
+      "step": 9327
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5808501812293857,
+      "learning_rate": 9.617325119961631e-06,
+      "loss": 1.8406,
+      "step": 9328
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5973991678002561,
+      "learning_rate": 9.612401400628162e-06,
+      "loss": 1.8617,
+      "step": 9329
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.559626654858846,
+      "learning_rate": 9.607478642005685e-06,
+      "loss": 2.1187,
+      "step": 9330
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5545233784869005,
+      "learning_rate": 9.602556844401553e-06,
+      "loss": 1.9134,
+      "step": 9331
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5700435269417434,
+      "learning_rate": 9.59763600812305e-06,
+      "loss": 1.8431,
+      "step": 9332
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6178587835056821,
+      "learning_rate": 9.592716133477406e-06,
+      "loss": 1.895,
+      "step": 9333
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5575293996099993,
+      "learning_rate": 9.587797220771789e-06,
+      "loss": 2.1126,
+      "step": 9334
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5992280849496696,
+      "learning_rate": 9.582879270313298e-06,
+      "loss": 1.8719,
+      "step": 9335
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6058746809145183,
+      "learning_rate": 9.577962282409002e-06,
+      "loss": 1.9174,
+      "step": 9336
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6047944422993576,
+      "learning_rate": 9.573046257365873e-06,
+      "loss": 1.9769,
+      "step": 9337
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6084504487046691,
+      "learning_rate": 9.56813119549084e-06,
+      "loss": 2.0983,
+      "step": 9338
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6121038646427984,
+      "learning_rate": 9.56321709709077e-06,
+      "loss": 1.873,
+      "step": 9339
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6689693776999579,
+      "learning_rate": 9.558303962472475e-06,
+      "loss": 1.8835,
+      "step": 9340
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6413289947369122,
+      "learning_rate": 9.553391791942701e-06,
+      "loss": 1.8679,
+      "step": 9341
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5876338706567029,
+      "learning_rate": 9.54848058580813e-06,
+      "loss": 2.0759,
+      "step": 9342
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5513303313729443,
+      "learning_rate": 9.543570344375394e-06,
+      "loss": 1.9421,
+      "step": 9343
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5783839120406689,
+      "learning_rate": 9.53866106795106e-06,
+      "loss": 1.8721,
+      "step": 9344
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5686676762625879,
+      "learning_rate": 9.53375275684163e-06,
+      "loss": 1.8483,
+      "step": 9345
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6792951813258494,
+      "learning_rate": 9.528845411353554e-06,
+      "loss": 2.0668,
+      "step": 9346
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5815055745084498,
+      "learning_rate": 9.523939031793214e-06,
+      "loss": 1.8526,
+      "step": 9347
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6181259903879921,
+      "learning_rate": 9.519033618466938e-06,
+      "loss": 1.8633,
+      "step": 9348
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5856358340516395,
+      "learning_rate": 9.514129171680992e-06,
+      "loss": 1.8275,
+      "step": 9349
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5724716323062503,
+      "learning_rate": 9.509225691741578e-06,
+      "loss": 2.1071,
+      "step": 9350
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6029574466175578,
+      "learning_rate": 9.504323178954852e-06,
+      "loss": 1.8944,
+      "step": 9351
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5814076117174487,
+      "learning_rate": 9.499421633626867e-06,
+      "loss": 1.8776,
+      "step": 9352
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6288234765318846,
+      "learning_rate": 9.494521056063679e-06,
+      "loss": 1.9064,
+      "step": 9353
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5662019462728898,
+      "learning_rate": 9.489621446571239e-06,
+      "loss": 2.0946,
+      "step": 9354
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6283084354313933,
+      "learning_rate": 9.484722805455452e-06,
+      "loss": 1.8854,
+      "step": 9355
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5574689220273532,
+      "learning_rate": 9.479825133022156e-06,
+      "loss": 1.9201,
+      "step": 9356
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5822509772374407,
+      "learning_rate": 9.474928429577135e-06,
+      "loss": 1.8668,
+      "step": 9357
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5515969106181204,
+      "learning_rate": 9.470032695426111e-06,
+      "loss": 2.0724,
+      "step": 9358
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6058671368309673,
+      "learning_rate": 9.465137930874757e-06,
+      "loss": 1.9054,
+      "step": 9359
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.572273402916584,
+      "learning_rate": 9.460244136228638e-06,
+      "loss": 1.845,
+      "step": 9360
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5944290046213113,
+      "learning_rate": 9.455351311793328e-06,
+      "loss": 1.8932,
+      "step": 9361
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6212275596774309,
+      "learning_rate": 9.450459457874305e-06,
+      "loss": 2.145,
+      "step": 9362
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.602064953299212,
+      "learning_rate": 9.445568574776958e-06,
+      "loss": 1.8713,
+      "step": 9363
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5613116881050803,
+      "learning_rate": 9.440678662806682e-06,
+      "loss": 1.9028,
+      "step": 9364
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6295597609600294,
+      "learning_rate": 9.435789722268745e-06,
+      "loss": 1.8922,
+      "step": 9365
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5943555245365125,
+      "learning_rate": 9.430901753468388e-06,
+      "loss": 2.0702,
+      "step": 9366
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6060005159332501,
+      "learning_rate": 9.426014756710813e-06,
+      "loss": 1.836,
+      "step": 9367
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5658496325921856,
+      "learning_rate": 9.421128732301098e-06,
+      "loss": 1.9566,
+      "step": 9368
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.587653914794449,
+      "learning_rate": 9.416243680544323e-06,
+      "loss": 1.8741,
+      "step": 9369
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5743503097097566,
+      "learning_rate": 9.411359601745484e-06,
+      "loss": 2.0568,
+      "step": 9370
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5399595927334069,
+      "learning_rate": 9.40647649620949e-06,
+      "loss": 1.8602,
+      "step": 9371
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6146152043421261,
+      "learning_rate": 9.401594364241242e-06,
+      "loss": 1.8741,
+      "step": 9372
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6778021536665929,
+      "learning_rate": 9.396713206145532e-06,
+      "loss": 1.8854,
+      "step": 9373
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5458681620257805,
+      "learning_rate": 9.391833022227106e-06,
+      "loss": 2.0699,
+      "step": 9374
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5714016278551034,
+      "learning_rate": 9.386953812790683e-06,
+      "loss": 1.9165,
+      "step": 9375
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5719024554631115,
+      "learning_rate": 9.382075578140862e-06,
+      "loss": 1.8661,
+      "step": 9376
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6901872939184617,
+      "learning_rate": 9.377198318582226e-06,
+      "loss": 1.8194,
+      "step": 9377
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6563965744280347,
+      "learning_rate": 9.372322034419276e-06,
+      "loss": 2.0907,
+      "step": 9378
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6260836635659269,
+      "learning_rate": 9.367446725956464e-06,
+      "loss": 1.843,
+      "step": 9379
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5835467090669153,
+      "learning_rate": 9.362572393498172e-06,
+      "loss": 1.8845,
+      "step": 9380
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5699337617864676,
+      "learning_rate": 9.357699037348724e-06,
+      "loss": 1.934,
+      "step": 9381
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6023244084027957,
+      "learning_rate": 9.35282665781239e-06,
+      "loss": 2.0561,
+      "step": 9382
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.621481098854153,
+      "learning_rate": 9.347955255193364e-06,
+      "loss": 1.9285,
+      "step": 9383
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5603901255737668,
+      "learning_rate": 9.343084829795793e-06,
+      "loss": 1.8745,
+      "step": 9384
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6488501978683392,
+      "learning_rate": 9.338215381923753e-06,
+      "loss": 1.8573,
+      "step": 9385
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5892705241135113,
+      "learning_rate": 9.333346911881271e-06,
+      "loss": 2.0976,
+      "step": 9386
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5760798551442249,
+      "learning_rate": 9.328479419972303e-06,
+      "loss": 1.9216,
+      "step": 9387
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6692091759960711,
+      "learning_rate": 9.32361290650074e-06,
+      "loss": 1.8401,
+      "step": 9388
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.621527679608654,
+      "learning_rate": 9.318747371770426e-06,
+      "loss": 1.8896,
+      "step": 9389
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6029148964668166,
+      "learning_rate": 9.313882816085135e-06,
+      "loss": 2.0533,
+      "step": 9390
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6068531747242455,
+      "learning_rate": 9.309019239748578e-06,
+      "loss": 1.913,
+      "step": 9391
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6044104142362767,
+      "learning_rate": 9.30415664306441e-06,
+      "loss": 1.8432,
+      "step": 9392
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5425562376020766,
+      "learning_rate": 9.299295026336225e-06,
+      "loss": 1.9906,
+      "step": 9393
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5785055803951957,
+      "learning_rate": 9.294434389867548e-06,
+      "loss": 2.0687,
+      "step": 9394
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6341989758770389,
+      "learning_rate": 9.289574733961856e-06,
+      "loss": 1.9035,
+      "step": 9395
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5692716145176596,
+      "learning_rate": 9.284716058922553e-06,
+      "loss": 1.8729,
+      "step": 9396
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5825671876383791,
+      "learning_rate": 9.279858365052985e-06,
+      "loss": 1.9146,
+      "step": 9397
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6695415546558015,
+      "learning_rate": 9.275001652656446e-06,
+      "loss": 2.0661,
+      "step": 9398
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5401738138764443,
+      "learning_rate": 9.270145922036139e-06,
+      "loss": 1.9267,
+      "step": 9399
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.58323333393003,
+      "learning_rate": 9.26529117349525e-06,
+      "loss": 1.8519,
+      "step": 9400
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6051823583841881,
+      "learning_rate": 9.260437407336872e-06,
+      "loss": 1.8358,
+      "step": 9401
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6234844133354178,
+      "learning_rate": 9.255584623864046e-06,
+      "loss": 2.0876,
+      "step": 9402
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5723569817697856,
+      "learning_rate": 9.250732823379754e-06,
+      "loss": 1.8554,
+      "step": 9403
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5681590229916659,
+      "learning_rate": 9.245882006186913e-06,
+      "loss": 1.8464,
+      "step": 9404
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5952711982025045,
+      "learning_rate": 9.241032172588373e-06,
+      "loss": 1.9631,
+      "step": 9405
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6163154862120638,
+      "learning_rate": 9.236183322886945e-06,
+      "loss": 2.1091,
+      "step": 9406
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6041059810488946,
+      "learning_rate": 9.231335457385335e-06,
+      "loss": 1.8495,
+      "step": 9407
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5921346946201703,
+      "learning_rate": 9.226488576386238e-06,
+      "loss": 1.8783,
+      "step": 9408
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5920007996975722,
+      "learning_rate": 9.22164268019227e-06,
+      "loss": 1.8993,
+      "step": 9409
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5962259412905768,
+      "learning_rate": 9.21679776910595e-06,
+      "loss": 2.0703,
+      "step": 9410
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6611801894038234,
+      "learning_rate": 9.211953843429799e-06,
+      "loss": 1.8377,
+      "step": 9411
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5846130304552307,
+      "learning_rate": 9.207110903466224e-06,
+      "loss": 1.9769,
+      "step": 9412
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6086267574917718,
+      "learning_rate": 9.202268949517584e-06,
+      "loss": 1.8348,
+      "step": 9413
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6175484212041576,
+      "learning_rate": 9.197427981886208e-06,
+      "loss": 2.0989,
+      "step": 9414
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6192458373781426,
+      "learning_rate": 9.192588000874314e-06,
+      "loss": 1.8728,
+      "step": 9415
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7007182080599461,
+      "learning_rate": 9.187749006784083e-06,
+      "loss": 1.8523,
+      "step": 9416
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6307553660134358,
+      "learning_rate": 9.182910999917655e-06,
+      "loss": 1.8979,
+      "step": 9417
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5083796241128526,
+      "learning_rate": 9.178073980577053e-06,
+      "loss": 1.9393,
+      "step": 9418
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6303539237793349,
+      "learning_rate": 9.173237949064308e-06,
+      "loss": 2.0958,
+      "step": 9419
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6209768820241488,
+      "learning_rate": 9.168402905681331e-06,
+      "loss": 1.9208,
+      "step": 9420
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5384370279887928,
+      "learning_rate": 9.163568850729986e-06,
+      "loss": 1.9146,
+      "step": 9421
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6571652833999045,
+      "learning_rate": 9.158735784512112e-06,
+      "loss": 2.0403,
+      "step": 9422
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6489007292984263,
+      "learning_rate": 9.153903707329437e-06,
+      "loss": 1.8428,
+      "step": 9423
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5703503555823981,
+      "learning_rate": 9.149072619483636e-06,
+      "loss": 2.0104,
+      "step": 9424
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5668344564488843,
+      "learning_rate": 9.144242521276369e-06,
+      "loss": 1.8733,
+      "step": 9425
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5746449095381313,
+      "learning_rate": 9.13941341300917e-06,
+      "loss": 2.0575,
+      "step": 9426
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6435069762237243,
+      "learning_rate": 9.13458529498355e-06,
+      "loss": 1.8776,
+      "step": 9427
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5907001994440858,
+      "learning_rate": 9.129758167500944e-06,
+      "loss": 1.8389,
+      "step": 9428
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6096899242007797,
+      "learning_rate": 9.124932030862735e-06,
+      "loss": 1.8674,
+      "step": 9429
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6442030602900712,
+      "learning_rate": 9.120106885370233e-06,
+      "loss": 1.9779,
+      "step": 9430
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6170124255336897,
+      "learning_rate": 9.115282731324696e-06,
+      "loss": 2.0336,
+      "step": 9431
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5595169007306818,
+      "learning_rate": 9.110459569027313e-06,
+      "loss": 1.8517,
+      "step": 9432
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5871197982990949,
+      "learning_rate": 9.105637398779216e-06,
+      "loss": 1.8355,
+      "step": 9433
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5660837130148821,
+      "learning_rate": 9.10081622088147e-06,
+      "loss": 2.0648,
+      "step": 9434
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5760349262808508,
+      "learning_rate": 9.095996035635084e-06,
+      "loss": 1.8565,
+      "step": 9435
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5729605771402125,
+      "learning_rate": 9.091176843341001e-06,
+      "loss": 1.9461,
+      "step": 9436
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6030907602604958,
+      "learning_rate": 9.086358644300102e-06,
+      "loss": 1.8891,
+      "step": 9437
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6453222172628725,
+      "learning_rate": 9.08154143881321e-06,
+      "loss": 2.0475,
+      "step": 9438
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5942308760362073,
+      "learning_rate": 9.07672522718108e-06,
+      "loss": 1.9126,
+      "step": 9439
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6188324557634042,
+      "learning_rate": 9.071910009704407e-06,
+      "loss": 1.86,
+      "step": 9440
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6245435240497579,
+      "learning_rate": 9.067095786683826e-06,
+      "loss": 1.8684,
+      "step": 9441
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6534368737723625,
+      "learning_rate": 9.06228255841991e-06,
+      "loss": 2.0437,
+      "step": 9442
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5507354831165161,
+      "learning_rate": 9.057470325213166e-06,
+      "loss": 1.941,
+      "step": 9443
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.639838833256376,
+      "learning_rate": 9.052659087364046e-06,
+      "loss": 1.9231,
+      "step": 9444
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6260917906716007,
+      "learning_rate": 9.04784884517294e-06,
+      "loss": 1.8535,
+      "step": 9445
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6576683699220596,
+      "learning_rate": 9.043039598940144e-06,
+      "loss": 2.054,
+      "step": 9446
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6235239081672191,
+      "learning_rate": 9.03823134896595e-06,
+      "loss": 1.8874,
+      "step": 9447
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5951418950763161,
+      "learning_rate": 9.033424095550556e-06,
+      "loss": 1.8911,
+      "step": 9448
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5845715382159836,
+      "learning_rate": 9.02861783899407e-06,
+      "loss": 1.9123,
+      "step": 9449
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6876808743205632,
+      "learning_rate": 9.023812579596588e-06,
+      "loss": 1.9128,
+      "step": 9450
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5611318276007188,
+      "learning_rate": 9.019008317658123e-06,
+      "loss": 2.0361,
+      "step": 9451
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6032761363567714,
+      "learning_rate": 9.014205053478619e-06,
+      "loss": 1.8789,
+      "step": 9452
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6347957666673468,
+      "learning_rate": 9.009402787357973e-06,
+      "loss": 1.8825,
+      "step": 9453
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5791982929822131,
+      "learning_rate": 9.004601519595988e-06,
+      "loss": 2.1024,
+      "step": 9454
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5729473772425203,
+      "learning_rate": 8.999801250492446e-06,
+      "loss": 1.971,
+      "step": 9455
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6257006091154033,
+      "learning_rate": 8.995001980347052e-06,
+      "loss": 1.8897,
+      "step": 9456
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6076539047162678,
+      "learning_rate": 8.990203709459417e-06,
+      "loss": 1.835,
+      "step": 9457
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5850401819996345,
+      "learning_rate": 8.985406438129142e-06,
+      "loss": 2.055,
+      "step": 9458
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5689146963705463,
+      "learning_rate": 8.980610166655742e-06,
+      "loss": 1.8846,
+      "step": 9459
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.575624075341681,
+      "learning_rate": 8.975814895338645e-06,
+      "loss": 1.8583,
+      "step": 9460
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6124471190544843,
+      "learning_rate": 8.971020624477267e-06,
+      "loss": 1.9483,
+      "step": 9461
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5870489599018549,
+      "learning_rate": 8.96622735437091e-06,
+      "loss": 1.8386,
+      "step": 9462
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6236758126541322,
+      "learning_rate": 8.961435085318841e-06,
+      "loss": 2.0685,
+      "step": 9463
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6193977701908088,
+      "learning_rate": 8.956643817620286e-06,
+      "loss": 1.8592,
+      "step": 9464
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5877995699443526,
+      "learning_rate": 8.951853551574346e-06,
+      "loss": 1.906,
+      "step": 9465
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6054047890924181,
+      "learning_rate": 8.947064287480134e-06,
+      "loss": 2.0601,
+      "step": 9466
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5390324628691262,
+      "learning_rate": 8.942276025636636e-06,
+      "loss": 1.939,
+      "step": 9467
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6175430859112587,
+      "learning_rate": 8.937488766342806e-06,
+      "loss": 1.8923,
+      "step": 9468
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6229227645876635,
+      "learning_rate": 8.932702509897556e-06,
+      "loss": 1.8753,
+      "step": 9469
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5736164626858687,
+      "learning_rate": 8.927917256599683e-06,
+      "loss": 1.8698,
+      "step": 9470
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6379059567881266,
+      "learning_rate": 8.923133006747955e-06,
+      "loss": 2.0279,
+      "step": 9471
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5760810313443809,
+      "learning_rate": 8.918349760641095e-06,
+      "loss": 1.8627,
+      "step": 9472
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5697705538618044,
+      "learning_rate": 8.913567518577718e-06,
+      "loss": 1.8896,
+      "step": 9473
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6981249684323197,
+      "learning_rate": 8.908786280856405e-06,
+      "loss": 1.9395,
+      "step": 9474
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6218000335349834,
+      "learning_rate": 8.904006047775668e-06,
+      "loss": 2.1116,
+      "step": 9475
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6178864632704987,
+      "learning_rate": 8.899226819633954e-06,
+      "loss": 1.9145,
+      "step": 9476
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7450733242307268,
+      "learning_rate": 8.89444859672966e-06,
+      "loss": 1.8561,
+      "step": 9477
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5841526035503444,
+      "learning_rate": 8.889671379361098e-06,
+      "loss": 2.0579,
+      "step": 9478
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6473474635039632,
+      "learning_rate": 8.884895167826537e-06,
+      "loss": 1.8439,
+      "step": 9479
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7216526059860939,
+      "learning_rate": 8.880119962424175e-06,
+      "loss": 1.9143,
+      "step": 9480
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6470226772967926,
+      "learning_rate": 8.875345763452142e-06,
+      "loss": 1.8636,
+      "step": 9481
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6166961930964023,
+      "learning_rate": 8.870572571208516e-06,
+      "loss": 1.8053,
+      "step": 9482
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.875935645520168,
+      "learning_rate": 8.865800385991307e-06,
+      "loss": 2.0708,
+      "step": 9483
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5715072908908646,
+      "learning_rate": 8.86102920809846e-06,
+      "loss": 1.857,
+      "step": 9484
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6951406734229545,
+      "learning_rate": 8.856259037827861e-06,
+      "loss": 1.8902,
+      "step": 9485
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7076093690015756,
+      "learning_rate": 8.851489875477328e-06,
+      "loss": 1.9391,
+      "step": 9486
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5949623925899069,
+      "learning_rate": 8.846721721344625e-06,
+      "loss": 2.047,
+      "step": 9487
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6090980475835315,
+      "learning_rate": 8.841954575727442e-06,
+      "loss": 1.824,
+      "step": 9488
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7688083262452369,
+      "learning_rate": 8.837188438923413e-06,
+      "loss": 1.8532,
+      "step": 9489
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6183425126155839,
+      "learning_rate": 8.832423311230112e-06,
+      "loss": 2.1082,
+      "step": 9490
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6694578889361724,
+      "learning_rate": 8.82765919294504e-06,
+      "loss": 1.8578,
+      "step": 9491
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7091615420809887,
+      "learning_rate": 8.82289608436564e-06,
+      "loss": 1.8628,
+      "step": 9492
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6218040623554549,
+      "learning_rate": 8.818133985789295e-06,
+      "loss": 1.8747,
+      "step": 9493
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6245583656484505,
+      "learning_rate": 8.813372897513325e-06,
+      "loss": 1.8746,
+      "step": 9494
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.729931058114289,
+      "learning_rate": 8.808612819834989e-06,
+      "loss": 2.0604,
+      "step": 9495
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6182482245011539,
+      "learning_rate": 8.803853753051452e-06,
+      "loss": 1.8452,
+      "step": 9496
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6247423435013503,
+      "learning_rate": 8.79909569745987e-06,
+      "loss": 1.8744,
+      "step": 9497
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.795580986524546,
+      "learning_rate": 8.794338653357297e-06,
+      "loss": 2.0337,
+      "step": 9498
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5750633129359688,
+      "learning_rate": 8.78958262104074e-06,
+      "loss": 1.9174,
+      "step": 9499
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.604489153943089,
+      "learning_rate": 8.784827600807139e-06,
+      "loss": 1.8853,
+      "step": 9500
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6949752365476995,
+      "learning_rate": 8.78007359295335e-06,
+      "loss": 1.874,
+      "step": 9501
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6135693159496811,
+      "learning_rate": 8.775320597776206e-06,
+      "loss": 1.8847,
+      "step": 9502
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6462318566111473,
+      "learning_rate": 8.770568615572461e-06,
+      "loss": 2.1086,
+      "step": 9503
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5733103533254105,
+      "learning_rate": 8.765817646638769e-06,
+      "loss": 1.8917,
+      "step": 9504
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5564097713988024,
+      "learning_rate": 8.761067691271785e-06,
+      "loss": 1.958,
+      "step": 9505
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6457923816094181,
+      "learning_rate": 8.756318749768061e-06,
+      "loss": 1.8115,
+      "step": 9506
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5834141388875931,
+      "learning_rate": 8.751570822424074e-06,
+      "loss": 2.0751,
+      "step": 9507
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6269065322068194,
+      "learning_rate": 8.746823909536286e-06,
+      "loss": 1.8994,
+      "step": 9508
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5626929307055278,
+      "learning_rate": 8.742078011401045e-06,
+      "loss": 1.8665,
+      "step": 9509
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5966242219412808,
+      "learning_rate": 8.737333128314653e-06,
+      "loss": 2.0732,
+      "step": 9510
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5810034332604023,
+      "learning_rate": 8.732589260573379e-06,
+      "loss": 1.9736,
+      "step": 9511
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5707036521543274,
+      "learning_rate": 8.727846408473377e-06,
+      "loss": 1.8542,
+      "step": 9512
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5879599653983297,
+      "learning_rate": 8.723104572310773e-06,
+      "loss": 1.878,
+      "step": 9513
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6057033813881587,
+      "learning_rate": 8.718363752381616e-06,
+      "loss": 1.8696,
+      "step": 9514
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5509816841704259,
+      "learning_rate": 8.713623948981892e-06,
+      "loss": 2.0485,
+      "step": 9515
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5338385562154905,
+      "learning_rate": 8.708885162407542e-06,
+      "loss": 1.8676,
+      "step": 9516
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5407978599157268,
+      "learning_rate": 8.704147392954412e-06,
+      "loss": 1.9647,
+      "step": 9517
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5859708347920605,
+      "learning_rate": 8.699410640918295e-06,
+      "loss": 1.8861,
+      "step": 9518
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5582291647799464,
+      "learning_rate": 8.694674906594954e-06,
+      "loss": 2.0295,
+      "step": 9519
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5877479785782478,
+      "learning_rate": 8.689940190280035e-06,
+      "loss": 1.8888,
+      "step": 9520
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6442823406633241,
+      "learning_rate": 8.685206492269155e-06,
+      "loss": 1.8627,
+      "step": 9521
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6453134998717286,
+      "learning_rate": 8.680473812857853e-06,
+      "loss": 2.0367,
+      "step": 9522
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5119894099689503,
+      "learning_rate": 8.675742152341615e-06,
+      "loss": 1.9228,
+      "step": 9523
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5429165080201868,
+      "learning_rate": 8.671011511015855e-06,
+      "loss": 1.8102,
+      "step": 9524
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6342332387544026,
+      "learning_rate": 8.666281889175928e-06,
+      "loss": 1.862,
+      "step": 9525
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5509788989570872,
+      "learning_rate": 8.661553287117125e-06,
+      "loss": 1.8252,
+      "step": 9526
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5627231021312027,
+      "learning_rate": 8.656825705134666e-06,
+      "loss": 2.0805,
+      "step": 9527
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5525287267445802,
+      "learning_rate": 8.652099143523721e-06,
+      "loss": 1.8514,
+      "step": 9528
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.55115614443392,
+      "learning_rate": 8.647373602579383e-06,
+      "loss": 1.9534,
+      "step": 9529
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5957817560656531,
+      "learning_rate": 8.642649082596693e-06,
+      "loss": 1.9548,
+      "step": 9530
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6657659770938509,
+      "learning_rate": 8.637925583870614e-06,
+      "loss": 2.0309,
+      "step": 9531
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5866233619020875,
+      "learning_rate": 8.633203106696062e-06,
+      "loss": 1.8447,
+      "step": 9532
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5808934747309141,
+      "learning_rate": 8.628481651367876e-06,
+      "loss": 1.8767,
+      "step": 9533
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.535185379678177,
+      "learning_rate": 8.623761218180834e-06,
+      "loss": 1.8669,
+      "step": 9534
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6063668267229426,
+      "learning_rate": 8.619041807429656e-06,
+      "loss": 2.07,
+      "step": 9535
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5902608817947048,
+      "learning_rate": 8.614323419408993e-06,
+      "loss": 1.9499,
+      "step": 9536
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5687856952619362,
+      "learning_rate": 8.60960605441343e-06,
+      "loss": 1.849,
+      "step": 9537
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5695324132287162,
+      "learning_rate": 8.604889712737496e-06,
+      "loss": 1.8757,
+      "step": 9538
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5526098141112786,
+      "learning_rate": 8.600174394675652e-06,
+      "loss": 2.0706,
+      "step": 9539
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5786330331838251,
+      "learning_rate": 8.59546010052229e-06,
+      "loss": 1.8599,
+      "step": 9540
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5602785682551447,
+      "learning_rate": 8.590746830571748e-06,
+      "loss": 1.8586,
+      "step": 9541
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5985409438282746,
+      "learning_rate": 8.5860345851183e-06,
+      "loss": 1.9022,
+      "step": 9542
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.592201552467059,
+      "learning_rate": 8.581323364456125e-06,
+      "loss": 2.0641,
+      "step": 9543
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5578314571771809,
+      "learning_rate": 8.576613168879392e-06,
+      "loss": 1.8497,
+      "step": 9544
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6244974886306275,
+      "learning_rate": 8.571903998682174e-06,
+      "loss": 1.871,
+      "step": 9545
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5907029026001209,
+      "learning_rate": 8.567195854158464e-06,
+      "loss": 1.8929,
+      "step": 9546
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.623254435928751,
+      "learning_rate": 8.56248873560224e-06,
+      "loss": 2.0496,
+      "step": 9547
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5260493697608941,
+      "learning_rate": 8.557782643307353e-06,
+      "loss": 1.9155,
+      "step": 9548
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.57905602423377,
+      "learning_rate": 8.55307757756765e-06,
+      "loss": 1.8251,
+      "step": 9549
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5434398927915,
+      "learning_rate": 8.54837353867689e-06,
+      "loss": 1.887,
+      "step": 9550
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5880842494468408,
+      "learning_rate": 8.543670526928738e-06,
+      "loss": 2.0498,
+      "step": 9551
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5738337689230462,
+      "learning_rate": 8.538968542616846e-06,
+      "loss": 1.8387,
+      "step": 9552
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5543706091646878,
+      "learning_rate": 8.53426758603478e-06,
+      "loss": 1.9034,
+      "step": 9553
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.545187432898989,
+      "learning_rate": 8.529567657476017e-06,
+      "loss": 1.9263,
+      "step": 9554
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5707560369565038,
+      "learning_rate": 8.524868757234022e-06,
+      "loss": 2.1097,
+      "step": 9555
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.58160244042883,
+      "learning_rate": 8.520170885602143e-06,
+      "loss": 1.8936,
+      "step": 9556
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.585963558994109,
+      "learning_rate": 8.515474042873691e-06,
+      "loss": 1.8756,
+      "step": 9557
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5746945313357992,
+      "learning_rate": 8.51077822934193e-06,
+      "loss": 1.8649,
+      "step": 9558
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6145953177314585,
+      "learning_rate": 8.506083445300015e-06,
+      "loss": 2.0292,
+      "step": 9559
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6025504246175644,
+      "learning_rate": 8.50138969104107e-06,
+      "loss": 1.89,
+      "step": 9560
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5796718589293484,
+      "learning_rate": 8.496696966858145e-06,
+      "loss": 2.0067,
+      "step": 9561
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6362908349276662,
+      "learning_rate": 8.492005273044214e-06,
+      "loss": 1.8651,
+      "step": 9562
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5986553542263581,
+      "learning_rate": 8.48731460989223e-06,
+      "loss": 2.0692,
+      "step": 9563
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6173634114156699,
+      "learning_rate": 8.482624977695023e-06,
+      "loss": 1.8443,
+      "step": 9564
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5768448676413439,
+      "learning_rate": 8.477936376745382e-06,
+      "loss": 1.8563,
+      "step": 9565
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5773784051898269,
+      "learning_rate": 8.473248807336067e-06,
+      "loss": 1.8736,
+      "step": 9566
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5840541693790218,
+      "learning_rate": 8.468562269759714e-06,
+      "loss": 2.1422,
+      "step": 9567
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5919013030193284,
+      "learning_rate": 8.463876764308928e-06,
+      "loss": 1.8795,
+      "step": 9568
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5927752171363201,
+      "learning_rate": 8.459192291276252e-06,
+      "loss": 1.8589,
+      "step": 9569
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6030607086938009,
+      "learning_rate": 8.454508850954151e-06,
+      "loss": 1.8474,
+      "step": 9570
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6274000771088933,
+      "learning_rate": 8.449826443635032e-06,
+      "loss": 2.0389,
+      "step": 9571
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5826636988271239,
+      "learning_rate": 8.445145069611237e-06,
+      "loss": 1.8763,
+      "step": 9572
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5066516616232706,
+      "learning_rate": 8.440464729175048e-06,
+      "loss": 1.9179,
+      "step": 9573
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6007634397741938,
+      "learning_rate": 8.435785422618672e-06,
+      "loss": 1.8771,
+      "step": 9574
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5789947552220854,
+      "learning_rate": 8.43110715023426e-06,
+      "loss": 2.0213,
+      "step": 9575
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5483313299968551,
+      "learning_rate": 8.426429912313896e-06,
+      "loss": 1.9292,
+      "step": 9576
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5629356730812739,
+      "learning_rate": 8.421753709149596e-06,
+      "loss": 1.8348,
+      "step": 9577
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5942744150441326,
+      "learning_rate": 8.417078541033319e-06,
+      "loss": 1.8496,
+      "step": 9578
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5961591159253581,
+      "learning_rate": 8.412404408256955e-06,
+      "loss": 2.0767,
+      "step": 9579
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5786137425216643,
+      "learning_rate": 8.407731311112324e-06,
+      "loss": 1.8441,
+      "step": 9580
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6185220628752255,
+      "learning_rate": 8.40305924989119e-06,
+      "loss": 1.9175,
+      "step": 9581
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5593484465853727,
+      "learning_rate": 8.398388224885251e-06,
+      "loss": 1.8667,
+      "step": 9582
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5969887176216477,
+      "learning_rate": 8.393718236386136e-06,
+      "loss": 2.0689,
+      "step": 9583
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5382555744946411,
+      "learning_rate": 8.38904928468541e-06,
+      "loss": 1.875,
+      "step": 9584
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.607727273528077,
+      "learning_rate": 8.384381370074578e-06,
+      "loss": 1.9763,
+      "step": 9585
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6004819773619966,
+      "learning_rate": 8.379714492845076e-06,
+      "loss": 1.8455,
+      "step": 9586
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5606449985766947,
+      "learning_rate": 8.375048653288275e-06,
+      "loss": 2.0609,
+      "step": 9587
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5899888914027556,
+      "learning_rate": 8.370383851695485e-06,
+      "loss": 1.9002,
+      "step": 9588
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5586240303395801,
+      "learning_rate": 8.365720088357956e-06,
+      "loss": 1.8756,
+      "step": 9589
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5565193793639786,
+      "learning_rate": 8.361057363566837e-06,
+      "loss": 1.8502,
+      "step": 9590
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6007831541356444,
+      "learning_rate": 8.356395677613274e-06,
+      "loss": 2.0462,
+      "step": 9591
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.531465543705732,
+      "learning_rate": 8.35173503078831e-06,
+      "loss": 1.9586,
+      "step": 9592
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5649096710271295,
+      "learning_rate": 8.347075423382902e-06,
+      "loss": 1.8504,
+      "step": 9593
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5834646403933313,
+      "learning_rate": 8.342416855688003e-06,
+      "loss": 1.8553,
+      "step": 9594
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.560449618071684,
+      "learning_rate": 8.337759327994447e-06,
+      "loss": 2.0834,
+      "step": 9595
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.570883296007682,
+      "learning_rate": 8.333102840593015e-06,
+      "loss": 1.8468,
+      "step": 9596
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6417575501851226,
+      "learning_rate": 8.328447393774458e-06,
+      "loss": 1.8318,
+      "step": 9597
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6370372213582371,
+      "learning_rate": 8.323792987829404e-06,
+      "loss": 1.9754,
+      "step": 9598
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5932703061201452,
+      "learning_rate": 8.319139623048469e-06,
+      "loss": 2.0881,
+      "step": 9599
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6287596935434779,
+      "learning_rate": 8.314487299722182e-06,
+      "loss": 1.8761,
+      "step": 9600
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5686017983670711,
+      "learning_rate": 8.309836018140982e-06,
+      "loss": 1.8558,
+      "step": 9601
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5786967322864862,
+      "learning_rate": 8.305185778595301e-06,
+      "loss": 1.873,
+      "step": 9602
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5813350082562291,
+      "learning_rate": 8.300536581375447e-06,
+      "loss": 2.0983,
+      "step": 9603
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5203254756064826,
+      "learning_rate": 8.295888426771693e-06,
+      "loss": 1.9227,
+      "step": 9604
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5848132054829336,
+      "learning_rate": 8.291241315074257e-06,
+      "loss": 1.8743,
+      "step": 9605
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5888031505444721,
+      "learning_rate": 8.286595246573262e-06,
+      "loss": 1.8805,
+      "step": 9606
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5994527790881932,
+      "learning_rate": 8.281950221558779e-06,
+      "loss": 2.0444,
+      "step": 9607
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5566278808405757,
+      "learning_rate": 8.277306240320838e-06,
+      "loss": 1.8583,
+      "step": 9608
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6227852225968724,
+      "learning_rate": 8.272663303149358e-06,
+      "loss": 1.8875,
+      "step": 9609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5633665756810162,
+      "learning_rate": 8.268021410334225e-06,
+      "loss": 1.8765,
+      "step": 9610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5652108792174515,
+      "learning_rate": 8.263380562165254e-06,
+      "loss": 2.0175,
+      "step": 9611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5492196933046195,
+      "learning_rate": 8.258740758932182e-06,
+      "loss": 1.8596,
+      "step": 9612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6070523775592923,
+      "learning_rate": 8.254102000924716e-06,
+      "loss": 1.9048,
+      "step": 9613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5796414910253367,
+      "learning_rate": 8.249464288432446e-06,
+      "loss": 1.873,
+      "step": 9614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.570817504933111,
+      "learning_rate": 8.244827621744936e-06,
+      "loss": 2.0529,
+      "step": 9615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5381811426702727,
+      "learning_rate": 8.240192001151668e-06,
+      "loss": 1.979,
+      "step": 9616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.595967925907073,
+      "learning_rate": 8.235557426942065e-06,
+      "loss": 1.831,
+      "step": 9617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5511386037989031,
+      "learning_rate": 8.23092389940548e-06,
+      "loss": 1.8653,
+      "step": 9618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5641650130504673,
+      "learning_rate": 8.226291418831206e-06,
+      "loss": 2.0384,
+      "step": 9619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5852091556456945,
+      "learning_rate": 8.221659985508468e-06,
+      "loss": 1.8671,
+      "step": 9620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5470918243119084,
+      "learning_rate": 8.217029599726422e-06,
+      "loss": 1.8894,
+      "step": 9621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.594198270327469,
+      "learning_rate": 8.212400261774166e-06,
+      "loss": 1.8649,
+      "step": 9622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.536672827691908,
+      "learning_rate": 8.207771971940729e-06,
+      "loss": 2.1247,
+      "step": 9623
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5904438585017769,
+      "learning_rate": 8.203144730515066e-06,
+      "loss": 1.947,
+      "step": 9624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5980538145827599,
+      "learning_rate": 8.198518537786087e-06,
+      "loss": 1.8866,
+      "step": 9625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5838714026412606,
+      "learning_rate": 8.193893394042614e-06,
+      "loss": 1.8535,
+      "step": 9626
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5414048183201188,
+      "learning_rate": 8.189269299573419e-06,
+      "loss": 2.0958,
+      "step": 9627
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6833080128211375,
+      "learning_rate": 8.18464625466721e-06,
+      "loss": 1.8503,
+      "step": 9628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4973246209478797,
+      "learning_rate": 8.180024259612598e-06,
+      "loss": 1.9884,
+      "step": 9629
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5896880089665866,
+      "learning_rate": 8.175403314698178e-06,
+      "loss": 1.8781,
+      "step": 9630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5791239060325994,
+      "learning_rate": 8.170783420212447e-06,
+      "loss": 2.0162,
+      "step": 9631
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5575897269130325,
+      "learning_rate": 8.16616457644384e-06,
+      "loss": 1.8666,
+      "step": 9632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5749575684852171,
+      "learning_rate": 8.161546783680738e-06,
+      "loss": 1.8645,
+      "step": 9633
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5654675519974227,
+      "learning_rate": 8.156930042211442e-06,
+      "loss": 1.8981,
+      "step": 9634
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5492550071909558,
+      "learning_rate": 8.152314352324194e-06,
+      "loss": 1.9606,
+      "step": 9635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5701287154088475,
+      "learning_rate": 8.147699714307186e-06,
+      "loss": 2.1076,
+      "step": 9636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5438180106521757,
+      "learning_rate": 8.143086128448496e-06,
+      "loss": 1.8455,
+      "step": 9637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5930548318201123,
+      "learning_rate": 8.1384735950362e-06,
+      "loss": 1.9113,
+      "step": 9638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5586758359959807,
+      "learning_rate": 8.133862114358274e-06,
+      "loss": 2.0901,
+      "step": 9639
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5688910209225688,
+      "learning_rate": 8.129251686702605e-06,
+      "loss": 1.8825,
+      "step": 9640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5335892056233568,
+      "learning_rate": 8.12464231235707e-06,
+      "loss": 1.9273,
+      "step": 9641
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.571242473452962,
+      "learning_rate": 8.120033991609449e-06,
+      "loss": 1.873,
+      "step": 9642
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5449500026848987,
+      "learning_rate": 8.115426724747436e-06,
+      "loss": 2.0696,
+      "step": 9643
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6168661471637013,
+      "learning_rate": 8.110820512058708e-06,
+      "loss": 1.8567,
+      "step": 9644
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5982839104189641,
+      "learning_rate": 8.106215353830825e-06,
+      "loss": 1.8543,
+      "step": 9645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5709563867642586,
+      "learning_rate": 8.101611250351327e-06,
+      "loss": 1.8973,
+      "step": 9646
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5805845267794946,
+      "learning_rate": 8.097008201907668e-06,
+      "loss": 2.0924,
+      "step": 9647
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6675423746985499,
+      "learning_rate": 8.092406208787206e-06,
+      "loss": 1.9715,
+      "step": 9648
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5567842990883047,
+      "learning_rate": 8.087805271277305e-06,
+      "loss": 1.8504,
+      "step": 9649
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5616935367192633,
+      "learning_rate": 8.083205389665186e-06,
+      "loss": 1.9171,
+      "step": 9650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6542306493533755,
+      "learning_rate": 8.078606564238042e-06,
+      "loss": 1.9939,
+      "step": 9651
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5811367268022334,
+      "learning_rate": 8.07400879528302e-06,
+      "loss": 1.8694,
+      "step": 9652
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5702946889745036,
+      "learning_rate": 8.069412083087158e-06,
+      "loss": 1.8269,
+      "step": 9653
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5782426372043452,
+      "learning_rate": 8.064816427937442e-06,
+      "loss": 1.9597,
+      "step": 9654
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5688237485072568,
+      "learning_rate": 8.060221830120823e-06,
+      "loss": 2.0419,
+      "step": 9655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.601790059969058,
+      "learning_rate": 8.05562828992414e-06,
+      "loss": 1.8492,
+      "step": 9656
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6176574832801204,
+      "learning_rate": 8.051035807634191e-06,
+      "loss": 1.8154,
+      "step": 9657
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5636739377178016,
+      "learning_rate": 8.046444383537704e-06,
+      "loss": 1.8865,
+      "step": 9658
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6104109326153345,
+      "learning_rate": 8.041854017921341e-06,
+      "loss": 2.083,
+      "step": 9659
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5989550837830616,
+      "learning_rate": 8.0372647110717e-06,
+      "loss": 1.9429,
+      "step": 9660
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.616621771430774,
+      "learning_rate": 8.032676463275304e-06,
+      "loss": 1.8406,
+      "step": 9661
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5978121767910628,
+      "learning_rate": 8.028089274818624e-06,
+      "loss": 1.8407,
+      "step": 9662
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5813324048841958,
+      "learning_rate": 8.023503145988048e-06,
+      "loss": 2.0978,
+      "step": 9663
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5794615904976046,
+      "learning_rate": 8.018918077069915e-06,
+      "loss": 1.8452,
+      "step": 9664
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5955023272413151,
+      "learning_rate": 8.014334068350487e-06,
+      "loss": 1.833,
+      "step": 9665
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5989621659558192,
+      "learning_rate": 8.00975112011596e-06,
+      "loss": 1.9333,
+      "step": 9666
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6141144551139448,
+      "learning_rate": 8.00516923265247e-06,
+      "loss": 1.8854,
+      "step": 9667
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6078873609887718,
+      "learning_rate": 8.000588406246081e-06,
+      "loss": 2.0883,
+      "step": 9668
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6111136433575022,
+      "learning_rate": 7.996008641182792e-06,
+      "loss": 1.8803,
+      "step": 9669
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5755441578275446,
+      "learning_rate": 7.99142993774854e-06,
+      "loss": 1.9041,
+      "step": 9670
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5787381342875569,
+      "learning_rate": 7.986852296229185e-06,
+      "loss": 2.0641,
+      "step": 9671
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5539860368666587,
+      "learning_rate": 7.98227571691054e-06,
+      "loss": 1.9362,
+      "step": 9672
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5889241734194053,
+      "learning_rate": 7.977700200078328e-06,
+      "loss": 1.8934,
+      "step": 9673
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5966601194218284,
+      "learning_rate": 7.973125746018222e-06,
+      "loss": 1.923,
+      "step": 9674
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6085391258984884,
+      "learning_rate": 7.968552355015832e-06,
+      "loss": 2.048,
+      "step": 9675
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6107953431120102,
+      "learning_rate": 7.963980027356671e-06,
+      "loss": 1.9115,
+      "step": 9676
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5722037609499528,
+      "learning_rate": 7.959408763326228e-06,
+      "loss": 1.825,
+      "step": 9677
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5698494172093421,
+      "learning_rate": 7.954838563209902e-06,
+      "loss": 1.855,
+      "step": 9678
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5817429264241074,
+      "learning_rate": 7.950269427293029e-06,
+      "loss": 1.9298,
+      "step": 9679
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6052188128890056,
+      "learning_rate": 7.945701355860876e-06,
+      "loss": 2.0255,
+      "step": 9680
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5814743663973506,
+      "learning_rate": 7.941134349198647e-06,
+      "loss": 1.8807,
+      "step": 9681
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5770736638403569,
+      "learning_rate": 7.936568407591482e-06,
+      "loss": 1.8497,
+      "step": 9682
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5300683017713927,
+      "learning_rate": 7.932003531324458e-06,
+      "loss": 2.0566,
+      "step": 9683
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7121966790739129,
+      "learning_rate": 7.927439720682555e-06,
+      "loss": 1.8543,
+      "step": 9684
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6295933228326592,
+      "learning_rate": 7.922876975950733e-06,
+      "loss": 1.9619,
+      "step": 9685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5867667112963422,
+      "learning_rate": 7.918315297413864e-06,
+      "loss": 1.8397,
+      "step": 9686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5956495850258585,
+      "learning_rate": 7.913754685356727e-06,
+      "loss": 1.8173,
+      "step": 9687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6122745384863749,
+      "learning_rate": 7.909195140064088e-06,
+      "loss": 2.0553,
+      "step": 9688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5679854273857808,
+      "learning_rate": 7.904636661820614e-06,
+      "loss": 1.8728,
+      "step": 9689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7021718818558582,
+      "learning_rate": 7.900079250910886e-06,
+      "loss": 1.9154,
+      "step": 9690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5972613644143542,
+      "learning_rate": 7.895522907619474e-06,
+      "loss": 1.8983,
+      "step": 9691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6005479026744721,
+      "learning_rate": 7.890967632230827e-06,
+      "loss": 2.0285,
+      "step": 9692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6709578902539571,
+      "learning_rate": 7.88641342502935e-06,
+      "loss": 1.8893,
+      "step": 9693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6394719621218592,
+      "learning_rate": 7.8818602862994e-06,
+      "loss": 1.8235,
+      "step": 9694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5776090049291386,
+      "learning_rate": 7.877308216325222e-06,
+      "loss": 2.064,
+      "step": 9695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6110683813273184,
+      "learning_rate": 7.872757215391047e-06,
+      "loss": 1.8333,
+      "step": 9696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6150996879712762,
+      "learning_rate": 7.868207283780992e-06,
+      "loss": 1.9594,
+      "step": 9697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5979223829977564,
+      "learning_rate": 7.86365842177913e-06,
+      "loss": 1.9047,
+      "step": 9698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6424338696092472,
+      "learning_rate": 7.859110629669487e-06,
+      "loss": 1.8873,
+      "step": 9699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6229086406716716,
+      "learning_rate": 7.854563907735971e-06,
+      "loss": 2.0732,
+      "step": 9700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5835406958390547,
+      "learning_rate": 7.850018256262461e-06,
+      "loss": 1.8864,
+      "step": 9701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5451495543382027,
+      "learning_rate": 7.84547367553278e-06,
+      "loss": 1.8416,
+      "step": 9702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5891973135618125,
+      "learning_rate": 7.840930165830644e-06,
+      "loss": 1.9485,
+      "step": 9703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6133954450114006,
+      "learning_rate": 7.836387727439728e-06,
+      "loss": 2.0584,
+      "step": 9704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5971132094306681,
+      "learning_rate": 7.831846360643635e-06,
+      "loss": 1.8663,
+      "step": 9705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.559862026843912,
+      "learning_rate": 7.827306065725903e-06,
+      "loss": 1.8273,
+      "step": 9706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6073417034894042,
+      "learning_rate": 7.82276684297e-06,
+      "loss": 2.1045,
+      "step": 9707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5859007018997301,
+      "learning_rate": 7.81822869265933e-06,
+      "loss": 1.873,
+      "step": 9708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6463676277298301,
+      "learning_rate": 7.813691615077227e-06,
+      "loss": 1.8535,
+      "step": 9709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5455862969665533,
+      "learning_rate": 7.80915561050696e-06,
+      "loss": 1.8422,
+      "step": 9710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5449465566255458,
+      "learning_rate": 7.804620679231727e-06,
+      "loss": 1.8465,
+      "step": 9711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6612260854950597,
+      "learning_rate": 7.800086821534669e-06,
+      "loss": 2.0853,
+      "step": 9712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5928490079703577,
+      "learning_rate": 7.79555403769885e-06,
+      "loss": 1.8618,
+      "step": 9713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6218533343621533,
+      "learning_rate": 7.791022328007267e-06,
+      "loss": 1.8742,
+      "step": 9714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.653408542850214,
+      "learning_rate": 7.786491692742856e-06,
+      "loss": 2.049,
+      "step": 9715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5662238959546546,
+      "learning_rate": 7.781962132188486e-06,
+      "loss": 1.9141,
+      "step": 9716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6121501601449149,
+      "learning_rate": 7.777433646626952e-06,
+      "loss": 1.8712,
+      "step": 9717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6222948474183657,
+      "learning_rate": 7.772906236340987e-06,
+      "loss": 1.8277,
+      "step": 9718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6139331526219693,
+      "learning_rate": 7.76837990161326e-06,
+      "loss": 1.8301,
+      "step": 9719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6163817830151918,
+      "learning_rate": 7.763854642726362e-06,
+      "loss": 2.048,
+      "step": 9720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6399370941854875,
+      "learning_rate": 7.759330459962827e-06,
+      "loss": 1.8919,
+      "step": 9721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5121582182416242,
+      "learning_rate": 7.754807353605117e-06,
+      "loss": 1.9126,
+      "step": 9722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.545924164595504,
+      "learning_rate": 7.75028532393563e-06,
+      "loss": 1.8336,
+      "step": 9723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6061016226510417,
+      "learning_rate": 7.745764371236693e-06,
+      "loss": 2.0902,
+      "step": 9724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5433844277571601,
+      "learning_rate": 7.741244495790578e-06,
+      "loss": 1.8588,
+      "step": 9725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6546356627910671,
+      "learning_rate": 7.736725697879454e-06,
+      "loss": 1.8949,
+      "step": 9726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5495456697213985,
+      "learning_rate": 7.732207977785472e-06,
+      "loss": 2.0562,
+      "step": 9727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5444023375842085,
+      "learning_rate": 7.727691335790682e-06,
+      "loss": 1.9151,
+      "step": 9728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5837129439414848,
+      "learning_rate": 7.72317577217708e-06,
+      "loss": 1.8481,
+      "step": 9729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5839048951679258,
+      "learning_rate": 7.718661287226595e-06,
+      "loss": 1.8934,
+      "step": 9730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6370023380745496,
+      "learning_rate": 7.714147881221065e-06,
+      "loss": 1.8901,
+      "step": 9731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5782269737825148,
+      "learning_rate": 7.709635554442304e-06,
+      "loss": 2.0652,
+      "step": 9732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.579088363364256,
+      "learning_rate": 7.705124307172032e-06,
+      "loss": 1.8941,
+      "step": 9733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5307138206736076,
+      "learning_rate": 7.700614139691883e-06,
+      "loss": 1.9473,
+      "step": 9734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5906331102066462,
+      "learning_rate": 7.69610505228347e-06,
+      "loss": 1.8922,
+      "step": 9735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5787807830362273,
+      "learning_rate": 7.691597045228313e-06,
+      "loss": 2.0765,
+      "step": 9736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.572947877888531,
+      "learning_rate": 7.687090118807839e-06,
+      "loss": 1.8036,
+      "step": 9737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5874298471430289,
+      "learning_rate": 7.682584273303469e-06,
+      "loss": 1.8566,
+      "step": 9738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6240506201421289,
+      "learning_rate": 7.678079508996497e-06,
+      "loss": 2.1029,
+      "step": 9739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6075277251535941,
+      "learning_rate": 7.673575826168175e-06,
+      "loss": 1.8466,
+      "step": 9740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.545044315614853,
+      "learning_rate": 7.669073225099707e-06,
+      "loss": 1.9593,
+      "step": 9741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5998179639304626,
+      "learning_rate": 7.664571706072179e-06,
+      "loss": 1.8515,
+      "step": 9742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5820251803700716,
+      "learning_rate": 7.660071269366669e-06,
+      "loss": 1.8923,
+      "step": 9743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6166619720779593,
+      "learning_rate": 7.65557191526414e-06,
+      "loss": 2.0312,
+      "step": 9744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6098554399864821,
+      "learning_rate": 7.651073644045498e-06,
+      "loss": 1.8669,
+      "step": 9745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5854523204816042,
+      "learning_rate": 7.646576455991617e-06,
+      "loss": 1.8429,
+      "step": 9746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5493494448527086,
+      "learning_rate": 7.642080351383249e-06,
+      "loss": 1.9263,
+      "step": 9747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5638602079774317,
+      "learning_rate": 7.637585330501108e-06,
+      "loss": 2.0687,
+      "step": 9748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5726613699603732,
+      "learning_rate": 7.633091393625852e-06,
+      "loss": 1.8277,
+      "step": 9749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5881831992311246,
+      "learning_rate": 7.62859854103804e-06,
+      "loss": 1.8824,
+      "step": 9750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5563221688713246,
+      "learning_rate": 7.624106773018183e-06,
+      "loss": 1.8518,
+      "step": 9751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5550242258484831,
+      "learning_rate": 7.619616089846723e-06,
+      "loss": 2.0594,
+      "step": 9752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5459083150519495,
+      "learning_rate": 7.615126491804034e-06,
+      "loss": 1.9124,
+      "step": 9753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5804658829412,
+      "learning_rate": 7.610637979170412e-06,
+      "loss": 1.8372,
+      "step": 9754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6008353982092726,
+      "learning_rate": 7.606150552226101e-06,
+      "loss": 1.865,
+      "step": 9755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5501898660385509,
+      "learning_rate": 7.601664211251267e-06,
+      "loss": 2.0516,
+      "step": 9756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5751352264385308,
+      "learning_rate": 7.59717895652601e-06,
+      "loss": 1.8731,
+      "step": 9757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6027609687475649,
+      "learning_rate": 7.592694788330363e-06,
+      "loss": 1.9024,
+      "step": 9758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5506082735635925,
+      "learning_rate": 7.588211706944293e-06,
+      "loss": 1.95,
+      "step": 9759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5600949765927711,
+      "learning_rate": 7.5837297126476915e-06,
+      "loss": 2.0792,
+      "step": 9760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6440422346153363,
+      "learning_rate": 7.579248805720396e-06,
+      "loss": 1.8502,
+      "step": 9761
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.610829375382557,
+      "learning_rate": 7.574768986442163e-06,
+      "loss": 1.8533,
+      "step": 9762
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5897168343803262,
+      "learning_rate": 7.570290255092685e-06,
+      "loss": 1.8644,
+      "step": 9763
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5772752792068511,
+      "learning_rate": 7.56581261195159e-06,
+      "loss": 2.1087,
+      "step": 9764
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5719931050046899,
+      "learning_rate": 7.561336057298435e-06,
+      "loss": 1.946,
+      "step": 9765
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5709006687761398,
+      "learning_rate": 7.5568605914127095e-06,
+      "loss": 1.9239,
+      "step": 9766
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5688497969778306,
+      "learning_rate": 7.552386214573837e-06,
+      "loss": 1.8844,
+      "step": 9767
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5958779531833449,
+      "learning_rate": 7.547912927061168e-06,
+      "loss": 2.0398,
+      "step": 9768
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6459584054559888,
+      "learning_rate": 7.543440729153992e-06,
+      "loss": 1.9062,
+      "step": 9769
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6115568693029245,
+      "learning_rate": 7.5389696211315255e-06,
+      "loss": 1.9047,
+      "step": 9770
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6244667792077003,
+      "learning_rate": 7.534499603272912e-06,
+      "loss": 1.8798,
+      "step": 9771
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.663040378961429,
+      "learning_rate": 7.530030675857253e-06,
+      "loss": 2.0619,
+      "step": 9772
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5577212686680397,
+      "learning_rate": 7.525562839163527e-06,
+      "loss": 1.8688,
+      "step": 9773
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5755058030475402,
+      "learning_rate": 7.521096093470709e-06,
+      "loss": 1.8774,
+      "step": 9774
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6109098626842512,
+      "learning_rate": 7.5166304390576645e-06,
+      "loss": 1.904,
+      "step": 9775
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5884879109454476,
+      "learning_rate": 7.5121658762032084e-06,
+      "loss": 2.073,
+      "step": 9776
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5743116729118976,
+      "learning_rate": 7.507702405186084e-06,
+      "loss": 1.9015,
+      "step": 9777
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5630612218224895,
+      "learning_rate": 7.503240026284947e-06,
+      "loss": 1.9045,
+      "step": 9778
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6215457547191732,
+      "learning_rate": 7.498778739778417e-06,
+      "loss": 1.8948,
+      "step": 9779
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5752771312826238,
+      "learning_rate": 7.494318545945037e-06,
+      "loss": 2.0286,
+      "step": 9780
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.597033133231822,
+      "learning_rate": 7.489859445063249e-06,
+      "loss": 1.8679,
+      "step": 9781
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5991020672115519,
+      "learning_rate": 7.48540143741148e-06,
+      "loss": 1.845,
+      "step": 9782
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6248990381958771,
+      "learning_rate": 7.480944523268055e-06,
+      "loss": 1.8684,
+      "step": 9783
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5210730829590403,
+      "learning_rate": 7.476488702911219e-06,
+      "loss": 2.1165,
+      "step": 9784
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6124024595590598,
+      "learning_rate": 7.472033976619197e-06,
+      "loss": 1.8835,
+      "step": 9785
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5606094281203028,
+      "learning_rate": 7.467580344670095e-06,
+      "loss": 1.8673,
+      "step": 9786
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5868345203915163,
+      "learning_rate": 7.463127807341966e-06,
+      "loss": 1.8819,
+      "step": 9787
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5882130574798546,
+      "learning_rate": 7.458676364912828e-06,
+      "loss": 2.0731,
+      "step": 9788
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6480282677175998,
+      "learning_rate": 7.4542260176605785e-06,
+      "loss": 1.8775,
+      "step": 9789
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5934217682097165,
+      "learning_rate": 7.449776765863079e-06,
+      "loss": 1.9618,
+      "step": 9790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6015078183139233,
+      "learning_rate": 7.445328609798113e-06,
+      "loss": 1.9043,
+      "step": 9791
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5997422970979445,
+      "learning_rate": 7.440881549743392e-06,
+      "loss": 2.0682,
+      "step": 9792
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.556133555298735,
+      "learning_rate": 7.436435585976587e-06,
+      "loss": 1.8984,
+      "step": 9793
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7484329735057513,
+      "learning_rate": 7.43199071877525e-06,
+      "loss": 1.8942,
+      "step": 9794
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5860758986796064,
+      "learning_rate": 7.427546948416897e-06,
+      "loss": 1.9382,
+      "step": 9795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.630075014559968,
+      "learning_rate": 7.423104275178994e-06,
+      "loss": 2.1123,
+      "step": 9796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6341688469944133,
+      "learning_rate": 7.41866269933889e-06,
+      "loss": 1.8937,
+      "step": 9797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5923632793244357,
+      "learning_rate": 7.414222221173903e-06,
+      "loss": 1.9039,
+      "step": 9798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5676605150522307,
+      "learning_rate": 7.409782840961263e-06,
+      "loss": 1.8915,
+      "step": 9799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6297880850959686,
+      "learning_rate": 7.4053445589781445e-06,
+      "loss": 2.0539,
+      "step": 9800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6033852085852077,
+      "learning_rate": 7.400907375501645e-06,
+      "loss": 1.8696,
+      "step": 9801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.56362465181915,
+      "learning_rate": 7.396471290808799e-06,
+      "loss": 1.8617,
+      "step": 9802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5312819220950944,
+      "learning_rate": 7.392036305176564e-06,
+      "loss": 1.8899,
+      "step": 9803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5669515056799804,
+      "learning_rate": 7.3876024188818395e-06,
+      "loss": 2.0494,
+      "step": 9804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6067697446509266,
+      "learning_rate": 7.3831696322014505e-06,
+      "loss": 1.8969,
+      "step": 9805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5524362610887664,
+      "learning_rate": 7.378737945412151e-06,
+      "loss": 1.8958,
+      "step": 9806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6457859036960178,
+      "learning_rate": 7.374307358790633e-06,
+      "loss": 1.8729,
+      "step": 9807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5969772096202524,
+      "learning_rate": 7.369877872613515e-06,
+      "loss": 2.0848,
+      "step": 9808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5109428898763667,
+      "learning_rate": 7.365449487157347e-06,
+      "loss": 1.9325,
+      "step": 9809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6627969106344849,
+      "learning_rate": 7.361022202698612e-06,
+      "loss": 1.8732,
+      "step": 9810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.606796263046673,
+      "learning_rate": 7.356596019513723e-06,
+      "loss": 1.8129,
+      "step": 9811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6056828403086492,
+      "learning_rate": 7.3521709378790275e-06,
+      "loss": 2.0809,
+      "step": 9812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.634480860933343,
+      "learning_rate": 7.347746958070798e-06,
+      "loss": 1.8592,
+      "step": 9813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6552062772959275,
+      "learning_rate": 7.343324080365244e-06,
+      "loss": 1.8457,
+      "step": 9814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5735452854076498,
+      "learning_rate": 7.338902305038503e-06,
+      "loss": 1.9174,
+      "step": 9815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5840527400736987,
+      "learning_rate": 7.334481632366643e-06,
+      "loss": 2.0661,
+      "step": 9816
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.552672251090426,
+      "learning_rate": 7.330062062625667e-06,
+      "loss": 1.8793,
+      "step": 9817
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5743133305934345,
+      "learning_rate": 7.325643596091508e-06,
+      "loss": 1.8126,
+      "step": 9818
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6042115091194048,
+      "learning_rate": 7.321226233040035e-06,
+      "loss": 1.8613,
+      "step": 9819
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5905324398901795,
+      "learning_rate": 7.316809973747016e-06,
+      "loss": 2.0659,
+      "step": 9820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6234941129279045,
+      "learning_rate": 7.312394818488205e-06,
+      "loss": 1.9728,
+      "step": 9821
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5773807809380678,
+      "learning_rate": 7.307980767539257e-06,
+      "loss": 1.8797,
+      "step": 9822
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5713096930685233,
+      "learning_rate": 7.303567821175733e-06,
+      "loss": 1.8893,
+      "step": 9823
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6245686703163551,
+      "learning_rate": 7.299155979673184e-06,
+      "loss": 2.0715,
+      "step": 9824
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5987778343956072,
+      "learning_rate": 7.294745243307027e-06,
+      "loss": 1.8755,
+      "step": 9825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5838754051737682,
+      "learning_rate": 7.290335612352669e-06,
+      "loss": 1.8576,
+      "step": 9826
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6420828747248806,
+      "learning_rate": 7.285927087085423e-06,
+      "loss": 1.8691,
+      "step": 9827
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6045700993834725,
+      "learning_rate": 7.281519667780501e-06,
+      "loss": 2.0805,
+      "step": 9828
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.589475220001574,
+      "learning_rate": 7.277113354713103e-06,
+      "loss": 1.825,
+      "step": 9829
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5477253206238963,
+      "learning_rate": 7.2727081481583335e-06,
+      "loss": 1.8102,
+      "step": 9830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6099778415126593,
+      "learning_rate": 7.268304048391205e-06,
+      "loss": 1.8386,
+      "step": 9831
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6169853212817328,
+      "learning_rate": 7.263901055686714e-06,
+      "loss": 2.0795,
+      "step": 9832
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5508458606721717,
+      "learning_rate": 7.259499170319734e-06,
+      "loss": 1.894,
+      "step": 9833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5349659496897745,
+      "learning_rate": 7.255098392565093e-06,
+      "loss": 1.9179,
+      "step": 9834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5857638243257839,
+      "learning_rate": 7.2506987226975716e-06,
+      "loss": 1.8731,
+      "step": 9835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5710302620260148,
+      "learning_rate": 7.246300160991839e-06,
+      "loss": 2.0431,
+      "step": 9836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5738535332957648,
+      "learning_rate": 7.241902707722511e-06,
+      "loss": 1.8598,
+      "step": 9837
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5996640473099134,
+      "learning_rate": 7.237506363164168e-06,
+      "loss": 1.882,
+      "step": 9838
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6237261172152307,
+      "learning_rate": 7.233111127591255e-06,
+      "loss": 1.859,
+      "step": 9839
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5218589257120632,
+      "learning_rate": 7.228717001278221e-06,
+      "loss": 2.0833,
+      "step": 9840
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6014920159700984,
+      "learning_rate": 7.224323984499379e-06,
+      "loss": 1.8829,
+      "step": 9841
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.581148539569195,
+      "learning_rate": 7.21993207752901e-06,
+      "loss": 1.894,
+      "step": 9842
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5489330794673478,
+      "learning_rate": 7.215541280641338e-06,
+      "loss": 1.8547,
+      "step": 9843
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6121026608985648,
+      "learning_rate": 7.211151594110477e-06,
+      "loss": 2.0526,
+      "step": 9844
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5817484792295209,
+      "learning_rate": 7.206763018210499e-06,
+      "loss": 1.8829,
+      "step": 9845
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5624682427572908,
+      "learning_rate": 7.202375553215404e-06,
+      "loss": 1.9599,
+      "step": 9846
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5884247706087061,
+      "learning_rate": 7.197989199399116e-06,
+      "loss": 1.8925,
+      "step": 9847
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6499740023749654,
+      "learning_rate": 7.1936039570354965e-06,
+      "loss": 2.0711,
+      "step": 9848
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5615400199454946,
+      "learning_rate": 7.189219826398333e-06,
+      "loss": 1.8745,
+      "step": 9849
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.590790370419666,
+      "learning_rate": 7.184836807761342e-06,
+      "loss": 1.8706,
+      "step": 9850
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6201380499080543,
+      "learning_rate": 7.180454901398176e-06,
+      "loss": 1.8563,
+      "step": 9851
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5389661428004826,
+      "learning_rate": 7.176074107582417e-06,
+      "loss": 1.8901,
+      "step": 9852
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5718947937785859,
+      "learning_rate": 7.171694426587572e-06,
+      "loss": 2.065,
+      "step": 9853
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.622479199982762,
+      "learning_rate": 7.167315858687085e-06,
+      "loss": 1.918,
+      "step": 9854
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6052910816377971,
+      "learning_rate": 7.1629384041543265e-06,
+      "loss": 1.8096,
+      "step": 9855
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6701438330129966,
+      "learning_rate": 7.158562063262603e-06,
+      "loss": 2.1117,
+      "step": 9856
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6168430298355714,
+      "learning_rate": 7.154186836285143e-06,
+      "loss": 1.8551,
+      "step": 9857
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5941065879165437,
+      "learning_rate": 7.149812723495109e-06,
+      "loss": 1.8708,
+      "step": 9858
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5585317496293507,
+      "learning_rate": 7.145439725165598e-06,
+      "loss": 1.9305,
+      "step": 9859
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.647558653147079,
+      "learning_rate": 7.1410678415696346e-06,
+      "loss": 2.0959,
+      "step": 9860
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5756354007410476,
+      "learning_rate": 7.136697072980172e-06,
+      "loss": 1.843,
+      "step": 9861
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6188633118354876,
+      "learning_rate": 7.132327419670095e-06,
+      "loss": 1.8542,
+      "step": 9862
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.602723005834421,
+      "learning_rate": 7.127958881912219e-06,
+      "loss": 1.8844,
+      "step": 9863
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6575052113461541,
+      "learning_rate": 7.12359145997929e-06,
+      "loss": 2.0612,
+      "step": 9864
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5371976658340089,
+      "learning_rate": 7.119225154143985e-06,
+      "loss": 1.9597,
+      "step": 9865
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6238552885204242,
+      "learning_rate": 7.114859964678918e-06,
+      "loss": 1.8646,
+      "step": 9866
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6448140663064231,
+      "learning_rate": 7.1104958918566e-06,
+      "loss": 1.859,
+      "step": 9867
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6313224309788756,
+      "learning_rate": 7.106132935949525e-06,
+      "loss": 2.0695,
+      "step": 9868
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5780677362809897,
+      "learning_rate": 7.10177109723009e-06,
+      "loss": 1.8542,
+      "step": 9869
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6567428839385522,
+      "learning_rate": 7.097410375970595e-06,
+      "loss": 1.8648,
+      "step": 9870
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5548950452897783,
+      "learning_rate": 7.093050772443324e-06,
+      "loss": 1.908,
+      "step": 9871
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5624506428149934,
+      "learning_rate": 7.088692286920456e-06,
+      "loss": 2.0273,
+      "step": 9872
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5788219561782851,
+      "learning_rate": 7.084334919674112e-06,
+      "loss": 1.9013,
+      "step": 9873
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6075821232200778,
+      "learning_rate": 7.079978670976345e-06,
+      "loss": 1.8772,
+      "step": 9874
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5760104755851339,
+      "learning_rate": 7.075623541099111e-06,
+      "loss": 1.8438,
+      "step": 9875
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5901380409973135,
+      "learning_rate": 7.071269530314342e-06,
+      "loss": 2.0426,
+      "step": 9876
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5599765420974949,
+      "learning_rate": 7.066916638893878e-06,
+      "loss": 1.9542,
+      "step": 9877
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5559067159582881,
+      "learning_rate": 7.062564867109464e-06,
+      "loss": 1.8511,
+      "step": 9878
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6438417692350227,
+      "learning_rate": 7.058214215232828e-06,
+      "loss": 1.8989,
+      "step": 9879
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6810140748723532,
+      "learning_rate": 7.05386468353558e-06,
+      "loss": 2.0964,
+      "step": 9880
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6039493022166044,
+      "learning_rate": 7.049516272289275e-06,
+      "loss": 1.8921,
+      "step": 9881
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6586785237096983,
+      "learning_rate": 7.045168981765426e-06,
+      "loss": 1.8629,
+      "step": 9882
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7018095238463288,
+      "learning_rate": 7.040822812235434e-06,
+      "loss": 1.9476,
+      "step": 9883
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6113648033187449,
+      "learning_rate": 7.036477763970642e-06,
+      "loss": 1.8693,
+      "step": 9884
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5861326403573981,
+      "learning_rate": 7.032133837242352e-06,
+      "loss": 2.0457,
+      "step": 9885
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.660770291217403,
+      "learning_rate": 7.027791032321757e-06,
+      "loss": 1.8811,
+      "step": 9886
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6476727127953458,
+      "learning_rate": 7.023449349479996e-06,
+      "loss": 1.8569,
+      "step": 9887
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5885927269379675,
+      "learning_rate": 7.019108788988143e-06,
+      "loss": 2.0204,
+      "step": 9888
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6160449409279809,
+      "learning_rate": 7.01476935111719e-06,
+      "loss": 1.8407,
+      "step": 9889
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6237541720839994,
+      "learning_rate": 7.010431036138085e-06,
+      "loss": 1.9501,
+      "step": 9890
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6145499552931796,
+      "learning_rate": 7.006093844321665e-06,
+      "loss": 1.8518,
+      "step": 9891
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6294071976101056,
+      "learning_rate": 7.001757775938728e-06,
+      "loss": 2.0135,
+      "step": 9892
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6610598908331337,
+      "learning_rate": 6.997422831259992e-06,
+      "loss": 1.9065,
+      "step": 9893
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5609240293097617,
+      "learning_rate": 6.993089010556103e-06,
+      "loss": 1.8445,
+      "step": 9894
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6073650543201157,
+      "learning_rate": 6.988756314097644e-06,
+      "loss": 1.8489,
+      "step": 9895
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6492080809728918,
+      "learning_rate": 6.984424742155119e-06,
+      "loss": 1.9524,
+      "step": 9896
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6164924151985212,
+      "learning_rate": 6.980094294998968e-06,
+      "loss": 2.0541,
+      "step": 9897
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5857045627558285,
+      "learning_rate": 6.975764972899556e-06,
+      "loss": 1.8433,
+      "step": 9898
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5855218376581135,
+      "learning_rate": 6.971436776127182e-06,
+      "loss": 1.8002,
+      "step": 9899
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6309242972311062,
+      "learning_rate": 6.967109704952077e-06,
+      "loss": 2.0417,
+      "step": 9900
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.584947267469249,
+      "learning_rate": 6.962783759644393e-06,
+      "loss": 1.8094,
+      "step": 9901
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5961253138212652,
+      "learning_rate": 6.9584589404742166e-06,
+      "loss": 1.9693,
+      "step": 9902
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.568661916869947,
+      "learning_rate": 6.954135247711563e-06,
+      "loss": 1.8323,
+      "step": 9903
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5825947746596803,
+      "learning_rate": 6.949812681626383e-06,
+      "loss": 1.8477,
+      "step": 9904
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5475381638221739,
+      "learning_rate": 6.9454912424885475e-06,
+      "loss": 2.0453,
+      "step": 9905
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5660396805192651,
+      "learning_rate": 6.941170930567864e-06,
+      "loss": 1.8565,
+      "step": 9906
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6140274907693337,
+      "learning_rate": 6.936851746134068e-06,
+      "loss": 1.8912,
+      "step": 9907
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5529963599960196,
+      "learning_rate": 6.932533689456822e-06,
+      "loss": 1.9491,
+      "step": 9908
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6078343002861833,
+      "learning_rate": 6.92821676080572e-06,
+      "loss": 2.0841,
+      "step": 9909
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5747974280690421,
+      "learning_rate": 6.923900960450289e-06,
+      "loss": 1.9019,
+      "step": 9910
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5718888992215727,
+      "learning_rate": 6.919586288659977e-06,
+      "loss": 1.8654,
+      "step": 9911
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5884662966864908,
+      "learning_rate": 6.915272745704168e-06,
+      "loss": 2.0349,
+      "step": 9912
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5813157646006861,
+      "learning_rate": 6.910960331852187e-06,
+      "loss": 1.8775,
+      "step": 9913
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.549247814829925,
+      "learning_rate": 6.906649047373246e-06,
+      "loss": 1.9923,
+      "step": 9914
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5920064263214806,
+      "learning_rate": 6.902338892536542e-06,
+      "loss": 1.8701,
+      "step": 9915
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5802551890479,
+      "learning_rate": 6.898029867611177e-06,
+      "loss": 1.8897,
+      "step": 9916
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5719162576192439,
+      "learning_rate": 6.893721972866154e-06,
+      "loss": 2.0842,
+      "step": 9917
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5752422800025054,
+      "learning_rate": 6.889415208570463e-06,
+      "loss": 1.8801,
+      "step": 9918
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5762258227887975,
+      "learning_rate": 6.885109574992984e-06,
+      "loss": 1.8074,
+      "step": 9919
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5635999605545179,
+      "learning_rate": 6.8808050724025185e-06,
+      "loss": 2.0545,
+      "step": 9920
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6084867791842884,
+      "learning_rate": 6.876501701067844e-06,
+      "loss": 1.9374,
+      "step": 9921
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.571157135239423,
+      "learning_rate": 6.872199461257606e-06,
+      "loss": 1.8836,
+      "step": 9922
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5904110741756465,
+      "learning_rate": 6.867898353240435e-06,
+      "loss": 1.8385,
+      "step": 9923
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5582102773559309,
+      "learning_rate": 6.863598377284863e-06,
+      "loss": 2.0694,
+      "step": 9924
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5941987551340875,
+      "learning_rate": 6.859299533659336e-06,
+      "loss": 1.8539,
+      "step": 9925
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5910099621134584,
+      "learning_rate": 6.855001822632278e-06,
+      "loss": 1.8286,
+      "step": 9926
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6149555828179953,
+      "learning_rate": 6.850705244471992e-06,
+      "loss": 1.9277,
+      "step": 9927
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6028207502760927,
+      "learning_rate": 6.846409799446729e-06,
+      "loss": 1.8903,
+      "step": 9928
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5428019352296549,
+      "learning_rate": 6.842115487824693e-06,
+      "loss": 2.0702,
+      "step": 9929
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6691235264557737,
+      "learning_rate": 6.837822309873976e-06,
+      "loss": 1.8943,
+      "step": 9930
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6329276741508363,
+      "learning_rate": 6.833530265862614e-06,
+      "loss": 1.8274,
+      "step": 9931
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6185793765716361,
+      "learning_rate": 6.8292393560586035e-06,
+      "loss": 2.0335,
+      "step": 9932
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5752119925983595,
+      "learning_rate": 6.82494958072982e-06,
+      "loss": 1.9542,
+      "step": 9933
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.579375854234088,
+      "learning_rate": 6.8206609401441e-06,
+      "loss": 1.899,
+      "step": 9934
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5940027930950578,
+      "learning_rate": 6.816373434569198e-06,
+      "loss": 1.8573,
+      "step": 9935
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6862348211290475,
+      "learning_rate": 6.8120870642727955e-06,
+      "loss": 1.8643,
+      "step": 9936
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6395998523386474,
+      "learning_rate": 6.807801829522534e-06,
+      "loss": 2.0391,
+      "step": 9937
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5597532720150011,
+      "learning_rate": 6.803517730585929e-06,
+      "loss": 1.8869,
+      "step": 9938
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5854461778465972,
+      "learning_rate": 6.799234767730464e-06,
+      "loss": 2.0131,
+      "step": 9939
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6222303392716293,
+      "learning_rate": 6.794952941223545e-06,
+      "loss": 1.8833,
+      "step": 9940
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.611920714170305,
+      "learning_rate": 6.790672251332503e-06,
+      "loss": 2.0961,
+      "step": 9941
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6002530474081406,
+      "learning_rate": 6.786392698324595e-06,
+      "loss": 1.8336,
+      "step": 9942
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6044973178823012,
+      "learning_rate": 6.782114282467017e-06,
+      "loss": 1.9133,
+      "step": 9943
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.612883647746902,
+      "learning_rate": 6.777837004026885e-06,
+      "loss": 2.0679,
+      "step": 9944
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5166189446993887,
+      "learning_rate": 6.773560863271247e-06,
+      "loss": 1.9493,
+      "step": 9945
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6223243861212813,
+      "learning_rate": 6.769285860467078e-06,
+      "loss": 1.8781,
+      "step": 9946
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6182267530224876,
+      "learning_rate": 6.76501199588129e-06,
+      "loss": 1.8873,
+      "step": 9947
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5544029840968173,
+      "learning_rate": 6.760739269780714e-06,
+      "loss": 1.8818,
+      "step": 9948
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5743640282013892,
+      "learning_rate": 6.75646768243211e-06,
+      "loss": 2.0683,
+      "step": 9949
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6017063720815752,
+      "learning_rate": 6.752197234102179e-06,
+      "loss": 1.856,
+      "step": 9950
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5936190725243424,
+      "learning_rate": 6.747927925057537e-06,
+      "loss": 1.82,
+      "step": 9951
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.587064800910742,
+      "learning_rate": 6.743659755564738e-06,
+      "loss": 1.9624,
+      "step": 9952
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5637034760619979,
+      "learning_rate": 6.739392725890256e-06,
+      "loss": 2.0994,
+      "step": 9953
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5717245413617875,
+      "learning_rate": 6.735126836300504e-06,
+      "loss": 1.8872,
+      "step": 9954
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5772121715304169,
+      "learning_rate": 6.730862087061818e-06,
+      "loss": 1.89,
+      "step": 9955
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.666442961052075,
+      "learning_rate": 6.7265984784404655e-06,
+      "loss": 1.9956,
+      "step": 9956
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5830694592813649,
+      "learning_rate": 6.722336010702634e-06,
+      "loss": 1.8649,
+      "step": 9957
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5619501115789374,
+      "learning_rate": 6.718074684114456e-06,
+      "loss": 1.9065,
+      "step": 9958
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5680654249010272,
+      "learning_rate": 6.713814498941978e-06,
+      "loss": 1.9055,
+      "step": 9959
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5780008396698069,
+      "learning_rate": 6.7095554554511916e-06,
+      "loss": 1.8141,
+      "step": 9960
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.608119798314736,
+      "learning_rate": 6.70529755390798e-06,
+      "loss": 2.0891,
+      "step": 9961
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5561305834915199,
+      "learning_rate": 6.701040794578206e-06,
+      "loss": 1.8134,
+      "step": 9962
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6067786735025787,
+      "learning_rate": 6.696785177727638e-06,
+      "loss": 1.8829,
+      "step": 9963
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5943768684151364,
+      "learning_rate": 6.692530703621947e-06,
+      "loss": 1.9586,
+      "step": 9964
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.557904373419838,
+      "learning_rate": 6.688277372526783e-06,
+      "loss": 2.0555,
+      "step": 9965
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6175311489517533,
+      "learning_rate": 6.684025184707693e-06,
+      "loss": 1.8625,
+      "step": 9966
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6023966439951527,
+      "learning_rate": 6.679774140430143e-06,
+      "loss": 1.9094,
+      "step": 9967
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5867447031494165,
+      "learning_rate": 6.675524239959566e-06,
+      "loss": 1.8962,
+      "step": 9968
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5933499724920428,
+      "learning_rate": 6.671275483561285e-06,
+      "loss": 2.0624,
+      "step": 9969
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.567088893904526,
+      "learning_rate": 6.667027871500562e-06,
+      "loss": 1.9079,
+      "step": 9970
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5785237753922738,
+      "learning_rate": 6.66278140404262e-06,
+      "loss": 1.8581,
+      "step": 9971
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6474861858884289,
+      "learning_rate": 6.658536081452549e-06,
+      "loss": 1.8548,
+      "step": 9972
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5314996654002195,
+      "learning_rate": 6.654291903995436e-06,
+      "loss": 2.0279,
+      "step": 9973
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5869809651041972,
+      "learning_rate": 6.650048871936237e-06,
+      "loss": 1.8561,
+      "step": 9974
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5830605481135058,
+      "learning_rate": 6.645806985539863e-06,
+      "loss": 1.8325,
+      "step": 9975
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.612416575592877,
+      "learning_rate": 6.6415662450711765e-06,
+      "loss": 2.0703,
+      "step": 9976
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5914717951755843,
+      "learning_rate": 6.637326650794917e-06,
+      "loss": 1.9208,
+      "step": 9977
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6242455548079016,
+      "learning_rate": 6.633088202975785e-06,
+      "loss": 1.919,
+      "step": 9978
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5772746947727049,
+      "learning_rate": 6.628850901878425e-06,
+      "loss": 1.9028,
+      "step": 9979
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5576240883809834,
+      "learning_rate": 6.6246147477673656e-06,
+      "loss": 1.8122,
+      "step": 9980
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6249005168028868,
+      "learning_rate": 6.620379740907095e-06,
+      "loss": 2.058,
+      "step": 9981
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5790472165284056,
+      "learning_rate": 6.6161458815620216e-06,
+      "loss": 1.8779,
+      "step": 9982
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.636280980716656,
+      "learning_rate": 6.611913169996487e-06,
+      "loss": 1.8997,
+      "step": 9983
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5848290117705705,
+      "learning_rate": 6.60768160647475e-06,
+      "loss": 1.8785,
+      "step": 9984
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5541827202225574,
+      "learning_rate": 6.603451191261009e-06,
+      "loss": 2.087,
+      "step": 9985
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5742661888654794,
+      "learning_rate": 6.599221924619378e-06,
+      "loss": 1.8494,
+      "step": 9986
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6299499531508945,
+      "learning_rate": 6.5949938068139285e-06,
+      "loss": 1.8677,
+      "step": 9987
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6773542594643112,
+      "learning_rate": 6.59076683810862e-06,
+      "loss": 1.865,
+      "step": 9988
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5700730851073591,
+      "learning_rate": 6.586541018767362e-06,
+      "loss": 2.115,
+      "step": 9989
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.552819991442068,
+      "learning_rate": 6.582316349053994e-06,
+      "loss": 1.8921,
+      "step": 9990
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6390164747217233,
+      "learning_rate": 6.578092829232277e-06,
+      "loss": 1.8666,
+      "step": 9991
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6055115445308528,
+      "learning_rate": 6.5738704595659065e-06,
+      "loss": 1.8626,
+      "step": 9992
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5979122394726587,
+      "learning_rate": 6.5696492403184984e-06,
+      "loss": 2.0472,
+      "step": 9993
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5779865783798013,
+      "learning_rate": 6.565429171753601e-06,
+      "loss": 1.9004,
+      "step": 9994
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.589463043358146,
+      "learning_rate": 6.561210254134692e-06,
+      "loss": 1.9451,
+      "step": 9995
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5716028717010009,
+      "learning_rate": 6.556992487725175e-06,
+      "loss": 1.8594,
+      "step": 9996
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5782449804425381,
+      "learning_rate": 6.5527758727883834e-06,
+      "loss": 2.0549,
+      "step": 9997
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6043053307236973,
+      "learning_rate": 6.548560409587576e-06,
+      "loss": 1.8392,
+      "step": 9998
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6028622129263858,
+      "learning_rate": 6.544346098385942e-06,
+      "loss": 1.8521,
+      "step": 9999
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5864628895617997,
+      "learning_rate": 6.540132939446603e-06,
+      "loss": 1.8805,
+      "step": 10000
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.586361992547937,
+      "learning_rate": 6.535920933032596e-06,
+      "loss": 2.0866,
+      "step": 10001
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5873368297972714,
+      "learning_rate": 6.531710079406906e-06,
+      "loss": 1.8355,
+      "step": 10002
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5791391711167567,
+      "learning_rate": 6.52750037883241e-06,
+      "loss": 1.895,
+      "step": 10003
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5485657013195775,
+      "learning_rate": 6.523291831571959e-06,
+      "loss": 1.8427,
+      "step": 10004
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5477465347979119,
+      "learning_rate": 6.519084437888304e-06,
+      "loss": 2.0585,
+      "step": 10005
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.575453858685787,
+      "learning_rate": 6.51487819804413e-06,
+      "loss": 1.9265,
+      "step": 10006
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5478010492156601,
+      "learning_rate": 6.5106731123020546e-06,
+      "loss": 1.8839,
+      "step": 10007
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5769967714588066,
+      "learning_rate": 6.506469180924599e-06,
+      "loss": 1.9629,
+      "step": 10008
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6034280670721905,
+      "learning_rate": 6.502266404174253e-06,
+      "loss": 2.0596,
+      "step": 10009
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6033345144568478,
+      "learning_rate": 6.498064782313415e-06,
+      "loss": 1.8825,
+      "step": 10010
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6067575839944989,
+      "learning_rate": 6.4938643156043835e-06,
+      "loss": 1.8883,
+      "step": 10011
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5867719907349904,
+      "learning_rate": 6.489665004309437e-06,
+      "loss": 1.8519,
+      "step": 10012
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5746031227098906,
+      "learning_rate": 6.485466848690752e-06,
+      "loss": 2.0812,
+      "step": 10013
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6096205149191006,
+      "learning_rate": 6.481269849010419e-06,
+      "loss": 1.9063,
+      "step": 10014
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6076453387857592,
+      "learning_rate": 6.477074005530498e-06,
+      "loss": 1.8952,
+      "step": 10015
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5946102402060761,
+      "learning_rate": 6.4728793185129336e-06,
+      "loss": 1.8409,
+      "step": 10016
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5968892988986049,
+      "learning_rate": 6.46868578821962e-06,
+      "loss": 2.0379,
+      "step": 10017
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5942751560893951,
+      "learning_rate": 6.464493414912396e-06,
+      "loss": 1.91,
+      "step": 10018
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5648972887726359,
+      "learning_rate": 6.460302198852977e-06,
+      "loss": 1.9067,
+      "step": 10019
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5942393337560976,
+      "learning_rate": 6.456112140303061e-06,
+      "loss": 1.9133,
+      "step": 10020
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6585322177155132,
+      "learning_rate": 6.451923239524254e-06,
+      "loss": 2.0709,
+      "step": 10021
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6141989158744526,
+      "learning_rate": 6.4477354967780626e-06,
+      "loss": 1.8884,
+      "step": 10022
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6002928844728935,
+      "learning_rate": 6.443548912325972e-06,
+      "loss": 1.8747,
+      "step": 10023
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5634394492724144,
+      "learning_rate": 6.439363486429348e-06,
+      "loss": 1.8797,
+      "step": 10024
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5761662899349146,
+      "learning_rate": 6.435179219349502e-06,
+      "loss": 2.0783,
+      "step": 10025
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5579647921737981,
+      "learning_rate": 6.430996111347698e-06,
+      "loss": 1.9562,
+      "step": 10026
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5659666406618951,
+      "learning_rate": 6.426814162685086e-06,
+      "loss": 1.8415,
+      "step": 10027
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5787391743069016,
+      "learning_rate": 6.422633373622763e-06,
+      "loss": 1.8564,
+      "step": 10028
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6630315907972386,
+      "learning_rate": 6.418453744421757e-06,
+      "loss": 2.0494,
+      "step": 10029
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6110796604399334,
+      "learning_rate": 6.414275275343018e-06,
+      "loss": 1.8535,
+      "step": 10030
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5794576965042313,
+      "learning_rate": 6.410097966647427e-06,
+      "loss": 1.8939,
+      "step": 10031
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5864510194271094,
+      "learning_rate": 6.4059218185957895e-06,
+      "loss": 1.9318,
+      "step": 10032
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.655566102467746,
+      "learning_rate": 6.401746831448838e-06,
+      "loss": 2.067,
+      "step": 10033
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5915265496063535,
+      "learning_rate": 6.397573005467236e-06,
+      "loss": 1.9018,
+      "step": 10034
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5551210299638268,
+      "learning_rate": 6.3934003409115705e-06,
+      "loss": 1.8719,
+      "step": 10035
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5871593044917669,
+      "learning_rate": 6.3892288380423605e-06,
+      "loss": 1.868,
+      "step": 10036
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5966944048283587,
+      "learning_rate": 6.38505849712005e-06,
+      "loss": 2.0476,
+      "step": 10037
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5842778970241503,
+      "learning_rate": 6.3808893184050086e-06,
+      "loss": 1.8378,
+      "step": 10038
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5544912691981234,
+      "learning_rate": 6.376721302157537e-06,
+      "loss": 1.9651,
+      "step": 10039
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5771080296047124,
+      "learning_rate": 6.37255444863786e-06,
+      "loss": 1.8301,
+      "step": 10040
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5491719794670327,
+      "learning_rate": 6.368388758106133e-06,
+      "loss": 2.0695,
+      "step": 10041
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5903850670669745,
+      "learning_rate": 6.3642242308224385e-06,
+      "loss": 1.8568,
+      "step": 10042
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5658938617819126,
+      "learning_rate": 6.36006086704678e-06,
+      "loss": 1.8107,
+      "step": 10043
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5305314551735982,
+      "learning_rate": 6.355898667039098e-06,
+      "loss": 1.8473,
+      "step": 10044
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5692333602334103,
+      "learning_rate": 6.351737631059251e-06,
+      "loss": 2.0614,
+      "step": 10045
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6000196452009824,
+      "learning_rate": 6.347577759367035e-06,
+      "loss": 1.8995,
+      "step": 10046
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5866873000412588,
+      "learning_rate": 6.343419052222166e-06,
+      "loss": 1.832,
+      "step": 10047
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5718061035122218,
+      "learning_rate": 6.339261509884287e-06,
+      "loss": 1.8479,
+      "step": 10048
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6466662187942003,
+      "learning_rate": 6.335105132612981e-06,
+      "loss": 2.0644,
+      "step": 10049
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5977666912694315,
+      "learning_rate": 6.330949920667725e-06,
+      "loss": 1.8454,
+      "step": 10050
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5484974019375556,
+      "learning_rate": 6.3267958743079644e-06,
+      "loss": 1.9043,
+      "step": 10051
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5749453761136436,
+      "learning_rate": 6.322642993793051e-06,
+      "loss": 1.8355,
+      "step": 10052
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5399241629411601,
+      "learning_rate": 6.318491279382263e-06,
+      "loss": 2.0221,
+      "step": 10053
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6115173243275183,
+      "learning_rate": 6.314340731334811e-06,
+      "loss": 1.9055,
+      "step": 10054
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5852342682935882,
+      "learning_rate": 6.31019134990983e-06,
+      "loss": 1.8264,
+      "step": 10055
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6147865791758184,
+      "learning_rate": 6.306043135366382e-06,
+      "loss": 1.8352,
+      "step": 10056
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5151068172060812,
+      "learning_rate": 6.301896087963463e-06,
+      "loss": 2.138,
+      "step": 10057
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6004863813288748,
+      "learning_rate": 6.297750207959973e-06,
+      "loss": 1.8971,
+      "step": 10058
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5557952775898698,
+      "learning_rate": 6.2936054956147734e-06,
+      "loss": 1.8587,
+      "step": 10059
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5763869969958483,
+      "learning_rate": 6.28946195118664e-06,
+      "loss": 1.8159,
+      "step": 10060
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5549057652601986,
+      "learning_rate": 6.285319574934248e-06,
+      "loss": 2.0673,
+      "step": 10061
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6102100694782228,
+      "learning_rate": 6.28117836711625e-06,
+      "loss": 1.9474,
+      "step": 10062
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5262440448497068,
+      "learning_rate": 6.277038327991178e-06,
+      "loss": 1.9861,
+      "step": 10063
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5857097821819889,
+      "learning_rate": 6.27289945781751e-06,
+      "loss": 1.8657,
+      "step": 10064
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.554190064908165,
+      "learning_rate": 6.268761756853678e-06,
+      "loss": 2.0046,
+      "step": 10065
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6066211580840386,
+      "learning_rate": 6.264625225357992e-06,
+      "loss": 1.8893,
+      "step": 10066
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6125863165231086,
+      "learning_rate": 6.260489863588714e-06,
+      "loss": 1.8816,
+      "step": 10067
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5681644085835287,
+      "learning_rate": 6.256355671804051e-06,
+      "loss": 1.8538,
+      "step": 10068
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5595138644142642,
+      "learning_rate": 6.252222650262088e-06,
+      "loss": 2.0469,
+      "step": 10069
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5808291677875814,
+      "learning_rate": 6.248090799220902e-06,
+      "loss": 1.9365,
+      "step": 10070
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5972158629384291,
+      "learning_rate": 6.24396011893843e-06,
+      "loss": 1.7938,
+      "step": 10071
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5568687597262374,
+      "learning_rate": 6.2398306096725725e-06,
+      "loss": 1.8651,
+      "step": 10072
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.593375835736498,
+      "learning_rate": 6.235702271681176e-06,
+      "loss": 2.0963,
+      "step": 10073
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6032003588416528,
+      "learning_rate": 6.231575105221965e-06,
+      "loss": 1.889,
+      "step": 10074
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5807758911649286,
+      "learning_rate": 6.227449110552622e-06,
+      "loss": 1.8522,
+      "step": 10075
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5355647110133122,
+      "learning_rate": 6.2233242879307485e-06,
+      "loss": 1.9854,
+      "step": 10076
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5517738752798058,
+      "learning_rate": 6.219200637613878e-06,
+      "loss": 2.0728,
+      "step": 10077
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5522783668122618,
+      "learning_rate": 6.215078159859467e-06,
+      "loss": 1.8672,
+      "step": 10078
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5545673671107626,
+      "learning_rate": 6.2109568549248955e-06,
+      "loss": 1.8258,
+      "step": 10079
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5912673635993275,
+      "learning_rate": 6.2068367230674746e-06,
+      "loss": 1.8731,
+      "step": 10080
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.634378555627443,
+      "learning_rate": 6.202717764544441e-06,
+      "loss": 2.0781,
+      "step": 10081
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5439353515640469,
+      "learning_rate": 6.1985999796129565e-06,
+      "loss": 1.9569,
+      "step": 10082
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5953419649431108,
+      "learning_rate": 6.194483368530116e-06,
+      "loss": 1.8085,
+      "step": 10083
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5688934715157574,
+      "learning_rate": 6.1903679315529345e-06,
+      "loss": 1.906,
+      "step": 10084
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5715189741768948,
+      "learning_rate": 6.1862536689383546e-06,
+      "loss": 2.0499,
+      "step": 10085
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6208629482778797,
+      "learning_rate": 6.1821405809432465e-06,
+      "loss": 1.8366,
+      "step": 10086
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5753310144701023,
+      "learning_rate": 6.178028667824407e-06,
+      "loss": 1.9116,
+      "step": 10087
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.49365565170200715,
+      "learning_rate": 6.1739179298385605e-06,
+      "loss": 1.9509,
+      "step": 10088
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5771349074060218,
+      "learning_rate": 6.1698083672423575e-06,
+      "loss": 1.9057,
+      "step": 10089
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5833238491934783,
+      "learning_rate": 6.165699980292375e-06,
+      "loss": 2.0414,
+      "step": 10090
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5954294323002491,
+      "learning_rate": 6.161592769245114e-06,
+      "loss": 1.8867,
+      "step": 10091
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5806750673585022,
+      "learning_rate": 6.157486734357007e-06,
+      "loss": 1.8525,
+      "step": 10092
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5945895819094865,
+      "learning_rate": 6.153381875884412e-06,
+      "loss": 2.0745,
+      "step": 10093
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5709817744439,
+      "learning_rate": 6.149278194083611e-06,
+      "loss": 1.9667,
+      "step": 10094
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6474232843319209,
+      "learning_rate": 6.145175689210808e-06,
+      "loss": 1.9097,
+      "step": 10095
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5598658412041255,
+      "learning_rate": 6.141074361522156e-06,
+      "loss": 1.847,
+      "step": 10096
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6146565546275735,
+      "learning_rate": 6.136974211273691e-06,
+      "loss": 2.0331,
+      "step": 10097
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5777107508715803,
+      "learning_rate": 6.132875238721425e-06,
+      "loss": 1.8992,
+      "step": 10098
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.630145675797826,
+      "learning_rate": 6.12877744412127e-06,
+      "loss": 1.9002,
+      "step": 10099
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6038323782580428,
+      "learning_rate": 6.1246808277290514e-06,
+      "loss": 1.8445,
+      "step": 10100
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5228287717230661,
+      "learning_rate": 6.120585389800559e-06,
+      "loss": 1.9604,
+      "step": 10101
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6381279711780531,
+      "learning_rate": 6.1164911305914785e-06,
+      "loss": 2.0549,
+      "step": 10102
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5851651295483327,
+      "learning_rate": 6.1123980503574314e-06,
+      "loss": 1.834,
+      "step": 10103
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5771326240035205,
+      "learning_rate": 6.108306149353973e-06,
+      "loss": 1.8883,
+      "step": 10104
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5689894770075464,
+      "learning_rate": 6.1042154278365595e-06,
+      "loss": 2.1256,
+      "step": 10105
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.575100721833533,
+      "learning_rate": 6.100125886060606e-06,
+      "loss": 1.8302,
+      "step": 10106
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5216443231597006,
+      "learning_rate": 6.096037524281445e-06,
+      "loss": 1.9294,
+      "step": 10107
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5852360694403315,
+      "learning_rate": 6.091950342754307e-06,
+      "loss": 1.8879,
+      "step": 10108
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5665430211497783,
+      "learning_rate": 6.0878643417344e-06,
+      "loss": 2.0706,
+      "step": 10109
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5664408951656804,
+      "learning_rate": 6.0837795214768065e-06,
+      "loss": 1.829,
+      "step": 10110
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5585570813029677,
+      "learning_rate": 6.079695882236561e-06,
+      "loss": 1.8622,
+      "step": 10111
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5863249716499088,
+      "learning_rate": 6.075613424268642e-06,
+      "loss": 1.953,
+      "step": 10112
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.576412366030406,
+      "learning_rate": 6.071532147827913e-06,
+      "loss": 1.9511,
+      "step": 10113
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6428748149986354,
+      "learning_rate": 6.067452053169184e-06,
+      "loss": 2.1085,
+      "step": 10114
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6100309206460723,
+      "learning_rate": 6.063373140547215e-06,
+      "loss": 1.8199,
+      "step": 10115
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6048292381368248,
+      "learning_rate": 6.059295410216642e-06,
+      "loss": 1.8836,
+      "step": 10116
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5347002340662244,
+      "learning_rate": 6.05521886243208e-06,
+      "loss": 1.9985,
+      "step": 10117
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6686727425979928,
+      "learning_rate": 6.0511434974480255e-06,
+      "loss": 1.8423,
+      "step": 10118
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6028589174735475,
+      "learning_rate": 6.047069315518916e-06,
+      "loss": 1.8959,
+      "step": 10119
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6132176449514787,
+      "learning_rate": 6.042996316899147e-06,
+      "loss": 1.8677,
+      "step": 10120
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6806251823939801,
+      "learning_rate": 6.0389245018429856e-06,
+      "loss": 1.8489,
+      "step": 10121
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5805168963197802,
+      "learning_rate": 6.034853870604665e-06,
+      "loss": 2.1149,
+      "step": 10122
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6969058015977357,
+      "learning_rate": 6.030784423438323e-06,
+      "loss": 1.8891,
+      "step": 10123
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.609700487912176,
+      "learning_rate": 6.026716160598042e-06,
+      "loss": 1.9001,
+      "step": 10124
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6152482489133086,
+      "learning_rate": 6.022649082337814e-06,
+      "loss": 2.0529,
+      "step": 10125
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6003211240117182,
+      "learning_rate": 6.018583188911564e-06,
+      "loss": 1.9437,
+      "step": 10126
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6002182451010295,
+      "learning_rate": 6.014518480573142e-06,
+      "loss": 1.836,
+      "step": 10127
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6660014445356606,
+      "learning_rate": 6.010454957576328e-06,
+      "loss": 1.8901,
+      "step": 10128
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6216928804887418,
+      "learning_rate": 6.00639262017482e-06,
+      "loss": 2.049,
+      "step": 10129
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5429054514471208,
+      "learning_rate": 6.002331468622249e-06,
+      "loss": 1.8567,
+      "step": 10130
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6413393093060423,
+      "learning_rate": 5.998271503172171e-06,
+      "loss": 1.8519,
+      "step": 10131
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5902031983831537,
+      "learning_rate": 5.99421272407806e-06,
+      "loss": 1.9589,
+      "step": 10132
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5941626227190078,
+      "learning_rate": 5.990155131593328e-06,
+      "loss": 1.8737,
+      "step": 10133
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5955417343391998,
+      "learning_rate": 5.986098725971307e-06,
+      "loss": 2.0747,
+      "step": 10134
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.552707163100088,
+      "learning_rate": 5.9820435074652515e-06,
+      "loss": 1.8385,
+      "step": 10135
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.677261042180884,
+      "learning_rate": 5.9779894763283495e-06,
+      "loss": 1.8682,
+      "step": 10136
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5713822323112319,
+      "learning_rate": 5.973936632813706e-06,
+      "loss": 2.0908,
+      "step": 10137
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5748302642485908,
+      "learning_rate": 5.9698849771743645e-06,
+      "loss": 1.9437,
+      "step": 10138
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.667278025961368,
+      "learning_rate": 5.965834509663276e-06,
+      "loss": 1.8593,
+      "step": 10139
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6520434595527309,
+      "learning_rate": 5.961785230533337e-06,
+      "loss": 1.8723,
+      "step": 10140
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7034314452073165,
+      "learning_rate": 5.957737140037356e-06,
+      "loss": 2.0831,
+      "step": 10141
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5838837759361781,
+      "learning_rate": 5.9536902384280725e-06,
+      "loss": 1.8429,
+      "step": 10142
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6326573336011698,
+      "learning_rate": 5.949644525958159e-06,
+      "loss": 1.8971,
+      "step": 10143
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5522052282123743,
+      "learning_rate": 5.945600002880186e-06,
+      "loss": 1.9224,
+      "step": 10144
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6074494212096242,
+      "learning_rate": 5.941556669446688e-06,
+      "loss": 1.8474,
+      "step": 10145
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6623475156410622,
+      "learning_rate": 5.937514525910107e-06,
+      "loss": 2.0108,
+      "step": 10146
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5992695229736276,
+      "learning_rate": 5.933473572522791e-06,
+      "loss": 1.8422,
+      "step": 10147
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5801299422822204,
+      "learning_rate": 5.929433809537055e-06,
+      "loss": 1.8989,
+      "step": 10148
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5457602110075374,
+      "learning_rate": 5.925395237205111e-06,
+      "loss": 2.1121,
+      "step": 10149
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6447570840681043,
+      "learning_rate": 5.9213578557791025e-06,
+      "loss": 1.9697,
+      "step": 10150
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6355062517673185,
+      "learning_rate": 5.917321665511105e-06,
+      "loss": 1.8199,
+      "step": 10151
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.57744702331984,
+      "learning_rate": 5.9132866666530975e-06,
+      "loss": 1.859,
+      "step": 10152
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5662753112895077,
+      "learning_rate": 5.9092528594570205e-06,
+      "loss": 1.8927,
+      "step": 10153
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6213024308491631,
+      "learning_rate": 5.905220244174719e-06,
+      "loss": 2.0644,
+      "step": 10154
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6821567518067191,
+      "learning_rate": 5.901188821057948e-06,
+      "loss": 1.8458,
+      "step": 10155
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6508438258807706,
+      "learning_rate": 5.897158590358434e-06,
+      "loss": 1.8065,
+      "step": 10156
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5853371132935865,
+      "learning_rate": 5.893129552327781e-06,
+      "loss": 1.934,
+      "step": 10157
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6669399420881194,
+      "learning_rate": 5.889101707217531e-06,
+      "loss": 2.1007,
+      "step": 10158
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6416726353851406,
+      "learning_rate": 5.885075055279188e-06,
+      "loss": 1.8645,
+      "step": 10159
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5765620991236448,
+      "learning_rate": 5.881049596764126e-06,
+      "loss": 1.8283,
+      "step": 10160
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6897390952731093,
+      "learning_rate": 5.877025331923677e-06,
+      "loss": 2.0477,
+      "step": 10161
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.605277413880483,
+      "learning_rate": 5.873002261009106e-06,
+      "loss": 1.8406,
+      "step": 10162
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5393009996615813,
+      "learning_rate": 5.868980384271575e-06,
+      "loss": 1.9112,
+      "step": 10163
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6269097657371308,
+      "learning_rate": 5.864959701962189e-06,
+      "loss": 1.887,
+      "step": 10164
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6095295341773123,
+      "learning_rate": 5.860940214331978e-06,
+      "loss": 1.8742,
+      "step": 10165
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5721410707849374,
+      "learning_rate": 5.856921921631889e-06,
+      "loss": 1.9884,
+      "step": 10166
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6392445124683205,
+      "learning_rate": 5.852904824112821e-06,
+      "loss": 1.8701,
+      "step": 10167
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5904792149223216,
+      "learning_rate": 5.848888922025553e-06,
+      "loss": 1.8669,
+      "step": 10168
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5273618143885774,
+      "learning_rate": 5.844874215620818e-06,
+      "loss": 1.9654,
+      "step": 10169
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5493000320003958,
+      "learning_rate": 5.8408607051492924e-06,
+      "loss": 2.0672,
+      "step": 10170
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5757970836402747,
+      "learning_rate": 5.836848390861533e-06,
+      "loss": 1.8803,
+      "step": 10171
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5844356893432947,
+      "learning_rate": 5.832837273008051e-06,
+      "loss": 1.8375,
+      "step": 10172
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.567491091008223,
+      "learning_rate": 5.828827351839281e-06,
+      "loss": 2.0416,
+      "step": 10173
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5832385356753257,
+      "learning_rate": 5.824818627605577e-06,
+      "loss": 1.8882,
+      "step": 10174
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5536219733833632,
+      "learning_rate": 5.820811100557217e-06,
+      "loss": 1.9457,
+      "step": 10175
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5688684334503991,
+      "learning_rate": 5.816804770944412e-06,
+      "loss": 1.8933,
+      "step": 10176
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5751088571899027,
+      "learning_rate": 5.812799639017291e-06,
+      "loss": 1.8959,
+      "step": 10177
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5346562415783215,
+      "learning_rate": 5.808795705025913e-06,
+      "loss": 2.0489,
+      "step": 10178
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.567526388646483,
+      "learning_rate": 5.8047929692202575e-06,
+      "loss": 1.8752,
+      "step": 10179
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5950583902507007,
+      "learning_rate": 5.800791431850233e-06,
+      "loss": 1.8292,
+      "step": 10180
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5553104342275026,
+      "learning_rate": 5.796791093165671e-06,
+      "loss": 1.9726,
+      "step": 10181
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5703071925588128,
+      "learning_rate": 5.792791953416329e-06,
+      "loss": 2.0853,
+      "step": 10182
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5771191189014386,
+      "learning_rate": 5.788794012851892e-06,
+      "loss": 1.9173,
+      "step": 10183
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5711228591058355,
+      "learning_rate": 5.784797271721967e-06,
+      "loss": 1.8595,
+      "step": 10184
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5812246908652152,
+      "learning_rate": 5.780801730276086e-06,
+      "loss": 1.8691,
+      "step": 10185
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6215626897561156,
+      "learning_rate": 5.776807388763706e-06,
+      "loss": 2.0689,
+      "step": 10186
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5774305712186452,
+      "learning_rate": 5.772814247434213e-06,
+      "loss": 1.8735,
+      "step": 10187
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5445873376729672,
+      "learning_rate": 5.768822306536914e-06,
+      "loss": 1.9326,
+      "step": 10188
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5452877879908641,
+      "learning_rate": 5.764831566321041e-06,
+      "loss": 1.8502,
+      "step": 10189
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5603654199093803,
+      "learning_rate": 5.760842027035762e-06,
+      "loss": 2.0599,
+      "step": 10190
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5841340235052108,
+      "learning_rate": 5.756853688930136e-06,
+      "loss": 1.8113,
+      "step": 10191
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5566513136453618,
+      "learning_rate": 5.7528665522531945e-06,
+      "loss": 1.8716,
+      "step": 10192
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5279703432641484,
+      "learning_rate": 5.7488806172538715e-06,
+      "loss": 2.038,
+      "step": 10193
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5368760779345237,
+      "learning_rate": 5.7448958841810005e-06,
+      "loss": 1.9449,
+      "step": 10194
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5515294187408957,
+      "learning_rate": 5.74091235328339e-06,
+      "loss": 1.8733,
+      "step": 10195
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5694584250660059,
+      "learning_rate": 5.736930024809747e-06,
+      "loss": 1.895,
+      "step": 10196
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5653134365590797,
+      "learning_rate": 5.732948899008683e-06,
+      "loss": 1.8986,
+      "step": 10197
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5749694234134166,
+      "learning_rate": 5.728968976128784e-06,
+      "loss": 2.0621,
+      "step": 10198
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5923613021910564,
+      "learning_rate": 5.724990256418505e-06,
+      "loss": 1.8973,
+      "step": 10199
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5283108617445471,
+      "learning_rate": 5.721012740126275e-06,
+      "loss": 1.9051,
+      "step": 10200
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5701399239105553,
+      "learning_rate": 5.717036427500428e-06,
+      "loss": 1.8832,
+      "step": 10201
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.597149923971932,
+      "learning_rate": 5.713061318789198e-06,
+      "loss": 2.0832,
+      "step": 10202
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5598766534175446,
+      "learning_rate": 5.709087414240797e-06,
+      "loss": 1.7911,
+      "step": 10203
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5507342236312438,
+      "learning_rate": 5.7051147141033105e-06,
+      "loss": 1.8805,
+      "step": 10204
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5412444472953101,
+      "learning_rate": 5.701143218624768e-06,
+      "loss": 1.8313,
+      "step": 10205
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5666461445034541,
+      "learning_rate": 5.697172928053154e-06,
+      "loss": 2.1236,
+      "step": 10206
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6036745699692719,
+      "learning_rate": 5.693203842636324e-06,
+      "loss": 1.8774,
+      "step": 10207
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6092923342936672,
+      "learning_rate": 5.689235962622086e-06,
+      "loss": 1.8851,
+      "step": 10208
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5542305622028801,
+      "learning_rate": 5.685269288258191e-06,
+      "loss": 1.9141,
+      "step": 10209
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5886553949991596,
+      "learning_rate": 5.6813038197922745e-06,
+      "loss": 2.0935,
+      "step": 10210
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6211429523901579,
+      "learning_rate": 5.677339557471923e-06,
+      "loss": 1.87,
+      "step": 10211
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5520842572665948,
+      "learning_rate": 5.673376501544641e-06,
+      "loss": 1.9255,
+      "step": 10212
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6052437487103176,
+      "learning_rate": 5.669414652257857e-06,
+      "loss": 1.9263,
+      "step": 10213
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5622522547144251,
+      "learning_rate": 5.665454009858939e-06,
+      "loss": 2.0954,
+      "step": 10214
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5889837654984621,
+      "learning_rate": 5.66149457459515e-06,
+      "loss": 1.8227,
+      "step": 10215
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6061514488626564,
+      "learning_rate": 5.6575363467136904e-06,
+      "loss": 1.8723,
+      "step": 10216
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5612073265682498,
+      "learning_rate": 5.6535793264617136e-06,
+      "loss": 1.9035,
+      "step": 10217
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5771406306708882,
+      "learning_rate": 5.649623514086247e-06,
+      "loss": 2.045,
+      "step": 10218
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.584967315559342,
+      "learning_rate": 5.6456689098342815e-06,
+      "loss": 1.9497,
+      "step": 10219
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6024156814405159,
+      "learning_rate": 5.641715513952711e-06,
+      "loss": 1.7953,
+      "step": 10220
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5769820008209147,
+      "learning_rate": 5.63776332668837e-06,
+      "loss": 1.8681,
+      "step": 10221
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5515671326118492,
+      "learning_rate": 5.633812348288006e-06,
+      "loss": 2.0348,
+      "step": 10222
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6110101064489594,
+      "learning_rate": 5.629862578998296e-06,
+      "loss": 1.8227,
+      "step": 10223
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6265852542030033,
+      "learning_rate": 5.625914019065836e-06,
+      "loss": 1.8829,
+      "step": 10224
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5913058524612007,
+      "learning_rate": 5.621966668737158e-06,
+      "loss": 1.9123,
+      "step": 10225
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5452853986978798,
+      "learning_rate": 5.618020528258708e-06,
+      "loss": 2.0632,
+      "step": 10226
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5585681239462417,
+      "learning_rate": 5.614075597876858e-06,
+      "loss": 1.8609,
+      "step": 10227
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5688685312968452,
+      "learning_rate": 5.61013187783791e-06,
+      "loss": 1.8949,
+      "step": 10228
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.592374676559332,
+      "learning_rate": 5.606189368388082e-06,
+      "loss": 1.8682,
+      "step": 10229
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6009408796425828,
+      "learning_rate": 5.602248069773525e-06,
+      "loss": 2.0481,
+      "step": 10230
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5492701933860198,
+      "learning_rate": 5.598307982240308e-06,
+      "loss": 1.9623,
+      "step": 10231
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5834691019948706,
+      "learning_rate": 5.594369106034431e-06,
+      "loss": 1.8692,
+      "step": 10232
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5432090300988969,
+      "learning_rate": 5.590431441401805e-06,
+      "loss": 1.8241,
+      "step": 10233
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5942845273288151,
+      "learning_rate": 5.586494988588287e-06,
+      "loss": 2.0979,
+      "step": 10234
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6139610389185082,
+      "learning_rate": 5.582559747839636e-06,
+      "loss": 1.8265,
+      "step": 10235
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.538122442200722,
+      "learning_rate": 5.578625719401548e-06,
+      "loss": 1.8358,
+      "step": 10236
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.580654600951496,
+      "learning_rate": 5.574692903519652e-06,
+      "loss": 1.9259,
+      "step": 10237
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6225132240193826,
+      "learning_rate": 5.570761300439462e-06,
+      "loss": 2.0463,
+      "step": 10238
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5581917644171344,
+      "learning_rate": 5.566830910406468e-06,
+      "loss": 1.9289,
+      "step": 10239
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5544339373049293,
+      "learning_rate": 5.562901733666062e-06,
+      "loss": 1.8166,
+      "step": 10240
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5939519999422949,
+      "learning_rate": 5.558973770463535e-06,
+      "loss": 1.8533,
+      "step": 10241
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5787973258444494,
+      "learning_rate": 5.555047021044152e-06,
+      "loss": 2.0659,
+      "step": 10242
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5746496768888303,
+      "learning_rate": 5.551121485653068e-06,
+      "loss": 1.9367,
+      "step": 10243
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5506921545418012,
+      "learning_rate": 5.547197164535356e-06,
+      "loss": 1.8886,
+      "step": 10244
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5731905075281262,
+      "learning_rate": 5.543274057936055e-06,
+      "loss": 1.8807,
+      "step": 10245
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5398101599746967,
+      "learning_rate": 5.539352166100068e-06,
+      "loss": 2.0526,
+      "step": 10246
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5745703364437389,
+      "learning_rate": 5.535431489272278e-06,
+      "loss": 1.8848,
+      "step": 10247
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5645346973901552,
+      "learning_rate": 5.531512027697469e-06,
+      "loss": 1.8675,
+      "step": 10248
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5659046088054787,
+      "learning_rate": 5.527593781620333e-06,
+      "loss": 1.832,
+      "step": 10249
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5266341694714294,
+      "learning_rate": 5.523676751285514e-06,
+      "loss": 2.0943,
+      "step": 10250
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5512121494694953,
+      "learning_rate": 5.519760936937576e-06,
+      "loss": 1.8815,
+      "step": 10251
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5818293062018851,
+      "learning_rate": 5.515846338820974e-06,
+      "loss": 1.8412,
+      "step": 10252
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5774122436165149,
+      "learning_rate": 5.511932957180143e-06,
+      "loss": 1.8504,
+      "step": 10253
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55915741770971,
+      "learning_rate": 5.508020792259388e-06,
+      "loss": 2.0508,
+      "step": 10254
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5837820683469448,
+      "learning_rate": 5.504109844302965e-06,
+      "loss": 1.8319,
+      "step": 10255
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5094465647589013,
+      "learning_rate": 5.500200113555071e-06,
+      "loss": 1.874,
+      "step": 10256
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5567812950145951,
+      "learning_rate": 5.496291600259781e-06,
+      "loss": 1.8742,
+      "step": 10257
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5528570780184897,
+      "learning_rate": 5.49238430466113e-06,
+      "loss": 2.035,
+      "step": 10258
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5731744766953656,
+      "learning_rate": 5.488478227003066e-06,
+      "loss": 1.8885,
+      "step": 10259
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.554468470868587,
+      "learning_rate": 5.484573367529464e-06,
+      "loss": 1.8441,
+      "step": 10260
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.584561113207535,
+      "learning_rate": 5.480669726484117e-06,
+      "loss": 1.8617,
+      "step": 10261
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5253242205105245,
+      "learning_rate": 5.476767304110744e-06,
+      "loss": 2.0885,
+      "step": 10262
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5727124144099541,
+      "learning_rate": 5.472866100652988e-06,
+      "loss": 1.8749,
+      "step": 10263
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5599191396672536,
+      "learning_rate": 5.468966116354432e-06,
+      "loss": 1.8963,
+      "step": 10264
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5746913336673557,
+      "learning_rate": 5.4650673514585525e-06,
+      "loss": 1.8766,
+      "step": 10265
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5809373332848577,
+      "learning_rate": 5.4611698062087685e-06,
+      "loss": 2.0741,
+      "step": 10266
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5537955606503521,
+      "learning_rate": 5.457273480848421e-06,
+      "loss": 1.8758,
+      "step": 10267
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5621255490953787,
+      "learning_rate": 5.453378375620771e-06,
+      "loss": 1.9321,
+      "step": 10268
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6335233778371688,
+      "learning_rate": 5.449484490769013e-06,
+      "loss": 1.8508,
+      "step": 10269
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.565445242133362,
+      "learning_rate": 5.44559182653625e-06,
+      "loss": 2.053,
+      "step": 10270
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5887890407133843,
+      "learning_rate": 5.441700383165524e-06,
+      "loss": 1.8463,
+      "step": 10271
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5899514262415477,
+      "learning_rate": 5.43781016089979e-06,
+      "loss": 1.8304,
+      "step": 10272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5901655907447991,
+      "learning_rate": 5.4339211599819305e-06,
+      "loss": 1.8753,
+      "step": 10273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5496382704505604,
+      "learning_rate": 5.430033380654751e-06,
+      "loss": 2.0473,
+      "step": 10274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5809429077436236,
+      "learning_rate": 5.426146823160985e-06,
+      "loss": 1.849,
+      "step": 10275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5957383967825098,
+      "learning_rate": 5.422261487743283e-06,
+      "loss": 1.8604,
+      "step": 10276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6129061434008679,
+      "learning_rate": 5.4183773746442236e-06,
+      "loss": 1.883,
+      "step": 10277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5565066566791456,
+      "learning_rate": 5.414494484106308e-06,
+      "loss": 2.0398,
+      "step": 10278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5982237011439069,
+      "learning_rate": 5.410612816371971e-06,
+      "loss": 1.8983,
+      "step": 10279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6127915927204217,
+      "learning_rate": 5.4067323716835345e-06,
+      "loss": 1.8712,
+      "step": 10280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5821930123210192,
+      "learning_rate": 5.402853150283294e-06,
+      "loss": 1.9474,
+      "step": 10281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.656713435681311,
+      "learning_rate": 5.398975152413438e-06,
+      "loss": 2.0953,
+      "step": 10282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5840902521479423,
+      "learning_rate": 5.395098378316088e-06,
+      "loss": 1.8489,
+      "step": 10283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.55350460863578,
+      "learning_rate": 5.391222828233283e-06,
+      "loss": 1.8798,
+      "step": 10284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6144365686094893,
+      "learning_rate": 5.387348502406994e-06,
+      "loss": 1.818,
+      "step": 10285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5814257581474069,
+      "learning_rate": 5.383475401079108e-06,
+      "loss": 2.0623,
+      "step": 10286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5293098189172737,
+      "learning_rate": 5.3796035244914475e-06,
+      "loss": 1.961,
+      "step": 10287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5636973968520742,
+      "learning_rate": 5.3757328728857244e-06,
+      "loss": 1.8499,
+      "step": 10288
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.560371910246276,
+      "learning_rate": 5.371863446503628e-06,
+      "loss": 1.788,
+      "step": 10289
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6198146161406368,
+      "learning_rate": 5.367995245586735e-06,
+      "loss": 2.0474,
+      "step": 10290
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5998493297919627,
+      "learning_rate": 5.3641282703765335e-06,
+      "loss": 1.8854,
+      "step": 10291
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5513247079457374,
+      "learning_rate": 5.360262521114487e-06,
+      "loss": 1.8743,
+      "step": 10292
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5830003159871909,
+      "learning_rate": 5.356397998041926e-06,
+      "loss": 1.9419,
+      "step": 10293
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6324548413570229,
+      "learning_rate": 5.35253470140013e-06,
+      "loss": 2.0546,
+      "step": 10294
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6026276940550389,
+      "learning_rate": 5.348672631430318e-06,
+      "loss": 1.8717,
+      "step": 10295
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5514225263636126,
+      "learning_rate": 5.344811788373591e-06,
+      "loss": 1.8382,
+      "step": 10296
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5671284110691961,
+      "learning_rate": 5.340952172471017e-06,
+      "loss": 1.8905,
+      "step": 10297
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5675155609848035,
+      "learning_rate": 5.3370937839635645e-06,
+      "loss": 2.039,
+      "step": 10298
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5305825517509315,
+      "learning_rate": 5.3332366230921136e-06,
+      "loss": 1.9334,
+      "step": 10299
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.54277746848529,
+      "learning_rate": 5.329380690097505e-06,
+      "loss": 1.8432,
+      "step": 10300
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5508172745350512,
+      "learning_rate": 5.325525985220461e-06,
+      "loss": 1.83,
+      "step": 10301
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5645998126572026,
+      "learning_rate": 5.32167250870165e-06,
+      "loss": 2.0421,
+      "step": 10302
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5578295076236371,
+      "learning_rate": 5.317820260781678e-06,
+      "loss": 1.8688,
+      "step": 10303
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5759097367056776,
+      "learning_rate": 5.313969241701039e-06,
+      "loss": 1.8703,
+      "step": 10304
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.574778720207803,
+      "learning_rate": 5.310119451700168e-06,
+      "loss": 1.8642,
+      "step": 10305
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5367015864686491,
+      "learning_rate": 5.30627089101943e-06,
+      "loss": 1.9388,
+      "step": 10306
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6049783532682824,
+      "learning_rate": 5.302423559899106e-06,
+      "loss": 2.0682,
+      "step": 10307
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6111634323387574,
+      "learning_rate": 5.298577458579396e-06,
+      "loss": 1.8124,
+      "step": 10308
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5836421462073448,
+      "learning_rate": 5.29473258730043e-06,
+      "loss": 1.8467,
+      "step": 10309
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5305833433133336,
+      "learning_rate": 5.290888946302255e-06,
+      "loss": 2.0496,
+      "step": 10310
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5445315896367134,
+      "learning_rate": 5.287046535824863e-06,
+      "loss": 1.9016,
+      "step": 10311
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.558590605602343,
+      "learning_rate": 5.283205356108134e-06,
+      "loss": 1.9273,
+      "step": 10312
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6059423219905423,
+      "learning_rate": 5.279365407391889e-06,
+      "loss": 1.8909,
+      "step": 10313
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5661399926850279,
+      "learning_rate": 5.27552668991588e-06,
+      "loss": 2.0573,
+      "step": 10314
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5675871387087331,
+      "learning_rate": 5.271689203919769e-06,
+      "loss": 1.8707,
+      "step": 10315
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5689459248486165,
+      "learning_rate": 5.267852949643146e-06,
+      "loss": 1.8222,
+      "step": 10316
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5765163887617202,
+      "learning_rate": 5.2640179273255235e-06,
+      "loss": 1.9105,
+      "step": 10317
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6306191073475599,
+      "learning_rate": 5.260184137206342e-06,
+      "loss": 1.9107,
+      "step": 10318
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6997031386173649,
+      "learning_rate": 5.256351579524957e-06,
+      "loss": 2.0389,
+      "step": 10319
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5807543663854395,
+      "learning_rate": 5.252520254520652e-06,
+      "loss": 1.8686,
+      "step": 10320
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5790411930285899,
+      "learning_rate": 5.248690162432632e-06,
+      "loss": 1.8863,
+      "step": 10321
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5862191611179373,
+      "learning_rate": 5.244861303500026e-06,
+      "loss": 2.0881,
+      "step": 10322
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5560918722047492,
+      "learning_rate": 5.241033677961884e-06,
+      "loss": 1.8146,
+      "step": 10323
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5430021711267736,
+      "learning_rate": 5.237207286057183e-06,
+      "loss": 1.9507,
+      "step": 10324
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5560114458057376,
+      "learning_rate": 5.2333821280248165e-06,
+      "loss": 1.8663,
+      "step": 10325
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6010762322339794,
+      "learning_rate": 5.229558204103615e-06,
+      "loss": 2.0784,
+      "step": 10326
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5592874855314791,
+      "learning_rate": 5.225735514532301e-06,
+      "loss": 1.7995,
+      "step": 10327
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6498090355589624,
+      "learning_rate": 5.221914059549559e-06,
+      "loss": 1.9245,
+      "step": 10328
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5893423299268513,
+      "learning_rate": 5.218093839393973e-06,
+      "loss": 1.8647,
+      "step": 10329
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5894203492278574,
+      "learning_rate": 5.214274854304055e-06,
+      "loss": 1.9527,
+      "step": 10330
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6373028922919814,
+      "learning_rate": 5.210457104518241e-06,
+      "loss": 2.0314,
+      "step": 10331
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6335270602257104,
+      "learning_rate": 5.206640590274886e-06,
+      "loss": 1.8584,
+      "step": 10332
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5496610046732414,
+      "learning_rate": 5.20282531181227e-06,
+      "loss": 1.8243,
+      "step": 10333
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5998920978285385,
+      "learning_rate": 5.199011269368609e-06,
+      "loss": 2.0822,
+      "step": 10334
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5983489863751883,
+      "learning_rate": 5.195198463182005e-06,
+      "loss": 1.9002,
+      "step": 10335
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5707070103568428,
+      "learning_rate": 5.191386893490527e-06,
+      "loss": 1.8556,
+      "step": 10336
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5744978603061615,
+      "learning_rate": 5.187576560532145e-06,
+      "loss": 1.914,
+      "step": 10337
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5672192672954681,
+      "learning_rate": 5.1837674645447395e-06,
+      "loss": 1.8454,
+      "step": 10338
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5913556339760355,
+      "learning_rate": 5.179959605766152e-06,
+      "loss": 2.0527,
+      "step": 10339
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.680219680864323,
+      "learning_rate": 5.176152984434102e-06,
+      "loss": 1.8865,
+      "step": 10340
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.584683757062755,
+      "learning_rate": 5.172347600786251e-06,
+      "loss": 1.842,
+      "step": 10341
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5760081367948665,
+      "learning_rate": 5.16854345506021e-06,
+      "loss": 2.0348,
+      "step": 10342
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5717842824326596,
+      "learning_rate": 5.16474054749346e-06,
+      "loss": 1.9429,
+      "step": 10343
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.68775606948644,
+      "learning_rate": 5.16093887832344e-06,
+      "loss": 1.8564,
+      "step": 10344
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6043264048554993,
+      "learning_rate": 5.157138447787518e-06,
+      "loss": 1.9089,
+      "step": 10345
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6394038609420872,
+      "learning_rate": 5.153339256122946e-06,
+      "loss": 2.0669,
+      "step": 10346
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.636460306375277,
+      "learning_rate": 5.149541303566951e-06,
+      "loss": 1.8703,
+      "step": 10347
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5820783354923312,
+      "learning_rate": 5.145744590356633e-06,
+      "loss": 1.8293,
+      "step": 10348
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5801644385919899,
+      "learning_rate": 5.141949116729036e-06,
+      "loss": 1.9309,
+      "step": 10349
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5601047242978071,
+      "learning_rate": 5.138154882921148e-06,
+      "loss": 1.8695,
+      "step": 10350
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6222478712100765,
+      "learning_rate": 5.134361889169839e-06,
+      "loss": 2.0389,
+      "step": 10351
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.577659493208015,
+      "learning_rate": 5.130570135711929e-06,
+      "loss": 1.8828,
+      "step": 10352
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5373195670886877,
+      "learning_rate": 5.126779622784147e-06,
+      "loss": 1.8101,
+      "step": 10353
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5292527344723461,
+      "learning_rate": 5.122990350623158e-06,
+      "loss": 2.0719,
+      "step": 10354
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5090977627271175,
+      "learning_rate": 5.119202319465538e-06,
+      "loss": 1.9228,
+      "step": 10355
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5594528704275471,
+      "learning_rate": 5.115415529547787e-06,
+      "loss": 1.844,
+      "step": 10356
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6029495574235477,
+      "learning_rate": 5.111629981106334e-06,
+      "loss": 1.8856,
+      "step": 10357
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6052338691824617,
+      "learning_rate": 5.107845674377526e-06,
+      "loss": 2.0608,
+      "step": 10358
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5491371500563812,
+      "learning_rate": 5.104062609597629e-06,
+      "loss": 1.8839,
+      "step": 10359
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5773457593640353,
+      "learning_rate": 5.100280787002839e-06,
+      "loss": 1.8259,
+      "step": 10360
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5649773987155331,
+      "learning_rate": 5.0965002068292685e-06,
+      "loss": 1.9423,
+      "step": 10361
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5785805466704477,
+      "learning_rate": 5.092720869312959e-06,
+      "loss": 1.873,
+      "step": 10362
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5676805001921199,
+      "learning_rate": 5.088942774689864e-06,
+      "loss": 2.0345,
+      "step": 10363
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5715183490442326,
+      "learning_rate": 5.08516592319587e-06,
+      "loss": 1.8918,
+      "step": 10364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6174112879230108,
+      "learning_rate": 5.081390315066778e-06,
+      "loss": 1.8792,
+      "step": 10365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5993335732602378,
+      "learning_rate": 5.077615950538317e-06,
+      "loss": 2.0485,
+      "step": 10366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5683376129831172,
+      "learning_rate": 5.073842829846134e-06,
+      "loss": 1.8389,
+      "step": 10367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.542394646173869,
+      "learning_rate": 5.070070953225803e-06,
+      "loss": 1.8919,
+      "step": 10368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6262665714458804,
+      "learning_rate": 5.066300320912817e-06,
+      "loss": 1.8945,
+      "step": 10369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.626755441614926,
+      "learning_rate": 5.062530933142593e-06,
+      "loss": 1.8764,
+      "step": 10370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5954528902464622,
+      "learning_rate": 5.058762790150465e-06,
+      "loss": 2.007,
+      "step": 10371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5930677156983201,
+      "learning_rate": 5.054995892171699e-06,
+      "loss": 1.869,
+      "step": 10372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5780747408663028,
+      "learning_rate": 5.051230239441482e-06,
+      "loss": 1.8801,
+      "step": 10373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5585698091130455,
+      "learning_rate": 5.047465832194897e-06,
+      "loss": 1.9302,
+      "step": 10374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5935062474674005,
+      "learning_rate": 5.043702670666995e-06,
+      "loss": 2.0592,
+      "step": 10375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6037233417699033,
+      "learning_rate": 5.039940755092726e-06,
+      "loss": 1.8571,
+      "step": 10376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5804962300112098,
+      "learning_rate": 5.03618008570694e-06,
+      "loss": 1.8558,
+      "step": 10377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5873479664586692,
+      "learning_rate": 5.03242066274445e-06,
+      "loss": 2.1213,
+      "step": 10378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5398112015742577,
+      "learning_rate": 5.028662486439967e-06,
+      "loss": 1.816,
+      "step": 10379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5314620328229054,
+      "learning_rate": 5.024905557028131e-06,
+      "loss": 1.9437,
+      "step": 10380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.562245946522114,
+      "learning_rate": 5.0211498747435056e-06,
+      "loss": 1.8473,
+      "step": 10381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5601921681062941,
+      "learning_rate": 5.017395439820554e-06,
+      "loss": 1.8853,
+      "step": 10382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6182870045260961,
+      "learning_rate": 5.013642252493705e-06,
+      "loss": 2.0214,
+      "step": 10383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6224546570590392,
+      "learning_rate": 5.009890312997284e-06,
+      "loss": 1.8768,
+      "step": 10384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.640378831898014,
+      "learning_rate": 5.006139621565517e-06,
+      "loss": 1.8567,
+      "step": 10385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5551483050093223,
+      "learning_rate": 5.002390178432603e-06,
+      "loss": 1.9441,
+      "step": 10386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5719679927224324,
+      "learning_rate": 4.998641983832617e-06,
+      "loss": 2.0436,
+      "step": 10387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5670876010467094,
+      "learning_rate": 4.9948950379995725e-06,
+      "loss": 1.8826,
+      "step": 10388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5727792955927588,
+      "learning_rate": 4.991149341167425e-06,
+      "loss": 1.8285,
+      "step": 10389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5724534693483202,
+      "learning_rate": 4.987404893570019e-06,
+      "loss": 2.0774,
+      "step": 10390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.603741929050136,
+      "learning_rate": 4.983661695441133e-06,
+      "loss": 1.9009,
+      "step": 10391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5281196676362316,
+      "learning_rate": 4.979919747014486e-06,
+      "loss": 1.9007,
+      "step": 10392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5728356756697508,
+      "learning_rate": 4.976179048523685e-06,
+      "loss": 1.8816,
+      "step": 10393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6421015162427863,
+      "learning_rate": 4.972439600202297e-06,
+      "loss": 1.8389,
+      "step": 10394
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.637883448070874,
+      "learning_rate": 4.968701402283773e-06,
+      "loss": 2.0585,
+      "step": 10395
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5505554823735205,
+      "learning_rate": 4.964964455001505e-06,
+      "loss": 1.8754,
+      "step": 10396
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5519493623918355,
+      "learning_rate": 4.961228758588824e-06,
+      "loss": 1.8408,
+      "step": 10397
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6057503850916376,
+      "learning_rate": 4.957494313278949e-06,
+      "loss": 2.0569,
+      "step": 10398
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5477108863437177,
+      "learning_rate": 4.953761119305031e-06,
+      "loss": 1.9866,
+      "step": 10399
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.585291830808401,
+      "learning_rate": 4.950029176900172e-06,
+      "loss": 1.8872,
+      "step": 10400
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.578500486052067,
+      "learning_rate": 4.9462984862973545e-06,
+      "loss": 1.8129,
+      "step": 10401
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5874258755013646,
+      "learning_rate": 4.942569047729506e-06,
+      "loss": 1.84,
+      "step": 10402
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5399540475919716,
+      "learning_rate": 4.938840861429467e-06,
+      "loss": 2.0339,
+      "step": 10403
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5579303258079316,
+      "learning_rate": 4.9351139276300085e-06,
+      "loss": 1.8504,
+      "step": 10404
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5244675783005178,
+      "learning_rate": 4.931388246563815e-06,
+      "loss": 1.8727,
+      "step": 10405
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5791870072121792,
+      "learning_rate": 4.927663818463496e-06,
+      "loss": 1.8567,
+      "step": 10406
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6195623670387986,
+      "learning_rate": 4.923940643561583e-06,
+      "loss": 2.0648,
+      "step": 10407
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5944153120036526,
+      "learning_rate": 4.920218722090533e-06,
+      "loss": 1.8782,
+      "step": 10408
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5413637079012079,
+      "learning_rate": 4.916498054282717e-06,
+      "loss": 1.7945,
+      "step": 10409
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5368396943795246,
+      "learning_rate": 4.912778640370433e-06,
+      "loss": 2.0301,
+      "step": 10410
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5278498142011506,
+      "learning_rate": 4.9090604805859e-06,
+      "loss": 1.9584,
+      "step": 10411
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5501669685041819,
+      "learning_rate": 4.905343575161256e-06,
+      "loss": 1.8437,
+      "step": 10412
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5233635488254041,
+      "learning_rate": 4.901627924328561e-06,
+      "loss": 1.821,
+      "step": 10413
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5766796825557069,
+      "learning_rate": 4.897913528319803e-06,
+      "loss": 1.8523,
+      "step": 10414
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.551979371801613,
+      "learning_rate": 4.894200387366884e-06,
+      "loss": 2.0413,
+      "step": 10415
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5836250799455329,
+      "learning_rate": 4.890488501701634e-06,
+      "loss": 1.8462,
+      "step": 10416
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5950701799724265,
+      "learning_rate": 4.886777871555798e-06,
+      "loss": 1.9546,
+      "step": 10417
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5826279314693502,
+      "learning_rate": 4.883068497161045e-06,
+      "loss": 1.8413,
+      "step": 10418
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5561326880613706,
+      "learning_rate": 4.879360378748968e-06,
+      "loss": 2.0547,
+      "step": 10419
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5741954912855409,
+      "learning_rate": 4.875653516551087e-06,
+      "loss": 1.8776,
+      "step": 10420
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5858028268982356,
+      "learning_rate": 4.871947910798818e-06,
+      "loss": 1.8612,
+      "step": 10421
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5646329302176061,
+      "learning_rate": 4.868243561723535e-06,
+      "loss": 1.8294,
+      "step": 10422
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5745474944816826,
+      "learning_rate": 4.864540469556517e-06,
+      "loss": 2.1368,
+      "step": 10423
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5901469417854697,
+      "learning_rate": 4.860838634528941e-06,
+      "loss": 1.8662,
+      "step": 10424
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.602480374412365,
+      "learning_rate": 4.857138056871951e-06,
+      "loss": 1.8809,
+      "step": 10425
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5743160014398665,
+      "learning_rate": 4.853438736816582e-06,
+      "loss": 1.8233,
+      "step": 10426
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5670704660474511,
+      "learning_rate": 4.849740674593795e-06,
+      "loss": 2.0468,
+      "step": 10427
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5857919946391883,
+      "learning_rate": 4.846043870434483e-06,
+      "loss": 1.8831,
+      "step": 10428
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5751834023916403,
+      "learning_rate": 4.842348324569435e-06,
+      "loss": 1.8539,
+      "step": 10429
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5440895214545866,
+      "learning_rate": 4.838654037229395e-06,
+      "loss": 1.9342,
+      "step": 10430
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5792815689905975,
+      "learning_rate": 4.83496100864502e-06,
+      "loss": 2.065,
+      "step": 10431
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5886922179406396,
+      "learning_rate": 4.831269239046851e-06,
+      "loss": 1.8552,
+      "step": 10432
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5836859869292756,
+      "learning_rate": 4.827578728665408e-06,
+      "loss": 1.8887,
+      "step": 10433
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5818632893246573,
+      "learning_rate": 4.8238894777311e-06,
+      "loss": 1.8484,
+      "step": 10434
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5230721166385053,
+      "learning_rate": 4.820201486474246e-06,
+      "loss": 1.9947,
+      "step": 10435
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5177893117169098,
+      "learning_rate": 4.816514755125126e-06,
+      "loss": 1.961,
+      "step": 10436
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5637150078055186,
+      "learning_rate": 4.8128292839139e-06,
+      "loss": 1.8317,
+      "step": 10437
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5691140861968128,
+      "learning_rate": 4.809145073070667e-06,
+      "loss": 1.8061,
+      "step": 10438
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5718479059451684,
+      "learning_rate": 4.805462122825463e-06,
+      "loss": 2.0331,
+      "step": 10439
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5818875503730755,
+      "learning_rate": 4.801780433408215e-06,
+      "loss": 1.9082,
+      "step": 10440
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.599549074130894,
+      "learning_rate": 4.798100005048791e-06,
+      "loss": 1.8758,
+      "step": 10441
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5083303564810173,
+      "learning_rate": 4.7944208379769766e-06,
+      "loss": 1.9687,
+      "step": 10442
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5749447036694735,
+      "learning_rate": 4.790742932422468e-06,
+      "loss": 2.0679,
+      "step": 10443
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5390275564723571,
+      "learning_rate": 4.787066288614914e-06,
+      "loss": 1.8556,
+      "step": 10444
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5713121176561802,
+      "learning_rate": 4.783390906783841e-06,
+      "loss": 1.8419,
+      "step": 10445
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5581697193853977,
+      "learning_rate": 4.77971678715872e-06,
+      "loss": 1.8443,
+      "step": 10446
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6283991937112697,
+      "learning_rate": 4.77604392996896e-06,
+      "loss": 2.082,
+      "step": 10447
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5738822714978264,
+      "learning_rate": 4.772372335443851e-06,
+      "loss": 1.9904,
+      "step": 10448
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5563085336015282,
+      "learning_rate": 4.768702003812636e-06,
+      "loss": 1.8566,
+      "step": 10449
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5285736616919252,
+      "learning_rate": 4.765032935304467e-06,
+      "loss": 1.8808,
+      "step": 10450
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.621729538054919,
+      "learning_rate": 4.761365130148418e-06,
+      "loss": 2.0317,
+      "step": 10451
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5617033943441008,
+      "learning_rate": 4.757698588573487e-06,
+      "loss": 1.8598,
+      "step": 10452
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5682788021599456,
+      "learning_rate": 4.754033310808589e-06,
+      "loss": 1.8582,
+      "step": 10453
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5658360035958425,
+      "learning_rate": 4.750369297082566e-06,
+      "loss": 1.8712,
+      "step": 10454
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5395946876281712,
+      "learning_rate": 4.746706547624174e-06,
+      "loss": 2.0595,
+      "step": 10455
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5644131661364578,
+      "learning_rate": 4.743045062662096e-06,
+      "loss": 1.877,
+      "step": 10456
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5340837464456463,
+      "learning_rate": 4.739384842424935e-06,
+      "loss": 1.848,
+      "step": 10457
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6144970831514104,
+      "learning_rate": 4.735725887141207e-06,
+      "loss": 1.8686,
+      "step": 10458
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5434290523420635,
+      "learning_rate": 4.7320681970393605e-06,
+      "loss": 2.0927,
+      "step": 10459
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5851139938577856,
+      "learning_rate": 4.72841177234776e-06,
+      "loss": 1.8149,
+      "step": 10460
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5356937969411537,
+      "learning_rate": 4.72475661329469e-06,
+      "loss": 1.8551,
+      "step": 10461
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5821718114167291,
+      "learning_rate": 4.721102720108356e-06,
+      "loss": 1.8609,
+      "step": 10462
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5778070883959584,
+      "learning_rate": 4.717450093016889e-06,
+      "loss": 2.0455,
+      "step": 10463
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.545703023020577,
+      "learning_rate": 4.713798732248334e-06,
+      "loss": 1.8449,
+      "step": 10464
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5576551028034836,
+      "learning_rate": 4.710148638030665e-06,
+      "loss": 1.8547,
+      "step": 10465
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5751212877928249,
+      "learning_rate": 4.7064998105917666e-06,
+      "loss": 1.8656,
+      "step": 10466
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5718368178092039,
+      "learning_rate": 4.702852250159451e-06,
+      "loss": 2.1307,
+      "step": 10467
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5622218784145107,
+      "learning_rate": 4.699205956961455e-06,
+      "loss": 1.9162,
+      "step": 10468
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5472202939953829,
+      "learning_rate": 4.695560931225429e-06,
+      "loss": 1.83,
+      "step": 10469
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5559056806275278,
+      "learning_rate": 4.691917173178953e-06,
+      "loss": 1.8336,
+      "step": 10470
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5554318272060033,
+      "learning_rate": 4.688274683049501e-06,
+      "loss": 2.0599,
+      "step": 10471
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5631269649243916,
+      "learning_rate": 4.684633461064511e-06,
+      "loss": 1.8143,
+      "step": 10472
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5213349474067015,
+      "learning_rate": 4.680993507451317e-06,
+      "loss": 1.9708,
+      "step": 10473
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.589386748457972,
+      "learning_rate": 4.677354822437161e-06,
+      "loss": 1.8629,
+      "step": 10474
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.551180884404929,
+      "learning_rate": 4.673717406249239e-06,
+      "loss": 2.0454,
+      "step": 10475
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5598026545608378,
+      "learning_rate": 4.6700812591146335e-06,
+      "loss": 1.9079,
+      "step": 10476
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5759549581278354,
+      "learning_rate": 4.666446381260378e-06,
+      "loss": 1.8469,
+      "step": 10477
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5898736000653801,
+      "learning_rate": 4.6628127729134105e-06,
+      "loss": 1.8376,
+      "step": 10478
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5802252246443256,
+      "learning_rate": 4.659180434300581e-06,
+      "loss": 2.1074,
+      "step": 10479
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5552607861964837,
+      "learning_rate": 4.6555493656486815e-06,
+      "loss": 1.8412,
+      "step": 10480
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5390279090910323,
+      "learning_rate": 4.651919567184421e-06,
+      "loss": 1.8018,
+      "step": 10481
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5696023229054242,
+      "learning_rate": 4.648291039134397e-06,
+      "loss": 1.8407,
+      "step": 10482
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5878588761577364,
+      "learning_rate": 4.644663781725187e-06,
+      "loss": 2.077,
+      "step": 10483
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5595398778147734,
+      "learning_rate": 4.641037795183231e-06,
+      "loss": 1.8689,
+      "step": 10484
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5832328246729406,
+      "learning_rate": 4.6374130797349175e-06,
+      "loss": 1.7857,
+      "step": 10485
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6016599359661842,
+      "learning_rate": 4.63378963560657e-06,
+      "loss": 1.9533,
+      "step": 10486
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5699129715939523,
+      "learning_rate": 4.630167463024393e-06,
+      "loss": 2.0349,
+      "step": 10487
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5873996622885543,
+      "learning_rate": 4.626546562214546e-06,
+      "loss": 1.8121,
+      "step": 10488
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6251358619305442,
+      "learning_rate": 4.622926933403091e-06,
+      "loss": 1.8198,
+      "step": 10489
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6082653078986949,
+      "learning_rate": 4.619308576816012e-06,
+      "loss": 1.8298,
+      "step": 10490
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5633985137890262,
+      "learning_rate": 4.615691492679239e-06,
+      "loss": 2.0661,
+      "step": 10491
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5154932160367471,
+      "learning_rate": 4.612075681218578e-06,
+      "loss": 1.9036,
+      "step": 10492
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5992158380181931,
+      "learning_rate": 4.608461142659784e-06,
+      "loss": 1.8709,
+      "step": 10493
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5833879023377221,
+      "learning_rate": 4.604847877228543e-06,
+      "loss": 1.8265,
+      "step": 10494
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5421769690340095,
+      "learning_rate": 4.6012358851504285e-06,
+      "loss": 2.1087,
+      "step": 10495
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5889878707401719,
+      "learning_rate": 4.5976251666509565e-06,
+      "loss": 1.8288,
+      "step": 10496
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5672418726249308,
+      "learning_rate": 4.594015721955561e-06,
+      "loss": 1.9037,
+      "step": 10497
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5553774836258919,
+      "learning_rate": 4.5904075512895964e-06,
+      "loss": 1.9864,
+      "step": 10498
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5682713802690531,
+      "learning_rate": 4.5868006548783316e-06,
+      "loss": 2.0856,
+      "step": 10499
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.579594831237363,
+      "learning_rate": 4.58319503294696e-06,
+      "loss": 1.9021,
+      "step": 10500
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5577872742786512,
+      "learning_rate": 4.579590685720597e-06,
+      "loss": 1.8465,
+      "step": 10501
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5811784312294996,
+      "learning_rate": 4.575987613424279e-06,
+      "loss": 1.8962,
+      "step": 10502
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.563951559438157,
+      "learning_rate": 4.572385816282956e-06,
+      "loss": 2.0676,
+      "step": 10503
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5452488235748394,
+      "learning_rate": 4.568785294521505e-06,
+      "loss": 1.9255,
+      "step": 10504
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.540224968107299,
+      "learning_rate": 4.565186048364722e-06,
+      "loss": 1.8362,
+      "step": 10505
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5661015197357013,
+      "learning_rate": 4.5615880780373235e-06,
+      "loss": 1.8333,
+      "step": 10506
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6071440240362667,
+      "learning_rate": 4.557991383763943e-06,
+      "loss": 2.067,
+      "step": 10507
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5530019723039468,
+      "learning_rate": 4.55439596576914e-06,
+      "loss": 1.8678,
+      "step": 10508
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5743756792914757,
+      "learning_rate": 4.55080182427739e-06,
+      "loss": 1.8947,
+      "step": 10509
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5965351939529518,
+      "learning_rate": 4.547208959513085e-06,
+      "loss": 1.9512,
+      "step": 10510
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5634071259324567,
+      "learning_rate": 4.543617371700551e-06,
+      "loss": 2.1034,
+      "step": 10511
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5429486144623774,
+      "learning_rate": 4.54002706106402e-06,
+      "loss": 1.8489,
+      "step": 10512
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6010379119475018,
+      "learning_rate": 4.536438027827652e-06,
+      "loss": 1.7957,
+      "step": 10513
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5582237640091071,
+      "learning_rate": 4.5328502722155244e-06,
+      "loss": 1.8773,
+      "step": 10514
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6214373048028072,
+      "learning_rate": 4.529263794451635e-06,
+      "loss": 2.0235,
+      "step": 10515
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5390223991927653,
+      "learning_rate": 4.525678594759903e-06,
+      "loss": 1.847,
+      "step": 10516
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5310958657562963,
+      "learning_rate": 4.5220946733641716e-06,
+      "loss": 1.9537,
+      "step": 10517
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6392939906769093,
+      "learning_rate": 4.518512030488184e-06,
+      "loss": 1.8614,
+      "step": 10518
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5624030554027067,
+      "learning_rate": 4.514930666355638e-06,
+      "loss": 2.0752,
+      "step": 10519
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6107495245275211,
+      "learning_rate": 4.51135058119013e-06,
+      "loss": 1.9063,
+      "step": 10520
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5958147860329133,
+      "learning_rate": 4.507771775215161e-06,
+      "loss": 1.8578,
+      "step": 10521
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6077220240074688,
+      "learning_rate": 4.504194248654198e-06,
+      "loss": 1.8405,
+      "step": 10522
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5892603901999128,
+      "learning_rate": 4.5006180017305745e-06,
+      "loss": 1.9627,
+      "step": 10523
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.55879207097425,
+      "learning_rate": 4.497043034667589e-06,
+      "loss": 2.0468,
+      "step": 10524
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5759871833829338,
+      "learning_rate": 4.493469347688442e-06,
+      "loss": 1.8877,
+      "step": 10525
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5856719915663197,
+      "learning_rate": 4.489896941016233e-06,
+      "loss": 1.8726,
+      "step": 10526
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5959876767390332,
+      "learning_rate": 4.486325814874023e-06,
+      "loss": 2.0274,
+      "step": 10527
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5982217904457898,
+      "learning_rate": 4.48275596948477e-06,
+      "loss": 1.8946,
+      "step": 10528
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.609111479823473,
+      "learning_rate": 4.4791874050713365e-06,
+      "loss": 1.9245,
+      "step": 10529
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5896932177228011,
+      "learning_rate": 4.475620121856547e-06,
+      "loss": 1.8745,
+      "step": 10530
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.565351407453564,
+      "learning_rate": 4.472054120063104e-06,
+      "loss": 2.0897,
+      "step": 10531
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5373687706260235,
+      "learning_rate": 4.468489399913648e-06,
+      "loss": 1.8591,
+      "step": 10532
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6110272412522735,
+      "learning_rate": 4.464925961630753e-06,
+      "loss": 1.8503,
+      "step": 10533
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6148657184463854,
+      "learning_rate": 4.461363805436889e-06,
+      "loss": 1.8377,
+      "step": 10534
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5443570187087716,
+      "learning_rate": 4.457802931554453e-06,
+      "loss": 1.9492,
+      "step": 10535
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6184282258689743,
+      "learning_rate": 4.454243340205771e-06,
+      "loss": 2.0631,
+      "step": 10536
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.572016278642415,
+      "learning_rate": 4.450685031613083e-06,
+      "loss": 1.8403,
+      "step": 10537
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5856567995080311,
+      "learning_rate": 4.447128005998547e-06,
+      "loss": 1.8818,
+      "step": 10538
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5454554922355379,
+      "learning_rate": 4.443572263584242e-06,
+      "loss": 2.0648,
+      "step": 10539
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5820548089774745,
+      "learning_rate": 4.440017804592161e-06,
+      "loss": 1.8422,
+      "step": 10540
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6047407887122879,
+      "learning_rate": 4.436464629244244e-06,
+      "loss": 1.9451,
+      "step": 10541
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6140644233269725,
+      "learning_rate": 4.432912737762312e-06,
+      "loss": 1.8327,
+      "step": 10542
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5568272860779881,
+      "learning_rate": 4.429362130368128e-06,
+      "loss": 2.0586,
+      "step": 10543
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5523858173090604,
+      "learning_rate": 4.4258128072833745e-06,
+      "loss": 1.9109,
+      "step": 10544
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5665358581469662,
+      "learning_rate": 4.422264768729648e-06,
+      "loss": 1.8393,
+      "step": 10545
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5656893694996717,
+      "learning_rate": 4.418718014928466e-06,
+      "loss": 1.8616,
+      "step": 10546
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5642854776490852,
+      "learning_rate": 4.41517254610127e-06,
+      "loss": 2.0375,
+      "step": 10547
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5690070827515783,
+      "learning_rate": 4.411628362469417e-06,
+      "loss": 1.9732,
+      "step": 10548
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.564564307647766,
+      "learning_rate": 4.408085464254183e-06,
+      "loss": 1.8156,
+      "step": 10549
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5537797760222043,
+      "learning_rate": 4.404543851676768e-06,
+      "loss": 1.8613,
+      "step": 10550
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5445309394615668,
+      "learning_rate": 4.401003524958286e-06,
+      "loss": 2.0653,
+      "step": 10551
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5608439116844629,
+      "learning_rate": 4.39746448431978e-06,
+      "loss": 1.8558,
+      "step": 10552
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5891345587663599,
+      "learning_rate": 4.393926729982204e-06,
+      "loss": 1.8296,
+      "step": 10553
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5557733066866404,
+      "learning_rate": 4.39039026216643e-06,
+      "loss": 1.9472,
+      "step": 10554
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5431614137210012,
+      "learning_rate": 4.38685508109326e-06,
+      "loss": 1.8638,
+      "step": 10555
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6253208720823549,
+      "learning_rate": 4.383321186983411e-06,
+      "loss": 2.0757,
+      "step": 10556
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5606981555762648,
+      "learning_rate": 4.379788580057515e-06,
+      "loss": 1.8714,
+      "step": 10557
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5902668514626077,
+      "learning_rate": 4.3762572605361264e-06,
+      "loss": 1.9062,
+      "step": 10558
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5442211779161195,
+      "learning_rate": 4.37272722863972e-06,
+      "loss": 2.0565,
+      "step": 10559
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5428041391839626,
+      "learning_rate": 4.3691984845886934e-06,
+      "loss": 1.9226,
+      "step": 10560
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5543268175278592,
+      "learning_rate": 4.36567102860336e-06,
+      "loss": 1.8203,
+      "step": 10561
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5752307068894305,
+      "learning_rate": 4.3621448609039525e-06,
+      "loss": 1.8843,
+      "step": 10562
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5926579422871165,
+      "learning_rate": 4.358619981710621e-06,
+      "loss": 2.0152,
+      "step": 10563
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5792831571610885,
+      "learning_rate": 4.355096391243449e-06,
+      "loss": 1.8817,
+      "step": 10564
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5731650021611585,
+      "learning_rate": 4.351574089722408e-06,
+      "loss": 1.9217,
+      "step": 10565
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5156083002728282,
+      "learning_rate": 4.348053077367428e-06,
+      "loss": 1.9291,
+      "step": 10566
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6094471595873626,
+      "learning_rate": 4.344533354398342e-06,
+      "loss": 1.8604,
+      "step": 10567
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5379255466176427,
+      "learning_rate": 4.341014921034881e-06,
+      "loss": 2.0517,
+      "step": 10568
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.598855412568698,
+      "learning_rate": 4.337497777496738e-06,
+      "loss": 1.8574,
+      "step": 10569
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5742304056283951,
+      "learning_rate": 4.3339819240034885e-06,
+      "loss": 1.8847,
+      "step": 10570
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5746026109802733,
+      "learning_rate": 4.33046736077464e-06,
+      "loss": 2.0352,
+      "step": 10571
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.550613946682847,
+      "learning_rate": 4.326954088029636e-06,
+      "loss": 1.9641,
+      "step": 10572
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6041230013151062,
+      "learning_rate": 4.3234421059878045e-06,
+      "loss": 1.8474,
+      "step": 10573
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5893225723839458,
+      "learning_rate": 4.3199314148684304e-06,
+      "loss": 1.8612,
+      "step": 10574
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.601748834173107,
+      "learning_rate": 4.3164220148907e-06,
+      "loss": 2.0793,
+      "step": 10575
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5681197801509726,
+      "learning_rate": 4.312913906273703e-06,
+      "loss": 1.8371,
+      "step": 10576
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5699461870048294,
+      "learning_rate": 4.309407089236486e-06,
+      "loss": 1.8715,
+      "step": 10577
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6395778649460857,
+      "learning_rate": 4.305901563997977e-06,
+      "loss": 1.8728,
+      "step": 10578
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5654437413902124,
+      "learning_rate": 4.302397330777039e-06,
+      "loss": 1.9314,
+      "step": 10579
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5878659707735111,
+      "learning_rate": 4.298894389792479e-06,
+      "loss": 2.0977,
+      "step": 10580
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6028337652154623,
+      "learning_rate": 4.295392741262977e-06,
+      "loss": 1.8658,
+      "step": 10581
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5625756455965717,
+      "learning_rate": 4.291892385407156e-06,
+      "loss": 1.8498,
+      "step": 10582
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.631036590586062,
+      "learning_rate": 4.288393322443579e-06,
+      "loss": 2.0587,
+      "step": 10583
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6055564674823034,
+      "learning_rate": 4.2848955525906846e-06,
+      "loss": 1.8628,
+      "step": 10584
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6304497022301477,
+      "learning_rate": 4.281399076066861e-06,
+      "loss": 1.9385,
+      "step": 10585
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.611662023560275,
+      "learning_rate": 4.277903893090407e-06,
+      "loss": 1.8918,
+      "step": 10586
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5944052306495856,
+      "learning_rate": 4.274410003879534e-06,
+      "loss": 1.859,
+      "step": 10587
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6610418643533845,
+      "learning_rate": 4.2709174086524e-06,
+      "loss": 2.0596,
+      "step": 10588
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6087664896515093,
+      "learning_rate": 4.267426107627046e-06,
+      "loss": 1.9043,
+      "step": 10589
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5550199618821093,
+      "learning_rate": 4.263936101021448e-06,
+      "loss": 1.8859,
+      "step": 10590
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5806254836161225,
+      "learning_rate": 4.260447389053507e-06,
+      "loss": 1.9557,
+      "step": 10591
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6157435715465363,
+      "learning_rate": 4.256959971941038e-06,
+      "loss": 2.0248,
+      "step": 10592
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5612962236954546,
+      "learning_rate": 4.253473849901768e-06,
+      "loss": 1.8265,
+      "step": 10593
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5925141038601712,
+      "learning_rate": 4.249989023153358e-06,
+      "loss": 1.852,
+      "step": 10594
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5740414024905831,
+      "learning_rate": 4.246505491913375e-06,
+      "loss": 2.0621,
+      "step": 10595
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6304525744570076,
+      "learning_rate": 4.243023256399312e-06,
+      "loss": 1.836,
+      "step": 10596
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5720337605326904,
+      "learning_rate": 4.2395423168285765e-06,
+      "loss": 1.9533,
+      "step": 10597
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6709423768403865,
+      "learning_rate": 4.236062673418503e-06,
+      "loss": 1.9178,
+      "step": 10598
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5946918594321177,
+      "learning_rate": 4.232584326386335e-06,
+      "loss": 1.8735,
+      "step": 10599
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5824161634228078,
+      "learning_rate": 4.229107275949245e-06,
+      "loss": 2.0783,
+      "step": 10600
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5697304777629438,
+      "learning_rate": 4.225631522324317e-06,
+      "loss": 1.8634,
+      "step": 10601
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5947652246841143,
+      "learning_rate": 4.222157065728552e-06,
+      "loss": 1.8837,
+      "step": 10602
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5709404250396083,
+      "learning_rate": 4.2186839063788895e-06,
+      "loss": 2.0695,
+      "step": 10603
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5113522847744275,
+      "learning_rate": 4.215212044492148e-06,
+      "loss": 1.9255,
+      "step": 10604
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5568738428940869,
+      "learning_rate": 4.211741480285112e-06,
+      "loss": 1.8672,
+      "step": 10605
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5475215485222347,
+      "learning_rate": 4.208272213974454e-06,
+      "loss": 1.865,
+      "step": 10606
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5508275450451428,
+      "learning_rate": 4.2048042457767775e-06,
+      "loss": 2.0345,
+      "step": 10607
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6170877164647876,
+      "learning_rate": 4.201337575908601e-06,
+      "loss": 1.8203,
+      "step": 10608
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.543294493269657,
+      "learning_rate": 4.1978722045863616e-06,
+      "loss": 1.8665,
+      "step": 10609
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5486194732218859,
+      "learning_rate": 4.1944081320264176e-06,
+      "loss": 1.9353,
+      "step": 10610
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6080563575543125,
+      "learning_rate": 4.190945358445053e-06,
+      "loss": 1.9004,
+      "step": 10611
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5832549123247058,
+      "learning_rate": 4.187483884058441e-06,
+      "loss": 2.0686,
+      "step": 10612
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.560764645460689,
+      "learning_rate": 4.184023709082718e-06,
+      "loss": 1.8479,
+      "step": 10613
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5765711989157728,
+      "learning_rate": 4.1805648337339135e-06,
+      "loss": 1.8794,
+      "step": 10614
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5644684676203171,
+      "learning_rate": 4.177107258227964e-06,
+      "loss": 2.0754,
+      "step": 10615
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5636783934960637,
+      "learning_rate": 4.1736509827807634e-06,
+      "loss": 1.9728,
+      "step": 10616
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5632071127042587,
+      "learning_rate": 4.170196007608082e-06,
+      "loss": 1.8444,
+      "step": 10617
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5614474201714444,
+      "learning_rate": 4.16674233292563e-06,
+      "loss": 1.7933,
+      "step": 10618
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5696355733475515,
+      "learning_rate": 4.1632899589490485e-06,
+      "loss": 1.8119,
+      "step": 10619
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5820098392482016,
+      "learning_rate": 4.159838885893866e-06,
+      "loss": 2.0465,
+      "step": 10620
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.541708654382157,
+      "learning_rate": 4.156389113975559e-06,
+      "loss": 1.8205,
+      "step": 10621
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5089678322528497,
+      "learning_rate": 4.152940643409514e-06,
+      "loss": 1.9418,
+      "step": 10622
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5644980771877202,
+      "learning_rate": 4.149493474411012e-06,
+      "loss": 1.8693,
+      "step": 10623
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5887673251328183,
+      "learning_rate": 4.146047607195302e-06,
+      "loss": 2.0303,
+      "step": 10624
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5873937234005123,
+      "learning_rate": 4.142603041977505e-06,
+      "loss": 1.8806,
+      "step": 10625
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5542495182549052,
+      "learning_rate": 4.139159778972679e-06,
+      "loss": 1.8346,
+      "step": 10626
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6154343886627305,
+      "learning_rate": 4.135717818395818e-06,
+      "loss": 2.073,
+      "step": 10627
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5116603399246111,
+      "learning_rate": 4.1322771604618e-06,
+      "loss": 1.9651,
+      "step": 10628
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5724575021791194,
+      "learning_rate": 4.128837805385438e-06,
+      "loss": 1.8553,
+      "step": 10629
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6022243427740438,
+      "learning_rate": 4.125399753381487e-06,
+      "loss": 1.8254,
+      "step": 10630
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5612549249224116,
+      "learning_rate": 4.121963004664578e-06,
+      "loss": 1.8825,
+      "step": 10631
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5802257401342035,
+      "learning_rate": 4.118527559449287e-06,
+      "loss": 2.0873,
+      "step": 10632
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5841035115219498,
+      "learning_rate": 4.115093417950102e-06,
+      "loss": 1.8762,
+      "step": 10633
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.575701519642263,
+      "learning_rate": 4.111660580381435e-06,
+      "loss": 1.802,
+      "step": 10634
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5328275065702764,
+      "learning_rate": 4.108229046957609e-06,
+      "loss": 1.9027,
+      "step": 10635
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5591962355437892,
+      "learning_rate": 4.104798817892869e-06,
+      "loss": 2.041,
+      "step": 10636
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6006509975482716,
+      "learning_rate": 4.10136989340138e-06,
+      "loss": 1.9045,
+      "step": 10637
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5700619826788629,
+      "learning_rate": 4.09794227369722e-06,
+      "loss": 1.8795,
+      "step": 10638
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5570069764791588,
+      "learning_rate": 4.094515958994391e-06,
+      "loss": 1.8577,
+      "step": 10639
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5798105674632692,
+      "learning_rate": 4.091090949506815e-06,
+      "loss": 1.9932,
+      "step": 10640
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5382403147617756,
+      "learning_rate": 4.087667245448326e-06,
+      "loss": 1.9242,
+      "step": 10641
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5450971594536743,
+      "learning_rate": 4.084244847032678e-06,
+      "loss": 1.8647,
+      "step": 10642
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6067586082974338,
+      "learning_rate": 4.080823754473554e-06,
+      "loss": 1.8473,
+      "step": 10643
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6361510777469066,
+      "learning_rate": 4.077403967984536e-06,
+      "loss": 2.0681,
+      "step": 10644
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6102597959052373,
+      "learning_rate": 4.073985487779141e-06,
+      "loss": 1.8577,
+      "step": 10645
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5634807330602346,
+      "learning_rate": 4.070568314070802e-06,
+      "loss": 1.8892,
+      "step": 10646
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5483968184630161,
+      "learning_rate": 4.0671524470728615e-06,
+      "loss": 1.8988,
+      "step": 10647
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6037163714275811,
+      "learning_rate": 4.063737886998587e-06,
+      "loss": 2.0665,
+      "step": 10648
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5803500412220749,
+      "learning_rate": 4.060324634061166e-06,
+      "loss": 1.8924,
+      "step": 10649
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6439519683100701,
+      "learning_rate": 4.056912688473704e-06,
+      "loss": 1.9049,
+      "step": 10650
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6001678568741484,
+      "learning_rate": 4.053502050449209e-06,
+      "loss": 1.8642,
+      "step": 10651
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5898285140265045,
+      "learning_rate": 4.050092720200638e-06,
+      "loss": 2.0217,
+      "step": 10652
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5286710434074449,
+      "learning_rate": 4.046684697940847e-06,
+      "loss": 1.8705,
+      "step": 10653
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5460732152789862,
+      "learning_rate": 4.043277983882598e-06,
+      "loss": 1.8931,
+      "step": 10654
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5559422028337601,
+      "learning_rate": 4.039872578238602e-06,
+      "loss": 1.8296,
+      "step": 10655
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6630854926405113,
+      "learning_rate": 4.03646848122147e-06,
+      "loss": 2.0648,
+      "step": 10656
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5929550745571056,
+      "learning_rate": 4.033065693043728e-06,
+      "loss": 1.8743,
+      "step": 10657
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5892262969771478,
+      "learning_rate": 4.029664213917836e-06,
+      "loss": 1.8921,
+      "step": 10658
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5296473977046164,
+      "learning_rate": 4.026264044056144e-06,
+      "loss": 1.9301,
+      "step": 10659
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5276090018150382,
+      "learning_rate": 4.022865183670954e-06,
+      "loss": 2.0714,
+      "step": 10660
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5709064409267968,
+      "learning_rate": 4.019467632974475e-06,
+      "loss": 1.829,
+      "step": 10661
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.608692827250539,
+      "learning_rate": 4.016071392178808e-06,
+      "loss": 1.8517,
+      "step": 10662
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6022284470369109,
+      "learning_rate": 4.012676461496015e-06,
+      "loss": 1.8832,
+      "step": 10663
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6088131575344443,
+      "learning_rate": 4.009282841138057e-06,
+      "loss": 2.1101,
+      "step": 10664
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5970617794871753,
+      "learning_rate": 4.005890531316789e-06,
+      "loss": 1.8285,
+      "step": 10665
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5392640046202631,
+      "learning_rate": 4.002499532244033e-06,
+      "loss": 1.9196,
+      "step": 10666
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.538572824491611,
+      "learning_rate": 3.999109844131485e-06,
+      "loss": 1.8297,
+      "step": 10667
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6138075043850049,
+      "learning_rate": 3.995721467190777e-06,
+      "loss": 2.041,
+      "step": 10668
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.571605445608972,
+      "learning_rate": 3.992334401633477e-06,
+      "loss": 1.8457,
+      "step": 10669
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6403212187680287,
+      "learning_rate": 3.988948647671031e-06,
+      "loss": 1.8729,
+      "step": 10670
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.565269087068171,
+      "learning_rate": 3.9855642055148496e-06,
+      "loss": 1.8737,
+      "step": 10671
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.537088129794888,
+      "learning_rate": 3.982181075376218e-06,
+      "loss": 2.0976,
+      "step": 10672
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5805221163471836,
+      "learning_rate": 3.978799257466356e-06,
+      "loss": 1.843,
+      "step": 10673
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5966064014751619,
+      "learning_rate": 3.975418751996429e-06,
+      "loss": 1.8961,
+      "step": 10674
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6045539483690441,
+      "learning_rate": 3.972039559177473e-06,
+      "loss": 1.9182,
+      "step": 10675
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.625646862038081,
+      "learning_rate": 3.968661679220468e-06,
+      "loss": 2.0668,
+      "step": 10676
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6233525712289661,
+      "learning_rate": 3.965285112336326e-06,
+      "loss": 1.8189,
+      "step": 10677
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5390807768627728,
+      "learning_rate": 3.961909858735838e-06,
+      "loss": 1.9383,
+      "step": 10678
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5727035669787839,
+      "learning_rate": 3.958535918629744e-06,
+      "loss": 1.8585,
+      "step": 10679
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5707310984824652,
+      "learning_rate": 3.955163292228695e-06,
+      "loss": 2.0756,
+      "step": 10680
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5608016359449933,
+      "learning_rate": 3.951791979743255e-06,
+      "loss": 1.8043,
+      "step": 10681
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6299898186736491,
+      "learning_rate": 3.948421981383913e-06,
+      "loss": 1.8583,
+      "step": 10682
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6301800931983205,
+      "learning_rate": 3.945053297361065e-06,
+      "loss": 1.8624,
+      "step": 10683
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5194553934447379,
+      "learning_rate": 3.941685927885036e-06,
+      "loss": 2.0464,
+      "step": 10684
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5795581959263472,
+      "learning_rate": 3.9383198731660655e-06,
+      "loss": 1.8038,
+      "step": 10685
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6058273957992596,
+      "learning_rate": 3.934955133414309e-06,
+      "loss": 1.8635,
+      "step": 10686
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6143540863082474,
+      "learning_rate": 3.9315917088398405e-06,
+      "loss": 1.8206,
+      "step": 10687
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5691748800863341,
+      "learning_rate": 3.928229599652653e-06,
+      "loss": 2.0207,
+      "step": 10688
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5670368091645882,
+      "learning_rate": 3.924868806062657e-06,
+      "loss": 1.8942,
+      "step": 10689
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.526738411069201,
+      "learning_rate": 3.92150932827968e-06,
+      "loss": 1.9671,
+      "step": 10690
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5419907575417912,
+      "learning_rate": 3.918151166513465e-06,
+      "loss": 1.8159,
+      "step": 10691
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5598445857364877,
+      "learning_rate": 3.914794320973683e-06,
+      "loss": 2.0529,
+      "step": 10692
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6126120280339779,
+      "learning_rate": 3.911438791869909e-06,
+      "loss": 1.8707,
+      "step": 10693
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5383020458193987,
+      "learning_rate": 3.908084579411647e-06,
+      "loss": 1.899,
+      "step": 10694
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.563856598439715,
+      "learning_rate": 3.904731683808313e-06,
+      "loss": 1.8763,
+      "step": 10695
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5456713368450398,
+      "learning_rate": 3.901380105269242e-06,
+      "loss": 2.0243,
+      "step": 10696
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5269004386001627,
+      "learning_rate": 3.898029844003684e-06,
+      "loss": 1.9599,
+      "step": 10697
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6062287795023077,
+      "learning_rate": 3.894680900220812e-06,
+      "loss": 1.8332,
+      "step": 10698
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6244027539377481,
+      "learning_rate": 3.891333274129716e-06,
+      "loss": 1.858,
+      "step": 10699
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.567995497445232,
+      "learning_rate": 3.887986965939405e-06,
+      "loss": 2.0448,
+      "step": 10700
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5861379639235234,
+      "learning_rate": 3.8846419758587884e-06,
+      "loss": 1.8497,
+      "step": 10701
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5934809415454316,
+      "learning_rate": 3.8812983040967195e-06,
+      "loss": 1.852,
+      "step": 10702
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5191989313707627,
+      "learning_rate": 3.877955950861958e-06,
+      "loss": 1.9266,
+      "step": 10703
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5363355264282338,
+      "learning_rate": 3.8746149163631785e-06,
+      "loss": 2.0707,
+      "step": 10704
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5832172674021526,
+      "learning_rate": 3.8712752008089795e-06,
+      "loss": 1.843,
+      "step": 10705
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5687438601743268,
+      "learning_rate": 3.8679368044078565e-06,
+      "loss": 1.8778,
+      "step": 10706
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5837821123838463,
+      "learning_rate": 3.864599727368257e-06,
+      "loss": 1.8834,
+      "step": 10707
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5449700873798506,
+      "learning_rate": 3.861263969898529e-06,
+      "loss": 2.0545,
+      "step": 10708
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.531165510768281,
+      "learning_rate": 3.857929532206922e-06,
+      "loss": 1.8991,
+      "step": 10709
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5958949169788778,
+      "learning_rate": 3.854596414501633e-06,
+      "loss": 1.8997,
+      "step": 10710
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6351689814939461,
+      "learning_rate": 3.851264616990763e-06,
+      "loss": 1.8516,
+      "step": 10711
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5573929354426366,
+      "learning_rate": 3.847934139882311e-06,
+      "loss": 2.1147,
+      "step": 10712
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5503276909580921,
+      "learning_rate": 3.8446049833842366e-06,
+      "loss": 1.9063,
+      "step": 10713
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5894230886448758,
+      "learning_rate": 3.841277147704378e-06,
+      "loss": 1.8588,
+      "step": 10714
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6304056089298457,
+      "learning_rate": 3.837950633050505e-06,
+      "loss": 1.9215,
+      "step": 10715
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5722506324458648,
+      "learning_rate": 3.834625439630318e-06,
+      "loss": 2.0479,
+      "step": 10716
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5665467464368783,
+      "learning_rate": 3.83130156765141e-06,
+      "loss": 1.8004,
+      "step": 10717
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5643343684276644,
+      "learning_rate": 3.827979017321309e-06,
+      "loss": 1.8476,
+      "step": 10718
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6208498023858295,
+      "learning_rate": 3.824657788847452e-06,
+      "loss": 1.9089,
+      "step": 10719
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5602600700380348,
+      "learning_rate": 3.821337882437198e-06,
+      "loss": 2.0319,
+      "step": 10720
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5103762411744561,
+      "learning_rate": 3.818019298297834e-06,
+      "loss": 1.9146,
+      "step": 10721
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6206672104740065,
+      "learning_rate": 3.814702036636536e-06,
+      "loss": 1.8843,
+      "step": 10722
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5904050295468254,
+      "learning_rate": 3.811386097660416e-06,
+      "loss": 1.8494,
+      "step": 10723
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5897503173551677,
+      "learning_rate": 3.8080714815765196e-06,
+      "loss": 2.0914,
+      "step": 10724
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.570551204299024,
+      "learning_rate": 3.8047581885917727e-06,
+      "loss": 1.8968,
+      "step": 10725
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.578588601196136,
+      "learning_rate": 3.8014462189130434e-06,
+      "loss": 1.8782,
+      "step": 10726
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5738960214784489,
+      "learning_rate": 3.7981355727471135e-06,
+      "loss": 1.8017,
+      "step": 10727
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5574976927972151,
+      "learning_rate": 3.7948262503006787e-06,
+      "loss": 2.1131,
+      "step": 10728
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.542004617031652,
+      "learning_rate": 3.7915182517803544e-06,
+      "loss": 1.8234,
+      "step": 10729
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5699865695084094,
+      "learning_rate": 3.7882115773926726e-06,
+      "loss": 1.8617,
+      "step": 10730
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.573859240587759,
+      "learning_rate": 3.7849062273440815e-06,
+      "loss": 1.7971,
+      "step": 10731
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5536900172434109,
+      "learning_rate": 3.78160220184095e-06,
+      "loss": 2.0503,
+      "step": 10732
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5766615933089808,
+      "learning_rate": 3.7782995010895607e-06,
+      "loss": 1.8866,
+      "step": 10733
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5572580468141878,
+      "learning_rate": 3.7749981252961114e-06,
+      "loss": 1.9125,
+      "step": 10734
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6396995862758699,
+      "learning_rate": 3.771698074666727e-06,
+      "loss": 1.8355,
+      "step": 10735
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5987649765962956,
+      "learning_rate": 3.7683993494074367e-06,
+      "loss": 2.0659,
+      "step": 10736
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5787813025254561,
+      "learning_rate": 3.765101949724198e-06,
+      "loss": 1.8796,
+      "step": 10737
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6123504960491841,
+      "learning_rate": 3.7618058758228798e-06,
+      "loss": 1.8602,
+      "step": 10738
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6639818734042311,
+      "learning_rate": 3.75851112790927e-06,
+      "loss": 1.8905,
+      "step": 10739
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.554442448707175,
+      "learning_rate": 3.7552177061890736e-06,
+      "loss": 1.9199,
+      "step": 10740
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.544240338617265,
+      "learning_rate": 3.7519256108679097e-06,
+      "loss": 2.0012,
+      "step": 10741
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6049756624723479,
+      "learning_rate": 3.748634842151319e-06,
+      "loss": 1.8389,
+      "step": 10742
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5952696472087444,
+      "learning_rate": 3.74534540024476e-06,
+      "loss": 1.8304,
+      "step": 10743
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.538891829038343,
+      "learning_rate": 3.7420572853536008e-06,
+      "loss": 2.0192,
+      "step": 10744
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5593710980442993,
+      "learning_rate": 3.738770497683136e-06,
+      "loss": 1.8495,
+      "step": 10745
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5977112551711441,
+      "learning_rate": 3.735485037438574e-06,
+      "loss": 1.9231,
+      "step": 10746
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5894606584638741,
+      "learning_rate": 3.732200904825042e-06,
+      "loss": 1.851,
+      "step": 10747
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5791515076342623,
+      "learning_rate": 3.728918100047568e-06,
+      "loss": 2.0371,
+      "step": 10748
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5619769957405876,
+      "learning_rate": 3.7256366233111266e-06,
+      "loss": 1.9034,
+      "step": 10749
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5988008336224916,
+      "learning_rate": 3.722356474820593e-06,
+      "loss": 1.8153,
+      "step": 10750
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6287153649341175,
+      "learning_rate": 3.7190776547807447e-06,
+      "loss": 1.8467,
+      "step": 10751
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5754778825571893,
+      "learning_rate": 3.715800163396313e-06,
+      "loss": 2.0103,
+      "step": 10752
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5344265867023649,
+      "learning_rate": 3.712524000871903e-06,
+      "loss": 1.9268,
+      "step": 10753
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5389847035852853,
+      "learning_rate": 3.7092491674120766e-06,
+      "loss": 1.8182,
+      "step": 10754
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5807825516462479,
+      "learning_rate": 3.705975663221295e-06,
+      "loss": 1.8956,
+      "step": 10755
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6285476000003524,
+      "learning_rate": 3.7027034885039224e-06,
+      "loss": 2.0265,
+      "step": 10756
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5684692427237742,
+      "learning_rate": 3.699432643464265e-06,
+      "loss": 1.8601,
+      "step": 10757
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5688725012336092,
+      "learning_rate": 3.6961631283065374e-06,
+      "loss": 1.8645,
+      "step": 10758
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5695517873279012,
+      "learning_rate": 3.692894943234856e-06,
+      "loss": 1.9329,
+      "step": 10759
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5809810119288172,
+      "learning_rate": 3.6896280884532842e-06,
+      "loss": 2.0728,
+      "step": 10760
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5656820586016746,
+      "learning_rate": 3.6863625641657712e-06,
+      "loss": 1.8249,
+      "step": 10761
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5583207574363943,
+      "learning_rate": 3.683098370576196e-06,
+      "loss": 1.8458,
+      "step": 10762
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5400000532881264,
+      "learning_rate": 3.679835507888371e-06,
+      "loss": 1.861,
+      "step": 10763
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5895824726573278,
+      "learning_rate": 3.676573976305997e-06,
+      "loss": 2.0258,
+      "step": 10764
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5540433994483275,
+      "learning_rate": 3.6733137760327108e-06,
+      "loss": 1.962,
+      "step": 10765
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5439044586910192,
+      "learning_rate": 3.6700549072720554e-06,
+      "loss": 1.8099,
+      "step": 10766
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6100317532157468,
+      "learning_rate": 3.66679737022749e-06,
+      "loss": 1.8957,
+      "step": 10767
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5532694667341393,
+      "learning_rate": 3.663541165102419e-06,
+      "loss": 2.0533,
+      "step": 10768
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5763894349661339,
+      "learning_rate": 3.6602862921001178e-06,
+      "loss": 1.8437,
+      "step": 10769
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5574496609973877,
+      "learning_rate": 3.657032751423803e-06,
+      "loss": 1.8427,
+      "step": 10770
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5366708548972129,
+      "learning_rate": 3.6537805432766248e-06,
+      "loss": 1.9604,
+      "step": 10771
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5461627965779805,
+      "learning_rate": 3.650529667861613e-06,
+      "loss": 1.8717,
+      "step": 10772
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5939330798584653,
+      "learning_rate": 3.6472801253817412e-06,
+      "loss": 2.0636,
+      "step": 10773
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5828431606194739,
+      "learning_rate": 3.644031916039889e-06,
+      "loss": 1.8655,
+      "step": 10774
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5394274105516511,
+      "learning_rate": 3.6407850400388575e-06,
+      "loss": 1.8615,
+      "step": 10775
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5717235089222511,
+      "learning_rate": 3.637539497581363e-06,
+      "loss": 2.0144,
+      "step": 10776
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5220374457371912,
+      "learning_rate": 3.634295288870035e-06,
+      "loss": 1.9098,
+      "step": 10777
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5824677543095469,
+      "learning_rate": 3.6310524141074248e-06,
+      "loss": 1.8224,
+      "step": 10778
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5517912304144045,
+      "learning_rate": 3.6278108734959985e-06,
+      "loss": 1.8404,
+      "step": 10779
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5553743721089511,
+      "learning_rate": 3.624570667238139e-06,
+      "loss": 2.059,
+      "step": 10780
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5476866306798203,
+      "learning_rate": 3.6213317955361454e-06,
+      "loss": 1.878,
+      "step": 10781
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5596477394266737,
+      "learning_rate": 3.6180942585922334e-06,
+      "loss": 1.8362,
+      "step": 10782
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5437451009748437,
+      "learning_rate": 3.614858056608536e-06,
+      "loss": 1.825,
+      "step": 10783
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5873150764828959,
+      "learning_rate": 3.611623189787103e-06,
+      "loss": 1.928,
+      "step": 10784
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5337628838598454,
+      "learning_rate": 3.608389658329897e-06,
+      "loss": 2.0679,
+      "step": 10785
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5849488556801015,
+      "learning_rate": 3.6051574624388074e-06,
+      "loss": 1.8616,
+      "step": 10786
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5811216632282263,
+      "learning_rate": 3.6019266023156277e-06,
+      "loss": 1.882,
+      "step": 10787
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.564177530193364,
+      "learning_rate": 3.598697078162075e-06,
+      "loss": 2.0347,
+      "step": 10788
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5371775963769169,
+      "learning_rate": 3.5954688901797846e-06,
+      "loss": 1.8737,
+      "step": 10789
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5321897077376401,
+      "learning_rate": 3.592242038570301e-06,
+      "loss": 1.9171,
+      "step": 10790
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6619604305610112,
+      "learning_rate": 3.5890165235350938e-06,
+      "loss": 1.8731,
+      "step": 10791
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5602237946319516,
+      "learning_rate": 3.5857923452755433e-06,
+      "loss": 2.0482,
+      "step": 10792
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5595352119008782,
+      "learning_rate": 3.58256950399295e-06,
+      "loss": 1.8452,
+      "step": 10793
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.597463378281254,
+      "learning_rate": 3.57934799988853e-06,
+      "loss": 1.8423,
+      "step": 10794
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5951380225806212,
+      "learning_rate": 3.5761278331634037e-06,
+      "loss": 1.7885,
+      "step": 10795
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5163799770447567,
+      "learning_rate": 3.572909004018632e-06,
+      "loss": 1.9421,
+      "step": 10796
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5487157096653337,
+      "learning_rate": 3.569691512655182e-06,
+      "loss": 2.0615,
+      "step": 10797
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5929985441592112,
+      "learning_rate": 3.5664753592739188e-06,
+      "loss": 1.853,
+      "step": 10798
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6028244117953934,
+      "learning_rate": 3.563260544075661e-06,
+      "loss": 1.8583,
+      "step": 10799
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5502729670974442,
+      "learning_rate": 3.560047067261099e-06,
+      "loss": 2.0917,
+      "step": 10800
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5900503362596001,
+      "learning_rate": 3.556834929030883e-06,
+      "loss": 1.8284,
+      "step": 10801
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5593725769303093,
+      "learning_rate": 3.553624129585559e-06,
+      "loss": 1.9733,
+      "step": 10802
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5769159843991529,
+      "learning_rate": 3.5504146691255736e-06,
+      "loss": 1.7989,
+      "step": 10803
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6188408346993773,
+      "learning_rate": 3.547206547851323e-06,
+      "loss": 1.8711,
+      "step": 10804
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5774724604516033,
+      "learning_rate": 3.543999765963102e-06,
+      "loss": 2.0039,
+      "step": 10805
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5727024566900277,
+      "learning_rate": 3.540794323661109e-06,
+      "loss": 1.8579,
+      "step": 10806
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5377008856995557,
+      "learning_rate": 3.5375902211454925e-06,
+      "loss": 1.8773,
+      "step": 10807
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5173820121251498,
+      "learning_rate": 3.534387458616284e-06,
+      "loss": 1.9311,
+      "step": 10808
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6088263802567797,
+      "learning_rate": 3.5311860362734454e-06,
+      "loss": 2.0766,
+      "step": 10809
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5589596116059125,
+      "learning_rate": 3.5279859543168674e-06,
+      "loss": 1.8362,
+      "step": 10810
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6205291753948583,
+      "learning_rate": 3.5247872129463318e-06,
+      "loss": 1.8313,
+      "step": 10811
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5648615650274053,
+      "learning_rate": 3.5215898123615477e-06,
+      "loss": 2.0192,
+      "step": 10812
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5620175775536919,
+      "learning_rate": 3.518393752762156e-06,
+      "loss": 1.8636,
+      "step": 10813
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5921677924134402,
+      "learning_rate": 3.5151990343476853e-06,
+      "loss": 1.8834,
+      "step": 10814
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5027932497238712,
+      "learning_rate": 3.512005657317602e-06,
+      "loss": 1.9328,
+      "step": 10815
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6179834376305374,
+      "learning_rate": 3.5088136218712817e-06,
+      "loss": 1.8615,
+      "step": 10816
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5851861744880031,
+      "learning_rate": 3.5056229282080077e-06,
+      "loss": 2.0146,
+      "step": 10817
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5495288408140643,
+      "learning_rate": 3.502433576527006e-06,
+      "loss": 1.8705,
+      "step": 10818
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5778685820041513,
+      "learning_rate": 3.4992455670273843e-06,
+      "loss": 1.8614,
+      "step": 10819
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5452358348438684,
+      "learning_rate": 3.4960588999081916e-06,
+      "loss": 2.0577,
+      "step": 10820
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5407436032092618,
+      "learning_rate": 3.49287357536838e-06,
+      "loss": 1.9434,
+      "step": 10821
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5451299243195878,
+      "learning_rate": 3.4896895936068263e-06,
+      "loss": 1.8468,
+      "step": 10822
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5892697917447388,
+      "learning_rate": 3.486506954822316e-06,
+      "loss": 1.8974,
+      "step": 10823
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5441851278320088,
+      "learning_rate": 3.483325659213557e-06,
+      "loss": 1.8046,
+      "step": 10824
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5295188915764267,
+      "learning_rate": 3.480145706979171e-06,
+      "loss": 2.0737,
+      "step": 10825
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6030719770605287,
+      "learning_rate": 3.4769670983176935e-06,
+      "loss": 1.8629,
+      "step": 10826
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6035899084310105,
+      "learning_rate": 3.4737898334275792e-06,
+      "loss": 1.9344,
+      "step": 10827
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5473661709414861,
+      "learning_rate": 3.4706139125071977e-06,
+      "loss": 1.8696,
+      "step": 10828
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5739439369620827,
+      "learning_rate": 3.4674393357548322e-06,
+      "loss": 2.069,
+      "step": 10829
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5413005008842366,
+      "learning_rate": 3.464266103368688e-06,
+      "loss": 1.8865,
+      "step": 10830
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5906839263018648,
+      "learning_rate": 3.461094215546884e-06,
+      "loss": 1.8743,
+      "step": 10831
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5576929035833064,
+      "learning_rate": 3.457923672487451e-06,
+      "loss": 2.0499,
+      "step": 10832
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5293207939269674,
+      "learning_rate": 3.4547544743883385e-06,
+      "loss": 1.9605,
+      "step": 10833
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5572364052476882,
+      "learning_rate": 3.451586621447417e-06,
+      "loss": 1.8204,
+      "step": 10834
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5455729339537025,
+      "learning_rate": 3.4484201138624632e-06,
+      "loss": 1.8502,
+      "step": 10835
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5810245520922116,
+      "learning_rate": 3.445254951831181e-06,
+      "loss": 1.8398,
+      "step": 10836
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5660997967884946,
+      "learning_rate": 3.4420911355511785e-06,
+      "loss": 2.0328,
+      "step": 10837
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5545514371393105,
+      "learning_rate": 3.4389286652199896e-06,
+      "loss": 1.8644,
+      "step": 10838
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5361064017113826,
+      "learning_rate": 3.435767541035059e-06,
+      "loss": 1.9717,
+      "step": 10839
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5702882641858916,
+      "learning_rate": 3.4326077631937485e-06,
+      "loss": 1.8423,
+      "step": 10840
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5600441781492409,
+      "learning_rate": 3.429449331893342e-06,
+      "loss": 2.0623,
+      "step": 10841
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5745624838647883,
+      "learning_rate": 3.426292247331017e-06,
+      "loss": 1.8127,
+      "step": 10842
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.571814145647379,
+      "learning_rate": 3.4231365097039e-06,
+      "loss": 1.8425,
+      "step": 10843
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5445758035333338,
+      "learning_rate": 3.419982119209017e-06,
+      "loss": 2.0191,
+      "step": 10844
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5649250270253655,
+      "learning_rate": 3.4168290760432902e-06,
+      "loss": 1.9191,
+      "step": 10845
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5562775792943725,
+      "learning_rate": 3.413677380403596e-06,
+      "loss": 1.922,
+      "step": 10846
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.553313119925402,
+      "learning_rate": 3.4105270324867073e-06,
+      "loss": 1.8749,
+      "step": 10847
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5644913423287873,
+      "learning_rate": 3.407378032489297e-06,
+      "loss": 1.8885,
+      "step": 10848
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5338045614090035,
+      "learning_rate": 3.4042303806079922e-06,
+      "loss": 2.0532,
+      "step": 10849
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5535319692979269,
+      "learning_rate": 3.401084077039293e-06,
+      "loss": 1.8533,
+      "step": 10850
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.548198985056893,
+      "learning_rate": 3.3979391219796487e-06,
+      "loss": 1.8851,
+      "step": 10851
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5213757869730109,
+      "learning_rate": 3.3947955156254153e-06,
+      "loss": 1.9287,
+      "step": 10852
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5571967257297079,
+      "learning_rate": 3.3916532581728414e-06,
+      "loss": 2.0456,
+      "step": 10853
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5415759166943671,
+      "learning_rate": 3.3885123498181348e-06,
+      "loss": 1.8124,
+      "step": 10854
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5567354341952665,
+      "learning_rate": 3.3853727907573796e-06,
+      "loss": 1.8581,
+      "step": 10855
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.542965612653651,
+      "learning_rate": 3.382234581186591e-06,
+      "loss": 1.8745,
+      "step": 10856
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5594755061050475,
+      "learning_rate": 3.3790977213017187e-06,
+      "loss": 2.0588,
+      "step": 10857
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5408739885797237,
+      "learning_rate": 3.375962211298589e-06,
+      "loss": 1.9478,
+      "step": 10858
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5603631408160092,
+      "learning_rate": 3.372828051372967e-06,
+      "loss": 1.8481,
+      "step": 10859
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.537985529359681,
+      "learning_rate": 3.369695241720547e-06,
+      "loss": 1.8287,
+      "step": 10860
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5668723929917758,
+      "learning_rate": 3.366563782536905e-06,
+      "loss": 2.0508,
+      "step": 10861
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5451123759197846,
+      "learning_rate": 3.3634336740175605e-06,
+      "loss": 1.8814,
+      "step": 10862
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5806161231990192,
+      "learning_rate": 3.360304916357937e-06,
+      "loss": 1.8504,
+      "step": 10863
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5286821665757154,
+      "learning_rate": 3.3571775097533704e-06,
+      "loss": 1.9745,
+      "step": 10864
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5312105031970817,
+      "learning_rate": 3.3540514543991347e-06,
+      "loss": 2.0771,
+      "step": 10865
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5456644626425059,
+      "learning_rate": 3.3509267504903853e-06,
+      "loss": 1.8138,
+      "step": 10866
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6022251467290242,
+      "learning_rate": 3.347803398222213e-06,
+      "loss": 1.8542,
+      "step": 10867
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5652011274413492,
+      "learning_rate": 3.344681397789626e-06,
+      "loss": 1.8681,
+      "step": 10868
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5216526325442508,
+      "learning_rate": 3.34156074938754e-06,
+      "loss": 2.0074,
+      "step": 10869
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4881695947940889,
+      "learning_rate": 3.3384414532107918e-06,
+      "loss": 1.9275,
+      "step": 10870
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.589856056191756,
+      "learning_rate": 3.335323509454133e-06,
+      "loss": 1.8702,
+      "step": 10871
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5820299323170798,
+      "learning_rate": 3.332206918312225e-06,
+      "loss": 1.8092,
+      "step": 10872
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5421502780921352,
+      "learning_rate": 3.3290916799796567e-06,
+      "loss": 2.051,
+      "step": 10873
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5819498483380261,
+      "learning_rate": 3.3259777946509195e-06,
+      "loss": 1.851,
+      "step": 10874
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.583393186932019,
+      "learning_rate": 3.3228652625204276e-06,
+      "loss": 1.9099,
+      "step": 10875
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5471643139005614,
+      "learning_rate": 3.3197540837825085e-06,
+      "loss": 2.1232,
+      "step": 10876
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5257979191058662,
+      "learning_rate": 3.316644258631407e-06,
+      "loss": 1.8862,
+      "step": 10877
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5764249499473517,
+      "learning_rate": 3.313535787261282e-06,
+      "loss": 1.8363,
+      "step": 10878
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.563350269811615,
+      "learning_rate": 3.3104286698662083e-06,
+      "loss": 1.8594,
+      "step": 10879
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6021398286114723,
+      "learning_rate": 3.3073229066401756e-06,
+      "loss": 1.7979,
+      "step": 10880
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5760971456272964,
+      "learning_rate": 3.3042184977770898e-06,
+      "loss": 2.008,
+      "step": 10881
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5710239034222614,
+      "learning_rate": 3.301115443470773e-06,
+      "loss": 1.9209,
+      "step": 10882
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5725731292578958,
+      "learning_rate": 3.2980137439149574e-06,
+      "loss": 1.9172,
+      "step": 10883
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5988762070048066,
+      "learning_rate": 3.2949133993032984e-06,
+      "loss": 1.8472,
+      "step": 10884
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5632641812432931,
+      "learning_rate": 3.2918144098293635e-06,
+      "loss": 2.1158,
+      "step": 10885
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5721633369808168,
+      "learning_rate": 3.2887167756866345e-06,
+      "loss": 1.8525,
+      "step": 10886
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5819395697415632,
+      "learning_rate": 3.285620497068509e-06,
+      "loss": 1.8494,
+      "step": 10887
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6221131603716856,
+      "learning_rate": 3.2825255741683077e-06,
+      "loss": 1.886,
+      "step": 10888
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6126319192034867,
+      "learning_rate": 3.2794320071792394e-06,
+      "loss": 2.0642,
+      "step": 10889
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6057442093698397,
+      "learning_rate": 3.2763397962944695e-06,
+      "loss": 1.9056,
+      "step": 10890
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5712629758685175,
+      "learning_rate": 3.273248941707052e-06,
+      "loss": 1.8698,
+      "step": 10891
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5912102428183791,
+      "learning_rate": 3.2701594436099493e-06,
+      "loss": 1.8882,
+      "step": 10892
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6044906752970196,
+      "learning_rate": 3.2670713021960647e-06,
+      "loss": 2.0466,
+      "step": 10893
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6025411656744994,
+      "learning_rate": 3.2639845176582056e-06,
+      "loss": 1.8497,
+      "step": 10894
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5228589107430774,
+      "learning_rate": 3.2608990901890762e-06,
+      "loss": 1.9279,
+      "step": 10895
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5645994358684547,
+      "learning_rate": 3.2578150199813363e-06,
+      "loss": 1.8453,
+      "step": 10896
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5939542135803495,
+      "learning_rate": 3.254732307227509e-06,
+      "loss": 2.056,
+      "step": 10897
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5699327238498245,
+      "learning_rate": 3.251650952120083e-06,
+      "loss": 1.901,
+      "step": 10898
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5802531639070976,
+      "learning_rate": 3.2485709548514405e-06,
+      "loss": 1.9231,
+      "step": 10899
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5811166266384296,
+      "learning_rate": 3.2454923156138544e-06,
+      "loss": 1.8618,
+      "step": 10900
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5001782263775642,
+      "learning_rate": 3.242415034599566e-06,
+      "loss": 2.0583,
+      "step": 10901
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5450268345550175,
+      "learning_rate": 3.239339112000686e-06,
+      "loss": 1.8481,
+      "step": 10902
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5675184993638455,
+      "learning_rate": 3.2362645480092517e-06,
+      "loss": 1.8528,
+      "step": 10903
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5609127780451145,
+      "learning_rate": 3.233191342817243e-06,
+      "loss": 1.8931,
+      "step": 10904
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5432441480349907,
+      "learning_rate": 3.2301194966165127e-06,
+      "loss": 2.055,
+      "step": 10905
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5601793007087711,
+      "learning_rate": 3.227049009598848e-06,
+      "loss": 1.8685,
+      "step": 10906
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5745000688554834,
+      "learning_rate": 3.2239798819559713e-06,
+      "loss": 1.8877,
+      "step": 10907
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5694937797266293,
+      "learning_rate": 3.2209121138794846e-06,
+      "loss": 1.917,
+      "step": 10908
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5494321123759563,
+      "learning_rate": 3.2178457055609234e-06,
+      "loss": 2.0821,
+      "step": 10909
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5613103130268279,
+      "learning_rate": 3.214780657191738e-06,
+      "loss": 1.8897,
+      "step": 10910
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5613196158389909,
+      "learning_rate": 3.2117169689632935e-06,
+      "loss": 1.9026,
+      "step": 10911
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5763871917601698,
+      "learning_rate": 3.208654641066866e-06,
+      "loss": 1.8172,
+      "step": 10912
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6088420855459398,
+      "learning_rate": 3.2055936736936493e-06,
+      "loss": 2.0673,
+      "step": 10913
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5670884902561305,
+      "learning_rate": 3.202534067034757e-06,
+      "loss": 1.916,
+      "step": 10914
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5546009904752476,
+      "learning_rate": 3.199475821281206e-06,
+      "loss": 1.8339,
+      "step": 10915
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6141315829959625,
+      "learning_rate": 3.1964189366239377e-06,
+      "loss": 1.8576,
+      "step": 10916
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5363785658436857,
+      "learning_rate": 3.1933634132538083e-06,
+      "loss": 2.0606,
+      "step": 10917
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5980089163191169,
+      "learning_rate": 3.190309251361584e-06,
+      "loss": 1.862,
+      "step": 10918
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5687030982623018,
+      "learning_rate": 3.1872564511379517e-06,
+      "loss": 1.8173,
+      "step": 10919
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5407838935298391,
+      "learning_rate": 3.1842050127735057e-06,
+      "loss": 1.9234,
+      "step": 10920
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5858651323103572,
+      "learning_rate": 3.1811549364587634e-06,
+      "loss": 2.062,
+      "step": 10921
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6159956817345205,
+      "learning_rate": 3.1781062223841528e-06,
+      "loss": 1.7894,
+      "step": 10922
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.566405201962037,
+      "learning_rate": 3.1750588707400162e-06,
+      "loss": 1.868,
+      "step": 10923
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5644229331228829,
+      "learning_rate": 3.1720128817166156e-06,
+      "loss": 1.8439,
+      "step": 10924
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6059954353350857,
+      "learning_rate": 3.1689682555041208e-06,
+      "loss": 2.0537,
+      "step": 10925
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4957434842587024,
+      "learning_rate": 3.165924992292621e-06,
+      "loss": 1.9416,
+      "step": 10926
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5608216883341663,
+      "learning_rate": 3.1628830922721202e-06,
+      "loss": 1.8501,
+      "step": 10927
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5648481454963536,
+      "learning_rate": 3.159842555632539e-06,
+      "loss": 1.9111,
+      "step": 10928
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5336754229406827,
+      "learning_rate": 3.1568033825637082e-06,
+      "loss": 2.0868,
+      "step": 10929
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5313713704557796,
+      "learning_rate": 3.1537655732553768e-06,
+      "loss": 1.8309,
+      "step": 10930
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5927933687774304,
+      "learning_rate": 3.150729127897206e-06,
+      "loss": 1.8666,
+      "step": 10931
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5860486204504295,
+      "learning_rate": 3.147694046678776e-06,
+      "loss": 1.8593,
+      "step": 10932
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5233292949254459,
+      "learning_rate": 3.1446603297895784e-06,
+      "loss": 2.0888,
+      "step": 10933
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5306765012175834,
+      "learning_rate": 3.1416279774190177e-06,
+      "loss": 1.8644,
+      "step": 10934
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.542430463344358,
+      "learning_rate": 3.138596989756426e-06,
+      "loss": 1.8481,
+      "step": 10935
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5812676091764443,
+      "learning_rate": 3.135567366991024e-06,
+      "loss": 1.8621,
+      "step": 10936
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5591068149272558,
+      "learning_rate": 3.1325391093119745e-06,
+      "loss": 2.0399,
+      "step": 10937
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5547368034731127,
+      "learning_rate": 3.1295122169083506e-06,
+      "loss": 1.8667,
+      "step": 10938
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.548716081856602,
+      "learning_rate": 3.126486689969113e-06,
+      "loss": 1.935,
+      "step": 10939
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5694557164154117,
+      "learning_rate": 3.123462528683177e-06,
+      "loss": 1.8678,
+      "step": 10940
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5401284296612855,
+      "learning_rate": 3.1204397332393497e-06,
+      "loss": 2.0801,
+      "step": 10941
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5625482705614778,
+      "learning_rate": 3.117418303826344e-06,
+      "loss": 1.84,
+      "step": 10942
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5467805514481275,
+      "learning_rate": 3.114398240632818e-06,
+      "loss": 1.8563,
+      "step": 10943
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5606917962000476,
+      "learning_rate": 3.1113795438473146e-06,
+      "loss": 1.8411,
+      "step": 10944
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5234531034048913,
+      "learning_rate": 3.1083622136583028e-06,
+      "loss": 2.0926,
+      "step": 10945
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5821915255224257,
+      "learning_rate": 3.1053462502541793e-06,
+      "loss": 1.8482,
+      "step": 10946
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5831788725882885,
+      "learning_rate": 3.102331653823226e-06,
+      "loss": 1.8606,
+      "step": 10947
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5679157632610933,
+      "learning_rate": 3.0993184245536767e-06,
+      "loss": 1.8617,
+      "step": 10948
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5772493333324574,
+      "learning_rate": 3.0963065626336414e-06,
+      "loss": 2.0526,
+      "step": 10949
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6189221110985661,
+      "learning_rate": 3.093296068251164e-06,
+      "loss": 1.8168,
+      "step": 10950
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5508559474949465,
+      "learning_rate": 3.0902869415942197e-06,
+      "loss": 1.9296,
+      "step": 10951
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5825031704800143,
+      "learning_rate": 3.0872791828506635e-06,
+      "loss": 1.8297,
+      "step": 10952
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6086337161606319,
+      "learning_rate": 3.0842727922082844e-06,
+      "loss": 2.0893,
+      "step": 10953
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5409128068430991,
+      "learning_rate": 3.0812677698547955e-06,
+      "loss": 1.8471,
+      "step": 10954
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5657734012788627,
+      "learning_rate": 3.0782641159777998e-06,
+      "loss": 1.8651,
+      "step": 10955
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5812652155657768,
+      "learning_rate": 3.0752618307648308e-06,
+      "loss": 1.8285,
+      "step": 10956
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5378682695544696,
+      "learning_rate": 3.0722609144033356e-06,
+      "loss": 1.9144,
+      "step": 10957
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5585393706339652,
+      "learning_rate": 3.0692613670806724e-06,
+      "loss": 2.0567,
+      "step": 10958
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5799617958904439,
+      "learning_rate": 3.0662631889841143e-06,
+      "loss": 1.9037,
+      "step": 10959
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6310044991387034,
+      "learning_rate": 3.0632663803008555e-06,
+      "loss": 1.84,
+      "step": 10960
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5530864289185683,
+      "learning_rate": 3.0602709412179853e-06,
+      "loss": 2.0856,
+      "step": 10961
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5511277973527723,
+      "learning_rate": 3.057276871922543e-06,
+      "loss": 1.8509,
+      "step": 10962
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.534423446839176,
+      "learning_rate": 3.054284172601446e-06,
+      "loss": 1.8204,
+      "step": 10963
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6004772203396654,
+      "learning_rate": 3.0512928434415387e-06,
+      "loss": 1.9541,
+      "step": 10964
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5555100380601801,
+      "learning_rate": 3.048302884629592e-06,
+      "loss": 2.0494,
+      "step": 10965
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.556165659204902,
+      "learning_rate": 3.045314296352272e-06,
+      "loss": 1.832,
+      "step": 10966
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5470158456892792,
+      "learning_rate": 3.0423270787961727e-06,
+      "loss": 1.8445,
+      "step": 10967
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5374116575970441,
+      "learning_rate": 3.039341232147799e-06,
+      "loss": 1.8538,
+      "step": 10968
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5631786532645231,
+      "learning_rate": 3.0363567565935695e-06,
+      "loss": 2.0561,
+      "step": 10969
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5373128964890701,
+      "learning_rate": 3.0333736523198182e-06,
+      "loss": 1.9225,
+      "step": 10970
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6310472502490602,
+      "learning_rate": 3.0303919195127872e-06,
+      "loss": 1.8468,
+      "step": 10971
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5652327565739999,
+      "learning_rate": 3.027411558358645e-06,
+      "loss": 1.7827,
+      "step": 10972
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5759118657274799,
+      "learning_rate": 3.0244325690434643e-06,
+      "loss": 2.0348,
+      "step": 10973
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.60265365008215,
+      "learning_rate": 3.021454951753236e-06,
+      "loss": 1.8139,
+      "step": 10974
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5787671158801412,
+      "learning_rate": 3.0184787066738634e-06,
+      "loss": 1.8254,
+      "step": 10975
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5131282797751109,
+      "learning_rate": 3.015503833991171e-06,
+      "loss": 1.9287,
+      "step": 10976
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5752678084997072,
+      "learning_rate": 3.0125303338908895e-06,
+      "loss": 2.083,
+      "step": 10977
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5763334798925462,
+      "learning_rate": 3.0095582065586574e-06,
+      "loss": 1.8385,
+      "step": 10978
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5873987680843143,
+      "learning_rate": 3.006587452180054e-06,
+      "loss": 1.8466,
+      "step": 10979
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5741686458833027,
+      "learning_rate": 3.003618070940542e-06,
+      "loss": 1.8635,
+      "step": 10980
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5907973108532687,
+      "learning_rate": 3.0006500630255204e-06,
+      "loss": 2.1192,
+      "step": 10981
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5079033073657315,
+      "learning_rate": 2.997683428620296e-06,
+      "loss": 1.9007,
+      "step": 10982
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.578098589806834,
+      "learning_rate": 2.994718167910071e-06,
+      "loss": 1.864,
+      "step": 10983
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5473025274105937,
+      "learning_rate": 2.9917542810799976e-06,
+      "loss": 1.9249,
+      "step": 10984
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5811269206084355,
+      "learning_rate": 2.988791768315122e-06,
+      "loss": 2.0584,
+      "step": 10985
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5400387978958406,
+      "learning_rate": 2.98583062980039e-06,
+      "loss": 1.8277,
+      "step": 10986
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.610797745623506,
+      "learning_rate": 2.982870865720691e-06,
+      "loss": 1.8161,
+      "step": 10987
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5767610086673044,
+      "learning_rate": 2.979912476260821e-06,
+      "loss": 1.9089,
+      "step": 10988
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5506393541426384,
+      "learning_rate": 2.9769554616054627e-06,
+      "loss": 1.8206,
+      "step": 10989
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.555791523176043,
+      "learning_rate": 2.9739998219392608e-06,
+      "loss": 2.0572,
+      "step": 10990
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5455962531658465,
+      "learning_rate": 2.9710455574467277e-06,
+      "loss": 1.8515,
+      "step": 10991
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6155663929220291,
+      "learning_rate": 2.968092668312314e-06,
+      "loss": 1.866,
+      "step": 10992
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5563076759270154,
+      "learning_rate": 2.9651411547203937e-06,
+      "loss": 2.0706,
+      "step": 10993
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5879381629948147,
+      "learning_rate": 2.962191016855223e-06,
+      "loss": 1.81,
+      "step": 10994
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5216280954920404,
+      "learning_rate": 2.959242254901007e-06,
+      "loss": 1.8923,
+      "step": 10995
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5590860226410004,
+      "learning_rate": 2.956294869041845e-06,
+      "loss": 1.8979,
+      "step": 10996
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5498133373839367,
+      "learning_rate": 2.9533488594617435e-06,
+      "loss": 2.045,
+      "step": 10997
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5495819613580668,
+      "learning_rate": 2.9504042263446522e-06,
+      "loss": 1.8799,
+      "step": 10998
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5731031295335662,
+      "learning_rate": 2.947460969874402e-06,
+      "loss": 1.9329,
+      "step": 10999
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5467644516566684,
+      "learning_rate": 2.9445190902347543e-06,
+      "loss": 1.8432,
+      "step": 11000
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5379478976555605,
+      "learning_rate": 2.9415785876093947e-06,
+      "loss": 1.9561,
+      "step": 11001
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5411904796126497,
+      "learning_rate": 2.938639462181897e-06,
+      "loss": 2.0329,
+      "step": 11002
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5676258839137424,
+      "learning_rate": 2.935701714135769e-06,
+      "loss": 1.8489,
+      "step": 11003
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5795091086435591,
+      "learning_rate": 2.9327653436544226e-06,
+      "loss": 1.9103,
+      "step": 11004
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5523509532403,
+      "learning_rate": 2.929830350921192e-06,
+      "loss": 2.0373,
+      "step": 11005
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5790198262462639,
+      "learning_rate": 2.9268967361193193e-06,
+      "loss": 1.824,
+      "step": 11006
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5201229156966158,
+      "learning_rate": 2.923964499431964e-06,
+      "loss": 1.9696,
+      "step": 11007
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5720206230304521,
+      "learning_rate": 2.9210336410421924e-06,
+      "loss": 1.8691,
+      "step": 11008
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5365668277764181,
+      "learning_rate": 2.918104161132995e-06,
+      "loss": 2.0197,
+      "step": 11009
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5939603982323544,
+      "learning_rate": 2.91517605988727e-06,
+      "loss": 1.8698,
+      "step": 11010
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5519467082479718,
+      "learning_rate": 2.9122493374878295e-06,
+      "loss": 1.8179,
+      "step": 11011
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5525193006052546,
+      "learning_rate": 2.9093239941173994e-06,
+      "loss": 1.8397,
+      "step": 11012
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5273563961587224,
+      "learning_rate": 2.906400029958625e-06,
+      "loss": 1.9815,
+      "step": 11013
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5924625918768084,
+      "learning_rate": 2.9034774451940612e-06,
+      "loss": 2.0666,
+      "step": 11014
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5778011474494228,
+      "learning_rate": 2.900556240006172e-06,
+      "loss": 1.8578,
+      "step": 11015
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6480074084187046,
+      "learning_rate": 2.8976364145773455e-06,
+      "loss": 1.8581,
+      "step": 11016
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5763887235236398,
+      "learning_rate": 2.8947179690898773e-06,
+      "loss": 2.0487,
+      "step": 11017
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5742864038819305,
+      "learning_rate": 2.8918009037259746e-06,
+      "loss": 1.8369,
+      "step": 11018
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.540195170980317,
+      "learning_rate": 2.888885218667767e-06,
+      "loss": 1.9255,
+      "step": 11019
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5701069033442724,
+      "learning_rate": 2.8859709140972886e-06,
+      "loss": 1.8987,
+      "step": 11020
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.554268675967514,
+      "learning_rate": 2.8830579901964916e-06,
+      "loss": 1.8862,
+      "step": 11021
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.590149951011285,
+      "learning_rate": 2.8801464471472446e-06,
+      "loss": 2.0442,
+      "step": 11022
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5372193084564831,
+      "learning_rate": 2.877236285131324e-06,
+      "loss": 1.8532,
+      "step": 11023
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5839194672203313,
+      "learning_rate": 2.8743275043304318e-06,
+      "loss": 1.8678,
+      "step": 11024
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5454407550459932,
+      "learning_rate": 2.8714201049261562e-06,
+      "loss": 2.051,
+      "step": 11025
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5287070490814902,
+      "learning_rate": 2.8685140871000355e-06,
+      "loss": 1.9101,
+      "step": 11026
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5658380179881011,
+      "learning_rate": 2.8656094510335046e-06,
+      "loss": 1.802,
+      "step": 11027
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.586186235560385,
+      "learning_rate": 2.8627061969078967e-06,
+      "loss": 1.8452,
+      "step": 11028
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5434137414814252,
+      "learning_rate": 2.8598043249044916e-06,
+      "loss": 2.0927,
+      "step": 11029
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5818504002713629,
+      "learning_rate": 2.856903835204447e-06,
+      "loss": 1.8513,
+      "step": 11030
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5455938290187642,
+      "learning_rate": 2.8540047279888652e-06,
+      "loss": 1.8966,
+      "step": 11031
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5180221728074166,
+      "learning_rate": 2.851107003438755e-06,
+      "loss": 1.9282,
+      "step": 11032
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5802955978263918,
+      "learning_rate": 2.8482106617350125e-06,
+      "loss": 1.8015,
+      "step": 11033
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.543842652437521,
+      "learning_rate": 2.8453157030584847e-06,
+      "loss": 2.0644,
+      "step": 11034
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5894892583123816,
+      "learning_rate": 2.8424221275899165e-06,
+      "loss": 1.8706,
+      "step": 11035
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5479315470375944,
+      "learning_rate": 2.839529935509952e-06,
+      "loss": 1.854,
+      "step": 11036
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.546320347066512,
+      "learning_rate": 2.83663912699918e-06,
+      "loss": 2.074,
+      "step": 11037
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5693772512103398,
+      "learning_rate": 2.8337497022380698e-06,
+      "loss": 1.9029,
+      "step": 11038
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5787814168398973,
+      "learning_rate": 2.8308616614070248e-06,
+      "loss": 1.8227,
+      "step": 11039
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5663286254562697,
+      "learning_rate": 2.827975004686367e-06,
+      "loss": 1.8353,
+      "step": 11040
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5330453057641804,
+      "learning_rate": 2.825089732256311e-06,
+      "loss": 1.8034,
+      "step": 11041
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5325817204145,
+      "learning_rate": 2.8222058442969963e-06,
+      "loss": 2.0201,
+      "step": 11042
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.552892674884264,
+      "learning_rate": 2.819323340988486e-06,
+      "loss": 1.8741,
+      "step": 11043
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5596738376664717,
+      "learning_rate": 2.8164422225107345e-06,
+      "loss": 1.9557,
+      "step": 11044
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5512342164996215,
+      "learning_rate": 2.8135624890436335e-06,
+      "loss": 1.9056,
+      "step": 11045
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5943153710304545,
+      "learning_rate": 2.8106841407669673e-06,
+      "loss": 2.0774,
+      "step": 11046
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5800654507401186,
+      "learning_rate": 2.8078071778604416e-06,
+      "loss": 1.9021,
+      "step": 11047
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5718435752845573,
+      "learning_rate": 2.8049316005036906e-06,
+      "loss": 1.8718,
+      "step": 11048
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5714808889154996,
+      "learning_rate": 2.8020574088762326e-06,
+      "loss": 2.0614,
+      "step": 11049
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5344693755388021,
+      "learning_rate": 2.7991846031575257e-06,
+      "loss": 1.9715,
+      "step": 11050
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5713569354584445,
+      "learning_rate": 2.7963131835269247e-06,
+      "loss": 1.8571,
+      "step": 11051
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5644392717326377,
+      "learning_rate": 2.7934431501637076e-06,
+      "loss": 1.8638,
+      "step": 11052
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5820972981252023,
+      "learning_rate": 2.790574503247062e-06,
+      "loss": 1.8757,
+      "step": 11053
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5154199781045713,
+      "learning_rate": 2.787707242956086e-06,
+      "loss": 2.0486,
+      "step": 11054
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5784381480450455,
+      "learning_rate": 2.784841369469801e-06,
+      "loss": 1.8639,
+      "step": 11055
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5704443151540566,
+      "learning_rate": 2.78197688296713e-06,
+      "loss": 1.8681,
+      "step": 11056
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5520884808935981,
+      "learning_rate": 2.779113783626916e-06,
+      "loss": 1.9462,
+      "step": 11057
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5697594945467133,
+      "learning_rate": 2.7762520716279135e-06,
+      "loss": 2.0659,
+      "step": 11058
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.579430920283695,
+      "learning_rate": 2.773391747148793e-06,
+      "loss": 1.897,
+      "step": 11059
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.593009555462676,
+      "learning_rate": 2.7705328103681317e-06,
+      "loss": 1.8401,
+      "step": 11060
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5733792092830888,
+      "learning_rate": 2.767675261464431e-06,
+      "loss": 2.0061,
+      "step": 11061
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.553163924029067,
+      "learning_rate": 2.764819100616095e-06,
+      "loss": 1.8295,
+      "step": 11062
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5909269760666099,
+      "learning_rate": 2.7619643280014486e-06,
+      "loss": 1.9441,
+      "step": 11063
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6127396813111059,
+      "learning_rate": 2.759110943798726e-06,
+      "loss": 1.8773,
+      "step": 11064
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5813109143152801,
+      "learning_rate": 2.756258948186072e-06,
+      "loss": 1.8449,
+      "step": 11065
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5424026098165504,
+      "learning_rate": 2.753408341341554e-06,
+      "loss": 2.0664,
+      "step": 11066
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5736207285773557,
+      "learning_rate": 2.7505591234431414e-06,
+      "loss": 1.8714,
+      "step": 11067
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5671986559764755,
+      "learning_rate": 2.7477112946687283e-06,
+      "loss": 1.8405,
+      "step": 11068
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5592837508433316,
+      "learning_rate": 2.744864855196114e-06,
+      "loss": 1.9559,
+      "step": 11069
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5642971883243525,
+      "learning_rate": 2.742019805203014e-06,
+      "loss": 2.0829,
+      "step": 11070
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5550203508228634,
+      "learning_rate": 2.7391761448670587e-06,
+      "loss": 1.8859,
+      "step": 11071
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5598563273713845,
+      "learning_rate": 2.736333874365776e-06,
+      "loss": 1.8646,
+      "step": 11072
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6150124921818244,
+      "learning_rate": 2.733492993876638e-06,
+      "loss": 1.8223,
+      "step": 11073
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5560040065382178,
+      "learning_rate": 2.7306535035770077e-06,
+      "loss": 2.0464,
+      "step": 11074
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5695290748116774,
+      "learning_rate": 2.727815403644157e-06,
+      "loss": 1.9434,
+      "step": 11075
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5874578429537222,
+      "learning_rate": 2.7249786942552895e-06,
+      "loss": 1.8526,
+      "step": 11076
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6023082165974242,
+      "learning_rate": 2.72214337558751e-06,
+      "loss": 1.8827,
+      "step": 11077
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.59503292452373,
+      "learning_rate": 2.7193094478178417e-06,
+      "loss": 2.01,
+      "step": 11078
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6195899635284864,
+      "learning_rate": 2.7164769111232195e-06,
+      "loss": 1.8103,
+      "step": 11079
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5808438458319892,
+      "learning_rate": 2.713645765680478e-06,
+      "loss": 1.8664,
+      "step": 11080
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6030336493853367,
+      "learning_rate": 2.7108160116663893e-06,
+      "loss": 2.0332,
+      "step": 11081
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4978833017158907,
+      "learning_rate": 2.707987649257629e-06,
+      "loss": 1.9048,
+      "step": 11082
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5501691564461823,
+      "learning_rate": 2.705160678630769e-06,
+      "loss": 1.829,
+      "step": 11083
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5842797071934983,
+      "learning_rate": 2.7023350999623246e-06,
+      "loss": 1.85,
+      "step": 11084
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6227716556008069,
+      "learning_rate": 2.699510913428699e-06,
+      "loss": 1.903,
+      "step": 11085
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5997720983786747,
+      "learning_rate": 2.6966881192062116e-06,
+      "loss": 2.0441,
+      "step": 11086
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5842314093446619,
+      "learning_rate": 2.693866717471122e-06,
+      "loss": 1.9234,
+      "step": 11087
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49736657373171883,
+      "learning_rate": 2.691046708399561e-06,
+      "loss": 1.8658,
+      "step": 11088
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5608805639727138,
+      "learning_rate": 2.6882280921675996e-06,
+      "loss": 1.832,
+      "step": 11089
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5547555250748492,
+      "learning_rate": 2.685410868951227e-06,
+      "loss": 2.0387,
+      "step": 11090
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.545393927578426,
+      "learning_rate": 2.682595038926314e-06,
+      "loss": 1.8974,
+      "step": 11091
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5538317197948686,
+      "learning_rate": 2.6797806022686834e-06,
+      "loss": 1.8561,
+      "step": 11092
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5974283220348613,
+      "learning_rate": 2.6769675591540394e-06,
+      "loss": 2.022,
+      "step": 11093
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5221108121006991,
+      "learning_rate": 2.674155909758011e-06,
+      "loss": 1.9385,
+      "step": 11094
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6037087757415877,
+      "learning_rate": 2.6713456542561575e-06,
+      "loss": 1.821,
+      "step": 11095
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5604649967570153,
+      "learning_rate": 2.668536792823917e-06,
+      "loss": 1.8948,
+      "step": 11096
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6123391002984477,
+      "learning_rate": 2.665729325636665e-06,
+      "loss": 1.7996,
+      "step": 11097
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.537079244148139,
+      "learning_rate": 2.6629232528696814e-06,
+      "loss": 2.0488,
+      "step": 11098
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5524344482271178,
+      "learning_rate": 2.6601185746981644e-06,
+      "loss": 1.8586,
+      "step": 11099
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5460961107505126,
+      "learning_rate": 2.657315291297219e-06,
+      "loss": 1.9524,
+      "step": 11100
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5326339762230258,
+      "learning_rate": 2.654513402841863e-06,
+      "loss": 1.8351,
+      "step": 11101
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5703173537087283,
+      "learning_rate": 2.651712909507034e-06,
+      "loss": 2.056,
+      "step": 11102
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5439627309345092,
+      "learning_rate": 2.6489138114675795e-06,
+      "loss": 1.8461,
+      "step": 11103
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.596770569540055,
+      "learning_rate": 2.6461161088982555e-06,
+      "loss": 1.8309,
+      "step": 11104
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5574797731060114,
+      "learning_rate": 2.643319801973734e-06,
+      "loss": 1.822,
+      "step": 11105
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5144228978523118,
+      "learning_rate": 2.6405248908686008e-06,
+      "loss": 2.0973,
+      "step": 11106
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5613189547104983,
+      "learning_rate": 2.637731375757352e-06,
+      "loss": 1.8585,
+      "step": 11107
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5770588056935186,
+      "learning_rate": 2.634939256814403e-06,
+      "loss": 1.8428,
+      "step": 11108
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6226534630888009,
+      "learning_rate": 2.632148534214074e-06,
+      "loss": 1.8736,
+      "step": 11109
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.61972989152355,
+      "learning_rate": 2.6293592081305994e-06,
+      "loss": 2.018,
+      "step": 11110
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5801818192202404,
+      "learning_rate": 2.6265712787381297e-06,
+      "loss": 1.8536,
+      "step": 11111
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5564235419488669,
+      "learning_rate": 2.6237847462107305e-06,
+      "loss": 1.8799,
+      "step": 11112
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5395678256829634,
+      "learning_rate": 2.6209996107223718e-06,
+      "loss": 1.9146,
+      "step": 11113
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6091093964342706,
+      "learning_rate": 2.6182158724469413e-06,
+      "loss": 2.0359,
+      "step": 11114
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5667788613242687,
+      "learning_rate": 2.6154335315582423e-06,
+      "loss": 1.8653,
+      "step": 11115
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5751331994344332,
+      "learning_rate": 2.612652588229986e-06,
+      "loss": 1.8962,
+      "step": 11116
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5315577363396237,
+      "learning_rate": 2.609873042635799e-06,
+      "loss": 1.8471,
+      "step": 11117
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.527203059812803,
+      "learning_rate": 2.607094894949222e-06,
+      "loss": 2.0455,
+      "step": 11118
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5589480608292238,
+      "learning_rate": 2.6043181453436953e-06,
+      "loss": 1.9284,
+      "step": 11119
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5503262045370972,
+      "learning_rate": 2.6015427939925923e-06,
+      "loss": 1.9212,
+      "step": 11120
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5866505630335165,
+      "learning_rate": 2.598768841069196e-06,
+      "loss": 1.8739,
+      "step": 11121
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5673783842449909,
+      "learning_rate": 2.595996286746674e-06,
+      "loss": 2.0257,
+      "step": 11122
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5654470605003413,
+      "learning_rate": 2.5932251311981513e-06,
+      "loss": 1.8353,
+      "step": 11123
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5656856236489031,
+      "learning_rate": 2.590455374596634e-06,
+      "loss": 1.9151,
+      "step": 11124
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5199376742175843,
+      "learning_rate": 2.58768701711504e-06,
+      "loss": 1.9037,
+      "step": 11125
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5839023796593644,
+      "learning_rate": 2.584920058926227e-06,
+      "loss": 2.0741,
+      "step": 11126
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.581743288866878,
+      "learning_rate": 2.5821545002029298e-06,
+      "loss": 1.865,
+      "step": 11127
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5656500166113706,
+      "learning_rate": 2.5793903411178268e-06,
+      "loss": 1.8409,
+      "step": 11128
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5799791227023938,
+      "learning_rate": 2.5766275818434936e-06,
+      "loss": 1.8291,
+      "step": 11129
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5750515017697811,
+      "learning_rate": 2.5738662225524096e-06,
+      "loss": 2.0255,
+      "step": 11130
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5020256537431663,
+      "learning_rate": 2.5711062634169975e-06,
+      "loss": 1.8861,
+      "step": 11131
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.542742324875227,
+      "learning_rate": 2.5683477046095557e-06,
+      "loss": 1.8566,
+      "step": 11132
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5690100595693123,
+      "learning_rate": 2.5655905463023127e-06,
+      "loss": 1.8812,
+      "step": 11133
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5718892803191697,
+      "learning_rate": 2.5628347886674232e-06,
+      "loss": 2.055,
+      "step": 11134
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5714850464725609,
+      "learning_rate": 2.5600804318769295e-06,
+      "loss": 1.8915,
+      "step": 11135
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5487985890329029,
+      "learning_rate": 2.5573274761027944e-06,
+      "loss": 1.8847,
+      "step": 11136
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5385141133625567,
+      "learning_rate": 2.5545759215169106e-06,
+      "loss": 1.9681,
+      "step": 11137
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5571988797925721,
+      "learning_rate": 2.551825768291055e-06,
+      "loss": 2.0664,
+      "step": 11138
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5812598552715662,
+      "learning_rate": 2.5490770165969375e-06,
+      "loss": 1.8618,
+      "step": 11139
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.546920929007709,
+      "learning_rate": 2.546329666606173e-06,
+      "loss": 1.8501,
+      "step": 11140
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5826021627139846,
+      "learning_rate": 2.543583718490283e-06,
+      "loss": 1.8611,
+      "step": 11141
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5394600820489752,
+      "learning_rate": 2.540839172420725e-06,
+      "loss": 2.03,
+      "step": 11142
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5624079003924163,
+      "learning_rate": 2.538096028568837e-06,
+      "loss": 1.8547,
+      "step": 11143
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5166128284060002,
+      "learning_rate": 2.5353542871058844e-06,
+      "loss": 1.8952,
+      "step": 11144
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5802555013582236,
+      "learning_rate": 2.5326139482030638e-06,
+      "loss": 1.8134,
+      "step": 11145
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5471388468595368,
+      "learning_rate": 2.5298750120314437e-06,
+      "loss": 2.0419,
+      "step": 11146
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5428048842060942,
+      "learning_rate": 2.527137478762037e-06,
+      "loss": 1.8362,
+      "step": 11147
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5641489798165573,
+      "learning_rate": 2.524401348565761e-06,
+      "loss": 1.7885,
+      "step": 11148
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5509013513429136,
+      "learning_rate": 2.521666621613439e-06,
+      "loss": 1.8314,
+      "step": 11149
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49930698100206855,
+      "learning_rate": 2.5189332980758124e-06,
+      "loss": 2.026,
+      "step": 11150
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.607406993017461,
+      "learning_rate": 2.5162013781235337e-06,
+      "loss": 1.9186,
+      "step": 11151
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5829267850116784,
+      "learning_rate": 2.5134708619271722e-06,
+      "loss": 1.8205,
+      "step": 11152
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6433692587466404,
+      "learning_rate": 2.5107417496572e-06,
+      "loss": 1.8933,
+      "step": 11153
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5270570066146664,
+      "learning_rate": 2.508014041484008e-06,
+      "loss": 2.0295,
+      "step": 11154
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5701922517705831,
+      "learning_rate": 2.5052877375778995e-06,
+      "loss": 1.8695,
+      "step": 11155
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5192785253260583,
+      "learning_rate": 2.5025628381090877e-06,
+      "loss": 1.9016,
+      "step": 11156
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5560929580792876,
+      "learning_rate": 2.499839343247698e-06,
+      "loss": 1.8535,
+      "step": 11157
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5930083717615772,
+      "learning_rate": 2.4971172531637747e-06,
+      "loss": 2.0489,
+      "step": 11158
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5545289073908967,
+      "learning_rate": 2.494396568027263e-06,
+      "loss": 1.8454,
+      "step": 11159
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.584851600785787,
+      "learning_rate": 2.491677288008029e-06,
+      "loss": 1.8323,
+      "step": 11160
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5569525378811834,
+      "learning_rate": 2.488959413275849e-06,
+      "loss": 1.8851,
+      "step": 11161
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5395363285090012,
+      "learning_rate": 2.4862429440004108e-06,
+      "loss": 2.0596,
+      "step": 11162
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5730595459105783,
+      "learning_rate": 2.4835278803513136e-06,
+      "loss": 1.8707,
+      "step": 11163
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5550563237059293,
+      "learning_rate": 2.4808142224980733e-06,
+      "loss": 1.8547,
+      "step": 11164
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5399600083213264,
+      "learning_rate": 2.478101970610114e-06,
+      "loss": 1.8785,
+      "step": 11165
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6212810292663378,
+      "learning_rate": 2.475391124856763e-06,
+      "loss": 2.065,
+      "step": 11166
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5650863642847908,
+      "learning_rate": 2.4726816854072797e-06,
+      "loss": 1.8569,
+      "step": 11167
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.537784248189289,
+      "learning_rate": 2.469973652430832e-06,
+      "loss": 1.913,
+      "step": 11168
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5544963826747129,
+      "learning_rate": 2.467267026096473e-06,
+      "loss": 1.8637,
+      "step": 11169
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5882843999607376,
+      "learning_rate": 2.464561806573204e-06,
+      "loss": 2.0303,
+      "step": 11170
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5344342368678255,
+      "learning_rate": 2.4618579940299256e-06,
+      "loss": 1.866,
+      "step": 11171
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5335356737719169,
+      "learning_rate": 2.459155588635434e-06,
+      "loss": 1.8494,
+      "step": 11172
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5591720025866959,
+      "learning_rate": 2.4564545905584657e-06,
+      "loss": 1.8787,
+      "step": 11173
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5624590075536152,
+      "learning_rate": 2.453754999967639e-06,
+      "loss": 2.0874,
+      "step": 11174
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5660977564009094,
+      "learning_rate": 2.451056817031516e-06,
+      "loss": 1.8929,
+      "step": 11175
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5927977742228033,
+      "learning_rate": 2.4483600419185542e-06,
+      "loss": 1.8572,
+      "step": 11176
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5549569209061838,
+      "learning_rate": 2.4456646747971102e-06,
+      "loss": 1.867,
+      "step": 11177
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5562491030038597,
+      "learning_rate": 2.4429707158354853e-06,
+      "loss": 2.0772,
+      "step": 11178
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6960403182335471,
+      "learning_rate": 2.440278165201859e-06,
+      "loss": 1.8104,
+      "step": 11179
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5572973378499211,
+      "learning_rate": 2.4375870230643417e-06,
+      "loss": 1.8433,
+      "step": 11180
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5408198478186312,
+      "learning_rate": 2.434897289590965e-06,
+      "loss": 1.9341,
+      "step": 11181
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5521888656464179,
+      "learning_rate": 2.432208964949642e-06,
+      "loss": 2.0475,
+      "step": 11182
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5415081898986202,
+      "learning_rate": 2.429522049308222e-06,
+      "loss": 1.8598,
+      "step": 11183
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5739269056097686,
+      "learning_rate": 2.4268365428344736e-06,
+      "loss": 1.8756,
+      "step": 11184
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5607125664400467,
+      "learning_rate": 2.4241524456960457e-06,
+      "loss": 1.7849,
+      "step": 11185
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5317577450246797,
+      "learning_rate": 2.4214697580605237e-06,
+      "loss": 2.084,
+      "step": 11186
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.49680180249323114,
+      "learning_rate": 2.418788480095402e-06,
+      "loss": 1.9178,
+      "step": 11187
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5510305202390415,
+      "learning_rate": 2.416108611968079e-06,
+      "loss": 1.8763,
+      "step": 11188
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5638856389418918,
+      "learning_rate": 2.4134301538458747e-06,
+      "loss": 1.8331,
+      "step": 11189
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5402537285074656,
+      "learning_rate": 2.4107531058960104e-06,
+      "loss": 2.0089,
+      "step": 11190
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5493474489942526,
+      "learning_rate": 2.408077468285627e-06,
+      "loss": 1.8438,
+      "step": 11191
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5497207178117967,
+      "learning_rate": 2.405403241181786e-06,
+      "loss": 1.8629,
+      "step": 11192
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5228326010400871,
+      "learning_rate": 2.402730424751437e-06,
+      "loss": 1.9137,
+      "step": 11193
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5415726696041544,
+      "learning_rate": 2.4000590191614574e-06,
+      "loss": 2.0364,
+      "step": 11194
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5532820321452955,
+      "learning_rate": 2.397389024578639e-06,
+      "loss": 1.8601,
+      "step": 11195
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5767398956765146,
+      "learning_rate": 2.394720441169676e-06,
+      "loss": 1.895,
+      "step": 11196
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5523924828731,
+      "learning_rate": 2.3920532691011825e-06,
+      "loss": 1.8137,
+      "step": 11197
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5371665340955613,
+      "learning_rate": 2.389387508539678e-06,
+      "loss": 2.0274,
+      "step": 11198
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5144604742738368,
+      "learning_rate": 2.3867231596515964e-06,
+      "loss": 1.9241,
+      "step": 11199
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.563427837684983,
+      "learning_rate": 2.3840602226032875e-06,
+      "loss": 1.8656,
+      "step": 11200
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.543917409932888,
+      "learning_rate": 2.3813986975610077e-06,
+      "loss": 1.8236,
+      "step": 11201
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5466573688013141,
+      "learning_rate": 2.378738584690926e-06,
+      "loss": 2.0644,
+      "step": 11202
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5434375361939563,
+      "learning_rate": 2.3760798841591243e-06,
+      "loss": 1.8917,
+      "step": 11203
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5573268266755599,
+      "learning_rate": 2.373422596131597e-06,
+      "loss": 1.834,
+      "step": 11204
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5551949520423707,
+      "learning_rate": 2.3707667207742506e-06,
+      "loss": 1.8884,
+      "step": 11205
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5108744672462963,
+      "learning_rate": 2.3681122582529e-06,
+      "loss": 1.9497,
+      "step": 11206
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5636372601658254,
+      "learning_rate": 2.3654592087332755e-06,
+      "loss": 2.056,
+      "step": 11207
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5650551694457353,
+      "learning_rate": 2.362807572381018e-06,
+      "loss": 1.8548,
+      "step": 11208
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5544478434957251,
+      "learning_rate": 2.36015734936168e-06,
+      "loss": 1.8559,
+      "step": 11209
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5408471781778638,
+      "learning_rate": 2.3575085398407247e-06,
+      "loss": 2.0323,
+      "step": 11210
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.549856534880969,
+      "learning_rate": 2.3548611439835305e-06,
+      "loss": 1.8444,
+      "step": 11211
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5227804950916602,
+      "learning_rate": 2.3522151619553875e-06,
+      "loss": 1.944,
+      "step": 11212
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6072816110239908,
+      "learning_rate": 2.34957059392148e-06,
+      "loss": 1.8504,
+      "step": 11213
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5289102551760602,
+      "learning_rate": 2.3469274400469366e-06,
+      "loss": 2.0188,
+      "step": 11214
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5399085851550032,
+      "learning_rate": 2.344285700496779e-06,
+      "loss": 1.8554,
+      "step": 11215
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5679989989771184,
+      "learning_rate": 2.3416453754359273e-06,
+      "loss": 1.8708,
+      "step": 11216
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5481679765093012,
+      "learning_rate": 2.3390064650292447e-06,
+      "loss": 1.8268,
+      "step": 11217
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5532323108883072,
+      "learning_rate": 2.336368969441485e-06,
+      "loss": 1.9146,
+      "step": 11218
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5825307857881219,
+      "learning_rate": 2.333732888837306e-06,
+      "loss": 2.0575,
+      "step": 11219
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5613955901073903,
+      "learning_rate": 2.3310982233813063e-06,
+      "loss": 1.8782,
+      "step": 11220
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5576089643400929,
+      "learning_rate": 2.3284649732379677e-06,
+      "loss": 1.8228,
+      "step": 11221
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5413130156732641,
+      "learning_rate": 2.3258331385716926e-06,
+      "loss": 2.0829,
+      "step": 11222
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.625205475949688,
+      "learning_rate": 2.3232027195468104e-06,
+      "loss": 1.8306,
+      "step": 11223
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5702348429046165,
+      "learning_rate": 2.3205737163275343e-06,
+      "loss": 1.9641,
+      "step": 11224
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5606726303290873,
+      "learning_rate": 2.3179461290780137e-06,
+      "loss": 1.8508,
+      "step": 11225
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5429020737250992,
+      "learning_rate": 2.3153199579623e-06,
+      "loss": 2.0305,
+      "step": 11226
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5759606799846586,
+      "learning_rate": 2.3126952031443467e-06,
+      "loss": 1.8796,
+      "step": 11227
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5431037186086397,
+      "learning_rate": 2.310071864788041e-06,
+      "loss": 1.8805,
+      "step": 11228
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5793859567745122,
+      "learning_rate": 2.307449943057155e-06,
+      "loss": 1.8562,
+      "step": 11229
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5391712122048643,
+      "learning_rate": 2.3048294381153885e-06,
+      "loss": 2.048,
+      "step": 11230
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5368238269759552,
+      "learning_rate": 2.3022103501263657e-06,
+      "loss": 1.9706,
+      "step": 11231
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5889449249856246,
+      "learning_rate": 2.299592679253587e-06,
+      "loss": 1.805,
+      "step": 11232
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5723530325559715,
+      "learning_rate": 2.2969764256604965e-06,
+      "loss": 1.8663,
+      "step": 11233
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5322989001498653,
+      "learning_rate": 2.2943615895104325e-06,
+      "loss": 2.0234,
+      "step": 11234
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5472885639284986,
+      "learning_rate": 2.291748170966651e-06,
+      "loss": 1.8409,
+      "step": 11235
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5765461213881032,
+      "learning_rate": 2.2891361701923213e-06,
+      "loss": 1.8413,
+      "step": 11236
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5162892566322802,
+      "learning_rate": 2.2865255873505185e-06,
+      "loss": 1.9394,
+      "step": 11237
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5544108274293472,
+      "learning_rate": 2.2839164226042266e-06,
+      "loss": 1.8491,
+      "step": 11238
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5530816195654853,
+      "learning_rate": 2.2813086761163614e-06,
+      "loss": 2.0496,
+      "step": 11239
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5559668185007612,
+      "learning_rate": 2.2787023480497246e-06,
+      "loss": 1.863,
+      "step": 11240
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5561916753836837,
+      "learning_rate": 2.2760974385670377e-06,
+      "loss": 1.8567,
+      "step": 11241
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5433597473236561,
+      "learning_rate": 2.2734939478309432e-06,
+      "loss": 2.0422,
+      "step": 11242
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5310572430934506,
+      "learning_rate": 2.270891876003983e-06,
+      "loss": 1.9369,
+      "step": 11243
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5518141067798691,
+      "learning_rate": 2.2682912232486143e-06,
+      "loss": 1.776,
+      "step": 11244
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5677971958347475,
+      "learning_rate": 2.2656919897272115e-06,
+      "loss": 1.8395,
+      "step": 11245
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.537189654853249,
+      "learning_rate": 2.2630941756020512e-06,
+      "loss": 2.0076,
+      "step": 11246
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5532704582541089,
+      "learning_rate": 2.2604977810353286e-06,
+      "loss": 1.8821,
+      "step": 11247
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5453766876422301,
+      "learning_rate": 2.2579028061891465e-06,
+      "loss": 1.8514,
+      "step": 11248
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5185142381893495,
+      "learning_rate": 2.2553092512255153e-06,
+      "loss": 1.9292,
+      "step": 11249
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5655559500521713,
+      "learning_rate": 2.2527171163063687e-06,
+      "loss": 1.8736,
+      "step": 11250
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5403050542895828,
+      "learning_rate": 2.2501264015935387e-06,
+      "loss": 2.0669,
+      "step": 11251
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5315193236127808,
+      "learning_rate": 2.2475371072487762e-06,
+      "loss": 1.7833,
+      "step": 11252
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5524684059112607,
+      "learning_rate": 2.2449492334337436e-06,
+      "loss": 1.8304,
+      "step": 11253
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5562949745471019,
+      "learning_rate": 2.2423627803100145e-06,
+      "loss": 2.0566,
+      "step": 11254
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5431313701969532,
+      "learning_rate": 2.239777748039057e-06,
+      "loss": 1.977,
+      "step": 11255
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5889820961085118,
+      "learning_rate": 2.237194136782281e-06,
+      "loss": 1.8436,
+      "step": 11256
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5430057266518863,
+      "learning_rate": 2.234611946700987e-06,
+      "loss": 1.8809,
+      "step": 11257
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5540410499123491,
+      "learning_rate": 2.2320311779563924e-06,
+      "loss": 1.8056,
+      "step": 11258
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5392564061816609,
+      "learning_rate": 2.229451830709625e-06,
+      "loss": 2.0326,
+      "step": 11259
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5704134989746348,
+      "learning_rate": 2.2268739051217236e-06,
+      "loss": 1.8539,
+      "step": 11260
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5519489409415285,
+      "learning_rate": 2.2242974013536365e-06,
+      "loss": 1.8195,
+      "step": 11261
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5405882761437483,
+      "learning_rate": 2.221722319566233e-06,
+      "loss": 1.9229,
+      "step": 11262
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5465540579041684,
+      "learning_rate": 2.2191486599202726e-06,
+      "loss": 2.0346,
+      "step": 11263
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5618174097076192,
+      "learning_rate": 2.2165764225764524e-06,
+      "loss": 1.805,
+      "step": 11264
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5387537874727099,
+      "learning_rate": 2.2140056076953658e-06,
+      "loss": 1.8417,
+      "step": 11265
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.557298248759387,
+      "learning_rate": 2.211436215437507e-06,
+      "loss": 2.0192,
+      "step": 11266
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5592222717797685,
+      "learning_rate": 2.2088682459633137e-06,
+      "loss": 1.8461,
+      "step": 11267
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5176376167307225,
+      "learning_rate": 2.206301699433097e-06,
+      "loss": 1.9796,
+      "step": 11268
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5484814386352039,
+      "learning_rate": 2.203736576007104e-06,
+      "loss": 1.8503,
+      "step": 11269
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5614535391894523,
+      "learning_rate": 2.201172875845492e-06,
+      "loss": 1.8505,
+      "step": 11270
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5659117374812458,
+      "learning_rate": 2.1986105991083084e-06,
+      "loss": 2.0148,
+      "step": 11271
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5308201278777612,
+      "learning_rate": 2.196049745955539e-06,
+      "loss": 1.8477,
+      "step": 11272
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6116829807195332,
+      "learning_rate": 2.19349031654707e-06,
+      "loss": 1.8185,
+      "step": 11273
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5429642187792052,
+      "learning_rate": 2.190932311042684e-06,
+      "loss": 1.8634,
+      "step": 11274
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5489604407823094,
+      "learning_rate": 2.1883757296021045e-06,
+      "loss": 2.0837,
+      "step": 11275
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5470825973465622,
+      "learning_rate": 2.1858205723849362e-06,
+      "loss": 1.9031,
+      "step": 11276
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5579022900911772,
+      "learning_rate": 2.1832668395507074e-06,
+      "loss": 1.8191,
+      "step": 11277
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.572529990047102,
+      "learning_rate": 2.180714531258873e-06,
+      "loss": 2.0263,
+      "step": 11278
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5616079358340998,
+      "learning_rate": 2.178163647668771e-06,
+      "loss": 1.8397,
+      "step": 11279
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5269903301876449,
+      "learning_rate": 2.1756141889396647e-06,
+      "loss": 1.9306,
+      "step": 11280
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5694279824480117,
+      "learning_rate": 2.1730661552307323e-06,
+      "loss": 1.8678,
+      "step": 11281
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5545374310775893,
+      "learning_rate": 2.1705195467010525e-06,
+      "loss": 1.8395,
+      "step": 11282
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5872291465127009,
+      "learning_rate": 2.1679743635096255e-06,
+      "loss": 2.0756,
+      "step": 11283
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5864343754941461,
+      "learning_rate": 2.1654306058153522e-06,
+      "loss": 1.8858,
+      "step": 11284
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.565070031796905,
+      "learning_rate": 2.1628882737770555e-06,
+      "loss": 1.8763,
+      "step": 11285
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5125167296322625,
+      "learning_rate": 2.1603473675534614e-06,
+      "loss": 1.9787,
+      "step": 11286
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5446003377478872,
+      "learning_rate": 2.1578078873032063e-06,
+      "loss": 2.0932,
+      "step": 11287
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.568133999438589,
+      "learning_rate": 2.1552698331848447e-06,
+      "loss": 1.9141,
+      "step": 11288
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5488982915893751,
+      "learning_rate": 2.1527332053568357e-06,
+      "loss": 1.8652,
+      "step": 11289
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5873104977773866,
+      "learning_rate": 2.150198003977552e-06,
+      "loss": 1.8005,
+      "step": 11290
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.52261890273868,
+      "learning_rate": 2.1476642292052763e-06,
+      "loss": 2.047,
+      "step": 11291
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5647540322864588,
+      "learning_rate": 2.1451318811982006e-06,
+      "loss": 1.8114,
+      "step": 11292
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.535383911330196,
+      "learning_rate": 2.1426009601144326e-06,
+      "loss": 1.959,
+      "step": 11293
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.537429994868881,
+      "learning_rate": 2.1400714661119897e-06,
+      "loss": 1.7978,
+      "step": 11294
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5743541758929745,
+      "learning_rate": 2.137543399348793e-06,
+      "loss": 2.0562,
+      "step": 11295
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5513022731371093,
+      "learning_rate": 2.135016759982683e-06,
+      "loss": 1.8234,
+      "step": 11296
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5576755971816011,
+      "learning_rate": 2.1324915481714113e-06,
+      "loss": 1.8453,
+      "step": 11297
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.559126085397488,
+      "learning_rate": 2.129967764072632e-06,
+      "loss": 2.0561,
+      "step": 11298
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5393568575597373,
+      "learning_rate": 2.127445407843917e-06,
+      "loss": 1.9609,
+      "step": 11299
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5512449096638727,
+      "learning_rate": 2.1249244796427508e-06,
+      "loss": 1.8465,
+      "step": 11300
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5332588795434223,
+      "learning_rate": 2.1224049796265246e-06,
+      "loss": 1.7936,
+      "step": 11301
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5378823733753211,
+      "learning_rate": 2.119886907952531e-06,
+      "loss": 1.8513,
+      "step": 11302
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5529996483875359,
+      "learning_rate": 2.117370264777996e-06,
+      "loss": 2.0297,
+      "step": 11303
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5598368815471738,
+      "learning_rate": 2.1148550502600396e-06,
+      "loss": 1.8486,
+      "step": 11304
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49884275269143374,
+      "learning_rate": 2.112341264555698e-06,
+      "loss": 1.9286,
+      "step": 11305
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5471970783075354,
+      "learning_rate": 2.1098289078219147e-06,
+      "loss": 1.8486,
+      "step": 11306
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5603608659261473,
+      "learning_rate": 2.107317980215548e-06,
+      "loss": 2.073,
+      "step": 11307
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5838193588011032,
+      "learning_rate": 2.104808481893367e-06,
+      "loss": 1.8544,
+      "step": 11308
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6172612241126606,
+      "learning_rate": 2.102300413012051e-06,
+      "loss": 1.8497,
+      "step": 11309
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5350741290922535,
+      "learning_rate": 2.099793773728176e-06,
+      "loss": 2.0337,
+      "step": 11310
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5498345428351682,
+      "learning_rate": 2.09728856419826e-06,
+      "loss": 1.954,
+      "step": 11311
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5452153391717481,
+      "learning_rate": 2.094784784578707e-06,
+      "loss": 1.9139,
+      "step": 11312
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5623295867506422,
+      "learning_rate": 2.09228243502583e-06,
+      "loss": 1.8498,
+      "step": 11313
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5969262846712321,
+      "learning_rate": 2.089781515695877e-06,
+      "loss": 1.8558,
+      "step": 11314
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5631674729019587,
+      "learning_rate": 2.0872820267449745e-06,
+      "loss": 2.0479,
+      "step": 11315
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5304841242386363,
+      "learning_rate": 2.0847839683291792e-06,
+      "loss": 1.8174,
+      "step": 11316
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.552456689601667,
+      "learning_rate": 2.082287340604469e-06,
+      "loss": 1.9302,
+      "step": 11317
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5686911682245245,
+      "learning_rate": 2.0797921437267055e-06,
+      "loss": 1.8951,
+      "step": 11318
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5544318525301177,
+      "learning_rate": 2.077298377851669e-06,
+      "loss": 2.0505,
+      "step": 11319
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.572812764271841,
+      "learning_rate": 2.0748060431350747e-06,
+      "loss": 1.8521,
+      "step": 11320
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5645169750106035,
+      "learning_rate": 2.0723151397325086e-06,
+      "loss": 1.8578,
+      "step": 11321
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5966568233554009,
+      "learning_rate": 2.0698256677995077e-06,
+      "loss": 1.8113,
+      "step": 11322
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5262893152324377,
+      "learning_rate": 2.0673376274914836e-06,
+      "loss": 2.0343,
+      "step": 11323
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5323368165568843,
+      "learning_rate": 2.0648510189637785e-06,
+      "loss": 1.9887,
+      "step": 11324
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5478733957343552,
+      "learning_rate": 2.0623658423716496e-06,
+      "loss": 1.8484,
+      "step": 11325
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5558995215599773,
+      "learning_rate": 2.05988209787025e-06,
+      "loss": 1.8338,
+      "step": 11326
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5793835482879174,
+      "learning_rate": 2.057399785614653e-06,
+      "loss": 2.0891,
+      "step": 11327
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.544927015199699,
+      "learning_rate": 2.0549189057598337e-06,
+      "loss": 1.8477,
+      "step": 11328
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5780353365902464,
+      "learning_rate": 2.052439458460689e-06,
+      "loss": 1.8575,
+      "step": 11329
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5453804382469871,
+      "learning_rate": 2.049961443872023e-06,
+      "loss": 1.8991,
+      "step": 11330
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5523543164279991,
+      "learning_rate": 2.047484862148541e-06,
+      "loss": 2.0741,
+      "step": 11331
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.576603720609261,
+      "learning_rate": 2.045009713444873e-06,
+      "loss": 1.8801,
+      "step": 11332
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5861415410715166,
+      "learning_rate": 2.042535997915551e-06,
+      "loss": 1.8513,
+      "step": 11333
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5689708095546199,
+      "learning_rate": 2.0400637157150156e-06,
+      "loss": 1.8432,
+      "step": 11334
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.567836307375821,
+      "learning_rate": 2.0375928669976256e-06,
+      "loss": 2.0546,
+      "step": 11335
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5188205952028175,
+      "learning_rate": 2.035123451917645e-06,
+      "loss": 1.9323,
+      "step": 11336
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5630712092179299,
+      "learning_rate": 2.0326554706292465e-06,
+      "loss": 1.862,
+      "step": 11337
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5653816110602845,
+      "learning_rate": 2.030188923286522e-06,
+      "loss": 1.8151,
+      "step": 11338
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5406949830404786,
+      "learning_rate": 2.0277238100434666e-06,
+      "loss": 2.0524,
+      "step": 11339
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5676194855851181,
+      "learning_rate": 2.0252601310539825e-06,
+      "loss": 1.8532,
+      "step": 11340
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.558816523811979,
+      "learning_rate": 2.0227978864718944e-06,
+      "loss": 1.8615,
+      "step": 11341
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5446730467179306,
+      "learning_rate": 2.0203370764509262e-06,
+      "loss": 1.9245,
+      "step": 11342
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5594999543495495,
+      "learning_rate": 2.0178777011447157e-06,
+      "loss": 2.0295,
+      "step": 11343
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.563304431702629,
+      "learning_rate": 2.0154197607068133e-06,
+      "loss": 1.8574,
+      "step": 11344
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5596852370277634,
+      "learning_rate": 2.0129632552906787e-06,
+      "loss": 1.8692,
+      "step": 11345
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5440368394663957,
+      "learning_rate": 2.0105081850496836e-06,
+      "loss": 1.9005,
+      "step": 11346
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.556728443812638,
+      "learning_rate": 2.0080545501371033e-06,
+      "loss": 2.077,
+      "step": 11347
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5515716062235766,
+      "learning_rate": 2.0056023507061343e-06,
+      "loss": 1.9558,
+      "step": 11348
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5791199292798315,
+      "learning_rate": 2.0031515869098655e-06,
+      "loss": 1.8771,
+      "step": 11349
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5434476139362671,
+      "learning_rate": 2.000702258901324e-06,
+      "loss": 1.8385,
+      "step": 11350
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.559490711483806,
+      "learning_rate": 1.9982543668334244e-06,
+      "loss": 2.0258,
+      "step": 11351
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.548634427973004,
+      "learning_rate": 1.9958079108589906e-06,
+      "loss": 1.851,
+      "step": 11352
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5698944808313946,
+      "learning_rate": 1.993362891130779e-06,
+      "loss": 1.8042,
+      "step": 11353
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5711369165580861,
+      "learning_rate": 1.990919307801434e-06,
+      "loss": 1.8625,
+      "step": 11354
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52956372077662,
+      "learning_rate": 1.9884771610235193e-06,
+      "loss": 2.0948,
+      "step": 11355
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5867014579665674,
+      "learning_rate": 1.986036450949516e-06,
+      "loss": 1.8711,
+      "step": 11356
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5320978489803737,
+      "learning_rate": 1.9835971777317913e-06,
+      "loss": 1.8458,
+      "step": 11357
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5656434533262092,
+      "learning_rate": 1.981159341522651e-06,
+      "loss": 1.8417,
+      "step": 11358
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5370699392084854,
+      "learning_rate": 1.9787229424743035e-06,
+      "loss": 2.006,
+      "step": 11359
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5621554495264767,
+      "learning_rate": 1.9762879807388444e-06,
+      "loss": 1.8233,
+      "step": 11360
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5114883561620484,
+      "learning_rate": 1.9738544564683214e-06,
+      "loss": 1.9193,
+      "step": 11361
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5634356465972353,
+      "learning_rate": 1.9714223698146543e-06,
+      "loss": 1.7716,
+      "step": 11362
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5413141255429869,
+      "learning_rate": 1.9689917209296887e-06,
+      "loss": 2.0252,
+      "step": 11363
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5646520316611936,
+      "learning_rate": 1.9665625099651917e-06,
+      "loss": 1.8821,
+      "step": 11364
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5588443832031327,
+      "learning_rate": 1.9641347370728174e-06,
+      "loss": 1.8671,
+      "step": 11365
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5776577099415924,
+      "learning_rate": 1.961708402404139e-06,
+      "loss": 1.8839,
+      "step": 11366
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4730394318420244,
+      "learning_rate": 1.9592835061106595e-06,
+      "loss": 2.0708,
+      "step": 11367
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5601752312403557,
+      "learning_rate": 1.9568600483437533e-06,
+      "loss": 1.8701,
+      "step": 11368
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5624520862470677,
+      "learning_rate": 1.9544380292547486e-06,
+      "loss": 1.8635,
+      "step": 11369
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5524109785465701,
+      "learning_rate": 1.9520174489948473e-06,
+      "loss": 1.8733,
+      "step": 11370
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5763411800158951,
+      "learning_rate": 1.9495983077151726e-06,
+      "loss": 2.063,
+      "step": 11371
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5834554644109189,
+      "learning_rate": 1.947180605566781e-06,
+      "loss": 1.8621,
+      "step": 11372
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.553018427423777,
+      "learning_rate": 1.9447643427006e-06,
+      "loss": 1.921,
+      "step": 11373
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5554544631154585,
+      "learning_rate": 1.9423495192674945e-06,
+      "loss": 1.8736,
+      "step": 11374
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5592957752143144,
+      "learning_rate": 1.9399361354182354e-06,
+      "loss": 2.0579,
+      "step": 11375
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5917400585614372,
+      "learning_rate": 1.937524191303497e-06,
+      "loss": 1.8608,
+      "step": 11376
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5682246420780163,
+      "learning_rate": 1.9351136870738614e-06,
+      "loss": 1.8746,
+      "step": 11377
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5812271034965071,
+      "learning_rate": 1.9327046228798358e-06,
+      "loss": 1.8705,
+      "step": 11378
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4914333441989122,
+      "learning_rate": 1.9302969988718195e-06,
+      "loss": 2.0685,
+      "step": 11379
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5499568516794129,
+      "learning_rate": 1.927890815200137e-06,
+      "loss": 1.8482,
+      "step": 11380
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.590792570107143,
+      "learning_rate": 1.925486072015012e-06,
+      "loss": 1.8396,
+      "step": 11381
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5607462689493338,
+      "learning_rate": 1.923082769466586e-06,
+      "loss": 1.8812,
+      "step": 11382
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5427023693670741,
+      "learning_rate": 1.9206809077049025e-06,
+      "loss": 2.0409,
+      "step": 11383
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5502759350673243,
+      "learning_rate": 1.9182804868799254e-06,
+      "loss": 1.822,
+      "step": 11384
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6185817430145331,
+      "learning_rate": 1.9158815071415147e-06,
+      "loss": 1.9241,
+      "step": 11385
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5644341920983977,
+      "learning_rate": 1.9134839686394568e-06,
+      "loss": 1.9634,
+      "step": 11386
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5484151840681039,
+      "learning_rate": 1.911087871523437e-06,
+      "loss": 2.0262,
+      "step": 11387
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6193101845250805,
+      "learning_rate": 1.9086932159430527e-06,
+      "loss": 1.864,
+      "step": 11388
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5808121623256574,
+      "learning_rate": 1.9063000020478117e-06,
+      "loss": 1.8507,
+      "step": 11389
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5582539100275166,
+      "learning_rate": 1.903908229987131e-06,
+      "loss": 1.8784,
+      "step": 11390
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5365706460361583,
+      "learning_rate": 1.9015178999103433e-06,
+      "loss": 2.0549,
+      "step": 11391
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5205814788034887,
+      "learning_rate": 1.8991290119666828e-06,
+      "loss": 1.9109,
+      "step": 11392
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5679481654679627,
+      "learning_rate": 1.896741566305299e-06,
+      "loss": 1.8288,
+      "step": 11393
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5418815057964814,
+      "learning_rate": 1.8943555630752507e-06,
+      "loss": 1.8601,
+      "step": 11394
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5862655515690436,
+      "learning_rate": 1.8919710024255105e-06,
+      "loss": 2.0581,
+      "step": 11395
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5640665429497063,
+      "learning_rate": 1.8895878845049398e-06,
+      "loss": 1.8747,
+      "step": 11396
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5455842901947854,
+      "learning_rate": 1.8872062094623416e-06,
+      "loss": 1.8234,
+      "step": 11397
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5312684942535182,
+      "learning_rate": 1.884825977446414e-06,
+      "loss": 1.944,
+      "step": 11398
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5432389523187544,
+      "learning_rate": 1.8824471886057516e-06,
+      "loss": 2.0851,
+      "step": 11399
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5791278155183416,
+      "learning_rate": 1.8800698430888857e-06,
+      "loss": 1.8549,
+      "step": 11400
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5599541645812369,
+      "learning_rate": 1.8776939410442397e-06,
+      "loss": 1.8802,
+      "step": 11401
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5446443623878265,
+      "learning_rate": 1.875319482620147e-06,
+      "loss": 1.8289,
+      "step": 11402
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5573991198873848,
+      "learning_rate": 1.8729464679648618e-06,
+      "loss": 2.0435,
+      "step": 11403
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5030148175000844,
+      "learning_rate": 1.870574897226529e-06,
+      "loss": 1.8941,
+      "step": 11404
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5576267883753275,
+      "learning_rate": 1.8682047705532308e-06,
+      "loss": 1.8351,
+      "step": 11405
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5480264130596546,
+      "learning_rate": 1.8658360880929371e-06,
+      "loss": 1.7954,
+      "step": 11406
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.577334551147996,
+      "learning_rate": 1.863468849993527e-06,
+      "loss": 2.0625,
+      "step": 11407
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5517706061202078,
+      "learning_rate": 1.8611030564028075e-06,
+      "loss": 1.8786,
+      "step": 11408
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5760873933821362,
+      "learning_rate": 1.8587387074684852e-06,
+      "loss": 1.9031,
+      "step": 11409
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5928241719947808,
+      "learning_rate": 1.8563758033381645e-06,
+      "loss": 1.8635,
+      "step": 11410
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5049704377248493,
+      "learning_rate": 1.8540143441593854e-06,
+      "loss": 2.1061,
+      "step": 11411
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5885543974706761,
+      "learning_rate": 1.8516543300795746e-06,
+      "loss": 1.8573,
+      "step": 11412
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5434842731844524,
+      "learning_rate": 1.8492957612460754e-06,
+      "loss": 1.9061,
+      "step": 11413
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6006982994514282,
+      "learning_rate": 1.846938637806156e-06,
+      "loss": 1.7901,
+      "step": 11414
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5305207585592315,
+      "learning_rate": 1.8445829599069682e-06,
+      "loss": 2.0208,
+      "step": 11415
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5634056904571927,
+      "learning_rate": 1.8422287276955886e-06,
+      "loss": 1.834,
+      "step": 11416
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5138256777073237,
+      "learning_rate": 1.8398759413190053e-06,
+      "loss": 1.9281,
+      "step": 11417
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5794606665028862,
+      "learning_rate": 1.8375246009241092e-06,
+      "loss": 1.875,
+      "step": 11418
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5374531556585735,
+      "learning_rate": 1.8351747066577134e-06,
+      "loss": 2.0518,
+      "step": 11419
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5708448603297026,
+      "learning_rate": 1.8328262586665173e-06,
+      "loss": 1.8389,
+      "step": 11420
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5434416526072174,
+      "learning_rate": 1.8304792570971508e-06,
+      "loss": 1.8417,
+      "step": 11421
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5487952362585242,
+      "learning_rate": 1.828133702096152e-06,
+      "loss": 1.9191,
+      "step": 11422
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5023233947685983,
+      "learning_rate": 1.8257895938099568e-06,
+      "loss": 1.9448,
+      "step": 11423
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5217867061043434,
+      "learning_rate": 1.8234469323849201e-06,
+      "loss": 2.0017,
+      "step": 11424
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5360032944416055,
+      "learning_rate": 1.8211057179673003e-06,
+      "loss": 1.8718,
+      "step": 11425
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5630404258046613,
+      "learning_rate": 1.8187659507032718e-06,
+      "loss": 1.835,
+      "step": 11426
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5449949965008825,
+      "learning_rate": 1.8164276307389178e-06,
+      "loss": 2.0378,
+      "step": 11427
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5432824272167419,
+      "learning_rate": 1.8140907582202271e-06,
+      "loss": 1.8485,
+      "step": 11428
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5268602587517842,
+      "learning_rate": 1.8117553332930997e-06,
+      "loss": 1.8888,
+      "step": 11429
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5839837550403904,
+      "learning_rate": 1.8094213561033469e-06,
+      "loss": 1.8385,
+      "step": 11430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5310612007541586,
+      "learning_rate": 1.8070888267966878e-06,
+      "loss": 2.0793,
+      "step": 11431
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5337712370840574,
+      "learning_rate": 1.8047577455187537e-06,
+      "loss": 1.8606,
+      "step": 11432
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5718593742179925,
+      "learning_rate": 1.8024281124150805e-06,
+      "loss": 1.8169,
+      "step": 11433
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5447043727200745,
+      "learning_rate": 1.8000999276311187e-06,
+      "loss": 1.8874,
+      "step": 11434
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.517376371045215,
+      "learning_rate": 1.7977731913122242e-06,
+      "loss": 1.9434,
+      "step": 11435
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.537247692656742,
+      "learning_rate": 1.7954479036036698e-06,
+      "loss": 2.0057,
+      "step": 11436
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.558218381539825,
+      "learning_rate": 1.7931240646506254e-06,
+      "loss": 1.8266,
+      "step": 11437
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5632202271515055,
+      "learning_rate": 1.790801674598186e-06,
+      "loss": 1.7982,
+      "step": 11438
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5868016870077111,
+      "learning_rate": 1.7884807335913412e-06,
+      "loss": 1.9939,
+      "step": 11439
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5621041248749206,
+      "learning_rate": 1.786161241775e-06,
+      "loss": 1.8937,
+      "step": 11440
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6043670282233546,
+      "learning_rate": 1.7838431992939775e-06,
+      "loss": 1.8092,
+      "step": 11441
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5236691426667504,
+      "learning_rate": 1.7815266062929992e-06,
+      "loss": 1.921,
+      "step": 11442
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5348056290663287,
+      "learning_rate": 1.7792114629166967e-06,
+      "loss": 2.063,
+      "step": 11443
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5660441627646823,
+      "learning_rate": 1.7768977693096184e-06,
+      "loss": 1.8692,
+      "step": 11444
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5796437953649689,
+      "learning_rate": 1.774585525616218e-06,
+      "loss": 1.8092,
+      "step": 11445
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5477984053334748,
+      "learning_rate": 1.7722747319808469e-06,
+      "loss": 1.8704,
+      "step": 11446
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5251831251568906,
+      "learning_rate": 1.7699653885477923e-06,
+      "loss": 2.0335,
+      "step": 11447
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5026893527417187,
+      "learning_rate": 1.7676574954612335e-06,
+      "loss": 1.8634,
+      "step": 11448
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5377440378348431,
+      "learning_rate": 1.7653510528652494e-06,
+      "loss": 1.8886,
+      "step": 11449
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.547145007729135,
+      "learning_rate": 1.7630460609038556e-06,
+      "loss": 1.8442,
+      "step": 11450
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5469870296677756,
+      "learning_rate": 1.760742519720951e-06,
+      "loss": 2.0614,
+      "step": 11451
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.594287691495631,
+      "learning_rate": 1.7584404294603618e-06,
+      "loss": 1.8553,
+      "step": 11452
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5539632393921048,
+      "learning_rate": 1.7561397902658205e-06,
+      "loss": 1.8254,
+      "step": 11453
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5817113916503119,
+      "learning_rate": 1.753840602280954e-06,
+      "loss": 1.9122,
+      "step": 11454
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5498871052031432,
+      "learning_rate": 1.7515428656493167e-06,
+      "loss": 1.826,
+      "step": 11455
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5349252140993577,
+      "learning_rate": 1.7492465805143715e-06,
+      "loss": 1.9992,
+      "step": 11456
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5698287777477677,
+      "learning_rate": 1.7469517470194707e-06,
+      "loss": 1.8563,
+      "step": 11457
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6149504626125878,
+      "learning_rate": 1.7446583653079052e-06,
+      "loss": 1.8079,
+      "step": 11458
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5711284169011605,
+      "learning_rate": 1.742366435522852e-06,
+      "loss": 2.04,
+      "step": 11459
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5119927951453412,
+      "learning_rate": 1.7400759578073994e-06,
+      "loss": 1.9854,
+      "step": 11460
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5780208662192877,
+      "learning_rate": 1.7377869323045693e-06,
+      "loss": 1.8439,
+      "step": 11461
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5811806932942469,
+      "learning_rate": 1.7354993591572583e-06,
+      "loss": 1.8628,
+      "step": 11462
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.52653393976474,
+      "learning_rate": 1.7332132385082966e-06,
+      "loss": 1.9811,
+      "step": 11463
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5545719825569585,
+      "learning_rate": 1.7309285705004174e-06,
+      "loss": 1.7913,
+      "step": 11464
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5519025613258323,
+      "learning_rate": 1.728645355276251e-06,
+      "loss": 1.8902,
+      "step": 11465
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5177559789773891,
+      "learning_rate": 1.7263635929783666e-06,
+      "loss": 1.9415,
+      "step": 11466
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5385086436625898,
+      "learning_rate": 1.7240832837492089e-06,
+      "loss": 1.8887,
+      "step": 11467
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5442795652765817,
+      "learning_rate": 1.7218044277311468e-06,
+      "loss": 2.0151,
+      "step": 11468
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5479580565307639,
+      "learning_rate": 1.7195270250664725e-06,
+      "loss": 1.8735,
+      "step": 11469
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5864212348747019,
+      "learning_rate": 1.7172510758973608e-06,
+      "loss": 1.8612,
+      "step": 11470
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.575159541732696,
+      "learning_rate": 1.7149765803659124e-06,
+      "loss": 2.0793,
+      "step": 11471
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.533729930601059,
+      "learning_rate": 1.7127035386141326e-06,
+      "loss": 1.8003,
+      "step": 11472
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5089293084138778,
+      "learning_rate": 1.710431950783939e-06,
+      "loss": 1.9449,
+      "step": 11473
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5476500538044772,
+      "learning_rate": 1.7081618170171538e-06,
+      "loss": 1.8473,
+      "step": 11474
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5704101526407619,
+      "learning_rate": 1.7058931374555108e-06,
+      "loss": 1.801,
+      "step": 11475
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5662651394140729,
+      "learning_rate": 1.7036259122406557e-06,
+      "loss": 2.0124,
+      "step": 11476
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.549025468735945,
+      "learning_rate": 1.7013601415141383e-06,
+      "loss": 1.8221,
+      "step": 11477
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5502669478535362,
+      "learning_rate": 1.6990958254174182e-06,
+      "loss": 1.8643,
+      "step": 11478
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5412541320252907,
+      "learning_rate": 1.6968329640918712e-06,
+      "loss": 1.9434,
+      "step": 11479
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5721593338532074,
+      "learning_rate": 1.694571557678773e-06,
+      "loss": 2.0734,
+      "step": 11480
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5630047601013076,
+      "learning_rate": 1.6923116063193107e-06,
+      "loss": 1.7974,
+      "step": 11481
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5553723584718063,
+      "learning_rate": 1.690053110154588e-06,
+      "loss": 1.9089,
+      "step": 11482
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5585254013197967,
+      "learning_rate": 1.687796069325609e-06,
+      "loss": 2.1024,
+      "step": 11483
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5636307723415748,
+      "learning_rate": 1.6855404839732913e-06,
+      "loss": 1.843,
+      "step": 11484
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5511016877619674,
+      "learning_rate": 1.6832863542384585e-06,
+      "loss": 1.9488,
+      "step": 11485
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5599689550315416,
+      "learning_rate": 1.6810336802618453e-06,
+      "loss": 1.8306,
+      "step": 11486
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5537086824748512,
+      "learning_rate": 1.6787824621840976e-06,
+      "loss": 1.8575,
+      "step": 11487
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5536654417187133,
+      "learning_rate": 1.6765327001457665e-06,
+      "loss": 2.0624,
+      "step": 11488
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5438282667619144,
+      "learning_rate": 1.674284394287312e-06,
+      "loss": 1.8376,
+      "step": 11489
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5323617698369226,
+      "learning_rate": 1.672037544749111e-06,
+      "loss": 1.8746,
+      "step": 11490
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5631009029239705,
+      "learning_rate": 1.6697921516714372e-06,
+      "loss": 1.945,
+      "step": 11491
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5884476755481036,
+      "learning_rate": 1.6675482151944866e-06,
+      "loss": 2.0714,
+      "step": 11492
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5872801935769651,
+      "learning_rate": 1.6653057354583445e-06,
+      "loss": 1.8592,
+      "step": 11493
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5703552451578564,
+      "learning_rate": 1.6630647126030325e-06,
+      "loss": 1.8877,
+      "step": 11494
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5367921988048878,
+      "learning_rate": 1.6608251467684632e-06,
+      "loss": 2.0715,
+      "step": 11495
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5788993200993916,
+      "learning_rate": 1.658587038094453e-06,
+      "loss": 1.8256,
+      "step": 11496
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5302187426392724,
+      "learning_rate": 1.6563503867207481e-06,
+      "loss": 1.9384,
+      "step": 11497
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5715673241539011,
+      "learning_rate": 1.6541151927869868e-06,
+      "loss": 1.8412,
+      "step": 11498
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5790838234986355,
+      "learning_rate": 1.6518814564327134e-06,
+      "loss": 1.8505,
+      "step": 11499
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5350456843446728,
+      "learning_rate": 1.6496491777974077e-06,
+      "loss": 2.0565,
+      "step": 11500
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5759006085601842,
+      "learning_rate": 1.647418357020422e-06,
+      "loss": 1.8292,
+      "step": 11501
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5306123612452441,
+      "learning_rate": 1.6451889942410453e-06,
+      "loss": 1.8654,
+      "step": 11502
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5426362792230239,
+      "learning_rate": 1.6429610895984688e-06,
+      "loss": 2.0804,
+      "step": 11503
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5216841658217515,
+      "learning_rate": 1.6407346432317755e-06,
+      "loss": 1.8741,
+      "step": 11504
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5667584330820248,
+      "learning_rate": 1.6385096552799906e-06,
+      "loss": 1.8444,
+      "step": 11505
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6019417627288464,
+      "learning_rate": 1.636286125882014e-06,
+      "loss": 1.8824,
+      "step": 11506
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5803286184776754,
+      "learning_rate": 1.6340640551766706e-06,
+      "loss": 1.8879,
+      "step": 11507
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5595180063103656,
+      "learning_rate": 1.6318434433027051e-06,
+      "loss": 2.0683,
+      "step": 11508
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.564393284386757,
+      "learning_rate": 1.6296242903987514e-06,
+      "loss": 1.8471,
+      "step": 11509
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.526819182177988,
+      "learning_rate": 1.6274065966033592e-06,
+      "loss": 1.9064,
+      "step": 11510
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5797477153724305,
+      "learning_rate": 1.6251903620549901e-06,
+      "loss": 1.8588,
+      "step": 11511
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5282103382720872,
+      "learning_rate": 1.6229755868920116e-06,
+      "loss": 1.9987,
+      "step": 11512
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5686101759770306,
+      "learning_rate": 1.6207622712527043e-06,
+      "loss": 1.8895,
+      "step": 11513
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5798136941689981,
+      "learning_rate": 1.6185504152752523e-06,
+      "loss": 1.8486,
+      "step": 11514
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.578283250504507,
+      "learning_rate": 1.616340019097748e-06,
+      "loss": 2.0023,
+      "step": 11515
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5522000020238053,
+      "learning_rate": 1.6141310828582058e-06,
+      "loss": 1.9585,
+      "step": 11516
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5671364730243942,
+      "learning_rate": 1.6119236066945264e-06,
+      "loss": 1.8699,
+      "step": 11517
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5521964335085521,
+      "learning_rate": 1.6097175907445389e-06,
+      "loss": 1.8465,
+      "step": 11518
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5544334709279299,
+      "learning_rate": 1.6075130351459715e-06,
+      "loss": 1.8755,
+      "step": 11519
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5406042487548138,
+      "learning_rate": 1.6053099400364645e-06,
+      "loss": 2.0455,
+      "step": 11520
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.527987913374905,
+      "learning_rate": 1.6031083055535662e-06,
+      "loss": 1.8534,
+      "step": 11521
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5154940508786238,
+      "learning_rate": 1.6009081318347302e-06,
+      "loss": 1.9454,
+      "step": 11522
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5219694964022139,
+      "learning_rate": 1.5987094190173273e-06,
+      "loss": 1.8532,
+      "step": 11523
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5672431341750163,
+      "learning_rate": 1.5965121672386313e-06,
+      "loss": 2.1037,
+      "step": 11524
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5613193408313942,
+      "learning_rate": 1.5943163766358238e-06,
+      "loss": 1.81,
+      "step": 11525
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5686138602445893,
+      "learning_rate": 1.592122047345998e-06,
+      "loss": 1.848,
+      "step": 11526
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5236972679363534,
+      "learning_rate": 1.5899291795061554e-06,
+      "loss": 2.001,
+      "step": 11527
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5166000909253287,
+      "learning_rate": 1.5877377732532033e-06,
+      "loss": 1.9114,
+      "step": 11528
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5556321395789722,
+      "learning_rate": 1.5855478287239628e-06,
+      "loss": 1.866,
+      "step": 11529
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5464626028019876,
+      "learning_rate": 1.5833593460551606e-06,
+      "loss": 1.8368,
+      "step": 11530
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5407746105954445,
+      "learning_rate": 1.581172325383437e-06,
+      "loss": 1.822,
+      "step": 11531
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5333676049003504,
+      "learning_rate": 1.5789867668453223e-06,
+      "loss": 2.0637,
+      "step": 11532
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5396263740254015,
+      "learning_rate": 1.5768026705772848e-06,
+      "loss": 1.834,
+      "step": 11533
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5438394018269245,
+      "learning_rate": 1.5746200367156823e-06,
+      "loss": 1.8543,
+      "step": 11534
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5321757616084442,
+      "learning_rate": 1.572438865396783e-06,
+      "loss": 1.9551,
+      "step": 11535
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.541679020778305,
+      "learning_rate": 1.5702591567567705e-06,
+      "loss": 2.0459,
+      "step": 11536
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5683657345046792,
+      "learning_rate": 1.5680809109317295e-06,
+      "loss": 1.8654,
+      "step": 11537
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5805774880487938,
+      "learning_rate": 1.5659041280576576e-06,
+      "loss": 1.8293,
+      "step": 11538
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5705951698548497,
+      "learning_rate": 1.5637288082704649e-06,
+      "loss": 1.8518,
+      "step": 11539
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5400605409766198,
+      "learning_rate": 1.5615549517059513e-06,
+      "loss": 2.0092,
+      "step": 11540
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5350277584267021,
+      "learning_rate": 1.5593825584998556e-06,
+      "loss": 1.8851,
+      "step": 11541
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5958600529362325,
+      "learning_rate": 1.5572116287878081e-06,
+      "loss": 1.8933,
+      "step": 11542
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5643169319921733,
+      "learning_rate": 1.5550421627053336e-06,
+      "loss": 1.8949,
+      "step": 11543
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5688988531788705,
+      "learning_rate": 1.5528741603879016e-06,
+      "loss": 2.0592,
+      "step": 11544
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5569108556299291,
+      "learning_rate": 1.5507076219708538e-06,
+      "loss": 1.876,
+      "step": 11545
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5608668024932708,
+      "learning_rate": 1.5485425475894572e-06,
+      "loss": 1.829,
+      "step": 11546
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5186299500402831,
+      "learning_rate": 1.5463789373788978e-06,
+      "loss": 1.9358,
+      "step": 11547
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6071739448379364,
+      "learning_rate": 1.5442167914742455e-06,
+      "loss": 2.0678,
+      "step": 11548
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5662753644720798,
+      "learning_rate": 1.5420561100105007e-06,
+      "loss": 1.8703,
+      "step": 11549
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5635858042631295,
+      "learning_rate": 1.5398968931225639e-06,
+      "loss": 1.8335,
+      "step": 11550
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5759985110604045,
+      "learning_rate": 1.5377391409452324e-06,
+      "loss": 1.841,
+      "step": 11551
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5429365477523836,
+      "learning_rate": 1.5355828536132405e-06,
+      "loss": 2.0888,
+      "step": 11552
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5408100568461693,
+      "learning_rate": 1.5334280312612e-06,
+      "loss": 1.9563,
+      "step": 11553
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5577544197795093,
+      "learning_rate": 1.5312746740236471e-06,
+      "loss": 1.8509,
+      "step": 11554
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5462128136671481,
+      "learning_rate": 1.529122782035039e-06,
+      "loss": 1.8493,
+      "step": 11555
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5372412256542201,
+      "learning_rate": 1.5269723554297095e-06,
+      "loss": 2.0328,
+      "step": 11556
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5428278282472243,
+      "learning_rate": 1.5248233943419205e-06,
+      "loss": 1.8162,
+      "step": 11557
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5694209568403608,
+      "learning_rate": 1.5226758989058565e-06,
+      "loss": 1.8687,
+      "step": 11558
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5586634864506735,
+      "learning_rate": 1.5205298692555769e-06,
+      "loss": 1.8877,
+      "step": 11559
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4931589729173802,
+      "learning_rate": 1.5183853055250746e-06,
+      "loss": 2.1076,
+      "step": 11560
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5618606027167793,
+      "learning_rate": 1.5162422078482424e-06,
+      "loss": 1.846,
+      "step": 11561
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5478665746683364,
+      "learning_rate": 1.5141005763588845e-06,
+      "loss": 1.8341,
+      "step": 11562
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5391092299472615,
+      "learning_rate": 1.5119604111907077e-06,
+      "loss": 1.8563,
+      "step": 11563
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5154712713242152,
+      "learning_rate": 1.5098217124773356e-06,
+      "loss": 2.0463,
+      "step": 11564
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5399775172940153,
+      "learning_rate": 1.5076844803522922e-06,
+      "loss": 1.8257,
+      "step": 11565
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.505914977858674,
+      "learning_rate": 1.505548714949015e-06,
+      "loss": 1.8391,
+      "step": 11566
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5647101880926146,
+      "learning_rate": 1.503414416400853e-06,
+      "loss": 1.8434,
+      "step": 11567
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5565693985751698,
+      "learning_rate": 1.5012815848410523e-06,
+      "loss": 2.0487,
+      "step": 11568
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5825650948006594,
+      "learning_rate": 1.4991502204027786e-06,
+      "loss": 1.8467,
+      "step": 11569
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5404904882481903,
+      "learning_rate": 1.4970203232190977e-06,
+      "loss": 1.792,
+      "step": 11570
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5325039152037033,
+      "learning_rate": 1.494891893422995e-06,
+      "loss": 1.8435,
+      "step": 11571
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5016319729314044,
+      "learning_rate": 1.4927649311473501e-06,
+      "loss": 2.0959,
+      "step": 11572
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5341816761030543,
+      "learning_rate": 1.4906394365249627e-06,
+      "loss": 1.8666,
+      "step": 11573
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5328364186729707,
+      "learning_rate": 1.4885154096885318e-06,
+      "loss": 1.8636,
+      "step": 11574
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5541914577813613,
+      "learning_rate": 1.486392850770671e-06,
+      "loss": 1.8676,
+      "step": 11575
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5382688032761922,
+      "learning_rate": 1.4842717599039047e-06,
+      "loss": 2.0766,
+      "step": 11576
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5569870313632911,
+      "learning_rate": 1.4821521372206547e-06,
+      "loss": 1.8406,
+      "step": 11577
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5217034421803733,
+      "learning_rate": 1.4800339828532627e-06,
+      "loss": 1.9297,
+      "step": 11578
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5428979127909535,
+      "learning_rate": 1.4779172969339672e-06,
+      "loss": 1.8456,
+      "step": 11579
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5481496322229213,
+      "learning_rate": 1.475802079594929e-06,
+      "loss": 2.0423,
+      "step": 11580
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5434713311108774,
+      "learning_rate": 1.4736883309682065e-06,
+      "loss": 1.8353,
+      "step": 11581
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5706542306496056,
+      "learning_rate": 1.471576051185769e-06,
+      "loss": 1.8321,
+      "step": 11582
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5665692413734926,
+      "learning_rate": 1.4694652403794972e-06,
+      "loss": 1.8212,
+      "step": 11583
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.500725732836912,
+      "learning_rate": 1.4673558986811776e-06,
+      "loss": 2.126,
+      "step": 11584
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5564945456815288,
+      "learning_rate": 1.4652480262225015e-06,
+      "loss": 1.8164,
+      "step": 11585
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5628586916486585,
+      "learning_rate": 1.4631416231350782e-06,
+      "loss": 1.8236,
+      "step": 11586
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5373099881686948,
+      "learning_rate": 1.4610366895504074e-06,
+      "loss": 1.8577,
+      "step": 11587
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5370315493794655,
+      "learning_rate": 1.4589332255999205e-06,
+      "loss": 2.0491,
+      "step": 11588
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5786041662723317,
+      "learning_rate": 1.4568312314149456e-06,
+      "loss": 1.8325,
+      "step": 11589
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5399534401068655,
+      "learning_rate": 1.4547307071267086e-06,
+      "loss": 1.8105,
+      "step": 11590
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5409039572752965,
+      "learning_rate": 1.4526316528663653e-06,
+      "loss": 1.9102,
+      "step": 11591
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5588041357744535,
+      "learning_rate": 1.450534068764961e-06,
+      "loss": 2.0673,
+      "step": 11592
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5550623775895449,
+      "learning_rate": 1.448437954953452e-06,
+      "loss": 1.8847,
+      "step": 11593
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5754586748824821,
+      "learning_rate": 1.4463433115627229e-06,
+      "loss": 1.8457,
+      "step": 11594
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5609369312532884,
+      "learning_rate": 1.444250138723538e-06,
+      "loss": 1.887,
+      "step": 11595
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5536132673369919,
+      "learning_rate": 1.4421584365665847e-06,
+      "loss": 2.0327,
+      "step": 11596
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5121870789793983,
+      "learning_rate": 1.4400682052224613e-06,
+      "loss": 1.9637,
+      "step": 11597
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5425302361243759,
+      "learning_rate": 1.4379794448216633e-06,
+      "loss": 1.8411,
+      "step": 11598
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5469752757658726,
+      "learning_rate": 1.4358921554946087e-06,
+      "loss": 1.8819,
+      "step": 11599
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5362345275848386,
+      "learning_rate": 1.433806337371607e-06,
+      "loss": 2.0141,
+      "step": 11600
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.546044911409612,
+      "learning_rate": 1.431721990582885e-06,
+      "loss": 1.8677,
+      "step": 11601
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5554629976181261,
+      "learning_rate": 1.429639115258588e-06,
+      "loss": 1.8623,
+      "step": 11602
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5253589802437095,
+      "learning_rate": 1.4275577115287458e-06,
+      "loss": 1.8856,
+      "step": 11603
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5395582732106576,
+      "learning_rate": 1.4254777795233098e-06,
+      "loss": 2.0669,
+      "step": 11604
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5429214973083368,
+      "learning_rate": 1.4233993193721513e-06,
+      "loss": 1.8088,
+      "step": 11605
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5437350852995847,
+      "learning_rate": 1.4213223312050216e-06,
+      "loss": 1.8232,
+      "step": 11606
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5605509528789246,
+      "learning_rate": 1.4192468151516064e-06,
+      "loss": 1.7867,
+      "step": 11607
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.555345067347659,
+      "learning_rate": 1.4171727713414824e-06,
+      "loss": 2.1308,
+      "step": 11608
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5017882847950005,
+      "learning_rate": 1.415100199904143e-06,
+      "loss": 1.9585,
+      "step": 11609
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5699754789317231,
+      "learning_rate": 1.4130291009689879e-06,
+      "loss": 1.8357,
+      "step": 11610
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.550954707692324,
+      "learning_rate": 1.4109594746653243e-06,
+      "loss": 1.8447,
+      "step": 11611
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5496112979347849,
+      "learning_rate": 1.408891321122366e-06,
+      "loss": 2.0417,
+      "step": 11612
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5744276835341443,
+      "learning_rate": 1.406824640469237e-06,
+      "loss": 1.8653,
+      "step": 11613
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5914070211002466,
+      "learning_rate": 1.404759432834968e-06,
+      "loss": 1.8646,
+      "step": 11614
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5227930174030866,
+      "learning_rate": 1.4026956983485024e-06,
+      "loss": 1.9122,
+      "step": 11615
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5945733806069876,
+      "learning_rate": 1.400633437138682e-06,
+      "loss": 2.0478,
+      "step": 11616
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5677315986787986,
+      "learning_rate": 1.3985726493342649e-06,
+      "loss": 1.8489,
+      "step": 11617
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5269670738629265,
+      "learning_rate": 1.3965133350639149e-06,
+      "loss": 1.8789,
+      "step": 11618
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5400614841820163,
+      "learning_rate": 1.3944554944562038e-06,
+      "loss": 1.8268,
+      "step": 11619
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5632628281370748,
+      "learning_rate": 1.3923991276396097e-06,
+      "loss": 2.0724,
+      "step": 11620
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.562864145863654,
+      "learning_rate": 1.3903442347425188e-06,
+      "loss": 1.8104,
+      "step": 11621
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49793932970079546,
+      "learning_rate": 1.3882908158932311e-06,
+      "loss": 1.9099,
+      "step": 11622
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5600569635142791,
+      "learning_rate": 1.3862388712199442e-06,
+      "loss": 1.865,
+      "step": 11623
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5484038013626298,
+      "learning_rate": 1.384188400850775e-06,
+      "loss": 2.0173,
+      "step": 11624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5638574969012546,
+      "learning_rate": 1.3821394049137405e-06,
+      "loss": 1.8346,
+      "step": 11625
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5373387734770145,
+      "learning_rate": 1.3800918835367633e-06,
+      "loss": 1.8105,
+      "step": 11626
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5398784543311672,
+      "learning_rate": 1.3780458368476856e-06,
+      "loss": 1.889,
+      "step": 11627
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5550276347529988,
+      "learning_rate": 1.3760012649742499e-06,
+      "loss": 2.1092,
+      "step": 11628
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5396034582180304,
+      "learning_rate": 1.373958168044101e-06,
+      "loss": 1.8298,
+      "step": 11629
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5451938993298576,
+      "learning_rate": 1.3719165461848038e-06,
+      "loss": 1.8466,
+      "step": 11630
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.590680006946242,
+      "learning_rate": 1.3698763995238228e-06,
+      "loss": 1.811,
+      "step": 11631
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5212617300719344,
+      "learning_rate": 1.3678377281885312e-06,
+      "loss": 2.0822,
+      "step": 11632
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5284491757533188,
+      "learning_rate": 1.3658005323062218e-06,
+      "loss": 1.8538,
+      "step": 11633
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49128686588545867,
+      "learning_rate": 1.3637648120040675e-06,
+      "loss": 1.9444,
+      "step": 11634
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5294896628727015,
+      "learning_rate": 1.361730567409178e-06,
+      "loss": 1.83,
+      "step": 11635
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5443466021337594,
+      "learning_rate": 1.3596977986485627e-06,
+      "loss": 2.0661,
+      "step": 11636
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5445264210530161,
+      "learning_rate": 1.3576665058491229e-06,
+      "loss": 1.8738,
+      "step": 11637
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5671393281054745,
+      "learning_rate": 1.3556366891376932e-06,
+      "loss": 1.8182,
+      "step": 11638
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5590281380217906,
+      "learning_rate": 1.353608348641e-06,
+      "loss": 1.8231,
+      "step": 11639
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5314347333718293,
+      "learning_rate": 1.3515814844856756e-06,
+      "loss": 1.9337,
+      "step": 11640
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5351449113856471,
+      "learning_rate": 1.349556096798274e-06,
+      "loss": 2.0345,
+      "step": 11641
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5696918765728142,
+      "learning_rate": 1.3475321857052386e-06,
+      "loss": 1.7705,
+      "step": 11642
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5399951073542054,
+      "learning_rate": 1.3455097513329352e-06,
+      "loss": 1.9177,
+      "step": 11643
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5609708560243463,
+      "learning_rate": 1.3434887938076373e-06,
+      "loss": 2.0378,
+      "step": 11644
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5716074265111633,
+      "learning_rate": 1.341469313255514e-06,
+      "loss": 1.8565,
+      "step": 11645
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4970874654286937,
+      "learning_rate": 1.3394513098026585e-06,
+      "loss": 1.95,
+      "step": 11646
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5404093075670032,
+      "learning_rate": 1.3374347835750534e-06,
+      "loss": 1.8409,
+      "step": 11647
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5148236035340942,
+      "learning_rate": 1.335419734698598e-06,
+      "loss": 2.0339,
+      "step": 11648
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.545245452747753,
+      "learning_rate": 1.333406163299114e-06,
+      "loss": 1.8663,
+      "step": 11649
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5406087523417343,
+      "learning_rate": 1.3313940695023065e-06,
+      "loss": 1.8486,
+      "step": 11650
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5475893700078736,
+      "learning_rate": 1.329383453433794e-06,
+      "loss": 1.8387,
+      "step": 11651
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5335378735327975,
+      "learning_rate": 1.327374315219121e-06,
+      "loss": 2.0847,
+      "step": 11652
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5138035852166473,
+      "learning_rate": 1.3253666549837174e-06,
+      "loss": 1.915,
+      "step": 11653
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5635210039457367,
+      "learning_rate": 1.3233604728529303e-06,
+      "loss": 1.8059,
+      "step": 11654
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.604284014446974,
+      "learning_rate": 1.321355768952015e-06,
+      "loss": 1.8535,
+      "step": 11655
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5427330845038708,
+      "learning_rate": 1.3193525434061327e-06,
+      "loss": 2.0185,
+      "step": 11656
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5603510476231374,
+      "learning_rate": 1.3173507963403526e-06,
+      "loss": 1.7882,
+      "step": 11657
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5449415385601569,
+      "learning_rate": 1.315350527879655e-06,
+      "loss": 1.8209,
+      "step": 11658
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49752840259731257,
+      "learning_rate": 1.3133517381489241e-06,
+      "loss": 1.8991,
+      "step": 11659
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5679971385295358,
+      "learning_rate": 1.3113544272729483e-06,
+      "loss": 2.0728,
+      "step": 11660
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5516762220737355,
+      "learning_rate": 1.3093585953764337e-06,
+      "loss": 1.8588,
+      "step": 11661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5615364809411306,
+      "learning_rate": 1.3073642425839833e-06,
+      "loss": 1.8555,
+      "step": 11662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5645680627661618,
+      "learning_rate": 1.3053713690201169e-06,
+      "loss": 1.8519,
+      "step": 11663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5618147188517575,
+      "learning_rate": 1.3033799748092545e-06,
+      "loss": 2.0133,
+      "step": 11664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5089274696036804,
+      "learning_rate": 1.3013900600757301e-06,
+      "loss": 1.9022,
+      "step": 11665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5982750960852059,
+      "learning_rate": 1.2994016249437802e-06,
+      "loss": 1.8409,
+      "step": 11666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5474835445239772,
+      "learning_rate": 1.29741466953755e-06,
+      "loss": 1.8468,
+      "step": 11667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5720836908376999,
+      "learning_rate": 1.2954291939810959e-06,
+      "loss": 2.0688,
+      "step": 11668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5502129279239272,
+      "learning_rate": 1.293445198398377e-06,
+      "loss": 1.8325,
+      "step": 11669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5445088075760147,
+      "learning_rate": 1.2914626829132637e-06,
+      "loss": 1.8304,
+      "step": 11670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.49086490734368843,
+      "learning_rate": 1.2894816476495319e-06,
+      "loss": 1.9582,
+      "step": 11671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5596714564502115,
+      "learning_rate": 1.2875020927308661e-06,
+      "loss": 1.8509,
+      "step": 11672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5506238006737157,
+      "learning_rate": 1.2855240182808593e-06,
+      "loss": 2.0923,
+      "step": 11673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5852959940018374,
+      "learning_rate": 1.2835474244230095e-06,
+      "loss": 1.7989,
+      "step": 11674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5407076481321933,
+      "learning_rate": 1.2815723112807266e-06,
+      "loss": 1.8466,
+      "step": 11675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5177528869089557,
+      "learning_rate": 1.2795986789773145e-06,
+      "loss": 2.0896,
+      "step": 11676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.493932006825372,
+      "learning_rate": 1.2776265276360055e-06,
+      "loss": 1.91,
+      "step": 11677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.540197842995922,
+      "learning_rate": 1.275655857379926e-06,
+      "loss": 1.8271,
+      "step": 11678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.551498650963775,
+      "learning_rate": 1.2736866683321164e-06,
+      "loss": 1.8192,
+      "step": 11679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5517703114815048,
+      "learning_rate": 1.27171896061552e-06,
+      "loss": 2.0621,
+      "step": 11680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5264024798795706,
+      "learning_rate": 1.2697527343529802e-06,
+      "loss": 1.8793,
+      "step": 11681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5444054298611706,
+      "learning_rate": 1.2677879896672656e-06,
+      "loss": 1.8478,
+      "step": 11682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5374627600267693,
+      "learning_rate": 1.2658247266810475e-06,
+      "loss": 1.8385,
+      "step": 11683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5115284993837694,
+      "learning_rate": 1.263862945516886e-06,
+      "loss": 1.8877,
+      "step": 11684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5485831110715437,
+      "learning_rate": 1.2619026462972721e-06,
+      "loss": 2.0309,
+      "step": 11685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5550212097352878,
+      "learning_rate": 1.2599438291446026e-06,
+      "loss": 1.8081,
+      "step": 11686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.561576066890958,
+      "learning_rate": 1.257986494181157e-06,
+      "loss": 1.83,
+      "step": 11687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5693421753245838,
+      "learning_rate": 1.2560306415291546e-06,
+      "loss": 2.1009,
+      "step": 11688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5574440831779182,
+      "learning_rate": 1.254076271310703e-06,
+      "loss": 1.8411,
+      "step": 11689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5141377125876563,
+      "learning_rate": 1.252123383647813e-06,
+      "loss": 1.9309,
+      "step": 11690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5282351216418917,
+      "learning_rate": 1.250171978662426e-06,
+      "loss": 1.8565,
+      "step": 11691
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5655292069881842,
+      "learning_rate": 1.248222056476367e-06,
+      "loss": 1.8624,
+      "step": 11692
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5475046202705339,
+      "learning_rate": 1.2462736172113798e-06,
+      "loss": 2.0698,
+      "step": 11693
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5568818256302587,
+      "learning_rate": 1.2443266609891147e-06,
+      "loss": 1.8501,
+      "step": 11694
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5423448451050904,
+      "learning_rate": 1.2423811879311215e-06,
+      "loss": 1.7978,
+      "step": 11695
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.508833125413515,
+      "learning_rate": 1.2404371981588753e-06,
+      "loss": 1.8741,
+      "step": 11696
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5277871966599968,
+      "learning_rate": 1.2384946917937402e-06,
+      "loss": 2.0685,
+      "step": 11697
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.529910598485824,
+      "learning_rate": 1.2365536689569912e-06,
+      "loss": 1.833,
+      "step": 11698
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5434716499213597,
+      "learning_rate": 1.2346141297698288e-06,
+      "loss": 1.8591,
+      "step": 11699
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.540306294605111,
+      "learning_rate": 1.232676074353331e-06,
+      "loss": 2.0499,
+      "step": 11700
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5552632078108508,
+      "learning_rate": 1.2307395028285067e-06,
+      "loss": 1.8912,
+      "step": 11701
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5373591383581019,
+      "learning_rate": 1.2288044153162593e-06,
+      "loss": 1.9035,
+      "step": 11702
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5587410217514533,
+      "learning_rate": 1.2268708119374083e-06,
+      "loss": 1.8817,
+      "step": 11703
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5687774803904874,
+      "learning_rate": 1.2249386928126744e-06,
+      "loss": 1.836,
+      "step": 11704
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5164803666724173,
+      "learning_rate": 1.2230080580626885e-06,
+      "loss": 2.0257,
+      "step": 11705
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5431744424935566,
+      "learning_rate": 1.2210789078079876e-06,
+      "loss": 1.8537,
+      "step": 11706
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5543007448868885,
+      "learning_rate": 1.219151242169017e-06,
+      "loss": 1.8623,
+      "step": 11707
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5328856847781371,
+      "learning_rate": 1.217225061266128e-06,
+      "loss": 2.0623,
+      "step": 11708
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.540747534150304,
+      "learning_rate": 1.215300365219582e-06,
+      "loss": 1.9069,
+      "step": 11709
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5311373694961518,
+      "learning_rate": 1.213377154149542e-06,
+      "loss": 1.8586,
+      "step": 11710
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5701537270693223,
+      "learning_rate": 1.2114554281760866e-06,
+      "loss": 1.8764,
+      "step": 11711
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5382629987349338,
+      "learning_rate": 1.2095351874191946e-06,
+      "loss": 2.0619,
+      "step": 11712
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5529203820074862,
+      "learning_rate": 1.207616431998751e-06,
+      "loss": 1.8775,
+      "step": 11713
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5471952414976659,
+      "learning_rate": 1.2056991620345599e-06,
+      "loss": 1.8319,
+      "step": 11714
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.535681572745494,
+      "learning_rate": 1.203783377646317e-06,
+      "loss": 1.9151,
+      "step": 11715
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5675284741828543,
+      "learning_rate": 1.2018690789536352e-06,
+      "loss": 1.8395,
+      "step": 11716
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5304234249325372,
+      "learning_rate": 1.19995626607603e-06,
+      "loss": 2.0355,
+      "step": 11717
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.558426551540738,
+      "learning_rate": 1.1980449391329307e-06,
+      "loss": 1.8874,
+      "step": 11718
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5679184444588375,
+      "learning_rate": 1.196135098243667e-06,
+      "loss": 1.877,
+      "step": 11719
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5652704495527306,
+      "learning_rate": 1.1942267435274767e-06,
+      "loss": 2.0452,
+      "step": 11720
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5105423601508826,
+      "learning_rate": 1.1923198751035064e-06,
+      "loss": 1.9018,
+      "step": 11721
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.589169380400336,
+      "learning_rate": 1.1904144930908161e-06,
+      "loss": 1.8704,
+      "step": 11722
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5453583277882987,
+      "learning_rate": 1.1885105976083555e-06,
+      "loss": 1.8728,
+      "step": 11723
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5318337947410641,
+      "learning_rate": 1.1866081887749986e-06,
+      "loss": 1.842,
+      "step": 11724
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5404784524096817,
+      "learning_rate": 1.1847072667095256e-06,
+      "loss": 2.0476,
+      "step": 11725
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.563705007895485,
+      "learning_rate": 1.1828078315306052e-06,
+      "loss": 1.8918,
+      "step": 11726
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4945931238775916,
+      "learning_rate": 1.1809098833568426e-06,
+      "loss": 1.9327,
+      "step": 11727
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5627656930134513,
+      "learning_rate": 1.179013422306724e-06,
+      "loss": 1.8128,
+      "step": 11728
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.569377198273101,
+      "learning_rate": 1.177118448498657e-06,
+      "loss": 2.0869,
+      "step": 11729
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5350003569583077,
+      "learning_rate": 1.1752249620509553e-06,
+      "loss": 1.8696,
+      "step": 11730
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.568220048848639,
+      "learning_rate": 1.1733329630818273e-06,
+      "loss": 1.8396,
+      "step": 11731
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5397075551396808,
+      "learning_rate": 1.1714424517094065e-06,
+      "loss": 1.9979,
+      "step": 11732
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.518590728379184,
+      "learning_rate": 1.1695534280517283e-06,
+      "loss": 1.8887,
+      "step": 11733
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5525872491023931,
+      "learning_rate": 1.1676658922267215e-06,
+      "loss": 1.8226,
+      "step": 11734
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5787253552890166,
+      "learning_rate": 1.1657798443522466e-06,
+      "loss": 1.8767,
+      "step": 11735
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5826549138136291,
+      "learning_rate": 1.1638952845460432e-06,
+      "loss": 1.8482,
+      "step": 11736
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5625548084775526,
+      "learning_rate": 1.1620122129257783e-06,
+      "loss": 2.0367,
+      "step": 11737
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5314618488795448,
+      "learning_rate": 1.1601306296090243e-06,
+      "loss": 1.8234,
+      "step": 11738
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5505654203767429,
+      "learning_rate": 1.1582505347132482e-06,
+      "loss": 1.8673,
+      "step": 11739
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.50946198351235,
+      "learning_rate": 1.1563719283558371e-06,
+      "loss": 1.911,
+      "step": 11740
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5279208645648353,
+      "learning_rate": 1.1544948106540775e-06,
+      "loss": 2.0436,
+      "step": 11741
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5302752674669353,
+      "learning_rate": 1.1526191817251618e-06,
+      "loss": 1.78,
+      "step": 11742
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5365834237020748,
+      "learning_rate": 1.1507450416862075e-06,
+      "loss": 1.8179,
+      "step": 11743
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5765674355993516,
+      "learning_rate": 1.1488723906542098e-06,
+      "loss": 2.0456,
+      "step": 11744
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5583160738885791,
+      "learning_rate": 1.1470012287460918e-06,
+      "loss": 1.8891,
+      "step": 11745
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4981355220775474,
+      "learning_rate": 1.1451315560786796e-06,
+      "loss": 1.9219,
+      "step": 11746
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.537244029985671,
+      "learning_rate": 1.143263372768702e-06,
+      "loss": 1.7573,
+      "step": 11747
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5670799208266605,
+      "learning_rate": 1.141396678932799e-06,
+      "loss": 1.8452,
+      "step": 11748
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5628178483534041,
+      "learning_rate": 1.1395314746875134e-06,
+      "loss": 2.0636,
+      "step": 11749
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5561326788896732,
+      "learning_rate": 1.1376677601492996e-06,
+      "loss": 1.8317,
+      "step": 11750
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5712972328181289,
+      "learning_rate": 1.1358055354345172e-06,
+      "loss": 1.816,
+      "step": 11751
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.505380192796172,
+      "learning_rate": 1.1339448006594282e-06,
+      "loss": 1.9051,
+      "step": 11752
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5248917854481459,
+      "learning_rate": 1.1320855559402127e-06,
+      "loss": 2.0336,
+      "step": 11753
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5396657551550129,
+      "learning_rate": 1.1302278013929468e-06,
+      "loss": 1.8935,
+      "step": 11754
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5739422266499928,
+      "learning_rate": 1.1283715371336212e-06,
+      "loss": 1.8355,
+      "step": 11755
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5331158887257462,
+      "learning_rate": 1.1265167632781238e-06,
+      "loss": 1.8548,
+      "step": 11756
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.534123068524729,
+      "learning_rate": 1.1246634799422617e-06,
+      "loss": 2.01,
+      "step": 11757
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4985067594614268,
+      "learning_rate": 1.1228116872417398e-06,
+      "loss": 1.9783,
+      "step": 11758
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5682062384155223,
+      "learning_rate": 1.120961385292177e-06,
+      "loss": 1.8863,
+      "step": 11759
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5769116928791308,
+      "learning_rate": 1.1191125742090885e-06,
+      "loss": 1.8365,
+      "step": 11760
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5190870138042077,
+      "learning_rate": 1.1172652541079103e-06,
+      "loss": 2.0264,
+      "step": 11761
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.532632312095928,
+      "learning_rate": 1.1154194251039752e-06,
+      "loss": 1.8304,
+      "step": 11762
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5539867629519758,
+      "learning_rate": 1.113575087312524e-06,
+      "loss": 1.8593,
+      "step": 11763
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5203881259700491,
+      "learning_rate": 1.1117322408487064e-06,
+      "loss": 1.9438,
+      "step": 11764
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5354982316190441,
+      "learning_rate": 1.109890885827583e-06,
+      "loss": 2.0468,
+      "step": 11765
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5548755198085497,
+      "learning_rate": 1.108051022364115e-06,
+      "loss": 1.8531,
+      "step": 11766
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5369463412592698,
+      "learning_rate": 1.1062126505731713e-06,
+      "loss": 1.813,
+      "step": 11767
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5457424466873846,
+      "learning_rate": 1.1043757705695297e-06,
+      "loss": 1.8733,
+      "step": 11768
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5427023109282663,
+      "learning_rate": 1.102540382467876e-06,
+      "loss": 2.0437,
+      "step": 11769
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5457317006538139,
+      "learning_rate": 1.1007064863827966e-06,
+      "loss": 1.8539,
+      "step": 11770
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.510378003769033,
+      "learning_rate": 1.098874082428794e-06,
+      "loss": 1.945,
+      "step": 11771
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5535389413191191,
+      "learning_rate": 1.0970431707202717e-06,
+      "loss": 1.8788,
+      "step": 11772
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5452996261570157,
+      "learning_rate": 1.0952137513715348e-06,
+      "loss": 2.0343,
+      "step": 11773
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.544547770610094,
+      "learning_rate": 1.093385824496815e-06,
+      "loss": 1.8433,
+      "step": 11774
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5557503081773384,
+      "learning_rate": 1.0915593902102227e-06,
+      "loss": 1.8576,
+      "step": 11775
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5642614401685496,
+      "learning_rate": 1.0897344486257955e-06,
+      "loss": 1.8029,
+      "step": 11776
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5114328009568181,
+      "learning_rate": 1.0879109998574804e-06,
+      "loss": 2.0922,
+      "step": 11777
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5535535634538401,
+      "learning_rate": 1.0860890440191035e-06,
+      "loss": 1.841,
+      "step": 11778
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.564030046211814,
+      "learning_rate": 1.0842685812244347e-06,
+      "loss": 1.8548,
+      "step": 11779
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51640556280171,
+      "learning_rate": 1.0824496115871303e-06,
+      "loss": 1.8047,
+      "step": 11780
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.554209636544499,
+      "learning_rate": 1.0806321352207439e-06,
+      "loss": 2.0592,
+      "step": 11781
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5818740071796774,
+      "learning_rate": 1.0788161522387651e-06,
+      "loss": 1.8419,
+      "step": 11782
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5232021561864928,
+      "learning_rate": 1.0770016627545593e-06,
+      "loss": 1.9036,
+      "step": 11783
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5569178603423323,
+      "learning_rate": 1.075188666881416e-06,
+      "loss": 1.849,
+      "step": 11784
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5489386621755648,
+      "learning_rate": 1.0733771647325363e-06,
+      "loss": 2.0182,
+      "step": 11785
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5512114216480695,
+      "learning_rate": 1.0715671564210106e-06,
+      "loss": 1.884,
+      "step": 11786
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5556669596772178,
+      "learning_rate": 1.0697586420598427e-06,
+      "loss": 1.8795,
+      "step": 11787
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5540836288405323,
+      "learning_rate": 1.067951621761959e-06,
+      "loss": 1.8172,
+      "step": 11788
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5082737828472793,
+      "learning_rate": 1.0661460956401664e-06,
+      "loss": 2.1239,
+      "step": 11789
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5452476690803758,
+      "learning_rate": 1.0643420638071971e-06,
+      "loss": 1.8652,
+      "step": 11790
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5463291548225684,
+      "learning_rate": 1.0625395263756805e-06,
+      "loss": 1.8522,
+      "step": 11791
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5582753436705997,
+      "learning_rate": 1.060738483458157e-06,
+      "loss": 1.8517,
+      "step": 11792
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5411743462484173,
+      "learning_rate": 1.0589389351670809e-06,
+      "loss": 2.0387,
+      "step": 11793
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5494928086444191,
+      "learning_rate": 1.0571408816147987e-06,
+      "loss": 1.8261,
+      "step": 11794
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.51612670444131,
+      "learning_rate": 1.055344322913568e-06,
+      "loss": 1.9503,
+      "step": 11795
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5774723610747716,
+      "learning_rate": 1.0535492591755597e-06,
+      "loss": 1.8094,
+      "step": 11796
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5319660726300384,
+      "learning_rate": 1.0517556905128457e-06,
+      "loss": 2.0339,
+      "step": 11797
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5421727360932004,
+      "learning_rate": 1.0499636170374056e-06,
+      "loss": 1.8485,
+      "step": 11798
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5670375665884982,
+      "learning_rate": 1.048173038861125e-06,
+      "loss": 1.8999,
+      "step": 11799
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5772303193809998,
+      "learning_rate": 1.0463839560958006e-06,
+      "loss": 1.8618,
+      "step": 11800
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5480437310431202,
+      "learning_rate": 1.0445963688531264e-06,
+      "loss": 2.0734,
+      "step": 11801
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.517455140826079,
+      "learning_rate": 1.042810277244713e-06,
+      "loss": 1.8693,
+      "step": 11802
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5783830105799179,
+      "learning_rate": 1.041025681382074e-06,
+      "loss": 1.8309,
+      "step": 11803
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5635013311611135,
+      "learning_rate": 1.0392425813766287e-06,
+      "loss": 1.8114,
+      "step": 11804
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.575997976180365,
+      "learning_rate": 1.0374609773396992e-06,
+      "loss": 2.0573,
+      "step": 11805
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.548188256073481,
+      "learning_rate": 1.0356808693825244e-06,
+      "loss": 1.8728,
+      "step": 11806
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5502040637247195,
+      "learning_rate": 1.03390225761624e-06,
+      "loss": 1.8539,
+      "step": 11807
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5188579662928685,
+      "learning_rate": 1.0321251421518935e-06,
+      "loss": 1.9308,
+      "step": 11808
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5457352640913005,
+      "learning_rate": 1.0303495231004324e-06,
+      "loss": 2.0722,
+      "step": 11809
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5609458793380842,
+      "learning_rate": 1.0285754005727233e-06,
+      "loss": 1.8498,
+      "step": 11810
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5482076933283306,
+      "learning_rate": 1.0268027746795307e-06,
+      "loss": 1.8252,
+      "step": 11811
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5931409980376064,
+      "learning_rate": 1.025031645531524e-06,
+      "loss": 1.8747,
+      "step": 11812
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5587364256945818,
+      "learning_rate": 1.0232620132392817e-06,
+      "loss": 2.0968,
+      "step": 11813
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5222389838168767,
+      "learning_rate": 1.0214938779132932e-06,
+      "loss": 1.9463,
+      "step": 11814
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5329884277189912,
+      "learning_rate": 1.0197272396639478e-06,
+      "loss": 1.8673,
+      "step": 11815
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5403492472536453,
+      "learning_rate": 1.0179620986015465e-06,
+      "loss": 1.8685,
+      "step": 11816
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5545107649082348,
+      "learning_rate": 1.0161984548362869e-06,
+      "loss": 2.0621,
+      "step": 11817
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5415432863081561,
+      "learning_rate": 1.0144363084782864e-06,
+      "loss": 1.859,
+      "step": 11818
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.567183814006836,
+      "learning_rate": 1.0126756596375686e-06,
+      "loss": 1.8533,
+      "step": 11819
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4962614085235359,
+      "learning_rate": 1.0109165084240424e-06,
+      "loss": 1.9117,
+      "step": 11820
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5123647679954968,
+      "learning_rate": 1.0091588549475534e-06,
+      "loss": 2.0395,
+      "step": 11821
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5399521632202037,
+      "learning_rate": 1.007402699317836e-06,
+      "loss": 1.8251,
+      "step": 11822
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5698949193600872,
+      "learning_rate": 1.0056480416445273e-06,
+      "loss": 1.8861,
+      "step": 11823
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5317784965506607,
+      "learning_rate": 1.0038948820371874e-06,
+      "loss": 1.8325,
+      "step": 11824
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5310758506657064,
+      "learning_rate": 1.0021432206052645e-06,
+      "loss": 2.0241,
+      "step": 11825
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5221094901156447,
+      "learning_rate": 1.0003930574581266e-06,
+      "loss": 1.9432,
+      "step": 11826
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5571145825080576,
+      "learning_rate": 9.986443927050475e-07,
+      "loss": 1.895,
+      "step": 11827
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5262985987607308,
+      "learning_rate": 9.96897226455193e-07,
+      "loss": 1.9005,
+      "step": 11828
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.56579582424519,
+      "learning_rate": 9.951515588176558e-07,
+      "loss": 2.0806,
+      "step": 11829
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5497881730011629,
+      "learning_rate": 9.934073899014213e-07,
+      "loss": 1.8488,
+      "step": 11830
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5471937685333395,
+      "learning_rate": 9.9166471981538e-07,
+      "loss": 1.8683,
+      "step": 11831
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5391937218653906,
+      "learning_rate": 9.899235486683451e-07,
+      "loss": 1.8482,
+      "step": 11832
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5224584781116529,
+      "learning_rate": 9.881838765690154e-07,
+      "loss": 2.0845,
+      "step": 11833
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5753857611989643,
+      "learning_rate": 9.864457036260072e-07,
+      "loss": 1.8834,
+      "step": 11834
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5418116486071205,
+      "learning_rate": 9.8470902994785e-07,
+      "loss": 1.902,
+      "step": 11835
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5563870468251938,
+      "learning_rate": 9.829738556429629e-07,
+      "loss": 1.8204,
+      "step": 11836
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.528432009452502,
+      "learning_rate": 9.812401808196813e-07,
+      "loss": 2.0238,
+      "step": 11837
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5306562901430418,
+      "learning_rate": 9.795080055862466e-07,
+      "loss": 1.8485,
+      "step": 11838
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5152606913602151,
+      "learning_rate": 9.777773300508026e-07,
+      "loss": 1.9189,
+      "step": 11839
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5265339910541998,
+      "learning_rate": 9.760481543214128e-07,
+      "loss": 1.8243,
+      "step": 11840
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.533441625616188,
+      "learning_rate": 9.743204785060246e-07,
+      "loss": 2.0289,
+      "step": 11841
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5363355485539634,
+      "learning_rate": 9.725943027125124e-07,
+      "loss": 1.8521,
+      "step": 11842
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5691911052026314,
+      "learning_rate": 9.70869627048643e-07,
+      "loss": 1.8488,
+      "step": 11843
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.544980388832805,
+      "learning_rate": 9.691464516220966e-07,
+      "loss": 1.84,
+      "step": 11844
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5081780620891033,
+      "learning_rate": 9.674247765404571e-07,
+      "loss": 2.0888,
+      "step": 11845
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5600623798900044,
+      "learning_rate": 9.657046019112188e-07,
+      "loss": 1.8304,
+      "step": 11846
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5476236120841461,
+      "learning_rate": 9.639859278417761e-07,
+      "loss": 1.87,
+      "step": 11847
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5410034801548375,
+      "learning_rate": 9.622687544394382e-07,
+      "loss": 1.8196,
+      "step": 11848
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5711445190650498,
+      "learning_rate": 9.605530818114077e-07,
+      "loss": 2.0827,
+      "step": 11849
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5370827074356536,
+      "learning_rate": 9.58838910064805e-07,
+      "loss": 1.8848,
+      "step": 11850
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.522896154780989,
+      "learning_rate": 9.571262393066548e-07,
+      "loss": 1.8907,
+      "step": 11851
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.549417599896585,
+      "learning_rate": 9.554150696438834e-07,
+      "loss": 1.8545,
+      "step": 11852
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5608399590410367,
+      "learning_rate": 9.537054011833247e-07,
+      "loss": 2.0427,
+      "step": 11853
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5605400075515798,
+      "learning_rate": 9.519972340317262e-07,
+      "loss": 1.896,
+      "step": 11854
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5587087769314447,
+      "learning_rate": 9.50290568295728e-07,
+      "loss": 1.8249,
+      "step": 11855
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5425341471811116,
+      "learning_rate": 9.48585404081892e-07,
+      "loss": 1.8339,
+      "step": 11856
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5388478651779817,
+      "learning_rate": 9.468817414966719e-07,
+      "loss": 2.0671,
+      "step": 11857
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.505354736284435,
+      "learning_rate": 9.451795806464381e-07,
+      "loss": 1.94,
+      "step": 11858
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5663372866657157,
+      "learning_rate": 9.43478921637464e-07,
+      "loss": 1.8341,
+      "step": 11859
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5885815708016894,
+      "learning_rate": 9.417797645759285e-07,
+      "loss": 1.8658,
+      "step": 11860
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5469144714152501,
+      "learning_rate": 9.400821095679136e-07,
+      "loss": 2.0374,
+      "step": 11861
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5473432494973925,
+      "learning_rate": 9.383859567194148e-07,
+      "loss": 1.8638,
+      "step": 11862
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5469174168439362,
+      "learning_rate": 9.36691306136328e-07,
+      "loss": 1.8536,
+      "step": 11863
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4974083892842534,
+      "learning_rate": 9.349981579244549e-07,
+      "loss": 1.9522,
+      "step": 11864
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.574733731721297,
+      "learning_rate": 9.333065121895107e-07,
+      "loss": 2.1142,
+      "step": 11865
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5607069446456302,
+      "learning_rate": 9.31616369037111e-07,
+      "loss": 1.8515,
+      "step": 11866
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5805434259357729,
+      "learning_rate": 9.299277285727714e-07,
+      "loss": 1.81,
+      "step": 11867
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5495817125195068,
+      "learning_rate": 9.282405909019298e-07,
+      "loss": 1.8622,
+      "step": 11868
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5443258688246629,
+      "learning_rate": 9.265549561299213e-07,
+      "loss": 2.0641,
+      "step": 11869
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5226504797415402,
+      "learning_rate": 9.248708243619758e-07,
+      "loss": 1.922,
+      "step": 11870
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5643166882521794,
+      "learning_rate": 9.231881957032534e-07,
+      "loss": 1.8333,
+      "step": 11871
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5662526832419851,
+      "learning_rate": 9.21507070258798e-07,
+      "loss": 1.847,
+      "step": 11872
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5338919919606523,
+      "learning_rate": 9.198274481335728e-07,
+      "loss": 2.0022,
+      "step": 11873
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5620204242449471,
+      "learning_rate": 9.181493294324494e-07,
+      "loss": 1.8835,
+      "step": 11874
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5277450015660122,
+      "learning_rate": 9.164727142601859e-07,
+      "loss": 1.8417,
+      "step": 11875
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5012193863953166,
+      "learning_rate": 9.147976027214761e-07,
+      "loss": 1.8792,
+      "step": 11876
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5795859162532453,
+      "learning_rate": 9.131239949208947e-07,
+      "loss": 2.0434,
+      "step": 11877
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5418394861459361,
+      "learning_rate": 9.114518909629278e-07,
+      "loss": 1.8316,
+      "step": 11878
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5660283943019464,
+      "learning_rate": 9.097812909519859e-07,
+      "loss": 1.8387,
+      "step": 11879
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5736851482552975,
+      "learning_rate": 9.081121949923582e-07,
+      "loss": 1.8834,
+      "step": 11880
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5757140368323163,
+      "learning_rate": 9.064446031882556e-07,
+      "loss": 2.0582,
+      "step": 11881
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5180111595949544,
+      "learning_rate": 9.047785156438033e-07,
+      "loss": 1.9046,
+      "step": 11882
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5560004338699833,
+      "learning_rate": 9.031139324630095e-07,
+      "loss": 1.8751,
+      "step": 11883
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.544977212663069,
+      "learning_rate": 9.014508537498051e-07,
+      "loss": 1.8412,
+      "step": 11884
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5382486691967211,
+      "learning_rate": 8.997892796080209e-07,
+      "loss": 2.097,
+      "step": 11885
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.556459628801389,
+      "learning_rate": 8.981292101414019e-07,
+      "loss": 1.8018,
+      "step": 11886
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5458398133795308,
+      "learning_rate": 8.96470645453587e-07,
+      "loss": 1.8675,
+      "step": 11887
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5384218146489703,
+      "learning_rate": 8.948135856481298e-07,
+      "loss": 1.8591,
+      "step": 11888
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5335689472800059,
+      "learning_rate": 8.931580308284893e-07,
+      "loss": 1.9734,
+      "step": 11889
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5448979333650018,
+      "learning_rate": 8.915039810980242e-07,
+      "loss": 2.0614,
+      "step": 11890
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5341996357206228,
+      "learning_rate": 8.898514365600074e-07,
+      "loss": 1.8323,
+      "step": 11891
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5462844441674541,
+      "learning_rate": 8.882003973176095e-07,
+      "loss": 1.797,
+      "step": 11892
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5686052216123942,
+      "learning_rate": 8.865508634739172e-07,
+      "loss": 2.0506,
+      "step": 11893
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5421570968389228,
+      "learning_rate": 8.84902835131915e-07,
+      "loss": 1.8453,
+      "step": 11894
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5246606184418561,
+      "learning_rate": 8.832563123944926e-07,
+      "loss": 1.9431,
+      "step": 11895
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5710368911481135,
+      "learning_rate": 8.816112953644567e-07,
+      "loss": 1.8928,
+      "step": 11896
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5574967825633839,
+      "learning_rate": 8.799677841445059e-07,
+      "loss": 2.0443,
+      "step": 11897
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5468648614169509,
+      "learning_rate": 8.78325778837255e-07,
+      "loss": 1.8276,
+      "step": 11898
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5980513193997323,
+      "learning_rate": 8.766852795452224e-07,
+      "loss": 1.8486,
+      "step": 11899
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5638881637125096,
+      "learning_rate": 8.750462863708259e-07,
+      "loss": 1.8658,
+      "step": 11900
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5021591731341097,
+      "learning_rate": 8.734087994163975e-07,
+      "loss": 1.9167,
+      "step": 11901
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5465755382603655,
+      "learning_rate": 8.717728187841722e-07,
+      "loss": 2.0215,
+      "step": 11902
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5423741351545511,
+      "learning_rate": 8.701383445762934e-07,
+      "loss": 1.8442,
+      "step": 11903
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5428343966907364,
+      "learning_rate": 8.685053768948043e-07,
+      "loss": 1.8557,
+      "step": 11904
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.534695345026071,
+      "learning_rate": 8.668739158416622e-07,
+      "loss": 2.0782,
+      "step": 11905
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5719366583091761,
+      "learning_rate": 8.652439615187163e-07,
+      "loss": 1.8724,
+      "step": 11906
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5087360655115176,
+      "learning_rate": 8.636155140277408e-07,
+      "loss": 1.883,
+      "step": 11907
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5816074291386367,
+      "learning_rate": 8.619885734704042e-07,
+      "loss": 1.8206,
+      "step": 11908
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5646659408652039,
+      "learning_rate": 8.603631399482809e-07,
+      "loss": 1.878,
+      "step": 11909
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5314669057588279,
+      "learning_rate": 8.587392135628591e-07,
+      "loss": 2.0513,
+      "step": 11910
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5532194759434288,
+      "learning_rate": 8.57116794415519e-07,
+      "loss": 1.8029,
+      "step": 11911
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.539268647887158,
+      "learning_rate": 8.554958826075599e-07,
+      "loss": 1.8628,
+      "step": 11912
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5081376630713487,
+      "learning_rate": 8.538764782401842e-07,
+      "loss": 1.9415,
+      "step": 11913
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5603904436695119,
+      "learning_rate": 8.52258581414489e-07,
+      "loss": 2.0655,
+      "step": 11914
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5476354524819602,
+      "learning_rate": 8.506421922314934e-07,
+      "loss": 1.8395,
+      "step": 11915
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5638888659668961,
+      "learning_rate": 8.490273107921165e-07,
+      "loss": 1.8492,
+      "step": 11916
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5228077267876022,
+      "learning_rate": 8.474139371971751e-07,
+      "loss": 2.0645,
+      "step": 11917
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5409271899646221,
+      "learning_rate": 8.45802071547408e-07,
+      "loss": 1.8687,
+      "step": 11918
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5621099302541483,
+      "learning_rate": 8.44191713943443e-07,
+      "loss": 1.839,
+      "step": 11919
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5502732220891616,
+      "learning_rate": 8.425828644858219e-07,
+      "loss": 1.9343,
+      "step": 11920
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5857710001146665,
+      "learning_rate": 8.409755232749977e-07,
+      "loss": 1.8051,
+      "step": 11921
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5695933943707873,
+      "learning_rate": 8.393696904113152e-07,
+      "loss": 2.0803,
+      "step": 11922
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5618243614538307,
+      "learning_rate": 8.377653659950441e-07,
+      "loss": 1.8803,
+      "step": 11923
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.531793837697941,
+      "learning_rate": 8.361625501263348e-07,
+      "loss": 1.8341,
+      "step": 11924
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.552410194383956,
+      "learning_rate": 8.345612429052658e-07,
+      "loss": 2.0645,
+      "step": 11925
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5021487485531813,
+      "learning_rate": 8.329614444318179e-07,
+      "loss": 1.9366,
+      "step": 11926
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5400174975334813,
+      "learning_rate": 8.313631548058642e-07,
+      "loss": 1.8499,
+      "step": 11927
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5691426724455668,
+      "learning_rate": 8.297663741271916e-07,
+      "loss": 1.8464,
+      "step": 11928
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5134475907701236,
+      "learning_rate": 8.281711024955063e-07,
+      "loss": 2.0638,
+      "step": 11929
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5432226467773476,
+      "learning_rate": 8.265773400103954e-07,
+      "loss": 1.8662,
+      "step": 11930
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5298302401537774,
+      "learning_rate": 8.249850867713682e-07,
+      "loss": 1.8365,
+      "step": 11931
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5060669420050612,
+      "learning_rate": 8.233943428778368e-07,
+      "loss": 1.9298,
+      "step": 11932
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5486658629518246,
+      "learning_rate": 8.218051084291162e-07,
+      "loss": 1.8796,
+      "step": 11933
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5153681845088866,
+      "learning_rate": 8.202173835244271e-07,
+      "loss": 2.0416,
+      "step": 11934
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5483279979116691,
+      "learning_rate": 8.186311682629011e-07,
+      "loss": 1.8554,
+      "step": 11935
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5314938246671458,
+      "learning_rate": 8.170464627435704e-07,
+      "loss": 1.82,
+      "step": 11936
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5298649424089256,
+      "learning_rate": 8.15463267065375e-07,
+      "loss": 2.0815,
+      "step": 11937
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5278995215211619,
+      "learning_rate": 8.138815813271611e-07,
+      "loss": 1.9195,
+      "step": 11938
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5718884048293611,
+      "learning_rate": 8.123014056276801e-07,
+      "loss": 1.8503,
+      "step": 11939
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5672190365499018,
+      "learning_rate": 8.107227400655864e-07,
+      "loss": 1.8631,
+      "step": 11940
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5472847075074475,
+      "learning_rate": 8.09145584739443e-07,
+      "loss": 1.854,
+      "step": 11941
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5397064844627799,
+      "learning_rate": 8.07569939747721e-07,
+      "loss": 2.0831,
+      "step": 11942
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5576858655157813,
+      "learning_rate": 8.059958051887917e-07,
+      "loss": 1.8548,
+      "step": 11943
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5005606177798906,
+      "learning_rate": 8.044231811609376e-07,
+      "loss": 1.8836,
+      "step": 11944
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5706739615983839,
+      "learning_rate": 8.028520677623414e-07,
+      "loss": 1.8601,
+      "step": 11945
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5290625460223916,
+      "learning_rate": 8.012824650910938e-07,
+      "loss": 2.0629,
+      "step": 11946
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5576930280306315,
+      "learning_rate": 7.997143732451945e-07,
+      "loss": 1.8738,
+      "step": 11947
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5630809797159464,
+      "learning_rate": 7.981477923225428e-07,
+      "loss": 1.8955,
+      "step": 11948
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5546566379075027,
+      "learning_rate": 7.965827224209466e-07,
+      "loss": 2.0761,
+      "step": 11949
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5702444145629301,
+      "learning_rate": 7.950191636381249e-07,
+      "loss": 1.8768,
+      "step": 11950
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5116258652672493,
+      "learning_rate": 7.934571160716914e-07,
+      "loss": 1.9289,
+      "step": 11951
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5583553447553387,
+      "learning_rate": 7.918965798191763e-07,
+      "loss": 1.8486,
+      "step": 11952
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5521591979988606,
+      "learning_rate": 7.90337554977999e-07,
+      "loss": 1.8504,
+      "step": 11953
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5431407955011353,
+      "learning_rate": 7.887800416455093e-07,
+      "loss": 2.0314,
+      "step": 11954
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6003466095351941,
+      "learning_rate": 7.872240399189434e-07,
+      "loss": 1.8956,
+      "step": 11955
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5541574981327462,
+      "learning_rate": 7.856695498954486e-07,
+      "loss": 1.8989,
+      "step": 11956
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5086403327007514,
+      "learning_rate": 7.841165716720833e-07,
+      "loss": 1.876,
+      "step": 11957
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5613833546116672,
+      "learning_rate": 7.825651053457949e-07,
+      "loss": 2.0115,
+      "step": 11958
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5538188415848389,
+      "learning_rate": 7.810151510134589e-07,
+      "loss": 1.8556,
+      "step": 11959
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5390133573782259,
+      "learning_rate": 7.794667087718421e-07,
+      "loss": 1.7937,
+      "step": 11960
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5089928894748439,
+      "learning_rate": 7.779197787176145e-07,
+      "loss": 2.0045,
+      "step": 11961
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5459011013537541,
+      "learning_rate": 7.763743609473628e-07,
+      "loss": 1.8717,
+      "step": 11962
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5186150708865325,
+      "learning_rate": 7.748304555575792e-07,
+      "loss": 1.9197,
+      "step": 11963
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5549381498410941,
+      "learning_rate": 7.73288062644642e-07,
+      "loss": 1.8279,
+      "step": 11964
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5466436116645752,
+      "learning_rate": 7.717471823048605e-07,
+      "loss": 1.9041,
+      "step": 11965
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5474183394380007,
+      "learning_rate": 7.702078146344354e-07,
+      "loss": 2.1042,
+      "step": 11966
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5713697551234194,
+      "learning_rate": 7.686699597294705e-07,
+      "loss": 1.8527,
+      "step": 11967
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5911695677904448,
+      "learning_rate": 7.671336176859917e-07,
+      "loss": 1.8762,
+      "step": 11968
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.542545247927804,
+      "learning_rate": 7.655987885999056e-07,
+      "loss": 1.9272,
+      "step": 11969
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5542683880007954,
+      "learning_rate": 7.640654725670465e-07,
+      "loss": 2.0525,
+      "step": 11970
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5647088679849149,
+      "learning_rate": 7.625336696831464e-07,
+      "loss": 1.8235,
+      "step": 11971
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5472405972053979,
+      "learning_rate": 7.610033800438344e-07,
+      "loss": 1.8649,
+      "step": 11972
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5640594451569386,
+      "learning_rate": 7.594746037446642e-07,
+      "loss": 1.8706,
+      "step": 11973
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5311015609978653,
+      "learning_rate": 7.579473408810739e-07,
+      "loss": 2.0539,
+      "step": 11974
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5284631743060324,
+      "learning_rate": 7.564215915484146e-07,
+      "loss": 1.9215,
+      "step": 11975
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5402489815762331,
+      "learning_rate": 7.548973558419603e-07,
+      "loss": 1.8788,
+      "step": 11976
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5405191594318214,
+      "learning_rate": 7.533746338568599e-07,
+      "loss": 1.8656,
+      "step": 11977
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5359344702495117,
+      "learning_rate": 7.518534256881898e-07,
+      "loss": 2.0282,
+      "step": 11978
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5406650486089863,
+      "learning_rate": 7.503337314309217e-07,
+      "loss": 1.8385,
+      "step": 11979
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5429358016034893,
+      "learning_rate": 7.488155511799433e-07,
+      "loss": 1.8745,
+      "step": 11980
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5114506174523652,
+      "learning_rate": 7.472988850300317e-07,
+      "loss": 2.0855,
+      "step": 11981
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5296394043509212,
+      "learning_rate": 7.45783733075886e-07,
+      "loss": 1.9833,
+      "step": 11982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.548173780356348,
+      "learning_rate": 7.442700954121001e-07,
+      "loss": 1.8445,
+      "step": 11983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5317315758945761,
+      "learning_rate": 7.42757972133179e-07,
+      "loss": 1.7907,
+      "step": 11984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5706128891721373,
+      "learning_rate": 7.412473633335277e-07,
+      "loss": 1.868,
+      "step": 11985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5618555972535536,
+      "learning_rate": 7.397382691074595e-07,
+      "loss": 2.0272,
+      "step": 11986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5499407966878038,
+      "learning_rate": 7.382306895491964e-07,
+      "loss": 1.8174,
+      "step": 11987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5200596198104859,
+      "learning_rate": 7.367246247528603e-07,
+      "loss": 1.9104,
+      "step": 11988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5436961977209263,
+      "learning_rate": 7.352200748124816e-07,
+      "loss": 1.8366,
+      "step": 11989
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.532953175263536,
+      "learning_rate": 7.337170398219962e-07,
+      "loss": 2.0576,
+      "step": 11990
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.534701341634024,
+      "learning_rate": 7.322155198752456e-07,
+      "loss": 1.8574,
+      "step": 11991
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5433682535268415,
+      "learning_rate": 7.307155150659744e-07,
+      "loss": 1.8851,
+      "step": 11992
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5719929165540277,
+      "learning_rate": 7.292170254878328e-07,
+      "loss": 1.83,
+      "step": 11993
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5096531979117557,
+      "learning_rate": 7.277200512343818e-07,
+      "loss": 2.0602,
+      "step": 11994
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5396676452755502,
+      "learning_rate": 7.262245923990773e-07,
+      "loss": 1.8827,
+      "step": 11995
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5474877917638199,
+      "learning_rate": 7.247306490752948e-07,
+      "loss": 1.8967,
+      "step": 11996
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5404838591180459,
+      "learning_rate": 7.232382213563011e-07,
+      "loss": 1.8592,
+      "step": 11997
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5297255502773219,
+      "learning_rate": 7.217473093352745e-07,
+      "loss": 2.0471,
+      "step": 11998
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5999482569534146,
+      "learning_rate": 7.202579131053072e-07,
+      "loss": 1.8287,
+      "step": 11999
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5157188688486453,
+      "learning_rate": 7.187700327593749e-07,
+      "loss": 1.9667,
+      "step": 12000
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5399714605368815,
+      "learning_rate": 7.172836683903811e-07,
+      "loss": 1.8492,
+      "step": 12001
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5670168143656699,
+      "learning_rate": 7.157988200911291e-07,
+      "loss": 2.06,
+      "step": 12002
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5443737022249971,
+      "learning_rate": 7.143154879543118e-07,
+      "loss": 1.8194,
+      "step": 12003
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5778497121678692,
+      "learning_rate": 7.12833672072552e-07,
+      "loss": 1.88,
+      "step": 12004
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5742933519939378,
+      "learning_rate": 7.113533725383564e-07,
+      "loss": 1.8713,
+      "step": 12005
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4891852658328119,
+      "learning_rate": 7.098745894441511e-07,
+      "loss": 2.0514,
+      "step": 12006
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5432824662067035,
+      "learning_rate": 7.083973228822648e-07,
+      "loss": 1.8328,
+      "step": 12007
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5380265985228503,
+      "learning_rate": 7.069215729449186e-07,
+      "loss": 1.8358,
+      "step": 12008
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5668338340377961,
+      "learning_rate": 7.054473397242634e-07,
+      "loss": 1.8835,
+      "step": 12009
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5732653691430536,
+      "learning_rate": 7.03974623312334e-07,
+      "loss": 2.0706,
+      "step": 12010
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5487969385206273,
+      "learning_rate": 7.025034238010764e-07,
+      "loss": 1.8164,
+      "step": 12011
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5357722996604882,
+      "learning_rate": 7.010337412823531e-07,
+      "loss": 1.8167,
+      "step": 12012
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5136236043033853,
+      "learning_rate": 6.995655758479103e-07,
+      "loss": 1.923,
+      "step": 12013
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5157797273743138,
+      "learning_rate": 6.980989275894162e-07,
+      "loss": 2.0702,
+      "step": 12014
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5532182062532611,
+      "learning_rate": 6.966337965984476e-07,
+      "loss": 1.8432,
+      "step": 12015
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.539683100410497,
+      "learning_rate": 6.951701829664675e-07,
+      "loss": 1.8699,
+      "step": 12016
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5603943534852046,
+      "learning_rate": 6.937080867848583e-07,
+      "loss": 1.8639,
+      "step": 12017
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.559670081500926,
+      "learning_rate": 6.922475081449082e-07,
+      "loss": 2.06,
+      "step": 12018
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.499788380330511,
+      "learning_rate": 6.907884471378023e-07,
+      "loss": 1.9303,
+      "step": 12019
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5373315425909709,
+      "learning_rate": 6.893309038546431e-07,
+      "loss": 1.8847,
+      "step": 12020
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5554722734027194,
+      "learning_rate": 6.878748783864242e-07,
+      "loss": 1.8624,
+      "step": 12021
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.548067132090641,
+      "learning_rate": 6.864203708240507e-07,
+      "loss": 2.0301,
+      "step": 12022
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5509567734675266,
+      "learning_rate": 6.849673812583391e-07,
+      "loss": 1.8359,
+      "step": 12023
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5798341510670232,
+      "learning_rate": 6.835159097800026e-07,
+      "loss": 1.8936,
+      "step": 12024
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5129686984181393,
+      "learning_rate": 6.820659564796605e-07,
+      "loss": 1.9308,
+      "step": 12025
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5655841540613078,
+      "learning_rate": 6.806175214478433e-07,
+      "loss": 2.0763,
+      "step": 12026
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5525845501780215,
+      "learning_rate": 6.791706047749785e-07,
+      "loss": 1.8715,
+      "step": 12027
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5596378698238804,
+      "learning_rate": 6.777252065514078e-07,
+      "loss": 1.8456,
+      "step": 12028
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5615912898264964,
+      "learning_rate": 6.762813268673701e-07,
+      "loss": 1.8172,
+      "step": 12029
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.532558233627528,
+      "learning_rate": 6.748389658130128e-07,
+      "loss": 2.054,
+      "step": 12030
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5270721388356946,
+      "learning_rate": 6.733981234783887e-07,
+      "loss": 1.9429,
+      "step": 12031
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5644681120940344,
+      "learning_rate": 6.719587999534565e-07,
+      "loss": 1.8606,
+      "step": 12032
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5283696269313857,
+      "learning_rate": 6.705209953280778e-07,
+      "loss": 1.8082,
+      "step": 12033
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5394656269774584,
+      "learning_rate": 6.690847096920222e-07,
+      "loss": 2.0301,
+      "step": 12034
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5517927351674337,
+      "learning_rate": 6.676499431349625e-07,
+      "loss": 1.911,
+      "step": 12035
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.562096439790101,
+      "learning_rate": 6.662166957464772e-07,
+      "loss": 1.8715,
+      "step": 12036
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5552496653630097,
+      "learning_rate": 6.647849676160472e-07,
+      "loss": 1.8506,
+      "step": 12037
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.49935203096553693,
+      "learning_rate": 6.633547588330652e-07,
+      "loss": 2.0622,
+      "step": 12038
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5599250506696739,
+      "learning_rate": 6.619260694868235e-07,
+      "loss": 1.8735,
+      "step": 12039
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5456614907798658,
+      "learning_rate": 6.604988996665173e-07,
+      "loss": 1.8643,
+      "step": 12040
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5536663243623563,
+      "learning_rate": 6.590732494612562e-07,
+      "loss": 1.8719,
+      "step": 12041
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5463697326516083,
+      "learning_rate": 6.576491189600465e-07,
+      "loss": 2.0502,
+      "step": 12042
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5562017834588785,
+      "learning_rate": 6.562265082518059e-07,
+      "loss": 1.8859,
+      "step": 12043
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5320660893448803,
+      "learning_rate": 6.54805417425347e-07,
+      "loss": 1.9479,
+      "step": 12044
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5448159752183019,
+      "learning_rate": 6.533858465694015e-07,
+      "loss": 1.856,
+      "step": 12045
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5157673866739633,
+      "learning_rate": 6.519677957725956e-07,
+      "loss": 2.0269,
+      "step": 12046
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5448171398172151,
+      "learning_rate": 6.505512651234613e-07,
+      "loss": 1.8996,
+      "step": 12047
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5646446400961729,
+      "learning_rate": 6.491362547104446e-07,
+      "loss": 1.8423,
+      "step": 12048
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5313713198139957,
+      "learning_rate": 6.477227646218887e-07,
+      "loss": 1.9068,
+      "step": 12049
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5110933774164131,
+      "learning_rate": 6.463107949460395e-07,
+      "loss": 2.136,
+      "step": 12050
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5560249762278169,
+      "learning_rate": 6.449003457710546e-07,
+      "loss": 1.8414,
+      "step": 12051
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5378572424290625,
+      "learning_rate": 6.434914171849965e-07,
+      "loss": 1.8422,
+      "step": 12052
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5481566825593707,
+      "learning_rate": 6.420840092758257e-07,
+      "loss": 1.8436,
+      "step": 12053
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5565134971772293,
+      "learning_rate": 6.40678122131419e-07,
+      "loss": 2.0849,
+      "step": 12054
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5413968469171467,
+      "learning_rate": 6.392737558395423e-07,
+      "loss": 1.8559,
+      "step": 12055
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5034657565372168,
+      "learning_rate": 6.378709104878839e-07,
+      "loss": 1.9399,
+      "step": 12056
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.551262151656272,
+      "learning_rate": 6.364695861640318e-07,
+      "loss": 1.8214,
+      "step": 12057
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5337161582732907,
+      "learning_rate": 6.350697829554636e-07,
+      "loss": 2.0487,
+      "step": 12058
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5366577107713069,
+      "learning_rate": 6.33671500949587e-07,
+      "loss": 1.8142,
+      "step": 12059
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5456591215677576,
+      "learning_rate": 6.322747402336959e-07,
+      "loss": 1.8366,
+      "step": 12060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5318763837532458,
+      "learning_rate": 6.308795008949958e-07,
+      "loss": 1.832,
+      "step": 12061
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5027404758948059,
+      "learning_rate": 6.294857830206058e-07,
+      "loss": 2.104,
+      "step": 12062
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5495700252760988,
+      "learning_rate": 6.280935866975313e-07,
+      "loss": 1.853,
+      "step": 12063
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5527659919622202,
+      "learning_rate": 6.267029120126944e-07,
+      "loss": 1.8391,
+      "step": 12064
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5389161929513338,
+      "learning_rate": 6.253137590529257e-07,
+      "loss": 1.7941,
+      "step": 12065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5190827140714885,
+      "learning_rate": 6.239261279049529e-07,
+      "loss": 2.0275,
+      "step": 12066
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5547715258785623,
+      "learning_rate": 6.225400186554098e-07,
+      "loss": 1.8141,
+      "step": 12067
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.540224719139451,
+      "learning_rate": 6.21155431390838e-07,
+      "loss": 1.8219,
+      "step": 12068
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5085356653522338,
+      "learning_rate": 6.197723661976824e-07,
+      "loss": 1.9397,
+      "step": 12069
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.546295112903078,
+      "learning_rate": 6.183908231622986e-07,
+      "loss": 2.073,
+      "step": 12070
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5262887184138038,
+      "learning_rate": 6.170108023709348e-07,
+      "loss": 1.8918,
+      "step": 12071
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5632086587309858,
+      "learning_rate": 6.15632303909755e-07,
+      "loss": 1.8636,
+      "step": 12072
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5524100704268221,
+      "learning_rate": 6.142553278648239e-07,
+      "loss": 1.852,
+      "step": 12073
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5279465094978973,
+      "learning_rate": 6.128798743221143e-07,
+      "loss": 2.1131,
+      "step": 12074
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5096687476904901,
+      "learning_rate": 6.115059433674963e-07,
+      "loss": 1.9601,
+      "step": 12075
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5510228145494964,
+      "learning_rate": 6.101335350867515e-07,
+      "loss": 1.8322,
+      "step": 12076
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5529306529549411,
+      "learning_rate": 6.087626495655696e-07,
+      "loss": 1.8869,
+      "step": 12077
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5379775757040772,
+      "learning_rate": 6.073932868895349e-07,
+      "loss": 2.0862,
+      "step": 12078
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5583679009194233,
+      "learning_rate": 6.060254471441456e-07,
+      "loss": 1.8529,
+      "step": 12079
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5473126156099183,
+      "learning_rate": 6.046591304148003e-07,
+      "loss": 1.8528,
+      "step": 12080
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5186878496649145,
+      "learning_rate": 6.032943367868027e-07,
+      "loss": 1.9441,
+      "step": 12081
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5389580000196879,
+      "learning_rate": 6.019310663453653e-07,
+      "loss": 2.0365,
+      "step": 12082
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6111879803525038,
+      "learning_rate": 6.005693191756006e-07,
+      "loss": 1.8427,
+      "step": 12083
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5498179434010276,
+      "learning_rate": 5.992090953625295e-07,
+      "loss": 1.8574,
+      "step": 12084
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5493380725347365,
+      "learning_rate": 5.978503949910757e-07,
+      "loss": 1.8433,
+      "step": 12085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.519289108908545,
+      "learning_rate": 5.964932181460658e-07,
+      "loss": 1.9947,
+      "step": 12086
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5216815187795476,
+      "learning_rate": 5.951375649122404e-07,
+      "loss": 1.9278,
+      "step": 12087
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5695668700667512,
+      "learning_rate": 5.937834353742316e-07,
+      "loss": 1.8564,
+      "step": 12088
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5295866250049729,
+      "learning_rate": 5.924308296165859e-07,
+      "loss": 1.7958,
+      "step": 12089
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5129824341499704,
+      "learning_rate": 5.910797477237551e-07,
+      "loss": 2.0275,
+      "step": 12090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5554984307209867,
+      "learning_rate": 5.897301897800883e-07,
+      "loss": 1.8665,
+      "step": 12091
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5217822294880273,
+      "learning_rate": 5.883821558698432e-07,
+      "loss": 1.8822,
+      "step": 12092
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5038629141065102,
+      "learning_rate": 5.870356460771915e-07,
+      "loss": 1.9113,
+      "step": 12093
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5222649185787611,
+      "learning_rate": 5.856906604861906e-07,
+      "loss": 2.0156,
+      "step": 12094
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.547598621198585,
+      "learning_rate": 5.84347199180818e-07,
+      "loss": 1.8406,
+      "step": 12095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5473717519739777,
+      "learning_rate": 5.830052622449566e-07,
+      "loss": 1.8839,
+      "step": 12096
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5436525485233353,
+      "learning_rate": 5.816648497623783e-07,
+      "loss": 1.8307,
+      "step": 12097
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5383867410736927,
+      "learning_rate": 5.8032596181678e-07,
+      "loss": 2.0667,
+      "step": 12098
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5630590629347338,
+      "learning_rate": 5.789885984917504e-07,
+      "loss": 1.864,
+      "step": 12099
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5056727428791555,
+      "learning_rate": 5.776527598707837e-07,
+      "loss": 1.9184,
+      "step": 12100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5333257597445358,
+      "learning_rate": 5.763184460372884e-07,
+      "loss": 1.8192,
+      "step": 12101
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5225802417729573,
+      "learning_rate": 5.749856570745643e-07,
+      "loss": 2.0243,
+      "step": 12102
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5792814458973582,
+      "learning_rate": 5.736543930658284e-07,
+      "loss": 1.8291,
+      "step": 12103
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5608060018255706,
+      "learning_rate": 5.723246540941946e-07,
+      "loss": 1.8407,
+      "step": 12104
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5602728477966785,
+      "learning_rate": 5.709964402426826e-07,
+      "loss": 1.8661,
+      "step": 12105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5064918229671801,
+      "learning_rate": 5.696697515942235e-07,
+      "loss": 1.8765,
+      "step": 12106
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5194614451807128,
+      "learning_rate": 5.683445882316396e-07,
+      "loss": 2.004,
+      "step": 12107
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.57016010923494,
+      "learning_rate": 5.670209502376706e-07,
+      "loss": 1.8499,
+      "step": 12108
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.54206056839908,
+      "learning_rate": 5.656988376949613e-07,
+      "loss": 1.8657,
+      "step": 12109
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5465392592248804,
+      "learning_rate": 5.643782506860484e-07,
+      "loss": 2.0576,
+      "step": 12110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5509752857808684,
+      "learning_rate": 5.630591892933857e-07,
+      "loss": 1.8976,
+      "step": 12111
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4984076012332594,
+      "learning_rate": 5.617416535993291e-07,
+      "loss": 1.8952,
+      "step": 12112
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5809744286810843,
+      "learning_rate": 5.604256436861355e-07,
+      "loss": 1.8262,
+      "step": 12113
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5374952332161287,
+      "learning_rate": 5.591111596359666e-07,
+      "loss": 2.014,
+      "step": 12114
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5332028291482898,
+      "learning_rate": 5.577982015308958e-07,
+      "loss": 1.8591,
+      "step": 12115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5585024938748564,
+      "learning_rate": 5.564867694528935e-07,
+      "loss": 1.8533,
+      "step": 12116
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5560685959676128,
+      "learning_rate": 5.551768634838417e-07,
+      "loss": 1.8669,
+      "step": 12117
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5440056492524306,
+      "learning_rate": 5.538684837055163e-07,
+      "loss": 1.9523,
+      "step": 12118
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5121292914098255,
+      "learning_rate": 5.525616301996078e-07,
+      "loss": 2.0361,
+      "step": 12119
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5462760916854598,
+      "learning_rate": 5.512563030477146e-07,
+      "loss": 1.821,
+      "step": 12120
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5418468983622584,
+      "learning_rate": 5.499525023313273e-07,
+      "loss": 1.8454,
+      "step": 12121
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5346413962763046,
+      "learning_rate": 5.486502281318473e-07,
+      "loss": 2.0669,
+      "step": 12122
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5398009281824697,
+      "learning_rate": 5.473494805305818e-07,
+      "loss": 1.8483,
+      "step": 12123
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5003600787040667,
+      "learning_rate": 5.460502596087408e-07,
+      "loss": 1.9661,
+      "step": 12124
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5622621099640973,
+      "learning_rate": 5.447525654474428e-07,
+      "loss": 1.8474,
+      "step": 12125
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5426874842493473,
+      "learning_rate": 5.434563981277063e-07,
+      "loss": 1.8621,
+      "step": 12126
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5202740840655923,
+      "learning_rate": 5.421617577304583e-07,
+      "loss": 2.0296,
+      "step": 12127
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5584009958630821,
+      "learning_rate": 5.408686443365257e-07,
+      "loss": 1.8683,
+      "step": 12128
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5326150135490557,
+      "learning_rate": 5.39577058026644e-07,
+      "loss": 1.8318,
+      "step": 12129
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5064788744843992,
+      "learning_rate": 5.382869988814543e-07,
+      "loss": 2.0582,
+      "step": 12130
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5045057071303688,
+      "learning_rate": 5.369984669814976e-07,
+      "loss": 1.9598,
+      "step": 12131
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.563094415580324,
+      "learning_rate": 5.357114624072235e-07,
+      "loss": 1.8792,
+      "step": 12132
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5425507622075342,
+      "learning_rate": 5.344259852389844e-07,
+      "loss": 1.8703,
+      "step": 12133
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5453864811909958,
+      "learning_rate": 5.331420355570382e-07,
+      "loss": 2.0554,
+      "step": 12134
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5407871471121478,
+      "learning_rate": 5.31859613441546e-07,
+      "loss": 1.8406,
+      "step": 12135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5592319995955283,
+      "learning_rate": 5.305787189725769e-07,
+      "loss": 1.8804,
+      "step": 12136
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5074730849703841,
+      "learning_rate": 5.292993522301005e-07,
+      "loss": 1.9213,
+      "step": 12137
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5455377662819488,
+      "learning_rate": 5.280215132939942e-07,
+      "loss": 1.8096,
+      "step": 12138
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5436737766226606,
+      "learning_rate": 5.267452022440389e-07,
+      "loss": 2.0154,
+      "step": 12139
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5421312487172725,
+      "learning_rate": 5.254704191599208e-07,
+      "loss": 1.829,
+      "step": 12140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5774860151621288,
+      "learning_rate": 5.241971641212234e-07,
+      "loss": 1.8712,
+      "step": 12141
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5295691503089527,
+      "learning_rate": 5.229254372074471e-07,
+      "loss": 2.0486,
+      "step": 12142
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5112865990683462,
+      "learning_rate": 5.21655238497995e-07,
+      "loss": 1.9115,
+      "step": 12143
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5491790652722256,
+      "learning_rate": 5.203865680721593e-07,
+      "loss": 1.8245,
+      "step": 12144
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5339669367896346,
+      "learning_rate": 5.191194260091597e-07,
+      "loss": 1.8649,
+      "step": 12145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5189749243388174,
+      "learning_rate": 5.178538123881055e-07,
+      "loss": 1.9766,
+      "step": 12146
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5475449812726021,
+      "learning_rate": 5.165897272880082e-07,
+      "loss": 1.8079,
+      "step": 12147
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5409235159761037,
+      "learning_rate": 5.153271707877994e-07,
+      "loss": 1.8517,
+      "step": 12148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5255110994406206,
+      "learning_rate": 5.140661429662963e-07,
+      "loss": 1.9023,
+      "step": 12149
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5399704780944399,
+      "learning_rate": 5.128066439022361e-07,
+      "loss": 1.8641,
+      "step": 12150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5615741487356377,
+      "learning_rate": 5.115486736742559e-07,
+      "loss": 2.0317,
+      "step": 12151
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.572540203570783,
+      "learning_rate": 5.102922323608872e-07,
+      "loss": 1.856,
+      "step": 12152
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5576647822411424,
+      "learning_rate": 5.090373200405868e-07,
+      "loss": 1.8761,
+      "step": 12153
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5357490382229382,
+      "learning_rate": 5.077839367916948e-07,
+      "loss": 2.0908,
+      "step": 12154
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.48986963106455633,
+      "learning_rate": 5.06532082692468e-07,
+      "loss": 1.9065,
+      "step": 12155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5482354454489646,
+      "learning_rate": 5.052817578210661e-07,
+      "loss": 1.8325,
+      "step": 12156
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5298763947650685,
+      "learning_rate": 5.040329622555517e-07,
+      "loss": 1.8998,
+      "step": 12157
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5187499227749742,
+      "learning_rate": 5.027856960738875e-07,
+      "loss": 1.8549,
+      "step": 12158
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5404049222300696,
+      "learning_rate": 5.015399593539555e-07,
+      "loss": 2.0579,
+      "step": 12159
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5470673906702084,
+      "learning_rate": 5.002957521735213e-07,
+      "loss": 1.8497,
+      "step": 12160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5641814186955232,
+      "learning_rate": 4.990530746102729e-07,
+      "loss": 1.8228,
+      "step": 12161
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5319470538022956,
+      "learning_rate": 4.978119267417925e-07,
+      "loss": 1.9395,
+      "step": 12162
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5422591456966249,
+      "learning_rate": 4.965723086455709e-07,
+      "loss": 2.0241,
+      "step": 12163
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5647452775183375,
+      "learning_rate": 4.95334220399002e-07,
+      "loss": 1.8166,
+      "step": 12164
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5504909963979242,
+      "learning_rate": 4.940976620793875e-07,
+      "loss": 1.823,
+      "step": 12165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5185672040985381,
+      "learning_rate": 4.928626337639241e-07,
+      "loss": 2.0392,
+      "step": 12166
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5586598039786779,
+      "learning_rate": 4.916291355297309e-07,
+      "loss": 1.8082,
+      "step": 12167
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5266276729778374,
+      "learning_rate": 4.9039716745381e-07,
+      "loss": 1.9233,
+      "step": 12168
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5730368155960067,
+      "learning_rate": 4.891667296130802e-07,
+      "loss": 1.8899,
+      "step": 12169
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5512821852406627,
+      "learning_rate": 4.879378220843667e-07,
+      "loss": 1.8265,
+      "step": 12170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5144916815875704,
+      "learning_rate": 4.867104449443938e-07,
+      "loss": 2.0208,
+      "step": 12171
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.524171546956164,
+      "learning_rate": 4.854845982697892e-07,
+      "loss": 1.8305,
+      "step": 12172
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5479290579506326,
+      "learning_rate": 4.842602821370917e-07,
+      "loss": 1.8427,
+      "step": 12173
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4999135790266691,
+      "learning_rate": 4.830374966227346e-07,
+      "loss": 1.9056,
+      "step": 12174
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5423823762111387,
+      "learning_rate": 4.81816241803068e-07,
+      "loss": 2.0381,
+      "step": 12175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5362492005951296,
+      "learning_rate": 4.805965177543365e-07,
+      "loss": 1.8068,
+      "step": 12176
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5421371683272366,
+      "learning_rate": 4.7937832455269e-07,
+      "loss": 1.8633,
+      "step": 12177
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5657536872965847,
+      "learning_rate": 4.781616622741903e-07,
+      "loss": 2.0864,
+      "step": 12178
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5266896015864112,
+      "learning_rate": 4.769465309947957e-07,
+      "loss": 1.8672,
+      "step": 12179
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5122677548849798,
+      "learning_rate": 4.7573293079037076e-07,
+      "loss": 1.8986,
+      "step": 12180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5413571229997403,
+      "learning_rate": 4.745208617366881e-07,
+      "loss": 1.8546,
+      "step": 12181
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5809757271057946,
+      "learning_rate": 4.7331032390942333e-07,
+      "loss": 1.8566,
+      "step": 12182
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5157749460031795,
+      "learning_rate": 4.721013173841493e-07,
+      "loss": 2.0547,
+      "step": 12183
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5520335341724514,
+      "learning_rate": 4.7089384223635567e-07,
+      "loss": 1.7968,
+      "step": 12184
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.564075199584124,
+      "learning_rate": 4.696878985414266e-07,
+      "loss": 1.8278,
+      "step": 12185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5569996639026346,
+      "learning_rate": 4.6848348637465735e-07,
+      "loss": 2.0797,
+      "step": 12186
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5191816331003827,
+      "learning_rate": 4.672806058112406e-07,
+      "loss": 1.9974,
+      "step": 12187
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5765193637109733,
+      "learning_rate": 4.660792569262773e-07,
+      "loss": 1.8405,
+      "step": 12188
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5386959376417068,
+      "learning_rate": 4.6487943979477424e-07,
+      "loss": 1.8342,
+      "step": 12189
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5510767576306024,
+      "learning_rate": 4.6368115449164364e-07,
+      "loss": 1.8199,
+      "step": 12190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5587408109876917,
+      "learning_rate": 4.6248440109169235e-07,
+      "loss": 2.0511,
+      "step": 12191
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5524100859254958,
+      "learning_rate": 4.6128917966964393e-07,
+      "loss": 1.8395,
+      "step": 12192
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5241241427212794,
+      "learning_rate": 4.60095490300122e-07,
+      "loss": 1.9507,
+      "step": 12193
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.537602196274474,
+      "learning_rate": 4.5890333305764475e-07,
+      "loss": 1.913,
+      "step": 12194
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5211663766292899,
+      "learning_rate": 4.5771270801665824e-07,
+      "loss": 2.0403,
+      "step": 12195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5478911134978154,
+      "learning_rate": 4.5652361525148345e-07,
+      "loss": 1.901,
+      "step": 12196
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.548952006633524,
+      "learning_rate": 4.5533605483636666e-07,
+      "loss": 1.8567,
+      "step": 12197
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5539852273587849,
+      "learning_rate": 4.5415002684545404e-07,
+      "loss": 2.0462,
+      "step": 12198
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5013202601978779,
+      "learning_rate": 4.52965531352792e-07,
+      "loss": 1.9344,
+      "step": 12199
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5392221144311241,
+      "learning_rate": 4.517825684323324e-07,
+      "loss": 1.8405,
+      "step": 12200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5577884990522757,
+      "learning_rate": 4.5060113815793556e-07,
+      "loss": 1.8669,
+      "step": 12201
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5391962976494958,
+      "learning_rate": 4.494212406033593e-07,
+      "loss": 1.799,
+      "step": 12202
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.555755275843181,
+      "learning_rate": 4.482428758422724e-07,
+      "loss": 2.0166,
+      "step": 12203
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5518085292901486,
+      "learning_rate": 4.470660439482438e-07,
+      "loss": 1.8602,
+      "step": 12204
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5230656526445953,
+      "learning_rate": 4.4589074499474525e-07,
+      "loss": 1.9116,
+      "step": 12205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.568985452113987,
+      "learning_rate": 4.447169790551625e-07,
+      "loss": 1.8377,
+      "step": 12206
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5314224309406822,
+      "learning_rate": 4.435447462027731e-07,
+      "loss": 2.0176,
+      "step": 12207
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5590341387344524,
+      "learning_rate": 4.4237404651076565e-07,
+      "loss": 1.8626,
+      "step": 12208
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5551050586448392,
+      "learning_rate": 4.4120488005223173e-07,
+      "loss": 1.8873,
+      "step": 12209
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5603297600317924,
+      "learning_rate": 4.4003724690016566e-07,
+      "loss": 1.846,
+      "step": 12210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49990366928934526,
+      "learning_rate": 4.3887114712747035e-07,
+      "loss": 2.0786,
+      "step": 12211
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5505149694964356,
+      "learning_rate": 4.377065808069486e-07,
+      "loss": 1.8472,
+      "step": 12212
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5415482672162248,
+      "learning_rate": 4.365435480113089e-07,
+      "loss": 1.8593,
+      "step": 12213
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5403853125018282,
+      "learning_rate": 4.3538204881316824e-07,
+      "loss": 1.8149,
+      "step": 12214
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5257227583946817,
+      "learning_rate": 4.3422208328503523e-07,
+      "loss": 2.0478,
+      "step": 12215
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.559025423293466,
+      "learning_rate": 4.330636514993408e-07,
+      "loss": 1.8964,
+      "step": 12216
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5637217293375276,
+      "learning_rate": 4.3190675352840216e-07,
+      "loss": 1.8179,
+      "step": 12217
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5104559733500768,
+      "learning_rate": 4.3075138944445595e-07,
+      "loss": 1.9275,
+      "step": 12218
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5289990924999646,
+      "learning_rate": 4.2959755931963055e-07,
+      "loss": 2.0687,
+      "step": 12219
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5617302931137816,
+      "learning_rate": 4.284452632259711e-07,
+      "loss": 1.8438,
+      "step": 12220
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5614950693369076,
+      "learning_rate": 4.272945012354146e-07,
+      "loss": 1.8278,
+      "step": 12221
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5690435567735518,
+      "learning_rate": 4.2614527341980906e-07,
+      "loss": 1.8742,
+      "step": 12222
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5391416957886751,
+      "learning_rate": 4.249975798509054e-07,
+      "loss": 2.0235,
+      "step": 12223
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.49598629779309555,
+      "learning_rate": 4.238514206003602e-07,
+      "loss": 1.9298,
+      "step": 12224
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5598855292605013,
+      "learning_rate": 4.2270679573973296e-07,
+      "loss": 1.8675,
+      "step": 12225
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5377604720596268,
+      "learning_rate": 4.215637053404858e-07,
+      "loss": 1.8927,
+      "step": 12226
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5243668154673904,
+      "learning_rate": 4.2042214947398684e-07,
+      "loss": 2.081,
+      "step": 12227
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5512437084153002,
+      "learning_rate": 4.192821282115095e-07,
+      "loss": 1.9157,
+      "step": 12228
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5310905822238625,
+      "learning_rate": 4.181436416242301e-07,
+      "loss": 1.8376,
+      "step": 12229
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5016822173488662,
+      "learning_rate": 4.1700668978322533e-07,
+      "loss": 1.9272,
+      "step": 12230
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5835360070773158,
+      "learning_rate": 4.1587127275948266e-07,
+      "loss": 2.0966,
+      "step": 12231
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5636751502306065,
+      "learning_rate": 4.1473739062389273e-07,
+      "loss": 1.8681,
+      "step": 12232
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5389633755504954,
+      "learning_rate": 4.1360504344724327e-07,
+      "loss": 1.8612,
+      "step": 12233
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5591933203759875,
+      "learning_rate": 4.124742313002361e-07,
+      "loss": 1.8127,
+      "step": 12234
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.58168621871857,
+      "learning_rate": 4.113449542534703e-07,
+      "loss": 2.0429,
+      "step": 12235
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5028949709261705,
+      "learning_rate": 4.1021721237745337e-07,
+      "loss": 1.9433,
+      "step": 12236
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5713537806442365,
+      "learning_rate": 4.0909100574259284e-07,
+      "loss": 1.8276,
+      "step": 12237
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5685058728049907,
+      "learning_rate": 4.0796633441920196e-07,
+      "loss": 1.8422,
+      "step": 12238
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5159785340553931,
+      "learning_rate": 4.0684319847749954e-07,
+      "loss": 2.0439,
+      "step": 12239
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5789110531664214,
+      "learning_rate": 4.057215979876072e-07,
+      "loss": 1.8515,
+      "step": 12240
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5468434558351187,
+      "learning_rate": 4.0460153301954964e-07,
+      "loss": 1.8555,
+      "step": 12241
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.520424528174259,
+      "learning_rate": 4.034830036432652e-07,
+      "loss": 1.8924,
+      "step": 12242
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5177848054482475,
+      "learning_rate": 4.0236600992857585e-07,
+      "loss": 1.993,
+      "step": 12243
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5410933823546056,
+      "learning_rate": 4.012505519452259e-07,
+      "loss": 1.86,
+      "step": 12244
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5461110229319871,
+      "learning_rate": 4.0013662976286226e-07,
+      "loss": 1.8738,
+      "step": 12245
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5509420023329147,
+      "learning_rate": 3.9902424345102384e-07,
+      "loss": 1.8659,
+      "step": 12246
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5303491835729459,
+      "learning_rate": 3.979133930791662e-07,
+      "loss": 1.9885,
+      "step": 12247
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.554539817806879,
+      "learning_rate": 3.9680407871664496e-07,
+      "loss": 1.8653,
+      "step": 12248
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5142982438360112,
+      "learning_rate": 3.956963004327158e-07,
+      "loss": 1.9383,
+      "step": 12249
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5693899343743858,
+      "learning_rate": 3.945900582965456e-07,
+      "loss": 1.8727,
+      "step": 12250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5378790521144408,
+      "learning_rate": 3.9348535237719863e-07,
+      "loss": 2.0658,
+      "step": 12251
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.54164503752832,
+      "learning_rate": 3.923821827436447e-07,
+      "loss": 1.8345,
+      "step": 12252
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5369444453511621,
+      "learning_rate": 3.912805494647648e-07,
+      "loss": 1.817,
+      "step": 12253
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5482704559172112,
+      "learning_rate": 3.9018045260933445e-07,
+      "loss": 1.8346,
+      "step": 12254
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.529781017295403,
+      "learning_rate": 3.8908189224603763e-07,
+      "loss": 2.069,
+      "step": 12255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5509203384048785,
+      "learning_rate": 3.8798486844346393e-07,
+      "loss": 1.8348,
+      "step": 12256
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5495009492182514,
+      "learning_rate": 3.868893812701002e-07,
+      "loss": 1.911,
+      "step": 12257
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5653906395793388,
+      "learning_rate": 3.857954307943501e-07,
+      "loss": 1.9123,
+      "step": 12258
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5340609782581943,
+      "learning_rate": 3.847030170845062e-07,
+      "loss": 2.0473,
+      "step": 12259
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5434235608959901,
+      "learning_rate": 3.83612140208775e-07,
+      "loss": 1.8683,
+      "step": 12260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5271745985071611,
+      "learning_rate": 3.82522800235266e-07,
+      "loss": 1.9213,
+      "step": 12261
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5295137234310723,
+      "learning_rate": 3.8143499723198863e-07,
+      "loss": 1.8267,
+      "step": 12262
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5313218753984338,
+      "learning_rate": 3.803487312668608e-07,
+      "loss": 2.0484,
+      "step": 12263
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5771844660821385,
+      "learning_rate": 3.7926400240770044e-07,
+      "loss": 1.882,
+      "step": 12264
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5523301716143051,
+      "learning_rate": 3.78180810722234e-07,
+      "loss": 1.8582,
+      "step": 12265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5394392939218419,
+      "learning_rate": 3.7709915627808787e-07,
+      "loss": 1.8433,
+      "step": 12266
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4899183774861811,
+      "learning_rate": 3.7601903914279425e-07,
+      "loss": 2.0708,
+      "step": 12267
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5519238246461585,
+      "learning_rate": 3.7494045938379075e-07,
+      "loss": 1.8505,
+      "step": 12268
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5627441240378865,
+      "learning_rate": 3.7386341706841797e-07,
+      "loss": 1.877,
+      "step": 12269
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5594633713796201,
+      "learning_rate": 3.727879122639166e-07,
+      "loss": 1.857,
+      "step": 12270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5301094687640432,
+      "learning_rate": 3.717139450374385e-07,
+      "loss": 2.0582,
+      "step": 12271
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5698654002278227,
+      "learning_rate": 3.706415154560328e-07,
+      "loss": 1.8756,
+      "step": 12272
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5221582840636342,
+      "learning_rate": 3.695706235866597e-07,
+      "loss": 1.9205,
+      "step": 12273
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5465714589680125,
+      "learning_rate": 3.6850126949617427e-07,
+      "loss": 1.834,
+      "step": 12274
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5080209074639347,
+      "learning_rate": 3.6743345325134515e-07,
+      "loss": 2.0151,
+      "step": 12275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5659795046138616,
+      "learning_rate": 3.6636717491883856e-07,
+      "loss": 1.8135,
+      "step": 12276
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5668541237269371,
+      "learning_rate": 3.6530243456522615e-07,
+      "loss": 1.8756,
+      "step": 12277
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5418833038128792,
+      "learning_rate": 3.6423923225698265e-07,
+      "loss": 1.8742,
+      "step": 12278
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5194763297402295,
+      "learning_rate": 3.6317756806049374e-07,
+      "loss": 2.085,
+      "step": 12279
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4996968547903169,
+      "learning_rate": 3.6211744204203703e-07,
+      "loss": 1.9095,
+      "step": 12280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5346625108320361,
+      "learning_rate": 3.610588542678039e-07,
+      "loss": 1.7892,
+      "step": 12281
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5359006757533082,
+      "learning_rate": 3.600018048038861e-07,
+      "loss": 1.8787,
+      "step": 12282
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5397848866104188,
+      "learning_rate": 3.5894629371628073e-07,
+      "loss": 2.0539,
+      "step": 12283
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5493073636740198,
+      "learning_rate": 3.5789232107088786e-07,
+      "loss": 1.8393,
+      "step": 12284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5549291537615051,
+      "learning_rate": 3.568398869335049e-07,
+      "loss": 1.8239,
+      "step": 12285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5191194307834666,
+      "learning_rate": 3.557889913698459e-07,
+      "loss": 1.9295,
+      "step": 12286
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.550780643054991,
+      "learning_rate": 3.547396344455223e-07,
+      "loss": 2.106,
+      "step": 12287
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.532555199114257,
+      "learning_rate": 3.536918162260483e-07,
+      "loss": 1.8294,
+      "step": 12288
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5536490952472499,
+      "learning_rate": 3.5264553677684386e-07,
+      "loss": 1.883,
+      "step": 12289
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5538876429713857,
+      "learning_rate": 3.516007961632317e-07,
+      "loss": 1.8884,
+      "step": 12290
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5402184041294837,
+      "learning_rate": 3.5055759445044025e-07,
+      "loss": 2.0005,
+      "step": 12291
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4989007938093657,
+      "learning_rate": 3.4951593170360066e-07,
+      "loss": 1.9393,
+      "step": 12292
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5640278942238962,
+      "learning_rate": 3.4847580798774983e-07,
+      "loss": 1.8135,
+      "step": 12293
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.526503125647013,
+      "learning_rate": 3.474372233678219e-07,
+      "loss": 1.8382,
+      "step": 12294
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5370124172405291,
+      "learning_rate": 3.4640017790866506e-07,
+      "loss": 2.0725,
+      "step": 12295
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5458650143859223,
+      "learning_rate": 3.453646716750247e-07,
+      "loss": 1.8321,
+      "step": 12296
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5810408513485574,
+      "learning_rate": 3.4433070473155194e-07,
+      "loss": 1.8435,
+      "step": 12297
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.508044175636359,
+      "learning_rate": 3.4329827714280336e-07,
+      "loss": 1.9192,
+      "step": 12298
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5245742263355347,
+      "learning_rate": 3.422673889732303e-07,
+      "loss": 2.0712,
+      "step": 12299
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5554381885030606,
+      "learning_rate": 3.412380402872062e-07,
+      "loss": 1.8641,
+      "step": 12300
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5508865719830663,
+      "learning_rate": 3.40210231148988e-07,
+      "loss": 1.7974,
+      "step": 12301
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5603484641934283,
+      "learning_rate": 3.3918396162275214e-07,
+      "loss": 1.8769,
+      "step": 12302
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5285344672910878,
+      "learning_rate": 3.3815923177256956e-07,
+      "loss": 2.1095,
+      "step": 12303
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5047771148356289,
+      "learning_rate": 3.3713604166241973e-07,
+      "loss": 1.922,
+      "step": 12304
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5493741563796235,
+      "learning_rate": 3.361143913561848e-07,
+      "loss": 1.8481,
+      "step": 12305
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5149484210637101,
+      "learning_rate": 3.3509428091764996e-07,
+      "loss": 1.8607,
+      "step": 12306
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5254544785840605,
+      "learning_rate": 3.3407571041050577e-07,
+      "loss": 2.0649,
+      "step": 12307
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5284349277228865,
+      "learning_rate": 3.3305867989834596e-07,
+      "loss": 1.8172,
+      "step": 12308
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5528527644226342,
+      "learning_rate": 3.3204318944466406e-07,
+      "loss": 1.806,
+      "step": 12309
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5477552595381143,
+      "learning_rate": 3.310292391128678e-07,
+      "loss": 1.9241,
+      "step": 12310
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4989331953621159,
+      "learning_rate": 3.300168289662564e-07,
+      "loss": 2.0885,
+      "step": 12311
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.551681746850437,
+      "learning_rate": 3.2900595906804053e-07,
+      "loss": 1.8621,
+      "step": 12312
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5506843293759881,
+      "learning_rate": 3.2799662948133635e-07,
+      "loss": 1.8343,
+      "step": 12313
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5886823406736493,
+      "learning_rate": 3.2698884026915734e-07,
+      "loss": 1.876,
+      "step": 12314
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5604859639607,
+      "learning_rate": 3.2598259149442254e-07,
+      "loss": 2.0219,
+      "step": 12315
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5779869988479575,
+      "learning_rate": 3.2497788321995957e-07,
+      "loss": 1.8455,
+      "step": 12316
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.502363936274882,
+      "learning_rate": 3.239747155084932e-07,
+      "loss": 1.9543,
+      "step": 12317
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.553722107028103,
+      "learning_rate": 3.229730884226567e-07,
+      "loss": 1.8599,
+      "step": 12318
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5629480882485736,
+      "learning_rate": 3.219730020249889e-07,
+      "loss": 2.0727,
+      "step": 12319
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.587033926305675,
+      "learning_rate": 3.2097445637792324e-07,
+      "loss": 1.8629,
+      "step": 12320
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5472509987159287,
+      "learning_rate": 3.1997745154380977e-07,
+      "loss": 1.8708,
+      "step": 12321
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5751421174249743,
+      "learning_rate": 3.189819875848876e-07,
+      "loss": 1.8544,
+      "step": 12322
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5402861685040083,
+      "learning_rate": 3.179880645633182e-07,
+      "loss": 1.9086,
+      "step": 12323
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5241695954677373,
+      "learning_rate": 3.169956825411435e-07,
+      "loss": 2.0229,
+      "step": 12324
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5430333294428568,
+      "learning_rate": 3.160048415803335e-07,
+      "loss": 1.8735,
+      "step": 12325
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5504624968284116,
+      "learning_rate": 3.150155417427442e-07,
+      "loss": 1.8777,
+      "step": 12326
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5337205091858703,
+      "learning_rate": 3.140277830901428e-07,
+      "loss": 2.0682,
+      "step": 12327
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5847126676620163,
+      "learning_rate": 3.1304156568419675e-07,
+      "loss": 1.8632,
+      "step": 12328
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5057182113660615,
+      "learning_rate": 3.1205688958648725e-07,
+      "loss": 1.9295,
+      "step": 12329
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5211825750872021,
+      "learning_rate": 3.110737548584847e-07,
+      "loss": 1.8267,
+      "step": 12330
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5182513614885402,
+      "learning_rate": 3.1009216156157315e-07,
+      "loss": 2.0463,
+      "step": 12331
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5537858737262462,
+      "learning_rate": 3.091121097570343e-07,
+      "loss": 1.85,
+      "step": 12332
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5690584144719656,
+      "learning_rate": 3.08133599506058e-07,
+      "loss": 1.7827,
+      "step": 12333
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5473953630463927,
+      "learning_rate": 3.0715663086974265e-07,
+      "loss": 1.8585,
+      "step": 12334
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5607094569621096,
+      "learning_rate": 3.0618120390907555e-07,
+      "loss": 2.0552,
+      "step": 12335
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4985698285304345,
+      "learning_rate": 3.0520731868496355e-07,
+      "loss": 1.8852,
+      "step": 12336
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.579430602208256,
+      "learning_rate": 3.04234975258208e-07,
+      "loss": 1.8359,
+      "step": 12337
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5783297342988696,
+      "learning_rate": 3.032641736895103e-07,
+      "loss": 1.9002,
+      "step": 12338
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5155027762659877,
+      "learning_rate": 3.022949140394915e-07,
+      "loss": 1.9951,
+      "step": 12339
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5263107517911539,
+      "learning_rate": 3.0132719636866146e-07,
+      "loss": 1.8384,
+      "step": 12340
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5398907210331203,
+      "learning_rate": 3.0036102073743855e-07,
+      "loss": 1.8755,
+      "step": 12341
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5041202731640542,
+      "learning_rate": 2.993963872061467e-07,
+      "loss": 1.9217,
+      "step": 12342
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5451679361804138,
+      "learning_rate": 2.9843329583501e-07,
+      "loss": 1.848,
+      "step": 12343
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5274908179825102,
+      "learning_rate": 2.974717466841581e-07,
+      "loss": 2.0717,
+      "step": 12344
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5598256165934916,
+      "learning_rate": 2.965117398136291e-07,
+      "loss": 1.8405,
+      "step": 12345
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5737466474565918,
+      "learning_rate": 2.955532752833501e-07,
+      "loss": 1.8285,
+      "step": 12346
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5370546368731641,
+      "learning_rate": 2.945963531531759e-07,
+      "loss": 2.0828,
+      "step": 12347
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5228934012992893,
+      "learning_rate": 2.936409734828394e-07,
+      "loss": 1.8933,
+      "step": 12348
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5694115245380343,
+      "learning_rate": 2.9268713633199276e-07,
+      "loss": 1.895,
+      "step": 12349
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.577088092875283,
+      "learning_rate": 2.9173484176019393e-07,
+      "loss": 1.8677,
+      "step": 12350
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5231379914295355,
+      "learning_rate": 2.907840898268871e-07,
+      "loss": 2.0329,
+      "step": 12351
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5520524327905233,
+      "learning_rate": 2.8983488059144135e-07,
+      "loss": 1.803,
+      "step": 12352
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5858015789056435,
+      "learning_rate": 2.8888721411311494e-07,
+      "loss": 1.8046,
+      "step": 12353
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5491024458431032,
+      "learning_rate": 2.879410904510743e-07,
+      "loss": 1.9226,
+      "step": 12354
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5810116934631,
+      "learning_rate": 2.8699650966439174e-07,
+      "loss": 1.8539,
+      "step": 12355
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5426844998355125,
+      "learning_rate": 2.8605347181203946e-07,
+      "loss": 2.0453,
+      "step": 12356
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5569202869658801,
+      "learning_rate": 2.8511197695289815e-07,
+      "loss": 1.8935,
+      "step": 12357
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5423199332722081,
+      "learning_rate": 2.841720251457458e-07,
+      "loss": 1.836,
+      "step": 12358
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5448797739159761,
+      "learning_rate": 2.8323361644926883e-07,
+      "loss": 2.1123,
+      "step": 12359
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4761314574841633,
+      "learning_rate": 2.822967509220592e-07,
+      "loss": 1.9265,
+      "step": 12360
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5420007857759062,
+      "learning_rate": 2.813614286226035e-07,
+      "loss": 1.8603,
+      "step": 12361
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5652302429272086,
+      "learning_rate": 2.804276496093022e-07,
+      "loss": 1.8996,
+      "step": 12362
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5186137867521708,
+      "learning_rate": 2.794954139404532e-07,
+      "loss": 2.0704,
+      "step": 12363
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5674649644703988,
+      "learning_rate": 2.7856472167425707e-07,
+      "loss": 1.8336,
+      "step": 12364
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5600444592478302,
+      "learning_rate": 2.776355728688257e-07,
+      "loss": 1.8927,
+      "step": 12365
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5541777553761364,
+      "learning_rate": 2.767079675821682e-07,
+      "loss": 1.8561,
+      "step": 12366
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5018088116267028,
+      "learning_rate": 2.757819058721967e-07,
+      "loss": 1.9305,
+      "step": 12367
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.528825069283403,
+      "learning_rate": 2.7485738779673143e-07,
+      "loss": 2.0703,
+      "step": 12368
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5407507583267065,
+      "learning_rate": 2.7393441341349304e-07,
+      "loss": 1.8262,
+      "step": 12369
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5603465324800216,
+      "learning_rate": 2.7301298278010755e-07,
+      "loss": 1.8728,
+      "step": 12370
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5182275020324781,
+      "learning_rate": 2.720930959540985e-07,
+      "loss": 2.0472,
+      "step": 12371
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5566520161880496,
+      "learning_rate": 2.7117475299290594e-07,
+      "loss": 1.8426,
+      "step": 12372
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5137081713735653,
+      "learning_rate": 2.702579539538619e-07,
+      "loss": 1.9067,
+      "step": 12373
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5379390680171702,
+      "learning_rate": 2.6934269889420384e-07,
+      "loss": 1.8573,
+      "step": 12374
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5805139441556773,
+      "learning_rate": 2.6842898787107773e-07,
+      "loss": 1.8504,
+      "step": 12375
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5328653307281224,
+      "learning_rate": 2.675168209415324e-07,
+      "loss": 2.0335,
+      "step": 12376
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.553313503276388,
+      "learning_rate": 2.6660619816251397e-07,
+      "loss": 1.8511,
+      "step": 12377
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5497424013014071,
+      "learning_rate": 2.6569711959087693e-07,
+      "loss": 1.8623,
+      "step": 12378
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5449162646944041,
+      "learning_rate": 2.647895852833815e-07,
+      "loss": 1.9373,
+      "step": 12379
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5414848391700594,
+      "learning_rate": 2.6388359529668506e-07,
+      "loss": 2.0285,
+      "step": 12380
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.544323035405987,
+      "learning_rate": 2.6297914968735625e-07,
+      "loss": 1.7962,
+      "step": 12381
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5334283845142075,
+      "learning_rate": 2.6207624851185833e-07,
+      "loss": 1.8634,
+      "step": 12382
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5460798667873793,
+      "learning_rate": 2.6117489182656827e-07,
+      "loss": 2.0472,
+      "step": 12383
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5602045842838083,
+      "learning_rate": 2.6027507968776067e-07,
+      "loss": 1.8211,
+      "step": 12384
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5066887380268822,
+      "learning_rate": 2.593768121516099e-07,
+      "loss": 1.9263,
+      "step": 12385
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5424878535430804,
+      "learning_rate": 2.584800892742073e-07,
+      "loss": 1.8474,
+      "step": 12386
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.536058109031376,
+      "learning_rate": 2.575849111115275e-07,
+      "loss": 1.8758,
+      "step": 12387
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5500069813474334,
+      "learning_rate": 2.5669127771946743e-07,
+      "loss": 2.0468,
+      "step": 12388
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.542400335799354,
+      "learning_rate": 2.557991891538214e-07,
+      "loss": 1.8158,
+      "step": 12389
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5089344621594065,
+      "learning_rate": 2.549086454702837e-07,
+      "loss": 1.8096,
+      "step": 12390
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5108176060258455,
+      "learning_rate": 2.5401964672445154e-07,
+      "loss": 1.895,
+      "step": 12391
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5298504557373275,
+      "learning_rate": 2.531321929718333e-07,
+      "loss": 2.0107,
+      "step": 12392
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5352394143456641,
+      "learning_rate": 2.522462842678347e-07,
+      "loss": 1.8577,
+      "step": 12393
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5390564275935895,
+      "learning_rate": 2.513619206677698e-07,
+      "loss": 1.8268,
+      "step": 12394
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5288484907912087,
+      "learning_rate": 2.504791022268471e-07,
+      "loss": 2.093,
+      "step": 12395
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5619047506054939,
+      "learning_rate": 2.495978290001866e-07,
+      "loss": 1.8471,
+      "step": 12396
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5651827248256176,
+      "learning_rate": 2.4871810104281356e-07,
+      "loss": 1.925,
+      "step": 12397
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4992166862084879,
+      "learning_rate": 2.4783991840964805e-07,
+      "loss": 1.9487,
+      "step": 12398
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5230245578010629,
+      "learning_rate": 2.469632811555239e-07,
+      "loss": 1.8326,
+      "step": 12399
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5348566496717568,
+      "learning_rate": 2.4608818933516686e-07,
+      "loss": 2.0406,
+      "step": 12400
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5240812360986895,
+      "learning_rate": 2.452146430032165e-07,
+      "loss": 1.8626,
+      "step": 12401
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5448488850190394,
+      "learning_rate": 2.4434264221421255e-07,
+      "loss": 1.8479,
+      "step": 12402
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.528819997492905,
+      "learning_rate": 2.434721870225948e-07,
+      "loss": 2.0754,
+      "step": 12403
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.521779962255133,
+      "learning_rate": 2.4260327748271136e-07,
+      "loss": 1.9587,
+      "step": 12404
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5457535017597165,
+      "learning_rate": 2.4173591364881066e-07,
+      "loss": 1.8263,
+      "step": 12405
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.556643670626656,
+      "learning_rate": 2.408700955750492e-07,
+      "loss": 1.8918,
+      "step": 12406
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5809544877309927,
+      "learning_rate": 2.400058233154784e-07,
+      "loss": 1.8611,
+      "step": 12407
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5382186129127595,
+      "learning_rate": 2.3914309692406045e-07,
+      "loss": 2.0274,
+      "step": 12408
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5224938864866695,
+      "learning_rate": 2.382819164546579e-07,
+      "loss": 1.8325,
+      "step": 12409
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5420614482794511,
+      "learning_rate": 2.3742228196104166e-07,
+      "loss": 1.91,
+      "step": 12410
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5381463122715112,
+      "learning_rate": 2.3656419349687986e-07,
+      "loss": 1.8296,
+      "step": 12411
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5261606163973161,
+      "learning_rate": 2.3570765111574354e-07,
+      "loss": 2.0603,
+      "step": 12412
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.557720406692031,
+      "learning_rate": 2.3485265487111497e-07,
+      "loss": 1.8317,
+      "step": 12413
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5059603519031224,
+      "learning_rate": 2.3399920481637083e-07,
+      "loss": 1.7887,
+      "step": 12414
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5556067719900546,
+      "learning_rate": 2.331473010047991e-07,
+      "loss": 2.0533,
+      "step": 12415
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5352671376490844,
+      "learning_rate": 2.3229694348958497e-07,
+      "loss": 1.9135,
+      "step": 12416
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5460479944231387,
+      "learning_rate": 2.314481323238249e-07,
+      "loss": 1.8527,
+      "step": 12417
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5762529193680177,
+      "learning_rate": 2.3060086756050426e-07,
+      "loss": 1.8838,
+      "step": 12418
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5794252363952456,
+      "learning_rate": 2.2975514925253072e-07,
+      "loss": 1.8713,
+      "step": 12419
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5316220642173061,
+      "learning_rate": 2.2891097745270097e-07,
+      "loss": 2.0842,
+      "step": 12420
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5396891263817313,
+      "learning_rate": 2.2806835221372003e-07,
+      "loss": 1.8976,
+      "step": 12421
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5018710322250157,
+      "learning_rate": 2.2722727358819584e-07,
+      "loss": 1.9232,
+      "step": 12422
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5320132030177137,
+      "learning_rate": 2.2638774162864751e-07,
+      "loss": 1.8492,
+      "step": 12423
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5182018472798056,
+      "learning_rate": 2.2554975638747754e-07,
+      "loss": 2.0256,
+      "step": 12424
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.549548814503514,
+      "learning_rate": 2.2471331791701912e-07,
+      "loss": 1.8569,
+      "step": 12425
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5434820488029828,
+      "learning_rate": 2.2387842626948319e-07,
+      "loss": 1.8361,
+      "step": 12426
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.554275302082161,
+      "learning_rate": 2.230450814970031e-07,
+      "loss": 1.8847,
+      "step": 12427
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5276322185752452,
+      "learning_rate": 2.2221328365160386e-07,
+      "loss": 2.0158,
+      "step": 12428
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5062758494390275,
+      "learning_rate": 2.2138303278521622e-07,
+      "loss": 1.9021,
+      "step": 12429
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5523816308607804,
+      "learning_rate": 2.20554328949682e-07,
+      "loss": 1.8757,
+      "step": 12430
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5417471527920115,
+      "learning_rate": 2.1972717219673754e-07,
+      "loss": 1.8609,
+      "step": 12431
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5010203023610377,
+      "learning_rate": 2.1890156257802495e-07,
+      "loss": 2.0392,
+      "step": 12432
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5671128469426724,
+      "learning_rate": 2.1807750014509186e-07,
+      "loss": 1.8746,
+      "step": 12433
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5450920054437463,
+      "learning_rate": 2.1725498494938602e-07,
+      "loss": 1.8084,
+      "step": 12434
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.506297053819699,
+      "learning_rate": 2.1643401704226075e-07,
+      "loss": 1.9425,
+      "step": 12435
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5185994738775137,
+      "learning_rate": 2.1561459647497783e-07,
+      "loss": 2.0042,
+      "step": 12436
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.548419339855689,
+      "learning_rate": 2.1479672329868805e-07,
+      "loss": 1.8826,
+      "step": 12437
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5537267920297948,
+      "learning_rate": 2.1398039756445886e-07,
+      "loss": 1.805,
+      "step": 12438
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5423706764836805,
+      "learning_rate": 2.1316561932325785e-07,
+      "loss": 1.8168,
+      "step": 12439
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.513332141458772,
+      "learning_rate": 2.1235238862595541e-07,
+      "loss": 2.0412,
+      "step": 12440
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5121199465188533,
+      "learning_rate": 2.1154070552332206e-07,
+      "loss": 1.9127,
+      "step": 12441
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.545177019456379,
+      "learning_rate": 2.1073057006603669e-07,
+      "loss": 1.8367,
+      "step": 12442
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5396070560126897,
+      "learning_rate": 2.0992198230467829e-07,
+      "loss": 1.8216,
+      "step": 12443
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5089123989148564,
+      "learning_rate": 2.091149422897315e-07,
+      "loss": 2.0106,
+      "step": 12444
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5256701337816259,
+      "learning_rate": 2.08309450071581e-07,
+      "loss": 1.8689,
+      "step": 12445
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5253200391424355,
+      "learning_rate": 2.0750550570051984e-07,
+      "loss": 1.8227,
+      "step": 12446
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5052653570887029,
+      "learning_rate": 2.067031092267385e-07,
+      "loss": 1.9541,
+      "step": 12447
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5224144641889229,
+      "learning_rate": 2.059022607003358e-07,
+      "loss": 2.0869,
+      "step": 12448
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5717522517304106,
+      "learning_rate": 2.0510296017131336e-07,
+      "loss": 1.8272,
+      "step": 12449
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5423504368693773,
+      "learning_rate": 2.0430520768957295e-07,
+      "loss": 1.8448,
+      "step": 12450
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5559752937273479,
+      "learning_rate": 2.0350900330491917e-07,
+      "loss": 1.8296,
+      "step": 12451
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5219547363193888,
+      "learning_rate": 2.0271434706706782e-07,
+      "loss": 2.0336,
+      "step": 12452
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5079057157754275,
+      "learning_rate": 2.0192123902562642e-07,
+      "loss": 1.9552,
+      "step": 12453
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5485193374667694,
+      "learning_rate": 2.011296792301165e-07,
+      "loss": 1.8395,
+      "step": 12454
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5293042047432247,
+      "learning_rate": 2.003396677299596e-07,
+      "loss": 1.8367,
+      "step": 12455
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5118128830368989,
+      "learning_rate": 1.9955120457447462e-07,
+      "loss": 2.0296,
+      "step": 12456
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5561386644196412,
+      "learning_rate": 1.9876428981288887e-07,
+      "loss": 1.87,
+      "step": 12457
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.546218140041735,
+      "learning_rate": 1.97978923494338e-07,
+      "loss": 1.8283,
+      "step": 12458
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.534924277059728,
+      "learning_rate": 1.9719510566784948e-07,
+      "loss": 1.8438,
+      "step": 12459
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5437760333591966,
+      "learning_rate": 1.9641283638236474e-07,
+      "loss": 2.0904,
+      "step": 12460
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5519887971950597,
+      "learning_rate": 1.9563211568672247e-07,
+      "loss": 1.8316,
+      "step": 12461
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5343767342809388,
+      "learning_rate": 1.948529436296642e-07,
+      "loss": 1.8202,
+      "step": 12462
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5475635817389152,
+      "learning_rate": 1.9407532025983997e-07,
+      "loss": 1.85,
+      "step": 12463
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5381371053068487,
+      "learning_rate": 1.9329924562579705e-07,
+      "loss": 2.0429,
+      "step": 12464
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5456808466454736,
+      "learning_rate": 1.9252471977599385e-07,
+      "loss": 1.8589,
+      "step": 12465
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5102659653264047,
+      "learning_rate": 1.917517427587834e-07,
+      "loss": 1.9421,
+      "step": 12466
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.568506163859508,
+      "learning_rate": 1.9098031462242705e-07,
+      "loss": 1.8757,
+      "step": 12467
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.533655515387091,
+      "learning_rate": 1.9021043541508354e-07,
+      "loss": 1.9989,
+      "step": 12468
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5743923118102602,
+      "learning_rate": 1.8944210518482552e-07,
+      "loss": 1.8672,
+      "step": 12469
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5570272160654681,
+      "learning_rate": 1.8867532397962573e-07,
+      "loss": 1.8704,
+      "step": 12470
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.558709378767941,
+      "learning_rate": 1.8791009184734588e-07,
+      "loss": 1.8772,
+      "step": 12471
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5356752272922317,
+      "learning_rate": 1.871464088357755e-07,
+      "loss": 2.1065,
+      "step": 12472
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5583683539766308,
+      "learning_rate": 1.863842749925848e-07,
+      "loss": 1.8405,
+      "step": 12473
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.533271416335019,
+      "learning_rate": 1.8562369036536076e-07,
+      "loss": 1.8053,
+      "step": 12474
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5319187372987666,
+      "learning_rate": 1.848646550015931e-07,
+      "loss": 1.818,
+      "step": 12475
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5440419798774808,
+      "learning_rate": 1.8410716894866342e-07,
+      "loss": 2.0316,
+      "step": 12476
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5462511133741799,
+      "learning_rate": 1.8335123225387275e-07,
+      "loss": 1.8281,
+      "step": 12477
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4862759817667618,
+      "learning_rate": 1.8259684496441387e-07,
+      "loss": 1.8953,
+      "step": 12478
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5691828856888608,
+      "learning_rate": 1.8184400712738525e-07,
+      "loss": 1.8863,
+      "step": 12479
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.530777394843586,
+      "learning_rate": 1.8109271878979372e-07,
+      "loss": 2.0429,
+      "step": 12480
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5742917196912182,
+      "learning_rate": 1.8034297999854066e-07,
+      "loss": 1.7977,
+      "step": 12481
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.537871279138477,
+      "learning_rate": 1.7959479080043305e-07,
+      "loss": 1.8537,
+      "step": 12482
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5769397914140664,
+      "learning_rate": 1.788481512421919e-07,
+      "loss": 1.8751,
+      "step": 12483
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5067905128618029,
+      "learning_rate": 1.7810306137042987e-07,
+      "loss": 2.1057,
+      "step": 12484
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5509915554759925,
+      "learning_rate": 1.7735952123165978e-07,
+      "loss": 1.9035,
+      "step": 12485
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5519980230005824,
+      "learning_rate": 1.7661753087231115e-07,
+      "loss": 1.8594,
+      "step": 12486
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5292472963154756,
+      "learning_rate": 1.7587709033870803e-07,
+      "loss": 1.8472,
+      "step": 12487
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5157220036224456,
+      "learning_rate": 1.7513819967707735e-07,
+      "loss": 2.0532,
+      "step": 12488
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5377902554672803,
+      "learning_rate": 1.7440085893355163e-07,
+      "loss": 1.8323,
+      "step": 12489
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5715956149480098,
+      "learning_rate": 1.7366506815416627e-07,
+      "loss": 1.7942,
+      "step": 12490
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4998533636995358,
+      "learning_rate": 1.7293082738486232e-07,
+      "loss": 1.8844,
+      "step": 12491
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5110151907441478,
+      "learning_rate": 1.7219813667147532e-07,
+      "loss": 2.0146,
+      "step": 12492
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5357589423193935,
+      "learning_rate": 1.714669960597548e-07,
+      "loss": 1.8153,
+      "step": 12493
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.540694750369853,
+      "learning_rate": 1.7073740559535033e-07,
+      "loss": 1.864,
+      "step": 12494
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5152038171756893,
+      "learning_rate": 1.7000936532380884e-07,
+      "loss": 1.8593,
+      "step": 12495
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5327718587610514,
+      "learning_rate": 1.6928287529058563e-07,
+      "loss": 2.0654,
+      "step": 12496
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5011354476819369,
+      "learning_rate": 1.685579355410416e-07,
+      "loss": 1.898,
+      "step": 12497
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5593724148558932,
+      "learning_rate": 1.6783454612043502e-07,
+      "loss": 1.8614,
+      "step": 12498
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5398587042589721,
+      "learning_rate": 1.671127070739298e-07,
+      "loss": 1.8156,
+      "step": 12499
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5092281612917596,
+      "learning_rate": 1.6639241844659537e-07,
+      "loss": 2.0143,
+      "step": 12500
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5489207537679166,
+      "learning_rate": 1.6567368028340135e-07,
+      "loss": 1.8882,
+      "step": 12501
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5502652506950222,
+      "learning_rate": 1.6495649262922297e-07,
+      "loss": 1.8519,
+      "step": 12502
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5101195734327651,
+      "learning_rate": 1.6424085552883274e-07,
+      "loss": 1.8702,
+      "step": 12503
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5190898083988714,
+      "learning_rate": 1.635267690269171e-07,
+      "loss": 2.0807,
+      "step": 12504
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5240690807551348,
+      "learning_rate": 1.6281423316805434e-07,
+      "loss": 1.8883,
+      "step": 12505
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5585956554202949,
+      "learning_rate": 1.6210324799673382e-07,
+      "loss": 1.8834,
+      "step": 12506
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.540225742463416,
+      "learning_rate": 1.613938135573395e-07,
+      "loss": 1.8514,
+      "step": 12507
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5396760004924651,
+      "learning_rate": 1.6068592989417485e-07,
+      "loss": 2.0752,
+      "step": 12508
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5030728033992318,
+      "learning_rate": 1.5997959705142673e-07,
+      "loss": 1.9776,
+      "step": 12509
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5551317564585917,
+      "learning_rate": 1.5927481507319875e-07,
+      "loss": 1.8449,
+      "step": 12510
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.545625308832388,
+      "learning_rate": 1.585715840034918e-07,
+      "loss": 1.8684,
+      "step": 12511
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5301462697166901,
+      "learning_rate": 1.5786990388621247e-07,
+      "loss": 2.0515,
+      "step": 12512
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5312824407391193,
+      "learning_rate": 1.5716977476516737e-07,
+      "loss": 1.8124,
+      "step": 12513
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5524569144586571,
+      "learning_rate": 1.5647119668407429e-07,
+      "loss": 1.8629,
+      "step": 12514
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5511747793035497,
+      "learning_rate": 1.5577416968654e-07,
+      "loss": 1.8754,
+      "step": 12515
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.512940529710257,
+      "learning_rate": 1.550786938160881e-07,
+      "loss": 2.0836,
+      "step": 12516
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5852547267474459,
+      "learning_rate": 1.543847691161393e-07,
+      "loss": 1.836,
+      "step": 12517
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5355785276558362,
+      "learning_rate": 1.5369239563001736e-07,
+      "loss": 1.8866,
+      "step": 12518
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5487670181302974,
+      "learning_rate": 1.5300157340095156e-07,
+      "loss": 1.825,
+      "step": 12519
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5091948934643001,
+      "learning_rate": 1.523123024720713e-07,
+      "loss": 2.055,
+      "step": 12520
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5175750687676114,
+      "learning_rate": 1.5162458288640878e-07,
+      "loss": 1.829,
+      "step": 12521
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5090117129478666,
+      "learning_rate": 1.5093841468690472e-07,
+      "loss": 1.9673,
+      "step": 12522
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5327583228384077,
+      "learning_rate": 1.5025379791639705e-07,
+      "loss": 1.8166,
+      "step": 12523
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5393044911573703,
+      "learning_rate": 1.4957073261763212e-07,
+      "loss": 2.099,
+      "step": 12524
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5356423405696384,
+      "learning_rate": 1.4888921883325357e-07,
+      "loss": 1.8642,
+      "step": 12525
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5392096903622182,
+      "learning_rate": 1.4820925660581075e-07,
+      "loss": 1.8182,
+      "step": 12526
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5359680310615048,
+      "learning_rate": 1.4753084597776133e-07,
+      "loss": 1.8436,
+      "step": 12527
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5153624859697878,
+      "learning_rate": 1.4685398699145482e-07,
+      "loss": 2.064,
+      "step": 12528
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5812000556549394,
+      "learning_rate": 1.4617867968915178e-07,
+      "loss": 1.8328,
+      "step": 12529
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5578181879258931,
+      "learning_rate": 1.4550492411301851e-07,
+      "loss": 1.8122,
+      "step": 12530
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5374745210501191,
+      "learning_rate": 1.4483272030511585e-07,
+      "loss": 1.8452,
+      "step": 12531
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5299271136333544,
+      "learning_rate": 1.4416206830741298e-07,
+      "loss": 2.0439,
+      "step": 12532
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5420342024211667,
+      "learning_rate": 1.4349296816178746e-07,
+      "loss": 1.7912,
+      "step": 12533
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5325794389655181,
+      "learning_rate": 1.4282541991000598e-07,
+      "loss": 1.9735,
+      "step": 12534
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5306285657268817,
+      "learning_rate": 1.4215942359374902e-07,
+      "loss": 1.821,
+      "step": 12535
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5312574800939726,
+      "learning_rate": 1.414949792545972e-07,
+      "loss": 2.0612,
+      "step": 12536
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5526942710250652,
+      "learning_rate": 1.4083208693403683e-07,
+      "loss": 1.8297,
+      "step": 12537
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.524278994127449,
+      "learning_rate": 1.4017074667345144e-07,
+      "loss": 1.8396,
+      "step": 12538
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5653408535402199,
+      "learning_rate": 1.3951095851413576e-07,
+      "loss": 1.8231,
+      "step": 12539
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4859839720252603,
+      "learning_rate": 1.3885272249727632e-07,
+      "loss": 1.9353,
+      "step": 12540
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5166221049756597,
+      "learning_rate": 1.3819603866397635e-07,
+      "loss": 2.0634,
+      "step": 12541
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5386256100596942,
+      "learning_rate": 1.375409070552336e-07,
+      "loss": 1.8795,
+      "step": 12542
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5714181341589131,
+      "learning_rate": 1.3688732771194868e-07,
+      "loss": 1.8362,
+      "step": 12543
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5083934969048498,
+      "learning_rate": 1.3623530067492785e-07,
+      "loss": 2.0956,
+      "step": 12544
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.534113148459979,
+      "learning_rate": 1.3558482598487743e-07,
+      "loss": 1.8619,
+      "step": 12545
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5500580739872959,
+      "learning_rate": 1.3493590368241493e-07,
+      "loss": 1.8139,
+      "step": 12546
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5529064406659795,
+      "learning_rate": 1.342885338080524e-07,
+      "loss": 1.928,
+      "step": 12547
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5449751043424931,
+      "learning_rate": 1.3364271640220748e-07,
+      "loss": 2.0957,
+      "step": 12548
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5576994275258095,
+      "learning_rate": 1.3299845150520074e-07,
+      "loss": 1.821,
+      "step": 12549
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5547620307417171,
+      "learning_rate": 1.323557391572555e-07,
+      "loss": 1.844,
+      "step": 12550
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5462297236665662,
+      "learning_rate": 1.3171457939850086e-07,
+      "loss": 1.8459,
+      "step": 12551
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5120830552148097,
+      "learning_rate": 1.310749722689686e-07,
+      "loss": 2.0574,
+      "step": 12552
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.49356176899517235,
+      "learning_rate": 1.3043691780859068e-07,
+      "loss": 1.9089,
+      "step": 12553
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5597195045515584,
+      "learning_rate": 1.2980041605719917e-07,
+      "loss": 1.8275,
+      "step": 12554
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5291635827550134,
+      "learning_rate": 1.2916546705453725e-07,
+      "loss": 1.8555,
+      "step": 12555
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5096586409762475,
+      "learning_rate": 1.2853207084025098e-07,
+      "loss": 1.9616,
+      "step": 12556
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5358808220035992,
+      "learning_rate": 1.279002274538782e-07,
+      "loss": 1.8707,
+      "step": 12557
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5254971260091508,
+      "learning_rate": 1.2726993693487067e-07,
+      "loss": 1.8466,
+      "step": 12558
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5153441807618191,
+      "learning_rate": 1.26641199322583e-07,
+      "loss": 1.9579,
+      "step": 12559
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5475401381067887,
+      "learning_rate": 1.2601401465626716e-07,
+      "loss": 1.8133,
+      "step": 12560
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5395166355395358,
+      "learning_rate": 1.253883829750835e-07,
+      "loss": 2.0474,
+      "step": 12561
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5673586689545557,
+      "learning_rate": 1.2476430431808682e-07,
+      "loss": 1.8737,
+      "step": 12562
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5242772064899629,
+      "learning_rate": 1.2414177872424603e-07,
+      "loss": 1.8382,
+      "step": 12563
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5247384427323458,
+      "learning_rate": 1.2352080623242724e-07,
+      "loss": 2.0358,
+      "step": 12564
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5178375965707774,
+      "learning_rate": 1.229013868813994e-07,
+      "loss": 1.8789,
+      "step": 12565
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5515400217574341,
+      "learning_rate": 1.2228352070983719e-07,
+      "loss": 1.8514,
+      "step": 12566
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.531798453110105,
+      "learning_rate": 1.2166720775631247e-07,
+      "loss": 1.8277,
+      "step": 12567
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5312651652487288,
+      "learning_rate": 1.210524480593056e-07,
+      "loss": 1.9952,
+      "step": 12568
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5492669528373674,
+      "learning_rate": 1.2043924165720255e-07,
+      "loss": 1.8466,
+      "step": 12569
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5412493092558326,
+      "learning_rate": 1.1982758858828658e-07,
+      "loss": 1.899,
+      "step": 12570
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4885499120743349,
+      "learning_rate": 1.1921748889074104e-07,
+      "loss": 1.917,
+      "step": 12571
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5384729953621615,
+      "learning_rate": 1.186089426026632e-07,
+      "loss": 1.8693,
+      "step": 12572
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5248666451650283,
+      "learning_rate": 1.1800194976204215e-07,
+      "loss": 2.0541,
+      "step": 12573
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5723669480282808,
+      "learning_rate": 1.1739651040677812e-07,
+      "loss": 1.8354,
+      "step": 12574
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5457631398202833,
+      "learning_rate": 1.1679262457467144e-07,
+      "loss": 1.819,
+      "step": 12575
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5180906612387542,
+      "learning_rate": 1.161902923034197e-07,
+      "loss": 1.9986,
+      "step": 12576
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5371651065631734,
+      "learning_rate": 1.1558951363064008e-07,
+      "loss": 1.8462,
+      "step": 12577
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.516921171311841,
+      "learning_rate": 1.1499028859383031e-07,
+      "loss": 1.8756,
+      "step": 12578
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5459777778827176,
+      "learning_rate": 1.1439261723040773e-07,
+      "loss": 1.8503,
+      "step": 12579
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5180000826287885,
+      "learning_rate": 1.137964995776869e-07,
+      "loss": 2.0285,
+      "step": 12580
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5323573443097662,
+      "learning_rate": 1.1320193567288529e-07,
+      "loss": 1.8159,
+      "step": 12581
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5450359683393713,
+      "learning_rate": 1.1260892555312597e-07,
+      "loss": 1.8563,
+      "step": 12582
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5612941595208045,
+      "learning_rate": 1.1201746925542933e-07,
+      "loss": 1.8445,
+      "step": 12583
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5023515488564199,
+      "learning_rate": 1.1142756681672417e-07,
+      "loss": 1.9386,
+      "step": 12584
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5185690796785887,
+      "learning_rate": 1.1083921827384214e-07,
+      "loss": 2.0632,
+      "step": 12585
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.559092574513715,
+      "learning_rate": 1.1025242366351496e-07,
+      "loss": 1.8496,
+      "step": 12586
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5505028740963146,
+      "learning_rate": 1.0966718302237999e-07,
+      "loss": 1.8386,
+      "step": 12587
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5375084409045378,
+      "learning_rate": 1.0908349638697191e-07,
+      "loss": 2.0352,
+      "step": 12588
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5573371652510921,
+      "learning_rate": 1.0850136379373654e-07,
+      "loss": 1.8089,
+      "step": 12589
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5463213396848147,
+      "learning_rate": 1.0792078527901983e-07,
+      "loss": 1.9531,
+      "step": 12590
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.562011590350056,
+      "learning_rate": 1.07341760879065e-07,
+      "loss": 1.862,
+      "step": 12591
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5691895403734718,
+      "learning_rate": 1.0676429063002646e-07,
+      "loss": 1.874,
+      "step": 12592
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5246059769537014,
+      "learning_rate": 1.0618837456795872e-07,
+      "loss": 2.0259,
+      "step": 12593
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5547160969649736,
+      "learning_rate": 1.0561401272881633e-07,
+      "loss": 1.867,
+      "step": 12594
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5401859237227011,
+      "learning_rate": 1.050412051484595e-07,
+      "loss": 1.8659,
+      "step": 12595
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4948326888696587,
+      "learning_rate": 1.044699518626513e-07,
+      "loss": 1.8602,
+      "step": 12596
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5654582025215356,
+      "learning_rate": 1.0390025290705485e-07,
+      "loss": 2.06,
+      "step": 12597
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5610644156053208,
+      "learning_rate": 1.0333210831724449e-07,
+      "loss": 1.8882,
+      "step": 12598
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.535859326193958,
+      "learning_rate": 1.0276551812868906e-07,
+      "loss": 1.8899,
+      "step": 12599
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5432551680255661,
+      "learning_rate": 1.0220048237676305e-07,
+      "loss": 2.0705,
+      "step": 12600
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5563337790798661,
+      "learning_rate": 1.0163700109674101e-07,
+      "loss": 1.9066,
+      "step": 12601
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.49674691099510326,
+      "learning_rate": 1.0107507432380591e-07,
+      "loss": 1.9466,
+      "step": 12602
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5515358471326531,
+      "learning_rate": 1.0051470209304359e-07,
+      "loss": 1.8315,
+      "step": 12603
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.544961555038401,
+      "learning_rate": 9.995588443943715e-08,
+      "loss": 1.8383,
+      "step": 12604
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5488002659085539,
+      "learning_rate": 9.939862139787814e-08,
+      "loss": 2.0881,
+      "step": 12605
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5428586615860987,
+      "learning_rate": 9.88429130031554e-08,
+      "loss": 1.8387,
+      "step": 12606
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5472002396991176,
+      "learning_rate": 9.828875928996617e-08,
+      "loss": 1.8417,
+      "step": 12607
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5319973909105155,
+      "learning_rate": 9.773616029291055e-08,
+      "loss": 2.0772,
+      "step": 12608
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.49273799229513887,
+      "learning_rate": 9.718511604648595e-08,
+      "loss": 1.9355,
+      "step": 12609
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5443128044985188,
+      "learning_rate": 9.663562658509817e-08,
+      "loss": 1.8679,
+      "step": 12610
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5350359420197642,
+      "learning_rate": 9.608769194305312e-08,
+      "loss": 1.8265,
+      "step": 12611
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5165794021836044,
+      "learning_rate": 9.55413121545623e-08,
+      "loss": 1.988,
+      "step": 12612
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5293335817613162,
+      "learning_rate": 9.499648725373456e-08,
+      "loss": 1.8485,
+      "step": 12613
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5394531034378792,
+      "learning_rate": 9.445321727459266e-08,
+      "loss": 1.8071,
+      "step": 12614
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5195230801665499,
+      "learning_rate": 9.391150225104838e-08,
+      "loss": 1.9285,
+      "step": 12615
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5449366559439456,
+      "learning_rate": 9.337134221692745e-08,
+      "loss": 1.8346,
+      "step": 12616
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5406203819873967,
+      "learning_rate": 9.283273720595287e-08,
+      "loss": 2.0912,
+      "step": 12617
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.526284757663565,
+      "learning_rate": 9.229568725174776e-08,
+      "loss": 1.7782,
+      "step": 12618
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5416142510966998,
+      "learning_rate": 9.176019238785194e-08,
+      "loss": 1.8835,
+      "step": 12619
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5332863809409796,
+      "learning_rate": 9.122625264768869e-08,
+      "loss": 2.0115,
+      "step": 12620
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.49979229266951447,
+      "learning_rate": 9.069386806460078e-08,
+      "loss": 1.9268,
+      "step": 12621
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5364194228608096,
+      "learning_rate": 9.016303867182274e-08,
+      "loss": 1.8509,
+      "step": 12622
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5586636663630387,
+      "learning_rate": 8.96337645024975e-08,
+      "loss": 1.8639,
+      "step": 12623
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5538907095114045,
+      "learning_rate": 8.910604558967361e-08,
+      "loss": 1.8214,
+      "step": 12624
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5431647550571213,
+      "learning_rate": 8.857988196629697e-08,
+      "loss": 2.0741,
+      "step": 12625
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5458643774152034,
+      "learning_rate": 8.805527366521349e-08,
+      "loss": 1.82,
+      "step": 12626
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5006335486984427,
+      "learning_rate": 8.753222071918032e-08,
+      "loss": 1.9412,
+      "step": 12627
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5407675028444235,
+      "learning_rate": 8.701072316085468e-08,
+      "loss": 1.8168,
+      "step": 12628
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5261728593233176,
+      "learning_rate": 8.649078102279662e-08,
+      "loss": 2.0601,
+      "step": 12629
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5370374367185273,
+      "learning_rate": 8.597239433746351e-08,
+      "loss": 1.8244,
+      "step": 12630
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5707207883258855,
+      "learning_rate": 8.545556313722391e-08,
+      "loss": 1.8486,
+      "step": 12631
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5229009867842098,
+      "learning_rate": 8.494028745434368e-08,
+      "loss": 1.9996,
+      "step": 12632
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5195900136527069,
+      "learning_rate": 8.442656732099707e-08,
+      "loss": 1.9094,
+      "step": 12633
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5350028033686787,
+      "learning_rate": 8.391440276925566e-08,
+      "loss": 1.8415,
+      "step": 12634
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5577082776266667,
+      "learning_rate": 8.340379383109387e-08,
+      "loss": 1.8343,
+      "step": 12635
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5382150867940216,
+      "learning_rate": 8.289474053839452e-08,
+      "loss": 1.7985,
+      "step": 12636
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.532893845302689,
+      "learning_rate": 8.238724292294053e-08,
+      "loss": 2.052,
+      "step": 12637
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.567900878615726,
+      "learning_rate": 8.188130101641212e-08,
+      "loss": 1.874,
+      "step": 12638
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5527065720761866,
+      "learning_rate": 8.137691485040066e-08,
+      "loss": 1.8213,
+      "step": 12639
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49983910040330676,
+      "learning_rate": 8.087408445639766e-08,
+      "loss": 1.9008,
+      "step": 12640
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.51717154516016,
+      "learning_rate": 8.037280986579466e-08,
+      "loss": 2.0688,
+      "step": 12641
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5689292508001577,
+      "learning_rate": 7.987309110989161e-08,
+      "loss": 1.8857,
+      "step": 12642
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5358937832116302,
+      "learning_rate": 7.93749282198858e-08,
+      "loss": 1.8644,
+      "step": 12643
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.519379638150312,
+      "learning_rate": 7.887832122687732e-08,
+      "loss": 1.8693,
+      "step": 12644
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5233268286120923,
+      "learning_rate": 7.838327016187474e-08,
+      "loss": 2.0274,
+      "step": 12645
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4973262696283313,
+      "learning_rate": 7.788977505578666e-08,
+      "loss": 1.9418,
+      "step": 12646
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5518111053824067,
+      "learning_rate": 7.739783593942174e-08,
+      "loss": 1.8502,
+      "step": 12647
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5725631126749992,
+      "learning_rate": 7.690745284349432e-08,
+      "loss": 1.807,
+      "step": 12648
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.535070471072652,
+      "learning_rate": 7.64186257986188e-08,
+      "loss": 2.0662,
+      "step": 12649
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5620079190844915,
+      "learning_rate": 7.593135483532077e-08,
+      "loss": 1.8727,
+      "step": 12650
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5259601592682241,
+      "learning_rate": 7.544563998401755e-08,
+      "loss": 1.8083,
+      "step": 12651
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5291452998839081,
+      "learning_rate": 7.496148127503488e-08,
+      "loss": 1.9502,
+      "step": 12652
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5383358448263236,
+      "learning_rate": 7.447887873860415e-08,
+      "loss": 2.0279,
+      "step": 12653
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.540220288201958,
+      "learning_rate": 7.399783240485125e-08,
+      "loss": 1.8525,
+      "step": 12654
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5495777913048978,
+      "learning_rate": 7.351834230381327e-08,
+      "loss": 1.8702,
+      "step": 12655
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5473260683270397,
+      "learning_rate": 7.304040846542459e-08,
+      "loss": 1.9092,
+      "step": 12656
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5188596261148773,
+      "learning_rate": 7.256403091952523e-08,
+      "loss": 2.0571,
+      "step": 12657
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4812111618076212,
+      "learning_rate": 7.208920969586085e-08,
+      "loss": 1.9482,
+      "step": 12658
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.556614725038666,
+      "learning_rate": 7.161594482407164e-08,
+      "loss": 1.8053,
+      "step": 12659
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5718698252619451,
+      "learning_rate": 7.114423633370615e-08,
+      "loss": 1.811,
+      "step": 12660
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5309689179370667,
+      "learning_rate": 7.067408425421862e-08,
+      "loss": 2.0472,
+      "step": 12661
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5471651863457686,
+      "learning_rate": 7.020548861495779e-08,
+      "loss": 1.8587,
+      "step": 12662
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5371848841488011,
+      "learning_rate": 6.973844944518637e-08,
+      "loss": 1.8301,
+      "step": 12663
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5200849316253853,
+      "learning_rate": 6.927296677405604e-08,
+      "loss": 2.0161,
+      "step": 12664
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5224383581107138,
+      "learning_rate": 6.880904063063243e-08,
+      "loss": 1.9337,
+      "step": 12665
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5508369577985587,
+      "learning_rate": 6.834667104388403e-08,
+      "loss": 1.8609,
+      "step": 12666
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5438198142178579,
+      "learning_rate": 6.788585804267112e-08,
+      "loss": 1.8763,
+      "step": 12667
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5343235896165646,
+      "learning_rate": 6.742660165576786e-08,
+      "loss": 1.816,
+      "step": 12668
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5168653020761007,
+      "learning_rate": 6.696890191184856e-08,
+      "loss": 2.0808,
+      "step": 12669
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5459212451722081,
+      "learning_rate": 6.65127588394876e-08,
+      "loss": 1.918,
+      "step": 12670
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5018438398293998,
+      "learning_rate": 6.605817246716495e-08,
+      "loss": 1.9623,
+      "step": 12671
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5539811303810094,
+      "learning_rate": 6.560514282326347e-08,
+      "loss": 1.8767,
+      "step": 12672
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5200526372878406,
+      "learning_rate": 6.515366993606332e-08,
+      "loss": 1.9968,
+      "step": 12673
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5736847151433727,
+      "learning_rate": 6.470375383375582e-08,
+      "loss": 1.8673,
+      "step": 12674
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5591678346149859,
+      "learning_rate": 6.425539454442963e-08,
+      "loss": 1.8409,
+      "step": 12675
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5468278237356234,
+      "learning_rate": 6.380859209607626e-08,
+      "loss": 1.8718,
+      "step": 12676
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.48315293383017843,
+      "learning_rate": 6.33633465165956e-08,
+      "loss": 2.0875,
+      "step": 12677
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5397592932382222,
+      "learning_rate": 6.291965783377929e-08,
+      "loss": 1.8753,
+      "step": 12678
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5525657239818212,
+      "learning_rate": 6.247752607533575e-08,
+      "loss": 1.8797,
+      "step": 12679
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5417268485491037,
+      "learning_rate": 6.203695126886788e-08,
+      "loss": 1.8334,
+      "step": 12680
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5182953659560007,
+      "learning_rate": 6.15979334418787e-08,
+      "loss": 2.095,
+      "step": 12681
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5443430182078263,
+      "learning_rate": 6.11604726217796e-08,
+      "loss": 1.8308,
+      "step": 12682
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.513913194390248,
+      "learning_rate": 6.072456883588485e-08,
+      "loss": 1.9282,
+      "step": 12683
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5491914494730041,
+      "learning_rate": 6.0290222111406e-08,
+      "loss": 1.7579,
+      "step": 12684
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5271174798744462,
+      "learning_rate": 5.985743247546582e-08,
+      "loss": 2.0379,
+      "step": 12685
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5438864693030997,
+      "learning_rate": 5.942619995508159e-08,
+      "loss": 1.8065,
+      "step": 12686
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5470665758471396,
+      "learning_rate": 5.8996524577178967e-08,
+      "loss": 1.8731,
+      "step": 12687
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5660181968909646,
+      "learning_rate": 5.856840636858374e-08,
+      "loss": 1.8047,
+      "step": 12688
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49110596949710383,
+      "learning_rate": 5.814184535602452e-08,
+      "loss": 2.0712,
+      "step": 12689
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5480895925612354,
+      "learning_rate": 5.771684156613277e-08,
+      "loss": 1.8166,
+      "step": 12690
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5523892303984304,
+      "learning_rate": 5.7293395025445617e-08,
+      "loss": 1.8763,
+      "step": 12691
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5361268346768417,
+      "learning_rate": 5.687150576039746e-08,
+      "loss": 1.8493,
+      "step": 12692
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5253058924694681,
+      "learning_rate": 5.645117379733111e-08,
+      "loss": 2.0683,
+      "step": 12693
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5518078771811902,
+      "learning_rate": 5.603239916248948e-08,
+      "loss": 1.8244,
+      "step": 12694
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5589974677716034,
+      "learning_rate": 5.5615181882015534e-08,
+      "loss": 1.8401,
+      "step": 12695
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49192212574870675,
+      "learning_rate": 5.519952198196066e-08,
+      "loss": 1.9287,
+      "step": 12696
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5426366164477189,
+      "learning_rate": 5.478541948827631e-08,
+      "loss": 2.0957,
+      "step": 12697
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5561824100639771,
+      "learning_rate": 5.4372874426814045e-08,
+      "loss": 1.8399,
+      "step": 12698
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5607261188328716,
+      "learning_rate": 5.3961886823333806e-08,
+      "loss": 1.8499,
+      "step": 12699
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49683726883633333,
+      "learning_rate": 5.3552456703495626e-08,
+      "loss": 1.8472,
+      "step": 12700
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5489112065707532,
+      "learning_rate": 5.314458409285683e-08,
+      "loss": 2.034,
+      "step": 12701
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49833393531273956,
+      "learning_rate": 5.273826901688872e-08,
+      "loss": 1.9565,
+      "step": 12702
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5382170386068984,
+      "learning_rate": 5.2333511500954334e-08,
+      "loss": 1.8432,
+      "step": 12703
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5552571614438986,
+      "learning_rate": 5.193031157032513e-08,
+      "loss": 1.8412,
+      "step": 12704
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5499559579250465,
+      "learning_rate": 5.1528669250178184e-08,
+      "loss": 2.0328,
+      "step": 12705
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5443377080605588,
+      "learning_rate": 5.1128584565585115e-08,
+      "loss": 1.812,
+      "step": 12706
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5437971121918732,
+      "learning_rate": 5.073005754152871e-08,
+      "loss": 1.8017,
+      "step": 12707
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5155693935909547,
+      "learning_rate": 5.033308820289184e-08,
+      "loss": 1.9271,
+      "step": 12708
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5639006346933892,
+      "learning_rate": 4.993767657445192e-08,
+      "loss": 2.0794,
+      "step": 12709
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5385293324501685,
+      "learning_rate": 4.95438226809003e-08,
+      "loss": 1.8203,
+      "step": 12710
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5398241014775447,
+      "learning_rate": 4.915152654682842e-08,
+      "loss": 1.8872,
+      "step": 12711
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5888667607431607,
+      "learning_rate": 4.8760788196727804e-08,
+      "loss": 1.8844,
+      "step": 12712
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5250310451161447,
+      "learning_rate": 4.8371607654992825e-08,
+      "loss": 2.0716,
+      "step": 12713
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.49539355719705613,
+      "learning_rate": 4.798398494592071e-08,
+      "loss": 1.9101,
+      "step": 12714
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5352536356496731,
+      "learning_rate": 4.759792009371711e-08,
+      "loss": 1.8494,
+      "step": 12715
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5463177850110065,
+      "learning_rate": 4.72134131224794e-08,
+      "loss": 1.846,
+      "step": 12716
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5268101760204572,
+      "learning_rate": 4.683046405621894e-08,
+      "loss": 2.0437,
+      "step": 12717
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5339729818327588,
+      "learning_rate": 4.64490729188416e-08,
+      "loss": 1.9023,
+      "step": 12718
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5390969025439178,
+      "learning_rate": 4.6069239734161665e-08,
+      "loss": 1.8271,
+      "step": 12719
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.490460920450286,
+      "learning_rate": 4.5690964525890726e-08,
+      "loss": 1.9334,
+      "step": 12720
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5330683969267866,
+      "learning_rate": 4.531424731764877e-08,
+      "loss": 2.0577,
+      "step": 12721
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5812444356767384,
+      "learning_rate": 4.493908813295311e-08,
+      "loss": 1.8555,
+      "step": 12722
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5768701402542648,
+      "learning_rate": 4.4565486995229446e-08,
+      "loss": 1.8807,
+      "step": 12723
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5321720368405246,
+      "learning_rate": 4.419344392780356e-08,
+      "loss": 1.8588,
+      "step": 12724
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5291896766039336,
+      "learning_rate": 4.382295895389854e-08,
+      "loss": 2.0183,
+      "step": 12725
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5532818790852237,
+      "learning_rate": 4.3454032096651446e-08,
+      "loss": 1.8614,
+      "step": 12726
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5140013341804782,
+      "learning_rate": 4.308666337909106e-08,
+      "loss": 1.9338,
+      "step": 12727
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5476791640202888,
+      "learning_rate": 4.272085282415739e-08,
+      "loss": 1.8071,
+      "step": 12728
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5211579821670043,
+      "learning_rate": 4.23566004546877e-08,
+      "loss": 2.0501,
+      "step": 12729
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5343809462811061,
+      "learning_rate": 4.199390629342492e-08,
+      "loss": 1.8399,
+      "step": 12730
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5404780585593314,
+      "learning_rate": 4.1632770363012056e-08,
+      "loss": 1.8227,
+      "step": 12731
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5253215361773808,
+      "learning_rate": 4.127319268599494e-08,
+      "loss": 1.8566,
+      "step": 12732
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4780602218067452,
+      "learning_rate": 4.0915173284827857e-08,
+      "loss": 2.0651,
+      "step": 12733
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5581575483723127,
+      "learning_rate": 4.055871218185958e-08,
+      "loss": 1.8485,
+      "step": 12734
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5377137220996736,
+      "learning_rate": 4.020380939934732e-08,
+      "loss": 1.8824,
+      "step": 12735
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5487743107257379,
+      "learning_rate": 3.985046495944833e-08,
+      "loss": 1.8545,
+      "step": 12736
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5402240659215451,
+      "learning_rate": 3.949867888422554e-08,
+      "loss": 2.0564,
+      "step": 12737
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5396436311355212,
+      "learning_rate": 3.914845119563915e-08,
+      "loss": 1.8447,
+      "step": 12738
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4928139754264298,
+      "learning_rate": 3.8799781915555e-08,
+      "loss": 1.9578,
+      "step": 12739
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5679757530552128,
+      "learning_rate": 3.845267106574735e-08,
+      "loss": 1.7783,
+      "step": 12740
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5412951693886414,
+      "learning_rate": 3.810711866788219e-08,
+      "loss": 2.0156,
+      "step": 12741
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.552954387511485,
+      "learning_rate": 3.776312474353394e-08,
+      "loss": 1.8231,
+      "step": 12742
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5495456785466195,
+      "learning_rate": 3.742068931418263e-08,
+      "loss": 1.8587,
+      "step": 12743
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.562644544699006,
+      "learning_rate": 3.707981240120839e-08,
+      "loss": 1.8713,
+      "step": 12744
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5040839986101762,
+      "learning_rate": 3.674049402588864e-08,
+      "loss": 2.06,
+      "step": 12745
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5381314260734263,
+      "learning_rate": 3.640273420941198e-08,
+      "loss": 1.8638,
+      "step": 12746
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5216463037844793,
+      "learning_rate": 3.606653297286711e-08,
+      "loss": 1.817,
+      "step": 12747
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5577135071554226,
+      "learning_rate": 3.5731890337242776e-08,
+      "loss": 1.8524,
+      "step": 12748
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5322586414052444,
+      "learning_rate": 3.53988063234334e-08,
+      "loss": 1.9986,
+      "step": 12749
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5477515667008829,
+      "learning_rate": 3.506728095223066e-08,
+      "loss": 1.8136,
+      "step": 12750
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5119090284384246,
+      "learning_rate": 3.473731424433746e-08,
+      "loss": 1.9188,
+      "step": 12751
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5367923286069458,
+      "learning_rate": 3.440890622035398e-08,
+      "loss": 1.8416,
+      "step": 12752
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5351785567275853,
+      "learning_rate": 3.4082056900783254e-08,
+      "loss": 1.9968,
+      "step": 12753
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5358556438066955,
+      "learning_rate": 3.37567663060312e-08,
+      "loss": 1.8075,
+      "step": 12754
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5415079954579507,
+      "learning_rate": 3.3433034456412106e-08,
+      "loss": 1.9122,
+      "step": 12755
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.540279277493834,
+      "learning_rate": 3.3110861372129264e-08,
+      "loss": 1.8225,
+      "step": 12756
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5070501218706311,
+      "learning_rate": 3.279024707330547e-08,
+      "loss": 2.0261,
+      "step": 12757
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5302924634080722,
+      "learning_rate": 3.2471191579952485e-08,
+      "loss": 1.9096,
+      "step": 12758
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.605153847892772,
+      "learning_rate": 3.2153694911990494e-08,
+      "loss": 1.8467,
+      "step": 12759
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5431018642030828,
+      "learning_rate": 3.183775708924808e-08,
+      "loss": 1.8638,
+      "step": 12760
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5582861872441743,
+      "learning_rate": 3.1523378131442796e-08,
+      "loss": 2.0233,
+      "step": 12761
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5424774883403826,
+      "learning_rate": 3.1210558058208937e-08,
+      "loss": 1.8802,
+      "step": 12762
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5575470057551708,
+      "learning_rate": 3.089929688907256e-08,
+      "loss": 1.8521,
+      "step": 12763
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5116598775666753,
+      "learning_rate": 3.0589594643468114e-08,
+      "loss": 1.962,
+      "step": 12764
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5226371891507903,
+      "learning_rate": 3.02814513407329e-08,
+      "loss": 2.0378,
+      "step": 12765
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5606416762952409,
+      "learning_rate": 2.997486700010432e-08,
+      "loss": 1.89,
+      "step": 12766
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.550864330832561,
+      "learning_rate": 2.9669841640725393e-08,
+      "loss": 1.84,
+      "step": 12767
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.543953206855594,
+      "learning_rate": 2.9366375281639213e-08,
+      "loss": 1.8224,
+      "step": 12768
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5366439724100214,
+      "learning_rate": 2.906446794179174e-08,
+      "loss": 2.052,
+      "step": 12769
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5024024512682185,
+      "learning_rate": 2.8764119640031783e-08,
+      "loss": 1.9314,
+      "step": 12770
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5378747490599656,
+      "learning_rate": 2.8465330395111012e-08,
+      "loss": 1.8749,
+      "step": 12771
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5551855842379617,
+      "learning_rate": 2.816810022568672e-08,
+      "loss": 1.887,
+      "step": 12772
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5220511198792261,
+      "learning_rate": 2.7872429150316294e-08,
+      "loss": 2.0715,
+      "step": 12773
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.571609528831424,
+      "learning_rate": 2.757831718745718e-08,
+      "loss": 1.8691,
+      "step": 12774
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5494269113386022,
+      "learning_rate": 2.7285764355472475e-08,
+      "loss": 1.8,
+      "step": 12775
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4988895170506658,
+      "learning_rate": 2.6994770672628123e-08,
+      "loss": 1.9143,
+      "step": 12776
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5465917675273891,
+      "learning_rate": 2.670533615709292e-08,
+      "loss": 1.8395,
+      "step": 12777
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5114986090501923,
+      "learning_rate": 2.6417460826935746e-08,
+      "loss": 2.0511,
+      "step": 12778
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5576799581000161,
+      "learning_rate": 2.613114470012834e-08,
+      "loss": 1.8487,
+      "step": 12779
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5363373919308031,
+      "learning_rate": 2.5846387794550842e-08,
+      "loss": 1.8412,
+      "step": 12780
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5395181084629698,
+      "learning_rate": 2.5563190127980696e-08,
+      "loss": 2.0086,
+      "step": 12781
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5174136543638879,
+      "learning_rate": 2.5281551718098208e-08,
+      "loss": 1.9393,
+      "step": 12782
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5556289835232506,
+      "learning_rate": 2.500147258248653e-08,
+      "loss": 1.8183,
+      "step": 12783
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5559679702027939,
+      "learning_rate": 2.4722952738631676e-08,
+      "loss": 1.86,
+      "step": 12784
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.562802808749891,
+      "learning_rate": 2.4445992203925293e-08,
+      "loss": 2.0791,
+      "step": 12785
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5539333060397166,
+      "learning_rate": 2.417059099565633e-08,
+      "loss": 1.8789,
+      "step": 12786
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5751472653169857,
+      "learning_rate": 2.3896749131022135e-08,
+      "loss": 1.8365,
+      "step": 12787
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5485465960917203,
+      "learning_rate": 2.362446662711737e-08,
+      "loss": 1.8577,
+      "step": 12788
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5034027923825843,
+      "learning_rate": 2.3353743500942325e-08,
+      "loss": 1.9211,
+      "step": 12789
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5572957006449073,
+      "learning_rate": 2.308457976940015e-08,
+      "loss": 2.0417,
+      "step": 12790
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5312749746377146,
+      "learning_rate": 2.281697544929684e-08,
+      "loss": 1.8282,
+      "step": 12791
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.555257056049469,
+      "learning_rate": 2.2550930557335704e-08,
+      "loss": 1.8038,
+      "step": 12792
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5103850517755295,
+      "learning_rate": 2.2286445110131225e-08,
+      "loss": 2.0027,
+      "step": 12793
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5326093446501065,
+      "learning_rate": 2.20235191241952e-08,
+      "loss": 1.8296,
+      "step": 12794
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5135536995616966,
+      "learning_rate": 2.1762152615942277e-08,
+      "loss": 1.8821,
+      "step": 12795
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.540209221417365,
+      "learning_rate": 2.150234560169273e-08,
+      "loss": 1.8356,
+      "step": 12796
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5309298960346736,
+      "learning_rate": 2.124409809766692e-08,
+      "loss": 1.9961,
+      "step": 12797
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5561346650899518,
+      "learning_rate": 2.0987410119985284e-08,
+      "loss": 1.8623,
+      "step": 12798
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.542254420812681,
+      "learning_rate": 2.073228168467667e-08,
+      "loss": 1.8488,
+      "step": 12799
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5565537588407203,
+      "learning_rate": 2.0478712807667222e-08,
+      "loss": 1.7944,
+      "step": 12800
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5131197592858376,
+      "learning_rate": 2.02267035047915e-08,
+      "loss": 1.9616,
+      "step": 12801
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5359541476718073,
+      "learning_rate": 1.997625379178414e-08,
+      "loss": 2.054,
+      "step": 12802
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5885334090488491,
+      "learning_rate": 1.9727363684277077e-08,
+      "loss": 1.8489,
+      "step": 12803
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5662611088438113,
+      "learning_rate": 1.948003319781344e-08,
+      "loss": 1.7935,
+      "step": 12804
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.52930347528441,
+      "learning_rate": 1.9234262347833652e-08,
+      "loss": 2.0759,
+      "step": 12805
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5419895160582128,
+      "learning_rate": 1.8990051149683774e-08,
+      "loss": 1.8791,
+      "step": 12806
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5081769097193298,
+      "learning_rate": 1.8747399618607164e-08,
+      "loss": 1.958,
+      "step": 12807
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5274914532840101,
+      "learning_rate": 1.850630776975837e-08,
+      "loss": 1.827,
+      "step": 12808
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5476653170661598,
+      "learning_rate": 1.826677561818646e-08,
+      "loss": 1.8268,
+      "step": 12809
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5528649562007844,
+      "learning_rate": 1.8028803178846142e-08,
+      "loss": 2.0665,
+      "step": 12810
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5472353927704439,
+      "learning_rate": 1.7792390466597754e-08,
+      "loss": 1.8691,
+      "step": 12811
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5671658498593329,
+      "learning_rate": 1.7557537496198927e-08,
+      "loss": 1.8198,
+      "step": 12812
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5273872028797206,
+      "learning_rate": 1.732424428231294e-08,
+      "loss": 2.1192,
+      "step": 12813
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5108209403648459,
+      "learning_rate": 1.7092510839505914e-08,
+      "loss": 1.9177,
+      "step": 12814
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.557838702541044,
+      "learning_rate": 1.6862337182246835e-08,
+      "loss": 1.8563,
+      "step": 12815
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5572753071706266,
+      "learning_rate": 1.663372332490476e-08,
+      "loss": 1.8785,
+      "step": 12816
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5287800529999468,
+      "learning_rate": 1.6406669281754382e-08,
+      "loss": 2.0596,
+      "step": 12817
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5336721475028753,
+      "learning_rate": 1.618117506697048e-08,
+      "loss": 1.8819,
+      "step": 12818
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5365070177425443,
+      "learning_rate": 1.5957240694630672e-08,
+      "loss": 1.8576,
+      "step": 12819
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5145073357756669,
+      "learning_rate": 1.5734866178718223e-08,
+      "loss": 1.9196,
+      "step": 12820
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5246799975326919,
+      "learning_rate": 1.5514051533116468e-08,
+      "loss": 1.8061,
+      "step": 12821
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5399867104724179,
+      "learning_rate": 1.529479677161161e-08,
+      "loss": 2.0693,
+      "step": 12822
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5453435124435558,
+      "learning_rate": 1.507710190788991e-08,
+      "loss": 1.8097,
+      "step": 12823
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5446354099529966,
+      "learning_rate": 1.4860966955548838e-08,
+      "loss": 1.8353,
+      "step": 12824
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5387154863444586,
+      "learning_rate": 1.4646391928077595e-08,
+      "loss": 2.0569,
+      "step": 12825
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.49982161676903925,
+      "learning_rate": 1.4433376838873801e-08,
+      "loss": 1.8912,
+      "step": 12826
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5498456077376449,
+      "learning_rate": 1.4221921701237928e-08,
+      "loss": 1.8672,
+      "step": 12827
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5509377876659737,
+      "learning_rate": 1.4012026528370525e-08,
+      "loss": 1.8488,
+      "step": 12828
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5497211119502855,
+      "learning_rate": 1.3803691333377777e-08,
+      "loss": 2.052,
+      "step": 12829
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5382089975719664,
+      "learning_rate": 1.3596916129268721e-08,
+      "loss": 1.8808,
+      "step": 12830
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5422248182104982,
+      "learning_rate": 1.3391700928949703e-08,
+      "loss": 1.8315,
+      "step": 12831
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5103107601285621,
+      "learning_rate": 1.3188045745232691e-08,
+      "loss": 1.9446,
+      "step": 12832
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5749850218982084,
+      "learning_rate": 1.2985950590835294e-08,
+      "loss": 1.8522,
+      "step": 12833
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5302559700026642,
+      "learning_rate": 1.2785415478375196e-08,
+      "loss": 2.0538,
+      "step": 12834
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5515767994926126,
+      "learning_rate": 1.2586440420372936e-08,
+      "loss": 1.8736,
+      "step": 12835
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5333634540863439,
+      "learning_rate": 1.238902542924636e-08,
+      "loss": 1.8794,
+      "step": 12836
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5448599623840182,
+      "learning_rate": 1.2193170517327268e-08,
+      "loss": 2.0072,
+      "step": 12837
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5055703725745492,
+      "learning_rate": 1.1998875696839218e-08,
+      "loss": 1.9062,
+      "step": 12838
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5438334034623047,
+      "learning_rate": 1.1806140979916947e-08,
+      "loss": 1.9356,
+      "step": 12839
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5440237899180227,
+      "learning_rate": 1.161496637858972e-08,
+      "loss": 1.8395,
+      "step": 12840
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5403451737465077,
+      "learning_rate": 1.1425351904795212e-08,
+      "loss": 1.8121,
+      "step": 12841
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5040399856048634,
+      "learning_rate": 1.1237297570371175e-08,
+      "loss": 2.0805,
+      "step": 12842
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5361409237252649,
+      "learning_rate": 1.1050803387060992e-08,
+      "loss": 1.8551,
+      "step": 12843
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5511620807969663,
+      "learning_rate": 1.0865869366505355e-08,
+      "loss": 1.7945,
+      "step": 12844
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.507633089375155,
+      "learning_rate": 1.068249552025058e-08,
+      "loss": 1.874,
+      "step": 12845
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5549417732088375,
+      "learning_rate": 1.0500681859745842e-08,
+      "loss": 2.0484,
+      "step": 12846
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5378655826097648,
+      "learning_rate": 1.032042839634595e-08,
+      "loss": 1.7979,
+      "step": 12847
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5470753199781028,
+      "learning_rate": 1.014173514129746e-08,
+      "loss": 1.8902,
+      "step": 12848
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5032611999544041,
+      "learning_rate": 9.964602105763665e-09,
+      "loss": 2.0236,
+      "step": 12849
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5495371835638619,
+      "learning_rate": 9.789029300802389e-09,
+      "loss": 1.8618,
+      "step": 12850
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5339553827663366,
+      "learning_rate": 9.615016737374306e-09,
+      "loss": 1.9248,
+      "step": 12851
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5411082800945316,
+      "learning_rate": 9.442564426342949e-09,
+      "loss": 1.7471,
+      "step": 12852
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5430415430204144,
+      "learning_rate": 9.271672378474706e-09,
+      "loss": 1.8393,
+      "step": 12853
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5322541222375715,
+      "learning_rate": 9.102340604441595e-09,
+      "loss": 2.0216,
+      "step": 12854
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5798996038023844,
+      "learning_rate": 8.934569114815717e-09,
+      "loss": 1.8642,
+      "step": 12855
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.540324213626176,
+      "learning_rate": 8.768357920069247e-09,
+      "loss": 1.8543,
+      "step": 12856
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5259371568034433,
+      "learning_rate": 8.603707030582774e-09,
+      "loss": 1.9317,
+      "step": 12857
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5332460163335814,
+      "learning_rate": 8.440616456631411e-09,
+      "loss": 2.0635,
+      "step": 12858
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5427477069813599,
+      "learning_rate": 8.279086208401454e-09,
+      "loss": 1.8299,
+      "step": 12859
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5343159031207851,
+      "learning_rate": 8.119116295976503e-09,
+      "loss": 1.8659,
+      "step": 12860
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5434889605822667,
+      "learning_rate": 7.96070672934579e-09,
+      "loss": 1.8206,
+      "step": 12861
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5668502349485895,
+      "learning_rate": 7.803857518395852e-09,
+      "loss": 2.0587,
+      "step": 12862
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5116010395545514,
+      "learning_rate": 7.64856867292163e-09,
+      "loss": 1.9073,
+      "step": 12863
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5439204149447824,
+      "learning_rate": 7.494840202620922e-09,
+      "loss": 1.8806,
+      "step": 12864
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5394396024011466,
+      "learning_rate": 7.342672117086058e-09,
+      "loss": 1.8655,
+      "step": 12865
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5151057475224693,
+      "learning_rate": 7.1920644258205436e-09,
+      "loss": 2.0685,
+      "step": 12866
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.530427163020553,
+      "learning_rate": 7.0430171382307455e-09,
+      "loss": 1.8967,
+      "step": 12867
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5585150485696962,
+      "learning_rate": 6.895530263617555e-09,
+      "loss": 1.8691,
+      "step": 12868
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5091573650165916,
+      "learning_rate": 6.749603811190275e-09,
+      "loss": 1.9402,
+      "step": 12869
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5225382375092924,
+      "learning_rate": 6.605237790058283e-09,
+      "loss": 2.0152,
+      "step": 12870
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5216198249999051,
+      "learning_rate": 6.462432209239367e-09,
+      "loss": 1.8616,
+      "step": 12871
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5236206213815553,
+      "learning_rate": 6.32118707764584e-09,
+      "loss": 1.8691,
+      "step": 12872
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5653374398981833,
+      "learning_rate": 6.181502404098427e-09,
+      "loss": 1.848,
+      "step": 12873
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5240068455431693,
+      "learning_rate": 6.043378197315153e-09,
+      "loss": 2.0557,
+      "step": 12874
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5535647209309493,
+      "learning_rate": 5.90681446592245e-09,
+      "loss": 1.8326,
+      "step": 12875
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5421142193506108,
+      "learning_rate": 5.771811218444056e-09,
+      "loss": 1.9204,
+      "step": 12876
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5651087921344566,
+      "learning_rate": 5.638368463312116e-09,
+      "loss": 1.8378,
+      "step": 12877
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5081701422759135,
+      "learning_rate": 5.506486208856076e-09,
+      "loss": 2.0438,
+      "step": 12878
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5450160909175693,
+      "learning_rate": 5.376164463311018e-09,
+      "loss": 1.8624,
+      "step": 12879
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5314499709272762,
+      "learning_rate": 5.247403234809323e-09,
+      "loss": 1.8817,
+      "step": 12880
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5574501845343649,
+      "learning_rate": 5.120202531397333e-09,
+      "loss": 2.0645,
+      "step": 12881
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.48722957981921344,
+      "learning_rate": 4.994562361010368e-09,
+      "loss": 1.9237,
+      "step": 12882
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.542611995529748,
+      "learning_rate": 4.87048273149493e-09,
+      "loss": 1.8454,
+      "step": 12883
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5330092863172161,
+      "learning_rate": 4.747963650597598e-09,
+      "loss": 1.8638,
+      "step": 12884
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5608851744263007,
+      "learning_rate": 4.627005125967809e-09,
+      "loss": 1.864,
+      "step": 12885
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5277984886168451,
+      "learning_rate": 4.507607165157857e-09,
+      "loss": 2.0985,
+      "step": 12886
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5839726330849501,
+      "learning_rate": 4.3897697756228874e-09,
+      "loss": 1.8729,
+      "step": 12887
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4972330756009016,
+      "learning_rate": 4.2734929647181285e-09,
+      "loss": 1.923,
+      "step": 12888
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5364795677232849,
+      "learning_rate": 4.15877673970444e-09,
+      "loss": 1.8419,
+      "step": 12889
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5186413746557103,
+      "learning_rate": 4.045621107745534e-09,
+      "loss": 2.0454,
+      "step": 12890
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5387920096109056,
+      "learning_rate": 3.93402607590243e-09,
+      "loss": 1.8888,
+      "step": 12891
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5312001458061415,
+      "learning_rate": 3.823991651144554e-09,
+      "loss": 1.8191,
+      "step": 12892
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5388291101410658,
+      "learning_rate": 3.7155178403441848e-09,
+      "loss": 1.814,
+      "step": 12893
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5047745144893855,
+      "learning_rate": 3.6086046502681326e-09,
+      "loss": 2.0588,
+      "step": 12894
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5627606259668229,
+      "learning_rate": 3.5032520875971643e-09,
+      "loss": 1.8592,
+      "step": 12895
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5541122171074505,
+      "learning_rate": 3.3994601589038e-09,
+      "loss": 1.8253,
+      "step": 12896
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5513142822874707,
+      "learning_rate": 3.2972288706717425e-09,
+      "loss": 1.8622,
+      "step": 12897
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5081497117862622,
+      "learning_rate": 3.196558229284774e-09,
+      "loss": 2.0356,
+      "step": 12898
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5395541136465374,
+      "learning_rate": 3.0974482410239813e-09,
+      "loss": 1.8324,
+      "step": 12899
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5028777404217911,
+      "learning_rate": 2.9998989120788578e-09,
+      "loss": 1.9053,
+      "step": 12900
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5483321901645064,
+      "learning_rate": 2.9039102485417524e-09,
+      "loss": 1.8417,
+      "step": 12901
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5625779688427248,
+      "learning_rate": 2.8094822564050936e-09,
+      "loss": 1.9986,
+      "step": 12902
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5501153195081722,
+      "learning_rate": 2.7166149415613905e-09,
+      "loss": 1.858,
+      "step": 12903
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5385904350419389,
+      "learning_rate": 2.6253083098115584e-09,
+      "loss": 1.849,
+      "step": 12904
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5515099757028045,
+      "learning_rate": 2.5355623668538164e-09,
+      "loss": 1.8753,
+      "step": 12905
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5529170780677597,
+      "learning_rate": 2.4473771182947914e-09,
+      "loss": 2.0499,
+      "step": 12906
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5137424625825454,
+      "learning_rate": 2.3607525696384135e-09,
+      "loss": 1.9149,
+      "step": 12907
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5259261199510776,
+      "learning_rate": 2.275688726294245e-09,
+      "loss": 1.8352,
+      "step": 12908
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5733449561476708,
+      "learning_rate": 2.1921855935719272e-09,
+      "loss": 1.8858,
+      "step": 12909
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5181933314992972,
+      "learning_rate": 2.110243176683957e-09,
+      "loss": 2.0265,
+      "step": 12910
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5415346639630024,
+      "learning_rate": 2.0298614807456874e-09,
+      "loss": 1.8162,
+      "step": 12911
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5366704449611802,
+      "learning_rate": 1.951040510780877e-09,
+      "loss": 1.8396,
+      "step": 12912
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5151198861679401,
+      "learning_rate": 1.873780271705039e-09,
+      "loss": 1.937,
+      "step": 12913
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5335693682582073,
+      "learning_rate": 1.7980807683448674e-09,
+      "loss": 2.0364,
+      "step": 12914
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5419626636876316,
+      "learning_rate": 1.7239420054243615e-09,
+      "loss": 1.8178,
+      "step": 12915
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5523098733901055,
+      "learning_rate": 1.651363987575927e-09,
+      "loss": 1.8896,
+      "step": 12916
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5362504726965033,
+      "learning_rate": 1.5803467193264977e-09,
+      "loss": 1.8363,
+      "step": 12917
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5446112410298122,
+      "learning_rate": 1.5108902051141903e-09,
+      "loss": 2.0733,
+      "step": 12918
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5106676437203901,
+      "learning_rate": 1.4429944492716507e-09,
+      "loss": 1.9181,
+      "step": 12919
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5372676542601247,
+      "learning_rate": 1.3766594560399305e-09,
+      "loss": 1.8278,
+      "step": 12920
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5683255149306662,
+      "learning_rate": 1.3118852295601613e-09,
+      "loss": 1.8624,
+      "step": 12921
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5194736067535166,
+      "learning_rate": 1.248671773876331e-09,
+      "loss": 2.0548,
+      "step": 12922
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5382111895933023,
+      "learning_rate": 1.187019092935282e-09,
+      "loss": 1.8067,
+      "step": 12923
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5270498365092458,
+      "learning_rate": 1.1269271905867128e-09,
+      "loss": 1.8474,
+      "step": 12924
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5219585103045453,
+      "learning_rate": 1.0683960705831774e-09,
+      "loss": 1.9149,
+      "step": 12925
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5208685240656405,
+      "learning_rate": 1.0114257365773094e-09,
+      "loss": 2.028,
+      "step": 12926
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5525641344438285,
+      "learning_rate": 9.56016192124598e-10,
+      "loss": 1.9042,
+      "step": 12927
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5554333474483573,
+      "learning_rate": 9.021674406889391e-10,
+      "loss": 1.8162,
+      "step": 12928
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5481007471207056,
+      "learning_rate": 8.498794856287573e-10,
+      "loss": 1.9002,
+      "step": 12929
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5302882133059988,
+      "learning_rate": 7.991523302108839e-10,
+      "loss": 2.0671,
+      "step": 12930
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.508259019058269,
+      "learning_rate": 7.499859775994544e-10,
+      "loss": 1.9218,
+      "step": 12931
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5611338073183052,
+      "learning_rate": 7.023804308670112e-10,
+      "loss": 1.8314,
+      "step": 12932
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5397483938328642,
+      "learning_rate": 6.563356929834009e-10,
+      "loss": 1.7805,
+      "step": 12933
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5279416094867868,
+      "learning_rate": 6.118517668268764e-10,
+      "loss": 2.0409,
+      "step": 12934
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5424628778356461,
+      "learning_rate": 5.689286551702199e-10,
+      "loss": 1.8519,
+      "step": 12935
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5465559577822899,
+      "learning_rate": 5.275663606973958e-10,
+      "loss": 1.8336,
+      "step": 12936
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5806456675669902,
+      "learning_rate": 4.877648859868966e-10,
+      "loss": 1.8561,
+      "step": 12937
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4915816112944975,
+      "learning_rate": 4.495242335256222e-10,
+      "loss": 2.074,
+      "step": 12938
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5238544703738774,
+      "learning_rate": 4.128444057033276e-10,
+      "loss": 1.8602,
+      "step": 12939
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5440292299768047,
+      "learning_rate": 3.7772540480707217e-10,
+      "loss": 1.8325,
+      "step": 12940
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.551005928058127,
+      "learning_rate": 3.44167233032322e-10,
+      "loss": 1.7959,
+      "step": 12941
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.556787073139806,
+      "learning_rate": 3.1216989246907193e-10,
+      "loss": 2.0652,
+      "step": 12942
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5290659297564679,
+      "learning_rate": 2.8173338512127445e-10,
+      "loss": 1.8164,
+      "step": 12943
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5088968466529408,
+      "learning_rate": 2.5285771288463546e-10,
+      "loss": 1.9008,
+      "step": 12944
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5558595456533397,
+      "learning_rate": 2.255428775632673e-10,
+      "loss": 1.8267,
+      "step": 12945
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5157993397084163,
+      "learning_rate": 1.9978888086413795e-10,
+      "loss": 2.0419,
+      "step": 12946
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5342265901200998,
+      "learning_rate": 1.7559572439429516e-10,
+      "loss": 1.8417,
+      "step": 12947
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5382101598106954,
+      "learning_rate": 1.529634096636423e-10,
+      "loss": 1.831,
+      "step": 12948
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5776222194645314,
+      "learning_rate": 1.3189193808493815e-10,
+      "loss": 1.8744,
+      "step": 12949
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5188828639426037,
+      "learning_rate": 1.12381310973797e-10,
+      "loss": 2.1104,
+      "step": 12950
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5279330734845321,
+      "learning_rate": 9.443152954868861e-11,
+      "loss": 1.8617,
+      "step": 12951
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.534475577994914,
+      "learning_rate": 7.804259493093824e-11,
+      "loss": 1.8338,
+      "step": 12952
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.542408574656326,
+      "learning_rate": 6.321450814195107e-11,
+      "loss": 1.8309,
+      "step": 12953
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5495583978918248,
+      "learning_rate": 4.994727011153888e-11,
+      "loss": 2.0997,
+      "step": 12954
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5372520342820748,
+      "learning_rate": 3.8240881664042275e-11,
+      "loss": 1.8736,
+      "step": 12955
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5011767299602624,
+      "learning_rate": 2.8095343529432882e-11,
+      "loss": 1.8717,
+      "step": 12956
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5497651832940555,
+      "learning_rate": 1.9510656346088953e-11,
+      "loss": 1.806,
+      "step": 12957
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5328127220106561,
+      "learning_rate": 1.248682064691753e-11,
+      "loss": 2.08,
+      "step": 12958
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5335184717219762,
+      "learning_rate": 7.0238368704567124e-12,
+      "loss": 1.8916,
+      "step": 12959
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5320203093793444,
+      "learning_rate": 3.121705355324522e-12,
+      "loss": 1.8631,
+      "step": 12960
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5560938197837585,
+      "learning_rate": 7.804263513211397e-13,
+      "loss": 1.8308,
+      "step": 12961
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.49048502620299095,
+      "learning_rate": 0.0,
+      "loss": 1.9668,
+      "step": 12962
+    },
+    {
+      "epoch": 1.0,
+      "step": 12962,
+      "total_flos": 0.0,
+      "train_loss": 0.23017303778275058,
+      "train_runtime": 7116.4219,
+      "train_samples_per_second": 1865.25,
+      "train_steps_per_second": 1.821
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 12962,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}