diff --git "a/Baseline/trainer_state.json" "b/Baseline/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/Baseline/trainer_state.json"
@@ -0,0 +1,53447 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999099150744839,
+  "eval_steps": 500,
+  "global_step": 7631,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.114334106445312,
+      "learning_rate": 4.366812227074236e-08,
+      "loss": 1.369,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.827465057373047,
+      "learning_rate": 8.733624454148472e-08,
+      "loss": 1.4185,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.591986656188965,
+      "learning_rate": 1.3100436681222707e-07,
+      "loss": 1.3864,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.766721725463867,
+      "learning_rate": 1.7467248908296944e-07,
+      "loss": 1.3985,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.706441879272461,
+      "learning_rate": 2.183406113537118e-07,
+      "loss": 1.379,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8751120567321777,
+      "learning_rate": 2.6200873362445414e-07,
+      "loss": 1.2724,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 15.535872459411621,
+      "learning_rate": 3.056768558951965e-07,
+      "loss": 1.3777,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.917743682861328,
+      "learning_rate": 3.4934497816593887e-07,
+      "loss": 1.4449,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 13.040949821472168,
+      "learning_rate": 3.930131004366813e-07,
+      "loss": 1.4199,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.67154312133789,
+      "learning_rate": 4.366812227074236e-07,
+      "loss": 1.4543,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 12.787934303283691,
+      "learning_rate": 4.80349344978166e-07,
+      "loss": 1.3675,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.8540256023406982,
+      "learning_rate": 5.240174672489083e-07,
+      "loss": 1.2868,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.9865798950195312,
+      "learning_rate": 5.676855895196507e-07,
+      "loss": 1.2863,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 11.627809524536133,
+      "learning_rate": 6.11353711790393e-07,
+      "loss": 1.3723,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 10.4605131149292,
+      "learning_rate": 6.550218340611354e-07,
+      "loss": 1.4158,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.573136329650879,
+      "learning_rate": 6.986899563318777e-07,
+      "loss": 1.3623,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.221022605895996,
+      "learning_rate": 7.423580786026202e-07,
+      "loss": 1.4387,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.131499290466309,
+      "learning_rate": 7.860262008733626e-07,
+      "loss": 1.4321,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.746452808380127,
+      "learning_rate": 8.296943231441049e-07,
+      "loss": 1.3574,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.5743231773376465,
+      "learning_rate": 8.733624454148472e-07,
+      "loss": 1.345,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.95228385925293,
+      "learning_rate": 9.170305676855896e-07,
+      "loss": 1.3388,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.297760009765625,
+      "learning_rate": 9.60698689956332e-07,
+      "loss": 1.3601,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.944675445556641,
+      "learning_rate": 1.0043668122270742e-06,
+      "loss": 1.3503,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.965779781341553,
+      "learning_rate": 1.0480349344978166e-06,
+      "loss": 1.3378,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.62257981300354,
+      "learning_rate": 1.091703056768559e-06,
+      "loss": 1.2562,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.957221031188965,
+      "learning_rate": 1.1353711790393014e-06,
+      "loss": 1.3447,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5104503631591797,
+      "learning_rate": 1.1790393013100437e-06,
+      "loss": 1.2786,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.110462188720703,
+      "learning_rate": 1.222707423580786e-06,
+      "loss": 1.3664,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.548264980316162,
+      "learning_rate": 1.2663755458515283e-06,
+      "loss": 1.302,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.483218193054199,
+      "learning_rate": 1.3100436681222709e-06,
+      "loss": 1.2981,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.3175454139709473,
+      "learning_rate": 1.3537117903930134e-06,
+      "loss": 1.3088,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.815730571746826,
+      "learning_rate": 1.3973799126637555e-06,
+      "loss": 1.2298,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 3.017672538757324,
+      "learning_rate": 1.441048034934498e-06,
+      "loss": 1.258,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.487450122833252,
+      "learning_rate": 1.4847161572052403e-06,
+      "loss": 1.2196,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5065386295318604,
+      "learning_rate": 1.5283842794759826e-06,
+      "loss": 1.2706,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.5095653533935547,
+      "learning_rate": 1.5720524017467252e-06,
+      "loss": 1.2906,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.880254030227661,
+      "learning_rate": 1.6157205240174673e-06,
+      "loss": 1.2518,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 2.3712377548217773,
+      "learning_rate": 1.6593886462882098e-06,
+      "loss": 1.2829,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.4269745349884033,
+      "learning_rate": 1.703056768558952e-06,
+      "loss": 1.1674,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.411029100418091,
+      "learning_rate": 1.7467248908296944e-06,
+      "loss": 1.2548,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.6733808517456055,
+      "learning_rate": 1.790393013100437e-06,
+      "loss": 1.3065,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.349301338195801,
+      "learning_rate": 1.8340611353711792e-06,
+      "loss": 1.2092,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.260509729385376,
+      "learning_rate": 1.8777292576419216e-06,
+      "loss": 1.1718,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.1078643798828125,
+      "learning_rate": 1.921397379912664e-06,
+      "loss": 1.2152,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0145347118377686,
+      "learning_rate": 1.965065502183406e-06,
+      "loss": 1.1948,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0033299922943115,
+      "learning_rate": 2.0087336244541485e-06,
+      "loss": 1.1944,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0396435260772705,
+      "learning_rate": 2.052401746724891e-06,
+      "loss": 1.2126,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7730211019515991,
+      "learning_rate": 2.096069868995633e-06,
+      "loss": 1.1416,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.73203444480896,
+      "learning_rate": 2.1397379912663756e-06,
+      "loss": 1.261,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.589874267578125,
+      "learning_rate": 2.183406113537118e-06,
+      "loss": 1.2721,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7610877752304077,
+      "learning_rate": 2.2270742358078603e-06,
+      "loss": 1.1456,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.764288306236267,
+      "learning_rate": 2.270742358078603e-06,
+      "loss": 1.1363,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.199096441268921,
+      "learning_rate": 2.3144104803493453e-06,
+      "loss": 1.2663,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.701950192451477,
+      "learning_rate": 2.3580786026200874e-06,
+      "loss": 1.1818,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7686148881912231,
+      "learning_rate": 2.40174672489083e-06,
+      "loss": 1.1661,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7591885328292847,
+      "learning_rate": 2.445414847161572e-06,
+      "loss": 1.1473,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0681238174438477,
+      "learning_rate": 2.4890829694323146e-06,
+      "loss": 1.1038,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7116835117340088,
+      "learning_rate": 2.5327510917030567e-06,
+      "loss": 1.1757,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5723899602890015,
+      "learning_rate": 2.576419213973799e-06,
+      "loss": 1.1591,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6173808574676514,
+      "learning_rate": 2.6200873362445417e-06,
+      "loss": 1.1065,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.0891618728637695,
+      "learning_rate": 2.6637554585152842e-06,
+      "loss": 1.1831,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4882575273513794,
+      "learning_rate": 2.7074235807860268e-06,
+      "loss": 1.1489,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6471997499465942,
+      "learning_rate": 2.7510917030567684e-06,
+      "loss": 1.2598,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6566061973571777,
+      "learning_rate": 2.794759825327511e-06,
+      "loss": 1.2368,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.527205228805542,
+      "learning_rate": 2.8384279475982535e-06,
+      "loss": 1.1786,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6906688213348389,
+      "learning_rate": 2.882096069868996e-06,
+      "loss": 1.1281,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.850226640701294,
+      "learning_rate": 2.9257641921397385e-06,
+      "loss": 1.087,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5384325981140137,
+      "learning_rate": 2.9694323144104806e-06,
+      "loss": 1.1091,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7883366346359253,
+      "learning_rate": 3.0131004366812227e-06,
+      "loss": 1.1442,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5563234090805054,
+      "learning_rate": 3.0567685589519653e-06,
+      "loss": 1.1118,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.454598307609558,
+      "learning_rate": 3.100436681222708e-06,
+      "loss": 1.0765,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6705197095870972,
+      "learning_rate": 3.1441048034934503e-06,
+      "loss": 1.1367,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.466437816619873,
+      "learning_rate": 3.1877729257641924e-06,
+      "loss": 1.1073,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.506177306175232,
+      "learning_rate": 3.2314410480349345e-06,
+      "loss": 1.1264,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5004005432128906,
+      "learning_rate": 3.275109170305677e-06,
+      "loss": 1.0939,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4757541418075562,
+      "learning_rate": 3.3187772925764196e-06,
+      "loss": 1.1508,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.184744119644165,
+      "learning_rate": 3.362445414847162e-06,
+      "loss": 1.2361,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5279091596603394,
+      "learning_rate": 3.406113537117904e-06,
+      "loss": 1.0901,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.100316047668457,
+      "learning_rate": 3.4497816593886467e-06,
+      "loss": 1.2447,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0338648557662964,
+      "learning_rate": 3.493449781659389e-06,
+      "loss": 1.2331,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6067360639572144,
+      "learning_rate": 3.5371179039301313e-06,
+      "loss": 1.1379,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.632983922958374,
+      "learning_rate": 3.580786026200874e-06,
+      "loss": 1.1066,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8384180068969727,
+      "learning_rate": 3.624454148471616e-06,
+      "loss": 1.1371,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8944818377494812,
+      "learning_rate": 3.6681222707423585e-06,
+      "loss": 1.2308,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.542994737625122,
+      "learning_rate": 3.7117903930131006e-06,
+      "loss": 1.0687,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8230865001678467,
+      "learning_rate": 3.755458515283843e-06,
+      "loss": 1.2156,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5007416009902954,
+      "learning_rate": 3.7991266375545856e-06,
+      "loss": 1.1085,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8410069942474365,
+      "learning_rate": 3.842794759825328e-06,
+      "loss": 1.223,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7863571047782898,
+      "learning_rate": 3.88646288209607e-06,
+      "loss": 1.2075,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7148558497428894,
+      "learning_rate": 3.930131004366812e-06,
+      "loss": 1.2217,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.697709321975708,
+      "learning_rate": 3.9737991266375545e-06,
+      "loss": 1.212,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.6653743982315063,
+      "learning_rate": 4.017467248908297e-06,
+      "loss": 1.2249,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.6029561758041382,
+      "learning_rate": 4.0611353711790395e-06,
+      "loss": 1.0649,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.653239369392395,
+      "learning_rate": 4.104803493449782e-06,
+      "loss": 1.0887,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7746754884719849,
+      "learning_rate": 4.1484716157205246e-06,
+      "loss": 1.0646,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.699591338634491,
+      "learning_rate": 4.192139737991266e-06,
+      "loss": 1.2141,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4821548461914062,
+      "learning_rate": 4.235807860262009e-06,
+      "loss": 1.056,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4838083982467651,
+      "learning_rate": 4.279475982532751e-06,
+      "loss": 1.053,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7875638008117676,
+      "learning_rate": 4.323144104803494e-06,
+      "loss": 1.2137,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5186035633087158,
+      "learning_rate": 4.366812227074236e-06,
+      "loss": 1.0808,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4975703954696655,
+      "learning_rate": 4.410480349344978e-06,
+      "loss": 1.032,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4499491453170776,
+      "learning_rate": 4.4541484716157205e-06,
+      "loss": 1.0501,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.484489917755127,
+      "learning_rate": 4.497816593886463e-06,
+      "loss": 1.1067,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3885366916656494,
+      "learning_rate": 4.541484716157206e-06,
+      "loss": 1.0859,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3516345024108887,
+      "learning_rate": 4.585152838427948e-06,
+      "loss": 1.0501,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8306176662445068,
+      "learning_rate": 4.628820960698691e-06,
+      "loss": 1.1823,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.171459913253784,
+      "learning_rate": 4.672489082969432e-06,
+      "loss": 1.0203,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7086100578308105,
+      "learning_rate": 4.716157205240175e-06,
+      "loss": 1.2144,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.9315613508224487,
+      "learning_rate": 4.759825327510917e-06,
+      "loss": 1.0498,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.485727310180664,
+      "learning_rate": 4.80349344978166e-06,
+      "loss": 1.0778,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.700325846672058,
+      "learning_rate": 4.847161572052402e-06,
+      "loss": 1.0652,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4874215126037598,
+      "learning_rate": 4.890829694323144e-06,
+      "loss": 1.118,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5805107355117798,
+      "learning_rate": 4.934497816593887e-06,
+      "loss": 1.0631,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0520724058151245,
+      "learning_rate": 4.978165938864629e-06,
+      "loss": 1.2091,
+      "step": 114
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8394785523414612,
+      "learning_rate": 5.021834061135371e-06,
+      "loss": 1.1918,
+      "step": 115
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5445523262023926,
+      "learning_rate": 5.065502183406113e-06,
+      "loss": 1.0477,
+      "step": 116
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6541285514831543,
+      "learning_rate": 5.109170305676856e-06,
+      "loss": 1.0459,
+      "step": 117
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8316869735717773,
+      "learning_rate": 5.152838427947598e-06,
+      "loss": 1.21,
+      "step": 118
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4341267347335815,
+      "learning_rate": 5.196506550218341e-06,
+      "loss": 1.0812,
+      "step": 119
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5050230026245117,
+      "learning_rate": 5.2401746724890834e-06,
+      "loss": 1.0665,
+      "step": 120
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.096174478530884,
+      "learning_rate": 5.283842794759826e-06,
+      "loss": 1.0662,
+      "step": 121
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4701011180877686,
+      "learning_rate": 5.3275109170305685e-06,
+      "loss": 1.0404,
+      "step": 122
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5522528886795044,
+      "learning_rate": 5.371179039301311e-06,
+      "loss": 1.0034,
+      "step": 123
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3812271356582642,
+      "learning_rate": 5.4148471615720535e-06,
+      "loss": 1.0769,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7800785303115845,
+      "learning_rate": 5.458515283842796e-06,
+      "loss": 1.0727,
+      "step": 125
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9726714491844177,
+      "learning_rate": 5.502183406113537e-06,
+      "loss": 1.2173,
+      "step": 126
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4145596027374268,
+      "learning_rate": 5.545851528384279e-06,
+      "loss": 1.0082,
+      "step": 127
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7721313238143921,
+      "learning_rate": 5.589519650655022e-06,
+      "loss": 1.1921,
+      "step": 128
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5143922567367554,
+      "learning_rate": 5.6331877729257645e-06,
+      "loss": 1.1072,
+      "step": 129
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6113042831420898,
+      "learning_rate": 5.676855895196507e-06,
+      "loss": 1.0638,
+      "step": 130
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9427786469459534,
+      "learning_rate": 5.7205240174672495e-06,
+      "loss": 1.1999,
+      "step": 131
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8960140347480774,
+      "learning_rate": 5.764192139737992e-06,
+      "loss": 1.2015,
+      "step": 132
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4322227239608765,
+      "learning_rate": 5.8078602620087346e-06,
+      "loss": 1.0406,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6556090712547302,
+      "learning_rate": 5.851528384279477e-06,
+      "loss": 1.1988,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7828678488731384,
+      "learning_rate": 5.89519650655022e-06,
+      "loss": 1.2024,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.537335753440857,
+      "learning_rate": 5.938864628820961e-06,
+      "loss": 1.0105,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.53865647315979,
+      "learning_rate": 5.982532751091703e-06,
+      "loss": 1.0589,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8548715114593506,
+      "learning_rate": 6.0262008733624455e-06,
+      "loss": 1.2223,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5609651803970337,
+      "learning_rate": 6.069868995633188e-06,
+      "loss": 1.0516,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4039651155471802,
+      "learning_rate": 6.1135371179039305e-06,
+      "loss": 1.0479,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.515905737876892,
+      "learning_rate": 6.157205240174673e-06,
+      "loss": 1.0526,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4653518199920654,
+      "learning_rate": 6.200873362445416e-06,
+      "loss": 1.0059,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9291254281997681,
+      "learning_rate": 6.244541484716158e-06,
+      "loss": 1.2051,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7633457183837891,
+      "learning_rate": 6.288209606986901e-06,
+      "loss": 1.204,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5056383609771729,
+      "learning_rate": 6.331877729257643e-06,
+      "loss": 1.0234,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.796854019165039,
+      "learning_rate": 6.375545851528385e-06,
+      "loss": 0.9925,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5027351379394531,
+      "learning_rate": 6.419213973799127e-06,
+      "loss": 1.0867,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6836470365524292,
+      "learning_rate": 6.462882096069869e-06,
+      "loss": 1.0116,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.469018578529358,
+      "learning_rate": 6.5065502183406116e-06,
+      "loss": 1.0149,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4953076839447021,
+      "learning_rate": 6.550218340611354e-06,
+      "loss": 1.0036,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4721965789794922,
+      "learning_rate": 6.593886462882097e-06,
+      "loss": 1.0949,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3856313228607178,
+      "learning_rate": 6.637554585152839e-06,
+      "loss": 0.9974,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3665705919265747,
+      "learning_rate": 6.681222707423582e-06,
+      "loss": 1.2098,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0376454591751099,
+      "learning_rate": 6.724890829694324e-06,
+      "loss": 1.1764,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.580572485923767,
+      "learning_rate": 6.768558951965066e-06,
+      "loss": 1.0552,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5477577447891235,
+      "learning_rate": 6.812227074235808e-06,
+      "loss": 1.036,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6403752565383911,
+      "learning_rate": 6.855895196506551e-06,
+      "loss": 1.0335,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8613828420639038,
+      "learning_rate": 6.8995633187772934e-06,
+      "loss": 1.1675,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7371251583099365,
+      "learning_rate": 6.943231441048035e-06,
+      "loss": 1.0755,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5106327533721924,
+      "learning_rate": 6.986899563318778e-06,
+      "loss": 1.0319,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6614861488342285,
+      "learning_rate": 7.03056768558952e-06,
+      "loss": 1.0599,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2880473136901855,
+      "learning_rate": 7.074235807860263e-06,
+      "loss": 1.1821,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.7639179229736328,
+      "learning_rate": 7.117903930131005e-06,
+      "loss": 1.0037,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.8456499576568604,
+      "learning_rate": 7.161572052401748e-06,
+      "loss": 1.021,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5289312601089478,
+      "learning_rate": 7.205240174672489e-06,
+      "loss": 1.036,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5389316082000732,
+      "learning_rate": 7.248908296943232e-06,
+      "loss": 1.0436,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.514217734336853,
+      "learning_rate": 7.2925764192139745e-06,
+      "loss": 0.9855,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.663690209388733,
+      "learning_rate": 7.336244541484717e-06,
+      "loss": 0.9891,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5568764209747314,
+      "learning_rate": 7.3799126637554595e-06,
+      "loss": 1.0307,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4854170083999634,
+      "learning_rate": 7.423580786026201e-06,
+      "loss": 1.0357,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.47286856174469,
+      "learning_rate": 7.467248908296944e-06,
+      "loss": 0.999,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6005613803863525,
+      "learning_rate": 7.510917030567686e-06,
+      "loss": 1.0289,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.484571933746338,
+      "learning_rate": 7.554585152838429e-06,
+      "loss": 1.1862,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.2075533866882324,
+      "learning_rate": 7.598253275109171e-06,
+      "loss": 1.2108,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7499205470085144,
+      "learning_rate": 7.641921397379913e-06,
+      "loss": 1.2121,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.807909369468689,
+      "learning_rate": 7.685589519650655e-06,
+      "loss": 1.0457,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3225926160812378,
+      "learning_rate": 7.729257641921398e-06,
+      "loss": 1.1975,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.6531331539154053,
+      "learning_rate": 7.77292576419214e-06,
+      "loss": 1.0691,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5647892951965332,
+      "learning_rate": 7.816593886462883e-06,
+      "loss": 1.0226,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4981244802474976,
+      "learning_rate": 7.860262008733624e-06,
+      "loss": 1.1632,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5906938314437866,
+      "learning_rate": 7.903930131004366e-06,
+      "loss": 1.024,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5964020490646362,
+      "learning_rate": 7.947598253275109e-06,
+      "loss": 1.0567,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.4128762483596802,
+      "learning_rate": 7.991266375545851e-06,
+      "loss": 1.0098,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.5638844966888428,
+      "learning_rate": 8.034934497816594e-06,
+      "loss": 1.0003,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.845028281211853,
+      "learning_rate": 8.078602620087337e-06,
+      "loss": 1.1834,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7634133696556091,
+      "learning_rate": 8.122270742358079e-06,
+      "loss": 1.1995,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7181407809257507,
+      "learning_rate": 8.165938864628822e-06,
+      "loss": 1.1768,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7330619096755981,
+      "learning_rate": 8.209606986899564e-06,
+      "loss": 1.1663,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8280870914459229,
+      "learning_rate": 8.253275109170307e-06,
+      "loss": 1.18,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 2.0398659706115723,
+      "learning_rate": 8.296943231441049e-06,
+      "loss": 1.0643,
+      "step": 190
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8338253498077393,
+      "learning_rate": 8.34061135371179e-06,
+      "loss": 1.0534,
+      "step": 191
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4531362056732178,
+      "learning_rate": 8.384279475982532e-06,
+      "loss": 1.0017,
+      "step": 192
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5866163969039917,
+      "learning_rate": 8.427947598253275e-06,
+      "loss": 1.0174,
+      "step": 193
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.2356362342834473,
+      "learning_rate": 8.471615720524018e-06,
+      "loss": 1.0106,
+      "step": 194
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6008371114730835,
+      "learning_rate": 8.51528384279476e-06,
+      "loss": 1.0096,
+      "step": 195
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5470542907714844,
+      "learning_rate": 8.558951965065503e-06,
+      "loss": 0.9423,
+      "step": 196
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6005938053131104,
+      "learning_rate": 8.602620087336245e-06,
+      "loss": 0.9953,
+      "step": 197
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5316663980484009,
+      "learning_rate": 8.646288209606988e-06,
+      "loss": 1.0275,
+      "step": 198
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.8333423137664795,
+      "learning_rate": 8.68995633187773e-06,
+      "loss": 0.9841,
+      "step": 199
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3432679176330566,
+      "learning_rate": 8.733624454148473e-06,
+      "loss": 1.208,
+      "step": 200
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5995087623596191,
+      "learning_rate": 8.777292576419215e-06,
+      "loss": 0.9943,
+      "step": 201
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6491607427597046,
+      "learning_rate": 8.820960698689956e-06,
+      "loss": 0.9287,
+      "step": 202
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5003896951675415,
+      "learning_rate": 8.864628820960699e-06,
+      "loss": 1.0114,
+      "step": 203
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9981769919395447,
+      "learning_rate": 8.908296943231441e-06,
+      "loss": 1.1633,
+      "step": 204
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.1375489234924316,
+      "learning_rate": 8.951965065502184e-06,
+      "loss": 1.0395,
+      "step": 205
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7915791273117065,
+      "learning_rate": 8.995633187772926e-06,
+      "loss": 1.1763,
+      "step": 206
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.7662112712860107,
+      "learning_rate": 9.039301310043669e-06,
+      "loss": 1.0048,
+      "step": 207
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5124331712722778,
+      "learning_rate": 9.082969432314411e-06,
+      "loss": 0.9883,
+      "step": 208
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4944243431091309,
+      "learning_rate": 9.126637554585154e-06,
+      "loss": 1.0164,
+      "step": 209
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6241964101791382,
+      "learning_rate": 9.170305676855896e-06,
+      "loss": 0.9605,
+      "step": 210
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4862587451934814,
+      "learning_rate": 9.213973799126639e-06,
+      "loss": 1.0327,
+      "step": 211
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.676472783088684,
+      "learning_rate": 9.257641921397381e-06,
+      "loss": 1.0313,
+      "step": 212
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6227633953094482,
+      "learning_rate": 9.301310043668122e-06,
+      "loss": 1.0147,
+      "step": 213
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0895848274230957,
+      "learning_rate": 9.344978165938865e-06,
+      "loss": 1.1834,
+      "step": 214
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6997036933898926,
+      "learning_rate": 9.388646288209607e-06,
+      "loss": 1.0454,
+      "step": 215
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8218080997467041,
+      "learning_rate": 9.43231441048035e-06,
+      "loss": 1.171,
+      "step": 216
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8271385431289673,
+      "learning_rate": 9.475982532751092e-06,
+      "loss": 1.1599,
+      "step": 217
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5108028650283813,
+      "learning_rate": 9.519650655021835e-06,
+      "loss": 1.0272,
+      "step": 218
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5309537649154663,
+      "learning_rate": 9.563318777292577e-06,
+      "loss": 0.9693,
+      "step": 219
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4888508319854736,
+      "learning_rate": 9.60698689956332e-06,
+      "loss": 1.0062,
+      "step": 220
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5289844274520874,
+      "learning_rate": 9.650655021834062e-06,
+      "loss": 0.9666,
+      "step": 221
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5640686750411987,
+      "learning_rate": 9.694323144104805e-06,
+      "loss": 0.9836,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5706342458724976,
+      "learning_rate": 9.737991266375547e-06,
+      "loss": 0.9839,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.25709068775177,
+      "learning_rate": 9.781659388646288e-06,
+      "loss": 1.1755,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6188373565673828,
+      "learning_rate": 9.82532751091703e-06,
+      "loss": 1.0019,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5536643266677856,
+      "learning_rate": 9.868995633187773e-06,
+      "loss": 1.0251,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.620069980621338,
+      "learning_rate": 9.912663755458516e-06,
+      "loss": 1.0491,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.010317087173462,
+      "learning_rate": 9.956331877729258e-06,
+      "loss": 1.1699,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5573339462280273,
+      "learning_rate": 1e-05,
+      "loss": 0.9156,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3798346519470215,
+      "learning_rate": 9.999999549658898e-06,
+      "loss": 0.9482,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.566415548324585,
+      "learning_rate": 9.999998198635674e-06,
+      "loss": 0.9987,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9352176189422607,
+      "learning_rate": 9.999995946930572e-06,
+      "loss": 1.1749,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4207650423049927,
+      "learning_rate": 9.999992794543995e-06,
+      "loss": 1.0131,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4835906028747559,
+      "learning_rate": 9.999988741476512e-06,
+      "loss": 0.8996,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5416728258132935,
+      "learning_rate": 9.999983787728855e-06,
+      "loss": 1.0404,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3962267637252808,
+      "learning_rate": 9.999977933301912e-06,
+      "loss": 0.9977,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9215784072875977,
+      "learning_rate": 9.999971178196743e-06,
+      "loss": 1.1827,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4465010166168213,
+      "learning_rate": 9.999963522414561e-06,
+      "loss": 1.044,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4311773777008057,
+      "learning_rate": 9.999954965956747e-06,
+      "loss": 0.9418,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3910574913024902,
+      "learning_rate": 9.999945508824842e-06,
+      "loss": 0.9762,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7730720639228821,
+      "learning_rate": 9.999935151020548e-06,
+      "loss": 1.1719,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5372891426086426,
+      "learning_rate": 9.999923892545734e-06,
+      "loss": 1.0308,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3889086246490479,
+      "learning_rate": 9.999911733402426e-06,
+      "loss": 1.0016,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7555667757987976,
+      "learning_rate": 9.999898673592812e-06,
+      "loss": 1.1465,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.537814974784851,
+      "learning_rate": 9.999884713119248e-06,
+      "loss": 0.9952,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.466597080230713,
+      "learning_rate": 9.99986985198425e-06,
+      "loss": 0.9597,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.475334882736206,
+      "learning_rate": 9.99985409019049e-06,
+      "loss": 0.968,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8309423923492432,
+      "learning_rate": 9.99983742774081e-06,
+      "loss": 1.1604,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.398297905921936,
+      "learning_rate": 9.999819864638214e-06,
+      "loss": 1.0342,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.467044472694397,
+      "learning_rate": 9.999801400885862e-06,
+      "loss": 1.028,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4388614892959595,
+      "learning_rate": 9.999782036487082e-06,
+      "loss": 1.0409,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.0385165214538574,
+      "learning_rate": 9.99976177144536e-06,
+      "loss": 1.1597,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7305343747138977,
+      "learning_rate": 9.99974060576435e-06,
+      "loss": 1.1648,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8584312200546265,
+      "learning_rate": 9.99971853944786e-06,
+      "loss": 1.1762,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.9326953291893005,
+      "learning_rate": 9.99969557249987e-06,
+      "loss": 1.1722,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6565356254577637,
+      "learning_rate": 9.999671704924514e-06,
+      "loss": 0.9538,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4468778371810913,
+      "learning_rate": 9.999646936726094e-06,
+      "loss": 0.9219,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4158809185028076,
+      "learning_rate": 9.999621267909068e-06,
+      "loss": 0.9934,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.540346622467041,
+      "learning_rate": 9.999594698478063e-06,
+      "loss": 1.0143,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4745497703552246,
+      "learning_rate": 9.999567228437864e-06,
+      "loss": 0.9885,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4921648502349854,
+      "learning_rate": 9.999538857793419e-06,
+      "loss": 1.0347,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.5970346927642822,
+      "learning_rate": 9.999509586549837e-06,
+      "loss": 0.9819,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.6705199480056763,
+      "learning_rate": 9.999479414712395e-06,
+      "loss": 1.0132,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.3813551664352417,
+      "learning_rate": 9.999448342286526e-06,
+      "loss": 0.9779,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.4913301467895508,
+      "learning_rate": 9.999416369277825e-06,
+      "loss": 1.003,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 2.0298080444335938,
+      "learning_rate": 9.999383495692056e-06,
+      "loss": 0.9534,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 1.474820852279663,
+      "learning_rate": 9.999349721535136e-06,
+      "loss": 0.9667,
+      "step": 267
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6020655632019043,
+      "learning_rate": 9.999315046813153e-06,
+      "loss": 1.0276,
+      "step": 268
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3962788581848145,
+      "learning_rate": 9.99927947153235e-06,
+      "loss": 0.9295,
+      "step": 269
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5710713863372803,
+      "learning_rate": 9.99924299569914e-06,
+      "loss": 0.9859,
+      "step": 270
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4353978633880615,
+      "learning_rate": 9.999205619320087e-06,
+      "loss": 0.9444,
+      "step": 271
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4774583578109741,
+      "learning_rate": 9.99916734240193e-06,
+      "loss": 1.1723,
+      "step": 272
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.079396367073059,
+      "learning_rate": 9.999128164951559e-06,
+      "loss": 1.1835,
+      "step": 273
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8113067746162415,
+      "learning_rate": 9.999088086976034e-06,
+      "loss": 1.1473,
+      "step": 274
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.150009274482727,
+      "learning_rate": 9.999047108482576e-06,
+      "loss": 1.1731,
+      "step": 275
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.770877718925476,
+      "learning_rate": 9.999005229478561e-06,
+      "loss": 1.0045,
+      "step": 276
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6352653503417969,
+      "learning_rate": 9.998962449971542e-06,
+      "loss": 0.9938,
+      "step": 277
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.282896637916565,
+      "learning_rate": 9.998918769969216e-06,
+      "loss": 1.1578,
+      "step": 278
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5645115375518799,
+      "learning_rate": 9.998874189479457e-06,
+      "loss": 1.0212,
+      "step": 279
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5804671049118042,
+      "learning_rate": 9.998828708510295e-06,
+      "loss": 0.9684,
+      "step": 280
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0627777576446533,
+      "learning_rate": 9.99878232706992e-06,
+      "loss": 1.1412,
+      "step": 281
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6583425998687744,
+      "learning_rate": 9.99873504516669e-06,
+      "loss": 0.9803,
+      "step": 282
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5497288703918457,
+      "learning_rate": 9.998686862809119e-06,
+      "loss": 0.9956,
+      "step": 283
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8791226744651794,
+      "learning_rate": 9.998637780005889e-06,
+      "loss": 1.1827,
+      "step": 284
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8969788551330566,
+      "learning_rate": 9.998587796765842e-06,
+      "loss": 0.9741,
+      "step": 285
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5783872604370117,
+      "learning_rate": 9.99853691309798e-06,
+      "loss": 1.0512,
+      "step": 286
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.41765558719635,
+      "learning_rate": 9.998485129011468e-06,
+      "loss": 0.9823,
+      "step": 287
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4332833290100098,
+      "learning_rate": 9.998432444515638e-06,
+      "loss": 1.0656,
+      "step": 288
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3859519958496094,
+      "learning_rate": 9.998378859619978e-06,
+      "loss": 0.9629,
+      "step": 289
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4948663711547852,
+      "learning_rate": 9.99832437433414e-06,
+      "loss": 0.9897,
+      "step": 290
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4854594469070435,
+      "learning_rate": 9.99826898866794e-06,
+      "loss": 0.9975,
+      "step": 291
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3559926748275757,
+      "learning_rate": 9.998212702631355e-06,
+      "loss": 0.9924,
+      "step": 292
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7533085346221924,
+      "learning_rate": 9.998155516234523e-06,
+      "loss": 0.9473,
+      "step": 293
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.603756070137024,
+      "learning_rate": 9.998097429487748e-06,
+      "loss": 0.9708,
+      "step": 294
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.3760790824890137,
+      "learning_rate": 9.99803844240149e-06,
+      "loss": 0.9189,
+      "step": 295
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4437580108642578,
+      "learning_rate": 9.997978554986376e-06,
+      "loss": 0.9729,
+      "step": 296
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4473764896392822,
+      "learning_rate": 9.997917767253195e-06,
+      "loss": 0.9616,
+      "step": 297
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.121743083000183,
+      "learning_rate": 9.997856079212896e-06,
+      "loss": 1.1513,
+      "step": 298
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.908117949962616,
+      "learning_rate": 9.997793490876593e-06,
+      "loss": 1.161,
+      "step": 299
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6347765922546387,
+      "learning_rate": 9.997730002255559e-06,
+      "loss": 1.0198,
+      "step": 300
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5136479139328003,
+      "learning_rate": 9.99766561336123e-06,
+      "loss": 1.0278,
+      "step": 301
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8063300848007202,
+      "learning_rate": 9.997600324205204e-06,
+      "loss": 0.9889,
+      "step": 302
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5346884727478027,
+      "learning_rate": 9.997534134799244e-06,
+      "loss": 0.9751,
+      "step": 303
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8673632144927979,
+      "learning_rate": 9.997467045155276e-06,
+      "loss": 1.1586,
+      "step": 304
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5194854736328125,
+      "learning_rate": 9.997399055285379e-06,
+      "loss": 1.1744,
+      "step": 305
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7975331544876099,
+      "learning_rate": 9.997330165201805e-06,
+      "loss": 0.9435,
+      "step": 306
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.999853491783142,
+      "learning_rate": 9.99726037491696e-06,
+      "loss": 0.9459,
+      "step": 307
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1616649627685547,
+      "learning_rate": 9.997189684443418e-06,
+      "loss": 1.1819,
+      "step": 308
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.182125449180603,
+      "learning_rate": 9.997118093793912e-06,
+      "loss": 1.1742,
+      "step": 309
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9045939445495605,
+      "learning_rate": 9.99704560298134e-06,
+      "loss": 1.1707,
+      "step": 310
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8500620722770691,
+      "learning_rate": 9.996972212018758e-06,
+      "loss": 1.1865,
+      "step": 311
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.013197660446167,
+      "learning_rate": 9.996897920919388e-06,
+      "loss": 1.1602,
+      "step": 312
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.111116886138916,
+      "learning_rate": 9.996822729696613e-06,
+      "loss": 0.9365,
+      "step": 313
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7581537961959839,
+      "learning_rate": 9.996746638363976e-06,
+      "loss": 0.971,
+      "step": 314
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7266359329223633,
+      "learning_rate": 9.996669646935185e-06,
+      "loss": 0.9092,
+      "step": 315
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6151282787322998,
+      "learning_rate": 9.996591755424105e-06,
+      "loss": 1.0398,
+      "step": 316
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.797358512878418,
+      "learning_rate": 9.996512963844774e-06,
+      "loss": 0.9798,
+      "step": 317
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.583309531211853,
+      "learning_rate": 9.996433272211381e-06,
+      "loss": 0.9758,
+      "step": 318
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4960753917694092,
+      "learning_rate": 9.996352680538281e-06,
+      "loss": 0.9373,
+      "step": 319
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5167077779769897,
+      "learning_rate": 9.996271188839994e-06,
+      "loss": 0.9902,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8606681823730469,
+      "learning_rate": 9.996188797131197e-06,
+      "loss": 1.1761,
+      "step": 321
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.8522230386734009,
+      "learning_rate": 9.996105505426733e-06,
+      "loss": 0.9795,
+      "step": 322
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6095331907272339,
+      "learning_rate": 9.996021313741606e-06,
+      "loss": 1.0061,
+      "step": 323
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5219907760620117,
+      "learning_rate": 9.995936222090982e-06,
+      "loss": 0.9424,
+      "step": 324
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.5091710090637207,
+      "learning_rate": 9.995850230490188e-06,
+      "loss": 0.9604,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.7116498947143555,
+      "learning_rate": 9.995763338954715e-06,
+      "loss": 0.955,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.638999104499817,
+      "learning_rate": 9.995675547500216e-06,
+      "loss": 0.9345,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.0840184688568115,
+      "learning_rate": 9.995586856142507e-06,
+      "loss": 1.159,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6375458240509033,
+      "learning_rate": 9.99549726489756e-06,
+      "loss": 0.9632,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.6365212202072144,
+      "learning_rate": 9.995406773781518e-06,
+      "loss": 0.9508,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4284478425979614,
+      "learning_rate": 9.995315382810677e-06,
+      "loss": 0.9323,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8914143443107605,
+      "learning_rate": 9.995223092001505e-06,
+      "loss": 1.1592,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.425621509552002,
+      "learning_rate": 9.995129901370624e-06,
+      "loss": 0.8805,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.556113600730896,
+      "learning_rate": 9.995035810934821e-06,
+      "loss": 1.003,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7444076538085938,
+      "learning_rate": 9.994940820711045e-06,
+      "loss": 1.1372,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6744773387908936,
+      "learning_rate": 9.99484493071641e-06,
+      "loss": 1.1524,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.436039924621582,
+      "learning_rate": 9.994748140968187e-06,
+      "loss": 0.9453,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.421215534210205,
+      "learning_rate": 9.99465045148381e-06,
+      "loss": 0.944,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.437152624130249,
+      "learning_rate": 9.994551862280879e-06,
+      "loss": 1.0471,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.2827507257461548,
+      "learning_rate": 9.994452373377152e-06,
+      "loss": 0.9577,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.1961654424667358,
+      "learning_rate": 9.994351984790552e-06,
+      "loss": 1.1509,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8754168748855591,
+      "learning_rate": 9.994250696539162e-06,
+      "loss": 1.1664,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.4683219194412231,
+      "learning_rate": 9.994148508641228e-06,
+      "loss": 0.8941,
+      "step": 343
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4538320302963257,
+      "learning_rate": 9.994045421115155e-06,
+      "loss": 0.9556,
+      "step": 344
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6988738775253296,
+      "learning_rate": 9.993941433979516e-06,
+      "loss": 0.9773,
+      "step": 345
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6813734769821167,
+      "learning_rate": 9.993836547253041e-06,
+      "loss": 1.1855,
+      "step": 346
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5717681646347046,
+      "learning_rate": 9.993730760954628e-06,
+      "loss": 1.151,
+      "step": 347
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5430364608764648,
+      "learning_rate": 9.993624075103328e-06,
+      "loss": 1.0315,
+      "step": 348
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.575338363647461,
+      "learning_rate": 9.99351648971836e-06,
+      "loss": 0.9762,
+      "step": 349
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.272899866104126,
+      "learning_rate": 9.993408004819104e-06,
+      "loss": 1.1379,
+      "step": 350
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4215993881225586,
+      "learning_rate": 9.993298620425106e-06,
+      "loss": 1.0101,
+      "step": 351
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.461519479751587,
+      "learning_rate": 9.993188336556065e-06,
+      "loss": 0.8754,
+      "step": 352
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.094693660736084,
+      "learning_rate": 9.99307715323185e-06,
+      "loss": 1.1444,
+      "step": 353
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5604984760284424,
+      "learning_rate": 9.992965070472487e-06,
+      "loss": 0.9661,
+      "step": 354
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5438612699508667,
+      "learning_rate": 9.992852088298171e-06,
+      "loss": 0.9327,
+      "step": 355
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5264252424240112,
+      "learning_rate": 9.992738206729247e-06,
+      "loss": 0.9432,
+      "step": 356
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6323277950286865,
+      "learning_rate": 9.992623425786235e-06,
+      "loss": 0.9796,
+      "step": 357
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.2774513959884644,
+      "learning_rate": 9.992507745489808e-06,
+      "loss": 1.1593,
+      "step": 358
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.4815845489501953,
+      "learning_rate": 9.992391165860806e-06,
+      "loss": 0.9491,
+      "step": 359
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5447276830673218,
+      "learning_rate": 9.992273686920229e-06,
+      "loss": 1.0076,
+      "step": 360
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.8468621969223022,
+      "learning_rate": 9.992155308689239e-06,
+      "loss": 0.8996,
+      "step": 361
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.0651299953460693,
+      "learning_rate": 9.992036031189161e-06,
+      "loss": 0.9523,
+      "step": 362
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.587266206741333,
+      "learning_rate": 9.991915854441478e-06,
+      "loss": 0.936,
+      "step": 363
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5927507877349854,
+      "learning_rate": 9.991794778467843e-06,
+      "loss": 0.9648,
+      "step": 364
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.556699514389038,
+      "learning_rate": 9.991672803290061e-06,
+      "loss": 0.9554,
+      "step": 365
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0772273540496826,
+      "learning_rate": 9.991549928930111e-06,
+      "loss": 1.1705,
+      "step": 366
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6042507886886597,
+      "learning_rate": 9.99142615541012e-06,
+      "loss": 0.9356,
+      "step": 367
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5132592916488647,
+      "learning_rate": 9.99130148275239e-06,
+      "loss": 0.9732,
+      "step": 368
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4475903511047363,
+      "learning_rate": 9.991175910979374e-06,
+      "loss": 0.9708,
+      "step": 369
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4826067686080933,
+      "learning_rate": 9.991049440113693e-06,
+      "loss": 0.9566,
+      "step": 370
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6107072830200195,
+      "learning_rate": 9.990922070178133e-06,
+      "loss": 0.9615,
+      "step": 371
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5381810665130615,
+      "learning_rate": 9.990793801195636e-06,
+      "loss": 0.9987,
+      "step": 372
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1827740669250488,
+      "learning_rate": 9.990664633189306e-06,
+      "loss": 1.1385,
+      "step": 373
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4979099035263062,
+      "learning_rate": 9.99053456618241e-06,
+      "loss": 0.9068,
+      "step": 374
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.579780101776123,
+      "learning_rate": 9.990403600198383e-06,
+      "loss": 0.9593,
+      "step": 375
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6342265605926514,
+      "learning_rate": 9.990271735260813e-06,
+      "loss": 0.9313,
+      "step": 376
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4864715337753296,
+      "learning_rate": 9.990138971393456e-06,
+      "loss": 0.9002,
+      "step": 377
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8929136991500854,
+      "learning_rate": 9.990005308620223e-06,
+      "loss": 1.1616,
+      "step": 378
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.577991247177124,
+      "learning_rate": 9.989870746965195e-06,
+      "loss": 0.9962,
+      "step": 379
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.417180061340332,
+      "learning_rate": 9.989735286452613e-06,
+      "loss": 0.9317,
+      "step": 380
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4584673643112183,
+      "learning_rate": 9.989598927106874e-06,
+      "loss": 0.9691,
+      "step": 381
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3864586353302002,
+      "learning_rate": 9.989461668952545e-06,
+      "loss": 0.946,
+      "step": 382
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5059852600097656,
+      "learning_rate": 9.989323512014351e-06,
+      "loss": 1.0177,
+      "step": 383
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.509445071220398,
+      "learning_rate": 9.989184456317176e-06,
+      "loss": 0.9592,
+      "step": 384
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5963958501815796,
+      "learning_rate": 9.989044501886072e-06,
+      "loss": 0.8835,
+      "step": 385
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4955120086669922,
+      "learning_rate": 9.98890364874625e-06,
+      "loss": 0.988,
+      "step": 386
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5209298133850098,
+      "learning_rate": 9.988761896923078e-06,
+      "loss": 0.9444,
+      "step": 387
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4584766626358032,
+      "learning_rate": 9.988619246442098e-06,
+      "loss": 0.9836,
+      "step": 388
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5144188404083252,
+      "learning_rate": 9.988475697328999e-06,
+      "loss": 0.9588,
+      "step": 389
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4474812746047974,
+      "learning_rate": 9.988331249609646e-06,
+      "loss": 0.9515,
+      "step": 390
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0216593742370605,
+      "learning_rate": 9.988185903310058e-06,
+      "loss": 1.1497,
+      "step": 391
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3893070220947266,
+      "learning_rate": 9.988039658456411e-06,
+      "loss": 0.9234,
+      "step": 392
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5299872159957886,
+      "learning_rate": 9.987892515075058e-06,
+      "loss": 0.9455,
+      "step": 393
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.6663475036621094,
+      "learning_rate": 9.9877444731925e-06,
+      "loss": 0.9943,
+      "step": 394
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.629140853881836,
+      "learning_rate": 9.987595532835403e-06,
+      "loss": 0.9193,
+      "step": 395
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3945270776748657,
+      "learning_rate": 9.987445694030601e-06,
+      "loss": 0.9482,
+      "step": 396
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5609408617019653,
+      "learning_rate": 9.987294956805082e-06,
+      "loss": 0.9976,
+      "step": 397
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.1717299222946167,
+      "learning_rate": 9.987143321186002e-06,
+      "loss": 1.126,
+      "step": 398
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 2.2488107681274414,
+      "learning_rate": 9.986990787200675e-06,
+      "loss": 0.9767,
+      "step": 399
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4843134880065918,
+      "learning_rate": 9.986837354876578e-06,
+      "loss": 0.9684,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.496118426322937,
+      "learning_rate": 9.986683024241348e-06,
+      "loss": 0.9726,
+      "step": 401
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4857666492462158,
+      "learning_rate": 9.986527795322787e-06,
+      "loss": 0.9521,
+      "step": 402
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.3628171682357788,
+      "learning_rate": 9.986371668148858e-06,
+      "loss": 0.9631,
+      "step": 403
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.058052659034729,
+      "learning_rate": 9.986214642747685e-06,
+      "loss": 1.1568,
+      "step": 404
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.470193862915039,
+      "learning_rate": 9.986056719147553e-06,
+      "loss": 1.0039,
+      "step": 405
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7840606570243835,
+      "learning_rate": 9.98589789737691e-06,
+      "loss": 1.1366,
+      "step": 406
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.7688361406326294,
+      "learning_rate": 9.985738177464367e-06,
+      "loss": 0.9518,
+      "step": 407
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.559470772743225,
+      "learning_rate": 9.985577559438694e-06,
+      "loss": 0.89,
+      "step": 408
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.0118401050567627,
+      "learning_rate": 9.985416043328823e-06,
+      "loss": 1.1565,
+      "step": 409
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.449958086013794,
+      "learning_rate": 9.985253629163853e-06,
+      "loss": 1.0087,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4975261688232422,
+      "learning_rate": 9.985090316973035e-06,
+      "loss": 0.9345,
+      "step": 411
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.9507182836532593,
+      "learning_rate": 9.984926106785795e-06,
+      "loss": 1.0151,
+      "step": 412
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.584429383277893,
+      "learning_rate": 9.984760998631706e-06,
+      "loss": 0.9803,
+      "step": 413
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.419223427772522,
+      "learning_rate": 9.984594992540513e-06,
+      "loss": 0.9119,
+      "step": 414
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.530431866645813,
+      "learning_rate": 9.984428088542121e-06,
+      "loss": 0.9308,
+      "step": 415
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4944473505020142,
+      "learning_rate": 9.984260286666592e-06,
+      "loss": 0.9299,
+      "step": 416
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4353196620941162,
+      "learning_rate": 9.984091586944158e-06,
+      "loss": 0.9011,
+      "step": 417
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4353175163269043,
+      "learning_rate": 9.983921989405203e-06,
+      "loss": 1.1504,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.4474848508834839,
+      "learning_rate": 9.98375149408028e-06,
+      "loss": 0.8919,
+      "step": 419
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.7379816770553589,
+      "learning_rate": 9.983580101000103e-06,
+      "loss": 0.9565,
+      "step": 420
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8258067965507507,
+      "learning_rate": 9.983407810195544e-06,
+      "loss": 1.1297,
+      "step": 421
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.538323998451233,
+      "learning_rate": 9.983234621697639e-06,
+      "loss": 0.9742,
+      "step": 422
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3819758892059326,
+      "learning_rate": 9.983060535537587e-06,
+      "loss": 0.9333,
+      "step": 423
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.473607063293457,
+      "learning_rate": 9.982885551746745e-06,
+      "loss": 0.9213,
+      "step": 424
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3530857563018799,
+      "learning_rate": 9.982709670356636e-06,
+      "loss": 0.9709,
+      "step": 425
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5649759769439697,
+      "learning_rate": 9.982532891398941e-06,
+      "loss": 0.9462,
+      "step": 426
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0598127841949463,
+      "learning_rate": 9.982355214905506e-06,
+      "loss": 1.136,
+      "step": 427
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6548621654510498,
+      "learning_rate": 9.982176640908337e-06,
+      "loss": 0.9512,
+      "step": 428
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.729656457901001,
+      "learning_rate": 9.981997169439599e-06,
+      "loss": 1.1475,
+      "step": 429
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.476915717124939,
+      "learning_rate": 9.981816800531623e-06,
+      "loss": 1.0,
+      "step": 430
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.378775715827942,
+      "learning_rate": 9.981635534216902e-06,
+      "loss": 0.9218,
+      "step": 431
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4408196210861206,
+      "learning_rate": 9.981453370528085e-06,
+      "loss": 1.0227,
+      "step": 432
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.376600980758667,
+      "learning_rate": 9.98127030949799e-06,
+      "loss": 0.918,
+      "step": 433
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4468235969543457,
+      "learning_rate": 9.98108635115959e-06,
+      "loss": 1.1571,
+      "step": 434
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4071303606033325,
+      "learning_rate": 9.980901495546025e-06,
+      "loss": 0.9831,
+      "step": 435
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3805649280548096,
+      "learning_rate": 9.980715742690592e-06,
+      "loss": 0.9199,
+      "step": 436
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3507908582687378,
+      "learning_rate": 9.980529092626754e-06,
+      "loss": 0.923,
+      "step": 437
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6634057760238647,
+      "learning_rate": 9.980341545388132e-06,
+      "loss": 0.9174,
+      "step": 438
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9817489385604858,
+      "learning_rate": 9.980153101008509e-06,
+      "loss": 1.1577,
+      "step": 439
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.8682441711425781,
+      "learning_rate": 9.979963759521833e-06,
+      "loss": 1.0022,
+      "step": 440
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5420656204223633,
+      "learning_rate": 9.979773520962212e-06,
+      "loss": 0.9684,
+      "step": 441
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.0556681156158447,
+      "learning_rate": 9.979582385363912e-06,
+      "loss": 0.9731,
+      "step": 442
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6466350555419922,
+      "learning_rate": 9.979390352761364e-06,
+      "loss": 0.9323,
+      "step": 443
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4357584714889526,
+      "learning_rate": 9.979197423189163e-06,
+      "loss": 0.8852,
+      "step": 444
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5189789533615112,
+      "learning_rate": 9.979003596682058e-06,
+      "loss": 0.8733,
+      "step": 445
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3932422399520874,
+      "learning_rate": 9.978808873274968e-06,
+      "loss": 0.9042,
+      "step": 446
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.604669213294983,
+      "learning_rate": 9.978613253002968e-06,
+      "loss": 0.9553,
+      "step": 447
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4281587600708008,
+      "learning_rate": 9.9784167359013e-06,
+      "loss": 0.9765,
+      "step": 448
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4735482931137085,
+      "learning_rate": 9.978219322005358e-06,
+      "loss": 0.9363,
+      "step": 449
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.094127893447876,
+      "learning_rate": 9.978021011350706e-06,
+      "loss": 1.1654,
+      "step": 450
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4267462491989136,
+      "learning_rate": 9.977821803973067e-06,
+      "loss": 0.9488,
+      "step": 451
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7310751080513,
+      "learning_rate": 9.977621699908327e-06,
+      "loss": 1.159,
+      "step": 452
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5487116575241089,
+      "learning_rate": 9.97742069919253e-06,
+      "loss": 0.9546,
+      "step": 453
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5542573928833008,
+      "learning_rate": 9.977218801861884e-06,
+      "loss": 0.9506,
+      "step": 454
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4567456245422363,
+      "learning_rate": 9.97701600795276e-06,
+      "loss": 0.9582,
+      "step": 455
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3822704553604126,
+      "learning_rate": 9.976812317501685e-06,
+      "loss": 0.9369,
+      "step": 456
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3387315273284912,
+      "learning_rate": 9.976607730545354e-06,
+      "loss": 1.156,
+      "step": 457
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.755492091178894,
+      "learning_rate": 9.97640224712062e-06,
+      "loss": 0.9587,
+      "step": 458
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.382556438446045,
+      "learning_rate": 9.976195867264496e-06,
+      "loss": 0.9458,
+      "step": 459
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.485973596572876,
+      "learning_rate": 9.97598859101416e-06,
+      "loss": 0.9418,
+      "step": 460
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5449447631835938,
+      "learning_rate": 9.975780418406953e-06,
+      "loss": 0.9827,
+      "step": 461
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3587554693222046,
+      "learning_rate": 9.975571349480368e-06,
+      "loss": 0.9269,
+      "step": 462
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3224170207977295,
+      "learning_rate": 9.975361384272073e-06,
+      "loss": 0.9674,
+      "step": 463
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4727439880371094,
+      "learning_rate": 9.975150522819886e-06,
+      "loss": 0.9435,
+      "step": 464
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5488736629486084,
+      "learning_rate": 9.97493876516179e-06,
+      "loss": 0.9159,
+      "step": 465
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4395999908447266,
+      "learning_rate": 9.974726111335932e-06,
+      "loss": 0.9646,
+      "step": 466
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.015296220779419,
+      "learning_rate": 9.974512561380622e-06,
+      "loss": 1.1357,
+      "step": 467
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4091243743896484,
+      "learning_rate": 9.974298115334322e-06,
+      "loss": 0.9068,
+      "step": 468
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.807292103767395,
+      "learning_rate": 9.974082773235665e-06,
+      "loss": 0.9317,
+      "step": 469
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9084365963935852,
+      "learning_rate": 9.973866535123442e-06,
+      "loss": 1.1345,
+      "step": 470
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7847452759742737,
+      "learning_rate": 9.973649401036605e-06,
+      "loss": 1.1444,
+      "step": 471
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5080116987228394,
+      "learning_rate": 9.973431371014266e-06,
+      "loss": 0.9899,
+      "step": 472
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7479475140571594,
+      "learning_rate": 9.973212445095704e-06,
+      "loss": 1.1157,
+      "step": 473
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4061172008514404,
+      "learning_rate": 9.97299262332035e-06,
+      "loss": 0.887,
+      "step": 474
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4733768701553345,
+      "learning_rate": 9.972771905727807e-06,
+      "loss": 0.9209,
+      "step": 475
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8450043201446533,
+      "learning_rate": 9.972550292357833e-06,
+      "loss": 1.1411,
+      "step": 476
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7580406069755554,
+      "learning_rate": 9.972327783250348e-06,
+      "loss": 1.1413,
+      "step": 477
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.67802894115448,
+      "learning_rate": 9.972104378445434e-06,
+      "loss": 1.1431,
+      "step": 478
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4709429740905762,
+      "learning_rate": 9.971880077983334e-06,
+      "loss": 0.9352,
+      "step": 479
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.427835464477539,
+      "learning_rate": 9.97165488190445e-06,
+      "loss": 0.951,
+      "step": 480
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3926692008972168,
+      "learning_rate": 9.971428790249355e-06,
+      "loss": 0.9801,
+      "step": 481
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0937511920928955,
+      "learning_rate": 9.971201803058772e-06,
+      "loss": 1.1435,
+      "step": 482
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3851370811462402,
+      "learning_rate": 9.97097392037359e-06,
+      "loss": 1.019,
+      "step": 483
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.3514741659164429,
+      "learning_rate": 9.970745142234859e-06,
+      "loss": 0.9644,
+      "step": 484
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5331555604934692,
+      "learning_rate": 9.970515468683789e-06,
+      "loss": 0.9636,
+      "step": 485
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.6411426067352295,
+      "learning_rate": 9.970284899761756e-06,
+      "loss": 0.9618,
+      "step": 486
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.0324474573135376,
+      "learning_rate": 9.97005343551029e-06,
+      "loss": 1.1271,
+      "step": 487
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8424718976020813,
+      "learning_rate": 9.96982107597109e-06,
+      "loss": 1.1458,
+      "step": 488
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5709595680236816,
+      "learning_rate": 9.969587821186008e-06,
+      "loss": 0.9681,
+      "step": 489
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5103217363357544,
+      "learning_rate": 9.969353671197064e-06,
+      "loss": 0.9219,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4588223695755005,
+      "learning_rate": 9.969118626046439e-06,
+      "loss": 0.9302,
+      "step": 491
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.5540438890457153,
+      "learning_rate": 9.96888268577647e-06,
+      "loss": 1.1576,
+      "step": 492
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.573871374130249,
+      "learning_rate": 9.968645850429659e-06,
+      "loss": 0.9168,
+      "step": 493
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.326905608177185,
+      "learning_rate": 9.968408120048671e-06,
+      "loss": 0.8864,
+      "step": 494
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.9657586216926575,
+      "learning_rate": 9.968169494676327e-06,
+      "loss": 1.139,
+      "step": 495
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.4620364904403687,
+      "learning_rate": 9.967929974355615e-06,
+      "loss": 0.9255,
+      "step": 496
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6144969463348389,
+      "learning_rate": 9.967689559129677e-06,
+      "loss": 1.0273,
+      "step": 497
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5554896593093872,
+      "learning_rate": 9.967448249041825e-06,
+      "loss": 0.9636,
+      "step": 498
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5357083082199097,
+      "learning_rate": 9.967206044135526e-06,
+      "loss": 1.1277,
+      "step": 499
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.425915241241455,
+      "learning_rate": 9.966962944454408e-06,
+      "loss": 0.9001,
+      "step": 500
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4305177927017212,
+      "learning_rate": 9.966718950042267e-06,
+      "loss": 0.8808,
+      "step": 501
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8771573901176453,
+      "learning_rate": 9.96647406094305e-06,
+      "loss": 1.1306,
+      "step": 502
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7713894248008728,
+      "learning_rate": 9.966228277200874e-06,
+      "loss": 1.1297,
+      "step": 503
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4764586687088013,
+      "learning_rate": 9.965981598860012e-06,
+      "loss": 0.9352,
+      "step": 504
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.7074099779129028,
+      "learning_rate": 9.965734025964899e-06,
+      "loss": 0.9836,
+      "step": 505
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2787067890167236,
+      "learning_rate": 9.965485558560134e-06,
+      "loss": 1.1257,
+      "step": 506
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1242448091506958,
+      "learning_rate": 9.965236196690475e-06,
+      "loss": 1.1641,
+      "step": 507
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4829778671264648,
+      "learning_rate": 9.96498594040084e-06,
+      "loss": 0.923,
+      "step": 508
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.9968745708465576,
+      "learning_rate": 9.964734789736307e-06,
+      "loss": 1.1454,
+      "step": 509
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4719913005828857,
+      "learning_rate": 9.964482744742123e-06,
+      "loss": 0.916,
+      "step": 510
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1045069694519043,
+      "learning_rate": 9.964229805463685e-06,
+      "loss": 1.1425,
+      "step": 511
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4275307655334473,
+      "learning_rate": 9.96397597194656e-06,
+      "loss": 1.0004,
+      "step": 512
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.463562250137329,
+      "learning_rate": 9.963721244236473e-06,
+      "loss": 0.9614,
+      "step": 513
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.0080337524414062,
+      "learning_rate": 9.963465622379308e-06,
+      "loss": 1.1224,
+      "step": 514
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3887238502502441,
+      "learning_rate": 9.963209106421109e-06,
+      "loss": 0.9097,
+      "step": 515
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.351333737373352,
+      "learning_rate": 9.962951696408089e-06,
+      "loss": 0.8782,
+      "step": 516
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4044694900512695,
+      "learning_rate": 9.962693392386615e-06,
+      "loss": 0.9424,
+      "step": 517
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3451911211013794,
+      "learning_rate": 9.962434194403217e-06,
+      "loss": 0.9411,
+      "step": 518
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4138976335525513,
+      "learning_rate": 9.962174102504586e-06,
+      "loss": 0.9826,
+      "step": 519
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4317078590393066,
+      "learning_rate": 9.961913116737573e-06,
+      "loss": 0.9713,
+      "step": 520
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.612273931503296,
+      "learning_rate": 9.961651237149194e-06,
+      "loss": 0.9617,
+      "step": 521
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5424280166625977,
+      "learning_rate": 9.961388463786619e-06,
+      "loss": 0.9794,
+      "step": 522
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.423339605331421,
+      "learning_rate": 9.961124796697185e-06,
+      "loss": 0.9928,
+      "step": 523
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4458669424057007,
+      "learning_rate": 9.96086023592839e-06,
+      "loss": 0.887,
+      "step": 524
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.384997844696045,
+      "learning_rate": 9.960594781527887e-06,
+      "loss": 0.9011,
+      "step": 525
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4113718271255493,
+      "learning_rate": 9.960328433543499e-06,
+      "loss": 0.9411,
+      "step": 526
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3857786655426025,
+      "learning_rate": 9.9600611920232e-06,
+      "loss": 0.8922,
+      "step": 527
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2150624990463257,
+      "learning_rate": 9.959793057015133e-06,
+      "loss": 1.1288,
+      "step": 528
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3422139883041382,
+      "learning_rate": 9.959524028567598e-06,
+      "loss": 0.9422,
+      "step": 529
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5169236660003662,
+      "learning_rate": 9.959254106729057e-06,
+      "loss": 0.9259,
+      "step": 530
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7655589580535889,
+      "learning_rate": 9.958983291548131e-06,
+      "loss": 1.1385,
+      "step": 531
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6829167604446411,
+      "learning_rate": 9.958711583073608e-06,
+      "loss": 0.9464,
+      "step": 532
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3683396577835083,
+      "learning_rate": 9.958438981354427e-06,
+      "loss": 0.8892,
+      "step": 533
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3807895183563232,
+      "learning_rate": 9.9581654864397e-06,
+      "loss": 0.9231,
+      "step": 534
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4451568126678467,
+      "learning_rate": 9.957891098378688e-06,
+      "loss": 0.9814,
+      "step": 535
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5388907194137573,
+      "learning_rate": 9.957615817220818e-06,
+      "loss": 0.9575,
+      "step": 536
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.2202492952346802,
+      "learning_rate": 9.957339643015682e-06,
+      "loss": 1.1169,
+      "step": 537
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8963650465011597,
+      "learning_rate": 9.957062575813027e-06,
+      "loss": 1.1127,
+      "step": 538
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.420667052268982,
+      "learning_rate": 9.956784615662764e-06,
+      "loss": 1.0222,
+      "step": 539
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1137043237686157,
+      "learning_rate": 9.956505762614963e-06,
+      "loss": 1.1478,
+      "step": 540
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.366085410118103,
+      "learning_rate": 9.956226016719855e-06,
+      "loss": 0.8931,
+      "step": 541
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.449710488319397,
+      "learning_rate": 9.955945378027832e-06,
+      "loss": 0.932,
+      "step": 542
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3646560907363892,
+      "learning_rate": 9.955663846589448e-06,
+      "loss": 0.9454,
+      "step": 543
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3915393352508545,
+      "learning_rate": 9.955381422455418e-06,
+      "loss": 0.9119,
+      "step": 544
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.1647049188613892,
+      "learning_rate": 9.955098105676617e-06,
+      "loss": 1.1441,
+      "step": 545
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3468338251113892,
+      "learning_rate": 9.954813896304078e-06,
+      "loss": 0.8892,
+      "step": 546
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.498673677444458,
+      "learning_rate": 9.954528794389002e-06,
+      "loss": 0.9358,
+      "step": 547
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7803265452384949,
+      "learning_rate": 9.954242799982742e-06,
+      "loss": 1.15,
+      "step": 548
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.715859591960907,
+      "learning_rate": 9.953955913136817e-06,
+      "loss": 1.1175,
+      "step": 549
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.486336350440979,
+      "learning_rate": 9.953668133902907e-06,
+      "loss": 0.9056,
+      "step": 550
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5462275743484497,
+      "learning_rate": 9.95337946233285e-06,
+      "loss": 0.9738,
+      "step": 551
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.6007275581359863,
+      "learning_rate": 9.95308989847865e-06,
+      "loss": 0.948,
+      "step": 552
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3717137575149536,
+      "learning_rate": 9.952799442392462e-06,
+      "loss": 0.9236,
+      "step": 553
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4301596879959106,
+      "learning_rate": 9.952508094126611e-06,
+      "loss": 0.9016,
+      "step": 554
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3156275749206543,
+      "learning_rate": 9.952215853733581e-06,
+      "loss": 1.1524,
+      "step": 555
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4383926391601562,
+      "learning_rate": 9.951922721266013e-06,
+      "loss": 0.914,
+      "step": 556
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5146205425262451,
+      "learning_rate": 9.95162869677671e-06,
+      "loss": 0.9929,
+      "step": 557
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3907339572906494,
+      "learning_rate": 9.95133378031864e-06,
+      "loss": 0.9848,
+      "step": 558
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3659486770629883,
+      "learning_rate": 9.951037971944923e-06,
+      "loss": 0.9012,
+      "step": 559
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.5088067054748535,
+      "learning_rate": 9.95074127170885e-06,
+      "loss": 0.9556,
+      "step": 560
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3314696550369263,
+      "learning_rate": 9.950443679663867e-06,
+      "loss": 0.9303,
+      "step": 561
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3800345659255981,
+      "learning_rate": 9.950145195863577e-06,
+      "loss": 0.9215,
+      "step": 562
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8306305408477783,
+      "learning_rate": 9.94984582036175e-06,
+      "loss": 0.9612,
+      "step": 563
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4110983610153198,
+      "learning_rate": 9.949545553212317e-06,
+      "loss": 0.9267,
+      "step": 564
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.3865903615951538,
+      "learning_rate": 9.949244394469362e-06,
+      "loss": 0.9067,
+      "step": 565
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.104546070098877,
+      "learning_rate": 9.948942344187141e-06,
+      "loss": 1.1489,
+      "step": 566
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4215689897537231,
+      "learning_rate": 9.948639402420058e-06,
+      "loss": 0.9418,
+      "step": 567
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7267431616783142,
+      "learning_rate": 9.948335569222687e-06,
+      "loss": 1.1273,
+      "step": 568
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7639335989952087,
+      "learning_rate": 9.94803084464976e-06,
+      "loss": 1.1269,
+      "step": 569
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4126269817352295,
+      "learning_rate": 9.947725228756168e-06,
+      "loss": 0.9424,
+      "step": 570
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.4219353199005127,
+      "learning_rate": 9.947418721596963e-06,
+      "loss": 0.9355,
+      "step": 571
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 1.8723633289337158,
+      "learning_rate": 9.94711132322736e-06,
+      "loss": 0.965,
+      "step": 572
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4134798049926758,
+      "learning_rate": 9.94680303370273e-06,
+      "loss": 0.9541,
+      "step": 573
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4761064052581787,
+      "learning_rate": 9.946493853078609e-06,
+      "loss": 0.9683,
+      "step": 574
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.95396888256073,
+      "learning_rate": 9.946183781410692e-06,
+      "loss": 0.9445,
+      "step": 575
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6169358491897583,
+      "learning_rate": 9.945872818754831e-06,
+      "loss": 0.96,
+      "step": 576
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.410517692565918,
+      "learning_rate": 9.945560965167047e-06,
+      "loss": 0.885,
+      "step": 577
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.318150281906128,
+      "learning_rate": 9.945248220703513e-06,
+      "loss": 0.9563,
+      "step": 578
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4391018152236938,
+      "learning_rate": 9.944934585420565e-06,
+      "loss": 0.9093,
+      "step": 579
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.35861074924469,
+      "learning_rate": 9.944620059374702e-06,
+      "loss": 0.937,
+      "step": 580
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4941349029541016,
+      "learning_rate": 9.94430464262258e-06,
+      "loss": 0.9814,
+      "step": 581
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3723918199539185,
+      "learning_rate": 9.94398833522102e-06,
+      "loss": 0.9062,
+      "step": 582
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4743592739105225,
+      "learning_rate": 9.943671137226996e-06,
+      "loss": 0.9274,
+      "step": 583
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5117233991622925,
+      "learning_rate": 9.943353048697649e-06,
+      "loss": 1.146,
+      "step": 584
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.437294602394104,
+      "learning_rate": 9.94303406969028e-06,
+      "loss": 0.9632,
+      "step": 585
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3592242002487183,
+      "learning_rate": 9.942714200262347e-06,
+      "loss": 0.9506,
+      "step": 586
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.345357060432434,
+      "learning_rate": 9.94239344047147e-06,
+      "loss": 0.9293,
+      "step": 587
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3968546390533447,
+      "learning_rate": 9.94207179037543e-06,
+      "loss": 0.9611,
+      "step": 588
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3991177082061768,
+      "learning_rate": 9.941749250032168e-06,
+      "loss": 0.9392,
+      "step": 589
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4542577266693115,
+      "learning_rate": 9.941425819499787e-06,
+      "loss": 0.994,
+      "step": 590
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3680903911590576,
+      "learning_rate": 9.941101498836544e-06,
+      "loss": 0.8873,
+      "step": 591
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3767004013061523,
+      "learning_rate": 9.940776288100865e-06,
+      "loss": 0.9428,
+      "step": 592
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.438422679901123,
+      "learning_rate": 9.940450187351332e-06,
+      "loss": 0.9436,
+      "step": 593
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.322338581085205,
+      "learning_rate": 9.940123196646686e-06,
+      "loss": 1.1528,
+      "step": 594
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5043398141860962,
+      "learning_rate": 9.93979531604583e-06,
+      "loss": 0.8826,
+      "step": 595
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.8179361820220947,
+      "learning_rate": 9.939466545607827e-06,
+      "loss": 1.0,
+      "step": 596
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3657214641571045,
+      "learning_rate": 9.939136885391903e-06,
+      "loss": 0.9075,
+      "step": 597
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5685930252075195,
+      "learning_rate": 9.938806335457441e-06,
+      "loss": 0.9312,
+      "step": 598
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4781017303466797,
+      "learning_rate": 9.938474895863984e-06,
+      "loss": 0.9731,
+      "step": 599
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5386683940887451,
+      "learning_rate": 9.938142566671234e-06,
+      "loss": 0.9461,
+      "step": 600
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.416353702545166,
+      "learning_rate": 9.93780934793906e-06,
+      "loss": 0.9221,
+      "step": 601
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4591283798217773,
+      "learning_rate": 9.937475239727486e-06,
+      "loss": 0.9259,
+      "step": 602
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4638112783432007,
+      "learning_rate": 9.937140242096697e-06,
+      "loss": 0.9849,
+      "step": 603
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3391438722610474,
+      "learning_rate": 9.936804355107035e-06,
+      "loss": 0.9653,
+      "step": 604
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3830032348632812,
+      "learning_rate": 9.93646757881901e-06,
+      "loss": 0.9708,
+      "step": 605
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3517951965332031,
+      "learning_rate": 9.936129913293284e-06,
+      "loss": 0.9688,
+      "step": 606
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5770984888076782,
+      "learning_rate": 9.935791358590684e-06,
+      "loss": 0.9615,
+      "step": 607
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3737425804138184,
+      "learning_rate": 9.935451914772199e-06,
+      "loss": 0.9238,
+      "step": 608
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3757994174957275,
+      "learning_rate": 9.93511158189897e-06,
+      "loss": 0.9413,
+      "step": 609
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4307546615600586,
+      "learning_rate": 9.93477036003231e-06,
+      "loss": 1.1189,
+      "step": 610
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4145463705062866,
+      "learning_rate": 9.934428249233676e-06,
+      "loss": 0.9802,
+      "step": 611
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3678884506225586,
+      "learning_rate": 9.934085249564705e-06,
+      "loss": 0.8923,
+      "step": 612
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7151396870613098,
+      "learning_rate": 9.933741361087177e-06,
+      "loss": 1.1292,
+      "step": 613
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.706983208656311,
+      "learning_rate": 9.933396583863039e-06,
+      "loss": 1.1266,
+      "step": 614
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4221751689910889,
+      "learning_rate": 9.933050917954404e-06,
+      "loss": 0.9937,
+      "step": 615
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8166865110397339,
+      "learning_rate": 9.932704363423531e-06,
+      "loss": 1.1212,
+      "step": 616
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6971117258071899,
+      "learning_rate": 9.932356920332852e-06,
+      "loss": 1.12,
+      "step": 617
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6175451278686523,
+      "learning_rate": 9.932008588744955e-06,
+      "loss": 0.9015,
+      "step": 618
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8130108714103699,
+      "learning_rate": 9.931659368722583e-06,
+      "loss": 1.1109,
+      "step": 619
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4336563348770142,
+      "learning_rate": 9.931309260328645e-06,
+      "loss": 0.9334,
+      "step": 620
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5936589241027832,
+      "learning_rate": 9.930958263626212e-06,
+      "loss": 0.8914,
+      "step": 621
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.816307008266449,
+      "learning_rate": 9.930606378678504e-06,
+      "loss": 1.1282,
+      "step": 622
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.376326322555542,
+      "learning_rate": 9.930253605548914e-06,
+      "loss": 0.9012,
+      "step": 623
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4268866777420044,
+      "learning_rate": 9.929899944300988e-06,
+      "loss": 0.9263,
+      "step": 624
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3812365531921387,
+      "learning_rate": 9.929545394998433e-06,
+      "loss": 0.9479,
+      "step": 625
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.404974102973938,
+      "learning_rate": 9.929189957705116e-06,
+      "loss": 0.9768,
+      "step": 626
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.9409897923469543,
+      "learning_rate": 9.928833632485064e-06,
+      "loss": 1.159,
+      "step": 627
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.8492332696914673,
+      "learning_rate": 9.928476419402466e-06,
+      "loss": 1.0938,
+      "step": 628
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6055244207382202,
+      "learning_rate": 9.928118318521666e-06,
+      "loss": 0.9398,
+      "step": 629
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.434980869293213,
+      "learning_rate": 9.927759329907174e-06,
+      "loss": 0.9356,
+      "step": 630
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.374350905418396,
+      "learning_rate": 9.927399453623655e-06,
+      "loss": 0.9209,
+      "step": 631
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.471172571182251,
+      "learning_rate": 9.927038689735934e-06,
+      "loss": 0.9532,
+      "step": 632
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.29723060131073,
+      "learning_rate": 9.926677038309005e-06,
+      "loss": 1.1262,
+      "step": 633
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.0589560270309448,
+      "learning_rate": 9.926314499408007e-06,
+      "loss": 1.1562,
+      "step": 634
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7575719952583313,
+      "learning_rate": 9.92595107309825e-06,
+      "loss": 1.105,
+      "step": 635
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.81711345911026,
+      "learning_rate": 9.925586759445199e-06,
+      "loss": 1.1003,
+      "step": 636
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6560413837432861,
+      "learning_rate": 9.92522155851448e-06,
+      "loss": 0.9848,
+      "step": 637
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.1626784801483154,
+      "learning_rate": 9.924855470371879e-06,
+      "loss": 1.1568,
+      "step": 638
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.535831093788147,
+      "learning_rate": 9.924488495083345e-06,
+      "loss": 0.9633,
+      "step": 639
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4600175619125366,
+      "learning_rate": 9.92412063271498e-06,
+      "loss": 0.9263,
+      "step": 640
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.6926287412643433,
+      "learning_rate": 9.92375188333305e-06,
+      "loss": 0.9429,
+      "step": 641
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4561275243759155,
+      "learning_rate": 9.92338224700398e-06,
+      "loss": 0.9016,
+      "step": 642
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3561327457427979,
+      "learning_rate": 9.923011723794358e-06,
+      "loss": 1.1295,
+      "step": 643
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.452518105506897,
+      "learning_rate": 9.922640313770926e-06,
+      "loss": 0.948,
+      "step": 644
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.4014123678207397,
+      "learning_rate": 9.922268017000589e-06,
+      "loss": 0.9537,
+      "step": 645
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.824941873550415,
+      "learning_rate": 9.92189483355041e-06,
+      "loss": 1.1388,
+      "step": 646
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3943184614181519,
+      "learning_rate": 9.921520763487614e-06,
+      "loss": 0.8776,
+      "step": 647
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.5196688175201416,
+      "learning_rate": 9.921145806879586e-06,
+      "loss": 0.9391,
+      "step": 648
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.644792914390564,
+      "learning_rate": 9.920769963793868e-06,
+      "loss": 0.9608,
+      "step": 649
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.462834358215332,
+      "learning_rate": 9.920393234298163e-06,
+      "loss": 0.8576,
+      "step": 650
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.1263771057128906,
+      "learning_rate": 9.920015618460333e-06,
+      "loss": 1.1138,
+      "step": 651
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8842138648033142,
+      "learning_rate": 9.919637116348402e-06,
+      "loss": 1.1237,
+      "step": 652
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4538922309875488,
+      "learning_rate": 9.91925772803055e-06,
+      "loss": 0.9114,
+      "step": 653
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.443524956703186,
+      "learning_rate": 9.918877453575121e-06,
+      "loss": 0.9226,
+      "step": 654
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6096255779266357,
+      "learning_rate": 9.918496293050615e-06,
+      "loss": 1.1308,
+      "step": 655
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.449823021888733,
+      "learning_rate": 9.918114246525693e-06,
+      "loss": 1.1211,
+      "step": 656
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3837579488754272,
+      "learning_rate": 9.917731314069176e-06,
+      "loss": 0.9147,
+      "step": 657
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.369454026222229,
+      "learning_rate": 9.917347495750043e-06,
+      "loss": 0.9744,
+      "step": 658
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0645726919174194,
+      "learning_rate": 9.916962791637434e-06,
+      "loss": 1.1127,
+      "step": 659
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.320002555847168,
+      "learning_rate": 9.91657720180065e-06,
+      "loss": 0.8664,
+      "step": 660
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3326737880706787,
+      "learning_rate": 9.916190726309147e-06,
+      "loss": 0.9033,
+      "step": 661
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5223101377487183,
+      "learning_rate": 9.915803365232546e-06,
+      "loss": 0.9933,
+      "step": 662
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8164642453193665,
+      "learning_rate": 9.915415118640623e-06,
+      "loss": 1.1113,
+      "step": 663
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5073466300964355,
+      "learning_rate": 9.915025986603315e-06,
+      "loss": 0.8748,
+      "step": 664
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4350310564041138,
+      "learning_rate": 9.91463596919072e-06,
+      "loss": 0.883,
+      "step": 665
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.75862717628479,
+      "learning_rate": 9.914245066473095e-06,
+      "loss": 1.1214,
+      "step": 666
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.461933970451355,
+      "learning_rate": 9.913853278520855e-06,
+      "loss": 0.8833,
+      "step": 667
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4822319746017456,
+      "learning_rate": 9.913460605404574e-06,
+      "loss": 0.9405,
+      "step": 668
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5620428323745728,
+      "learning_rate": 9.913067047194988e-06,
+      "loss": 0.9104,
+      "step": 669
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4057612419128418,
+      "learning_rate": 9.912672603962992e-06,
+      "loss": 0.9588,
+      "step": 670
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4753210544586182,
+      "learning_rate": 9.912277275779639e-06,
+      "loss": 0.9466,
+      "step": 671
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5751234292984009,
+      "learning_rate": 9.91188106271614e-06,
+      "loss": 0.9201,
+      "step": 672
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5821892023086548,
+      "learning_rate": 9.911483964843872e-06,
+      "loss": 0.8657,
+      "step": 673
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3740191459655762,
+      "learning_rate": 9.91108598223436e-06,
+      "loss": 0.9415,
+      "step": 674
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3343149423599243,
+      "learning_rate": 9.910687114959302e-06,
+      "loss": 0.9265,
+      "step": 675
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4451647996902466,
+      "learning_rate": 9.910287363090544e-06,
+      "loss": 0.8947,
+      "step": 676
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4009990692138672,
+      "learning_rate": 9.909886726700099e-06,
+      "loss": 0.9563,
+      "step": 677
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4082438945770264,
+      "learning_rate": 9.909485205860134e-06,
+      "loss": 0.9497,
+      "step": 678
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3920165300369263,
+      "learning_rate": 9.909082800642978e-06,
+      "loss": 0.932,
+      "step": 679
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5418907403945923,
+      "learning_rate": 9.90867951112112e-06,
+      "loss": 0.9311,
+      "step": 680
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4368867874145508,
+      "learning_rate": 9.908275337367205e-06,
+      "loss": 0.9299,
+      "step": 681
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3839168548583984,
+      "learning_rate": 9.907870279454042e-06,
+      "loss": 0.9592,
+      "step": 682
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9564417004585266,
+      "learning_rate": 9.907464337454595e-06,
+      "loss": 1.114,
+      "step": 683
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7685620784759521,
+      "learning_rate": 9.907057511441988e-06,
+      "loss": 1.1227,
+      "step": 684
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4570655822753906,
+      "learning_rate": 9.906649801489507e-06,
+      "loss": 0.9294,
+      "step": 685
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.940588653087616,
+      "learning_rate": 9.906241207670597e-06,
+      "loss": 1.1327,
+      "step": 686
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5451198816299438,
+      "learning_rate": 9.905831730058857e-06,
+      "loss": 0.9369,
+      "step": 687
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.360522747039795,
+      "learning_rate": 9.905421368728049e-06,
+      "loss": 0.9128,
+      "step": 688
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3612407445907593,
+      "learning_rate": 9.905010123752096e-06,
+      "loss": 0.8746,
+      "step": 689
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4158837795257568,
+      "learning_rate": 9.90459799520508e-06,
+      "loss": 0.9295,
+      "step": 690
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3428480625152588,
+      "learning_rate": 9.904184983161236e-06,
+      "loss": 0.8896,
+      "step": 691
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8808942437171936,
+      "learning_rate": 9.903771087694965e-06,
+      "loss": 1.1136,
+      "step": 692
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4396038055419922,
+      "learning_rate": 9.903356308880824e-06,
+      "loss": 0.9212,
+      "step": 693
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.906672477722168,
+      "learning_rate": 9.90294064679353e-06,
+      "loss": 0.962,
+      "step": 694
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4450615644454956,
+      "learning_rate": 9.902524101507959e-06,
+      "loss": 0.9399,
+      "step": 695
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7781892418861389,
+      "learning_rate": 9.902106673099146e-06,
+      "loss": 1.1197,
+      "step": 696
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.445783019065857,
+      "learning_rate": 9.901688361642283e-06,
+      "loss": 0.8836,
+      "step": 697
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4756804704666138,
+      "learning_rate": 9.901269167212726e-06,
+      "loss": 0.9811,
+      "step": 698
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3984458446502686,
+      "learning_rate": 9.900849089885987e-06,
+      "loss": 0.9532,
+      "step": 699
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8606259226799011,
+      "learning_rate": 9.900428129737737e-06,
+      "loss": 1.0782,
+      "step": 700
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.463690161705017,
+      "learning_rate": 9.900006286843804e-06,
+      "loss": 0.932,
+      "step": 701
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4584020376205444,
+      "learning_rate": 9.89958356128018e-06,
+      "loss": 0.9224,
+      "step": 702
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4571337699890137,
+      "learning_rate": 9.899159953123012e-06,
+      "loss": 0.9156,
+      "step": 703
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3630081415176392,
+      "learning_rate": 9.898735462448608e-06,
+      "loss": 0.9139,
+      "step": 704
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.372493028640747,
+      "learning_rate": 9.898310089333433e-06,
+      "loss": 0.9163,
+      "step": 705
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.551344871520996,
+      "learning_rate": 9.897883833854113e-06,
+      "loss": 0.9459,
+      "step": 706
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0029802322387695,
+      "learning_rate": 9.897456696087433e-06,
+      "loss": 1.108,
+      "step": 707
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7271108031272888,
+      "learning_rate": 9.897028676110337e-06,
+      "loss": 1.1188,
+      "step": 708
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.5299679040908813,
+      "learning_rate": 9.896599773999922e-06,
+      "loss": 0.9407,
+      "step": 709
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.359800100326538,
+      "learning_rate": 9.896169989833453e-06,
+      "loss": 0.9406,
+      "step": 710
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4756405353546143,
+      "learning_rate": 9.895739323688348e-06,
+      "loss": 0.9381,
+      "step": 711
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4864418506622314,
+      "learning_rate": 9.895307775642187e-06,
+      "loss": 0.8842,
+      "step": 712
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.3680310249328613,
+      "learning_rate": 9.894875345772707e-06,
+      "loss": 0.9252,
+      "step": 713
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.713040828704834,
+      "learning_rate": 9.894442034157806e-06,
+      "loss": 1.125,
+      "step": 714
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4732533693313599,
+      "learning_rate": 9.894007840875537e-06,
+      "loss": 0.8904,
+      "step": 715
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.0964598655700684,
+      "learning_rate": 9.893572766004113e-06,
+      "loss": 1.1148,
+      "step": 716
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6082347631454468,
+      "learning_rate": 9.893136809621911e-06,
+      "loss": 0.9282,
+      "step": 717
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.1417346000671387,
+      "learning_rate": 9.89269997180746e-06,
+      "loss": 0.9359,
+      "step": 718
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4872990846633911,
+      "learning_rate": 9.892262252639448e-06,
+      "loss": 0.9137,
+      "step": 719
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6158307790756226,
+      "learning_rate": 9.891823652196728e-06,
+      "loss": 0.9558,
+      "step": 720
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4410362243652344,
+      "learning_rate": 9.891384170558306e-06,
+      "loss": 0.9595,
+      "step": 721
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.4146114587783813,
+      "learning_rate": 9.890943807803353e-06,
+      "loss": 0.8655,
+      "step": 722
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.484069585800171,
+      "learning_rate": 9.890502564011186e-06,
+      "loss": 0.9454,
+      "step": 723
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.43441641330719,
+      "learning_rate": 9.890060439261298e-06,
+      "loss": 0.9482,
+      "step": 724
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 2.045081853866577,
+      "learning_rate": 9.889617433633326e-06,
+      "loss": 1.1032,
+      "step": 725
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6579293012619019,
+      "learning_rate": 9.889173547207072e-06,
+      "loss": 1.1429,
+      "step": 726
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5138592720031738,
+      "learning_rate": 9.888728780062496e-06,
+      "loss": 0.9602,
+      "step": 727
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6800189018249512,
+      "learning_rate": 9.888283132279719e-06,
+      "loss": 0.9712,
+      "step": 728
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5768169164657593,
+      "learning_rate": 9.887836603939019e-06,
+      "loss": 0.8676,
+      "step": 729
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4296081066131592,
+      "learning_rate": 9.887389195120829e-06,
+      "loss": 0.9589,
+      "step": 730
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3505432605743408,
+      "learning_rate": 9.886940905905745e-06,
+      "loss": 0.9454,
+      "step": 731
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5894566774368286,
+      "learning_rate": 9.886491736374521e-06,
+      "loss": 1.1268,
+      "step": 732
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3556098937988281,
+      "learning_rate": 9.886041686608067e-06,
+      "loss": 1.1112,
+      "step": 733
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7911111116409302,
+      "learning_rate": 9.885590756687456e-06,
+      "loss": 0.9108,
+      "step": 734
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8626328706741333,
+      "learning_rate": 9.885138946693912e-06,
+      "loss": 0.9157,
+      "step": 735
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6134271621704102,
+      "learning_rate": 9.884686256708829e-06,
+      "loss": 0.898,
+      "step": 736
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5621552467346191,
+      "learning_rate": 9.884232686813748e-06,
+      "loss": 0.9216,
+      "step": 737
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5403895378112793,
+      "learning_rate": 9.883778237090375e-06,
+      "loss": 0.9524,
+      "step": 738
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4030793905258179,
+      "learning_rate": 9.883322907620575e-06,
+      "loss": 1.0034,
+      "step": 739
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.427384614944458,
+      "learning_rate": 9.882866698486365e-06,
+      "loss": 1.0014,
+      "step": 740
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2645798921585083,
+      "learning_rate": 9.88240960976993e-06,
+      "loss": 1.1393,
+      "step": 741
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9799103140830994,
+      "learning_rate": 9.881951641553604e-06,
+      "loss": 1.0986,
+      "step": 742
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8831274509429932,
+      "learning_rate": 9.881492793919885e-06,
+      "loss": 0.9453,
+      "step": 743
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.685616135597229,
+      "learning_rate": 9.881033066951432e-06,
+      "loss": 0.8896,
+      "step": 744
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.509330153465271,
+      "learning_rate": 9.88057246073105e-06,
+      "loss": 0.8984,
+      "step": 745
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5033503770828247,
+      "learning_rate": 9.880110975341721e-06,
+      "loss": 0.9427,
+      "step": 746
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6860713958740234,
+      "learning_rate": 9.879648610866567e-06,
+      "loss": 0.8875,
+      "step": 747
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4275321960449219,
+      "learning_rate": 9.879185367388883e-06,
+      "loss": 0.8938,
+      "step": 748
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5005210638046265,
+      "learning_rate": 9.878721244992113e-06,
+      "loss": 0.9548,
+      "step": 749
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5131465196609497,
+      "learning_rate": 9.878256243759863e-06,
+      "loss": 0.8728,
+      "step": 750
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5526535511016846,
+      "learning_rate": 9.877790363775894e-06,
+      "loss": 0.9091,
+      "step": 751
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.716649055480957,
+      "learning_rate": 9.877323605124131e-06,
+      "loss": 1.1541,
+      "step": 752
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.6066560745239258,
+      "learning_rate": 9.876855967888656e-06,
+      "loss": 0.9301,
+      "step": 753
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.51528000831604,
+      "learning_rate": 9.876387452153702e-06,
+      "loss": 0.9286,
+      "step": 754
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5050122737884521,
+      "learning_rate": 9.87591805800367e-06,
+      "loss": 0.9249,
+      "step": 755
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4074193239212036,
+      "learning_rate": 9.875447785523116e-06,
+      "loss": 0.8664,
+      "step": 756
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.419379472732544,
+      "learning_rate": 9.874976634796747e-06,
+      "loss": 0.9247,
+      "step": 757
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5030075311660767,
+      "learning_rate": 9.874504605909441e-06,
+      "loss": 0.8643,
+      "step": 758
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5689328908920288,
+      "learning_rate": 9.874031698946225e-06,
+      "loss": 0.8384,
+      "step": 759
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1931895017623901,
+      "learning_rate": 9.873557913992286e-06,
+      "loss": 1.1142,
+      "step": 760
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5059504508972168,
+      "learning_rate": 9.87308325113297e-06,
+      "loss": 0.9126,
+      "step": 761
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4416589736938477,
+      "learning_rate": 9.872607710453784e-06,
+      "loss": 0.9375,
+      "step": 762
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4374290704727173,
+      "learning_rate": 9.872131292040387e-06,
+      "loss": 0.9188,
+      "step": 763
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8626406788825989,
+      "learning_rate": 9.871653995978599e-06,
+      "loss": 1.1022,
+      "step": 764
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5024367570877075,
+      "learning_rate": 9.871175822354401e-06,
+      "loss": 0.9199,
+      "step": 765
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.560868263244629,
+      "learning_rate": 9.87069677125393e-06,
+      "loss": 0.9636,
+      "step": 766
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8639456033706665,
+      "learning_rate": 9.870216842763476e-06,
+      "loss": 1.127,
+      "step": 767
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3551703691482544,
+      "learning_rate": 9.869736036969495e-06,
+      "loss": 0.8997,
+      "step": 768
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7061930298805237,
+      "learning_rate": 9.869254353958597e-06,
+      "loss": 1.1384,
+      "step": 769
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.367272138595581,
+      "learning_rate": 9.86877179381755e-06,
+      "loss": 0.9354,
+      "step": 770
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3472838401794434,
+      "learning_rate": 9.868288356633284e-06,
+      "loss": 0.9034,
+      "step": 771
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.8082624673843384,
+      "learning_rate": 9.867804042492878e-06,
+      "loss": 0.9405,
+      "step": 772
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4717350006103516,
+      "learning_rate": 9.86731885148358e-06,
+      "loss": 0.9091,
+      "step": 773
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3814733028411865,
+      "learning_rate": 9.866832783692788e-06,
+      "loss": 0.8955,
+      "step": 774
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.406077265739441,
+      "learning_rate": 9.86634583920806e-06,
+      "loss": 0.8706,
+      "step": 775
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.2865548133850098,
+      "learning_rate": 9.865858018117113e-06,
+      "loss": 1.1203,
+      "step": 776
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5438820123672485,
+      "learning_rate": 9.865369320507823e-06,
+      "loss": 0.9385,
+      "step": 777
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3504446744918823,
+      "learning_rate": 9.86487974646822e-06,
+      "loss": 0.8641,
+      "step": 778
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.745499312877655,
+      "learning_rate": 9.864389296086494e-06,
+      "loss": 1.1354,
+      "step": 779
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4633314609527588,
+      "learning_rate": 9.863897969450994e-06,
+      "loss": 0.9234,
+      "step": 780
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3779075145721436,
+      "learning_rate": 9.863405766650228e-06,
+      "loss": 0.8953,
+      "step": 781
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9630442261695862,
+      "learning_rate": 9.862912687772857e-06,
+      "loss": 1.1094,
+      "step": 782
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7089740633964539,
+      "learning_rate": 9.862418732907702e-06,
+      "loss": 1.1308,
+      "step": 783
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5074201822280884,
+      "learning_rate": 9.861923902143742e-06,
+      "loss": 0.9259,
+      "step": 784
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.472191333770752,
+      "learning_rate": 9.861428195570116e-06,
+      "loss": 0.9505,
+      "step": 785
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.7149107456207275,
+      "learning_rate": 9.860931613276118e-06,
+      "loss": 0.8216,
+      "step": 786
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.1636314392089844,
+      "learning_rate": 9.860434155351202e-06,
+      "loss": 1.1398,
+      "step": 787
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.326488733291626,
+      "learning_rate": 9.859935821884975e-06,
+      "loss": 0.9037,
+      "step": 788
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5029767751693726,
+      "learning_rate": 9.859436612967208e-06,
+      "loss": 0.8663,
+      "step": 789
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4110857248306274,
+      "learning_rate": 9.858936528687826e-06,
+      "loss": 0.8632,
+      "step": 790
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4305750131607056,
+      "learning_rate": 9.85843556913691e-06,
+      "loss": 0.8814,
+      "step": 791
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5750113725662231,
+      "learning_rate": 9.857933734404705e-06,
+      "loss": 0.8754,
+      "step": 792
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3338792324066162,
+      "learning_rate": 9.857431024581607e-06,
+      "loss": 0.8823,
+      "step": 793
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9354731440544128,
+      "learning_rate": 9.856927439758174e-06,
+      "loss": 1.1254,
+      "step": 794
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.388304591178894,
+      "learning_rate": 9.856422980025119e-06,
+      "loss": 0.9038,
+      "step": 795
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.5699182748794556,
+      "learning_rate": 9.855917645473314e-06,
+      "loss": 0.8658,
+      "step": 796
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3996248245239258,
+      "learning_rate": 9.855411436193788e-06,
+      "loss": 0.9042,
+      "step": 797
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4551799297332764,
+      "learning_rate": 9.854904352277725e-06,
+      "loss": 0.9219,
+      "step": 798
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 2.2575182914733887,
+      "learning_rate": 9.854396393816474e-06,
+      "loss": 0.9342,
+      "step": 799
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.3463671207427979,
+      "learning_rate": 9.853887560901537e-06,
+      "loss": 0.9363,
+      "step": 800
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 1.4415009021759033,
+      "learning_rate": 9.853377853624568e-06,
+      "loss": 0.9093,
+      "step": 801
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3184658288955688,
+      "learning_rate": 9.852867272077387e-06,
+      "loss": 0.9442,
+      "step": 802
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.0735880136489868,
+      "learning_rate": 9.85235581635197e-06,
+      "loss": 1.1314,
+      "step": 803
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3737776279449463,
+      "learning_rate": 9.851843486540447e-06,
+      "loss": 0.9267,
+      "step": 804
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5813853740692139,
+      "learning_rate": 9.851330282735106e-06,
+      "loss": 0.8974,
+      "step": 805
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4731910228729248,
+      "learning_rate": 9.850816205028397e-06,
+      "loss": 0.9485,
+      "step": 806
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.603867769241333,
+      "learning_rate": 9.85030125351292e-06,
+      "loss": 0.9724,
+      "step": 807
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5065815448760986,
+      "learning_rate": 9.84978542828144e-06,
+      "loss": 0.8434,
+      "step": 808
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2322756052017212,
+      "learning_rate": 9.849268729426874e-06,
+      "loss": 1.1124,
+      "step": 809
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.403719425201416,
+      "learning_rate": 9.8487511570423e-06,
+      "loss": 0.8791,
+      "step": 810
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3078422546386719,
+      "learning_rate": 9.84823271122095e-06,
+      "loss": 0.9445,
+      "step": 811
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7694791555404663,
+      "learning_rate": 9.847713392056218e-06,
+      "loss": 1.1043,
+      "step": 812
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7324251532554626,
+      "learning_rate": 9.847193199641648e-06,
+      "loss": 1.1104,
+      "step": 813
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4696495532989502,
+      "learning_rate": 9.846672134070947e-06,
+      "loss": 0.9408,
+      "step": 814
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8858897686004639,
+      "learning_rate": 9.846150195437978e-06,
+      "loss": 0.9111,
+      "step": 815
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5261526107788086,
+      "learning_rate": 9.845627383836763e-06,
+      "loss": 0.8889,
+      "step": 816
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.333091378211975,
+      "learning_rate": 9.845103699361477e-06,
+      "loss": 0.9144,
+      "step": 817
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.344914436340332,
+      "learning_rate": 9.844579142106456e-06,
+      "loss": 0.8814,
+      "step": 818
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3431506156921387,
+      "learning_rate": 9.844053712166192e-06,
+      "loss": 0.9242,
+      "step": 819
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5108554363250732,
+      "learning_rate": 9.843527409635334e-06,
+      "loss": 0.9462,
+      "step": 820
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4665024280548096,
+      "learning_rate": 9.843000234608686e-06,
+      "loss": 0.9946,
+      "step": 821
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3748857975006104,
+      "learning_rate": 9.842472187181214e-06,
+      "loss": 0.9277,
+      "step": 822
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.469646692276001,
+      "learning_rate": 9.84194326744804e-06,
+      "loss": 0.9608,
+      "step": 823
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2906383275985718,
+      "learning_rate": 9.841413475504438e-06,
+      "loss": 1.1178,
+      "step": 824
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.8606929779052734,
+      "learning_rate": 9.840882811445844e-06,
+      "loss": 0.925,
+      "step": 825
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3798640966415405,
+      "learning_rate": 9.84035127536785e-06,
+      "loss": 0.9604,
+      "step": 826
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3181670904159546,
+      "learning_rate": 9.839818867366208e-06,
+      "loss": 0.8974,
+      "step": 827
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3647445440292358,
+      "learning_rate": 9.83928558753682e-06,
+      "loss": 0.8728,
+      "step": 828
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4647878408432007,
+      "learning_rate": 9.838751435975751e-06,
+      "loss": 0.9291,
+      "step": 829
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4797368049621582,
+      "learning_rate": 9.83821641277922e-06,
+      "loss": 0.921,
+      "step": 830
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3563878536224365,
+      "learning_rate": 9.837680518043605e-06,
+      "loss": 1.124,
+      "step": 831
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.445198655128479,
+      "learning_rate": 9.837143751865442e-06,
+      "loss": 0.9178,
+      "step": 832
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.397397756576538,
+      "learning_rate": 9.836606114341418e-06,
+      "loss": 0.9393,
+      "step": 833
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.424719214439392,
+      "learning_rate": 9.836067605568384e-06,
+      "loss": 0.9203,
+      "step": 834
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7073420286178589,
+      "learning_rate": 9.835528225643346e-06,
+      "loss": 1.1274,
+      "step": 835
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7356659770011902,
+      "learning_rate": 9.834987974663461e-06,
+      "loss": 1.0885,
+      "step": 836
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5254230499267578,
+      "learning_rate": 9.834446852726053e-06,
+      "loss": 0.9193,
+      "step": 837
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3715060949325562,
+      "learning_rate": 9.833904859928599e-06,
+      "loss": 0.9453,
+      "step": 838
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7997551560401917,
+      "learning_rate": 9.833361996368725e-06,
+      "loss": 1.1065,
+      "step": 839
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3234626054763794,
+      "learning_rate": 9.832818262144226e-06,
+      "loss": 0.9449,
+      "step": 840
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6916645169258118,
+      "learning_rate": 9.832273657353046e-06,
+      "loss": 1.1081,
+      "step": 841
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7047054767608643,
+      "learning_rate": 9.831728182093289e-06,
+      "loss": 1.1229,
+      "step": 842
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3821524381637573,
+      "learning_rate": 9.831181836463216e-06,
+      "loss": 0.9362,
+      "step": 843
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4429839849472046,
+      "learning_rate": 9.83063462056124e-06,
+      "loss": 0.9241,
+      "step": 844
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3593006134033203,
+      "learning_rate": 9.83008653448594e-06,
+      "loss": 0.8616,
+      "step": 845
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3238955736160278,
+      "learning_rate": 9.82953757833604e-06,
+      "loss": 0.8785,
+      "step": 846
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2875698804855347,
+      "learning_rate": 9.828987752210432e-06,
+      "loss": 0.8657,
+      "step": 847
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.36510169506073,
+      "learning_rate": 9.828437056208158e-06,
+      "loss": 0.937,
+      "step": 848
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2714747190475464,
+      "learning_rate": 9.827885490428419e-06,
+      "loss": 0.8857,
+      "step": 849
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4878358840942383,
+      "learning_rate": 9.827333054970572e-06,
+      "loss": 0.9214,
+      "step": 850
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2913289070129395,
+      "learning_rate": 9.826779749934128e-06,
+      "loss": 0.8844,
+      "step": 851
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3984339237213135,
+      "learning_rate": 9.826225575418762e-06,
+      "loss": 0.9847,
+      "step": 852
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.5177282094955444,
+      "learning_rate": 9.825670531524295e-06,
+      "loss": 0.8786,
+      "step": 853
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3851971626281738,
+      "learning_rate": 9.82511461835072e-06,
+      "loss": 0.8794,
+      "step": 854
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2874341011047363,
+      "learning_rate": 9.824557835998168e-06,
+      "loss": 0.9152,
+      "step": 855
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4814056158065796,
+      "learning_rate": 9.82400018456694e-06,
+      "loss": 0.8856,
+      "step": 856
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2996644973754883,
+      "learning_rate": 9.823441664157489e-06,
+      "loss": 0.8541,
+      "step": 857
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2745521068572998,
+      "learning_rate": 9.822882274870424e-06,
+      "loss": 1.1105,
+      "step": 858
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.6215871572494507,
+      "learning_rate": 9.822322016806514e-06,
+      "loss": 0.8941,
+      "step": 859
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4659969806671143,
+      "learning_rate": 9.821760890066677e-06,
+      "loss": 0.8941,
+      "step": 860
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3405988216400146,
+      "learning_rate": 9.821198894751998e-06,
+      "loss": 0.8436,
+      "step": 861
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.8484325408935547,
+      "learning_rate": 9.820636030963708e-06,
+      "loss": 1.1145,
+      "step": 862
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4241193532943726,
+      "learning_rate": 9.820072298803203e-06,
+      "loss": 0.9316,
+      "step": 863
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7278963923454285,
+      "learning_rate": 9.819507698372028e-06,
+      "loss": 1.092,
+      "step": 864
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.2696492671966553,
+      "learning_rate": 9.818942229771892e-06,
+      "loss": 0.9236,
+      "step": 865
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4283007383346558,
+      "learning_rate": 9.818375893104654e-06,
+      "loss": 0.9454,
+      "step": 866
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.9323930740356445,
+      "learning_rate": 9.817808688472331e-06,
+      "loss": 1.1121,
+      "step": 867
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.4268368482589722,
+      "learning_rate": 9.817240615977102e-06,
+      "loss": 0.9009,
+      "step": 868
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7397623062133789,
+      "learning_rate": 9.816671675721292e-06,
+      "loss": 1.1072,
+      "step": 869
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.3941733837127686,
+      "learning_rate": 9.816101867807393e-06,
+      "loss": 0.919,
+      "step": 870
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7309621572494507,
+      "learning_rate": 9.815531192338042e-06,
+      "loss": 1.0827,
+      "step": 871
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7381539344787598,
+      "learning_rate": 9.814959649416044e-06,
+      "loss": 1.1003,
+      "step": 872
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.376697301864624,
+      "learning_rate": 9.814387239144352e-06,
+      "loss": 0.9166,
+      "step": 873
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 1.424964189529419,
+      "learning_rate": 9.813813961626078e-06,
+      "loss": 0.8941,
+      "step": 874
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7128735184669495,
+      "learning_rate": 9.81323981696449e-06,
+      "loss": 1.1109,
+      "step": 875
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.693276584148407,
+      "learning_rate": 9.812664805263015e-06,
+      "loss": 1.0896,
+      "step": 876
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6537851691246033,
+      "learning_rate": 9.812088926625231e-06,
+      "loss": 1.0961,
+      "step": 877
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6513513326644897,
+      "learning_rate": 9.811512181154876e-06,
+      "loss": 1.1179,
+      "step": 878
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3623323440551758,
+      "learning_rate": 9.810934568955842e-06,
+      "loss": 0.8943,
+      "step": 879
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4124168157577515,
+      "learning_rate": 9.810356090132176e-06,
+      "loss": 0.9178,
+      "step": 880
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6425068378448486,
+      "learning_rate": 9.809776744788086e-06,
+      "loss": 0.8845,
+      "step": 881
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3479130268096924,
+      "learning_rate": 9.809196533027934e-06,
+      "loss": 0.902,
+      "step": 882
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8339334726333618,
+      "learning_rate": 9.808615454956235e-06,
+      "loss": 1.1075,
+      "step": 883
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4327218532562256,
+      "learning_rate": 9.808033510677663e-06,
+      "loss": 0.929,
+      "step": 884
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3831000328063965,
+      "learning_rate": 9.807450700297048e-06,
+      "loss": 0.9169,
+      "step": 885
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5067532062530518,
+      "learning_rate": 9.806867023919376e-06,
+      "loss": 0.955,
+      "step": 886
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7550895810127258,
+      "learning_rate": 9.806282481649785e-06,
+      "loss": 1.1401,
+      "step": 887
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3361260890960693,
+      "learning_rate": 9.805697073593577e-06,
+      "loss": 0.9368,
+      "step": 888
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3764623403549194,
+      "learning_rate": 9.805110799856201e-06,
+      "loss": 1.0001,
+      "step": 889
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3273553848266602,
+      "learning_rate": 9.804523660543269e-06,
+      "loss": 0.9094,
+      "step": 890
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3496049642562866,
+      "learning_rate": 9.803935655760546e-06,
+      "loss": 0.901,
+      "step": 891
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8810259699821472,
+      "learning_rate": 9.803346785613952e-06,
+      "loss": 1.0877,
+      "step": 892
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7477895617485046,
+      "learning_rate": 9.802757050209563e-06,
+      "loss": 1.0951,
+      "step": 893
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.639329433441162,
+      "learning_rate": 9.802166449653616e-06,
+      "loss": 0.9126,
+      "step": 894
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3972188234329224,
+      "learning_rate": 9.801574984052495e-06,
+      "loss": 0.9952,
+      "step": 895
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.387281060218811,
+      "learning_rate": 9.800982653512747e-06,
+      "loss": 0.9068,
+      "step": 896
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4690560102462769,
+      "learning_rate": 9.800389458141073e-06,
+      "loss": 0.9235,
+      "step": 897
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4229357242584229,
+      "learning_rate": 9.799795398044328e-06,
+      "loss": 0.921,
+      "step": 898
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3221145868301392,
+      "learning_rate": 9.799200473329522e-06,
+      "loss": 0.9352,
+      "step": 899
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3082382678985596,
+      "learning_rate": 9.798604684103824e-06,
+      "loss": 0.9531,
+      "step": 900
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.435999870300293,
+      "learning_rate": 9.79800803047456e-06,
+      "loss": 0.9424,
+      "step": 901
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4516382217407227,
+      "learning_rate": 9.797410512549207e-06,
+      "loss": 0.9108,
+      "step": 902
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.44174325466156,
+      "learning_rate": 9.796812130435398e-06,
+      "loss": 0.9109,
+      "step": 903
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3258553743362427,
+      "learning_rate": 9.796212884240925e-06,
+      "loss": 0.9305,
+      "step": 904
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3849319219589233,
+      "learning_rate": 9.795612774073737e-06,
+      "loss": 0.933,
+      "step": 905
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3578015565872192,
+      "learning_rate": 9.79501180004193e-06,
+      "loss": 0.9191,
+      "step": 906
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3510295152664185,
+      "learning_rate": 9.794409962253764e-06,
+      "loss": 0.8292,
+      "step": 907
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.2173515558242798,
+      "learning_rate": 9.793807260817653e-06,
+      "loss": 0.875,
+      "step": 908
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3375475406646729,
+      "learning_rate": 9.793203695842163e-06,
+      "loss": 0.9271,
+      "step": 909
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3301331996917725,
+      "learning_rate": 9.792599267436022e-06,
+      "loss": 0.8941,
+      "step": 910
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4985271692276,
+      "learning_rate": 9.791993975708107e-06,
+      "loss": 0.8857,
+      "step": 911
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3625617027282715,
+      "learning_rate": 9.791387820767454e-06,
+      "loss": 0.9407,
+      "step": 912
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5562024116516113,
+      "learning_rate": 9.79078080272325e-06,
+      "loss": 0.8784,
+      "step": 913
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4168256521224976,
+      "learning_rate": 9.790172921684844e-06,
+      "loss": 0.941,
+      "step": 914
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3001397848129272,
+      "learning_rate": 9.78956417776174e-06,
+      "loss": 0.874,
+      "step": 915
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4899840354919434,
+      "learning_rate": 9.78895457106359e-06,
+      "loss": 0.8993,
+      "step": 916
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3441619873046875,
+      "learning_rate": 9.78834410170021e-06,
+      "loss": 0.9128,
+      "step": 917
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.7627578973770142,
+      "learning_rate": 9.787732769781566e-06,
+      "loss": 1.1013,
+      "step": 918
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3577990531921387,
+      "learning_rate": 9.787120575417783e-06,
+      "loss": 0.8726,
+      "step": 919
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4691054821014404,
+      "learning_rate": 9.786507518719138e-06,
+      "loss": 0.8893,
+      "step": 920
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.341252088546753,
+      "learning_rate": 9.785893599796064e-06,
+      "loss": 0.8858,
+      "step": 921
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3775274753570557,
+      "learning_rate": 9.785278818759153e-06,
+      "loss": 0.9578,
+      "step": 922
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3146746158599854,
+      "learning_rate": 9.784663175719145e-06,
+      "loss": 0.8979,
+      "step": 923
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3292315006256104,
+      "learning_rate": 9.784046670786944e-06,
+      "loss": 0.8876,
+      "step": 924
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.0918813943862915,
+      "learning_rate": 9.783429304073604e-06,
+      "loss": 1.116,
+      "step": 925
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7856582403182983,
+      "learning_rate": 9.782811075690335e-06,
+      "loss": 1.1151,
+      "step": 926
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3710891008377075,
+      "learning_rate": 9.782191985748502e-06,
+      "loss": 0.9234,
+      "step": 927
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5680114030838013,
+      "learning_rate": 9.781572034359624e-06,
+      "loss": 0.9305,
+      "step": 928
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3606529235839844,
+      "learning_rate": 9.780951221635379e-06,
+      "loss": 0.914,
+      "step": 929
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4433451890945435,
+      "learning_rate": 9.7803295476876e-06,
+      "loss": 0.93,
+      "step": 930
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3727461099624634,
+      "learning_rate": 9.77970701262827e-06,
+      "loss": 0.9357,
+      "step": 931
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.9287233352661133,
+      "learning_rate": 9.77908361656953e-06,
+      "loss": 1.1128,
+      "step": 932
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3438667058944702,
+      "learning_rate": 9.778459359623678e-06,
+      "loss": 0.9269,
+      "step": 933
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3997734785079956,
+      "learning_rate": 9.777834241903166e-06,
+      "loss": 0.9452,
+      "step": 934
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.415816307067871,
+      "learning_rate": 9.777208263520597e-06,
+      "loss": 0.9405,
+      "step": 935
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5025745630264282,
+      "learning_rate": 9.776581424588736e-06,
+      "loss": 0.9232,
+      "step": 936
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3108644485473633,
+      "learning_rate": 9.7759537252205e-06,
+      "loss": 0.9118,
+      "step": 937
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.5168485641479492,
+      "learning_rate": 9.775325165528957e-06,
+      "loss": 0.9122,
+      "step": 938
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3798491954803467,
+      "learning_rate": 9.774695745627335e-06,
+      "loss": 0.9019,
+      "step": 939
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4065024852752686,
+      "learning_rate": 9.774065465629017e-06,
+      "loss": 0.9527,
+      "step": 940
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.1282289028167725,
+      "learning_rate": 9.773434325647537e-06,
+      "loss": 1.1335,
+      "step": 941
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4264854192733765,
+      "learning_rate": 9.77280232579659e-06,
+      "loss": 0.8707,
+      "step": 942
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.405255913734436,
+      "learning_rate": 9.772169466190017e-06,
+      "loss": 0.9094,
+      "step": 943
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.336747646331787,
+      "learning_rate": 9.771535746941822e-06,
+      "loss": 0.8754,
+      "step": 944
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.358632206916809,
+      "learning_rate": 9.770901168166163e-06,
+      "loss": 0.9158,
+      "step": 945
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.6276514530181885,
+      "learning_rate": 9.770265729977347e-06,
+      "loss": 0.8966,
+      "step": 946
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.4485509395599365,
+      "learning_rate": 9.769629432489842e-06,
+      "loss": 1.114,
+      "step": 947
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.378295660018921,
+      "learning_rate": 9.768992275818267e-06,
+      "loss": 0.9061,
+      "step": 948
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3635905981063843,
+      "learning_rate": 9.768354260077398e-06,
+      "loss": 0.9029,
+      "step": 949
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7868611812591553,
+      "learning_rate": 9.767715385382163e-06,
+      "loss": 1.1054,
+      "step": 950
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7272782325744629,
+      "learning_rate": 9.76707565184765e-06,
+      "loss": 1.1041,
+      "step": 951
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.3503079414367676,
+      "learning_rate": 9.766435059589096e-06,
+      "loss": 0.9045,
+      "step": 952
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 1.422147512435913,
+      "learning_rate": 9.765793608721894e-06,
+      "loss": 0.9008,
+      "step": 953
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3593039512634277,
+      "learning_rate": 9.765151299361596e-06,
+      "loss": 0.9155,
+      "step": 954
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2013765573501587,
+      "learning_rate": 9.764508131623902e-06,
+      "loss": 1.0921,
+      "step": 955
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2972534894943237,
+      "learning_rate": 9.763864105624673e-06,
+      "loss": 0.8471,
+      "step": 956
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3935850858688354,
+      "learning_rate": 9.763219221479919e-06,
+      "loss": 0.92,
+      "step": 957
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.379590630531311,
+      "learning_rate": 9.762573479305807e-06,
+      "loss": 0.9953,
+      "step": 958
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8828725814819336,
+      "learning_rate": 9.761926879218661e-06,
+      "loss": 1.1181,
+      "step": 959
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7993552088737488,
+      "learning_rate": 9.761279421334957e-06,
+      "loss": 1.1241,
+      "step": 960
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.7418361902236938,
+      "learning_rate": 9.760631105771322e-06,
+      "loss": 0.8522,
+      "step": 961
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5241304636001587,
+      "learning_rate": 9.759981932644547e-06,
+      "loss": 0.8886,
+      "step": 962
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.305220127105713,
+      "learning_rate": 9.759331902071569e-06,
+      "loss": 0.9384,
+      "step": 963
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.1453157663345337,
+      "learning_rate": 9.75868101416948e-06,
+      "loss": 1.1162,
+      "step": 964
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.035523772239685,
+      "learning_rate": 9.758029269055531e-06,
+      "loss": 1.0998,
+      "step": 965
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5871840715408325,
+      "learning_rate": 9.757376666847125e-06,
+      "loss": 0.879,
+      "step": 966
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8631168603897095,
+      "learning_rate": 9.756723207661818e-06,
+      "loss": 1.0986,
+      "step": 967
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0033209323883057,
+      "learning_rate": 9.756068891617323e-06,
+      "loss": 1.1212,
+      "step": 968
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7400749325752258,
+      "learning_rate": 9.755413718831507e-06,
+      "loss": 1.0996,
+      "step": 969
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4170284271240234,
+      "learning_rate": 9.75475768942239e-06,
+      "loss": 0.8806,
+      "step": 970
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4158636331558228,
+      "learning_rate": 9.754100803508146e-06,
+      "loss": 0.9416,
+      "step": 971
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3944331407546997,
+      "learning_rate": 9.753443061207104e-06,
+      "loss": 0.899,
+      "step": 972
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4060157537460327,
+      "learning_rate": 9.752784462637748e-06,
+      "loss": 0.9319,
+      "step": 973
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3008254766464233,
+      "learning_rate": 9.752125007918714e-06,
+      "loss": 0.9115,
+      "step": 974
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5550199747085571,
+      "learning_rate": 9.751464697168798e-06,
+      "loss": 0.959,
+      "step": 975
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.8273433446884155,
+      "learning_rate": 9.75080353050694e-06,
+      "loss": 1.1075,
+      "step": 976
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4079643487930298,
+      "learning_rate": 9.750141508052246e-06,
+      "loss": 0.9157,
+      "step": 977
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4804089069366455,
+      "learning_rate": 9.749478629923968e-06,
+      "loss": 0.8774,
+      "step": 978
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4623254537582397,
+      "learning_rate": 9.748814896241512e-06,
+      "loss": 0.9288,
+      "step": 979
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.6720458269119263,
+      "learning_rate": 9.748150307124444e-06,
+      "loss": 0.955,
+      "step": 980
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3912584781646729,
+      "learning_rate": 9.74748486269248e-06,
+      "loss": 0.8748,
+      "step": 981
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2628130912780762,
+      "learning_rate": 9.746818563065491e-06,
+      "loss": 0.861,
+      "step": 982
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8581674098968506,
+      "learning_rate": 9.7461514083635e-06,
+      "loss": 1.1196,
+      "step": 983
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5322763919830322,
+      "learning_rate": 9.74548339870669e-06,
+      "loss": 0.8846,
+      "step": 984
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4993070363998413,
+      "learning_rate": 9.744814534215388e-06,
+      "loss": 0.8809,
+      "step": 985
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.41949462890625,
+      "learning_rate": 9.744144815010085e-06,
+      "loss": 0.8782,
+      "step": 986
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3912094831466675,
+      "learning_rate": 9.743474241211422e-06,
+      "loss": 0.9015,
+      "step": 987
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4038804769515991,
+      "learning_rate": 9.742802812940191e-06,
+      "loss": 0.9585,
+      "step": 988
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3601959943771362,
+      "learning_rate": 9.742130530317342e-06,
+      "loss": 0.9008,
+      "step": 989
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5303606986999512,
+      "learning_rate": 9.741457393463977e-06,
+      "loss": 0.9012,
+      "step": 990
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4507466554641724,
+      "learning_rate": 9.740783402501356e-06,
+      "loss": 0.9416,
+      "step": 991
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2820522785186768,
+      "learning_rate": 9.740108557550885e-06,
+      "loss": 1.1388,
+      "step": 992
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.410462737083435,
+      "learning_rate": 9.739432858734131e-06,
+      "loss": 0.948,
+      "step": 993
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3880784511566162,
+      "learning_rate": 9.73875630617281e-06,
+      "loss": 0.8953,
+      "step": 994
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7712681889533997,
+      "learning_rate": 9.738078899988794e-06,
+      "loss": 1.1135,
+      "step": 995
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7149209380149841,
+      "learning_rate": 9.737400640304108e-06,
+      "loss": 1.0818,
+      "step": 996
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.442094326019287,
+      "learning_rate": 9.736721527240935e-06,
+      "loss": 0.9149,
+      "step": 997
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.416501522064209,
+      "learning_rate": 9.736041560921604e-06,
+      "loss": 0.9548,
+      "step": 998
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5174809694290161,
+      "learning_rate": 9.735360741468604e-06,
+      "loss": 0.9271,
+      "step": 999
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.350113034248352,
+      "learning_rate": 9.734679069004575e-06,
+      "loss": 0.8787,
+      "step": 1000
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0986210107803345,
+      "learning_rate": 9.73399654365231e-06,
+      "loss": 1.0986,
+      "step": 1001
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 3.3665270805358887,
+      "learning_rate": 9.733313165534757e-06,
+      "loss": 0.881,
+      "step": 1002
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3017349243164062,
+      "learning_rate": 9.732628934775017e-06,
+      "loss": 0.9035,
+      "step": 1003
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7666157484054565,
+      "learning_rate": 9.731943851496347e-06,
+      "loss": 1.1082,
+      "step": 1004
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7171239852905273,
+      "learning_rate": 9.731257915822152e-06,
+      "loss": 1.0908,
+      "step": 1005
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4561336040496826,
+      "learning_rate": 9.730571127875997e-06,
+      "loss": 0.8664,
+      "step": 1006
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.328704833984375,
+      "learning_rate": 9.729883487781595e-06,
+      "loss": 0.924,
+      "step": 1007
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.414703607559204,
+      "learning_rate": 9.729194995662818e-06,
+      "loss": 0.9244,
+      "step": 1008
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3965975046157837,
+      "learning_rate": 9.728505651643687e-06,
+      "loss": 0.9777,
+      "step": 1009
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3188437223434448,
+      "learning_rate": 9.727815455848376e-06,
+      "loss": 0.8579,
+      "step": 1010
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.0789718627929688,
+      "learning_rate": 9.727124408401221e-06,
+      "loss": 0.9701,
+      "step": 1011
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3225479125976562,
+      "learning_rate": 9.726432509426697e-06,
+      "loss": 1.1237,
+      "step": 1012
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5081555843353271,
+      "learning_rate": 9.725739759049444e-06,
+      "loss": 0.9217,
+      "step": 1013
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4641574621200562,
+      "learning_rate": 9.72504615739425e-06,
+      "loss": 0.8788,
+      "step": 1014
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2886466979980469,
+      "learning_rate": 9.72435170458606e-06,
+      "loss": 0.9033,
+      "step": 1015
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4555450677871704,
+      "learning_rate": 9.72365640074997e-06,
+      "loss": 0.8935,
+      "step": 1016
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.478292465209961,
+      "learning_rate": 9.722960246011226e-06,
+      "loss": 0.9429,
+      "step": 1017
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3475677967071533,
+      "learning_rate": 9.722263240495236e-06,
+      "loss": 0.9105,
+      "step": 1018
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.340890645980835,
+      "learning_rate": 9.721565384327554e-06,
+      "loss": 0.9136,
+      "step": 1019
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2854143381118774,
+      "learning_rate": 9.720866677633889e-06,
+      "loss": 1.1121,
+      "step": 1020
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4526268243789673,
+      "learning_rate": 9.720167120540102e-06,
+      "loss": 0.9267,
+      "step": 1021
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4795323610305786,
+      "learning_rate": 9.719466713172212e-06,
+      "loss": 0.9524,
+      "step": 1022
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.4914432764053345,
+      "learning_rate": 9.718765455656385e-06,
+      "loss": 0.8823,
+      "step": 1023
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.558433175086975,
+      "learning_rate": 9.718063348118945e-06,
+      "loss": 0.853,
+      "step": 1024
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.364861011505127,
+      "learning_rate": 9.717360390686366e-06,
+      "loss": 0.9605,
+      "step": 1025
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.2552236318588257,
+      "learning_rate": 9.716656583485276e-06,
+      "loss": 1.0932,
+      "step": 1026
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.9541007876396179,
+      "learning_rate": 9.715951926642458e-06,
+      "loss": 1.1109,
+      "step": 1027
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.5047439336776733,
+      "learning_rate": 9.715246420284845e-06,
+      "loss": 0.8356,
+      "step": 1028
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.432947039604187,
+      "learning_rate": 9.714540064539524e-06,
+      "loss": 0.9509,
+      "step": 1029
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.3897539377212524,
+      "learning_rate": 9.713832859533737e-06,
+      "loss": 0.8623,
+      "step": 1030
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3628628253936768,
+      "learning_rate": 9.713124805394877e-06,
+      "loss": 0.8027,
+      "step": 1031
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3933042287826538,
+      "learning_rate": 9.71241590225049e-06,
+      "loss": 0.8686,
+      "step": 1032
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4417505264282227,
+      "learning_rate": 9.711706150228273e-06,
+      "loss": 0.9236,
+      "step": 1033
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 2.7045676708221436,
+      "learning_rate": 9.710995549456082e-06,
+      "loss": 1.1284,
+      "step": 1034
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4331904649734497,
+      "learning_rate": 9.71028410006192e-06,
+      "loss": 0.8965,
+      "step": 1035
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4205878973007202,
+      "learning_rate": 9.709571802173947e-06,
+      "loss": 0.9178,
+      "step": 1036
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6267415285110474,
+      "learning_rate": 9.70885865592047e-06,
+      "loss": 0.9086,
+      "step": 1037
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3848646879196167,
+      "learning_rate": 9.708144661429956e-06,
+      "loss": 1.111,
+      "step": 1038
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8326174020767212,
+      "learning_rate": 9.70742981883102e-06,
+      "loss": 1.0821,
+      "step": 1039
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5743353366851807,
+      "learning_rate": 9.706714128252433e-06,
+      "loss": 0.945,
+      "step": 1040
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5336658954620361,
+      "learning_rate": 9.705997589823112e-06,
+      "loss": 0.9687,
+      "step": 1041
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5602487325668335,
+      "learning_rate": 9.705280203672139e-06,
+      "loss": 0.9009,
+      "step": 1042
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.508573293685913,
+      "learning_rate": 9.704561969928736e-06,
+      "loss": 0.9163,
+      "step": 1043
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3893829584121704,
+      "learning_rate": 9.703842888722284e-06,
+      "loss": 0.9617,
+      "step": 1044
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4410080909729004,
+      "learning_rate": 9.703122960182316e-06,
+      "loss": 0.9142,
+      "step": 1045
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5045729875564575,
+      "learning_rate": 9.70240218443852e-06,
+      "loss": 0.9124,
+      "step": 1046
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.448928713798523,
+      "learning_rate": 9.70168056162073e-06,
+      "loss": 1.0018,
+      "step": 1047
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4437127113342285,
+      "learning_rate": 9.700958091858937e-06,
+      "loss": 0.8978,
+      "step": 1048
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3614842891693115,
+      "learning_rate": 9.700234775283287e-06,
+      "loss": 0.9166,
+      "step": 1049
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.330217957496643,
+      "learning_rate": 9.699510612024074e-06,
+      "loss": 0.874,
+      "step": 1050
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4388245344161987,
+      "learning_rate": 9.698785602211745e-06,
+      "loss": 0.8954,
+      "step": 1051
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.246807813644409,
+      "learning_rate": 9.698059745976905e-06,
+      "loss": 1.1414,
+      "step": 1052
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5520516633987427,
+      "learning_rate": 9.697333043450301e-06,
+      "loss": 0.9428,
+      "step": 1053
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3263925313949585,
+      "learning_rate": 9.696605494762842e-06,
+      "loss": 0.8338,
+      "step": 1054
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.462982416152954,
+      "learning_rate": 9.695877100045586e-06,
+      "loss": 0.8747,
+      "step": 1055
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2923386096954346,
+      "learning_rate": 9.695147859429744e-06,
+      "loss": 0.9052,
+      "step": 1056
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4018656015396118,
+      "learning_rate": 9.694417773046679e-06,
+      "loss": 0.8801,
+      "step": 1057
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.410710096359253,
+      "learning_rate": 9.693686841027902e-06,
+      "loss": 0.9566,
+      "step": 1058
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3415939807891846,
+      "learning_rate": 9.692955063505084e-06,
+      "loss": 0.8774,
+      "step": 1059
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3373653888702393,
+      "learning_rate": 9.692222440610046e-06,
+      "loss": 0.8997,
+      "step": 1060
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.9869592189788818,
+      "learning_rate": 9.691488972474758e-06,
+      "loss": 1.1005,
+      "step": 1061
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3699198961257935,
+      "learning_rate": 9.690754659231346e-06,
+      "loss": 0.9052,
+      "step": 1062
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3815295696258545,
+      "learning_rate": 9.690019501012085e-06,
+      "loss": 0.9203,
+      "step": 1063
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6508132219314575,
+      "learning_rate": 9.689283497949401e-06,
+      "loss": 0.933,
+      "step": 1064
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4232162237167358,
+      "learning_rate": 9.688546650175883e-06,
+      "loss": 0.9064,
+      "step": 1065
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4092766046524048,
+      "learning_rate": 9.687808957824256e-06,
+      "loss": 0.9479,
+      "step": 1066
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3794867992401123,
+      "learning_rate": 9.68707042102741e-06,
+      "loss": 0.918,
+      "step": 1067
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3532946109771729,
+      "learning_rate": 9.686331039918381e-06,
+      "loss": 1.0862,
+      "step": 1068
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5370250940322876,
+      "learning_rate": 9.685590814630356e-06,
+      "loss": 0.9192,
+      "step": 1069
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3860273361206055,
+      "learning_rate": 9.684849745296682e-06,
+      "loss": 0.9433,
+      "step": 1070
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.7155805826187134,
+      "learning_rate": 9.684107832050849e-06,
+      "loss": 0.8954,
+      "step": 1071
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.7394401431083679,
+      "learning_rate": 9.683365075026503e-06,
+      "loss": 1.1163,
+      "step": 1072
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.349814772605896,
+      "learning_rate": 9.682621474357441e-06,
+      "loss": 0.8975,
+      "step": 1073
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3953816890716553,
+      "learning_rate": 9.681877030177614e-06,
+      "loss": 0.9546,
+      "step": 1074
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.375501036643982,
+      "learning_rate": 9.681131742621123e-06,
+      "loss": 0.8591,
+      "step": 1075
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3920036554336548,
+      "learning_rate": 9.68038561182222e-06,
+      "loss": 0.8947,
+      "step": 1076
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.808283269405365,
+      "learning_rate": 9.679638637915314e-06,
+      "loss": 1.0805,
+      "step": 1077
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3495595455169678,
+      "learning_rate": 9.678890821034959e-06,
+      "loss": 0.8741,
+      "step": 1078
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6649292707443237,
+      "learning_rate": 9.678142161315864e-06,
+      "loss": 1.0947,
+      "step": 1079
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3772448301315308,
+      "learning_rate": 9.677392658892892e-06,
+      "loss": 0.9132,
+      "step": 1080
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3991271257400513,
+      "learning_rate": 9.676642313901055e-06,
+      "loss": 0.9172,
+      "step": 1081
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.2858245372772217,
+      "learning_rate": 9.675891126475514e-06,
+      "loss": 0.8687,
+      "step": 1082
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3640477657318115,
+      "learning_rate": 9.675139096751591e-06,
+      "loss": 0.8457,
+      "step": 1083
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3537641763687134,
+      "learning_rate": 9.674386224864752e-06,
+      "loss": 0.888,
+      "step": 1084
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.1004773378372192,
+      "learning_rate": 9.673632510950615e-06,
+      "loss": 1.0998,
+      "step": 1085
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6651397943496704,
+      "learning_rate": 9.672877955144952e-06,
+      "loss": 0.9348,
+      "step": 1086
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5398797988891602,
+      "learning_rate": 9.672122557583685e-06,
+      "loss": 0.8295,
+      "step": 1087
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.558078646659851,
+      "learning_rate": 9.671366318402892e-06,
+      "loss": 0.9344,
+      "step": 1088
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8531787395477295,
+      "learning_rate": 9.670609237738795e-06,
+      "loss": 1.117,
+      "step": 1089
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4031476974487305,
+      "learning_rate": 9.669851315727777e-06,
+      "loss": 0.9222,
+      "step": 1090
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4014407396316528,
+      "learning_rate": 9.669092552506361e-06,
+      "loss": 0.892,
+      "step": 1091
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3723554611206055,
+      "learning_rate": 9.668332948211234e-06,
+      "loss": 0.9178,
+      "step": 1092
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.5300356149673462,
+      "learning_rate": 9.667572502979223e-06,
+      "loss": 0.9074,
+      "step": 1093
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.6450855731964111,
+      "learning_rate": 9.666811216947315e-06,
+      "loss": 0.9464,
+      "step": 1094
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.916716992855072,
+      "learning_rate": 9.666049090252647e-06,
+      "loss": 1.1032,
+      "step": 1095
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8336219787597656,
+      "learning_rate": 9.665286123032502e-06,
+      "loss": 1.0971,
+      "step": 1096
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3464208841323853,
+      "learning_rate": 9.664522315424321e-06,
+      "loss": 0.9086,
+      "step": 1097
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3830885887145996,
+      "learning_rate": 9.663757667565691e-06,
+      "loss": 0.8856,
+      "step": 1098
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8823848962783813,
+      "learning_rate": 9.662992179594355e-06,
+      "loss": 1.1034,
+      "step": 1099
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3900983333587646,
+      "learning_rate": 9.662225851648205e-06,
+      "loss": 0.8633,
+      "step": 1100
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4113588333129883,
+      "learning_rate": 9.661458683865283e-06,
+      "loss": 0.9075,
+      "step": 1101
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.8360231518745422,
+      "learning_rate": 9.660690676383786e-06,
+      "loss": 1.1111,
+      "step": 1102
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6912524104118347,
+      "learning_rate": 9.65992182934206e-06,
+      "loss": 1.102,
+      "step": 1103
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6594443917274475,
+      "learning_rate": 9.659152142878601e-06,
+      "loss": 1.0761,
+      "step": 1104
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.446264624595642,
+      "learning_rate": 9.658381617132058e-06,
+      "loss": 0.9073,
+      "step": 1105
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.3242789506912231,
+      "learning_rate": 9.657610252241232e-06,
+      "loss": 0.8705,
+      "step": 1106
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3375641107559204,
+      "learning_rate": 9.656838048345072e-06,
+      "loss": 0.8977,
+      "step": 1107
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3886852264404297,
+      "learning_rate": 9.65606500558268e-06,
+      "loss": 0.8439,
+      "step": 1108
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.1885432004928589,
+      "learning_rate": 9.655291124093311e-06,
+      "loss": 1.0992,
+      "step": 1109
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.363637089729309,
+      "learning_rate": 9.65451640401637e-06,
+      "loss": 0.9245,
+      "step": 1110
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4045805931091309,
+      "learning_rate": 9.653740845491408e-06,
+      "loss": 0.9154,
+      "step": 1111
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.765995442867279,
+      "learning_rate": 9.652964448658135e-06,
+      "loss": 1.1038,
+      "step": 1112
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 3.0874032974243164,
+      "learning_rate": 9.652187213656409e-06,
+      "loss": 0.8949,
+      "step": 1113
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7427797317504883,
+      "learning_rate": 9.651409140626236e-06,
+      "loss": 1.0769,
+      "step": 1114
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7499036192893982,
+      "learning_rate": 9.650630229707775e-06,
+      "loss": 1.1038,
+      "step": 1115
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4427802562713623,
+      "learning_rate": 9.649850481041338e-06,
+      "loss": 0.9401,
+      "step": 1116
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3859816789627075,
+      "learning_rate": 9.649069894767386e-06,
+      "loss": 0.8661,
+      "step": 1117
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4529129266738892,
+      "learning_rate": 9.64828847102653e-06,
+      "loss": 0.8878,
+      "step": 1118
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3162460327148438,
+      "learning_rate": 9.647506209959534e-06,
+      "loss": 0.8321,
+      "step": 1119
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.6959104537963867,
+      "learning_rate": 9.64672311170731e-06,
+      "loss": 0.9676,
+      "step": 1120
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3590779304504395,
+      "learning_rate": 9.645939176410924e-06,
+      "loss": 0.9096,
+      "step": 1121
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5125510692596436,
+      "learning_rate": 9.64515440421159e-06,
+      "loss": 0.9948,
+      "step": 1122
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5956114530563354,
+      "learning_rate": 9.644368795250675e-06,
+      "loss": 0.9569,
+      "step": 1123
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3924094438552856,
+      "learning_rate": 9.643582349669698e-06,
+      "loss": 0.8983,
+      "step": 1124
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9817600250244141,
+      "learning_rate": 9.642795067610324e-06,
+      "loss": 1.098,
+      "step": 1125
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8505191802978516,
+      "learning_rate": 9.64200694921437e-06,
+      "loss": 1.0923,
+      "step": 1126
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.475404977798462,
+      "learning_rate": 9.641217994623808e-06,
+      "loss": 0.9087,
+      "step": 1127
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4378615617752075,
+      "learning_rate": 9.640428203980755e-06,
+      "loss": 0.906,
+      "step": 1128
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.46708345413208,
+      "learning_rate": 9.639637577427483e-06,
+      "loss": 0.9025,
+      "step": 1129
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2481263875961304,
+      "learning_rate": 9.63884611510641e-06,
+      "loss": 1.101,
+      "step": 1130
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.619655966758728,
+      "learning_rate": 9.63805381716011e-06,
+      "loss": 0.926,
+      "step": 1131
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3848029375076294,
+      "learning_rate": 9.637260683731303e-06,
+      "loss": 0.9181,
+      "step": 1132
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.805594563484192,
+      "learning_rate": 9.63646671496286e-06,
+      "loss": 0.8694,
+      "step": 1133
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3090288639068604,
+      "learning_rate": 9.635671910997808e-06,
+      "loss": 0.8304,
+      "step": 1134
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4123964309692383,
+      "learning_rate": 9.634876271979317e-06,
+      "loss": 0.8418,
+      "step": 1135
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5823867321014404,
+      "learning_rate": 9.63407979805071e-06,
+      "loss": 0.9122,
+      "step": 1136
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4758652448654175,
+      "learning_rate": 9.633282489355463e-06,
+      "loss": 0.8733,
+      "step": 1137
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.494917869567871,
+      "learning_rate": 9.6324843460372e-06,
+      "loss": 0.833,
+      "step": 1138
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.279958963394165,
+      "learning_rate": 9.631685368239695e-06,
+      "loss": 0.9359,
+      "step": 1139
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4909703731536865,
+      "learning_rate": 9.630885556106873e-06,
+      "loss": 0.9229,
+      "step": 1140
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.7972108125686646,
+      "learning_rate": 9.630084909782809e-06,
+      "loss": 0.9272,
+      "step": 1141
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9773024916648865,
+      "learning_rate": 9.629283429411731e-06,
+      "loss": 1.105,
+      "step": 1142
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3453388214111328,
+      "learning_rate": 9.628481115138011e-06,
+      "loss": 0.8591,
+      "step": 1143
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6910380721092224,
+      "learning_rate": 9.627677967106178e-06,
+      "loss": 1.0872,
+      "step": 1144
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.402666687965393,
+      "learning_rate": 9.626873985460908e-06,
+      "loss": 0.9609,
+      "step": 1145
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4157836437225342,
+      "learning_rate": 9.626069170347022e-06,
+      "loss": 0.8788,
+      "step": 1146
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.364540696144104,
+      "learning_rate": 9.625263521909508e-06,
+      "loss": 0.9074,
+      "step": 1147
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.2653732299804688,
+      "learning_rate": 9.62445704029348e-06,
+      "loss": 0.8735,
+      "step": 1148
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.373826265335083,
+      "learning_rate": 9.623649725644225e-06,
+      "loss": 0.9064,
+      "step": 1149
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3788669109344482,
+      "learning_rate": 9.622841578107163e-06,
+      "loss": 0.8066,
+      "step": 1150
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4701579809188843,
+      "learning_rate": 9.622032597827873e-06,
+      "loss": 0.9095,
+      "step": 1151
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.2773241996765137,
+      "learning_rate": 9.621222784952082e-06,
+      "loss": 1.11,
+      "step": 1152
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.474960207939148,
+      "learning_rate": 9.620412139625666e-06,
+      "loss": 0.9231,
+      "step": 1153
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7250705361366272,
+      "learning_rate": 9.619600661994655e-06,
+      "loss": 1.0887,
+      "step": 1154
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.5495281219482422,
+      "learning_rate": 9.618788352205221e-06,
+      "loss": 0.9165,
+      "step": 1155
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.488939881324768,
+      "learning_rate": 9.617975210403694e-06,
+      "loss": 0.9203,
+      "step": 1156
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.455804705619812,
+      "learning_rate": 9.617161236736549e-06,
+      "loss": 0.8544,
+      "step": 1157
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4940812587738037,
+      "learning_rate": 9.616346431350413e-06,
+      "loss": 0.8816,
+      "step": 1158
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3697607517242432,
+      "learning_rate": 9.61553079439206e-06,
+      "loss": 0.9158,
+      "step": 1159
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3249434232711792,
+      "learning_rate": 9.614714326008419e-06,
+      "loss": 0.9099,
+      "step": 1160
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3767491579055786,
+      "learning_rate": 9.613897026346563e-06,
+      "loss": 0.8919,
+      "step": 1161
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4638699293136597,
+      "learning_rate": 9.61307889555372e-06,
+      "loss": 0.8502,
+      "step": 1162
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3708561658859253,
+      "learning_rate": 9.612259933777265e-06,
+      "loss": 0.8784,
+      "step": 1163
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4481416940689087,
+      "learning_rate": 9.61144014116472e-06,
+      "loss": 0.9004,
+      "step": 1164
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3279197216033936,
+      "learning_rate": 9.610619517863763e-06,
+      "loss": 0.8165,
+      "step": 1165
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.369691014289856,
+      "learning_rate": 9.609798064022215e-06,
+      "loss": 0.9816,
+      "step": 1166
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.324994444847107,
+      "learning_rate": 9.608975779788054e-06,
+      "loss": 0.9329,
+      "step": 1167
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 2.3443236351013184,
+      "learning_rate": 9.608152665309399e-06,
+      "loss": 0.8523,
+      "step": 1168
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.8809269666671753,
+      "learning_rate": 9.607328720734523e-06,
+      "loss": 1.098,
+      "step": 1169
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4859977960586548,
+      "learning_rate": 9.606503946211852e-06,
+      "loss": 0.8703,
+      "step": 1170
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3937709331512451,
+      "learning_rate": 9.605678341889955e-06,
+      "loss": 0.9275,
+      "step": 1171
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3247096538543701,
+      "learning_rate": 9.604851907917555e-06,
+      "loss": 0.9051,
+      "step": 1172
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.741723358631134,
+      "learning_rate": 9.60402464444352e-06,
+      "loss": 1.1015,
+      "step": 1173
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7663747072219849,
+      "learning_rate": 9.603196551616877e-06,
+      "loss": 1.0816,
+      "step": 1174
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7672956585884094,
+      "learning_rate": 9.602367629586789e-06,
+      "loss": 1.0928,
+      "step": 1175
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4169869422912598,
+      "learning_rate": 9.601537878502577e-06,
+      "loss": 0.884,
+      "step": 1176
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3338158130645752,
+      "learning_rate": 9.60070729851371e-06,
+      "loss": 0.9448,
+      "step": 1177
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.4944127798080444,
+      "learning_rate": 9.599875889769806e-06,
+      "loss": 0.9159,
+      "step": 1178
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9337764978408813,
+      "learning_rate": 9.59904365242063e-06,
+      "loss": 1.0917,
+      "step": 1179
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.3863800764083862,
+      "learning_rate": 9.598210586616101e-06,
+      "loss": 0.8544,
+      "step": 1180
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.331653118133545,
+      "learning_rate": 9.597376692506282e-06,
+      "loss": 0.8718,
+      "step": 1181
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8445698618888855,
+      "learning_rate": 9.596541970241392e-06,
+      "loss": 1.1184,
+      "step": 1182
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4977306127548218,
+      "learning_rate": 9.595706419971788e-06,
+      "loss": 0.9167,
+      "step": 1183
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4046216011047363,
+      "learning_rate": 9.594870041847988e-06,
+      "loss": 0.8843,
+      "step": 1184
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7470315098762512,
+      "learning_rate": 9.594032836020654e-06,
+      "loss": 1.117,
+      "step": 1185
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5406763553619385,
+      "learning_rate": 9.593194802640596e-06,
+      "loss": 0.8345,
+      "step": 1186
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5581947565078735,
+      "learning_rate": 9.592355941858775e-06,
+      "loss": 0.9336,
+      "step": 1187
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3667434453964233,
+      "learning_rate": 9.591516253826301e-06,
+      "loss": 0.8874,
+      "step": 1188
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3359529972076416,
+      "learning_rate": 9.590675738694431e-06,
+      "loss": 0.8936,
+      "step": 1189
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.308421015739441,
+      "learning_rate": 9.58983439661457e-06,
+      "loss": 0.8775,
+      "step": 1190
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9949820637702942,
+      "learning_rate": 9.588992227738281e-06,
+      "loss": 1.0803,
+      "step": 1191
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3594329357147217,
+      "learning_rate": 9.588149232217264e-06,
+      "loss": 0.8869,
+      "step": 1192
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5813730955123901,
+      "learning_rate": 9.587305410203376e-06,
+      "loss": 0.9397,
+      "step": 1193
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3873496055603027,
+      "learning_rate": 9.586460761848617e-06,
+      "loss": 0.9163,
+      "step": 1194
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3416510820388794,
+      "learning_rate": 9.585615287305144e-06,
+      "loss": 0.8947,
+      "step": 1195
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3933900594711304,
+      "learning_rate": 9.58476898672525e-06,
+      "loss": 0.9064,
+      "step": 1196
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.0750415325164795,
+      "learning_rate": 9.583921860261393e-06,
+      "loss": 1.0798,
+      "step": 1197
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.388221263885498,
+      "learning_rate": 9.583073908066166e-06,
+      "loss": 0.905,
+      "step": 1198
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4209089279174805,
+      "learning_rate": 9.582225130292319e-06,
+      "loss": 0.9531,
+      "step": 1199
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4553166627883911,
+      "learning_rate": 9.581375527092744e-06,
+      "loss": 0.9092,
+      "step": 1200
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4526365995407104,
+      "learning_rate": 9.58052509862049e-06,
+      "loss": 0.8156,
+      "step": 1201
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8385547995567322,
+      "learning_rate": 9.579673845028747e-06,
+      "loss": 1.0647,
+      "step": 1202
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.360856533050537,
+      "learning_rate": 9.578821766470859e-06,
+      "loss": 0.8624,
+      "step": 1203
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3089284896850586,
+      "learning_rate": 9.577968863100315e-06,
+      "loss": 0.904,
+      "step": 1204
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3375340700149536,
+      "learning_rate": 9.577115135070753e-06,
+      "loss": 0.889,
+      "step": 1205
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8006930947303772,
+      "learning_rate": 9.576260582535963e-06,
+      "loss": 1.0944,
+      "step": 1206
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3802123069763184,
+      "learning_rate": 9.575405205649881e-06,
+      "loss": 0.8842,
+      "step": 1207
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6319462060928345,
+      "learning_rate": 9.574549004566589e-06,
+      "loss": 0.8568,
+      "step": 1208
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.747957170009613,
+      "learning_rate": 9.573691979440321e-06,
+      "loss": 1.0806,
+      "step": 1209
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7468222379684448,
+      "learning_rate": 9.57283413042546e-06,
+      "loss": 1.1026,
+      "step": 1210
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6660601496696472,
+      "learning_rate": 9.571975457676534e-06,
+      "loss": 1.0998,
+      "step": 1211
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6712894439697266,
+      "learning_rate": 9.571115961348222e-06,
+      "loss": 1.0918,
+      "step": 1212
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5691310167312622,
+      "learning_rate": 9.570255641595352e-06,
+      "loss": 0.8902,
+      "step": 1213
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.432307243347168,
+      "learning_rate": 9.569394498572895e-06,
+      "loss": 0.9362,
+      "step": 1214
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3805128335952759,
+      "learning_rate": 9.568532532435979e-06,
+      "loss": 0.897,
+      "step": 1215
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3140202760696411,
+      "learning_rate": 9.567669743339872e-06,
+      "loss": 0.8278,
+      "step": 1216
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.434285283088684,
+      "learning_rate": 9.566806131439995e-06,
+      "loss": 0.8465,
+      "step": 1217
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2680917978286743,
+      "learning_rate": 9.565941696891917e-06,
+      "loss": 1.1034,
+      "step": 1218
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4904155731201172,
+      "learning_rate": 9.565076439851354e-06,
+      "loss": 0.9443,
+      "step": 1219
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9193599224090576,
+      "learning_rate": 9.564210360474166e-06,
+      "loss": 1.0904,
+      "step": 1220
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.508672833442688,
+      "learning_rate": 9.563343458916372e-06,
+      "loss": 0.902,
+      "step": 1221
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3924235105514526,
+      "learning_rate": 9.562475735334128e-06,
+      "loss": 0.9028,
+      "step": 1222
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4033690690994263,
+      "learning_rate": 9.561607189883746e-06,
+      "loss": 0.8805,
+      "step": 1223
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5970468521118164,
+      "learning_rate": 9.56073782272168e-06,
+      "loss": 0.9226,
+      "step": 1224
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5870928764343262,
+      "learning_rate": 9.559867634004535e-06,
+      "loss": 0.9345,
+      "step": 1225
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4338129758834839,
+      "learning_rate": 9.558996623889065e-06,
+      "loss": 0.9291,
+      "step": 1226
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4185922145843506,
+      "learning_rate": 9.558124792532169e-06,
+      "loss": 0.8623,
+      "step": 1227
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.337969422340393,
+      "learning_rate": 9.557252140090897e-06,
+      "loss": 0.888,
+      "step": 1228
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.35490882396698,
+      "learning_rate": 9.556378666722445e-06,
+      "loss": 1.0978,
+      "step": 1229
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9877997040748596,
+      "learning_rate": 9.555504372584158e-06,
+      "loss": 1.0967,
+      "step": 1230
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4375364780426025,
+      "learning_rate": 9.554629257833528e-06,
+      "loss": 0.8163,
+      "step": 1231
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5591329336166382,
+      "learning_rate": 9.553753322628192e-06,
+      "loss": 0.8567,
+      "step": 1232
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.952489972114563,
+      "learning_rate": 9.552876567125943e-06,
+      "loss": 0.9199,
+      "step": 1233
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.688155174255371,
+      "learning_rate": 9.551998991484715e-06,
+      "loss": 0.9176,
+      "step": 1234
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7590664625167847,
+      "learning_rate": 9.551120595862588e-06,
+      "loss": 1.0885,
+      "step": 1235
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.7940524816513062,
+      "learning_rate": 9.550241380417797e-06,
+      "loss": 0.8805,
+      "step": 1236
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.532044768333435,
+      "learning_rate": 9.54936134530872e-06,
+      "loss": 0.8753,
+      "step": 1237
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.2320048809051514,
+      "learning_rate": 9.54848049069388e-06,
+      "loss": 1.0923,
+      "step": 1238
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5129653215408325,
+      "learning_rate": 9.547598816731956e-06,
+      "loss": 0.9121,
+      "step": 1239
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4998801946640015,
+      "learning_rate": 9.546716323581766e-06,
+      "loss": 0.8578,
+      "step": 1240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.8871206641197205,
+      "learning_rate": 9.545833011402281e-06,
+      "loss": 1.0958,
+      "step": 1241
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5186933279037476,
+      "learning_rate": 9.544948880352618e-06,
+      "loss": 0.8767,
+      "step": 1242
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4512568712234497,
+      "learning_rate": 9.54406393059204e-06,
+      "loss": 0.9077,
+      "step": 1243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4753965139389038,
+      "learning_rate": 9.54317816227996e-06,
+      "loss": 0.9149,
+      "step": 1244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.891757071018219,
+      "learning_rate": 9.542291575575936e-06,
+      "loss": 1.0837,
+      "step": 1245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3680521249771118,
+      "learning_rate": 9.541404170639672e-06,
+      "loss": 0.8484,
+      "step": 1246
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3836685419082642,
+      "learning_rate": 9.540515947631028e-06,
+      "loss": 0.9387,
+      "step": 1247
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.5988337993621826,
+      "learning_rate": 9.539626906710002e-06,
+      "loss": 0.8965,
+      "step": 1248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3837143182754517,
+      "learning_rate": 9.538737048036741e-06,
+      "loss": 0.9025,
+      "step": 1249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4585869312286377,
+      "learning_rate": 9.537846371771544e-06,
+      "loss": 0.9329,
+      "step": 1250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.9857527017593384,
+      "learning_rate": 9.536954878074853e-06,
+      "loss": 1.0812,
+      "step": 1251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3606339693069458,
+      "learning_rate": 9.536062567107258e-06,
+      "loss": 0.8661,
+      "step": 1252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.3730188608169556,
+      "learning_rate": 9.5351694390295e-06,
+      "loss": 0.8625,
+      "step": 1253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.4044581651687622,
+      "learning_rate": 9.534275494002457e-06,
+      "loss": 0.9289,
+      "step": 1254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7495524883270264,
+      "learning_rate": 9.533380732187168e-06,
+      "loss": 1.0785,
+      "step": 1255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7182850241661072,
+      "learning_rate": 9.532485153744808e-06,
+      "loss": 1.0903,
+      "step": 1256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.6741694211959839,
+      "learning_rate": 9.531588758836706e-06,
+      "loss": 0.9287,
+      "step": 1257
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.366279125213623,
+      "learning_rate": 9.530691547624333e-06,
+      "loss": 0.8409,
+      "step": 1258
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 1.44547700881958,
+      "learning_rate": 9.529793520269312e-06,
+      "loss": 0.8933,
+      "step": 1259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.451137900352478,
+      "learning_rate": 9.528894676933408e-06,
+      "loss": 0.8635,
+      "step": 1260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.861094057559967,
+      "learning_rate": 9.527995017778538e-06,
+      "loss": 1.1019,
+      "step": 1261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4732885360717773,
+      "learning_rate": 9.52709454296676e-06,
+      "loss": 0.8531,
+      "step": 1262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.573150396347046,
+      "learning_rate": 9.526193252660285e-06,
+      "loss": 0.897,
+      "step": 1263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3591346740722656,
+      "learning_rate": 9.525291147021467e-06,
+      "loss": 0.9165,
+      "step": 1264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.829187273979187,
+      "learning_rate": 9.52438822621281e-06,
+      "loss": 1.0805,
+      "step": 1265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3724899291992188,
+      "learning_rate": 9.523484490396961e-06,
+      "loss": 0.8843,
+      "step": 1266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4280927181243896,
+      "learning_rate": 9.522579939736716e-06,
+      "loss": 0.9113,
+      "step": 1267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.346091866493225,
+      "learning_rate": 9.521674574395016e-06,
+      "loss": 0.9079,
+      "step": 1268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4053744077682495,
+      "learning_rate": 9.520768394534955e-06,
+      "loss": 0.9244,
+      "step": 1269
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3842873573303223,
+      "learning_rate": 9.519861400319765e-06,
+      "loss": 0.9374,
+      "step": 1270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.1958885192871094,
+      "learning_rate": 9.51895359191283e-06,
+      "loss": 1.1132,
+      "step": 1271
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8445179462432861,
+      "learning_rate": 9.518044969477678e-06,
+      "loss": 1.0998,
+      "step": 1272
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6674192547798157,
+      "learning_rate": 9.517135533177989e-06,
+      "loss": 1.0859,
+      "step": 1273
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.478468418121338,
+      "learning_rate": 9.516225283177579e-06,
+      "loss": 0.8818,
+      "step": 1274
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6212321519851685,
+      "learning_rate": 9.515314219640424e-06,
+      "loss": 0.8956,
+      "step": 1275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5678879022598267,
+      "learning_rate": 9.514402342730636e-06,
+      "loss": 0.9645,
+      "step": 1276
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4018826484680176,
+      "learning_rate": 9.513489652612475e-06,
+      "loss": 0.8809,
+      "step": 1277
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3716280460357666,
+      "learning_rate": 9.512576149450356e-06,
+      "loss": 0.9191,
+      "step": 1278
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7175655364990234,
+      "learning_rate": 9.511661833408831e-06,
+      "loss": 0.9021,
+      "step": 1279
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5780271291732788,
+      "learning_rate": 9.5107467046526e-06,
+      "loss": 1.12,
+      "step": 1280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2926362752914429,
+      "learning_rate": 9.509830763346514e-06,
+      "loss": 1.1022,
+      "step": 1281
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6977616548538208,
+      "learning_rate": 9.508914009655565e-06,
+      "loss": 0.9045,
+      "step": 1282
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7150018215179443,
+      "learning_rate": 9.507996443744896e-06,
+      "loss": 0.9051,
+      "step": 1283
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.6555160284042358,
+      "learning_rate": 9.507078065779793e-06,
+      "loss": 0.9012,
+      "step": 1284
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2266383171081543,
+      "learning_rate": 9.506158875925689e-06,
+      "loss": 1.0617,
+      "step": 1285
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.456619381904602,
+      "learning_rate": 9.505238874348161e-06,
+      "loss": 0.9301,
+      "step": 1286
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4059298038482666,
+      "learning_rate": 9.50431806121294e-06,
+      "loss": 0.909,
+      "step": 1287
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5388344526290894,
+      "learning_rate": 9.503396436685896e-06,
+      "loss": 0.8536,
+      "step": 1288
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4148207902908325,
+      "learning_rate": 9.502474000933046e-06,
+      "loss": 0.8579,
+      "step": 1289
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4283761978149414,
+      "learning_rate": 9.501550754120557e-06,
+      "loss": 0.9027,
+      "step": 1290
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5679874420166016,
+      "learning_rate": 9.500626696414734e-06,
+      "loss": 0.8888,
+      "step": 1291
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2868530750274658,
+      "learning_rate": 9.49970182798204e-06,
+      "loss": 0.9314,
+      "step": 1292
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7737815976142883,
+      "learning_rate": 9.498776148989073e-06,
+      "loss": 1.0853,
+      "step": 1293
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.439746618270874,
+      "learning_rate": 9.497849659602582e-06,
+      "loss": 0.8516,
+      "step": 1294
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5074148178100586,
+      "learning_rate": 9.496922359989464e-06,
+      "loss": 0.8918,
+      "step": 1295
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4026669263839722,
+      "learning_rate": 9.495994250316756e-06,
+      "loss": 0.8445,
+      "step": 1296
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.376091480255127,
+      "learning_rate": 9.495065330751649e-06,
+      "loss": 0.8491,
+      "step": 1297
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.36817467212677,
+      "learning_rate": 9.494135601461469e-06,
+      "loss": 0.8926,
+      "step": 1298
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3092468976974487,
+      "learning_rate": 9.4932050626137e-06,
+      "loss": 0.8719,
+      "step": 1299
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.71452796459198,
+      "learning_rate": 9.492273714375963e-06,
+      "loss": 1.0977,
+      "step": 1300
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6875912547111511,
+      "learning_rate": 9.49134155691603e-06,
+      "loss": 1.093,
+      "step": 1301
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6646436452865601,
+      "learning_rate": 9.490408590401814e-06,
+      "loss": 1.0956,
+      "step": 1302
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5104724168777466,
+      "learning_rate": 9.489474815001377e-06,
+      "loss": 0.9024,
+      "step": 1303
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5155185461044312,
+      "learning_rate": 9.488540230882925e-06,
+      "loss": 0.9458,
+      "step": 1304
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3984274864196777,
+      "learning_rate": 9.487604838214812e-06,
+      "loss": 0.8499,
+      "step": 1305
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.36191725730896,
+      "learning_rate": 9.486668637165536e-06,
+      "loss": 0.9111,
+      "step": 1306
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2850044965744019,
+      "learning_rate": 9.485731627903743e-06,
+      "loss": 0.8724,
+      "step": 1307
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.281544804573059,
+      "learning_rate": 9.484793810598218e-06,
+      "loss": 0.8306,
+      "step": 1308
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3364924192428589,
+      "learning_rate": 9.4838551854179e-06,
+      "loss": 0.8172,
+      "step": 1309
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4317471981048584,
+      "learning_rate": 9.482915752531868e-06,
+      "loss": 1.0881,
+      "step": 1310
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.321443796157837,
+      "learning_rate": 9.481975512109349e-06,
+      "loss": 0.8572,
+      "step": 1311
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.390732765197754,
+      "learning_rate": 9.481034464319713e-06,
+      "loss": 0.8476,
+      "step": 1312
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3966625928878784,
+      "learning_rate": 9.480092609332477e-06,
+      "loss": 0.8668,
+      "step": 1313
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.767971932888031,
+      "learning_rate": 9.479149947317306e-06,
+      "loss": 1.0892,
+      "step": 1314
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4850214719772339,
+      "learning_rate": 9.478206478444006e-06,
+      "loss": 0.947,
+      "step": 1315
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8024415969848633,
+      "learning_rate": 9.477262202882528e-06,
+      "loss": 1.097,
+      "step": 1316
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6817196011543274,
+      "learning_rate": 9.476317120802974e-06,
+      "loss": 1.1142,
+      "step": 1317
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6716673374176025,
+      "learning_rate": 9.475371232375588e-06,
+      "loss": 1.0798,
+      "step": 1318
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3505096435546875,
+      "learning_rate": 9.474424537770755e-06,
+      "loss": 0.8969,
+      "step": 1319
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.745694637298584,
+      "learning_rate": 9.473477037159012e-06,
+      "loss": 0.8879,
+      "step": 1320
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3778886795043945,
+      "learning_rate": 9.472528730711039e-06,
+      "loss": 0.9307,
+      "step": 1321
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3326694965362549,
+      "learning_rate": 9.471579618597657e-06,
+      "loss": 0.8738,
+      "step": 1322
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4053153991699219,
+      "learning_rate": 9.47062970098984e-06,
+      "loss": 0.8665,
+      "step": 1323
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.3729604482650757,
+      "learning_rate": 9.4696789780587e-06,
+      "loss": 1.0726,
+      "step": 1324
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.0426945686340332,
+      "learning_rate": 9.4687274499755e-06,
+      "loss": 1.0653,
+      "step": 1325
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4546116590499878,
+      "learning_rate": 9.46777511691164e-06,
+      "loss": 0.8726,
+      "step": 1326
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4583364725112915,
+      "learning_rate": 9.466821979038675e-06,
+      "loss": 0.8984,
+      "step": 1327
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.7383337020874023,
+      "learning_rate": 9.465868036528295e-06,
+      "loss": 0.9101,
+      "step": 1328
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4172909259796143,
+      "learning_rate": 9.464913289552342e-06,
+      "loss": 0.8162,
+      "step": 1329
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9685325622558594,
+      "learning_rate": 9.463957738282802e-06,
+      "loss": 1.0883,
+      "step": 1330
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.382623314857483,
+      "learning_rate": 9.463001382891805e-06,
+      "loss": 0.9598,
+      "step": 1331
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.4497069120407104,
+      "learning_rate": 9.462044223551621e-06,
+      "loss": 0.8893,
+      "step": 1332
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.9082592725753784,
+      "learning_rate": 9.461086260434674e-06,
+      "loss": 0.9184,
+      "step": 1333
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.8044934272766113,
+      "learning_rate": 9.460127493713526e-06,
+      "loss": 0.8863,
+      "step": 1334
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.5492290258407593,
+      "learning_rate": 9.459167923560886e-06,
+      "loss": 0.8706,
+      "step": 1335
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5222903490066528,
+      "learning_rate": 9.458207550149608e-06,
+      "loss": 0.9343,
+      "step": 1336
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4332942962646484,
+      "learning_rate": 9.45724637365269e-06,
+      "loss": 0.9132,
+      "step": 1337
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9093112349510193,
+      "learning_rate": 9.456284394243273e-06,
+      "loss": 1.0725,
+      "step": 1338
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4749616384506226,
+      "learning_rate": 9.455321612094649e-06,
+      "loss": 0.9475,
+      "step": 1339
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3656666278839111,
+      "learning_rate": 9.454358027380246e-06,
+      "loss": 0.9001,
+      "step": 1340
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4315742254257202,
+      "learning_rate": 9.453393640273642e-06,
+      "loss": 0.893,
+      "step": 1341
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.58907151222229,
+      "learning_rate": 9.452428450948562e-06,
+      "loss": 0.9024,
+      "step": 1342
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4966959953308105,
+      "learning_rate": 9.451462459578864e-06,
+      "loss": 0.9696,
+      "step": 1343
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.105455756187439,
+      "learning_rate": 9.450495666338563e-06,
+      "loss": 1.0737,
+      "step": 1344
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.7757881879806519,
+      "learning_rate": 9.449528071401816e-06,
+      "loss": 0.9145,
+      "step": 1345
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5635273456573486,
+      "learning_rate": 9.448559674942916e-06,
+      "loss": 0.9105,
+      "step": 1346
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3736313581466675,
+      "learning_rate": 9.447590477136312e-06,
+      "loss": 0.9136,
+      "step": 1347
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7640069723129272,
+      "learning_rate": 9.446620478156589e-06,
+      "loss": 1.0808,
+      "step": 1348
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3646495342254639,
+      "learning_rate": 9.445649678178482e-06,
+      "loss": 0.8851,
+      "step": 1349
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7911260724067688,
+      "learning_rate": 9.444678077376862e-06,
+      "loss": 1.0936,
+      "step": 1350
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7448989748954773,
+      "learning_rate": 9.443705675926755e-06,
+      "loss": 1.0484,
+      "step": 1351
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.460054636001587,
+      "learning_rate": 9.442732474003323e-06,
+      "loss": 0.924,
+      "step": 1352
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3592928647994995,
+      "learning_rate": 9.441758471781876e-06,
+      "loss": 0.879,
+      "step": 1353
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4203429222106934,
+      "learning_rate": 9.440783669437868e-06,
+      "loss": 0.8826,
+      "step": 1354
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9128873348236084,
+      "learning_rate": 9.439808067146895e-06,
+      "loss": 1.1197,
+      "step": 1355
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8230389356613159,
+      "learning_rate": 9.4388316650847e-06,
+      "loss": 1.0753,
+      "step": 1356
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.2840808629989624,
+      "learning_rate": 9.437854463427168e-06,
+      "loss": 0.8768,
+      "step": 1357
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4542776346206665,
+      "learning_rate": 9.43687646235033e-06,
+      "loss": 0.8925,
+      "step": 1358
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8328933715820312,
+      "learning_rate": 9.435897662030354e-06,
+      "loss": 1.0797,
+      "step": 1359
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3390213251113892,
+      "learning_rate": 9.434918062643564e-06,
+      "loss": 0.8924,
+      "step": 1360
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7426146864891052,
+      "learning_rate": 9.433937664366419e-06,
+      "loss": 1.0817,
+      "step": 1361
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6534510850906372,
+      "learning_rate": 9.432956467375525e-06,
+      "loss": 0.9531,
+      "step": 1362
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3052959442138672,
+      "learning_rate": 9.431974471847633e-06,
+      "loss": 0.8903,
+      "step": 1363
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3234848976135254,
+      "learning_rate": 9.430991677959632e-06,
+      "loss": 0.9129,
+      "step": 1364
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9204829931259155,
+      "learning_rate": 9.430008085888563e-06,
+      "loss": 1.1096,
+      "step": 1365
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7500194907188416,
+      "learning_rate": 9.429023695811603e-06,
+      "loss": 1.0866,
+      "step": 1366
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3860403299331665,
+      "learning_rate": 9.428038507906079e-06,
+      "loss": 0.8582,
+      "step": 1367
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7688742876052856,
+      "learning_rate": 9.42705252234946e-06,
+      "loss": 1.0618,
+      "step": 1368
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.444994330406189,
+      "learning_rate": 9.426065739319356e-06,
+      "loss": 0.8719,
+      "step": 1369
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6418555974960327,
+      "learning_rate": 9.425078158993524e-06,
+      "loss": 0.9056,
+      "step": 1370
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4516242742538452,
+      "learning_rate": 9.424089781549862e-06,
+      "loss": 0.9028,
+      "step": 1371
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4327666759490967,
+      "learning_rate": 9.423100607166413e-06,
+      "loss": 0.8992,
+      "step": 1372
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.482205867767334,
+      "learning_rate": 9.422110636021365e-06,
+      "loss": 0.906,
+      "step": 1373
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3948973417282104,
+      "learning_rate": 9.421119868293045e-06,
+      "loss": 0.8449,
+      "step": 1374
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4532030820846558,
+      "learning_rate": 9.420128304159927e-06,
+      "loss": 0.8947,
+      "step": 1375
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.539316177368164,
+      "learning_rate": 9.41913594380063e-06,
+      "loss": 0.9068,
+      "step": 1376
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5254619121551514,
+      "learning_rate": 9.418142787393913e-06,
+      "loss": 0.8874,
+      "step": 1377
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4765911102294922,
+      "learning_rate": 9.417148835118681e-06,
+      "loss": 0.8998,
+      "step": 1378
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4571938514709473,
+      "learning_rate": 9.416154087153978e-06,
+      "loss": 0.9407,
+      "step": 1379
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.374051809310913,
+      "learning_rate": 9.415158543678998e-06,
+      "loss": 0.9003,
+      "step": 1380
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0537854433059692,
+      "learning_rate": 9.41416220487307e-06,
+      "loss": 1.1126,
+      "step": 1381
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8440405130386353,
+      "learning_rate": 9.413165070915676e-06,
+      "loss": 1.0963,
+      "step": 1382
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5474193096160889,
+      "learning_rate": 9.412167141986433e-06,
+      "loss": 0.925,
+      "step": 1383
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.604691743850708,
+      "learning_rate": 9.411168418265105e-06,
+      "loss": 0.9252,
+      "step": 1384
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4557229280471802,
+      "learning_rate": 9.410168899931599e-06,
+      "loss": 0.864,
+      "step": 1385
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.46400785446167,
+      "learning_rate": 9.409168587165964e-06,
+      "loss": 0.8364,
+      "step": 1386
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.6388304233551025,
+      "learning_rate": 9.408167480148394e-06,
+      "loss": 1.105,
+      "step": 1387
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4678694009780884,
+      "learning_rate": 9.407165579059224e-06,
+      "loss": 0.9092,
+      "step": 1388
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5385395288467407,
+      "learning_rate": 9.406162884078932e-06,
+      "loss": 0.8821,
+      "step": 1389
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.0952048301696777,
+      "learning_rate": 9.405159395388141e-06,
+      "loss": 1.0983,
+      "step": 1390
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.417761206626892,
+      "learning_rate": 9.404155113167617e-06,
+      "loss": 0.8502,
+      "step": 1391
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.43987238407135,
+      "learning_rate": 9.403150037598266e-06,
+      "loss": 0.9145,
+      "step": 1392
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4155848026275635,
+      "learning_rate": 9.402144168861139e-06,
+      "loss": 0.8691,
+      "step": 1393
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4633870124816895,
+      "learning_rate": 9.401137507137428e-06,
+      "loss": 0.8821,
+      "step": 1394
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.453886866569519,
+      "learning_rate": 9.400130052608473e-06,
+      "loss": 0.9468,
+      "step": 1395
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 5.635190486907959,
+      "learning_rate": 9.399121805455753e-06,
+      "loss": 0.866,
+      "step": 1396
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4808626174926758,
+      "learning_rate": 9.398112765860887e-06,
+      "loss": 1.1007,
+      "step": 1397
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4928480386734009,
+      "learning_rate": 9.397102934005642e-06,
+      "loss": 0.9151,
+      "step": 1398
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.291694164276123,
+      "learning_rate": 9.396092310071924e-06,
+      "loss": 0.9059,
+      "step": 1399
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9766984581947327,
+      "learning_rate": 9.395080894241786e-06,
+      "loss": 1.0966,
+      "step": 1400
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7692593932151794,
+      "learning_rate": 9.394068686697418e-06,
+      "loss": 1.1148,
+      "step": 1401
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7152684330940247,
+      "learning_rate": 9.393055687621155e-06,
+      "loss": 1.0971,
+      "step": 1402
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8813909888267517,
+      "learning_rate": 9.392041897195478e-06,
+      "loss": 1.0762,
+      "step": 1403
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 2.2798526287078857,
+      "learning_rate": 9.391027315603007e-06,
+      "loss": 0.8757,
+      "step": 1404
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.5632692575454712,
+      "learning_rate": 9.390011943026505e-06,
+      "loss": 0.9383,
+      "step": 1405
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.4570841789245605,
+      "learning_rate": 9.388995779648875e-06,
+      "loss": 0.9076,
+      "step": 1406
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.1996790170669556,
+      "learning_rate": 9.387978825653169e-06,
+      "loss": 1.0769,
+      "step": 1407
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 1.3912982940673828,
+      "learning_rate": 9.386961081222572e-06,
+      "loss": 0.9453,
+      "step": 1408
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.874078631401062,
+      "learning_rate": 9.38594254654042e-06,
+      "loss": 1.052,
+      "step": 1409
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.78174889087677,
+      "learning_rate": 9.384923221790192e-06,
+      "loss": 1.0828,
+      "step": 1410
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8021097779273987,
+      "learning_rate": 9.3839031071555e-06,
+      "loss": 1.0632,
+      "step": 1411
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3891319036483765,
+      "learning_rate": 9.382882202820107e-06,
+      "loss": 0.8775,
+      "step": 1412
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8603055477142334,
+      "learning_rate": 9.381860508967913e-06,
+      "loss": 1.1126,
+      "step": 1413
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8401925563812256,
+      "learning_rate": 9.380838025782964e-06,
+      "loss": 1.0908,
+      "step": 1414
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.392820119857788,
+      "learning_rate": 9.379814753449445e-06,
+      "loss": 0.8725,
+      "step": 1415
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.373021125793457,
+      "learning_rate": 9.378790692151685e-06,
+      "loss": 0.901,
+      "step": 1416
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5193902254104614,
+      "learning_rate": 9.377765842074155e-06,
+      "loss": 0.8969,
+      "step": 1417
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3316855430603027,
+      "learning_rate": 9.376740203401468e-06,
+      "loss": 0.8986,
+      "step": 1418
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4249229431152344,
+      "learning_rate": 9.375713776318381e-06,
+      "loss": 0.8939,
+      "step": 1419
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2849328517913818,
+      "learning_rate": 9.374686561009786e-06,
+      "loss": 0.8955,
+      "step": 1420
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5859466791152954,
+      "learning_rate": 9.373658557660726e-06,
+      "loss": 0.8781,
+      "step": 1421
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.39251708984375,
+      "learning_rate": 9.37262976645638e-06,
+      "loss": 0.8931,
+      "step": 1422
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4505990743637085,
+      "learning_rate": 9.37160018758207e-06,
+      "loss": 1.0813,
+      "step": 1423
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4108266830444336,
+      "learning_rate": 9.370569821223264e-06,
+      "loss": 0.9068,
+      "step": 1424
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.9337676167488098,
+      "learning_rate": 9.369538667565566e-06,
+      "loss": 1.0965,
+      "step": 1425
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5195279121398926,
+      "learning_rate": 9.368506726794725e-06,
+      "loss": 0.9106,
+      "step": 1426
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5921649932861328,
+      "learning_rate": 9.36747399909663e-06,
+      "loss": 0.864,
+      "step": 1427
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4300220012664795,
+      "learning_rate": 9.366440484657316e-06,
+      "loss": 0.8593,
+      "step": 1428
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5511733293533325,
+      "learning_rate": 9.365406183662952e-06,
+      "loss": 0.8645,
+      "step": 1429
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5490772724151611,
+      "learning_rate": 9.364371096299855e-06,
+      "loss": 0.8496,
+      "step": 1430
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3683109283447266,
+      "learning_rate": 9.363335222754486e-06,
+      "loss": 0.9034,
+      "step": 1431
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5000412464141846,
+      "learning_rate": 9.36229856321344e-06,
+      "loss": 0.8757,
+      "step": 1432
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3616200685501099,
+      "learning_rate": 9.361261117863457e-06,
+      "loss": 0.932,
+      "step": 1433
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4639949798583984,
+      "learning_rate": 9.360222886891417e-06,
+      "loss": 0.9063,
+      "step": 1434
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.611850619316101,
+      "learning_rate": 9.359183870484347e-06,
+      "loss": 0.8813,
+      "step": 1435
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.664322018623352,
+      "learning_rate": 9.35814406882941e-06,
+      "loss": 1.0687,
+      "step": 1436
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3912842273712158,
+      "learning_rate": 9.357103482113912e-06,
+      "loss": 0.9227,
+      "step": 1437
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4821115732192993,
+      "learning_rate": 9.356062110525302e-06,
+      "loss": 0.8724,
+      "step": 1438
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.388408899307251,
+      "learning_rate": 9.355019954251165e-06,
+      "loss": 0.8743,
+      "step": 1439
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8800217509269714,
+      "learning_rate": 9.353977013479238e-06,
+      "loss": 1.0804,
+      "step": 1440
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5791373252868652,
+      "learning_rate": 9.352933288397387e-06,
+      "loss": 0.9799,
+      "step": 1441
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.5450150966644287,
+      "learning_rate": 9.351888779193626e-06,
+      "loss": 0.9279,
+      "step": 1442
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3928768634796143,
+      "learning_rate": 9.35084348605611e-06,
+      "loss": 0.933,
+      "step": 1443
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.294081687927246,
+      "learning_rate": 9.349797409173136e-06,
+      "loss": 0.8725,
+      "step": 1444
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2952684164047241,
+      "learning_rate": 9.348750548733138e-06,
+      "loss": 0.8766,
+      "step": 1445
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3956830501556396,
+      "learning_rate": 9.347702904924694e-06,
+      "loss": 0.8717,
+      "step": 1446
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3926458358764648,
+      "learning_rate": 9.346654477936523e-06,
+      "loss": 1.0737,
+      "step": 1447
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4245588779449463,
+      "learning_rate": 9.345605267957487e-06,
+      "loss": 0.8825,
+      "step": 1448
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4079110622406006,
+      "learning_rate": 9.344555275176585e-06,
+      "loss": 0.8923,
+      "step": 1449
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8620107173919678,
+      "learning_rate": 9.343504499782958e-06,
+      "loss": 1.0804,
+      "step": 1450
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4184831380844116,
+      "learning_rate": 9.342452941965892e-06,
+      "loss": 0.8602,
+      "step": 1451
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6957417130470276,
+      "learning_rate": 9.341400601914807e-06,
+      "loss": 1.0763,
+      "step": 1452
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3495999574661255,
+      "learning_rate": 9.340347479819272e-06,
+      "loss": 0.8722,
+      "step": 1453
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7774455547332764,
+      "learning_rate": 9.339293575868988e-06,
+      "loss": 1.065,
+      "step": 1454
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2873965501785278,
+      "learning_rate": 9.338238890253805e-06,
+      "loss": 0.9118,
+      "step": 1455
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3604868650436401,
+      "learning_rate": 9.33718342316371e-06,
+      "loss": 0.8486,
+      "step": 1456
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8723037242889404,
+      "learning_rate": 9.33612717478883e-06,
+      "loss": 1.0941,
+      "step": 1457
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7710858583450317,
+      "learning_rate": 9.335070145319433e-06,
+      "loss": 1.0783,
+      "step": 1458
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6487324237823486,
+      "learning_rate": 9.334012334945932e-06,
+      "loss": 1.0752,
+      "step": 1459
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7826824188232422,
+      "learning_rate": 9.332953743858874e-06,
+      "loss": 1.0829,
+      "step": 1460
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3682681322097778,
+      "learning_rate": 9.33189437224895e-06,
+      "loss": 0.903,
+      "step": 1461
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4611834287643433,
+      "learning_rate": 9.330834220306995e-06,
+      "loss": 0.9184,
+      "step": 1462
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3819847106933594,
+      "learning_rate": 9.329773288223976e-06,
+      "loss": 0.9095,
+      "step": 1463
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3947621583938599,
+      "learning_rate": 9.328711576191007e-06,
+      "loss": 0.8813,
+      "step": 1464
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3553097248077393,
+      "learning_rate": 9.327649084399341e-06,
+      "loss": 0.8219,
+      "step": 1465
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3520272970199585,
+      "learning_rate": 9.326585813040377e-06,
+      "loss": 0.9042,
+      "step": 1466
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4227609634399414,
+      "learning_rate": 9.325521762305639e-06,
+      "loss": 0.8748,
+      "step": 1467
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4682427644729614,
+      "learning_rate": 9.324456932386809e-06,
+      "loss": 0.8379,
+      "step": 1468
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4175275564193726,
+      "learning_rate": 9.323391323475699e-06,
+      "loss": 0.9637,
+      "step": 1469
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3851326704025269,
+      "learning_rate": 9.322324935764264e-06,
+      "loss": 0.9154,
+      "step": 1470
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4792391061782837,
+      "learning_rate": 9.321257769444598e-06,
+      "loss": 0.9022,
+      "step": 1471
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.348631501197815,
+      "learning_rate": 9.32018982470894e-06,
+      "loss": 0.8327,
+      "step": 1472
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3692190647125244,
+      "learning_rate": 9.319121101749664e-06,
+      "loss": 0.9296,
+      "step": 1473
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4083888530731201,
+      "learning_rate": 9.318051600759284e-06,
+      "loss": 0.9036,
+      "step": 1474
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.2941473722457886,
+      "learning_rate": 9.316981321930459e-06,
+      "loss": 0.9273,
+      "step": 1475
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3001328706741333,
+      "learning_rate": 9.315910265455983e-06,
+      "loss": 0.8833,
+      "step": 1476
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.412145733833313,
+      "learning_rate": 9.314838431528794e-06,
+      "loss": 0.8735,
+      "step": 1477
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3927537202835083,
+      "learning_rate": 9.313765820341968e-06,
+      "loss": 0.8476,
+      "step": 1478
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4199415445327759,
+      "learning_rate": 9.31269243208872e-06,
+      "loss": 1.0704,
+      "step": 1479
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4436583518981934,
+      "learning_rate": 9.311618266962408e-06,
+      "loss": 0.8533,
+      "step": 1480
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4382548332214355,
+      "learning_rate": 9.310543325156529e-06,
+      "loss": 0.9139,
+      "step": 1481
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4787451028823853,
+      "learning_rate": 9.309467606864715e-06,
+      "loss": 0.9136,
+      "step": 1482
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.3607559204101562,
+      "learning_rate": 9.308391112280747e-06,
+      "loss": 0.9007,
+      "step": 1483
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7760105729103088,
+      "learning_rate": 9.307313841598538e-06,
+      "loss": 1.0847,
+      "step": 1484
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7858136296272278,
+      "learning_rate": 9.306235795012144e-06,
+      "loss": 1.1057,
+      "step": 1485
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6864528059959412,
+      "learning_rate": 9.305156972715762e-06,
+      "loss": 1.0699,
+      "step": 1486
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 1.4545220136642456,
+      "learning_rate": 9.304077374903726e-06,
+      "loss": 0.9386,
+      "step": 1487
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8210257887840271,
+      "learning_rate": 9.302997001770511e-06,
+      "loss": 1.0658,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3217891454696655,
+      "learning_rate": 9.301915853510733e-06,
+      "loss": 0.8893,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3626608848571777,
+      "learning_rate": 9.300833930319142e-06,
+      "loss": 0.9489,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3336788415908813,
+      "learning_rate": 9.299751232390638e-06,
+      "loss": 0.9114,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8489395976066589,
+      "learning_rate": 9.298667759920248e-06,
+      "loss": 1.1035,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3911775350570679,
+      "learning_rate": 9.29758351310315e-06,
+      "loss": 0.929,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4109838008880615,
+      "learning_rate": 9.296498492134653e-06,
+      "loss": 0.8751,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3559167385101318,
+      "learning_rate": 9.29541269721021e-06,
+      "loss": 0.9021,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3628898859024048,
+      "learning_rate": 9.294326128525413e-06,
+      "loss": 0.9021,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4358856678009033,
+      "learning_rate": 9.29323878627599e-06,
+      "loss": 0.9117,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.605065941810608,
+      "learning_rate": 9.292150670657817e-06,
+      "loss": 0.8435,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3332918882369995,
+      "learning_rate": 9.291061781866896e-06,
+      "loss": 0.8858,
+      "step": 1499
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4154071807861328,
+      "learning_rate": 9.28997212009938e-06,
+      "loss": 0.8786,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2835302352905273,
+      "learning_rate": 9.288881685551554e-06,
+      "loss": 0.8954,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.159651517868042,
+      "learning_rate": 9.287790478419849e-06,
+      "loss": 0.8651,
+      "step": 1502
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.9815153479576111,
+      "learning_rate": 9.286698498900825e-06,
+      "loss": 1.0787,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5184550285339355,
+      "learning_rate": 9.285605747191193e-06,
+      "loss": 0.9159,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4485242366790771,
+      "learning_rate": 9.284512223487797e-06,
+      "loss": 0.8414,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3552143573760986,
+      "learning_rate": 9.283417927987616e-06,
+      "loss": 0.9217,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3659883737564087,
+      "learning_rate": 9.282322860887778e-06,
+      "loss": 0.9313,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8740627765655518,
+      "learning_rate": 9.28122702238554e-06,
+      "loss": 1.0947,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5149987936019897,
+      "learning_rate": 9.280130412678306e-06,
+      "loss": 0.9065,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7317695617675781,
+      "learning_rate": 9.27903303196361e-06,
+      "loss": 1.1128,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3492783308029175,
+      "learning_rate": 9.277934880439138e-06,
+      "loss": 0.915,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3747704029083252,
+      "learning_rate": 9.276835958302702e-06,
+      "loss": 0.882,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3464101552963257,
+      "learning_rate": 9.275736265752257e-06,
+      "loss": 0.8512,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3705521821975708,
+      "learning_rate": 9.2746358029859e-06,
+      "loss": 0.9382,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3838802576065063,
+      "learning_rate": 9.273534570201867e-06,
+      "loss": 0.9315,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2976751327514648,
+      "learning_rate": 9.272432567598523e-06,
+      "loss": 0.8847,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.306828260421753,
+      "learning_rate": 9.271329795374386e-06,
+      "loss": 0.886,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.292664647102356,
+      "learning_rate": 9.270226253728104e-06,
+      "loss": 0.9211,
+      "step": 1518
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4017263650894165,
+      "learning_rate": 9.26912194285846e-06,
+      "loss": 0.8544,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.300333023071289,
+      "learning_rate": 9.268016862964386e-06,
+      "loss": 0.8051,
+      "step": 1520
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.2590912580490112,
+      "learning_rate": 9.266911014244946e-06,
+      "loss": 0.7678,
+      "step": 1521
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3055387735366821,
+      "learning_rate": 9.26580439689934e-06,
+      "loss": 0.8247,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.495570421218872,
+      "learning_rate": 9.264697011126916e-06,
+      "loss": 1.0898,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0813838243484497,
+      "learning_rate": 9.263588857127152e-06,
+      "loss": 1.1009,
+      "step": 1524
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.369903564453125,
+      "learning_rate": 9.262479935099665e-06,
+      "loss": 0.8494,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.982667088508606,
+      "learning_rate": 9.261370245244214e-06,
+      "loss": 0.8519,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5194462537765503,
+      "learning_rate": 9.260259787760695e-06,
+      "loss": 0.8808,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4303983449935913,
+      "learning_rate": 9.25914856284914e-06,
+      "loss": 0.9409,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.427566647529602,
+      "learning_rate": 9.258036570709724e-06,
+      "loss": 0.8989,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.9340436458587646,
+      "learning_rate": 9.256923811542755e-06,
+      "loss": 1.1164,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3260940313339233,
+      "learning_rate": 9.255810285548684e-06,
+      "loss": 0.8363,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.7749673128128052,
+      "learning_rate": 9.254695992928094e-06,
+      "loss": 0.8399,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.462605595588684,
+      "learning_rate": 9.253580933881711e-06,
+      "loss": 0.8718,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4949884414672852,
+      "learning_rate": 9.252465108610397e-06,
+      "loss": 0.8999,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4554671049118042,
+      "learning_rate": 9.251348517315158e-06,
+      "loss": 0.8884,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3538055419921875,
+      "learning_rate": 9.250231160197127e-06,
+      "loss": 0.8476,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3548195362091064,
+      "learning_rate": 9.249113037457582e-06,
+      "loss": 0.9187,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.373862624168396,
+      "learning_rate": 9.247994149297938e-06,
+      "loss": 0.8825,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0401777029037476,
+      "learning_rate": 9.246874495919749e-06,
+      "loss": 1.1043,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.527527928352356,
+      "learning_rate": 9.245754077524702e-06,
+      "loss": 0.915,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3834123611450195,
+      "learning_rate": 9.244632894314628e-06,
+      "loss": 0.886,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.746333658695221,
+      "learning_rate": 9.243510946491494e-06,
+      "loss": 1.1025,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4447660446166992,
+      "learning_rate": 9.242388234257402e-06,
+      "loss": 0.891,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3532090187072754,
+      "learning_rate": 9.241264757814591e-06,
+      "loss": 0.856,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3899813890457153,
+      "learning_rate": 9.240140517365445e-06,
+      "loss": 0.8911,
+      "step": 1545
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.030625343322754,
+      "learning_rate": 9.239015513112476e-06,
+      "loss": 1.0742,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7995862364768982,
+      "learning_rate": 9.237889745258342e-06,
+      "loss": 1.0623,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6377469301223755,
+      "learning_rate": 9.236763214005832e-06,
+      "loss": 1.1079,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3890396356582642,
+      "learning_rate": 9.235635919557876e-06,
+      "loss": 0.8545,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3755342960357666,
+      "learning_rate": 9.234507862117542e-06,
+      "loss": 0.8611,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.500388264656067,
+      "learning_rate": 9.233379041888034e-06,
+      "loss": 0.8634,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3243639469146729,
+      "learning_rate": 9.232249459072693e-06,
+      "loss": 0.9361,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.6002277135849,
+      "learning_rate": 9.231119113874998e-06,
+      "loss": 1.0831,
+      "step": 1553
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3572239875793457,
+      "learning_rate": 9.229988006498564e-06,
+      "loss": 1.0958,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8556057810783386,
+      "learning_rate": 9.228856137147149e-06,
+      "loss": 1.0578,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.5188156366348267,
+      "learning_rate": 9.22772350602464e-06,
+      "loss": 0.8799,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.503766417503357,
+      "learning_rate": 9.226590113335065e-06,
+      "loss": 0.9006,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.742444396018982,
+      "learning_rate": 9.225455959282592e-06,
+      "loss": 0.9094,
+      "step": 1558
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4099773168563843,
+      "learning_rate": 9.224321044071522e-06,
+      "loss": 0.8695,
+      "step": 1559
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.319807529449463,
+      "learning_rate": 9.223185367906293e-06,
+      "loss": 0.8262,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.3830478191375732,
+      "learning_rate": 9.222048930991485e-06,
+      "loss": 1.11,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4534220695495605,
+      "learning_rate": 9.22091173353181e-06,
+      "loss": 0.8658,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4679570198059082,
+      "learning_rate": 9.21977377573212e-06,
+      "loss": 0.8231,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.4076354503631592,
+      "learning_rate": 9.218635057797399e-06,
+      "loss": 0.9201,
+      "step": 1564
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3884029388427734,
+      "learning_rate": 9.217495579932775e-06,
+      "loss": 0.8507,
+      "step": 1565
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4839295148849487,
+      "learning_rate": 9.216355342343508e-06,
+      "loss": 0.9037,
+      "step": 1566
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5924797058105469,
+      "learning_rate": 9.215214345234998e-06,
+      "loss": 0.8232,
+      "step": 1567
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.6230063438415527,
+      "learning_rate": 9.214072588812779e-06,
+      "loss": 1.1009,
+      "step": 1568
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4138773679733276,
+      "learning_rate": 9.21293007328252e-06,
+      "loss": 0.8873,
+      "step": 1569
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3834443092346191,
+      "learning_rate": 9.211786798850036e-06,
+      "loss": 0.9379,
+      "step": 1570
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4402408599853516,
+      "learning_rate": 9.210642765721267e-06,
+      "loss": 0.854,
+      "step": 1571
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0731405019760132,
+      "learning_rate": 9.209497974102298e-06,
+      "loss": 1.0645,
+      "step": 1572
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4113227128982544,
+      "learning_rate": 9.208352424199347e-06,
+      "loss": 0.9337,
+      "step": 1573
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2471121549606323,
+      "learning_rate": 9.207206116218769e-06,
+      "loss": 0.892,
+      "step": 1574
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8204759955406189,
+      "learning_rate": 9.206059050367055e-06,
+      "loss": 1.066,
+      "step": 1575
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7964875102043152,
+      "learning_rate": 9.204911226850835e-06,
+      "loss": 1.062,
+      "step": 1576
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3801742792129517,
+      "learning_rate": 9.203762645876872e-06,
+      "loss": 0.8995,
+      "step": 1577
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.382960557937622,
+      "learning_rate": 9.202613307652068e-06,
+      "loss": 0.8765,
+      "step": 1578
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.496346116065979,
+      "learning_rate": 9.201463212383463e-06,
+      "loss": 0.9036,
+      "step": 1579
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3555471897125244,
+      "learning_rate": 9.200312360278227e-06,
+      "loss": 0.8856,
+      "step": 1580
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.423231840133667,
+      "learning_rate": 9.199160751543674e-06,
+      "loss": 0.8737,
+      "step": 1581
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.432507872581482,
+      "learning_rate": 9.19800838638725e-06,
+      "loss": 0.9131,
+      "step": 1582
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0106227397918701,
+      "learning_rate": 9.196855265016535e-06,
+      "loss": 1.0838,
+      "step": 1583
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8685385584831238,
+      "learning_rate": 9.195701387639253e-06,
+      "loss": 1.071,
+      "step": 1584
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6778897643089294,
+      "learning_rate": 9.194546754463255e-06,
+      "loss": 1.097,
+      "step": 1585
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.352664589881897,
+      "learning_rate": 9.193391365696534e-06,
+      "loss": 0.8571,
+      "step": 1586
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4248247146606445,
+      "learning_rate": 9.192235221547218e-06,
+      "loss": 0.9174,
+      "step": 1587
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.0059901475906372,
+      "learning_rate": 9.191078322223572e-06,
+      "loss": 1.0611,
+      "step": 1588
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.5397629737854004,
+      "learning_rate": 9.189920667933991e-06,
+      "loss": 0.9103,
+      "step": 1589
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.375553011894226,
+      "learning_rate": 9.188762258887016e-06,
+      "loss": 0.8699,
+      "step": 1590
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8771775960922241,
+      "learning_rate": 9.187603095291318e-06,
+      "loss": 1.0831,
+      "step": 1591
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4854158163070679,
+      "learning_rate": 9.186443177355702e-06,
+      "loss": 0.877,
+      "step": 1592
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3738826513290405,
+      "learning_rate": 9.185282505289113e-06,
+      "loss": 0.9343,
+      "step": 1593
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8099785447120667,
+      "learning_rate": 9.184121079300629e-06,
+      "loss": 1.0855,
+      "step": 1594
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.353387475013733,
+      "learning_rate": 9.182958899599466e-06,
+      "loss": 0.8358,
+      "step": 1595
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4026700258255005,
+      "learning_rate": 9.181795966394977e-06,
+      "loss": 0.9039,
+      "step": 1596
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2954083681106567,
+      "learning_rate": 9.180632279896646e-06,
+      "loss": 0.8678,
+      "step": 1597
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4031389951705933,
+      "learning_rate": 9.179467840314096e-06,
+      "loss": 0.9239,
+      "step": 1598
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4593234062194824,
+      "learning_rate": 9.178302647857085e-06,
+      "loss": 0.8684,
+      "step": 1599
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3668723106384277,
+      "learning_rate": 9.177136702735507e-06,
+      "loss": 0.8547,
+      "step": 1600
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3391920328140259,
+      "learning_rate": 9.17597000515939e-06,
+      "loss": 0.832,
+      "step": 1601
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.422023057937622,
+      "learning_rate": 9.174802555338901e-06,
+      "loss": 0.9145,
+      "step": 1602
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.050050973892212,
+      "learning_rate": 9.17363435348434e-06,
+      "loss": 1.0798,
+      "step": 1603
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4436248540878296,
+      "learning_rate": 9.172465399806141e-06,
+      "loss": 0.9453,
+      "step": 1604
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.396449327468872,
+      "learning_rate": 9.171295694514875e-06,
+      "loss": 0.874,
+      "step": 1605
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4348831176757812,
+      "learning_rate": 9.17012523782125e-06,
+      "loss": 0.858,
+      "step": 1606
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7905135750770569,
+      "learning_rate": 9.168954029936105e-06,
+      "loss": 1.1058,
+      "step": 1607
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4348032474517822,
+      "learning_rate": 9.167782071070423e-06,
+      "loss": 0.8395,
+      "step": 1608
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3510074615478516,
+      "learning_rate": 9.16660936143531e-06,
+      "loss": 0.9025,
+      "step": 1609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3749746084213257,
+      "learning_rate": 9.165435901242019e-06,
+      "loss": 0.8696,
+      "step": 1610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4665056467056274,
+      "learning_rate": 9.16426169070193e-06,
+      "loss": 0.9023,
+      "step": 1611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9933146834373474,
+      "learning_rate": 9.163086730026562e-06,
+      "loss": 1.0823,
+      "step": 1612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2851201295852661,
+      "learning_rate": 9.16191101942757e-06,
+      "loss": 0.8429,
+      "step": 1613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2621235847473145,
+      "learning_rate": 9.160734559116738e-06,
+      "loss": 0.8441,
+      "step": 1614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.34567391872406,
+      "learning_rate": 9.159557349305991e-06,
+      "loss": 0.9173,
+      "step": 1615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4764485359191895,
+      "learning_rate": 9.158379390207389e-06,
+      "loss": 0.8809,
+      "step": 1616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4421414136886597,
+      "learning_rate": 9.157200682033126e-06,
+      "loss": 0.9651,
+      "step": 1617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4001466035842896,
+      "learning_rate": 9.156021224995526e-06,
+      "loss": 0.8531,
+      "step": 1618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.312121868133545,
+      "learning_rate": 9.154841019307057e-06,
+      "loss": 0.8407,
+      "step": 1619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.274419903755188,
+      "learning_rate": 9.153660065180314e-06,
+      "loss": 0.8329,
+      "step": 1620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3573837280273438,
+      "learning_rate": 9.15247836282803e-06,
+      "loss": 0.8593,
+      "step": 1621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3484364748001099,
+      "learning_rate": 9.151295912463076e-06,
+      "loss": 0.9227,
+      "step": 1622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9985756278038025,
+      "learning_rate": 9.150112714298451e-06,
+      "loss": 1.082,
+      "step": 1623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3726718425750732,
+      "learning_rate": 9.148928768547292e-06,
+      "loss": 0.8929,
+      "step": 1624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3635265827178955,
+      "learning_rate": 9.147744075422874e-06,
+      "loss": 0.8927,
+      "step": 1625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4111502170562744,
+      "learning_rate": 9.1465586351386e-06,
+      "loss": 0.928,
+      "step": 1626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 2.7528836727142334,
+      "learning_rate": 9.145372447908011e-06,
+      "loss": 0.8371,
+      "step": 1627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.9010924696922302,
+      "learning_rate": 9.144185513944785e-06,
+      "loss": 1.0718,
+      "step": 1628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3650360107421875,
+      "learning_rate": 9.142997833462733e-06,
+      "loss": 0.8429,
+      "step": 1629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7302423715591431,
+      "learning_rate": 9.141809406675794e-06,
+      "loss": 1.0986,
+      "step": 1630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.2866642475128174,
+      "learning_rate": 9.140620233798053e-06,
+      "loss": 0.8217,
+      "step": 1631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7973983287811279,
+      "learning_rate": 9.139430315043718e-06,
+      "loss": 1.0585,
+      "step": 1632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3965994119644165,
+      "learning_rate": 9.13823965062714e-06,
+      "loss": 0.9253,
+      "step": 1633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3784170150756836,
+      "learning_rate": 9.137048240762802e-06,
+      "loss": 0.8585,
+      "step": 1634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8198153376579285,
+      "learning_rate": 9.135856085665317e-06,
+      "loss": 1.0752,
+      "step": 1635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.3953344821929932,
+      "learning_rate": 9.13466318554944e-06,
+      "loss": 0.9315,
+      "step": 1636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 1.4435492753982544,
+      "learning_rate": 9.133469540630051e-06,
+      "loss": 0.8481,
+      "step": 1637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7648250460624695,
+      "learning_rate": 9.132275151122173e-06,
+      "loss": 1.0759,
+      "step": 1638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6732589602470398,
+      "learning_rate": 9.131080017240956e-06,
+      "loss": 1.0752,
+      "step": 1639
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6185219287872314,
+      "learning_rate": 9.129884139201688e-06,
+      "loss": 1.0683,
+      "step": 1640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0411877632141113,
+      "learning_rate": 9.12868751721979e-06,
+      "loss": 0.8905,
+      "step": 1641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5077942609786987,
+      "learning_rate": 9.127490151510819e-06,
+      "loss": 0.9128,
+      "step": 1642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2355434894561768,
+      "learning_rate": 9.126292042290464e-06,
+      "loss": 0.9007,
+      "step": 1643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3023957014083862,
+      "learning_rate": 9.125093189774545e-06,
+      "loss": 0.9232,
+      "step": 1644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2591451406478882,
+      "learning_rate": 9.123893594179022e-06,
+      "loss": 0.822,
+      "step": 1645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4061049222946167,
+      "learning_rate": 9.122693255719986e-06,
+      "loss": 0.906,
+      "step": 1646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4762135744094849,
+      "learning_rate": 9.121492174613659e-06,
+      "loss": 0.891,
+      "step": 1647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4361082315444946,
+      "learning_rate": 9.120290351076402e-06,
+      "loss": 0.8941,
+      "step": 1648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4688752889633179,
+      "learning_rate": 9.119087785324708e-06,
+      "loss": 0.9154,
+      "step": 1649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9649577736854553,
+      "learning_rate": 9.117884477575198e-06,
+      "loss": 1.0463,
+      "step": 1650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4406909942626953,
+      "learning_rate": 9.116680428044638e-06,
+      "loss": 0.9096,
+      "step": 1651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7095515727996826,
+      "learning_rate": 9.115475636949916e-06,
+      "loss": 1.079,
+      "step": 1652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4496010541915894,
+      "learning_rate": 9.114270104508063e-06,
+      "loss": 0.9527,
+      "step": 1653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.0250868797302246,
+      "learning_rate": 9.113063830936234e-06,
+      "loss": 0.8284,
+      "step": 1654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3792530298233032,
+      "learning_rate": 9.11185681645173e-06,
+      "loss": 0.871,
+      "step": 1655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3666672706604004,
+      "learning_rate": 9.110649061271971e-06,
+      "loss": 0.82,
+      "step": 1656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3423936367034912,
+      "learning_rate": 9.109440565614522e-06,
+      "loss": 0.8906,
+      "step": 1657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.0301214456558228,
+      "learning_rate": 9.108231329697077e-06,
+      "loss": 1.0966,
+      "step": 1658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3555583953857422,
+      "learning_rate": 9.107021353737463e-06,
+      "loss": 0.8875,
+      "step": 1659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.505197286605835,
+      "learning_rate": 9.10581063795364e-06,
+      "loss": 0.8613,
+      "step": 1660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7300689816474915,
+      "learning_rate": 9.104599182563701e-06,
+      "loss": 1.0924,
+      "step": 1661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3723580837249756,
+      "learning_rate": 9.103386987785876e-06,
+      "loss": 0.9102,
+      "step": 1662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.35226309299469,
+      "learning_rate": 9.102174053838523e-06,
+      "loss": 0.872,
+      "step": 1663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8183510899543762,
+      "learning_rate": 9.100960380940138e-06,
+      "loss": 1.0905,
+      "step": 1664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.448512077331543,
+      "learning_rate": 9.099745969309346e-06,
+      "loss": 0.9444,
+      "step": 1665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4092351198196411,
+      "learning_rate": 9.098530819164907e-06,
+      "loss": 0.896,
+      "step": 1666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.341110348701477,
+      "learning_rate": 9.097314930725717e-06,
+      "loss": 0.8591,
+      "step": 1667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4970694780349731,
+      "learning_rate": 9.096098304210794e-06,
+      "loss": 0.953,
+      "step": 1668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.368952989578247,
+      "learning_rate": 9.094880939839303e-06,
+      "loss": 0.88,
+      "step": 1669
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5930558443069458,
+      "learning_rate": 9.093662837830537e-06,
+      "loss": 0.9407,
+      "step": 1670
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.904436469078064,
+      "learning_rate": 9.092443998403915e-06,
+      "loss": 0.8265,
+      "step": 1671
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8439142107963562,
+      "learning_rate": 9.091224421779e-06,
+      "loss": 1.066,
+      "step": 1672
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3692028522491455,
+      "learning_rate": 9.090004108175476e-06,
+      "loss": 0.8946,
+      "step": 1673
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3541215658187866,
+      "learning_rate": 9.088783057813172e-06,
+      "loss": 0.8652,
+      "step": 1674
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.47551691532135,
+      "learning_rate": 9.08756127091204e-06,
+      "loss": 0.8692,
+      "step": 1675
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4315286874771118,
+      "learning_rate": 9.08633874769217e-06,
+      "loss": 0.82,
+      "step": 1676
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5274277925491333,
+      "learning_rate": 9.085115488373783e-06,
+      "loss": 0.8277,
+      "step": 1677
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9247099757194519,
+      "learning_rate": 9.08389149317723e-06,
+      "loss": 1.0752,
+      "step": 1678
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4582631587982178,
+      "learning_rate": 9.082666762323e-06,
+      "loss": 0.8701,
+      "step": 1679
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4599374532699585,
+      "learning_rate": 9.081441296031713e-06,
+      "loss": 0.8588,
+      "step": 1680
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5061408281326294,
+      "learning_rate": 9.080215094524116e-06,
+      "loss": 0.8581,
+      "step": 1681
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.297629475593567,
+      "learning_rate": 9.078988158021094e-06,
+      "loss": 0.8308,
+      "step": 1682
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4117028713226318,
+      "learning_rate": 9.077760486743663e-06,
+      "loss": 0.8954,
+      "step": 1683
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.9670777320861816,
+      "learning_rate": 9.076532080912974e-06,
+      "loss": 1.0893,
+      "step": 1684
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.349969744682312,
+      "learning_rate": 9.075302940750304e-06,
+      "loss": 0.8985,
+      "step": 1685
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2970057725906372,
+      "learning_rate": 9.074073066477068e-06,
+      "loss": 0.83,
+      "step": 1686
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5461856126785278,
+      "learning_rate": 9.07284245831481e-06,
+      "loss": 0.8732,
+      "step": 1687
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.50886070728302,
+      "learning_rate": 9.071611116485209e-06,
+      "loss": 0.8671,
+      "step": 1688
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.357362151145935,
+      "learning_rate": 9.070379041210072e-06,
+      "loss": 0.8655,
+      "step": 1689
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4223928451538086,
+      "learning_rate": 9.069146232711343e-06,
+      "loss": 0.8377,
+      "step": 1690
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.44011390209198,
+      "learning_rate": 9.067912691211095e-06,
+      "loss": 0.8619,
+      "step": 1691
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.442953109741211,
+      "learning_rate": 9.066678416931533e-06,
+      "loss": 0.9094,
+      "step": 1692
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.6780025959014893,
+      "learning_rate": 9.065443410094997e-06,
+      "loss": 0.8718,
+      "step": 1693
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3386861085891724,
+      "learning_rate": 9.064207670923953e-06,
+      "loss": 0.8282,
+      "step": 1694
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.4780431985855103,
+      "learning_rate": 9.062971199641007e-06,
+      "loss": 0.8482,
+      "step": 1695
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5060054063796997,
+      "learning_rate": 9.061733996468889e-06,
+      "loss": 0.8882,
+      "step": 1696
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3938137292861938,
+      "learning_rate": 9.060496061630466e-06,
+      "loss": 0.8757,
+      "step": 1697
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 2.576958179473877,
+      "learning_rate": 9.059257395348736e-06,
+      "loss": 1.075,
+      "step": 1698
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.409332275390625,
+      "learning_rate": 9.058017997846825e-06,
+      "loss": 0.8681,
+      "step": 1699
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3238976001739502,
+      "learning_rate": 9.056777869347995e-06,
+      "loss": 0.8905,
+      "step": 1700
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3043630123138428,
+      "learning_rate": 9.05553701007564e-06,
+      "loss": 0.8477,
+      "step": 1701
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.384853482246399,
+      "learning_rate": 9.054295420253283e-06,
+      "loss": 0.8653,
+      "step": 1702
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3462328910827637,
+      "learning_rate": 9.05305310010458e-06,
+      "loss": 0.9346,
+      "step": 1703
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6672659516334534,
+      "learning_rate": 9.051810049853316e-06,
+      "loss": 1.0686,
+      "step": 1704
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6844951510429382,
+      "learning_rate": 9.050566269723412e-06,
+      "loss": 1.0815,
+      "step": 1705
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3779035806655884,
+      "learning_rate": 9.049321759938918e-06,
+      "loss": 0.91,
+      "step": 1706
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.670836329460144,
+      "learning_rate": 9.048076520724014e-06,
+      "loss": 0.8588,
+      "step": 1707
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2981195449829102,
+      "learning_rate": 9.046830552303012e-06,
+      "loss": 0.8669,
+      "step": 1708
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.418195366859436,
+      "learning_rate": 9.04558385490036e-06,
+      "loss": 0.8864,
+      "step": 1709
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7332640886306763,
+      "learning_rate": 9.044336428740632e-06,
+      "loss": 1.0628,
+      "step": 1710
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.5718847513198853,
+      "learning_rate": 9.043088274048535e-06,
+      "loss": 0.9284,
+      "step": 1711
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3900251388549805,
+      "learning_rate": 9.041839391048905e-06,
+      "loss": 0.9053,
+      "step": 1712
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.33114492893219,
+      "learning_rate": 9.040589779966716e-06,
+      "loss": 0.8849,
+      "step": 1713
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.405131220817566,
+      "learning_rate": 9.039339441027064e-06,
+      "loss": 0.887,
+      "step": 1714
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.3704776763916016,
+      "learning_rate": 9.038088374455181e-06,
+      "loss": 0.8915,
+      "step": 1715
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.2749296426773071,
+      "learning_rate": 9.036836580476432e-06,
+      "loss": 0.8485,
+      "step": 1716
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 1.340626835823059,
+      "learning_rate": 9.03558405931631e-06,
+      "loss": 0.8826,
+      "step": 1717
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.466511607170105,
+      "learning_rate": 9.034330811200438e-06,
+      "loss": 0.9045,
+      "step": 1718
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4100117683410645,
+      "learning_rate": 9.033076836354575e-06,
+      "loss": 0.8966,
+      "step": 1719
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3510836362838745,
+      "learning_rate": 9.031822135004604e-06,
+      "loss": 0.8952,
+      "step": 1720
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4781301021575928,
+      "learning_rate": 9.030566707376545e-06,
+      "loss": 0.9021,
+      "step": 1721
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7973368763923645,
+      "learning_rate": 9.029310553696545e-06,
+      "loss": 1.0773,
+      "step": 1722
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4822598695755005,
+      "learning_rate": 9.028053674190883e-06,
+      "loss": 0.894,
+      "step": 1723
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4074392318725586,
+      "learning_rate": 9.02679606908597e-06,
+      "loss": 0.8358,
+      "step": 1724
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.409791350364685,
+      "learning_rate": 9.025537738608344e-06,
+      "loss": 0.8725,
+      "step": 1725
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5821796655654907,
+      "learning_rate": 9.024278682984679e-06,
+      "loss": 0.8635,
+      "step": 1726
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.397955298423767,
+      "learning_rate": 9.023018902441774e-06,
+      "loss": 0.8808,
+      "step": 1727
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.7465827465057373,
+      "learning_rate": 9.021758397206565e-06,
+      "loss": 0.9361,
+      "step": 1728
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4271962642669678,
+      "learning_rate": 9.020497167506113e-06,
+      "loss": 0.8906,
+      "step": 1729
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3114113807678223,
+      "learning_rate": 9.01923521356761e-06,
+      "loss": 0.8797,
+      "step": 1730
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.351205825805664,
+      "learning_rate": 9.017972535618381e-06,
+      "loss": 0.8894,
+      "step": 1731
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2779748439788818,
+      "learning_rate": 9.01670913388588e-06,
+      "loss": 0.884,
+      "step": 1732
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.9739753007888794,
+      "learning_rate": 9.015445008597694e-06,
+      "loss": 0.8707,
+      "step": 1733
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3414291143417358,
+      "learning_rate": 9.014180159981536e-06,
+      "loss": 0.875,
+      "step": 1734
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4370055198669434,
+      "learning_rate": 9.01291458826525e-06,
+      "loss": 0.9397,
+      "step": 1735
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3338063955307007,
+      "learning_rate": 9.011648293676816e-06,
+      "loss": 0.8434,
+      "step": 1736
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3232618570327759,
+      "learning_rate": 9.010381276444336e-06,
+      "loss": 0.8932,
+      "step": 1737
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 2.968594551086426,
+      "learning_rate": 9.009113536796046e-06,
+      "loss": 0.91,
+      "step": 1738
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4628615379333496,
+      "learning_rate": 9.007845074960314e-06,
+      "loss": 0.8773,
+      "step": 1739
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9686388373374939,
+      "learning_rate": 9.006575891165633e-06,
+      "loss": 1.0803,
+      "step": 1740
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7599841356277466,
+      "learning_rate": 9.005305985640633e-06,
+      "loss": 1.0809,
+      "step": 1741
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6825365424156189,
+      "learning_rate": 9.00403535861407e-06,
+      "loss": 1.0607,
+      "step": 1742
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5153968334197998,
+      "learning_rate": 9.002764010314828e-06,
+      "loss": 0.9072,
+      "step": 1743
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9079862833023071,
+      "learning_rate": 9.001491940971922e-06,
+      "loss": 1.0722,
+      "step": 1744
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4535881280899048,
+      "learning_rate": 9.000219150814504e-06,
+      "loss": 0.8589,
+      "step": 1745
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.571048378944397,
+      "learning_rate": 8.998945640071844e-06,
+      "loss": 0.8951,
+      "step": 1746
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4458436965942383,
+      "learning_rate": 8.99767140897335e-06,
+      "loss": 0.8527,
+      "step": 1747
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0627233982086182,
+      "learning_rate": 8.996396457748556e-06,
+      "loss": 1.0874,
+      "step": 1748
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.534490704536438,
+      "learning_rate": 8.995120786627129e-06,
+      "loss": 0.895,
+      "step": 1749
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7990822792053223,
+      "learning_rate": 8.993844395838865e-06,
+      "loss": 1.0534,
+      "step": 1750
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5025299787521362,
+      "learning_rate": 8.992567285613686e-06,
+      "loss": 0.8542,
+      "step": 1751
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3729281425476074,
+      "learning_rate": 8.991289456181647e-06,
+      "loss": 0.7983,
+      "step": 1752
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9743713736534119,
+      "learning_rate": 8.99001090777293e-06,
+      "loss": 1.0548,
+      "step": 1753
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.5838356018066406,
+      "learning_rate": 8.98873164061785e-06,
+      "loss": 0.859,
+      "step": 1754
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4201873540878296,
+      "learning_rate": 8.98745165494685e-06,
+      "loss": 0.8651,
+      "step": 1755
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.52068293094635,
+      "learning_rate": 8.986170950990502e-06,
+      "loss": 0.8562,
+      "step": 1756
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.402377963066101,
+      "learning_rate": 8.984889528979505e-06,
+      "loss": 0.8773,
+      "step": 1757
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3955827951431274,
+      "learning_rate": 8.983607389144692e-06,
+      "loss": 0.8311,
+      "step": 1758
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4212640523910522,
+      "learning_rate": 8.982324531717024e-06,
+      "loss": 0.9458,
+      "step": 1759
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3713598251342773,
+      "learning_rate": 8.981040956927588e-06,
+      "loss": 0.8296,
+      "step": 1760
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3667383193969727,
+      "learning_rate": 8.979756665007604e-06,
+      "loss": 0.9,
+      "step": 1761
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.8398536443710327,
+      "learning_rate": 8.97847165618842e-06,
+      "loss": 0.8773,
+      "step": 1762
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6709331274032593,
+      "learning_rate": 8.977185930701513e-06,
+      "loss": 0.8822,
+      "step": 1763
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.306581974029541,
+      "learning_rate": 8.975899488778487e-06,
+      "loss": 0.8909,
+      "step": 1764
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.366308331489563,
+      "learning_rate": 8.974612330651078e-06,
+      "loss": 0.8855,
+      "step": 1765
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3681169748306274,
+      "learning_rate": 8.973324456551152e-06,
+      "loss": 0.9149,
+      "step": 1766
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4292452335357666,
+      "learning_rate": 8.972035866710699e-06,
+      "loss": 0.9014,
+      "step": 1767
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3034169673919678,
+      "learning_rate": 8.970746561361842e-06,
+      "loss": 0.9301,
+      "step": 1768
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4062244892120361,
+      "learning_rate": 8.969456540736835e-06,
+      "loss": 0.8563,
+      "step": 1769
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.399153709411621,
+      "learning_rate": 8.968165805068054e-06,
+      "loss": 0.9342,
+      "step": 1770
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8736134171485901,
+      "learning_rate": 8.966874354588008e-06,
+      "loss": 1.0763,
+      "step": 1771
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.380345344543457,
+      "learning_rate": 8.965582189529336e-06,
+      "loss": 0.8839,
+      "step": 1772
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.338686227798462,
+      "learning_rate": 8.964289310124803e-06,
+      "loss": 0.8969,
+      "step": 1773
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3952672481536865,
+      "learning_rate": 8.962995716607304e-06,
+      "loss": 0.8876,
+      "step": 1774
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.891251802444458,
+      "learning_rate": 8.96170140920986e-06,
+      "loss": 0.9155,
+      "step": 1775
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.356103539466858,
+      "learning_rate": 8.960406388165626e-06,
+      "loss": 0.9079,
+      "step": 1776
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.2437160015106201,
+      "learning_rate": 8.959110653707883e-06,
+      "loss": 0.8667,
+      "step": 1777
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4234168529510498,
+      "learning_rate": 8.957814206070037e-06,
+      "loss": 0.8834,
+      "step": 1778
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8960672616958618,
+      "learning_rate": 8.956517045485626e-06,
+      "loss": 1.0763,
+      "step": 1779
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3880730867385864,
+      "learning_rate": 8.955219172188319e-06,
+      "loss": 0.9479,
+      "step": 1780
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.3599027395248413,
+      "learning_rate": 8.953920586411906e-06,
+      "loss": 0.8519,
+      "step": 1781
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.746292233467102,
+      "learning_rate": 8.952621288390313e-06,
+      "loss": 1.0689,
+      "step": 1782
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7193189263343811,
+      "learning_rate": 8.951321278357589e-06,
+      "loss": 1.0664,
+      "step": 1783
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6427579522132874,
+      "learning_rate": 8.950020556547913e-06,
+      "loss": 1.052,
+      "step": 1784
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.6580779552459717,
+      "learning_rate": 8.948719123195596e-06,
+      "loss": 0.8343,
+      "step": 1785
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.375563144683838,
+      "learning_rate": 8.947416978535069e-06,
+      "loss": 0.839,
+      "step": 1786
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.4837244749069214,
+      "learning_rate": 8.946114122800896e-06,
+      "loss": 0.881,
+      "step": 1787
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.0056952238082886,
+      "learning_rate": 8.944810556227771e-06,
+      "loss": 1.0633,
+      "step": 1788
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.685058355331421,
+      "learning_rate": 8.943506279050515e-06,
+      "loss": 0.8602,
+      "step": 1789
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.498584270477295,
+      "learning_rate": 8.942201291504071e-06,
+      "loss": 0.9288,
+      "step": 1790
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7663305401802063,
+      "learning_rate": 8.940895593823519e-06,
+      "loss": 1.0666,
+      "step": 1791
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 1.358616828918457,
+      "learning_rate": 8.939589186244063e-06,
+      "loss": 0.8245,
+      "step": 1792
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.7684771418571472,
+      "learning_rate": 8.938282069001031e-06,
+      "loss": 1.0893,
+      "step": 1793
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7145503759384155,
+      "learning_rate": 8.936974242329887e-06,
+      "loss": 1.0646,
+      "step": 1794
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3257697820663452,
+      "learning_rate": 8.935665706466214e-06,
+      "loss": 0.8317,
+      "step": 1795
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7133659720420837,
+      "learning_rate": 8.93435646164573e-06,
+      "loss": 1.0564,
+      "step": 1796
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.370383381843567,
+      "learning_rate": 8.933046508104275e-06,
+      "loss": 0.9175,
+      "step": 1797
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.452759027481079,
+      "learning_rate": 8.93173584607782e-06,
+      "loss": 0.8845,
+      "step": 1798
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4403610229492188,
+      "learning_rate": 8.930424475802466e-06,
+      "loss": 0.8856,
+      "step": 1799
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2971696853637695,
+      "learning_rate": 8.929112397514436e-06,
+      "loss": 0.9154,
+      "step": 1800
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3849505186080933,
+      "learning_rate": 8.927799611450083e-06,
+      "loss": 0.8381,
+      "step": 1801
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8526045083999634,
+      "learning_rate": 8.926486117845886e-06,
+      "loss": 1.0813,
+      "step": 1802
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.353899598121643,
+      "learning_rate": 8.925171916938459e-06,
+      "loss": 0.8703,
+      "step": 1803
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3929617404937744,
+      "learning_rate": 8.923857008964534e-06,
+      "loss": 0.8066,
+      "step": 1804
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3753849267959595,
+      "learning_rate": 8.92254139416097e-06,
+      "loss": 0.866,
+      "step": 1805
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.25497567653656,
+      "learning_rate": 8.921225072764764e-06,
+      "loss": 0.8742,
+      "step": 1806
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.441863775253296,
+      "learning_rate": 8.919908045013028e-06,
+      "loss": 0.8215,
+      "step": 1807
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.074110269546509,
+      "learning_rate": 8.91859031114301e-06,
+      "loss": 0.8616,
+      "step": 1808
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2311004400253296,
+      "learning_rate": 8.917271871392079e-06,
+      "loss": 0.8784,
+      "step": 1809
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7557865977287292,
+      "learning_rate": 8.915952725997736e-06,
+      "loss": 1.0765,
+      "step": 1810
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3809040784835815,
+      "learning_rate": 8.914632875197608e-06,
+      "loss": 0.8638,
+      "step": 1811
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6441102623939514,
+      "learning_rate": 8.913312319229448e-06,
+      "loss": 1.0776,
+      "step": 1812
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6683152914047241,
+      "learning_rate": 8.911991058331133e-06,
+      "loss": 1.0799,
+      "step": 1813
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2906626462936401,
+      "learning_rate": 8.910669092740675e-06,
+      "loss": 0.8523,
+      "step": 1814
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.345818281173706,
+      "learning_rate": 8.909346422696204e-06,
+      "loss": 0.8673,
+      "step": 1815
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4519542455673218,
+      "learning_rate": 8.908023048435983e-06,
+      "loss": 0.9385,
+      "step": 1816
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3366880416870117,
+      "learning_rate": 8.906698970198402e-06,
+      "loss": 0.8647,
+      "step": 1817
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8976703882217407,
+      "learning_rate": 8.905374188221971e-06,
+      "loss": 1.0785,
+      "step": 1818
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7508203387260437,
+      "learning_rate": 8.904048702745336e-06,
+      "loss": 1.0691,
+      "step": 1819
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6612090468406677,
+      "learning_rate": 8.902722514007262e-06,
+      "loss": 1.0675,
+      "step": 1820
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.386809229850769,
+      "learning_rate": 8.901395622246645e-06,
+      "loss": 0.8934,
+      "step": 1821
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8148211240768433,
+      "learning_rate": 8.900068027702508e-06,
+      "loss": 1.0581,
+      "step": 1822
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2772516012191772,
+      "learning_rate": 8.898739730613996e-06,
+      "loss": 0.9176,
+      "step": 1823
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3755404949188232,
+      "learning_rate": 8.897410731220388e-06,
+      "loss": 0.8686,
+      "step": 1824
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3851432800292969,
+      "learning_rate": 8.896081029761081e-06,
+      "loss": 0.8526,
+      "step": 1825
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3238402605056763,
+      "learning_rate": 8.894750626475605e-06,
+      "loss": 0.8601,
+      "step": 1826
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3392366170883179,
+      "learning_rate": 8.893419521603615e-06,
+      "loss": 0.8443,
+      "step": 1827
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3164211511611938,
+      "learning_rate": 8.892087715384888e-06,
+      "loss": 0.8422,
+      "step": 1828
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 2.129117250442505,
+      "learning_rate": 8.890755208059334e-06,
+      "loss": 0.9005,
+      "step": 1829
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4278264045715332,
+      "learning_rate": 8.889421999866985e-06,
+      "loss": 0.8751,
+      "step": 1830
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3012850284576416,
+      "learning_rate": 8.888088091048001e-06,
+      "loss": 0.9295,
+      "step": 1831
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3334646224975586,
+      "learning_rate": 8.886753481842667e-06,
+      "loss": 0.876,
+      "step": 1832
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.0416276454925537,
+      "learning_rate": 8.885418172491396e-06,
+      "loss": 1.0916,
+      "step": 1833
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3888019323349,
+      "learning_rate": 8.884082163234722e-06,
+      "loss": 0.8464,
+      "step": 1834
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6335923671722412,
+      "learning_rate": 8.882745454313313e-06,
+      "loss": 0.8892,
+      "step": 1835
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6880072355270386,
+      "learning_rate": 8.88140804596796e-06,
+      "loss": 1.0631,
+      "step": 1836
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4918798208236694,
+      "learning_rate": 8.880069938439573e-06,
+      "loss": 0.9197,
+      "step": 1837
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5533642768859863,
+      "learning_rate": 8.878731131969199e-06,
+      "loss": 0.875,
+      "step": 1838
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.416382908821106,
+      "learning_rate": 8.877391626798004e-06,
+      "loss": 0.9172,
+      "step": 1839
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.7412391901016235,
+      "learning_rate": 8.876051423167281e-06,
+      "loss": 0.9216,
+      "step": 1840
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.8811568021774292,
+      "learning_rate": 8.874710521318452e-06,
+      "loss": 1.0876,
+      "step": 1841
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3886911869049072,
+      "learning_rate": 8.87336892149306e-06,
+      "loss": 0.8556,
+      "step": 1842
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3045520782470703,
+      "learning_rate": 8.872026623932779e-06,
+      "loss": 0.8306,
+      "step": 1843
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4403804540634155,
+      "learning_rate": 8.870683628879401e-06,
+      "loss": 0.8141,
+      "step": 1844
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.789621114730835,
+      "learning_rate": 8.869339936574852e-06,
+      "loss": 1.078,
+      "step": 1845
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3173651695251465,
+      "learning_rate": 8.867995547261179e-06,
+      "loss": 0.8946,
+      "step": 1846
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2830300331115723,
+      "learning_rate": 8.866650461180557e-06,
+      "loss": 0.9135,
+      "step": 1847
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6987025737762451,
+      "learning_rate": 8.865304678575281e-06,
+      "loss": 1.0488,
+      "step": 1848
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6966933012008667,
+      "learning_rate": 8.863958199687779e-06,
+      "loss": 0.9184,
+      "step": 1849
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4413715600967407,
+      "learning_rate": 8.862611024760598e-06,
+      "loss": 0.8885,
+      "step": 1850
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3802939653396606,
+      "learning_rate": 8.861263154036417e-06,
+      "loss": 0.9181,
+      "step": 1851
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4522321224212646,
+      "learning_rate": 8.859914587758035e-06,
+      "loss": 0.8684,
+      "step": 1852
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7728126049041748,
+      "learning_rate": 8.858565326168377e-06,
+      "loss": 1.0849,
+      "step": 1853
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6879520416259766,
+      "learning_rate": 8.857215369510496e-06,
+      "loss": 1.0721,
+      "step": 1854
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5033268928527832,
+      "learning_rate": 8.855864718027565e-06,
+      "loss": 0.97,
+      "step": 1855
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3005163669586182,
+      "learning_rate": 8.85451337196289e-06,
+      "loss": 0.9013,
+      "step": 1856
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.31600821018219,
+      "learning_rate": 8.853161331559894e-06,
+      "loss": 0.884,
+      "step": 1857
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3804577589035034,
+      "learning_rate": 8.851808597062131e-06,
+      "loss": 0.8969,
+      "step": 1858
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.2594002485275269,
+      "learning_rate": 8.850455168713277e-06,
+      "loss": 0.8665,
+      "step": 1859
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4975119829177856,
+      "learning_rate": 8.849101046757134e-06,
+      "loss": 0.8745,
+      "step": 1860
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4257913827896118,
+      "learning_rate": 8.847746231437629e-06,
+      "loss": 0.9256,
+      "step": 1861
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.5152971744537354,
+      "learning_rate": 8.84639072299881e-06,
+      "loss": 0.9333,
+      "step": 1862
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.275823950767517,
+      "learning_rate": 8.845034521684859e-06,
+      "loss": 1.0755,
+      "step": 1863
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4606250524520874,
+      "learning_rate": 8.843677627740075e-06,
+      "loss": 0.9081,
+      "step": 1864
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.4124798774719238,
+      "learning_rate": 8.842320041408882e-06,
+      "loss": 0.8483,
+      "step": 1865
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7253978252410889,
+      "learning_rate": 8.840961762935836e-06,
+      "loss": 1.0721,
+      "step": 1866
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.393515706062317,
+      "learning_rate": 8.839602792565606e-06,
+      "loss": 0.8276,
+      "step": 1867
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.6245272159576416,
+      "learning_rate": 8.838243130542995e-06,
+      "loss": 0.8681,
+      "step": 1868
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 1.3611037731170654,
+      "learning_rate": 8.83688277711293e-06,
+      "loss": 0.8264,
+      "step": 1869
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.9836920499801636,
+      "learning_rate": 8.835521732520459e-06,
+      "loss": 1.0955,
+      "step": 1870
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7313248515129089,
+      "learning_rate": 8.834159997010751e-06,
+      "loss": 1.0397,
+      "step": 1871
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4687851667404175,
+      "learning_rate": 8.83279757082911e-06,
+      "loss": 0.8946,
+      "step": 1872
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.411566138267517,
+      "learning_rate": 8.831434454220957e-06,
+      "loss": 0.8913,
+      "step": 1873
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4347537755966187,
+      "learning_rate": 8.830070647431839e-06,
+      "loss": 0.8794,
+      "step": 1874
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3193713426589966,
+      "learning_rate": 8.828706150707424e-06,
+      "loss": 0.8422,
+      "step": 1875
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5208511352539062,
+      "learning_rate": 8.827340964293514e-06,
+      "loss": 1.0708,
+      "step": 1876
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3423826694488525,
+      "learning_rate": 8.825975088436023e-06,
+      "loss": 0.89,
+      "step": 1877
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.500097393989563,
+      "learning_rate": 8.824608523380998e-06,
+      "loss": 0.9611,
+      "step": 1878
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3479200601577759,
+      "learning_rate": 8.823241269374606e-06,
+      "loss": 0.8274,
+      "step": 1879
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4322720766067505,
+      "learning_rate": 8.82187332666314e-06,
+      "loss": 0.8815,
+      "step": 1880
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.753732442855835,
+      "learning_rate": 8.820504695493018e-06,
+      "loss": 1.0685,
+      "step": 1881
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5466588735580444,
+      "learning_rate": 8.819135376110776e-06,
+      "loss": 0.879,
+      "step": 1882
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.353731393814087,
+      "learning_rate": 8.817765368763081e-06,
+      "loss": 0.9276,
+      "step": 1883
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6865143775939941,
+      "learning_rate": 8.816394673696722e-06,
+      "loss": 1.0832,
+      "step": 1884
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4016969203948975,
+      "learning_rate": 8.815023291158608e-06,
+      "loss": 0.8855,
+      "step": 1885
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3458102941513062,
+      "learning_rate": 8.813651221395779e-06,
+      "loss": 0.8437,
+      "step": 1886
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7619431018829346,
+      "learning_rate": 8.812278464655394e-06,
+      "loss": 1.0727,
+      "step": 1887
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4230338335037231,
+      "learning_rate": 8.810905021184732e-06,
+      "loss": 0.8486,
+      "step": 1888
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4361132383346558,
+      "learning_rate": 8.809530891231206e-06,
+      "loss": 0.9215,
+      "step": 1889
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.7007415294647217,
+      "learning_rate": 8.808156075042343e-06,
+      "loss": 0.9032,
+      "step": 1890
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4330562353134155,
+      "learning_rate": 8.806780572865801e-06,
+      "loss": 0.8505,
+      "step": 1891
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3375364542007446,
+      "learning_rate": 8.805404384949354e-06,
+      "loss": 0.8226,
+      "step": 1892
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7868821620941162,
+      "learning_rate": 8.804027511540906e-06,
+      "loss": 1.0531,
+      "step": 1893
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3550751209259033,
+      "learning_rate": 8.802649952888483e-06,
+      "loss": 0.868,
+      "step": 1894
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3924355506896973,
+      "learning_rate": 8.80127170924023e-06,
+      "loss": 0.9213,
+      "step": 1895
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7095077037811279,
+      "learning_rate": 8.799892780844424e-06,
+      "loss": 1.0678,
+      "step": 1896
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6830340623855591,
+      "learning_rate": 8.798513167949453e-06,
+      "loss": 1.0628,
+      "step": 1897
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3669883012771606,
+      "learning_rate": 8.797132870803844e-06,
+      "loss": 0.8846,
+      "step": 1898
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.0693278312683105,
+      "learning_rate": 8.795751889656232e-06,
+      "loss": 0.822,
+      "step": 1899
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4809867143630981,
+      "learning_rate": 8.794370224755386e-06,
+      "loss": 0.8699,
+      "step": 1900
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4493680000305176,
+      "learning_rate": 8.792987876350194e-06,
+      "loss": 0.8859,
+      "step": 1901
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5297424793243408,
+      "learning_rate": 8.791604844689664e-06,
+      "loss": 0.8358,
+      "step": 1902
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3584353923797607,
+      "learning_rate": 8.790221130022934e-06,
+      "loss": 0.8593,
+      "step": 1903
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6102445125579834,
+      "learning_rate": 8.78883673259926e-06,
+      "loss": 0.88,
+      "step": 1904
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 2.127807378768921,
+      "learning_rate": 8.787451652668023e-06,
+      "loss": 0.9035,
+      "step": 1905
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1817377805709839,
+      "learning_rate": 8.786065890478725e-06,
+      "loss": 1.0933,
+      "step": 1906
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.375930666923523,
+      "learning_rate": 8.784679446280994e-06,
+      "loss": 0.8518,
+      "step": 1907
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3648995161056519,
+      "learning_rate": 8.783292320324578e-06,
+      "loss": 0.898,
+      "step": 1908
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3694229125976562,
+      "learning_rate": 8.781904512859348e-06,
+      "loss": 0.8286,
+      "step": 1909
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8736174702644348,
+      "learning_rate": 8.780516024135301e-06,
+      "loss": 1.071,
+      "step": 1910
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.447430968284607,
+      "learning_rate": 8.779126854402554e-06,
+      "loss": 0.8025,
+      "step": 1911
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.350923776626587,
+      "learning_rate": 8.777737003911346e-06,
+      "loss": 0.8573,
+      "step": 1912
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.73961341381073,
+      "learning_rate": 8.776346472912039e-06,
+      "loss": 1.0636,
+      "step": 1913
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3616633415222168,
+      "learning_rate": 8.77495526165512e-06,
+      "loss": 0.8369,
+      "step": 1914
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6748552322387695,
+      "learning_rate": 8.773563370391198e-06,
+      "loss": 1.0913,
+      "step": 1915
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7434237599372864,
+      "learning_rate": 8.772170799371e-06,
+      "loss": 1.0889,
+      "step": 1916
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.413402795791626,
+      "learning_rate": 8.77077754884538e-06,
+      "loss": 0.8558,
+      "step": 1917
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.544804334640503,
+      "learning_rate": 8.769383619065315e-06,
+      "loss": 0.8701,
+      "step": 1918
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.939536690711975,
+      "learning_rate": 8.767989010281899e-06,
+      "loss": 0.8393,
+      "step": 1919
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3224928379058838,
+      "learning_rate": 8.766593722746355e-06,
+      "loss": 0.8336,
+      "step": 1920
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4183446168899536,
+      "learning_rate": 8.765197756710025e-06,
+      "loss": 0.8164,
+      "step": 1921
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.0609568357467651,
+      "learning_rate": 8.763801112424372e-06,
+      "loss": 1.0631,
+      "step": 1922
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4024368524551392,
+      "learning_rate": 8.762403790140982e-06,
+      "loss": 0.8356,
+      "step": 1923
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5415114164352417,
+      "learning_rate": 8.761005790111567e-06,
+      "loss": 0.8957,
+      "step": 1924
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.351745367050171,
+      "learning_rate": 8.759607112587956e-06,
+      "loss": 0.8745,
+      "step": 1925
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4216660261154175,
+      "learning_rate": 8.7582077578221e-06,
+      "loss": 0.7682,
+      "step": 1926
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7781400680541992,
+      "learning_rate": 8.756807726066073e-06,
+      "loss": 1.0538,
+      "step": 1927
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7759211659431458,
+      "learning_rate": 8.755407017572075e-06,
+      "loss": 1.0648,
+      "step": 1928
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.63298100233078,
+      "learning_rate": 8.754005632592425e-06,
+      "loss": 1.0636,
+      "step": 1929
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.5753194093704224,
+      "learning_rate": 8.752603571379559e-06,
+      "loss": 0.8395,
+      "step": 1930
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3418915271759033,
+      "learning_rate": 8.751200834186044e-06,
+      "loss": 0.8334,
+      "step": 1931
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.458600640296936,
+      "learning_rate": 8.749797421264563e-06,
+      "loss": 0.8928,
+      "step": 1932
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3107020854949951,
+      "learning_rate": 8.748393332867919e-06,
+      "loss": 0.8949,
+      "step": 1933
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.2591869831085205,
+      "learning_rate": 8.746988569249044e-06,
+      "loss": 1.0464,
+      "step": 1934
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4105409383773804,
+      "learning_rate": 8.745583130660983e-06,
+      "loss": 0.8577,
+      "step": 1935
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.3790479898452759,
+      "learning_rate": 8.74417701735691e-06,
+      "loss": 0.8923,
+      "step": 1936
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4177829027175903,
+      "learning_rate": 8.742770229590113e-06,
+      "loss": 0.8691,
+      "step": 1937
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.388040542602539,
+      "learning_rate": 8.741362767614012e-06,
+      "loss": 0.8745,
+      "step": 1938
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7726190090179443,
+      "learning_rate": 8.739954631682136e-06,
+      "loss": 1.075,
+      "step": 1939
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4041719436645508,
+      "learning_rate": 8.738545822048144e-06,
+      "loss": 0.8764,
+      "step": 1940
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6956908106803894,
+      "learning_rate": 8.737136338965816e-06,
+      "loss": 1.0785,
+      "step": 1941
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.4248274564743042,
+      "learning_rate": 8.735726182689048e-06,
+      "loss": 0.8131,
+      "step": 1942
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6773179173469543,
+      "learning_rate": 8.734315353471863e-06,
+      "loss": 1.0514,
+      "step": 1943
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.347542405128479,
+      "learning_rate": 8.732903851568401e-06,
+      "loss": 0.8819,
+      "step": 1944
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.441172480583191,
+      "learning_rate": 8.731491677232926e-06,
+      "loss": 0.8344,
+      "step": 1945
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7205271124839783,
+      "learning_rate": 8.73007883071982e-06,
+      "loss": 1.0587,
+      "step": 1946
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4342392683029175,
+      "learning_rate": 8.728665312283593e-06,
+      "loss": 0.9187,
+      "step": 1947
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6366753578186035,
+      "learning_rate": 8.727251122178866e-06,
+      "loss": 1.0656,
+      "step": 1948
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4351943731307983,
+      "learning_rate": 8.725836260660387e-06,
+      "loss": 0.9517,
+      "step": 1949
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3143811225891113,
+      "learning_rate": 8.724420727983028e-06,
+      "loss": 0.9354,
+      "step": 1950
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4659693241119385,
+      "learning_rate": 8.723004524401774e-06,
+      "loss": 0.9109,
+      "step": 1951
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3133201599121094,
+      "learning_rate": 8.721587650171737e-06,
+      "loss": 0.8496,
+      "step": 1952
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.482869029045105,
+      "learning_rate": 8.720170105548145e-06,
+      "loss": 0.89,
+      "step": 1953
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.852698802947998,
+      "learning_rate": 8.718751890786354e-06,
+      "loss": 1.0466,
+      "step": 1954
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6997069120407104,
+      "learning_rate": 8.717333006141831e-06,
+      "loss": 1.0439,
+      "step": 1955
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6433049440383911,
+      "learning_rate": 8.715913451870173e-06,
+      "loss": 1.0699,
+      "step": 1956
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3456803560256958,
+      "learning_rate": 8.71449322822709e-06,
+      "loss": 0.9173,
+      "step": 1957
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7878799438476562,
+      "learning_rate": 8.71307233546842e-06,
+      "loss": 1.0585,
+      "step": 1958
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7846865653991699,
+      "learning_rate": 8.711650773850114e-06,
+      "loss": 1.0678,
+      "step": 1959
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3340226411819458,
+      "learning_rate": 8.710228543628248e-06,
+      "loss": 0.7944,
+      "step": 1960
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4635143280029297,
+      "learning_rate": 8.708805645059019e-06,
+      "loss": 0.9225,
+      "step": 1961
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7575302124023438,
+      "learning_rate": 8.70738207839874e-06,
+      "loss": 1.0598,
+      "step": 1962
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.9257854223251343,
+      "learning_rate": 8.705957843903852e-06,
+      "loss": 0.9135,
+      "step": 1963
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7308155298233032,
+      "learning_rate": 8.704532941830905e-06,
+      "loss": 1.0584,
+      "step": 1964
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.316074252128601,
+      "learning_rate": 8.703107372436583e-06,
+      "loss": 0.8524,
+      "step": 1965
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4237297773361206,
+      "learning_rate": 8.701681135977678e-06,
+      "loss": 0.8644,
+      "step": 1966
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7871028780937195,
+      "learning_rate": 8.700254232711107e-06,
+      "loss": 1.0806,
+      "step": 1967
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.244920015335083,
+      "learning_rate": 8.69882666289391e-06,
+      "loss": 0.8726,
+      "step": 1968
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6663981676101685,
+      "learning_rate": 8.697398426783243e-06,
+      "loss": 1.0377,
+      "step": 1969
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.617922842502594,
+      "learning_rate": 8.695969524636383e-06,
+      "loss": 1.0838,
+      "step": 1970
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2909467220306396,
+      "learning_rate": 8.694539956710728e-06,
+      "loss": 0.8086,
+      "step": 1971
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3245768547058105,
+      "learning_rate": 8.693109723263794e-06,
+      "loss": 0.8432,
+      "step": 1972
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3860341310501099,
+      "learning_rate": 8.691678824553223e-06,
+      "loss": 0.8992,
+      "step": 1973
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7223508954048157,
+      "learning_rate": 8.690247260836765e-06,
+      "loss": 1.0579,
+      "step": 1974
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.391032338142395,
+      "learning_rate": 8.688815032372302e-06,
+      "loss": 0.9129,
+      "step": 1975
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7094430923461914,
+      "learning_rate": 8.687382139417827e-06,
+      "loss": 1.0994,
+      "step": 1976
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.479591965675354,
+      "learning_rate": 8.685948582231458e-06,
+      "loss": 0.8516,
+      "step": 1977
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.938592791557312,
+      "learning_rate": 8.68451436107143e-06,
+      "loss": 0.9373,
+      "step": 1978
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7373561263084412,
+      "learning_rate": 8.683079476196101e-06,
+      "loss": 1.0707,
+      "step": 1979
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.406818151473999,
+      "learning_rate": 8.681643927863944e-06,
+      "loss": 0.8552,
+      "step": 1980
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.729547381401062,
+      "learning_rate": 8.680207716333552e-06,
+      "loss": 0.8547,
+      "step": 1981
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7417411804199219,
+      "learning_rate": 8.678770841863644e-06,
+      "loss": 1.0777,
+      "step": 1982
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3678401708602905,
+      "learning_rate": 8.677333304713048e-06,
+      "loss": 0.8607,
+      "step": 1983
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6682898998260498,
+      "learning_rate": 8.67589510514072e-06,
+      "loss": 1.053,
+      "step": 1984
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.382290005683899,
+      "learning_rate": 8.674456243405732e-06,
+      "loss": 0.8393,
+      "step": 1985
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6607999205589294,
+      "learning_rate": 8.673016719767274e-06,
+      "loss": 1.0756,
+      "step": 1986
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.212644100189209,
+      "learning_rate": 8.671576534484655e-06,
+      "loss": 0.8331,
+      "step": 1987
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.323799967765808,
+      "learning_rate": 8.670135687817312e-06,
+      "loss": 0.8194,
+      "step": 1988
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4182924032211304,
+      "learning_rate": 8.668694180024785e-06,
+      "loss": 0.8842,
+      "step": 1989
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7785651683807373,
+      "learning_rate": 8.66725201136675e-06,
+      "loss": 1.0502,
+      "step": 1990
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3411190509796143,
+      "learning_rate": 8.665809182102987e-06,
+      "loss": 0.9095,
+      "step": 1991
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4019577503204346,
+      "learning_rate": 8.664365692493408e-06,
+      "loss": 0.8763,
+      "step": 1992
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.7649356126785278,
+      "learning_rate": 8.662921542798032e-06,
+      "loss": 0.8534,
+      "step": 1993
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4061756134033203,
+      "learning_rate": 8.661476733277011e-06,
+      "loss": 0.8292,
+      "step": 1994
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4497069120407104,
+      "learning_rate": 8.660031264190603e-06,
+      "loss": 0.8793,
+      "step": 1995
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3088041543960571,
+      "learning_rate": 8.658585135799188e-06,
+      "loss": 0.8915,
+      "step": 1996
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4936143159866333,
+      "learning_rate": 8.65713834836327e-06,
+      "loss": 0.8756,
+      "step": 1997
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.768703818321228,
+      "learning_rate": 8.655690902143466e-06,
+      "loss": 0.8823,
+      "step": 1998
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.38458251953125,
+      "learning_rate": 8.654242797400516e-06,
+      "loss": 0.8953,
+      "step": 1999
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3408067226409912,
+      "learning_rate": 8.652794034395274e-06,
+      "loss": 0.8857,
+      "step": 2000
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4233827590942383,
+      "learning_rate": 8.651344613388714e-06,
+      "loss": 0.8557,
+      "step": 2001
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.33262038230896,
+      "learning_rate": 8.649894534641933e-06,
+      "loss": 0.8143,
+      "step": 2002
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8509079813957214,
+      "learning_rate": 8.648443798416142e-06,
+      "loss": 1.0611,
+      "step": 2003
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7345284819602966,
+      "learning_rate": 8.646992404972671e-06,
+      "loss": 1.0837,
+      "step": 2004
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3927648067474365,
+      "learning_rate": 8.645540354572969e-06,
+      "loss": 0.883,
+      "step": 2005
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4448801279067993,
+      "learning_rate": 8.644087647478602e-06,
+      "loss": 0.9232,
+      "step": 2006
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9943041205406189,
+      "learning_rate": 8.642634283951257e-06,
+      "loss": 1.0593,
+      "step": 2007
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4191185235977173,
+      "learning_rate": 8.641180264252738e-06,
+      "loss": 0.8943,
+      "step": 2008
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4717671871185303,
+      "learning_rate": 8.639725588644965e-06,
+      "loss": 0.8963,
+      "step": 2009
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3369464874267578,
+      "learning_rate": 8.63827025738998e-06,
+      "loss": 0.8458,
+      "step": 2010
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.6330853700637817,
+      "learning_rate": 8.63681427074994e-06,
+      "loss": 0.9317,
+      "step": 2011
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.435011625289917,
+      "learning_rate": 8.635357628987122e-06,
+      "loss": 0.8692,
+      "step": 2012
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5842838287353516,
+      "learning_rate": 8.63390033236392e-06,
+      "loss": 0.8622,
+      "step": 2013
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.228672742843628,
+      "learning_rate": 8.632442381142846e-06,
+      "loss": 0.7893,
+      "step": 2014
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.2601556777954102,
+      "learning_rate": 8.630983775586532e-06,
+      "loss": 0.8479,
+      "step": 2015
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.9485021829605103,
+      "learning_rate": 8.629524515957724e-06,
+      "loss": 1.0279,
+      "step": 2016
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3314533233642578,
+      "learning_rate": 8.628064602519287e-06,
+      "loss": 0.8514,
+      "step": 2017
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.5839899778366089,
+      "learning_rate": 8.626604035534205e-06,
+      "loss": 0.8759,
+      "step": 2018
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.3172168731689453,
+      "learning_rate": 8.625142815265581e-06,
+      "loss": 0.8131,
+      "step": 2019
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.419778823852539,
+      "learning_rate": 8.623680941976634e-06,
+      "loss": 0.892,
+      "step": 2020
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.4386640787124634,
+      "learning_rate": 8.622218415930698e-06,
+      "loss": 0.8941,
+      "step": 2021
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 1.587243676185608,
+      "learning_rate": 8.62075523739123e-06,
+      "loss": 0.9217,
+      "step": 2022
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.059421420097351,
+      "learning_rate": 8.619291406621801e-06,
+      "loss": 1.0582,
+      "step": 2023
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4295897483825684,
+      "learning_rate": 8.617826923886102e-06,
+      "loss": 0.8841,
+      "step": 2024
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5077108144760132,
+      "learning_rate": 8.616361789447935e-06,
+      "loss": 0.9392,
+      "step": 2025
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4502383470535278,
+      "learning_rate": 8.614896003571227e-06,
+      "loss": 0.8255,
+      "step": 2026
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3102778196334839,
+      "learning_rate": 8.61342956652002e-06,
+      "loss": 0.8842,
+      "step": 2027
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3786203861236572,
+      "learning_rate": 8.611962478558473e-06,
+      "loss": 0.9297,
+      "step": 2028
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.1232157945632935,
+      "learning_rate": 8.61049473995086e-06,
+      "loss": 1.0758,
+      "step": 2029
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.8021550178527832,
+      "learning_rate": 8.609026350961576e-06,
+      "loss": 1.0632,
+      "step": 2030
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.441365122795105,
+      "learning_rate": 8.60755731185513e-06,
+      "loss": 0.8453,
+      "step": 2031
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6146366596221924,
+      "learning_rate": 8.60608762289615e-06,
+      "loss": 0.8758,
+      "step": 2032
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7268567085266113,
+      "learning_rate": 8.604617284349381e-06,
+      "loss": 0.8926,
+      "step": 2033
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4320820569992065,
+      "learning_rate": 8.603146296479683e-06,
+      "loss": 0.8799,
+      "step": 2034
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7665687799453735,
+      "learning_rate": 8.601674659552038e-06,
+      "loss": 0.8765,
+      "step": 2035
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4184983968734741,
+      "learning_rate": 8.600202373831538e-06,
+      "loss": 0.8562,
+      "step": 2036
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4768610000610352,
+      "learning_rate": 8.598729439583398e-06,
+      "loss": 0.8937,
+      "step": 2037
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5518685579299927,
+      "learning_rate": 8.597255857072944e-06,
+      "loss": 0.8617,
+      "step": 2038
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4641942977905273,
+      "learning_rate": 8.595781626565624e-06,
+      "loss": 0.8522,
+      "step": 2039
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5545706748962402,
+      "learning_rate": 8.594306748327e-06,
+      "loss": 0.8606,
+      "step": 2040
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3208602666854858,
+      "learning_rate": 8.592831222622752e-06,
+      "loss": 0.8962,
+      "step": 2041
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 2.3387627601623535,
+      "learning_rate": 8.591355049718675e-06,
+      "loss": 1.0672,
+      "step": 2042
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.8943548202514648,
+      "learning_rate": 8.589878229880682e-06,
+      "loss": 1.0692,
+      "step": 2043
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4522082805633545,
+      "learning_rate": 8.588400763374802e-06,
+      "loss": 0.8538,
+      "step": 2044
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7448051571846008,
+      "learning_rate": 8.586922650467182e-06,
+      "loss": 1.0319,
+      "step": 2045
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.926529049873352,
+      "learning_rate": 8.58544389142408e-06,
+      "loss": 1.068,
+      "step": 2046
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.719583511352539,
+      "learning_rate": 8.58396448651188e-06,
+      "loss": 0.8848,
+      "step": 2047
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5139647722244263,
+      "learning_rate": 8.582484435997071e-06,
+      "loss": 0.8808,
+      "step": 2048
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6658366918563843,
+      "learning_rate": 8.581003740146269e-06,
+      "loss": 0.877,
+      "step": 2049
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3294460773468018,
+      "learning_rate": 8.579522399226196e-06,
+      "loss": 0.9062,
+      "step": 2050
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.417796015739441,
+      "learning_rate": 8.578040413503702e-06,
+      "loss": 0.8802,
+      "step": 2051
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7349704504013062,
+      "learning_rate": 8.576557783245743e-06,
+      "loss": 0.8769,
+      "step": 2052
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3909212350845337,
+      "learning_rate": 8.575074508719392e-06,
+      "loss": 0.8835,
+      "step": 2053
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4662184715270996,
+      "learning_rate": 8.573590590191844e-06,
+      "loss": 0.8719,
+      "step": 2054
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4855685234069824,
+      "learning_rate": 8.572106027930409e-06,
+      "loss": 0.8826,
+      "step": 2055
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7683361768722534,
+      "learning_rate": 8.570620822202507e-06,
+      "loss": 1.0827,
+      "step": 2056
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4792038202285767,
+      "learning_rate": 8.569134973275678e-06,
+      "loss": 0.8918,
+      "step": 2057
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.625923752784729,
+      "learning_rate": 8.567648481417581e-06,
+      "loss": 0.9008,
+      "step": 2058
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9278605580329895,
+      "learning_rate": 8.566161346895983e-06,
+      "loss": 1.0548,
+      "step": 2059
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5978384017944336,
+      "learning_rate": 8.564673569978774e-06,
+      "loss": 0.9117,
+      "step": 2060
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.39349365234375,
+      "learning_rate": 8.563185150933955e-06,
+      "loss": 0.8297,
+      "step": 2061
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.087948203086853,
+      "learning_rate": 8.561696090029645e-06,
+      "loss": 1.0599,
+      "step": 2062
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0584356784820557,
+      "learning_rate": 8.560206387534078e-06,
+      "loss": 1.0402,
+      "step": 2063
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.80888432264328,
+      "learning_rate": 8.558716043715606e-06,
+      "loss": 1.0628,
+      "step": 2064
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.728522539138794,
+      "learning_rate": 8.557225058842691e-06,
+      "loss": 0.8881,
+      "step": 2065
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5370346307754517,
+      "learning_rate": 8.555733433183915e-06,
+      "loss": 0.8516,
+      "step": 2066
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6702052354812622,
+      "learning_rate": 8.554241167007976e-06,
+      "loss": 0.9052,
+      "step": 2067
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.2845031023025513,
+      "learning_rate": 8.55274826058368e-06,
+      "loss": 1.0518,
+      "step": 2068
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3327728509902954,
+      "learning_rate": 8.551254714179961e-06,
+      "loss": 0.8596,
+      "step": 2069
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9371837973594666,
+      "learning_rate": 8.549760528065855e-06,
+      "loss": 1.0371,
+      "step": 2070
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6674561500549316,
+      "learning_rate": 8.548265702510524e-06,
+      "loss": 1.0605,
+      "step": 2071
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7766066193580627,
+      "learning_rate": 8.54677023778324e-06,
+      "loss": 1.0805,
+      "step": 2072
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.692778468132019,
+      "learning_rate": 8.545274134153386e-06,
+      "loss": 0.8727,
+      "step": 2073
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.089470624923706,
+      "learning_rate": 8.54377739189047e-06,
+      "loss": 1.065,
+      "step": 2074
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.881123960018158,
+      "learning_rate": 8.542280011264108e-06,
+      "loss": 1.0589,
+      "step": 2075
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7202778458595276,
+      "learning_rate": 8.540781992544032e-06,
+      "loss": 1.0712,
+      "step": 2076
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4170852899551392,
+      "learning_rate": 8.539283336000091e-06,
+      "loss": 0.8745,
+      "step": 2077
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.9860764741897583,
+      "learning_rate": 8.537784041902247e-06,
+      "loss": 1.0515,
+      "step": 2078
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3346116542816162,
+      "learning_rate": 8.536284110520577e-06,
+      "loss": 0.9203,
+      "step": 2079
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4569605588912964,
+      "learning_rate": 8.534783542125276e-06,
+      "loss": 0.8493,
+      "step": 2080
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7789324522018433,
+      "learning_rate": 8.533282336986646e-06,
+      "loss": 0.8607,
+      "step": 2081
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3355717658996582,
+      "learning_rate": 8.531780495375112e-06,
+      "loss": 0.8254,
+      "step": 2082
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3036370277404785,
+      "learning_rate": 8.530278017561214e-06,
+      "loss": 0.8228,
+      "step": 2083
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6045470237731934,
+      "learning_rate": 8.528774903815596e-06,
+      "loss": 0.8177,
+      "step": 2084
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3292607069015503,
+      "learning_rate": 8.527271154409027e-06,
+      "loss": 0.885,
+      "step": 2085
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.0896612405776978,
+      "learning_rate": 8.525766769612387e-06,
+      "loss": 1.0596,
+      "step": 2086
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4412721395492554,
+      "learning_rate": 8.52426174969667e-06,
+      "loss": 0.9013,
+      "step": 2087
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.366882562637329,
+      "learning_rate": 8.522756094932987e-06,
+      "loss": 0.8289,
+      "step": 2088
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7415438294410706,
+      "learning_rate": 8.521249805592558e-06,
+      "loss": 1.0801,
+      "step": 2089
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.5601248741149902,
+      "learning_rate": 8.519742881946723e-06,
+      "loss": 0.8567,
+      "step": 2090
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.7161226272583008,
+      "learning_rate": 8.518235324266931e-06,
+      "loss": 0.7966,
+      "step": 2091
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3612483739852905,
+      "learning_rate": 8.51672713282475e-06,
+      "loss": 0.827,
+      "step": 2092
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3061974048614502,
+      "learning_rate": 8.515218307891864e-06,
+      "loss": 0.8547,
+      "step": 2093
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4842816591262817,
+      "learning_rate": 8.513708849740061e-06,
+      "loss": 0.8884,
+      "step": 2094
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.4746569395065308,
+      "learning_rate": 8.512198758641252e-06,
+      "loss": 0.8114,
+      "step": 2095
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.3216726779937744,
+      "learning_rate": 8.510688034867461e-06,
+      "loss": 0.8408,
+      "step": 2096
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.019812822341919,
+      "learning_rate": 8.509176678690823e-06,
+      "loss": 1.0551,
+      "step": 2097
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.6495743989944458,
+      "learning_rate": 8.507664690383585e-06,
+      "loss": 0.8632,
+      "step": 2098
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6988558769226074,
+      "learning_rate": 8.506152070218119e-06,
+      "loss": 1.0695,
+      "step": 2099
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.582718849182129,
+      "learning_rate": 8.504638818466895e-06,
+      "loss": 0.8301,
+      "step": 2100
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.384694218635559,
+      "learning_rate": 8.50312493540251e-06,
+      "loss": 0.8521,
+      "step": 2101
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4113938808441162,
+      "learning_rate": 8.501610421297666e-06,
+      "loss": 0.8206,
+      "step": 2102
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3278846740722656,
+      "learning_rate": 8.500095276425187e-06,
+      "loss": 0.8432,
+      "step": 2103
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3527781963348389,
+      "learning_rate": 8.498579501058e-06,
+      "loss": 0.832,
+      "step": 2104
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3032948970794678,
+      "learning_rate": 8.497063095469155e-06,
+      "loss": 0.8854,
+      "step": 2105
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3846054077148438,
+      "learning_rate": 8.49554605993181e-06,
+      "loss": 0.8684,
+      "step": 2106
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.409877896308899,
+      "learning_rate": 8.494028394719241e-06,
+      "loss": 0.9172,
+      "step": 2107
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3258568048477173,
+      "learning_rate": 8.492510100104833e-06,
+      "loss": 0.835,
+      "step": 2108
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3367030620574951,
+      "learning_rate": 8.490991176362086e-06,
+      "loss": 1.0568,
+      "step": 2109
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3547948598861694,
+      "learning_rate": 8.489471623764613e-06,
+      "loss": 0.8608,
+      "step": 2110
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3892825841903687,
+      "learning_rate": 8.487951442586141e-06,
+      "loss": 0.8784,
+      "step": 2111
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.444891333580017,
+      "learning_rate": 8.486430633100513e-06,
+      "loss": 0.8652,
+      "step": 2112
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.783378541469574,
+      "learning_rate": 8.484909195581682e-06,
+      "loss": 1.0429,
+      "step": 2113
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4134604930877686,
+      "learning_rate": 8.483387130303708e-06,
+      "loss": 0.8121,
+      "step": 2114
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3825933933258057,
+      "learning_rate": 8.481864437540777e-06,
+      "loss": 0.9589,
+      "step": 2115
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8058869242668152,
+      "learning_rate": 8.480341117567177e-06,
+      "loss": 1.07,
+      "step": 2116
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3997387886047363,
+      "learning_rate": 8.478817170657319e-06,
+      "loss": 0.904,
+      "step": 2117
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4259949922561646,
+      "learning_rate": 8.477292597085714e-06,
+      "loss": 0.8562,
+      "step": 2118
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.385130524635315,
+      "learning_rate": 8.475767397126999e-06,
+      "loss": 0.8902,
+      "step": 2119
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3821345567703247,
+      "learning_rate": 8.474241571055916e-06,
+      "loss": 0.9182,
+      "step": 2120
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.358797311782837,
+      "learning_rate": 8.472715119147324e-06,
+      "loss": 0.8861,
+      "step": 2121
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3467940092086792,
+      "learning_rate": 8.47118804167619e-06,
+      "loss": 0.917,
+      "step": 2122
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9338010549545288,
+      "learning_rate": 8.469660338917596e-06,
+      "loss": 1.0688,
+      "step": 2123
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7762277126312256,
+      "learning_rate": 8.468132011146739e-06,
+      "loss": 0.9162,
+      "step": 2124
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3138870000839233,
+      "learning_rate": 8.466603058638925e-06,
+      "loss": 0.9208,
+      "step": 2125
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3601100444793701,
+      "learning_rate": 8.465073481669574e-06,
+      "loss": 0.8862,
+      "step": 2126
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4122940301895142,
+      "learning_rate": 8.46354328051422e-06,
+      "loss": 0.8709,
+      "step": 2127
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2548487186431885,
+      "learning_rate": 8.462012455448507e-06,
+      "loss": 0.9003,
+      "step": 2128
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.29526686668396,
+      "learning_rate": 8.460481006748194e-06,
+      "loss": 0.8882,
+      "step": 2129
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.355535864830017,
+      "learning_rate": 8.458948934689145e-06,
+      "loss": 0.8511,
+      "step": 2130
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.47536039352417,
+      "learning_rate": 8.457416239547352e-06,
+      "loss": 0.8756,
+      "step": 2131
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3462799787521362,
+      "learning_rate": 8.4558829215989e-06,
+      "loss": 0.9024,
+      "step": 2132
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.9130178689956665,
+      "learning_rate": 8.45434898112e-06,
+      "loss": 1.0758,
+      "step": 2133
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4181673526763916,
+      "learning_rate": 8.45281441838697e-06,
+      "loss": 0.9019,
+      "step": 2134
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5330225229263306,
+      "learning_rate": 8.45127923367624e-06,
+      "loss": 0.8625,
+      "step": 2135
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3262348175048828,
+      "learning_rate": 8.449743427264351e-06,
+      "loss": 0.8704,
+      "step": 2136
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.572252869606018,
+      "learning_rate": 8.448206999427962e-06,
+      "loss": 0.825,
+      "step": 2137
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4548803567886353,
+      "learning_rate": 8.446669950443836e-06,
+      "loss": 0.8291,
+      "step": 2138
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2795658111572266,
+      "learning_rate": 8.445132280588853e-06,
+      "loss": 0.8246,
+      "step": 2139
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.1095552444458008,
+      "learning_rate": 8.443593990140001e-06,
+      "loss": 1.0581,
+      "step": 2140
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3607463836669922,
+      "learning_rate": 8.442055079374389e-06,
+      "loss": 0.8661,
+      "step": 2141
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3116782903671265,
+      "learning_rate": 8.440515548569224e-06,
+      "loss": 0.7983,
+      "step": 2142
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4075897932052612,
+      "learning_rate": 8.438975398001834e-06,
+      "loss": 0.8635,
+      "step": 2143
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.35330331325531,
+      "learning_rate": 8.437434627949655e-06,
+      "loss": 0.8573,
+      "step": 2144
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8683505654335022,
+      "learning_rate": 8.435893238690238e-06,
+      "loss": 1.0758,
+      "step": 2145
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.5015854835510254,
+      "learning_rate": 8.434351230501242e-06,
+      "loss": 0.8898,
+      "step": 2146
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6982920169830322,
+      "learning_rate": 8.432808603660439e-06,
+      "loss": 1.0732,
+      "step": 2147
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.469725489616394,
+      "learning_rate": 8.431265358445713e-06,
+      "loss": 0.8324,
+      "step": 2148
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4404841661453247,
+      "learning_rate": 8.429721495135057e-06,
+      "loss": 0.8823,
+      "step": 2149
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2917494773864746,
+      "learning_rate": 8.42817701400658e-06,
+      "loss": 0.8689,
+      "step": 2150
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7910402417182922,
+      "learning_rate": 8.426631915338497e-06,
+      "loss": 1.0843,
+      "step": 2151
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7247693538665771,
+      "learning_rate": 8.425086199409135e-06,
+      "loss": 1.0761,
+      "step": 2152
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.484450340270996,
+      "learning_rate": 8.423539866496938e-06,
+      "loss": 0.848,
+      "step": 2153
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.349365472793579,
+      "learning_rate": 8.421992916880454e-06,
+      "loss": 0.8779,
+      "step": 2154
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3048125505447388,
+      "learning_rate": 8.420445350838346e-06,
+      "loss": 0.8351,
+      "step": 2155
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3846960067749023,
+      "learning_rate": 8.418897168649388e-06,
+      "loss": 0.8623,
+      "step": 2156
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3761154413223267,
+      "learning_rate": 8.417348370592462e-06,
+      "loss": 0.8312,
+      "step": 2157
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4627491235733032,
+      "learning_rate": 8.415798956946564e-06,
+      "loss": 0.8534,
+      "step": 2158
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3888540267944336,
+      "learning_rate": 8.414248927990797e-06,
+      "loss": 0.8588,
+      "step": 2159
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2354793548583984,
+      "learning_rate": 8.412698284004385e-06,
+      "loss": 1.0617,
+      "step": 2160
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3521515130996704,
+      "learning_rate": 8.411147025266649e-06,
+      "loss": 0.8735,
+      "step": 2161
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.33030366897583,
+      "learning_rate": 8.409595152057028e-06,
+      "loss": 0.8631,
+      "step": 2162
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.799965500831604,
+      "learning_rate": 8.408042664655071e-06,
+      "loss": 1.0602,
+      "step": 2163
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.385288953781128,
+      "learning_rate": 8.406489563340442e-06,
+      "loss": 0.8471,
+      "step": 2164
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.515330195426941,
+      "learning_rate": 8.404935848392905e-06,
+      "loss": 0.8523,
+      "step": 2165
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8810166716575623,
+      "learning_rate": 8.403381520092346e-06,
+      "loss": 1.0564,
+      "step": 2166
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7872344255447388,
+      "learning_rate": 8.401826578718752e-06,
+      "loss": 1.0647,
+      "step": 2167
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3966506719589233,
+      "learning_rate": 8.400271024552225e-06,
+      "loss": 0.8912,
+      "step": 2168
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.405947208404541,
+      "learning_rate": 8.39871485787298e-06,
+      "loss": 0.8844,
+      "step": 2169
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.4039008617401123,
+      "learning_rate": 8.397158078961336e-06,
+      "loss": 0.876,
+      "step": 2170
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3969320058822632,
+      "learning_rate": 8.39560068809773e-06,
+      "loss": 0.8649,
+      "step": 2171
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.6204932928085327,
+      "learning_rate": 8.394042685562699e-06,
+      "loss": 1.0662,
+      "step": 2172
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.2936007976531982,
+      "learning_rate": 8.3924840716369e-06,
+      "loss": 0.8827,
+      "step": 2173
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.7266877889633179,
+      "learning_rate": 8.390924846601095e-06,
+      "loss": 0.8486,
+      "step": 2174
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 1.3402459621429443,
+      "learning_rate": 8.389365010736157e-06,
+      "loss": 0.8641,
+      "step": 2175
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.770187258720398,
+      "learning_rate": 8.38780456432307e-06,
+      "loss": 1.0525,
+      "step": 2176
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7449145317077637,
+      "learning_rate": 8.386243507642926e-06,
+      "loss": 1.0381,
+      "step": 2177
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3693112134933472,
+      "learning_rate": 8.384681840976932e-06,
+      "loss": 0.8612,
+      "step": 2178
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.366475224494934,
+      "learning_rate": 8.383119564606394e-06,
+      "loss": 0.84,
+      "step": 2179
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9532701373100281,
+      "learning_rate": 8.381556678812742e-06,
+      "loss": 1.0634,
+      "step": 2180
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2555240392684937,
+      "learning_rate": 8.379993183877506e-06,
+      "loss": 0.8131,
+      "step": 2181
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.41875422000885,
+      "learning_rate": 8.378429080082327e-06,
+      "loss": 0.9027,
+      "step": 2182
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4229464530944824,
+      "learning_rate": 8.376864367708958e-06,
+      "loss": 0.8794,
+      "step": 2183
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4112945795059204,
+      "learning_rate": 8.375299047039263e-06,
+      "loss": 0.8251,
+      "step": 2184
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3529011011123657,
+      "learning_rate": 8.37373311835521e-06,
+      "loss": 0.8629,
+      "step": 2185
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4291563034057617,
+      "learning_rate": 8.372166581938882e-06,
+      "loss": 0.9527,
+      "step": 2186
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9990508556365967,
+      "learning_rate": 8.370599438072465e-06,
+      "loss": 1.0476,
+      "step": 2187
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3876240253448486,
+      "learning_rate": 8.369031687038267e-06,
+      "loss": 0.8523,
+      "step": 2188
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5071697235107422,
+      "learning_rate": 8.367463329118688e-06,
+      "loss": 0.9249,
+      "step": 2189
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.8721572160720825,
+      "learning_rate": 8.365894364596253e-06,
+      "loss": 0.8915,
+      "step": 2190
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7996845245361328,
+      "learning_rate": 8.364324793753587e-06,
+      "loss": 1.0821,
+      "step": 2191
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4171340465545654,
+      "learning_rate": 8.362754616873428e-06,
+      "loss": 0.8609,
+      "step": 2192
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4332002401351929,
+      "learning_rate": 8.36118383423862e-06,
+      "loss": 0.8833,
+      "step": 2193
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.287490725517273,
+      "learning_rate": 8.35961244613212e-06,
+      "loss": 0.888,
+      "step": 2194
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3308405876159668,
+      "learning_rate": 8.358040452836991e-06,
+      "loss": 0.8786,
+      "step": 2195
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7000333070755005,
+      "learning_rate": 8.356467854636408e-06,
+      "loss": 1.056,
+      "step": 2196
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.35874342918396,
+      "learning_rate": 8.354894651813653e-06,
+      "loss": 0.8854,
+      "step": 2197
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7063946723937988,
+      "learning_rate": 8.353320844652114e-06,
+      "loss": 1.0327,
+      "step": 2198
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3658970594406128,
+      "learning_rate": 8.351746433435295e-06,
+      "loss": 0.8906,
+      "step": 2199
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.6276720762252808,
+      "learning_rate": 8.350171418446803e-06,
+      "loss": 0.9337,
+      "step": 2200
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 3.3413889408111572,
+      "learning_rate": 8.348595799970356e-06,
+      "loss": 0.8849,
+      "step": 2201
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7854729890823364,
+      "learning_rate": 8.347019578289781e-06,
+      "loss": 1.0611,
+      "step": 2202
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4045149087905884,
+      "learning_rate": 8.34544275368901e-06,
+      "loss": 0.8752,
+      "step": 2203
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6719449758529663,
+      "learning_rate": 8.34386532645209e-06,
+      "loss": 1.0671,
+      "step": 2204
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6399966478347778,
+      "learning_rate": 8.342287296863173e-06,
+      "loss": 1.0715,
+      "step": 2205
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4103572368621826,
+      "learning_rate": 8.340708665206517e-06,
+      "loss": 0.8359,
+      "step": 2206
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3307595252990723,
+      "learning_rate": 8.339129431766492e-06,
+      "loss": 0.8774,
+      "step": 2207
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8235159516334534,
+      "learning_rate": 8.337549596827578e-06,
+      "loss": 1.0522,
+      "step": 2208
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3351058959960938,
+      "learning_rate": 8.335969160674357e-06,
+      "loss": 0.8581,
+      "step": 2209
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2698413133621216,
+      "learning_rate": 8.334388123591528e-06,
+      "loss": 0.8978,
+      "step": 2210
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3445113897323608,
+      "learning_rate": 8.332806485863888e-06,
+      "loss": 0.8791,
+      "step": 2211
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5070133209228516,
+      "learning_rate": 8.331224247776351e-06,
+      "loss": 0.8747,
+      "step": 2212
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3162689208984375,
+      "learning_rate": 8.329641409613933e-06,
+      "loss": 0.8454,
+      "step": 2213
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.702490508556366,
+      "learning_rate": 8.328057971661765e-06,
+      "loss": 1.0467,
+      "step": 2214
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3856401443481445,
+      "learning_rate": 8.326473934205077e-06,
+      "loss": 0.8701,
+      "step": 2215
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5132057666778564,
+      "learning_rate": 8.324889297529217e-06,
+      "loss": 0.8872,
+      "step": 2216
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2707164287567139,
+      "learning_rate": 8.323304061919633e-06,
+      "loss": 0.8503,
+      "step": 2217
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7070263028144836,
+      "learning_rate": 8.321718227661881e-06,
+      "loss": 1.0857,
+      "step": 2218
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.963502287864685,
+      "learning_rate": 8.320131795041633e-06,
+      "loss": 0.8429,
+      "step": 2219
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3099638223648071,
+      "learning_rate": 8.31854476434466e-06,
+      "loss": 0.8256,
+      "step": 2220
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2819503545761108,
+      "learning_rate": 8.316957135856845e-06,
+      "loss": 0.7611,
+      "step": 2221
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.30851149559021,
+      "learning_rate": 8.315368909864177e-06,
+      "loss": 0.8877,
+      "step": 2222
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.572088360786438,
+      "learning_rate": 8.313780086652753e-06,
+      "loss": 0.8294,
+      "step": 2223
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3793156147003174,
+      "learning_rate": 8.31219066650878e-06,
+      "loss": 0.8491,
+      "step": 2224
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7720226645469666,
+      "learning_rate": 8.31060064971857e-06,
+      "loss": 1.0604,
+      "step": 2225
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2822519540786743,
+      "learning_rate": 8.30901003656854e-06,
+      "loss": 0.8086,
+      "step": 2226
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4399689435958862,
+      "learning_rate": 8.307418827345224e-06,
+      "loss": 0.8804,
+      "step": 2227
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6532321572303772,
+      "learning_rate": 8.305827022335249e-06,
+      "loss": 1.0454,
+      "step": 2228
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.401733160018921,
+      "learning_rate": 8.30423462182536e-06,
+      "loss": 0.8789,
+      "step": 2229
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4192389249801636,
+      "learning_rate": 8.30264162610241e-06,
+      "loss": 0.889,
+      "step": 2230
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3178857564926147,
+      "learning_rate": 8.30104803545335e-06,
+      "loss": 0.84,
+      "step": 2231
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7783831357955933,
+      "learning_rate": 8.299453850165247e-06,
+      "loss": 1.0706,
+      "step": 2232
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4116562604904175,
+      "learning_rate": 8.297859070525271e-06,
+      "loss": 0.8446,
+      "step": 2233
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.357462763786316,
+      "learning_rate": 8.296263696820701e-06,
+      "loss": 0.8923,
+      "step": 2234
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3446985483169556,
+      "learning_rate": 8.29466772933892e-06,
+      "loss": 0.8618,
+      "step": 2235
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.2808020114898682,
+      "learning_rate": 8.293071168367422e-06,
+      "loss": 0.889,
+      "step": 2236
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4724957942962646,
+      "learning_rate": 8.291474014193803e-06,
+      "loss": 0.8875,
+      "step": 2237
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.374625563621521,
+      "learning_rate": 8.289876267105771e-06,
+      "loss": 0.8631,
+      "step": 2238
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.380990982055664,
+      "learning_rate": 8.288277927391138e-06,
+      "loss": 0.8971,
+      "step": 2239
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4264053106307983,
+      "learning_rate": 8.286678995337823e-06,
+      "loss": 0.8947,
+      "step": 2240
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3692878484725952,
+      "learning_rate": 8.285079471233852e-06,
+      "loss": 0.8792,
+      "step": 2241
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3494952917099,
+      "learning_rate": 8.283479355367357e-06,
+      "loss": 0.8595,
+      "step": 2242
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4638673067092896,
+      "learning_rate": 8.281878648026577e-06,
+      "loss": 0.8634,
+      "step": 2243
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.5175824165344238,
+      "learning_rate": 8.280277349499863e-06,
+      "loss": 0.8561,
+      "step": 2244
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.4309653043746948,
+      "learning_rate": 8.278675460075658e-06,
+      "loss": 0.8884,
+      "step": 2245
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3614115715026855,
+      "learning_rate": 8.277072980042528e-06,
+      "loss": 0.8375,
+      "step": 2246
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.9764857292175293,
+      "learning_rate": 8.275469909689134e-06,
+      "loss": 1.0505,
+      "step": 2247
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.330025553703308,
+      "learning_rate": 8.27386624930425e-06,
+      "loss": 0.8599,
+      "step": 2248
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6691585779190063,
+      "learning_rate": 8.27226199917675e-06,
+      "loss": 1.0615,
+      "step": 2249
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.7028941512107849,
+      "learning_rate": 8.270657159595623e-06,
+      "loss": 1.0347,
+      "step": 2250
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 1.3520692586898804,
+      "learning_rate": 8.269051730849955e-06,
+      "loss": 0.7978,
+      "step": 2251
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6978772282600403,
+      "learning_rate": 8.267445713228943e-06,
+      "loss": 1.0489,
+      "step": 2252
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.394895076751709,
+      "learning_rate": 8.265839107021893e-06,
+      "loss": 0.8786,
+      "step": 2253
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.592198133468628,
+      "learning_rate": 8.264231912518207e-06,
+      "loss": 0.9044,
+      "step": 2254
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7334381937980652,
+      "learning_rate": 8.262624130007404e-06,
+      "loss": 1.0713,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7637110352516174,
+      "learning_rate": 8.261015759779103e-06,
+      "loss": 1.0538,
+      "step": 2256
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8418129682540894,
+      "learning_rate": 8.259406802123029e-06,
+      "loss": 0.8501,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3038573265075684,
+      "learning_rate": 8.257797257329017e-06,
+      "loss": 0.8537,
+      "step": 2258
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7001702189445496,
+      "learning_rate": 8.256187125687001e-06,
+      "loss": 1.0549,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3077365159988403,
+      "learning_rate": 8.254576407487026e-06,
+      "loss": 0.8233,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.387817621231079,
+      "learning_rate": 8.25296510301924e-06,
+      "loss": 0.8548,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3940025568008423,
+      "learning_rate": 8.251353212573899e-06,
+      "loss": 0.7893,
+      "step": 2262
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2991691827774048,
+      "learning_rate": 8.249740736441362e-06,
+      "loss": 0.8135,
+      "step": 2263
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.867573618888855,
+      "learning_rate": 8.248127674912097e-06,
+      "loss": 1.0834,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.334493637084961,
+      "learning_rate": 8.246514028276675e-06,
+      "loss": 0.9115,
+      "step": 2265
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.265126347541809,
+      "learning_rate": 8.244899796825767e-06,
+      "loss": 0.8739,
+      "step": 2266
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2710767984390259,
+      "learning_rate": 8.243284980850164e-06,
+      "loss": 0.8305,
+      "step": 2267
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4627249240875244,
+      "learning_rate": 8.241669580640746e-06,
+      "loss": 0.8584,
+      "step": 2268
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4345078468322754,
+      "learning_rate": 8.240053596488507e-06,
+      "loss": 0.8574,
+      "step": 2269
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3705072402954102,
+      "learning_rate": 8.238437028684546e-06,
+      "loss": 0.8913,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3475830554962158,
+      "learning_rate": 8.236819877520064e-06,
+      "loss": 0.8408,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3350614309310913,
+      "learning_rate": 8.235202143286371e-06,
+      "loss": 0.8897,
+      "step": 2272
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.8533003330230713,
+      "learning_rate": 8.23358382627488e-06,
+      "loss": 0.9167,
+      "step": 2273
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8875615000724792,
+      "learning_rate": 8.231964926777106e-06,
+      "loss": 1.0332,
+      "step": 2274
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.440656065940857,
+      "learning_rate": 8.230345445084675e-06,
+      "loss": 0.8662,
+      "step": 2275
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3593724966049194,
+      "learning_rate": 8.228725381489312e-06,
+      "loss": 0.8643,
+      "step": 2276
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6643288135528564,
+      "learning_rate": 8.227104736282851e-06,
+      "loss": 1.0691,
+      "step": 2277
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.439435601234436,
+      "learning_rate": 8.225483509757231e-06,
+      "loss": 0.842,
+      "step": 2278
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4375998973846436,
+      "learning_rate": 8.22386170220449e-06,
+      "loss": 0.8679,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3005706071853638,
+      "learning_rate": 8.222239313916776e-06,
+      "loss": 0.8823,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4286876916885376,
+      "learning_rate": 8.220616345186342e-06,
+      "loss": 0.8552,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4202978610992432,
+      "learning_rate": 8.218992796305544e-06,
+      "loss": 0.8953,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9888331294059753,
+      "learning_rate": 8.217368667566838e-06,
+      "loss": 1.0546,
+      "step": 2283
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.42974853515625,
+      "learning_rate": 8.215743959262793e-06,
+      "loss": 0.8841,
+      "step": 2284
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3745731115341187,
+      "learning_rate": 8.214118671686078e-06,
+      "loss": 0.902,
+      "step": 2285
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7238791584968567,
+      "learning_rate": 8.212492805129465e-06,
+      "loss": 1.0635,
+      "step": 2286
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.578402519226074,
+      "learning_rate": 8.21086635988583e-06,
+      "loss": 0.8811,
+      "step": 2287
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7576668858528137,
+      "learning_rate": 8.209239336248159e-06,
+      "loss": 1.0469,
+      "step": 2288
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6693471074104309,
+      "learning_rate": 8.207611734509537e-06,
+      "loss": 1.0559,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6674351692199707,
+      "learning_rate": 8.205983554963152e-06,
+      "loss": 1.0557,
+      "step": 2290
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.7547013759613037,
+      "learning_rate": 8.204354797902304e-06,
+      "loss": 0.8743,
+      "step": 2291
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2705607414245605,
+      "learning_rate": 8.202725463620384e-06,
+      "loss": 0.8832,
+      "step": 2292
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9185460209846497,
+      "learning_rate": 8.2010955524109e-06,
+      "loss": 1.0539,
+      "step": 2293
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.210677981376648,
+      "learning_rate": 8.199465064567457e-06,
+      "loss": 0.8477,
+      "step": 2294
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3373026847839355,
+      "learning_rate": 8.197834000383766e-06,
+      "loss": 0.8241,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.477432131767273,
+      "learning_rate": 8.19620236015364e-06,
+      "loss": 0.9371,
+      "step": 2296
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4587979316711426,
+      "learning_rate": 8.194570144170997e-06,
+      "loss": 0.9129,
+      "step": 2297
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3262968063354492,
+      "learning_rate": 8.192937352729858e-06,
+      "loss": 0.8958,
+      "step": 2298
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9128394722938538,
+      "learning_rate": 8.191303986124347e-06,
+      "loss": 1.0331,
+      "step": 2299
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.331095576286316,
+      "learning_rate": 8.189670044648696e-06,
+      "loss": 0.8455,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7053428292274475,
+      "learning_rate": 8.188035528597237e-06,
+      "loss": 1.0607,
+      "step": 2301
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4467360973358154,
+      "learning_rate": 8.186400438264405e-06,
+      "loss": 0.8791,
+      "step": 2302
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.525076150894165,
+      "learning_rate": 8.184764773944737e-06,
+      "loss": 0.8021,
+      "step": 2303
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.469259262084961,
+      "learning_rate": 8.183128535932882e-06,
+      "loss": 0.8732,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4053525924682617,
+      "learning_rate": 8.18149172452358e-06,
+      "loss": 0.8747,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.518134355545044,
+      "learning_rate": 8.179854340011683e-06,
+      "loss": 0.9209,
+      "step": 2306
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3835525512695312,
+      "learning_rate": 8.178216382692143e-06,
+      "loss": 0.8746,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4152497053146362,
+      "learning_rate": 8.176577852860015e-06,
+      "loss": 0.8522,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2004226446151733,
+      "learning_rate": 8.174938750810461e-06,
+      "loss": 1.0572,
+      "step": 2309
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9180067181587219,
+      "learning_rate": 8.173299076838739e-06,
+      "loss": 1.0454,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7919889688491821,
+      "learning_rate": 8.171658831240216e-06,
+      "loss": 1.0458,
+      "step": 2311
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7796300649642944,
+      "learning_rate": 8.170018014310361e-06,
+      "loss": 1.0405,
+      "step": 2312
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.513171672821045,
+      "learning_rate": 8.168376626344743e-06,
+      "loss": 0.9232,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4488352537155151,
+      "learning_rate": 8.166734667639039e-06,
+      "loss": 0.8885,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3799651861190796,
+      "learning_rate": 8.16509213848902e-06,
+      "loss": 0.8146,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.473968744277954,
+      "learning_rate": 8.163449039190569e-06,
+      "loss": 1.0712,
+      "step": 2316
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3480080366134644,
+      "learning_rate": 8.161805370039671e-06,
+      "loss": 0.8247,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.6661397218704224,
+      "learning_rate": 8.160161131332404e-06,
+      "loss": 0.8182,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.270523190498352,
+      "learning_rate": 8.15851632336496e-06,
+      "loss": 0.8434,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.2598754167556763,
+      "learning_rate": 8.156870946433628e-06,
+      "loss": 0.8622,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3360073566436768,
+      "learning_rate": 8.1552250008348e-06,
+      "loss": 0.8763,
+      "step": 2321
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4238051176071167,
+      "learning_rate": 8.153578486864967e-06,
+      "loss": 0.8721,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.3249090909957886,
+      "learning_rate": 8.151931404820734e-06,
+      "loss": 0.8743,
+      "step": 2323
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.1962590217590332,
+      "learning_rate": 8.150283754998795e-06,
+      "loss": 1.0565,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.8864197731018066,
+      "learning_rate": 8.148635537695955e-06,
+      "loss": 1.0538,
+      "step": 2325
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7142506241798401,
+      "learning_rate": 8.146986753209114e-06,
+      "loss": 1.0413,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 1.4882700443267822,
+      "learning_rate": 8.14533740183528e-06,
+      "loss": 0.8899,
+      "step": 2327
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4054632186889648,
+      "learning_rate": 8.143687483871563e-06,
+      "loss": 0.835,
+      "step": 2328
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5190225839614868,
+      "learning_rate": 8.142036999615171e-06,
+      "loss": 0.8766,
+      "step": 2329
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3997687101364136,
+      "learning_rate": 8.14038594936342e-06,
+      "loss": 0.8317,
+      "step": 2330
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3671146631240845,
+      "learning_rate": 8.138734333413715e-06,
+      "loss": 0.8654,
+      "step": 2331
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2739046812057495,
+      "learning_rate": 8.137082152063584e-06,
+      "loss": 0.8835,
+      "step": 2332
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.395334005355835,
+      "learning_rate": 8.13542940561064e-06,
+      "loss": 0.8947,
+      "step": 2333
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4119561910629272,
+      "learning_rate": 8.133776094352602e-06,
+      "loss": 0.8918,
+      "step": 2334
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4013473987579346,
+      "learning_rate": 8.13212221858729e-06,
+      "loss": 0.913,
+      "step": 2335
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.389649510383606,
+      "learning_rate": 8.130467778612634e-06,
+      "loss": 0.8849,
+      "step": 2336
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4662131071090698,
+      "learning_rate": 8.128812774726651e-06,
+      "loss": 0.8739,
+      "step": 2337
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3381640911102295,
+      "learning_rate": 8.127157207227473e-06,
+      "loss": 0.8519,
+      "step": 2338
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2455556392669678,
+      "learning_rate": 8.125501076413325e-06,
+      "loss": 0.8702,
+      "step": 2339
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.269357919692993,
+      "learning_rate": 8.123844382582538e-06,
+      "loss": 1.0539,
+      "step": 2340
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.9220174551010132,
+      "learning_rate": 8.122187126033542e-06,
+      "loss": 1.086,
+      "step": 2341
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.369940996170044,
+      "learning_rate": 8.12052930706487e-06,
+      "loss": 0.8669,
+      "step": 2342
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4550926685333252,
+      "learning_rate": 8.118870925975158e-06,
+      "loss": 0.8592,
+      "step": 2343
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.426878809928894,
+      "learning_rate": 8.117211983063134e-06,
+      "loss": 0.9016,
+      "step": 2344
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.417916178703308,
+      "learning_rate": 8.115552478627642e-06,
+      "loss": 0.9156,
+      "step": 2345
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3465172052383423,
+      "learning_rate": 8.113892412967614e-06,
+      "loss": 0.7878,
+      "step": 2346
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3681279420852661,
+      "learning_rate": 8.11223178638209e-06,
+      "loss": 0.8689,
+      "step": 2347
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3823853731155396,
+      "learning_rate": 8.11057059917021e-06,
+      "loss": 0.884,
+      "step": 2348
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.377400279045105,
+      "learning_rate": 8.108908851631213e-06,
+      "loss": 0.8638,
+      "step": 2349
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3906524181365967,
+      "learning_rate": 8.10724654406444e-06,
+      "loss": 0.8531,
+      "step": 2350
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4404971599578857,
+      "learning_rate": 8.105583676769336e-06,
+      "loss": 0.9215,
+      "step": 2351
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4389725923538208,
+      "learning_rate": 8.103920250045442e-06,
+      "loss": 0.8959,
+      "step": 2352
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.4590914249420166,
+      "learning_rate": 8.1022562641924e-06,
+      "loss": 1.0755,
+      "step": 2353
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.489874005317688,
+      "learning_rate": 8.100591719509959e-06,
+      "loss": 0.8951,
+      "step": 2354
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3509197235107422,
+      "learning_rate": 8.09892661629796e-06,
+      "loss": 0.896,
+      "step": 2355
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4670807123184204,
+      "learning_rate": 8.09726095485635e-06,
+      "loss": 1.052,
+      "step": 2356
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.9160798192024231,
+      "learning_rate": 8.095594735485177e-06,
+      "loss": 1.0639,
+      "step": 2357
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8568662405014038,
+      "learning_rate": 8.093927958484585e-06,
+      "loss": 1.0447,
+      "step": 2358
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6072900295257568,
+      "learning_rate": 8.092260624154822e-06,
+      "loss": 0.9054,
+      "step": 2359
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.564530372619629,
+      "learning_rate": 8.09059273279624e-06,
+      "loss": 0.9324,
+      "step": 2360
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6262537240982056,
+      "learning_rate": 8.08892428470928e-06,
+      "loss": 0.9068,
+      "step": 2361
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 2.182525157928467,
+      "learning_rate": 8.087255280194494e-06,
+      "loss": 1.0522,
+      "step": 2362
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4464929103851318,
+      "learning_rate": 8.08558571955253e-06,
+      "loss": 0.8564,
+      "step": 2363
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.7512189149856567,
+      "learning_rate": 8.083915603084138e-06,
+      "loss": 0.9424,
+      "step": 2364
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6079978942871094,
+      "learning_rate": 8.082244931090167e-06,
+      "loss": 0.8449,
+      "step": 2365
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.482356071472168,
+      "learning_rate": 8.080573703871563e-06,
+      "loss": 0.8626,
+      "step": 2366
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5486292839050293,
+      "learning_rate": 8.078901921729377e-06,
+      "loss": 0.8988,
+      "step": 2367
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5163346529006958,
+      "learning_rate": 8.077229584964757e-06,
+      "loss": 1.06,
+      "step": 2368
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.099661946296692,
+      "learning_rate": 8.075556693878954e-06,
+      "loss": 1.051,
+      "step": 2369
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7769708037376404,
+      "learning_rate": 8.073883248773313e-06,
+      "loss": 1.0587,
+      "step": 2370
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.501251220703125,
+      "learning_rate": 8.072209249949285e-06,
+      "loss": 0.8975,
+      "step": 2371
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5075386762619019,
+      "learning_rate": 8.07053469770842e-06,
+      "loss": 0.8485,
+      "step": 2372
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.5272488594055176,
+      "learning_rate": 8.068859592352359e-06,
+      "loss": 1.0678,
+      "step": 2373
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4914487600326538,
+      "learning_rate": 8.067183934182856e-06,
+      "loss": 0.8835,
+      "step": 2374
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.6101833581924438,
+      "learning_rate": 8.065507723501757e-06,
+      "loss": 0.8686,
+      "step": 2375
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.332781195640564,
+      "learning_rate": 8.063830960611005e-06,
+      "loss": 0.9001,
+      "step": 2376
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3252266645431519,
+      "learning_rate": 8.062153645812651e-06,
+      "loss": 0.8481,
+      "step": 2377
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.386142373085022,
+      "learning_rate": 8.060475779408839e-06,
+      "loss": 0.8874,
+      "step": 2378
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4343831539154053,
+      "learning_rate": 8.058797361701812e-06,
+      "loss": 0.8288,
+      "step": 2379
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4223213195800781,
+      "learning_rate": 8.057118392993916e-06,
+      "loss": 0.8435,
+      "step": 2380
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4389142990112305,
+      "learning_rate": 8.055438873587593e-06,
+      "loss": 0.8233,
+      "step": 2381
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3330577611923218,
+      "learning_rate": 8.053758803785387e-06,
+      "loss": 0.8077,
+      "step": 2382
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4836986064910889,
+      "learning_rate": 8.052078183889938e-06,
+      "loss": 1.0693,
+      "step": 2383
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.38245689868927,
+      "learning_rate": 8.05039701420399e-06,
+      "loss": 0.7937,
+      "step": 2384
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.087321162223816,
+      "learning_rate": 8.04871529503038e-06,
+      "loss": 1.0415,
+      "step": 2385
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.8594996929168701,
+      "learning_rate": 8.047033026672049e-06,
+      "loss": 1.0305,
+      "step": 2386
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3257992267608643,
+      "learning_rate": 8.045350209432033e-06,
+      "loss": 0.8697,
+      "step": 2387
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.4538222551345825,
+      "learning_rate": 8.04366684361347e-06,
+      "loss": 0.8225,
+      "step": 2388
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.047473669052124,
+      "learning_rate": 8.041982929519595e-06,
+      "loss": 1.0901,
+      "step": 2389
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.897099256515503,
+      "learning_rate": 8.040298467453742e-06,
+      "loss": 0.7993,
+      "step": 2390
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.0937689542770386,
+      "learning_rate": 8.038613457719344e-06,
+      "loss": 1.0707,
+      "step": 2391
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3952254056930542,
+      "learning_rate": 8.036927900619932e-06,
+      "loss": 0.8374,
+      "step": 2392
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3553649187088013,
+      "learning_rate": 8.035241796459142e-06,
+      "loss": 0.8855,
+      "step": 2393
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3433171510696411,
+      "learning_rate": 8.033555145540693e-06,
+      "loss": 0.8964,
+      "step": 2394
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3145830631256104,
+      "learning_rate": 8.031867948168418e-06,
+      "loss": 0.8876,
+      "step": 2395
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.321035385131836,
+      "learning_rate": 8.030180204646245e-06,
+      "loss": 0.8413,
+      "step": 2396
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3518011569976807,
+      "learning_rate": 8.028491915278193e-06,
+      "loss": 0.836,
+      "step": 2397
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.2974841594696045,
+      "learning_rate": 8.026803080368387e-06,
+      "loss": 0.855,
+      "step": 2398
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.057131052017212,
+      "learning_rate": 8.025113700221047e-06,
+      "loss": 1.0626,
+      "step": 2399
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3061957359313965,
+      "learning_rate": 8.023423775140494e-06,
+      "loss": 0.8743,
+      "step": 2400
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3297899961471558,
+      "learning_rate": 8.021733305431141e-06,
+      "loss": 0.859,
+      "step": 2401
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.474126935005188,
+      "learning_rate": 8.020042291397507e-06,
+      "loss": 0.9285,
+      "step": 2402
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 1.3225809335708618,
+      "learning_rate": 8.018350733344203e-06,
+      "loss": 0.9028,
+      "step": 2403
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7394471168518066,
+      "learning_rate": 8.016658631575942e-06,
+      "loss": 1.0257,
+      "step": 2404
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7132319211959839,
+      "learning_rate": 8.01496598639753e-06,
+      "loss": 1.0544,
+      "step": 2405
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5568954944610596,
+      "learning_rate": 8.01327279811388e-06,
+      "loss": 0.8228,
+      "step": 2406
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3915327787399292,
+      "learning_rate": 8.01157906702999e-06,
+      "loss": 0.857,
+      "step": 2407
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4026364088058472,
+      "learning_rate": 8.009884793450968e-06,
+      "loss": 0.8549,
+      "step": 2408
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7554807066917419,
+      "learning_rate": 8.008189977682013e-06,
+      "loss": 1.0576,
+      "step": 2409
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.417267918586731,
+      "learning_rate": 8.00649462002842e-06,
+      "loss": 0.8035,
+      "step": 2410
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.767094612121582,
+      "learning_rate": 8.004798720795588e-06,
+      "loss": 0.8994,
+      "step": 2411
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.484433889389038,
+      "learning_rate": 8.00310228028901e-06,
+      "loss": 0.89,
+      "step": 2412
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8331546187400818,
+      "learning_rate": 8.001405298814275e-06,
+      "loss": 1.0571,
+      "step": 2413
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.420613408088684,
+      "learning_rate": 7.999707776677072e-06,
+      "loss": 0.8919,
+      "step": 2414
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2654937505722046,
+      "learning_rate": 7.998009714183188e-06,
+      "loss": 0.8461,
+      "step": 2415
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4899762868881226,
+      "learning_rate": 7.996311111638505e-06,
+      "loss": 0.86,
+      "step": 2416
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3116148710250854,
+      "learning_rate": 7.994611969349004e-06,
+      "loss": 0.9469,
+      "step": 2417
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7337781190872192,
+      "learning_rate": 7.99291228762076e-06,
+      "loss": 1.0381,
+      "step": 2418
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.270693063735962,
+      "learning_rate": 7.991212066759949e-06,
+      "loss": 0.9055,
+      "step": 2419
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5247607231140137,
+      "learning_rate": 7.989511307072842e-06,
+      "loss": 0.858,
+      "step": 2420
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4520014524459839,
+      "learning_rate": 7.98781000886581e-06,
+      "loss": 0.9087,
+      "step": 2421
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3829913139343262,
+      "learning_rate": 7.986108172445318e-06,
+      "loss": 0.8741,
+      "step": 2422
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7727935314178467,
+      "learning_rate": 7.984405798117928e-06,
+      "loss": 1.0417,
+      "step": 2423
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2924716472625732,
+      "learning_rate": 7.982702886190299e-06,
+      "loss": 0.8462,
+      "step": 2424
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6649379134178162,
+      "learning_rate": 7.980999436969187e-06,
+      "loss": 1.0515,
+      "step": 2425
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2940984964370728,
+      "learning_rate": 7.97929545076145e-06,
+      "loss": 0.8849,
+      "step": 2426
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2897077798843384,
+      "learning_rate": 7.977590927874033e-06,
+      "loss": 0.8479,
+      "step": 2427
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6642120480537415,
+      "learning_rate": 7.975885868613985e-06,
+      "loss": 1.0799,
+      "step": 2428
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.384598970413208,
+      "learning_rate": 7.974180273288447e-06,
+      "loss": 0.8785,
+      "step": 2429
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4265908002853394,
+      "learning_rate": 7.972474142204663e-06,
+      "loss": 0.8404,
+      "step": 2430
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2268563508987427,
+      "learning_rate": 7.970767475669966e-06,
+      "loss": 0.8061,
+      "step": 2431
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2447587251663208,
+      "learning_rate": 7.969060273991787e-06,
+      "loss": 0.7967,
+      "step": 2432
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.236411452293396,
+      "learning_rate": 7.96735253747766e-06,
+      "loss": 0.8378,
+      "step": 2433
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7313517332077026,
+      "learning_rate": 7.965644266435209e-06,
+      "loss": 1.0406,
+      "step": 2434
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7292114496231079,
+      "learning_rate": 7.963935461172156e-06,
+      "loss": 1.0566,
+      "step": 2435
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.303958535194397,
+      "learning_rate": 7.962226121996317e-06,
+      "loss": 0.851,
+      "step": 2436
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2984391450881958,
+      "learning_rate": 7.960516249215609e-06,
+      "loss": 0.9122,
+      "step": 2437
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3215768337249756,
+      "learning_rate": 7.95880584313804e-06,
+      "loss": 0.8881,
+      "step": 2438
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2628270387649536,
+      "learning_rate": 7.957094904071718e-06,
+      "loss": 0.8754,
+      "step": 2439
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2914276123046875,
+      "learning_rate": 7.955383432324846e-06,
+      "loss": 0.8811,
+      "step": 2440
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.2223663330078125,
+      "learning_rate": 7.95367142820572e-06,
+      "loss": 0.8647,
+      "step": 2441
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9081702828407288,
+      "learning_rate": 7.951958892022738e-06,
+      "loss": 1.0604,
+      "step": 2442
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.8501556515693665,
+      "learning_rate": 7.950245824084386e-06,
+      "loss": 1.0553,
+      "step": 2443
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3779456615447998,
+      "learning_rate": 7.948532224699253e-06,
+      "loss": 0.8816,
+      "step": 2444
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3788257837295532,
+      "learning_rate": 7.946818094176019e-06,
+      "loss": 0.8599,
+      "step": 2445
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.324237585067749,
+      "learning_rate": 7.945103432823463e-06,
+      "loss": 0.7894,
+      "step": 2446
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3806028366088867,
+      "learning_rate": 7.943388240950456e-06,
+      "loss": 0.9001,
+      "step": 2447
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2323811054229736,
+      "learning_rate": 7.941672518865969e-06,
+      "loss": 1.0642,
+      "step": 2448
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0110344886779785,
+      "learning_rate": 7.939956266879063e-06,
+      "loss": 1.0736,
+      "step": 2449
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3843166828155518,
+      "learning_rate": 7.9382394852989e-06,
+      "loss": 0.812,
+      "step": 2450
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.5421262979507446,
+      "learning_rate": 7.936522174434736e-06,
+      "loss": 0.94,
+      "step": 2451
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0039469003677368,
+      "learning_rate": 7.934804334595915e-06,
+      "loss": 1.0402,
+      "step": 2452
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4370644092559814,
+      "learning_rate": 7.93308596609189e-06,
+      "loss": 0.8907,
+      "step": 2453
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3175280094146729,
+      "learning_rate": 7.9313670692322e-06,
+      "loss": 0.8525,
+      "step": 2454
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3068184852600098,
+      "learning_rate": 7.929647644326477e-06,
+      "loss": 0.8951,
+      "step": 2455
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3654981851577759,
+      "learning_rate": 7.927927691684457e-06,
+      "loss": 0.8468,
+      "step": 2456
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4025169610977173,
+      "learning_rate": 7.926207211615962e-06,
+      "loss": 0.9047,
+      "step": 2457
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4432016611099243,
+      "learning_rate": 7.924486204430917e-06,
+      "loss": 0.8619,
+      "step": 2458
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.381432056427002,
+      "learning_rate": 7.922764670439335e-06,
+      "loss": 0.8557,
+      "step": 2459
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3716864585876465,
+      "learning_rate": 7.921042609951328e-06,
+      "loss": 0.8819,
+      "step": 2460
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3042714595794678,
+      "learning_rate": 7.919320023277101e-06,
+      "loss": 0.8563,
+      "step": 2461
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3130065202713013,
+      "learning_rate": 7.917596910726958e-06,
+      "loss": 0.8584,
+      "step": 2462
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3375167846679688,
+      "learning_rate": 7.91587327261129e-06,
+      "loss": 0.8583,
+      "step": 2463
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3982305526733398,
+      "learning_rate": 7.914149109240592e-06,
+      "loss": 0.9069,
+      "step": 2464
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3929204940795898,
+      "learning_rate": 7.912424420925444e-06,
+      "loss": 0.8381,
+      "step": 2465
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.0996073484420776,
+      "learning_rate": 7.910699207976527e-06,
+      "loss": 1.0522,
+      "step": 2466
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3251118659973145,
+      "learning_rate": 7.908973470704614e-06,
+      "loss": 0.8583,
+      "step": 2467
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.252397060394287,
+      "learning_rate": 7.907247209420575e-06,
+      "loss": 0.8273,
+      "step": 2468
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7277053594589233,
+      "learning_rate": 7.905520424435371e-06,
+      "loss": 1.064,
+      "step": 2469
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3573458194732666,
+      "learning_rate": 7.90379311606006e-06,
+      "loss": 0.7705,
+      "step": 2470
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6909373998641968,
+      "learning_rate": 7.90206528460579e-06,
+      "loss": 1.0625,
+      "step": 2471
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7077136635780334,
+      "learning_rate": 7.90033693038381e-06,
+      "loss": 1.047,
+      "step": 2472
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3941906690597534,
+      "learning_rate": 7.89860805370546e-06,
+      "loss": 0.8944,
+      "step": 2473
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3941609859466553,
+      "learning_rate": 7.89687865488217e-06,
+      "loss": 0.8052,
+      "step": 2474
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4106205701828003,
+      "learning_rate": 7.895148734225472e-06,
+      "loss": 0.8567,
+      "step": 2475
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.4376025199890137,
+      "learning_rate": 7.893418292046984e-06,
+      "loss": 0.8421,
+      "step": 2476
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.2935487031936646,
+      "learning_rate": 7.891687328658423e-06,
+      "loss": 0.8842,
+      "step": 2477
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3297611474990845,
+      "learning_rate": 7.889955844371601e-06,
+      "loss": 0.8851,
+      "step": 2478
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 1.3833492994308472,
+      "learning_rate": 7.88822383949842e-06,
+      "loss": 0.8828,
+      "step": 2479
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.9494116902351379,
+      "learning_rate": 7.886491314350874e-06,
+      "loss": 1.0498,
+      "step": 2480
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.310686707496643,
+      "learning_rate": 7.884758269241057e-06,
+      "loss": 0.8324,
+      "step": 2481
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7221014499664307,
+      "learning_rate": 7.883024704481154e-06,
+      "loss": 1.0322,
+      "step": 2482
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3975856304168701,
+      "learning_rate": 7.881290620383442e-06,
+      "loss": 0.8847,
+      "step": 2483
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3217295408248901,
+      "learning_rate": 7.879556017260293e-06,
+      "loss": 0.8576,
+      "step": 2484
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7412630319595337,
+      "learning_rate": 7.87782089542417e-06,
+      "loss": 1.0393,
+      "step": 2485
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4124795198440552,
+      "learning_rate": 7.876085255187638e-06,
+      "loss": 0.8926,
+      "step": 2486
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.326843500137329,
+      "learning_rate": 7.874349096863341e-06,
+      "loss": 0.911,
+      "step": 2487
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7309504747390747,
+      "learning_rate": 7.872612420764031e-06,
+      "loss": 1.053,
+      "step": 2488
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7032256722450256,
+      "learning_rate": 7.870875227202544e-06,
+      "loss": 1.055,
+      "step": 2489
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6449598073959351,
+      "learning_rate": 7.86913751649181e-06,
+      "loss": 1.0316,
+      "step": 2490
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.353677749633789,
+      "learning_rate": 7.867399288944855e-06,
+      "loss": 0.8431,
+      "step": 2491
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3616108894348145,
+      "learning_rate": 7.8656605448748e-06,
+      "loss": 0.7752,
+      "step": 2492
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.552804946899414,
+      "learning_rate": 7.863921284594856e-06,
+      "loss": 0.8614,
+      "step": 2493
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3504129648208618,
+      "learning_rate": 7.86218150841832e-06,
+      "loss": 0.8894,
+      "step": 2494
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3156113624572754,
+      "learning_rate": 7.8604412166586e-06,
+      "loss": 0.8098,
+      "step": 2495
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3730769157409668,
+      "learning_rate": 7.858700409629177e-06,
+      "loss": 0.8681,
+      "step": 2496
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.0599480867385864,
+      "learning_rate": 7.856959087643638e-06,
+      "loss": 1.0277,
+      "step": 2497
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3462780714035034,
+      "learning_rate": 7.855217251015658e-06,
+      "loss": 0.877,
+      "step": 2498
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3449459075927734,
+      "learning_rate": 7.853474900059005e-06,
+      "loss": 0.7931,
+      "step": 2499
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.290866494178772,
+      "learning_rate": 7.851732035087537e-06,
+      "loss": 0.822,
+      "step": 2500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.47977876663208,
+      "learning_rate": 7.849988656415214e-06,
+      "loss": 0.8431,
+      "step": 2501
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7629051208496094,
+      "learning_rate": 7.848244764356076e-06,
+      "loss": 1.0575,
+      "step": 2502
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3270139694213867,
+      "learning_rate": 7.846500359224266e-06,
+      "loss": 0.8275,
+      "step": 2503
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3538626432418823,
+      "learning_rate": 7.844755441334012e-06,
+      "loss": 0.8866,
+      "step": 2504
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4356274604797363,
+      "learning_rate": 7.843010010999636e-06,
+      "loss": 0.8287,
+      "step": 2505
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2324256896972656,
+      "learning_rate": 7.841264068535559e-06,
+      "loss": 0.8483,
+      "step": 2506
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.5315293073654175,
+      "learning_rate": 7.839517614256282e-06,
+      "loss": 0.8802,
+      "step": 2507
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.260432243347168,
+      "learning_rate": 7.837770648476411e-06,
+      "loss": 0.8683,
+      "step": 2508
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3747426271438599,
+      "learning_rate": 7.836023171510637e-06,
+      "loss": 0.876,
+      "step": 2509
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.409818172454834,
+      "learning_rate": 7.83427518367374e-06,
+      "loss": 0.8488,
+      "step": 2510
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4334816932678223,
+      "learning_rate": 7.832526685280601e-06,
+      "loss": 0.8218,
+      "step": 2511
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.350540280342102,
+      "learning_rate": 7.830777676646187e-06,
+      "loss": 0.8719,
+      "step": 2512
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.31666100025177,
+      "learning_rate": 7.829028158085556e-06,
+      "loss": 0.8313,
+      "step": 2513
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4702039957046509,
+      "learning_rate": 7.827278129913865e-06,
+      "loss": 0.8233,
+      "step": 2514
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4699398279190063,
+      "learning_rate": 7.825527592446353e-06,
+      "loss": 0.813,
+      "step": 2515
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3092087507247925,
+      "learning_rate": 7.823776545998357e-06,
+      "loss": 0.9029,
+      "step": 2516
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.371852993965149,
+      "learning_rate": 7.822024990885304e-06,
+      "loss": 0.8396,
+      "step": 2517
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.845367968082428,
+      "learning_rate": 7.820272927422717e-06,
+      "loss": 1.0886,
+      "step": 2518
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7091947197914124,
+      "learning_rate": 7.818520355926198e-06,
+      "loss": 1.0596,
+      "step": 2519
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.6656858921051025,
+      "learning_rate": 7.816767276711458e-06,
+      "loss": 0.8581,
+      "step": 2520
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3614180088043213,
+      "learning_rate": 7.815013690094284e-06,
+      "loss": 0.8105,
+      "step": 2521
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9445258975028992,
+      "learning_rate": 7.813259596390565e-06,
+      "loss": 1.0702,
+      "step": 2522
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7913612723350525,
+      "learning_rate": 7.811504995916275e-06,
+      "loss": 1.0374,
+      "step": 2523
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.361954927444458,
+      "learning_rate": 7.809749888987482e-06,
+      "loss": 0.9021,
+      "step": 2524
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 6.004542827606201,
+      "learning_rate": 7.807994275920345e-06,
+      "loss": 1.0399,
+      "step": 2525
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3536205291748047,
+      "learning_rate": 7.806238157031112e-06,
+      "loss": 0.8717,
+      "step": 2526
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.2976326942443848,
+      "learning_rate": 7.804481532636128e-06,
+      "loss": 0.779,
+      "step": 2527
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.353508710861206,
+      "learning_rate": 7.80272440305182e-06,
+      "loss": 0.7898,
+      "step": 2528
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3044395446777344,
+      "learning_rate": 7.800966768594715e-06,
+      "loss": 0.8307,
+      "step": 2529
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3902554512023926,
+      "learning_rate": 7.799208629581426e-06,
+      "loss": 0.879,
+      "step": 2530
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3574479818344116,
+      "learning_rate": 7.797449986328657e-06,
+      "loss": 0.8778,
+      "step": 2531
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.450240135192871,
+      "learning_rate": 7.795690839153205e-06,
+      "loss": 1.0635,
+      "step": 2532
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3868077993392944,
+      "learning_rate": 7.793931188371955e-06,
+      "loss": 0.8559,
+      "step": 2533
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.795284628868103,
+      "learning_rate": 7.792171034301885e-06,
+      "loss": 1.0771,
+      "step": 2534
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.427801251411438,
+      "learning_rate": 7.790410377260065e-06,
+      "loss": 0.9299,
+      "step": 2535
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.477254867553711,
+      "learning_rate": 7.78864921756365e-06,
+      "loss": 0.873,
+      "step": 2536
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3029768466949463,
+      "learning_rate": 7.78688755552989e-06,
+      "loss": 0.9093,
+      "step": 2537
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.224205493927002,
+      "learning_rate": 7.785125391476125e-06,
+      "loss": 0.7856,
+      "step": 2538
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3681889772415161,
+      "learning_rate": 7.783362725719788e-06,
+      "loss": 0.8987,
+      "step": 2539
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3986605405807495,
+      "learning_rate": 7.781599558578394e-06,
+      "loss": 1.0615,
+      "step": 2540
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4630681276321411,
+      "learning_rate": 7.779835890369555e-06,
+      "loss": 0.9138,
+      "step": 2541
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4934000968933105,
+      "learning_rate": 7.778071721410973e-06,
+      "loss": 0.8254,
+      "step": 2542
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4273444414138794,
+      "learning_rate": 7.776307052020442e-06,
+      "loss": 0.8464,
+      "step": 2543
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4138139486312866,
+      "learning_rate": 7.774541882515839e-06,
+      "loss": 0.8668,
+      "step": 2544
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 2.070120096206665,
+      "learning_rate": 7.772776213215136e-06,
+      "loss": 0.8673,
+      "step": 2545
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4061793088912964,
+      "learning_rate": 7.771010044436393e-06,
+      "loss": 0.894,
+      "step": 2546
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.415271520614624,
+      "learning_rate": 7.769243376497766e-06,
+      "loss": 0.8631,
+      "step": 2547
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4617646932601929,
+      "learning_rate": 7.767476209717495e-06,
+      "loss": 0.8795,
+      "step": 2548
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7926614284515381,
+      "learning_rate": 7.765708544413907e-06,
+      "loss": 1.0624,
+      "step": 2549
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4563400745391846,
+      "learning_rate": 7.763940380905427e-06,
+      "loss": 0.8401,
+      "step": 2550
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.4023265838623047,
+      "learning_rate": 7.762171719510564e-06,
+      "loss": 0.8438,
+      "step": 2551
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.269331932067871,
+      "learning_rate": 7.760402560547919e-06,
+      "loss": 0.8125,
+      "step": 2552
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3498727083206177,
+      "learning_rate": 7.758632904336183e-06,
+      "loss": 0.8394,
+      "step": 2553
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.577934980392456,
+      "learning_rate": 7.756862751194132e-06,
+      "loss": 0.8438,
+      "step": 2554
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7184392809867859,
+      "learning_rate": 7.755092101440638e-06,
+      "loss": 1.0394,
+      "step": 2555
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 1.3743963241577148,
+      "learning_rate": 7.75332095539466e-06,
+      "loss": 0.8708,
+      "step": 2556
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2932240962982178,
+      "learning_rate": 7.751549313375246e-06,
+      "loss": 0.8467,
+      "step": 2557
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4450010061264038,
+      "learning_rate": 7.74977717570153e-06,
+      "loss": 0.8954,
+      "step": 2558
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6921409964561462,
+      "learning_rate": 7.748004542692742e-06,
+      "loss": 1.0367,
+      "step": 2559
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4179620742797852,
+      "learning_rate": 7.746231414668195e-06,
+      "loss": 0.8875,
+      "step": 2560
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3014779090881348,
+      "learning_rate": 7.744457791947297e-06,
+      "loss": 0.9091,
+      "step": 2561
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.729714572429657,
+      "learning_rate": 7.742683674849541e-06,
+      "loss": 1.0438,
+      "step": 2562
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4273179769515991,
+      "learning_rate": 7.74090906369451e-06,
+      "loss": 0.8536,
+      "step": 2563
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6835248470306396,
+      "learning_rate": 7.739133958801874e-06,
+      "loss": 1.0451,
+      "step": 2564
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3288394212722778,
+      "learning_rate": 7.737358360491396e-06,
+      "loss": 0.8943,
+      "step": 2565
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.308853030204773,
+      "learning_rate": 7.735582269082928e-06,
+      "loss": 0.8548,
+      "step": 2566
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5454413890838623,
+      "learning_rate": 7.733805684896407e-06,
+      "loss": 0.8154,
+      "step": 2567
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7432546019554138,
+      "learning_rate": 7.732028608251859e-06,
+      "loss": 1.0364,
+      "step": 2568
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.269213318824768,
+      "learning_rate": 7.730251039469401e-06,
+      "loss": 0.8035,
+      "step": 2569
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3544975519180298,
+      "learning_rate": 7.728472978869239e-06,
+      "loss": 0.8928,
+      "step": 2570
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2708725929260254,
+      "learning_rate": 7.726694426771667e-06,
+      "loss": 0.8649,
+      "step": 2571
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5775200128555298,
+      "learning_rate": 7.724915383497064e-06,
+      "loss": 0.8463,
+      "step": 2572
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4566773176193237,
+      "learning_rate": 7.723135849365906e-06,
+      "loss": 0.8491,
+      "step": 2573
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3554047346115112,
+      "learning_rate": 7.721355824698746e-06,
+      "loss": 0.8327,
+      "step": 2574
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4650299549102783,
+      "learning_rate": 7.719575309816232e-06,
+      "loss": 0.8858,
+      "step": 2575
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3480695486068726,
+      "learning_rate": 7.717794305039105e-06,
+      "loss": 0.9237,
+      "step": 2576
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3538190126419067,
+      "learning_rate": 7.716012810688183e-06,
+      "loss": 0.8477,
+      "step": 2577
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3943167924880981,
+      "learning_rate": 7.714230827084383e-06,
+      "loss": 0.8889,
+      "step": 2578
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4195367097854614,
+      "learning_rate": 7.7124483545487e-06,
+      "loss": 0.898,
+      "step": 2579
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3809503316879272,
+      "learning_rate": 7.710665393402226e-06,
+      "loss": 0.8774,
+      "step": 2580
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8960800170898438,
+      "learning_rate": 7.708881943966135e-06,
+      "loss": 1.0652,
+      "step": 2581
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3422744274139404,
+      "learning_rate": 7.707098006561692e-06,
+      "loss": 0.8708,
+      "step": 2582
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3768349885940552,
+      "learning_rate": 7.70531358151025e-06,
+      "loss": 0.8247,
+      "step": 2583
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3594311475753784,
+      "learning_rate": 7.703528669133248e-06,
+      "loss": 0.8369,
+      "step": 2584
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3158868551254272,
+      "learning_rate": 7.701743269752217e-06,
+      "loss": 0.8544,
+      "step": 2585
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2710018157958984,
+      "learning_rate": 7.699957383688767e-06,
+      "loss": 0.8428,
+      "step": 2586
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4012783765792847,
+      "learning_rate": 7.698171011264604e-06,
+      "loss": 0.798,
+      "step": 2587
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5327163934707642,
+      "learning_rate": 7.696384152801519e-06,
+      "loss": 0.8862,
+      "step": 2588
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5342034101486206,
+      "learning_rate": 7.694596808621389e-06,
+      "loss": 0.8779,
+      "step": 2589
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9401656985282898,
+      "learning_rate": 7.692808979046182e-06,
+      "loss": 1.0439,
+      "step": 2590
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.9817523956298828,
+      "learning_rate": 7.69102066439795e-06,
+      "loss": 0.8388,
+      "step": 2591
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4086576700210571,
+      "learning_rate": 7.689231864998832e-06,
+      "loss": 0.8755,
+      "step": 2592
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6250082850456238,
+      "learning_rate": 7.687442581171059e-06,
+      "loss": 1.0362,
+      "step": 2593
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6397867202758789,
+      "learning_rate": 7.685652813236946e-06,
+      "loss": 1.0354,
+      "step": 2594
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4222298860549927,
+      "learning_rate": 7.683862561518892e-06,
+      "loss": 0.8933,
+      "step": 2595
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7133262753486633,
+      "learning_rate": 7.68207182633939e-06,
+      "loss": 1.0707,
+      "step": 2596
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6727280616760254,
+      "learning_rate": 7.680280608021014e-06,
+      "loss": 1.0405,
+      "step": 2597
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4373284578323364,
+      "learning_rate": 7.67848890688643e-06,
+      "loss": 0.8395,
+      "step": 2598
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3010265827178955,
+      "learning_rate": 7.676696723258387e-06,
+      "loss": 0.8222,
+      "step": 2599
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.447676181793213,
+      "learning_rate": 7.674904057459726e-06,
+      "loss": 0.9172,
+      "step": 2600
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3810360431671143,
+      "learning_rate": 7.673110909813366e-06,
+      "loss": 0.8454,
+      "step": 2601
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9866548776626587,
+      "learning_rate": 7.671317280642323e-06,
+      "loss": 1.042,
+      "step": 2602
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2790600061416626,
+      "learning_rate": 7.66952317026969e-06,
+      "loss": 0.8668,
+      "step": 2603
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.293752908706665,
+      "learning_rate": 7.66772857901866e-06,
+      "loss": 0.7954,
+      "step": 2604
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3384727239608765,
+      "learning_rate": 7.665933507212495e-06,
+      "loss": 0.8141,
+      "step": 2605
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3754874467849731,
+      "learning_rate": 7.664137955174558e-06,
+      "loss": 0.8767,
+      "step": 2606
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.5661600828170776,
+      "learning_rate": 7.66234192322829e-06,
+      "loss": 0.8385,
+      "step": 2607
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4407860040664673,
+      "learning_rate": 7.660545411697227e-06,
+      "loss": 0.8576,
+      "step": 2608
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3400001525878906,
+      "learning_rate": 7.658748420904983e-06,
+      "loss": 0.9283,
+      "step": 2609
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3061017990112305,
+      "learning_rate": 7.65695095117526e-06,
+      "loss": 0.8493,
+      "step": 2610
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2518014907836914,
+      "learning_rate": 7.655153002831849e-06,
+      "loss": 0.8785,
+      "step": 2611
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2850465774536133,
+      "learning_rate": 7.653354576198629e-06,
+      "loss": 0.8819,
+      "step": 2612
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.4009685516357422,
+      "learning_rate": 7.651555671599557e-06,
+      "loss": 0.8697,
+      "step": 2613
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3776490688323975,
+      "learning_rate": 7.649756289358686e-06,
+      "loss": 0.8804,
+      "step": 2614
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.35089910030365,
+      "learning_rate": 7.647956429800147e-06,
+      "loss": 0.9017,
+      "step": 2615
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3477543592453003,
+      "learning_rate": 7.646156093248162e-06,
+      "loss": 0.8527,
+      "step": 2616
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3532179594039917,
+      "learning_rate": 7.644355280027037e-06,
+      "loss": 0.8531,
+      "step": 2617
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3406792879104614,
+      "learning_rate": 7.642553990461165e-06,
+      "loss": 0.823,
+      "step": 2618
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3695142269134521,
+      "learning_rate": 7.640752224875022e-06,
+      "loss": 0.9363,
+      "step": 2619
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3072868585586548,
+      "learning_rate": 7.63894998359317e-06,
+      "loss": 0.9043,
+      "step": 2620
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.0123707056045532,
+      "learning_rate": 7.637147266940264e-06,
+      "loss": 1.0464,
+      "step": 2621
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3528279066085815,
+      "learning_rate": 7.635344075241034e-06,
+      "loss": 0.8849,
+      "step": 2622
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.238257884979248,
+      "learning_rate": 7.633540408820301e-06,
+      "loss": 0.8178,
+      "step": 2623
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2728708982467651,
+      "learning_rate": 7.631736268002975e-06,
+      "loss": 0.8788,
+      "step": 2624
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.8042746186256409,
+      "learning_rate": 7.629931653114044e-06,
+      "loss": 1.0644,
+      "step": 2625
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2526775598526,
+      "learning_rate": 7.628126564478584e-06,
+      "loss": 0.8229,
+      "step": 2626
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3482106924057007,
+      "learning_rate": 7.626321002421761e-06,
+      "loss": 0.8786,
+      "step": 2627
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.306247353553772,
+      "learning_rate": 7.624514967268819e-06,
+      "loss": 0.8946,
+      "step": 2628
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2784656286239624,
+      "learning_rate": 7.622708459345093e-06,
+      "loss": 0.7991,
+      "step": 2629
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.3026020526885986,
+      "learning_rate": 7.6209014789759995e-06,
+      "loss": 0.8161,
+      "step": 2630
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.386157751083374,
+      "learning_rate": 7.619094026487042e-06,
+      "loss": 0.8216,
+      "step": 2631
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.7146445512771606,
+      "learning_rate": 7.617286102203809e-06,
+      "loss": 0.8251,
+      "step": 2632
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.309522271156311,
+      "learning_rate": 7.615477706451975e-06,
+      "loss": 0.8683,
+      "step": 2633
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7100332975387573,
+      "learning_rate": 7.613668839557294e-06,
+      "loss": 1.025,
+      "step": 2634
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4721834659576416,
+      "learning_rate": 7.611859501845611e-06,
+      "loss": 0.8943,
+      "step": 2635
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.303832769393921,
+      "learning_rate": 7.6100496936428545e-06,
+      "loss": 0.8788,
+      "step": 2636
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3066720962524414,
+      "learning_rate": 7.6082394152750365e-06,
+      "loss": 0.925,
+      "step": 2637
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3687832355499268,
+      "learning_rate": 7.606428667068253e-06,
+      "loss": 0.862,
+      "step": 2638
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.374218463897705,
+      "learning_rate": 7.6046174493486854e-06,
+      "loss": 0.8515,
+      "step": 2639
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.479307770729065,
+      "learning_rate": 7.6028057624426025e-06,
+      "loss": 0.8192,
+      "step": 2640
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.330735683441162,
+      "learning_rate": 7.600993606676355e-06,
+      "loss": 0.8565,
+      "step": 2641
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3329397439956665,
+      "learning_rate": 7.599180982376373e-06,
+      "loss": 0.8063,
+      "step": 2642
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7208575010299683,
+      "learning_rate": 7.597367889869183e-06,
+      "loss": 1.0446,
+      "step": 2643
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2814151048660278,
+      "learning_rate": 7.595554329481385e-06,
+      "loss": 0.7819,
+      "step": 2644
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3018085956573486,
+      "learning_rate": 7.593740301539667e-06,
+      "loss": 0.8911,
+      "step": 2645
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3742337226867676,
+      "learning_rate": 7.5919258063708055e-06,
+      "loss": 0.8452,
+      "step": 2646
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2883790731430054,
+      "learning_rate": 7.590110844301655e-06,
+      "loss": 0.915,
+      "step": 2647
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7362564206123352,
+      "learning_rate": 7.588295415659154e-06,
+      "loss": 1.0335,
+      "step": 2648
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2879751920700073,
+      "learning_rate": 7.586479520770329e-06,
+      "loss": 0.8307,
+      "step": 2649
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.430144190788269,
+      "learning_rate": 7.58466315996229e-06,
+      "loss": 0.8793,
+      "step": 2650
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3625682592391968,
+      "learning_rate": 7.582846333562231e-06,
+      "loss": 0.8862,
+      "step": 2651
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 2.025439739227295,
+      "learning_rate": 7.581029041897423e-06,
+      "loss": 0.8283,
+      "step": 2652
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.6079237461090088,
+      "learning_rate": 7.579211285295231e-06,
+      "loss": 0.8199,
+      "step": 2653
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8230974674224854,
+      "learning_rate": 7.577393064083101e-06,
+      "loss": 1.0234,
+      "step": 2654
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6980593204498291,
+      "learning_rate": 7.575574378588555e-06,
+      "loss": 1.0368,
+      "step": 2655
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4290401935577393,
+      "learning_rate": 7.573755229139208e-06,
+      "loss": 0.8641,
+      "step": 2656
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3582427501678467,
+      "learning_rate": 7.571935616062756e-06,
+      "loss": 0.9198,
+      "step": 2657
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.716372013092041,
+      "learning_rate": 7.570115539686975e-06,
+      "loss": 1.072,
+      "step": 2658
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2504347562789917,
+      "learning_rate": 7.568295000339729e-06,
+      "loss": 0.875,
+      "step": 2659
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3007079362869263,
+      "learning_rate": 7.566473998348963e-06,
+      "loss": 0.8606,
+      "step": 2660
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7114502191543579,
+      "learning_rate": 7.5646525340427054e-06,
+      "loss": 1.0592,
+      "step": 2661
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2763556241989136,
+      "learning_rate": 7.56283060774907e-06,
+      "loss": 0.846,
+      "step": 2662
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6550308465957642,
+      "learning_rate": 7.561008219796249e-06,
+      "loss": 1.0458,
+      "step": 2663
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2474700212478638,
+      "learning_rate": 7.559185370512523e-06,
+      "loss": 0.8547,
+      "step": 2664
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.45805025100708,
+      "learning_rate": 7.557362060226254e-06,
+      "loss": 0.8929,
+      "step": 2665
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3269504308700562,
+      "learning_rate": 7.555538289265885e-06,
+      "loss": 0.8811,
+      "step": 2666
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7343219518661499,
+      "learning_rate": 7.553714057959944e-06,
+      "loss": 1.0583,
+      "step": 2667
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2863682508468628,
+      "learning_rate": 7.551889366637043e-06,
+      "loss": 0.8499,
+      "step": 2668
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2767258882522583,
+      "learning_rate": 7.550064215625874e-06,
+      "loss": 0.9046,
+      "step": 2669
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2958149909973145,
+      "learning_rate": 7.548238605255215e-06,
+      "loss": 0.8906,
+      "step": 2670
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.316394329071045,
+      "learning_rate": 7.5464125358539216e-06,
+      "loss": 0.8665,
+      "step": 2671
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5174118280410767,
+      "learning_rate": 7.544586007750938e-06,
+      "loss": 0.8525,
+      "step": 2672
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.392342448234558,
+      "learning_rate": 7.542759021275289e-06,
+      "loss": 0.8609,
+      "step": 2673
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4050655364990234,
+      "learning_rate": 7.540931576756079e-06,
+      "loss": 0.889,
+      "step": 2674
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.372880220413208,
+      "learning_rate": 7.539103674522499e-06,
+      "loss": 0.8974,
+      "step": 2675
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7680037617683411,
+      "learning_rate": 7.53727531490382e-06,
+      "loss": 1.0363,
+      "step": 2676
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6916270852088928,
+      "learning_rate": 7.5354464982293975e-06,
+      "loss": 1.0564,
+      "step": 2677
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.5370872020721436,
+      "learning_rate": 7.533617224828667e-06,
+      "loss": 0.8806,
+      "step": 2678
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4327971935272217,
+      "learning_rate": 7.531787495031148e-06,
+      "loss": 0.8544,
+      "step": 2679
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3589788675308228,
+      "learning_rate": 7.52995730916644e-06,
+      "loss": 0.8896,
+      "step": 2680
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3501620292663574,
+      "learning_rate": 7.528126667564226e-06,
+      "loss": 0.8658,
+      "step": 2681
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4265851974487305,
+      "learning_rate": 7.526295570554274e-06,
+      "loss": 0.8261,
+      "step": 2682
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9273315072059631,
+      "learning_rate": 7.524464018466428e-06,
+      "loss": 1.0696,
+      "step": 2683
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.382622241973877,
+      "learning_rate": 7.522632011630622e-06,
+      "loss": 0.8393,
+      "step": 2684
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7250391840934753,
+      "learning_rate": 7.520799550376863e-06,
+      "loss": 1.0224,
+      "step": 2685
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.375396728515625,
+      "learning_rate": 7.518966635035246e-06,
+      "loss": 0.8389,
+      "step": 2686
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.354806661605835,
+      "learning_rate": 7.517133265935943e-06,
+      "loss": 0.8485,
+      "step": 2687
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.2948566675186157,
+      "learning_rate": 7.515299443409214e-06,
+      "loss": 0.803,
+      "step": 2688
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3654017448425293,
+      "learning_rate": 7.513465167785394e-06,
+      "loss": 0.8997,
+      "step": 2689
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.0329875946044922,
+      "learning_rate": 7.511630439394908e-06,
+      "loss": 1.0516,
+      "step": 2690
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3983458280563354,
+      "learning_rate": 7.509795258568251e-06,
+      "loss": 0.8861,
+      "step": 2691
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7460485696792603,
+      "learning_rate": 7.507959625636012e-06,
+      "loss": 1.0535,
+      "step": 2692
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.446847915649414,
+      "learning_rate": 7.5061235409288515e-06,
+      "loss": 0.8821,
+      "step": 2693
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4474363327026367,
+      "learning_rate": 7.504287004777517e-06,
+      "loss": 0.9023,
+      "step": 2694
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.8918709754943848,
+      "learning_rate": 7.5024500175128336e-06,
+      "loss": 1.0617,
+      "step": 2695
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3235818147659302,
+      "learning_rate": 7.500612579465711e-06,
+      "loss": 0.8357,
+      "step": 2696
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.4496994018554688,
+      "learning_rate": 7.498774690967139e-06,
+      "loss": 0.8955,
+      "step": 2697
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 3.493685245513916,
+      "learning_rate": 7.49693635234819e-06,
+      "loss": 0.8244,
+      "step": 2698
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3069632053375244,
+      "learning_rate": 7.49509756394001e-06,
+      "loss": 0.8802,
+      "step": 2699
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.7680505514144897,
+      "learning_rate": 7.493258326073838e-06,
+      "loss": 1.0409,
+      "step": 2700
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6761111617088318,
+      "learning_rate": 7.491418639080985e-06,
+      "loss": 1.0643,
+      "step": 2701
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.392865538597107,
+      "learning_rate": 7.489578503292847e-06,
+      "loss": 0.873,
+      "step": 2702
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.430578351020813,
+      "learning_rate": 7.487737919040897e-06,
+      "loss": 0.8964,
+      "step": 2703
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3908814191818237,
+      "learning_rate": 7.4858968866566925e-06,
+      "loss": 0.8724,
+      "step": 2704
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3965988159179688,
+      "learning_rate": 7.484055406471872e-06,
+      "loss": 0.9006,
+      "step": 2705
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3792040348052979,
+      "learning_rate": 7.482213478818151e-06,
+      "loss": 0.8578,
+      "step": 2706
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.3608440160751343,
+      "learning_rate": 7.480371104027331e-06,
+      "loss": 0.8328,
+      "step": 2707
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9528942108154297,
+      "learning_rate": 7.478528282431287e-06,
+      "loss": 1.0544,
+      "step": 2708
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.353542447090149,
+      "learning_rate": 7.4766850143619795e-06,
+      "loss": 0.8597,
+      "step": 2709
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.695218801498413,
+      "learning_rate": 7.47484130015145e-06,
+      "loss": 0.8294,
+      "step": 2710
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5607125759124756,
+      "learning_rate": 7.472997140131816e-06,
+      "loss": 0.8292,
+      "step": 2711
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3883063793182373,
+      "learning_rate": 7.471152534635281e-06,
+      "loss": 0.8038,
+      "step": 2712
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5132087469100952,
+      "learning_rate": 7.469307483994122e-06,
+      "loss": 0.9035,
+      "step": 2713
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2980084419250488,
+      "learning_rate": 7.467461988540704e-06,
+      "loss": 0.8715,
+      "step": 2714
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7726878523826599,
+      "learning_rate": 7.465616048607464e-06,
+      "loss": 1.0428,
+      "step": 2715
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3518482446670532,
+      "learning_rate": 7.463769664526926e-06,
+      "loss": 0.8998,
+      "step": 2716
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6591354012489319,
+      "learning_rate": 7.461922836631688e-06,
+      "loss": 1.0596,
+      "step": 2717
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4403034448623657,
+      "learning_rate": 7.4600755652544344e-06,
+      "loss": 0.8445,
+      "step": 2718
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.37944757938385,
+      "learning_rate": 7.458227850727925e-06,
+      "loss": 0.896,
+      "step": 2719
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3255246877670288,
+      "learning_rate": 7.456379693384999e-06,
+      "loss": 0.869,
+      "step": 2720
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7270732522010803,
+      "learning_rate": 7.454531093558579e-06,
+      "loss": 1.0594,
+      "step": 2721
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.35533607006073,
+      "learning_rate": 7.4526820515816635e-06,
+      "loss": 0.8434,
+      "step": 2722
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3315578699111938,
+      "learning_rate": 7.450832567787333e-06,
+      "loss": 0.8183,
+      "step": 2723
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4529958963394165,
+      "learning_rate": 7.448982642508746e-06,
+      "loss": 0.8547,
+      "step": 2724
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6418157815933228,
+      "learning_rate": 7.447132276079143e-06,
+      "loss": 1.0326,
+      "step": 2725
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4115114212036133,
+      "learning_rate": 7.445281468831843e-06,
+      "loss": 0.9081,
+      "step": 2726
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3697843551635742,
+      "learning_rate": 7.443430221100242e-06,
+      "loss": 0.7934,
+      "step": 2727
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6389558911323547,
+      "learning_rate": 7.441578533217817e-06,
+      "loss": 1.0334,
+      "step": 2728
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3788126707077026,
+      "learning_rate": 7.439726405518126e-06,
+      "loss": 0.8575,
+      "step": 2729
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6302975416183472,
+      "learning_rate": 7.437873838334805e-06,
+      "loss": 1.0504,
+      "step": 2730
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.8755111694335938,
+      "learning_rate": 7.436020832001567e-06,
+      "loss": 0.8444,
+      "step": 2731
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2926311492919922,
+      "learning_rate": 7.434167386852206e-06,
+      "loss": 0.8825,
+      "step": 2732
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6445984840393066,
+      "learning_rate": 7.432313503220597e-06,
+      "loss": 0.8414,
+      "step": 2733
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6824275851249695,
+      "learning_rate": 7.43045918144069e-06,
+      "loss": 1.0168,
+      "step": 2734
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3683998584747314,
+      "learning_rate": 7.428604421846517e-06,
+      "loss": 0.8806,
+      "step": 2735
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6775426864624023,
+      "learning_rate": 7.426749224772188e-06,
+      "loss": 1.0369,
+      "step": 2736
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4259032011032104,
+      "learning_rate": 7.424893590551891e-06,
+      "loss": 0.8145,
+      "step": 2737
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2641359567642212,
+      "learning_rate": 7.423037519519892e-06,
+      "loss": 0.853,
+      "step": 2738
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2669713497161865,
+      "learning_rate": 7.42118101201054e-06,
+      "loss": 0.8838,
+      "step": 2739
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3224544525146484,
+      "learning_rate": 7.4193240683582576e-06,
+      "loss": 0.7707,
+      "step": 2740
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7085390686988831,
+      "learning_rate": 7.417466688897549e-06,
+      "loss": 1.0342,
+      "step": 2741
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6368933320045471,
+      "learning_rate": 7.415608873962995e-06,
+      "loss": 1.025,
+      "step": 2742
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6317013502120972,
+      "learning_rate": 7.413750623889257e-06,
+      "loss": 1.0561,
+      "step": 2743
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4235260486602783,
+      "learning_rate": 7.411891939011073e-06,
+      "loss": 0.8911,
+      "step": 2744
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7227402925491333,
+      "learning_rate": 7.410032819663259e-06,
+      "loss": 0.8407,
+      "step": 2745
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9678303003311157,
+      "learning_rate": 7.40817326618071e-06,
+      "loss": 0.8748,
+      "step": 2746
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8445633053779602,
+      "learning_rate": 7.406313278898402e-06,
+      "loss": 1.0376,
+      "step": 2747
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.326100468635559,
+      "learning_rate": 7.404452858151383e-06,
+      "loss": 0.8349,
+      "step": 2748
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6295932531356812,
+      "learning_rate": 7.402592004274785e-06,
+      "loss": 0.8587,
+      "step": 2749
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3698469400405884,
+      "learning_rate": 7.4007307176038156e-06,
+      "loss": 0.8909,
+      "step": 2750
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6812220215797424,
+      "learning_rate": 7.398868998473758e-06,
+      "loss": 1.0298,
+      "step": 2751
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.697009265422821,
+      "learning_rate": 7.397006847219978e-06,
+      "loss": 1.0662,
+      "step": 2752
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6571000814437866,
+      "learning_rate": 7.395144264177918e-06,
+      "loss": 1.0559,
+      "step": 2753
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4516921043395996,
+      "learning_rate": 7.393281249683093e-06,
+      "loss": 0.8961,
+      "step": 2754
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3776955604553223,
+      "learning_rate": 7.391417804071104e-06,
+      "loss": 0.9155,
+      "step": 2755
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3492079973220825,
+      "learning_rate": 7.389553927677622e-06,
+      "loss": 0.8511,
+      "step": 2756
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4248158931732178,
+      "learning_rate": 7.3876896208384e-06,
+      "loss": 0.8155,
+      "step": 2757
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.343787431716919,
+      "learning_rate": 7.38582488388927e-06,
+      "loss": 0.813,
+      "step": 2758
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3672984838485718,
+      "learning_rate": 7.383959717166136e-06,
+      "loss": 0.8472,
+      "step": 2759
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.6396207809448242,
+      "learning_rate": 7.382094121004983e-06,
+      "loss": 0.8815,
+      "step": 2760
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3950896263122559,
+      "learning_rate": 7.380228095741875e-06,
+      "loss": 0.8534,
+      "step": 2761
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3495616912841797,
+      "learning_rate": 7.378361641712948e-06,
+      "loss": 0.8578,
+      "step": 2762
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3874928951263428,
+      "learning_rate": 7.376494759254422e-06,
+      "loss": 0.8208,
+      "step": 2763
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.383642554283142,
+      "learning_rate": 7.3746274487025885e-06,
+      "loss": 0.8627,
+      "step": 2764
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.763833999633789,
+      "learning_rate": 7.3727597103938175e-06,
+      "loss": 0.872,
+      "step": 2765
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.305888056755066,
+      "learning_rate": 7.370891544664558e-06,
+      "loss": 0.8369,
+      "step": 2766
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3667426109313965,
+      "learning_rate": 7.369022951851333e-06,
+      "loss": 0.8469,
+      "step": 2767
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3130909204483032,
+      "learning_rate": 7.367153932290748e-06,
+      "loss": 0.81,
+      "step": 2768
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.4133235216140747,
+      "learning_rate": 7.365284486319477e-06,
+      "loss": 0.966,
+      "step": 2769
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8673095107078552,
+      "learning_rate": 7.363414614274278e-06,
+      "loss": 1.0312,
+      "step": 2770
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.362581729888916,
+      "learning_rate": 7.361544316491982e-06,
+      "loss": 0.8304,
+      "step": 2771
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6979987025260925,
+      "learning_rate": 7.3596735933095e-06,
+      "loss": 1.0511,
+      "step": 2772
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3687434196472168,
+      "learning_rate": 7.357802445063814e-06,
+      "loss": 0.8163,
+      "step": 2773
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.9125434160232544,
+      "learning_rate": 7.35593087209199e-06,
+      "loss": 0.8067,
+      "step": 2774
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3310447931289673,
+      "learning_rate": 7.354058874731161e-06,
+      "loss": 0.8299,
+      "step": 2775
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3089749813079834,
+      "learning_rate": 7.352186453318546e-06,
+      "loss": 0.8412,
+      "step": 2776
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7536232471466064,
+      "learning_rate": 7.350313608191436e-06,
+      "loss": 1.055,
+      "step": 2777
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3357516527175903,
+      "learning_rate": 7.3484403396871975e-06,
+      "loss": 0.8702,
+      "step": 2778
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3849122524261475,
+      "learning_rate": 7.346566648143275e-06,
+      "loss": 0.8787,
+      "step": 2779
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3967736959457397,
+      "learning_rate": 7.344692533897188e-06,
+      "loss": 0.8381,
+      "step": 2780
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3468022346496582,
+      "learning_rate": 7.3428179972865345e-06,
+      "loss": 0.8497,
+      "step": 2781
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.3095555305480957,
+      "learning_rate": 7.340943038648986e-06,
+      "loss": 0.8479,
+      "step": 2782
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.2956345081329346,
+      "learning_rate": 7.33906765832229e-06,
+      "loss": 0.8598,
+      "step": 2783
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.328594446182251,
+      "learning_rate": 7.33719185664427e-06,
+      "loss": 0.8672,
+      "step": 2784
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8649559020996094,
+      "learning_rate": 7.33531563395283e-06,
+      "loss": 1.0544,
+      "step": 2785
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3298004865646362,
+      "learning_rate": 7.333438990585942e-06,
+      "loss": 0.8603,
+      "step": 2786
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.501159906387329,
+      "learning_rate": 7.331561926881661e-06,
+      "loss": 0.8377,
+      "step": 2787
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.307759165763855,
+      "learning_rate": 7.329684443178112e-06,
+      "loss": 0.87,
+      "step": 2788
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4254789352416992,
+      "learning_rate": 7.3278065398135e-06,
+      "loss": 0.8832,
+      "step": 2789
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4418158531188965,
+      "learning_rate": 7.3259282171261035e-06,
+      "loss": 0.7641,
+      "step": 2790
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7772379517555237,
+      "learning_rate": 7.324049475454277e-06,
+      "loss": 1.0465,
+      "step": 2791
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3320952653884888,
+      "learning_rate": 7.3221703151364485e-06,
+      "loss": 0.8544,
+      "step": 2792
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4189234972000122,
+      "learning_rate": 7.320290736511125e-06,
+      "loss": 0.8081,
+      "step": 2793
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.333329200744629,
+      "learning_rate": 7.318410739916888e-06,
+      "loss": 0.8155,
+      "step": 2794
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4019490480422974,
+      "learning_rate": 7.316530325692392e-06,
+      "loss": 0.9057,
+      "step": 2795
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.33206045627594,
+      "learning_rate": 7.314649494176368e-06,
+      "loss": 0.8628,
+      "step": 2796
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7263073325157166,
+      "learning_rate": 7.312768245707622e-06,
+      "loss": 1.039,
+      "step": 2797
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4380720853805542,
+      "learning_rate": 7.310886580625036e-06,
+      "loss": 0.8373,
+      "step": 2798
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.417195439338684,
+      "learning_rate": 7.309004499267567e-06,
+      "loss": 0.8697,
+      "step": 2799
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4434254169464111,
+      "learning_rate": 7.307122001974245e-06,
+      "loss": 0.7776,
+      "step": 2800
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.34926438331604,
+      "learning_rate": 7.305239089084179e-06,
+      "loss": 0.8531,
+      "step": 2801
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6816434264183044,
+      "learning_rate": 7.303355760936547e-06,
+      "loss": 1.0696,
+      "step": 2802
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6304277181625366,
+      "learning_rate": 7.301472017870607e-06,
+      "loss": 1.0409,
+      "step": 2803
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5791465044021606,
+      "learning_rate": 7.2995878602256885e-06,
+      "loss": 0.8023,
+      "step": 2804
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3728855848312378,
+      "learning_rate": 7.297703288341199e-06,
+      "loss": 0.7828,
+      "step": 2805
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6820299029350281,
+      "learning_rate": 7.295818302556617e-06,
+      "loss": 1.0624,
+      "step": 2806
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3523592948913574,
+      "learning_rate": 7.293932903211497e-06,
+      "loss": 0.8117,
+      "step": 2807
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3359668254852295,
+      "learning_rate": 7.292047090645469e-06,
+      "loss": 0.8503,
+      "step": 2808
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3351532220840454,
+      "learning_rate": 7.290160865198234e-06,
+      "loss": 0.8119,
+      "step": 2809
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.659393846988678,
+      "learning_rate": 7.288274227209575e-06,
+      "loss": 1.0377,
+      "step": 2810
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6453802585601807,
+      "learning_rate": 7.286387177019341e-06,
+      "loss": 1.0586,
+      "step": 2811
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2365527153015137,
+      "learning_rate": 7.284499714967458e-06,
+      "loss": 0.895,
+      "step": 2812
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4041041135787964,
+      "learning_rate": 7.2826118413939275e-06,
+      "loss": 0.8522,
+      "step": 2813
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5146857500076294,
+      "learning_rate": 7.2807235566388245e-06,
+      "loss": 0.896,
+      "step": 2814
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7044409513473511,
+      "learning_rate": 7.278834861042299e-06,
+      "loss": 1.0483,
+      "step": 2815
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2575976848602295,
+      "learning_rate": 7.276945754944571e-06,
+      "loss": 0.8574,
+      "step": 2816
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4431843757629395,
+      "learning_rate": 7.275056238685939e-06,
+      "loss": 0.8893,
+      "step": 2817
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.499857783317566,
+      "learning_rate": 7.2731663126067745e-06,
+      "loss": 0.8383,
+      "step": 2818
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3650764226913452,
+      "learning_rate": 7.271275977047521e-06,
+      "loss": 0.8539,
+      "step": 2819
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.2734078168869019,
+      "learning_rate": 7.269385232348697e-06,
+      "loss": 0.8635,
+      "step": 2820
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3388481140136719,
+      "learning_rate": 7.267494078850894e-06,
+      "loss": 0.8982,
+      "step": 2821
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3006469011306763,
+      "learning_rate": 7.265602516894778e-06,
+      "loss": 0.8687,
+      "step": 2822
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.364147663116455,
+      "learning_rate": 7.26371054682109e-06,
+      "loss": 0.8218,
+      "step": 2823
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6657305955886841,
+      "learning_rate": 7.261818168970642e-06,
+      "loss": 1.0385,
+      "step": 2824
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6348033547401428,
+      "learning_rate": 7.259925383684318e-06,
+      "loss": 1.0296,
+      "step": 2825
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3742433786392212,
+      "learning_rate": 7.25803219130308e-06,
+      "loss": 0.8377,
+      "step": 2826
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6385548114776611,
+      "learning_rate": 7.256138592167958e-06,
+      "loss": 1.0313,
+      "step": 2827
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3743634223937988,
+      "learning_rate": 7.254244586620061e-06,
+      "loss": 0.793,
+      "step": 2828
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4888485670089722,
+      "learning_rate": 7.252350175000569e-06,
+      "loss": 0.8608,
+      "step": 2829
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.6118617057800293,
+      "learning_rate": 7.250455357650732e-06,
+      "loss": 0.8573,
+      "step": 2830
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.346218466758728,
+      "learning_rate": 7.2485601349118775e-06,
+      "loss": 0.8383,
+      "step": 2831
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4184136390686035,
+      "learning_rate": 7.246664507125403e-06,
+      "loss": 0.8585,
+      "step": 2832
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4453375339508057,
+      "learning_rate": 7.244768474632782e-06,
+      "loss": 0.8346,
+      "step": 2833
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.424199104309082,
+      "learning_rate": 7.242872037775557e-06,
+      "loss": 0.8451,
+      "step": 2834
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3555575609207153,
+      "learning_rate": 7.240975196895346e-06,
+      "loss": 0.8143,
+      "step": 2835
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.401351809501648,
+      "learning_rate": 7.23907795233384e-06,
+      "loss": 0.8888,
+      "step": 2836
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3153897523880005,
+      "learning_rate": 7.2371803044328e-06,
+      "loss": 0.941,
+      "step": 2837
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3766552209854126,
+      "learning_rate": 7.2352822535340635e-06,
+      "loss": 0.8329,
+      "step": 2838
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8387898206710815,
+      "learning_rate": 7.233383799979537e-06,
+      "loss": 1.0367,
+      "step": 2839
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7573778033256531,
+      "learning_rate": 7.2314849441112024e-06,
+      "loss": 1.0337,
+      "step": 2840
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3407062292099,
+      "learning_rate": 7.2295856862711125e-06,
+      "loss": 0.8207,
+      "step": 2841
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3766282796859741,
+      "learning_rate": 7.2276860268013925e-06,
+      "loss": 0.8333,
+      "step": 2842
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7829431295394897,
+      "learning_rate": 7.225785966044242e-06,
+      "loss": 1.0305,
+      "step": 2843
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7840604782104492,
+      "learning_rate": 7.223885504341927e-06,
+      "loss": 0.7722,
+      "step": 2844
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7923754453659058,
+      "learning_rate": 7.2219846420367956e-06,
+      "loss": 1.0348,
+      "step": 2845
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6944705843925476,
+      "learning_rate": 7.2200833794712574e-06,
+      "loss": 1.0376,
+      "step": 2846
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6556015610694885,
+      "learning_rate": 7.218181716987803e-06,
+      "loss": 1.0615,
+      "step": 2847
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4235773086547852,
+      "learning_rate": 7.216279654928988e-06,
+      "loss": 0.8902,
+      "step": 2848
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7483575940132141,
+      "learning_rate": 7.2143771936374445e-06,
+      "loss": 1.0465,
+      "step": 2849
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.440616488456726,
+      "learning_rate": 7.212474333455875e-06,
+      "loss": 0.897,
+      "step": 2850
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3291040658950806,
+      "learning_rate": 7.210571074727053e-06,
+      "loss": 0.8794,
+      "step": 2851
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.7642383575439453,
+      "learning_rate": 7.208667417793827e-06,
+      "loss": 0.8191,
+      "step": 2852
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3302717208862305,
+      "learning_rate": 7.206763362999114e-06,
+      "loss": 0.8163,
+      "step": 2853
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7860089540481567,
+      "learning_rate": 7.204858910685902e-06,
+      "loss": 1.029,
+      "step": 2854
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.5824140310287476,
+      "learning_rate": 7.202954061197254e-06,
+      "loss": 0.8706,
+      "step": 2855
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7060274481773376,
+      "learning_rate": 7.201048814876303e-06,
+      "loss": 1.0229,
+      "step": 2856
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6640135645866394,
+      "learning_rate": 7.199143172066252e-06,
+      "loss": 1.0626,
+      "step": 2857
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3625178337097168,
+      "learning_rate": 7.197237133110377e-06,
+      "loss": 0.7891,
+      "step": 2858
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.4109846353530884,
+      "learning_rate": 7.195330698352026e-06,
+      "loss": 0.8863,
+      "step": 2859
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 1.3346434831619263,
+      "learning_rate": 7.193423868134616e-06,
+      "loss": 0.8217,
+      "step": 2860
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.8488196730613708,
+      "learning_rate": 7.191516642801639e-06,
+      "loss": 1.0623,
+      "step": 2861
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.402040719985962,
+      "learning_rate": 7.189609022696653e-06,
+      "loss": 0.8183,
+      "step": 2862
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3209065198898315,
+      "learning_rate": 7.18770100816329e-06,
+      "loss": 0.8425,
+      "step": 2863
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3935573101043701,
+      "learning_rate": 7.185792599545256e-06,
+      "loss": 0.9232,
+      "step": 2864
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7448511123657227,
+      "learning_rate": 7.183883797186322e-06,
+      "loss": 1.0578,
+      "step": 2865
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3773747682571411,
+      "learning_rate": 7.181974601430335e-06,
+      "loss": 0.8263,
+      "step": 2866
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.285701870918274,
+      "learning_rate": 7.180065012621208e-06,
+      "loss": 0.838,
+      "step": 2867
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6519359350204468,
+      "learning_rate": 7.178155031102929e-06,
+      "loss": 1.0424,
+      "step": 2868
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.266383409500122,
+      "learning_rate": 7.176244657219556e-06,
+      "loss": 0.8809,
+      "step": 2869
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3942114114761353,
+      "learning_rate": 7.174333891315216e-06,
+      "loss": 0.8501,
+      "step": 2870
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7163624167442322,
+      "learning_rate": 7.172422733734109e-06,
+      "loss": 1.028,
+      "step": 2871
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3235827684402466,
+      "learning_rate": 7.170511184820502e-06,
+      "loss": 0.8548,
+      "step": 2872
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6475024819374084,
+      "learning_rate": 7.168599244918736e-06,
+      "loss": 1.0425,
+      "step": 2873
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4485671520233154,
+      "learning_rate": 7.166686914373221e-06,
+      "loss": 0.8617,
+      "step": 2874
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3524842262268066,
+      "learning_rate": 7.164774193528436e-06,
+      "loss": 0.9134,
+      "step": 2875
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7082595825195312,
+      "learning_rate": 7.162861082728934e-06,
+      "loss": 1.0413,
+      "step": 2876
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3555501699447632,
+      "learning_rate": 7.160947582319335e-06,
+      "loss": 0.8911,
+      "step": 2877
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2604918479919434,
+      "learning_rate": 7.15903369264433e-06,
+      "loss": 0.7835,
+      "step": 2878
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3629612922668457,
+      "learning_rate": 7.157119414048679e-06,
+      "loss": 0.867,
+      "step": 2879
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6873379349708557,
+      "learning_rate": 7.155204746877214e-06,
+      "loss": 1.0411,
+      "step": 2880
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.660346508026123,
+      "learning_rate": 7.153289691474839e-06,
+      "loss": 1.0304,
+      "step": 2881
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.820460081100464,
+      "learning_rate": 7.151374248186521e-06,
+      "loss": 0.838,
+      "step": 2882
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6616886258125305,
+      "learning_rate": 7.149458417357305e-06,
+      "loss": 1.0456,
+      "step": 2883
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.422995686531067,
+      "learning_rate": 7.147542199332298e-06,
+      "loss": 0.9307,
+      "step": 2884
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3398622274398804,
+      "learning_rate": 7.1456255944566856e-06,
+      "loss": 0.9174,
+      "step": 2885
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.5448558330535889,
+      "learning_rate": 7.1437086030757144e-06,
+      "loss": 0.9036,
+      "step": 2886
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7370859384536743,
+      "learning_rate": 7.1417912255347054e-06,
+      "loss": 1.0333,
+      "step": 2887
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3291311264038086,
+      "learning_rate": 7.139873462179048e-06,
+      "loss": 0.8814,
+      "step": 2888
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3646959066390991,
+      "learning_rate": 7.137955313354202e-06,
+      "loss": 0.9087,
+      "step": 2889
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3189733028411865,
+      "learning_rate": 7.136036779405695e-06,
+      "loss": 0.8226,
+      "step": 2890
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3417359590530396,
+      "learning_rate": 7.1341178606791265e-06,
+      "loss": 0.8555,
+      "step": 2891
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.7489418983459473,
+      "learning_rate": 7.13219855752016e-06,
+      "loss": 0.7789,
+      "step": 2892
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3065528869628906,
+      "learning_rate": 7.130278870274538e-06,
+      "loss": 0.8301,
+      "step": 2893
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7133085131645203,
+      "learning_rate": 7.12835879928806e-06,
+      "loss": 1.0381,
+      "step": 2894
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6731514930725098,
+      "learning_rate": 7.126438344906603e-06,
+      "loss": 1.054,
+      "step": 2895
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2599256038665771,
+      "learning_rate": 7.124517507476113e-06,
+      "loss": 0.7967,
+      "step": 2896
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6369399428367615,
+      "learning_rate": 7.1225962873426e-06,
+      "loss": 1.032,
+      "step": 2897
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.475805640220642,
+      "learning_rate": 7.120674684852147e-06,
+      "loss": 0.9092,
+      "step": 2898
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6825120449066162,
+      "learning_rate": 7.118752700350904e-06,
+      "loss": 1.0313,
+      "step": 2899
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6676730513572693,
+      "learning_rate": 7.116830334185091e-06,
+      "loss": 1.0339,
+      "step": 2900
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.287333607673645,
+      "learning_rate": 7.1149075867009965e-06,
+      "loss": 0.8349,
+      "step": 2901
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2918767929077148,
+      "learning_rate": 7.112984458244976e-06,
+      "loss": 0.8392,
+      "step": 2902
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.430222988128662,
+      "learning_rate": 7.111060949163456e-06,
+      "loss": 0.825,
+      "step": 2903
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.8009207844734192,
+      "learning_rate": 7.109137059802931e-06,
+      "loss": 1.0322,
+      "step": 2904
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3817150592803955,
+      "learning_rate": 7.107212790509962e-06,
+      "loss": 0.8147,
+      "step": 2905
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3390651941299438,
+      "learning_rate": 7.1052881416311815e-06,
+      "loss": 0.8512,
+      "step": 2906
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3314878940582275,
+      "learning_rate": 7.103363113513288e-06,
+      "loss": 0.7868,
+      "step": 2907
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7095843553543091,
+      "learning_rate": 7.101437706503052e-06,
+      "loss": 1.0488,
+      "step": 2908
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.276295781135559,
+      "learning_rate": 7.099511920947305e-06,
+      "loss": 0.8304,
+      "step": 2909
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6807549595832825,
+      "learning_rate": 7.0975857571929545e-06,
+      "loss": 1.0412,
+      "step": 2910
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2858105897903442,
+      "learning_rate": 7.09565921558697e-06,
+      "loss": 0.8689,
+      "step": 2911
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3686367273330688,
+      "learning_rate": 7.093732296476394e-06,
+      "loss": 0.8311,
+      "step": 2912
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7068115472793579,
+      "learning_rate": 7.091805000208336e-06,
+      "loss": 1.0493,
+      "step": 2913
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2506954669952393,
+      "learning_rate": 7.089877327129969e-06,
+      "loss": 0.8342,
+      "step": 2914
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.519206166267395,
+      "learning_rate": 7.087949277588539e-06,
+      "loss": 0.8233,
+      "step": 2915
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.2508326768875122,
+      "learning_rate": 7.086020851931359e-06,
+      "loss": 0.7858,
+      "step": 2916
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.351481318473816,
+      "learning_rate": 7.084092050505807e-06,
+      "loss": 0.8282,
+      "step": 2917
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.352946162223816,
+      "learning_rate": 7.08216287365933e-06,
+      "loss": 0.8732,
+      "step": 2918
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3868889808654785,
+      "learning_rate": 7.080233321739443e-06,
+      "loss": 0.8239,
+      "step": 2919
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7724853157997131,
+      "learning_rate": 7.078303395093731e-06,
+      "loss": 1.0612,
+      "step": 2920
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3749061822891235,
+      "learning_rate": 7.076373094069842e-06,
+      "loss": 0.8464,
+      "step": 2921
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6669034361839294,
+      "learning_rate": 7.074442419015496e-06,
+      "loss": 1.0207,
+      "step": 2922
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3977878093719482,
+      "learning_rate": 7.072511370278474e-06,
+      "loss": 0.8576,
+      "step": 2923
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.411033034324646,
+      "learning_rate": 7.0705799482066305e-06,
+      "loss": 0.8546,
+      "step": 2924
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.426568627357483,
+      "learning_rate": 7.068648153147886e-06,
+      "loss": 0.8791,
+      "step": 2925
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3721860647201538,
+      "learning_rate": 7.0667159854502254e-06,
+      "loss": 0.8398,
+      "step": 2926
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.9258052706718445,
+      "learning_rate": 7.064783445461705e-06,
+      "loss": 1.0325,
+      "step": 2927
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.321649432182312,
+      "learning_rate": 7.062850533530441e-06,
+      "loss": 0.8553,
+      "step": 2928
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.384859323501587,
+      "learning_rate": 7.060917250004627e-06,
+      "loss": 0.8128,
+      "step": 2929
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4308022260665894,
+      "learning_rate": 7.058983595232512e-06,
+      "loss": 0.9134,
+      "step": 2930
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.7576515078544617,
+      "learning_rate": 7.057049569562422e-06,
+      "loss": 1.0602,
+      "step": 2931
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.3255155086517334,
+      "learning_rate": 7.0551151733427446e-06,
+      "loss": 0.7873,
+      "step": 2932
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4258767366409302,
+      "learning_rate": 7.053180406921934e-06,
+      "loss": 0.8726,
+      "step": 2933
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.203108549118042,
+      "learning_rate": 7.051245270648514e-06,
+      "loss": 0.7939,
+      "step": 2934
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.643866777420044,
+      "learning_rate": 7.049309764871071e-06,
+      "loss": 1.0292,
+      "step": 2935
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.374314785003662,
+      "learning_rate": 7.04737388993826e-06,
+      "loss": 0.8338,
+      "step": 2936
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.567413330078125,
+      "learning_rate": 7.045437646198806e-06,
+      "loss": 0.9176,
+      "step": 2937
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4062788486480713,
+      "learning_rate": 7.043501034001493e-06,
+      "loss": 0.8602,
+      "step": 2938
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6574814915657043,
+      "learning_rate": 7.041564053695178e-06,
+      "loss": 1.0388,
+      "step": 2939
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6606649160385132,
+      "learning_rate": 7.039626705628779e-06,
+      "loss": 1.0462,
+      "step": 2940
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.336669683456421,
+      "learning_rate": 7.037688990151288e-06,
+      "loss": 0.8591,
+      "step": 2941
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3303583860397339,
+      "learning_rate": 7.035750907611752e-06,
+      "loss": 0.8464,
+      "step": 2942
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6479679346084595,
+      "learning_rate": 7.033812458359295e-06,
+      "loss": 1.0513,
+      "step": 2943
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.352297067642212,
+      "learning_rate": 7.0318736427431e-06,
+      "loss": 0.8451,
+      "step": 2944
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6984426975250244,
+      "learning_rate": 7.029934461112419e-06,
+      "loss": 1.0154,
+      "step": 2945
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3184033632278442,
+      "learning_rate": 7.02799491381657e-06,
+      "loss": 0.9153,
+      "step": 2946
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5022743940353394,
+      "learning_rate": 7.026055001204935e-06,
+      "loss": 0.8858,
+      "step": 2947
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 2.029230833053589,
+      "learning_rate": 7.024114723626962e-06,
+      "loss": 0.8319,
+      "step": 2948
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.381248116493225,
+      "learning_rate": 7.0221740814321695e-06,
+      "loss": 0.8071,
+      "step": 2949
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6783046126365662,
+      "learning_rate": 7.020233074970134e-06,
+      "loss": 1.0312,
+      "step": 2950
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6776041388511658,
+      "learning_rate": 7.018291704590503e-06,
+      "loss": 1.0392,
+      "step": 2951
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3681085109710693,
+      "learning_rate": 7.0163499706429885e-06,
+      "loss": 0.7722,
+      "step": 2952
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.37394380569458,
+      "learning_rate": 7.014407873477367e-06,
+      "loss": 0.8555,
+      "step": 2953
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3130197525024414,
+      "learning_rate": 7.012465413443481e-06,
+      "loss": 0.8477,
+      "step": 2954
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7013691663742065,
+      "learning_rate": 7.010522590891238e-06,
+      "loss": 1.027,
+      "step": 2955
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6809448599815369,
+      "learning_rate": 7.008579406170612e-06,
+      "loss": 1.0356,
+      "step": 2956
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6191347241401672,
+      "learning_rate": 7.006635859631641e-06,
+      "loss": 1.0421,
+      "step": 2957
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3415883779525757,
+      "learning_rate": 7.004691951624429e-06,
+      "loss": 0.8289,
+      "step": 2958
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6875634789466858,
+      "learning_rate": 7.002747682499141e-06,
+      "loss": 1.0369,
+      "step": 2959
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3284116983413696,
+      "learning_rate": 7.000803052606018e-06,
+      "loss": 0.7984,
+      "step": 2960
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3456472158432007,
+      "learning_rate": 6.998858062295352e-06,
+      "loss": 0.7962,
+      "step": 2961
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6490214467048645,
+      "learning_rate": 6.996912711917509e-06,
+      "loss": 1.0402,
+      "step": 2962
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2926368713378906,
+      "learning_rate": 6.994967001822918e-06,
+      "loss": 0.8971,
+      "step": 2963
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.388733983039856,
+      "learning_rate": 6.9930209323620735e-06,
+      "loss": 0.8925,
+      "step": 2964
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6734509468078613,
+      "learning_rate": 6.99107450388553e-06,
+      "loss": 1.0429,
+      "step": 2965
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3752799034118652,
+      "learning_rate": 6.989127716743913e-06,
+      "loss": 0.8628,
+      "step": 2966
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3803635835647583,
+      "learning_rate": 6.987180571287908e-06,
+      "loss": 0.81,
+      "step": 2967
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4296681880950928,
+      "learning_rate": 6.985233067868269e-06,
+      "loss": 0.9276,
+      "step": 2968
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6417984366416931,
+      "learning_rate": 6.98328520683581e-06,
+      "loss": 1.0344,
+      "step": 2969
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4403409957885742,
+      "learning_rate": 6.981336988541414e-06,
+      "loss": 0.8375,
+      "step": 2970
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6408445835113525,
+      "learning_rate": 6.979388413336025e-06,
+      "loss": 1.0344,
+      "step": 2971
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4187450408935547,
+      "learning_rate": 6.977439481570651e-06,
+      "loss": 0.8074,
+      "step": 2972
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3260083198547363,
+      "learning_rate": 6.9754901935963695e-06,
+      "loss": 0.846,
+      "step": 2973
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6537084579467773,
+      "learning_rate": 6.973540549764315e-06,
+      "loss": 1.0265,
+      "step": 2974
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6467183828353882,
+      "learning_rate": 6.971590550425688e-06,
+      "loss": 1.0417,
+      "step": 2975
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3017154932022095,
+      "learning_rate": 6.96964019593176e-06,
+      "loss": 0.8208,
+      "step": 2976
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3448435068130493,
+      "learning_rate": 6.967689486633856e-06,
+      "loss": 0.8861,
+      "step": 2977
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6503695249557495,
+      "learning_rate": 6.965738422883372e-06,
+      "loss": 1.0336,
+      "step": 2978
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.281365156173706,
+      "learning_rate": 6.963787005031765e-06,
+      "loss": 0.8446,
+      "step": 2979
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3493332862854004,
+      "learning_rate": 6.961835233430556e-06,
+      "loss": 0.8011,
+      "step": 2980
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.638970136642456,
+      "learning_rate": 6.959883108431331e-06,
+      "loss": 1.0202,
+      "step": 2981
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2928389310836792,
+      "learning_rate": 6.957930630385738e-06,
+      "loss": 0.8909,
+      "step": 2982
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3547143936157227,
+      "learning_rate": 6.955977799645491e-06,
+      "loss": 0.8705,
+      "step": 2983
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.32655930519104,
+      "learning_rate": 6.954024616562365e-06,
+      "loss": 0.8383,
+      "step": 2984
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7332683801651001,
+      "learning_rate": 6.952071081488199e-06,
+      "loss": 1.0447,
+      "step": 2985
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6387076377868652,
+      "learning_rate": 6.9501171947748955e-06,
+      "loss": 1.0361,
+      "step": 2986
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5498719215393066,
+      "learning_rate": 6.9481629567744224e-06,
+      "loss": 0.8313,
+      "step": 2987
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.8569600582122803,
+      "learning_rate": 6.94620836783881e-06,
+      "loss": 0.8475,
+      "step": 2988
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3797787427902222,
+      "learning_rate": 6.944253428320148e-06,
+      "loss": 0.8042,
+      "step": 2989
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7253808975219727,
+      "learning_rate": 6.942298138570594e-06,
+      "loss": 1.0268,
+      "step": 2990
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6385308504104614,
+      "learning_rate": 6.940342498942365e-06,
+      "loss": 1.037,
+      "step": 2991
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2986605167388916,
+      "learning_rate": 6.938386509787745e-06,
+      "loss": 0.8762,
+      "step": 2992
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3656010627746582,
+      "learning_rate": 6.936430171459079e-06,
+      "loss": 0.8493,
+      "step": 2993
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6855315566062927,
+      "learning_rate": 6.934473484308775e-06,
+      "loss": 1.0208,
+      "step": 2994
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3936793804168701,
+      "learning_rate": 6.9325164486893015e-06,
+      "loss": 0.8707,
+      "step": 2995
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6738913059234619,
+      "learning_rate": 6.930559064953194e-06,
+      "loss": 1.0212,
+      "step": 2996
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.271657943725586,
+      "learning_rate": 6.928601333453049e-06,
+      "loss": 0.9044,
+      "step": 2997
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2747575044631958,
+      "learning_rate": 6.926643254541522e-06,
+      "loss": 0.8445,
+      "step": 2998
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.327525019645691,
+      "learning_rate": 6.924684828571338e-06,
+      "loss": 0.8498,
+      "step": 2999
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4311046600341797,
+      "learning_rate": 6.922726055895278e-06,
+      "loss": 0.9052,
+      "step": 3000
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4792977571487427,
+      "learning_rate": 6.92076693686619e-06,
+      "loss": 0.8534,
+      "step": 3001
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3159003257751465,
+      "learning_rate": 6.918807471836985e-06,
+      "loss": 0.8713,
+      "step": 3002
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.331673264503479,
+      "learning_rate": 6.916847661160628e-06,
+      "loss": 0.8813,
+      "step": 3003
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.308565616607666,
+      "learning_rate": 6.9148875051901584e-06,
+      "loss": 0.803,
+      "step": 3004
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3787950277328491,
+      "learning_rate": 6.912927004278669e-06,
+      "loss": 0.8683,
+      "step": 3005
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3756848573684692,
+      "learning_rate": 6.9109661587793176e-06,
+      "loss": 0.9144,
+      "step": 3006
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3846133947372437,
+      "learning_rate": 6.909004969045323e-06,
+      "loss": 0.8372,
+      "step": 3007
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.4814780950546265,
+      "learning_rate": 6.9070434354299685e-06,
+      "loss": 0.86,
+      "step": 3008
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2655317783355713,
+      "learning_rate": 6.905081558286598e-06,
+      "loss": 0.8517,
+      "step": 3009
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3090980052947998,
+      "learning_rate": 6.903119337968615e-06,
+      "loss": 0.8323,
+      "step": 3010
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.3724325895309448,
+      "learning_rate": 6.90115677482949e-06,
+      "loss": 0.8687,
+      "step": 3011
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.5243370532989502,
+      "learning_rate": 6.899193869222749e-06,
+      "loss": 0.8218,
+      "step": 3012
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8169469833374023,
+      "learning_rate": 6.8972306215019846e-06,
+      "loss": 1.0363,
+      "step": 3013
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.2801480293273926,
+      "learning_rate": 6.895267032020849e-06,
+      "loss": 0.8488,
+      "step": 3014
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3161693811416626,
+      "learning_rate": 6.893303101133055e-06,
+      "loss": 0.8563,
+      "step": 3015
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3066635131835938,
+      "learning_rate": 6.89133882919238e-06,
+      "loss": 0.8649,
+      "step": 3016
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6743606925010681,
+      "learning_rate": 6.889374216552661e-06,
+      "loss": 1.0578,
+      "step": 3017
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.7082346677780151,
+      "learning_rate": 6.887409263567795e-06,
+      "loss": 0.8718,
+      "step": 3018
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6133765578269958,
+      "learning_rate": 6.885443970591741e-06,
+      "loss": 1.0252,
+      "step": 3019
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4338957071304321,
+      "learning_rate": 6.883478337978522e-06,
+      "loss": 0.8846,
+      "step": 3020
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3341633081436157,
+      "learning_rate": 6.881512366082219e-06,
+      "loss": 0.8545,
+      "step": 3021
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2813103199005127,
+      "learning_rate": 6.8795460552569735e-06,
+      "loss": 0.8362,
+      "step": 3022
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7072163820266724,
+      "learning_rate": 6.877579405856993e-06,
+      "loss": 1.0213,
+      "step": 3023
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3853099346160889,
+      "learning_rate": 6.8756124182365405e-06,
+      "loss": 0.7994,
+      "step": 3024
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3649141788482666,
+      "learning_rate": 6.8736450927499446e-06,
+      "loss": 0.8404,
+      "step": 3025
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4385353326797485,
+      "learning_rate": 6.871677429751588e-06,
+      "loss": 0.8627,
+      "step": 3026
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.328565239906311,
+      "learning_rate": 6.869709429595922e-06,
+      "loss": 0.8277,
+      "step": 3027
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4015023708343506,
+      "learning_rate": 6.867741092637454e-06,
+      "loss": 0.8527,
+      "step": 3028
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4202708005905151,
+      "learning_rate": 6.865772419230754e-06,
+      "loss": 0.871,
+      "step": 3029
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.389575719833374,
+      "learning_rate": 6.863803409730452e-06,
+      "loss": 0.8569,
+      "step": 3030
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4312865734100342,
+      "learning_rate": 6.861834064491235e-06,
+      "loss": 0.8817,
+      "step": 3031
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7674356698989868,
+      "learning_rate": 6.859864383867858e-06,
+      "loss": 1.0395,
+      "step": 3032
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3018871545791626,
+      "learning_rate": 6.85789436821513e-06,
+      "loss": 0.833,
+      "step": 3033
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.264054775238037,
+      "learning_rate": 6.855924017887925e-06,
+      "loss": 0.8426,
+      "step": 3034
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4010534286499023,
+      "learning_rate": 6.853953333241172e-06,
+      "loss": 0.9368,
+      "step": 3035
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.671296238899231,
+      "learning_rate": 6.8519823146298645e-06,
+      "loss": 1.036,
+      "step": 3036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2584319114685059,
+      "learning_rate": 6.850010962409055e-06,
+      "loss": 0.8259,
+      "step": 3037
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2965195178985596,
+      "learning_rate": 6.848039276933855e-06,
+      "loss": 0.8431,
+      "step": 3038
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3588910102844238,
+      "learning_rate": 6.8460672585594365e-06,
+      "loss": 0.8543,
+      "step": 3039
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2917330265045166,
+      "learning_rate": 6.844094907641033e-06,
+      "loss": 0.8446,
+      "step": 3040
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7057825922966003,
+      "learning_rate": 6.842122224533937e-06,
+      "loss": 1.0394,
+      "step": 3041
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.316014051437378,
+      "learning_rate": 6.8401492095935e-06,
+      "loss": 0.9019,
+      "step": 3042
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3660557270050049,
+      "learning_rate": 6.838175863175134e-06,
+      "loss": 0.916,
+      "step": 3043
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3658499717712402,
+      "learning_rate": 6.8362021856343104e-06,
+      "loss": 0.9023,
+      "step": 3044
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3558987379074097,
+      "learning_rate": 6.834228177326561e-06,
+      "loss": 0.8798,
+      "step": 3045
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3921420574188232,
+      "learning_rate": 6.832253838607475e-06,
+      "loss": 0.8833,
+      "step": 3046
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3124773502349854,
+      "learning_rate": 6.830279169832705e-06,
+      "loss": 0.8775,
+      "step": 3047
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4600332975387573,
+      "learning_rate": 6.82830417135796e-06,
+      "loss": 0.8899,
+      "step": 3048
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3557510375976562,
+      "learning_rate": 6.826328843539009e-06,
+      "loss": 0.8089,
+      "step": 3049
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4486881494522095,
+      "learning_rate": 6.824353186731681e-06,
+      "loss": 0.8192,
+      "step": 3050
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7204334735870361,
+      "learning_rate": 6.822377201291863e-06,
+      "loss": 1.0305,
+      "step": 3051
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.277127742767334,
+      "learning_rate": 6.8204008875755024e-06,
+      "loss": 0.8532,
+      "step": 3052
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.039100408554077,
+      "learning_rate": 6.818424245938605e-06,
+      "loss": 0.8413,
+      "step": 3053
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.380968451499939,
+      "learning_rate": 6.816447276737236e-06,
+      "loss": 0.8861,
+      "step": 3054
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2877129316329956,
+      "learning_rate": 6.814469980327519e-06,
+      "loss": 0.8401,
+      "step": 3055
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3522393703460693,
+      "learning_rate": 6.812492357065641e-06,
+      "loss": 0.8255,
+      "step": 3056
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.273059606552124,
+      "learning_rate": 6.810514407307838e-06,
+      "loss": 0.8382,
+      "step": 3057
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.295587182044983,
+      "learning_rate": 6.808536131410416e-06,
+      "loss": 0.7987,
+      "step": 3058
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3038007020950317,
+      "learning_rate": 6.8065575297297305e-06,
+      "loss": 0.8452,
+      "step": 3059
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5074701309204102,
+      "learning_rate": 6.804578602622202e-06,
+      "loss": 0.85,
+      "step": 3060
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4338033199310303,
+      "learning_rate": 6.802599350444306e-06,
+      "loss": 0.8948,
+      "step": 3061
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7278188467025757,
+      "learning_rate": 6.800619773552582e-06,
+      "loss": 1.0164,
+      "step": 3062
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.27864408493042,
+      "learning_rate": 6.798639872303618e-06,
+      "loss": 0.8358,
+      "step": 3063
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3325200080871582,
+      "learning_rate": 6.7966596470540715e-06,
+      "loss": 0.874,
+      "step": 3064
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.2975480556488037,
+      "learning_rate": 6.794679098160649e-06,
+      "loss": 0.8672,
+      "step": 3065
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6797429323196411,
+      "learning_rate": 6.792698225980122e-06,
+      "loss": 1.061,
+      "step": 3066
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2412481307983398,
+      "learning_rate": 6.790717030869319e-06,
+      "loss": 0.8274,
+      "step": 3067
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6182728409767151,
+      "learning_rate": 6.788735513185122e-06,
+      "loss": 1.0355,
+      "step": 3068
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6337783336639404,
+      "learning_rate": 6.786753673284476e-06,
+      "loss": 1.0361,
+      "step": 3069
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6406527757644653,
+      "learning_rate": 6.784771511524384e-06,
+      "loss": 1.02,
+      "step": 3070
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3307441473007202,
+      "learning_rate": 6.7827890282619035e-06,
+      "loss": 0.844,
+      "step": 3071
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.266629695892334,
+      "learning_rate": 6.780806223854154e-06,
+      "loss": 0.8546,
+      "step": 3072
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.307648777961731,
+      "learning_rate": 6.778823098658309e-06,
+      "loss": 0.8214,
+      "step": 3073
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6924213767051697,
+      "learning_rate": 6.776839653031602e-06,
+      "loss": 1.04,
+      "step": 3074
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.361074686050415,
+      "learning_rate": 6.774855887331325e-06,
+      "loss": 0.8018,
+      "step": 3075
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3784397840499878,
+      "learning_rate": 6.772871801914826e-06,
+      "loss": 0.8502,
+      "step": 3076
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4367430210113525,
+      "learning_rate": 6.7708873971395095e-06,
+      "loss": 0.8208,
+      "step": 3077
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.5022015571594238,
+      "learning_rate": 6.768902673362841e-06,
+      "loss": 0.9036,
+      "step": 3078
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3044697046279907,
+      "learning_rate": 6.766917630942341e-06,
+      "loss": 0.8554,
+      "step": 3079
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3401353359222412,
+      "learning_rate": 6.764932270235587e-06,
+      "loss": 0.8269,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2881014347076416,
+      "learning_rate": 6.762946591600218e-06,
+      "loss": 0.8176,
+      "step": 3081
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3233104944229126,
+      "learning_rate": 6.7609605953939225e-06,
+      "loss": 0.827,
+      "step": 3082
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.4115384817123413,
+      "learning_rate": 6.758974281974454e-06,
+      "loss": 0.8777,
+      "step": 3083
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3443444967269897,
+      "learning_rate": 6.756987651699618e-06,
+      "loss": 0.8477,
+      "step": 3084
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.380110740661621,
+      "learning_rate": 6.755000704927281e-06,
+      "loss": 0.8565,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3361269235610962,
+      "learning_rate": 6.753013442015365e-06,
+      "loss": 0.8238,
+      "step": 3086
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.3312931060791016,
+      "learning_rate": 6.751025863321845e-06,
+      "loss": 0.866,
+      "step": 3087
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.429092288017273,
+      "learning_rate": 6.749037969204759e-06,
+      "loss": 0.8584,
+      "step": 3088
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2971727848052979,
+      "learning_rate": 6.747049760022197e-06,
+      "loss": 0.8222,
+      "step": 3089
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.265570044517517,
+      "learning_rate": 6.745061236132311e-06,
+      "loss": 0.8087,
+      "step": 3090
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.8327125906944275,
+      "learning_rate": 6.743072397893307e-06,
+      "loss": 1.0394,
+      "step": 3091
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7312641143798828,
+      "learning_rate": 6.741083245663443e-06,
+      "loss": 1.0505,
+      "step": 3092
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3563973903656006,
+      "learning_rate": 6.739093779801039e-06,
+      "loss": 0.831,
+      "step": 3093
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3776118755340576,
+      "learning_rate": 6.737104000664474e-06,
+      "loss": 0.8336,
+      "step": 3094
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7334795594215393,
+      "learning_rate": 6.735113908612176e-06,
+      "loss": 1.0389,
+      "step": 3095
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3704447746276855,
+      "learning_rate": 6.733123504002635e-06,
+      "loss": 0.8783,
+      "step": 3096
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.318499207496643,
+      "learning_rate": 6.731132787194394e-06,
+      "loss": 0.8245,
+      "step": 3097
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7119347453117371,
+      "learning_rate": 6.729141758546054e-06,
+      "loss": 1.0363,
+      "step": 3098
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2610827684402466,
+      "learning_rate": 6.727150418416273e-06,
+      "loss": 0.809,
+      "step": 3099
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3379114866256714,
+      "learning_rate": 6.7251587671637616e-06,
+      "loss": 0.8328,
+      "step": 3100
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2893261909484863,
+      "learning_rate": 6.723166805147292e-06,
+      "loss": 0.8237,
+      "step": 3101
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3390871286392212,
+      "learning_rate": 6.721174532725686e-06,
+      "loss": 0.8836,
+      "step": 3102
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3372972011566162,
+      "learning_rate": 6.719181950257826e-06,
+      "loss": 0.8695,
+      "step": 3103
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.729439914226532,
+      "learning_rate": 6.717189058102649e-06,
+      "loss": 1.0374,
+      "step": 3104
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6926238536834717,
+      "learning_rate": 6.715195856619146e-06,
+      "loss": 1.035,
+      "step": 3105
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.619774580001831,
+      "learning_rate": 6.713202346166368e-06,
+      "loss": 1.0135,
+      "step": 3106
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6444083452224731,
+      "learning_rate": 6.7112085271034154e-06,
+      "loss": 0.8984,
+      "step": 3107
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6890362501144409,
+      "learning_rate": 6.70921439978945e-06,
+      "loss": 1.0297,
+      "step": 3108
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4300448894500732,
+      "learning_rate": 6.707219964583688e-06,
+      "loss": 0.8925,
+      "step": 3109
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.679058313369751,
+      "learning_rate": 6.705225221845395e-06,
+      "loss": 1.0302,
+      "step": 3110
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3401165008544922,
+      "learning_rate": 6.703230171933902e-06,
+      "loss": 0.8039,
+      "step": 3111
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4105561971664429,
+      "learning_rate": 6.701234815208587e-06,
+      "loss": 0.8241,
+      "step": 3112
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4878029823303223,
+      "learning_rate": 6.699239152028888e-06,
+      "loss": 0.8829,
+      "step": 3113
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4990525245666504,
+      "learning_rate": 6.697243182754297e-06,
+      "loss": 0.8669,
+      "step": 3114
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2641690969467163,
+      "learning_rate": 6.69524690774436e-06,
+      "loss": 0.7848,
+      "step": 3115
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.360312581062317,
+      "learning_rate": 6.693250327358678e-06,
+      "loss": 0.8812,
+      "step": 3116
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3991992473602295,
+      "learning_rate": 6.691253441956909e-06,
+      "loss": 0.8278,
+      "step": 3117
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3914247751235962,
+      "learning_rate": 6.689256251898763e-06,
+      "loss": 0.8746,
+      "step": 3118
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3918375968933105,
+      "learning_rate": 6.687258757544012e-06,
+      "loss": 0.8701,
+      "step": 3119
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3864266872406006,
+      "learning_rate": 6.6852609592524706e-06,
+      "loss": 0.8591,
+      "step": 3120
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4073928594589233,
+      "learning_rate": 6.68326285738402e-06,
+      "loss": 0.8909,
+      "step": 3121
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.371484637260437,
+      "learning_rate": 6.6812644522985885e-06,
+      "loss": 0.8366,
+      "step": 3122
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3283365964889526,
+      "learning_rate": 6.679265744356164e-06,
+      "loss": 0.7436,
+      "step": 3123
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4416142702102661,
+      "learning_rate": 6.677266733916784e-06,
+      "loss": 0.8282,
+      "step": 3124
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3063918352127075,
+      "learning_rate": 6.675267421340544e-06,
+      "loss": 0.875,
+      "step": 3125
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3312994241714478,
+      "learning_rate": 6.673267806987594e-06,
+      "loss": 0.8701,
+      "step": 3126
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3071624040603638,
+      "learning_rate": 6.671267891218137e-06,
+      "loss": 0.8,
+      "step": 3127
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2381222248077393,
+      "learning_rate": 6.669267674392432e-06,
+      "loss": 0.8412,
+      "step": 3128
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3942906856536865,
+      "learning_rate": 6.667267156870788e-06,
+      "loss": 0.8693,
+      "step": 3129
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3170876502990723,
+      "learning_rate": 6.665266339013573e-06,
+      "loss": 0.8445,
+      "step": 3130
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9697448015213013,
+      "learning_rate": 6.663265221181205e-06,
+      "loss": 1.0473,
+      "step": 3131
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4311827421188354,
+      "learning_rate": 6.661263803734163e-06,
+      "loss": 0.8092,
+      "step": 3132
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3658009767532349,
+      "learning_rate": 6.659262087032972e-06,
+      "loss": 0.9013,
+      "step": 3133
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4006996154785156,
+      "learning_rate": 6.657260071438213e-06,
+      "loss": 0.8405,
+      "step": 3134
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3130370378494263,
+      "learning_rate": 6.6552577573105236e-06,
+      "loss": 0.9133,
+      "step": 3135
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.25845205783844,
+      "learning_rate": 6.653255145010594e-06,
+      "loss": 0.8508,
+      "step": 3136
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3364442586898804,
+      "learning_rate": 6.651252234899167e-06,
+      "loss": 0.8011,
+      "step": 3137
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7493829727172852,
+      "learning_rate": 6.64924902733704e-06,
+      "loss": 1.0555,
+      "step": 3138
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3973716497421265,
+      "learning_rate": 6.647245522685062e-06,
+      "loss": 0.8725,
+      "step": 3139
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.436095118522644,
+      "learning_rate": 6.6452417213041394e-06,
+      "loss": 0.8211,
+      "step": 3140
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6948164701461792,
+      "learning_rate": 6.643237623555228e-06,
+      "loss": 1.0573,
+      "step": 3141
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3968842029571533,
+      "learning_rate": 6.641233229799342e-06,
+      "loss": 0.7877,
+      "step": 3142
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6744611859321594,
+      "learning_rate": 6.639228540397542e-06,
+      "loss": 1.0277,
+      "step": 3143
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2659611701965332,
+      "learning_rate": 6.6372235557109475e-06,
+      "loss": 0.8758,
+      "step": 3144
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7219061851501465,
+      "learning_rate": 6.6352182761007286e-06,
+      "loss": 1.0423,
+      "step": 3145
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4072785377502441,
+      "learning_rate": 6.63321270192811e-06,
+      "loss": 0.8427,
+      "step": 3146
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.295092225074768,
+      "learning_rate": 6.6312068335543685e-06,
+      "loss": 0.8414,
+      "step": 3147
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.375829815864563,
+      "learning_rate": 6.629200671340834e-06,
+      "loss": 0.9107,
+      "step": 3148
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.5170754194259644,
+      "learning_rate": 6.627194215648889e-06,
+      "loss": 0.7687,
+      "step": 3149
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.437765121459961,
+      "learning_rate": 6.625187466839969e-06,
+      "loss": 0.8794,
+      "step": 3150
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3426730632781982,
+      "learning_rate": 6.623180425275565e-06,
+      "loss": 0.9098,
+      "step": 3151
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2567600011825562,
+      "learning_rate": 6.621173091317216e-06,
+      "loss": 0.7965,
+      "step": 3152
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3674216270446777,
+      "learning_rate": 6.619165465326516e-06,
+      "loss": 0.8528,
+      "step": 3153
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.9597799181938171,
+      "learning_rate": 6.617157547665112e-06,
+      "loss": 1.0208,
+      "step": 3154
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2993170022964478,
+      "learning_rate": 6.615149338694704e-06,
+      "loss": 0.8231,
+      "step": 3155
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3016778230667114,
+      "learning_rate": 6.613140838777041e-06,
+      "loss": 0.8688,
+      "step": 3156
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.34652578830719,
+      "learning_rate": 6.611132048273932e-06,
+      "loss": 0.8357,
+      "step": 3157
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3508219718933105,
+      "learning_rate": 6.609122967547228e-06,
+      "loss": 0.86,
+      "step": 3158
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2808936834335327,
+      "learning_rate": 6.607113596958841e-06,
+      "loss": 0.8304,
+      "step": 3159
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.6240317821502686,
+      "learning_rate": 6.605103936870731e-06,
+      "loss": 0.8978,
+      "step": 3160
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.4742474555969238,
+      "learning_rate": 6.60309398764491e-06,
+      "loss": 0.897,
+      "step": 3161
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7429847717285156,
+      "learning_rate": 6.601083749643444e-06,
+      "loss": 1.0252,
+      "step": 3162
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.389528512954712,
+      "learning_rate": 6.599073223228449e-06,
+      "loss": 0.8857,
+      "step": 3163
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6645315885543823,
+      "learning_rate": 6.5970624087620975e-06,
+      "loss": 1.0702,
+      "step": 3164
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.2525591850280762,
+      "learning_rate": 6.595051306606606e-06,
+      "loss": 0.8657,
+      "step": 3165
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3176701068878174,
+      "learning_rate": 6.593039917124251e-06,
+      "loss": 0.8326,
+      "step": 3166
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 1.3404972553253174,
+      "learning_rate": 6.591028240677355e-06,
+      "loss": 0.8051,
+      "step": 3167
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2860705852508545,
+      "learning_rate": 6.589016277628294e-06,
+      "loss": 0.7948,
+      "step": 3168
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7382155060768127,
+      "learning_rate": 6.587004028339496e-06,
+      "loss": 1.0493,
+      "step": 3169
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4142343997955322,
+      "learning_rate": 6.584991493173443e-06,
+      "loss": 0.8644,
+      "step": 3170
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3349246978759766,
+      "learning_rate": 6.582978672492662e-06,
+      "loss": 0.8635,
+      "step": 3171
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3921685218811035,
+      "learning_rate": 6.580965566659739e-06,
+      "loss": 0.9126,
+      "step": 3172
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.660760223865509,
+      "learning_rate": 6.5789521760373035e-06,
+      "loss": 1.0099,
+      "step": 3173
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2773680686950684,
+      "learning_rate": 6.576938500988045e-06,
+      "loss": 0.9054,
+      "step": 3174
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6619707942008972,
+      "learning_rate": 6.5749245418746965e-06,
+      "loss": 1.0554,
+      "step": 3175
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.346888542175293,
+      "learning_rate": 6.5729102990600466e-06,
+      "loss": 0.8384,
+      "step": 3176
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6714622378349304,
+      "learning_rate": 6.5708957729069355e-06,
+      "loss": 1.0554,
+      "step": 3177
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.639369010925293,
+      "learning_rate": 6.568880963778249e-06,
+      "loss": 1.0218,
+      "step": 3178
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.322191596031189,
+      "learning_rate": 6.56686587203693e-06,
+      "loss": 0.8276,
+      "step": 3179
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4192583560943604,
+      "learning_rate": 6.564850498045971e-06,
+      "loss": 0.8635,
+      "step": 3180
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4554405212402344,
+      "learning_rate": 6.562834842168413e-06,
+      "loss": 0.8957,
+      "step": 3181
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2535560131072998,
+      "learning_rate": 6.560818904767349e-06,
+      "loss": 0.8236,
+      "step": 3182
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3238840103149414,
+      "learning_rate": 6.558802686205922e-06,
+      "loss": 0.8759,
+      "step": 3183
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7163515090942383,
+      "learning_rate": 6.556786186847328e-06,
+      "loss": 1.0408,
+      "step": 3184
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2823340892791748,
+      "learning_rate": 6.554769407054811e-06,
+      "loss": 0.8587,
+      "step": 3185
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2907580137252808,
+      "learning_rate": 6.552752347191667e-06,
+      "loss": 0.8788,
+      "step": 3186
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3079754114151,
+      "learning_rate": 6.5507350076212405e-06,
+      "loss": 0.82,
+      "step": 3187
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3553670644760132,
+      "learning_rate": 6.548717388706931e-06,
+      "loss": 0.8561,
+      "step": 3188
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4185081720352173,
+      "learning_rate": 6.546699490812184e-06,
+      "loss": 0.8838,
+      "step": 3189
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7799632549285889,
+      "learning_rate": 6.544681314300494e-06,
+      "loss": 1.0441,
+      "step": 3190
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2896445989608765,
+      "learning_rate": 6.542662859535411e-06,
+      "loss": 0.8486,
+      "step": 3191
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3660210371017456,
+      "learning_rate": 6.540644126880531e-06,
+      "loss": 0.8679,
+      "step": 3192
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2553318738937378,
+      "learning_rate": 6.5386251166995006e-06,
+      "loss": 0.7885,
+      "step": 3193
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4636603593826294,
+      "learning_rate": 6.53660582935602e-06,
+      "loss": 0.8214,
+      "step": 3194
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.063786506652832,
+      "learning_rate": 6.534586265213834e-06,
+      "loss": 0.8872,
+      "step": 3195
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6883736848831177,
+      "learning_rate": 6.53256642463674e-06,
+      "loss": 1.0437,
+      "step": 3196
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4040820598602295,
+      "learning_rate": 6.530546307988586e-06,
+      "loss": 0.8718,
+      "step": 3197
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.356772541999817,
+      "learning_rate": 6.528525915633267e-06,
+      "loss": 0.8327,
+      "step": 3198
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2550774812698364,
+      "learning_rate": 6.526505247934729e-06,
+      "loss": 0.8309,
+      "step": 3199
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.385721206665039,
+      "learning_rate": 6.5244843052569696e-06,
+      "loss": 0.8483,
+      "step": 3200
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3030407428741455,
+      "learning_rate": 6.522463087964033e-06,
+      "loss": 0.8037,
+      "step": 3201
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6997559070587158,
+      "learning_rate": 6.520441596420015e-06,
+      "loss": 1.0339,
+      "step": 3202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.307430624961853,
+      "learning_rate": 6.51841983098906e-06,
+      "loss": 0.8356,
+      "step": 3203
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.335957646369934,
+      "learning_rate": 6.516397792035361e-06,
+      "loss": 0.8366,
+      "step": 3204
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4911177158355713,
+      "learning_rate": 6.51437547992316e-06,
+      "loss": 0.8649,
+      "step": 3205
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4027776718139648,
+      "learning_rate": 6.512352895016751e-06,
+      "loss": 0.7814,
+      "step": 3206
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6665273308753967,
+      "learning_rate": 6.510330037680475e-06,
+      "loss": 1.0599,
+      "step": 3207
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.351499319076538,
+      "learning_rate": 6.5083069082787205e-06,
+      "loss": 0.8849,
+      "step": 3208
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.283685564994812,
+      "learning_rate": 6.506283507175929e-06,
+      "loss": 0.8005,
+      "step": 3209
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.331523060798645,
+      "learning_rate": 6.504259834736587e-06,
+      "loss": 0.8923,
+      "step": 3210
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6509556174278259,
+      "learning_rate": 6.502235891325233e-06,
+      "loss": 1.0266,
+      "step": 3211
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6315797567367554,
+      "learning_rate": 6.500211677306451e-06,
+      "loss": 1.0216,
+      "step": 3212
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.8757507801055908,
+      "learning_rate": 6.4981871930448795e-06,
+      "loss": 0.8672,
+      "step": 3213
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6221902370452881,
+      "learning_rate": 6.496162438905198e-06,
+      "loss": 1.0337,
+      "step": 3214
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.380060076713562,
+      "learning_rate": 6.49413741525214e-06,
+      "loss": 0.8696,
+      "step": 3215
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2807421684265137,
+      "learning_rate": 6.492112122450487e-06,
+      "loss": 0.8291,
+      "step": 3216
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.292570948600769,
+      "learning_rate": 6.4900865608650666e-06,
+      "loss": 0.894,
+      "step": 3217
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3720440864562988,
+      "learning_rate": 6.488060730860757e-06,
+      "loss": 0.799,
+      "step": 3218
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7914753556251526,
+      "learning_rate": 6.486034632802484e-06,
+      "loss": 1.0296,
+      "step": 3219
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7689070701599121,
+      "learning_rate": 6.48400826705522e-06,
+      "loss": 1.0316,
+      "step": 3220
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2813210487365723,
+      "learning_rate": 6.481981633983989e-06,
+      "loss": 0.8353,
+      "step": 3221
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.286820888519287,
+      "learning_rate": 6.479954733953864e-06,
+      "loss": 0.8805,
+      "step": 3222
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3354796171188354,
+      "learning_rate": 6.477927567329958e-06,
+      "loss": 0.872,
+      "step": 3223
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4155309200286865,
+      "learning_rate": 6.475900134477441e-06,
+      "loss": 0.8093,
+      "step": 3224
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7962488532066345,
+      "learning_rate": 6.473872435761527e-06,
+      "loss": 1.019,
+      "step": 3225
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 4.723483085632324,
+      "learning_rate": 6.471844471547479e-06,
+      "loss": 0.8426,
+      "step": 3226
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3764568567276,
+      "learning_rate": 6.469816242200606e-06,
+      "loss": 0.8711,
+      "step": 3227
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3903069496154785,
+      "learning_rate": 6.467787748086266e-06,
+      "loss": 0.82,
+      "step": 3228
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3425043821334839,
+      "learning_rate": 6.465758989569866e-06,
+      "loss": 0.918,
+      "step": 3229
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7074387073516846,
+      "learning_rate": 6.463729967016858e-06,
+      "loss": 1.0574,
+      "step": 3230
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6414318084716797,
+      "learning_rate": 6.461700680792745e-06,
+      "loss": 1.0581,
+      "step": 3231
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2837061882019043,
+      "learning_rate": 6.459671131263073e-06,
+      "loss": 0.8428,
+      "step": 3232
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3715265989303589,
+      "learning_rate": 6.4576413187934394e-06,
+      "loss": 0.9186,
+      "step": 3233
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.4247117042541504,
+      "learning_rate": 6.455611243749486e-06,
+      "loss": 0.842,
+      "step": 3234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.390359878540039,
+      "learning_rate": 6.453580906496905e-06,
+      "loss": 0.8186,
+      "step": 3235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2908896207809448,
+      "learning_rate": 6.451550307401432e-06,
+      "loss": 0.8667,
+      "step": 3236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3175994157791138,
+      "learning_rate": 6.449519446828855e-06,
+      "loss": 0.8414,
+      "step": 3237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8706154227256775,
+      "learning_rate": 6.4474883251450015e-06,
+      "loss": 1.0366,
+      "step": 3238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.518818736076355,
+      "learning_rate": 6.445456942715756e-06,
+      "loss": 0.8299,
+      "step": 3239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.3407732248306274,
+      "learning_rate": 6.4434252999070406e-06,
+      "loss": 0.7933,
+      "step": 3240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.2757612466812134,
+      "learning_rate": 6.441393397084827e-06,
+      "loss": 0.8158,
+      "step": 3241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.731271505355835,
+      "learning_rate": 6.439361234615139e-06,
+      "loss": 1.0216,
+      "step": 3242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.459439754486084,
+      "learning_rate": 6.437328812864041e-06,
+      "loss": 0.855,
+      "step": 3243
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2869137525558472,
+      "learning_rate": 6.435296132197647e-06,
+      "loss": 0.8851,
+      "step": 3244
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2933931350708008,
+      "learning_rate": 6.433263192982116e-06,
+      "loss": 0.8402,
+      "step": 3245
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7046028971672058,
+      "learning_rate": 6.431229995583656e-06,
+      "loss": 1.0437,
+      "step": 3246
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2870112657546997,
+      "learning_rate": 6.429196540368516e-06,
+      "loss": 0.8431,
+      "step": 3247
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3182005882263184,
+      "learning_rate": 6.427162827702999e-06,
+      "loss": 0.847,
+      "step": 3248
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3368850946426392,
+      "learning_rate": 6.425128857953451e-06,
+      "loss": 0.8438,
+      "step": 3249
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.511538505554199,
+      "learning_rate": 6.423094631486264e-06,
+      "loss": 0.865,
+      "step": 3250
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.712445855140686,
+      "learning_rate": 6.421060148667874e-06,
+      "loss": 1.0351,
+      "step": 3251
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4412076473236084,
+      "learning_rate": 6.419025409864766e-06,
+      "loss": 0.8917,
+      "step": 3252
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 2.43279767036438,
+      "learning_rate": 6.416990415443471e-06,
+      "loss": 0.9042,
+      "step": 3253
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.298100233078003,
+      "learning_rate": 6.414955165770569e-06,
+      "loss": 0.8327,
+      "step": 3254
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3330491781234741,
+      "learning_rate": 6.4129196612126765e-06,
+      "loss": 0.8083,
+      "step": 3255
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6171782612800598,
+      "learning_rate": 6.4108839021364665e-06,
+      "loss": 1.0132,
+      "step": 3256
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3283662796020508,
+      "learning_rate": 6.4088478889086525e-06,
+      "loss": 0.813,
+      "step": 3257
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3321255445480347,
+      "learning_rate": 6.406811621895994e-06,
+      "loss": 0.8062,
+      "step": 3258
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4067027568817139,
+      "learning_rate": 6.404775101465298e-06,
+      "loss": 0.8479,
+      "step": 3259
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2701798677444458,
+      "learning_rate": 6.402738327983413e-06,
+      "loss": 0.8033,
+      "step": 3260
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6613016128540039,
+      "learning_rate": 6.40070130181724e-06,
+      "loss": 1.0209,
+      "step": 3261
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2848504781723022,
+      "learning_rate": 6.398664023333719e-06,
+      "loss": 0.7794,
+      "step": 3262
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4074174165725708,
+      "learning_rate": 6.39662649289984e-06,
+      "loss": 0.9259,
+      "step": 3263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3017449378967285,
+      "learning_rate": 6.394588710882635e-06,
+      "loss": 0.837,
+      "step": 3264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6514301300048828,
+      "learning_rate": 6.392550677649183e-06,
+      "loss": 1.0482,
+      "step": 3265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6395043730735779,
+      "learning_rate": 6.390512393566609e-06,
+      "loss": 1.0156,
+      "step": 3266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5955634117126465,
+      "learning_rate": 6.388473859002082e-06,
+      "loss": 1.0254,
+      "step": 3267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4880129098892212,
+      "learning_rate": 6.3864350743228165e-06,
+      "loss": 0.854,
+      "step": 3268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6286242604255676,
+      "learning_rate": 6.38439603989607e-06,
+      "loss": 1.0512,
+      "step": 3269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3253425359725952,
+      "learning_rate": 6.382356756089149e-06,
+      "loss": 0.8152,
+      "step": 3270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.613697350025177,
+      "learning_rate": 6.380317223269401e-06,
+      "loss": 1.0492,
+      "step": 3271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3275458812713623,
+      "learning_rate": 6.3782774418042216e-06,
+      "loss": 0.8917,
+      "step": 3272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2623478174209595,
+      "learning_rate": 6.376237412061051e-06,
+      "loss": 0.8262,
+      "step": 3273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2314285039901733,
+      "learning_rate": 6.37419713440737e-06,
+      "loss": 0.8024,
+      "step": 3274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.6525194644927979,
+      "learning_rate": 6.372156609210709e-06,
+      "loss": 0.8314,
+      "step": 3275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3007644414901733,
+      "learning_rate": 6.370115836838642e-06,
+      "loss": 0.8316,
+      "step": 3276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8484784960746765,
+      "learning_rate": 6.368074817658783e-06,
+      "loss": 1.0276,
+      "step": 3277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3717372417449951,
+      "learning_rate": 6.366033552038796e-06,
+      "loss": 0.9083,
+      "step": 3278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4231020212173462,
+      "learning_rate": 6.363992040346388e-06,
+      "loss": 0.9123,
+      "step": 3279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3271902799606323,
+      "learning_rate": 6.361950282949306e-06,
+      "loss": 0.838,
+      "step": 3280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4000310897827148,
+      "learning_rate": 6.359908280215349e-06,
+      "loss": 0.8624,
+      "step": 3281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7044742703437805,
+      "learning_rate": 6.357866032512355e-06,
+      "loss": 1.0319,
+      "step": 3282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.245535969734192,
+      "learning_rate": 6.355823540208207e-06,
+      "loss": 0.8489,
+      "step": 3283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3733680248260498,
+      "learning_rate": 6.3537808036708314e-06,
+      "loss": 0.8804,
+      "step": 3284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3527578115463257,
+      "learning_rate": 6.3517378232682005e-06,
+      "loss": 0.8572,
+      "step": 3285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4243533611297607,
+      "learning_rate": 6.34969459936833e-06,
+      "loss": 0.8455,
+      "step": 3286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3123029470443726,
+      "learning_rate": 6.347651132339279e-06,
+      "loss": 0.8118,
+      "step": 3287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6993917226791382,
+      "learning_rate": 6.345607422549148e-06,
+      "loss": 1.0293,
+      "step": 3288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2953420877456665,
+      "learning_rate": 6.3435634703660855e-06,
+      "loss": 0.8196,
+      "step": 3289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.381770372390747,
+      "learning_rate": 6.341519276158282e-06,
+      "loss": 0.8546,
+      "step": 3290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3473315238952637,
+      "learning_rate": 6.339474840293971e-06,
+      "loss": 0.8382,
+      "step": 3291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4145386219024658,
+      "learning_rate": 6.337430163141431e-06,
+      "loss": 0.8872,
+      "step": 3292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4052764177322388,
+      "learning_rate": 6.335385245068979e-06,
+      "loss": 0.8112,
+      "step": 3293
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3603298664093018,
+      "learning_rate": 6.333340086444984e-06,
+      "loss": 0.877,
+      "step": 3294
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7319486141204834,
+      "learning_rate": 6.33129468763785e-06,
+      "loss": 1.057,
+      "step": 3295
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4854880571365356,
+      "learning_rate": 6.329249049016032e-06,
+      "loss": 0.9088,
+      "step": 3296
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6760543584823608,
+      "learning_rate": 6.32720317094802e-06,
+      "loss": 1.0546,
+      "step": 3297
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6318080425262451,
+      "learning_rate": 6.325157053802354e-06,
+      "loss": 1.007,
+      "step": 3298
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2917273044586182,
+      "learning_rate": 6.323110697947611e-06,
+      "loss": 0.8041,
+      "step": 3299
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3409204483032227,
+      "learning_rate": 6.321064103752415e-06,
+      "loss": 0.8096,
+      "step": 3300
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3815393447875977,
+      "learning_rate": 6.319017271585437e-06,
+      "loss": 0.907,
+      "step": 3301
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3223551511764526,
+      "learning_rate": 6.31697020181538e-06,
+      "loss": 0.8512,
+      "step": 3302
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3874106407165527,
+      "learning_rate": 6.314922894810999e-06,
+      "loss": 0.8766,
+      "step": 3303
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9824295043945312,
+      "learning_rate": 6.312875350941088e-06,
+      "loss": 1.0274,
+      "step": 3304
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7947071194648743,
+      "learning_rate": 6.310827570574484e-06,
+      "loss": 1.03,
+      "step": 3305
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6405401229858398,
+      "learning_rate": 6.3087795540800664e-06,
+      "loss": 1.0557,
+      "step": 3306
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3464752435684204,
+      "learning_rate": 6.306731301826759e-06,
+      "loss": 0.8349,
+      "step": 3307
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8482257127761841,
+      "learning_rate": 6.304682814183526e-06,
+      "loss": 1.0326,
+      "step": 3308
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.8591811656951904,
+      "learning_rate": 6.302634091519374e-06,
+      "loss": 1.0465,
+      "step": 3309
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.317479133605957,
+      "learning_rate": 6.300585134203354e-06,
+      "loss": 0.7918,
+      "step": 3310
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2591843605041504,
+      "learning_rate": 6.298535942604556e-06,
+      "loss": 0.8121,
+      "step": 3311
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2829535007476807,
+      "learning_rate": 6.296486517092116e-06,
+      "loss": 0.8266,
+      "step": 3312
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3655019998550415,
+      "learning_rate": 6.2944368580352085e-06,
+      "loss": 0.7926,
+      "step": 3313
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7998020648956299,
+      "learning_rate": 6.292386965803054e-06,
+      "loss": 1.0471,
+      "step": 3314
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.7384310960769653,
+      "learning_rate": 6.290336840764913e-06,
+      "loss": 1.0163,
+      "step": 3315
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3165277242660522,
+      "learning_rate": 6.288286483290082e-06,
+      "loss": 0.816,
+      "step": 3316
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.3659716844558716,
+      "learning_rate": 6.286235893747913e-06,
+      "loss": 0.8369,
+      "step": 3317
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4330832958221436,
+      "learning_rate": 6.284185072507786e-06,
+      "loss": 0.8541,
+      "step": 3318
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.4362763166427612,
+      "learning_rate": 6.28213401993913e-06,
+      "loss": 0.8714,
+      "step": 3319
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.365130066871643,
+      "learning_rate": 6.280082736411418e-06,
+      "loss": 0.8497,
+      "step": 3320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3538832664489746,
+      "learning_rate": 6.278031222294157e-06,
+      "loss": 0.8112,
+      "step": 3321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4163798093795776,
+      "learning_rate": 6.275979477956899e-06,
+      "loss": 0.8381,
+      "step": 3322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3702069520950317,
+      "learning_rate": 6.2739275037692394e-06,
+      "loss": 0.8506,
+      "step": 3323
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4380626678466797,
+      "learning_rate": 6.2718753001008145e-06,
+      "loss": 0.8508,
+      "step": 3324
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4252172708511353,
+      "learning_rate": 6.2698228673213005e-06,
+      "loss": 0.8084,
+      "step": 3325
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5210602283477783,
+      "learning_rate": 6.2677702058004145e-06,
+      "loss": 0.8921,
+      "step": 3326
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3524378538131714,
+      "learning_rate": 6.265717315907915e-06,
+      "loss": 0.8099,
+      "step": 3327
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8520245552062988,
+      "learning_rate": 6.263664198013604e-06,
+      "loss": 1.0457,
+      "step": 3328
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4320323467254639,
+      "learning_rate": 6.261610852487323e-06,
+      "loss": 0.8326,
+      "step": 3329
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.241868495941162,
+      "learning_rate": 6.259557279698952e-06,
+      "loss": 0.8796,
+      "step": 3330
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4740241765975952,
+      "learning_rate": 6.257503480018417e-06,
+      "loss": 0.81,
+      "step": 3331
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.434856653213501,
+      "learning_rate": 6.255449453815679e-06,
+      "loss": 0.8134,
+      "step": 3332
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.335726261138916,
+      "learning_rate": 6.253395201460746e-06,
+      "loss": 0.7786,
+      "step": 3333
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.700243592262268,
+      "learning_rate": 6.251340723323663e-06,
+      "loss": 0.9158,
+      "step": 3334
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7394315600395203,
+      "learning_rate": 6.249286019774517e-06,
+      "loss": 1.0142,
+      "step": 3335
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4873994588851929,
+      "learning_rate": 6.247231091183432e-06,
+      "loss": 0.8749,
+      "step": 3336
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.8140939474105835,
+      "learning_rate": 6.245175937920579e-06,
+      "loss": 0.8781,
+      "step": 3337
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.44718337059021,
+      "learning_rate": 6.2431205603561626e-06,
+      "loss": 0.8516,
+      "step": 3338
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3269323110580444,
+      "learning_rate": 6.241064958860435e-06,
+      "loss": 0.9077,
+      "step": 3339
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2965173721313477,
+      "learning_rate": 6.239009133803682e-06,
+      "loss": 0.8217,
+      "step": 3340
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2706760168075562,
+      "learning_rate": 6.2369530855562335e-06,
+      "loss": 0.7987,
+      "step": 3341
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8180785775184631,
+      "learning_rate": 6.234896814488459e-06,
+      "loss": 1.0424,
+      "step": 3342
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.359400749206543,
+      "learning_rate": 6.232840320970769e-06,
+      "loss": 0.8791,
+      "step": 3343
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6518474221229553,
+      "learning_rate": 6.2307836053736096e-06,
+      "loss": 1.023,
+      "step": 3344
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4058072566986084,
+      "learning_rate": 6.228726668067473e-06,
+      "loss": 0.8191,
+      "step": 3345
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3543418645858765,
+      "learning_rate": 6.2266695094228866e-06,
+      "loss": 0.8289,
+      "step": 3346
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3507784605026245,
+      "learning_rate": 6.224612129810422e-06,
+      "loss": 0.8499,
+      "step": 3347
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8200409412384033,
+      "learning_rate": 6.222554529600689e-06,
+      "loss": 1.0346,
+      "step": 3348
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5729326009750366,
+      "learning_rate": 6.220496709164332e-06,
+      "loss": 0.8601,
+      "step": 3349
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3616877794265747,
+      "learning_rate": 6.218438668872042e-06,
+      "loss": 0.8711,
+      "step": 3350
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.358131766319275,
+      "learning_rate": 6.2163804090945465e-06,
+      "loss": 0.8521,
+      "step": 3351
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3093593120574951,
+      "learning_rate": 6.214321930202615e-06,
+      "loss": 0.8895,
+      "step": 3352
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2938817739486694,
+      "learning_rate": 6.212263232567053e-06,
+      "loss": 0.8612,
+      "step": 3353
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3405755758285522,
+      "learning_rate": 6.210204316558707e-06,
+      "loss": 0.8673,
+      "step": 3354
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3076634407043457,
+      "learning_rate": 6.208145182548462e-06,
+      "loss": 0.849,
+      "step": 3355
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3676338195800781,
+      "learning_rate": 6.2060858309072445e-06,
+      "loss": 0.8243,
+      "step": 3356
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2943880558013916,
+      "learning_rate": 6.204026262006017e-06,
+      "loss": 0.8463,
+      "step": 3357
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7443543672561646,
+      "learning_rate": 6.201966476215788e-06,
+      "loss": 1.0524,
+      "step": 3358
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6887378692626953,
+      "learning_rate": 6.199906473907593e-06,
+      "loss": 1.0443,
+      "step": 3359
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6345992684364319,
+      "learning_rate": 6.197846255452516e-06,
+      "loss": 1.0296,
+      "step": 3360
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3509758710861206,
+      "learning_rate": 6.195785821221679e-06,
+      "loss": 0.8634,
+      "step": 3361
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6542819142341614,
+      "learning_rate": 6.193725171586241e-06,
+      "loss": 1.0486,
+      "step": 3362
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3407121896743774,
+      "learning_rate": 6.191664306917399e-06,
+      "loss": 0.8896,
+      "step": 3363
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.439468264579773,
+      "learning_rate": 6.189603227586389e-06,
+      "loss": 0.875,
+      "step": 3364
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.345211148262024,
+      "learning_rate": 6.187541933964489e-06,
+      "loss": 0.8472,
+      "step": 3365
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.261631727218628,
+      "learning_rate": 6.18548042642301e-06,
+      "loss": 0.802,
+      "step": 3366
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3623533248901367,
+      "learning_rate": 6.1834187053333095e-06,
+      "loss": 0.8899,
+      "step": 3367
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4643363952636719,
+      "learning_rate": 6.181356771066773e-06,
+      "loss": 0.8088,
+      "step": 3368
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3589471578598022,
+      "learning_rate": 6.179294623994833e-06,
+      "loss": 0.8405,
+      "step": 3369
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3298689126968384,
+      "learning_rate": 6.177232264488958e-06,
+      "loss": 0.8219,
+      "step": 3370
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.4132862091064453,
+      "learning_rate": 6.175169692920653e-06,
+      "loss": 0.9079,
+      "step": 3371
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 4.319317817687988,
+      "learning_rate": 6.173106909661461e-06,
+      "loss": 0.9065,
+      "step": 3372
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2624664306640625,
+      "learning_rate": 6.171043915082967e-06,
+      "loss": 0.8149,
+      "step": 3373
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3070234060287476,
+      "learning_rate": 6.168980709556789e-06,
+      "loss": 0.8348,
+      "step": 3374
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3258695602416992,
+      "learning_rate": 6.1669172934545875e-06,
+      "loss": 0.8048,
+      "step": 3375
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.44307279586792,
+      "learning_rate": 6.164853667148059e-06,
+      "loss": 0.8842,
+      "step": 3376
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3151425123214722,
+      "learning_rate": 6.162789831008935e-06,
+      "loss": 0.8663,
+      "step": 3377
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5204044580459595,
+      "learning_rate": 6.160725785408991e-06,
+      "loss": 0.8255,
+      "step": 3378
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.391797661781311,
+      "learning_rate": 6.158661530720035e-06,
+      "loss": 0.8692,
+      "step": 3379
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.9435542225837708,
+      "learning_rate": 6.156597067313914e-06,
+      "loss": 1.037,
+      "step": 3380
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.5590109825134277,
+      "learning_rate": 6.154532395562515e-06,
+      "loss": 0.8289,
+      "step": 3381
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.700984537601471,
+      "learning_rate": 6.152467515837759e-06,
+      "loss": 1.0405,
+      "step": 3382
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.2765984535217285,
+      "learning_rate": 6.150402428511607e-06,
+      "loss": 0.8269,
+      "step": 3383
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3979624509811401,
+      "learning_rate": 6.1483371339560545e-06,
+      "loss": 0.8214,
+      "step": 3384
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7408748269081116,
+      "learning_rate": 6.146271632543142e-06,
+      "loss": 1.0157,
+      "step": 3385
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7344962358474731,
+      "learning_rate": 6.144205924644934e-06,
+      "loss": 1.0235,
+      "step": 3386
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3100961446762085,
+      "learning_rate": 6.1421400106335435e-06,
+      "loss": 0.7976,
+      "step": 3387
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6463127136230469,
+      "learning_rate": 6.140073890881117e-06,
+      "loss": 1.016,
+      "step": 3388
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3137173652648926,
+      "learning_rate": 6.138007565759838e-06,
+      "loss": 0.8401,
+      "step": 3389
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3597760200500488,
+      "learning_rate": 6.135941035641927e-06,
+      "loss": 0.7934,
+      "step": 3390
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8691713213920593,
+      "learning_rate": 6.133874300899641e-06,
+      "loss": 1.0317,
+      "step": 3391
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.3387244939804077,
+      "learning_rate": 6.1318073619052735e-06,
+      "loss": 0.8323,
+      "step": 3392
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.403876543045044,
+      "learning_rate": 6.129740219031156e-06,
+      "loss": 0.8505,
+      "step": 3393
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.356537938117981,
+      "learning_rate": 6.127672872649657e-06,
+      "loss": 0.8514,
+      "step": 3394
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7481333017349243,
+      "learning_rate": 6.12560532313318e-06,
+      "loss": 1.0462,
+      "step": 3395
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7296017408370972,
+      "learning_rate": 6.123537570854168e-06,
+      "loss": 1.0395,
+      "step": 3396
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.315605640411377,
+      "learning_rate": 6.121469616185096e-06,
+      "loss": 0.864,
+      "step": 3397
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4000250101089478,
+      "learning_rate": 6.119401459498479e-06,
+      "loss": 0.8644,
+      "step": 3398
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3762993812561035,
+      "learning_rate": 6.117333101166867e-06,
+      "loss": 0.8306,
+      "step": 3399
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7053033113479614,
+      "learning_rate": 6.115264541562847e-06,
+      "loss": 1.0388,
+      "step": 3400
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.350176453590393,
+      "learning_rate": 6.1131957810590425e-06,
+      "loss": 0.868,
+      "step": 3401
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.292807936668396,
+      "learning_rate": 6.1111268200281115e-06,
+      "loss": 0.8494,
+      "step": 3402
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2523484230041504,
+      "learning_rate": 6.1090576588427495e-06,
+      "loss": 0.813,
+      "step": 3403
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2803305387496948,
+      "learning_rate": 6.106988297875691e-06,
+      "loss": 0.8373,
+      "step": 3404
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3280142545700073,
+      "learning_rate": 6.104918737499699e-06,
+      "loss": 0.7893,
+      "step": 3405
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2900004386901855,
+      "learning_rate": 6.102848978087576e-06,
+      "loss": 0.83,
+      "step": 3406
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2500079870224,
+      "learning_rate": 6.100779020012167e-06,
+      "loss": 0.8368,
+      "step": 3407
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7538713216781616,
+      "learning_rate": 6.0987088636463406e-06,
+      "loss": 1.0263,
+      "step": 3408
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3551980257034302,
+      "learning_rate": 6.096638509363013e-06,
+      "loss": 0.8739,
+      "step": 3409
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.341577172279358,
+      "learning_rate": 6.0945679575351255e-06,
+      "loss": 0.8525,
+      "step": 3410
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5069959163665771,
+      "learning_rate": 6.092497208535662e-06,
+      "loss": 0.8384,
+      "step": 3411
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4761996269226074,
+      "learning_rate": 6.090426262737639e-06,
+      "loss": 0.9304,
+      "step": 3412
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3790476322174072,
+      "learning_rate": 6.088355120514113e-06,
+      "loss": 0.8377,
+      "step": 3413
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3916600942611694,
+      "learning_rate": 6.086283782238168e-06,
+      "loss": 0.8998,
+      "step": 3414
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4340887069702148,
+      "learning_rate": 6.084212248282928e-06,
+      "loss": 0.8129,
+      "step": 3415
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.293046236038208,
+      "learning_rate": 6.0821405190215535e-06,
+      "loss": 0.843,
+      "step": 3416
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4140983819961548,
+      "learning_rate": 6.080068594827238e-06,
+      "loss": 0.8603,
+      "step": 3417
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.361955165863037,
+      "learning_rate": 6.0779964760732095e-06,
+      "loss": 0.8344,
+      "step": 3418
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3747769594192505,
+      "learning_rate": 6.075924163132734e-06,
+      "loss": 0.7979,
+      "step": 3419
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.787880003452301,
+      "learning_rate": 6.073851656379107e-06,
+      "loss": 1.035,
+      "step": 3420
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6463401317596436,
+      "learning_rate": 6.071778956185666e-06,
+      "loss": 0.8251,
+      "step": 3421
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4804004430770874,
+      "learning_rate": 6.069706062925777e-06,
+      "loss": 0.8616,
+      "step": 3422
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4498074054718018,
+      "learning_rate": 6.067632976972848e-06,
+      "loss": 0.8398,
+      "step": 3423
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3441599607467651,
+      "learning_rate": 6.065559698700312e-06,
+      "loss": 0.8169,
+      "step": 3424
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.636210560798645,
+      "learning_rate": 6.063486228481645e-06,
+      "loss": 1.0174,
+      "step": 3425
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6382579803466797,
+      "learning_rate": 6.061412566690355e-06,
+      "loss": 1.0311,
+      "step": 3426
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3073716163635254,
+      "learning_rate": 6.059338713699983e-06,
+      "loss": 0.8215,
+      "step": 3427
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3787322044372559,
+      "learning_rate": 6.057264669884105e-06,
+      "loss": 0.872,
+      "step": 3428
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6547699570655823,
+      "learning_rate": 6.055190435616333e-06,
+      "loss": 1.0623,
+      "step": 3429
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.321929931640625,
+      "learning_rate": 6.053116011270311e-06,
+      "loss": 0.8726,
+      "step": 3430
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6489963531494141,
+      "learning_rate": 6.051041397219719e-06,
+      "loss": 1.0422,
+      "step": 3431
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3689433336257935,
+      "learning_rate": 6.048966593838272e-06,
+      "loss": 0.8438,
+      "step": 3432
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2622747421264648,
+      "learning_rate": 6.0468916014997156e-06,
+      "loss": 0.8048,
+      "step": 3433
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5004572868347168,
+      "learning_rate": 6.0448164205778315e-06,
+      "loss": 0.8989,
+      "step": 3434
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2914190292358398,
+      "learning_rate": 6.042741051446437e-06,
+      "loss": 0.8962,
+      "step": 3435
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.28981351852417,
+      "learning_rate": 6.040665494479381e-06,
+      "loss": 0.9483,
+      "step": 3436
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.251471757888794,
+      "learning_rate": 6.038589750050546e-06,
+      "loss": 0.8186,
+      "step": 3437
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7371683120727539,
+      "learning_rate": 6.036513818533852e-06,
+      "loss": 1.0128,
+      "step": 3438
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3631740808486938,
+      "learning_rate": 6.034437700303246e-06,
+      "loss": 0.7736,
+      "step": 3439
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6714648604393005,
+      "learning_rate": 6.0323613957327156e-06,
+      "loss": 1.0305,
+      "step": 3440
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2955166101455688,
+      "learning_rate": 6.0302849051962765e-06,
+      "loss": 0.8156,
+      "step": 3441
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4473786354064941,
+      "learning_rate": 6.028208229067983e-06,
+      "loss": 0.8824,
+      "step": 3442
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.6991379261016846,
+      "learning_rate": 6.026131367721919e-06,
+      "loss": 0.8218,
+      "step": 3443
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3496953248977661,
+      "learning_rate": 6.024054321532201e-06,
+      "loss": 0.8892,
+      "step": 3444
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2964987754821777,
+      "learning_rate": 6.021977090872983e-06,
+      "loss": 0.82,
+      "step": 3445
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.284592628479004,
+      "learning_rate": 6.0198996761184515e-06,
+      "loss": 0.8206,
+      "step": 3446
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3249496221542358,
+      "learning_rate": 6.017822077642821e-06,
+      "loss": 0.9051,
+      "step": 3447
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3280819654464722,
+      "learning_rate": 6.015744295820343e-06,
+      "loss": 0.8514,
+      "step": 3448
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2578198909759521,
+      "learning_rate": 6.013666331025303e-06,
+      "loss": 0.7871,
+      "step": 3449
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 2.6774086952209473,
+      "learning_rate": 6.011588183632019e-06,
+      "loss": 0.8349,
+      "step": 3450
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2893459796905518,
+      "learning_rate": 6.009509854014841e-06,
+      "loss": 0.876,
+      "step": 3451
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.297751545906067,
+      "learning_rate": 6.0074313425481504e-06,
+      "loss": 0.8674,
+      "step": 3452
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.280259370803833,
+      "learning_rate": 6.005352649606362e-06,
+      "loss": 0.8285,
+      "step": 3453
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.283220648765564,
+      "learning_rate": 6.003273775563927e-06,
+      "loss": 0.8344,
+      "step": 3454
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3379662036895752,
+      "learning_rate": 6.001194720795324e-06,
+      "loss": 0.9369,
+      "step": 3455
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.358204960823059,
+      "learning_rate": 5.999115485675069e-06,
+      "loss": 0.8328,
+      "step": 3456
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3193854093551636,
+      "learning_rate": 5.997036070577705e-06,
+      "loss": 0.8202,
+      "step": 3457
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.4021680355072021,
+      "learning_rate": 5.994956475877814e-06,
+      "loss": 0.8417,
+      "step": 3458
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.5006083250045776,
+      "learning_rate": 5.992876701950003e-06,
+      "loss": 0.8397,
+      "step": 3459
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2983436584472656,
+      "learning_rate": 5.990796749168919e-06,
+      "loss": 0.8296,
+      "step": 3460
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2915081977844238,
+      "learning_rate": 5.988716617909233e-06,
+      "loss": 0.868,
+      "step": 3461
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3860969543457031,
+      "learning_rate": 5.986636308545655e-06,
+      "loss": 0.9229,
+      "step": 3462
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.380147099494934,
+      "learning_rate": 5.984555821452923e-06,
+      "loss": 0.8425,
+      "step": 3463
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.838549017906189,
+      "learning_rate": 5.982475157005812e-06,
+      "loss": 1.037,
+      "step": 3464
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.7314153909683228,
+      "learning_rate": 5.9803943155791225e-06,
+      "loss": 1.037,
+      "step": 3465
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6554968953132629,
+      "learning_rate": 5.978313297547688e-06,
+      "loss": 1.0113,
+      "step": 3466
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3763246536254883,
+      "learning_rate": 5.9762321032863816e-06,
+      "loss": 0.8702,
+      "step": 3467
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3521801233291626,
+      "learning_rate": 5.974150733170096e-06,
+      "loss": 0.8446,
+      "step": 3468
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3299239873886108,
+      "learning_rate": 5.972069187573766e-06,
+      "loss": 0.8079,
+      "step": 3469
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3296985626220703,
+      "learning_rate": 5.969987466872354e-06,
+      "loss": 0.8479,
+      "step": 3470
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.2809885740280151,
+      "learning_rate": 5.967905571440849e-06,
+      "loss": 0.865,
+      "step": 3471
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 1.3928624391555786,
+      "learning_rate": 5.9658235016542806e-06,
+      "loss": 0.844,
+      "step": 3472
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.328762173652649,
+      "learning_rate": 5.963741257887704e-06,
+      "loss": 0.8091,
+      "step": 3473
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4836841821670532,
+      "learning_rate": 5.961658840516209e-06,
+      "loss": 0.895,
+      "step": 3474
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3463692665100098,
+      "learning_rate": 5.959576249914913e-06,
+      "loss": 0.7698,
+      "step": 3475
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5247139930725098,
+      "learning_rate": 5.957493486458965e-06,
+      "loss": 1.0503,
+      "step": 3476
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.367148518562317,
+      "learning_rate": 5.95541055052355e-06,
+      "loss": 0.8792,
+      "step": 3477
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4359627962112427,
+      "learning_rate": 5.953327442483878e-06,
+      "loss": 0.8308,
+      "step": 3478
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3952529430389404,
+      "learning_rate": 5.951244162715194e-06,
+      "loss": 0.8198,
+      "step": 3479
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7326706647872925,
+      "learning_rate": 5.949160711592773e-06,
+      "loss": 1.0279,
+      "step": 3480
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4419689178466797,
+      "learning_rate": 5.947077089491919e-06,
+      "loss": 0.8106,
+      "step": 3481
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4076647758483887,
+      "learning_rate": 5.9449932967879686e-06,
+      "loss": 0.8848,
+      "step": 3482
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7936524152755737,
+      "learning_rate": 5.9429093338562895e-06,
+      "loss": 1.0424,
+      "step": 3483
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.596879005432129,
+      "learning_rate": 5.94082520107228e-06,
+      "loss": 0.7842,
+      "step": 3484
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.660452127456665,
+      "learning_rate": 5.9387408988113656e-06,
+      "loss": 1.0403,
+      "step": 3485
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6252462267875671,
+      "learning_rate": 5.936656427449009e-06,
+      "loss": 1.0133,
+      "step": 3486
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3403382301330566,
+      "learning_rate": 5.934571787360696e-06,
+      "loss": 0.8337,
+      "step": 3487
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3704341650009155,
+      "learning_rate": 5.932486978921948e-06,
+      "loss": 0.8614,
+      "step": 3488
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3411246538162231,
+      "learning_rate": 5.930402002508315e-06,
+      "loss": 0.8082,
+      "step": 3489
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2620586156845093,
+      "learning_rate": 5.928316858495377e-06,
+      "loss": 0.8692,
+      "step": 3490
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3408715724945068,
+      "learning_rate": 5.926231547258743e-06,
+      "loss": 0.7726,
+      "step": 3491
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.002680778503418,
+      "learning_rate": 5.924146069174055e-06,
+      "loss": 1.0364,
+      "step": 3492
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5142691135406494,
+      "learning_rate": 5.922060424616985e-06,
+      "loss": 0.8181,
+      "step": 3493
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7294716238975525,
+      "learning_rate": 5.91997461396323e-06,
+      "loss": 1.0239,
+      "step": 3494
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4176379442214966,
+      "learning_rate": 5.9178886375885235e-06,
+      "loss": 0.8632,
+      "step": 3495
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6914899349212646,
+      "learning_rate": 5.915802495868624e-06,
+      "loss": 1.041,
+      "step": 3496
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2679376602172852,
+      "learning_rate": 5.913716189179324e-06,
+      "loss": 0.8439,
+      "step": 3497
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8243005871772766,
+      "learning_rate": 5.911629717896441e-06,
+      "loss": 1.0461,
+      "step": 3498
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3273820877075195,
+      "learning_rate": 5.909543082395826e-06,
+      "loss": 0.826,
+      "step": 3499
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3412779569625854,
+      "learning_rate": 5.907456283053357e-06,
+      "loss": 0.8513,
+      "step": 3500
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3808259963989258,
+      "learning_rate": 5.905369320244942e-06,
+      "loss": 0.8796,
+      "step": 3501
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.346027135848999,
+      "learning_rate": 5.903282194346522e-06,
+      "loss": 0.8417,
+      "step": 3502
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.7626582384109497,
+      "learning_rate": 5.9011949057340624e-06,
+      "loss": 0.8282,
+      "step": 3503
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2821601629257202,
+      "learning_rate": 5.899107454783558e-06,
+      "loss": 0.8181,
+      "step": 3504
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8804784417152405,
+      "learning_rate": 5.897019841871039e-06,
+      "loss": 1.0232,
+      "step": 3505
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3602901697158813,
+      "learning_rate": 5.894932067372556e-06,
+      "loss": 0.8311,
+      "step": 3506
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7414374947547913,
+      "learning_rate": 5.892844131664199e-06,
+      "loss": 1.0187,
+      "step": 3507
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2932366132736206,
+      "learning_rate": 5.890756035122076e-06,
+      "loss": 0.8948,
+      "step": 3508
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.683745265007019,
+      "learning_rate": 5.888667778122331e-06,
+      "loss": 1.0576,
+      "step": 3509
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6773361563682556,
+      "learning_rate": 5.886579361041136e-06,
+      "loss": 1.024,
+      "step": 3510
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.373472809791565,
+      "learning_rate": 5.884490784254691e-06,
+      "loss": 0.8412,
+      "step": 3511
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7002538442611694,
+      "learning_rate": 5.882402048139224e-06,
+      "loss": 1.0332,
+      "step": 3512
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4015532732009888,
+      "learning_rate": 5.880313153070992e-06,
+      "loss": 0.89,
+      "step": 3513
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.351362705230713,
+      "learning_rate": 5.878224099426283e-06,
+      "loss": 0.811,
+      "step": 3514
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3236969709396362,
+      "learning_rate": 5.87613488758141e-06,
+      "loss": 0.7852,
+      "step": 3515
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3049277067184448,
+      "learning_rate": 5.874045517912717e-06,
+      "loss": 0.8114,
+      "step": 3516
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.324050784111023,
+      "learning_rate": 5.871955990796575e-06,
+      "loss": 0.8756,
+      "step": 3517
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3592160940170288,
+      "learning_rate": 5.869866306609385e-06,
+      "loss": 0.8088,
+      "step": 3518
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2463502883911133,
+      "learning_rate": 5.867776465727574e-06,
+      "loss": 0.8437,
+      "step": 3519
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2984036207199097,
+      "learning_rate": 5.865686468527599e-06,
+      "loss": 0.8363,
+      "step": 3520
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3316372632980347,
+      "learning_rate": 5.863596315385947e-06,
+      "loss": 0.889,
+      "step": 3521
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8308576941490173,
+      "learning_rate": 5.861506006679125e-06,
+      "loss": 1.033,
+      "step": 3522
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7312998175621033,
+      "learning_rate": 5.859415542783678e-06,
+      "loss": 1.0207,
+      "step": 3523
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3431591987609863,
+      "learning_rate": 5.857324924076175e-06,
+      "loss": 0.7941,
+      "step": 3524
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.465315818786621,
+      "learning_rate": 5.8552341509332125e-06,
+      "loss": 0.8509,
+      "step": 3525
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.355861783027649,
+      "learning_rate": 5.8531432237314135e-06,
+      "loss": 0.8047,
+      "step": 3526
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3827518224716187,
+      "learning_rate": 5.85105214284743e-06,
+      "loss": 0.8313,
+      "step": 3527
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8654420971870422,
+      "learning_rate": 5.8489609086579425e-06,
+      "loss": 1.0335,
+      "step": 3528
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2836018800735474,
+      "learning_rate": 5.846869521539659e-06,
+      "loss": 0.8568,
+      "step": 3529
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6745295524597168,
+      "learning_rate": 5.844777981869315e-06,
+      "loss": 1.0445,
+      "step": 3530
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.2716827392578125,
+      "learning_rate": 5.842686290023671e-06,
+      "loss": 0.8296,
+      "step": 3531
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4249364137649536,
+      "learning_rate": 5.840594446379518e-06,
+      "loss": 0.8895,
+      "step": 3532
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.5207853317260742,
+      "learning_rate": 5.838502451313675e-06,
+      "loss": 0.8438,
+      "step": 3533
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3234974145889282,
+      "learning_rate": 5.836410305202983e-06,
+      "loss": 0.8593,
+      "step": 3534
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3436635732650757,
+      "learning_rate": 5.834318008424316e-06,
+      "loss": 0.8201,
+      "step": 3535
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3479371070861816,
+      "learning_rate": 5.832225561354573e-06,
+      "loss": 0.8172,
+      "step": 3536
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3427575826644897,
+      "learning_rate": 5.8301329643706794e-06,
+      "loss": 0.8341,
+      "step": 3537
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3457437753677368,
+      "learning_rate": 5.828040217849589e-06,
+      "loss": 0.9005,
+      "step": 3538
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.924049973487854,
+      "learning_rate": 5.825947322168281e-06,
+      "loss": 1.024,
+      "step": 3539
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3646290302276611,
+      "learning_rate": 5.823854277703761e-06,
+      "loss": 0.8327,
+      "step": 3540
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.368123173713684,
+      "learning_rate": 5.821761084833064e-06,
+      "loss": 0.8014,
+      "step": 3541
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7127262353897095,
+      "learning_rate": 5.819667743933251e-06,
+      "loss": 1.0352,
+      "step": 3542
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4121381044387817,
+      "learning_rate": 5.8175742553814065e-06,
+      "loss": 0.792,
+      "step": 3543
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3485805988311768,
+      "learning_rate": 5.815480619554646e-06,
+      "loss": 0.8376,
+      "step": 3544
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.526260495185852,
+      "learning_rate": 5.81338683683011e-06,
+      "loss": 0.8262,
+      "step": 3545
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.3742763996124268,
+      "learning_rate": 5.811292907584963e-06,
+      "loss": 0.8677,
+      "step": 3546
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4578189849853516,
+      "learning_rate": 5.809198832196401e-06,
+      "loss": 0.8461,
+      "step": 3547
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 1.4834810495376587,
+      "learning_rate": 5.80710461104164e-06,
+      "loss": 0.899,
+      "step": 3548
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.269546627998352,
+      "learning_rate": 5.805010244497928e-06,
+      "loss": 0.8567,
+      "step": 3549
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3742201328277588,
+      "learning_rate": 5.802915732942536e-06,
+      "loss": 0.8429,
+      "step": 3550
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4553416967391968,
+      "learning_rate": 5.800821076752761e-06,
+      "loss": 0.8127,
+      "step": 3551
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9723564982414246,
+      "learning_rate": 5.798726276305927e-06,
+      "loss": 1.0396,
+      "step": 3552
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 2.0938947200775146,
+      "learning_rate": 5.796631331979385e-06,
+      "loss": 0.8989,
+      "step": 3553
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4659239053726196,
+      "learning_rate": 5.794536244150512e-06,
+      "loss": 0.8194,
+      "step": 3554
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.623982310295105,
+      "learning_rate": 5.792441013196706e-06,
+      "loss": 0.8639,
+      "step": 3555
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.32310152053833,
+      "learning_rate": 5.790345639495396e-06,
+      "loss": 0.8599,
+      "step": 3556
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7368608713150024,
+      "learning_rate": 5.788250123424037e-06,
+      "loss": 1.0469,
+      "step": 3557
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7489502429962158,
+      "learning_rate": 5.786154465360107e-06,
+      "loss": 1.029,
+      "step": 3558
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4041391611099243,
+      "learning_rate": 5.784058665681108e-06,
+      "loss": 0.8245,
+      "step": 3559
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3604360818862915,
+      "learning_rate": 5.781962724764572e-06,
+      "loss": 0.8041,
+      "step": 3560
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4064821004867554,
+      "learning_rate": 5.779866642988055e-06,
+      "loss": 0.8117,
+      "step": 3561
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3647047281265259,
+      "learning_rate": 5.777770420729135e-06,
+      "loss": 0.8351,
+      "step": 3562
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6933575868606567,
+      "learning_rate": 5.775674058365423e-06,
+      "loss": 1.0172,
+      "step": 3563
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7592675685882568,
+      "learning_rate": 5.773577556274543e-06,
+      "loss": 0.8667,
+      "step": 3564
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6879668235778809,
+      "learning_rate": 5.771480914834158e-06,
+      "loss": 1.0486,
+      "step": 3565
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4414547681808472,
+      "learning_rate": 5.769384134421945e-06,
+      "loss": 0.8459,
+      "step": 3566
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.322039008140564,
+      "learning_rate": 5.767287215415613e-06,
+      "loss": 0.8167,
+      "step": 3567
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6900490522384644,
+      "learning_rate": 5.765190158192895e-06,
+      "loss": 1.0374,
+      "step": 3568
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6373037099838257,
+      "learning_rate": 5.7630929631315424e-06,
+      "loss": 1.0186,
+      "step": 3569
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6209146976470947,
+      "learning_rate": 5.76099563060934e-06,
+      "loss": 1.0147,
+      "step": 3570
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2745126485824585,
+      "learning_rate": 5.758898161004095e-06,
+      "loss": 0.8148,
+      "step": 3571
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7550437450408936,
+      "learning_rate": 5.756800554693636e-06,
+      "loss": 0.8347,
+      "step": 3572
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.712481677532196,
+      "learning_rate": 5.754702812055819e-06,
+      "loss": 1.0199,
+      "step": 3573
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.6060924530029297,
+      "learning_rate": 5.752604933468523e-06,
+      "loss": 0.8072,
+      "step": 3574
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7209165096282959,
+      "learning_rate": 5.750506919309652e-06,
+      "loss": 1.0236,
+      "step": 3575
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.313083291053772,
+      "learning_rate": 5.748408769957137e-06,
+      "loss": 0.8235,
+      "step": 3576
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2852624654769897,
+      "learning_rate": 5.746310485788931e-06,
+      "loss": 0.8598,
+      "step": 3577
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.397993564605713,
+      "learning_rate": 5.744212067183008e-06,
+      "loss": 0.8843,
+      "step": 3578
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3954493999481201,
+      "learning_rate": 5.742113514517375e-06,
+      "loss": 0.8629,
+      "step": 3579
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6775407791137695,
+      "learning_rate": 5.740014828170053e-06,
+      "loss": 1.0422,
+      "step": 3580
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.259825348854065,
+      "learning_rate": 5.737916008519093e-06,
+      "loss": 0.7701,
+      "step": 3581
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.465469241142273,
+      "learning_rate": 5.735817055942572e-06,
+      "loss": 0.8569,
+      "step": 3582
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4610146284103394,
+      "learning_rate": 5.733717970818584e-06,
+      "loss": 0.9033,
+      "step": 3583
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.37160062789917,
+      "learning_rate": 5.731618753525252e-06,
+      "loss": 0.8502,
+      "step": 3584
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3237327337265015,
+      "learning_rate": 5.729519404440722e-06,
+      "loss": 0.8372,
+      "step": 3585
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6208661794662476,
+      "learning_rate": 5.727419923943163e-06,
+      "loss": 1.0403,
+      "step": 3586
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2761006355285645,
+      "learning_rate": 5.7253203124107695e-06,
+      "loss": 0.8898,
+      "step": 3587
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6147862076759338,
+      "learning_rate": 5.7232205702217545e-06,
+      "loss": 1.0256,
+      "step": 3588
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6279263496398926,
+      "learning_rate": 5.721120697754361e-06,
+      "loss": 1.0384,
+      "step": 3589
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3230286836624146,
+      "learning_rate": 5.719020695386851e-06,
+      "loss": 0.797,
+      "step": 3590
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.318723440170288,
+      "learning_rate": 5.716920563497514e-06,
+      "loss": 0.8472,
+      "step": 3591
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3286292552947998,
+      "learning_rate": 5.714820302464656e-06,
+      "loss": 0.8547,
+      "step": 3592
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.398119330406189,
+      "learning_rate": 5.712719912666615e-06,
+      "loss": 0.9271,
+      "step": 3593
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.7316741943359375,
+      "learning_rate": 5.710619394481745e-06,
+      "loss": 0.8054,
+      "step": 3594
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2913641929626465,
+      "learning_rate": 5.708518748288425e-06,
+      "loss": 0.8281,
+      "step": 3595
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2941827774047852,
+      "learning_rate": 5.706417974465061e-06,
+      "loss": 0.8483,
+      "step": 3596
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7892493605613708,
+      "learning_rate": 5.7043170733900775e-06,
+      "loss": 1.0422,
+      "step": 3597
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7052578330039978,
+      "learning_rate": 5.702216045441924e-06,
+      "loss": 1.0549,
+      "step": 3598
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3368233442306519,
+      "learning_rate": 5.700114890999071e-06,
+      "loss": 0.8235,
+      "step": 3599
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6304470896720886,
+      "learning_rate": 5.698013610440012e-06,
+      "loss": 1.0334,
+      "step": 3600
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.353500485420227,
+      "learning_rate": 5.695912204143268e-06,
+      "loss": 0.8608,
+      "step": 3601
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4156051874160767,
+      "learning_rate": 5.6938106724873745e-06,
+      "loss": 0.7996,
+      "step": 3602
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.38575279712677,
+      "learning_rate": 5.691709015850897e-06,
+      "loss": 0.7978,
+      "step": 3603
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2951754331588745,
+      "learning_rate": 5.689607234612418e-06,
+      "loss": 0.8522,
+      "step": 3604
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9363430142402649,
+      "learning_rate": 5.687505329150548e-06,
+      "loss": 1.0215,
+      "step": 3605
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4136312007904053,
+      "learning_rate": 5.685403299843914e-06,
+      "loss": 0.8462,
+      "step": 3606
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3226245641708374,
+      "learning_rate": 5.68330114707117e-06,
+      "loss": 0.8664,
+      "step": 3607
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4125845432281494,
+      "learning_rate": 5.681198871210988e-06,
+      "loss": 0.8765,
+      "step": 3608
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6958194971084595,
+      "learning_rate": 5.679096472642066e-06,
+      "loss": 1.0437,
+      "step": 3609
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2911268472671509,
+      "learning_rate": 5.676993951743124e-06,
+      "loss": 0.7968,
+      "step": 3610
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2995867729187012,
+      "learning_rate": 5.6748913088929005e-06,
+      "loss": 0.8161,
+      "step": 3611
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3301914930343628,
+      "learning_rate": 5.6727885444701584e-06,
+      "loss": 0.7744,
+      "step": 3612
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2912594079971313,
+      "learning_rate": 5.670685658853682e-06,
+      "loss": 0.8804,
+      "step": 3613
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.277772307395935,
+      "learning_rate": 5.6685826524222785e-06,
+      "loss": 0.8197,
+      "step": 3614
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.310802698135376,
+      "learning_rate": 5.666479525554777e-06,
+      "loss": 0.854,
+      "step": 3615
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3218085765838623,
+      "learning_rate": 5.664376278630025e-06,
+      "loss": 0.8333,
+      "step": 3616
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2771422863006592,
+      "learning_rate": 5.662272912026894e-06,
+      "loss": 0.8402,
+      "step": 3617
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.2747981548309326,
+      "learning_rate": 5.66016942612428e-06,
+      "loss": 0.8368,
+      "step": 3618
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8045984506607056,
+      "learning_rate": 5.658065821301093e-06,
+      "loss": 1.0426,
+      "step": 3619
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3061320781707764,
+      "learning_rate": 5.6559620979362736e-06,
+      "loss": 0.8013,
+      "step": 3620
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.330458402633667,
+      "learning_rate": 5.653858256408775e-06,
+      "loss": 0.8212,
+      "step": 3621
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.8157769441604614,
+      "learning_rate": 5.651754297097577e-06,
+      "loss": 0.9181,
+      "step": 3622
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.4760971069335938,
+      "learning_rate": 5.64965022038168e-06,
+      "loss": 0.792,
+      "step": 3623
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 1.3216371536254883,
+      "learning_rate": 5.647546026640106e-06,
+      "loss": 0.7829,
+      "step": 3624
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6248553395271301,
+      "learning_rate": 5.6454417162518935e-06,
+      "loss": 1.01,
+      "step": 3625
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6177812814712524,
+      "learning_rate": 5.6433372895961095e-06,
+      "loss": 1.0498,
+      "step": 3626
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6041745543479919,
+      "learning_rate": 5.641232747051834e-06,
+      "loss": 1.0188,
+      "step": 3627
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3069376945495605,
+      "learning_rate": 5.639128088998175e-06,
+      "loss": 0.8115,
+      "step": 3628
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6171747446060181,
+      "learning_rate": 5.637023315814257e-06,
+      "loss": 0.9982,
+      "step": 3629
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3531867265701294,
+      "learning_rate": 5.634918427879225e-06,
+      "loss": 0.8951,
+      "step": 3630
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6361227631568909,
+      "learning_rate": 5.632813425572247e-06,
+      "loss": 1.0237,
+      "step": 3631
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3402373790740967,
+      "learning_rate": 5.630708309272513e-06,
+      "loss": 0.8929,
+      "step": 3632
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.392063021659851,
+      "learning_rate": 5.62860307935923e-06,
+      "loss": 0.8221,
+      "step": 3633
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3576931953430176,
+      "learning_rate": 5.626497736211622e-06,
+      "loss": 0.8483,
+      "step": 3634
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2520755529403687,
+      "learning_rate": 5.624392280208943e-06,
+      "loss": 0.8087,
+      "step": 3635
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2964617013931274,
+      "learning_rate": 5.622286711730461e-06,
+      "loss": 0.8591,
+      "step": 3636
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2535372972488403,
+      "learning_rate": 5.620181031155467e-06,
+      "loss": 0.8292,
+      "step": 3637
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6736599206924438,
+      "learning_rate": 5.618075238863269e-06,
+      "loss": 1.0422,
+      "step": 3638
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3792369365692139,
+      "learning_rate": 5.615969335233197e-06,
+      "loss": 0.843,
+      "step": 3639
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3498426675796509,
+      "learning_rate": 5.613863320644601e-06,
+      "loss": 0.8592,
+      "step": 3640
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3033041954040527,
+      "learning_rate": 5.611757195476852e-06,
+      "loss": 0.816,
+      "step": 3641
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.307948112487793,
+      "learning_rate": 5.609650960109338e-06,
+      "loss": 0.8895,
+      "step": 3642
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2746773958206177,
+      "learning_rate": 5.607544614921472e-06,
+      "loss": 0.8863,
+      "step": 3643
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.442594051361084,
+      "learning_rate": 5.60543816029268e-06,
+      "loss": 0.8531,
+      "step": 3644
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6473795175552368,
+      "learning_rate": 5.603331596602412e-06,
+      "loss": 1.0258,
+      "step": 3645
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.437896966934204,
+      "learning_rate": 5.601224924230139e-06,
+      "loss": 0.8468,
+      "step": 3646
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2501376867294312,
+      "learning_rate": 5.5991181435553476e-06,
+      "loss": 0.8566,
+      "step": 3647
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4065022468566895,
+      "learning_rate": 5.5970112549575465e-06,
+      "loss": 0.8855,
+      "step": 3648
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2237094640731812,
+      "learning_rate": 5.594904258816262e-06,
+      "loss": 0.8126,
+      "step": 3649
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4009263515472412,
+      "learning_rate": 5.592797155511043e-06,
+      "loss": 0.8543,
+      "step": 3650
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3468719720840454,
+      "learning_rate": 5.590689945421452e-06,
+      "loss": 0.8536,
+      "step": 3651
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3408335447311401,
+      "learning_rate": 5.58858262892708e-06,
+      "loss": 0.8378,
+      "step": 3652
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.269748330116272,
+      "learning_rate": 5.586475206407527e-06,
+      "loss": 0.8583,
+      "step": 3653
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.721798300743103,
+      "learning_rate": 5.5843676782424176e-06,
+      "loss": 1.053,
+      "step": 3654
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.261269450187683,
+      "learning_rate": 5.5822600448113926e-06,
+      "loss": 0.8752,
+      "step": 3655
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3836113214492798,
+      "learning_rate": 5.5801523064941175e-06,
+      "loss": 0.8117,
+      "step": 3656
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6971721053123474,
+      "learning_rate": 5.578044463670271e-06,
+      "loss": 1.0305,
+      "step": 3657
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6556771397590637,
+      "learning_rate": 5.575936516719552e-06,
+      "loss": 1.0507,
+      "step": 3658
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3241426944732666,
+      "learning_rate": 5.573828466021679e-06,
+      "loss": 0.8971,
+      "step": 3659
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5056077241897583,
+      "learning_rate": 5.5717203119563875e-06,
+      "loss": 0.8485,
+      "step": 3660
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6969035267829895,
+      "learning_rate": 5.569612054903436e-06,
+      "loss": 1.0153,
+      "step": 3661
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3310418128967285,
+      "learning_rate": 5.567503695242595e-06,
+      "loss": 0.8377,
+      "step": 3662
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7301118969917297,
+      "learning_rate": 5.5653952333536575e-06,
+      "loss": 1.0455,
+      "step": 3663
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6424595713615417,
+      "learning_rate": 5.563286669616435e-06,
+      "loss": 1.0139,
+      "step": 3664
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6066001653671265,
+      "learning_rate": 5.561178004410758e-06,
+      "loss": 1.0258,
+      "step": 3665
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6516397595405579,
+      "learning_rate": 5.559069238116472e-06,
+      "loss": 1.0361,
+      "step": 3666
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4296576976776123,
+      "learning_rate": 5.556960371113445e-06,
+      "loss": 0.7381,
+      "step": 3667
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3629016876220703,
+      "learning_rate": 5.5548514037815556e-06,
+      "loss": 0.847,
+      "step": 3668
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.238357424736023,
+      "learning_rate": 5.552742336500711e-06,
+      "loss": 0.866,
+      "step": 3669
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.426792025566101,
+      "learning_rate": 5.55063316965083e-06,
+      "loss": 0.8351,
+      "step": 3670
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.236164927482605,
+      "learning_rate": 5.5485239036118495e-06,
+      "loss": 0.8083,
+      "step": 3671
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.358920693397522,
+      "learning_rate": 5.546414538763725e-06,
+      "loss": 0.8542,
+      "step": 3672
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2971534729003906,
+      "learning_rate": 5.54430507548643e-06,
+      "loss": 0.8498,
+      "step": 3673
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2511790990829468,
+      "learning_rate": 5.542195514159956e-06,
+      "loss": 0.8038,
+      "step": 3674
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1972150802612305,
+      "learning_rate": 5.540085855164312e-06,
+      "loss": 0.8226,
+      "step": 3675
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.9039009809494019,
+      "learning_rate": 5.5379760988795265e-06,
+      "loss": 1.0089,
+      "step": 3676
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3186486959457397,
+      "learning_rate": 5.5358662456856385e-06,
+      "loss": 0.8432,
+      "step": 3677
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4438211917877197,
+      "learning_rate": 5.533756295962715e-06,
+      "loss": 0.8296,
+      "step": 3678
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4684638977050781,
+      "learning_rate": 5.531646250090829e-06,
+      "loss": 0.8551,
+      "step": 3679
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3856970071792603,
+      "learning_rate": 5.5295361084500806e-06,
+      "loss": 0.8219,
+      "step": 3680
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.633507490158081,
+      "learning_rate": 5.527425871420584e-06,
+      "loss": 0.8292,
+      "step": 3681
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7369523644447327,
+      "learning_rate": 5.5253155393824676e-06,
+      "loss": 1.0348,
+      "step": 3682
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.470842719078064,
+      "learning_rate": 5.523205112715878e-06,
+      "loss": 0.9117,
+      "step": 3683
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3368284702301025,
+      "learning_rate": 5.521094591800983e-06,
+      "loss": 0.8112,
+      "step": 3684
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2587209939956665,
+      "learning_rate": 5.5189839770179645e-06,
+      "loss": 0.817,
+      "step": 3685
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.354111909866333,
+      "learning_rate": 5.5168732687470175e-06,
+      "loss": 0.8112,
+      "step": 3686
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2850233316421509,
+      "learning_rate": 5.51476246736836e-06,
+      "loss": 0.8515,
+      "step": 3687
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.272871732711792,
+      "learning_rate": 5.512651573262223e-06,
+      "loss": 0.8664,
+      "step": 3688
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.4370983839035034,
+      "learning_rate": 5.510540586808857e-06,
+      "loss": 0.821,
+      "step": 3689
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.363971471786499,
+      "learning_rate": 5.508429508388528e-06,
+      "loss": 0.878,
+      "step": 3690
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.294964075088501,
+      "learning_rate": 5.506318338381516e-06,
+      "loss": 0.8237,
+      "step": 3691
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2917009592056274,
+      "learning_rate": 5.50420707716812e-06,
+      "loss": 0.8815,
+      "step": 3692
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7435228228569031,
+      "learning_rate": 5.502095725128656e-06,
+      "loss": 1.0329,
+      "step": 3693
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.312154769897461,
+      "learning_rate": 5.499984282643456e-06,
+      "loss": 0.7948,
+      "step": 3694
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.2558175325393677,
+      "learning_rate": 5.497872750092866e-06,
+      "loss": 0.84,
+      "step": 3695
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.0843679904937744,
+      "learning_rate": 5.495761127857251e-06,
+      "loss": 0.8244,
+      "step": 3696
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3311784267425537,
+      "learning_rate": 5.493649416316991e-06,
+      "loss": 0.871,
+      "step": 3697
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7049567699432373,
+      "learning_rate": 5.491537615852483e-06,
+      "loss": 1.0203,
+      "step": 3698
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.3249729871749878,
+      "learning_rate": 5.489425726844138e-06,
+      "loss": 0.8353,
+      "step": 3699
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.310448408126831,
+      "learning_rate": 5.487313749672385e-06,
+      "loss": 0.8293,
+      "step": 3700
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6315441131591797,
+      "learning_rate": 5.485201684717667e-06,
+      "loss": 1.0329,
+      "step": 3701
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3548450469970703,
+      "learning_rate": 5.483089532360444e-06,
+      "loss": 0.8967,
+      "step": 3702
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3625388145446777,
+      "learning_rate": 5.480977292981192e-06,
+      "loss": 0.7972,
+      "step": 3703
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.401428461074829,
+      "learning_rate": 5.4788649669604044e-06,
+      "loss": 0.8168,
+      "step": 3704
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3701895475387573,
+      "learning_rate": 5.476752554678584e-06,
+      "loss": 0.8665,
+      "step": 3705
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.335404872894287,
+      "learning_rate": 5.474640056516256e-06,
+      "loss": 0.7929,
+      "step": 3706
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2477437257766724,
+      "learning_rate": 5.472527472853957e-06,
+      "loss": 0.8882,
+      "step": 3707
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.292777419090271,
+      "learning_rate": 5.470414804072243e-06,
+      "loss": 0.8329,
+      "step": 3708
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2894599437713623,
+      "learning_rate": 5.468302050551678e-06,
+      "loss": 0.8469,
+      "step": 3709
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2940865755081177,
+      "learning_rate": 5.466189212672849e-06,
+      "loss": 0.844,
+      "step": 3710
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7607514262199402,
+      "learning_rate": 5.464076290816354e-06,
+      "loss": 1.0401,
+      "step": 3711
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3446763753890991,
+      "learning_rate": 5.461963285362809e-06,
+      "loss": 0.8466,
+      "step": 3712
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6290137767791748,
+      "learning_rate": 5.459850196692843e-06,
+      "loss": 1.0307,
+      "step": 3713
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6296639442443848,
+      "learning_rate": 5.457737025187098e-06,
+      "loss": 1.0294,
+      "step": 3714
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6466906070709229,
+      "learning_rate": 5.455623771226234e-06,
+      "loss": 1.0105,
+      "step": 3715
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3536709547042847,
+      "learning_rate": 5.453510435190925e-06,
+      "loss": 0.8261,
+      "step": 3716
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3959273099899292,
+      "learning_rate": 5.451397017461862e-06,
+      "loss": 0.8305,
+      "step": 3717
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3524836301803589,
+      "learning_rate": 5.4492835184197465e-06,
+      "loss": 0.8903,
+      "step": 3718
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4383624792099,
+      "learning_rate": 5.447169938445297e-06,
+      "loss": 0.8113,
+      "step": 3719
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.363406777381897,
+      "learning_rate": 5.445056277919247e-06,
+      "loss": 0.8711,
+      "step": 3720
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3324573040008545,
+      "learning_rate": 5.442942537222342e-06,
+      "loss": 0.8805,
+      "step": 3721
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7934686541557312,
+      "learning_rate": 5.440828716735347e-06,
+      "loss": 1.0456,
+      "step": 3722
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8095865249633789,
+      "learning_rate": 5.438714816839035e-06,
+      "loss": 0.9827,
+      "step": 3723
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2644269466400146,
+      "learning_rate": 5.436600837914195e-06,
+      "loss": 0.8957,
+      "step": 3724
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3248255252838135,
+      "learning_rate": 5.4344867803416365e-06,
+      "loss": 0.8172,
+      "step": 3725
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6560685634613037,
+      "learning_rate": 5.4323726445021755e-06,
+      "loss": 1.0274,
+      "step": 3726
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4481163024902344,
+      "learning_rate": 5.430258430776645e-06,
+      "loss": 0.7884,
+      "step": 3727
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.8855162858963013,
+      "learning_rate": 5.428144139545892e-06,
+      "loss": 0.8563,
+      "step": 3728
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3654735088348389,
+      "learning_rate": 5.426029771190777e-06,
+      "loss": 0.8557,
+      "step": 3729
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3649619817733765,
+      "learning_rate": 5.4239153260921764e-06,
+      "loss": 0.8899,
+      "step": 3730
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3197462558746338,
+      "learning_rate": 5.421800804630978e-06,
+      "loss": 0.8868,
+      "step": 3731
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.613131046295166,
+      "learning_rate": 5.419686207188084e-06,
+      "loss": 0.8376,
+      "step": 3732
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2690292596817017,
+      "learning_rate": 5.417571534144409e-06,
+      "loss": 0.8314,
+      "step": 3733
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3917152881622314,
+      "learning_rate": 5.415456785880885e-06,
+      "loss": 0.7991,
+      "step": 3734
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.337519645690918,
+      "learning_rate": 5.4133419627784545e-06,
+      "loss": 0.8526,
+      "step": 3735
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8356655836105347,
+      "learning_rate": 5.411227065218074e-06,
+      "loss": 1.0066,
+      "step": 3736
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4522819519042969,
+      "learning_rate": 5.409112093580714e-06,
+      "loss": 0.8214,
+      "step": 3737
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3180354833602905,
+      "learning_rate": 5.406997048247357e-06,
+      "loss": 0.7729,
+      "step": 3738
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2769821882247925,
+      "learning_rate": 5.404881929599001e-06,
+      "loss": 0.8052,
+      "step": 3739
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.670427143573761,
+      "learning_rate": 5.402766738016654e-06,
+      "loss": 1.039,
+      "step": 3740
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3691045045852661,
+      "learning_rate": 5.400651473881342e-06,
+      "loss": 0.8151,
+      "step": 3741
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3110222816467285,
+      "learning_rate": 5.398536137574099e-06,
+      "loss": 0.8215,
+      "step": 3742
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2451181411743164,
+      "learning_rate": 5.396420729475974e-06,
+      "loss": 0.8419,
+      "step": 3743
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6902169585227966,
+      "learning_rate": 5.394305249968029e-06,
+      "loss": 1.0207,
+      "step": 3744
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6813607215881348,
+      "learning_rate": 5.39218969943134e-06,
+      "loss": 1.0318,
+      "step": 3745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3421639204025269,
+      "learning_rate": 5.390074078246997e-06,
+      "loss": 0.8242,
+      "step": 3746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4017010927200317,
+      "learning_rate": 5.387958386796096e-06,
+      "loss": 0.837,
+      "step": 3747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2861559391021729,
+      "learning_rate": 5.385842625459752e-06,
+      "loss": 0.8096,
+      "step": 3748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2766779661178589,
+      "learning_rate": 5.383726794619091e-06,
+      "loss": 0.8844,
+      "step": 3749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2742575407028198,
+      "learning_rate": 5.381610894655251e-06,
+      "loss": 0.8404,
+      "step": 3750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2834339141845703,
+      "learning_rate": 5.379494925949383e-06,
+      "loss": 0.8856,
+      "step": 3751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3430041074752808,
+      "learning_rate": 5.37737888888265e-06,
+      "loss": 0.83,
+      "step": 3752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9709500074386597,
+      "learning_rate": 5.375262783836226e-06,
+      "loss": 1.0401,
+      "step": 3753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.357181429862976,
+      "learning_rate": 5.3731466111913e-06,
+      "loss": 0.8648,
+      "step": 3754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3622322082519531,
+      "learning_rate": 5.371030371329073e-06,
+      "loss": 0.8428,
+      "step": 3755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3352361917495728,
+      "learning_rate": 5.368914064630755e-06,
+      "loss": 0.8594,
+      "step": 3756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.5551927089691162,
+      "learning_rate": 5.366797691477568e-06,
+      "loss": 0.8421,
+      "step": 3757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.383959412574768,
+      "learning_rate": 5.364681252250754e-06,
+      "loss": 0.811,
+      "step": 3758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3082183599472046,
+      "learning_rate": 5.362564747331556e-06,
+      "loss": 0.8326,
+      "step": 3759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3905582427978516,
+      "learning_rate": 5.360448177101234e-06,
+      "loss": 0.8809,
+      "step": 3760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.751606285572052,
+      "learning_rate": 5.358331541941062e-06,
+      "loss": 1.0269,
+      "step": 3761
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2701141834259033,
+      "learning_rate": 5.356214842232319e-06,
+      "loss": 0.8211,
+      "step": 3762
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3759490251541138,
+      "learning_rate": 5.354098078356304e-06,
+      "loss": 0.8222,
+      "step": 3763
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3108404874801636,
+      "learning_rate": 5.3519812506943215e-06,
+      "loss": 0.9119,
+      "step": 3764
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6374039053916931,
+      "learning_rate": 5.349864359627689e-06,
+      "loss": 1.0567,
+      "step": 3765
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3655786514282227,
+      "learning_rate": 5.347747405537735e-06,
+      "loss": 0.7994,
+      "step": 3766
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.626589834690094,
+      "learning_rate": 5.345630388805802e-06,
+      "loss": 1.0248,
+      "step": 3767
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4753987789154053,
+      "learning_rate": 5.343513309813239e-06,
+      "loss": 0.8034,
+      "step": 3768
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2402867078781128,
+      "learning_rate": 5.341396168941415e-06,
+      "loss": 0.8164,
+      "step": 3769
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.2574771642684937,
+      "learning_rate": 5.339278966571697e-06,
+      "loss": 0.8681,
+      "step": 3770
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3564869165420532,
+      "learning_rate": 5.337161703085475e-06,
+      "loss": 0.8362,
+      "step": 3771
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6857696771621704,
+      "learning_rate": 5.3350443788641434e-06,
+      "loss": 0.994,
+      "step": 3772
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4206448793411255,
+      "learning_rate": 5.332926994289109e-06,
+      "loss": 0.8908,
+      "step": 3773
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6902830004692078,
+      "learning_rate": 5.330809549741793e-06,
+      "loss": 1.0231,
+      "step": 3774
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.4200329780578613,
+      "learning_rate": 5.32869204560362e-06,
+      "loss": 0.8622,
+      "step": 3775
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6081632375717163,
+      "learning_rate": 5.326574482256034e-06,
+      "loss": 1.0261,
+      "step": 3776
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 1.3439714908599854,
+      "learning_rate": 5.324456860080481e-06,
+      "loss": 0.9066,
+      "step": 3777
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7308283448219299,
+      "learning_rate": 5.322339179458427e-06,
+      "loss": 1.0098,
+      "step": 3778
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7095532417297363,
+      "learning_rate": 5.320221440771337e-06,
+      "loss": 1.0315,
+      "step": 3779
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4999314546585083,
+      "learning_rate": 5.318103644400699e-06,
+      "loss": 0.81,
+      "step": 3780
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3513941764831543,
+      "learning_rate": 5.315985790728002e-06,
+      "loss": 0.8423,
+      "step": 3781
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6359423995018005,
+      "learning_rate": 5.313867880134749e-06,
+      "loss": 1.0141,
+      "step": 3782
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.6970624923706055,
+      "learning_rate": 5.3117499130024534e-06,
+      "loss": 0.8251,
+      "step": 3783
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6595423817634583,
+      "learning_rate": 5.3096318897126375e-06,
+      "loss": 1.0267,
+      "step": 3784
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3494014739990234,
+      "learning_rate": 5.307513810646836e-06,
+      "loss": 0.8068,
+      "step": 3785
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4564857482910156,
+      "learning_rate": 5.3053956761865895e-06,
+      "loss": 0.8758,
+      "step": 3786
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2863088846206665,
+      "learning_rate": 5.303277486713454e-06,
+      "loss": 0.8353,
+      "step": 3787
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2747610807418823,
+      "learning_rate": 5.301159242608992e-06,
+      "loss": 0.8305,
+      "step": 3788
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2929139137268066,
+      "learning_rate": 5.2990409442547755e-06,
+      "loss": 0.8505,
+      "step": 3789
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.733697235584259,
+      "learning_rate": 5.296922592032388e-06,
+      "loss": 1.0346,
+      "step": 3790
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7235943675041199,
+      "learning_rate": 5.29480418632342e-06,
+      "loss": 1.0187,
+      "step": 3791
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2845617532730103,
+      "learning_rate": 5.292685727509478e-06,
+      "loss": 0.8112,
+      "step": 3792
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6176550388336182,
+      "learning_rate": 5.2905672159721685e-06,
+      "loss": 1.0505,
+      "step": 3793
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3830406665802002,
+      "learning_rate": 5.288448652093115e-06,
+      "loss": 0.8367,
+      "step": 3794
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.451873779296875,
+      "learning_rate": 5.28633003625395e-06,
+      "loss": 0.7685,
+      "step": 3795
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4007667303085327,
+      "learning_rate": 5.2842113688363086e-06,
+      "loss": 0.8133,
+      "step": 3796
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4470592737197876,
+      "learning_rate": 5.282092650221845e-06,
+      "loss": 0.7906,
+      "step": 3797
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3050628900527954,
+      "learning_rate": 5.279973880792214e-06,
+      "loss": 0.8668,
+      "step": 3798
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.5152907371520996,
+      "learning_rate": 5.277855060929086e-06,
+      "loss": 0.8158,
+      "step": 3799
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4483537673950195,
+      "learning_rate": 5.275736191014136e-06,
+      "loss": 0.8664,
+      "step": 3800
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3442317247390747,
+      "learning_rate": 5.273617271429049e-06,
+      "loss": 0.7893,
+      "step": 3801
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3145052194595337,
+      "learning_rate": 5.2714983025555226e-06,
+      "loss": 0.8398,
+      "step": 3802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8376743793487549,
+      "learning_rate": 5.269379284775256e-06,
+      "loss": 1.0379,
+      "step": 3803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4137893915176392,
+      "learning_rate": 5.267260218469964e-06,
+      "loss": 0.8647,
+      "step": 3804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6950117945671082,
+      "learning_rate": 5.265141104021369e-06,
+      "loss": 1.0358,
+      "step": 3805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6742233633995056,
+      "learning_rate": 5.263021941811197e-06,
+      "loss": 1.0001,
+      "step": 3806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4727212190628052,
+      "learning_rate": 5.260902732221192e-06,
+      "loss": 0.8718,
+      "step": 3807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.256888508796692,
+      "learning_rate": 5.258783475633094e-06,
+      "loss": 0.8636,
+      "step": 3808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.280696988105774,
+      "learning_rate": 5.256664172428664e-06,
+      "loss": 0.872,
+      "step": 3809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3074501752853394,
+      "learning_rate": 5.254544822989662e-06,
+      "loss": 0.8304,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.531716227531433,
+      "learning_rate": 5.252425427697863e-06,
+      "loss": 0.7781,
+      "step": 3811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8964094519615173,
+      "learning_rate": 5.250305986935045e-06,
+      "loss": 1.0269,
+      "step": 3812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8239060640335083,
+      "learning_rate": 5.2481865010829966e-06,
+      "loss": 1.0181,
+      "step": 3813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7129186987876892,
+      "learning_rate": 5.246066970523517e-06,
+      "loss": 1.0281,
+      "step": 3814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.361136794090271,
+      "learning_rate": 5.243947395638408e-06,
+      "loss": 0.8198,
+      "step": 3815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3215460777282715,
+      "learning_rate": 5.241827776809485e-06,
+      "loss": 0.8252,
+      "step": 3816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8773756623268127,
+      "learning_rate": 5.239708114418566e-06,
+      "loss": 1.0374,
+      "step": 3817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.436217188835144,
+      "learning_rate": 5.237588408847479e-06,
+      "loss": 0.9215,
+      "step": 3818
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.7819355726242065,
+      "learning_rate": 5.235468660478065e-06,
+      "loss": 1.0266,
+      "step": 3819
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3606842756271362,
+      "learning_rate": 5.233348869692162e-06,
+      "loss": 0.8072,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.39474356174469,
+      "learning_rate": 5.2312290368716256e-06,
+      "loss": 0.7958,
+      "step": 3821
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8085846900939941,
+      "learning_rate": 5.229109162398313e-06,
+      "loss": 1.0216,
+      "step": 3822
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.500139832496643,
+      "learning_rate": 5.226989246654091e-06,
+      "loss": 0.8519,
+      "step": 3823
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3123340606689453,
+      "learning_rate": 5.224869290020833e-06,
+      "loss": 0.8382,
+      "step": 3824
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3251389265060425,
+      "learning_rate": 5.2227492928804235e-06,
+      "loss": 0.8123,
+      "step": 3825
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3975962400436401,
+      "learning_rate": 5.220629255614747e-06,
+      "loss": 0.8179,
+      "step": 3826
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4623315334320068,
+      "learning_rate": 5.218509178605704e-06,
+      "loss": 0.7921,
+      "step": 3827
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3051825761795044,
+      "learning_rate": 5.216389062235193e-06,
+      "loss": 0.8043,
+      "step": 3828
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4014796018600464,
+      "learning_rate": 5.214268906885128e-06,
+      "loss": 0.8312,
+      "step": 3829
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.284649133682251,
+      "learning_rate": 5.212148712937425e-06,
+      "loss": 0.8362,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8701838850975037,
+      "learning_rate": 5.210028480774008e-06,
+      "loss": 1.0321,
+      "step": 3831
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.353187084197998,
+      "learning_rate": 5.207908210776807e-06,
+      "loss": 0.8444,
+      "step": 3832
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.2183918952941895,
+      "learning_rate": 5.205787903327761e-06,
+      "loss": 0.8076,
+      "step": 3833
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3453452587127686,
+      "learning_rate": 5.203667558808815e-06,
+      "loss": 0.8232,
+      "step": 3834
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.009061574935913,
+      "learning_rate": 5.20154717760192e-06,
+      "loss": 0.8562,
+      "step": 3835
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6595736145973206,
+      "learning_rate": 5.1994267600890336e-06,
+      "loss": 1.0233,
+      "step": 3836
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3245079517364502,
+      "learning_rate": 5.1973063066521205e-06,
+      "loss": 0.8315,
+      "step": 3837
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3240495920181274,
+      "learning_rate": 5.195185817673151e-06,
+      "loss": 0.8323,
+      "step": 3838
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4195181131362915,
+      "learning_rate": 5.193065293534104e-06,
+      "loss": 0.8274,
+      "step": 3839
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.703628420829773,
+      "learning_rate": 5.1909447346169614e-06,
+      "loss": 1.04,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6601074934005737,
+      "learning_rate": 5.188824141303715e-06,
+      "loss": 1.0104,
+      "step": 3841
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3337470293045044,
+      "learning_rate": 5.18670351397636e-06,
+      "loss": 0.8276,
+      "step": 3842
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5930667519569397,
+      "learning_rate": 5.184582853016897e-06,
+      "loss": 1.019,
+      "step": 3843
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6144682168960571,
+      "learning_rate": 5.182462158807338e-06,
+      "loss": 1.0228,
+      "step": 3844
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3117213249206543,
+      "learning_rate": 5.180341431729693e-06,
+      "loss": 0.8393,
+      "step": 3845
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3398014307022095,
+      "learning_rate": 5.178220672165985e-06,
+      "loss": 0.8066,
+      "step": 3846
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3400202989578247,
+      "learning_rate": 5.176099880498239e-06,
+      "loss": 0.8596,
+      "step": 3847
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4993066787719727,
+      "learning_rate": 5.173979057108487e-06,
+      "loss": 0.8315,
+      "step": 3848
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3250102996826172,
+      "learning_rate": 5.171858202378769e-06,
+      "loss": 0.8818,
+      "step": 3849
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.8339300751686096,
+      "learning_rate": 5.169737316691125e-06,
+      "loss": 1.0198,
+      "step": 3850
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.309587001800537,
+      "learning_rate": 5.167616400427604e-06,
+      "loss": 0.9141,
+      "step": 3851
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3050309419631958,
+      "learning_rate": 5.165495453970262e-06,
+      "loss": 0.869,
+      "step": 3852
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.4239184856414795,
+      "learning_rate": 5.163374477701159e-06,
+      "loss": 0.8706,
+      "step": 3853
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.3765474557876587,
+      "learning_rate": 5.161253472002358e-06,
+      "loss": 0.8994,
+      "step": 3854
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6308039426803589,
+      "learning_rate": 5.1591324372559294e-06,
+      "loss": 0.8211,
+      "step": 3855
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3718910217285156,
+      "learning_rate": 5.157011373843951e-06,
+      "loss": 0.8295,
+      "step": 3856
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4121801853179932,
+      "learning_rate": 5.154890282148502e-06,
+      "loss": 0.8044,
+      "step": 3857
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2951492071151733,
+      "learning_rate": 5.152769162551671e-06,
+      "loss": 0.8192,
+      "step": 3858
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3662571907043457,
+      "learning_rate": 5.150648015435544e-06,
+      "loss": 0.8292,
+      "step": 3859
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3410199880599976,
+      "learning_rate": 5.148526841182222e-06,
+      "loss": 0.8079,
+      "step": 3860
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3206990957260132,
+      "learning_rate": 5.146405640173802e-06,
+      "loss": 0.7557,
+      "step": 3861
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3364975452423096,
+      "learning_rate": 5.1442844127923915e-06,
+      "loss": 0.8516,
+      "step": 3862
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4406942129135132,
+      "learning_rate": 5.142163159420102e-06,
+      "loss": 0.8672,
+      "step": 3863
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2747406959533691,
+      "learning_rate": 5.140041880439045e-06,
+      "loss": 0.8592,
+      "step": 3864
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.413133144378662,
+      "learning_rate": 5.137920576231342e-06,
+      "loss": 0.8662,
+      "step": 3865
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3318698406219482,
+      "learning_rate": 5.135799247179118e-06,
+      "loss": 0.8494,
+      "step": 3866
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4411975145339966,
+      "learning_rate": 5.133677893664501e-06,
+      "loss": 0.8496,
+      "step": 3867
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6753718852996826,
+      "learning_rate": 5.131556516069625e-06,
+      "loss": 0.9892,
+      "step": 3868
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.273563265800476,
+      "learning_rate": 5.129435114776624e-06,
+      "loss": 0.8234,
+      "step": 3869
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3791841268539429,
+      "learning_rate": 5.127313690167643e-06,
+      "loss": 0.8432,
+      "step": 3870
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3896933794021606,
+      "learning_rate": 5.125192242624828e-06,
+      "loss": 0.8776,
+      "step": 3871
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5599554777145386,
+      "learning_rate": 5.123070772530328e-06,
+      "loss": 0.849,
+      "step": 3872
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2850061655044556,
+      "learning_rate": 5.1209492802662966e-06,
+      "loss": 0.8396,
+      "step": 3873
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3423724174499512,
+      "learning_rate": 5.118827766214891e-06,
+      "loss": 0.8433,
+      "step": 3874
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3125001192092896,
+      "learning_rate": 5.116706230758276e-06,
+      "loss": 0.8205,
+      "step": 3875
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5583226680755615,
+      "learning_rate": 5.114584674278616e-06,
+      "loss": 0.9033,
+      "step": 3876
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5517805814743042,
+      "learning_rate": 5.11246309715808e-06,
+      "loss": 0.8645,
+      "step": 3877
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3959169387817383,
+      "learning_rate": 5.1103414997788424e-06,
+      "loss": 0.8505,
+      "step": 3878
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6987668871879578,
+      "learning_rate": 5.10821988252308e-06,
+      "loss": 1.0364,
+      "step": 3879
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3665434122085571,
+      "learning_rate": 5.106098245772974e-06,
+      "loss": 0.8649,
+      "step": 3880
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3128818273544312,
+      "learning_rate": 5.1039765899107055e-06,
+      "loss": 0.8426,
+      "step": 3881
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6010353565216064,
+      "learning_rate": 5.101854915318466e-06,
+      "loss": 1.0157,
+      "step": 3882
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2982473373413086,
+      "learning_rate": 5.099733222378443e-06,
+      "loss": 0.8238,
+      "step": 3883
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2590123414993286,
+      "learning_rate": 5.097611511472833e-06,
+      "loss": 0.8613,
+      "step": 3884
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6481634378433228,
+      "learning_rate": 5.095489782983833e-06,
+      "loss": 1.0421,
+      "step": 3885
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.217660903930664,
+      "learning_rate": 5.0933680372936445e-06,
+      "loss": 0.8373,
+      "step": 3886
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2708512544631958,
+      "learning_rate": 5.091246274784468e-06,
+      "loss": 0.8584,
+      "step": 3887
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.253760576248169,
+      "learning_rate": 5.089124495838513e-06,
+      "loss": 0.8587,
+      "step": 3888
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.255526065826416,
+      "learning_rate": 5.087002700837988e-06,
+      "loss": 0.8222,
+      "step": 3889
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2585415840148926,
+      "learning_rate": 5.084880890165107e-06,
+      "loss": 0.8572,
+      "step": 3890
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3847754001617432,
+      "learning_rate": 5.0827590642020856e-06,
+      "loss": 0.8243,
+      "step": 3891
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.8209028840065002,
+      "learning_rate": 5.080637223331138e-06,
+      "loss": 1.0197,
+      "step": 3892
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3961610794067383,
+      "learning_rate": 5.07851536793449e-06,
+      "loss": 0.8285,
+      "step": 3893
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3621591329574585,
+      "learning_rate": 5.0763934983943634e-06,
+      "loss": 0.8664,
+      "step": 3894
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.8327065706253052,
+      "learning_rate": 5.074271615092984e-06,
+      "loss": 0.8308,
+      "step": 3895
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.282901644706726,
+      "learning_rate": 5.072149718412579e-06,
+      "loss": 0.833,
+      "step": 3896
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.280216097831726,
+      "learning_rate": 5.07002780873538e-06,
+      "loss": 0.8169,
+      "step": 3897
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3539273738861084,
+      "learning_rate": 5.067905886443621e-06,
+      "loss": 0.8115,
+      "step": 3898
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3717031478881836,
+      "learning_rate": 5.0657839519195396e-06,
+      "loss": 0.8764,
+      "step": 3899
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.391037940979004,
+      "learning_rate": 5.06366200554537e-06,
+      "loss": 0.8413,
+      "step": 3900
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4699841737747192,
+      "learning_rate": 5.061540047703354e-06,
+      "loss": 0.8346,
+      "step": 3901
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3891441822052002,
+      "learning_rate": 5.0594180787757305e-06,
+      "loss": 0.8223,
+      "step": 3902
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2522720098495483,
+      "learning_rate": 5.057296099144747e-06,
+      "loss": 0.8374,
+      "step": 3903
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2750271558761597,
+      "learning_rate": 5.055174109192649e-06,
+      "loss": 0.8234,
+      "step": 3904
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7385058403015137,
+      "learning_rate": 5.053052109301682e-06,
+      "loss": 1.0144,
+      "step": 3905
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6805347204208374,
+      "learning_rate": 5.0509300998540974e-06,
+      "loss": 1.0096,
+      "step": 3906
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2753539085388184,
+      "learning_rate": 5.048808081232146e-06,
+      "loss": 0.8327,
+      "step": 3907
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.332140564918518,
+      "learning_rate": 5.046686053818081e-06,
+      "loss": 0.8441,
+      "step": 3908
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4605653285980225,
+      "learning_rate": 5.044564017994157e-06,
+      "loss": 0.84,
+      "step": 3909
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.480402946472168,
+      "learning_rate": 5.042441974142628e-06,
+      "loss": 0.7962,
+      "step": 3910
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2810715436935425,
+      "learning_rate": 5.040319922645752e-06,
+      "loss": 0.8132,
+      "step": 3911
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.6536731719970703,
+      "learning_rate": 5.03819786388579e-06,
+      "loss": 0.891,
+      "step": 3912
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.81514972448349,
+      "learning_rate": 5.0360757982450005e-06,
+      "loss": 1.0257,
+      "step": 3913
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3320950269699097,
+      "learning_rate": 5.033953726105646e-06,
+      "loss": 0.8242,
+      "step": 3914
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.384708046913147,
+      "learning_rate": 5.031831647849987e-06,
+      "loss": 0.8305,
+      "step": 3915
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.704196035861969,
+      "learning_rate": 5.0297095638602885e-06,
+      "loss": 1.0161,
+      "step": 3916
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6178955435752869,
+      "learning_rate": 5.0275874745188145e-06,
+      "loss": 1.0283,
+      "step": 3917
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4516880512237549,
+      "learning_rate": 5.025465380207832e-06,
+      "loss": 0.8056,
+      "step": 3918
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4788706302642822,
+      "learning_rate": 5.023343281309607e-06,
+      "loss": 0.8617,
+      "step": 3919
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2611174583435059,
+      "learning_rate": 5.021221178206405e-06,
+      "loss": 0.8335,
+      "step": 3920
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2356899976730347,
+      "learning_rate": 5.019099071280496e-06,
+      "loss": 0.8772,
+      "step": 3921
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.360358715057373,
+      "learning_rate": 5.016976960914149e-06,
+      "loss": 0.8167,
+      "step": 3922
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3407092094421387,
+      "learning_rate": 5.014854847489633e-06,
+      "loss": 0.9167,
+      "step": 3923
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5450173616409302,
+      "learning_rate": 5.0127327313892164e-06,
+      "loss": 0.8548,
+      "step": 3924
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.2415906190872192,
+      "learning_rate": 5.0106106129951704e-06,
+      "loss": 0.8122,
+      "step": 3925
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4024080038070679,
+      "learning_rate": 5.008488492689767e-06,
+      "loss": 0.7808,
+      "step": 3926
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.3625411987304688,
+      "learning_rate": 5.006366370855277e-06,
+      "loss": 0.8176,
+      "step": 3927
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.5211148262023926,
+      "learning_rate": 5.0042442478739715e-06,
+      "loss": 0.8171,
+      "step": 3928
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.4084467887878418,
+      "learning_rate": 5.002122124128122e-06,
+      "loss": 0.8331,
+      "step": 3929
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 1.05301833152771,
+      "learning_rate": 5e-06,
+      "loss": 1.0216,
+      "step": 3930
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3367832899093628,
+      "learning_rate": 4.99787787587188e-06,
+      "loss": 0.8745,
+      "step": 3931
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4201254844665527,
+      "learning_rate": 4.995755752126029e-06,
+      "loss": 0.9013,
+      "step": 3932
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6894169449806213,
+      "learning_rate": 4.993633629144726e-06,
+      "loss": 1.0122,
+      "step": 3933
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4049252271652222,
+      "learning_rate": 4.991511507310233e-06,
+      "loss": 0.825,
+      "step": 3934
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3398224115371704,
+      "learning_rate": 4.989389387004831e-06,
+      "loss": 0.8428,
+      "step": 3935
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 5.251138210296631,
+      "learning_rate": 4.987267268610786e-06,
+      "loss": 0.8152,
+      "step": 3936
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8620211482048035,
+      "learning_rate": 4.985145152510369e-06,
+      "loss": 1.0445,
+      "step": 3937
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3737879991531372,
+      "learning_rate": 4.9830230390858525e-06,
+      "loss": 0.8001,
+      "step": 3938
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4497098922729492,
+      "learning_rate": 4.9809009287195045e-06,
+      "loss": 0.8225,
+      "step": 3939
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7447367310523987,
+      "learning_rate": 4.978778821793596e-06,
+      "loss": 1.0175,
+      "step": 3940
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3564335107803345,
+      "learning_rate": 4.976656718690395e-06,
+      "loss": 0.8596,
+      "step": 3941
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6632445454597473,
+      "learning_rate": 4.974534619792169e-06,
+      "loss": 1.0365,
+      "step": 3942
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4841822385787964,
+      "learning_rate": 4.972412525481186e-06,
+      "loss": 0.8464,
+      "step": 3943
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6667802929878235,
+      "learning_rate": 4.9702904361397114e-06,
+      "loss": 1.0264,
+      "step": 3944
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.455933928489685,
+      "learning_rate": 4.968168352150014e-06,
+      "loss": 0.8824,
+      "step": 3945
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2805722951889038,
+      "learning_rate": 4.966046273894356e-06,
+      "loss": 0.8246,
+      "step": 3946
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3417890071868896,
+      "learning_rate": 4.9639242017549995e-06,
+      "loss": 0.8593,
+      "step": 3947
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3238110542297363,
+      "learning_rate": 4.961802136114211e-06,
+      "loss": 0.8185,
+      "step": 3948
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4058899879455566,
+      "learning_rate": 4.95968007735425e-06,
+      "loss": 0.7961,
+      "step": 3949
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.557210087776184,
+      "learning_rate": 4.957558025857374e-06,
+      "loss": 0.8506,
+      "step": 3950
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7666508555412292,
+      "learning_rate": 4.9554359820058465e-06,
+      "loss": 1.0371,
+      "step": 3951
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3396961688995361,
+      "learning_rate": 4.95331394618192e-06,
+      "loss": 0.8001,
+      "step": 3952
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2780379056930542,
+      "learning_rate": 4.951191918767855e-06,
+      "loss": 0.8453,
+      "step": 3953
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3899905681610107,
+      "learning_rate": 4.949069900145904e-06,
+      "loss": 0.8437,
+      "step": 3954
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3829156160354614,
+      "learning_rate": 4.946947890698318e-06,
+      "loss": 0.8223,
+      "step": 3955
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.287413239479065,
+      "learning_rate": 4.944825890807353e-06,
+      "loss": 0.834,
+      "step": 3956
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.416698932647705,
+      "learning_rate": 4.942703900855254e-06,
+      "loss": 0.8907,
+      "step": 3957
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3943370580673218,
+      "learning_rate": 4.94058192122427e-06,
+      "loss": 0.8624,
+      "step": 3958
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3214038610458374,
+      "learning_rate": 4.938459952296649e-06,
+      "loss": 0.8193,
+      "step": 3959
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2850064039230347,
+      "learning_rate": 4.936337994454632e-06,
+      "loss": 0.8079,
+      "step": 3960
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3598051071166992,
+      "learning_rate": 4.934216048080462e-06,
+      "loss": 0.829,
+      "step": 3961
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3689836263656616,
+      "learning_rate": 4.9320941135563786e-06,
+      "loss": 0.7841,
+      "step": 3962
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3982913494110107,
+      "learning_rate": 4.929972191264621e-06,
+      "loss": 0.8743,
+      "step": 3963
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.293536901473999,
+      "learning_rate": 4.927850281587424e-06,
+      "loss": 0.8467,
+      "step": 3964
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3041386604309082,
+      "learning_rate": 4.925728384907018e-06,
+      "loss": 0.8514,
+      "step": 3965
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7224178314208984,
+      "learning_rate": 4.923606501605639e-06,
+      "loss": 1.0258,
+      "step": 3966
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4466078281402588,
+      "learning_rate": 4.92148463206551e-06,
+      "loss": 0.8397,
+      "step": 3967
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.331717848777771,
+      "learning_rate": 4.9193627766688625e-06,
+      "loss": 0.8536,
+      "step": 3968
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3347874879837036,
+      "learning_rate": 4.917240935797918e-06,
+      "loss": 0.7636,
+      "step": 3969
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4095672369003296,
+      "learning_rate": 4.915119109834894e-06,
+      "loss": 0.7972,
+      "step": 3970
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2683875560760498,
+      "learning_rate": 4.9129972991620135e-06,
+      "loss": 0.7594,
+      "step": 3971
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6824325919151306,
+      "learning_rate": 4.910875504161488e-06,
+      "loss": 1.027,
+      "step": 3972
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.5134729146957397,
+      "learning_rate": 4.9087537252155335e-06,
+      "loss": 0.8266,
+      "step": 3973
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6272510290145874,
+      "learning_rate": 4.906631962706359e-06,
+      "loss": 1.0318,
+      "step": 3974
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3449962139129639,
+      "learning_rate": 4.904510217016168e-06,
+      "loss": 0.8291,
+      "step": 3975
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6615743041038513,
+      "learning_rate": 4.9023884885271675e-06,
+      "loss": 1.0385,
+      "step": 3976
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3351545333862305,
+      "learning_rate": 4.900266777621559e-06,
+      "loss": 0.8127,
+      "step": 3977
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.272688388824463,
+      "learning_rate": 4.898145084681536e-06,
+      "loss": 0.7863,
+      "step": 3978
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.302107810974121,
+      "learning_rate": 4.896023410089295e-06,
+      "loss": 0.8859,
+      "step": 3979
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6795272827148438,
+      "learning_rate": 4.893901754227028e-06,
+      "loss": 1.0309,
+      "step": 3980
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6583009958267212,
+      "learning_rate": 4.891780117476921e-06,
+      "loss": 1.0027,
+      "step": 3981
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6425154805183411,
+      "learning_rate": 4.889658500221159e-06,
+      "loss": 1.0237,
+      "step": 3982
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6199939250946045,
+      "learning_rate": 4.887536902841921e-06,
+      "loss": 1.0101,
+      "step": 3983
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3318008184432983,
+      "learning_rate": 4.885415325721386e-06,
+      "loss": 0.8402,
+      "step": 3984
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3620553016662598,
+      "learning_rate": 4.883293769241725e-06,
+      "loss": 0.8257,
+      "step": 3985
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3206923007965088,
+      "learning_rate": 4.88117223378511e-06,
+      "loss": 0.8209,
+      "step": 3986
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4468532800674438,
+      "learning_rate": 4.879050719733707e-06,
+      "loss": 0.8129,
+      "step": 3987
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4193493127822876,
+      "learning_rate": 4.876929227469673e-06,
+      "loss": 0.8169,
+      "step": 3988
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3434323072433472,
+      "learning_rate": 4.874807757375174e-06,
+      "loss": 0.7697,
+      "step": 3989
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.8227210640907288,
+      "learning_rate": 4.8726863098323566e-06,
+      "loss": 1.0028,
+      "step": 3990
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.323317527770996,
+      "learning_rate": 4.870564885223377e-06,
+      "loss": 0.8329,
+      "step": 3991
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6562132835388184,
+      "learning_rate": 4.868443483930379e-06,
+      "loss": 1.0182,
+      "step": 3992
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.382500171661377,
+      "learning_rate": 4.8663221063355e-06,
+      "loss": 0.8566,
+      "step": 3993
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6817378997802734,
+      "learning_rate": 4.8642007528208835e-06,
+      "loss": 1.041,
+      "step": 3994
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.462205171585083,
+      "learning_rate": 4.862079423768658e-06,
+      "loss": 0.8044,
+      "step": 3995
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2861301898956299,
+      "learning_rate": 4.859958119560956e-06,
+      "loss": 0.7731,
+      "step": 3996
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.268129587173462,
+      "learning_rate": 4.8578368405799e-06,
+      "loss": 0.8832,
+      "step": 3997
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7111806273460388,
+      "learning_rate": 4.8557155872076085e-06,
+      "loss": 1.0325,
+      "step": 3998
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6528083682060242,
+      "learning_rate": 4.8535943598261984e-06,
+      "loss": 1.0372,
+      "step": 3999
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.4004019498825073,
+      "learning_rate": 4.8514731588177784e-06,
+      "loss": 0.7956,
+      "step": 4000
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.364266037940979,
+      "learning_rate": 4.849351984564457e-06,
+      "loss": 0.8647,
+      "step": 4001
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2836891412734985,
+      "learning_rate": 4.847230837448331e-06,
+      "loss": 0.8238,
+      "step": 4002
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6866990327835083,
+      "learning_rate": 4.845109717851498e-06,
+      "loss": 1.0359,
+      "step": 4003
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3348984718322754,
+      "learning_rate": 4.842988626156051e-06,
+      "loss": 0.8125,
+      "step": 4004
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.2708165645599365,
+      "learning_rate": 4.840867562744073e-06,
+      "loss": 0.8211,
+      "step": 4005
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.3245742321014404,
+      "learning_rate": 4.8387465279976446e-06,
+      "loss": 0.8352,
+      "step": 4006
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6885313987731934,
+      "learning_rate": 4.836625522298844e-06,
+      "loss": 1.0593,
+      "step": 4007
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3454704284667969,
+      "learning_rate": 4.834504546029739e-06,
+      "loss": 0.8247,
+      "step": 4008
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3581111431121826,
+      "learning_rate": 4.832383599572397e-06,
+      "loss": 0.8415,
+      "step": 4009
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4140245914459229,
+      "learning_rate": 4.830262683308878e-06,
+      "loss": 0.8371,
+      "step": 4010
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6363005638122559,
+      "learning_rate": 4.828141797621232e-06,
+      "loss": 1.0211,
+      "step": 4011
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3079111576080322,
+      "learning_rate": 4.826020942891514e-06,
+      "loss": 0.8926,
+      "step": 4012
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3770140409469604,
+      "learning_rate": 4.823900119501762e-06,
+      "loss": 0.7796,
+      "step": 4013
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6183586716651917,
+      "learning_rate": 4.821779327834017e-06,
+      "loss": 1.0424,
+      "step": 4014
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3253816366195679,
+      "learning_rate": 4.81965856827031e-06,
+      "loss": 0.8583,
+      "step": 4015
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 3.112436532974243,
+      "learning_rate": 4.817537841192664e-06,
+      "loss": 0.8506,
+      "step": 4016
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3304872512817383,
+      "learning_rate": 4.815417146983104e-06,
+      "loss": 0.9113,
+      "step": 4017
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6266931295394897,
+      "learning_rate": 4.813296486023642e-06,
+      "loss": 1.0157,
+      "step": 4018
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2865760326385498,
+      "learning_rate": 4.811175858696286e-06,
+      "loss": 0.791,
+      "step": 4019
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6167840361595154,
+      "learning_rate": 4.809055265383039e-06,
+      "loss": 1.0564,
+      "step": 4020
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4523619413375854,
+      "learning_rate": 4.806934706465897e-06,
+      "loss": 0.8202,
+      "step": 4021
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4445677995681763,
+      "learning_rate": 4.80481418232685e-06,
+      "loss": 0.8627,
+      "step": 4022
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6053305864334106,
+      "learning_rate": 4.80269369334788e-06,
+      "loss": 1.0142,
+      "step": 4023
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.260496735572815,
+      "learning_rate": 4.800573239910968e-06,
+      "loss": 0.8387,
+      "step": 4024
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6394516825675964,
+      "learning_rate": 4.798452822398083e-06,
+      "loss": 1.0383,
+      "step": 4025
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2678650617599487,
+      "learning_rate": 4.796332441191185e-06,
+      "loss": 0.783,
+      "step": 4026
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.30374014377594,
+      "learning_rate": 4.7942120966722405e-06,
+      "loss": 0.8743,
+      "step": 4027
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2553057670593262,
+      "learning_rate": 4.792091789223193e-06,
+      "loss": 0.866,
+      "step": 4028
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3888201713562012,
+      "learning_rate": 4.789971519225994e-06,
+      "loss": 0.7617,
+      "step": 4029
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6976369619369507,
+      "learning_rate": 4.7878512870625775e-06,
+      "loss": 1.0373,
+      "step": 4030
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7655280232429504,
+      "learning_rate": 4.785731093114873e-06,
+      "loss": 1.0423,
+      "step": 4031
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6381654739379883,
+      "learning_rate": 4.783610937764808e-06,
+      "loss": 1.0344,
+      "step": 4032
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3802751302719116,
+      "learning_rate": 4.781490821394299e-06,
+      "loss": 0.8386,
+      "step": 4033
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.475807547569275,
+      "learning_rate": 4.7793707443852536e-06,
+      "loss": 0.8394,
+      "step": 4034
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4612258672714233,
+      "learning_rate": 4.77725070711958e-06,
+      "loss": 0.7642,
+      "step": 4035
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3135768175125122,
+      "learning_rate": 4.775130709979167e-06,
+      "loss": 0.8318,
+      "step": 4036
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3534700870513916,
+      "learning_rate": 4.7730107533459115e-06,
+      "loss": 0.8409,
+      "step": 4037
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4282468557357788,
+      "learning_rate": 4.770890837601691e-06,
+      "loss": 0.8624,
+      "step": 4038
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3767695426940918,
+      "learning_rate": 4.768770963128376e-06,
+      "loss": 0.8531,
+      "step": 4039
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.32289719581604,
+      "learning_rate": 4.76665113030784e-06,
+      "loss": 0.9121,
+      "step": 4040
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.396311640739441,
+      "learning_rate": 4.764531339521938e-06,
+      "loss": 0.8296,
+      "step": 4041
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.320469856262207,
+      "learning_rate": 4.762411591152522e-06,
+      "loss": 0.84,
+      "step": 4042
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3581181764602661,
+      "learning_rate": 4.7602918855814365e-06,
+      "loss": 0.8336,
+      "step": 4043
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.0390969514846802,
+      "learning_rate": 4.7581722231905165e-06,
+      "loss": 1.0265,
+      "step": 4044
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4781440496444702,
+      "learning_rate": 4.756052604361594e-06,
+      "loss": 0.8779,
+      "step": 4045
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7185244560241699,
+      "learning_rate": 4.753933029476484e-06,
+      "loss": 1.0128,
+      "step": 4046
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3943967819213867,
+      "learning_rate": 4.751813498917004e-06,
+      "loss": 0.8248,
+      "step": 4047
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.368621587753296,
+      "learning_rate": 4.749694013064958e-06,
+      "loss": 0.8658,
+      "step": 4048
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2613086700439453,
+      "learning_rate": 4.747574572302138e-06,
+      "loss": 0.9137,
+      "step": 4049
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9728686809539795,
+      "learning_rate": 4.74545517701034e-06,
+      "loss": 0.9958,
+      "step": 4050
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8921542763710022,
+      "learning_rate": 4.743335827571337e-06,
+      "loss": 1.011,
+      "step": 4051
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2716790437698364,
+      "learning_rate": 4.7412165243669065e-06,
+      "loss": 0.8721,
+      "step": 4052
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.607926607131958,
+      "learning_rate": 4.739097267778811e-06,
+      "loss": 1.0383,
+      "step": 4053
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.738983690738678,
+      "learning_rate": 4.736978058188803e-06,
+      "loss": 1.0229,
+      "step": 4054
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2745238542556763,
+      "learning_rate": 4.734858895978633e-06,
+      "loss": 0.8444,
+      "step": 4055
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.7353696227073669,
+      "learning_rate": 4.732739781530036e-06,
+      "loss": 1.0456,
+      "step": 4056
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3939924240112305,
+      "learning_rate": 4.730620715224745e-06,
+      "loss": 0.8559,
+      "step": 4057
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 4.983041763305664,
+      "learning_rate": 4.72850169744448e-06,
+      "loss": 0.8431,
+      "step": 4058
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3250755071640015,
+      "learning_rate": 4.726382728570952e-06,
+      "loss": 0.8589,
+      "step": 4059
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.358657717704773,
+      "learning_rate": 4.724263808985866e-06,
+      "loss": 0.8556,
+      "step": 4060
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3141719102859497,
+      "learning_rate": 4.722144939070915e-06,
+      "loss": 0.8062,
+      "step": 4061
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4172829389572144,
+      "learning_rate": 4.720026119207787e-06,
+      "loss": 0.8238,
+      "step": 4062
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2795403003692627,
+      "learning_rate": 4.717907349778156e-06,
+      "loss": 0.7903,
+      "step": 4063
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4383169412612915,
+      "learning_rate": 4.715788631163692e-06,
+      "loss": 0.7794,
+      "step": 4064
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3356307744979858,
+      "learning_rate": 4.713669963746053e-06,
+      "loss": 0.8223,
+      "step": 4065
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.878689169883728,
+      "learning_rate": 4.7115513479068866e-06,
+      "loss": 1.0506,
+      "step": 4066
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.756655752658844,
+      "learning_rate": 4.709432784027832e-06,
+      "loss": 1.0285,
+      "step": 4067
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6374886631965637,
+      "learning_rate": 4.707314272490524e-06,
+      "loss": 1.0195,
+      "step": 4068
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3372243642807007,
+      "learning_rate": 4.70519581367658e-06,
+      "loss": 0.8863,
+      "step": 4069
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.2909003496170044,
+      "learning_rate": 4.703077407967614e-06,
+      "loss": 0.8316,
+      "step": 4070
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4493216276168823,
+      "learning_rate": 4.700959055745227e-06,
+      "loss": 0.8345,
+      "step": 4071
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4760961532592773,
+      "learning_rate": 4.698840757391009e-06,
+      "loss": 0.8094,
+      "step": 4072
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.4248872995376587,
+      "learning_rate": 4.696722513286547e-06,
+      "loss": 0.8882,
+      "step": 4073
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.018182396888733,
+      "learning_rate": 4.6946043238134104e-06,
+      "loss": 1.0161,
+      "step": 4074
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3319047689437866,
+      "learning_rate": 4.692486189353166e-06,
+      "loss": 0.8771,
+      "step": 4075
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3316787481307983,
+      "learning_rate": 4.690368110287365e-06,
+      "loss": 0.8233,
+      "step": 4076
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6435905694961548,
+      "learning_rate": 4.688250086997547e-06,
+      "loss": 1.0045,
+      "step": 4077
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6283195614814758,
+      "learning_rate": 4.6861321198652535e-06,
+      "loss": 0.9871,
+      "step": 4078
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6394302248954773,
+      "learning_rate": 4.684014209272e-06,
+      "loss": 1.0194,
+      "step": 4079
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3946595191955566,
+      "learning_rate": 4.681896355599302e-06,
+      "loss": 0.8262,
+      "step": 4080
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.3713865280151367,
+      "learning_rate": 4.679778559228664e-06,
+      "loss": 0.7966,
+      "step": 4081
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 1.426530361175537,
+      "learning_rate": 4.677660820541575e-06,
+      "loss": 0.84,
+      "step": 4082
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4160168170928955,
+      "learning_rate": 4.67554313991952e-06,
+      "loss": 0.8823,
+      "step": 4083
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3186007738113403,
+      "learning_rate": 4.673425517743967e-06,
+      "loss": 0.8653,
+      "step": 4084
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3193562030792236,
+      "learning_rate": 4.67130795439638e-06,
+      "loss": 0.8093,
+      "step": 4085
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4900401830673218,
+      "learning_rate": 4.6691904502582095e-06,
+      "loss": 0.8042,
+      "step": 4086
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4051368236541748,
+      "learning_rate": 4.66707300571089e-06,
+      "loss": 0.8888,
+      "step": 4087
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9184963703155518,
+      "learning_rate": 4.664955621135858e-06,
+      "loss": 1.0461,
+      "step": 4088
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8213314414024353,
+      "learning_rate": 4.662838296914525e-06,
+      "loss": 1.0181,
+      "step": 4089
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.652876079082489,
+      "learning_rate": 4.6607210334283036e-06,
+      "loss": 1.0372,
+      "step": 4090
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.434249997138977,
+      "learning_rate": 4.658603831058588e-06,
+      "loss": 0.8482,
+      "step": 4091
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5398845672607422,
+      "learning_rate": 4.65648669018676e-06,
+      "loss": 0.8438,
+      "step": 4092
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.43696928024292,
+      "learning_rate": 4.6543696111942e-06,
+      "loss": 0.8161,
+      "step": 4093
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4839134216308594,
+      "learning_rate": 4.6522525944622675e-06,
+      "loss": 0.8763,
+      "step": 4094
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4816465377807617,
+      "learning_rate": 4.650135640372313e-06,
+      "loss": 0.8681,
+      "step": 4095
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3422729969024658,
+      "learning_rate": 4.648018749305681e-06,
+      "loss": 0.8742,
+      "step": 4096
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6135857105255127,
+      "learning_rate": 4.645901921643697e-06,
+      "loss": 0.8,
+      "step": 4097
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.357039451599121,
+      "learning_rate": 4.643785157767682e-06,
+      "loss": 0.8378,
+      "step": 4098
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2687069177627563,
+      "learning_rate": 4.641668458058941e-06,
+      "loss": 0.8032,
+      "step": 4099
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4521101713180542,
+      "learning_rate": 4.6395518228987676e-06,
+      "loss": 0.8262,
+      "step": 4100
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4351773262023926,
+      "learning_rate": 4.637435252668446e-06,
+      "loss": 0.8561,
+      "step": 4101
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2777185440063477,
+      "learning_rate": 4.635318747749247e-06,
+      "loss": 1.0376,
+      "step": 4102
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5062410831451416,
+      "learning_rate": 4.633202308522432e-06,
+      "loss": 0.7974,
+      "step": 4103
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5924856662750244,
+      "learning_rate": 4.631085935369247e-06,
+      "loss": 0.823,
+      "step": 4104
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3228423595428467,
+      "learning_rate": 4.6289696286709284e-06,
+      "loss": 0.8547,
+      "step": 4105
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7318580150604248,
+      "learning_rate": 4.626853388808702e-06,
+      "loss": 0.9998,
+      "step": 4106
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4289250373840332,
+      "learning_rate": 4.624737216163774e-06,
+      "loss": 0.8576,
+      "step": 4107
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4694030284881592,
+      "learning_rate": 4.622621111117352e-06,
+      "loss": 0.8249,
+      "step": 4108
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6730291247367859,
+      "learning_rate": 4.620505074050619e-06,
+      "loss": 1.0063,
+      "step": 4109
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6552523374557495,
+      "learning_rate": 4.6183891053447495e-06,
+      "loss": 1.0131,
+      "step": 4110
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7094762325286865,
+      "learning_rate": 4.616273205380911e-06,
+      "loss": 0.8229,
+      "step": 4111
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6363412737846375,
+      "learning_rate": 4.614157374540248e-06,
+      "loss": 1.0272,
+      "step": 4112
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3801544904708862,
+      "learning_rate": 4.6120416132039055e-06,
+      "loss": 0.8433,
+      "step": 4113
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.269727110862732,
+      "learning_rate": 4.6099259217530055e-06,
+      "loss": 0.7769,
+      "step": 4114
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.341106653213501,
+      "learning_rate": 4.607810300568659e-06,
+      "loss": 0.8628,
+      "step": 4115
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2763521671295166,
+      "learning_rate": 4.605694750031972e-06,
+      "loss": 0.824,
+      "step": 4116
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4020854234695435,
+      "learning_rate": 4.603579270524028e-06,
+      "loss": 0.8312,
+      "step": 4117
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.333797574043274,
+      "learning_rate": 4.601463862425903e-06,
+      "loss": 0.7941,
+      "step": 4118
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.7599210739135742,
+      "learning_rate": 4.59934852611866e-06,
+      "loss": 0.8818,
+      "step": 4119
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7922137975692749,
+      "learning_rate": 4.597233261983347e-06,
+      "loss": 1.0066,
+      "step": 4120
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6649919748306274,
+      "learning_rate": 4.595118070401001e-06,
+      "loss": 0.8417,
+      "step": 4121
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2961335182189941,
+      "learning_rate": 4.593002951752645e-06,
+      "loss": 0.8316,
+      "step": 4122
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.359747290611267,
+      "learning_rate": 4.590887906419288e-06,
+      "loss": 0.8568,
+      "step": 4123
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3989167213439941,
+      "learning_rate": 4.588772934781929e-06,
+      "loss": 0.8737,
+      "step": 4124
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3229578733444214,
+      "learning_rate": 4.586658037221546e-06,
+      "loss": 0.8748,
+      "step": 4125
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4282320737838745,
+      "learning_rate": 4.584543214119117e-06,
+      "loss": 0.8243,
+      "step": 4126
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.7118299603462219,
+      "learning_rate": 4.582428465855594e-06,
+      "loss": 1.0607,
+      "step": 4127
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.398364543914795,
+      "learning_rate": 4.5803137928119176e-06,
+      "loss": 0.8339,
+      "step": 4128
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.8385459184646606,
+      "learning_rate": 4.5781991953690245e-06,
+      "loss": 0.8826,
+      "step": 4129
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2830767631530762,
+      "learning_rate": 4.576084673907824e-06,
+      "loss": 0.8816,
+      "step": 4130
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2907601594924927,
+      "learning_rate": 4.573970228809224e-06,
+      "loss": 0.8434,
+      "step": 4131
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6380224823951721,
+      "learning_rate": 4.571855860454111e-06,
+      "loss": 1.042,
+      "step": 4132
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3639442920684814,
+      "learning_rate": 4.569741569223356e-06,
+      "loss": 0.8502,
+      "step": 4133
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4702032804489136,
+      "learning_rate": 4.567627355497827e-06,
+      "loss": 0.8898,
+      "step": 4134
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2936769723892212,
+      "learning_rate": 4.565513219658364e-06,
+      "loss": 0.8409,
+      "step": 4135
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6859135031700134,
+      "learning_rate": 4.563399162085806e-06,
+      "loss": 1.0375,
+      "step": 4136
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6366080641746521,
+      "learning_rate": 4.5612851831609684e-06,
+      "loss": 1.0282,
+      "step": 4137
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3298513889312744,
+      "learning_rate": 4.559171283264654e-06,
+      "loss": 0.871,
+      "step": 4138
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3260397911071777,
+      "learning_rate": 4.5570574627776584e-06,
+      "loss": 0.7908,
+      "step": 4139
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3849670886993408,
+      "learning_rate": 4.554943722080754e-06,
+      "loss": 0.8256,
+      "step": 4140
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.313706636428833,
+      "learning_rate": 4.5528300615547045e-06,
+      "loss": 0.8445,
+      "step": 4141
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.8219835758209229,
+      "learning_rate": 4.550716481580254e-06,
+      "loss": 1.0181,
+      "step": 4142
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3113648891448975,
+      "learning_rate": 4.548602982538139e-06,
+      "loss": 0.707,
+      "step": 4143
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4594861268997192,
+      "learning_rate": 4.546489564809076e-06,
+      "loss": 0.8416,
+      "step": 4144
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.655335009098053,
+      "learning_rate": 4.544376228773766e-06,
+      "loss": 1.0307,
+      "step": 4145
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6083617806434631,
+      "learning_rate": 4.542262974812904e-06,
+      "loss": 1.0075,
+      "step": 4146
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6133086085319519,
+      "learning_rate": 4.54014980330716e-06,
+      "loss": 1.0105,
+      "step": 4147
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6384602189064026,
+      "learning_rate": 4.538036714637191e-06,
+      "loss": 1.025,
+      "step": 4148
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4676947593688965,
+      "learning_rate": 4.5359237091836465e-06,
+      "loss": 0.7924,
+      "step": 4149
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.5387489795684814,
+      "learning_rate": 4.533810787327153e-06,
+      "loss": 0.8655,
+      "step": 4150
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4091217517852783,
+      "learning_rate": 4.531697949448323e-06,
+      "loss": 0.8155,
+      "step": 4151
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6424919962882996,
+      "learning_rate": 4.529585195927761e-06,
+      "loss": 1.0005,
+      "step": 4152
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.3407306671142578,
+      "learning_rate": 4.527472527146043e-06,
+      "loss": 0.8447,
+      "step": 4153
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6301578283309937,
+      "learning_rate": 4.5253599434837456e-06,
+      "loss": 1.0088,
+      "step": 4154
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.6031091213226318,
+      "learning_rate": 4.523247445321418e-06,
+      "loss": 0.7983,
+      "step": 4155
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.434720754623413,
+      "learning_rate": 4.521135033039597e-06,
+      "loss": 0.8374,
+      "step": 4156
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2168371677398682,
+      "learning_rate": 4.519022707018809e-06,
+      "loss": 0.8071,
+      "step": 4157
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.4249513149261475,
+      "learning_rate": 4.516910467639556e-06,
+      "loss": 0.8128,
+      "step": 4158
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 1.2287558317184448,
+      "learning_rate": 4.514798315282335e-06,
+      "loss": 0.7943,
+      "step": 4159
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.339540719985962,
+      "learning_rate": 4.512686250327618e-06,
+      "loss": 0.8188,
+      "step": 4160
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2660397291183472,
+      "learning_rate": 4.510574273155864e-06,
+      "loss": 0.7975,
+      "step": 4161
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3106536865234375,
+      "learning_rate": 4.5084623841475184e-06,
+      "loss": 0.8047,
+      "step": 4162
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3063781261444092,
+      "learning_rate": 4.50635058368301e-06,
+      "loss": 0.84,
+      "step": 4163
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7524858117103577,
+      "learning_rate": 4.504238872142751e-06,
+      "loss": 1.026,
+      "step": 4164
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4013645648956299,
+      "learning_rate": 4.502127249907136e-06,
+      "loss": 0.8033,
+      "step": 4165
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6568412184715271,
+      "learning_rate": 4.500015717356545e-06,
+      "loss": 1.0377,
+      "step": 4166
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6365652084350586,
+      "learning_rate": 4.497904274871346e-06,
+      "loss": 1.0459,
+      "step": 4167
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4199609756469727,
+      "learning_rate": 4.495792922831881e-06,
+      "loss": 0.829,
+      "step": 4168
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3671139478683472,
+      "learning_rate": 4.493681661618487e-06,
+      "loss": 0.8997,
+      "step": 4169
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3898780345916748,
+      "learning_rate": 4.491570491611475e-06,
+      "loss": 0.8598,
+      "step": 4170
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6855762004852295,
+      "learning_rate": 4.489459413191143e-06,
+      "loss": 1.0144,
+      "step": 4171
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.31523597240448,
+      "learning_rate": 4.487348426737779e-06,
+      "loss": 0.8547,
+      "step": 4172
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2353326082229614,
+      "learning_rate": 4.485237532631641e-06,
+      "loss": 0.8352,
+      "step": 4173
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3571525812149048,
+      "learning_rate": 4.483126731252984e-06,
+      "loss": 0.8717,
+      "step": 4174
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3070884943008423,
+      "learning_rate": 4.481016022982039e-06,
+      "loss": 0.8027,
+      "step": 4175
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.288332462310791,
+      "learning_rate": 4.478905408199017e-06,
+      "loss": 0.7947,
+      "step": 4176
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.7558302879333496,
+      "learning_rate": 4.476794887284123e-06,
+      "loss": 1.0509,
+      "step": 4177
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6645744442939758,
+      "learning_rate": 4.474684460617536e-06,
+      "loss": 1.0097,
+      "step": 4178
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3246041536331177,
+      "learning_rate": 4.4725741285794175e-06,
+      "loss": 0.8088,
+      "step": 4179
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3158137798309326,
+      "learning_rate": 4.47046389154992e-06,
+      "loss": 0.7814,
+      "step": 4180
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3966747522354126,
+      "learning_rate": 4.468353749909172e-06,
+      "loss": 0.8372,
+      "step": 4181
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3291912078857422,
+      "learning_rate": 4.466243704037287e-06,
+      "loss": 0.7885,
+      "step": 4182
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.6096843481063843,
+      "learning_rate": 4.464133754314363e-06,
+      "loss": 0.8225,
+      "step": 4183
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2729610204696655,
+      "learning_rate": 4.462023901120476e-06,
+      "loss": 0.7739,
+      "step": 4184
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4452991485595703,
+      "learning_rate": 4.459914144835689e-06,
+      "loss": 0.8173,
+      "step": 4185
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.8345254063606262,
+      "learning_rate": 4.457804485840044e-06,
+      "loss": 1.0276,
+      "step": 4186
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2552493810653687,
+      "learning_rate": 4.4556949245135715e-06,
+      "loss": 0.8437,
+      "step": 4187
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6930148005485535,
+      "learning_rate": 4.453585461236277e-06,
+      "loss": 1.0004,
+      "step": 4188
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.383614182472229,
+      "learning_rate": 4.451476096388151e-06,
+      "loss": 0.808,
+      "step": 4189
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.268728256225586,
+      "learning_rate": 4.449366830349172e-06,
+      "loss": 0.8062,
+      "step": 4190
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3235994577407837,
+      "learning_rate": 4.447257663499289e-06,
+      "loss": 0.8644,
+      "step": 4191
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3183071613311768,
+      "learning_rate": 4.445148596218445e-06,
+      "loss": 0.8757,
+      "step": 4192
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.318776249885559,
+      "learning_rate": 4.443039628886559e-06,
+      "loss": 0.8161,
+      "step": 4193
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3548153638839722,
+      "learning_rate": 4.4409307618835284e-06,
+      "loss": 0.8474,
+      "step": 4194
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.929246723651886,
+      "learning_rate": 4.4388219955892436e-06,
+      "loss": 1.0243,
+      "step": 4195
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.432762622833252,
+      "learning_rate": 4.4367133303835646e-06,
+      "loss": 0.7749,
+      "step": 4196
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4133985042572021,
+      "learning_rate": 4.434604766646344e-06,
+      "loss": 0.8205,
+      "step": 4197
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2982913255691528,
+      "learning_rate": 4.432496304757408e-06,
+      "loss": 0.8539,
+      "step": 4198
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5951471328735352,
+      "learning_rate": 4.430387945096566e-06,
+      "loss": 0.8614,
+      "step": 4199
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6904889941215515,
+      "learning_rate": 4.428279688043613e-06,
+      "loss": 1.0626,
+      "step": 4200
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3546949625015259,
+      "learning_rate": 4.426171533978321e-06,
+      "loss": 0.8641,
+      "step": 4201
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.265763521194458,
+      "learning_rate": 4.4240634832804496e-06,
+      "loss": 0.8637,
+      "step": 4202
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6204572916030884,
+      "learning_rate": 4.42195553632973e-06,
+      "loss": 1.0371,
+      "step": 4203
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.5417243242263794,
+      "learning_rate": 4.419847693505883e-06,
+      "loss": 0.8094,
+      "step": 4204
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2851718664169312,
+      "learning_rate": 4.417739955188608e-06,
+      "loss": 0.7994,
+      "step": 4205
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2794259786605835,
+      "learning_rate": 4.415632321757583e-06,
+      "loss": 0.846,
+      "step": 4206
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6364712119102478,
+      "learning_rate": 4.413524793592475e-06,
+      "loss": 1.023,
+      "step": 4207
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2683773040771484,
+      "learning_rate": 4.411417371072922e-06,
+      "loss": 0.8398,
+      "step": 4208
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6102322936058044,
+      "learning_rate": 4.409310054578547e-06,
+      "loss": 1.0277,
+      "step": 4209
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3595421314239502,
+      "learning_rate": 4.407202844488959e-06,
+      "loss": 0.7737,
+      "step": 4210
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2549363374710083,
+      "learning_rate": 4.40509574118374e-06,
+      "loss": 0.768,
+      "step": 4211
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3466054201126099,
+      "learning_rate": 4.402988745042454e-06,
+      "loss": 0.8278,
+      "step": 4212
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4317834377288818,
+      "learning_rate": 4.400881856444654e-06,
+      "loss": 0.8341,
+      "step": 4213
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4033308029174805,
+      "learning_rate": 4.398775075769862e-06,
+      "loss": 0.8366,
+      "step": 4214
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4795957803726196,
+      "learning_rate": 4.396668403397589e-06,
+      "loss": 0.8691,
+      "step": 4215
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.638575553894043,
+      "learning_rate": 4.394561839707323e-06,
+      "loss": 1.0256,
+      "step": 4216
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3773423433303833,
+      "learning_rate": 4.392455385078529e-06,
+      "loss": 0.8322,
+      "step": 4217
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3602685928344727,
+      "learning_rate": 4.3903490398906634e-06,
+      "loss": 0.8033,
+      "step": 4218
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3364691734313965,
+      "learning_rate": 4.388242804523149e-06,
+      "loss": 0.8035,
+      "step": 4219
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.609992504119873,
+      "learning_rate": 4.3861366793554e-06,
+      "loss": 1.0167,
+      "step": 4220
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3597116470336914,
+      "learning_rate": 4.3840306647668055e-06,
+      "loss": 0.8507,
+      "step": 4221
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3565317392349243,
+      "learning_rate": 4.381924761136733e-06,
+      "loss": 0.8294,
+      "step": 4222
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6215677857398987,
+      "learning_rate": 4.379818968844534e-06,
+      "loss": 1.0188,
+      "step": 4223
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3673412799835205,
+      "learning_rate": 4.377713288269539e-06,
+      "loss": 0.8957,
+      "step": 4224
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3446564674377441,
+      "learning_rate": 4.375607719791058e-06,
+      "loss": 0.8809,
+      "step": 4225
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2942662239074707,
+      "learning_rate": 4.37350226378838e-06,
+      "loss": 0.8511,
+      "step": 4226
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2689645290374756,
+      "learning_rate": 4.371396920640773e-06,
+      "loss": 0.8418,
+      "step": 4227
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.240302324295044,
+      "learning_rate": 4.369291690727489e-06,
+      "loss": 0.7963,
+      "step": 4228
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.332343578338623,
+      "learning_rate": 4.367186574427752e-06,
+      "loss": 0.8513,
+      "step": 4229
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3315887451171875,
+      "learning_rate": 4.365081572120776e-06,
+      "loss": 0.8355,
+      "step": 4230
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.2896432876586914,
+      "learning_rate": 4.362976684185746e-06,
+      "loss": 0.7776,
+      "step": 4231
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.3390673398971558,
+      "learning_rate": 4.360871911001826e-06,
+      "loss": 0.8253,
+      "step": 4232
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.315057396888733,
+      "learning_rate": 4.358767252948167e-06,
+      "loss": 0.8004,
+      "step": 4233
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6586228609085083,
+      "learning_rate": 4.356662710403891e-06,
+      "loss": 1.0035,
+      "step": 4234
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 1.4889296293258667,
+      "learning_rate": 4.354558283748107e-06,
+      "loss": 0.8593,
+      "step": 4235
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5606088638305664,
+      "learning_rate": 4.352453973359897e-06,
+      "loss": 0.8621,
+      "step": 4236
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3700438737869263,
+      "learning_rate": 4.3503497796183204e-06,
+      "loss": 0.8733,
+      "step": 4237
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3233450651168823,
+      "learning_rate": 4.348245702902425e-06,
+      "loss": 0.8106,
+      "step": 4238
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3507822751998901,
+      "learning_rate": 4.346141743591228e-06,
+      "loss": 0.8465,
+      "step": 4239
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6578850150108337,
+      "learning_rate": 4.344037902063728e-06,
+      "loss": 1.0469,
+      "step": 4240
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3149641752243042,
+      "learning_rate": 4.341934178698908e-06,
+      "loss": 0.8103,
+      "step": 4241
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.405917763710022,
+      "learning_rate": 4.339830573875722e-06,
+      "loss": 0.8428,
+      "step": 4242
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3472779989242554,
+      "learning_rate": 4.337727087973107e-06,
+      "loss": 0.8532,
+      "step": 4243
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6435452699661255,
+      "learning_rate": 4.335623721369977e-06,
+      "loss": 1.0363,
+      "step": 4244
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.8307856321334839,
+      "learning_rate": 4.333520474445225e-06,
+      "loss": 0.8773,
+      "step": 4245
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5997069478034973,
+      "learning_rate": 4.331417347577723e-06,
+      "loss": 1.0275,
+      "step": 4246
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2676888704299927,
+      "learning_rate": 4.329314341146319e-06,
+      "loss": 0.7926,
+      "step": 4247
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6248051524162292,
+      "learning_rate": 4.327211455529844e-06,
+      "loss": 1.01,
+      "step": 4248
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5871606469154358,
+      "learning_rate": 4.325108691107103e-06,
+      "loss": 1.0062,
+      "step": 4249
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3558568954467773,
+      "learning_rate": 4.323006048256878e-06,
+      "loss": 0.8875,
+      "step": 4250
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6118272542953491,
+      "learning_rate": 4.320903527357936e-06,
+      "loss": 1.0497,
+      "step": 4251
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.358523964881897,
+      "learning_rate": 4.318801128789013e-06,
+      "loss": 0.84,
+      "step": 4252
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6004492044448853,
+      "learning_rate": 4.316698852928832e-06,
+      "loss": 1.0312,
+      "step": 4253
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3548332452774048,
+      "learning_rate": 4.314596700156088e-06,
+      "loss": 0.8179,
+      "step": 4254
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3188755512237549,
+      "learning_rate": 4.312494670849453e-06,
+      "loss": 0.8318,
+      "step": 4255
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3838889598846436,
+      "learning_rate": 4.310392765387583e-06,
+      "loss": 0.7883,
+      "step": 4256
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3776048421859741,
+      "learning_rate": 4.3082909841491034e-06,
+      "loss": 0.8427,
+      "step": 4257
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2795385122299194,
+      "learning_rate": 4.306189327512626e-06,
+      "loss": 0.8502,
+      "step": 4258
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2738443613052368,
+      "learning_rate": 4.304087795856735e-06,
+      "loss": 0.8368,
+      "step": 4259
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6768737435340881,
+      "learning_rate": 4.301986389559989e-06,
+      "loss": 1.013,
+      "step": 4260
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6432403922080994,
+      "learning_rate": 4.299885109000931e-06,
+      "loss": 1.0367,
+      "step": 4261
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6549636125564575,
+      "learning_rate": 4.297783954558077e-06,
+      "loss": 0.9833,
+      "step": 4262
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4415335655212402,
+      "learning_rate": 4.295682926609923e-06,
+      "loss": 0.8626,
+      "step": 4263
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.5405739545822144,
+      "learning_rate": 4.29358202553494e-06,
+      "loss": 0.8195,
+      "step": 4264
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6505914330482483,
+      "learning_rate": 4.291481251711575e-06,
+      "loss": 1.0033,
+      "step": 4265
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.460250735282898,
+      "learning_rate": 4.289380605518258e-06,
+      "loss": 0.8259,
+      "step": 4266
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6513702273368835,
+      "learning_rate": 4.287280087333388e-06,
+      "loss": 1.0122,
+      "step": 4267
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6416940093040466,
+      "learning_rate": 4.285179697535345e-06,
+      "loss": 1.0181,
+      "step": 4268
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2901531457901,
+      "learning_rate": 4.283079436502489e-06,
+      "loss": 0.8505,
+      "step": 4269
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3473260402679443,
+      "learning_rate": 4.280979304613149e-06,
+      "loss": 0.8437,
+      "step": 4270
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3963547945022583,
+      "learning_rate": 4.278879302245641e-06,
+      "loss": 0.8628,
+      "step": 4271
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3363667726516724,
+      "learning_rate": 4.276779429778248e-06,
+      "loss": 0.8109,
+      "step": 4272
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4822109937667847,
+      "learning_rate": 4.274679687589232e-06,
+      "loss": 0.8649,
+      "step": 4273
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6789964437484741,
+      "learning_rate": 4.272580076056838e-06,
+      "loss": 1.0148,
+      "step": 4274
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3023569583892822,
+      "learning_rate": 4.270480595559278e-06,
+      "loss": 0.8712,
+      "step": 4275
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.300182819366455,
+      "learning_rate": 4.268381246474749e-06,
+      "loss": 0.894,
+      "step": 4276
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6211379766464233,
+      "learning_rate": 4.266282029181419e-06,
+      "loss": 1.01,
+      "step": 4277
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3280271291732788,
+      "learning_rate": 4.264182944057429e-06,
+      "loss": 0.8287,
+      "step": 4278
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.241971492767334,
+      "learning_rate": 4.262083991480908e-06,
+      "loss": 0.772,
+      "step": 4279
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6273932456970215,
+      "learning_rate": 4.259985171829949e-06,
+      "loss": 1.0071,
+      "step": 4280
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6108824610710144,
+      "learning_rate": 4.257886485482627e-06,
+      "loss": 0.9951,
+      "step": 4281
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.366946816444397,
+      "learning_rate": 4.255787932816993e-06,
+      "loss": 0.852,
+      "step": 4282
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6157119870185852,
+      "learning_rate": 4.253689514211071e-06,
+      "loss": 1.0085,
+      "step": 4283
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3397266864776611,
+      "learning_rate": 4.251591230042865e-06,
+      "loss": 0.8347,
+      "step": 4284
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3109440803527832,
+      "learning_rate": 4.249493080690348e-06,
+      "loss": 0.8295,
+      "step": 4285
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.4004814624786377,
+      "learning_rate": 4.247395066531479e-06,
+      "loss": 0.8509,
+      "step": 4286
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.9078677892684937,
+      "learning_rate": 4.245297187944184e-06,
+      "loss": 0.7884,
+      "step": 4287
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3146623373031616,
+      "learning_rate": 4.243199445306365e-06,
+      "loss": 0.8088,
+      "step": 4288
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.684705913066864,
+      "learning_rate": 4.241101838995907e-06,
+      "loss": 1.0108,
+      "step": 4289
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3249350786209106,
+      "learning_rate": 4.239004369390659e-06,
+      "loss": 0.8762,
+      "step": 4290
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.338261365890503,
+      "learning_rate": 4.236907036868458e-06,
+      "loss": 0.8489,
+      "step": 4291
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2691092491149902,
+      "learning_rate": 4.234809841807108e-06,
+      "loss": 0.8023,
+      "step": 4292
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6231737732887268,
+      "learning_rate": 4.2327127845843874e-06,
+      "loss": 1.0296,
+      "step": 4293
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3584355115890503,
+      "learning_rate": 4.230615865578057e-06,
+      "loss": 0.8426,
+      "step": 4294
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3296090364456177,
+      "learning_rate": 4.228519085165846e-06,
+      "loss": 0.8719,
+      "step": 4295
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.239111065864563,
+      "learning_rate": 4.226422443725458e-06,
+      "loss": 0.8381,
+      "step": 4296
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.3574250936508179,
+      "learning_rate": 4.2243259416345805e-06,
+      "loss": 0.8485,
+      "step": 4297
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7003766298294067,
+      "learning_rate": 4.222229579270865e-06,
+      "loss": 1.0025,
+      "step": 4298
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.371049404144287,
+      "learning_rate": 4.220133357011946e-06,
+      "loss": 0.838,
+      "step": 4299
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6275076866149902,
+      "learning_rate": 4.218037275235429e-06,
+      "loss": 1.0069,
+      "step": 4300
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.53190279006958,
+      "learning_rate": 4.2159413343188935e-06,
+      "loss": 0.827,
+      "step": 4301
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.2883942127227783,
+      "learning_rate": 4.213845534639895e-06,
+      "loss": 0.8164,
+      "step": 4302
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.709102213382721,
+      "learning_rate": 4.211749876575964e-06,
+      "loss": 1.0256,
+      "step": 4303
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7301206588745117,
+      "learning_rate": 4.209654360504605e-06,
+      "loss": 1.0351,
+      "step": 4304
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6341770887374878,
+      "learning_rate": 4.207558986803295e-06,
+      "loss": 1.0386,
+      "step": 4305
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7403583526611328,
+      "learning_rate": 4.20546375584949e-06,
+      "loss": 0.8284,
+      "step": 4306
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6325705051422119,
+      "learning_rate": 4.203368668020616e-06,
+      "loss": 1.0078,
+      "step": 4307
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.7281314134597778,
+      "learning_rate": 4.201273723694073e-06,
+      "loss": 0.8944,
+      "step": 4308
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7434083223342896,
+      "learning_rate": 4.199178923247241e-06,
+      "loss": 1.0204,
+      "step": 4309
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.7111446261405945,
+      "learning_rate": 4.197084267057466e-06,
+      "loss": 1.0259,
+      "step": 4310
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6176974773406982,
+      "learning_rate": 4.194989755502073e-06,
+      "loss": 1.0095,
+      "step": 4311
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6001999378204346,
+      "learning_rate": 4.192895388958361e-06,
+      "loss": 1.0209,
+      "step": 4312
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3805128335952759,
+      "learning_rate": 4.1908011678036e-06,
+      "loss": 0.8852,
+      "step": 4313
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.378409504890442,
+      "learning_rate": 4.188707092415038e-06,
+      "loss": 0.8783,
+      "step": 4314
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.5236696004867554,
+      "learning_rate": 4.186613163169892e-06,
+      "loss": 0.8216,
+      "step": 4315
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3334980010986328,
+      "learning_rate": 4.184519380445355e-06,
+      "loss": 0.8413,
+      "step": 4316
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7312199473381042,
+      "learning_rate": 4.182425744618595e-06,
+      "loss": 1.013,
+      "step": 4317
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6895466446876526,
+      "learning_rate": 4.18033225606675e-06,
+      "loss": 1.0228,
+      "step": 4318
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2864160537719727,
+      "learning_rate": 4.178238915166937e-06,
+      "loss": 0.8067,
+      "step": 4319
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.328148603439331,
+      "learning_rate": 4.176145722296241e-06,
+      "loss": 0.8717,
+      "step": 4320
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6152690052986145,
+      "learning_rate": 4.174052677831722e-06,
+      "loss": 1.0297,
+      "step": 4321
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3907384872436523,
+      "learning_rate": 4.171959782150412e-06,
+      "loss": 0.8637,
+      "step": 4322
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2744190692901611,
+      "learning_rate": 4.1698670356293205e-06,
+      "loss": 0.8167,
+      "step": 4323
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3148995637893677,
+      "learning_rate": 4.167774438645428e-06,
+      "loss": 0.7878,
+      "step": 4324
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3924752473831177,
+      "learning_rate": 4.165681991575686e-06,
+      "loss": 0.8325,
+      "step": 4325
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3674817085266113,
+      "learning_rate": 4.163589694797018e-06,
+      "loss": 0.8505,
+      "step": 4326
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.340960144996643,
+      "learning_rate": 4.161497548686327e-06,
+      "loss": 0.8446,
+      "step": 4327
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4394768476486206,
+      "learning_rate": 4.1594055536204835e-06,
+      "loss": 0.854,
+      "step": 4328
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.7100018858909607,
+      "learning_rate": 4.15731370997633e-06,
+      "loss": 1.0319,
+      "step": 4329
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2768704891204834,
+      "learning_rate": 4.1552220181306875e-06,
+      "loss": 0.7963,
+      "step": 4330
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2834656238555908,
+      "learning_rate": 4.153130478460341e-06,
+      "loss": 0.8163,
+      "step": 4331
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.325893759727478,
+      "learning_rate": 4.151039091342058e-06,
+      "loss": 0.8667,
+      "step": 4332
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.298163890838623,
+      "learning_rate": 4.148947857152572e-06,
+      "loss": 0.841,
+      "step": 4333
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3414536714553833,
+      "learning_rate": 4.146856776268588e-06,
+      "loss": 0.8682,
+      "step": 4334
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.449724793434143,
+      "learning_rate": 4.14476584906679e-06,
+      "loss": 0.8048,
+      "step": 4335
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2798950672149658,
+      "learning_rate": 4.142675075923825e-06,
+      "loss": 0.8181,
+      "step": 4336
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6494815349578857,
+      "learning_rate": 4.1405844572163225e-06,
+      "loss": 1.0371,
+      "step": 4337
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.312556266784668,
+      "learning_rate": 4.138493993320877e-06,
+      "loss": 0.8038,
+      "step": 4338
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2548518180847168,
+      "learning_rate": 4.136403684614055e-06,
+      "loss": 0.8192,
+      "step": 4339
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3458620309829712,
+      "learning_rate": 4.1343135314724026e-06,
+      "loss": 0.8187,
+      "step": 4340
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3538552522659302,
+      "learning_rate": 4.132223534272427e-06,
+      "loss": 0.872,
+      "step": 4341
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6135764122009277,
+      "learning_rate": 4.130133693390616e-06,
+      "loss": 1.0072,
+      "step": 4342
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4513514041900635,
+      "learning_rate": 4.128044009203427e-06,
+      "loss": 0.9051,
+      "step": 4343
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.561801791191101,
+      "learning_rate": 4.125954482087284e-06,
+      "loss": 0.8421,
+      "step": 4344
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3489491939544678,
+      "learning_rate": 4.123865112418593e-06,
+      "loss": 0.8801,
+      "step": 4345
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3173381090164185,
+      "learning_rate": 4.121775900573718e-06,
+      "loss": 0.7986,
+      "step": 4346
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.309416651725769,
+      "learning_rate": 4.119686846929009e-06,
+      "loss": 0.8289,
+      "step": 4347
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6241777539253235,
+      "learning_rate": 4.117597951860779e-06,
+      "loss": 1.0338,
+      "step": 4348
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2869277000427246,
+      "learning_rate": 4.11550921574531e-06,
+      "loss": 0.9299,
+      "step": 4349
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6143568754196167,
+      "learning_rate": 4.113420638958866e-06,
+      "loss": 1.015,
+      "step": 4350
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3531816005706787,
+      "learning_rate": 4.111332221877669e-06,
+      "loss": 0.882,
+      "step": 4351
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2701150178909302,
+      "learning_rate": 4.109243964877926e-06,
+      "loss": 0.8711,
+      "step": 4352
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.0288004875183105,
+      "learning_rate": 4.1071558683358035e-06,
+      "loss": 0.8936,
+      "step": 4353
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6160069704055786,
+      "learning_rate": 4.105067932627443e-06,
+      "loss": 1.034,
+      "step": 4354
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.29619300365448,
+      "learning_rate": 4.102980158128963e-06,
+      "loss": 0.8196,
+      "step": 4355
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.304351806640625,
+      "learning_rate": 4.1008925452164435e-06,
+      "loss": 0.7819,
+      "step": 4356
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.289900302886963,
+      "learning_rate": 4.098805094265939e-06,
+      "loss": 0.8478,
+      "step": 4357
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.556833267211914,
+      "learning_rate": 4.09671780565348e-06,
+      "loss": 0.8938,
+      "step": 4358
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3781344890594482,
+      "learning_rate": 4.094630679755058e-06,
+      "loss": 0.7491,
+      "step": 4359
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.330119252204895,
+      "learning_rate": 4.092543716946644e-06,
+      "loss": 0.8257,
+      "step": 4360
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.463096022605896,
+      "learning_rate": 4.090456917604176e-06,
+      "loss": 0.8321,
+      "step": 4361
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4322071075439453,
+      "learning_rate": 4.08837028210356e-06,
+      "loss": 0.8504,
+      "step": 4362
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.4320124387741089,
+      "learning_rate": 4.086283810820677e-06,
+      "loss": 0.8344,
+      "step": 4363
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6648374795913696,
+      "learning_rate": 4.0841975041313766e-06,
+      "loss": 1.0183,
+      "step": 4364
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6323179602622986,
+      "learning_rate": 4.082111362411478e-06,
+      "loss": 1.018,
+      "step": 4365
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3145687580108643,
+      "learning_rate": 4.0800253860367725e-06,
+      "loss": 0.8232,
+      "step": 4366
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3488386869430542,
+      "learning_rate": 4.077939575383017e-06,
+      "loss": 0.8848,
+      "step": 4367
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2911027669906616,
+      "learning_rate": 4.075853930825947e-06,
+      "loss": 0.8338,
+      "step": 4368
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3238348960876465,
+      "learning_rate": 4.073768452741258e-06,
+      "loss": 0.8242,
+      "step": 4369
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.285727620124817,
+      "learning_rate": 4.071683141504625e-06,
+      "loss": 0.8204,
+      "step": 4370
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2794731855392456,
+      "learning_rate": 4.069597997491688e-06,
+      "loss": 0.8277,
+      "step": 4371
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3096588850021362,
+      "learning_rate": 4.0675130210780525e-06,
+      "loss": 0.8458,
+      "step": 4372
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3423362970352173,
+      "learning_rate": 4.065428212639306e-06,
+      "loss": 0.8453,
+      "step": 4373
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.8356714248657227,
+      "learning_rate": 4.063343572550992e-06,
+      "loss": 0.8147,
+      "step": 4374
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2753221988677979,
+      "learning_rate": 4.061259101188635e-06,
+      "loss": 0.8689,
+      "step": 4375
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 2.143887758255005,
+      "learning_rate": 4.0591747989277225e-06,
+      "loss": 0.822,
+      "step": 4376
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3400251865386963,
+      "learning_rate": 4.057090666143711e-06,
+      "loss": 0.8119,
+      "step": 4377
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6729621291160583,
+      "learning_rate": 4.055006703212033e-06,
+      "loss": 1.0109,
+      "step": 4378
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2725101709365845,
+      "learning_rate": 4.052922910508083e-06,
+      "loss": 0.7692,
+      "step": 4379
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.457771897315979,
+      "learning_rate": 4.0508392884072285e-06,
+      "loss": 0.8791,
+      "step": 4380
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.2828466892242432,
+      "learning_rate": 4.048755837284807e-06,
+      "loss": 0.8213,
+      "step": 4381
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.9227179288864136,
+      "learning_rate": 4.0466725575161235e-06,
+      "loss": 0.8484,
+      "step": 4382
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6188197135925293,
+      "learning_rate": 4.044589449476451e-06,
+      "loss": 1.0267,
+      "step": 4383
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3623535633087158,
+      "learning_rate": 4.0425065135410364e-06,
+      "loss": 0.8715,
+      "step": 4384
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3626925945281982,
+      "learning_rate": 4.040423750085089e-06,
+      "loss": 0.8355,
+      "step": 4385
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6144638061523438,
+      "learning_rate": 4.038341159483793e-06,
+      "loss": 1.0252,
+      "step": 4386
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.867006540298462,
+      "learning_rate": 4.036258742112296e-06,
+      "loss": 0.8117,
+      "step": 4387
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6093570590019226,
+      "learning_rate": 4.03417649834572e-06,
+      "loss": 1.0338,
+      "step": 4388
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6007617712020874,
+      "learning_rate": 4.032094428559153e-06,
+      "loss": 1.0166,
+      "step": 4389
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3975183963775635,
+      "learning_rate": 4.030012533127649e-06,
+      "loss": 0.8197,
+      "step": 4390
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3937207460403442,
+      "learning_rate": 4.027930812426236e-06,
+      "loss": 0.8843,
+      "step": 4391
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4817568063735962,
+      "learning_rate": 4.0258492668299045e-06,
+      "loss": 0.8776,
+      "step": 4392
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5946435332298279,
+      "learning_rate": 4.023767896713621e-06,
+      "loss": 1.014,
+      "step": 4393
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3733021020889282,
+      "learning_rate": 4.021686702452313e-06,
+      "loss": 0.8648,
+      "step": 4394
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2985998392105103,
+      "learning_rate": 4.01960568442088e-06,
+      "loss": 0.8785,
+      "step": 4395
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3544597625732422,
+      "learning_rate": 4.017524842994191e-06,
+      "loss": 0.8342,
+      "step": 4396
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6148921251296997,
+      "learning_rate": 4.015444178547076e-06,
+      "loss": 1.0082,
+      "step": 4397
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3223884105682373,
+      "learning_rate": 4.013363691454347e-06,
+      "loss": 0.8102,
+      "step": 4398
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5061655044555664,
+      "learning_rate": 4.01128338209077e-06,
+      "loss": 0.8027,
+      "step": 4399
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4727212190628052,
+      "learning_rate": 4.009203250831083e-06,
+      "loss": 0.8479,
+      "step": 4400
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3038647174835205,
+      "learning_rate": 4.007123298049998e-06,
+      "loss": 0.7874,
+      "step": 4401
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.316727638244629,
+      "learning_rate": 4.0050435241221876e-06,
+      "loss": 0.8709,
+      "step": 4402
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.621134340763092,
+      "learning_rate": 4.002963929422296e-06,
+      "loss": 0.9948,
+      "step": 4403
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3540233373641968,
+      "learning_rate": 4.0008845143249325e-06,
+      "loss": 0.8584,
+      "step": 4404
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3163962364196777,
+      "learning_rate": 3.998805279204676e-06,
+      "loss": 0.8359,
+      "step": 4405
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5958545207977295,
+      "learning_rate": 3.996726224436075e-06,
+      "loss": 1.0146,
+      "step": 4406
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3500150442123413,
+      "learning_rate": 3.994647350393638e-06,
+      "loss": 0.8261,
+      "step": 4407
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.4569056034088135,
+      "learning_rate": 3.992568657451852e-06,
+      "loss": 0.8625,
+      "step": 4408
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2885924577713013,
+      "learning_rate": 3.990490145985162e-06,
+      "loss": 0.7759,
+      "step": 4409
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.210986852645874,
+      "learning_rate": 3.988411816367981e-06,
+      "loss": 0.8244,
+      "step": 4410
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6103920936584473,
+      "learning_rate": 3.986333668974698e-06,
+      "loss": 1.0369,
+      "step": 4411
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3924704790115356,
+      "learning_rate": 3.9842557041796595e-06,
+      "loss": 0.831,
+      "step": 4412
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.510992169380188,
+      "learning_rate": 3.982177922357181e-06,
+      "loss": 0.8254,
+      "step": 4413
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6157448887825012,
+      "learning_rate": 3.980100323881551e-06,
+      "loss": 1.0241,
+      "step": 4414
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3057377338409424,
+      "learning_rate": 3.978022909127017e-06,
+      "loss": 0.8058,
+      "step": 4415
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3238290548324585,
+      "learning_rate": 3.9759456784678e-06,
+      "loss": 0.8511,
+      "step": 4416
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3604916334152222,
+      "learning_rate": 3.973868632278084e-06,
+      "loss": 0.824,
+      "step": 4417
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5881941914558411,
+      "learning_rate": 3.971791770932018e-06,
+      "loss": 1.0276,
+      "step": 4418
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3219891786575317,
+      "learning_rate": 3.969715094803724e-06,
+      "loss": 0.8326,
+      "step": 4419
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5984787940979004,
+      "learning_rate": 3.967638604267286e-06,
+      "loss": 1.0094,
+      "step": 4420
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6019412279129028,
+      "learning_rate": 3.965562299696754e-06,
+      "loss": 1.0236,
+      "step": 4421
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.288964033126831,
+      "learning_rate": 3.963486181466151e-06,
+      "loss": 0.7951,
+      "step": 4422
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3492441177368164,
+      "learning_rate": 3.961410249949455e-06,
+      "loss": 0.8177,
+      "step": 4423
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3169221878051758,
+      "learning_rate": 3.95933450552062e-06,
+      "loss": 0.7808,
+      "step": 4424
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3775008916854858,
+      "learning_rate": 3.957258948553563e-06,
+      "loss": 0.8767,
+      "step": 4425
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5963233113288879,
+      "learning_rate": 3.95518357942217e-06,
+      "loss": 1.0344,
+      "step": 4426
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2551335096359253,
+      "learning_rate": 3.953108398500287e-06,
+      "loss": 0.7857,
+      "step": 4427
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3815860748291016,
+      "learning_rate": 3.951033406161729e-06,
+      "loss": 0.8492,
+      "step": 4428
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3717334270477295,
+      "learning_rate": 3.948958602780283e-06,
+      "loss": 0.8339,
+      "step": 4429
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3378628492355347,
+      "learning_rate": 3.9468839887296895e-06,
+      "loss": 0.8091,
+      "step": 4430
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6423127055168152,
+      "learning_rate": 3.944809564383669e-06,
+      "loss": 1.029,
+      "step": 4431
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3467776775360107,
+      "learning_rate": 3.942735330115897e-06,
+      "loss": 0.9254,
+      "step": 4432
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.351557731628418,
+      "learning_rate": 3.940661286300018e-06,
+      "loss": 0.7934,
+      "step": 4433
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3477897644042969,
+      "learning_rate": 3.938587433309646e-06,
+      "loss": 0.8396,
+      "step": 4434
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2517309188842773,
+      "learning_rate": 3.936513771518354e-06,
+      "loss": 0.8017,
+      "step": 4435
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.489848017692566,
+      "learning_rate": 3.934440301299689e-06,
+      "loss": 0.8584,
+      "step": 4436
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2756319046020508,
+      "learning_rate": 3.932367023027155e-06,
+      "loss": 0.8315,
+      "step": 4437
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.487576961517334,
+      "learning_rate": 3.930293937074224e-06,
+      "loss": 0.8582,
+      "step": 4438
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.300359845161438,
+      "learning_rate": 3.928221043814336e-06,
+      "loss": 0.8269,
+      "step": 4439
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6180042624473572,
+      "learning_rate": 3.926148343620896e-06,
+      "loss": 1.0239,
+      "step": 4440
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.6352462768554688,
+      "learning_rate": 3.924075836867268e-06,
+      "loss": 0.8321,
+      "step": 4441
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6129339933395386,
+      "learning_rate": 3.922003523926791e-06,
+      "loss": 1.0088,
+      "step": 4442
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3214023113250732,
+      "learning_rate": 3.919931405172763e-06,
+      "loss": 0.8268,
+      "step": 4443
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.260480523109436,
+      "learning_rate": 3.917859480978447e-06,
+      "loss": 0.8197,
+      "step": 4444
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3734797239303589,
+      "learning_rate": 3.915787751717073e-06,
+      "loss": 0.7929,
+      "step": 4445
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.390723705291748,
+      "learning_rate": 3.9137162177618335e-06,
+      "loss": 0.868,
+      "step": 4446
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2769676446914673,
+      "learning_rate": 3.911644879485889e-06,
+      "loss": 0.8536,
+      "step": 4447
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3294411897659302,
+      "learning_rate": 3.90957373726236e-06,
+      "loss": 0.7961,
+      "step": 4448
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2597644329071045,
+      "learning_rate": 3.907502791464339e-06,
+      "loss": 0.8298,
+      "step": 4449
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.291412353515625,
+      "learning_rate": 3.905432042464877e-06,
+      "loss": 0.8571,
+      "step": 4450
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3872605562210083,
+      "learning_rate": 3.903361490636989e-06,
+      "loss": 0.7899,
+      "step": 4451
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6899526119232178,
+      "learning_rate": 3.90129113635366e-06,
+      "loss": 1.013,
+      "step": 4452
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3385249376296997,
+      "learning_rate": 3.899220979987834e-06,
+      "loss": 0.8069,
+      "step": 4453
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.2895727157592773,
+      "learning_rate": 3.8971510219124246e-06,
+      "loss": 0.847,
+      "step": 4454
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.638158917427063,
+      "learning_rate": 3.8950812625003044e-06,
+      "loss": 1.0156,
+      "step": 4455
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3420958518981934,
+      "learning_rate": 3.893011702124311e-06,
+      "loss": 0.7754,
+      "step": 4456
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.338675856590271,
+      "learning_rate": 3.890942341157251e-06,
+      "loss": 0.873,
+      "step": 4457
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.616894006729126,
+      "learning_rate": 3.8888731799718885e-06,
+      "loss": 1.0111,
+      "step": 4458
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6334955096244812,
+      "learning_rate": 3.886804218940959e-06,
+      "loss": 1.031,
+      "step": 4459
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6233447194099426,
+      "learning_rate": 3.884735458437155e-06,
+      "loss": 1.0028,
+      "step": 4460
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6182667016983032,
+      "learning_rate": 3.882666898833135e-06,
+      "loss": 1.0314,
+      "step": 4461
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.5492700338363647,
+      "learning_rate": 3.880598540501523e-06,
+      "loss": 0.8592,
+      "step": 4462
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.612374484539032,
+      "learning_rate": 3.878530383814905e-06,
+      "loss": 1.0302,
+      "step": 4463
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.3140034675598145,
+      "learning_rate": 3.8764624291458346e-06,
+      "loss": 0.8108,
+      "step": 4464
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6031137704849243,
+      "learning_rate": 3.874394676866821e-06,
+      "loss": 1.0289,
+      "step": 4465
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3509628772735596,
+      "learning_rate": 3.872327127350344e-06,
+      "loss": 0.8042,
+      "step": 4466
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3143104314804077,
+      "learning_rate": 3.870259780968846e-06,
+      "loss": 0.8067,
+      "step": 4467
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6119428873062134,
+      "learning_rate": 3.868192638094727e-06,
+      "loss": 1.0357,
+      "step": 4468
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3307017087936401,
+      "learning_rate": 3.866125699100361e-06,
+      "loss": 0.8857,
+      "step": 4469
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6579880714416504,
+      "learning_rate": 3.864058964358075e-06,
+      "loss": 1.0329,
+      "step": 4470
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.322471261024475,
+      "learning_rate": 3.8619924342401625e-06,
+      "loss": 0.8279,
+      "step": 4471
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3062114715576172,
+      "learning_rate": 3.859926109118884e-06,
+      "loss": 0.8303,
+      "step": 4472
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6312189698219299,
+      "learning_rate": 3.857859989366459e-06,
+      "loss": 1.0088,
+      "step": 4473
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2772213220596313,
+      "learning_rate": 3.855794075355068e-06,
+      "loss": 0.8719,
+      "step": 4474
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3171578645706177,
+      "learning_rate": 3.853728367456862e-06,
+      "loss": 0.7873,
+      "step": 4475
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2810213565826416,
+      "learning_rate": 3.851662866043945e-06,
+      "loss": 0.8749,
+      "step": 4476
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3331599235534668,
+      "learning_rate": 3.849597571488395e-06,
+      "loss": 0.865,
+      "step": 4477
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3266642093658447,
+      "learning_rate": 3.847532484162244e-06,
+      "loss": 0.81,
+      "step": 4478
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3032063245773315,
+      "learning_rate": 3.845467604437486e-06,
+      "loss": 0.8057,
+      "step": 4479
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6823522448539734,
+      "learning_rate": 3.843402932686088e-06,
+      "loss": 1.0139,
+      "step": 4480
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6724558472633362,
+      "learning_rate": 3.841338469279967e-06,
+      "loss": 1.0157,
+      "step": 4481
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4054315090179443,
+      "learning_rate": 3.839274214591011e-06,
+      "loss": 0.8574,
+      "step": 4482
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.601162314414978,
+      "learning_rate": 3.8372101689910666e-06,
+      "loss": 1.0271,
+      "step": 4483
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4559149742126465,
+      "learning_rate": 3.835146332851943e-06,
+      "loss": 0.7971,
+      "step": 4484
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3009883165359497,
+      "learning_rate": 3.833082706545414e-06,
+      "loss": 0.8554,
+      "step": 4485
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.317020297050476,
+      "learning_rate": 3.831019290443211e-06,
+      "loss": 0.8895,
+      "step": 4486
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6592246890068054,
+      "learning_rate": 3.828956084917035e-06,
+      "loss": 1.0067,
+      "step": 4487
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4607181549072266,
+      "learning_rate": 3.826893090338541e-06,
+      "loss": 0.887,
+      "step": 4488
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2919692993164062,
+      "learning_rate": 3.824830307079348e-06,
+      "loss": 0.7807,
+      "step": 4489
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3474196195602417,
+      "learning_rate": 3.822767735511043e-06,
+      "loss": 0.8431,
+      "step": 4490
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2831926345825195,
+      "learning_rate": 3.820705376005166e-06,
+      "loss": 0.7564,
+      "step": 4491
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6524324417114258,
+      "learning_rate": 3.818643228933228e-06,
+      "loss": 1.038,
+      "step": 4492
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.364243745803833,
+      "learning_rate": 3.816581294666693e-06,
+      "loss": 0.7951,
+      "step": 4493
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6040434241294861,
+      "learning_rate": 3.814519573576989e-06,
+      "loss": 1.0317,
+      "step": 4494
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.602460503578186,
+      "learning_rate": 3.8124580660355128e-06,
+      "loss": 1.0402,
+      "step": 4495
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3370012044906616,
+      "learning_rate": 3.810396772413611e-06,
+      "loss": 0.8451,
+      "step": 4496
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3899962902069092,
+      "learning_rate": 3.8083356930826025e-06,
+      "loss": 0.7678,
+      "step": 4497
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3455889225006104,
+      "learning_rate": 3.8062748284137612e-06,
+      "loss": 0.8347,
+      "step": 4498
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3711481094360352,
+      "learning_rate": 3.8042141787783215e-06,
+      "loss": 0.9054,
+      "step": 4499
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3899433612823486,
+      "learning_rate": 3.8021537445474854e-06,
+      "loss": 0.8363,
+      "step": 4500
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6539288759231567,
+      "learning_rate": 3.80009352609241e-06,
+      "loss": 1.0262,
+      "step": 4501
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3395696878433228,
+      "learning_rate": 3.798033523784214e-06,
+      "loss": 0.8533,
+      "step": 4502
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3214420080184937,
+      "learning_rate": 3.795973737993983e-06,
+      "loss": 0.8232,
+      "step": 4503
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2818608283996582,
+      "learning_rate": 3.7939141690927567e-06,
+      "loss": 0.7463,
+      "step": 4504
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6383938193321228,
+      "learning_rate": 3.79185481745154e-06,
+      "loss": 1.0313,
+      "step": 4505
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6326167583465576,
+      "learning_rate": 3.789795683441295e-06,
+      "loss": 1.0236,
+      "step": 4506
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.23043954372406,
+      "learning_rate": 3.7877367674329484e-06,
+      "loss": 0.8725,
+      "step": 4507
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.5408155918121338,
+      "learning_rate": 3.785678069797387e-06,
+      "loss": 0.8237,
+      "step": 4508
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.602781355381012,
+      "learning_rate": 3.7836195909054534e-06,
+      "loss": 1.0135,
+      "step": 4509
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3712421655654907,
+      "learning_rate": 3.78156133112796e-06,
+      "loss": 0.8292,
+      "step": 4510
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3395766019821167,
+      "learning_rate": 3.779503290835671e-06,
+      "loss": 0.7648,
+      "step": 4511
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6402292251586914,
+      "learning_rate": 3.777445470399313e-06,
+      "loss": 1.0189,
+      "step": 4512
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.341883659362793,
+      "learning_rate": 3.7753878701895792e-06,
+      "loss": 0.8922,
+      "step": 4513
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2909839153289795,
+      "learning_rate": 3.773330490577113e-06,
+      "loss": 0.8893,
+      "step": 4514
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4953252077102661,
+      "learning_rate": 3.7712733319325288e-06,
+      "loss": 0.821,
+      "step": 4515
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6154786944389343,
+      "learning_rate": 3.7692163946263925e-06,
+      "loss": 1.011,
+      "step": 4516
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6136094927787781,
+      "learning_rate": 3.767159679029233e-06,
+      "loss": 1.0279,
+      "step": 4517
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3658922910690308,
+      "learning_rate": 3.765103185511543e-06,
+      "loss": 0.8025,
+      "step": 4518
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5944488048553467,
+      "learning_rate": 3.7630469144437664e-06,
+      "loss": 0.9928,
+      "step": 4519
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.297043800354004,
+      "learning_rate": 3.760990866196319e-06,
+      "loss": 0.8529,
+      "step": 4520
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6172482371330261,
+      "learning_rate": 3.7589350411395674e-06,
+      "loss": 1.0176,
+      "step": 4521
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2667033672332764,
+      "learning_rate": 3.7568794396438387e-06,
+      "loss": 0.8019,
+      "step": 4522
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3568720817565918,
+      "learning_rate": 3.7548240620794228e-06,
+      "loss": 0.84,
+      "step": 4523
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.503808617591858,
+      "learning_rate": 3.7527689088165687e-06,
+      "loss": 0.8306,
+      "step": 4524
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.47563898563385,
+      "learning_rate": 3.7507139802254855e-06,
+      "loss": 0.7944,
+      "step": 4525
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3580389022827148,
+      "learning_rate": 3.7486592766763375e-06,
+      "loss": 0.8018,
+      "step": 4526
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.3279876708984375,
+      "learning_rate": 3.746604798539254e-06,
+      "loss": 0.8381,
+      "step": 4527
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.321740746498108,
+      "learning_rate": 3.7445505461843225e-06,
+      "loss": 0.8531,
+      "step": 4528
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2569340467453003,
+      "learning_rate": 3.7424965199815866e-06,
+      "loss": 0.779,
+      "step": 4529
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2761088609695435,
+      "learning_rate": 3.7404427203010497e-06,
+      "loss": 0.7811,
+      "step": 4530
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2859938144683838,
+      "learning_rate": 3.73838914751268e-06,
+      "loss": 0.8365,
+      "step": 4531
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.465101957321167,
+      "learning_rate": 3.736335801986396e-06,
+      "loss": 0.791,
+      "step": 4532
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.2292996644973755,
+      "learning_rate": 3.7342826840920863e-06,
+      "loss": 0.804,
+      "step": 4533
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7550904750823975,
+      "learning_rate": 3.7322297941995884e-06,
+      "loss": 1.0232,
+      "step": 4534
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7003620266914368,
+      "learning_rate": 3.7301771326787008e-06,
+      "loss": 1.0178,
+      "step": 4535
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.434975028038025,
+      "learning_rate": 3.728124699899187e-06,
+      "loss": 0.8845,
+      "step": 4536
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.4802097082138062,
+      "learning_rate": 3.7260724962307605e-06,
+      "loss": 0.8396,
+      "step": 4537
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 7.973137378692627,
+      "learning_rate": 3.7240205220431025e-06,
+      "loss": 0.7626,
+      "step": 4538
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.397182583808899,
+      "learning_rate": 3.7219687777058468e-06,
+      "loss": 0.7936,
+      "step": 4539
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6543964147567749,
+      "learning_rate": 3.719917263588584e-06,
+      "loss": 1.0303,
+      "step": 4540
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3018274307250977,
+      "learning_rate": 3.7178659800608706e-06,
+      "loss": 0.8431,
+      "step": 4541
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3500641584396362,
+      "learning_rate": 3.715814927492216e-06,
+      "loss": 0.7882,
+      "step": 4542
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4149081707000732,
+      "learning_rate": 3.7137641062520892e-06,
+      "loss": 0.9136,
+      "step": 4543
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6284478902816772,
+      "learning_rate": 3.7117135167099194e-06,
+      "loss": 1.0211,
+      "step": 4544
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6697445511817932,
+      "learning_rate": 3.7096631592350895e-06,
+      "loss": 1.0049,
+      "step": 4545
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.316197156906128,
+      "learning_rate": 3.7076130341969474e-06,
+      "loss": 0.8073,
+      "step": 4546
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.369402289390564,
+      "learning_rate": 3.705563141964791e-06,
+      "loss": 0.7805,
+      "step": 4547
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.358659267425537,
+      "learning_rate": 3.7035134829078844e-06,
+      "loss": 0.8687,
+      "step": 4548
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3290830850601196,
+      "learning_rate": 3.7014640573954454e-06,
+      "loss": 0.9019,
+      "step": 4549
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.359511137008667,
+      "learning_rate": 3.6994148657966465e-06,
+      "loss": 0.8346,
+      "step": 4550
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3233821392059326,
+      "learning_rate": 3.697365908480627e-06,
+      "loss": 0.8476,
+      "step": 4551
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5329838991165161,
+      "learning_rate": 3.6953171858164745e-06,
+      "loss": 0.7847,
+      "step": 4552
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8176588416099548,
+      "learning_rate": 3.693268698173242e-06,
+      "loss": 1.0045,
+      "step": 4553
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3795976638793945,
+      "learning_rate": 3.6912204459199352e-06,
+      "loss": 0.8597,
+      "step": 4554
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 4.039713382720947,
+      "learning_rate": 3.689172429425517e-06,
+      "loss": 0.8308,
+      "step": 4555
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3498338460922241,
+      "learning_rate": 3.6871246490589143e-06,
+      "loss": 0.7836,
+      "step": 4556
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.36821711063385,
+      "learning_rate": 3.6850771051890035e-06,
+      "loss": 0.8164,
+      "step": 4557
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4597225189208984,
+      "learning_rate": 3.6830297981846206e-06,
+      "loss": 0.8217,
+      "step": 4558
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4431318044662476,
+      "learning_rate": 3.6809827284145655e-06,
+      "loss": 0.7994,
+      "step": 4559
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6535739302635193,
+      "learning_rate": 3.678935896247584e-06,
+      "loss": 1.0134,
+      "step": 4560
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.439640760421753,
+      "learning_rate": 3.676889302052391e-06,
+      "loss": 0.8574,
+      "step": 4561
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3824115991592407,
+      "learning_rate": 3.6748429461976493e-06,
+      "loss": 0.8612,
+      "step": 4562
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2881293296813965,
+      "learning_rate": 3.6727968290519816e-06,
+      "loss": 0.8449,
+      "step": 4563
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4003130197525024,
+      "learning_rate": 3.6707509509839696e-06,
+      "loss": 0.8559,
+      "step": 4564
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4031902551651,
+      "learning_rate": 3.6687053123621497e-06,
+      "loss": 0.7582,
+      "step": 4565
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3651176691055298,
+      "learning_rate": 3.6666599135550172e-06,
+      "loss": 0.8524,
+      "step": 4566
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.413513422012329,
+      "learning_rate": 3.6646147549310217e-06,
+      "loss": 0.856,
+      "step": 4567
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3568941354751587,
+      "learning_rate": 3.662569836858571e-06,
+      "loss": 0.8762,
+      "step": 4568
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.291898250579834,
+      "learning_rate": 3.660525159706031e-06,
+      "loss": 0.7551,
+      "step": 4569
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6549327969551086,
+      "learning_rate": 3.6584807238417185e-06,
+      "loss": 1.0142,
+      "step": 4570
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.8122671842575073,
+      "learning_rate": 3.6564365296339153e-06,
+      "loss": 0.8562,
+      "step": 4571
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2402050495147705,
+      "learning_rate": 3.6543925774508546e-06,
+      "loss": 0.7618,
+      "step": 4572
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6355399489402771,
+      "learning_rate": 3.6523488676607235e-06,
+      "loss": 1.0075,
+      "step": 4573
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3952125310897827,
+      "learning_rate": 3.650305400631672e-06,
+      "loss": 0.8358,
+      "step": 4574
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.251120686531067,
+      "learning_rate": 3.6482621767317995e-06,
+      "loss": 0.8083,
+      "step": 4575
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6189517378807068,
+      "learning_rate": 3.64621919632917e-06,
+      "loss": 1.0281,
+      "step": 4576
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6077256798744202,
+      "learning_rate": 3.6441764597917954e-06,
+      "loss": 1.0026,
+      "step": 4577
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6235523223876953,
+      "learning_rate": 3.642133967487646e-06,
+      "loss": 1.0216,
+      "step": 4578
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.303275227546692,
+      "learning_rate": 3.640091719784653e-06,
+      "loss": 0.8095,
+      "step": 4579
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4019825458526611,
+      "learning_rate": 3.6380497170506957e-06,
+      "loss": 0.7961,
+      "step": 4580
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.356007695198059,
+      "learning_rate": 3.636007959653615e-06,
+      "loss": 0.8343,
+      "step": 4581
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.337503433227539,
+      "learning_rate": 3.6339664479612064e-06,
+      "loss": 0.8035,
+      "step": 4582
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.6551592350006104,
+      "learning_rate": 3.6319251823412187e-06,
+      "loss": 0.8323,
+      "step": 4583
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2385071516036987,
+      "learning_rate": 3.6298841631613598e-06,
+      "loss": 0.7602,
+      "step": 4584
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3069499731063843,
+      "learning_rate": 3.6278433907892906e-06,
+      "loss": 0.8096,
+      "step": 4585
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2796006202697754,
+      "learning_rate": 3.62580286559263e-06,
+      "loss": 0.8562,
+      "step": 4586
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.413069725036621,
+      "learning_rate": 3.623762587938951e-06,
+      "loss": 0.9148,
+      "step": 4587
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.300134301185608,
+      "learning_rate": 3.6217225581957784e-06,
+      "loss": 0.8438,
+      "step": 4588
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3840373754501343,
+      "learning_rate": 3.6196827767306003e-06,
+      "loss": 0.8602,
+      "step": 4589
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2964576482772827,
+      "learning_rate": 3.6176432439108545e-06,
+      "loss": 0.8246,
+      "step": 4590
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2731820344924927,
+      "learning_rate": 3.6156039601039315e-06,
+      "loss": 0.8271,
+      "step": 4591
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3204954862594604,
+      "learning_rate": 3.6135649256771865e-06,
+      "loss": 0.8177,
+      "step": 4592
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3466482162475586,
+      "learning_rate": 3.6115261409979183e-06,
+      "loss": 0.835,
+      "step": 4593
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6562184691429138,
+      "learning_rate": 3.6094876064333917e-06,
+      "loss": 1.01,
+      "step": 4594
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.288568377494812,
+      "learning_rate": 3.607449322350819e-06,
+      "loss": 0.8569,
+      "step": 4595
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5194250345230103,
+      "learning_rate": 3.6054112891173657e-06,
+      "loss": 0.8568,
+      "step": 4596
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4001011848449707,
+      "learning_rate": 3.603373507100162e-06,
+      "loss": 0.8525,
+      "step": 4597
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2644989490509033,
+      "learning_rate": 3.6013359766662814e-06,
+      "loss": 0.881,
+      "step": 4598
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3310467004776,
+      "learning_rate": 3.5992986981827615e-06,
+      "loss": 0.8158,
+      "step": 4599
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.385895013809204,
+      "learning_rate": 3.597261672016589e-06,
+      "loss": 0.8471,
+      "step": 4600
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.7915438413619995,
+      "learning_rate": 3.5952248985347037e-06,
+      "loss": 0.8271,
+      "step": 4601
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3392870426177979,
+      "learning_rate": 3.5931883781040078e-06,
+      "loss": 0.7748,
+      "step": 4602
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.396795392036438,
+      "learning_rate": 3.5911521110913483e-06,
+      "loss": 0.8575,
+      "step": 4603
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3277795314788818,
+      "learning_rate": 3.5891160978635348e-06,
+      "loss": 0.8062,
+      "step": 4604
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3029422760009766,
+      "learning_rate": 3.5870803387873243e-06,
+      "loss": 0.8374,
+      "step": 4605
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.64401775598526,
+      "learning_rate": 3.585044834229433e-06,
+      "loss": 0.9873,
+      "step": 4606
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6350314021110535,
+      "learning_rate": 3.5830095845565297e-06,
+      "loss": 1.0312,
+      "step": 4607
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.2219479084014893,
+      "learning_rate": 3.5809745901352344e-06,
+      "loss": 0.8139,
+      "step": 4608
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3548474311828613,
+      "learning_rate": 3.5789398513321284e-06,
+      "loss": 0.814,
+      "step": 4609
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.5909271240234375,
+      "learning_rate": 3.576905368513739e-06,
+      "loss": 0.8158,
+      "step": 4610
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.4447249174118042,
+      "learning_rate": 3.5748711420465487e-06,
+      "loss": 0.8644,
+      "step": 4611
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6408455967903137,
+      "learning_rate": 3.572837172297001e-06,
+      "loss": 1.0181,
+      "step": 4612
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6266711950302124,
+      "learning_rate": 3.570803459631484e-06,
+      "loss": 1.0079,
+      "step": 4613
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3668224811553955,
+      "learning_rate": 3.5687700044163464e-06,
+      "loss": 0.8954,
+      "step": 4614
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.2534055709838867,
+      "learning_rate": 3.566736807017886e-06,
+      "loss": 0.8037,
+      "step": 4615
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.3454375267028809,
+      "learning_rate": 3.5647038678023537e-06,
+      "loss": 0.8878,
+      "step": 4616
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.39839506149292,
+      "learning_rate": 3.5626711871359598e-06,
+      "loss": 0.8616,
+      "step": 4617
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3028982877731323,
+      "learning_rate": 3.5606387653848627e-06,
+      "loss": 0.8149,
+      "step": 4618
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3417296409606934,
+      "learning_rate": 3.5586066029151735e-06,
+      "loss": 0.8307,
+      "step": 4619
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.344285011291504,
+      "learning_rate": 3.556574700092963e-06,
+      "loss": 0.8476,
+      "step": 4620
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2841858863830566,
+      "learning_rate": 3.5545430572842467e-06,
+      "loss": 0.8306,
+      "step": 4621
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6727075576782227,
+      "learning_rate": 3.5525116748549994e-06,
+      "loss": 1.0183,
+      "step": 4622
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6373836994171143,
+      "learning_rate": 3.550480553171148e-06,
+      "loss": 0.8357,
+      "step": 4623
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2692135572433472,
+      "learning_rate": 3.548449692598569e-06,
+      "loss": 0.8224,
+      "step": 4624
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3525139093399048,
+      "learning_rate": 3.5464190935030967e-06,
+      "loss": 0.8268,
+      "step": 4625
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3046064376831055,
+      "learning_rate": 3.5443887562505153e-06,
+      "loss": 0.8109,
+      "step": 4626
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4012454748153687,
+      "learning_rate": 3.5423586812065626e-06,
+      "loss": 0.8405,
+      "step": 4627
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4131269454956055,
+      "learning_rate": 3.5403288687369296e-06,
+      "loss": 0.8661,
+      "step": 4628
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2731375694274902,
+      "learning_rate": 3.538299319207256e-06,
+      "loss": 0.8085,
+      "step": 4629
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.491817831993103,
+      "learning_rate": 3.5362700329831436e-06,
+      "loss": 0.792,
+      "step": 4630
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.329742431640625,
+      "learning_rate": 3.5342410104301346e-06,
+      "loss": 0.8444,
+      "step": 4631
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2676736116409302,
+      "learning_rate": 3.5322122519137354e-06,
+      "loss": 0.8353,
+      "step": 4632
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3370088338851929,
+      "learning_rate": 3.530183757799397e-06,
+      "loss": 0.8517,
+      "step": 4633
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3376989364624023,
+      "learning_rate": 3.5281555284525227e-06,
+      "loss": 0.8456,
+      "step": 4634
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4641149044036865,
+      "learning_rate": 3.5261275642384745e-06,
+      "loss": 0.8442,
+      "step": 4635
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3010023832321167,
+      "learning_rate": 3.5240998655225593e-06,
+      "loss": 0.7832,
+      "step": 4636
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6626308560371399,
+      "learning_rate": 3.5220724326700434e-06,
+      "loss": 1.0308,
+      "step": 4637
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3869633674621582,
+      "learning_rate": 3.5200452660461394e-06,
+      "loss": 0.8684,
+      "step": 4638
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6372787356376648,
+      "learning_rate": 3.5180183660160106e-06,
+      "loss": 1.0354,
+      "step": 4639
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 4.244260311126709,
+      "learning_rate": 3.5159917329447814e-06,
+      "loss": 0.8547,
+      "step": 4640
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3031554222106934,
+      "learning_rate": 3.5139653671975175e-06,
+      "loss": 0.7794,
+      "step": 4641
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6242287158966064,
+      "learning_rate": 3.511939269139244e-06,
+      "loss": 1.0528,
+      "step": 4642
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4186909198760986,
+      "learning_rate": 3.5099134391349347e-06,
+      "loss": 0.8523,
+      "step": 4643
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2847511768341064,
+      "learning_rate": 3.507887877549514e-06,
+      "loss": 0.8074,
+      "step": 4644
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3129985332489014,
+      "learning_rate": 3.505862584747861e-06,
+      "loss": 0.8462,
+      "step": 4645
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3030050992965698,
+      "learning_rate": 3.5038375610948037e-06,
+      "loss": 0.867,
+      "step": 4646
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4097628593444824,
+      "learning_rate": 3.501812806955122e-06,
+      "loss": 0.8552,
+      "step": 4647
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6595415472984314,
+      "learning_rate": 3.4997883226935503e-06,
+      "loss": 1.0255,
+      "step": 4648
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6543489098548889,
+      "learning_rate": 3.497764108674768e-06,
+      "loss": 1.0014,
+      "step": 4649
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2627880573272705,
+      "learning_rate": 3.4957401652634147e-06,
+      "loss": 0.81,
+      "step": 4650
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6104167699813843,
+      "learning_rate": 3.493716492824074e-06,
+      "loss": 0.8419,
+      "step": 4651
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6146339774131775,
+      "learning_rate": 3.4916930917212803e-06,
+      "loss": 1.0092,
+      "step": 4652
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2588109970092773,
+      "learning_rate": 3.4896699623195274e-06,
+      "loss": 0.7635,
+      "step": 4653
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6232293248176575,
+      "learning_rate": 3.4876471049832495e-06,
+      "loss": 0.9931,
+      "step": 4654
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4359604120254517,
+      "learning_rate": 3.4856245200768413e-06,
+      "loss": 0.7807,
+      "step": 4655
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3402637243270874,
+      "learning_rate": 3.4836022079646414e-06,
+      "loss": 0.9142,
+      "step": 4656
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4700820446014404,
+      "learning_rate": 3.481580169010941e-06,
+      "loss": 0.7809,
+      "step": 4657
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3438403606414795,
+      "learning_rate": 3.479558403579987e-06,
+      "loss": 0.8517,
+      "step": 4658
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4852575063705444,
+      "learning_rate": 3.477536912035967e-06,
+      "loss": 0.844,
+      "step": 4659
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3178699016571045,
+      "learning_rate": 3.4755156947430325e-06,
+      "loss": 0.7979,
+      "step": 4660
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3317947387695312,
+      "learning_rate": 3.4734947520652735e-06,
+      "loss": 0.8115,
+      "step": 4661
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6174635887145996,
+      "learning_rate": 3.471474084366735e-06,
+      "loss": 1.0035,
+      "step": 4662
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6067750453948975,
+      "learning_rate": 3.4694536920114164e-06,
+      "loss": 1.0089,
+      "step": 4663
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2401914596557617,
+      "learning_rate": 3.467433575363261e-06,
+      "loss": 0.767,
+      "step": 4664
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3211578130722046,
+      "learning_rate": 3.4654137347861676e-06,
+      "loss": 0.8336,
+      "step": 4665
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3713635206222534,
+      "learning_rate": 3.463394170643981e-06,
+      "loss": 0.8283,
+      "step": 4666
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3351261615753174,
+      "learning_rate": 3.461374883300499e-06,
+      "loss": 0.8317,
+      "step": 4667
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.295675277709961,
+      "learning_rate": 3.459355873119471e-06,
+      "loss": 0.8388,
+      "step": 4668
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3811050653457642,
+      "learning_rate": 3.4573371404645895e-06,
+      "loss": 0.8048,
+      "step": 4669
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.317643642425537,
+      "learning_rate": 3.455318685699507e-06,
+      "loss": 0.8897,
+      "step": 4670
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.4199930429458618,
+      "learning_rate": 3.453300509187819e-06,
+      "loss": 0.81,
+      "step": 4671
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3659664392471313,
+      "learning_rate": 3.4512826112930696e-06,
+      "loss": 0.8441,
+      "step": 4672
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6537795066833496,
+      "learning_rate": 3.4492649923787603e-06,
+      "loss": 1.0069,
+      "step": 4673
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.31399405002594,
+      "learning_rate": 3.447247652808336e-06,
+      "loss": 0.9098,
+      "step": 4674
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6050794124603271,
+      "learning_rate": 3.44523059294519e-06,
+      "loss": 1.0269,
+      "step": 4675
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.2676349878311157,
+      "learning_rate": 3.4432138131526742e-06,
+      "loss": 0.7987,
+      "step": 4676
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.286391019821167,
+      "learning_rate": 3.441197313794079e-06,
+      "loss": 0.8305,
+      "step": 4677
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3909883499145508,
+      "learning_rate": 3.439181095232653e-06,
+      "loss": 0.7657,
+      "step": 4678
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3264198303222656,
+      "learning_rate": 3.437165157831589e-06,
+      "loss": 0.8269,
+      "step": 4679
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3631632328033447,
+      "learning_rate": 3.435149501954029e-06,
+      "loss": 0.8498,
+      "step": 4680
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.361540675163269,
+      "learning_rate": 3.43313412796307e-06,
+      "loss": 0.8408,
+      "step": 4681
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3546388149261475,
+      "learning_rate": 3.431119036221752e-06,
+      "loss": 0.8299,
+      "step": 4682
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3553348779678345,
+      "learning_rate": 3.429104227093066e-06,
+      "loss": 0.8423,
+      "step": 4683
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7122763991355896,
+      "learning_rate": 3.427089700939955e-06,
+      "loss": 1.0213,
+      "step": 4684
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3473111391067505,
+      "learning_rate": 3.4250754581253056e-06,
+      "loss": 0.7701,
+      "step": 4685
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6436505317687988,
+      "learning_rate": 3.4230614990119564e-06,
+      "loss": 1.0372,
+      "step": 4686
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.268938660621643,
+      "learning_rate": 3.421047823962697e-06,
+      "loss": 0.7839,
+      "step": 4687
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3639512062072754,
+      "learning_rate": 3.419034433340264e-06,
+      "loss": 0.809,
+      "step": 4688
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3041497468948364,
+      "learning_rate": 3.4170213275073406e-06,
+      "loss": 0.8238,
+      "step": 4689
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3672690391540527,
+      "learning_rate": 3.415008506826558e-06,
+      "loss": 0.8369,
+      "step": 4690
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3574023246765137,
+      "learning_rate": 3.4129959716605054e-06,
+      "loss": 0.8016,
+      "step": 4691
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.6162824630737305,
+      "learning_rate": 3.410983722371707e-06,
+      "loss": 0.8857,
+      "step": 4692
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 1.3075642585754395,
+      "learning_rate": 3.4089717593226467e-06,
+      "loss": 0.815,
+      "step": 4693
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3499786853790283,
+      "learning_rate": 3.406960082875751e-06,
+      "loss": 0.8057,
+      "step": 4694
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2842438220977783,
+      "learning_rate": 3.404948693393394e-06,
+      "loss": 0.8695,
+      "step": 4695
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3044676780700684,
+      "learning_rate": 3.4029375912379046e-06,
+      "loss": 0.7981,
+      "step": 4696
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.359805703163147,
+      "learning_rate": 3.4009267767715505e-06,
+      "loss": 0.8199,
+      "step": 4697
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3003567457199097,
+      "learning_rate": 3.3989162503565577e-06,
+      "loss": 0.9021,
+      "step": 4698
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7516499757766724,
+      "learning_rate": 3.396906012355092e-06,
+      "loss": 1.0402,
+      "step": 4699
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3046205043792725,
+      "learning_rate": 3.3948960631292705e-06,
+      "loss": 0.8134,
+      "step": 4700
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3192540407180786,
+      "learning_rate": 3.392886403041161e-06,
+      "loss": 0.7621,
+      "step": 4701
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6741209626197815,
+      "learning_rate": 3.3908770324527727e-06,
+      "loss": 1.0199,
+      "step": 4702
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3868402242660522,
+      "learning_rate": 3.388867951726069e-06,
+      "loss": 0.8805,
+      "step": 4703
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.294519066810608,
+      "learning_rate": 3.3868591612229595e-06,
+      "loss": 0.8649,
+      "step": 4704
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2851479053497314,
+      "learning_rate": 3.384850661305298e-06,
+      "loss": 0.8334,
+      "step": 4705
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2688935995101929,
+      "learning_rate": 3.38284245233489e-06,
+      "loss": 0.7914,
+      "step": 4706
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2992907762527466,
+      "learning_rate": 3.380834534673486e-06,
+      "loss": 0.8224,
+      "step": 4707
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3293728828430176,
+      "learning_rate": 3.3788269086827856e-06,
+      "loss": 0.848,
+      "step": 4708
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3774996995925903,
+      "learning_rate": 3.376819574724437e-06,
+      "loss": 0.8406,
+      "step": 4709
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4408485889434814,
+      "learning_rate": 3.374812533160031e-06,
+      "loss": 0.8449,
+      "step": 4710
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2688337564468384,
+      "learning_rate": 3.3728057843511125e-06,
+      "loss": 0.7928,
+      "step": 4711
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4023252725601196,
+      "learning_rate": 3.3707993286591683e-06,
+      "loss": 0.8599,
+      "step": 4712
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5173133611679077,
+      "learning_rate": 3.3687931664456323e-06,
+      "loss": 0.8189,
+      "step": 4713
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.483466386795044,
+      "learning_rate": 3.366787298071892e-06,
+      "loss": 0.8289,
+      "step": 4714
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.8283461928367615,
+      "learning_rate": 3.364781723899272e-06,
+      "loss": 1.0265,
+      "step": 4715
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3515925407409668,
+      "learning_rate": 3.362776444289054e-06,
+      "loss": 0.8169,
+      "step": 4716
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4825329780578613,
+      "learning_rate": 3.3607714596024602e-06,
+      "loss": 0.8501,
+      "step": 4717
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3704770803451538,
+      "learning_rate": 3.3587667702006594e-06,
+      "loss": 0.7968,
+      "step": 4718
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4086220264434814,
+      "learning_rate": 3.356762376444773e-06,
+      "loss": 0.7589,
+      "step": 4719
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7079104781150818,
+      "learning_rate": 3.354758278695861e-06,
+      "loss": 1.0202,
+      "step": 4720
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4399727582931519,
+      "learning_rate": 3.352754477314939e-06,
+      "loss": 0.8187,
+      "step": 4721
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6675034761428833,
+      "learning_rate": 3.3507509726629627e-06,
+      "loss": 1.0112,
+      "step": 4722
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4021878242492676,
+      "learning_rate": 3.3487477651008348e-06,
+      "loss": 0.9059,
+      "step": 4723
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6447230577468872,
+      "learning_rate": 3.3467448549894064e-06,
+      "loss": 1.0226,
+      "step": 4724
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2956660985946655,
+      "learning_rate": 3.3447422426894773e-06,
+      "loss": 0.831,
+      "step": 4725
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2901264429092407,
+      "learning_rate": 3.3427399285617884e-06,
+      "loss": 0.867,
+      "step": 4726
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7552719116210938,
+      "learning_rate": 3.34073791296703e-06,
+      "loss": 1.0375,
+      "step": 4727
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6715508699417114,
+      "learning_rate": 3.338736196265838e-06,
+      "loss": 0.991,
+      "step": 4728
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3434138298034668,
+      "learning_rate": 3.3367347788187953e-06,
+      "loss": 0.8889,
+      "step": 4729
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.498023271560669,
+      "learning_rate": 3.3347336609864276e-06,
+      "loss": 0.8051,
+      "step": 4730
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.5038093328475952,
+      "learning_rate": 3.3327328431292133e-06,
+      "loss": 0.7855,
+      "step": 4731
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4496270418167114,
+      "learning_rate": 3.3307323256075708e-06,
+      "loss": 0.8427,
+      "step": 4732
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.285479187965393,
+      "learning_rate": 3.328732108781863e-06,
+      "loss": 0.8447,
+      "step": 4733
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4326666593551636,
+      "learning_rate": 3.326732193012407e-06,
+      "loss": 0.8068,
+      "step": 4734
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7277170419692993,
+      "learning_rate": 3.3247325786594583e-06,
+      "loss": 1.0354,
+      "step": 4735
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4053516387939453,
+      "learning_rate": 3.3227332660832173e-06,
+      "loss": 0.8989,
+      "step": 4736
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3134891986846924,
+      "learning_rate": 3.3207342556438387e-06,
+      "loss": 0.865,
+      "step": 4737
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6344976425170898,
+      "learning_rate": 3.3187355477014114e-06,
+      "loss": 1.0143,
+      "step": 4738
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3240230083465576,
+      "learning_rate": 3.3167371426159816e-06,
+      "loss": 0.7956,
+      "step": 4739
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.7213025093078613,
+      "learning_rate": 3.314739040747531e-06,
+      "loss": 0.7605,
+      "step": 4740
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6313632130622864,
+      "learning_rate": 3.31274124245599e-06,
+      "loss": 0.9994,
+      "step": 4741
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3761707544326782,
+      "learning_rate": 3.310743748101237e-06,
+      "loss": 0.8036,
+      "step": 4742
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2407063245773315,
+      "learning_rate": 3.3087465580430927e-06,
+      "loss": 0.7869,
+      "step": 4743
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6645139455795288,
+      "learning_rate": 3.3067496726413236e-06,
+      "loss": 1.006,
+      "step": 4744
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2564560174942017,
+      "learning_rate": 3.3047530922556424e-06,
+      "loss": 0.8335,
+      "step": 4745
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.301190972328186,
+      "learning_rate": 3.3027568172457047e-06,
+      "loss": 0.8199,
+      "step": 4746
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6116386651992798,
+      "learning_rate": 3.300760847971114e-06,
+      "loss": 1.0168,
+      "step": 4747
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4430017471313477,
+      "learning_rate": 3.298765184791414e-06,
+      "loss": 0.8555,
+      "step": 4748
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6142794489860535,
+      "learning_rate": 3.2967698280660997e-06,
+      "loss": 1.0089,
+      "step": 4749
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4624792337417603,
+      "learning_rate": 3.2947747781546068e-06,
+      "loss": 0.7969,
+      "step": 4750
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3588416576385498,
+      "learning_rate": 3.2927800354163143e-06,
+      "loss": 0.7798,
+      "step": 4751
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4278782606124878,
+      "learning_rate": 3.290785600210551e-06,
+      "loss": 0.8055,
+      "step": 4752
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.356256365776062,
+      "learning_rate": 3.2887914728965845e-06,
+      "loss": 0.8069,
+      "step": 4753
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6160059571266174,
+      "learning_rate": 3.286797653833633e-06,
+      "loss": 1.0126,
+      "step": 4754
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3923919200897217,
+      "learning_rate": 3.2848041433808554e-06,
+      "loss": 0.8137,
+      "step": 4755
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.594460129737854,
+      "learning_rate": 3.282810941897352e-06,
+      "loss": 0.8206,
+      "step": 4756
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.454713225364685,
+      "learning_rate": 3.2808180497421755e-06,
+      "loss": 0.8046,
+      "step": 4757
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6531957983970642,
+      "learning_rate": 3.2788254672743146e-06,
+      "loss": 1.0124,
+      "step": 4758
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.4286932945251465,
+      "learning_rate": 3.27683319485271e-06,
+      "loss": 0.8124,
+      "step": 4759
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3595495223999023,
+      "learning_rate": 3.27484123283624e-06,
+      "loss": 0.8333,
+      "step": 4760
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.3541560173034668,
+      "learning_rate": 3.272849581583728e-06,
+      "loss": 0.7954,
+      "step": 4761
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6148074865341187,
+      "learning_rate": 3.2708582414539473e-06,
+      "loss": 0.9798,
+      "step": 4762
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.6227596998214722,
+      "learning_rate": 3.2688672128056086e-06,
+      "loss": 0.8721,
+      "step": 4763
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.41935396194458,
+      "learning_rate": 3.2668764959973677e-06,
+      "loss": 0.8423,
+      "step": 4764
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6264643669128418,
+      "learning_rate": 3.2648860913878254e-06,
+      "loss": 1.0211,
+      "step": 4765
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6390189528465271,
+      "learning_rate": 3.262895999335527e-06,
+      "loss": 1.0213,
+      "step": 4766
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5977368354797363,
+      "learning_rate": 3.2609062201989618e-06,
+      "loss": 1.0327,
+      "step": 4767
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 1.2958179712295532,
+      "learning_rate": 3.258916754336559e-06,
+      "loss": 0.7891,
+      "step": 4768
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.604960024356842,
+      "learning_rate": 3.2569276021066953e-06,
+      "loss": 1.0228,
+      "step": 4769
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3886616230010986,
+      "learning_rate": 3.25493876386769e-06,
+      "loss": 0.8097,
+      "step": 4770
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4419703483581543,
+      "learning_rate": 3.252950239977802e-06,
+      "loss": 0.8185,
+      "step": 4771
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4791879653930664,
+      "learning_rate": 3.250962030795243e-06,
+      "loss": 0.863,
+      "step": 4772
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.291500449180603,
+      "learning_rate": 3.248974136678158e-06,
+      "loss": 0.7928,
+      "step": 4773
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.621097207069397,
+      "learning_rate": 3.246986557984637e-06,
+      "loss": 0.9933,
+      "step": 4774
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3572916984558105,
+      "learning_rate": 3.2449992950727207e-06,
+      "loss": 0.8015,
+      "step": 4775
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.613478422164917,
+      "learning_rate": 3.2430123483003816e-06,
+      "loss": 0.8435,
+      "step": 4776
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2401251792907715,
+      "learning_rate": 3.2410257180255477e-06,
+      "loss": 0.7933,
+      "step": 4777
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3421186208724976,
+      "learning_rate": 3.2390394046060796e-06,
+      "loss": 0.8546,
+      "step": 4778
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2403465509414673,
+      "learning_rate": 3.237053408399784e-06,
+      "loss": 0.8743,
+      "step": 4779
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2963143587112427,
+      "learning_rate": 3.235067729764414e-06,
+      "loss": 0.8191,
+      "step": 4780
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3628915548324585,
+      "learning_rate": 3.2330823690576596e-06,
+      "loss": 0.8185,
+      "step": 4781
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2620139122009277,
+      "learning_rate": 3.2310973266371603e-06,
+      "loss": 0.8129,
+      "step": 4782
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4918959140777588,
+      "learning_rate": 3.2291126028604926e-06,
+      "loss": 0.826,
+      "step": 4783
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6295918822288513,
+      "learning_rate": 3.227128198085176e-06,
+      "loss": 1.003,
+      "step": 4784
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6313139796257019,
+      "learning_rate": 3.225144112668676e-06,
+      "loss": 0.992,
+      "step": 4785
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.333091378211975,
+      "learning_rate": 3.2231603469683982e-06,
+      "loss": 0.8146,
+      "step": 4786
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4556242227554321,
+      "learning_rate": 3.221176901341693e-06,
+      "loss": 0.7996,
+      "step": 4787
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.330678105354309,
+      "learning_rate": 3.219193776145849e-06,
+      "loss": 0.8001,
+      "step": 4788
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.284381628036499,
+      "learning_rate": 3.2172109717380973e-06,
+      "loss": 0.8521,
+      "step": 4789
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3190467357635498,
+      "learning_rate": 3.215228488475618e-06,
+      "loss": 0.7757,
+      "step": 4790
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2934666872024536,
+      "learning_rate": 3.213246326715526e-06,
+      "loss": 0.8001,
+      "step": 4791
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2618069648742676,
+      "learning_rate": 3.2112644868148794e-06,
+      "loss": 0.8367,
+      "step": 4792
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3650768995285034,
+      "learning_rate": 3.209282969130684e-06,
+      "loss": 0.8246,
+      "step": 4793
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3276420831680298,
+      "learning_rate": 3.2073017740198787e-06,
+      "loss": 0.8062,
+      "step": 4794
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.360693097114563,
+      "learning_rate": 3.205320901839353e-06,
+      "loss": 0.829,
+      "step": 4795
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3329849243164062,
+      "learning_rate": 3.203340352945932e-06,
+      "loss": 0.8175,
+      "step": 4796
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.279809832572937,
+      "learning_rate": 3.2013601276963825e-06,
+      "loss": 0.7966,
+      "step": 4797
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3330976963043213,
+      "learning_rate": 3.1993802264474205e-06,
+      "loss": 0.8326,
+      "step": 4798
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.373146653175354,
+      "learning_rate": 3.1974006495556936e-06,
+      "loss": 0.8637,
+      "step": 4799
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4144619703292847,
+      "learning_rate": 3.195421397377799e-06,
+      "loss": 0.7969,
+      "step": 4800
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.284619927406311,
+      "learning_rate": 3.193442470270272e-06,
+      "loss": 0.8471,
+      "step": 4801
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6687227487564087,
+      "learning_rate": 3.191463868589586e-06,
+      "loss": 1.0086,
+      "step": 4802
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6563835144042969,
+      "learning_rate": 3.189485592692163e-06,
+      "loss": 1.0321,
+      "step": 4803
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6003472805023193,
+      "learning_rate": 3.187507642934361e-06,
+      "loss": 1.0164,
+      "step": 4804
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6200007200241089,
+      "learning_rate": 3.1855300196724815e-06,
+      "loss": 1.0273,
+      "step": 4805
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.5080376863479614,
+      "learning_rate": 3.1835527232627657e-06,
+      "loss": 0.8118,
+      "step": 4806
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6311857104301453,
+      "learning_rate": 3.181575754061397e-06,
+      "loss": 1.0038,
+      "step": 4807
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6201167106628418,
+      "learning_rate": 3.1795991124244996e-06,
+      "loss": 1.0174,
+      "step": 4808
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3523809909820557,
+      "learning_rate": 3.1776227987081375e-06,
+      "loss": 0.8087,
+      "step": 4809
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4372189044952393,
+      "learning_rate": 3.1756468132683206e-06,
+      "loss": 0.808,
+      "step": 4810
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3827818632125854,
+      "learning_rate": 3.1736711564609924e-06,
+      "loss": 0.8531,
+      "step": 4811
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6519150137901306,
+      "learning_rate": 3.1716958286420397e-06,
+      "loss": 1.0221,
+      "step": 4812
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3562344312667847,
+      "learning_rate": 3.1697208301672957e-06,
+      "loss": 0.7291,
+      "step": 4813
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3226573467254639,
+      "learning_rate": 3.167746161392524e-06,
+      "loss": 0.7959,
+      "step": 4814
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2867292165756226,
+      "learning_rate": 3.16577182267344e-06,
+      "loss": 0.8207,
+      "step": 4815
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.744970679283142,
+      "learning_rate": 3.1637978143656912e-06,
+      "loss": 0.7761,
+      "step": 4816
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.6878563165664673,
+      "learning_rate": 3.161824136824867e-06,
+      "loss": 0.8313,
+      "step": 4817
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2710095643997192,
+      "learning_rate": 3.1598507904065013e-06,
+      "loss": 0.8496,
+      "step": 4818
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.452305793762207,
+      "learning_rate": 3.157877775466065e-06,
+      "loss": 0.8526,
+      "step": 4819
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3436832427978516,
+      "learning_rate": 3.155905092358968e-06,
+      "loss": 0.8964,
+      "step": 4820
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.648603618144989,
+      "learning_rate": 3.153932741440565e-06,
+      "loss": 1.0073,
+      "step": 4821
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6482113003730774,
+      "learning_rate": 3.1519607230661465e-06,
+      "loss": 1.0221,
+      "step": 4822
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3557124137878418,
+      "learning_rate": 3.1499890375909468e-06,
+      "loss": 0.8127,
+      "step": 4823
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3392332792282104,
+      "learning_rate": 3.148017685370137e-06,
+      "loss": 0.8236,
+      "step": 4824
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3720464706420898,
+      "learning_rate": 3.14604666675883e-06,
+      "loss": 0.7802,
+      "step": 4825
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3347316980361938,
+      "learning_rate": 3.1440759821120764e-06,
+      "loss": 0.8473,
+      "step": 4826
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6012058854103088,
+      "learning_rate": 3.1421056317848698e-06,
+      "loss": 1.044,
+      "step": 4827
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6047256588935852,
+      "learning_rate": 3.1401356161321433e-06,
+      "loss": 1.0095,
+      "step": 4828
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6142984628677368,
+      "learning_rate": 3.1381659355087666e-06,
+      "loss": 1.0202,
+      "step": 4829
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6324367523193359,
+      "learning_rate": 3.1361965902695503e-06,
+      "loss": 1.023,
+      "step": 4830
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2440532445907593,
+      "learning_rate": 3.134227580769248e-06,
+      "loss": 0.7779,
+      "step": 4831
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.511594533920288,
+      "learning_rate": 3.1322589073625464e-06,
+      "loss": 0.7769,
+      "step": 4832
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.261712670326233,
+      "learning_rate": 3.1302905704040794e-06,
+      "loss": 0.8015,
+      "step": 4833
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.306072473526001,
+      "learning_rate": 3.1283225702484145e-06,
+      "loss": 0.8326,
+      "step": 4834
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4852956533432007,
+      "learning_rate": 3.1263549072500575e-06,
+      "loss": 0.7747,
+      "step": 4835
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4121290445327759,
+      "learning_rate": 3.1243875817634607e-06,
+      "loss": 0.7691,
+      "step": 4836
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.3524577617645264,
+      "learning_rate": 3.1224205941430074e-06,
+      "loss": 0.787,
+      "step": 4837
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2823158502578735,
+      "learning_rate": 3.1204539447430273e-06,
+      "loss": 0.8173,
+      "step": 4838
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4631165266036987,
+      "learning_rate": 3.1184876339177837e-06,
+      "loss": 0.7955,
+      "step": 4839
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4134806394577026,
+      "learning_rate": 3.116521662021479e-06,
+      "loss": 0.7893,
+      "step": 4840
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7069225311279297,
+      "learning_rate": 3.1145560294082604e-06,
+      "loss": 1.0219,
+      "step": 4841
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.2298848628997803,
+      "learning_rate": 3.1125907364322074e-06,
+      "loss": 0.7712,
+      "step": 4842
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6845450401306152,
+      "learning_rate": 3.1106257834473397e-06,
+      "loss": 0.9979,
+      "step": 4843
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4189019203186035,
+      "learning_rate": 3.108661170807621e-06,
+      "loss": 0.7698,
+      "step": 4844
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.4309412240982056,
+      "learning_rate": 3.1066968988669456e-06,
+      "loss": 0.844,
+      "step": 4845
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 1.293274998664856,
+      "learning_rate": 3.1047329679791522e-06,
+      "loss": 0.7945,
+      "step": 4846
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2488340139389038,
+      "learning_rate": 3.1027693784980163e-06,
+      "loss": 0.852,
+      "step": 4847
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3999727964401245,
+      "learning_rate": 3.100806130777252e-06,
+      "loss": 0.8349,
+      "step": 4848
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2825648784637451,
+      "learning_rate": 3.0988432251705127e-06,
+      "loss": 0.7819,
+      "step": 4849
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.346997857093811,
+      "learning_rate": 3.096880662031385e-06,
+      "loss": 0.7813,
+      "step": 4850
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2601388692855835,
+      "learning_rate": 3.094918441713404e-06,
+      "loss": 0.7922,
+      "step": 4851
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6813905835151672,
+      "learning_rate": 3.092956564570033e-06,
+      "loss": 1.0283,
+      "step": 4852
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3331973552703857,
+      "learning_rate": 3.090995030954678e-06,
+      "loss": 0.8344,
+      "step": 4853
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.6171603202819824,
+      "learning_rate": 3.0890338412206845e-06,
+      "loss": 0.8198,
+      "step": 4854
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5217112302780151,
+      "learning_rate": 3.087072995721332e-06,
+      "loss": 0.8041,
+      "step": 4855
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 3.170427083969116,
+      "learning_rate": 3.0851124948098432e-06,
+      "loss": 0.7949,
+      "step": 4856
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.365759253501892,
+      "learning_rate": 3.0831523388393737e-06,
+      "loss": 0.8256,
+      "step": 4857
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6266351938247681,
+      "learning_rate": 3.0811925281630172e-06,
+      "loss": 1.0145,
+      "step": 4858
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6161625385284424,
+      "learning_rate": 3.0792330631338107e-06,
+      "loss": 1.0042,
+      "step": 4859
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3025333881378174,
+      "learning_rate": 3.077273944104723e-06,
+      "loss": 0.8894,
+      "step": 4860
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2996139526367188,
+      "learning_rate": 3.0753151714286644e-06,
+      "loss": 0.8111,
+      "step": 4861
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.346752405166626,
+      "learning_rate": 3.0733567454584805e-06,
+      "loss": 0.8586,
+      "step": 4862
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4310775995254517,
+      "learning_rate": 3.071398666546953e-06,
+      "loss": 0.7898,
+      "step": 4863
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2841379642486572,
+      "learning_rate": 3.069440935046808e-06,
+      "loss": 0.8491,
+      "step": 4864
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2918051481246948,
+      "learning_rate": 3.0674835513106993e-06,
+      "loss": 0.7293,
+      "step": 4865
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6315995454788208,
+      "learning_rate": 3.0655265156912274e-06,
+      "loss": 1.0064,
+      "step": 4866
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5999770164489746,
+      "learning_rate": 3.063569828540922e-06,
+      "loss": 1.0086,
+      "step": 4867
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5977880954742432,
+      "learning_rate": 3.061613490212256e-06,
+      "loss": 0.9963,
+      "step": 4868
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2756353616714478,
+      "learning_rate": 3.0596575010576368e-06,
+      "loss": 0.8986,
+      "step": 4869
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.322239637374878,
+      "learning_rate": 3.0577018614294074e-06,
+      "loss": 0.7926,
+      "step": 4870
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.47312331199646,
+      "learning_rate": 3.055746571679854e-06,
+      "loss": 0.8659,
+      "step": 4871
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6023451685905457,
+      "learning_rate": 3.0537916321611925e-06,
+      "loss": 1.0258,
+      "step": 4872
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2279183864593506,
+      "learning_rate": 3.051837043225577e-06,
+      "loss": 0.7631,
+      "step": 4873
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2654609680175781,
+      "learning_rate": 3.049882805225105e-06,
+      "loss": 0.8581,
+      "step": 4874
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2838726043701172,
+      "learning_rate": 3.0479289185118016e-06,
+      "loss": 0.8092,
+      "step": 4875
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3491116762161255,
+      "learning_rate": 3.0459753834376364e-06,
+      "loss": 0.7894,
+      "step": 4876
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3493160009384155,
+      "learning_rate": 3.0440222003545113e-06,
+      "loss": 0.8107,
+      "step": 4877
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4418549537658691,
+      "learning_rate": 3.042069369614262e-06,
+      "loss": 0.802,
+      "step": 4878
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2778400182724,
+      "learning_rate": 3.0401168915686707e-06,
+      "loss": 0.8296,
+      "step": 4879
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.356844186782837,
+      "learning_rate": 3.038164766569447e-06,
+      "loss": 0.818,
+      "step": 4880
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6388835906982422,
+      "learning_rate": 3.0362129949682368e-06,
+      "loss": 1.0532,
+      "step": 4881
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3911566734313965,
+      "learning_rate": 3.0342615771166303e-06,
+      "loss": 0.8318,
+      "step": 4882
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3930777311325073,
+      "learning_rate": 3.0323105133661457e-06,
+      "loss": 0.8385,
+      "step": 4883
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6341215968132019,
+      "learning_rate": 3.0303598040682413e-06,
+      "loss": 0.9954,
+      "step": 4884
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6254379153251648,
+      "learning_rate": 3.028409449574312e-06,
+      "loss": 1.0305,
+      "step": 4885
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6252160668373108,
+      "learning_rate": 3.026459450235687e-06,
+      "loss": 1.0314,
+      "step": 4886
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.31300950050354,
+      "learning_rate": 3.0245098064036317e-06,
+      "loss": 0.7776,
+      "step": 4887
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3570817708969116,
+      "learning_rate": 3.022560518429348e-06,
+      "loss": 0.8756,
+      "step": 4888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2680368423461914,
+      "learning_rate": 3.0206115866639762e-06,
+      "loss": 0.8035,
+      "step": 4889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6314951181411743,
+      "learning_rate": 3.018663011458588e-06,
+      "loss": 1.0062,
+      "step": 4890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3095383644104004,
+      "learning_rate": 3.0167147931641906e-06,
+      "loss": 0.8024,
+      "step": 4891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5088231563568115,
+      "learning_rate": 3.014766932131733e-06,
+      "loss": 0.796,
+      "step": 4892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6234979033470154,
+      "learning_rate": 3.0128194287120925e-06,
+      "loss": 1.0101,
+      "step": 4893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6084515452384949,
+      "learning_rate": 3.0108722832560887e-06,
+      "loss": 1.005,
+      "step": 4894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3821927309036255,
+      "learning_rate": 3.0089254961144725e-06,
+      "loss": 0.8114,
+      "step": 4895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6012381911277771,
+      "learning_rate": 3.0069790676379286e-06,
+      "loss": 1.001,
+      "step": 4896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3424474000930786,
+      "learning_rate": 3.0050329981770833e-06,
+      "loss": 0.8435,
+      "step": 4897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3620485067367554,
+      "learning_rate": 3.0030872880824914e-06,
+      "loss": 0.8424,
+      "step": 4898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6268346905708313,
+      "learning_rate": 3.0011419377046496e-06,
+      "loss": 1.0022,
+      "step": 4899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4352951049804688,
+      "learning_rate": 2.999196947393985e-06,
+      "loss": 0.8013,
+      "step": 4900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3222800493240356,
+      "learning_rate": 2.9972523175008584e-06,
+      "loss": 0.8367,
+      "step": 4901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6174086928367615,
+      "learning_rate": 2.9953080483755735e-06,
+      "loss": 0.9823,
+      "step": 4902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.293824315071106,
+      "learning_rate": 2.99336414036836e-06,
+      "loss": 0.8306,
+      "step": 4903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6324508190155029,
+      "learning_rate": 2.9914205938293883e-06,
+      "loss": 1.0085,
+      "step": 4904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.4049036502838135,
+      "learning_rate": 2.989477409108763e-06,
+      "loss": 0.7684,
+      "step": 4905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3129136562347412,
+      "learning_rate": 2.9875345865565198e-06,
+      "loss": 0.8982,
+      "step": 4906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.30680513381958,
+      "learning_rate": 2.9855921265226353e-06,
+      "loss": 0.8539,
+      "step": 4907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5989764928817749,
+      "learning_rate": 2.9836500293570123e-06,
+      "loss": 0.9796,
+      "step": 4908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6243683695793152,
+      "learning_rate": 2.981708295409498e-06,
+      "loss": 0.999,
+      "step": 4909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6205291748046875,
+      "learning_rate": 2.9797669250298677e-06,
+      "loss": 0.9967,
+      "step": 4910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2670722007751465,
+      "learning_rate": 2.9778259185678317e-06,
+      "loss": 0.7993,
+      "step": 4911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6099362373352051,
+      "learning_rate": 2.9758852763730385e-06,
+      "loss": 0.9968,
+      "step": 4912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3973487615585327,
+      "learning_rate": 2.9739449987950675e-06,
+      "loss": 0.8272,
+      "step": 4913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3659769296646118,
+      "learning_rate": 2.9720050861834315e-06,
+      "loss": 0.8642,
+      "step": 4914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6341552734375,
+      "learning_rate": 2.970065538887583e-06,
+      "loss": 1.005,
+      "step": 4915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.5455589294433594,
+      "learning_rate": 2.9681263572569007e-06,
+      "loss": 0.88,
+      "step": 4916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6473498940467834,
+      "learning_rate": 2.9661875416407064e-06,
+      "loss": 1.0047,
+      "step": 4917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6519238948822021,
+      "learning_rate": 2.9642490923882495e-06,
+      "loss": 1.0369,
+      "step": 4918
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.328171968460083,
+      "learning_rate": 2.962311009848714e-06,
+      "loss": 0.8173,
+      "step": 4919
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5807254314422607,
+      "learning_rate": 2.9603732943712215e-06,
+      "loss": 1.0065,
+      "step": 4920
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.3224952220916748,
+      "learning_rate": 2.958435946304823e-06,
+      "loss": 0.8541,
+      "step": 4921
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.2301652431488037,
+      "learning_rate": 2.9564989659985078e-06,
+      "loss": 0.8361,
+      "step": 4922
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3292477130889893,
+      "learning_rate": 2.954562353801196e-06,
+      "loss": 0.8386,
+      "step": 4923
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.653562068939209,
+      "learning_rate": 2.95262611006174e-06,
+      "loss": 1.0069,
+      "step": 4924
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4346710443496704,
+      "learning_rate": 2.95069023512893e-06,
+      "loss": 0.8071,
+      "step": 4925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4363809823989868,
+      "learning_rate": 2.9487547293514874e-06,
+      "loss": 0.8364,
+      "step": 4926
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3559000492095947,
+      "learning_rate": 2.946819593078067e-06,
+      "loss": 0.8338,
+      "step": 4927
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4723764657974243,
+      "learning_rate": 2.9448848266572562e-06,
+      "loss": 0.8229,
+      "step": 4928
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6044492125511169,
+      "learning_rate": 2.942950430437579e-06,
+      "loss": 1.0066,
+      "step": 4929
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4862444400787354,
+      "learning_rate": 2.9410164047674894e-06,
+      "loss": 0.8493,
+      "step": 4930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4400709867477417,
+      "learning_rate": 2.9390827499953746e-06,
+      "loss": 0.8772,
+      "step": 4931
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.317084550857544,
+      "learning_rate": 2.93714946646956e-06,
+      "loss": 0.7999,
+      "step": 4932
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2933286428451538,
+      "learning_rate": 2.9352165545382983e-06,
+      "loss": 0.768,
+      "step": 4933
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3963732719421387,
+      "learning_rate": 2.933284014549775e-06,
+      "loss": 0.7896,
+      "step": 4934
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.43902587890625,
+      "learning_rate": 2.931351846852115e-06,
+      "loss": 0.7042,
+      "step": 4935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3804640769958496,
+      "learning_rate": 2.929420051793371e-06,
+      "loss": 0.8496,
+      "step": 4936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3158613443374634,
+      "learning_rate": 2.9274886297215273e-06,
+      "loss": 0.8002,
+      "step": 4937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.385481357574463,
+      "learning_rate": 2.9255575809845067e-06,
+      "loss": 0.857,
+      "step": 4938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4030685424804688,
+      "learning_rate": 2.923626905930158e-06,
+      "loss": 0.8128,
+      "step": 4939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4193919897079468,
+      "learning_rate": 2.92169660490627e-06,
+      "loss": 0.8183,
+      "step": 4940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3215103149414062,
+      "learning_rate": 2.9197666782605588e-06,
+      "loss": 0.8774,
+      "step": 4941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6328803300857544,
+      "learning_rate": 2.917837126340671e-06,
+      "loss": 1.0484,
+      "step": 4942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6297498345375061,
+      "learning_rate": 2.9159079494941957e-06,
+      "loss": 1.0298,
+      "step": 4943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6360489726066589,
+      "learning_rate": 2.913979148068642e-06,
+      "loss": 1.0079,
+      "step": 4944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5986161828041077,
+      "learning_rate": 2.9120507224114613e-06,
+      "loss": 1.0346,
+      "step": 4945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.385038137435913,
+      "learning_rate": 2.910122672870032e-06,
+      "loss": 0.8251,
+      "step": 4946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3120555877685547,
+      "learning_rate": 2.9081949997916642e-06,
+      "loss": 0.8434,
+      "step": 4947
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6308524012565613,
+      "learning_rate": 2.906267703523606e-06,
+      "loss": 1.0006,
+      "step": 4948
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.351694941520691,
+      "learning_rate": 2.904340784413029e-06,
+      "loss": 0.7635,
+      "step": 4949
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2931016683578491,
+      "learning_rate": 2.9024142428070468e-06,
+      "loss": 0.8387,
+      "step": 4950
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4207395315170288,
+      "learning_rate": 2.9004880790526957e-06,
+      "loss": 0.8357,
+      "step": 4951
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.5119354724884033,
+      "learning_rate": 2.8985622934969483e-06,
+      "loss": 0.8048,
+      "step": 4952
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4665817022323608,
+      "learning_rate": 2.8966368864867116e-06,
+      "loss": 0.8883,
+      "step": 4953
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6308906078338623,
+      "learning_rate": 2.8947118583688193e-06,
+      "loss": 1.027,
+      "step": 4954
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2898422479629517,
+      "learning_rate": 2.8927872094900384e-06,
+      "loss": 0.7708,
+      "step": 4955
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6541871428489685,
+      "learning_rate": 2.8908629401970706e-06,
+      "loss": 1.0071,
+      "step": 4956
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4323322772979736,
+      "learning_rate": 2.8889390508365462e-06,
+      "loss": 0.8221,
+      "step": 4957
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4913926124572754,
+      "learning_rate": 2.8870155417550274e-06,
+      "loss": 0.8418,
+      "step": 4958
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3885773420333862,
+      "learning_rate": 2.8850924132990056e-06,
+      "loss": 0.8891,
+      "step": 4959
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6069929599761963,
+      "learning_rate": 2.883169665814911e-06,
+      "loss": 0.9911,
+      "step": 4960
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3788365125656128,
+      "learning_rate": 2.8812472996490993e-06,
+      "loss": 0.8217,
+      "step": 4961
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3320218324661255,
+      "learning_rate": 2.8793253151478553e-06,
+      "loss": 0.8094,
+      "step": 4962
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.2750722169876099,
+      "learning_rate": 2.8774037126574023e-06,
+      "loss": 0.7825,
+      "step": 4963
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6472107768058777,
+      "learning_rate": 2.8754824925238876e-06,
+      "loss": 1.0253,
+      "step": 4964
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6149535775184631,
+      "learning_rate": 2.8735616550933975e-06,
+      "loss": 1.0268,
+      "step": 4965
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.291786551475525,
+      "learning_rate": 2.8716412007119427e-06,
+      "loss": 0.8128,
+      "step": 4966
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3020669221878052,
+      "learning_rate": 2.869721129725464e-06,
+      "loss": 0.8596,
+      "step": 4967
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6125219464302063,
+      "learning_rate": 2.8678014424798406e-06,
+      "loss": 1.0217,
+      "step": 4968
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3404555320739746,
+      "learning_rate": 2.865882139320877e-06,
+      "loss": 0.8232,
+      "step": 4969
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4478986263275146,
+      "learning_rate": 2.863963220594306e-06,
+      "loss": 0.8352,
+      "step": 4970
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3486496210098267,
+      "learning_rate": 2.8620446866457995e-06,
+      "loss": 0.8148,
+      "step": 4971
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3803596496582031,
+      "learning_rate": 2.8601265378209523e-06,
+      "loss": 0.8223,
+      "step": 4972
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4437479972839355,
+      "learning_rate": 2.8582087744652962e-06,
+      "loss": 0.8151,
+      "step": 4973
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.4287673234939575,
+      "learning_rate": 2.8562913969242876e-06,
+      "loss": 0.8637,
+      "step": 4974
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3861509561538696,
+      "learning_rate": 2.8543744055433153e-06,
+      "loss": 0.838,
+      "step": 4975
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6452297568321228,
+      "learning_rate": 2.852457800667703e-06,
+      "loss": 1.0214,
+      "step": 4976
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6490718126296997,
+      "learning_rate": 2.8505415826426963e-06,
+      "loss": 1.0129,
+      "step": 4977
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6001209020614624,
+      "learning_rate": 2.8486257518134795e-06,
+      "loss": 1.0147,
+      "step": 4978
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3040956258773804,
+      "learning_rate": 2.8467103085251633e-06,
+      "loss": 0.7951,
+      "step": 4979
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3670496940612793,
+      "learning_rate": 2.844795253122786e-06,
+      "loss": 0.8117,
+      "step": 4980
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3323217630386353,
+      "learning_rate": 2.842880585951323e-06,
+      "loss": 0.7793,
+      "step": 4981
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3573691844940186,
+      "learning_rate": 2.8409663073556715e-06,
+      "loss": 0.8229,
+      "step": 4982
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3941248655319214,
+      "learning_rate": 2.8390524176806665e-06,
+      "loss": 0.8228,
+      "step": 4983
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3264124393463135,
+      "learning_rate": 2.8371389172710674e-06,
+      "loss": 0.8054,
+      "step": 4984
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.442017674446106,
+      "learning_rate": 2.8352258064715638e-06,
+      "loss": 0.7636,
+      "step": 4985
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3640235662460327,
+      "learning_rate": 2.83331308562678e-06,
+      "loss": 0.8437,
+      "step": 4986
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7381541132926941,
+      "learning_rate": 2.8314007550812635e-06,
+      "loss": 1.0067,
+      "step": 4987
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3235918283462524,
+      "learning_rate": 2.829488815179498e-06,
+      "loss": 0.7971,
+      "step": 4988
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3261983394622803,
+      "learning_rate": 2.827577266265892e-06,
+      "loss": 0.821,
+      "step": 4989
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3930546045303345,
+      "learning_rate": 2.825666108684783e-06,
+      "loss": 0.8614,
+      "step": 4990
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.318677544593811,
+      "learning_rate": 2.8237553427804443e-06,
+      "loss": 0.8171,
+      "step": 4991
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3070130348205566,
+      "learning_rate": 2.821844968897072e-06,
+      "loss": 0.8225,
+      "step": 4992
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.7421245574951172,
+      "learning_rate": 2.819934987378793e-06,
+      "loss": 0.9249,
+      "step": 4993
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.291574239730835,
+      "learning_rate": 2.818025398569667e-06,
+      "loss": 0.8435,
+      "step": 4994
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6331264972686768,
+      "learning_rate": 2.8161162028136803e-06,
+      "loss": 1.0264,
+      "step": 4995
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.352799415588379,
+      "learning_rate": 2.8142074004547447e-06,
+      "loss": 0.8393,
+      "step": 4996
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6074618101119995,
+      "learning_rate": 2.8122989918367106e-06,
+      "loss": 1.0217,
+      "step": 4997
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 1.3257712125778198,
+      "learning_rate": 2.81039097730335e-06,
+      "loss": 0.8564,
+      "step": 4998
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3491305112838745,
+      "learning_rate": 2.808483357198365e-06,
+      "loss": 0.8312,
+      "step": 4999
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5817853212356567,
+      "learning_rate": 2.806576131865385e-06,
+      "loss": 0.7716,
+      "step": 5000
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4658517837524414,
+      "learning_rate": 2.8046693016479763e-06,
+      "loss": 0.8196,
+      "step": 5001
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4774740934371948,
+      "learning_rate": 2.8027628668896263e-06,
+      "loss": 0.893,
+      "step": 5002
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3155412673950195,
+      "learning_rate": 2.80085682793375e-06,
+      "loss": 0.8333,
+      "step": 5003
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.5650501251220703,
+      "learning_rate": 2.7989511851237e-06,
+      "loss": 0.8488,
+      "step": 5004
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6185698509216309,
+      "learning_rate": 2.797045938802747e-06,
+      "loss": 1.0329,
+      "step": 5005
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.674322247505188,
+      "learning_rate": 2.7951410893140997e-06,
+      "loss": 1.0187,
+      "step": 5006
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.415811538696289,
+      "learning_rate": 2.793236637000889e-06,
+      "loss": 0.8332,
+      "step": 5007
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.33040189743042,
+      "learning_rate": 2.791332582206174e-06,
+      "loss": 0.8236,
+      "step": 5008
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.234292984008789,
+      "learning_rate": 2.789428925272948e-06,
+      "loss": 0.8621,
+      "step": 5009
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.317236304283142,
+      "learning_rate": 2.7875256665441256e-06,
+      "loss": 0.8346,
+      "step": 5010
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6113707423210144,
+      "learning_rate": 2.7856228063625568e-06,
+      "loss": 1.0127,
+      "step": 5011
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.245718240737915,
+      "learning_rate": 2.783720345071014e-06,
+      "loss": 0.8056,
+      "step": 5012
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2829910516738892,
+      "learning_rate": 2.7818182830121983e-06,
+      "loss": 0.8194,
+      "step": 5013
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6432583928108215,
+      "learning_rate": 2.779916620528744e-06,
+      "loss": 1.0262,
+      "step": 5014
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6141064763069153,
+      "learning_rate": 2.7780153579632052e-06,
+      "loss": 1.0285,
+      "step": 5015
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4274423122406006,
+      "learning_rate": 2.776114495658073e-06,
+      "loss": 0.818,
+      "step": 5016
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3296843767166138,
+      "learning_rate": 2.7742140339557607e-06,
+      "loss": 0.8101,
+      "step": 5017
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2791999578475952,
+      "learning_rate": 2.7723139731986084e-06,
+      "loss": 0.7926,
+      "step": 5018
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5910853147506714,
+      "learning_rate": 2.7704143137288887e-06,
+      "loss": 1.0303,
+      "step": 5019
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6037658452987671,
+      "learning_rate": 2.768515055888798e-06,
+      "loss": 1.0049,
+      "step": 5020
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.336174726486206,
+      "learning_rate": 2.7666162000204645e-06,
+      "loss": 0.7963,
+      "step": 5021
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3134047985076904,
+      "learning_rate": 2.764717746465939e-06,
+      "loss": 0.8067,
+      "step": 5022
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2636075019836426,
+      "learning_rate": 2.7628196955672004e-06,
+      "loss": 0.8553,
+      "step": 5023
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5952982902526855,
+      "learning_rate": 2.760922047666162e-06,
+      "loss": 1.0046,
+      "step": 5024
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4257116317749023,
+      "learning_rate": 2.759024803104656e-06,
+      "loss": 0.8205,
+      "step": 5025
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4888064861297607,
+      "learning_rate": 2.757127962224444e-06,
+      "loss": 0.8318,
+      "step": 5026
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.30685555934906,
+      "learning_rate": 2.755231525367219e-06,
+      "loss": 0.8522,
+      "step": 5027
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2892600297927856,
+      "learning_rate": 2.7533354928745958e-06,
+      "loss": 0.7549,
+      "step": 5028
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2622742652893066,
+      "learning_rate": 2.751439865088123e-06,
+      "loss": 0.8003,
+      "step": 5029
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.378831148147583,
+      "learning_rate": 2.749544642349269e-06,
+      "loss": 0.8376,
+      "step": 5030
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3238450288772583,
+      "learning_rate": 2.747649824999431e-06,
+      "loss": 0.8271,
+      "step": 5031
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.621729850769043,
+      "learning_rate": 2.745755413379939e-06,
+      "loss": 1.0059,
+      "step": 5032
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3352231979370117,
+      "learning_rate": 2.7438614078320413e-06,
+      "loss": 0.8342,
+      "step": 5033
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3855059146881104,
+      "learning_rate": 2.7419678086969216e-06,
+      "loss": 0.8022,
+      "step": 5034
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3615796566009521,
+      "learning_rate": 2.7400746163156837e-06,
+      "loss": 0.8226,
+      "step": 5035
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3168556690216064,
+      "learning_rate": 2.7381818310293604e-06,
+      "loss": 0.7969,
+      "step": 5036
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6325257420539856,
+      "learning_rate": 2.7362894531789104e-06,
+      "loss": 0.9988,
+      "step": 5037
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6191742420196533,
+      "learning_rate": 2.734397483105222e-06,
+      "loss": 1.022,
+      "step": 5038
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6104121804237366,
+      "learning_rate": 2.7325059211491077e-06,
+      "loss": 1.0123,
+      "step": 5039
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3626747131347656,
+      "learning_rate": 2.730614767651306e-06,
+      "loss": 0.8093,
+      "step": 5040
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3636294603347778,
+      "learning_rate": 2.7287240229524813e-06,
+      "loss": 0.8326,
+      "step": 5041
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6547920107841492,
+      "learning_rate": 2.7268336873932285e-06,
+      "loss": 1.0073,
+      "step": 5042
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2724179029464722,
+      "learning_rate": 2.7249437613140622e-06,
+      "loss": 0.8282,
+      "step": 5043
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3524243831634521,
+      "learning_rate": 2.7230542450554313e-06,
+      "loss": 0.8667,
+      "step": 5044
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.622654914855957,
+      "learning_rate": 2.7211651389577044e-06,
+      "loss": 1.0221,
+      "step": 5045
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3992115259170532,
+      "learning_rate": 2.7192764433611763e-06,
+      "loss": 0.8763,
+      "step": 5046
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6207564473152161,
+      "learning_rate": 2.7173881586060746e-06,
+      "loss": 1.0135,
+      "step": 5047
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3317540884017944,
+      "learning_rate": 2.7155002850325428e-06,
+      "loss": 0.8419,
+      "step": 5048
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3517400026321411,
+      "learning_rate": 2.7136128229806607e-06,
+      "loss": 0.8156,
+      "step": 5049
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.429842472076416,
+      "learning_rate": 2.7117257727904267e-06,
+      "loss": 0.816,
+      "step": 5050
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6116646528244019,
+      "learning_rate": 2.7098391348017656e-06,
+      "loss": 1.0146,
+      "step": 5051
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.1880654096603394,
+      "learning_rate": 2.707952909354533e-06,
+      "loss": 0.8602,
+      "step": 5052
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3644371032714844,
+      "learning_rate": 2.7060670967885062e-06,
+      "loss": 0.7826,
+      "step": 5053
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.5488429069519043,
+      "learning_rate": 2.7041816974433843e-06,
+      "loss": 0.8479,
+      "step": 5054
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4893066883087158,
+      "learning_rate": 2.7022967116588036e-06,
+      "loss": 0.8582,
+      "step": 5055
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.348783016204834,
+      "learning_rate": 2.7004121397743123e-06,
+      "loss": 0.8442,
+      "step": 5056
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.360163927078247,
+      "learning_rate": 2.6985279821293954e-06,
+      "loss": 0.8409,
+      "step": 5057
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.6587584018707275,
+      "learning_rate": 2.6966442390634557e-06,
+      "loss": 0.816,
+      "step": 5058
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.4231935739517212,
+      "learning_rate": 2.694760910915823e-06,
+      "loss": 0.8735,
+      "step": 5059
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3111622333526611,
+      "learning_rate": 2.6928779980257567e-06,
+      "loss": 0.8131,
+      "step": 5060
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.6436190605163574,
+      "learning_rate": 2.690995500732434e-06,
+      "loss": 0.8206,
+      "step": 5061
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6457180380821228,
+      "learning_rate": 2.6891134193749657e-06,
+      "loss": 0.9965,
+      "step": 5062
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3448041677474976,
+      "learning_rate": 2.687231754292381e-06,
+      "loss": 0.8139,
+      "step": 5063
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3881679773330688,
+      "learning_rate": 2.685350505823634e-06,
+      "loss": 0.843,
+      "step": 5064
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.290452480316162,
+      "learning_rate": 2.68346967430761e-06,
+      "loss": 0.8045,
+      "step": 5065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3976240158081055,
+      "learning_rate": 2.6815892600831124e-06,
+      "loss": 0.7699,
+      "step": 5066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3562473058700562,
+      "learning_rate": 2.6797092634888754e-06,
+      "loss": 0.8543,
+      "step": 5067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3875973224639893,
+      "learning_rate": 2.677829684863553e-06,
+      "loss": 0.8258,
+      "step": 5068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.3182785511016846,
+      "learning_rate": 2.675950524545724e-06,
+      "loss": 0.833,
+      "step": 5069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2347569465637207,
+      "learning_rate": 2.6740717828738973e-06,
+      "loss": 0.8033,
+      "step": 5070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6200498342514038,
+      "learning_rate": 2.6721934601864997e-06,
+      "loss": 1.0171,
+      "step": 5071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 1.2846399545669556,
+      "learning_rate": 2.670315556821888e-06,
+      "loss": 0.845,
+      "step": 5072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.1057684421539307,
+      "learning_rate": 2.6684380731183403e-06,
+      "loss": 0.8162,
+      "step": 5073
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6164913773536682,
+      "learning_rate": 2.666561009414057e-06,
+      "loss": 1.0123,
+      "step": 5074
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6121466755867004,
+      "learning_rate": 2.664684366047171e-06,
+      "loss": 1.0005,
+      "step": 5075
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3387515544891357,
+      "learning_rate": 2.6628081433557307e-06,
+      "loss": 0.7859,
+      "step": 5076
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2483487129211426,
+      "learning_rate": 2.660932341677713e-06,
+      "loss": 0.8063,
+      "step": 5077
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.447502613067627,
+      "learning_rate": 2.6590569613510152e-06,
+      "loss": 0.8297,
+      "step": 5078
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6445217728614807,
+      "learning_rate": 2.6571820027134663e-06,
+      "loss": 0.9888,
+      "step": 5079
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3971120119094849,
+      "learning_rate": 2.6553074661028133e-06,
+      "loss": 0.8658,
+      "step": 5080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4648059606552124,
+      "learning_rate": 2.6534333518567262e-06,
+      "loss": 0.7631,
+      "step": 5081
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4277344942092896,
+      "learning_rate": 2.6515596603128045e-06,
+      "loss": 0.8711,
+      "step": 5082
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5377624034881592,
+      "learning_rate": 2.6496863918085667e-06,
+      "loss": 0.8969,
+      "step": 5083
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3876975774765015,
+      "learning_rate": 2.6478135466814555e-06,
+      "loss": 0.8376,
+      "step": 5084
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.621476948261261,
+      "learning_rate": 2.6459411252688414e-06,
+      "loss": 1.0209,
+      "step": 5085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3826125860214233,
+      "learning_rate": 2.6440691279080144e-06,
+      "loss": 0.8349,
+      "step": 5086
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2347947359085083,
+      "learning_rate": 2.6421975549361867e-06,
+      "loss": 0.8214,
+      "step": 5087
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2921679019927979,
+      "learning_rate": 2.640326406690502e-06,
+      "loss": 0.801,
+      "step": 5088
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2642492055892944,
+      "learning_rate": 2.638455683508018e-06,
+      "loss": 0.8047,
+      "step": 5089
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2700910568237305,
+      "learning_rate": 2.6365853857257227e-06,
+      "loss": 0.8345,
+      "step": 5090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.41940438747406,
+      "learning_rate": 2.6347155136805247e-06,
+      "loss": 0.849,
+      "step": 5091
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4813578128814697,
+      "learning_rate": 2.6328460677092537e-06,
+      "loss": 0.8625,
+      "step": 5092
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3070703744888306,
+      "learning_rate": 2.630977048148668e-06,
+      "loss": 0.8266,
+      "step": 5093
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2964789867401123,
+      "learning_rate": 2.629108455335443e-06,
+      "loss": 0.7855,
+      "step": 5094
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3325375318527222,
+      "learning_rate": 2.6272402896061846e-06,
+      "loss": 0.8331,
+      "step": 5095
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6509270668029785,
+      "learning_rate": 2.6253725512974144e-06,
+      "loss": 0.9903,
+      "step": 5096
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.715661883354187,
+      "learning_rate": 2.623505240745579e-06,
+      "loss": 0.8111,
+      "step": 5097
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6284403800964355,
+      "learning_rate": 2.621638358287053e-06,
+      "loss": 1.0112,
+      "step": 5098
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6274008750915527,
+      "learning_rate": 2.619771904258126e-06,
+      "loss": 1.0423,
+      "step": 5099
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6015563607215881,
+      "learning_rate": 2.6179058789950184e-06,
+      "loss": 1.0243,
+      "step": 5100
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.607524037361145,
+      "learning_rate": 2.6160402828338673e-06,
+      "loss": 1.0119,
+      "step": 5101
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4275970458984375,
+      "learning_rate": 2.614175116110732e-06,
+      "loss": 0.8281,
+      "step": 5102
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3446180820465088,
+      "learning_rate": 2.612310379161601e-06,
+      "loss": 0.8428,
+      "step": 5103
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2160276174545288,
+      "learning_rate": 2.610446072322379e-06,
+      "loss": 0.7805,
+      "step": 5104
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6536328196525574,
+      "learning_rate": 2.608582195928898e-06,
+      "loss": 1.0066,
+      "step": 5105
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2868094444274902,
+      "learning_rate": 2.6067187503169085e-06,
+      "loss": 0.8608,
+      "step": 5106
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.314356803894043,
+      "learning_rate": 2.6048557358220826e-06,
+      "loss": 0.8579,
+      "step": 5107
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.32001793384552,
+      "learning_rate": 2.6029931527800217e-06,
+      "loss": 0.7657,
+      "step": 5108
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4785215854644775,
+      "learning_rate": 2.601131001526241e-06,
+      "loss": 0.827,
+      "step": 5109
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4845733642578125,
+      "learning_rate": 2.5992692823961853e-06,
+      "loss": 0.835,
+      "step": 5110
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.28475022315979,
+      "learning_rate": 2.597407995725216e-06,
+      "loss": 0.8231,
+      "step": 5111
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3387130498886108,
+      "learning_rate": 2.5955471418486167e-06,
+      "loss": 0.8813,
+      "step": 5112
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.639888346195221,
+      "learning_rate": 2.5936867211015993e-06,
+      "loss": 0.9972,
+      "step": 5113
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.60050368309021,
+      "learning_rate": 2.591826733819291e-06,
+      "loss": 0.8186,
+      "step": 5114
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6279332041740417,
+      "learning_rate": 2.589967180336742e-06,
+      "loss": 1.0029,
+      "step": 5115
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6085898876190186,
+      "learning_rate": 2.5881080609889286e-06,
+      "loss": 0.9978,
+      "step": 5116
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3709272146224976,
+      "learning_rate": 2.5862493761107448e-06,
+      "loss": 0.7684,
+      "step": 5117
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5860143303871155,
+      "learning_rate": 2.584391126037007e-06,
+      "loss": 1.0042,
+      "step": 5118
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4405105113983154,
+      "learning_rate": 2.582533311102452e-06,
+      "loss": 0.7776,
+      "step": 5119
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3655470609664917,
+      "learning_rate": 2.580675931641744e-06,
+      "loss": 0.8726,
+      "step": 5120
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4897855520248413,
+      "learning_rate": 2.5788189879894625e-06,
+      "loss": 0.8226,
+      "step": 5121
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4485275745391846,
+      "learning_rate": 2.576962480480109e-06,
+      "loss": 0.7841,
+      "step": 5122
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.331589937210083,
+      "learning_rate": 2.575106409448112e-06,
+      "loss": 0.8615,
+      "step": 5123
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.274507761001587,
+      "learning_rate": 2.5732507752278157e-06,
+      "loss": 0.8012,
+      "step": 5124
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6661547422409058,
+      "learning_rate": 2.571395578153485e-06,
+      "loss": 1.0176,
+      "step": 5125
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3695799112319946,
+      "learning_rate": 2.5695408185593126e-06,
+      "loss": 0.8113,
+      "step": 5126
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3779906034469604,
+      "learning_rate": 2.5676864967794047e-06,
+      "loss": 0.8227,
+      "step": 5127
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3368313312530518,
+      "learning_rate": 2.5658326131477955e-06,
+      "loss": 0.8102,
+      "step": 5128
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5472880601882935,
+      "learning_rate": 2.563979167998436e-06,
+      "loss": 0.85,
+      "step": 5129
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6160759925842285,
+      "learning_rate": 2.5621261616651965e-06,
+      "loss": 0.9915,
+      "step": 5130
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2558441162109375,
+      "learning_rate": 2.5602735944818754e-06,
+      "loss": 0.7882,
+      "step": 5131
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3268276453018188,
+      "learning_rate": 2.558421466782183e-06,
+      "loss": 0.7874,
+      "step": 5132
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5909254550933838,
+      "learning_rate": 2.5565697788997595e-06,
+      "loss": 1.0207,
+      "step": 5133
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3366445302963257,
+      "learning_rate": 2.5547185311681588e-06,
+      "loss": 0.8078,
+      "step": 5134
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.5942925214767456,
+      "learning_rate": 2.552867723920857e-06,
+      "loss": 0.8401,
+      "step": 5135
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.386388897895813,
+      "learning_rate": 2.5510173574912554e-06,
+      "loss": 0.8919,
+      "step": 5136
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2821911573410034,
+      "learning_rate": 2.549167432212668e-06,
+      "loss": 0.8036,
+      "step": 5137
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6098567843437195,
+      "learning_rate": 2.5473179484183386e-06,
+      "loss": 1.0137,
+      "step": 5138
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4557467699050903,
+      "learning_rate": 2.5454689064414233e-06,
+      "loss": 0.819,
+      "step": 5139
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4135608673095703,
+      "learning_rate": 2.5436203066150017e-06,
+      "loss": 0.8805,
+      "step": 5140
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.430972933769226,
+      "learning_rate": 2.541772149272077e-06,
+      "loss": 0.806,
+      "step": 5141
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.4420890808105469,
+      "learning_rate": 2.5399244347455677e-06,
+      "loss": 0.8275,
+      "step": 5142
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6326138973236084,
+      "learning_rate": 2.538077163368312e-06,
+      "loss": 1.0154,
+      "step": 5143
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.620720386505127,
+      "learning_rate": 2.5362303354730768e-06,
+      "loss": 1.0196,
+      "step": 5144
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3252296447753906,
+      "learning_rate": 2.534383951392536e-06,
+      "loss": 0.7788,
+      "step": 5145
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.3258590698242188,
+      "learning_rate": 2.532538011459298e-06,
+      "loss": 0.8508,
+      "step": 5146
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.340768814086914,
+      "learning_rate": 2.5306925160058794e-06,
+      "loss": 0.8155,
+      "step": 5147
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6038669347763062,
+      "learning_rate": 2.52884746536472e-06,
+      "loss": 1.0193,
+      "step": 5148
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.2898516654968262,
+      "learning_rate": 2.5270028598681846e-06,
+      "loss": 0.8096,
+      "step": 5149
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.310598373413086,
+      "learning_rate": 2.52515869984855e-06,
+      "loss": 0.8608,
+      "step": 5150
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 1.455422043800354,
+      "learning_rate": 2.523314985638021e-06,
+      "loss": 0.7533,
+      "step": 5151
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.32033109664917,
+      "learning_rate": 2.521471717568715e-06,
+      "loss": 0.8974,
+      "step": 5152
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3020669221878052,
+      "learning_rate": 2.51962889597267e-06,
+      "loss": 0.7567,
+      "step": 5153
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6425709128379822,
+      "learning_rate": 2.5177865211818486e-06,
+      "loss": 1.0017,
+      "step": 5154
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.7140452861785889,
+      "learning_rate": 2.515944593528129e-06,
+      "loss": 0.7915,
+      "step": 5155
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2681667804718018,
+      "learning_rate": 2.514103113343307e-06,
+      "loss": 0.7541,
+      "step": 5156
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 3.7633516788482666,
+      "learning_rate": 2.512262080959105e-06,
+      "loss": 0.8281,
+      "step": 5157
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6068690419197083,
+      "learning_rate": 2.5104214967071554e-06,
+      "loss": 1.0105,
+      "step": 5158
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4409894943237305,
+      "learning_rate": 2.5085813609190173e-06,
+      "loss": 0.8076,
+      "step": 5159
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2729004621505737,
+      "learning_rate": 2.506741673926163e-06,
+      "loss": 0.8278,
+      "step": 5160
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4426050186157227,
+      "learning_rate": 2.5049024360599914e-06,
+      "loss": 0.8305,
+      "step": 5161
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3860424757003784,
+      "learning_rate": 2.5030636476518145e-06,
+      "loss": 0.8305,
+      "step": 5162
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.243852138519287,
+      "learning_rate": 2.501225309032862e-06,
+      "loss": 0.8004,
+      "step": 5163
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3103060722351074,
+      "learning_rate": 2.499387420534291e-06,
+      "loss": 0.8443,
+      "step": 5164
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4832584857940674,
+      "learning_rate": 2.497549982487168e-06,
+      "loss": 0.7763,
+      "step": 5165
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5244121551513672,
+      "learning_rate": 2.495712995222486e-06,
+      "loss": 0.8201,
+      "step": 5166
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6295305490493774,
+      "learning_rate": 2.493876459071151e-06,
+      "loss": 0.9939,
+      "step": 5167
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6343121528625488,
+      "learning_rate": 2.4920403743639893e-06,
+      "loss": 1.0204,
+      "step": 5168
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.298376441001892,
+      "learning_rate": 2.4902047414317497e-06,
+      "loss": 0.8166,
+      "step": 5169
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5938814282417297,
+      "learning_rate": 2.4883695606050952e-06,
+      "loss": 1.0224,
+      "step": 5170
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2931874990463257,
+      "learning_rate": 2.4865348322146064e-06,
+      "loss": 0.7879,
+      "step": 5171
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.297902226448059,
+      "learning_rate": 2.4847005565907886e-06,
+      "loss": 0.8195,
+      "step": 5172
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3497867584228516,
+      "learning_rate": 2.482866734064058e-06,
+      "loss": 0.7869,
+      "step": 5173
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6077553629875183,
+      "learning_rate": 2.4810333649647565e-06,
+      "loss": 0.9961,
+      "step": 5174
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4396384954452515,
+      "learning_rate": 2.479200449623139e-06,
+      "loss": 0.8515,
+      "step": 5175
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.346609354019165,
+      "learning_rate": 2.4773679883693785e-06,
+      "loss": 0.8717,
+      "step": 5176
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3967982530593872,
+      "learning_rate": 2.475535981533572e-06,
+      "loss": 0.8661,
+      "step": 5177
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2742984294891357,
+      "learning_rate": 2.473704429445726e-06,
+      "loss": 0.7436,
+      "step": 5178
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6236284971237183,
+      "learning_rate": 2.4718733324357746e-06,
+      "loss": 0.989,
+      "step": 5179
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3161585330963135,
+      "learning_rate": 2.4700426908335627e-06,
+      "loss": 0.8015,
+      "step": 5180
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6530338525772095,
+      "learning_rate": 2.4682125049688536e-06,
+      "loss": 1.0219,
+      "step": 5181
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2839751243591309,
+      "learning_rate": 2.466382775171335e-06,
+      "loss": 0.833,
+      "step": 5182
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.344177484512329,
+      "learning_rate": 2.464553501770603e-06,
+      "loss": 0.7952,
+      "step": 5183
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2699261903762817,
+      "learning_rate": 2.4627246850961804e-06,
+      "loss": 0.8379,
+      "step": 5184
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3103183507919312,
+      "learning_rate": 2.4608963254775027e-06,
+      "loss": 0.8169,
+      "step": 5185
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3321808576583862,
+      "learning_rate": 2.4590684232439216e-06,
+      "loss": 0.8218,
+      "step": 5186
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2994030714035034,
+      "learning_rate": 2.4572409787247127e-06,
+      "loss": 0.8646,
+      "step": 5187
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6955392360687256,
+      "learning_rate": 2.4554139922490617e-06,
+      "loss": 1.0437,
+      "step": 5188
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.367482304573059,
+      "learning_rate": 2.453587464146079e-06,
+      "loss": 0.807,
+      "step": 5189
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.615735650062561,
+      "learning_rate": 2.4517613947447867e-06,
+      "loss": 1.0141,
+      "step": 5190
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3716678619384766,
+      "learning_rate": 2.449935784374125e-06,
+      "loss": 0.8571,
+      "step": 5191
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2912096977233887,
+      "learning_rate": 2.4481106333629573e-06,
+      "loss": 0.7759,
+      "step": 5192
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.608550488948822,
+      "learning_rate": 2.4462859420400547e-06,
+      "loss": 1.0119,
+      "step": 5193
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4171738624572754,
+      "learning_rate": 2.4444617107341157e-06,
+      "loss": 0.8094,
+      "step": 5194
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3461626768112183,
+      "learning_rate": 2.4426379397737474e-06,
+      "loss": 0.8129,
+      "step": 5195
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6014530658721924,
+      "learning_rate": 2.4408146294874785e-06,
+      "loss": 1.0083,
+      "step": 5196
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3905894756317139,
+      "learning_rate": 2.4389917802037518e-06,
+      "loss": 0.7966,
+      "step": 5197
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5957304835319519,
+      "learning_rate": 2.437169392250932e-06,
+      "loss": 1.024,
+      "step": 5198
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.4546098709106445,
+      "learning_rate": 2.4353474659572962e-06,
+      "loss": 0.8042,
+      "step": 5199
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.604885458946228,
+      "learning_rate": 2.43352600165104e-06,
+      "loss": 1.0182,
+      "step": 5200
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.5016204118728638,
+      "learning_rate": 2.4317049996602727e-06,
+      "loss": 0.8479,
+      "step": 5201
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3437436819076538,
+      "learning_rate": 2.429884460313027e-06,
+      "loss": 0.8463,
+      "step": 5202
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3741264343261719,
+      "learning_rate": 2.4280643839372476e-06,
+      "loss": 0.8069,
+      "step": 5203
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6216609477996826,
+      "learning_rate": 2.426244770860794e-06,
+      "loss": 0.9908,
+      "step": 5204
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6220255494117737,
+      "learning_rate": 2.424425621411448e-06,
+      "loss": 1.0197,
+      "step": 5205
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3496317863464355,
+      "learning_rate": 2.4226069359169015e-06,
+      "loss": 0.8259,
+      "step": 5206
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3032935857772827,
+      "learning_rate": 2.4207887147047696e-06,
+      "loss": 0.8719,
+      "step": 5207
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.348844289779663,
+      "learning_rate": 2.4189709581025794e-06,
+      "loss": 0.8136,
+      "step": 5208
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2894400358200073,
+      "learning_rate": 2.4171536664377714e-06,
+      "loss": 0.7801,
+      "step": 5209
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.326984167098999,
+      "learning_rate": 2.4153368400377114e-06,
+      "loss": 0.8336,
+      "step": 5210
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3627647161483765,
+      "learning_rate": 2.4135204792296714e-06,
+      "loss": 0.8178,
+      "step": 5211
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.271558165550232,
+      "learning_rate": 2.4117045843408476e-06,
+      "loss": 0.8151,
+      "step": 5212
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2289342880249023,
+      "learning_rate": 2.4098891556983483e-06,
+      "loss": 0.7746,
+      "step": 5213
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6123126149177551,
+      "learning_rate": 2.4080741936291953e-06,
+      "loss": 1.0024,
+      "step": 5214
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3973737955093384,
+      "learning_rate": 2.406259698460333e-06,
+      "loss": 0.8316,
+      "step": 5215
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.31827712059021,
+      "learning_rate": 2.404445670518616e-06,
+      "loss": 0.8487,
+      "step": 5216
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3417975902557373,
+      "learning_rate": 2.4026321101308186e-06,
+      "loss": 0.8058,
+      "step": 5217
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.290825366973877,
+      "learning_rate": 2.4008190176236287e-06,
+      "loss": 0.8255,
+      "step": 5218
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.3697993755340576,
+      "learning_rate": 2.3990063933236475e-06,
+      "loss": 0.7969,
+      "step": 5219
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3910930156707764,
+      "learning_rate": 2.397194237557399e-06,
+      "loss": 0.8503,
+      "step": 5220
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.341204285621643,
+      "learning_rate": 2.3953825506513145e-06,
+      "loss": 0.8115,
+      "step": 5221
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3574060201644897,
+      "learning_rate": 2.3935713329317486e-06,
+      "loss": 0.8543,
+      "step": 5222
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.3708261251449585,
+      "learning_rate": 2.391760584724966e-06,
+      "loss": 0.8437,
+      "step": 5223
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2844082117080688,
+      "learning_rate": 2.3899503063571463e-06,
+      "loss": 0.7996,
+      "step": 5224
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6186280846595764,
+      "learning_rate": 2.3881404981543898e-06,
+      "loss": 1.0059,
+      "step": 5225
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.2431610822677612,
+      "learning_rate": 2.3863311604427063e-06,
+      "loss": 0.7764,
+      "step": 5226
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 1.317060112953186,
+      "learning_rate": 2.3845222935480266e-06,
+      "loss": 0.8785,
+      "step": 5227
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3858994245529175,
+      "learning_rate": 2.3827138977961915e-06,
+      "loss": 0.7959,
+      "step": 5228
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2765476703643799,
+      "learning_rate": 2.3809059735129575e-06,
+      "loss": 0.8317,
+      "step": 5229
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3155895471572876,
+      "learning_rate": 2.379098521024001e-06,
+      "loss": 0.85,
+      "step": 5230
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2725523710250854,
+      "learning_rate": 2.3772915406549084e-06,
+      "loss": 0.8339,
+      "step": 5231
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.277816653251648,
+      "learning_rate": 2.375485032731181e-06,
+      "loss": 0.8069,
+      "step": 5232
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2988197803497314,
+      "learning_rate": 2.3736789975782404e-06,
+      "loss": 0.8357,
+      "step": 5233
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3711565732955933,
+      "learning_rate": 2.3718734355214157e-06,
+      "loss": 0.8262,
+      "step": 5234
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6305926442146301,
+      "learning_rate": 2.3700683468859576e-06,
+      "loss": 1.0175,
+      "step": 5235
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3001097440719604,
+      "learning_rate": 2.3682637319970265e-06,
+      "loss": 0.8639,
+      "step": 5236
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3018711805343628,
+      "learning_rate": 2.3664595911797e-06,
+      "loss": 0.8412,
+      "step": 5237
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3517817258834839,
+      "learning_rate": 2.364655924758967e-06,
+      "loss": 0.8499,
+      "step": 5238
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.258121371269226,
+      "learning_rate": 2.3628527330597385e-06,
+      "loss": 0.8678,
+      "step": 5239
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3850444555282593,
+      "learning_rate": 2.3610500164068317e-06,
+      "loss": 0.8741,
+      "step": 5240
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6410549283027649,
+      "learning_rate": 2.3592477751249827e-06,
+      "loss": 1.0232,
+      "step": 5241
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.7029433250427246,
+      "learning_rate": 2.3574460095388375e-06,
+      "loss": 0.8213,
+      "step": 5242
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3175973892211914,
+      "learning_rate": 2.3556447199729648e-06,
+      "loss": 0.834,
+      "step": 5243
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3003789186477661,
+      "learning_rate": 2.3538439067518388e-06,
+      "loss": 0.7454,
+      "step": 5244
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3197239637374878,
+      "learning_rate": 2.3520435701998544e-06,
+      "loss": 0.7674,
+      "step": 5245
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3671190738677979,
+      "learning_rate": 2.350243710641317e-06,
+      "loss": 0.853,
+      "step": 5246
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2875115871429443,
+      "learning_rate": 2.348444328400444e-06,
+      "loss": 0.8338,
+      "step": 5247
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5126450061798096,
+      "learning_rate": 2.3466454238013743e-06,
+      "loss": 0.8498,
+      "step": 5248
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.433475136756897,
+      "learning_rate": 2.3448469971681514e-06,
+      "loss": 0.8323,
+      "step": 5249
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6185484528541565,
+      "learning_rate": 2.3430490488247427e-06,
+      "loss": 1.0074,
+      "step": 5250
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.275639533996582,
+      "learning_rate": 2.3412515790950207e-06,
+      "loss": 0.7918,
+      "step": 5251
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3424420356750488,
+      "learning_rate": 2.3394545883027747e-06,
+      "loss": 0.9033,
+      "step": 5252
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3066604137420654,
+      "learning_rate": 2.337658076771711e-06,
+      "loss": 0.774,
+      "step": 5253
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3284944295883179,
+      "learning_rate": 2.3358620448254437e-06,
+      "loss": 0.8288,
+      "step": 5254
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5960077047348022,
+      "learning_rate": 2.3340664927875067e-06,
+      "loss": 1.0176,
+      "step": 5255
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.311362624168396,
+      "learning_rate": 2.3322714209813433e-06,
+      "loss": 0.8519,
+      "step": 5256
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2521941661834717,
+      "learning_rate": 2.3304768297303094e-06,
+      "loss": 0.8072,
+      "step": 5257
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.233693242073059,
+      "learning_rate": 2.328682719357679e-06,
+      "loss": 0.7859,
+      "step": 5258
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.280700922012329,
+      "learning_rate": 2.326889090186636e-06,
+      "loss": 0.8435,
+      "step": 5259
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2786080837249756,
+      "learning_rate": 2.325095942540276e-06,
+      "loss": 0.8152,
+      "step": 5260
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2635160684585571,
+      "learning_rate": 2.3233032767416136e-06,
+      "loss": 0.8455,
+      "step": 5261
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3172441720962524,
+      "learning_rate": 2.3215110931135702e-06,
+      "loss": 0.7882,
+      "step": 5262
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3532732725143433,
+      "learning_rate": 2.3197193919789873e-06,
+      "loss": 0.8088,
+      "step": 5263
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.272469162940979,
+      "learning_rate": 2.317928173660613e-06,
+      "loss": 0.7885,
+      "step": 5264
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6565998196601868,
+      "learning_rate": 2.316137438481109e-06,
+      "loss": 1.0014,
+      "step": 5265
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6215384006500244,
+      "learning_rate": 2.3143471867630563e-06,
+      "loss": 0.993,
+      "step": 5266
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.437565565109253,
+      "learning_rate": 2.3125574188289408e-06,
+      "loss": 0.8189,
+      "step": 5267
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3165427446365356,
+      "learning_rate": 2.3107681350011677e-06,
+      "loss": 0.8726,
+      "step": 5268
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2960718870162964,
+      "learning_rate": 2.308979335602052e-06,
+      "loss": 0.8073,
+      "step": 5269
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5966055989265442,
+      "learning_rate": 2.3071910209538177e-06,
+      "loss": 0.9945,
+      "step": 5270
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3009607791900635,
+      "learning_rate": 2.3054031913786113e-06,
+      "loss": 0.797,
+      "step": 5271
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6181526184082031,
+      "learning_rate": 2.303615847198481e-06,
+      "loss": 1.0142,
+      "step": 5272
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.334096074104309,
+      "learning_rate": 2.3018289887353964e-06,
+      "loss": 0.7735,
+      "step": 5273
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.554992437362671,
+      "learning_rate": 2.3000426163112342e-06,
+      "loss": 0.8113,
+      "step": 5274
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.404834508895874,
+      "learning_rate": 2.298256730247784e-06,
+      "loss": 0.9028,
+      "step": 5275
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3925808668136597,
+      "learning_rate": 2.2964713308667514e-06,
+      "loss": 0.7804,
+      "step": 5276
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3901764154434204,
+      "learning_rate": 2.2946864184897505e-06,
+      "loss": 0.8502,
+      "step": 5277
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2219728231430054,
+      "learning_rate": 2.2929019934383095e-06,
+      "loss": 0.8036,
+      "step": 5278
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4103384017944336,
+      "learning_rate": 2.291118056033866e-06,
+      "loss": 0.7615,
+      "step": 5279
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3690822124481201,
+      "learning_rate": 2.2893346065977764e-06,
+      "loss": 0.8504,
+      "step": 5280
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3612698316574097,
+      "learning_rate": 2.2875516454513023e-06,
+      "loss": 0.8793,
+      "step": 5281
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6168175339698792,
+      "learning_rate": 2.285769172915619e-06,
+      "loss": 1.0122,
+      "step": 5282
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3284567594528198,
+      "learning_rate": 2.2839871893118175e-06,
+      "loss": 0.8273,
+      "step": 5283
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.604800283908844,
+      "learning_rate": 2.282205694960897e-06,
+      "loss": 0.9947,
+      "step": 5284
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.480482578277588,
+      "learning_rate": 2.280424690183768e-06,
+      "loss": 0.8126,
+      "step": 5285
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3924355506896973,
+      "learning_rate": 2.2786441753012565e-06,
+      "loss": 0.8066,
+      "step": 5286
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3779308795928955,
+      "learning_rate": 2.2768641506340977e-06,
+      "loss": 0.858,
+      "step": 5287
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3273431062698364,
+      "learning_rate": 2.275084616502936e-06,
+      "loss": 0.7958,
+      "step": 5288
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.2926298379898071,
+      "learning_rate": 2.2733055732283356e-06,
+      "loss": 0.8214,
+      "step": 5289
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6058923006057739,
+      "learning_rate": 2.271527021130762e-06,
+      "loss": 1.0025,
+      "step": 5290
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3146458864212036,
+      "learning_rate": 2.2697489605306007e-06,
+      "loss": 0.8755,
+      "step": 5291
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.574099063873291,
+      "learning_rate": 2.2679713917481438e-06,
+      "loss": 0.8077,
+      "step": 5292
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.6561579704284668,
+      "learning_rate": 2.266194315103595e-06,
+      "loss": 0.8061,
+      "step": 5293
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4440559148788452,
+      "learning_rate": 2.2644177309170733e-06,
+      "loss": 0.7679,
+      "step": 5294
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.457335114479065,
+      "learning_rate": 2.262641639508603e-06,
+      "loss": 0.7706,
+      "step": 5295
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5984505414962769,
+      "learning_rate": 2.260866041198127e-06,
+      "loss": 1.0183,
+      "step": 5296
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3547741174697876,
+      "learning_rate": 2.259090936305493e-06,
+      "loss": 0.8688,
+      "step": 5297
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3099101781845093,
+      "learning_rate": 2.25731632515046e-06,
+      "loss": 0.8854,
+      "step": 5298
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6193394660949707,
+      "learning_rate": 2.255542208052704e-06,
+      "loss": 1.0127,
+      "step": 5299
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.5081169605255127,
+      "learning_rate": 2.2537685853318046e-06,
+      "loss": 0.8324,
+      "step": 5300
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.386731743812561,
+      "learning_rate": 2.2519954573072596e-06,
+      "loss": 0.8396,
+      "step": 5301
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4654008150100708,
+      "learning_rate": 2.250222824298472e-06,
+      "loss": 0.8358,
+      "step": 5302
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.3321863412857056,
+      "learning_rate": 2.2484506866247556e-06,
+      "loss": 0.832,
+      "step": 5303
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 1.4802926778793335,
+      "learning_rate": 2.2466790446053412e-06,
+      "loss": 0.8251,
+      "step": 5304
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.8578357696533203,
+      "learning_rate": 2.2449078985593614e-06,
+      "loss": 0.8045,
+      "step": 5305
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5232126712799072,
+      "learning_rate": 2.243137248805868e-06,
+      "loss": 0.8832,
+      "step": 5306
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.330433964729309,
+      "learning_rate": 2.2413670956638193e-06,
+      "loss": 0.8726,
+      "step": 5307
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6333154439926147,
+      "learning_rate": 2.239597439452081e-06,
+      "loss": 1.0163,
+      "step": 5308
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.466425895690918,
+      "learning_rate": 2.237828280489437e-06,
+      "loss": 0.8576,
+      "step": 5309
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2938052415847778,
+      "learning_rate": 2.2360596190945727e-06,
+      "loss": 0.8212,
+      "step": 5310
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3817752599716187,
+      "learning_rate": 2.2342914555860934e-06,
+      "loss": 0.8097,
+      "step": 5311
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6090506911277771,
+      "learning_rate": 2.232523790282507e-06,
+      "loss": 1.0091,
+      "step": 5312
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5899628400802612,
+      "learning_rate": 2.230756623502233e-06,
+      "loss": 1.011,
+      "step": 5313
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3758622407913208,
+      "learning_rate": 2.2289899555636062e-06,
+      "loss": 0.8311,
+      "step": 5314
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3845242261886597,
+      "learning_rate": 2.2272237867848664e-06,
+      "loss": 0.8406,
+      "step": 5315
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.355024814605713,
+      "learning_rate": 2.225458117484162e-06,
+      "loss": 0.8105,
+      "step": 5316
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3338241577148438,
+      "learning_rate": 2.2236929479795592e-06,
+      "loss": 0.7833,
+      "step": 5317
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.587446391582489,
+      "learning_rate": 2.2219282785890273e-06,
+      "loss": 0.9952,
+      "step": 5318
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3434629440307617,
+      "learning_rate": 2.220164109630447e-06,
+      "loss": 0.7996,
+      "step": 5319
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3224427700042725,
+      "learning_rate": 2.218400441421608e-06,
+      "loss": 0.8326,
+      "step": 5320
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3299753665924072,
+      "learning_rate": 2.216637274280215e-06,
+      "loss": 0.8005,
+      "step": 5321
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5913423895835876,
+      "learning_rate": 2.2148746085238763e-06,
+      "loss": 1.0254,
+      "step": 5322
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4036097526550293,
+      "learning_rate": 2.213112444470111e-06,
+      "loss": 0.7373,
+      "step": 5323
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3282718658447266,
+      "learning_rate": 2.211350782436353e-06,
+      "loss": 0.8421,
+      "step": 5324
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.329417109489441,
+      "learning_rate": 2.209589622739938e-06,
+      "loss": 0.8135,
+      "step": 5325
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5781421661376953,
+      "learning_rate": 2.2078289656981163e-06,
+      "loss": 0.8544,
+      "step": 5326
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.495559811592102,
+      "learning_rate": 2.206068811628047e-06,
+      "loss": 0.8388,
+      "step": 5327
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3361177444458008,
+      "learning_rate": 2.204309160846797e-06,
+      "loss": 0.819,
+      "step": 5328
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.5161248445510864,
+      "learning_rate": 2.2025500136713453e-06,
+      "loss": 0.7548,
+      "step": 5329
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5923037528991699,
+      "learning_rate": 2.2007913704185774e-06,
+      "loss": 0.9795,
+      "step": 5330
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2928913831710815,
+      "learning_rate": 2.1990332314052866e-06,
+      "loss": 0.765,
+      "step": 5331
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.415818214416504,
+      "learning_rate": 2.1972755969481823e-06,
+      "loss": 0.8315,
+      "step": 5332
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3356910943984985,
+      "learning_rate": 2.1955184673638745e-06,
+      "loss": 0.8382,
+      "step": 5333
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2912306785583496,
+      "learning_rate": 2.1937618429688896e-06,
+      "loss": 0.8462,
+      "step": 5334
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.252496361732483,
+      "learning_rate": 2.1920057240796583e-06,
+      "loss": 0.7408,
+      "step": 5335
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3019781112670898,
+      "learning_rate": 2.1902501110125196e-06,
+      "loss": 0.747,
+      "step": 5336
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3161519765853882,
+      "learning_rate": 2.188495004083727e-06,
+      "loss": 0.8058,
+      "step": 5337
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3423539400100708,
+      "learning_rate": 2.186740403609436e-06,
+      "loss": 0.8129,
+      "step": 5338
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3611446619033813,
+      "learning_rate": 2.1849863099057167e-06,
+      "loss": 0.8288,
+      "step": 5339
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3251062631607056,
+      "learning_rate": 2.1832327232885442e-06,
+      "loss": 0.8805,
+      "step": 5340
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2891658544540405,
+      "learning_rate": 2.1814796440738016e-06,
+      "loss": 0.8074,
+      "step": 5341
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3761727809906006,
+      "learning_rate": 2.1797270725772856e-06,
+      "loss": 0.7996,
+      "step": 5342
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.284893274307251,
+      "learning_rate": 2.177975009114695e-06,
+      "loss": 0.8731,
+      "step": 5343
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2866672277450562,
+      "learning_rate": 2.1762234540016434e-06,
+      "loss": 0.8613,
+      "step": 5344
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2231590747833252,
+      "learning_rate": 2.1744724075536488e-06,
+      "loss": 0.8323,
+      "step": 5345
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6149075031280518,
+      "learning_rate": 2.172721870086136e-06,
+      "loss": 1.0057,
+      "step": 5346
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4901950359344482,
+      "learning_rate": 2.170971841914444e-06,
+      "loss": 0.8694,
+      "step": 5347
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2609037160873413,
+      "learning_rate": 2.169222323353815e-06,
+      "loss": 0.839,
+      "step": 5348
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.415090560913086,
+      "learning_rate": 2.1674733147193995e-06,
+      "loss": 0.7637,
+      "step": 5349
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.588273823261261,
+      "learning_rate": 2.1657248163262607e-06,
+      "loss": 1.0297,
+      "step": 5350
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3720656633377075,
+      "learning_rate": 2.163976828489364e-06,
+      "loss": 0.8706,
+      "step": 5351
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.289925217628479,
+      "learning_rate": 2.1622293515235893e-06,
+      "loss": 0.8303,
+      "step": 5352
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4076344966888428,
+      "learning_rate": 2.1604823857437185e-06,
+      "loss": 0.8036,
+      "step": 5353
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5834177136421204,
+      "learning_rate": 2.1587359314644423e-06,
+      "loss": 1.0118,
+      "step": 5354
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.327070951461792,
+      "learning_rate": 2.156989989000364e-06,
+      "loss": 0.8355,
+      "step": 5355
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3094944953918457,
+      "learning_rate": 2.1552445586659904e-06,
+      "loss": 0.8017,
+      "step": 5356
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2584834098815918,
+      "learning_rate": 2.1534996407757354e-06,
+      "loss": 0.7886,
+      "step": 5357
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2579327821731567,
+      "learning_rate": 2.151755235643925e-06,
+      "loss": 0.8369,
+      "step": 5358
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3773833513259888,
+      "learning_rate": 2.150011343584788e-06,
+      "loss": 0.8172,
+      "step": 5359
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3841303586959839,
+      "learning_rate": 2.148267964912464e-06,
+      "loss": 0.84,
+      "step": 5360
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.3126728534698486,
+      "learning_rate": 2.1465250999409975e-06,
+      "loss": 0.7912,
+      "step": 5361
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3393559455871582,
+      "learning_rate": 2.144782748984345e-06,
+      "loss": 0.8006,
+      "step": 5362
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3354254961013794,
+      "learning_rate": 2.1430409123563652e-06,
+      "loss": 0.7896,
+      "step": 5363
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3231542110443115,
+      "learning_rate": 2.141299590370825e-06,
+      "loss": 0.7879,
+      "step": 5364
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.330003023147583,
+      "learning_rate": 2.1395587833414033e-06,
+      "loss": 0.8005,
+      "step": 5365
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4271845817565918,
+      "learning_rate": 2.1378184915816795e-06,
+      "loss": 0.7754,
+      "step": 5366
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3811886310577393,
+      "learning_rate": 2.136078715405147e-06,
+      "loss": 0.8628,
+      "step": 5367
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6241826415061951,
+      "learning_rate": 2.1343394551252018e-06,
+      "loss": 1.0305,
+      "step": 5368
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.6976467370986938,
+      "learning_rate": 2.132600711055145e-06,
+      "loss": 0.8374,
+      "step": 5369
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3167167901992798,
+      "learning_rate": 2.1308624835081924e-06,
+      "loss": 0.7868,
+      "step": 5370
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6187125444412231,
+      "learning_rate": 2.129124772797458e-06,
+      "loss": 1.0147,
+      "step": 5371
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.278370976448059,
+      "learning_rate": 2.1273875792359704e-06,
+      "loss": 0.8583,
+      "step": 5372
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3342353105545044,
+      "learning_rate": 2.12565090313666e-06,
+      "loss": 0.7716,
+      "step": 5373
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.338541865348816,
+      "learning_rate": 2.1239147448123637e-06,
+      "loss": 0.7998,
+      "step": 5374
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4982637166976929,
+      "learning_rate": 2.1221791045758306e-06,
+      "loss": 0.8484,
+      "step": 5375
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3329942226409912,
+      "learning_rate": 2.12044398273971e-06,
+      "loss": 0.846,
+      "step": 5376
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3335684537887573,
+      "learning_rate": 2.118709379616559e-06,
+      "loss": 0.7818,
+      "step": 5377
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3639403581619263,
+      "learning_rate": 2.1169752955188476e-06,
+      "loss": 0.8143,
+      "step": 5378
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.2878185510635376,
+      "learning_rate": 2.115241730758943e-06,
+      "loss": 0.819,
+      "step": 5379
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.3112413883209229,
+      "learning_rate": 2.1135086856491273e-06,
+      "loss": 0.8778,
+      "step": 5380
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.365369200706482,
+      "learning_rate": 2.111776160501583e-06,
+      "loss": 0.823,
+      "step": 5381
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5614254474639893,
+      "learning_rate": 2.1100441556283992e-06,
+      "loss": 0.8393,
+      "step": 5382
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.263627052307129,
+      "learning_rate": 2.1083126713415767e-06,
+      "loss": 0.8336,
+      "step": 5383
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.263931155204773,
+      "learning_rate": 2.106581707953016e-06,
+      "loss": 0.767,
+      "step": 5384
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.292508840560913,
+      "learning_rate": 2.104851265774529e-06,
+      "loss": 0.804,
+      "step": 5385
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2172366380691528,
+      "learning_rate": 2.1031213451178312e-06,
+      "loss": 0.7653,
+      "step": 5386
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6306527853012085,
+      "learning_rate": 2.101391946294541e-06,
+      "loss": 0.9996,
+      "step": 5387
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.628276526927948,
+      "learning_rate": 2.099663069616191e-06,
+      "loss": 1.0165,
+      "step": 5388
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4496091604232788,
+      "learning_rate": 2.0979347153942105e-06,
+      "loss": 0.8888,
+      "step": 5389
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3924834728240967,
+      "learning_rate": 2.0962068839399423e-06,
+      "loss": 0.832,
+      "step": 5390
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6015112400054932,
+      "learning_rate": 2.0944795755646315e-06,
+      "loss": 1.0216,
+      "step": 5391
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3590208292007446,
+      "learning_rate": 2.0927527905794258e-06,
+      "loss": 0.8065,
+      "step": 5392
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5958194732666016,
+      "learning_rate": 2.091026529295387e-06,
+      "loss": 0.9942,
+      "step": 5393
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.368297815322876,
+      "learning_rate": 2.089300792023473e-06,
+      "loss": 0.81,
+      "step": 5394
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.315234661102295,
+      "learning_rate": 2.0875755790745573e-06,
+      "loss": 0.8167,
+      "step": 5395
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5910853147506714,
+      "learning_rate": 2.0858508907594092e-06,
+      "loss": 1.0192,
+      "step": 5396
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3458459377288818,
+      "learning_rate": 2.0841267273887108e-06,
+      "loss": 0.7977,
+      "step": 5397
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3364261388778687,
+      "learning_rate": 2.082403089273042e-06,
+      "loss": 0.7821,
+      "step": 5398
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5849350690841675,
+      "learning_rate": 2.080679976722899e-06,
+      "loss": 1.009,
+      "step": 5399
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.618439257144928,
+      "learning_rate": 2.078957390048674e-06,
+      "loss": 1.0013,
+      "step": 5400
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5991705060005188,
+      "learning_rate": 2.0772353295606684e-06,
+      "loss": 1.0121,
+      "step": 5401
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2636933326721191,
+      "learning_rate": 2.075513795569085e-06,
+      "loss": 0.8321,
+      "step": 5402
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3466380834579468,
+      "learning_rate": 2.0737927883840397e-06,
+      "loss": 0.8055,
+      "step": 5403
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2843424081802368,
+      "learning_rate": 2.0720723083155464e-06,
+      "loss": 0.8198,
+      "step": 5404
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.36204195022583,
+      "learning_rate": 2.070352355673524e-06,
+      "loss": 0.828,
+      "step": 5405
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3503000736236572,
+      "learning_rate": 2.0686329307678025e-06,
+      "loss": 0.8347,
+      "step": 5406
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6347265839576721,
+      "learning_rate": 2.06691403390811e-06,
+      "loss": 0.9995,
+      "step": 5407
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3773717880249023,
+      "learning_rate": 2.0651956654040855e-06,
+      "loss": 0.8575,
+      "step": 5408
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3900822401046753,
+      "learning_rate": 2.0634778255652676e-06,
+      "loss": 0.7958,
+      "step": 5409
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.297010898590088,
+      "learning_rate": 2.061760514701101e-06,
+      "loss": 0.7836,
+      "step": 5410
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5473984479904175,
+      "learning_rate": 2.060043733120939e-06,
+      "loss": 0.8565,
+      "step": 5411
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6067873239517212,
+      "learning_rate": 2.0583274811340327e-06,
+      "loss": 1.0088,
+      "step": 5412
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5951722860336304,
+      "learning_rate": 2.0566117590495454e-06,
+      "loss": 0.9912,
+      "step": 5413
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3466870784759521,
+      "learning_rate": 2.05489656717654e-06,
+      "loss": 0.8191,
+      "step": 5414
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2781367301940918,
+      "learning_rate": 2.053181905823982e-06,
+      "loss": 0.7833,
+      "step": 5415
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3168739080429077,
+      "learning_rate": 2.051467775300749e-06,
+      "loss": 0.8413,
+      "step": 5416
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4019131660461426,
+      "learning_rate": 2.049754175915615e-06,
+      "loss": 0.8012,
+      "step": 5417
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4166545867919922,
+      "learning_rate": 2.0480411079772645e-06,
+      "loss": 0.8379,
+      "step": 5418
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3118915557861328,
+      "learning_rate": 2.0463285717942816e-06,
+      "loss": 0.8232,
+      "step": 5419
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3109101057052612,
+      "learning_rate": 2.044616567675155e-06,
+      "loss": 0.8358,
+      "step": 5420
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6357645392417908,
+      "learning_rate": 2.0429050959282832e-06,
+      "loss": 1.0014,
+      "step": 5421
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3855509757995605,
+      "learning_rate": 2.0411941568619602e-06,
+      "loss": 0.7878,
+      "step": 5422
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.245985507965088,
+      "learning_rate": 2.039483750784392e-06,
+      "loss": 0.8606,
+      "step": 5423
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.2535016536712646,
+      "learning_rate": 2.0377738780036843e-06,
+      "loss": 0.787,
+      "step": 5424
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3264219760894775,
+      "learning_rate": 2.036064538827845e-06,
+      "loss": 0.7499,
+      "step": 5425
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3441452980041504,
+      "learning_rate": 2.0343557335647916e-06,
+      "loss": 0.7966,
+      "step": 5426
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4628255367279053,
+      "learning_rate": 2.032647462522339e-06,
+      "loss": 0.8368,
+      "step": 5427
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.316988229751587,
+      "learning_rate": 2.030939726008213e-06,
+      "loss": 0.8789,
+      "step": 5428
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.468009352684021,
+      "learning_rate": 2.029232524330037e-06,
+      "loss": 0.8456,
+      "step": 5429
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.280771255493164,
+      "learning_rate": 2.0275258577953383e-06,
+      "loss": 0.8001,
+      "step": 5430
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3348088264465332,
+      "learning_rate": 2.0258197267115538e-06,
+      "loss": 0.8352,
+      "step": 5431
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.402309536933899,
+      "learning_rate": 2.024114131386017e-06,
+      "loss": 0.8326,
+      "step": 5432
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3554930686950684,
+      "learning_rate": 2.0224090721259675e-06,
+      "loss": 0.8222,
+      "step": 5433
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3824371099472046,
+      "learning_rate": 2.0207045492385514e-06,
+      "loss": 0.819,
+      "step": 5434
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6361715197563171,
+      "learning_rate": 2.0190005630308113e-06,
+      "loss": 1.0052,
+      "step": 5435
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3507695198059082,
+      "learning_rate": 2.017297113809702e-06,
+      "loss": 0.8349,
+      "step": 5436
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3313909769058228,
+      "learning_rate": 2.015594201882074e-06,
+      "loss": 0.8126,
+      "step": 5437
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.333404541015625,
+      "learning_rate": 2.013891827554684e-06,
+      "loss": 0.8584,
+      "step": 5438
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.507324457168579,
+      "learning_rate": 2.01218999113419e-06,
+      "loss": 0.8691,
+      "step": 5439
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3558796644210815,
+      "learning_rate": 2.010488692927159e-06,
+      "loss": 0.8249,
+      "step": 5440
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5867778658866882,
+      "learning_rate": 2.0087879332400533e-06,
+      "loss": 1.0045,
+      "step": 5441
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5977861285209656,
+      "learning_rate": 2.007087712379242e-06,
+      "loss": 1.032,
+      "step": 5442
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6121914982795715,
+      "learning_rate": 2.005388030650999e-06,
+      "loss": 1.0191,
+      "step": 5443
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3330552577972412,
+      "learning_rate": 2.003688888361497e-06,
+      "loss": 0.8057,
+      "step": 5444
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6030218005180359,
+      "learning_rate": 2.001990285816813e-06,
+      "loss": 1.0392,
+      "step": 5445
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4026437997817993,
+      "learning_rate": 2.0002922233229288e-06,
+      "loss": 0.8693,
+      "step": 5446
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4288228750228882,
+      "learning_rate": 1.9985947011857276e-06,
+      "loss": 0.8348,
+      "step": 5447
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.798517107963562,
+      "learning_rate": 1.9968977197109916e-06,
+      "loss": 0.7941,
+      "step": 5448
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6216334104537964,
+      "learning_rate": 1.995201279204414e-06,
+      "loss": 0.9763,
+      "step": 5449
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6113596558570862,
+      "learning_rate": 1.993505379971581e-06,
+      "loss": 1.0094,
+      "step": 5450
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.4969699382781982,
+      "learning_rate": 1.99181002231799e-06,
+      "loss": 0.8474,
+      "step": 5451
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.8296064138412476,
+      "learning_rate": 1.9901152065490337e-06,
+      "loss": 0.7538,
+      "step": 5452
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3926714658737183,
+      "learning_rate": 1.98842093297001e-06,
+      "loss": 0.8032,
+      "step": 5453
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.3152986764907837,
+      "learning_rate": 1.986727201886122e-06,
+      "loss": 0.8206,
+      "step": 5454
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5945865511894226,
+      "learning_rate": 1.9850340136024688e-06,
+      "loss": 1.0178,
+      "step": 5455
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5868383049964905,
+      "learning_rate": 1.9833413684240593e-06,
+      "loss": 1.0071,
+      "step": 5456
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5962748527526855,
+      "learning_rate": 1.9816492666557983e-06,
+      "loss": 1.0057,
+      "step": 5457
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5969836115837097,
+      "learning_rate": 1.979957708602494e-06,
+      "loss": 1.0185,
+      "step": 5458
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3262616395950317,
+      "learning_rate": 1.97826669456886e-06,
+      "loss": 0.8311,
+      "step": 5459
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3882522583007812,
+      "learning_rate": 1.976576224859509e-06,
+      "loss": 0.8387,
+      "step": 5460
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5719029903411865,
+      "learning_rate": 1.974886299778954e-06,
+      "loss": 0.9986,
+      "step": 5461
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3414729833602905,
+      "learning_rate": 1.9731969196316146e-06,
+      "loss": 0.8612,
+      "step": 5462
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.439680814743042,
+      "learning_rate": 1.971508084721808e-06,
+      "loss": 0.7693,
+      "step": 5463
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3215378522872925,
+      "learning_rate": 1.9698197953537566e-06,
+      "loss": 0.8806,
+      "step": 5464
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3195592164993286,
+      "learning_rate": 1.968132051831583e-06,
+      "loss": 0.8012,
+      "step": 5465
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.324954628944397,
+      "learning_rate": 1.9664448544593072e-06,
+      "loss": 0.8145,
+      "step": 5466
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.519609808921814,
+      "learning_rate": 1.964758203540861e-06,
+      "loss": 0.7929,
+      "step": 5467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3478657007217407,
+      "learning_rate": 1.9630720993800667e-06,
+      "loss": 0.8678,
+      "step": 5468
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3130627870559692,
+      "learning_rate": 1.961386542280657e-06,
+      "loss": 0.7619,
+      "step": 5469
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6036514639854431,
+      "learning_rate": 1.9597015325462597e-06,
+      "loss": 1.014,
+      "step": 5470
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.638012707233429,
+      "learning_rate": 1.9580170704804058e-06,
+      "loss": 1.0322,
+      "step": 5471
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3745239973068237,
+      "learning_rate": 1.956333156386531e-06,
+      "loss": 0.8303,
+      "step": 5472
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2426931858062744,
+      "learning_rate": 1.954649790567967e-06,
+      "loss": 0.7534,
+      "step": 5473
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6092899441719055,
+      "learning_rate": 1.952966973327952e-06,
+      "loss": 1.017,
+      "step": 5474
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2811999320983887,
+      "learning_rate": 1.951284704969621e-06,
+      "loss": 0.8291,
+      "step": 5475
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5933043956756592,
+      "learning_rate": 1.9496029857960103e-06,
+      "loss": 0.9988,
+      "step": 5476
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.329502820968628,
+      "learning_rate": 1.947921816110062e-06,
+      "loss": 0.7946,
+      "step": 5477
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2764917612075806,
+      "learning_rate": 1.946241196214615e-06,
+      "loss": 0.8172,
+      "step": 5478
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.621387243270874,
+      "learning_rate": 1.944561126412409e-06,
+      "loss": 1.0141,
+      "step": 5479
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5940325260162354,
+      "learning_rate": 1.942881607006086e-06,
+      "loss": 0.9843,
+      "step": 5480
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5867403745651245,
+      "learning_rate": 1.9412026382981897e-06,
+      "loss": 1.007,
+      "step": 5481
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5929911136627197,
+      "learning_rate": 1.9395242205911634e-06,
+      "loss": 1.0067,
+      "step": 5482
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.361215353012085,
+      "learning_rate": 1.9378463541873494e-06,
+      "loss": 0.8261,
+      "step": 5483
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 3.364744186401367,
+      "learning_rate": 1.936169039388996e-06,
+      "loss": 0.8216,
+      "step": 5484
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6003347039222717,
+      "learning_rate": 1.934492276498246e-06,
+      "loss": 1.015,
+      "step": 5485
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.7653372287750244,
+      "learning_rate": 1.932816065817145e-06,
+      "loss": 0.8205,
+      "step": 5486
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3726156949996948,
+      "learning_rate": 1.9311404076476433e-06,
+      "loss": 0.8436,
+      "step": 5487
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6165573000907898,
+      "learning_rate": 1.929465302291583e-06,
+      "loss": 1.0296,
+      "step": 5488
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.279543399810791,
+      "learning_rate": 1.9277907500507166e-06,
+      "loss": 0.7963,
+      "step": 5489
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4232475757598877,
+      "learning_rate": 1.9261167512266893e-06,
+      "loss": 0.7848,
+      "step": 5490
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4162862300872803,
+      "learning_rate": 1.9244433061210477e-06,
+      "loss": 0.7831,
+      "step": 5491
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3658372163772583,
+      "learning_rate": 1.9227704150352437e-06,
+      "loss": 0.8304,
+      "step": 5492
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5813940763473511,
+      "learning_rate": 1.921098078270625e-06,
+      "loss": 1.0249,
+      "step": 5493
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3309913873672485,
+      "learning_rate": 1.919426296128438e-06,
+      "loss": 0.8221,
+      "step": 5494
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.5014163255691528,
+      "learning_rate": 1.917755068909835e-06,
+      "loss": 0.8758,
+      "step": 5495
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5954284071922302,
+      "learning_rate": 1.9160843969158615e-06,
+      "loss": 1.0341,
+      "step": 5496
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.0812301635742188,
+      "learning_rate": 1.91441428044747e-06,
+      "loss": 0.8421,
+      "step": 5497
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.304259181022644,
+      "learning_rate": 1.9127447198055076e-06,
+      "loss": 0.7789,
+      "step": 5498
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.602641761302948,
+      "learning_rate": 1.9110757152907215e-06,
+      "loss": 1.008,
+      "step": 5499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3687268495559692,
+      "learning_rate": 1.909407267203763e-06,
+      "loss": 0.8287,
+      "step": 5500
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3050707578659058,
+      "learning_rate": 1.9077393758451777e-06,
+      "loss": 0.8644,
+      "step": 5501
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2315342426300049,
+      "learning_rate": 1.9060720415154165e-06,
+      "loss": 0.8502,
+      "step": 5502
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3317787647247314,
+      "learning_rate": 1.904405264514826e-06,
+      "loss": 0.8518,
+      "step": 5503
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.37966787815094,
+      "learning_rate": 1.9027390451436505e-06,
+      "loss": 0.8103,
+      "step": 5504
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2851195335388184,
+      "learning_rate": 1.9010733837020417e-06,
+      "loss": 0.815,
+      "step": 5505
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2725906372070312,
+      "learning_rate": 1.8994082804900422e-06,
+      "loss": 0.8035,
+      "step": 5506
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5912865400314331,
+      "learning_rate": 1.8977437358076002e-06,
+      "loss": 1.0344,
+      "step": 5507
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.4213303327560425,
+      "learning_rate": 1.8960797499545602e-06,
+      "loss": 0.8318,
+      "step": 5508
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3637361526489258,
+      "learning_rate": 1.8944163232306645e-06,
+      "loss": 0.814,
+      "step": 5509
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2558447122573853,
+      "learning_rate": 1.8927534559355603e-06,
+      "loss": 0.8514,
+      "step": 5510
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3477920293807983,
+      "learning_rate": 1.8910911483687872e-06,
+      "loss": 0.8344,
+      "step": 5511
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6148287057876587,
+      "learning_rate": 1.889429400829791e-06,
+      "loss": 1.0144,
+      "step": 5512
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6043081879615784,
+      "learning_rate": 1.8877682136179115e-06,
+      "loss": 1.0104,
+      "step": 5513
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3656575679779053,
+      "learning_rate": 1.8861075870323864e-06,
+      "loss": 0.7897,
+      "step": 5514
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.503416657447815,
+      "learning_rate": 1.8844475213723595e-06,
+      "loss": 0.828,
+      "step": 5515
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.26541268825531,
+      "learning_rate": 1.8827880169368667e-06,
+      "loss": 0.8249,
+      "step": 5516
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3459964990615845,
+      "learning_rate": 1.881129074024844e-06,
+      "loss": 0.7721,
+      "step": 5517
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.489195704460144,
+      "learning_rate": 1.8794706929351303e-06,
+      "loss": 0.8114,
+      "step": 5518
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3550517559051514,
+      "learning_rate": 1.8778128739664591e-06,
+      "loss": 0.8043,
+      "step": 5519
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3679630756378174,
+      "learning_rate": 1.8761556174174644e-06,
+      "loss": 0.8062,
+      "step": 5520
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3193665742874146,
+      "learning_rate": 1.8744989235866766e-06,
+      "loss": 0.8317,
+      "step": 5521
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3485053777694702,
+      "learning_rate": 1.8728427927725295e-06,
+      "loss": 0.8681,
+      "step": 5522
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3015291690826416,
+      "learning_rate": 1.8711872252733514e-06,
+      "loss": 0.8057,
+      "step": 5523
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.2906385660171509,
+      "learning_rate": 1.8695322213873684e-06,
+      "loss": 0.8337,
+      "step": 5524
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.618851363658905,
+      "learning_rate": 1.8678777814127108e-06,
+      "loss": 1.0129,
+      "step": 5525
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.318224310874939,
+      "learning_rate": 1.8662239056474013e-06,
+      "loss": 0.8039,
+      "step": 5526
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.612311601638794,
+      "learning_rate": 1.8645705943893615e-06,
+      "loss": 1.0015,
+      "step": 5527
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6027430295944214,
+      "learning_rate": 1.8629178479364174e-06,
+      "loss": 0.9894,
+      "step": 5528
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3102954626083374,
+      "learning_rate": 1.8612656665862843e-06,
+      "loss": 0.8297,
+      "step": 5529
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 1.3135167360305786,
+      "learning_rate": 1.8596140506365839e-06,
+      "loss": 0.8265,
+      "step": 5530
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5774848461151123,
+      "learning_rate": 1.8579630003848309e-06,
+      "loss": 0.9926,
+      "step": 5531
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6009293794631958,
+      "learning_rate": 1.8563125161284384e-06,
+      "loss": 1.0032,
+      "step": 5532
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3056671619415283,
+      "learning_rate": 1.8546625981647215e-06,
+      "loss": 0.7774,
+      "step": 5533
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3160566091537476,
+      "learning_rate": 1.8530132467908868e-06,
+      "loss": 0.8356,
+      "step": 5534
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.60333251953125,
+      "learning_rate": 1.8513644623040472e-06,
+      "loss": 1.0137,
+      "step": 5535
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3453757762908936,
+      "learning_rate": 1.8497162450012062e-06,
+      "loss": 0.8596,
+      "step": 5536
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5913205146789551,
+      "learning_rate": 1.8480685951792659e-06,
+      "loss": 0.9967,
+      "step": 5537
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7688844203948975,
+      "learning_rate": 1.8464215131350327e-06,
+      "loss": 0.876,
+      "step": 5538
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.780996322631836,
+      "learning_rate": 1.8447749991652014e-06,
+      "loss": 0.7472,
+      "step": 5539
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4480515718460083,
+      "learning_rate": 1.8431290535663737e-06,
+      "loss": 0.8401,
+      "step": 5540
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3164787292480469,
+      "learning_rate": 1.841483676635042e-06,
+      "loss": 0.8205,
+      "step": 5541
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.348060131072998,
+      "learning_rate": 1.8398388686675965e-06,
+      "loss": 0.8732,
+      "step": 5542
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.431010365486145,
+      "learning_rate": 1.8381946299603316e-06,
+      "loss": 0.8053,
+      "step": 5543
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2717070579528809,
+      "learning_rate": 1.83655096080943e-06,
+      "loss": 0.7827,
+      "step": 5544
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6197527647018433,
+      "learning_rate": 1.8349078615109805e-06,
+      "loss": 1.0043,
+      "step": 5545
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3442622423171997,
+      "learning_rate": 1.8332653323609638e-06,
+      "loss": 0.8458,
+      "step": 5546
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2826696634292603,
+      "learning_rate": 1.8316233736552568e-06,
+      "loss": 0.7917,
+      "step": 5547
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3466721773147583,
+      "learning_rate": 1.82998198568964e-06,
+      "loss": 0.8596,
+      "step": 5548
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4018661975860596,
+      "learning_rate": 1.8283411687597851e-06,
+      "loss": 0.8145,
+      "step": 5549
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.319158911705017,
+      "learning_rate": 1.826700923161262e-06,
+      "loss": 0.8661,
+      "step": 5550
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.34076988697052,
+      "learning_rate": 1.8250612491895415e-06,
+      "loss": 0.8492,
+      "step": 5551
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2676222324371338,
+      "learning_rate": 1.8234221471399849e-06,
+      "loss": 0.7866,
+      "step": 5552
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.610063910484314,
+      "learning_rate": 1.8217836173078584e-06,
+      "loss": 1.0017,
+      "step": 5553
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2787036895751953,
+      "learning_rate": 1.8201456599883188e-06,
+      "loss": 0.8056,
+      "step": 5554
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4180176258087158,
+      "learning_rate": 1.8185082754764205e-06,
+      "loss": 0.7921,
+      "step": 5555
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3433529138565063,
+      "learning_rate": 1.8168714640671193e-06,
+      "loss": 0.8995,
+      "step": 5556
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4944853782653809,
+      "learning_rate": 1.815235226055263e-06,
+      "loss": 0.825,
+      "step": 5557
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2519810199737549,
+      "learning_rate": 1.8135995617355957e-06,
+      "loss": 0.8478,
+      "step": 5558
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3531962633132935,
+      "learning_rate": 1.811964471402764e-06,
+      "loss": 0.8588,
+      "step": 5559
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5714420080184937,
+      "learning_rate": 1.8103299553513048e-06,
+      "loss": 0.8454,
+      "step": 5560
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.266432285308838,
+      "learning_rate": 1.8086960138756531e-06,
+      "loss": 0.7858,
+      "step": 5561
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6058846116065979,
+      "learning_rate": 1.8070626472701446e-06,
+      "loss": 1.0384,
+      "step": 5562
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6111336350440979,
+      "learning_rate": 1.8054298558290063e-06,
+      "loss": 0.9951,
+      "step": 5563
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2761151790618896,
+      "learning_rate": 1.803797639846363e-06,
+      "loss": 0.7952,
+      "step": 5564
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 4.375199317932129,
+      "learning_rate": 1.802165999616235e-06,
+      "loss": 0.8036,
+      "step": 5565
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6049579381942749,
+      "learning_rate": 1.8005349354325435e-06,
+      "loss": 0.9894,
+      "step": 5566
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2712759971618652,
+      "learning_rate": 1.7989044475891e-06,
+      "loss": 0.764,
+      "step": 5567
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3243943452835083,
+      "learning_rate": 1.7972745363796168e-06,
+      "loss": 0.8093,
+      "step": 5568
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.468772530555725,
+      "learning_rate": 1.795645202097699e-06,
+      "loss": 0.8389,
+      "step": 5569
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3058733940124512,
+      "learning_rate": 1.794016445036848e-06,
+      "loss": 0.8264,
+      "step": 5570
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4095447063446045,
+      "learning_rate": 1.792388265490465e-06,
+      "loss": 0.8098,
+      "step": 5571
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3091943264007568,
+      "learning_rate": 1.7907606637518415e-06,
+      "loss": 0.8585,
+      "step": 5572
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4442086219787598,
+      "learning_rate": 1.7891336401141712e-06,
+      "loss": 0.7747,
+      "step": 5573
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3673713207244873,
+      "learning_rate": 1.7875071948705386e-06,
+      "loss": 0.8217,
+      "step": 5574
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2366164922714233,
+      "learning_rate": 1.7858813283139237e-06,
+      "loss": 0.7984,
+      "step": 5575
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.278716802597046,
+      "learning_rate": 1.7842560407372083e-06,
+      "loss": 0.8003,
+      "step": 5576
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.321641206741333,
+      "learning_rate": 1.7826313324331646e-06,
+      "loss": 0.8504,
+      "step": 5577
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5957860350608826,
+      "learning_rate": 1.7810072036944588e-06,
+      "loss": 1.0089,
+      "step": 5578
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3900322914123535,
+      "learning_rate": 1.7793836548136595e-06,
+      "loss": 0.8161,
+      "step": 5579
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.384291172027588,
+      "learning_rate": 1.7777606860832242e-06,
+      "loss": 0.8448,
+      "step": 5580
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5921191573143005,
+      "learning_rate": 1.7761382977955117e-06,
+      "loss": 1.0062,
+      "step": 5581
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4023072719573975,
+      "learning_rate": 1.774516490242772e-06,
+      "loss": 0.7956,
+      "step": 5582
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4612479209899902,
+      "learning_rate": 1.7728952637171488e-06,
+      "loss": 0.8686,
+      "step": 5583
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6086737513542175,
+      "learning_rate": 1.7712746185106888e-06,
+      "loss": 0.9964,
+      "step": 5584
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.5536372661590576,
+      "learning_rate": 1.7696545549153255e-06,
+      "loss": 0.8116,
+      "step": 5585
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7600295543670654,
+      "learning_rate": 1.7680350732228945e-06,
+      "loss": 0.8293,
+      "step": 5586
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4102669954299927,
+      "learning_rate": 1.7664161737251217e-06,
+      "loss": 0.8273,
+      "step": 5587
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.473597764968872,
+      "learning_rate": 1.7647978567136292e-06,
+      "loss": 0.8018,
+      "step": 5588
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3396214246749878,
+      "learning_rate": 1.7631801224799367e-06,
+      "loss": 0.824,
+      "step": 5589
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2479586601257324,
+      "learning_rate": 1.7615629713154548e-06,
+      "loss": 0.8142,
+      "step": 5590
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5977891683578491,
+      "learning_rate": 1.7599464035114944e-06,
+      "loss": 1.0073,
+      "step": 5591
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3251757621765137,
+      "learning_rate": 1.7583304193592566e-06,
+      "loss": 0.7935,
+      "step": 5592
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.244616150856018,
+      "learning_rate": 1.7567150191498372e-06,
+      "loss": 0.7698,
+      "step": 5593
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5940034985542297,
+      "learning_rate": 1.7551002031742325e-06,
+      "loss": 1.0136,
+      "step": 5594
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3034143447875977,
+      "learning_rate": 1.753485971723326e-06,
+      "loss": 0.7931,
+      "step": 5595
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.492506742477417,
+      "learning_rate": 1.7518723250879027e-06,
+      "loss": 0.8167,
+      "step": 5596
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3438462018966675,
+      "learning_rate": 1.750259263558638e-06,
+      "loss": 0.8361,
+      "step": 5597
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5915213823318481,
+      "learning_rate": 1.7486467874261025e-06,
+      "loss": 1.0037,
+      "step": 5598
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.462377667427063,
+      "learning_rate": 1.747034896980761e-06,
+      "loss": 0.7887,
+      "step": 5599
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3660664558410645,
+      "learning_rate": 1.7454235925129764e-06,
+      "loss": 0.8038,
+      "step": 5600
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2605547904968262,
+      "learning_rate": 1.7438128743130018e-06,
+      "loss": 0.863,
+      "step": 5601
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3316082954406738,
+      "learning_rate": 1.7422027426709848e-06,
+      "loss": 0.8006,
+      "step": 5602
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.7046899795532227,
+      "learning_rate": 1.7405931978769719e-06,
+      "loss": 0.8252,
+      "step": 5603
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6090132594108582,
+      "learning_rate": 1.738984240220899e-06,
+      "loss": 1.0299,
+      "step": 5604
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.3240400552749634,
+      "learning_rate": 1.737375869992597e-06,
+      "loss": 0.8069,
+      "step": 5605
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.266689658164978,
+      "learning_rate": 1.7357680874817944e-06,
+      "loss": 0.8148,
+      "step": 5606
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.350319504737854,
+      "learning_rate": 1.7341608929781107e-06,
+      "loss": 0.8691,
+      "step": 5607
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.2574511766433716,
+      "learning_rate": 1.7325542867710576e-06,
+      "loss": 0.8094,
+      "step": 5608
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 1.4173221588134766,
+      "learning_rate": 1.7309482691500473e-06,
+      "loss": 0.7936,
+      "step": 5609
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.265723705291748,
+      "learning_rate": 1.72934284040438e-06,
+      "loss": 0.801,
+      "step": 5610
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3749773502349854,
+      "learning_rate": 1.7277380008232508e-06,
+      "loss": 0.7994,
+      "step": 5611
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.327202558517456,
+      "learning_rate": 1.7261337506957532e-06,
+      "loss": 0.7875,
+      "step": 5612
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6156080961227417,
+      "learning_rate": 1.7245300903108669e-06,
+      "loss": 0.9997,
+      "step": 5613
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5900075435638428,
+      "learning_rate": 1.7229270199574743e-06,
+      "loss": 0.8364,
+      "step": 5614
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6034912467002869,
+      "learning_rate": 1.7213245399243439e-06,
+      "loss": 1.0091,
+      "step": 5615
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2976627349853516,
+      "learning_rate": 1.719722650500139e-06,
+      "loss": 0.8385,
+      "step": 5616
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2722821235656738,
+      "learning_rate": 1.7181213519734225e-06,
+      "loss": 0.789,
+      "step": 5617
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5740172266960144,
+      "learning_rate": 1.7165206446326433e-06,
+      "loss": 1.0171,
+      "step": 5618
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.242266058921814,
+      "learning_rate": 1.7149205287661492e-06,
+      "loss": 0.7879,
+      "step": 5619
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4050815105438232,
+      "learning_rate": 1.713321004662179e-06,
+      "loss": 0.862,
+      "step": 5620
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6005438566207886,
+      "learning_rate": 1.7117220726088628e-06,
+      "loss": 0.9817,
+      "step": 5621
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5958085060119629,
+      "learning_rate": 1.7101237328942305e-06,
+      "loss": 0.9816,
+      "step": 5622
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4601460695266724,
+      "learning_rate": 1.7085259858061977e-06,
+      "loss": 0.8528,
+      "step": 5623
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.332800269126892,
+      "learning_rate": 1.7069288316325799e-06,
+      "loss": 0.8708,
+      "step": 5624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3957278728485107,
+      "learning_rate": 1.705332270661082e-06,
+      "loss": 0.8194,
+      "step": 5625
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.667022943496704,
+      "learning_rate": 1.7037363031793003e-06,
+      "loss": 0.8253,
+      "step": 5626
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.59517902135849,
+      "learning_rate": 1.7021409294747299e-06,
+      "loss": 1.03,
+      "step": 5627
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.340517520904541,
+      "learning_rate": 1.7005461498347532e-06,
+      "loss": 0.8505,
+      "step": 5628
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5802062153816223,
+      "learning_rate": 1.6989519645466507e-06,
+      "loss": 0.9945,
+      "step": 5629
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3621352910995483,
+      "learning_rate": 1.6973583738975925e-06,
+      "loss": 0.8586,
+      "step": 5630
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4260448217391968,
+      "learning_rate": 1.6957653781746397e-06,
+      "loss": 0.8067,
+      "step": 5631
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3204004764556885,
+      "learning_rate": 1.6941729776647524e-06,
+      "loss": 0.884,
+      "step": 5632
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5325778722763062,
+      "learning_rate": 1.6925811726547774e-06,
+      "loss": 0.8491,
+      "step": 5633
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3952596187591553,
+      "learning_rate": 1.6909899634314592e-06,
+      "loss": 0.889,
+      "step": 5634
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5426219701766968,
+      "learning_rate": 1.6893993502814316e-06,
+      "loss": 0.7928,
+      "step": 5635
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3271838426589966,
+      "learning_rate": 1.6878093334912199e-06,
+      "loss": 0.805,
+      "step": 5636
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.28501296043396,
+      "learning_rate": 1.6862199133472473e-06,
+      "loss": 0.8309,
+      "step": 5637
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.349823236465454,
+      "learning_rate": 1.6846310901358248e-06,
+      "loss": 0.8038,
+      "step": 5638
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3955975770950317,
+      "learning_rate": 1.683042864143158e-06,
+      "loss": 0.783,
+      "step": 5639
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6036735773086548,
+      "learning_rate": 1.6814552356553416e-06,
+      "loss": 1.0302,
+      "step": 5640
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3151493072509766,
+      "learning_rate": 1.6798682049583687e-06,
+      "loss": 0.8396,
+      "step": 5641
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3023802042007446,
+      "learning_rate": 1.6782817723381206e-06,
+      "loss": 0.8425,
+      "step": 5642
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.487183928489685,
+      "learning_rate": 1.6766959380803693e-06,
+      "loss": 0.8059,
+      "step": 5643
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6044960618019104,
+      "learning_rate": 1.6751107024707846e-06,
+      "loss": 1.0081,
+      "step": 5644
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3074182271957397,
+      "learning_rate": 1.6735260657949243e-06,
+      "loss": 0.8144,
+      "step": 5645
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.580496609210968,
+      "learning_rate": 1.6719420283382366e-06,
+      "loss": 1.0308,
+      "step": 5646
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.380036473274231,
+      "learning_rate": 1.6703585903860681e-06,
+      "loss": 0.8262,
+      "step": 5647
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3438615798950195,
+      "learning_rate": 1.668775752223652e-06,
+      "loss": 0.8416,
+      "step": 5648
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3598518371582031,
+      "learning_rate": 1.6671935141361134e-06,
+      "loss": 0.7979,
+      "step": 5649
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3248631954193115,
+      "learning_rate": 1.6656118764084745e-06,
+      "loss": 0.8602,
+      "step": 5650
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3201802968978882,
+      "learning_rate": 1.6640308393256427e-06,
+      "loss": 0.7998,
+      "step": 5651
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.328959584236145,
+      "learning_rate": 1.662450403172423e-06,
+      "loss": 0.8541,
+      "step": 5652
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3049330711364746,
+      "learning_rate": 1.6608705682335092e-06,
+      "loss": 0.7937,
+      "step": 5653
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2822679281234741,
+      "learning_rate": 1.6592913347934836e-06,
+      "loss": 0.8291,
+      "step": 5654
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3443666696548462,
+      "learning_rate": 1.657712703136829e-06,
+      "loss": 0.8914,
+      "step": 5655
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3467782735824585,
+      "learning_rate": 1.65613467354791e-06,
+      "loss": 0.7936,
+      "step": 5656
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.6473826169967651,
+      "learning_rate": 1.6545572463109904e-06,
+      "loss": 0.8074,
+      "step": 5657
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.295922040939331,
+      "learning_rate": 1.6529804217102214e-06,
+      "loss": 0.8277,
+      "step": 5658
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4261282682418823,
+      "learning_rate": 1.6514042000296448e-06,
+      "loss": 0.7529,
+      "step": 5659
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.329802393913269,
+      "learning_rate": 1.6498285815531984e-06,
+      "loss": 0.7687,
+      "step": 5660
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4890222549438477,
+      "learning_rate": 1.6482535665647053e-06,
+      "loss": 0.772,
+      "step": 5661
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.333763837814331,
+      "learning_rate": 1.6466791553478872e-06,
+      "loss": 0.8227,
+      "step": 5662
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3378708362579346,
+      "learning_rate": 1.6451053481863498e-06,
+      "loss": 0.8075,
+      "step": 5663
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.361269474029541,
+      "learning_rate": 1.6435321453635934e-06,
+      "loss": 0.7364,
+      "step": 5664
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.599864959716797,
+      "learning_rate": 1.6419595471630107e-06,
+      "loss": 0.8933,
+      "step": 5665
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3634365797042847,
+      "learning_rate": 1.6403875538678827e-06,
+      "loss": 0.8779,
+      "step": 5666
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5132228136062622,
+      "learning_rate": 1.6388161657613816e-06,
+      "loss": 0.8251,
+      "step": 5667
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3536571264266968,
+      "learning_rate": 1.6372453831265744e-06,
+      "loss": 0.928,
+      "step": 5668
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.1478469371795654,
+      "learning_rate": 1.6356752062464137e-06,
+      "loss": 0.8248,
+      "step": 5669
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6157410740852356,
+      "learning_rate": 1.6341056354037482e-06,
+      "loss": 1.0197,
+      "step": 5670
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4891843795776367,
+      "learning_rate": 1.6325366708813135e-06,
+      "loss": 0.8438,
+      "step": 5671
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6082231402397156,
+      "learning_rate": 1.630968312961735e-06,
+      "loss": 1.044,
+      "step": 5672
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3397574424743652,
+      "learning_rate": 1.6294005619275354e-06,
+      "loss": 0.8372,
+      "step": 5673
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.503509521484375,
+      "learning_rate": 1.62783341806112e-06,
+      "loss": 0.824,
+      "step": 5674
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.2896419763565063,
+      "learning_rate": 1.6262668816447918e-06,
+      "loss": 0.8451,
+      "step": 5675
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.5996801853179932,
+      "learning_rate": 1.6247009529607394e-06,
+      "loss": 0.8701,
+      "step": 5676
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5951268076896667,
+      "learning_rate": 1.6231356322910418e-06,
+      "loss": 0.9857,
+      "step": 5677
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.308688998222351,
+      "learning_rate": 1.6215709199176744e-06,
+      "loss": 0.7921,
+      "step": 5678
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3824912309646606,
+      "learning_rate": 1.6200068161224963e-06,
+      "loss": 0.8445,
+      "step": 5679
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6041085720062256,
+      "learning_rate": 1.6184433211872597e-06,
+      "loss": 1.0037,
+      "step": 5680
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.3647016286849976,
+      "learning_rate": 1.6168804353936063e-06,
+      "loss": 0.9008,
+      "step": 5681
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.309659719467163,
+      "learning_rate": 1.615318159023071e-06,
+      "loss": 0.8698,
+      "step": 5682
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4553090333938599,
+      "learning_rate": 1.613756492357076e-06,
+      "loss": 0.7873,
+      "step": 5683
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 1.4146724939346313,
+      "learning_rate": 1.6121954356769314e-06,
+      "loss": 0.7653,
+      "step": 5684
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5907643437385559,
+      "learning_rate": 1.6106349892638451e-06,
+      "loss": 1.0136,
+      "step": 5685
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.436120867729187,
+      "learning_rate": 1.6090751533989084e-06,
+      "loss": 0.8261,
+      "step": 5686
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.460119366645813,
+      "learning_rate": 1.607515928363102e-06,
+      "loss": 0.8472,
+      "step": 5687
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5815925002098083,
+      "learning_rate": 1.6059573144373035e-06,
+      "loss": 1.0031,
+      "step": 5688
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2822871208190918,
+      "learning_rate": 1.6043993119022722e-06,
+      "loss": 0.7895,
+      "step": 5689
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5898298025131226,
+      "learning_rate": 1.6028419210386648e-06,
+      "loss": 0.995,
+      "step": 5690
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3509808778762817,
+      "learning_rate": 1.601285142127022e-06,
+      "loss": 0.9003,
+      "step": 5691
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5870526432991028,
+      "learning_rate": 1.5997289754477758e-06,
+      "loss": 0.9931,
+      "step": 5692
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4222404956817627,
+      "learning_rate": 1.5981734212812506e-06,
+      "loss": 0.7917,
+      "step": 5693
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5830018520355225,
+      "learning_rate": 1.5966184799076567e-06,
+      "loss": 1.006,
+      "step": 5694
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5717180967330933,
+      "learning_rate": 1.5950641516070953e-06,
+      "loss": 1.0292,
+      "step": 5695
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4055287837982178,
+      "learning_rate": 1.59351043665956e-06,
+      "loss": 0.8711,
+      "step": 5696
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.415024995803833,
+      "learning_rate": 1.591957335344928e-06,
+      "loss": 0.8496,
+      "step": 5697
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3756482601165771,
+      "learning_rate": 1.5904048479429734e-06,
+      "loss": 0.7827,
+      "step": 5698
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2995874881744385,
+      "learning_rate": 1.5888529747333537e-06,
+      "loss": 0.8404,
+      "step": 5699
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3021905422210693,
+      "learning_rate": 1.5873017159956167e-06,
+      "loss": 0.7303,
+      "step": 5700
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3376107215881348,
+      "learning_rate": 1.585751072009203e-06,
+      "loss": 0.7968,
+      "step": 5701
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2737462520599365,
+      "learning_rate": 1.584201043053437e-06,
+      "loss": 0.7655,
+      "step": 5702
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3124929666519165,
+      "learning_rate": 1.5826516294075394e-06,
+      "loss": 0.8093,
+      "step": 5703
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3099980354309082,
+      "learning_rate": 1.5811028313506138e-06,
+      "loss": 0.8222,
+      "step": 5704
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5993737578392029,
+      "learning_rate": 1.5795546491616537e-06,
+      "loss": 1.0004,
+      "step": 5705
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.7160159349441528,
+      "learning_rate": 1.5780070831195466e-06,
+      "loss": 0.8406,
+      "step": 5706
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.390631079673767,
+      "learning_rate": 1.5764601335030615e-06,
+      "loss": 0.875,
+      "step": 5707
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2926608324050903,
+      "learning_rate": 1.5749138005908648e-06,
+      "loss": 0.7876,
+      "step": 5708
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6091787815093994,
+      "learning_rate": 1.573368084661505e-06,
+      "loss": 1.0016,
+      "step": 5709
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3286255598068237,
+      "learning_rate": 1.571822985993421e-06,
+      "loss": 0.8283,
+      "step": 5710
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.356781244277954,
+      "learning_rate": 1.5702785048649432e-06,
+      "loss": 0.8406,
+      "step": 5711
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5856958031654358,
+      "learning_rate": 1.568734641554287e-06,
+      "loss": 1.01,
+      "step": 5712
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5992536544799805,
+      "learning_rate": 1.5671913963395612e-06,
+      "loss": 0.9985,
+      "step": 5713
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3475677967071533,
+      "learning_rate": 1.5656487694987593e-06,
+      "loss": 0.9061,
+      "step": 5714
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5809083580970764,
+      "learning_rate": 1.5641067613097621e-06,
+      "loss": 1.0265,
+      "step": 5715
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3727073669433594,
+      "learning_rate": 1.5625653720503458e-06,
+      "loss": 0.7952,
+      "step": 5716
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3388718366622925,
+      "learning_rate": 1.561024601998168e-06,
+      "loss": 0.8587,
+      "step": 5717
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4089347124099731,
+      "learning_rate": 1.5594844514307767e-06,
+      "loss": 0.8512,
+      "step": 5718
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4586037397384644,
+      "learning_rate": 1.557944920625612e-06,
+      "loss": 0.8184,
+      "step": 5719
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.342061996459961,
+      "learning_rate": 1.5564060098599986e-06,
+      "loss": 0.8166,
+      "step": 5720
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4077059030532837,
+      "learning_rate": 1.5548677194111477e-06,
+      "loss": 0.8863,
+      "step": 5721
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.357832670211792,
+      "learning_rate": 1.5533300495561653e-06,
+      "loss": 0.8153,
+      "step": 5722
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3867276906967163,
+      "learning_rate": 1.5517930005720404e-06,
+      "loss": 0.8243,
+      "step": 5723
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3461307287216187,
+      "learning_rate": 1.550256572735651e-06,
+      "loss": 0.8755,
+      "step": 5724
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6420015692710876,
+      "learning_rate": 1.5487207663237625e-06,
+      "loss": 0.9995,
+      "step": 5725
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3797861337661743,
+      "learning_rate": 1.5471855816130328e-06,
+      "loss": 0.8296,
+      "step": 5726
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6168407797813416,
+      "learning_rate": 1.5456510188800023e-06,
+      "loss": 0.9979,
+      "step": 5727
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.44643235206604,
+      "learning_rate": 1.544117078401101e-06,
+      "loss": 0.7879,
+      "step": 5728
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4628093242645264,
+      "learning_rate": 1.5425837604526506e-06,
+      "loss": 0.8107,
+      "step": 5729
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4001802206039429,
+      "learning_rate": 1.541051065310854e-06,
+      "loss": 0.7887,
+      "step": 5730
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6090221405029297,
+      "learning_rate": 1.5395189932518085e-06,
+      "loss": 0.9955,
+      "step": 5731
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3001577854156494,
+      "learning_rate": 1.5379875445514947e-06,
+      "loss": 0.8113,
+      "step": 5732
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3705745935440063,
+      "learning_rate": 1.536456719485781e-06,
+      "loss": 0.8249,
+      "step": 5733
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5845023989677429,
+      "learning_rate": 1.5349265183304273e-06,
+      "loss": 1.0078,
+      "step": 5734
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3567553758621216,
+      "learning_rate": 1.5333969413610766e-06,
+      "loss": 0.8025,
+      "step": 5735
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.804337501525879,
+      "learning_rate": 1.5318679888532633e-06,
+      "loss": 0.7745,
+      "step": 5736
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.302914023399353,
+      "learning_rate": 1.5303396610824061e-06,
+      "loss": 0.8431,
+      "step": 5737
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3705129623413086,
+      "learning_rate": 1.528811958323812e-06,
+      "loss": 0.8066,
+      "step": 5738
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4893993139266968,
+      "learning_rate": 1.527284880852678e-06,
+      "loss": 0.8269,
+      "step": 5739
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3553133010864258,
+      "learning_rate": 1.5257584289440835e-06,
+      "loss": 0.8517,
+      "step": 5740
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6189367175102234,
+      "learning_rate": 1.5242326028730015e-06,
+      "loss": 0.9973,
+      "step": 5741
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6026320457458496,
+      "learning_rate": 1.5227074029142874e-06,
+      "loss": 1.012,
+      "step": 5742
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5995011329650879,
+      "learning_rate": 1.5211828293426833e-06,
+      "loss": 1.0102,
+      "step": 5743
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3245854377746582,
+      "learning_rate": 1.519658882432824e-06,
+      "loss": 0.8482,
+      "step": 5744
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.293561577796936,
+      "learning_rate": 1.5181355624592242e-06,
+      "loss": 0.8196,
+      "step": 5745
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5838345885276794,
+      "learning_rate": 1.5166128696962929e-06,
+      "loss": 1.017,
+      "step": 5746
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4642553329467773,
+      "learning_rate": 1.515090804418321e-06,
+      "loss": 0.7631,
+      "step": 5747
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3740122318267822,
+      "learning_rate": 1.5135693668994866e-06,
+      "loss": 0.8482,
+      "step": 5748
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6061688661575317,
+      "learning_rate": 1.5120485574138583e-06,
+      "loss": 1.0063,
+      "step": 5749
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.4049749374389648,
+      "learning_rate": 1.5105283762353866e-06,
+      "loss": 0.8188,
+      "step": 5750
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.230723261833191,
+      "learning_rate": 1.5090088236379151e-06,
+      "loss": 0.7725,
+      "step": 5751
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.2553048133850098,
+      "learning_rate": 1.5074898998951686e-06,
+      "loss": 0.8246,
+      "step": 5752
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3756577968597412,
+      "learning_rate": 1.5059716052807593e-06,
+      "loss": 0.8251,
+      "step": 5753
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3418726921081543,
+      "learning_rate": 1.5044539400681901e-06,
+      "loss": 0.8233,
+      "step": 5754
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.5075932741165161,
+      "learning_rate": 1.5029369045308466e-06,
+      "loss": 0.8046,
+      "step": 5755
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3551779985427856,
+      "learning_rate": 1.5014204989420005e-06,
+      "loss": 0.813,
+      "step": 5756
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6027165055274963,
+      "learning_rate": 1.4999047235748149e-06,
+      "loss": 0.9851,
+      "step": 5757
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3191052675247192,
+      "learning_rate": 1.4983895787023345e-06,
+      "loss": 0.7408,
+      "step": 5758
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 1.3125860691070557,
+      "learning_rate": 1.496875064597491e-06,
+      "loss": 0.7945,
+      "step": 5759
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5813114643096924,
+      "learning_rate": 1.495361181533106e-06,
+      "loss": 1.0083,
+      "step": 5760
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.592303454875946,
+      "learning_rate": 1.4938479297818837e-06,
+      "loss": 1.0204,
+      "step": 5761
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3009306192398071,
+      "learning_rate": 1.4923353096164144e-06,
+      "loss": 0.8479,
+      "step": 5762
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4317619800567627,
+      "learning_rate": 1.4908233213091793e-06,
+      "loss": 0.8182,
+      "step": 5763
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4305704832077026,
+      "learning_rate": 1.4893119651325405e-06,
+      "loss": 0.831,
+      "step": 5764
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6102326512336731,
+      "learning_rate": 1.4878012413587495e-06,
+      "loss": 0.9995,
+      "step": 5765
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2612279653549194,
+      "learning_rate": 1.4862911502599404e-06,
+      "loss": 0.8339,
+      "step": 5766
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.7905800342559814,
+      "learning_rate": 1.4847816921081382e-06,
+      "loss": 0.842,
+      "step": 5767
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4268862009048462,
+      "learning_rate": 1.4832728671752488e-06,
+      "loss": 0.8355,
+      "step": 5768
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5987620949745178,
+      "learning_rate": 1.48176467573307e-06,
+      "loss": 0.9973,
+      "step": 5769
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5935689210891724,
+      "learning_rate": 1.4802571180532798e-06,
+      "loss": 1.0156,
+      "step": 5770
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3791399002075195,
+      "learning_rate": 1.4787501944074435e-06,
+      "loss": 0.7879,
+      "step": 5771
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6060174703598022,
+      "learning_rate": 1.4772439050670157e-06,
+      "loss": 0.8098,
+      "step": 5772
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5831936001777649,
+      "learning_rate": 1.47573825030333e-06,
+      "loss": 1.0274,
+      "step": 5773
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5732036828994751,
+      "learning_rate": 1.4742332303876144e-06,
+      "loss": 1.0317,
+      "step": 5774
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3538316488265991,
+      "learning_rate": 1.472728845590975e-06,
+      "loss": 0.7587,
+      "step": 5775
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.600723147392273,
+      "learning_rate": 1.4712250961844054e-06,
+      "loss": 0.9949,
+      "step": 5776
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6032732725143433,
+      "learning_rate": 1.4697219824387887e-06,
+      "loss": 1.0123,
+      "step": 5777
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.390986680984497,
+      "learning_rate": 1.4682195046248871e-06,
+      "loss": 0.8602,
+      "step": 5778
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.781944751739502,
+      "learning_rate": 1.466717663013355e-06,
+      "loss": 0.909,
+      "step": 5779
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3581568002700806,
+      "learning_rate": 1.4652164578747274e-06,
+      "loss": 0.8047,
+      "step": 5780
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3372241258621216,
+      "learning_rate": 1.463715889479424e-06,
+      "loss": 0.7702,
+      "step": 5781
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.6417593955993652,
+      "learning_rate": 1.4622159580977546e-06,
+      "loss": 0.8177,
+      "step": 5782
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.527761459350586,
+      "learning_rate": 1.4607166639999114e-06,
+      "loss": 0.8428,
+      "step": 5783
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3010333776474,
+      "learning_rate": 1.4592180074559692e-06,
+      "loss": 0.7534,
+      "step": 5784
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2301329374313354,
+      "learning_rate": 1.4577199887358939e-06,
+      "loss": 0.7683,
+      "step": 5785
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6102602481842041,
+      "learning_rate": 1.4562226081095304e-06,
+      "loss": 1.0206,
+      "step": 5786
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3323019742965698,
+      "learning_rate": 1.4547258658466147e-06,
+      "loss": 0.7828,
+      "step": 5787
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.318537950515747,
+      "learning_rate": 1.453229762216763e-06,
+      "loss": 0.8387,
+      "step": 5788
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.376747965812683,
+      "learning_rate": 1.451734297489476e-06,
+      "loss": 0.8026,
+      "step": 5789
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.8250131607055664,
+      "learning_rate": 1.450239471934145e-06,
+      "loss": 0.8542,
+      "step": 5790
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2969990968704224,
+      "learning_rate": 1.4487452858200395e-06,
+      "loss": 0.8716,
+      "step": 5791
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3020683526992798,
+      "learning_rate": 1.44725173941632e-06,
+      "loss": 0.8775,
+      "step": 5792
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.360983967781067,
+      "learning_rate": 1.445758832992027e-06,
+      "loss": 0.7833,
+      "step": 5793
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6040380597114563,
+      "learning_rate": 1.4442665668160855e-06,
+      "loss": 1.0027,
+      "step": 5794
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.31201171875,
+      "learning_rate": 1.44277494115731e-06,
+      "loss": 0.8752,
+      "step": 5795
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5806965231895447,
+      "learning_rate": 1.4412839562843944e-06,
+      "loss": 1.0138,
+      "step": 5796
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6072664260864258,
+      "learning_rate": 1.4397936124659218e-06,
+      "loss": 0.9943,
+      "step": 5797
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4509943723678589,
+      "learning_rate": 1.438303909970356e-06,
+      "loss": 0.8369,
+      "step": 5798
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5862962603569031,
+      "learning_rate": 1.436814849066047e-06,
+      "loss": 1.0159,
+      "step": 5799
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2879165410995483,
+      "learning_rate": 1.4353264300212273e-06,
+      "loss": 0.8708,
+      "step": 5800
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2987159490585327,
+      "learning_rate": 1.4338386531040176e-06,
+      "loss": 0.8092,
+      "step": 5801
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2520283460617065,
+      "learning_rate": 1.4323515185824204e-06,
+      "loss": 0.7182,
+      "step": 5802
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3438979387283325,
+      "learning_rate": 1.4308650267243213e-06,
+      "loss": 0.8062,
+      "step": 5803
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4909369945526123,
+      "learning_rate": 1.4293791777974936e-06,
+      "loss": 0.8798,
+      "step": 5804
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2969924211502075,
+      "learning_rate": 1.4278939720695923e-06,
+      "loss": 0.8246,
+      "step": 5805
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3283579349517822,
+      "learning_rate": 1.4264094098081554e-06,
+      "loss": 0.7805,
+      "step": 5806
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2746118307113647,
+      "learning_rate": 1.4249254912806092e-06,
+      "loss": 0.7824,
+      "step": 5807
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3317487239837646,
+      "learning_rate": 1.4234422167542604e-06,
+      "loss": 0.8233,
+      "step": 5808
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3832323551177979,
+      "learning_rate": 1.4219595864962993e-06,
+      "loss": 0.7668,
+      "step": 5809
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4416002035140991,
+      "learning_rate": 1.420477600773804e-06,
+      "loss": 0.8706,
+      "step": 5810
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.385515570640564,
+      "learning_rate": 1.4189962598537338e-06,
+      "loss": 0.8392,
+      "step": 5811
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2730971574783325,
+      "learning_rate": 1.4175155640029298e-06,
+      "loss": 0.7985,
+      "step": 5812
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3022186756134033,
+      "learning_rate": 1.4160355134881227e-06,
+      "loss": 0.8304,
+      "step": 5813
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2113975286483765,
+      "learning_rate": 1.4145561085759201e-06,
+      "loss": 0.7912,
+      "step": 5814
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.349530577659607,
+      "learning_rate": 1.41307734953282e-06,
+      "loss": 0.8056,
+      "step": 5815
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2978026866912842,
+      "learning_rate": 1.4115992366251996e-06,
+      "loss": 0.7779,
+      "step": 5816
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2782918214797974,
+      "learning_rate": 1.4101217701193188e-06,
+      "loss": 0.8675,
+      "step": 5817
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3478323221206665,
+      "learning_rate": 1.4086449502813265e-06,
+      "loss": 0.8061,
+      "step": 5818
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3284629583358765,
+      "learning_rate": 1.4071687773772485e-06,
+      "loss": 0.7737,
+      "step": 5819
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6336382627487183,
+      "learning_rate": 1.405693251673001e-06,
+      "loss": 1.0141,
+      "step": 5820
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6116156578063965,
+      "learning_rate": 1.4042183734343779e-06,
+      "loss": 0.9977,
+      "step": 5821
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3799347877502441,
+      "learning_rate": 1.4027441429270566e-06,
+      "loss": 0.8891,
+      "step": 5822
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3028596639633179,
+      "learning_rate": 1.4012705604166038e-06,
+      "loss": 0.8088,
+      "step": 5823
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4613491296768188,
+      "learning_rate": 1.3997976261684615e-06,
+      "loss": 0.8426,
+      "step": 5824
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4435051679611206,
+      "learning_rate": 1.3983253404479625e-06,
+      "loss": 0.7857,
+      "step": 5825
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3417834043502808,
+      "learning_rate": 1.3968537035203173e-06,
+      "loss": 0.8584,
+      "step": 5826
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.400173306465149,
+      "learning_rate": 1.3953827156506194e-06,
+      "loss": 0.8218,
+      "step": 5827
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5918174982070923,
+      "learning_rate": 1.3939123771038515e-06,
+      "loss": 1.0298,
+      "step": 5828
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.2423192262649536,
+      "learning_rate": 1.392442688144871e-06,
+      "loss": 0.8005,
+      "step": 5829
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4185041189193726,
+      "learning_rate": 1.3909736490384258e-06,
+      "loss": 0.7991,
+      "step": 5830
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.3060165643692017,
+      "learning_rate": 1.3895052600491421e-06,
+      "loss": 0.8632,
+      "step": 5831
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.8556575775146484,
+      "learning_rate": 1.3880375214415281e-06,
+      "loss": 0.859,
+      "step": 5832
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.9625102281570435,
+      "learning_rate": 1.3865704334799807e-06,
+      "loss": 0.818,
+      "step": 5833
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.594008207321167,
+      "learning_rate": 1.3851039964287733e-06,
+      "loss": 0.9922,
+      "step": 5834
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.4959546327590942,
+      "learning_rate": 1.3836382105520663e-06,
+      "loss": 0.8562,
+      "step": 5835
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.5129352807998657,
+      "learning_rate": 1.3821730761139008e-06,
+      "loss": 0.7723,
+      "step": 5836
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6053160429000854,
+      "learning_rate": 1.3807085933781988e-06,
+      "loss": 1.0049,
+      "step": 5837
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.339548945426941,
+      "learning_rate": 1.3792447626087702e-06,
+      "loss": 0.8708,
+      "step": 5838
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5982996821403503,
+      "learning_rate": 1.3777815840693026e-06,
+      "loss": 1.0067,
+      "step": 5839
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2993459701538086,
+      "learning_rate": 1.3763190580233687e-06,
+      "loss": 0.7945,
+      "step": 5840
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5410149097442627,
+      "learning_rate": 1.37485718473442e-06,
+      "loss": 0.8757,
+      "step": 5841
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5439146757125854,
+      "learning_rate": 1.3733959644657964e-06,
+      "loss": 0.762,
+      "step": 5842
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3805087804794312,
+      "learning_rate": 1.3719353974807159e-06,
+      "loss": 0.7852,
+      "step": 5843
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8842169046401978,
+      "learning_rate": 1.370475484042278e-06,
+      "loss": 0.8302,
+      "step": 5844
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5950199961662292,
+      "learning_rate": 1.3690162244134697e-06,
+      "loss": 1.0261,
+      "step": 5845
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3064738512039185,
+      "learning_rate": 1.3675576188571548e-06,
+      "loss": 0.8194,
+      "step": 5846
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.266180157661438,
+      "learning_rate": 1.36609966763608e-06,
+      "loss": 0.7969,
+      "step": 5847
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3173664808273315,
+      "learning_rate": 1.3646423710128786e-06,
+      "loss": 0.7785,
+      "step": 5848
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5786139965057373,
+      "learning_rate": 1.363185729250061e-06,
+      "loss": 1.019,
+      "step": 5849
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4550254344940186,
+      "learning_rate": 1.3617297426100208e-06,
+      "loss": 0.8201,
+      "step": 5850
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.60068678855896,
+      "learning_rate": 1.360274411355036e-06,
+      "loss": 1.0058,
+      "step": 5851
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2844910621643066,
+      "learning_rate": 1.3588197357472628e-06,
+      "loss": 0.8034,
+      "step": 5852
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3397997617721558,
+      "learning_rate": 1.3573657160487436e-06,
+      "loss": 0.7813,
+      "step": 5853
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3401869535446167,
+      "learning_rate": 1.3559123525213997e-06,
+      "loss": 0.8398,
+      "step": 5854
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2869932651519775,
+      "learning_rate": 1.354459645427032e-06,
+      "loss": 0.7955,
+      "step": 5855
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3658721446990967,
+      "learning_rate": 1.3530075950273303e-06,
+      "loss": 0.8754,
+      "step": 5856
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6036465167999268,
+      "learning_rate": 1.3515562015838585e-06,
+      "loss": 1.0241,
+      "step": 5857
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.327297568321228,
+      "learning_rate": 1.3501054653580675e-06,
+      "loss": 0.8453,
+      "step": 5858
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5625548362731934,
+      "learning_rate": 1.3486553866112872e-06,
+      "loss": 0.8425,
+      "step": 5859
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.512234091758728,
+      "learning_rate": 1.347205965604728e-06,
+      "loss": 0.7999,
+      "step": 5860
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3563601970672607,
+      "learning_rate": 1.3457572025994864e-06,
+      "loss": 0.8741,
+      "step": 5861
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3279154300689697,
+      "learning_rate": 1.3443090978565344e-06,
+      "loss": 0.8272,
+      "step": 5862
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3852790594100952,
+      "learning_rate": 1.3428616516367316e-06,
+      "loss": 0.8162,
+      "step": 5863
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3672534227371216,
+      "learning_rate": 1.3414148642008134e-06,
+      "loss": 0.7911,
+      "step": 5864
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4050170183181763,
+      "learning_rate": 1.3399687358093987e-06,
+      "loss": 0.8024,
+      "step": 5865
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3350093364715576,
+      "learning_rate": 1.3385232667229904e-06,
+      "loss": 0.7916,
+      "step": 5866
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4185004234313965,
+      "learning_rate": 1.3370784572019663e-06,
+      "loss": 0.8407,
+      "step": 5867
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6063864231109619,
+      "learning_rate": 1.3356343075065937e-06,
+      "loss": 1.0224,
+      "step": 5868
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3074164390563965,
+      "learning_rate": 1.3341908178970141e-06,
+      "loss": 0.811,
+      "step": 5869
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6125591397285461,
+      "learning_rate": 1.3327479886332522e-06,
+      "loss": 1.0177,
+      "step": 5870
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4823209047317505,
+      "learning_rate": 1.3313058199752154e-06,
+      "loss": 0.782,
+      "step": 5871
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4765175580978394,
+      "learning_rate": 1.329864312182691e-06,
+      "loss": 0.8587,
+      "step": 5872
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3580210208892822,
+      "learning_rate": 1.3284234655153443e-06,
+      "loss": 0.7956,
+      "step": 5873
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2858961820602417,
+      "learning_rate": 1.326983280232728e-06,
+      "loss": 0.7934,
+      "step": 5874
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.332913875579834,
+      "learning_rate": 1.3255437565942692e-06,
+      "loss": 0.7984,
+      "step": 5875
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 2.141014814376831,
+      "learning_rate": 1.3241048948592805e-06,
+      "loss": 0.7995,
+      "step": 5876
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2462515830993652,
+      "learning_rate": 1.3226666952869532e-06,
+      "loss": 0.7847,
+      "step": 5877
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5925047993659973,
+      "learning_rate": 1.3212291581363567e-06,
+      "loss": 1.0233,
+      "step": 5878
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3347585201263428,
+      "learning_rate": 1.3197922836664473e-06,
+      "loss": 0.8386,
+      "step": 5879
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3936183452606201,
+      "learning_rate": 1.3183560721360577e-06,
+      "loss": 0.868,
+      "step": 5880
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3190162181854248,
+      "learning_rate": 1.316920523803899e-06,
+      "loss": 0.8563,
+      "step": 5881
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3426345586776733,
+      "learning_rate": 1.3154856389285702e-06,
+      "loss": 0.8862,
+      "step": 5882
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.363959550857544,
+      "learning_rate": 1.3140514177685438e-06,
+      "loss": 0.7813,
+      "step": 5883
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2116637229919434,
+      "learning_rate": 1.3126178605821754e-06,
+      "loss": 0.8007,
+      "step": 5884
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3671859502792358,
+      "learning_rate": 1.3111849676277e-06,
+      "loss": 0.7822,
+      "step": 5885
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3846714496612549,
+      "learning_rate": 1.3097527391632365e-06,
+      "loss": 0.8424,
+      "step": 5886
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.6170319318771362,
+      "learning_rate": 1.3083211754467801e-06,
+      "loss": 0.7586,
+      "step": 5887
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3460659980773926,
+      "learning_rate": 1.3068902767362057e-06,
+      "loss": 0.816,
+      "step": 5888
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5837749242782593,
+      "learning_rate": 1.3054600432892734e-06,
+      "loss": 1.0003,
+      "step": 5889
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.360946536064148,
+      "learning_rate": 1.304030475363618e-06,
+      "loss": 0.8056,
+      "step": 5890
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3833975791931152,
+      "learning_rate": 1.3026015732167586e-06,
+      "loss": 0.7954,
+      "step": 5891
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3008918762207031,
+      "learning_rate": 1.301173337106092e-06,
+      "loss": 0.7858,
+      "step": 5892
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 3.303854465484619,
+      "learning_rate": 1.299745767288894e-06,
+      "loss": 0.8486,
+      "step": 5893
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5302387475967407,
+      "learning_rate": 1.2983188640223248e-06,
+      "loss": 0.7841,
+      "step": 5894
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.27927565574646,
+      "learning_rate": 1.296892627563418e-06,
+      "loss": 0.7933,
+      "step": 5895
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3234707117080688,
+      "learning_rate": 1.2954670581690947e-06,
+      "loss": 0.807,
+      "step": 5896
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4694221019744873,
+      "learning_rate": 1.2940421560961501e-06,
+      "loss": 0.7753,
+      "step": 5897
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3109006881713867,
+      "learning_rate": 1.2926179216012596e-06,
+      "loss": 0.8481,
+      "step": 5898
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3849776983261108,
+      "learning_rate": 1.2911943549409827e-06,
+      "loss": 0.8283,
+      "step": 5899
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3566757440567017,
+      "learning_rate": 1.289771456371754e-06,
+      "loss": 0.821,
+      "step": 5900
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3315826654434204,
+      "learning_rate": 1.2883492261498876e-06,
+      "loss": 0.794,
+      "step": 5901
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.601540446281433,
+      "learning_rate": 1.286927664531582e-06,
+      "loss": 0.7982,
+      "step": 5902
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3049226999282837,
+      "learning_rate": 1.2855067717729103e-06,
+      "loss": 0.8154,
+      "step": 5903
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6097007989883423,
+      "learning_rate": 1.2840865481298288e-06,
+      "loss": 1.0217,
+      "step": 5904
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.7146130800247192,
+      "learning_rate": 1.2826669938581704e-06,
+      "loss": 0.8682,
+      "step": 5905
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3220407962799072,
+      "learning_rate": 1.2812481092136476e-06,
+      "loss": 0.8032,
+      "step": 5906
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.604347825050354,
+      "learning_rate": 1.2798298944518555e-06,
+      "loss": 0.9919,
+      "step": 5907
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.5475537776947021,
+      "learning_rate": 1.278412349828264e-06,
+      "loss": 0.8648,
+      "step": 5908
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3141980171203613,
+      "learning_rate": 1.2769954755982268e-06,
+      "loss": 0.8059,
+      "step": 5909
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3667486906051636,
+      "learning_rate": 1.2755792720169736e-06,
+      "loss": 0.8122,
+      "step": 5910
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3053284883499146,
+      "learning_rate": 1.2741637393396122e-06,
+      "loss": 0.8469,
+      "step": 5911
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.4900946617126465,
+      "learning_rate": 1.2727488778211356e-06,
+      "loss": 0.8148,
+      "step": 5912
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.2995854616165161,
+      "learning_rate": 1.271334687716408e-06,
+      "loss": 0.8239,
+      "step": 5913
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.3796800374984741,
+      "learning_rate": 1.2699211692801794e-06,
+      "loss": 0.8292,
+      "step": 5914
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5869297981262207,
+      "learning_rate": 1.2685083227670757e-06,
+      "loss": 0.9989,
+      "step": 5915
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3272398710250854,
+      "learning_rate": 1.2670961484315996e-06,
+      "loss": 0.7882,
+      "step": 5916
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5086859464645386,
+      "learning_rate": 1.265684646528138e-06,
+      "loss": 0.8574,
+      "step": 5917
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.432112455368042,
+      "learning_rate": 1.2642738173109531e-06,
+      "loss": 0.8305,
+      "step": 5918
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5807535648345947,
+      "learning_rate": 1.262863661034185e-06,
+      "loss": 1.0167,
+      "step": 5919
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5702880024909973,
+      "learning_rate": 1.2614541779518563e-06,
+      "loss": 0.9942,
+      "step": 5920
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3188611268997192,
+      "learning_rate": 1.2600453683178659e-06,
+      "loss": 0.8247,
+      "step": 5921
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.394139051437378,
+      "learning_rate": 1.2586372323859896e-06,
+      "loss": 0.8045,
+      "step": 5922
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3678505420684814,
+      "learning_rate": 1.2572297704098873e-06,
+      "loss": 0.8282,
+      "step": 5923
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3437687158584595,
+      "learning_rate": 1.2558229826430922e-06,
+      "loss": 0.756,
+      "step": 5924
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4134777784347534,
+      "learning_rate": 1.2544168693390191e-06,
+      "loss": 0.8096,
+      "step": 5925
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3730756044387817,
+      "learning_rate": 1.2530114307509577e-06,
+      "loss": 0.7629,
+      "step": 5926
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4106084108352661,
+      "learning_rate": 1.251606667132082e-06,
+      "loss": 0.8786,
+      "step": 5927
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4371874332427979,
+      "learning_rate": 1.25020257873544e-06,
+      "loss": 0.8247,
+      "step": 5928
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.596369206905365,
+      "learning_rate": 1.248799165813957e-06,
+      "loss": 1.0177,
+      "step": 5929
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.383608341217041,
+      "learning_rate": 1.2473964286204426e-06,
+      "loss": 0.8182,
+      "step": 5930
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6164594292640686,
+      "learning_rate": 1.245994367407577e-06,
+      "loss": 1.0161,
+      "step": 5931
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2828876972198486,
+      "learning_rate": 1.2445929824279262e-06,
+      "loss": 0.8144,
+      "step": 5932
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.338342547416687,
+      "learning_rate": 1.2431922739339293e-06,
+      "loss": 0.7848,
+      "step": 5933
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.595687210559845,
+      "learning_rate": 1.2417922421779027e-06,
+      "loss": 1.0182,
+      "step": 5934
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.358257532119751,
+      "learning_rate": 1.240392887412047e-06,
+      "loss": 0.8113,
+      "step": 5935
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2790395021438599,
+      "learning_rate": 1.2389942098884334e-06,
+      "loss": 0.8243,
+      "step": 5936
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6048648357391357,
+      "learning_rate": 1.2375962098590178e-06,
+      "loss": 1.0609,
+      "step": 5937
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3766522407531738,
+      "learning_rate": 1.2361988875756292e-06,
+      "loss": 0.7914,
+      "step": 5938
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3371638059616089,
+      "learning_rate": 1.2348022432899754e-06,
+      "loss": 0.7789,
+      "step": 5939
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3078798055648804,
+      "learning_rate": 1.2334062772536454e-06,
+      "loss": 0.8701,
+      "step": 5940
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3930343389511108,
+      "learning_rate": 1.232010989718101e-06,
+      "loss": 0.8574,
+      "step": 5941
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3504668474197388,
+      "learning_rate": 1.2306163809346867e-06,
+      "loss": 0.843,
+      "step": 5942
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5985310673713684,
+      "learning_rate": 1.2292224511546219e-06,
+      "loss": 1.023,
+      "step": 5943
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5158185958862305,
+      "learning_rate": 1.2278292006290015e-06,
+      "loss": 0.845,
+      "step": 5944
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2773357629776,
+      "learning_rate": 1.2264366296088043e-06,
+      "loss": 0.7592,
+      "step": 5945
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3796989917755127,
+      "learning_rate": 1.2250447383448798e-06,
+      "loss": 0.9085,
+      "step": 5946
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.4006367921829224,
+      "learning_rate": 1.2236535270879617e-06,
+      "loss": 0.863,
+      "step": 5947
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5889180302619934,
+      "learning_rate": 1.2222629960886561e-06,
+      "loss": 1.0132,
+      "step": 5948
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2997688055038452,
+      "learning_rate": 1.2208731455974466e-06,
+      "loss": 0.7849,
+      "step": 5949
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2351133823394775,
+      "learning_rate": 1.2194839758646999e-06,
+      "loss": 0.7585,
+      "step": 5950
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2780998945236206,
+      "learning_rate": 1.2180954871406514e-06,
+      "loss": 0.8034,
+      "step": 5951
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.9527373313903809,
+      "learning_rate": 1.216707679675423e-06,
+      "loss": 0.7752,
+      "step": 5952
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2994167804718018,
+      "learning_rate": 1.2153205537190077e-06,
+      "loss": 0.8468,
+      "step": 5953
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2525875568389893,
+      "learning_rate": 1.2139341095212754e-06,
+      "loss": 0.7661,
+      "step": 5954
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.303980827331543,
+      "learning_rate": 1.2125483473319782e-06,
+      "loss": 0.8269,
+      "step": 5955
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.326729655265808,
+      "learning_rate": 1.2111632674007412e-06,
+      "loss": 0.7739,
+      "step": 5956
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.339097261428833,
+      "learning_rate": 1.2097788699770658e-06,
+      "loss": 0.8221,
+      "step": 5957
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2846782207489014,
+      "learning_rate": 1.2083951553103362e-06,
+      "loss": 0.779,
+      "step": 5958
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3755793571472168,
+      "learning_rate": 1.2070121236498078e-06,
+      "loss": 0.8654,
+      "step": 5959
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 5.980408668518066,
+      "learning_rate": 1.2056297752446139e-06,
+      "loss": 0.8266,
+      "step": 5960
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.242810845375061,
+      "learning_rate": 1.2042481103437686e-06,
+      "loss": 0.7957,
+      "step": 5961
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2786715030670166,
+      "learning_rate": 1.202867129196158e-06,
+      "loss": 0.8458,
+      "step": 5962
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2768373489379883,
+      "learning_rate": 1.2014868320505468e-06,
+      "loss": 0.8176,
+      "step": 5963
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.403856635093689,
+      "learning_rate": 1.200107219155579e-06,
+      "loss": 0.8157,
+      "step": 5964
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5513718128204346,
+      "learning_rate": 1.198728290759772e-06,
+      "loss": 0.8435,
+      "step": 5965
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5978437066078186,
+      "learning_rate": 1.19735004711152e-06,
+      "loss": 1.0192,
+      "step": 5966
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3326002359390259,
+      "learning_rate": 1.1959724884590957e-06,
+      "loss": 0.7644,
+      "step": 5967
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2871556282043457,
+      "learning_rate": 1.1945956150506482e-06,
+      "loss": 0.7856,
+      "step": 5968
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3769466876983643,
+      "learning_rate": 1.1932194271342006e-06,
+      "loss": 0.8171,
+      "step": 5969
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3226675987243652,
+      "learning_rate": 1.191843924957658e-06,
+      "loss": 0.8781,
+      "step": 5970
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3040982484817505,
+      "learning_rate": 1.1904691087687964e-06,
+      "loss": 0.831,
+      "step": 5971
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.212841272354126,
+      "learning_rate": 1.189094978815269e-06,
+      "loss": 0.7964,
+      "step": 5972
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3154176473617554,
+      "learning_rate": 1.1877215353446092e-06,
+      "loss": 0.8241,
+      "step": 5973
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3073766231536865,
+      "learning_rate": 1.1863487786042215e-06,
+      "loss": 0.7929,
+      "step": 5974
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6204517483711243,
+      "learning_rate": 1.1849767088413926e-06,
+      "loss": 0.9766,
+      "step": 5975
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3649221658706665,
+      "learning_rate": 1.1836053263032804e-06,
+      "loss": 0.8248,
+      "step": 5976
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2739917039871216,
+      "learning_rate": 1.1822346312369198e-06,
+      "loss": 0.7901,
+      "step": 5977
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5999225378036499,
+      "learning_rate": 1.1808646238892257e-06,
+      "loss": 1.0235,
+      "step": 5978
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.5940029621124268,
+      "learning_rate": 1.1794953045069835e-06,
+      "loss": 0.7954,
+      "step": 5979
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5754348039627075,
+      "learning_rate": 1.1781266733368602e-06,
+      "loss": 1.0215,
+      "step": 5980
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2353698015213013,
+      "learning_rate": 1.1767587306253952e-06,
+      "loss": 0.787,
+      "step": 5981
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3652809858322144,
+      "learning_rate": 1.1753914766190028e-06,
+      "loss": 0.842,
+      "step": 5982
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5944556593894958,
+      "learning_rate": 1.174024911563978e-06,
+      "loss": 1.0217,
+      "step": 5983
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.3584318161010742,
+      "learning_rate": 1.172659035706487e-06,
+      "loss": 0.8277,
+      "step": 5984
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6046809554100037,
+      "learning_rate": 1.1712938492925757e-06,
+      "loss": 0.9867,
+      "step": 5985
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.381564974784851,
+      "learning_rate": 1.1699293525681632e-06,
+      "loss": 0.8125,
+      "step": 5986
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5982908606529236,
+      "learning_rate": 1.1685655457790435e-06,
+      "loss": 1.0124,
+      "step": 5987
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.303046703338623,
+      "learning_rate": 1.1672024291708906e-06,
+      "loss": 0.7971,
+      "step": 5988
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.316156268119812,
+      "learning_rate": 1.1658400029892502e-06,
+      "loss": 0.814,
+      "step": 5989
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2501730918884277,
+      "learning_rate": 1.1644782674795435e-06,
+      "loss": 0.7751,
+      "step": 5990
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3174421787261963,
+      "learning_rate": 1.1631172228870706e-06,
+      "loss": 0.854,
+      "step": 5991
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5789328813552856,
+      "learning_rate": 1.161756869457004e-06,
+      "loss": 1.0033,
+      "step": 5992
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5768986940383911,
+      "learning_rate": 1.160397207434395e-06,
+      "loss": 1.026,
+      "step": 5993
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3228554725646973,
+      "learning_rate": 1.1590382370641663e-06,
+      "loss": 0.7945,
+      "step": 5994
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.311160683631897,
+      "learning_rate": 1.1576799585911175e-06,
+      "loss": 0.8862,
+      "step": 5995
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.352142095565796,
+      "learning_rate": 1.1563223722599266e-06,
+      "loss": 0.8123,
+      "step": 5996
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4348658323287964,
+      "learning_rate": 1.154965478315141e-06,
+      "loss": 0.8127,
+      "step": 5997
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2782857418060303,
+      "learning_rate": 1.1536092770011898e-06,
+      "loss": 0.8311,
+      "step": 5998
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.312182068824768,
+      "learning_rate": 1.1522537685623735e-06,
+      "loss": 0.8007,
+      "step": 5999
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3165432214736938,
+      "learning_rate": 1.1508989532428683e-06,
+      "loss": 0.8235,
+      "step": 6000
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.299038290977478,
+      "learning_rate": 1.1495448312867235e-06,
+      "loss": 0.7921,
+      "step": 6001
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 3.8374814987182617,
+      "learning_rate": 1.1481914029378698e-06,
+      "loss": 0.8446,
+      "step": 6002
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.548584222793579,
+      "learning_rate": 1.1468386684401074e-06,
+      "loss": 0.8171,
+      "step": 6003
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5925439596176147,
+      "learning_rate": 1.1454866280371107e-06,
+      "loss": 0.812,
+      "step": 6004
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2500832080841064,
+      "learning_rate": 1.1441352819724355e-06,
+      "loss": 0.7928,
+      "step": 6005
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.312294602394104,
+      "learning_rate": 1.1427846304895063e-06,
+      "loss": 0.8246,
+      "step": 6006
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2796413898468018,
+      "learning_rate": 1.1414346738316234e-06,
+      "loss": 0.8331,
+      "step": 6007
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.6095136404037476,
+      "learning_rate": 1.1400854122419659e-06,
+      "loss": 0.7684,
+      "step": 6008
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2948733568191528,
+      "learning_rate": 1.138736845963584e-06,
+      "loss": 0.8443,
+      "step": 6009
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3759034872055054,
+      "learning_rate": 1.1373889752394018e-06,
+      "loss": 0.8032,
+      "step": 6010
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6157719492912292,
+      "learning_rate": 1.1360418003122231e-06,
+      "loss": 1.0295,
+      "step": 6011
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.343949794769287,
+      "learning_rate": 1.1346953214247197e-06,
+      "loss": 0.8401,
+      "step": 6012
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2912259101867676,
+      "learning_rate": 1.1333495388194454e-06,
+      "loss": 0.7622,
+      "step": 6013
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5946763753890991,
+      "learning_rate": 1.1320044527388223e-06,
+      "loss": 1.0016,
+      "step": 6014
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4409420490264893,
+      "learning_rate": 1.1306600634251486e-06,
+      "loss": 0.8428,
+      "step": 6015
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5697817206382751,
+      "learning_rate": 1.1293163711206e-06,
+      "loss": 0.9747,
+      "step": 6016
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5868133902549744,
+      "learning_rate": 1.1279733760672235e-06,
+      "loss": 1.0119,
+      "step": 6017
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.301956295967102,
+      "learning_rate": 1.1266310785069406e-06,
+      "loss": 0.7668,
+      "step": 6018
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4601460695266724,
+      "learning_rate": 1.1252894786815494e-06,
+      "loss": 0.8191,
+      "step": 6019
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3425614833831787,
+      "learning_rate": 1.1239485768327195e-06,
+      "loss": 0.7761,
+      "step": 6020
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5789660215377808,
+      "learning_rate": 1.1226083732019981e-06,
+      "loss": 1.0281,
+      "step": 6021
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.499480128288269,
+      "learning_rate": 1.1212688680308042e-06,
+      "loss": 0.8374,
+      "step": 6022
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5835797786712646,
+      "learning_rate": 1.1199300615604291e-06,
+      "loss": 1.0036,
+      "step": 6023
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.5365897417068481,
+      "learning_rate": 1.118591954032044e-06,
+      "loss": 0.8537,
+      "step": 6024
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3417580127716064,
+      "learning_rate": 1.1172545456866873e-06,
+      "loss": 0.8758,
+      "step": 6025
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3555368185043335,
+      "learning_rate": 1.115917836765279e-06,
+      "loss": 0.8298,
+      "step": 6026
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 2.479642391204834,
+      "learning_rate": 1.1145818275086068e-06,
+      "loss": 0.8105,
+      "step": 6027
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5825153589248657,
+      "learning_rate": 1.1132465181573332e-06,
+      "loss": 1.0296,
+      "step": 6028
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3221763372421265,
+      "learning_rate": 1.1119119089519997e-06,
+      "loss": 0.8635,
+      "step": 6029
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3680299520492554,
+      "learning_rate": 1.1105780001330147e-06,
+      "loss": 0.8054,
+      "step": 6030
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3455325365066528,
+      "learning_rate": 1.1092447919406662e-06,
+      "loss": 0.8432,
+      "step": 6031
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4549925327301025,
+      "learning_rate": 1.107912284615113e-06,
+      "loss": 0.8431,
+      "step": 6032
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5703139305114746,
+      "learning_rate": 1.1065804783963856e-06,
+      "loss": 0.9856,
+      "step": 6033
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2826157808303833,
+      "learning_rate": 1.1052493735243952e-06,
+      "loss": 0.8169,
+      "step": 6034
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4643388986587524,
+      "learning_rate": 1.1039189702389187e-06,
+      "loss": 0.782,
+      "step": 6035
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5757795572280884,
+      "learning_rate": 1.102589268779613e-06,
+      "loss": 1.0059,
+      "step": 6036
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2792742252349854,
+      "learning_rate": 1.1012602693860047e-06,
+      "loss": 0.8124,
+      "step": 6037
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2924542427062988,
+      "learning_rate": 1.0999319722974928e-06,
+      "loss": 0.8264,
+      "step": 6038
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2779085636138916,
+      "learning_rate": 1.0986043777533556e-06,
+      "loss": 0.7971,
+      "step": 6039
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3654447793960571,
+      "learning_rate": 1.0972774859927398e-06,
+      "loss": 0.8747,
+      "step": 6040
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5938294529914856,
+      "learning_rate": 1.095951297254665e-06,
+      "loss": 0.9926,
+      "step": 6041
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5946134328842163,
+      "learning_rate": 1.0946258117780294e-06,
+      "loss": 1.0364,
+      "step": 6042
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3384815454483032,
+      "learning_rate": 1.0933010298016e-06,
+      "loss": 0.8344,
+      "step": 6043
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3571865558624268,
+      "learning_rate": 1.0919769515640183e-06,
+      "loss": 0.8346,
+      "step": 6044
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5888329744338989,
+      "learning_rate": 1.0906535773037969e-06,
+      "loss": 1.0137,
+      "step": 6045
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5964393615722656,
+      "learning_rate": 1.0893309072593273e-06,
+      "loss": 1.0127,
+      "step": 6046
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5952311158180237,
+      "learning_rate": 1.0880089416688682e-06,
+      "loss": 1.0208,
+      "step": 6047
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3381268978118896,
+      "learning_rate": 1.086687680770554e-06,
+      "loss": 0.8141,
+      "step": 6048
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2436127662658691,
+      "learning_rate": 1.0853671248023934e-06,
+      "loss": 0.8009,
+      "step": 6049
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3080973625183105,
+      "learning_rate": 1.0840472740022657e-06,
+      "loss": 0.8684,
+      "step": 6050
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2822222709655762,
+      "learning_rate": 1.082728128607922e-06,
+      "loss": 0.7534,
+      "step": 6051
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5830941200256348,
+      "learning_rate": 1.0814096888569931e-06,
+      "loss": 0.99,
+      "step": 6052
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5673322081565857,
+      "learning_rate": 1.0800919549869737e-06,
+      "loss": 0.9951,
+      "step": 6053
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3665486574172974,
+      "learning_rate": 1.0787749272352381e-06,
+      "loss": 0.7753,
+      "step": 6054
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5742855668067932,
+      "learning_rate": 1.0774586058390314e-06,
+      "loss": 1.0092,
+      "step": 6055
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5778371691703796,
+      "learning_rate": 1.0761429910354682e-06,
+      "loss": 0.9971,
+      "step": 6056
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3542217016220093,
+      "learning_rate": 1.074828083061542e-06,
+      "loss": 0.8268,
+      "step": 6057
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5935133099555969,
+      "learning_rate": 1.0735138821541125e-06,
+      "loss": 1.0078,
+      "step": 6058
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2298269271850586,
+      "learning_rate": 1.0722003885499184e-06,
+      "loss": 0.7874,
+      "step": 6059
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3464362621307373,
+      "learning_rate": 1.0708876024855664e-06,
+      "loss": 0.8017,
+      "step": 6060
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6013810038566589,
+      "learning_rate": 1.0695755241975347e-06,
+      "loss": 1.0395,
+      "step": 6061
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.4284489154815674,
+      "learning_rate": 1.0682641539221805e-06,
+      "loss": 0.8215,
+      "step": 6062
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6000249981880188,
+      "learning_rate": 1.0669534918957258e-06,
+      "loss": 0.9833,
+      "step": 6063
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3104057312011719,
+      "learning_rate": 1.0656435383542724e-06,
+      "loss": 0.7904,
+      "step": 6064
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.58807373046875,
+      "learning_rate": 1.0643342935337876e-06,
+      "loss": 1.0083,
+      "step": 6065
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.3006675243377686,
+      "learning_rate": 1.0630257576701142e-06,
+      "loss": 0.8447,
+      "step": 6066
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 1.2369314432144165,
+      "learning_rate": 1.0617179309989688e-06,
+      "loss": 0.8849,
+      "step": 6067
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4004496335983276,
+      "learning_rate": 1.060410813755937e-06,
+      "loss": 0.8397,
+      "step": 6068
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3782739639282227,
+      "learning_rate": 1.0591044061764804e-06,
+      "loss": 0.8017,
+      "step": 6069
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5800756216049194,
+      "learning_rate": 1.0577987084959296e-06,
+      "loss": 1.0393,
+      "step": 6070
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5772849917411804,
+      "learning_rate": 1.0564937209494863e-06,
+      "loss": 0.9763,
+      "step": 6071
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3167890310287476,
+      "learning_rate": 1.0551894437722294e-06,
+      "loss": 0.8337,
+      "step": 6072
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5847127437591553,
+      "learning_rate": 1.0538858771991056e-06,
+      "loss": 0.9841,
+      "step": 6073
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.26302969455719,
+      "learning_rate": 1.052583021464933e-06,
+      "loss": 0.7713,
+      "step": 6074
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3494009971618652,
+      "learning_rate": 1.0512808768044063e-06,
+      "loss": 0.8155,
+      "step": 6075
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5817060470581055,
+      "learning_rate": 1.0499794434520867e-06,
+      "loss": 1.0191,
+      "step": 6076
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2420744895935059,
+      "learning_rate": 1.048678721642412e-06,
+      "loss": 0.7889,
+      "step": 6077
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5919626951217651,
+      "learning_rate": 1.0473787116096885e-06,
+      "loss": 0.7972,
+      "step": 6078
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5794653296470642,
+      "learning_rate": 1.0460794135880941e-06,
+      "loss": 0.9857,
+      "step": 6079
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4520626068115234,
+      "learning_rate": 1.0447808278116828e-06,
+      "loss": 0.816,
+      "step": 6080
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5771816372871399,
+      "learning_rate": 1.0434829545143748e-06,
+      "loss": 0.9985,
+      "step": 6081
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3638437986373901,
+      "learning_rate": 1.042185793929964e-06,
+      "loss": 0.7872,
+      "step": 6082
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5966333746910095,
+      "learning_rate": 1.0408893462921187e-06,
+      "loss": 1.0235,
+      "step": 6083
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.363856554031372,
+      "learning_rate": 1.0395936118343747e-06,
+      "loss": 0.8388,
+      "step": 6084
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4260228872299194,
+      "learning_rate": 1.0382985907901416e-06,
+      "loss": 0.8613,
+      "step": 6085
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2993106842041016,
+      "learning_rate": 1.037004283392698e-06,
+      "loss": 0.8436,
+      "step": 6086
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3976725339889526,
+      "learning_rate": 1.035710689875199e-06,
+      "loss": 0.8267,
+      "step": 6087
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.686564326286316,
+      "learning_rate": 1.034417810470666e-06,
+      "loss": 0.8285,
+      "step": 6088
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3440101146697998,
+      "learning_rate": 1.0331256454119927e-06,
+      "loss": 0.7886,
+      "step": 6089
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.36573326587677,
+      "learning_rate": 1.0318341949319478e-06,
+      "loss": 0.8554,
+      "step": 6090
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3654495477676392,
+      "learning_rate": 1.0305434592631658e-06,
+      "loss": 0.8035,
+      "step": 6091
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.427625298500061,
+      "learning_rate": 1.0292534386381581e-06,
+      "loss": 0.8401,
+      "step": 6092
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.265617847442627,
+      "learning_rate": 1.0279641332893032e-06,
+      "loss": 0.8286,
+      "step": 6093
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5982333421707153,
+      "learning_rate": 1.0266755434488501e-06,
+      "loss": 0.9999,
+      "step": 6094
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5717100501060486,
+      "learning_rate": 1.0253876693489234e-06,
+      "loss": 1.0342,
+      "step": 6095
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2664189338684082,
+      "learning_rate": 1.024100511221514e-06,
+      "loss": 0.8026,
+      "step": 6096
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5859978795051575,
+      "learning_rate": 1.022814069298489e-06,
+      "loss": 1.0065,
+      "step": 6097
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2678407430648804,
+      "learning_rate": 1.0215283438115813e-06,
+      "loss": 0.8183,
+      "step": 6098
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5760886669158936,
+      "learning_rate": 1.0202433349923957e-06,
+      "loss": 1.0188,
+      "step": 6099
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3500229120254517,
+      "learning_rate": 1.0189590430724123e-06,
+      "loss": 0.8288,
+      "step": 6100
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3371299505233765,
+      "learning_rate": 1.0176754682829759e-06,
+      "loss": 0.8521,
+      "step": 6101
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3993351459503174,
+      "learning_rate": 1.016392610855308e-06,
+      "loss": 0.8487,
+      "step": 6102
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3209290504455566,
+      "learning_rate": 1.0151104710204957e-06,
+      "loss": 0.8661,
+      "step": 6103
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5085529088974,
+      "learning_rate": 1.0138290490094992e-06,
+      "loss": 0.8357,
+      "step": 6104
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3051633834838867,
+      "learning_rate": 1.0125483450531509e-06,
+      "loss": 0.799,
+      "step": 6105
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.290930151939392,
+      "learning_rate": 1.011268359382151e-06,
+      "loss": 0.8477,
+      "step": 6106
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2717466354370117,
+      "learning_rate": 1.009989092227071e-06,
+      "loss": 0.7939,
+      "step": 6107
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2771821022033691,
+      "learning_rate": 1.0087105438183554e-06,
+      "loss": 0.8223,
+      "step": 6108
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4081933498382568,
+      "learning_rate": 1.007432714386315e-06,
+      "loss": 0.828,
+      "step": 6109
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2683502435684204,
+      "learning_rate": 1.0061556041611358e-06,
+      "loss": 0.7449,
+      "step": 6110
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3433830738067627,
+      "learning_rate": 1.0048792133728713e-06,
+      "loss": 0.7436,
+      "step": 6111
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3133749961853027,
+      "learning_rate": 1.003603542251444e-06,
+      "loss": 0.8144,
+      "step": 6112
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5898380279541016,
+      "learning_rate": 1.0023285910266517e-06,
+      "loss": 0.8183,
+      "step": 6113
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3059399127960205,
+      "learning_rate": 1.0010543599281569e-06,
+      "loss": 0.8036,
+      "step": 6114
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.274120569229126,
+      "learning_rate": 9.997808491854977e-07,
+      "loss": 0.7584,
+      "step": 6115
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4653551578521729,
+      "learning_rate": 9.985080590280783e-07,
+      "loss": 0.7988,
+      "step": 6116
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6270754337310791,
+      "learning_rate": 9.972359896851736e-07,
+      "loss": 1.0176,
+      "step": 6117
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2802248001098633,
+      "learning_rate": 9.95964641385932e-07,
+      "loss": 0.8046,
+      "step": 6118
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3072408437728882,
+      "learning_rate": 9.94694014359368e-07,
+      "loss": 0.8568,
+      "step": 6119
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.31790030002594,
+      "learning_rate": 9.934241088343671e-07,
+      "loss": 0.8214,
+      "step": 6120
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3104733228683472,
+      "learning_rate": 9.921549250396884e-07,
+      "loss": 0.7981,
+      "step": 6121
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6149839758872986,
+      "learning_rate": 9.90886463203956e-07,
+      "loss": 0.9942,
+      "step": 6122
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6079933643341064,
+      "learning_rate": 9.896187235556653e-07,
+      "loss": 1.0193,
+      "step": 6123
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5133967399597168,
+      "learning_rate": 9.88351706323185e-07,
+      "loss": 0.7927,
+      "step": 6124
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.590746283531189,
+      "learning_rate": 9.8708541173475e-07,
+      "loss": 1.011,
+      "step": 6125
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5870225429534912,
+      "learning_rate": 9.85819840018466e-07,
+      "loss": 0.9966,
+      "step": 6126
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5736048221588135,
+      "learning_rate": 9.845549914023067e-07,
+      "loss": 1.0032,
+      "step": 6127
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.385378360748291,
+      "learning_rate": 9.832908661141204e-07,
+      "loss": 0.8228,
+      "step": 6128
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5991426706314087,
+      "learning_rate": 9.820274643816197e-07,
+      "loss": 1.0142,
+      "step": 6129
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3813490867614746,
+      "learning_rate": 9.80764786432392e-07,
+      "loss": 0.7894,
+      "step": 6130
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.506265640258789,
+      "learning_rate": 9.795028324938899e-07,
+      "loss": 0.7686,
+      "step": 6131
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4224400520324707,
+      "learning_rate": 9.78241602793436e-07,
+      "loss": 0.8303,
+      "step": 6132
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.454015851020813,
+      "learning_rate": 9.769810975582268e-07,
+      "loss": 0.8921,
+      "step": 6133
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2824560403823853,
+      "learning_rate": 9.757213170153234e-07,
+      "loss": 0.7627,
+      "step": 6134
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6319940090179443,
+      "learning_rate": 9.744622613916572e-07,
+      "loss": 0.8153,
+      "step": 6135
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.5158947706222534,
+      "learning_rate": 9.73203930914033e-07,
+      "loss": 0.8324,
+      "step": 6136
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.4258753061294556,
+      "learning_rate": 9.719463258091182e-07,
+      "loss": 0.7672,
+      "step": 6137
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5811218619346619,
+      "learning_rate": 9.706894463034567e-07,
+      "loss": 1.0055,
+      "step": 6138
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3977712392807007,
+      "learning_rate": 9.694332926234567e-07,
+      "loss": 0.8885,
+      "step": 6139
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.363815188407898,
+      "learning_rate": 9.681778649953966e-07,
+      "loss": 0.8571,
+      "step": 6140
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.9657636880874634,
+      "learning_rate": 9.66923163645427e-07,
+      "loss": 0.7785,
+      "step": 6141
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.2675378322601318,
+      "learning_rate": 9.656691887995618e-07,
+      "loss": 0.8355,
+      "step": 6142
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.3264743089675903,
+      "learning_rate": 9.644159406836912e-07,
+      "loss": 0.851,
+      "step": 6143
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5859867930412292,
+      "learning_rate": 9.631634195235696e-07,
+      "loss": 1.0102,
+      "step": 6144
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.360632300376892,
+      "learning_rate": 9.619116255448196e-07,
+      "loss": 0.8652,
+      "step": 6145
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4244492053985596,
+      "learning_rate": 9.606605589729385e-07,
+      "loss": 0.8116,
+      "step": 6146
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.7090641260147095,
+      "learning_rate": 9.594102200332855e-07,
+      "loss": 0.843,
+      "step": 6147
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4951283931732178,
+      "learning_rate": 9.581606089510952e-07,
+      "loss": 0.7722,
+      "step": 6148
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2893919944763184,
+      "learning_rate": 9.569117259514666e-07,
+      "loss": 0.8176,
+      "step": 6149
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5806776881217957,
+      "learning_rate": 9.556635712593682e-07,
+      "loss": 1.0074,
+      "step": 6150
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3670458793640137,
+      "learning_rate": 9.544161450996402e-07,
+      "loss": 0.792,
+      "step": 6151
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2878901958465576,
+      "learning_rate": 9.531694476969872e-07,
+      "loss": 0.7975,
+      "step": 6152
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.380733847618103,
+      "learning_rate": 9.51923479275988e-07,
+      "loss": 0.8543,
+      "step": 6153
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3031367063522339,
+      "learning_rate": 9.506782400610842e-07,
+      "loss": 0.8633,
+      "step": 6154
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3618296384811401,
+      "learning_rate": 9.494337302765883e-07,
+      "loss": 0.8046,
+      "step": 6155
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5870642066001892,
+      "learning_rate": 9.481899501466846e-07,
+      "loss": 1.0075,
+      "step": 6156
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3888226747512817,
+      "learning_rate": 9.469468998954206e-07,
+      "loss": 0.8338,
+      "step": 6157
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.571599006652832,
+      "learning_rate": 9.457045797467173e-07,
+      "loss": 1.0037,
+      "step": 6158
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3069454431533813,
+      "learning_rate": 9.444629899243607e-07,
+      "loss": 0.7761,
+      "step": 6159
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.28342604637146,
+      "learning_rate": 9.43222130652005e-07,
+      "loss": 0.8187,
+      "step": 6160
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3682535886764526,
+      "learning_rate": 9.419820021531767e-07,
+      "loss": 0.8314,
+      "step": 6161
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3524099588394165,
+      "learning_rate": 9.407426046512669e-07,
+      "loss": 0.8201,
+      "step": 6162
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.391230583190918,
+      "learning_rate": 9.395039383695359e-07,
+      "loss": 0.7864,
+      "step": 6163
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2671313285827637,
+      "learning_rate": 9.382660035311125e-07,
+      "loss": 0.8309,
+      "step": 6164
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3553391695022583,
+      "learning_rate": 9.37028800358995e-07,
+      "loss": 0.7594,
+      "step": 6165
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5911656618118286,
+      "learning_rate": 9.357923290760484e-07,
+      "loss": 1.0068,
+      "step": 6166
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.616115391254425,
+      "learning_rate": 9.345565899050058e-07,
+      "loss": 1.0183,
+      "step": 6167
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4684375524520874,
+      "learning_rate": 9.333215830684678e-07,
+      "loss": 0.837,
+      "step": 6168
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3134355545043945,
+      "learning_rate": 9.320873087889071e-07,
+      "loss": 0.8271,
+      "step": 6169
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3466726541519165,
+      "learning_rate": 9.308537672886581e-07,
+      "loss": 0.8268,
+      "step": 6170
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3396931886672974,
+      "learning_rate": 9.296209587899291e-07,
+      "loss": 0.8106,
+      "step": 6171
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3510876893997192,
+      "learning_rate": 9.283888835147936e-07,
+      "loss": 0.7699,
+      "step": 6172
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.34946608543396,
+      "learning_rate": 9.271575416851908e-07,
+      "loss": 0.7886,
+      "step": 6173
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5683494210243225,
+      "learning_rate": 9.259269335229338e-07,
+      "loss": 0.9892,
+      "step": 6174
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6285390853881836,
+      "learning_rate": 9.246970592496968e-07,
+      "loss": 0.8397,
+      "step": 6175
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3199149370193481,
+      "learning_rate": 9.234679190870277e-07,
+      "loss": 0.8471,
+      "step": 6176
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5964332818984985,
+      "learning_rate": 9.222395132563378e-07,
+      "loss": 1.0157,
+      "step": 6177
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3382858037948608,
+      "learning_rate": 9.210118419789071e-07,
+      "loss": 0.8346,
+      "step": 6178
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3728625774383545,
+      "learning_rate": 9.197849054758861e-07,
+      "loss": 0.8726,
+      "step": 6179
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2452197074890137,
+      "learning_rate": 9.185587039682886e-07,
+      "loss": 0.8085,
+      "step": 6180
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.339928150177002,
+      "learning_rate": 9.173332376770006e-07,
+      "loss": 0.8286,
+      "step": 6181
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3761038780212402,
+      "learning_rate": 9.16108506822771e-07,
+      "loss": 0.8588,
+      "step": 6182
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.540544033050537,
+      "learning_rate": 9.148845116262184e-07,
+      "loss": 0.8039,
+      "step": 6183
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5890161395072937,
+      "learning_rate": 9.136612523078314e-07,
+      "loss": 0.9992,
+      "step": 6184
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3558428287506104,
+      "learning_rate": 9.124387290879605e-07,
+      "loss": 0.8334,
+      "step": 6185
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2946540117263794,
+      "learning_rate": 9.112169421868289e-07,
+      "loss": 0.7978,
+      "step": 6186
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4493640661239624,
+      "learning_rate": 9.099958918245249e-07,
+      "loss": 0.8718,
+      "step": 6187
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3711919784545898,
+      "learning_rate": 9.087755782210017e-07,
+      "loss": 0.8759,
+      "step": 6188
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3350707292556763,
+      "learning_rate": 9.075560015960855e-07,
+      "loss": 0.7926,
+      "step": 6189
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.4299321174621582,
+      "learning_rate": 9.06337162169465e-07,
+      "loss": 0.7624,
+      "step": 6190
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.564097285270691,
+      "learning_rate": 9.051190601606968e-07,
+      "loss": 0.755,
+      "step": 6191
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5246400833129883,
+      "learning_rate": 9.039016957892071e-07,
+      "loss": 0.8264,
+      "step": 6192
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3788528442382812,
+      "learning_rate": 9.026850692742856e-07,
+      "loss": 0.8112,
+      "step": 6193
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3003071546554565,
+      "learning_rate": 9.014691808350934e-07,
+      "loss": 0.8213,
+      "step": 6194
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5273720026016235,
+      "learning_rate": 9.002540306906549e-07,
+      "loss": 0.8746,
+      "step": 6195
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5933581590652466,
+      "learning_rate": 8.990396190598622e-07,
+      "loss": 0.7631,
+      "step": 6196
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5213404893875122,
+      "learning_rate": 8.978259461614769e-07,
+      "loss": 0.831,
+      "step": 6197
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6067389845848083,
+      "learning_rate": 8.966130122141242e-07,
+      "loss": 1.0095,
+      "step": 6198
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.429313063621521,
+      "learning_rate": 8.954008174362994e-07,
+      "loss": 0.8462,
+      "step": 6199
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2809960842132568,
+      "learning_rate": 8.94189362046362e-07,
+      "loss": 0.8745,
+      "step": 6200
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5922907590866089,
+      "learning_rate": 8.929786462625379e-07,
+      "loss": 1.0196,
+      "step": 6201
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3010764122009277,
+      "learning_rate": 8.917686703029233e-07,
+      "loss": 0.8001,
+      "step": 6202
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3498786687850952,
+      "learning_rate": 8.905594343854785e-07,
+      "loss": 0.8163,
+      "step": 6203
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3784815073013306,
+      "learning_rate": 8.893509387280302e-07,
+      "loss": 0.8653,
+      "step": 6204
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2520509958267212,
+      "learning_rate": 8.881431835482718e-07,
+      "loss": 0.7253,
+      "step": 6205
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3606568574905396,
+      "learning_rate": 8.869361690637668e-07,
+      "loss": 0.8649,
+      "step": 6206
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6330058574676514,
+      "learning_rate": 8.857298954919402e-07,
+      "loss": 0.8574,
+      "step": 6207
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.333374261856079,
+      "learning_rate": 8.84524363050085e-07,
+      "loss": 0.7658,
+      "step": 6208
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.5469938516616821,
+      "learning_rate": 8.833195719553645e-07,
+      "loss": 0.8031,
+      "step": 6209
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3189953565597534,
+      "learning_rate": 8.821155224248035e-07,
+      "loss": 0.8047,
+      "step": 6210
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3515925407409668,
+      "learning_rate": 8.809122146752946e-07,
+      "loss": 0.813,
+      "step": 6211
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.6532939672470093,
+      "learning_rate": 8.797096489235995e-07,
+      "loss": 0.8128,
+      "step": 6212
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.374282956123352,
+      "learning_rate": 8.785078253863422e-07,
+      "loss": 0.8629,
+      "step": 6213
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5820692181587219,
+      "learning_rate": 8.773067442800164e-07,
+      "loss": 1.017,
+      "step": 6214
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.3157607316970825,
+      "learning_rate": 8.7610640582098e-07,
+      "loss": 0.8407,
+      "step": 6215
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6080507636070251,
+      "learning_rate": 8.749068102254559e-07,
+      "loss": 0.9746,
+      "step": 6216
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5726568102836609,
+      "learning_rate": 8.737079577095386e-07,
+      "loss": 1.0208,
+      "step": 6217
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6143834590911865,
+      "learning_rate": 8.725098484891825e-07,
+      "loss": 0.9878,
+      "step": 6218
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5854786038398743,
+      "learning_rate": 8.713124827802105e-07,
+      "loss": 1.0039,
+      "step": 6219
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3165907859802246,
+      "learning_rate": 8.70115860798314e-07,
+      "loss": 0.8537,
+      "step": 6220
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3797430992126465,
+      "learning_rate": 8.689199827590455e-07,
+      "loss": 0.8316,
+      "step": 6221
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5928609371185303,
+      "learning_rate": 8.67724848877829e-07,
+      "loss": 0.9848,
+      "step": 6222
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3378517627716064,
+      "learning_rate": 8.665304593699503e-07,
+      "loss": 0.8217,
+      "step": 6223
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3513660430908203,
+      "learning_rate": 8.653368144505614e-07,
+      "loss": 0.8181,
+      "step": 6224
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.032189130783081,
+      "learning_rate": 8.64143914334683e-07,
+      "loss": 0.9181,
+      "step": 6225
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4680283069610596,
+      "learning_rate": 8.629517592371983e-07,
+      "loss": 0.8661,
+      "step": 6226
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.583757758140564,
+      "learning_rate": 8.617603493728604e-07,
+      "loss": 1.0196,
+      "step": 6227
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2344472408294678,
+      "learning_rate": 8.605696849562834e-07,
+      "loss": 0.7468,
+      "step": 6228
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.285304069519043,
+      "learning_rate": 8.593797662019493e-07,
+      "loss": 0.7503,
+      "step": 6229
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4358665943145752,
+      "learning_rate": 8.581905933242068e-07,
+      "loss": 0.8256,
+      "step": 6230
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6001828908920288,
+      "learning_rate": 8.570021665372686e-07,
+      "loss": 1.0096,
+      "step": 6231
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3508541584014893,
+      "learning_rate": 8.558144860552147e-07,
+      "loss": 0.8113,
+      "step": 6232
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4826107025146484,
+      "learning_rate": 8.546275520919894e-07,
+      "loss": 0.7943,
+      "step": 6233
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3513100147247314,
+      "learning_rate": 8.534413648614009e-07,
+      "loss": 0.8093,
+      "step": 6234
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.672287106513977,
+      "learning_rate": 8.522559245771272e-07,
+      "loss": 0.8258,
+      "step": 6235
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4167484045028687,
+      "learning_rate": 8.51071231452707e-07,
+      "loss": 0.888,
+      "step": 6236
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3172270059585571,
+      "learning_rate": 8.498872857015495e-07,
+      "loss": 0.7923,
+      "step": 6237
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5825258493423462,
+      "learning_rate": 8.487040875369251e-07,
+      "loss": 0.8298,
+      "step": 6238
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5990819931030273,
+      "learning_rate": 8.475216371719691e-07,
+      "loss": 1.0217,
+      "step": 6239
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5857210755348206,
+      "learning_rate": 8.463399348196871e-07,
+      "loss": 0.9776,
+      "step": 6240
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4516679048538208,
+      "learning_rate": 8.45158980692945e-07,
+      "loss": 0.7611,
+      "step": 6241
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.304887294769287,
+      "learning_rate": 8.439787750044742e-07,
+      "loss": 0.8279,
+      "step": 6242
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4698989391326904,
+      "learning_rate": 8.427993179668764e-07,
+      "loss": 0.7979,
+      "step": 6243
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3322829008102417,
+      "learning_rate": 8.416206097926122e-07,
+      "loss": 0.8131,
+      "step": 6244
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2242902517318726,
+      "learning_rate": 8.404426506940111e-07,
+      "loss": 0.8017,
+      "step": 6245
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2994779348373413,
+      "learning_rate": 8.39265440883264e-07,
+      "loss": 0.8082,
+      "step": 6246
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3634876012802124,
+      "learning_rate": 8.380889805724329e-07,
+      "loss": 0.8375,
+      "step": 6247
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3729331493377686,
+      "learning_rate": 8.369132699734394e-07,
+      "loss": 0.7983,
+      "step": 6248
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5377017259597778,
+      "learning_rate": 8.357383092980703e-07,
+      "loss": 0.8018,
+      "step": 6249
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.437077283859253,
+      "learning_rate": 8.345640987579817e-07,
+      "loss": 0.8073,
+      "step": 6250
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.399674415588379,
+      "learning_rate": 8.333906385646906e-07,
+      "loss": 0.737,
+      "step": 6251
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6103945970535278,
+      "learning_rate": 8.322179289295784e-07,
+      "loss": 0.8249,
+      "step": 6252
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.6593035459518433,
+      "learning_rate": 8.310459700638957e-07,
+      "loss": 0.8356,
+      "step": 6253
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3837299346923828,
+      "learning_rate": 8.298747621787517e-07,
+      "loss": 0.8445,
+      "step": 6254
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3118714094161987,
+      "learning_rate": 8.287043054851268e-07,
+      "loss": 0.8206,
+      "step": 6255
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 3.1803393363952637,
+      "learning_rate": 8.275346001938617e-07,
+      "loss": 0.8542,
+      "step": 6256
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5846434235572815,
+      "learning_rate": 8.263656465156611e-07,
+      "loss": 0.9946,
+      "step": 6257
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3135218620300293,
+      "learning_rate": 8.251974446610994e-07,
+      "loss": 0.7922,
+      "step": 6258
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.413478970527649,
+      "learning_rate": 8.240299948406094e-07,
+      "loss": 0.839,
+      "step": 6259
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3973323106765747,
+      "learning_rate": 8.22863297264494e-07,
+      "loss": 0.7852,
+      "step": 6260
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3667930364608765,
+      "learning_rate": 8.216973521429161e-07,
+      "loss": 0.8284,
+      "step": 6261
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3501793146133423,
+      "learning_rate": 8.205321596859045e-07,
+      "loss": 0.8032,
+      "step": 6262
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3733590841293335,
+      "learning_rate": 8.193677201033551e-07,
+      "loss": 0.8686,
+      "step": 6263
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5861020684242249,
+      "learning_rate": 8.182040336050234e-07,
+      "loss": 0.992,
+      "step": 6264
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2981445789337158,
+      "learning_rate": 8.170411004005341e-07,
+      "loss": 0.8398,
+      "step": 6265
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2351181507110596,
+      "learning_rate": 8.158789206993728e-07,
+      "loss": 0.786,
+      "step": 6266
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.606193482875824,
+      "learning_rate": 8.147174947108888e-07,
+      "loss": 0.9874,
+      "step": 6267
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5835267901420593,
+      "learning_rate": 8.135568226442997e-07,
+      "loss": 0.9852,
+      "step": 6268
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5855432152748108,
+      "learning_rate": 8.123969047086828e-07,
+      "loss": 1.0189,
+      "step": 6269
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4255956411361694,
+      "learning_rate": 8.11237741112984e-07,
+      "loss": 0.8245,
+      "step": 6270
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.585668683052063,
+      "learning_rate": 8.100793320660095e-07,
+      "loss": 1.028,
+      "step": 6271
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3715472221374512,
+      "learning_rate": 8.089216777764297e-07,
+      "loss": 0.8249,
+      "step": 6272
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.489250898361206,
+      "learning_rate": 8.077647784527831e-07,
+      "loss": 0.8033,
+      "step": 6273
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.269793152809143,
+      "learning_rate": 8.066086343034663e-07,
+      "loss": 0.8436,
+      "step": 6274
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5802743434906006,
+      "learning_rate": 8.054532455367459e-07,
+      "loss": 1.0019,
+      "step": 6275
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3322337865829468,
+      "learning_rate": 8.042986123607488e-07,
+      "loss": 0.8222,
+      "step": 6276
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3238270282745361,
+      "learning_rate": 8.031447349834643e-07,
+      "loss": 0.8109,
+      "step": 6277
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2703372240066528,
+      "learning_rate": 8.019916136127509e-07,
+      "loss": 0.8102,
+      "step": 6278
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.8262333869934082,
+      "learning_rate": 8.00839248456326e-07,
+      "loss": 0.8351,
+      "step": 6279
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5698643326759338,
+      "learning_rate": 7.996876397217723e-07,
+      "loss": 0.9976,
+      "step": 6280
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3427330255508423,
+      "learning_rate": 7.985367876165378e-07,
+      "loss": 0.8303,
+      "step": 6281
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5879999995231628,
+      "learning_rate": 7.973866923479323e-07,
+      "loss": 0.9992,
+      "step": 6282
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5803486108779907,
+      "learning_rate": 7.962373541231283e-07,
+      "loss": 1.0134,
+      "step": 6283
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.2519242763519287,
+      "learning_rate": 7.950887731491669e-07,
+      "loss": 0.8507,
+      "step": 6284
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4936035871505737,
+      "learning_rate": 7.939409496329465e-07,
+      "loss": 0.7938,
+      "step": 6285
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.3851596117019653,
+      "learning_rate": 7.927938837812338e-07,
+      "loss": 0.8709,
+      "step": 6286
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5817879438400269,
+      "learning_rate": 7.916475758006543e-07,
+      "loss": 0.9982,
+      "step": 6287
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5891947150230408,
+      "learning_rate": 7.905020258977031e-07,
+      "loss": 1.0103,
+      "step": 6288
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5798554420471191,
+      "learning_rate": 7.893572342787342e-07,
+      "loss": 0.7819,
+      "step": 6289
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.241498589515686,
+      "learning_rate": 7.882132011499654e-07,
+      "loss": 0.7695,
+      "step": 6290
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.596241295337677,
+      "learning_rate": 7.870699267174809e-07,
+      "loss": 1.0143,
+      "step": 6291
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5894591808319092,
+      "learning_rate": 7.859274111872228e-07,
+      "loss": 1.0177,
+      "step": 6292
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.567082941532135,
+      "learning_rate": 7.84785654765004e-07,
+      "loss": 0.9828,
+      "step": 6293
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.4573602676391602,
+      "learning_rate": 7.836446576564932e-07,
+      "loss": 0.8599,
+      "step": 6294
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5921211242675781,
+      "learning_rate": 7.82504420067226e-07,
+      "loss": 1.0086,
+      "step": 6295
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.37178373336792,
+      "learning_rate": 7.813649422026026e-07,
+      "loss": 0.7802,
+      "step": 6296
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2280619144439697,
+      "learning_rate": 7.802262242678815e-07,
+      "loss": 0.7808,
+      "step": 6297
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5778082609176636,
+      "learning_rate": 7.790882664681909e-07,
+      "loss": 1.0042,
+      "step": 6298
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2854833602905273,
+      "learning_rate": 7.779510690085157e-07,
+      "loss": 0.8311,
+      "step": 6299
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2369147539138794,
+      "learning_rate": 7.76814632093707e-07,
+      "loss": 0.788,
+      "step": 6300
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.563392162322998,
+      "learning_rate": 7.7567895592848e-07,
+      "loss": 1.0113,
+      "step": 6301
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2643022537231445,
+      "learning_rate": 7.745440407174088e-07,
+      "loss": 0.764,
+      "step": 6302
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5722072720527649,
+      "learning_rate": 7.734098866649358e-07,
+      "loss": 1.0016,
+      "step": 6303
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5687304735183716,
+      "learning_rate": 7.722764939753624e-07,
+      "loss": 1.0055,
+      "step": 6304
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6145217418670654,
+      "learning_rate": 7.711438628528523e-07,
+      "loss": 1.0179,
+      "step": 6305
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.278259038925171,
+      "learning_rate": 7.700119935014367e-07,
+      "loss": 0.8565,
+      "step": 6306
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5006990432739258,
+      "learning_rate": 7.688808861250047e-07,
+      "loss": 0.7961,
+      "step": 6307
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.387999176979065,
+      "learning_rate": 7.677505409273084e-07,
+      "loss": 0.7949,
+      "step": 6308
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.568187415599823,
+      "learning_rate": 7.666209581119677e-07,
+      "loss": 0.9893,
+      "step": 6309
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.7108820676803589,
+      "learning_rate": 7.654921378824582e-07,
+      "loss": 0.8504,
+      "step": 6310
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5726518630981445,
+      "learning_rate": 7.643640804421248e-07,
+      "loss": 0.9938,
+      "step": 6311
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5771254897117615,
+      "learning_rate": 7.632367859941698e-07,
+      "loss": 0.9951,
+      "step": 6312
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3006060123443604,
+      "learning_rate": 7.621102547416592e-07,
+      "loss": 0.7995,
+      "step": 6313
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6727641820907593,
+      "learning_rate": 7.609844868875249e-07,
+      "loss": 0.8552,
+      "step": 6314
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5702099800109863,
+      "learning_rate": 7.598594826345557e-07,
+      "loss": 1.0177,
+      "step": 6315
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2498176097869873,
+      "learning_rate": 7.587352421854088e-07,
+      "loss": 0.8406,
+      "step": 6316
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.8611772060394287,
+      "learning_rate": 7.576117657425997e-07,
+      "loss": 0.8051,
+      "step": 6317
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4514760971069336,
+      "learning_rate": 7.564890535085056e-07,
+      "loss": 0.8268,
+      "step": 6318
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4581880569458008,
+      "learning_rate": 7.553671056853712e-07,
+      "loss": 0.8388,
+      "step": 6319
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.358754277229309,
+      "learning_rate": 7.542459224752973e-07,
+      "loss": 0.8043,
+      "step": 6320
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5691114068031311,
+      "learning_rate": 7.53125504080252e-07,
+      "loss": 0.9913,
+      "step": 6321
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5858156085014343,
+      "learning_rate": 7.520058507020627e-07,
+      "loss": 1.0091,
+      "step": 6322
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3720883131027222,
+      "learning_rate": 7.508869625424198e-07,
+      "loss": 0.8399,
+      "step": 6323
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3807260990142822,
+      "learning_rate": 7.497688398028746e-07,
+      "loss": 0.7956,
+      "step": 6324
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.464770793914795,
+      "learning_rate": 7.48651482684844e-07,
+      "loss": 0.864,
+      "step": 6325
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5756046772003174,
+      "learning_rate": 7.47534891389603e-07,
+      "loss": 0.9757,
+      "step": 6326
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3699833154678345,
+      "learning_rate": 7.464190661182918e-07,
+      "loss": 0.8273,
+      "step": 6327
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.5645238161087036,
+      "learning_rate": 7.453040070719087e-07,
+      "loss": 0.7888,
+      "step": 6328
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3337962627410889,
+      "learning_rate": 7.441897144513188e-07,
+      "loss": 0.7682,
+      "step": 6329
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4281219244003296,
+      "learning_rate": 7.430761884572457e-07,
+      "loss": 0.8212,
+      "step": 6330
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.479784369468689,
+      "learning_rate": 7.419634292902772e-07,
+      "loss": 0.8435,
+      "step": 6331
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3940768241882324,
+      "learning_rate": 7.408514371508613e-07,
+      "loss": 0.8289,
+      "step": 6332
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3306623697280884,
+      "learning_rate": 7.397402122393066e-07,
+      "loss": 0.8071,
+      "step": 6333
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2465771436691284,
+      "learning_rate": 7.386297547557874e-07,
+      "loss": 0.7556,
+      "step": 6334
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.597827136516571,
+      "learning_rate": 7.375200649003372e-07,
+      "loss": 1.0107,
+      "step": 6335
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5751919746398926,
+      "learning_rate": 7.364111428728504e-07,
+      "loss": 1.0107,
+      "step": 6336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3323099613189697,
+      "learning_rate": 7.353029888730856e-07,
+      "loss": 0.8158,
+      "step": 6337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5890313386917114,
+      "learning_rate": 7.341956031006597e-07,
+      "loss": 1.0113,
+      "step": 6338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 4.9438862800598145,
+      "learning_rate": 7.330889857550566e-07,
+      "loss": 0.7788,
+      "step": 6339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4626573324203491,
+      "learning_rate": 7.319831370356162e-07,
+      "loss": 0.8484,
+      "step": 6340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4614814519882202,
+      "learning_rate": 7.308780571415408e-07,
+      "loss": 0.8053,
+      "step": 6341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2543790340423584,
+      "learning_rate": 7.297737462718985e-07,
+      "loss": 0.7434,
+      "step": 6342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2439271211624146,
+      "learning_rate": 7.286702046256139e-07,
+      "loss": 0.8605,
+      "step": 6343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3213014602661133,
+      "learning_rate": 7.275674324014764e-07,
+      "loss": 0.8248,
+      "step": 6344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3636424541473389,
+      "learning_rate": 7.264654297981355e-07,
+      "loss": 0.7634,
+      "step": 6345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4421420097351074,
+      "learning_rate": 7.253641970140995e-07,
+      "loss": 0.7996,
+      "step": 6346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2816957235336304,
+      "learning_rate": 7.242637342477437e-07,
+      "loss": 0.8196,
+      "step": 6347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3596967458724976,
+      "learning_rate": 7.231640416972996e-07,
+      "loss": 0.7779,
+      "step": 6348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.6733180284500122,
+      "learning_rate": 7.220651195608635e-07,
+      "loss": 0.7547,
+      "step": 6349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2310004234313965,
+      "learning_rate": 7.209669680363901e-07,
+      "loss": 0.7901,
+      "step": 6350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3406422138214111,
+      "learning_rate": 7.19869587321696e-07,
+      "loss": 0.8316,
+      "step": 6351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3810462951660156,
+      "learning_rate": 7.187729776144608e-07,
+      "loss": 0.7746,
+      "step": 6352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3300970792770386,
+      "learning_rate": 7.176771391122228e-07,
+      "loss": 0.852,
+      "step": 6353
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.568989634513855,
+      "learning_rate": 7.16582072012384e-07,
+      "loss": 1.0049,
+      "step": 6354
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3590538501739502,
+      "learning_rate": 7.154877765122048e-07,
+      "loss": 0.8674,
+      "step": 6355
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3175668716430664,
+      "learning_rate": 7.143942528088066e-07,
+      "loss": 0.7591,
+      "step": 6356
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4441062211990356,
+      "learning_rate": 7.13301501099175e-07,
+      "loss": 0.8108,
+      "step": 6357
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.36542546749115,
+      "learning_rate": 7.122095215801522e-07,
+      "loss": 0.8146,
+      "step": 6358
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6132822036743164,
+      "learning_rate": 7.111183144484462e-07,
+      "loss": 1.0101,
+      "step": 6359
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5772982239723206,
+      "learning_rate": 7.100278799006216e-07,
+      "loss": 1.0025,
+      "step": 6360
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3591578006744385,
+      "learning_rate": 7.089382181331045e-07,
+      "loss": 0.7676,
+      "step": 6361
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3024020195007324,
+      "learning_rate": 7.078493293421845e-07,
+      "loss": 0.8488,
+      "step": 6362
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3607925176620483,
+      "learning_rate": 7.067612137240098e-07,
+      "loss": 0.8429,
+      "step": 6363
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3959565162658691,
+      "learning_rate": 7.056738714745887e-07,
+      "loss": 0.7663,
+      "step": 6364
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2502104043960571,
+      "learning_rate": 7.045873027897909e-07,
+      "loss": 0.8445,
+      "step": 6365
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2833161354064941,
+      "learning_rate": 7.03501507865349e-07,
+      "loss": 0.8274,
+      "step": 6366
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.4003932476043701,
+      "learning_rate": 7.024164868968525e-07,
+      "loss": 0.8382,
+      "step": 6367
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5805695652961731,
+      "learning_rate": 7.01332240079754e-07,
+      "loss": 1.005,
+      "step": 6368
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.427797794342041,
+      "learning_rate": 7.002487676093645e-07,
+      "loss": 0.8523,
+      "step": 6369
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.2534410953521729,
+      "learning_rate": 6.991660696808594e-07,
+      "loss": 0.8427,
+      "step": 6370
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.594996988773346,
+      "learning_rate": 6.98084146489269e-07,
+      "loss": 0.9931,
+      "step": 6371
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 1.3664134740829468,
+      "learning_rate": 6.970029982294902e-07,
+      "loss": 0.8724,
+      "step": 6372
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5781664848327637,
+      "learning_rate": 6.959226250962759e-07,
+      "loss": 1.005,
+      "step": 6373
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5807703137397766,
+      "learning_rate": 6.948430272842388e-07,
+      "loss": 1.0271,
+      "step": 6374
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3513363599777222,
+      "learning_rate": 6.937642049878568e-07,
+      "loss": 0.8654,
+      "step": 6375
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2404065132141113,
+      "learning_rate": 6.926861584014627e-07,
+      "loss": 0.7918,
+      "step": 6376
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5927834510803223,
+      "learning_rate": 6.916088877192545e-07,
+      "loss": 1.0258,
+      "step": 6377
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3468518257141113,
+      "learning_rate": 6.905323931352864e-07,
+      "loss": 0.8583,
+      "step": 6378
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3962236642837524,
+      "learning_rate": 6.894566748434734e-07,
+      "loss": 0.8629,
+      "step": 6379
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3961809873580933,
+      "learning_rate": 6.883817330375931e-07,
+      "loss": 0.8177,
+      "step": 6380
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3649331331253052,
+      "learning_rate": 6.873075679112806e-07,
+      "loss": 0.7736,
+      "step": 6381
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3732868432998657,
+      "learning_rate": 6.862341796580335e-07,
+      "loss": 0.8182,
+      "step": 6382
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3829699754714966,
+      "learning_rate": 6.851615684712077e-07,
+      "loss": 0.7908,
+      "step": 6383
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5711863040924072,
+      "learning_rate": 6.840897345440173e-07,
+      "loss": 0.9934,
+      "step": 6384
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.26602303981781,
+      "learning_rate": 6.830186780695425e-07,
+      "loss": 0.8232,
+      "step": 6385
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3814504146575928,
+      "learning_rate": 6.819483992407166e-07,
+      "loss": 0.8123,
+      "step": 6386
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3371491432189941,
+      "learning_rate": 6.808788982503379e-07,
+      "loss": 0.8027,
+      "step": 6387
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2498112916946411,
+      "learning_rate": 6.798101752910613e-07,
+      "loss": 0.7419,
+      "step": 6388
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3299320936203003,
+      "learning_rate": 6.787422305554015e-07,
+      "loss": 0.7947,
+      "step": 6389
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.367149829864502,
+      "learning_rate": 6.776750642357372e-07,
+      "loss": 0.8529,
+      "step": 6390
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3140901327133179,
+      "learning_rate": 6.766086765243013e-07,
+      "loss": 0.8347,
+      "step": 6391
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3396869897842407,
+      "learning_rate": 6.755430676131919e-07,
+      "loss": 0.8004,
+      "step": 6392
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3976205587387085,
+      "learning_rate": 6.744782376943615e-07,
+      "loss": 0.8058,
+      "step": 6393
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.439739465713501,
+      "learning_rate": 6.73414186959625e-07,
+      "loss": 0.8112,
+      "step": 6394
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3746815919876099,
+      "learning_rate": 6.723509156006585e-07,
+      "loss": 0.8008,
+      "step": 6395
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.28885018825531,
+      "learning_rate": 6.712884238089945e-07,
+      "loss": 0.777,
+      "step": 6396
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5723810195922852,
+      "learning_rate": 6.702267117760253e-07,
+      "loss": 1.0021,
+      "step": 6397
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5894242525100708,
+      "learning_rate": 6.691657796930074e-07,
+      "loss": 0.9904,
+      "step": 6398
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4145236015319824,
+      "learning_rate": 6.681056277510495e-07,
+      "loss": 0.798,
+      "step": 6399
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3490396738052368,
+      "learning_rate": 6.670462561411262e-07,
+      "loss": 0.8193,
+      "step": 6400
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 5.67944860458374,
+      "learning_rate": 6.659876650540687e-07,
+      "loss": 0.8406,
+      "step": 6401
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5875470638275146,
+      "learning_rate": 6.64929854680566e-07,
+      "loss": 1.018,
+      "step": 6402
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.302955150604248,
+      "learning_rate": 6.638728252111704e-07,
+      "loss": 0.812,
+      "step": 6403
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3770462274551392,
+      "learning_rate": 6.62816576836291e-07,
+      "loss": 0.7985,
+      "step": 6404
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5835411548614502,
+      "learning_rate": 6.617611097461962e-07,
+      "loss": 1.0088,
+      "step": 6405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.409132480621338,
+      "learning_rate": 6.607064241310124e-07,
+      "loss": 0.7875,
+      "step": 6406
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5802074074745178,
+      "learning_rate": 6.596525201807302e-07,
+      "loss": 1.0076,
+      "step": 6407
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.36496102809906,
+      "learning_rate": 6.585993980851945e-07,
+      "loss": 0.9027,
+      "step": 6408
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3691805601119995,
+      "learning_rate": 6.575470580341098e-07,
+      "loss": 0.7566,
+      "step": 6409
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3886573314666748,
+      "learning_rate": 6.564955002170425e-07,
+      "loss": 0.8018,
+      "step": 6410
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2988970279693604,
+      "learning_rate": 6.55444724823417e-07,
+      "loss": 0.895,
+      "step": 6411
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5495002269744873,
+      "learning_rate": 6.543947320425137e-07,
+      "loss": 0.8317,
+      "step": 6412
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5856629610061646,
+      "learning_rate": 6.533455220634771e-07,
+      "loss": 1.0073,
+      "step": 6413
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4533867835998535,
+      "learning_rate": 6.522970950753066e-07,
+      "loss": 0.8915,
+      "step": 6414
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.336607575416565,
+      "learning_rate": 6.512494512668633e-07,
+      "loss": 0.7873,
+      "step": 6415
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5783938765525818,
+      "learning_rate": 6.502025908268655e-07,
+      "loss": 0.9939,
+      "step": 6416
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4092689752578735,
+      "learning_rate": 6.491565139438899e-07,
+      "loss": 0.8006,
+      "step": 6417
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.580971121788025,
+      "learning_rate": 6.48111220806375e-07,
+      "loss": 0.8535,
+      "step": 6418
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5852338671684265,
+      "learning_rate": 6.470667116026141e-07,
+      "loss": 1.0289,
+      "step": 6419
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3551125526428223,
+      "learning_rate": 6.460229865207635e-07,
+      "loss": 0.8449,
+      "step": 6420
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4241262674331665,
+      "learning_rate": 6.449800457488354e-07,
+      "loss": 0.8032,
+      "step": 6421
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5997257232666016,
+      "learning_rate": 6.439378894746995e-07,
+      "loss": 1.0225,
+      "step": 6422
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2480658292770386,
+      "learning_rate": 6.428965178860891e-07,
+      "loss": 0.7757,
+      "step": 6423
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5588975548744202,
+      "learning_rate": 6.418559311705919e-07,
+      "loss": 1.0112,
+      "step": 6424
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3769432306289673,
+      "learning_rate": 6.408161295156545e-07,
+      "loss": 0.7987,
+      "step": 6425
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4003040790557861,
+      "learning_rate": 6.397771131085845e-07,
+      "loss": 0.8177,
+      "step": 6426
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3459161520004272,
+      "learning_rate": 6.387388821365453e-07,
+      "loss": 0.7912,
+      "step": 6427
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3721985816955566,
+      "learning_rate": 6.37701436786562e-07,
+      "loss": 0.8075,
+      "step": 6428
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3293888568878174,
+      "learning_rate": 6.366647772455154e-07,
+      "loss": 0.8613,
+      "step": 6429
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5670851469039917,
+      "learning_rate": 6.35628903700144e-07,
+      "loss": 0.9972,
+      "step": 6430
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.408218502998352,
+      "learning_rate": 6.345938163370491e-07,
+      "loss": 0.7748,
+      "step": 6431
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4959794282913208,
+      "learning_rate": 6.335595153426854e-07,
+      "loss": 0.8105,
+      "step": 6432
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.449593424797058,
+      "learning_rate": 6.325260009033707e-07,
+      "loss": 0.8373,
+      "step": 6433
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3847328424453735,
+      "learning_rate": 6.314932732052764e-07,
+      "loss": 0.7874,
+      "step": 6434
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3407790660858154,
+      "learning_rate": 6.304613324344344e-07,
+      "loss": 0.8041,
+      "step": 6435
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.5036537647247314,
+      "learning_rate": 6.294301787767365e-07,
+      "loss": 0.8109,
+      "step": 6436
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.3485767841339111,
+      "learning_rate": 6.283998124179292e-07,
+      "loss": 0.8458,
+      "step": 6437
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.299645185470581,
+      "learning_rate": 6.27370233543621e-07,
+      "loss": 0.781,
+      "step": 6438
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.472183108329773,
+      "learning_rate": 6.263414423392755e-07,
+      "loss": 0.8786,
+      "step": 6439
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5846260786056519,
+      "learning_rate": 6.253134389902143e-07,
+      "loss": 1.0114,
+      "step": 6440
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2733937501907349,
+      "learning_rate": 6.242862236816205e-07,
+      "loss": 0.8323,
+      "step": 6441
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2466745376586914,
+      "learning_rate": 6.232597965985321e-07,
+      "loss": 0.8186,
+      "step": 6442
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5894896388053894,
+      "learning_rate": 6.222341579258451e-07,
+      "loss": 1.0406,
+      "step": 6443
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5745296478271484,
+      "learning_rate": 6.212093078483155e-07,
+      "loss": 1.0206,
+      "step": 6444
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5829235315322876,
+      "learning_rate": 6.201852465505564e-07,
+      "loss": 1.0036,
+      "step": 6445
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.4038894176483154,
+      "learning_rate": 6.19161974217038e-07,
+      "loss": 0.8385,
+      "step": 6446
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5772058367729187,
+      "learning_rate": 6.181394910320876e-07,
+      "loss": 0.9952,
+      "step": 6447
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5822814702987671,
+      "learning_rate": 6.171177971798942e-07,
+      "loss": 1.017,
+      "step": 6448
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3102089166641235,
+      "learning_rate": 6.16096892844501e-07,
+      "loss": 0.8571,
+      "step": 6449
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4402042627334595,
+      "learning_rate": 6.150767782098088e-07,
+      "loss": 0.8439,
+      "step": 6450
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3383033275604248,
+      "learning_rate": 6.1405745345958e-07,
+      "loss": 0.7536,
+      "step": 6451
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4279305934906006,
+      "learning_rate": 6.130389187774305e-07,
+      "loss": 0.8129,
+      "step": 6452
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.6856247186660767,
+      "learning_rate": 6.12021174346834e-07,
+      "loss": 0.7723,
+      "step": 6453
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5303469896316528,
+      "learning_rate": 6.110042203511268e-07,
+      "loss": 0.8599,
+      "step": 6454
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3442180156707764,
+      "learning_rate": 6.099880569734967e-07,
+      "loss": 0.8028,
+      "step": 6455
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.314329981803894,
+      "learning_rate": 6.08972684396994e-07,
+      "loss": 0.7666,
+      "step": 6456
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.585945188999176,
+      "learning_rate": 6.079581028045223e-07,
+      "loss": 0.9831,
+      "step": 6457
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4276623725891113,
+      "learning_rate": 6.069443123788449e-07,
+      "loss": 0.8486,
+      "step": 6458
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3428703546524048,
+      "learning_rate": 6.059313133025841e-07,
+      "loss": 0.7971,
+      "step": 6459
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5903027057647705,
+      "learning_rate": 6.049191057582155e-07,
+      "loss": 1.0165,
+      "step": 6460
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3449920415878296,
+      "learning_rate": 6.039076899280766e-07,
+      "loss": 0.8198,
+      "step": 6461
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3552418947219849,
+      "learning_rate": 6.028970659943601e-07,
+      "loss": 0.8392,
+      "step": 6462
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.351225733757019,
+      "learning_rate": 6.018872341391141e-07,
+      "loss": 0.7651,
+      "step": 6463
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3050165176391602,
+      "learning_rate": 6.008781945442488e-07,
+      "loss": 0.7776,
+      "step": 6464
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3632076978683472,
+      "learning_rate": 5.99869947391527e-07,
+      "loss": 0.8385,
+      "step": 6465
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4165266752243042,
+      "learning_rate": 5.988624928625725e-07,
+      "loss": 0.7938,
+      "step": 6466
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 2.4384212493896484,
+      "learning_rate": 5.978558311388639e-07,
+      "loss": 0.8591,
+      "step": 6467
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5711783766746521,
+      "learning_rate": 5.96849962401736e-07,
+      "loss": 1.0028,
+      "step": 6468
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5689610242843628,
+      "learning_rate": 5.958448868323846e-07,
+      "loss": 1.0108,
+      "step": 6469
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.594964861869812,
+      "learning_rate": 5.948406046118594e-07,
+      "loss": 1.0131,
+      "step": 6470
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.531395673751831,
+      "learning_rate": 5.938371159210687e-07,
+      "loss": 0.8199,
+      "step": 6471
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3424931764602661,
+      "learning_rate": 5.928344209407777e-07,
+      "loss": 0.7759,
+      "step": 6472
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4553825855255127,
+      "learning_rate": 5.918325198516067e-07,
+      "loss": 0.8502,
+      "step": 6473
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4742777347564697,
+      "learning_rate": 5.908314128340364e-07,
+      "loss": 0.8278,
+      "step": 6474
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4391109943389893,
+      "learning_rate": 5.89831100068401e-07,
+      "loss": 0.8073,
+      "step": 6475
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3292407989501953,
+      "learning_rate": 5.888315817348955e-07,
+      "loss": 0.8391,
+      "step": 6476
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3819475173950195,
+      "learning_rate": 5.878328580135679e-07,
+      "loss": 0.8573,
+      "step": 6477
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5568218231201172,
+      "learning_rate": 5.868349290843239e-07,
+      "loss": 0.8097,
+      "step": 6478
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5837796926498413,
+      "learning_rate": 5.858377951269295e-07,
+      "loss": 1.0143,
+      "step": 6479
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3851580619812012,
+      "learning_rate": 5.848414563210036e-07,
+      "loss": 0.8611,
+      "step": 6480
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.299323558807373,
+      "learning_rate": 5.838459128460216e-07,
+      "loss": 0.8429,
+      "step": 6481
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3782782554626465,
+      "learning_rate": 5.828511648813201e-07,
+      "loss": 0.8402,
+      "step": 6482
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5943530797958374,
+      "learning_rate": 5.818572126060873e-07,
+      "loss": 1.0423,
+      "step": 6483
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 3.4013466835021973,
+      "learning_rate": 5.8086405619937e-07,
+      "loss": 0.8306,
+      "step": 6484
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4495713710784912,
+      "learning_rate": 5.798716958400735e-07,
+      "loss": 0.8057,
+      "step": 6485
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3511207103729248,
+      "learning_rate": 5.788801317069575e-07,
+      "loss": 0.8331,
+      "step": 6486
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.327960729598999,
+      "learning_rate": 5.778893639786381e-07,
+      "loss": 0.8424,
+      "step": 6487
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3616670370101929,
+      "learning_rate": 5.768993928335887e-07,
+      "loss": 0.7978,
+      "step": 6488
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4936039447784424,
+      "learning_rate": 5.759102184501403e-07,
+      "loss": 0.7431,
+      "step": 6489
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3983556032180786,
+      "learning_rate": 5.749218410064789e-07,
+      "loss": 0.8368,
+      "step": 6490
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5783750414848328,
+      "learning_rate": 5.739342606806453e-07,
+      "loss": 1.0096,
+      "step": 6491
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 6.4604597091674805,
+      "learning_rate": 5.729474776505423e-07,
+      "loss": 0.8469,
+      "step": 6492
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.30471670627594,
+      "learning_rate": 5.719614920939215e-07,
+      "loss": 0.845,
+      "step": 6493
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.368909239768982,
+      "learning_rate": 5.709763041883987e-07,
+      "loss": 0.7446,
+      "step": 6494
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4058644771575928,
+      "learning_rate": 5.699919141114401e-07,
+      "loss": 0.8528,
+      "step": 6495
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.572044849395752,
+      "learning_rate": 5.690083220403692e-07,
+      "loss": 1.0,
+      "step": 6496
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.348859190940857,
+      "learning_rate": 5.680255281523694e-07,
+      "loss": 0.7763,
+      "step": 6497
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.401084065437317,
+      "learning_rate": 5.67043532624475e-07,
+      "loss": 0.8285,
+      "step": 6498
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4899523258209229,
+      "learning_rate": 5.660623356335815e-07,
+      "loss": 0.835,
+      "step": 6499
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.332526683807373,
+      "learning_rate": 5.650819373564376e-07,
+      "loss": 0.8305,
+      "step": 6500
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4622411727905273,
+      "learning_rate": 5.641023379696469e-07,
+      "loss": 0.8241,
+      "step": 6501
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5690934658050537,
+      "learning_rate": 5.631235376496735e-07,
+      "loss": 0.9951,
+      "step": 6502
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3728240728378296,
+      "learning_rate": 5.621455365728329e-07,
+      "loss": 0.8274,
+      "step": 6503
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3099931478500366,
+      "learning_rate": 5.611683349153013e-07,
+      "loss": 0.806,
+      "step": 6504
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3046438694000244,
+      "learning_rate": 5.601919328531064e-07,
+      "loss": 0.8235,
+      "step": 6505
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5758519768714905,
+      "learning_rate": 5.592163305621329e-07,
+      "loss": 0.9939,
+      "step": 6506
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.5837459564208984,
+      "learning_rate": 5.582415282181253e-07,
+      "loss": 0.7695,
+      "step": 6507
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.357968807220459,
+      "learning_rate": 5.572675259966775e-07,
+      "loss": 0.8065,
+      "step": 6508
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5887260437011719,
+      "learning_rate": 5.562943240732466e-07,
+      "loss": 1.0129,
+      "step": 6509
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2526670694351196,
+      "learning_rate": 5.553219226231393e-07,
+      "loss": 0.7857,
+      "step": 6510
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.4409524202346802,
+      "learning_rate": 5.543503218215196e-07,
+      "loss": 0.8052,
+      "step": 6511
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5744151473045349,
+      "learning_rate": 5.533795218434113e-07,
+      "loss": 1.0192,
+      "step": 6512
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3815536499023438,
+      "learning_rate": 5.524095228636889e-07,
+      "loss": 0.8866,
+      "step": 6513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.2907941341400146,
+      "learning_rate": 5.514403250570832e-07,
+      "loss": 0.8447,
+      "step": 6514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5775719881057739,
+      "learning_rate": 5.504719285981858e-07,
+      "loss": 1.0088,
+      "step": 6515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3026801347732544,
+      "learning_rate": 5.495043336614364e-07,
+      "loss": 0.8688,
+      "step": 6516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3023009300231934,
+      "learning_rate": 5.485375404211369e-07,
+      "loss": 0.7841,
+      "step": 6517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.29857337474823,
+      "learning_rate": 5.475715490514405e-07,
+      "loss": 0.8373,
+      "step": 6518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5646094679832458,
+      "learning_rate": 5.466063597263571e-07,
+      "loss": 1.0105,
+      "step": 6519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3693146705627441,
+      "learning_rate": 5.456419726197542e-07,
+      "loss": 0.7842,
+      "step": 6520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.9263923168182373,
+      "learning_rate": 5.446783879053513e-07,
+      "loss": 0.824,
+      "step": 6521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3086174726486206,
+      "learning_rate": 5.437156057567267e-07,
+      "loss": 0.8433,
+      "step": 6522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5698538422584534,
+      "learning_rate": 5.427536263473115e-07,
+      "loss": 0.9791,
+      "step": 6523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.227394938468933,
+      "learning_rate": 5.417924498503935e-07,
+      "loss": 0.8738,
+      "step": 6524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 1.3251091241836548,
+      "learning_rate": 5.408320764391146e-07,
+      "loss": 0.8325,
+      "step": 6525
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.57196044921875,
+      "learning_rate": 5.398725062864751e-07,
+      "loss": 1.0174,
+      "step": 6526
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5588911771774292,
+      "learning_rate": 5.389137395653271e-07,
+      "loss": 1.0122,
+      "step": 6527
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.313694953918457,
+      "learning_rate": 5.3795577644838e-07,
+      "loss": 0.8601,
+      "step": 6528
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5846872329711914,
+      "learning_rate": 5.36998617108197e-07,
+      "loss": 1.006,
+      "step": 6529
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5692394971847534,
+      "learning_rate": 5.36042261717199e-07,
+      "loss": 0.9996,
+      "step": 6530
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.263455867767334,
+      "learning_rate": 5.350867104476581e-07,
+      "loss": 0.8333,
+      "step": 6531
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3813941478729248,
+      "learning_rate": 5.341319634717063e-07,
+      "loss": 0.8165,
+      "step": 6532
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.483888030052185,
+      "learning_rate": 5.331780209613275e-07,
+      "loss": 0.8405,
+      "step": 6533
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.417080283164978,
+      "learning_rate": 5.322248830883603e-07,
+      "loss": 0.8651,
+      "step": 6534
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3976045846939087,
+      "learning_rate": 5.31272550024502e-07,
+      "loss": 0.8355,
+      "step": 6535
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2856484651565552,
+      "learning_rate": 5.303210219412996e-07,
+      "loss": 0.7604,
+      "step": 6536
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3513606786727905,
+      "learning_rate": 5.293702990101607e-07,
+      "loss": 0.8486,
+      "step": 6537
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5367333889007568,
+      "learning_rate": 5.284203814023436e-07,
+      "loss": 0.846,
+      "step": 6538
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3968180418014526,
+      "learning_rate": 5.274712692889627e-07,
+      "loss": 0.8031,
+      "step": 6539
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.575700581073761,
+      "learning_rate": 5.265229628409891e-07,
+      "loss": 0.9878,
+      "step": 6540
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3634839057922363,
+      "learning_rate": 5.255754622292464e-07,
+      "loss": 0.842,
+      "step": 6541
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5476608276367188,
+      "learning_rate": 5.246287676244138e-07,
+      "loss": 0.7509,
+      "step": 6542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2741957902908325,
+      "learning_rate": 5.236828791970266e-07,
+      "loss": 0.8094,
+      "step": 6543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3297604322433472,
+      "learning_rate": 5.227377971174718e-07,
+      "loss": 0.8044,
+      "step": 6544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5822710394859314,
+      "learning_rate": 5.217935215559961e-07,
+      "loss": 0.9932,
+      "step": 6545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3571847677230835,
+      "learning_rate": 5.208500526826954e-07,
+      "loss": 0.7348,
+      "step": 6546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4019980430603027,
+      "learning_rate": 5.199073906675234e-07,
+      "loss": 0.8089,
+      "step": 6547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.7561547756195068,
+      "learning_rate": 5.189655356802886e-07,
+      "loss": 0.8667,
+      "step": 6548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4587377309799194,
+      "learning_rate": 5.18024487890652e-07,
+      "loss": 0.864,
+      "step": 6549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3801897764205933,
+      "learning_rate": 5.170842474681326e-07,
+      "loss": 0.8206,
+      "step": 6550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.310421347618103,
+      "learning_rate": 5.161448145821013e-07,
+      "loss": 0.8501,
+      "step": 6551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5801116228103638,
+      "learning_rate": 5.152061894017823e-07,
+      "loss": 0.9993,
+      "step": 6552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5728558897972107,
+      "learning_rate": 5.142683720962588e-07,
+      "loss": 1.0262,
+      "step": 6553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5951876044273376,
+      "learning_rate": 5.133313628344638e-07,
+      "loss": 1.0015,
+      "step": 6554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.235339879989624,
+      "learning_rate": 5.123951617851891e-07,
+      "loss": 0.8234,
+      "step": 6555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2039707899093628,
+      "learning_rate": 5.11459769117077e-07,
+      "loss": 0.7415,
+      "step": 6556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5820724964141846,
+      "learning_rate": 5.105251849986249e-07,
+      "loss": 1.0107,
+      "step": 6557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.5062029361724854,
+      "learning_rate": 5.095914095981879e-07,
+      "loss": 0.8924,
+      "step": 6558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3756163120269775,
+      "learning_rate": 5.086584430839708e-07,
+      "loss": 0.8543,
+      "step": 6559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3592255115509033,
+      "learning_rate": 5.077262856240367e-07,
+      "loss": 0.7526,
+      "step": 6560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4620293378829956,
+      "learning_rate": 5.067949373863007e-07,
+      "loss": 0.8486,
+      "step": 6561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5653493404388428,
+      "learning_rate": 5.058643985385303e-07,
+      "loss": 1.007,
+      "step": 6562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3172587156295776,
+      "learning_rate": 5.049346692483525e-07,
+      "loss": 0.8298,
+      "step": 6563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5837941765785217,
+      "learning_rate": 5.040057496832446e-07,
+      "loss": 1.0156,
+      "step": 6564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5855484008789062,
+      "learning_rate": 5.03077640010538e-07,
+      "loss": 1.0414,
+      "step": 6565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5777463316917419,
+      "learning_rate": 5.021503403974187e-07,
+      "loss": 0.9996,
+      "step": 6566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3985306024551392,
+      "learning_rate": 5.01223851010929e-07,
+      "loss": 0.83,
+      "step": 6567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5666376948356628,
+      "learning_rate": 5.002981720179622e-07,
+      "loss": 0.9978,
+      "step": 6568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5770873427391052,
+      "learning_rate": 4.993733035852666e-07,
+      "loss": 1.0084,
+      "step": 6569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.291982650756836,
+      "learning_rate": 4.984492458794454e-07,
+      "loss": 0.8313,
+      "step": 6570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3643335103988647,
+      "learning_rate": 4.975259990669551e-07,
+      "loss": 0.8094,
+      "step": 6571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3416774272918701,
+      "learning_rate": 4.966035633141047e-07,
+      "loss": 0.8135,
+      "step": 6572
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.311350703239441,
+      "learning_rate": 4.956819387870604e-07,
+      "loss": 0.809,
+      "step": 6573
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.633307933807373,
+      "learning_rate": 4.947611256518392e-07,
+      "loss": 0.7962,
+      "step": 6574
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3600151538848877,
+      "learning_rate": 4.93841124074313e-07,
+      "loss": 0.8721,
+      "step": 6575
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.358518362045288,
+      "learning_rate": 4.929219342202091e-07,
+      "loss": 0.8246,
+      "step": 6576
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6467846632003784,
+      "learning_rate": 4.920035562551045e-07,
+      "loss": 0.8564,
+      "step": 6577
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3311866521835327,
+      "learning_rate": 4.910859903444354e-07,
+      "loss": 0.8387,
+      "step": 6578
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6013669967651367,
+      "learning_rate": 4.901692366534872e-07,
+      "loss": 1.0088,
+      "step": 6579
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3545663356781006,
+      "learning_rate": 4.892532953474e-07,
+      "loss": 0.8088,
+      "step": 6580
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3495978116989136,
+      "learning_rate": 4.883381665911708e-07,
+      "loss": 0.844,
+      "step": 6581
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3584758043289185,
+      "learning_rate": 4.874238505496442e-07,
+      "loss": 0.8526,
+      "step": 6582
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5968859195709229,
+      "learning_rate": 4.865103473875248e-07,
+      "loss": 1.0138,
+      "step": 6583
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5698787569999695,
+      "learning_rate": 4.855976572693666e-07,
+      "loss": 1.0104,
+      "step": 6584
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5900970697402954,
+      "learning_rate": 4.846857803595778e-07,
+      "loss": 1.0062,
+      "step": 6585
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2774763107299805,
+      "learning_rate": 4.837747168224222e-07,
+      "loss": 0.8661,
+      "step": 6586
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4126012325286865,
+      "learning_rate": 4.828644668220134e-07,
+      "loss": 0.7636,
+      "step": 6587
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.2967067956924438,
+      "learning_rate": 4.819550305223225e-07,
+      "loss": 0.7857,
+      "step": 6588
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5950155258178711,
+      "learning_rate": 4.810464080871719e-07,
+      "loss": 1.0193,
+      "step": 6589
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5953959822654724,
+      "learning_rate": 4.80138599680236e-07,
+      "loss": 1.0089,
+      "step": 6590
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5779624581336975,
+      "learning_rate": 4.792316054650458e-07,
+      "loss": 1.0029,
+      "step": 6591
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3855009078979492,
+      "learning_rate": 4.78325425604983e-07,
+      "loss": 0.8103,
+      "step": 6592
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.3548328876495361,
+      "learning_rate": 4.774200602632851e-07,
+      "loss": 0.8399,
+      "step": 6593
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.4260209798812866,
+      "learning_rate": 4.765155096030405e-07,
+      "loss": 0.7316,
+      "step": 6594
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.6277704238891602,
+      "learning_rate": 4.7561177378719016e-07,
+      "loss": 0.8213,
+      "step": 6595
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.244420289993286,
+      "learning_rate": 4.7470885297853265e-07,
+      "loss": 0.8075,
+      "step": 6596
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5754569172859192,
+      "learning_rate": 4.738067473397157e-07,
+      "loss": 1.0183,
+      "step": 6597
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.265709638595581,
+      "learning_rate": 4.7290545703324033e-07,
+      "loss": 0.8252,
+      "step": 6598
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5772496461868286,
+      "learning_rate": 4.720049822214634e-07,
+      "loss": 1.0039,
+      "step": 6599
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.259501576423645,
+      "learning_rate": 4.7110532306659184e-07,
+      "loss": 0.8279,
+      "step": 6600
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5901740789413452,
+      "learning_rate": 4.702064797306888e-07,
+      "loss": 1.013,
+      "step": 6601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4064513444900513,
+      "learning_rate": 4.693084523756675e-07,
+      "loss": 0.8269,
+      "step": 6602
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3586395978927612,
+      "learning_rate": 4.684112411632946e-07,
+      "loss": 0.7783,
+      "step": 6603
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3950225114822388,
+      "learning_rate": 4.675148462551926e-07,
+      "loss": 0.8223,
+      "step": 6604
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.544482946395874,
+      "learning_rate": 4.666192678128334e-07,
+      "loss": 0.8512,
+      "step": 6605
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5958073735237122,
+      "learning_rate": 4.6572450599754403e-07,
+      "loss": 0.9788,
+      "step": 6606
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2472193241119385,
+      "learning_rate": 4.6483056097050286e-07,
+      "loss": 0.744,
+      "step": 6607
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3286595344543457,
+      "learning_rate": 4.639374328927426e-07,
+      "loss": 0.8314,
+      "step": 6608
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2471060752868652,
+      "learning_rate": 4.630451219251486e-07,
+      "loss": 0.8014,
+      "step": 6609
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.424607515335083,
+      "learning_rate": 4.621536282284572e-07,
+      "loss": 0.7678,
+      "step": 6610
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3346346616744995,
+      "learning_rate": 4.6126295196326e-07,
+      "loss": 0.9032,
+      "step": 6611
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3138227462768555,
+      "learning_rate": 4.6037309329000077e-07,
+      "loss": 0.7831,
+      "step": 6612
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4867764711380005,
+      "learning_rate": 4.5948405236897366e-07,
+      "loss": 0.8173,
+      "step": 6613
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5763893127441406,
+      "learning_rate": 4.5859582936032877e-07,
+      "loss": 0.994,
+      "step": 6614
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3996405601501465,
+      "learning_rate": 4.577084244240665e-07,
+      "loss": 0.8267,
+      "step": 6615
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5836498737335205,
+      "learning_rate": 4.568218377200423e-07,
+      "loss": 1.0077,
+      "step": 6616
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5726943016052246,
+      "learning_rate": 4.5593606940796175e-07,
+      "loss": 0.9891,
+      "step": 6617
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4963549375534058,
+      "learning_rate": 4.550511196473828e-07,
+      "loss": 0.8111,
+      "step": 6618
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3497686386108398,
+      "learning_rate": 4.541669885977196e-07,
+      "loss": 0.814,
+      "step": 6619
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3202489614486694,
+      "learning_rate": 4.532836764182347e-07,
+      "loss": 0.8483,
+      "step": 6620
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.237574577331543,
+      "learning_rate": 4.524011832680458e-07,
+      "loss": 0.8603,
+      "step": 6621
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3934351205825806,
+      "learning_rate": 4.5151950930612144e-07,
+      "loss": 0.8072,
+      "step": 6622
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4235222339630127,
+      "learning_rate": 4.5063865469128276e-07,
+      "loss": 0.8604,
+      "step": 6623
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4091416597366333,
+      "learning_rate": 4.497586195822046e-07,
+      "loss": 0.768,
+      "step": 6624
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.347322940826416,
+      "learning_rate": 4.488794041374128e-07,
+      "loss": 0.8028,
+      "step": 6625
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4108129739761353,
+      "learning_rate": 4.480010085152875e-07,
+      "loss": 0.8736,
+      "step": 6626
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3151254653930664,
+      "learning_rate": 4.471234328740581e-07,
+      "loss": 0.7915,
+      "step": 6627
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3352562189102173,
+      "learning_rate": 4.4624667737180827e-07,
+      "loss": 0.8675,
+      "step": 6628
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.292142629623413,
+      "learning_rate": 4.4537074216647434e-07,
+      "loss": 0.801,
+      "step": 6629
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5678266882896423,
+      "learning_rate": 4.4449562741584353e-07,
+      "loss": 1.0131,
+      "step": 6630
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5777455568313599,
+      "learning_rate": 4.436213332775557e-07,
+      "loss": 1.0113,
+      "step": 6631
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3839454650878906,
+      "learning_rate": 4.42747859909104e-07,
+      "loss": 0.7614,
+      "step": 6632
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4230449199676514,
+      "learning_rate": 4.4187520746783177e-07,
+      "loss": 0.8196,
+      "step": 6633
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5652325749397278,
+      "learning_rate": 4.410033761109361e-07,
+      "loss": 1.0004,
+      "step": 6634
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.58445143699646,
+      "learning_rate": 4.401323659954665e-07,
+      "loss": 1.0276,
+      "step": 6635
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.48611319065094,
+      "learning_rate": 4.392621772783212e-07,
+      "loss": 0.8833,
+      "step": 6636
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.486163854598999,
+      "learning_rate": 4.383928101162549e-07,
+      "loss": 0.8058,
+      "step": 6637
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.384461522102356,
+      "learning_rate": 4.3752426466587174e-07,
+      "loss": 0.7817,
+      "step": 6638
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2952908277511597,
+      "learning_rate": 4.3665654108362877e-07,
+      "loss": 0.7903,
+      "step": 6639
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 2.0401663780212402,
+      "learning_rate": 4.357896395258343e-07,
+      "loss": 0.8037,
+      "step": 6640
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5613390803337097,
+      "learning_rate": 4.349235601486479e-07,
+      "loss": 0.9995,
+      "step": 6641
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2889909744262695,
+      "learning_rate": 4.340583031080836e-07,
+      "loss": 0.8079,
+      "step": 6642
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4698325395584106,
+      "learning_rate": 4.3319386856000565e-07,
+      "loss": 0.8408,
+      "step": 6643
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.356656789779663,
+      "learning_rate": 4.323302566601284e-07,
+      "loss": 0.8153,
+      "step": 6644
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3138017654418945,
+      "learning_rate": 4.3146746756402236e-07,
+      "loss": 0.7765,
+      "step": 6645
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4347097873687744,
+      "learning_rate": 4.30605501427106e-07,
+      "loss": 0.7678,
+      "step": 6646
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2583537101745605,
+      "learning_rate": 4.297443584046507e-07,
+      "loss": 0.8272,
+      "step": 6647
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.232646107673645,
+      "learning_rate": 4.28884038651779e-07,
+      "loss": 0.8525,
+      "step": 6648
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2890766859054565,
+      "learning_rate": 4.280245423234675e-07,
+      "loss": 0.8163,
+      "step": 6649
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.8059073686599731,
+      "learning_rate": 4.2716586957454174e-07,
+      "loss": 0.7362,
+      "step": 6650
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4294980764389038,
+      "learning_rate": 4.263080205596798e-07,
+      "loss": 0.8194,
+      "step": 6651
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3339768648147583,
+      "learning_rate": 4.254509954334124e-07,
+      "loss": 0.8487,
+      "step": 6652
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5736918449401855,
+      "learning_rate": 4.2459479435011995e-07,
+      "loss": 0.9966,
+      "step": 6653
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5806509256362915,
+      "learning_rate": 4.237394174640369e-07,
+      "loss": 0.98,
+      "step": 6654
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4130682945251465,
+      "learning_rate": 4.228848649292472e-07,
+      "loss": 0.8548,
+      "step": 6655
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3013921976089478,
+      "learning_rate": 4.22031136899686e-07,
+      "loss": 0.7888,
+      "step": 6656
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5785076022148132,
+      "learning_rate": 4.21178233529142e-07,
+      "loss": 0.9929,
+      "step": 6657
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3772045373916626,
+      "learning_rate": 4.2032615497125396e-07,
+      "loss": 0.8178,
+      "step": 6658
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6000142693519592,
+      "learning_rate": 4.1947490137951076e-07,
+      "loss": 1.0285,
+      "step": 6659
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.4070703983306885,
+      "learning_rate": 4.186244729072564e-07,
+      "loss": 0.8422,
+      "step": 6660
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5965368151664734,
+      "learning_rate": 4.177748697076822e-07,
+      "loss": 0.9984,
+      "step": 6661
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5701125860214233,
+      "learning_rate": 4.169260919338347e-07,
+      "loss": 1.0252,
+      "step": 6662
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3271576166152954,
+      "learning_rate": 4.160781397386082e-07,
+      "loss": 0.8123,
+      "step": 6663
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5736581683158875,
+      "learning_rate": 4.1523101327474945e-07,
+      "loss": 1.008,
+      "step": 6664
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3154222965240479,
+      "learning_rate": 4.143847126948586e-07,
+      "loss": 0.7941,
+      "step": 6665
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3449320793151855,
+      "learning_rate": 4.135392381513831e-07,
+      "loss": 0.7655,
+      "step": 6666
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2922335863113403,
+      "learning_rate": 4.126945897966256e-07,
+      "loss": 0.821,
+      "step": 6667
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3574022054672241,
+      "learning_rate": 4.118507677827371e-07,
+      "loss": 0.7743,
+      "step": 6668
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5885167717933655,
+      "learning_rate": 4.110077722617195e-07,
+      "loss": 0.9945,
+      "step": 6669
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.435173511505127,
+      "learning_rate": 4.1016560338542954e-07,
+      "loss": 0.8054,
+      "step": 6670
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5825942754745483,
+      "learning_rate": 4.093242613055709e-07,
+      "loss": 0.9925,
+      "step": 6671
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3037723302841187,
+      "learning_rate": 4.084837461737007e-07,
+      "loss": 0.8733,
+      "step": 6672
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.295261263847351,
+      "learning_rate": 4.0764405814122556e-07,
+      "loss": 0.8388,
+      "step": 6673
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.2873882055282593,
+      "learning_rate": 4.0680519735940394e-07,
+      "loss": 0.8154,
+      "step": 6674
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5703015327453613,
+      "learning_rate": 4.0596716397934665e-07,
+      "loss": 0.9786,
+      "step": 6675
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 1.3539124727249146,
+      "learning_rate": 4.0512995815201185e-07,
+      "loss": 0.7772,
+      "step": 6676
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5914307236671448,
+      "learning_rate": 4.0429358002821275e-07,
+      "loss": 1.0371,
+      "step": 6677
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2136893272399902,
+      "learning_rate": 4.034580297586105e-07,
+      "loss": 0.8294,
+      "step": 6678
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2806837558746338,
+      "learning_rate": 4.0262330749371805e-07,
+      "loss": 0.7871,
+      "step": 6679
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5669032335281372,
+      "learning_rate": 4.017894133839001e-07,
+      "loss": 1.0112,
+      "step": 6680
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.659733533859253,
+      "learning_rate": 4.009563475793704e-07,
+      "loss": 0.8559,
+      "step": 6681
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5894060730934143,
+      "learning_rate": 4.001241102301956e-07,
+      "loss": 0.9865,
+      "step": 6682
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2868345975875854,
+      "learning_rate": 3.992927014862913e-07,
+      "loss": 0.8759,
+      "step": 6683
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3112229108810425,
+      "learning_rate": 3.9846212149742433e-07,
+      "loss": 0.8151,
+      "step": 6684
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5609560012817383,
+      "learning_rate": 3.9763237041321233e-07,
+      "loss": 0.9837,
+      "step": 6685
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3978896141052246,
+      "learning_rate": 3.9680344838312456e-07,
+      "loss": 0.7832,
+      "step": 6686
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3391902446746826,
+      "learning_rate": 3.959753555564794e-07,
+      "loss": 0.8666,
+      "step": 6687
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2618529796600342,
+      "learning_rate": 3.9514809208244685e-07,
+      "loss": 0.7492,
+      "step": 6688
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3875435590744019,
+      "learning_rate": 3.943216581100462e-07,
+      "loss": 0.8333,
+      "step": 6689
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5726674199104309,
+      "learning_rate": 3.934960537881499e-07,
+      "loss": 0.9938,
+      "step": 6690
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4392648935317993,
+      "learning_rate": 3.9267127926547856e-07,
+      "loss": 0.8367,
+      "step": 6691
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2874377965927124,
+      "learning_rate": 3.918473346906032e-07,
+      "loss": 0.7774,
+      "step": 6692
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5791847705841064,
+      "learning_rate": 3.9102422021194853e-07,
+      "loss": 1.0162,
+      "step": 6693
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3086788654327393,
+      "learning_rate": 3.902019359777848e-07,
+      "loss": 0.8403,
+      "step": 6694
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3940050601959229,
+      "learning_rate": 3.893804821362379e-07,
+      "loss": 0.8152,
+      "step": 6695
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3108506202697754,
+      "learning_rate": 3.885598588352807e-07,
+      "loss": 0.8896,
+      "step": 6696
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.223982572555542,
+      "learning_rate": 3.8774006622273597e-07,
+      "loss": 0.8459,
+      "step": 6697
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2845499515533447,
+      "learning_rate": 3.869211044462806e-07,
+      "loss": 0.7992,
+      "step": 6698
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.778281807899475,
+      "learning_rate": 3.861029736534366e-07,
+      "loss": 0.8194,
+      "step": 6699
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4520373344421387,
+      "learning_rate": 3.8528567399158225e-07,
+      "loss": 0.8381,
+      "step": 6700
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3529504537582397,
+      "learning_rate": 3.844692056079413e-07,
+      "loss": 0.7957,
+      "step": 6701
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3336926698684692,
+      "learning_rate": 3.8365356864958903e-07,
+      "loss": 0.8476,
+      "step": 6702
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5619142055511475,
+      "learning_rate": 3.8283876326345283e-07,
+      "loss": 0.7981,
+      "step": 6703
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.5329457521438599,
+      "learning_rate": 3.8202478959630694e-07,
+      "loss": 0.9054,
+      "step": 6704
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.263308048248291,
+      "learning_rate": 3.812116477947803e-07,
+      "loss": 0.8058,
+      "step": 6705
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3427629470825195,
+      "learning_rate": 3.8039933800534724e-07,
+      "loss": 0.8177,
+      "step": 6706
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3593018054962158,
+      "learning_rate": 3.7958786037433413e-07,
+      "loss": 0.8308,
+      "step": 6707
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5958836674690247,
+      "learning_rate": 3.7877721504791963e-07,
+      "loss": 1.0257,
+      "step": 6708
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4012852907180786,
+      "learning_rate": 3.7796740217212804e-07,
+      "loss": 0.7932,
+      "step": 6709
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.581680178642273,
+      "learning_rate": 3.771584218928387e-07,
+      "loss": 0.9928,
+      "step": 6710
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3506653308868408,
+      "learning_rate": 3.7635027435577734e-07,
+      "loss": 0.7941,
+      "step": 6711
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3956079483032227,
+      "learning_rate": 3.755429597065197e-07,
+      "loss": 0.8324,
+      "step": 6712
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3487708568572998,
+      "learning_rate": 3.747364780904944e-07,
+      "loss": 0.8384,
+      "step": 6713
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.266674280166626,
+      "learning_rate": 3.7393082965297754e-07,
+      "loss": 0.7988,
+      "step": 6714
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3807626962661743,
+      "learning_rate": 3.731260145390941e-07,
+      "loss": 0.7944,
+      "step": 6715
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3838474750518799,
+      "learning_rate": 3.7232203289382374e-07,
+      "loss": 0.8523,
+      "step": 6716
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5783218145370483,
+      "learning_rate": 3.715188848619894e-07,
+      "loss": 1.0002,
+      "step": 6717
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.445943832397461,
+      "learning_rate": 3.707165705882704e-07,
+      "loss": 0.7776,
+      "step": 6718
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3075182437896729,
+      "learning_rate": 3.6991509021719163e-07,
+      "loss": 0.7801,
+      "step": 6719
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2799383401870728,
+      "learning_rate": 3.6911444389312766e-07,
+      "loss": 0.8173,
+      "step": 6720
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3365834951400757,
+      "learning_rate": 3.683146317603059e-07,
+      "loss": 0.7909,
+      "step": 6721
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4295783042907715,
+      "learning_rate": 3.6751565396280055e-07,
+      "loss": 0.7915,
+      "step": 6722
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4220350980758667,
+      "learning_rate": 3.6671751064453756e-07,
+      "loss": 0.8458,
+      "step": 6723
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5898751020431519,
+      "learning_rate": 3.659202019492908e-07,
+      "loss": 1.0206,
+      "step": 6724
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4205679893493652,
+      "learning_rate": 3.6512372802068486e-07,
+      "loss": 0.8912,
+      "step": 6725
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2325876951217651,
+      "learning_rate": 3.643280890021933e-07,
+      "loss": 0.7779,
+      "step": 6726
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5636802911758423,
+      "learning_rate": 3.6353328503714036e-07,
+      "loss": 1.0226,
+      "step": 6727
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2777782678604126,
+      "learning_rate": 3.6273931626869927e-07,
+      "loss": 0.8262,
+      "step": 6728
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5906734466552734,
+      "learning_rate": 3.6194618283989183e-07,
+      "loss": 0.9706,
+      "step": 6729
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4136848449707031,
+      "learning_rate": 3.6115388489359137e-07,
+      "loss": 0.828,
+      "step": 6730
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3682491779327393,
+      "learning_rate": 3.603624225725194e-07,
+      "loss": 0.8058,
+      "step": 6731
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5807645916938782,
+      "learning_rate": 3.5957179601924574e-07,
+      "loss": 1.0248,
+      "step": 6732
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.351361870765686,
+      "learning_rate": 3.587820053761931e-07,
+      "loss": 0.8436,
+      "step": 6733
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3193169832229614,
+      "learning_rate": 3.5799305078563104e-07,
+      "loss": 0.8233,
+      "step": 6734
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3026363849639893,
+      "learning_rate": 3.57204932389677e-07,
+      "loss": 0.793,
+      "step": 6735
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3742705583572388,
+      "learning_rate": 3.5641765033030294e-07,
+      "loss": 0.8025,
+      "step": 6736
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3736248016357422,
+      "learning_rate": 3.5563120474932435e-07,
+      "loss": 0.8251,
+      "step": 6737
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5742596983909607,
+      "learning_rate": 3.548455957884106e-07,
+      "loss": 1.0106,
+      "step": 6738
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4491777420043945,
+      "learning_rate": 3.54060823589078e-07,
+      "loss": 0.8063,
+      "step": 6739
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2857933044433594,
+      "learning_rate": 3.532768882926918e-07,
+      "loss": 0.7335,
+      "step": 6740
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5833033323287964,
+      "learning_rate": 3.5249379004046847e-07,
+      "loss": 1.0348,
+      "step": 6741
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3207073211669922,
+      "learning_rate": 3.5171152897347125e-07,
+      "loss": 0.7888,
+      "step": 6742
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5710121393203735,
+      "learning_rate": 3.5093010523261583e-07,
+      "loss": 1.0037,
+      "step": 6743
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5740703344345093,
+      "learning_rate": 3.5014951895866347e-07,
+      "loss": 1.0113,
+      "step": 6744
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3719851970672607,
+      "learning_rate": 3.4936977029222607e-07,
+      "loss": 0.8097,
+      "step": 6745
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5716421008110046,
+      "learning_rate": 3.485908593737658e-07,
+      "loss": 0.9981,
+      "step": 6746
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5837103724479675,
+      "learning_rate": 3.478127863435926e-07,
+      "loss": 1.0108,
+      "step": 6747
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3211997747421265,
+      "learning_rate": 3.4703555134186496e-07,
+      "loss": 0.7689,
+      "step": 6748
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5734263062477112,
+      "learning_rate": 3.462591545085925e-07,
+      "loss": 0.9822,
+      "step": 6749
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3799546957015991,
+      "learning_rate": 3.454835959836317e-07,
+      "loss": 0.8455,
+      "step": 6750
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.2330957651138306,
+      "learning_rate": 3.447088759066891e-07,
+      "loss": 0.7562,
+      "step": 6751
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.4328869581222534,
+      "learning_rate": 3.439349944173209e-07,
+      "loss": 0.8034,
+      "step": 6752
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.3302844762802124,
+      "learning_rate": 3.4316195165492947e-07,
+      "loss": 0.8624,
+      "step": 6753
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 1.380900263786316,
+      "learning_rate": 3.4238974775876954e-07,
+      "loss": 0.8312,
+      "step": 6754
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5682543516159058,
+      "learning_rate": 3.4161838286794256e-07,
+      "loss": 0.977,
+      "step": 6755
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3060401678085327,
+      "learning_rate": 3.4084785712139967e-07,
+      "loss": 0.8211,
+      "step": 6756
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4497214555740356,
+      "learning_rate": 3.4007817065794146e-07,
+      "loss": 0.875,
+      "step": 6757
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.338037133216858,
+      "learning_rate": 3.3930932361621426e-07,
+      "loss": 0.7985,
+      "step": 6758
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5826765894889832,
+      "learning_rate": 3.385413161347173e-07,
+      "loss": 1.0061,
+      "step": 6759
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5839805006980896,
+      "learning_rate": 3.37774148351796e-07,
+      "loss": 1.0065,
+      "step": 6760
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3464354276657104,
+      "learning_rate": 3.3700782040564594e-07,
+      "loss": 0.8402,
+      "step": 6761
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.566116988658905,
+      "learning_rate": 3.3624233243431004e-07,
+      "loss": 0.999,
+      "step": 6762
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.322358250617981,
+      "learning_rate": 3.3547768457568017e-07,
+      "loss": 0.7742,
+      "step": 6763
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4888125658035278,
+      "learning_rate": 3.3471387696749844e-07,
+      "loss": 0.7648,
+      "step": 6764
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2581281661987305,
+      "learning_rate": 3.3395090974735423e-07,
+      "loss": 0.7826,
+      "step": 6765
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5784320831298828,
+      "learning_rate": 3.331887830526853e-07,
+      "loss": 0.9902,
+      "step": 6766
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4463601112365723,
+      "learning_rate": 3.324274970207775e-07,
+      "loss": 0.7736,
+      "step": 6767
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.292305588722229,
+      "learning_rate": 3.3166705178876825e-07,
+      "loss": 0.8247,
+      "step": 6768
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.31169593334198,
+      "learning_rate": 3.309074474936402e-07,
+      "loss": 0.7859,
+      "step": 6769
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5933763384819031,
+      "learning_rate": 3.3014868427222513e-07,
+      "loss": 0.9848,
+      "step": 6770
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5754019618034363,
+      "learning_rate": 3.2939076226120524e-07,
+      "loss": 1.0041,
+      "step": 6771
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3519536256790161,
+      "learning_rate": 3.2863368159710975e-07,
+      "loss": 0.8206,
+      "step": 6772
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3812823295593262,
+      "learning_rate": 3.2787744241631556e-07,
+      "loss": 0.8345,
+      "step": 6773
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3603386878967285,
+      "learning_rate": 3.2712204485504994e-07,
+      "loss": 0.8339,
+      "step": 6774
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5822895169258118,
+      "learning_rate": 3.263674890493873e-07,
+      "loss": 0.9894,
+      "step": 6775
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3632402420043945,
+      "learning_rate": 3.256137751352495e-07,
+      "loss": 0.82,
+      "step": 6776
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5686249136924744,
+      "learning_rate": 3.2486090324840957e-07,
+      "loss": 1.0228,
+      "step": 6777
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3208601474761963,
+      "learning_rate": 3.241088735244857e-07,
+      "loss": 0.7781,
+      "step": 6778
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4004830121994019,
+      "learning_rate": 3.2335768609894735e-07,
+      "loss": 0.7762,
+      "step": 6779
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2899291515350342,
+      "learning_rate": 3.226073411071096e-07,
+      "loss": 0.8287,
+      "step": 6780
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.376309871673584,
+      "learning_rate": 3.21857838684137e-07,
+      "loss": 0.8594,
+      "step": 6781
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3831349611282349,
+      "learning_rate": 3.2110917896504277e-07,
+      "loss": 0.8323,
+      "step": 6782
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3931845426559448,
+      "learning_rate": 3.2036136208468684e-07,
+      "loss": 0.7878,
+      "step": 6783
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3016390800476074,
+      "learning_rate": 3.196143881777802e-07,
+      "loss": 0.8479,
+      "step": 6784
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5726263523101807,
+      "learning_rate": 3.188682573788787e-07,
+      "loss": 0.8475,
+      "step": 6785
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4436839818954468,
+      "learning_rate": 3.18122969822387e-07,
+      "loss": 0.7664,
+      "step": 6786
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.58822762966156,
+      "learning_rate": 3.173785256425599e-07,
+      "loss": 0.7766,
+      "step": 6787
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4259848594665527,
+      "learning_rate": 3.166349249734979e-07,
+      "loss": 0.8028,
+      "step": 6788
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5827158689498901,
+      "learning_rate": 3.158921679491522e-07,
+      "loss": 1.0059,
+      "step": 6789
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.5484415292739868,
+      "learning_rate": 3.151502547033186e-07,
+      "loss": 0.8375,
+      "step": 6790
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3576515913009644,
+      "learning_rate": 3.144091853696435e-07,
+      "loss": 0.8014,
+      "step": 6791
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3574150800704956,
+      "learning_rate": 3.1366896008162064e-07,
+      "loss": 0.8083,
+      "step": 6792
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.9565869569778442,
+      "learning_rate": 3.129295789725906e-07,
+      "loss": 0.7889,
+      "step": 6793
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5611687898635864,
+      "learning_rate": 3.121910421757446e-07,
+      "loss": 1.0,
+      "step": 6794
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5698570609092712,
+      "learning_rate": 3.1145334982411845e-07,
+      "loss": 0.9855,
+      "step": 6795
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5973178148269653,
+      "learning_rate": 3.1071650205059755e-07,
+      "loss": 0.9792,
+      "step": 6796
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5527148842811584,
+      "learning_rate": 3.0998049898791685e-07,
+      "loss": 0.9883,
+      "step": 6797
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4374927282333374,
+      "learning_rate": 3.0924534076865464e-07,
+      "loss": 0.8285,
+      "step": 6798
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3675631284713745,
+      "learning_rate": 3.0851102752524176e-07,
+      "loss": 0.8148,
+      "step": 6799
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3803668022155762,
+      "learning_rate": 3.07777559389954e-07,
+      "loss": 0.7963,
+      "step": 6800
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5835098624229431,
+      "learning_rate": 3.070449364949152e-07,
+      "loss": 1.0038,
+      "step": 6801
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2668713331222534,
+      "learning_rate": 3.0631315897209854e-07,
+      "loss": 0.8495,
+      "step": 6802
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 4.768804550170898,
+      "learning_rate": 3.055822269533237e-07,
+      "loss": 0.8248,
+      "step": 6803
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2721906900405884,
+      "learning_rate": 3.048521405702565e-07,
+      "loss": 0.8008,
+      "step": 6804
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5588178634643555,
+      "learning_rate": 3.0412289995441437e-07,
+      "loss": 1.0126,
+      "step": 6805
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3740745782852173,
+      "learning_rate": 3.0339450523715896e-07,
+      "loss": 0.8033,
+      "step": 6806
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5782884955406189,
+      "learning_rate": 3.026669565497009e-07,
+      "loss": 1.0009,
+      "step": 6807
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2627745866775513,
+      "learning_rate": 3.0194025402309757e-07,
+      "loss": 0.8482,
+      "step": 6808
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5822926163673401,
+      "learning_rate": 3.012143977882553e-07,
+      "loss": 1.0323,
+      "step": 6809
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.377341389656067,
+      "learning_rate": 3.00489387975928e-07,
+      "loss": 0.8336,
+      "step": 6810
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.357002854347229,
+      "learning_rate": 2.997652247167138e-07,
+      "loss": 0.8086,
+      "step": 6811
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3811140060424805,
+      "learning_rate": 2.990419081410639e-07,
+      "loss": 0.8669,
+      "step": 6812
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.351725697517395,
+      "learning_rate": 2.9831943837927257e-07,
+      "loss": 0.8233,
+      "step": 6813
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3707751035690308,
+      "learning_rate": 2.9759781556148224e-07,
+      "loss": 0.8013,
+      "step": 6814
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3664686679840088,
+      "learning_rate": 2.968770398176851e-07,
+      "loss": 0.8333,
+      "step": 6815
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2904728651046753,
+      "learning_rate": 2.961571112777173e-07,
+      "loss": 0.787,
+      "step": 6816
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5699200630187988,
+      "learning_rate": 2.9543803007126613e-07,
+      "loss": 1.0071,
+      "step": 6817
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.6002026796340942,
+      "learning_rate": 2.947197963278631e-07,
+      "loss": 0.7561,
+      "step": 6818
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.276980996131897,
+      "learning_rate": 2.940024101768879e-07,
+      "loss": 0.8015,
+      "step": 6819
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.323989987373352,
+      "learning_rate": 2.932858717475695e-07,
+      "loss": 0.8806,
+      "step": 6820
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.57606440782547,
+      "learning_rate": 2.9257018116898074e-07,
+      "loss": 1.0112,
+      "step": 6821
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5696162581443787,
+      "learning_rate": 2.918553385700451e-07,
+      "loss": 1.0257,
+      "step": 6822
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3403502702713013,
+      "learning_rate": 2.911413440795313e-07,
+      "loss": 0.8085,
+      "step": 6823
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3241205215454102,
+      "learning_rate": 2.904281978260548e-07,
+      "loss": 0.8228,
+      "step": 6824
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3769961595535278,
+      "learning_rate": 2.897158999380806e-07,
+      "loss": 0.8137,
+      "step": 6825
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5665758848190308,
+      "learning_rate": 2.8900445054391877e-07,
+      "loss": 1.0218,
+      "step": 6826
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3083233833312988,
+      "learning_rate": 2.8829384977172747e-07,
+      "loss": 0.808,
+      "step": 6827
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.3235533237457275,
+      "learning_rate": 2.875840977495126e-07,
+      "loss": 0.8174,
+      "step": 6828
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.571987509727478,
+      "learning_rate": 2.86875194605124e-07,
+      "loss": 0.9729,
+      "step": 6829
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.4195849895477295,
+      "learning_rate": 2.861671404662636e-07,
+      "loss": 0.8216,
+      "step": 6830
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3538826704025269,
+      "learning_rate": 2.85459935460477e-07,
+      "loss": 0.8326,
+      "step": 6831
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 2.154574394226074,
+      "learning_rate": 2.847535797151563e-07,
+      "loss": 0.8462,
+      "step": 6832
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3221744298934937,
+      "learning_rate": 2.8404807335754346e-07,
+      "loss": 0.7934,
+      "step": 6833
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3809453248977661,
+      "learning_rate": 2.8334341651472477e-07,
+      "loss": 0.8213,
+      "step": 6834
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3923174142837524,
+      "learning_rate": 2.8263960931363524e-07,
+      "loss": 0.9045,
+      "step": 6835
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3349593877792358,
+      "learning_rate": 2.819366518810568e-07,
+      "loss": 0.7283,
+      "step": 6836
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5698980093002319,
+      "learning_rate": 2.8123454434361597e-07,
+      "loss": 1.002,
+      "step": 6837
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3029247522354126,
+      "learning_rate": 2.8053328682778936e-07,
+      "loss": 0.8076,
+      "step": 6838
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2938110828399658,
+      "learning_rate": 2.7983287945989813e-07,
+      "loss": 0.8523,
+      "step": 6839
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.310308575630188,
+      "learning_rate": 2.7913332236611243e-07,
+      "loss": 0.8446,
+      "step": 6840
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5739769339561462,
+      "learning_rate": 2.784346156724471e-07,
+      "loss": 1.0021,
+      "step": 6841
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3116117715835571,
+      "learning_rate": 2.7773675950476366e-07,
+      "loss": 0.8268,
+      "step": 6842
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.294985055923462,
+      "learning_rate": 2.770397539887737e-07,
+      "loss": 0.7591,
+      "step": 6843
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2850916385650635,
+      "learning_rate": 2.763435992500318e-07,
+      "loss": 0.7875,
+      "step": 6844
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2768737077713013,
+      "learning_rate": 2.7564829541394057e-07,
+      "loss": 0.7414,
+      "step": 6845
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5822781324386597,
+      "learning_rate": 2.7495384260575077e-07,
+      "loss": 1.0015,
+      "step": 6846
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.6269055604934692,
+      "learning_rate": 2.7426024095055794e-07,
+      "loss": 0.8429,
+      "step": 6847
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5813559293746948,
+      "learning_rate": 2.7356749057330433e-07,
+      "loss": 1.0035,
+      "step": 6848
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4396553039550781,
+      "learning_rate": 2.7287559159878076e-07,
+      "loss": 0.8066,
+      "step": 6849
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.306899905204773,
+      "learning_rate": 2.72184544151623e-07,
+      "loss": 0.806,
+      "step": 6850
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4060832262039185,
+      "learning_rate": 2.7149434835631426e-07,
+      "loss": 0.8535,
+      "step": 6851
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3572025299072266,
+      "learning_rate": 2.708050043371824e-07,
+      "loss": 0.8266,
+      "step": 6852
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3424166440963745,
+      "learning_rate": 2.7011651221840574e-07,
+      "loss": 0.7785,
+      "step": 6853
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.317958950996399,
+      "learning_rate": 2.694288721240046e-07,
+      "loss": 0.8078,
+      "step": 6854
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4293237924575806,
+      "learning_rate": 2.687420841778493e-07,
+      "loss": 0.856,
+      "step": 6855
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3308484554290771,
+      "learning_rate": 2.6805614850365536e-07,
+      "loss": 0.8129,
+      "step": 6856
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2431128025054932,
+      "learning_rate": 2.673710652249839e-07,
+      "loss": 0.8346,
+      "step": 6857
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3587912321090698,
+      "learning_rate": 2.6668683446524456e-07,
+      "loss": 0.8292,
+      "step": 6858
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.5992048978805542,
+      "learning_rate": 2.660034563476921e-07,
+      "loss": 0.8449,
+      "step": 6859
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3364919424057007,
+      "learning_rate": 2.653209309954263e-07,
+      "loss": 0.8107,
+      "step": 6860
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.274157166481018,
+      "learning_rate": 2.6463925853139673e-07,
+      "loss": 0.7749,
+      "step": 6861
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5816813707351685,
+      "learning_rate": 2.639584390783956e-07,
+      "loss": 1.0283,
+      "step": 6862
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3831883668899536,
+      "learning_rate": 2.632784727590654e-07,
+      "loss": 0.8154,
+      "step": 6863
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2934236526489258,
+      "learning_rate": 2.6259935969589203e-07,
+      "loss": 0.7857,
+      "step": 6864
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2972288131713867,
+      "learning_rate": 2.6192110001120707e-07,
+      "loss": 0.795,
+      "step": 6865
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4726603031158447,
+      "learning_rate": 2.612436938271917e-07,
+      "loss": 0.8265,
+      "step": 6866
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4592944383621216,
+      "learning_rate": 2.605671412658706e-07,
+      "loss": 0.8624,
+      "step": 6867
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3764792680740356,
+      "learning_rate": 2.598914424491161e-07,
+      "loss": 0.8404,
+      "step": 6868
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.274970531463623,
+      "learning_rate": 2.592165974986455e-07,
+      "loss": 0.7999,
+      "step": 6869
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2845547199249268,
+      "learning_rate": 2.5854260653602306e-07,
+      "loss": 0.8051,
+      "step": 6870
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5706745386123657,
+      "learning_rate": 2.578694696826595e-07,
+      "loss": 1.0106,
+      "step": 6871
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.357849359512329,
+      "learning_rate": 2.5719718705981055e-07,
+      "loss": 0.7964,
+      "step": 6872
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.453639030456543,
+      "learning_rate": 2.5652575878858e-07,
+      "loss": 0.8311,
+      "step": 6873
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3325467109680176,
+      "learning_rate": 2.5585518498991604e-07,
+      "loss": 0.8534,
+      "step": 6874
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.331921100616455,
+      "learning_rate": 2.5518546578461256e-07,
+      "loss": 0.7603,
+      "step": 6875
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.346138596534729,
+      "learning_rate": 2.5451660129331204e-07,
+      "loss": 0.7859,
+      "step": 6876
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.374037265777588,
+      "learning_rate": 2.5384859163649966e-07,
+      "loss": 0.8433,
+      "step": 6877
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3412144184112549,
+      "learning_rate": 2.531814369345098e-07,
+      "loss": 0.8,
+      "step": 6878
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3663965463638306,
+      "learning_rate": 2.525151373075202e-07,
+      "loss": 0.8268,
+      "step": 6879
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3086154460906982,
+      "learning_rate": 2.5184969287555595e-07,
+      "loss": 0.8416,
+      "step": 6880
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2484643459320068,
+      "learning_rate": 2.5118510375848834e-07,
+      "loss": 0.773,
+      "step": 6881
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5754777193069458,
+      "learning_rate": 2.5052137007603337e-07,
+      "loss": 1.0092,
+      "step": 6882
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.370824933052063,
+      "learning_rate": 2.498584919477548e-07,
+      "loss": 0.8729,
+      "step": 6883
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.8598681688308716,
+      "learning_rate": 2.491964694930604e-07,
+      "loss": 0.848,
+      "step": 6884
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2854548692703247,
+      "learning_rate": 2.485353028312043e-07,
+      "loss": 0.7386,
+      "step": 6885
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5702117681503296,
+      "learning_rate": 2.4787499208128616e-07,
+      "loss": 0.9906,
+      "step": 6886
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.257894515991211,
+      "learning_rate": 2.472155373622537e-07,
+      "loss": 0.8091,
+      "step": 6887
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.359095811843872,
+      "learning_rate": 2.4655693879289787e-07,
+      "loss": 0.7959,
+      "step": 6888
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5989688038825989,
+      "learning_rate": 2.4589919649185556e-07,
+      "loss": 1.0272,
+      "step": 6889
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5671340227127075,
+      "learning_rate": 2.4524231057761136e-07,
+      "loss": 0.9582,
+      "step": 6890
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3591006994247437,
+      "learning_rate": 2.44586281168494e-07,
+      "loss": 0.8832,
+      "step": 6891
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4885302782058716,
+      "learning_rate": 2.439311083826779e-07,
+      "loss": 0.8427,
+      "step": 6892
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5809175372123718,
+      "learning_rate": 2.4327679233818294e-07,
+      "loss": 1.0134,
+      "step": 6893
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.608567714691162,
+      "learning_rate": 2.426233331528771e-07,
+      "loss": 0.8189,
+      "step": 6894
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2741724252700806,
+      "learning_rate": 2.419707309444702e-07,
+      "loss": 0.8318,
+      "step": 6895
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3790541887283325,
+      "learning_rate": 2.413189858305215e-07,
+      "loss": 0.7781,
+      "step": 6896
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3510183095932007,
+      "learning_rate": 2.406680979284337e-07,
+      "loss": 0.8152,
+      "step": 6897
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5664592385292053,
+      "learning_rate": 2.4001806735545365e-07,
+      "loss": 1.0017,
+      "step": 6898
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6047407984733582,
+      "learning_rate": 2.393688942286776e-07,
+      "loss": 0.9983,
+      "step": 6899
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.2971230745315552,
+      "learning_rate": 2.387205786650437e-07,
+      "loss": 0.8213,
+      "step": 6900
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3353599309921265,
+      "learning_rate": 2.3807312078133914e-07,
+      "loss": 0.9018,
+      "step": 6901
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5750572681427002,
+      "learning_rate": 2.3742652069419337e-07,
+      "loss": 1.0088,
+      "step": 6902
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3394012451171875,
+      "learning_rate": 2.3678077852008206e-07,
+      "loss": 0.8115,
+      "step": 6903
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5938864350318909,
+      "learning_rate": 2.3613589437532835e-07,
+      "loss": 1.021,
+      "step": 6904
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.3059725761413574,
+      "learning_rate": 2.3549186837609817e-07,
+      "loss": 0.8159,
+      "step": 6905
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.4141355752944946,
+      "learning_rate": 2.3484870063840483e-07,
+      "loss": 0.8175,
+      "step": 6906
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5680630803108215,
+      "learning_rate": 2.3420639127810618e-07,
+      "loss": 1.0299,
+      "step": 6907
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3419164419174194,
+      "learning_rate": 2.3356494041090527e-07,
+      "loss": 0.8357,
+      "step": 6908
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3455450534820557,
+      "learning_rate": 2.329243481523513e-07,
+      "loss": 0.8198,
+      "step": 6909
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5721918344497681,
+      "learning_rate": 2.3228461461783702e-07,
+      "loss": 1.0102,
+      "step": 6910
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5734810829162598,
+      "learning_rate": 2.3164573992260352e-07,
+      "loss": 1.0108,
+      "step": 6911
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3720628023147583,
+      "learning_rate": 2.3100772418173434e-07,
+      "loss": 0.837,
+      "step": 6912
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5856369137763977,
+      "learning_rate": 2.3037056751015864e-07,
+      "loss": 1.0057,
+      "step": 6913
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3635393381118774,
+      "learning_rate": 2.2973427002265348e-07,
+      "loss": 0.775,
+      "step": 6914
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3986259698867798,
+      "learning_rate": 2.2909883183383774e-07,
+      "loss": 0.8377,
+      "step": 6915
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3070060014724731,
+      "learning_rate": 2.2846425305817766e-07,
+      "loss": 0.789,
+      "step": 6916
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.446347713470459,
+      "learning_rate": 2.27830533809984e-07,
+      "loss": 0.8216,
+      "step": 6917
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5679118633270264,
+      "learning_rate": 2.2719767420341155e-07,
+      "loss": 1.0099,
+      "step": 6918
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3368020057678223,
+      "learning_rate": 2.2656567435246357e-07,
+      "loss": 0.778,
+      "step": 6919
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3643912076950073,
+      "learning_rate": 2.2593453437098457e-07,
+      "loss": 0.7996,
+      "step": 6920
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3949095010757446,
+      "learning_rate": 2.253042543726658e-07,
+      "loss": 0.8362,
+      "step": 6921
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3330509662628174,
+      "learning_rate": 2.2467483447104476e-07,
+      "loss": 0.8012,
+      "step": 6922
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.690533995628357,
+      "learning_rate": 2.2404627477950135e-07,
+      "loss": 0.8721,
+      "step": 6923
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2664989233016968,
+      "learning_rate": 2.2341857541126445e-07,
+      "loss": 0.7453,
+      "step": 6924
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.587594747543335,
+      "learning_rate": 2.2279173647940354e-07,
+      "loss": 1.0164,
+      "step": 6925
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3678851127624512,
+      "learning_rate": 2.2216575809683617e-07,
+      "loss": 0.8266,
+      "step": 6926
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5732263922691345,
+      "learning_rate": 2.215406403763226e-07,
+      "loss": 1.008,
+      "step": 6927
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3083815574645996,
+      "learning_rate": 2.2091638343047062e-07,
+      "loss": 0.7916,
+      "step": 6928
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3049906492233276,
+      "learning_rate": 2.2029298737173188e-07,
+      "loss": 0.8256,
+      "step": 6929
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.570612370967865,
+      "learning_rate": 2.1967045231240103e-07,
+      "loss": 1.0073,
+      "step": 6930
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.5049372911453247,
+      "learning_rate": 2.1904877836462059e-07,
+      "loss": 0.8352,
+      "step": 6931
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.306105375289917,
+      "learning_rate": 2.1842796564037706e-07,
+      "loss": 0.7846,
+      "step": 6932
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5779786109924316,
+      "learning_rate": 2.178080142514999e-07,
+      "loss": 1.0137,
+      "step": 6933
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3615912199020386,
+      "learning_rate": 2.1718892430966698e-07,
+      "loss": 0.789,
+      "step": 6934
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5831188559532166,
+      "learning_rate": 2.1657069592639747e-07,
+      "loss": 1.0007,
+      "step": 6935
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.330067753791809,
+      "learning_rate": 2.1595332921305666e-07,
+      "loss": 0.7918,
+      "step": 6936
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5693126320838928,
+      "learning_rate": 2.1533682428085568e-07,
+      "loss": 1.0186,
+      "step": 6937
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5926644802093506,
+      "learning_rate": 2.1472118124084895e-07,
+      "loss": 0.998,
+      "step": 6938
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3432369232177734,
+      "learning_rate": 2.1410640020393726e-07,
+      "loss": 0.844,
+      "step": 6939
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4235206842422485,
+      "learning_rate": 2.134924812808642e-07,
+      "loss": 0.799,
+      "step": 6940
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3670932054519653,
+      "learning_rate": 2.12879424582218e-07,
+      "loss": 0.7879,
+      "step": 6941
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3395298719406128,
+      "learning_rate": 2.1226723021843475e-07,
+      "loss": 0.7954,
+      "step": 6942
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4146174192428589,
+      "learning_rate": 2.1165589829979072e-07,
+      "loss": 0.8508,
+      "step": 6943
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5842981934547424,
+      "learning_rate": 2.1104542893641055e-07,
+      "loss": 0.9923,
+      "step": 6944
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.9800175428390503,
+      "learning_rate": 2.1043582223826185e-07,
+      "loss": 0.7499,
+      "step": 6945
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5864120721817017,
+      "learning_rate": 2.0982707831515625e-07,
+      "loss": 0.9996,
+      "step": 6946
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5783676505088806,
+      "learning_rate": 2.0921919727675154e-07,
+      "loss": 0.9981,
+      "step": 6947
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3143410682678223,
+      "learning_rate": 2.0861217923254906e-07,
+      "loss": 0.8213,
+      "step": 6948
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4688187837600708,
+      "learning_rate": 2.080060242918941e-07,
+      "loss": 0.7847,
+      "step": 6949
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2981607913970947,
+      "learning_rate": 2.0740073256397876e-07,
+      "loss": 0.798,
+      "step": 6950
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3174781799316406,
+      "learning_rate": 2.0679630415783637e-07,
+      "loss": 0.8571,
+      "step": 6951
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3458749055862427,
+      "learning_rate": 2.0619273918234762e-07,
+      "loss": 0.8335,
+      "step": 6952
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4123737812042236,
+      "learning_rate": 2.0559003774623664e-07,
+      "loss": 0.7791,
+      "step": 6953
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.635353684425354,
+      "learning_rate": 2.0498819995807107e-07,
+      "loss": 0.8218,
+      "step": 6954
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3848708868026733,
+      "learning_rate": 2.0438722592626524e-07,
+      "loss": 0.8114,
+      "step": 6955
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.323546051979065,
+      "learning_rate": 2.0378711575907428e-07,
+      "loss": 0.8527,
+      "step": 6956
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4352055788040161,
+      "learning_rate": 2.0318786956460223e-07,
+      "loss": 0.8371,
+      "step": 6957
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5598241686820984,
+      "learning_rate": 2.0258948745079444e-07,
+      "loss": 1.016,
+      "step": 6958
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3064714670181274,
+      "learning_rate": 2.019919695254402e-07,
+      "loss": 0.8478,
+      "step": 6959
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5873351693153381,
+      "learning_rate": 2.0139531589617567e-07,
+      "loss": 1.0057,
+      "step": 6960
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3471187353134155,
+      "learning_rate": 2.0079952667047875e-07,
+      "loss": 0.846,
+      "step": 6961
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.359866738319397,
+      "learning_rate": 2.002046019556736e-07,
+      "loss": 0.8659,
+      "step": 6962
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2357633113861084,
+      "learning_rate": 1.9961054185892836e-07,
+      "loss": 0.8106,
+      "step": 6963
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5674436688423157,
+      "learning_rate": 1.9901734648725303e-07,
+      "loss": 0.9562,
+      "step": 6964
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.27581787109375,
+      "learning_rate": 1.984250159475054e-07,
+      "loss": 0.8458,
+      "step": 6965
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3112863302230835,
+      "learning_rate": 1.9783355034638574e-07,
+      "loss": 0.8484,
+      "step": 6966
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3155311346054077,
+      "learning_rate": 1.972429497904377e-07,
+      "loss": 0.7943,
+      "step": 6967
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.2606512308120728,
+      "learning_rate": 1.9665321438605e-07,
+      "loss": 0.7831,
+      "step": 6968
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5712215900421143,
+      "learning_rate": 1.960643442394561e-07,
+      "loss": 0.9822,
+      "step": 6969
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.30588960647583,
+      "learning_rate": 1.9547633945673272e-07,
+      "loss": 0.7544,
+      "step": 6970
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3034456968307495,
+      "learning_rate": 1.948892001438002e-07,
+      "loss": 0.8167,
+      "step": 6971
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.359113335609436,
+      "learning_rate": 1.94302926406425e-07,
+      "loss": 0.7857,
+      "step": 6972
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.345962643623352,
+      "learning_rate": 1.9371751835021603e-07,
+      "loss": 0.8215,
+      "step": 6973
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5663273334503174,
+      "learning_rate": 1.9313297608062608e-07,
+      "loss": 1.0035,
+      "step": 6974
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4117906093597412,
+      "learning_rate": 1.925492997029532e-07,
+      "loss": 0.8261,
+      "step": 6975
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4127002954483032,
+      "learning_rate": 1.919664893223383e-07,
+      "loss": 0.8463,
+      "step": 6976
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3577985763549805,
+      "learning_rate": 1.9138454504376625e-07,
+      "loss": 0.7336,
+      "step": 6977
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.3674407005310059,
+      "learning_rate": 1.9080346697206765e-07,
+      "loss": 0.8444,
+      "step": 6978
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5862654447555542,
+      "learning_rate": 1.9022325521191432e-07,
+      "loss": 0.9865,
+      "step": 6979
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5800890922546387,
+      "learning_rate": 1.8964390986782545e-07,
+      "loss": 0.99,
+      "step": 6980
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 1.4235913753509521,
+      "learning_rate": 1.8906543104416086e-07,
+      "loss": 0.8335,
+      "step": 6981
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.574640154838562,
+      "learning_rate": 1.8848781884512613e-07,
+      "loss": 1.0159,
+      "step": 6982
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.578260064125061,
+      "learning_rate": 1.8791107337477022e-07,
+      "loss": 1.0014,
+      "step": 6983
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.54233980178833,
+      "learning_rate": 1.8733519473698557e-07,
+      "loss": 0.8391,
+      "step": 6984
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2860692739486694,
+      "learning_rate": 1.867601830355098e-07,
+      "loss": 0.8263,
+      "step": 6985
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5783355236053467,
+      "learning_rate": 1.8618603837392336e-07,
+      "loss": 0.992,
+      "step": 6986
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7027671337127686,
+      "learning_rate": 1.8561276085564906e-07,
+      "loss": 0.8199,
+      "step": 6987
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.349525809288025,
+      "learning_rate": 1.8504035058395708e-07,
+      "loss": 0.8252,
+      "step": 6988
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2855713367462158,
+      "learning_rate": 1.8446880766195828e-07,
+      "loss": 0.8674,
+      "step": 6989
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5794022083282471,
+      "learning_rate": 1.8389813219260865e-07,
+      "loss": 1.002,
+      "step": 6990
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.292633056640625,
+      "learning_rate": 1.8332832427870817e-07,
+      "loss": 0.854,
+      "step": 6991
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5829182863235474,
+      "learning_rate": 1.827593840228986e-07,
+      "loss": 1.0291,
+      "step": 6992
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3775451183319092,
+      "learning_rate": 1.8219131152766855e-07,
+      "loss": 0.8366,
+      "step": 6993
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5727830529212952,
+      "learning_rate": 1.816241068953467e-07,
+      "loss": 0.9996,
+      "step": 6994
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3630343675613403,
+      "learning_rate": 1.8105777022810911e-07,
+      "loss": 0.857,
+      "step": 6995
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5804916620254517,
+      "learning_rate": 1.8049230162797305e-07,
+      "loss": 1.0003,
+      "step": 6996
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.395096778869629,
+      "learning_rate": 1.799277011967987e-07,
+      "loss": 0.8938,
+      "step": 6997
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2780916690826416,
+      "learning_rate": 1.7936396903629304e-07,
+      "loss": 0.7591,
+      "step": 6998
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4264857769012451,
+      "learning_rate": 1.788011052480032e-07,
+      "loss": 0.8561,
+      "step": 6999
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2845243215560913,
+      "learning_rate": 1.7823910993332306e-07,
+      "loss": 0.7996,
+      "step": 7000
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5806745886802673,
+      "learning_rate": 1.776779831934877e-07,
+      "loss": 1.0137,
+      "step": 7001
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.582129955291748,
+      "learning_rate": 1.771177251295758e-07,
+      "loss": 1.011,
+      "step": 7002
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5912806987762451,
+      "learning_rate": 1.7655833584251157e-07,
+      "loss": 1.008,
+      "step": 7003
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3548052310943604,
+      "learning_rate": 1.759998154330611e-07,
+      "loss": 0.822,
+      "step": 7004
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.7419794797897339,
+      "learning_rate": 1.7544216400183277e-07,
+      "loss": 0.7758,
+      "step": 7005
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3420807123184204,
+      "learning_rate": 1.748853816492818e-07,
+      "loss": 0.845,
+      "step": 7006
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4267116785049438,
+      "learning_rate": 1.7432946847570408e-07,
+      "loss": 0.8364,
+      "step": 7007
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2984230518341064,
+      "learning_rate": 1.737744245812395e-07,
+      "loss": 0.8808,
+      "step": 7008
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3119744062423706,
+      "learning_rate": 1.7322025006587306e-07,
+      "loss": 0.7952,
+      "step": 7009
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.336135745048523,
+      "learning_rate": 1.7266694502943048e-07,
+      "loss": 0.7985,
+      "step": 7010
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5757025480270386,
+      "learning_rate": 1.721145095715826e-07,
+      "loss": 0.9995,
+      "step": 7011
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.431849479675293,
+      "learning_rate": 1.7156294379184257e-07,
+      "loss": 0.8398,
+      "step": 7012
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5739525556564331,
+      "learning_rate": 1.7101224778956927e-07,
+      "loss": 0.9993,
+      "step": 7013
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5841922163963318,
+      "learning_rate": 1.704624216639611e-07,
+      "loss": 1.017,
+      "step": 7014
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3402884006500244,
+      "learning_rate": 1.6991346551406275e-07,
+      "loss": 0.8541,
+      "step": 7015
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3185129165649414,
+      "learning_rate": 1.693653794387612e-07,
+      "loss": 0.8023,
+      "step": 7016
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.336058497428894,
+      "learning_rate": 1.6881816353678583e-07,
+      "loss": 0.7671,
+      "step": 7017
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5862839818000793,
+      "learning_rate": 1.682718179067122e-07,
+      "loss": 1.019,
+      "step": 7018
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2564630508422852,
+      "learning_rate": 1.6772634264695553e-07,
+      "loss": 0.7928,
+      "step": 7019
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5854629874229431,
+      "learning_rate": 1.6718173785577496e-07,
+      "loss": 1.0034,
+      "step": 7020
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5699886083602905,
+      "learning_rate": 1.6663800363127592e-07,
+      "loss": 1.0204,
+      "step": 7021
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3983044624328613,
+      "learning_rate": 1.6609514007140282e-07,
+      "loss": 0.8044,
+      "step": 7022
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.9021660089492798,
+      "learning_rate": 1.655531472739469e-07,
+      "loss": 0.8245,
+      "step": 7023
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3779178857803345,
+      "learning_rate": 1.6501202533653948e-07,
+      "loss": 0.8193,
+      "step": 7024
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3779515027999878,
+      "learning_rate": 1.644717743566565e-07,
+      "loss": 0.7697,
+      "step": 7025
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4523839950561523,
+      "learning_rate": 1.6393239443161735e-07,
+      "loss": 0.7664,
+      "step": 7026
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5878002047538757,
+      "learning_rate": 1.6339388565858317e-07,
+      "loss": 0.9743,
+      "step": 7027
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5812231302261353,
+      "learning_rate": 1.6285624813456025e-07,
+      "loss": 1.0045,
+      "step": 7028
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5762260556221008,
+      "learning_rate": 1.6231948195639613e-07,
+      "loss": 0.9848,
+      "step": 7029
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5080339908599854,
+      "learning_rate": 1.6178358722078125e-07,
+      "loss": 0.8201,
+      "step": 7030
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.329867959022522,
+      "learning_rate": 1.612485640242506e-07,
+      "loss": 0.7905,
+      "step": 7031
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5263837575912476,
+      "learning_rate": 1.60714412463181e-07,
+      "loss": 0.7976,
+      "step": 7032
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.57855623960495,
+      "learning_rate": 1.601811326337932e-07,
+      "loss": 0.9888,
+      "step": 7033
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5792664289474487,
+      "learning_rate": 1.5964872463214985e-07,
+      "loss": 1.0023,
+      "step": 7034
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5663550496101379,
+      "learning_rate": 1.5911718855415648e-07,
+      "loss": 0.9959,
+      "step": 7035
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5740136504173279,
+      "learning_rate": 1.5858652449556368e-07,
+      "loss": 0.9985,
+      "step": 7036
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3443156480789185,
+      "learning_rate": 1.5805673255196164e-07,
+      "loss": 0.7718,
+      "step": 7037
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.340481162071228,
+      "learning_rate": 1.5752781281878626e-07,
+      "loss": 0.7769,
+      "step": 7038
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3139017820358276,
+      "learning_rate": 1.5699976539131467e-07,
+      "loss": 0.8086,
+      "step": 7039
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.290483832359314,
+      "learning_rate": 1.56472590364668e-07,
+      "loss": 0.8274,
+      "step": 7040
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.609797477722168,
+      "learning_rate": 1.5594628783380917e-07,
+      "loss": 0.8123,
+      "step": 7041
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3159373998641968,
+      "learning_rate": 1.554208578935451e-07,
+      "loss": 0.7898,
+      "step": 7042
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5146372318267822,
+      "learning_rate": 1.5489630063852402e-07,
+      "loss": 0.7912,
+      "step": 7043
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3802191019058228,
+      "learning_rate": 1.5437261616323861e-07,
+      "loss": 0.7893,
+      "step": 7044
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5878559947013855,
+      "learning_rate": 1.5384980456202348e-07,
+      "loss": 0.9965,
+      "step": 7045
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.2958954572677612,
+      "learning_rate": 1.533278659290549e-07,
+      "loss": 0.8465,
+      "step": 7046
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5314971208572388,
+      "learning_rate": 1.5280680035835438e-07,
+      "loss": 0.7526,
+      "step": 7047
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.346713900566101,
+      "learning_rate": 1.522866079437846e-07,
+      "loss": 0.836,
+      "step": 7048
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4895890951156616,
+      "learning_rate": 1.5176728877905e-07,
+      "loss": 0.8774,
+      "step": 7049
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5482752323150635,
+      "learning_rate": 1.512488429577008e-07,
+      "loss": 0.7939,
+      "step": 7050
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.300065279006958,
+      "learning_rate": 1.507312705731262e-07,
+      "loss": 0.8267,
+      "step": 7051
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3903552293777466,
+      "learning_rate": 1.5021457171856103e-07,
+      "loss": 0.9174,
+      "step": 7052
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.331838846206665,
+      "learning_rate": 1.4969874648708082e-07,
+      "loss": 0.7737,
+      "step": 7053
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3618816137313843,
+      "learning_rate": 1.4918379497160517e-07,
+      "loss": 0.8442,
+      "step": 7054
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.5377382040023804,
+      "learning_rate": 1.4866971726489486e-07,
+      "loss": 0.7852,
+      "step": 7055
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.3526257276535034,
+      "learning_rate": 1.4815651345955474e-07,
+      "loss": 0.8196,
+      "step": 7056
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4396017789840698,
+      "learning_rate": 1.4764418364803134e-07,
+      "loss": 0.8405,
+      "step": 7057
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5771210789680481,
+      "learning_rate": 1.4713272792261368e-07,
+      "loss": 0.9962,
+      "step": 7058
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.4188508987426758,
+      "learning_rate": 1.466221463754336e-07,
+      "loss": 0.8438,
+      "step": 7059
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3459749221801758,
+      "learning_rate": 1.4611243909846529e-07,
+      "loss": 0.7993,
+      "step": 7060
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3092695474624634,
+      "learning_rate": 1.4560360618352643e-07,
+      "loss": 0.8045,
+      "step": 7061
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5742385983467102,
+      "learning_rate": 1.4509564772227535e-07,
+      "loss": 0.9862,
+      "step": 7062
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2884180545806885,
+      "learning_rate": 1.445885638062139e-07,
+      "loss": 0.8343,
+      "step": 7063
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.4537174701690674,
+      "learning_rate": 1.4408235452668783e-07,
+      "loss": 0.8208,
+      "step": 7064
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3814923763275146,
+      "learning_rate": 1.435770199748826e-07,
+      "loss": 0.8286,
+      "step": 7065
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5810520052909851,
+      "learning_rate": 1.4307256024182703e-07,
+      "loss": 1.0006,
+      "step": 7066
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5830573439598083,
+      "learning_rate": 1.4256897541839343e-07,
+      "loss": 0.9853,
+      "step": 7067
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3383824825286865,
+      "learning_rate": 1.420662655952959e-07,
+      "loss": 0.808,
+      "step": 7068
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 2.1456503868103027,
+      "learning_rate": 1.4156443086309034e-07,
+      "loss": 0.831,
+      "step": 7069
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.38627290725708,
+      "learning_rate": 1.4106347131217612e-07,
+      "loss": 0.823,
+      "step": 7070
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4228293895721436,
+      "learning_rate": 1.4056338703279326e-07,
+      "loss": 0.798,
+      "step": 7071
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5657803416252136,
+      "learning_rate": 1.4006417811502582e-07,
+      "loss": 1.0206,
+      "step": 7072
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.340654730796814,
+      "learning_rate": 1.3956584464879964e-07,
+      "loss": 0.837,
+      "step": 7073
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4171403646469116,
+      "learning_rate": 1.3906838672388233e-07,
+      "loss": 0.8449,
+      "step": 7074
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2789099216461182,
+      "learning_rate": 1.3857180442988506e-07,
+      "loss": 0.832,
+      "step": 7075
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5766944289207458,
+      "learning_rate": 1.3807609785625897e-07,
+      "loss": 0.9967,
+      "step": 7076
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3748680353164673,
+      "learning_rate": 1.3758126709229992e-07,
+      "loss": 0.8163,
+      "step": 7077
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5565219521522522,
+      "learning_rate": 1.370873122271449e-07,
+      "loss": 1.007,
+      "step": 7078
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5628466606140137,
+      "learning_rate": 1.3659423334977274e-07,
+      "loss": 0.9964,
+      "step": 7079
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3038126230239868,
+      "learning_rate": 1.3610203054900571e-07,
+      "loss": 0.8146,
+      "step": 7080
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3014346361160278,
+      "learning_rate": 1.3561070391350573e-07,
+      "loss": 0.8082,
+      "step": 7081
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3489060401916504,
+      "learning_rate": 1.351202535317808e-07,
+      "loss": 0.7999,
+      "step": 7082
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3451741933822632,
+      "learning_rate": 1.3463067949217756e-07,
+      "loss": 0.8551,
+      "step": 7083
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.5704402923583984,
+      "learning_rate": 1.3414198188288708e-07,
+      "loss": 0.8888,
+      "step": 7084
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2810419797897339,
+      "learning_rate": 1.336541607919406e-07,
+      "loss": 0.7848,
+      "step": 7085
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3466941118240356,
+      "learning_rate": 1.3316721630721285e-07,
+      "loss": 0.8861,
+      "step": 7086
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.562228798866272,
+      "learning_rate": 1.3268114851642032e-07,
+      "loss": 0.8599,
+      "step": 7087
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3096944093704224,
+      "learning_rate": 1.3219595750712133e-07,
+      "loss": 0.7699,
+      "step": 7088
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3655369281768799,
+      "learning_rate": 1.3171164336671705e-07,
+      "loss": 0.8584,
+      "step": 7089
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5771591663360596,
+      "learning_rate": 1.3122820618244936e-07,
+      "loss": 0.9907,
+      "step": 7090
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4859192371368408,
+      "learning_rate": 1.3074564604140362e-07,
+      "loss": 0.811,
+      "step": 7091
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.295251727104187,
+      "learning_rate": 1.3026396303050637e-07,
+      "loss": 0.8499,
+      "step": 7092
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3087033033370972,
+      "learning_rate": 1.2978315723652601e-07,
+      "loss": 0.8322,
+      "step": 7093
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.9704281091690063,
+      "learning_rate": 1.2930322874607272e-07,
+      "loss": 0.8127,
+      "step": 7094
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4151488542556763,
+      "learning_rate": 1.288241776455995e-07,
+      "loss": 0.7937,
+      "step": 7095
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.8659380674362183,
+      "learning_rate": 1.2834600402140128e-07,
+      "loss": 0.779,
+      "step": 7096
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.310221791267395,
+      "learning_rate": 1.278687079596147e-07,
+      "loss": 0.815,
+      "step": 7097
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2985776662826538,
+      "learning_rate": 1.2739228954621764e-07,
+      "loss": 0.8328,
+      "step": 7098
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5675316452980042,
+      "learning_rate": 1.2691674886702976e-07,
+      "loss": 0.9904,
+      "step": 7099
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2405496835708618,
+      "learning_rate": 1.2644208600771534e-07,
+      "loss": 0.8035,
+      "step": 7100
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2670533657073975,
+      "learning_rate": 1.25968301053776e-07,
+      "loss": 0.8849,
+      "step": 7101
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3451895713806152,
+      "learning_rate": 1.2549539409055955e-07,
+      "loss": 0.8463,
+      "step": 7102
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3484961986541748,
+      "learning_rate": 1.250233652032534e-07,
+      "loss": 0.7994,
+      "step": 7103
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3417425155639648,
+      "learning_rate": 1.2455221447688571e-07,
+      "loss": 0.8032,
+      "step": 7104
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.557817816734314,
+      "learning_rate": 1.2408194199633016e-07,
+      "loss": 1.0154,
+      "step": 7105
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5816439986228943,
+      "learning_rate": 1.2361254784629794e-07,
+      "loss": 0.9827,
+      "step": 7106
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6630351543426514,
+      "learning_rate": 1.2314403211134528e-07,
+      "loss": 0.7639,
+      "step": 7107
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5810190439224243,
+      "learning_rate": 1.2267639487586858e-07,
+      "loss": 1.006,
+      "step": 7108
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.348236322402954,
+      "learning_rate": 1.2220963622410598e-07,
+      "loss": 0.8601,
+      "step": 7109
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2723639011383057,
+      "learning_rate": 1.2174375624013858e-07,
+      "loss": 0.7939,
+      "step": 7110
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.1957337856292725,
+      "learning_rate": 1.2127875500788754e-07,
+      "loss": 0.766,
+      "step": 7111
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3027409315109253,
+      "learning_rate": 1.2081463261111704e-07,
+      "loss": 0.8691,
+      "step": 7112
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.405968189239502,
+      "learning_rate": 1.2035138913343236e-07,
+      "loss": 0.8555,
+      "step": 7113
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2207449674606323,
+      "learning_rate": 1.1988902465828013e-07,
+      "loss": 0.7696,
+      "step": 7114
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.4035671949386597,
+      "learning_rate": 1.194275392689498e-07,
+      "loss": 0.8133,
+      "step": 7115
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.364410400390625,
+      "learning_rate": 1.1896693304857044e-07,
+      "loss": 0.8609,
+      "step": 7116
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5627917647361755,
+      "learning_rate": 1.1850720608011513e-07,
+      "loss": 0.9971,
+      "step": 7117
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.423150897026062,
+      "learning_rate": 1.1804835844639761e-07,
+      "loss": 0.8377,
+      "step": 7118
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3952704668045044,
+      "learning_rate": 1.175903902300718e-07,
+      "loss": 0.8272,
+      "step": 7119
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3274112939834595,
+      "learning_rate": 1.1713330151363555e-07,
+      "loss": 0.7528,
+      "step": 7120
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3019875288009644,
+      "learning_rate": 1.1667709237942637e-07,
+      "loss": 0.8023,
+      "step": 7121
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3669962882995605,
+      "learning_rate": 1.1622176290962517e-07,
+      "loss": 0.8654,
+      "step": 7122
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2470808029174805,
+      "learning_rate": 1.1576731318625301e-07,
+      "loss": 0.8351,
+      "step": 7123
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3377126455307007,
+      "learning_rate": 1.1531374329117217e-07,
+      "loss": 0.8475,
+      "step": 7124
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2873101234436035,
+      "learning_rate": 1.1486105330608844e-07,
+      "loss": 0.8108,
+      "step": 7125
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.425157070159912,
+      "learning_rate": 1.1440924331254655e-07,
+      "loss": 0.823,
+      "step": 7126
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3538659811019897,
+      "learning_rate": 1.1395831339193419e-07,
+      "loss": 0.882,
+      "step": 7127
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5640295147895813,
+      "learning_rate": 1.1350826362548028e-07,
+      "loss": 1.0215,
+      "step": 7128
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2810288667678833,
+      "learning_rate": 1.1305909409425608e-07,
+      "loss": 0.8742,
+      "step": 7129
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3414403200149536,
+      "learning_rate": 1.1261080487917241e-07,
+      "loss": 0.8265,
+      "step": 7130
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.6079398393630981,
+      "learning_rate": 1.1216339606098247e-07,
+      "loss": 0.8573,
+      "step": 7131
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.2938934564590454,
+      "learning_rate": 1.1171686772028123e-07,
+      "loss": 0.8231,
+      "step": 7132
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5776501893997192,
+      "learning_rate": 1.112712199375049e-07,
+      "loss": 1.0034,
+      "step": 7133
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3261135816574097,
+      "learning_rate": 1.1082645279293036e-07,
+      "loss": 0.8783,
+      "step": 7134
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3880647420883179,
+      "learning_rate": 1.1038256636667687e-07,
+      "loss": 0.8369,
+      "step": 7135
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4096206426620483,
+      "learning_rate": 1.0993956073870438e-07,
+      "loss": 0.8499,
+      "step": 7136
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5663391351699829,
+      "learning_rate": 1.0949743598881401e-07,
+      "loss": 0.9971,
+      "step": 7137
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.361891746520996,
+      "learning_rate": 1.0905619219664932e-07,
+      "loss": 0.8252,
+      "step": 7138
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5809825658798218,
+      "learning_rate": 1.0861582944169335e-07,
+      "loss": 1.0076,
+      "step": 7139
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.418837070465088,
+      "learning_rate": 1.0817634780327213e-07,
+      "loss": 0.7883,
+      "step": 7140
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.380990982055664,
+      "learning_rate": 1.0773774736055287e-07,
+      "loss": 0.8143,
+      "step": 7141
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3121131658554077,
+      "learning_rate": 1.0730002819254236e-07,
+      "loss": 0.838,
+      "step": 7142
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.474678874015808,
+      "learning_rate": 1.0686319037809034e-07,
+      "loss": 0.8175,
+      "step": 7143
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3006105422973633,
+      "learning_rate": 1.0642723399588662e-07,
+      "loss": 0.7856,
+      "step": 7144
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4016886949539185,
+      "learning_rate": 1.0599215912446447e-07,
+      "loss": 0.8142,
+      "step": 7145
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.457997441291809,
+      "learning_rate": 1.0555796584219514e-07,
+      "loss": 0.7878,
+      "step": 7146
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5762951970100403,
+      "learning_rate": 1.0512465422729267e-07,
+      "loss": 1.0114,
+      "step": 7147
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.407472014427185,
+      "learning_rate": 1.0469222435781357e-07,
+      "loss": 0.7972,
+      "step": 7148
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5610100030899048,
+      "learning_rate": 1.0426067631165271e-07,
+      "loss": 0.9986,
+      "step": 7149
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4560203552246094,
+      "learning_rate": 1.0383001016654848e-07,
+      "loss": 0.8053,
+      "step": 7150
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.302699089050293,
+      "learning_rate": 1.0340022600007992e-07,
+      "loss": 0.8523,
+      "step": 7151
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5867680311203003,
+      "learning_rate": 1.0297132388966513e-07,
+      "loss": 0.9897,
+      "step": 7152
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4207581281661987,
+      "learning_rate": 1.0254330391256729e-07,
+      "loss": 0.8524,
+      "step": 7153
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3225408792495728,
+      "learning_rate": 1.0211616614588693e-07,
+      "loss": 0.8231,
+      "step": 7154
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3422138690948486,
+      "learning_rate": 1.0168991066656698e-07,
+      "loss": 0.826,
+      "step": 7155
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3533929586410522,
+      "learning_rate": 1.0126453755139266e-07,
+      "loss": 0.8194,
+      "step": 7156
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.36699640750885,
+      "learning_rate": 1.0084004687698823e-07,
+      "loss": 0.8525,
+      "step": 7157
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4186277389526367,
+      "learning_rate": 1.0041643871982032e-07,
+      "loss": 0.8729,
+      "step": 7158
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3652592897415161,
+      "learning_rate": 9.999371315619622e-08,
+      "loss": 0.7441,
+      "step": 7159
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4586576223373413,
+      "learning_rate": 9.957187026226389e-08,
+      "loss": 0.8199,
+      "step": 7160
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3454416990280151,
+      "learning_rate": 9.915091011401313e-08,
+      "loss": 0.8577,
+      "step": 7161
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5747032165527344,
+      "learning_rate": 9.87308327872738e-08,
+      "loss": 1.0018,
+      "step": 7162
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5635683536529541,
+      "learning_rate": 9.831163835771706e-08,
+      "loss": 1.0166,
+      "step": 7163
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5650964379310608,
+      "learning_rate": 9.789332690085585e-08,
+      "loss": 1.0165,
+      "step": 7164
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5588693618774414,
+      "learning_rate": 9.74758984920421e-08,
+      "loss": 1.0195,
+      "step": 7165
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2709215879440308,
+      "learning_rate": 9.70593532064712e-08,
+      "loss": 0.8165,
+      "step": 7166
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6973803043365479,
+      "learning_rate": 9.664369111917704e-08,
+      "loss": 0.8061,
+      "step": 7167
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3268229961395264,
+      "learning_rate": 9.622891230503584e-08,
+      "loss": 0.8331,
+      "step": 7168
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3721109628677368,
+      "learning_rate": 9.581501683876504e-08,
+      "loss": 0.8057,
+      "step": 7169
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2950713634490967,
+      "learning_rate": 9.540200479492112e-08,
+      "loss": 0.8666,
+      "step": 7170
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.706869125366211,
+      "learning_rate": 9.498987624790401e-08,
+      "loss": 0.8052,
+      "step": 7171
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4361755847930908,
+      "learning_rate": 9.457863127195099e-08,
+      "loss": 0.8528,
+      "step": 7172
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2851639986038208,
+      "learning_rate": 9.416826994114447e-08,
+      "loss": 0.775,
+      "step": 7173
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5619038939476013,
+      "learning_rate": 9.375879232940477e-08,
+      "loss": 0.9891,
+      "step": 7174
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2778117656707764,
+      "learning_rate": 9.335019851049288e-08,
+      "loss": 0.8274,
+      "step": 7175
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4557111263275146,
+      "learning_rate": 9.294248855801269e-08,
+      "loss": 0.7934,
+      "step": 7176
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2715038061141968,
+      "learning_rate": 9.253566254540603e-08,
+      "loss": 0.7758,
+      "step": 7177
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3603010177612305,
+      "learning_rate": 9.21297205459587e-08,
+      "loss": 0.8446,
+      "step": 7178
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5743131637573242,
+      "learning_rate": 9.172466263279555e-08,
+      "loss": 1.0134,
+      "step": 7179
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5807306170463562,
+      "learning_rate": 9.132048887888101e-08,
+      "loss": 1.0091,
+      "step": 7180
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.559709072113037,
+      "learning_rate": 9.091719935702236e-08,
+      "loss": 0.8219,
+      "step": 7181
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3749052286148071,
+      "learning_rate": 9.051479413986708e-08,
+      "loss": 0.7975,
+      "step": 7182
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5384252071380615,
+      "learning_rate": 9.01132732999016e-08,
+      "loss": 0.8045,
+      "step": 7183
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5756364464759827,
+      "learning_rate": 8.971263690945587e-08,
+      "loss": 1.0269,
+      "step": 7184
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2495030164718628,
+      "learning_rate": 8.931288504069879e-08,
+      "loss": 0.7977,
+      "step": 7185
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.2780921459197998,
+      "learning_rate": 8.891401776564001e-08,
+      "loss": 0.7951,
+      "step": 7186
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4080889225006104,
+      "learning_rate": 8.851603515613038e-08,
+      "loss": 0.7803,
+      "step": 7187
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5773888230323792,
+      "learning_rate": 8.811893728386033e-08,
+      "loss": 0.9779,
+      "step": 7188
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3581740856170654,
+      "learning_rate": 8.772272422036265e-08,
+      "loss": 0.8202,
+      "step": 7189
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3430535793304443,
+      "learning_rate": 8.732739603700857e-08,
+      "loss": 0.8228,
+      "step": 7190
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4329144954681396,
+      "learning_rate": 8.693295280501224e-08,
+      "loss": 0.8408,
+      "step": 7191
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3561562299728394,
+      "learning_rate": 8.653939459542682e-08,
+      "loss": 0.8016,
+      "step": 7192
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5802834630012512,
+      "learning_rate": 8.614672147914616e-08,
+      "loss": 0.9779,
+      "step": 7193
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3830281496047974,
+      "learning_rate": 8.57549335269059e-08,
+      "loss": 0.7655,
+      "step": 7194
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.339349627494812,
+      "learning_rate": 8.536403080928013e-08,
+      "loss": 0.8424,
+      "step": 7195
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3303862810134888,
+      "learning_rate": 8.497401339668531e-08,
+      "loss": 0.8236,
+      "step": 7196
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4477131366729736,
+      "learning_rate": 8.458488135937859e-08,
+      "loss": 0.8084,
+      "step": 7197
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3551524877548218,
+      "learning_rate": 8.419663476745499e-08,
+      "loss": 0.8056,
+      "step": 7198
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.577391505241394,
+      "learning_rate": 8.3809273690853e-08,
+      "loss": 1.0078,
+      "step": 7199
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3227890729904175,
+      "learning_rate": 8.342279819935073e-08,
+      "loss": 0.8189,
+      "step": 7200
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5827528834342957,
+      "learning_rate": 8.303720836256579e-08,
+      "loss": 0.9982,
+      "step": 7201
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.5076665878295898,
+      "learning_rate": 8.265250424995763e-08,
+      "loss": 0.8754,
+      "step": 7202
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.6078182458877563,
+      "learning_rate": 8.22686859308247e-08,
+      "loss": 0.8501,
+      "step": 7203
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.4223581552505493,
+      "learning_rate": 8.188575347430783e-08,
+      "loss": 0.8068,
+      "step": 7204
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3423521518707275,
+      "learning_rate": 8.150370694938626e-08,
+      "loss": 0.8555,
+      "step": 7205
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5680739879608154,
+      "learning_rate": 8.112254642487993e-08,
+      "loss": 1.0008,
+      "step": 7206
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3734145164489746,
+      "learning_rate": 8.074227196945061e-08,
+      "loss": 0.8044,
+      "step": 7207
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 1.3340086936950684,
+      "learning_rate": 8.03628836515996e-08,
+      "loss": 0.8203,
+      "step": 7208
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5709322094917297,
+      "learning_rate": 7.998438153966837e-08,
+      "loss": 1.0109,
+      "step": 7209
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5744004845619202,
+      "learning_rate": 7.960676570183901e-08,
+      "loss": 1.0048,
+      "step": 7210
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5686025023460388,
+      "learning_rate": 7.923003620613379e-08,
+      "loss": 1.0032,
+      "step": 7211
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3762454986572266,
+      "learning_rate": 7.885419312041509e-08,
+      "loss": 0.8085,
+      "step": 7212
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.494831919670105,
+      "learning_rate": 7.847923651238654e-08,
+      "loss": 0.7804,
+      "step": 7213
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3173680305480957,
+      "learning_rate": 7.810516644959131e-08,
+      "loss": 0.8025,
+      "step": 7214
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5559514760971069,
+      "learning_rate": 7.773198299941276e-08,
+      "loss": 1.0304,
+      "step": 7215
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5758073329925537,
+      "learning_rate": 7.735968622907486e-08,
+      "loss": 0.9936,
+      "step": 7216
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3000404834747314,
+      "learning_rate": 7.698827620564287e-08,
+      "loss": 0.8423,
+      "step": 7217
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.556128203868866,
+      "learning_rate": 7.661775299601936e-08,
+      "loss": 1.0127,
+      "step": 7218
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2955131530761719,
+      "learning_rate": 7.624811666695098e-08,
+      "loss": 0.7661,
+      "step": 7219
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.8861762285232544,
+      "learning_rate": 7.587936728502166e-08,
+      "loss": 0.7685,
+      "step": 7220
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5732319951057434,
+      "learning_rate": 7.55115049166566e-08,
+      "loss": 0.9975,
+      "step": 7221
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5679969787597656,
+      "learning_rate": 7.514452962812113e-08,
+      "loss": 1.0156,
+      "step": 7222
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4258828163146973,
+      "learning_rate": 7.477844148552127e-08,
+      "loss": 0.7945,
+      "step": 7223
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.569096565246582,
+      "learning_rate": 7.441324055480314e-08,
+      "loss": 1.0304,
+      "step": 7224
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2815585136413574,
+      "learning_rate": 7.404892690175247e-08,
+      "loss": 0.8436,
+      "step": 7225
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.327506422996521,
+      "learning_rate": 7.368550059199453e-08,
+      "loss": 0.8185,
+      "step": 7226
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.337364912033081,
+      "learning_rate": 7.332296169099695e-08,
+      "loss": 0.8163,
+      "step": 7227
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3204540014266968,
+      "learning_rate": 7.296131026406472e-08,
+      "loss": 0.7893,
+      "step": 7228
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3138407468795776,
+      "learning_rate": 7.260054637634628e-08,
+      "loss": 0.8271,
+      "step": 7229
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5680536031723022,
+      "learning_rate": 7.224067009282742e-08,
+      "loss": 1.0129,
+      "step": 7230
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5142368078231812,
+      "learning_rate": 7.188168147833464e-08,
+      "loss": 0.764,
+      "step": 7231
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3028082847595215,
+      "learning_rate": 7.152358059753506e-08,
+      "loss": 0.7952,
+      "step": 7232
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.262429118156433,
+      "learning_rate": 7.1166367514936e-08,
+      "loss": 0.7906,
+      "step": 7233
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5751594305038452,
+      "learning_rate": 7.081004229488431e-08,
+      "loss": 1.0041,
+      "step": 7234
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3055667877197266,
+      "learning_rate": 7.045460500156753e-08,
+      "loss": 0.8345,
+      "step": 7235
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2740387916564941,
+      "learning_rate": 7.010005569901223e-08,
+      "loss": 0.8165,
+      "step": 7236
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.390101432800293,
+      "learning_rate": 6.974639445108621e-08,
+      "loss": 0.8231,
+      "step": 7237
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.493383765220642,
+      "learning_rate": 6.93936213214963e-08,
+      "loss": 0.8066,
+      "step": 7238
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.282451868057251,
+      "learning_rate": 6.904173637378997e-08,
+      "loss": 0.8108,
+      "step": 7239
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4339659214019775,
+      "learning_rate": 6.869073967135487e-08,
+      "loss": 0.8143,
+      "step": 7240
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5851967334747314,
+      "learning_rate": 6.834063127741764e-08,
+      "loss": 1.0163,
+      "step": 7241
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2407392263412476,
+      "learning_rate": 6.799141125504671e-08,
+      "loss": 0.8212,
+      "step": 7242
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3835612535476685,
+      "learning_rate": 6.76430796671479e-08,
+      "loss": 0.8154,
+      "step": 7243
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.358812689781189,
+      "learning_rate": 6.729563657646931e-08,
+      "loss": 0.8554,
+      "step": 7244
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.274545431137085,
+      "learning_rate": 6.694908204559758e-08,
+      "loss": 0.7879,
+      "step": 7245
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3131309747695923,
+      "learning_rate": 6.660341613696053e-08,
+      "loss": 0.7942,
+      "step": 7246
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2728809118270874,
+      "learning_rate": 6.625863891282447e-08,
+      "loss": 0.7946,
+      "step": 7247
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5148429870605469,
+      "learning_rate": 6.591475043529638e-08,
+      "loss": 0.8461,
+      "step": 7248
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.584212601184845,
+      "learning_rate": 6.557175076632394e-08,
+      "loss": 1.003,
+      "step": 7249
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5701957941055298,
+      "learning_rate": 6.522963996769272e-08,
+      "loss": 1.0134,
+      "step": 7250
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4759835004806519,
+      "learning_rate": 6.488841810103008e-08,
+      "loss": 0.8244,
+      "step": 7251
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.396315574645996,
+      "learning_rate": 6.454808522780298e-08,
+      "loss": 0.8103,
+      "step": 7252
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2823814153671265,
+      "learning_rate": 6.420864140931682e-08,
+      "loss": 0.7984,
+      "step": 7253
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3679262399673462,
+      "learning_rate": 6.387008670671769e-08,
+      "loss": 0.8228,
+      "step": 7254
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3560951948165894,
+      "learning_rate": 6.353242118099234e-08,
+      "loss": 0.8226,
+      "step": 7255
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.399656891822815,
+      "learning_rate": 6.319564489296603e-08,
+      "loss": 0.8245,
+      "step": 7256
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4901663064956665,
+      "learning_rate": 6.285975790330522e-08,
+      "loss": 0.8439,
+      "step": 7257
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3877828121185303,
+      "learning_rate": 6.252476027251487e-08,
+      "loss": 0.7593,
+      "step": 7258
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3569422960281372,
+      "learning_rate": 6.219065206094e-08,
+      "loss": 0.8308,
+      "step": 7259
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.474320888519287,
+      "learning_rate": 6.185743332876637e-08,
+      "loss": 0.8604,
+      "step": 7260
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4671446084976196,
+      "learning_rate": 6.15251041360182e-08,
+      "loss": 0.8153,
+      "step": 7261
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3804396390914917,
+      "learning_rate": 6.119366454256037e-08,
+      "loss": 0.8132,
+      "step": 7262
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2971158027648926,
+      "learning_rate": 6.086311460809791e-08,
+      "loss": 0.7948,
+      "step": 7263
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3759453296661377,
+      "learning_rate": 6.053345439217318e-08,
+      "loss": 0.8834,
+      "step": 7264
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3247429132461548,
+      "learning_rate": 6.020468395417145e-08,
+      "loss": 0.8438,
+      "step": 7265
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3739254474639893,
+      "learning_rate": 5.987680335331592e-08,
+      "loss": 0.8022,
+      "step": 7266
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.302605152130127,
+      "learning_rate": 5.954981264866988e-08,
+      "loss": 0.835,
+      "step": 7267
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.308717131614685,
+      "learning_rate": 5.922371189913567e-08,
+      "loss": 0.8084,
+      "step": 7268
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2860863208770752,
+      "learning_rate": 5.889850116345686e-08,
+      "loss": 0.8014,
+      "step": 7269
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.409630298614502,
+      "learning_rate": 5.857418050021491e-08,
+      "loss": 0.7533,
+      "step": 7270
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2747135162353516,
+      "learning_rate": 5.825074996783253e-08,
+      "loss": 0.8449,
+      "step": 7271
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5781501531600952,
+      "learning_rate": 5.7928209624570356e-08,
+      "loss": 0.9818,
+      "step": 7272
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4005392789840698,
+      "learning_rate": 5.760655952853078e-08,
+      "loss": 0.8235,
+      "step": 7273
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2480911016464233,
+      "learning_rate": 5.728579973765413e-08,
+      "loss": 0.8169,
+      "step": 7274
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5665532946586609,
+      "learning_rate": 5.6965930309720866e-08,
+      "loss": 1.0176,
+      "step": 7275
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3272370100021362,
+      "learning_rate": 5.664695130235154e-08,
+      "loss": 0.8375,
+      "step": 7276
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.571286141872406,
+      "learning_rate": 5.63288627730052e-08,
+      "loss": 1.0257,
+      "step": 7277
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5375902652740479,
+      "learning_rate": 5.601166477898212e-08,
+      "loss": 0.863,
+      "step": 7278
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3892490863800049,
+      "learning_rate": 5.569535737742049e-08,
+      "loss": 0.7466,
+      "step": 7279
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.5542359352111816,
+      "learning_rate": 5.537994062529917e-08,
+      "loss": 0.8946,
+      "step": 7280
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.4803539514541626,
+      "learning_rate": 5.506541457943548e-08,
+      "loss": 0.842,
+      "step": 7281
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.570242702960968,
+      "learning_rate": 5.4751779296488006e-08,
+      "loss": 0.9945,
+      "step": 7282
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3205652236938477,
+      "learning_rate": 5.4439034832953764e-08,
+      "loss": 0.7722,
+      "step": 7283
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5680786967277527,
+      "learning_rate": 5.412718124516825e-08,
+      "loss": 0.9914,
+      "step": 7284
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.3292303085327148,
+      "learning_rate": 5.3816218589309297e-08,
+      "loss": 0.7916,
+      "step": 7285
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5747170448303223,
+      "learning_rate": 5.350614692139211e-08,
+      "loss": 0.9815,
+      "step": 7286
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.2824146747589111,
+      "learning_rate": 5.319696629727089e-08,
+      "loss": 0.8159,
+      "step": 7287
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5685046315193176,
+      "learning_rate": 5.2888676772641644e-08,
+      "loss": 1.0147,
+      "step": 7288
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3802536725997925,
+      "learning_rate": 5.258127840303773e-08,
+      "loss": 0.8353,
+      "step": 7289
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4881116151809692,
+      "learning_rate": 5.227477124383318e-08,
+      "loss": 0.8023,
+      "step": 7290
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3723633289337158,
+      "learning_rate": 5.196915535024105e-08,
+      "loss": 0.8497,
+      "step": 7291
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.6014131307601929,
+      "learning_rate": 5.1664430777313424e-08,
+      "loss": 0.8437,
+      "step": 7292
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2978477478027344,
+      "learning_rate": 5.1360597579943606e-08,
+      "loss": 0.8572,
+      "step": 7293
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5816636681556702,
+      "learning_rate": 5.105765581286115e-08,
+      "loss": 1.0097,
+      "step": 7294
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5774659514427185,
+      "learning_rate": 5.075560553063852e-08,
+      "loss": 1.0108,
+      "step": 7295
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.43397057056427,
+      "learning_rate": 5.045444678768496e-08,
+      "loss": 0.8052,
+      "step": 7296
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5787550210952759,
+      "learning_rate": 5.0154179638250424e-08,
+      "loss": 0.9989,
+      "step": 7297
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3450981378555298,
+      "learning_rate": 4.985480413642496e-08,
+      "loss": 0.7926,
+      "step": 7298
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3534338474273682,
+      "learning_rate": 4.955632033613544e-08,
+      "loss": 0.7655,
+      "step": 7299
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5648720860481262,
+      "learning_rate": 4.925872829114997e-08,
+      "loss": 1.0148,
+      "step": 7300
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2795815467834473,
+      "learning_rate": 4.896202805507677e-08,
+      "loss": 0.813,
+      "step": 7301
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.344794750213623,
+      "learning_rate": 4.866621968136142e-08,
+      "loss": 0.7799,
+      "step": 7302
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.385208249092102,
+      "learning_rate": 4.83713032232902e-08,
+      "loss": 0.8506,
+      "step": 7303
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.8709043264389038,
+      "learning_rate": 4.807727873398838e-08,
+      "loss": 0.7252,
+      "step": 7304
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5820430517196655,
+      "learning_rate": 4.7784146266419695e-08,
+      "loss": 1.0018,
+      "step": 7305
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5739922523498535,
+      "learning_rate": 4.7491905873389144e-08,
+      "loss": 0.9816,
+      "step": 7306
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5855177044868469,
+      "learning_rate": 4.72005576075385e-08,
+      "loss": 0.997,
+      "step": 7307
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7292118072509766,
+      "learning_rate": 4.691010152135189e-08,
+      "loss": 0.8255,
+      "step": 7308
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3670204877853394,
+      "learning_rate": 4.662053766714969e-08,
+      "loss": 0.7994,
+      "step": 7309
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2511003017425537,
+      "learning_rate": 4.6331866097093505e-08,
+      "loss": 0.7965,
+      "step": 7310
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.574091911315918,
+      "learning_rate": 4.6044086863183403e-08,
+      "loss": 1.0033,
+      "step": 7311
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3271342515945435,
+      "learning_rate": 4.575720001725903e-08,
+      "loss": 0.8443,
+      "step": 7312
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.323897361755371,
+      "learning_rate": 4.547120561099905e-08,
+      "loss": 0.7949,
+      "step": 7313
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5613582730293274,
+      "learning_rate": 4.518610369592169e-08,
+      "loss": 1.0285,
+      "step": 7314
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5601296424865723,
+      "learning_rate": 4.490189432338365e-08,
+      "loss": 0.9839,
+      "step": 7315
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.406463623046875,
+      "learning_rate": 4.461857754458232e-08,
+      "loss": 0.8505,
+      "step": 7316
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5877195596694946,
+      "learning_rate": 4.4336153410552416e-08,
+      "loss": 1.0069,
+      "step": 7317
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.302964210510254,
+      "learning_rate": 4.405462197216937e-08,
+      "loss": 0.7683,
+      "step": 7318
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3063809871673584,
+      "learning_rate": 4.377398328014704e-08,
+      "loss": 0.8414,
+      "step": 7319
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.272983431816101,
+      "learning_rate": 4.349423738503888e-08,
+      "loss": 0.8051,
+      "step": 7320
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5812650918960571,
+      "learning_rate": 4.321538433723682e-08,
+      "loss": 1.003,
+      "step": 7321
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5687074661254883,
+      "learning_rate": 4.293742418697344e-08,
+      "loss": 1.0064,
+      "step": 7322
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.659375786781311,
+      "learning_rate": 4.26603569843187e-08,
+      "loss": 0.8561,
+      "step": 7323
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3955751657485962,
+      "learning_rate": 4.238418277918266e-08,
+      "loss": 0.7787,
+      "step": 7324
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4403152465820312,
+      "learning_rate": 4.210890162131387e-08,
+      "loss": 0.829,
+      "step": 7325
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2716535329818726,
+      "learning_rate": 4.1834513560301545e-08,
+      "loss": 0.7597,
+      "step": 7326
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4702320098876953,
+      "learning_rate": 4.156101864557227e-08,
+      "loss": 0.838,
+      "step": 7327
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5768337249755859,
+      "learning_rate": 4.128841692639274e-08,
+      "loss": 1.0109,
+      "step": 7328
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4256778955459595,
+      "learning_rate": 4.101670845186812e-08,
+      "loss": 0.879,
+      "step": 7329
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3266053199768066,
+      "learning_rate": 4.074589327094369e-08,
+      "loss": 0.8471,
+      "step": 7330
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4748526811599731,
+      "learning_rate": 4.0475971432402675e-08,
+      "loss": 0.8596,
+      "step": 7331
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.4202111959457397,
+      "learning_rate": 4.020694298486727e-08,
+      "loss": 0.7902,
+      "step": 7332
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.367361307144165,
+      "learning_rate": 3.993880797679983e-08,
+      "loss": 0.8288,
+      "step": 7333
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2960621118545532,
+      "learning_rate": 3.967156645650172e-08,
+      "loss": 0.823,
+      "step": 7334
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5614935755729675,
+      "learning_rate": 3.940521847211221e-08,
+      "loss": 0.9999,
+      "step": 7335
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2758251428604126,
+      "learning_rate": 3.913976407161069e-08,
+      "loss": 0.8329,
+      "step": 7336
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3210186958312988,
+      "learning_rate": 3.8875203302815024e-08,
+      "loss": 0.8372,
+      "step": 7337
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3707447052001953,
+      "learning_rate": 3.861153621338154e-08,
+      "loss": 0.8213,
+      "step": 7338
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.358506679534912,
+      "learning_rate": 3.8348762850807794e-08,
+      "loss": 0.8541,
+      "step": 7339
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5737218260765076,
+      "learning_rate": 3.808688326242704e-08,
+      "loss": 1.0138,
+      "step": 7340
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.7930471897125244,
+      "learning_rate": 3.7825897495414874e-08,
+      "loss": 0.7324,
+      "step": 7341
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3621435165405273,
+      "learning_rate": 3.756580559678369e-08,
+      "loss": 0.8173,
+      "step": 7342
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3342434167861938,
+      "learning_rate": 3.730660761338545e-08,
+      "loss": 0.8095,
+      "step": 7343
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2769556045532227,
+      "learning_rate": 3.7048303591911695e-08,
+      "loss": 0.8191,
+      "step": 7344
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5623638033866882,
+      "learning_rate": 3.6790893578891874e-08,
+      "loss": 1.0002,
+      "step": 7345
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5662577748298645,
+      "learning_rate": 3.6534377620695e-08,
+      "loss": 1.0368,
+      "step": 7346
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3418277502059937,
+      "learning_rate": 3.627875576352857e-08,
+      "loss": 0.8548,
+      "step": 7347
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2949925661087036,
+      "learning_rate": 3.602402805344018e-08,
+      "loss": 0.7782,
+      "step": 7348
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.078336715698242,
+      "learning_rate": 3.5770194536314803e-08,
+      "loss": 0.8097,
+      "step": 7349
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.458436131477356,
+      "learning_rate": 3.551725525787808e-08,
+      "loss": 0.8201,
+      "step": 7350
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3321434259414673,
+      "learning_rate": 3.526521026369245e-08,
+      "loss": 0.7991,
+      "step": 7351
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3569639921188354,
+      "learning_rate": 3.50140595991616e-08,
+      "loss": 0.8139,
+      "step": 7352
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3150173425674438,
+      "learning_rate": 3.4763803309525446e-08,
+      "loss": 0.7617,
+      "step": 7353
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.2881014347076416,
+      "learning_rate": 3.451444143986571e-08,
+      "loss": 0.8019,
+      "step": 7354
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3805707693099976,
+      "learning_rate": 3.42659740351009e-08,
+      "loss": 0.7751,
+      "step": 7355
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.272660732269287,
+      "learning_rate": 3.401840113998855e-08,
+      "loss": 0.7638,
+      "step": 7356
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3103070259094238,
+      "learning_rate": 3.3771722799126885e-08,
+      "loss": 0.7629,
+      "step": 7357
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3644624948501587,
+      "learning_rate": 3.3525939056949786e-08,
+      "loss": 0.8028,
+      "step": 7358
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.523339867591858,
+      "learning_rate": 3.328104995773407e-08,
+      "loss": 0.7995,
+      "step": 7359
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5802314877510071,
+      "learning_rate": 3.303705554559167e-08,
+      "loss": 0.9989,
+      "step": 7360
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5733658075332642,
+      "learning_rate": 3.27939558644752e-08,
+      "loss": 1.0287,
+      "step": 7361
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.3685648441314697,
+      "learning_rate": 3.255175095817575e-08,
+      "loss": 0.8464,
+      "step": 7362
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5705839991569519,
+      "learning_rate": 3.2310440870323426e-08,
+      "loss": 1.0036,
+      "step": 7363
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5695980787277222,
+      "learning_rate": 3.207002564438677e-08,
+      "loss": 1.0267,
+      "step": 7364
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.361166000366211,
+      "learning_rate": 3.183050532367338e-08,
+      "loss": 0.7912,
+      "step": 7365
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.271673321723938,
+      "learning_rate": 3.159187995132984e-08,
+      "loss": 0.7454,
+      "step": 7366
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.315779447555542,
+      "learning_rate": 3.135414957034122e-08,
+      "loss": 0.8614,
+      "step": 7367
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2508132457733154,
+      "learning_rate": 3.1117314223531036e-08,
+      "loss": 0.7518,
+      "step": 7368
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2857836484909058,
+      "learning_rate": 3.088137395356239e-08,
+      "loss": 0.8176,
+      "step": 7369
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.7905291318893433,
+      "learning_rate": 3.0646328802936274e-08,
+      "loss": 0.8129,
+      "step": 7370
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.412611961364746,
+      "learning_rate": 3.0412178813993274e-08,
+      "loss": 0.8442,
+      "step": 7371
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5341377258300781,
+      "learning_rate": 3.017892402891243e-08,
+      "loss": 0.8095,
+      "step": 7372
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2921541929244995,
+      "learning_rate": 2.994656448971123e-08,
+      "loss": 0.797,
+      "step": 7373
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.36614990234375,
+      "learning_rate": 2.9715100238245642e-08,
+      "loss": 0.7856,
+      "step": 7374
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3098618984222412,
+      "learning_rate": 2.948453131621176e-08,
+      "loss": 0.7897,
+      "step": 7375
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3057392835617065,
+      "learning_rate": 2.9254857765142474e-08,
+      "loss": 0.838,
+      "step": 7376
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.3916311264038086,
+      "learning_rate": 2.902607962641135e-08,
+      "loss": 0.8273,
+      "step": 7377
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4252136945724487,
+      "learning_rate": 2.8798196941228762e-08,
+      "loss": 0.8675,
+      "step": 7378
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.321366548538208,
+      "learning_rate": 2.8571209750645202e-08,
+      "loss": 0.8523,
+      "step": 7379
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4162651300430298,
+      "learning_rate": 2.8345118095549073e-08,
+      "loss": 0.8224,
+      "step": 7380
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5182896852493286,
+      "learning_rate": 2.81199220166678e-08,
+      "loss": 0.8605,
+      "step": 7381
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5646988153457642,
+      "learning_rate": 2.789562155456782e-08,
+      "loss": 0.9823,
+      "step": 7382
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 2.9415056705474854,
+      "learning_rate": 2.7672216749652925e-08,
+      "loss": 0.8038,
+      "step": 7383
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2991664409637451,
+      "learning_rate": 2.744970764216759e-08,
+      "loss": 0.7843,
+      "step": 7384
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4203993082046509,
+      "learning_rate": 2.7228094272193086e-08,
+      "loss": 0.8398,
+      "step": 7385
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5662235617637634,
+      "learning_rate": 2.70073766796497e-08,
+      "loss": 1.0155,
+      "step": 7386
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.322136402130127,
+      "learning_rate": 2.6787554904297852e-08,
+      "loss": 0.8789,
+      "step": 7387
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5677968263626099,
+      "learning_rate": 2.65686289857342e-08,
+      "loss": 0.9988,
+      "step": 7388
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5834643244743347,
+      "learning_rate": 2.635059896339609e-08,
+      "loss": 1.0195,
+      "step": 7389
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5894080996513367,
+      "learning_rate": 2.6133464876558768e-08,
+      "loss": 1.0036,
+      "step": 7390
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4254993200302124,
+      "learning_rate": 2.591722676433539e-08,
+      "loss": 0.8271,
+      "step": 7391
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3272007703781128,
+      "learning_rate": 2.570188466567869e-08,
+      "loss": 0.7984,
+      "step": 7392
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5637536644935608,
+      "learning_rate": 2.5487438619379303e-08,
+      "loss": 0.9796,
+      "step": 7393
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3608824014663696,
+      "learning_rate": 2.5273888664067437e-08,
+      "loss": 0.8432,
+      "step": 7394
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3053202629089355,
+      "learning_rate": 2.5061234838210656e-08,
+      "loss": 0.8173,
+      "step": 7395
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4591819047927856,
+      "learning_rate": 2.4849477180115544e-08,
+      "loss": 0.8089,
+      "step": 7396
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5787189602851868,
+      "learning_rate": 2.463861572792825e-08,
+      "loss": 1.0272,
+      "step": 7397
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4180421829223633,
+      "learning_rate": 2.4428650519631725e-08,
+      "loss": 0.8465,
+      "step": 7398
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3419137001037598,
+      "learning_rate": 2.421958159304849e-08,
+      "loss": 0.8417,
+      "step": 7399
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2991780042648315,
+      "learning_rate": 2.4011408985839535e-08,
+      "loss": 0.7603,
+      "step": 7400
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.5789812803268433,
+      "learning_rate": 2.380413273550486e-08,
+      "loss": 0.817,
+      "step": 7401
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.335207462310791,
+      "learning_rate": 2.3597752879381818e-08,
+      "loss": 0.7734,
+      "step": 7402
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3507322072982788,
+      "learning_rate": 2.339226945464734e-08,
+      "loss": 0.8284,
+      "step": 7403
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4385559558868408,
+      "learning_rate": 2.318768249831571e-08,
+      "loss": 0.817,
+      "step": 7404
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5818255543708801,
+      "learning_rate": 2.2983992047241334e-08,
+      "loss": 0.9821,
+      "step": 7405
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5794207453727722,
+      "learning_rate": 2.278119813811652e-08,
+      "loss": 1.0336,
+      "step": 7406
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4759224653244019,
+      "learning_rate": 2.257930080747095e-08,
+      "loss": 0.824,
+      "step": 7407
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3141952753067017,
+      "learning_rate": 2.2378300091673853e-08,
+      "loss": 0.7917,
+      "step": 7408
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.575084388256073,
+      "learning_rate": 2.2178196026933495e-08,
+      "loss": 0.9959,
+      "step": 7409
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3192832469940186,
+      "learning_rate": 2.197898864929493e-08,
+      "loss": 0.8053,
+      "step": 7410
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3005483150482178,
+      "learning_rate": 2.1780677994643895e-08,
+      "loss": 0.817,
+      "step": 7411
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2831752300262451,
+      "learning_rate": 2.1583264098702374e-08,
+      "loss": 0.8144,
+      "step": 7412
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.315352439880371,
+      "learning_rate": 2.1386746997031914e-08,
+      "loss": 0.8125,
+      "step": 7413
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5863335132598877,
+      "learning_rate": 2.1191126725031984e-08,
+      "loss": 1.0136,
+      "step": 7414
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2498189210891724,
+      "learning_rate": 2.0996403317942172e-08,
+      "loss": 0.8107,
+      "step": 7415
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3401157855987549,
+      "learning_rate": 2.080257681083886e-08,
+      "loss": 0.8297,
+      "step": 7416
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3428221940994263,
+      "learning_rate": 2.0609647238636343e-08,
+      "loss": 0.8391,
+      "step": 7417
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.389541745185852,
+      "learning_rate": 2.04176146360896e-08,
+      "loss": 0.8062,
+      "step": 7418
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2537380456924438,
+      "learning_rate": 2.0226479037789292e-08,
+      "loss": 0.7625,
+      "step": 7419
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3641475439071655,
+      "learning_rate": 2.0036240478167323e-08,
+      "loss": 0.8586,
+      "step": 7420
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3720934391021729,
+      "learning_rate": 1.9846898991491283e-08,
+      "loss": 0.8171,
+      "step": 7421
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2874784469604492,
+      "learning_rate": 1.965845461186944e-08,
+      "loss": 0.7832,
+      "step": 7422
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2714325189590454,
+      "learning_rate": 1.947090737324686e-08,
+      "loss": 0.8013,
+      "step": 7423
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5744485855102539,
+      "learning_rate": 1.9284257309408194e-08,
+      "loss": 1.0179,
+      "step": 7424
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5722184777259827,
+      "learning_rate": 1.909850445397543e-08,
+      "loss": 1.0051,
+      "step": 7425
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5672627687454224,
+      "learning_rate": 1.8913648840410136e-08,
+      "loss": 1.011,
+      "step": 7426
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.521147608757019,
+      "learning_rate": 1.872969050201068e-08,
+      "loss": 0.8264,
+      "step": 7427
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3228318691253662,
+      "learning_rate": 1.854662947191499e-08,
+      "loss": 0.8643,
+      "step": 7428
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4553006887435913,
+      "learning_rate": 1.8364465783098918e-08,
+      "loss": 0.8172,
+      "step": 7429
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.334018349647522,
+      "learning_rate": 1.8183199468377322e-08,
+      "loss": 0.8017,
+      "step": 7430
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4204639196395874,
+      "learning_rate": 1.800283056040242e-08,
+      "loss": 0.7996,
+      "step": 7431
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3080437183380127,
+      "learning_rate": 1.7823359091664882e-08,
+      "loss": 0.772,
+      "step": 7432
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.302982211112976,
+      "learning_rate": 1.7644785094494964e-08,
+      "loss": 0.8068,
+      "step": 7433
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2705721855163574,
+      "learning_rate": 1.7467108601059713e-08,
+      "loss": 0.8041,
+      "step": 7434
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.4461288452148438,
+      "learning_rate": 1.729032964336519e-08,
+      "loss": 0.7404,
+      "step": 7435
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3784452676773071,
+      "learning_rate": 1.7114448253255924e-08,
+      "loss": 0.8181,
+      "step": 7436
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5686826109886169,
+      "learning_rate": 1.6939464462414347e-08,
+      "loss": 0.9858,
+      "step": 7437
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.2825384140014648,
+      "learning_rate": 1.6765378302361357e-08,
+      "loss": 0.837,
+      "step": 7438
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3995239734649658,
+      "learning_rate": 1.659218980445687e-08,
+      "loss": 0.8158,
+      "step": 7439
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.3779869079589844,
+      "learning_rate": 1.64198989998976e-08,
+      "loss": 0.7979,
+      "step": 7440
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3036407232284546,
+      "learning_rate": 1.6248505919719825e-08,
+      "loss": 0.8042,
+      "step": 7441
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3333873748779297,
+      "learning_rate": 1.6078010594797742e-08,
+      "loss": 0.7569,
+      "step": 7442
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7502716779708862,
+      "learning_rate": 1.590841305584345e-08,
+      "loss": 0.8732,
+      "step": 7443
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5745297074317932,
+      "learning_rate": 1.5739713333408068e-08,
+      "loss": 1.0038,
+      "step": 7444
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4186809062957764,
+      "learning_rate": 1.557191145788006e-08,
+      "loss": 0.8499,
+      "step": 7445
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3852026462554932,
+      "learning_rate": 1.5405007459487475e-08,
+      "loss": 0.778,
+      "step": 7446
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3527939319610596,
+      "learning_rate": 1.5239001368294593e-08,
+      "loss": 0.7465,
+      "step": 7447
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.7293758392333984,
+      "learning_rate": 1.5073893214206383e-08,
+      "loss": 0.8292,
+      "step": 7448
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3854871988296509,
+      "learning_rate": 1.490968302696405e-08,
+      "loss": 0.851,
+      "step": 7449
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5899322628974915,
+      "learning_rate": 1.4746370836147827e-08,
+      "loss": 0.9899,
+      "step": 7450
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2558088302612305,
+      "learning_rate": 1.4583956671176957e-08,
+      "loss": 0.7906,
+      "step": 7451
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2320153713226318,
+      "learning_rate": 1.4422440561307482e-08,
+      "loss": 0.7687,
+      "step": 7452
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2418181896209717,
+      "learning_rate": 1.4261822535633907e-08,
+      "loss": 0.802,
+      "step": 7453
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5782583355903625,
+      "learning_rate": 1.4102102623090863e-08,
+      "loss": 1.0128,
+      "step": 7454
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2582920789718628,
+      "learning_rate": 1.3943280852448116e-08,
+      "loss": 0.8473,
+      "step": 7455
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3660919666290283,
+      "learning_rate": 1.3785357252316666e-08,
+      "loss": 0.8729,
+      "step": 7456
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.257721185684204,
+      "learning_rate": 1.3628331851142651e-08,
+      "loss": 0.8225,
+      "step": 7457
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3104697465896606,
+      "learning_rate": 1.3472204677213995e-08,
+      "loss": 0.8541,
+      "step": 7458
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3469613790512085,
+      "learning_rate": 1.3316975758653761e-08,
+      "loss": 0.8218,
+      "step": 7459
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.31045401096344,
+      "learning_rate": 1.3162645123424022e-08,
+      "loss": 0.7885,
+      "step": 7460
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4976327419281006,
+      "learning_rate": 1.3009212799325877e-08,
+      "loss": 0.8372,
+      "step": 7461
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4997919797897339,
+      "learning_rate": 1.2856678813998325e-08,
+      "loss": 0.8702,
+      "step": 7462
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2831237316131592,
+      "learning_rate": 1.2705043194918276e-08,
+      "loss": 0.8614,
+      "step": 7463
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.9574114084243774,
+      "learning_rate": 1.2554305969399993e-08,
+      "loss": 0.8388,
+      "step": 7464
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5806486010551453,
+      "learning_rate": 1.2404467164597311e-08,
+      "loss": 1.0373,
+      "step": 7465
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3226702213287354,
+      "learning_rate": 1.2255526807501971e-08,
+      "loss": 0.7699,
+      "step": 7466
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2638635635375977,
+      "learning_rate": 1.210748492494307e-08,
+      "loss": 0.8124,
+      "step": 7467
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2721840143203735,
+      "learning_rate": 1.1960341543588716e-08,
+      "loss": 0.8423,
+      "step": 7468
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3827037811279297,
+      "learning_rate": 1.1814096689944932e-08,
+      "loss": 0.8118,
+      "step": 7469
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3617771863937378,
+      "learning_rate": 1.1668750390354533e-08,
+      "loss": 0.782,
+      "step": 7470
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2901818752288818,
+      "learning_rate": 1.152430267100102e-08,
+      "loss": 0.7746,
+      "step": 7471
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4737592935562134,
+      "learning_rate": 1.138075355790469e-08,
+      "loss": 0.8231,
+      "step": 7472
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3350062370300293,
+      "learning_rate": 1.1238103076922635e-08,
+      "loss": 0.8765,
+      "step": 7473
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4088101387023926,
+      "learning_rate": 1.1096351253752635e-08,
+      "loss": 0.7832,
+      "step": 7474
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6412208080291748,
+      "learning_rate": 1.0955498113929264e-08,
+      "loss": 0.8248,
+      "step": 7475
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2586804628372192,
+      "learning_rate": 1.0815543682825003e-08,
+      "loss": 0.8283,
+      "step": 7476
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5848299860954285,
+      "learning_rate": 1.0676487985650242e-08,
+      "loss": 0.9938,
+      "step": 7477
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3860018253326416,
+      "learning_rate": 1.0538331047454941e-08,
+      "loss": 0.7937,
+      "step": 7478
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.6070458889007568,
+      "learning_rate": 1.0401072893125863e-08,
+      "loss": 0.7845,
+      "step": 7479
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2494930028915405,
+      "learning_rate": 1.0264713547388227e-08,
+      "loss": 0.8125,
+      "step": 7480
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3632745742797852,
+      "learning_rate": 1.0129253034805165e-08,
+      "loss": 0.8277,
+      "step": 7481
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.5073801279067993,
+      "learning_rate": 9.99469137977771e-09,
+      "loss": 0.8766,
+      "step": 7482
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2439550161361694,
+      "learning_rate": 9.86102860654592e-09,
+      "loss": 0.8371,
+      "step": 7483
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3146297931671143,
+      "learning_rate": 9.728264739187198e-09,
+      "loss": 0.8147,
+      "step": 7484
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3731977939605713,
+      "learning_rate": 9.596399801616862e-09,
+      "loss": 0.8514,
+      "step": 7485
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3957005739212036,
+      "learning_rate": 9.465433817589243e-09,
+      "loss": 0.8417,
+      "step": 7486
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.564289391040802,
+      "learning_rate": 9.335366810695467e-09,
+      "loss": 1.0118,
+      "step": 7487
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3671952486038208,
+      "learning_rate": 9.206198804365684e-09,
+      "loss": 0.7898,
+      "step": 7488
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3746883869171143,
+      "learning_rate": 9.077929821867392e-09,
+      "loss": 0.8579,
+      "step": 7489
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.2853877544403076,
+      "learning_rate": 8.950559886306553e-09,
+      "loss": 0.8131,
+      "step": 7490
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3958454132080078,
+      "learning_rate": 8.82408902062759e-09,
+      "loss": 0.8545,
+      "step": 7491
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3111635446548462,
+      "learning_rate": 8.698517247612282e-09,
+      "loss": 0.746,
+      "step": 7492
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.375796914100647,
+      "learning_rate": 8.573844589880308e-09,
+      "loss": 0.8231,
+      "step": 7493
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3593924045562744,
+      "learning_rate": 8.450071069890375e-09,
+      "loss": 0.8007,
+      "step": 7494
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.273775339126587,
+      "learning_rate": 8.327196709937978e-09,
+      "loss": 0.8498,
+      "step": 7495
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.364674687385559,
+      "learning_rate": 8.205221532158192e-09,
+      "loss": 0.8054,
+      "step": 7496
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5763216018676758,
+      "learning_rate": 8.084145558522327e-09,
+      "loss": 0.9999,
+      "step": 7497
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3953659534454346,
+      "learning_rate": 7.963968810840162e-09,
+      "loss": 0.8014,
+      "step": 7498
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4147765636444092,
+      "learning_rate": 7.8446913107616e-09,
+      "loss": 0.7814,
+      "step": 7499
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3049609661102295,
+      "learning_rate": 7.726313079771119e-09,
+      "loss": 0.8591,
+      "step": 7500
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5683467388153076,
+      "learning_rate": 7.608834139194443e-09,
+      "loss": 1.0045,
+      "step": 7501
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3870481252670288,
+      "learning_rate": 7.49225451019242e-09,
+      "loss": 0.8,
+      "step": 7502
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.512754201889038,
+      "learning_rate": 7.37657421376603e-09,
+      "loss": 0.8297,
+      "step": 7503
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3875625133514404,
+      "learning_rate": 7.261793270753603e-09,
+      "loss": 0.7963,
+      "step": 7504
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.29007089138031,
+      "learning_rate": 7.147911701830823e-09,
+      "loss": 0.807,
+      "step": 7505
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3876044750213623,
+      "learning_rate": 7.034929527512946e-09,
+      "loss": 0.856,
+      "step": 7506
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3475338220596313,
+      "learning_rate": 6.9228467681509145e-09,
+      "loss": 0.8124,
+      "step": 7507
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5836284160614014,
+      "learning_rate": 6.811663443935801e-09,
+      "loss": 1.0032,
+      "step": 7508
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3852598667144775,
+      "learning_rate": 6.701379574894917e-09,
+      "loss": 0.8606,
+      "step": 7509
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5828558206558228,
+      "learning_rate": 6.591995180895705e-09,
+      "loss": 1.0254,
+      "step": 7510
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.4301782846450806,
+      "learning_rate": 6.483510281641292e-09,
+      "loss": 0.8511,
+      "step": 7511
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.567701518535614,
+      "learning_rate": 6.375924896673824e-09,
+      "loss": 0.9998,
+      "step": 7512
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3207554817199707,
+      "learning_rate": 6.2692390453733545e-09,
+      "loss": 0.8349,
+      "step": 7513
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.3908838033676147,
+      "learning_rate": 6.163452746958398e-09,
+      "loss": 0.7474,
+      "step": 7514
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5772576928138733,
+      "learning_rate": 6.058566020484824e-09,
+      "loss": 1.0242,
+      "step": 7515
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5759236812591553,
+      "learning_rate": 5.954578884845852e-09,
+      "loss": 1.0054,
+      "step": 7516
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.378363013267517,
+      "learning_rate": 5.8514913587737195e-09,
+      "loss": 0.8084,
+      "step": 7517
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5675086379051208,
+      "learning_rate": 5.749303460838573e-09,
+      "loss": 1.0111,
+      "step": 7518
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4325995445251465,
+      "learning_rate": 5.648015209448465e-09,
+      "loss": 0.8471,
+      "step": 7519
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5612800717353821,
+      "learning_rate": 5.547626622848246e-09,
+      "loss": 1.0083,
+      "step": 7520
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5648564696311951,
+      "learning_rate": 5.448137719121782e-09,
+      "loss": 1.0127,
+      "step": 7521
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5725412964820862,
+      "learning_rate": 5.34954851619085e-09,
+      "loss": 1.0124,
+      "step": 7522
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6196553707122803,
+      "learning_rate": 5.251859031814577e-09,
+      "loss": 0.7432,
+      "step": 7523
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5730879306793213,
+      "learning_rate": 5.155069283591108e-09,
+      "loss": 1.0059,
+      "step": 7524
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5798693895339966,
+      "learning_rate": 5.0591792889553845e-09,
+      "loss": 1.0149,
+      "step": 7525
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5912252068519592,
+      "learning_rate": 4.964189065180258e-09,
+      "loss": 1.0213,
+      "step": 7526
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5913587808609009,
+      "learning_rate": 4.870098629377596e-09,
+      "loss": 1.0211,
+      "step": 7527
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 2.075958490371704,
+      "learning_rate": 4.776907998496616e-09,
+      "loss": 0.8192,
+      "step": 7528
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5832540392875671,
+      "learning_rate": 4.684617189323892e-09,
+      "loss": 1.0187,
+      "step": 7529
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2495187520980835,
+      "learning_rate": 4.593226218483904e-09,
+      "loss": 0.8014,
+      "step": 7530
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.289039134979248,
+      "learning_rate": 4.502735102440703e-09,
+      "loss": 0.8069,
+      "step": 7531
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5606937408447266,
+      "learning_rate": 4.413143857494584e-09,
+      "loss": 1.0181,
+      "step": 7532
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.296994686126709,
+      "learning_rate": 4.324452499783749e-09,
+      "loss": 0.8255,
+      "step": 7533
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5951626896858215,
+      "learning_rate": 4.2366610452848625e-09,
+      "loss": 1.0152,
+      "step": 7534
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3407435417175293,
+      "learning_rate": 4.1497695098124954e-09,
+      "loss": 0.83,
+      "step": 7535
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5844160318374634,
+      "learning_rate": 4.063777909019129e-09,
+      "loss": 0.9835,
+      "step": 7536
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.37091863155365,
+      "learning_rate": 3.978686258395148e-09,
+      "loss": 0.8346,
+      "step": 7537
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.412817358970642,
+      "learning_rate": 3.8944945732682925e-09,
+      "loss": 0.7836,
+      "step": 7538
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3303238153457642,
+      "learning_rate": 3.811202868804209e-09,
+      "loss": 0.7915,
+      "step": 7539
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3070162534713745,
+      "learning_rate": 3.728811160007562e-09,
+      "loss": 0.8105,
+      "step": 7540
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4253180027008057,
+      "learning_rate": 3.6473194617198117e-09,
+      "loss": 0.8465,
+      "step": 7541
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5833593606948853,
+      "learning_rate": 3.5667277886203277e-09,
+      "loss": 1.0123,
+      "step": 7542
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3056970834732056,
+      "learning_rate": 3.4870361552269415e-09,
+      "loss": 0.767,
+      "step": 7543
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2700532674789429,
+      "learning_rate": 3.408244575894837e-09,
+      "loss": 0.7769,
+      "step": 7544
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3370598554611206,
+      "learning_rate": 3.3303530648171047e-09,
+      "loss": 0.8084,
+      "step": 7545
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.288732647895813,
+      "learning_rate": 3.253361636024743e-09,
+      "loss": 0.811,
+      "step": 7546
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.452989935874939,
+      "learning_rate": 3.1772703033877693e-09,
+      "loss": 0.8513,
+      "step": 7547
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5830979943275452,
+      "learning_rate": 3.1020790806118863e-09,
+      "loss": 0.9886,
+      "step": 7548
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5684618353843689,
+      "learning_rate": 3.0277879812418145e-09,
+      "loss": 1.0024,
+      "step": 7549
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3728957176208496,
+      "learning_rate": 2.9543970186607375e-09,
+      "loss": 0.8762,
+      "step": 7550
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.313248872756958,
+      "learning_rate": 2.881906206088081e-09,
+      "loss": 0.7912,
+      "step": 7551
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2940144538879395,
+      "learning_rate": 2.8103155565833982e-09,
+      "loss": 0.8217,
+      "step": 7552
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3697054386138916,
+      "learning_rate": 2.7396250830419302e-09,
+      "loss": 0.7806,
+      "step": 7553
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5300439596176147,
+      "learning_rate": 2.6698347981973794e-09,
+      "loss": 0.8097,
+      "step": 7554
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3593873977661133,
+      "learning_rate": 2.6009447146224664e-09,
+      "loss": 0.8071,
+      "step": 7555
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2878553867340088,
+      "learning_rate": 2.532954844725599e-09,
+      "loss": 0.8224,
+      "step": 7556
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2448266744613647,
+      "learning_rate": 2.465865200755313e-09,
+      "loss": 0.7398,
+      "step": 7557
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5850703120231628,
+      "learning_rate": 2.3996757947963855e-09,
+      "loss": 1.0098,
+      "step": 7558
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5814222693443298,
+      "learning_rate": 2.334386638771502e-09,
+      "loss": 1.0039,
+      "step": 7559
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.297684907913208,
+      "learning_rate": 2.26999774444292e-09,
+      "loss": 0.78,
+      "step": 7560
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5758689641952515,
+      "learning_rate": 2.2065091234080295e-09,
+      "loss": 0.9946,
+      "step": 7561
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5855284333229065,
+      "learning_rate": 2.1439207871043477e-09,
+      "loss": 1.025,
+      "step": 7562
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2601063251495361,
+      "learning_rate": 2.0822327468056347e-09,
+      "loss": 0.7954,
+      "step": 7563
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2300113439559937,
+      "learning_rate": 2.0214450136246676e-09,
+      "loss": 0.7909,
+      "step": 7564
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.6314345598220825,
+      "learning_rate": 1.9615575985115766e-09,
+      "loss": 0.8502,
+      "step": 7565
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4395487308502197,
+      "learning_rate": 1.902570512253843e-09,
+      "loss": 0.8847,
+      "step": 7566
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2474889755249023,
+      "learning_rate": 1.8444837654774118e-09,
+      "loss": 0.8087,
+      "step": 7567
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5894939303398132,
+      "learning_rate": 1.7872973686461348e-09,
+      "loss": 1.0062,
+      "step": 7568
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5579019784927368,
+      "learning_rate": 1.7310113320606613e-09,
+      "loss": 1.0047,
+      "step": 7569
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5736956596374512,
+      "learning_rate": 1.6756256658612135e-09,
+      "loss": 1.0174,
+      "step": 7570
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5907270908355713,
+      "learning_rate": 1.6211403800237003e-09,
+      "loss": 1.0218,
+      "step": 7571
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2604225873947144,
+      "learning_rate": 1.5675554843630481e-09,
+      "loss": 0.7697,
+      "step": 7572
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.385612964630127,
+      "learning_rate": 1.5148709885326463e-09,
+      "loss": 0.7449,
+      "step": 7573
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.283604621887207,
+      "learning_rate": 1.463086902021571e-09,
+      "loss": 0.8593,
+      "step": 7574
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2532423734664917,
+      "learning_rate": 1.4122032341595816e-09,
+      "loss": 0.8311,
+      "step": 7575
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5625503063201904,
+      "learning_rate": 1.3622199941115689e-09,
+      "loss": 1.0011,
+      "step": 7576
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5695516467094421,
+      "learning_rate": 1.3131371908819968e-09,
+      "loss": 1.0078,
+      "step": 7577
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5606665015220642,
+      "learning_rate": 1.2649548333115713e-09,
+      "loss": 1.0188,
+      "step": 7578
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.359728217124939,
+      "learning_rate": 1.2176729300811264e-09,
+      "loss": 0.7834,
+      "step": 7579
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2697001695632935,
+      "learning_rate": 1.1712914897060723e-09,
+      "loss": 0.8211,
+      "step": 7580
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5809470415115356,
+      "learning_rate": 1.125810520543058e-09,
+      "loss": 1.0115,
+      "step": 7581
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3373301029205322,
+      "learning_rate": 1.0812300307838642e-09,
+      "loss": 0.7704,
+      "step": 7582
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3449664115905762,
+      "learning_rate": 1.0375500284592888e-09,
+      "loss": 0.8376,
+      "step": 7583
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2924911975860596,
+      "learning_rate": 9.947705214380377e-10,
+      "loss": 0.8466,
+      "step": 7584
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3031929731369019,
+      "learning_rate": 9.528915174256137e-10,
+      "loss": 0.8191,
+      "step": 7585
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.306250810623169,
+      "learning_rate": 9.119130239665374e-10,
+      "loss": 0.8177,
+      "step": 7586
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3525761365890503,
+      "learning_rate": 8.718350484421267e-10,
+      "loss": 0.8303,
+      "step": 7587
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.4336652755737305,
+      "learning_rate": 8.326575980716067e-10,
+      "loss": 0.8315,
+      "step": 7588
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5881242752075195,
+      "learning_rate": 7.943806799132203e-10,
+      "loss": 1.0178,
+      "step": 7589
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.3366508483886719,
+      "learning_rate": 7.570043008614525e-10,
+      "loss": 0.7962,
+      "step": 7590
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.425206184387207,
+      "learning_rate": 7.205284676492508e-10,
+      "loss": 0.8276,
+      "step": 7591
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.5630438327789307,
+      "learning_rate": 6.849531868474702e-10,
+      "loss": 0.848,
+      "step": 7592
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 1.2719708681106567,
+      "learning_rate": 6.502784648637628e-10,
+      "loss": 0.7654,
+      "step": 7593
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3692731857299805,
+      "learning_rate": 6.165043079447986e-10,
+      "loss": 0.8056,
+      "step": 7594
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3577626943588257,
+      "learning_rate": 5.836307221751547e-10,
+      "loss": 0.8643,
+      "step": 7595
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5812339186668396,
+      "learning_rate": 5.516577134750955e-10,
+      "loss": 0.9925,
+      "step": 7596
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5579089522361755,
+      "learning_rate": 5.205852876055684e-10,
+      "loss": 0.9897,
+      "step": 7597
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4927219152450562,
+      "learning_rate": 4.904134501632074e-10,
+      "loss": 0.8474,
+      "step": 7598
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.331015944480896,
+      "learning_rate": 4.6114220658310946e-10,
+      "loss": 0.7458,
+      "step": 7599
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2737207412719727,
+      "learning_rate": 4.3277156213772377e-10,
+      "loss": 0.7551,
+      "step": 7600
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3418809175491333,
+      "learning_rate": 4.053015219379619e-10,
+      "loss": 0.7915,
+      "step": 7601
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.330378532409668,
+      "learning_rate": 3.787320909326431e-10,
+      "loss": 0.8019,
+      "step": 7602
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4333018064498901,
+      "learning_rate": 3.530632739073836e-10,
+      "loss": 0.8183,
+      "step": 7603
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5601443648338318,
+      "learning_rate": 3.282950754862624e-10,
+      "loss": 1.0031,
+      "step": 7604
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5738175511360168,
+      "learning_rate": 3.0442750013071065e-10,
+      "loss": 0.996,
+      "step": 7605
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4396812915802002,
+      "learning_rate": 2.8146055214006705e-10,
+      "loss": 0.8916,
+      "step": 7606
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4448179006576538,
+      "learning_rate": 2.5939423565157773e-10,
+      "loss": 0.8306,
+      "step": 7607
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2631139755249023,
+      "learning_rate": 2.382285546409513e-10,
+      "loss": 0.7589,
+      "step": 7608
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.369419813156128,
+      "learning_rate": 2.179635129195834e-10,
+      "loss": 0.8116,
+      "step": 7609
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.451481580734253,
+      "learning_rate": 1.9859911413899758e-10,
+      "loss": 0.825,
+      "step": 7610
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3897886276245117,
+      "learning_rate": 1.8013536178695946e-10,
+      "loss": 0.8389,
+      "step": 7611
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5548210144042969,
+      "learning_rate": 1.625722591896972e-10,
+      "loss": 0.7915,
+      "step": 7612
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3354486227035522,
+      "learning_rate": 1.4590980951079137e-10,
+      "loss": 0.7672,
+      "step": 7613
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.5399086475372314,
+      "learning_rate": 1.3014801575228497e-10,
+      "loss": 0.8138,
+      "step": 7614
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3669263124465942,
+      "learning_rate": 1.1528688075246319e-10,
+      "loss": 0.8041,
+      "step": 7615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3859407901763916,
+      "learning_rate": 1.0132640718862885e-10,
+      "loss": 0.8058,
+      "step": 7616
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.32549250125885,
+      "learning_rate": 8.82665975759922e-11,
+      "loss": 0.7932,
+      "step": 7617
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5687741637229919,
+      "learning_rate": 7.610745426711585e-11,
+      "loss": 1.0036,
+      "step": 7618
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5757688879966736,
+      "learning_rate": 6.484897945191471e-11,
+      "loss": 0.979,
+      "step": 7619
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2989665269851685,
+      "learning_rate": 5.449117515876623e-11,
+      "loss": 0.8029,
+      "step": 7620
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3851085901260376,
+      "learning_rate": 4.5034043253400214e-11,
+      "loss": 0.8322,
+      "step": 7621
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.3353502750396729,
+      "learning_rate": 3.64775854394539e-11,
+      "loss": 0.8254,
+      "step": 7622
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5688939094543457,
+      "learning_rate": 2.8821803257916835e-11,
+      "loss": 0.9969,
+      "step": 7623
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2350008487701416,
+      "learning_rate": 2.2066698087686022e-11,
+      "loss": 0.8355,
+      "step": 7624
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4670755863189697,
+      "learning_rate": 1.6212271146676116e-11,
+      "loss": 0.7968,
+      "step": 7625
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.583774209022522,
+      "learning_rate": 1.1258523488488771e-11,
+      "loss": 1.0074,
+      "step": 7626
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.2799921035766602,
+      "learning_rate": 7.205456006298406e-12,
+      "loss": 0.8037,
+      "step": 7627
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5878440141677856,
+      "learning_rate": 4.05306942952155e-12,
+      "loss": 0.981,
+      "step": 7628
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.417803406715393,
+      "learning_rate": 1.8013643260372804e-12,
+      "loss": 0.8671,
+      "step": 7629
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5596784353256226,
+      "learning_rate": 4.5034110163211243e-13,
+      "loss": 0.9902,
+      "step": 7630
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.497349500656128,
+      "learning_rate": 0.0,
+      "loss": 0.6985,
+      "step": 7631
+    },
+    {
+      "epoch": 1.0,
+      "step": 7631,
+      "total_flos": 8.505618448871588e+18,
+      "train_loss": 0.9054222581911768,
+      "train_runtime": 46835.4582,
+      "train_samples_per_second": 83.429,
+      "train_steps_per_second": 0.163
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 7631,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1200,
+  "total_flos": 8.505618448871588e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}