diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,57623 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 8226,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00024313153415998054,
+      "grad_norm": 38.5,
+      "learning_rate": 2.5000000000000002e-08,
+      "loss": 1.4348,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004862630683199611,
+      "grad_norm": 20.5,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 0.5402,
+      "step": 2
+    },
+    {
+      "epoch": 0.0007293946024799417,
+      "grad_norm": 37.0,
+      "learning_rate": 7.500000000000001e-08,
+      "loss": 0.9649,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009725261366399222,
+      "grad_norm": 31.125,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 1.6755,
+      "step": 4
+    },
+    {
+      "epoch": 0.0012156576707999028,
+      "grad_norm": 35.75,
+      "learning_rate": 1.2500000000000002e-07,
+      "loss": 1.5522,
+      "step": 5
+    },
+    {
+      "epoch": 0.0014587892049598833,
+      "grad_norm": 23.625,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 1.4726,
+      "step": 6
+    },
+    {
+      "epoch": 0.0017019207391198638,
+      "grad_norm": 26.25,
+      "learning_rate": 1.7500000000000002e-07,
+      "loss": 1.3913,
+      "step": 7
+    },
+    {
+      "epoch": 0.0019450522732798443,
+      "grad_norm": 31.125,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.578,
+      "step": 8
+    },
+    {
+      "epoch": 0.002188183807439825,
+      "grad_norm": 30.0,
+      "learning_rate": 2.2500000000000002e-07,
+      "loss": 1.1628,
+      "step": 9
+    },
+    {
+      "epoch": 0.0024313153415998056,
+      "grad_norm": 27.0,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 1.237,
+      "step": 10
+    },
+    {
+      "epoch": 0.002674446875759786,
+      "grad_norm": 32.25,
+      "learning_rate": 2.75e-07,
+      "loss": 1.5415,
+      "step": 11
+    },
+    {
+      "epoch": 0.0029175784099197666,
+      "grad_norm": 41.25,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 1.7788,
+      "step": 12
+    },
+    {
+      "epoch": 0.003160709944079747,
+      "grad_norm": 33.0,
+      "learning_rate": 3.25e-07,
+      "loss": 1.4964,
+      "step": 13
+    },
+    {
+      "epoch": 0.0034038414782397277,
+      "grad_norm": 21.25,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 1.2345,
+      "step": 14
+    },
+    {
+      "epoch": 0.0036469730123997084,
+      "grad_norm": 33.5,
+      "learning_rate": 3.75e-07,
+      "loss": 1.8139,
+      "step": 15
+    },
+    {
+      "epoch": 0.0038901045465596887,
+      "grad_norm": 28.0,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.3461,
+      "step": 16
+    },
+    {
+      "epoch": 0.004133236080719669,
+      "grad_norm": 29.125,
+      "learning_rate": 4.2500000000000006e-07,
+      "loss": 1.45,
+      "step": 17
+    },
+    {
+      "epoch": 0.00437636761487965,
+      "grad_norm": 43.75,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 1.8464,
+      "step": 18
+    },
+    {
+      "epoch": 0.0046194991490396305,
+      "grad_norm": 26.0,
+      "learning_rate": 4.7500000000000006e-07,
+      "loss": 1.2513,
+      "step": 19
+    },
+    {
+      "epoch": 0.004862630683199611,
+      "grad_norm": 31.375,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 1.8003,
+      "step": 20
+    },
+    {
+      "epoch": 0.005105762217359592,
+      "grad_norm": 27.625,
+      "learning_rate": 5.250000000000001e-07,
+      "loss": 1.2935,
+      "step": 21
+    },
+    {
+      "epoch": 0.005348893751519572,
+      "grad_norm": 24.125,
+      "learning_rate": 5.5e-07,
+      "loss": 1.2846,
+      "step": 22
+    },
+    {
+      "epoch": 0.0055920252856795525,
+      "grad_norm": 39.25,
+      "learning_rate": 5.750000000000001e-07,
+      "loss": 1.5984,
+      "step": 23
+    },
+    {
+      "epoch": 0.005835156819839533,
+      "grad_norm": 22.25,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.1517,
+      "step": 24
+    },
+    {
+      "epoch": 0.006078288353999514,
+      "grad_norm": 49.25,
+      "learning_rate": 6.25e-07,
+      "loss": 1.7331,
+      "step": 25
+    },
+    {
+      "epoch": 0.006321419888159494,
+      "grad_norm": 51.5,
+      "learning_rate": 6.5e-07,
+      "loss": 2.2625,
+      "step": 26
+    },
+    {
+      "epoch": 0.006564551422319475,
+      "grad_norm": 33.5,
+      "learning_rate": 6.750000000000001e-07,
+      "loss": 1.2859,
+      "step": 27
+    },
+    {
+      "epoch": 0.006807682956479455,
+      "grad_norm": 26.625,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 1.4909,
+      "step": 28
+    },
+    {
+      "epoch": 0.007050814490639436,
+      "grad_norm": 35.5,
+      "learning_rate": 7.25e-07,
+      "loss": 0.9688,
+      "step": 29
+    },
+    {
+      "epoch": 0.007293946024799417,
+      "grad_norm": 29.125,
+      "learning_rate": 7.5e-07,
+      "loss": 1.3298,
+      "step": 30
+    },
+    {
+      "epoch": 0.007537077558959397,
+      "grad_norm": 30.5,
+      "learning_rate": 7.750000000000001e-07,
+      "loss": 1.5472,
+      "step": 31
+    },
+    {
+      "epoch": 0.007780209093119377,
+      "grad_norm": 32.0,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.7422,
+      "step": 32
+    },
+    {
+      "epoch": 0.008023340627279359,
+      "grad_norm": 41.5,
+      "learning_rate": 8.250000000000001e-07,
+      "loss": 1.267,
+      "step": 33
+    },
+    {
+      "epoch": 0.008266472161439338,
+      "grad_norm": 24.125,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 0.8925,
+      "step": 34
+    },
+    {
+      "epoch": 0.008509603695599319,
+      "grad_norm": 22.0,
+      "learning_rate": 8.75e-07,
+      "loss": 1.4644,
+      "step": 35
+    },
+    {
+      "epoch": 0.0087527352297593,
+      "grad_norm": 26.5,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.6057,
+      "step": 36
+    },
+    {
+      "epoch": 0.00899586676391928,
+      "grad_norm": 23.25,
+      "learning_rate": 9.25e-07,
+      "loss": 1.0055,
+      "step": 37
+    },
+    {
+      "epoch": 0.009238998298079261,
+      "grad_norm": 28.375,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 1.5588,
+      "step": 38
+    },
+    {
+      "epoch": 0.009482129832239242,
+      "grad_norm": 21.125,
+      "learning_rate": 9.750000000000002e-07,
+      "loss": 1.6439,
+      "step": 39
+    },
+    {
+      "epoch": 0.009725261366399222,
+      "grad_norm": 26.375,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.2073,
+      "step": 40
+    },
+    {
+      "epoch": 0.009968392900559203,
+      "grad_norm": 25.0,
+      "learning_rate": 1.025e-06,
+      "loss": 1.1954,
+      "step": 41
+    },
+    {
+      "epoch": 0.010211524434719184,
+      "grad_norm": 33.25,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 1.4811,
+      "step": 42
+    },
+    {
+      "epoch": 0.010454655968879163,
+      "grad_norm": 29.5,
+      "learning_rate": 1.075e-06,
+      "loss": 1.5775,
+      "step": 43
+    },
+    {
+      "epoch": 0.010697787503039144,
+      "grad_norm": 22.0,
+      "learning_rate": 1.1e-06,
+      "loss": 1.1789,
+      "step": 44
+    },
+    {
+      "epoch": 0.010940919037199124,
+      "grad_norm": 53.75,
+      "learning_rate": 1.125e-06,
+      "loss": 1.3597,
+      "step": 45
+    },
+    {
+      "epoch": 0.011184050571359105,
+      "grad_norm": 28.875,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 1.3105,
+      "step": 46
+    },
+    {
+      "epoch": 0.011427182105519086,
+      "grad_norm": 21.0,
+      "learning_rate": 1.175e-06,
+      "loss": 1.0907,
+      "step": 47
+    },
+    {
+      "epoch": 0.011670313639679067,
+      "grad_norm": 28.125,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.7235,
+      "step": 48
+    },
+    {
+      "epoch": 0.011913445173839047,
+      "grad_norm": 24.875,
+      "learning_rate": 1.2250000000000001e-06,
+      "loss": 1.2318,
+      "step": 49
+    },
+    {
+      "epoch": 0.012156576707999028,
+      "grad_norm": 26.125,
+      "learning_rate": 1.25e-06,
+      "loss": 1.3413,
+      "step": 50
+    },
+    {
+      "epoch": 0.012399708242159009,
+      "grad_norm": 24.5,
+      "learning_rate": 1.275e-06,
+      "loss": 1.2784,
+      "step": 51
+    },
+    {
+      "epoch": 0.012642839776318988,
+      "grad_norm": 29.625,
+      "learning_rate": 1.3e-06,
+      "loss": 1.5526,
+      "step": 52
+    },
+    {
+      "epoch": 0.012885971310478968,
+      "grad_norm": 22.625,
+      "learning_rate": 1.3250000000000002e-06,
+      "loss": 0.7708,
+      "step": 53
+    },
+    {
+      "epoch": 0.01312910284463895,
+      "grad_norm": 19.75,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 1.1115,
+      "step": 54
+    },
+    {
+      "epoch": 0.01337223437879893,
+      "grad_norm": 22.375,
+      "learning_rate": 1.3750000000000002e-06,
+      "loss": 0.8853,
+      "step": 55
+    },
+    {
+      "epoch": 0.01361536591295891,
+      "grad_norm": 31.0,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.9406,
+      "step": 56
+    },
+    {
+      "epoch": 0.013858497447118891,
+      "grad_norm": 21.75,
+      "learning_rate": 1.425e-06,
+      "loss": 1.1941,
+      "step": 57
+    },
+    {
+      "epoch": 0.014101628981278872,
+      "grad_norm": 48.5,
+      "learning_rate": 1.45e-06,
+      "loss": 1.3161,
+      "step": 58
+    },
+    {
+      "epoch": 0.014344760515438853,
+      "grad_norm": 23.125,
+      "learning_rate": 1.475e-06,
+      "loss": 1.2449,
+      "step": 59
+    },
+    {
+      "epoch": 0.014587892049598834,
+      "grad_norm": 24.75,
+      "learning_rate": 1.5e-06,
+      "loss": 1.2071,
+      "step": 60
+    },
+    {
+      "epoch": 0.014831023583758814,
+      "grad_norm": 25.125,
+      "learning_rate": 1.525e-06,
+      "loss": 1.5179,
+      "step": 61
+    },
+    {
+      "epoch": 0.015074155117918793,
+      "grad_norm": 27.0,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 1.3866,
+      "step": 62
+    },
+    {
+      "epoch": 0.015317286652078774,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5750000000000002e-06,
+      "loss": 1.2359,
+      "step": 63
+    },
+    {
+      "epoch": 0.015560418186238755,
+      "grad_norm": 27.375,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.0784,
+      "step": 64
+    },
+    {
+      "epoch": 0.015803549720398737,
+      "grad_norm": 21.75,
+      "learning_rate": 1.6250000000000001e-06,
+      "loss": 1.4198,
+      "step": 65
+    },
+    {
+      "epoch": 0.016046681254558718,
+      "grad_norm": 22.125,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.6945,
+      "step": 66
+    },
+    {
+      "epoch": 0.016289812788718695,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.6750000000000003e-06,
+      "loss": 0.8836,
+      "step": 67
+    },
+    {
+      "epoch": 0.016532944322878676,
+      "grad_norm": 24.25,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.1728,
+      "step": 68
+    },
+    {
+      "epoch": 0.016776075857038657,
+      "grad_norm": 26.875,
+      "learning_rate": 1.725e-06,
+      "loss": 1.2629,
+      "step": 69
+    },
+    {
+      "epoch": 0.017019207391198637,
+      "grad_norm": 18.125,
+      "learning_rate": 1.75e-06,
+      "loss": 1.0292,
+      "step": 70
+    },
+    {
+      "epoch": 0.017262338925358618,
+      "grad_norm": 27.875,
+      "learning_rate": 1.7750000000000002e-06,
+      "loss": 1.4961,
+      "step": 71
+    },
+    {
+      "epoch": 0.0175054704595186,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.4118,
+      "step": 72
+    },
+    {
+      "epoch": 0.01774860199367858,
+      "grad_norm": 23.125,
+      "learning_rate": 1.825e-06,
+      "loss": 1.0297,
+      "step": 73
+    },
+    {
+      "epoch": 0.01799173352783856,
+      "grad_norm": 19.125,
+      "learning_rate": 1.85e-06,
+      "loss": 1.0484,
+      "step": 74
+    },
+    {
+      "epoch": 0.01823486506199854,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.9174,
+      "step": 75
+    },
+    {
+      "epoch": 0.018477996596158522,
+      "grad_norm": 31.0,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.2225,
+      "step": 76
+    },
+    {
+      "epoch": 0.018721128130318503,
+      "grad_norm": 22.375,
+      "learning_rate": 1.925e-06,
+      "loss": 1.0815,
+      "step": 77
+    },
+    {
+      "epoch": 0.018964259664478483,
+      "grad_norm": 38.5,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 1.2646,
+      "step": 78
+    },
+    {
+      "epoch": 0.019207391198638464,
+      "grad_norm": 23.375,
+      "learning_rate": 1.975e-06,
+      "loss": 1.1941,
+      "step": 79
+    },
+    {
+      "epoch": 0.019450522732798445,
+      "grad_norm": 23.375,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.0209,
+      "step": 80
+    },
+    {
+      "epoch": 0.019693654266958426,
+      "grad_norm": 22.0,
+      "learning_rate": 2.025e-06,
+      "loss": 1.1193,
+      "step": 81
+    },
+    {
+      "epoch": 0.019936785801118406,
+      "grad_norm": 17.875,
+      "learning_rate": 2.05e-06,
+      "loss": 0.7613,
+      "step": 82
+    },
+    {
+      "epoch": 0.020179917335278387,
+      "grad_norm": 20.25,
+      "learning_rate": 2.075e-06,
+      "loss": 1.0781,
+      "step": 83
+    },
+    {
+      "epoch": 0.020423048869438368,
+      "grad_norm": 21.75,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 1.1135,
+      "step": 84
+    },
+    {
+      "epoch": 0.020666180403598345,
+      "grad_norm": 24.125,
+      "learning_rate": 2.125e-06,
+      "loss": 1.4294,
+      "step": 85
+    },
+    {
+      "epoch": 0.020909311937758326,
+      "grad_norm": 25.625,
+      "learning_rate": 2.15e-06,
+      "loss": 1.3782,
+      "step": 86
+    },
+    {
+      "epoch": 0.021152443471918306,
+      "grad_norm": 23.875,
+      "learning_rate": 2.1750000000000004e-06,
+      "loss": 0.9395,
+      "step": 87
+    },
+    {
+      "epoch": 0.021395575006078287,
+      "grad_norm": 23.125,
+      "learning_rate": 2.2e-06,
+      "loss": 1.1842,
+      "step": 88
+    },
+    {
+      "epoch": 0.021638706540238268,
+      "grad_norm": 24.75,
+      "learning_rate": 2.2250000000000003e-06,
+      "loss": 1.1139,
+      "step": 89
+    },
+    {
+      "epoch": 0.02188183807439825,
+      "grad_norm": 26.25,
+      "learning_rate": 2.25e-06,
+      "loss": 1.0698,
+      "step": 90
+    },
+    {
+      "epoch": 0.02212496960855823,
+      "grad_norm": 24.125,
+      "learning_rate": 2.2750000000000002e-06,
+      "loss": 0.6632,
+      "step": 91
+    },
+    {
+      "epoch": 0.02236810114271821,
+      "grad_norm": 49.0,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 1.2604,
+      "step": 92
+    },
+    {
+      "epoch": 0.02261123267687819,
+      "grad_norm": 19.0,
+      "learning_rate": 2.325e-06,
+      "loss": 0.6398,
+      "step": 93
+    },
+    {
+      "epoch": 0.02285436421103817,
+      "grad_norm": 21.0,
+      "learning_rate": 2.35e-06,
+      "loss": 1.2111,
+      "step": 94
+    },
+    {
+      "epoch": 0.023097495745198152,
+      "grad_norm": 24.625,
+      "learning_rate": 2.375e-06,
+      "loss": 1.0793,
+      "step": 95
+    },
+    {
+      "epoch": 0.023340627279358133,
+      "grad_norm": 18.625,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.6156,
+      "step": 96
+    },
+    {
+      "epoch": 0.023583758813518114,
+      "grad_norm": 20.875,
+      "learning_rate": 2.425e-06,
+      "loss": 0.8457,
+      "step": 97
+    },
+    {
+      "epoch": 0.023826890347678095,
+      "grad_norm": 28.875,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 1.2779,
+      "step": 98
+    },
+    {
+      "epoch": 0.024070021881838075,
+      "grad_norm": 22.625,
+      "learning_rate": 2.475e-06,
+      "loss": 1.4615,
+      "step": 99
+    },
+    {
+      "epoch": 0.024313153415998056,
+      "grad_norm": 23.125,
+      "learning_rate": 2.5e-06,
+      "loss": 1.1989,
+      "step": 100
+    },
+    {
+      "epoch": 0.024556284950158037,
+      "grad_norm": 22.375,
+      "learning_rate": 2.499999906582956e-06,
+      "loss": 1.3843,
+      "step": 101
+    },
+    {
+      "epoch": 0.024799416484318017,
+      "grad_norm": 19.5,
+      "learning_rate": 2.4999996263318378e-06,
+      "loss": 0.7177,
+      "step": 102
+    },
+    {
+      "epoch": 0.025042548018477998,
+      "grad_norm": 21.875,
+      "learning_rate": 2.4999991592466867e-06,
+      "loss": 1.0586,
+      "step": 103
+    },
+    {
+      "epoch": 0.025285679552637975,
+      "grad_norm": 22.375,
+      "learning_rate": 2.4999985053275737e-06,
+      "loss": 1.253,
+      "step": 104
+    },
+    {
+      "epoch": 0.025528811086797956,
+      "grad_norm": 25.5,
+      "learning_rate": 2.499997664574595e-06,
+      "loss": 1.1569,
+      "step": 105
+    },
+    {
+      "epoch": 0.025771942620957937,
+      "grad_norm": 23.0,
+      "learning_rate": 2.499996636987878e-06,
+      "loss": 1.1424,
+      "step": 106
+    },
+    {
+      "epoch": 0.026015074155117918,
+      "grad_norm": 19.25,
+      "learning_rate": 2.499995422567575e-06,
+      "loss": 1.1592,
+      "step": 107
+    },
+    {
+      "epoch": 0.0262582056892779,
+      "grad_norm": 20.0,
+      "learning_rate": 2.499994021313868e-06,
+      "loss": 0.9183,
+      "step": 108
+    },
+    {
+      "epoch": 0.02650133722343788,
+      "grad_norm": 15.75,
+      "learning_rate": 2.499992433226966e-06,
+      "loss": 0.815,
+      "step": 109
+    },
+    {
+      "epoch": 0.02674446875759786,
+      "grad_norm": 23.875,
+      "learning_rate": 2.499990658307107e-06,
+      "loss": 0.9233,
+      "step": 110
+    },
+    {
+      "epoch": 0.02698760029175784,
+      "grad_norm": 19.625,
+      "learning_rate": 2.499988696554556e-06,
+      "loss": 1.0855,
+      "step": 111
+    },
+    {
+      "epoch": 0.02723073182591782,
+      "grad_norm": 23.375,
+      "learning_rate": 2.499986547969607e-06,
+      "loss": 1.3179,
+      "step": 112
+    },
+    {
+      "epoch": 0.027473863360077802,
+      "grad_norm": 18.75,
+      "learning_rate": 2.499984212552579e-06,
+      "loss": 1.1547,
+      "step": 113
+    },
+    {
+      "epoch": 0.027716994894237783,
+      "grad_norm": 16.625,
+      "learning_rate": 2.4999816903038236e-06,
+      "loss": 0.8664,
+      "step": 114
+    },
+    {
+      "epoch": 0.027960126428397764,
+      "grad_norm": 23.75,
+      "learning_rate": 2.499978981223716e-06,
+      "loss": 1.032,
+      "step": 115
+    },
+    {
+      "epoch": 0.028203257962557744,
+      "grad_norm": 20.25,
+      "learning_rate": 2.499976085312662e-06,
+      "loss": 0.8514,
+      "step": 116
+    },
+    {
+      "epoch": 0.028446389496717725,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4999730025710945e-06,
+      "loss": 1.1174,
+      "step": 117
+    },
+    {
+      "epoch": 0.028689521030877706,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.4999697329994736e-06,
+      "loss": 1.0144,
+      "step": 118
+    },
+    {
+      "epoch": 0.028932652565037686,
+      "grad_norm": 22.125,
+      "learning_rate": 2.4999662765982884e-06,
+      "loss": 1.0733,
+      "step": 119
+    },
+    {
+      "epoch": 0.029175784099197667,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4999626333680554e-06,
+      "loss": 0.8249,
+      "step": 120
+    },
+    {
+      "epoch": 0.029418915633357648,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.49995880330932e-06,
+      "loss": 0.7594,
+      "step": 121
+    },
+    {
+      "epoch": 0.02966204716751763,
+      "grad_norm": 18.0,
+      "learning_rate": 2.4999547864226532e-06,
+      "loss": 1.0326,
+      "step": 122
+    },
+    {
+      "epoch": 0.029905178701677606,
+      "grad_norm": 26.875,
+      "learning_rate": 2.499950582708656e-06,
+      "loss": 0.809,
+      "step": 123
+    },
+    {
+      "epoch": 0.030148310235837587,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4999461921679567e-06,
+      "loss": 0.9553,
+      "step": 124
+    },
+    {
+      "epoch": 0.030391441769997567,
+      "grad_norm": 22.75,
+      "learning_rate": 2.4999416148012122e-06,
+      "loss": 1.0211,
+      "step": 125
+    },
+    {
+      "epoch": 0.030634573304157548,
+      "grad_norm": 25.0,
+      "learning_rate": 2.499936850609106e-06,
+      "loss": 1.1137,
+      "step": 126
+    },
+    {
+      "epoch": 0.03087770483831753,
+      "grad_norm": 21.5,
+      "learning_rate": 2.4999318995923507e-06,
+      "loss": 0.7679,
+      "step": 127
+    },
+    {
+      "epoch": 0.03112083637247751,
+      "grad_norm": 44.25,
+      "learning_rate": 2.499926761751685e-06,
+      "loss": 1.2556,
+      "step": 128
+    },
+    {
+      "epoch": 0.031363967906637494,
+      "grad_norm": 21.5,
+      "learning_rate": 2.499921437087878e-06,
+      "loss": 0.8554,
+      "step": 129
+    },
+    {
+      "epoch": 0.031607099440797475,
+      "grad_norm": 19.25,
+      "learning_rate": 2.499915925601726e-06,
+      "loss": 0.9156,
+      "step": 130
+    },
+    {
+      "epoch": 0.031850230974957455,
+      "grad_norm": 21.125,
+      "learning_rate": 2.4999102272940516e-06,
+      "loss": 1.0176,
+      "step": 131
+    },
+    {
+      "epoch": 0.032093362509117436,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4999043421657075e-06,
+      "loss": 0.9295,
+      "step": 132
+    },
+    {
+      "epoch": 0.03233649404327741,
+      "grad_norm": 20.125,
+      "learning_rate": 2.499898270217572e-06,
+      "loss": 0.8986,
+      "step": 133
+    },
+    {
+      "epoch": 0.03257962557743739,
+      "grad_norm": 26.0,
+      "learning_rate": 2.499892011450554e-06,
+      "loss": 1.1765,
+      "step": 134
+    },
+    {
+      "epoch": 0.03282275711159737,
+      "grad_norm": 20.875,
+      "learning_rate": 2.499885565865589e-06,
+      "loss": 1.0191,
+      "step": 135
+    },
+    {
+      "epoch": 0.03306588864575735,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.4998789334636393e-06,
+      "loss": 0.9031,
+      "step": 136
+    },
+    {
+      "epoch": 0.03330902017991733,
+      "grad_norm": 19.0,
+      "learning_rate": 2.499872114245697e-06,
+      "loss": 1.0915,
+      "step": 137
+    },
+    {
+      "epoch": 0.03355215171407731,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4998651082127815e-06,
+      "loss": 0.7411,
+      "step": 138
+    },
+    {
+      "epoch": 0.033795283248237294,
+      "grad_norm": 22.75,
+      "learning_rate": 2.4998579153659393e-06,
+      "loss": 1.328,
+      "step": 139
+    },
+    {
+      "epoch": 0.034038414782397275,
+      "grad_norm": 16.25,
+      "learning_rate": 2.4998505357062457e-06,
+      "loss": 0.7537,
+      "step": 140
+    },
+    {
+      "epoch": 0.034281546316557256,
+      "grad_norm": 20.125,
+      "learning_rate": 2.499842969234804e-06,
+      "loss": 1.1528,
+      "step": 141
+    },
+    {
+      "epoch": 0.034524677850717236,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4998352159527458e-06,
+      "loss": 1.0924,
+      "step": 142
+    },
+    {
+      "epoch": 0.03476780938487722,
+      "grad_norm": 19.25,
+      "learning_rate": 2.499827275861228e-06,
+      "loss": 1.127,
+      "step": 143
+    },
+    {
+      "epoch": 0.0350109409190372,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4998191489614393e-06,
+      "loss": 1.0547,
+      "step": 144
+    },
+    {
+      "epoch": 0.03525407245319718,
+      "grad_norm": 29.25,
+      "learning_rate": 2.4998108352545933e-06,
+      "loss": 1.1,
+      "step": 145
+    },
+    {
+      "epoch": 0.03549720398735716,
+      "grad_norm": 20.625,
+      "learning_rate": 2.499802334741933e-06,
+      "loss": 0.8825,
+      "step": 146
+    },
+    {
+      "epoch": 0.03574033552151714,
+      "grad_norm": 18.75,
+      "learning_rate": 2.499793647424729e-06,
+      "loss": 0.9249,
+      "step": 147
+    },
+    {
+      "epoch": 0.03598346705567712,
+      "grad_norm": 20.25,
+      "learning_rate": 2.49978477330428e-06,
+      "loss": 1.0968,
+      "step": 148
+    },
+    {
+      "epoch": 0.0362265985898371,
+      "grad_norm": 26.375,
+      "learning_rate": 2.4997757123819117e-06,
+      "loss": 0.6389,
+      "step": 149
+    },
+    {
+      "epoch": 0.03646973012399708,
+      "grad_norm": 18.875,
+      "learning_rate": 2.499766464658979e-06,
+      "loss": 1.0685,
+      "step": 150
+    },
+    {
+      "epoch": 0.03671286165815706,
+      "grad_norm": 17.375,
+      "learning_rate": 2.499757030136864e-06,
+      "loss": 1.1539,
+      "step": 151
+    },
+    {
+      "epoch": 0.036955993192317044,
+      "grad_norm": 35.0,
+      "learning_rate": 2.4997474088169764e-06,
+      "loss": 1.4932,
+      "step": 152
+    },
+    {
+      "epoch": 0.037199124726477024,
+      "grad_norm": 19.625,
+      "learning_rate": 2.499737600700755e-06,
+      "loss": 0.8824,
+      "step": 153
+    },
+    {
+      "epoch": 0.037442256260637005,
+      "grad_norm": 26.25,
+      "learning_rate": 2.4997276057896656e-06,
+      "loss": 0.9399,
+      "step": 154
+    },
+    {
+      "epoch": 0.037685387794796986,
+      "grad_norm": 28.125,
+      "learning_rate": 2.499717424085202e-06,
+      "loss": 1.3844,
+      "step": 155
+    },
+    {
+      "epoch": 0.03792851932895697,
+      "grad_norm": 22.625,
+      "learning_rate": 2.4997070555888855e-06,
+      "loss": 1.4123,
+      "step": 156
+    },
+    {
+      "epoch": 0.03817165086311695,
+      "grad_norm": 35.0,
+      "learning_rate": 2.4996965003022667e-06,
+      "loss": 0.8678,
+      "step": 157
+    },
+    {
+      "epoch": 0.03841478239727693,
+      "grad_norm": 16.0,
+      "learning_rate": 2.499685758226923e-06,
+      "loss": 0.7204,
+      "step": 158
+    },
+    {
+      "epoch": 0.03865791393143691,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4996748293644597e-06,
+      "loss": 1.0738,
+      "step": 159
+    },
+    {
+      "epoch": 0.03890104546559689,
+      "grad_norm": 23.25,
+      "learning_rate": 2.4996637137165106e-06,
+      "loss": 0.9177,
+      "step": 160
+    },
+    {
+      "epoch": 0.03914417699975687,
+      "grad_norm": 21.375,
+      "learning_rate": 2.499652411284737e-06,
+      "loss": 1.3205,
+      "step": 161
+    },
+    {
+      "epoch": 0.03938730853391685,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4996409220708282e-06,
+      "loss": 0.6068,
+      "step": 162
+    },
+    {
+      "epoch": 0.03963044006807683,
+      "grad_norm": 21.5,
+      "learning_rate": 2.499629246076502e-06,
+      "loss": 1.1155,
+      "step": 163
+    },
+    {
+      "epoch": 0.03987357160223681,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4996173833035027e-06,
+      "loss": 1.0647,
+      "step": 164
+    },
+    {
+      "epoch": 0.04011670313639679,
+      "grad_norm": 25.125,
+      "learning_rate": 2.4996053337536043e-06,
+      "loss": 1.0785,
+      "step": 165
+    },
+    {
+      "epoch": 0.040359834670556774,
+      "grad_norm": 21.75,
+      "learning_rate": 2.4995930974286067e-06,
+      "loss": 1.1023,
+      "step": 166
+    },
+    {
+      "epoch": 0.040602966204716755,
+      "grad_norm": 21.25,
+      "learning_rate": 2.4995806743303396e-06,
+      "loss": 1.0521,
+      "step": 167
+    },
+    {
+      "epoch": 0.040846097738876735,
+      "grad_norm": 28.625,
+      "learning_rate": 2.49956806446066e-06,
+      "loss": 1.0485,
+      "step": 168
+    },
+    {
+      "epoch": 0.041089229273036716,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4995552678214523e-06,
+      "loss": 1.2728,
+      "step": 169
+    },
+    {
+      "epoch": 0.04133236080719669,
+      "grad_norm": 18.25,
+      "learning_rate": 2.499542284414629e-06,
+      "loss": 0.8282,
+      "step": 170
+    },
+    {
+      "epoch": 0.04157549234135667,
+      "grad_norm": 22.5,
+      "learning_rate": 2.4995291142421315e-06,
+      "loss": 1.1808,
+      "step": 171
+    },
+    {
+      "epoch": 0.04181862387551665,
+      "grad_norm": 16.125,
+      "learning_rate": 2.4995157573059274e-06,
+      "loss": 0.6348,
+      "step": 172
+    },
+    {
+      "epoch": 0.04206175540967663,
+      "grad_norm": 25.125,
+      "learning_rate": 2.499502213608013e-06,
+      "loss": 1.0446,
+      "step": 173
+    },
+    {
+      "epoch": 0.04230488694383661,
+      "grad_norm": 15.875,
+      "learning_rate": 2.499488483150414e-06,
+      "loss": 0.6286,
+      "step": 174
+    },
+    {
+      "epoch": 0.042548018477996594,
+      "grad_norm": 22.625,
+      "learning_rate": 2.4994745659351815e-06,
+      "loss": 1.149,
+      "step": 175
+    },
+    {
+      "epoch": 0.042791150012156574,
+      "grad_norm": 21.75,
+      "learning_rate": 2.4994604619643957e-06,
+      "loss": 1.2704,
+      "step": 176
+    },
+    {
+      "epoch": 0.043034281546316555,
+      "grad_norm": 22.125,
+      "learning_rate": 2.4994461712401652e-06,
+      "loss": 1.3536,
+      "step": 177
+    },
+    {
+      "epoch": 0.043277413080476536,
+      "grad_norm": 16.875,
+      "learning_rate": 2.4994316937646258e-06,
+      "loss": 0.7183,
+      "step": 178
+    },
+    {
+      "epoch": 0.04352054461463652,
+      "grad_norm": 23.375,
+      "learning_rate": 2.499417029539941e-06,
+      "loss": 0.9442,
+      "step": 179
+    },
+    {
+      "epoch": 0.0437636761487965,
+      "grad_norm": 16.0,
+      "learning_rate": 2.499402178568303e-06,
+      "loss": 0.968,
+      "step": 180
+    },
+    {
+      "epoch": 0.04400680768295648,
+      "grad_norm": 19.875,
+      "learning_rate": 2.499387140851932e-06,
+      "loss": 1.123,
+      "step": 181
+    },
+    {
+      "epoch": 0.04424993921711646,
+      "grad_norm": 18.25,
+      "learning_rate": 2.4993719163930745e-06,
+      "loss": 0.7866,
+      "step": 182
+    },
+    {
+      "epoch": 0.04449307075127644,
+      "grad_norm": 26.5,
+      "learning_rate": 2.4993565051940072e-06,
+      "loss": 1.1286,
+      "step": 183
+    },
+    {
+      "epoch": 0.04473620228543642,
+      "grad_norm": 16.5,
+      "learning_rate": 2.4993409072570328e-06,
+      "loss": 0.7304,
+      "step": 184
+    },
+    {
+      "epoch": 0.0449793338195964,
+      "grad_norm": 24.5,
+      "learning_rate": 2.4993251225844826e-06,
+      "loss": 1.2202,
+      "step": 185
+    },
+    {
+      "epoch": 0.04522246535375638,
+      "grad_norm": 23.125,
+      "learning_rate": 2.499309151178717e-06,
+      "loss": 1.2155,
+      "step": 186
+    },
+    {
+      "epoch": 0.04546559688791636,
+      "grad_norm": 16.25,
+      "learning_rate": 2.4992929930421215e-06,
+      "loss": 0.9387,
+      "step": 187
+    },
+    {
+      "epoch": 0.04570872842207634,
+      "grad_norm": 16.375,
+      "learning_rate": 2.499276648177113e-06,
+      "loss": 0.6564,
+      "step": 188
+    },
+    {
+      "epoch": 0.045951859956236324,
+      "grad_norm": 20.0,
+      "learning_rate": 2.4992601165861334e-06,
+      "loss": 0.8013,
+      "step": 189
+    },
+    {
+      "epoch": 0.046194991490396305,
+      "grad_norm": 20.75,
+      "learning_rate": 2.499243398271654e-06,
+      "loss": 1.3058,
+      "step": 190
+    },
+    {
+      "epoch": 0.046438123024556285,
+      "grad_norm": 25.0,
+      "learning_rate": 2.499226493236173e-06,
+      "loss": 1.3281,
+      "step": 191
+    },
+    {
+      "epoch": 0.046681254558716266,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4992094014822182e-06,
+      "loss": 0.7938,
+      "step": 192
+    },
+    {
+      "epoch": 0.04692438609287625,
+      "grad_norm": 19.625,
+      "learning_rate": 2.499192123012344e-06,
+      "loss": 1.0102,
+      "step": 193
+    },
+    {
+      "epoch": 0.04716751762703623,
+      "grad_norm": 28.75,
+      "learning_rate": 2.499174657829132e-06,
+      "loss": 1.401,
+      "step": 194
+    },
+    {
+      "epoch": 0.04741064916119621,
+      "grad_norm": 24.0,
+      "learning_rate": 2.499157005935194e-06,
+      "loss": 0.6036,
+      "step": 195
+    },
+    {
+      "epoch": 0.04765378069535619,
+      "grad_norm": 21.375,
+      "learning_rate": 2.499139167333168e-06,
+      "loss": 1.1887,
+      "step": 196
+    },
+    {
+      "epoch": 0.04789691222951617,
+      "grad_norm": 23.0,
+      "learning_rate": 2.4991211420257195e-06,
+      "loss": 1.0193,
+      "step": 197
+    },
+    {
+      "epoch": 0.04814004376367615,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.4991029300155432e-06,
+      "loss": 0.7501,
+      "step": 198
+    },
+    {
+      "epoch": 0.04838317529783613,
+      "grad_norm": 18.125,
+      "learning_rate": 2.499084531305361e-06,
+      "loss": 1.0581,
+      "step": 199
+    },
+    {
+      "epoch": 0.04862630683199611,
+      "grad_norm": 19.375,
+      "learning_rate": 2.499065945897924e-06,
+      "loss": 0.9109,
+      "step": 200
+    },
+    {
+      "epoch": 0.04886943836615609,
+      "grad_norm": 18.125,
+      "learning_rate": 2.4990471737960086e-06,
+      "loss": 0.59,
+      "step": 201
+    },
+    {
+      "epoch": 0.04911256990031607,
+      "grad_norm": 22.75,
+      "learning_rate": 2.499028215002422e-06,
+      "loss": 1.4277,
+      "step": 202
+    },
+    {
+      "epoch": 0.049355701434476054,
+      "grad_norm": 18.25,
+      "learning_rate": 2.4990090695199964e-06,
+      "loss": 0.7634,
+      "step": 203
+    },
+    {
+      "epoch": 0.049598832968636035,
+      "grad_norm": 18.25,
+      "learning_rate": 2.498989737351595e-06,
+      "loss": 1.0516,
+      "step": 204
+    },
+    {
+      "epoch": 0.049841964502796016,
+      "grad_norm": 20.0,
+      "learning_rate": 2.498970218500106e-06,
+      "loss": 0.9042,
+      "step": 205
+    },
+    {
+      "epoch": 0.050085096036955996,
+      "grad_norm": 16.375,
+      "learning_rate": 2.4989505129684473e-06,
+      "loss": 0.8591,
+      "step": 206
+    },
+    {
+      "epoch": 0.05032822757111598,
+      "grad_norm": 19.625,
+      "learning_rate": 2.498930620759565e-06,
+      "loss": 0.9474,
+      "step": 207
+    },
+    {
+      "epoch": 0.05057135910527595,
+      "grad_norm": 19.0,
+      "learning_rate": 2.498910541876431e-06,
+      "loss": 0.8818,
+      "step": 208
+    },
+    {
+      "epoch": 0.05081449063943593,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4988902763220472e-06,
+      "loss": 0.8594,
+      "step": 209
+    },
+    {
+      "epoch": 0.05105762217359591,
+      "grad_norm": 27.25,
+      "learning_rate": 2.4988698240994427e-06,
+      "loss": 1.1321,
+      "step": 210
+    },
+    {
+      "epoch": 0.05130075370775589,
+      "grad_norm": 25.375,
+      "learning_rate": 2.498849185211674e-06,
+      "loss": 1.2577,
+      "step": 211
+    },
+    {
+      "epoch": 0.051543885241915874,
+      "grad_norm": 15.375,
+      "learning_rate": 2.4988283596618264e-06,
+      "loss": 0.5451,
+      "step": 212
+    },
+    {
+      "epoch": 0.051787016776075855,
+      "grad_norm": 20.25,
+      "learning_rate": 2.498807347453012e-06,
+      "loss": 0.8714,
+      "step": 213
+    },
+    {
+      "epoch": 0.052030148310235835,
+      "grad_norm": 30.375,
+      "learning_rate": 2.4987861485883726e-06,
+      "loss": 1.3122,
+      "step": 214
+    },
+    {
+      "epoch": 0.052273279844395816,
+      "grad_norm": 20.375,
+      "learning_rate": 2.4987647630710757e-06,
+      "loss": 0.9451,
+      "step": 215
+    },
+    {
+      "epoch": 0.0525164113785558,
+      "grad_norm": 21.25,
+      "learning_rate": 2.498743190904318e-06,
+      "loss": 0.9435,
+      "step": 216
+    },
+    {
+      "epoch": 0.05275954291271578,
+      "grad_norm": 23.875,
+      "learning_rate": 2.498721432091324e-06,
+      "loss": 1.2594,
+      "step": 217
+    },
+    {
+      "epoch": 0.05300267444687576,
+      "grad_norm": 21.375,
+      "learning_rate": 2.498699486635346e-06,
+      "loss": 1.0688,
+      "step": 218
+    },
+    {
+      "epoch": 0.05324580598103574,
+      "grad_norm": 20.0,
+      "learning_rate": 2.498677354539663e-06,
+      "loss": 0.9037,
+      "step": 219
+    },
+    {
+      "epoch": 0.05348893751519572,
+      "grad_norm": 20.125,
+      "learning_rate": 2.498655035807585e-06,
+      "loss": 0.8801,
+      "step": 220
+    },
+    {
+      "epoch": 0.0537320690493557,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4986325304424465e-06,
+      "loss": 1.0117,
+      "step": 221
+    },
+    {
+      "epoch": 0.05397520058351568,
+      "grad_norm": 24.375,
+      "learning_rate": 2.4986098384476116e-06,
+      "loss": 0.8751,
+      "step": 222
+    },
+    {
+      "epoch": 0.05421833211767566,
+      "grad_norm": 16.25,
+      "learning_rate": 2.4985869598264724e-06,
+      "loss": 0.8845,
+      "step": 223
+    },
+    {
+      "epoch": 0.05446146365183564,
+      "grad_norm": 16.75,
+      "learning_rate": 2.498563894582448e-06,
+      "loss": 0.7487,
+      "step": 224
+    },
+    {
+      "epoch": 0.05470459518599562,
+      "grad_norm": 20.125,
+      "learning_rate": 2.498540642718986e-06,
+      "loss": 0.7439,
+      "step": 225
+    },
+    {
+      "epoch": 0.054947726720155604,
+      "grad_norm": 25.0,
+      "learning_rate": 2.4985172042395617e-06,
+      "loss": 1.1016,
+      "step": 226
+    },
+    {
+      "epoch": 0.055190858254315585,
+      "grad_norm": 15.625,
+      "learning_rate": 2.498493579147679e-06,
+      "loss": 0.9374,
+      "step": 227
+    },
+    {
+      "epoch": 0.055433989788475566,
+      "grad_norm": 28.0,
+      "learning_rate": 2.4984697674468688e-06,
+      "loss": 1.1315,
+      "step": 228
+    },
+    {
+      "epoch": 0.055677121322635546,
+      "grad_norm": 24.625,
+      "learning_rate": 2.4984457691406896e-06,
+      "loss": 1.1034,
+      "step": 229
+    },
+    {
+      "epoch": 0.05592025285679553,
+      "grad_norm": 19.125,
+      "learning_rate": 2.498421584232729e-06,
+      "loss": 0.92,
+      "step": 230
+    },
+    {
+      "epoch": 0.05616338439095551,
+      "grad_norm": 16.875,
+      "learning_rate": 2.4983972127266015e-06,
+      "loss": 0.7087,
+      "step": 231
+    },
+    {
+      "epoch": 0.05640651592511549,
+      "grad_norm": 20.75,
+      "learning_rate": 2.49837265462595e-06,
+      "loss": 1.2363,
+      "step": 232
+    },
+    {
+      "epoch": 0.05664964745927547,
+      "grad_norm": 21.75,
+      "learning_rate": 2.4983479099344454e-06,
+      "loss": 0.8479,
+      "step": 233
+    },
+    {
+      "epoch": 0.05689277899343545,
+      "grad_norm": 20.375,
+      "learning_rate": 2.498322978655786e-06,
+      "loss": 1.3034,
+      "step": 234
+    },
+    {
+      "epoch": 0.05713591052759543,
+      "grad_norm": 17.0,
+      "learning_rate": 2.498297860793698e-06,
+      "loss": 0.9552,
+      "step": 235
+    },
+    {
+      "epoch": 0.05737904206175541,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.4982725563519357e-06,
+      "loss": 0.5749,
+      "step": 236
+    },
+    {
+      "epoch": 0.05762217359591539,
+      "grad_norm": 20.875,
+      "learning_rate": 2.4982470653342816e-06,
+      "loss": 1.4229,
+      "step": 237
+    },
+    {
+      "epoch": 0.05786530513007537,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4982213877445456e-06,
+      "loss": 0.966,
+      "step": 238
+    },
+    {
+      "epoch": 0.058108436664235354,
+      "grad_norm": 19.625,
+      "learning_rate": 2.4981955235865657e-06,
+      "loss": 0.9087,
+      "step": 239
+    },
+    {
+      "epoch": 0.058351568198395334,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4981694728642077e-06,
+      "loss": 1.1955,
+      "step": 240
+    },
+    {
+      "epoch": 0.058594699732555315,
+      "grad_norm": 19.5,
+      "learning_rate": 2.498143235581365e-06,
+      "loss": 0.8822,
+      "step": 241
+    },
+    {
+      "epoch": 0.058837831266715296,
+      "grad_norm": 20.25,
+      "learning_rate": 2.49811681174196e-06,
+      "loss": 1.0489,
+      "step": 242
+    },
+    {
+      "epoch": 0.05908096280087528,
+      "grad_norm": 22.5,
+      "learning_rate": 2.4980902013499417e-06,
+      "loss": 1.1603,
+      "step": 243
+    },
+    {
+      "epoch": 0.05932409433503526,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4980634044092877e-06,
+      "loss": 0.8474,
+      "step": 244
+    },
+    {
+      "epoch": 0.05956722586919524,
+      "grad_norm": 21.625,
+      "learning_rate": 2.498036420924003e-06,
+      "loss": 0.9012,
+      "step": 245
+    },
+    {
+      "epoch": 0.05981035740335521,
+      "grad_norm": 18.625,
+      "learning_rate": 2.4980092508981204e-06,
+      "loss": 1.0037,
+      "step": 246
+    },
+    {
+      "epoch": 0.06005348893751519,
+      "grad_norm": 18.125,
+      "learning_rate": 2.497981894335702e-06,
+      "loss": 0.7192,
+      "step": 247
+    },
+    {
+      "epoch": 0.06029662047167517,
+      "grad_norm": 17.375,
+      "learning_rate": 2.4979543512408353e-06,
+      "loss": 1.0596,
+      "step": 248
+    },
+    {
+      "epoch": 0.060539752005835154,
+      "grad_norm": 21.75,
+      "learning_rate": 2.497926621617639e-06,
+      "loss": 0.8915,
+      "step": 249
+    },
+    {
+      "epoch": 0.060782883539995135,
+      "grad_norm": 28.375,
+      "learning_rate": 2.497898705470256e-06,
+      "loss": 1.4206,
+      "step": 250
+    },
+    {
+      "epoch": 0.061026015074155116,
+      "grad_norm": 23.125,
+      "learning_rate": 2.4978706028028595e-06,
+      "loss": 1.0814,
+      "step": 251
+    },
+    {
+      "epoch": 0.061269146608315096,
+      "grad_norm": 22.125,
+      "learning_rate": 2.49784231361965e-06,
+      "loss": 0.898,
+      "step": 252
+    },
+    {
+      "epoch": 0.06151227814247508,
+      "grad_norm": 17.5,
+      "learning_rate": 2.4978138379248555e-06,
+      "loss": 1.0512,
+      "step": 253
+    },
+    {
+      "epoch": 0.06175540967663506,
+      "grad_norm": 26.625,
+      "learning_rate": 2.4977851757227327e-06,
+      "loss": 1.4677,
+      "step": 254
+    },
+    {
+      "epoch": 0.06199854121079504,
+      "grad_norm": 16.125,
+      "learning_rate": 2.497756327017566e-06,
+      "loss": 0.4791,
+      "step": 255
+    },
+    {
+      "epoch": 0.06224167274495502,
+      "grad_norm": 24.625,
+      "learning_rate": 2.4977272918136656e-06,
+      "loss": 1.4746,
+      "step": 256
+    },
+    {
+      "epoch": 0.062484804279115,
+      "grad_norm": 20.5,
+      "learning_rate": 2.497698070115373e-06,
+      "loss": 0.7946,
+      "step": 257
+    },
+    {
+      "epoch": 0.06272793581327499,
+      "grad_norm": 23.875,
+      "learning_rate": 2.4976686619270555e-06,
+      "loss": 1.0674,
+      "step": 258
+    },
+    {
+      "epoch": 0.06297106734743496,
+      "grad_norm": 20.5,
+      "learning_rate": 2.4976390672531082e-06,
+      "loss": 1.1499,
+      "step": 259
+    },
+    {
+      "epoch": 0.06321419888159495,
+      "grad_norm": 26.5,
+      "learning_rate": 2.497609286097955e-06,
+      "loss": 1.1468,
+      "step": 260
+    },
+    {
+      "epoch": 0.06345733041575492,
+      "grad_norm": 17.75,
+      "learning_rate": 2.497579318466047e-06,
+      "loss": 0.8431,
+      "step": 261
+    },
+    {
+      "epoch": 0.06370046194991491,
+      "grad_norm": 15.125,
+      "learning_rate": 2.4975491643618633e-06,
+      "loss": 0.6385,
+      "step": 262
+    },
+    {
+      "epoch": 0.06394359348407488,
+      "grad_norm": 20.0,
+      "learning_rate": 2.497518823789911e-06,
+      "loss": 1.1233,
+      "step": 263
+    },
+    {
+      "epoch": 0.06418672501823487,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4974882967547255e-06,
+      "loss": 1.2095,
+      "step": 264
+    },
+    {
+      "epoch": 0.06442985655239485,
+      "grad_norm": 14.125,
+      "learning_rate": 2.497457583260869e-06,
+      "loss": 0.4845,
+      "step": 265
+    },
+    {
+      "epoch": 0.06467298808655482,
+      "grad_norm": 16.125,
+      "learning_rate": 2.497426683312932e-06,
+      "loss": 0.7712,
+      "step": 266
+    },
+    {
+      "epoch": 0.06491611962071481,
+      "grad_norm": 21.5,
+      "learning_rate": 2.497395596915534e-06,
+      "loss": 1.0351,
+      "step": 267
+    },
+    {
+      "epoch": 0.06515925115487478,
+      "grad_norm": 27.625,
+      "learning_rate": 2.49736432407332e-06,
+      "loss": 0.9552,
+      "step": 268
+    },
+    {
+      "epoch": 0.06540238268903477,
+      "grad_norm": 32.5,
+      "learning_rate": 2.4973328647909657e-06,
+      "loss": 1.0366,
+      "step": 269
+    },
+    {
+      "epoch": 0.06564551422319474,
+      "grad_norm": 34.25,
+      "learning_rate": 2.4973012190731723e-06,
+      "loss": 1.2298,
+      "step": 270
+    },
+    {
+      "epoch": 0.06588864575735473,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.49726938692467e-06,
+      "loss": 0.7068,
+      "step": 271
+    },
+    {
+      "epoch": 0.0661317772915147,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.497237368350217e-06,
+      "loss": 0.4992,
+      "step": 272
+    },
+    {
+      "epoch": 0.06637490882567469,
+      "grad_norm": 18.375,
+      "learning_rate": 2.4972051633545987e-06,
+      "loss": 0.6491,
+      "step": 273
+    },
+    {
+      "epoch": 0.06661804035983467,
+      "grad_norm": 18.375,
+      "learning_rate": 2.4971727719426282e-06,
+      "loss": 1.1035,
+      "step": 274
+    },
+    {
+      "epoch": 0.06686117189399465,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.497140194119148e-06,
+      "loss": 0.481,
+      "step": 275
+    },
+    {
+      "epoch": 0.06710430342815463,
+      "grad_norm": 19.125,
+      "learning_rate": 2.497107429889027e-06,
+      "loss": 1.1583,
+      "step": 276
+    },
+    {
+      "epoch": 0.06734743496231461,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.4970744792571622e-06,
+      "loss": 0.6236,
+      "step": 277
+    },
+    {
+      "epoch": 0.06759056649647459,
+      "grad_norm": 25.375,
+      "learning_rate": 2.497041342228479e-06,
+      "loss": 1.2272,
+      "step": 278
+    },
+    {
+      "epoch": 0.06783369803063458,
+      "grad_norm": 17.75,
+      "learning_rate": 2.4970080188079297e-06,
+      "loss": 1.0533,
+      "step": 279
+    },
+    {
+      "epoch": 0.06807682956479455,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4969745090004952e-06,
+      "loss": 0.9864,
+      "step": 280
+    },
+    {
+      "epoch": 0.06831996109895454,
+      "grad_norm": 37.5,
+      "learning_rate": 2.496940812811185e-06,
+      "loss": 1.1633,
+      "step": 281
+    },
+    {
+      "epoch": 0.06856309263311451,
+      "grad_norm": 27.875,
+      "learning_rate": 2.4969069302450345e-06,
+      "loss": 1.5523,
+      "step": 282
+    },
+    {
+      "epoch": 0.0688062241672745,
+      "grad_norm": 23.375,
+      "learning_rate": 2.4968728613071086e-06,
+      "loss": 1.1659,
+      "step": 283
+    },
+    {
+      "epoch": 0.06904935570143447,
+      "grad_norm": 17.875,
+      "learning_rate": 2.496838606002499e-06,
+      "loss": 0.8856,
+      "step": 284
+    },
+    {
+      "epoch": 0.06929248723559446,
+      "grad_norm": 16.375,
+      "learning_rate": 2.4968041643363265e-06,
+      "loss": 0.5519,
+      "step": 285
+    },
+    {
+      "epoch": 0.06953561876975443,
+      "grad_norm": 16.75,
+      "learning_rate": 2.4967695363137385e-06,
+      "loss": 0.8209,
+      "step": 286
+    },
+    {
+      "epoch": 0.06977875030391442,
+      "grad_norm": 18.625,
+      "learning_rate": 2.4967347219399108e-06,
+      "loss": 1.0672,
+      "step": 287
+    },
+    {
+      "epoch": 0.0700218818380744,
+      "grad_norm": 18.125,
+      "learning_rate": 2.496699721220047e-06,
+      "loss": 0.9386,
+      "step": 288
+    },
+    {
+      "epoch": 0.07026501337223438,
+      "grad_norm": 18.625,
+      "learning_rate": 2.4966645341593786e-06,
+      "loss": 1.0656,
+      "step": 289
+    },
+    {
+      "epoch": 0.07050814490639436,
+      "grad_norm": 17.625,
+      "learning_rate": 2.496629160763165e-06,
+      "loss": 0.8377,
+      "step": 290
+    },
+    {
+      "epoch": 0.07075127644055434,
+      "grad_norm": 16.625,
+      "learning_rate": 2.4965936010366934e-06,
+      "loss": 0.7864,
+      "step": 291
+    },
+    {
+      "epoch": 0.07099440797471432,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4965578549852786e-06,
+      "loss": 0.8277,
+      "step": 292
+    },
+    {
+      "epoch": 0.0712375395088743,
+      "grad_norm": 20.875,
+      "learning_rate": 2.4965219226142635e-06,
+      "loss": 1.3112,
+      "step": 293
+    },
+    {
+      "epoch": 0.07148067104303428,
+      "grad_norm": 21.875,
+      "learning_rate": 2.496485803929019e-06,
+      "loss": 1.0034,
+      "step": 294
+    },
+    {
+      "epoch": 0.07172380257719427,
+      "grad_norm": 20.75,
+      "learning_rate": 2.4964494989349437e-06,
+      "loss": 0.9285,
+      "step": 295
+    },
+    {
+      "epoch": 0.07196693411135424,
+      "grad_norm": 21.0,
+      "learning_rate": 2.4964130076374632e-06,
+      "loss": 0.8082,
+      "step": 296
+    },
+    {
+      "epoch": 0.07221006564551423,
+      "grad_norm": 20.125,
+      "learning_rate": 2.496376330042033e-06,
+      "loss": 1.1866,
+      "step": 297
+    },
+    {
+      "epoch": 0.0724531971796742,
+      "grad_norm": 22.875,
+      "learning_rate": 2.4963394661541345e-06,
+      "loss": 1.0004,
+      "step": 298
+    },
+    {
+      "epoch": 0.07269632871383419,
+      "grad_norm": 17.0,
+      "learning_rate": 2.4963024159792778e-06,
+      "loss": 1.0495,
+      "step": 299
+    },
+    {
+      "epoch": 0.07293946024799416,
+      "grad_norm": 15.875,
+      "learning_rate": 2.496265179523e-06,
+      "loss": 0.7691,
+      "step": 300
+    },
+    {
+      "epoch": 0.07318259178215415,
+      "grad_norm": 25.375,
+      "learning_rate": 2.496227756790868e-06,
+      "loss": 0.7636,
+      "step": 301
+    },
+    {
+      "epoch": 0.07342572331631413,
+      "grad_norm": 18.0,
+      "learning_rate": 2.496190147788475e-06,
+      "loss": 0.7876,
+      "step": 302
+    },
+    {
+      "epoch": 0.0736688548504741,
+      "grad_norm": 14.625,
+      "learning_rate": 2.4961523525214414e-06,
+      "loss": 0.8633,
+      "step": 303
+    },
+    {
+      "epoch": 0.07391198638463409,
+      "grad_norm": 20.375,
+      "learning_rate": 2.4961143709954174e-06,
+      "loss": 1.1787,
+      "step": 304
+    },
+    {
+      "epoch": 0.07415511791879406,
+      "grad_norm": 20.375,
+      "learning_rate": 2.496076203216079e-06,
+      "loss": 0.9555,
+      "step": 305
+    },
+    {
+      "epoch": 0.07439824945295405,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4960378491891317e-06,
+      "loss": 0.9444,
+      "step": 306
+    },
+    {
+      "epoch": 0.07464138098711402,
+      "grad_norm": 16.875,
+      "learning_rate": 2.4959993089203084e-06,
+      "loss": 0.6795,
+      "step": 307
+    },
+    {
+      "epoch": 0.07488451252127401,
+      "grad_norm": 26.625,
+      "learning_rate": 2.4959605824153687e-06,
+      "loss": 0.9794,
+      "step": 308
+    },
+    {
+      "epoch": 0.07512764405543398,
+      "grad_norm": 24.375,
+      "learning_rate": 2.4959216696801012e-06,
+      "loss": 0.8844,
+      "step": 309
+    },
+    {
+      "epoch": 0.07537077558959397,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.4958825707203234e-06,
+      "loss": 0.6024,
+      "step": 310
+    },
+    {
+      "epoch": 0.07561390712375395,
+      "grad_norm": 18.0,
+      "learning_rate": 2.4958432855418776e-06,
+      "loss": 0.7961,
+      "step": 311
+    },
+    {
+      "epoch": 0.07585703865791393,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4958038141506363e-06,
+      "loss": 1.1005,
+      "step": 312
+    },
+    {
+      "epoch": 0.07610017019207391,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.495764156552499e-06,
+      "loss": 0.8304,
+      "step": 313
+    },
+    {
+      "epoch": 0.0763433017262339,
+      "grad_norm": 25.5,
+      "learning_rate": 2.495724312753394e-06,
+      "loss": 1.1424,
+      "step": 314
+    },
+    {
+      "epoch": 0.07658643326039387,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4956842827592757e-06,
+      "loss": 1.0434,
+      "step": 315
+    },
+    {
+      "epoch": 0.07682956479455386,
+      "grad_norm": 23.0,
+      "learning_rate": 2.495644066576128e-06,
+      "loss": 1.0894,
+      "step": 316
+    },
+    {
+      "epoch": 0.07707269632871383,
+      "grad_norm": 20.875,
+      "learning_rate": 2.4956036642099613e-06,
+      "loss": 1.2238,
+      "step": 317
+    },
+    {
+      "epoch": 0.07731582786287382,
+      "grad_norm": 29.125,
+      "learning_rate": 2.4955630756668143e-06,
+      "loss": 1.6621,
+      "step": 318
+    },
+    {
+      "epoch": 0.07755895939703379,
+      "grad_norm": 26.0,
+      "learning_rate": 2.495522300952754e-06,
+      "loss": 1.5402,
+      "step": 319
+    },
+    {
+      "epoch": 0.07780209093119378,
+      "grad_norm": 16.875,
+      "learning_rate": 2.4954813400738754e-06,
+      "loss": 0.5566,
+      "step": 320
+    },
+    {
+      "epoch": 0.07804522246535375,
+      "grad_norm": 25.625,
+      "learning_rate": 2.4954401930363003e-06,
+      "loss": 0.8753,
+      "step": 321
+    },
+    {
+      "epoch": 0.07828835399951374,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4953988598461788e-06,
+      "loss": 1.0414,
+      "step": 322
+    },
+    {
+      "epoch": 0.07853148553367371,
+      "grad_norm": 20.375,
+      "learning_rate": 2.4953573405096886e-06,
+      "loss": 0.88,
+      "step": 323
+    },
+    {
+      "epoch": 0.0787746170678337,
+      "grad_norm": 15.625,
+      "learning_rate": 2.495315635033036e-06,
+      "loss": 0.79,
+      "step": 324
+    },
+    {
+      "epoch": 0.07901774860199368,
+      "grad_norm": 20.75,
+      "learning_rate": 2.4952737434224545e-06,
+      "loss": 0.9045,
+      "step": 325
+    },
+    {
+      "epoch": 0.07926088013615366,
+      "grad_norm": 22.25,
+      "learning_rate": 2.4952316656842055e-06,
+      "loss": 1.2891,
+      "step": 326
+    },
+    {
+      "epoch": 0.07950401167031364,
+      "grad_norm": 35.25,
+      "learning_rate": 2.495189401824578e-06,
+      "loss": 1.3177,
+      "step": 327
+    },
+    {
+      "epoch": 0.07974714320447363,
+      "grad_norm": 52.5,
+      "learning_rate": 2.495146951849889e-06,
+      "loss": 1.2862,
+      "step": 328
+    },
+    {
+      "epoch": 0.0799902747386336,
+      "grad_norm": 16.625,
+      "learning_rate": 2.495104315766484e-06,
+      "loss": 0.6278,
+      "step": 329
+    },
+    {
+      "epoch": 0.08023340627279359,
+      "grad_norm": 18.0,
+      "learning_rate": 2.495061493580735e-06,
+      "loss": 0.8916,
+      "step": 330
+    },
+    {
+      "epoch": 0.08047653780695356,
+      "grad_norm": 14.625,
+      "learning_rate": 2.495018485299043e-06,
+      "loss": 0.7179,
+      "step": 331
+    },
+    {
+      "epoch": 0.08071966934111355,
+      "grad_norm": 17.0,
+      "learning_rate": 2.4949752909278363e-06,
+      "loss": 1.0449,
+      "step": 332
+    },
+    {
+      "epoch": 0.08096280087527352,
+      "grad_norm": 16.875,
+      "learning_rate": 2.494931910473571e-06,
+      "loss": 0.7221,
+      "step": 333
+    },
+    {
+      "epoch": 0.08120593240943351,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.4948883439427305e-06,
+      "loss": 0.7525,
+      "step": 334
+    },
+    {
+      "epoch": 0.08144906394359348,
+      "grad_norm": 18.375,
+      "learning_rate": 2.4948445913418272e-06,
+      "loss": 0.8925,
+      "step": 335
+    },
+    {
+      "epoch": 0.08169219547775347,
+      "grad_norm": 26.5,
+      "learning_rate": 2.4948006526774003e-06,
+      "loss": 1.4271,
+      "step": 336
+    },
+    {
+      "epoch": 0.08193532701191344,
+      "grad_norm": 24.0,
+      "learning_rate": 2.4947565279560183e-06,
+      "loss": 1.5347,
+      "step": 337
+    },
+    {
+      "epoch": 0.08217845854607343,
+      "grad_norm": 21.75,
+      "learning_rate": 2.4947122171842747e-06,
+      "loss": 0.8424,
+      "step": 338
+    },
+    {
+      "epoch": 0.0824215900802334,
+      "grad_norm": 22.125,
+      "learning_rate": 2.4946677203687933e-06,
+      "loss": 1.273,
+      "step": 339
+    },
+    {
+      "epoch": 0.08266472161439338,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.494623037516225e-06,
+      "loss": 0.5317,
+      "step": 340
+    },
+    {
+      "epoch": 0.08290785314855337,
+      "grad_norm": 16.875,
+      "learning_rate": 2.494578168633249e-06,
+      "loss": 0.8341,
+      "step": 341
+    },
+    {
+      "epoch": 0.08315098468271334,
+      "grad_norm": 24.5,
+      "learning_rate": 2.49453311372657e-06,
+      "loss": 0.9029,
+      "step": 342
+    },
+    {
+      "epoch": 0.08339411621687333,
+      "grad_norm": 16.125,
+      "learning_rate": 2.494487872802924e-06,
+      "loss": 0.815,
+      "step": 343
+    },
+    {
+      "epoch": 0.0836372477510333,
+      "grad_norm": 19.375,
+      "learning_rate": 2.4944424458690727e-06,
+      "loss": 0.8027,
+      "step": 344
+    },
+    {
+      "epoch": 0.08388037928519329,
+      "grad_norm": 21.25,
+      "learning_rate": 2.4943968329318046e-06,
+      "loss": 1.3753,
+      "step": 345
+    },
+    {
+      "epoch": 0.08412351081935326,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4943510339979394e-06,
+      "loss": 0.7825,
+      "step": 346
+    },
+    {
+      "epoch": 0.08436664235351325,
+      "grad_norm": 19.125,
+      "learning_rate": 2.4943050490743208e-06,
+      "loss": 1.1817,
+      "step": 347
+    },
+    {
+      "epoch": 0.08460977388767323,
+      "grad_norm": 21.0,
+      "learning_rate": 2.4942588781678227e-06,
+      "loss": 0.9534,
+      "step": 348
+    },
+    {
+      "epoch": 0.08485290542183321,
+      "grad_norm": 20.375,
+      "learning_rate": 2.4942125212853465e-06,
+      "loss": 1.1751,
+      "step": 349
+    },
+    {
+      "epoch": 0.08509603695599319,
+      "grad_norm": 27.25,
+      "learning_rate": 2.494165978433821e-06,
+      "loss": 1.0186,
+      "step": 350
+    },
+    {
+      "epoch": 0.08533916849015317,
+      "grad_norm": 23.375,
+      "learning_rate": 2.4941192496202016e-06,
+      "loss": 1.0297,
+      "step": 351
+    },
+    {
+      "epoch": 0.08558230002431315,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.4940723348514746e-06,
+      "loss": 0.7368,
+      "step": 352
+    },
+    {
+      "epoch": 0.08582543155847314,
+      "grad_norm": 20.75,
+      "learning_rate": 2.4940252341346503e-06,
+      "loss": 1.2031,
+      "step": 353
+    },
+    {
+      "epoch": 0.08606856309263311,
+      "grad_norm": 31.125,
+      "learning_rate": 2.4939779474767706e-06,
+      "loss": 1.6584,
+      "step": 354
+    },
+    {
+      "epoch": 0.0863116946267931,
+      "grad_norm": 33.5,
+      "learning_rate": 2.493930474884902e-06,
+      "loss": 1.4518,
+      "step": 355
+    },
+    {
+      "epoch": 0.08655482616095307,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4938828163661405e-06,
+      "loss": 1.035,
+      "step": 356
+    },
+    {
+      "epoch": 0.08679795769511306,
+      "grad_norm": 56.75,
+      "learning_rate": 2.4938349719276096e-06,
+      "loss": 1.2927,
+      "step": 357
+    },
+    {
+      "epoch": 0.08704108922927303,
+      "grad_norm": 20.625,
+      "learning_rate": 2.4937869415764602e-06,
+      "loss": 0.8661,
+      "step": 358
+    },
+    {
+      "epoch": 0.08728422076343302,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4937387253198715e-06,
+      "loss": 0.5965,
+      "step": 359
+    },
+    {
+      "epoch": 0.087527352297593,
+      "grad_norm": 22.875,
+      "learning_rate": 2.4936903231650504e-06,
+      "loss": 1.2066,
+      "step": 360
+    },
+    {
+      "epoch": 0.08777048383175298,
+      "grad_norm": 19.125,
+      "learning_rate": 2.493641735119231e-06,
+      "loss": 0.9761,
+      "step": 361
+    },
+    {
+      "epoch": 0.08801361536591296,
+      "grad_norm": 16.375,
+      "learning_rate": 2.493592961189676e-06,
+      "loss": 0.4858,
+      "step": 362
+    },
+    {
+      "epoch": 0.08825674690007294,
+      "grad_norm": 27.125,
+      "learning_rate": 2.493544001383675e-06,
+      "loss": 0.8465,
+      "step": 363
+    },
+    {
+      "epoch": 0.08849987843423292,
+      "grad_norm": 21.75,
+      "learning_rate": 2.493494855708547e-06,
+      "loss": 0.8336,
+      "step": 364
+    },
+    {
+      "epoch": 0.0887430099683929,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4934455241716365e-06,
+      "loss": 1.1203,
+      "step": 365
+    },
+    {
+      "epoch": 0.08898614150255288,
+      "grad_norm": 19.75,
+      "learning_rate": 2.493396006780317e-06,
+      "loss": 1.0841,
+      "step": 366
+    },
+    {
+      "epoch": 0.08922927303671287,
+      "grad_norm": 35.5,
+      "learning_rate": 2.493346303541991e-06,
+      "loss": 1.5637,
+      "step": 367
+    },
+    {
+      "epoch": 0.08947240457087284,
+      "grad_norm": 37.25,
+      "learning_rate": 2.4932964144640858e-06,
+      "loss": 1.0716,
+      "step": 368
+    },
+    {
+      "epoch": 0.08971553610503283,
+      "grad_norm": 20.0,
+      "learning_rate": 2.493246339554059e-06,
+      "loss": 1.3607,
+      "step": 369
+    },
+    {
+      "epoch": 0.0899586676391928,
+      "grad_norm": 15.625,
+      "learning_rate": 2.4931960788193956e-06,
+      "loss": 0.6747,
+      "step": 370
+    },
+    {
+      "epoch": 0.09020179917335279,
+      "grad_norm": 16.875,
+      "learning_rate": 2.493145632267607e-06,
+      "loss": 0.7467,
+      "step": 371
+    },
+    {
+      "epoch": 0.09044493070751276,
+      "grad_norm": 21.75,
+      "learning_rate": 2.493094999906234e-06,
+      "loss": 1.1146,
+      "step": 372
+    },
+    {
+      "epoch": 0.09068806224167275,
+      "grad_norm": 20.5,
+      "learning_rate": 2.4930441817428443e-06,
+      "loss": 1.0361,
+      "step": 373
+    },
+    {
+      "epoch": 0.09093119377583272,
+      "grad_norm": 18.25,
+      "learning_rate": 2.4929931777850338e-06,
+      "loss": 0.9021,
+      "step": 374
+    },
+    {
+      "epoch": 0.09117432530999271,
+      "grad_norm": 20.625,
+      "learning_rate": 2.4929419880404252e-06,
+      "loss": 0.8993,
+      "step": 375
+    },
+    {
+      "epoch": 0.09141745684415269,
+      "grad_norm": 23.375,
+      "learning_rate": 2.4928906125166703e-06,
+      "loss": 1.1707,
+      "step": 376
+    },
+    {
+      "epoch": 0.09166058837831267,
+      "grad_norm": 17.0,
+      "learning_rate": 2.492839051221448e-06,
+      "loss": 0.9641,
+      "step": 377
+    },
+    {
+      "epoch": 0.09190371991247265,
+      "grad_norm": 21.25,
+      "learning_rate": 2.4927873041624645e-06,
+      "loss": 0.966,
+      "step": 378
+    },
+    {
+      "epoch": 0.09214685144663262,
+      "grad_norm": 25.25,
+      "learning_rate": 2.492735371347455e-06,
+      "loss": 1.3207,
+      "step": 379
+    },
+    {
+      "epoch": 0.09238998298079261,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4926832527841815e-06,
+      "loss": 1.0189,
+      "step": 380
+    },
+    {
+      "epoch": 0.09263311451495258,
+      "grad_norm": 14.5,
+      "learning_rate": 2.492630948480434e-06,
+      "loss": 0.5323,
+      "step": 381
+    },
+    {
+      "epoch": 0.09287624604911257,
+      "grad_norm": 29.5,
+      "learning_rate": 2.4925784584440304e-06,
+      "loss": 1.523,
+      "step": 382
+    },
+    {
+      "epoch": 0.09311937758327254,
+      "grad_norm": 25.625,
+      "learning_rate": 2.4925257826828157e-06,
+      "loss": 1.3599,
+      "step": 383
+    },
+    {
+      "epoch": 0.09336250911743253,
+      "grad_norm": 16.625,
+      "learning_rate": 2.492472921204664e-06,
+      "loss": 0.6617,
+      "step": 384
+    },
+    {
+      "epoch": 0.0936056406515925,
+      "grad_norm": 17.125,
+      "learning_rate": 2.492419874017476e-06,
+      "loss": 0.6534,
+      "step": 385
+    },
+    {
+      "epoch": 0.0938487721857525,
+      "grad_norm": 19.125,
+      "learning_rate": 2.4923666411291802e-06,
+      "loss": 1.0552,
+      "step": 386
+    },
+    {
+      "epoch": 0.09409190371991247,
+      "grad_norm": 17.0,
+      "learning_rate": 2.4923132225477336e-06,
+      "loss": 1.0621,
+      "step": 387
+    },
+    {
+      "epoch": 0.09433503525407246,
+      "grad_norm": 20.25,
+      "learning_rate": 2.4922596182811206e-06,
+      "loss": 1.003,
+      "step": 388
+    },
+    {
+      "epoch": 0.09457816678823243,
+      "grad_norm": 22.75,
+      "learning_rate": 2.492205828337353e-06,
+      "loss": 1.0473,
+      "step": 389
+    },
+    {
+      "epoch": 0.09482129832239242,
+      "grad_norm": 33.5,
+      "learning_rate": 2.4921518527244705e-06,
+      "loss": 1.5681,
+      "step": 390
+    },
+    {
+      "epoch": 0.09506442985655239,
+      "grad_norm": 16.25,
+      "learning_rate": 2.492097691450541e-06,
+      "loss": 0.9737,
+      "step": 391
+    },
+    {
+      "epoch": 0.09530756139071238,
+      "grad_norm": 19.375,
+      "learning_rate": 2.4920433445236596e-06,
+      "loss": 0.7637,
+      "step": 392
+    },
+    {
+      "epoch": 0.09555069292487235,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4919888119519496e-06,
+      "loss": 1.172,
+      "step": 393
+    },
+    {
+      "epoch": 0.09579382445903234,
+      "grad_norm": 16.0,
+      "learning_rate": 2.491934093743562e-06,
+      "loss": 0.7444,
+      "step": 394
+    },
+    {
+      "epoch": 0.09603695599319231,
+      "grad_norm": 24.25,
+      "learning_rate": 2.491879189906675e-06,
+      "loss": 0.971,
+      "step": 395
+    },
+    {
+      "epoch": 0.0962800875273523,
+      "grad_norm": 16.375,
+      "learning_rate": 2.491824100449495e-06,
+      "loss": 0.8551,
+      "step": 396
+    },
+    {
+      "epoch": 0.09652321906151227,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4917688253802563e-06,
+      "loss": 0.8836,
+      "step": 397
+    },
+    {
+      "epoch": 0.09676635059567226,
+      "grad_norm": 26.625,
+      "learning_rate": 2.4917133647072204e-06,
+      "loss": 1.2094,
+      "step": 398
+    },
+    {
+      "epoch": 0.09700948212983224,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.4916577184386775e-06,
+      "loss": 0.5406,
+      "step": 399
+    },
+    {
+      "epoch": 0.09725261366399222,
+      "grad_norm": 21.375,
+      "learning_rate": 2.491601886582944e-06,
+      "loss": 1.1714,
+      "step": 400
+    },
+    {
+      "epoch": 0.0974957451981522,
+      "grad_norm": 18.25,
+      "learning_rate": 2.491545869148365e-06,
+      "loss": 0.9329,
+      "step": 401
+    },
+    {
+      "epoch": 0.09773887673231219,
+      "grad_norm": 20.5,
+      "learning_rate": 2.491489666143314e-06,
+      "loss": 0.8319,
+      "step": 402
+    },
+    {
+      "epoch": 0.09798200826647216,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.491433277576191e-06,
+      "loss": 0.6509,
+      "step": 403
+    },
+    {
+      "epoch": 0.09822513980063215,
+      "grad_norm": 23.875,
+      "learning_rate": 2.491376703455425e-06,
+      "loss": 1.1198,
+      "step": 404
+    },
+    {
+      "epoch": 0.09846827133479212,
+      "grad_norm": 20.0,
+      "learning_rate": 2.4913199437894705e-06,
+      "loss": 0.7157,
+      "step": 405
+    },
+    {
+      "epoch": 0.09871140286895211,
+      "grad_norm": 16.5,
+      "learning_rate": 2.4912629985868126e-06,
+      "loss": 0.7598,
+      "step": 406
+    },
+    {
+      "epoch": 0.09895453440311208,
+      "grad_norm": 24.125,
+      "learning_rate": 2.491205867855962e-06,
+      "loss": 1.2696,
+      "step": 407
+    },
+    {
+      "epoch": 0.09919766593727207,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4911485516054577e-06,
+      "loss": 0.8278,
+      "step": 408
+    },
+    {
+      "epoch": 0.09944079747143204,
+      "grad_norm": 15.25,
+      "learning_rate": 2.4910910498438674e-06,
+      "loss": 0.4924,
+      "step": 409
+    },
+    {
+      "epoch": 0.09968392900559203,
+      "grad_norm": 20.875,
+      "learning_rate": 2.4910333625797856e-06,
+      "loss": 1.1013,
+      "step": 410
+    },
+    {
+      "epoch": 0.099927060539752,
+      "grad_norm": 17.25,
+      "learning_rate": 2.490975489821834e-06,
+      "loss": 0.7482,
+      "step": 411
+    },
+    {
+      "epoch": 0.10017019207391199,
+      "grad_norm": 17.5,
+      "learning_rate": 2.490917431578663e-06,
+      "loss": 0.6875,
+      "step": 412
+    },
+    {
+      "epoch": 0.10041332360807197,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4908591878589507e-06,
+      "loss": 1.0509,
+      "step": 413
+    },
+    {
+      "epoch": 0.10065645514223195,
+      "grad_norm": 23.25,
+      "learning_rate": 2.4908007586714023e-06,
+      "loss": 1.0468,
+      "step": 414
+    },
+    {
+      "epoch": 0.10089958667639193,
+      "grad_norm": 22.125,
+      "learning_rate": 2.490742144024751e-06,
+      "loss": 0.7809,
+      "step": 415
+    },
+    {
+      "epoch": 0.1011427182105519,
+      "grad_norm": 28.375,
+      "learning_rate": 2.4906833439277577e-06,
+      "loss": 0.9365,
+      "step": 416
+    },
+    {
+      "epoch": 0.10138584974471189,
+      "grad_norm": 15.75,
+      "learning_rate": 2.490624358389212e-06,
+      "loss": 0.6247,
+      "step": 417
+    },
+    {
+      "epoch": 0.10162898127887186,
+      "grad_norm": 16.0,
+      "learning_rate": 2.4905651874179294e-06,
+      "loss": 0.7028,
+      "step": 418
+    },
+    {
+      "epoch": 0.10187211281303185,
+      "grad_norm": 17.375,
+      "learning_rate": 2.4905058310227536e-06,
+      "loss": 0.9292,
+      "step": 419
+    },
+    {
+      "epoch": 0.10211524434719182,
+      "grad_norm": 18.0,
+      "learning_rate": 2.490446289212558e-06,
+      "loss": 0.9055,
+      "step": 420
+    },
+    {
+      "epoch": 0.10235837588135181,
+      "grad_norm": 18.25,
+      "learning_rate": 2.4903865619962405e-06,
+      "loss": 1.0531,
+      "step": 421
+    },
+    {
+      "epoch": 0.10260150741551179,
+      "grad_norm": 21.25,
+      "learning_rate": 2.4903266493827294e-06,
+      "loss": 1.2379,
+      "step": 422
+    },
+    {
+      "epoch": 0.10284463894967177,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4902665513809793e-06,
+      "loss": 1.206,
+      "step": 423
+    },
+    {
+      "epoch": 0.10308777048383175,
+      "grad_norm": 18.875,
+      "learning_rate": 2.490206267999973e-06,
+      "loss": 0.9579,
+      "step": 424
+    },
+    {
+      "epoch": 0.10333090201799174,
+      "grad_norm": 24.125,
+      "learning_rate": 2.490145799248721e-06,
+      "loss": 1.1271,
+      "step": 425
+    },
+    {
+      "epoch": 0.10357403355215171,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4900851451362612e-06,
+      "loss": 1.0493,
+      "step": 426
+    },
+    {
+      "epoch": 0.1038171650863117,
+      "grad_norm": 20.5,
+      "learning_rate": 2.4900243056716593e-06,
+      "loss": 1.2056,
+      "step": 427
+    },
+    {
+      "epoch": 0.10406029662047167,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.489963280864009e-06,
+      "loss": 0.7268,
+      "step": 428
+    },
+    {
+      "epoch": 0.10430342815463166,
+      "grad_norm": 20.125,
+      "learning_rate": 2.489902070722431e-06,
+      "loss": 1.0395,
+      "step": 429
+    },
+    {
+      "epoch": 0.10454655968879163,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.4898406752560756e-06,
+      "loss": 0.7524,
+      "step": 430
+    },
+    {
+      "epoch": 0.10478969122295162,
+      "grad_norm": 21.5,
+      "learning_rate": 2.489779094474118e-06,
+      "loss": 1.1313,
+      "step": 431
+    },
+    {
+      "epoch": 0.1050328227571116,
+      "grad_norm": 21.0,
+      "learning_rate": 2.4897173283857628e-06,
+      "loss": 1.0249,
+      "step": 432
+    },
+    {
+      "epoch": 0.10527595429127158,
+      "grad_norm": 19.5,
+      "learning_rate": 2.4896553770002425e-06,
+      "loss": 1.0617,
+      "step": 433
+    },
+    {
+      "epoch": 0.10551908582543156,
+      "grad_norm": 23.0,
+      "learning_rate": 2.4895932403268165e-06,
+      "loss": 1.1214,
+      "step": 434
+    },
+    {
+      "epoch": 0.10576221735959154,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4895309183747725e-06,
+      "loss": 0.982,
+      "step": 435
+    },
+    {
+      "epoch": 0.10600534889375152,
+      "grad_norm": 19.375,
+      "learning_rate": 2.4894684111534247e-06,
+      "loss": 1.0424,
+      "step": 436
+    },
+    {
+      "epoch": 0.1062484804279115,
+      "grad_norm": 28.5,
+      "learning_rate": 2.489405718672117e-06,
+      "loss": 1.1515,
+      "step": 437
+    },
+    {
+      "epoch": 0.10649161196207148,
+      "grad_norm": 23.125,
+      "learning_rate": 2.489342840940219e-06,
+      "loss": 1.0771,
+      "step": 438
+    },
+    {
+      "epoch": 0.10673474349623147,
+      "grad_norm": 23.75,
+      "learning_rate": 2.4892797779671298e-06,
+      "loss": 1.128,
+      "step": 439
+    },
+    {
+      "epoch": 0.10697787503039144,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4892165297622737e-06,
+      "loss": 0.7404,
+      "step": 440
+    },
+    {
+      "epoch": 0.10722100656455143,
+      "grad_norm": 16.5,
+      "learning_rate": 2.489153096335106e-06,
+      "loss": 0.825,
+      "step": 441
+    },
+    {
+      "epoch": 0.1074641380987114,
+      "grad_norm": 20.5,
+      "learning_rate": 2.489089477695107e-06,
+      "loss": 0.9519,
+      "step": 442
+    },
+    {
+      "epoch": 0.10770726963287139,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4890256738517853e-06,
+      "loss": 1.0615,
+      "step": 443
+    },
+    {
+      "epoch": 0.10795040116703136,
+      "grad_norm": 22.25,
+      "learning_rate": 2.4889616848146786e-06,
+      "loss": 1.0737,
+      "step": 444
+    },
+    {
+      "epoch": 0.10819353270119135,
+      "grad_norm": 19.875,
+      "learning_rate": 2.48889751059335e-06,
+      "loss": 0.9922,
+      "step": 445
+    },
+    {
+      "epoch": 0.10843666423535132,
+      "grad_norm": 24.75,
+      "learning_rate": 2.4888331511973924e-06,
+      "loss": 1.0417,
+      "step": 446
+    },
+    {
+      "epoch": 0.10867979576951131,
+      "grad_norm": 25.375,
+      "learning_rate": 2.4887686066364246e-06,
+      "loss": 1.2682,
+      "step": 447
+    },
+    {
+      "epoch": 0.10892292730367129,
+      "grad_norm": 22.5,
+      "learning_rate": 2.4887038769200943e-06,
+      "loss": 0.9191,
+      "step": 448
+    },
+    {
+      "epoch": 0.10916605883783127,
+      "grad_norm": 24.25,
+      "learning_rate": 2.4886389620580763e-06,
+      "loss": 1.0753,
+      "step": 449
+    },
+    {
+      "epoch": 0.10940919037199125,
+      "grad_norm": 19.125,
+      "learning_rate": 2.4885738620600737e-06,
+      "loss": 0.8045,
+      "step": 450
+    },
+    {
+      "epoch": 0.10965232190615123,
+      "grad_norm": 17.375,
+      "learning_rate": 2.4885085769358166e-06,
+      "loss": 0.7605,
+      "step": 451
+    },
+    {
+      "epoch": 0.10989545344031121,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4884431066950626e-06,
+      "loss": 1.004,
+      "step": 452
+    },
+    {
+      "epoch": 0.11013858497447118,
+      "grad_norm": 20.0,
+      "learning_rate": 2.488377451347598e-06,
+      "loss": 0.8924,
+      "step": 453
+    },
+    {
+      "epoch": 0.11038171650863117,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4883116109032352e-06,
+      "loss": 0.7186,
+      "step": 454
+    },
+    {
+      "epoch": 0.11062484804279114,
+      "grad_norm": 16.5,
+      "learning_rate": 2.4882455853718164e-06,
+      "loss": 0.7337,
+      "step": 455
+    },
+    {
+      "epoch": 0.11086797957695113,
+      "grad_norm": 28.75,
+      "learning_rate": 2.488179374763209e-06,
+      "loss": 1.0111,
+      "step": 456
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 23.5,
+      "learning_rate": 2.4881129790873106e-06,
+      "loss": 1.238,
+      "step": 457
+    },
+    {
+      "epoch": 0.11135424264527109,
+      "grad_norm": 24.375,
+      "learning_rate": 2.488046398354044e-06,
+      "loss": 1.1166,
+      "step": 458
+    },
+    {
+      "epoch": 0.11159737417943107,
+      "grad_norm": 20.0,
+      "learning_rate": 2.4879796325733612e-06,
+      "loss": 0.9251,
+      "step": 459
+    },
+    {
+      "epoch": 0.11184050571359105,
+      "grad_norm": 18.625,
+      "learning_rate": 2.487912681755242e-06,
+      "loss": 0.9987,
+      "step": 460
+    },
+    {
+      "epoch": 0.11208363724775103,
+      "grad_norm": 16.5,
+      "learning_rate": 2.4878455459096936e-06,
+      "loss": 0.8793,
+      "step": 461
+    },
+    {
+      "epoch": 0.11232676878191102,
+      "grad_norm": 16.625,
+      "learning_rate": 2.4877782250467493e-06,
+      "loss": 0.7242,
+      "step": 462
+    },
+    {
+      "epoch": 0.11256990031607099,
+      "grad_norm": 23.375,
+      "learning_rate": 2.487710719176472e-06,
+      "loss": 1.1936,
+      "step": 463
+    },
+    {
+      "epoch": 0.11281303185023098,
+      "grad_norm": 19.125,
+      "learning_rate": 2.487643028308952e-06,
+      "loss": 0.8931,
+      "step": 464
+    },
+    {
+      "epoch": 0.11305616338439095,
+      "grad_norm": 23.75,
+      "learning_rate": 2.4875751524543067e-06,
+      "loss": 1.3118,
+      "step": 465
+    },
+    {
+      "epoch": 0.11329929491855094,
+      "grad_norm": 23.75,
+      "learning_rate": 2.487507091622681e-06,
+      "loss": 1.5804,
+      "step": 466
+    },
+    {
+      "epoch": 0.11354242645271091,
+      "grad_norm": 20.875,
+      "learning_rate": 2.487438845824248e-06,
+      "loss": 0.9065,
+      "step": 467
+    },
+    {
+      "epoch": 0.1137855579868709,
+      "grad_norm": 13.375,
+      "learning_rate": 2.487370415069208e-06,
+      "loss": 0.5431,
+      "step": 468
+    },
+    {
+      "epoch": 0.11402868952103087,
+      "grad_norm": 20.125,
+      "learning_rate": 2.48730179936779e-06,
+      "loss": 1.0669,
+      "step": 469
+    },
+    {
+      "epoch": 0.11427182105519086,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4872329987302484e-06,
+      "loss": 0.8652,
+      "step": 470
+    },
+    {
+      "epoch": 0.11451495258935084,
+      "grad_norm": 20.75,
+      "learning_rate": 2.487164013166868e-06,
+      "loss": 1.491,
+      "step": 471
+    },
+    {
+      "epoch": 0.11475808412351082,
+      "grad_norm": 25.625,
+      "learning_rate": 2.487094842687959e-06,
+      "loss": 1.1311,
+      "step": 472
+    },
+    {
+      "epoch": 0.1150012156576708,
+      "grad_norm": 22.375,
+      "learning_rate": 2.4870254873038602e-06,
+      "loss": 1.1097,
+      "step": 473
+    },
+    {
+      "epoch": 0.11524434719183078,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.4869559470249384e-06,
+      "loss": 0.6992,
+      "step": 474
+    },
+    {
+      "epoch": 0.11548747872599076,
+      "grad_norm": 19.875,
+      "learning_rate": 2.486886221861587e-06,
+      "loss": 0.8973,
+      "step": 475
+    },
+    {
+      "epoch": 0.11573061026015075,
+      "grad_norm": 20.75,
+      "learning_rate": 2.4868163118242283e-06,
+      "loss": 1.2272,
+      "step": 476
+    },
+    {
+      "epoch": 0.11597374179431072,
+      "grad_norm": 17.625,
+      "learning_rate": 2.486746216923311e-06,
+      "loss": 1.1184,
+      "step": 477
+    },
+    {
+      "epoch": 0.11621687332847071,
+      "grad_norm": 16.75,
+      "learning_rate": 2.4866759371693127e-06,
+      "loss": 0.5163,
+      "step": 478
+    },
+    {
+      "epoch": 0.11646000486263068,
+      "grad_norm": 23.875,
+      "learning_rate": 2.486605472572737e-06,
+      "loss": 1.2058,
+      "step": 479
+    },
+    {
+      "epoch": 0.11670313639679067,
+      "grad_norm": 16.375,
+      "learning_rate": 2.4865348231441168e-06,
+      "loss": 0.7734,
+      "step": 480
+    },
+    {
+      "epoch": 0.11694626793095064,
+      "grad_norm": 20.5,
+      "learning_rate": 2.486463988894011e-06,
+      "loss": 0.7897,
+      "step": 481
+    },
+    {
+      "epoch": 0.11718939946511063,
+      "grad_norm": 29.125,
+      "learning_rate": 2.4863929698330085e-06,
+      "loss": 1.7795,
+      "step": 482
+    },
+    {
+      "epoch": 0.1174325309992706,
+      "grad_norm": 23.5,
+      "learning_rate": 2.486321765971723e-06,
+      "loss": 1.096,
+      "step": 483
+    },
+    {
+      "epoch": 0.11767566253343059,
+      "grad_norm": 24.125,
+      "learning_rate": 2.4862503773207973e-06,
+      "loss": 1.3117,
+      "step": 484
+    },
+    {
+      "epoch": 0.11791879406759057,
+      "grad_norm": 21.25,
+      "learning_rate": 2.486178803890902e-06,
+      "loss": 1.0737,
+      "step": 485
+    },
+    {
+      "epoch": 0.11816192560175055,
+      "grad_norm": 20.625,
+      "learning_rate": 2.486107045692735e-06,
+      "loss": 1.2179,
+      "step": 486
+    },
+    {
+      "epoch": 0.11840505713591053,
+      "grad_norm": 21.5,
+      "learning_rate": 2.486035102737022e-06,
+      "loss": 0.756,
+      "step": 487
+    },
+    {
+      "epoch": 0.11864818867007051,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4859629750345154e-06,
+      "loss": 0.7913,
+      "step": 488
+    },
+    {
+      "epoch": 0.11889132020423049,
+      "grad_norm": 28.0,
+      "learning_rate": 2.4858906625959965e-06,
+      "loss": 1.277,
+      "step": 489
+    },
+    {
+      "epoch": 0.11913445173839048,
+      "grad_norm": 21.25,
+      "learning_rate": 2.4858181654322737e-06,
+      "loss": 1.3129,
+      "step": 490
+    },
+    {
+      "epoch": 0.11937758327255045,
+      "grad_norm": 19.75,
+      "learning_rate": 2.485745483554183e-06,
+      "loss": 0.7163,
+      "step": 491
+    },
+    {
+      "epoch": 0.11962071480671042,
+      "grad_norm": 18.5,
+      "learning_rate": 2.485672616972587e-06,
+      "loss": 0.9383,
+      "step": 492
+    },
+    {
+      "epoch": 0.11986384634087041,
+      "grad_norm": 14.625,
+      "learning_rate": 2.4855995656983782e-06,
+      "loss": 0.4747,
+      "step": 493
+    },
+    {
+      "epoch": 0.12010697787503039,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4855263297424742e-06,
+      "loss": 0.8701,
+      "step": 494
+    },
+    {
+      "epoch": 0.12035010940919037,
+      "grad_norm": 23.125,
+      "learning_rate": 2.4854529091158224e-06,
+      "loss": 0.8861,
+      "step": 495
+    },
+    {
+      "epoch": 0.12059324094335035,
+      "grad_norm": 23.5,
+      "learning_rate": 2.485379303829396e-06,
+      "loss": 1.0566,
+      "step": 496
+    },
+    {
+      "epoch": 0.12083637247751033,
+      "grad_norm": 17.25,
+      "learning_rate": 2.485305513894197e-06,
+      "loss": 0.4624,
+      "step": 497
+    },
+    {
+      "epoch": 0.12107950401167031,
+      "grad_norm": 19.625,
+      "learning_rate": 2.4852315393212547e-06,
+      "loss": 0.8714,
+      "step": 498
+    },
+    {
+      "epoch": 0.1213226355458303,
+      "grad_norm": 20.375,
+      "learning_rate": 2.4851573801216254e-06,
+      "loss": 0.9453,
+      "step": 499
+    },
+    {
+      "epoch": 0.12156576707999027,
+      "grad_norm": 19.375,
+      "learning_rate": 2.485083036306394e-06,
+      "loss": 0.953,
+      "step": 500
+    },
+    {
+      "epoch": 0.12180889861415026,
+      "grad_norm": 17.625,
+      "learning_rate": 2.485008507886672e-06,
+      "loss": 0.9461,
+      "step": 501
+    },
+    {
+      "epoch": 0.12205203014831023,
+      "grad_norm": 25.625,
+      "learning_rate": 2.484933794873599e-06,
+      "loss": 0.9967,
+      "step": 502
+    },
+    {
+      "epoch": 0.12229516168247022,
+      "grad_norm": 27.25,
+      "learning_rate": 2.4848588972783426e-06,
+      "loss": 1.2476,
+      "step": 503
+    },
+    {
+      "epoch": 0.12253829321663019,
+      "grad_norm": 14.25,
+      "learning_rate": 2.4847838151120974e-06,
+      "loss": 0.6132,
+      "step": 504
+    },
+    {
+      "epoch": 0.12278142475079018,
+      "grad_norm": 13.75,
+      "learning_rate": 2.4847085483860854e-06,
+      "loss": 0.5193,
+      "step": 505
+    },
+    {
+      "epoch": 0.12302455628495015,
+      "grad_norm": 21.0,
+      "learning_rate": 2.4846330971115563e-06,
+      "loss": 1.1056,
+      "step": 506
+    },
+    {
+      "epoch": 0.12326768781911014,
+      "grad_norm": 22.0,
+      "learning_rate": 2.4845574612997887e-06,
+      "loss": 1.1884,
+      "step": 507
+    },
+    {
+      "epoch": 0.12351081935327012,
+      "grad_norm": 16.125,
+      "learning_rate": 2.4844816409620863e-06,
+      "loss": 0.7524,
+      "step": 508
+    },
+    {
+      "epoch": 0.1237539508874301,
+      "grad_norm": 19.375,
+      "learning_rate": 2.484405636109783e-06,
+      "loss": 0.8291,
+      "step": 509
+    },
+    {
+      "epoch": 0.12399708242159008,
+      "grad_norm": 19.875,
+      "learning_rate": 2.484329446754238e-06,
+      "loss": 0.8617,
+      "step": 510
+    },
+    {
+      "epoch": 0.12424021395575006,
+      "grad_norm": 21.75,
+      "learning_rate": 2.48425307290684e-06,
+      "loss": 1.0044,
+      "step": 511
+    },
+    {
+      "epoch": 0.12448334548991004,
+      "grad_norm": 22.875,
+      "learning_rate": 2.4841765145790034e-06,
+      "loss": 1.0024,
+      "step": 512
+    },
+    {
+      "epoch": 0.12472647702407003,
+      "grad_norm": 23.5,
+      "learning_rate": 2.484099771782172e-06,
+      "loss": 1.1801,
+      "step": 513
+    },
+    {
+      "epoch": 0.12496960855823,
+      "grad_norm": 15.625,
+      "learning_rate": 2.484022844527816e-06,
+      "loss": 0.5301,
+      "step": 514
+    },
+    {
+      "epoch": 0.12521274009238997,
+      "grad_norm": 28.625,
+      "learning_rate": 2.483945732827434e-06,
+      "loss": 1.3416,
+      "step": 515
+    },
+    {
+      "epoch": 0.12545587162654998,
+      "grad_norm": 23.875,
+      "learning_rate": 2.483868436692551e-06,
+      "loss": 1.2757,
+      "step": 516
+    },
+    {
+      "epoch": 0.12569900316070995,
+      "grad_norm": 21.5,
+      "learning_rate": 2.4837909561347202e-06,
+      "loss": 1.1423,
+      "step": 517
+    },
+    {
+      "epoch": 0.12594213469486992,
+      "grad_norm": 17.0,
+      "learning_rate": 2.483713291165523e-06,
+      "loss": 0.8155,
+      "step": 518
+    },
+    {
+      "epoch": 0.1261852662290299,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4836354417965675e-06,
+      "loss": 0.9583,
+      "step": 519
+    },
+    {
+      "epoch": 0.1264283977631899,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.483557408039489e-06,
+      "loss": 0.4277,
+      "step": 520
+    },
+    {
+      "epoch": 0.12667152929734987,
+      "grad_norm": 23.625,
+      "learning_rate": 2.4834791899059524e-06,
+      "loss": 0.829,
+      "step": 521
+    },
+    {
+      "epoch": 0.12691466083150985,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4834007874076475e-06,
+      "loss": 0.8476,
+      "step": 522
+    },
+    {
+      "epoch": 0.12715779236566982,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.4833222005562936e-06,
+      "loss": 0.5097,
+      "step": 523
+    },
+    {
+      "epoch": 0.12740092389982982,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4832434293636364e-06,
+      "loss": 1.2364,
+      "step": 524
+    },
+    {
+      "epoch": 0.1276440554339898,
+      "grad_norm": 17.875,
+      "learning_rate": 2.48316447384145e-06,
+      "loss": 0.8209,
+      "step": 525
+    },
+    {
+      "epoch": 0.12788718696814977,
+      "grad_norm": 20.25,
+      "learning_rate": 2.483085334001535e-06,
+      "loss": 0.8799,
+      "step": 526
+    },
+    {
+      "epoch": 0.12813031850230974,
+      "grad_norm": 22.625,
+      "learning_rate": 2.4830060098557217e-06,
+      "loss": 0.9055,
+      "step": 527
+    },
+    {
+      "epoch": 0.12837345003646974,
+      "grad_norm": 29.375,
+      "learning_rate": 2.482926501415865e-06,
+      "loss": 1.4084,
+      "step": 528
+    },
+    {
+      "epoch": 0.12861658157062972,
+      "grad_norm": 18.375,
+      "learning_rate": 2.482846808693849e-06,
+      "loss": 0.9058,
+      "step": 529
+    },
+    {
+      "epoch": 0.1288597131047897,
+      "grad_norm": 29.0,
+      "learning_rate": 2.4827669317015857e-06,
+      "loss": 1.0377,
+      "step": 530
+    },
+    {
+      "epoch": 0.12910284463894967,
+      "grad_norm": 20.0,
+      "learning_rate": 2.4826868704510137e-06,
+      "loss": 1.447,
+      "step": 531
+    },
+    {
+      "epoch": 0.12934597617310964,
+      "grad_norm": 38.0,
+      "learning_rate": 2.4826066249540997e-06,
+      "loss": 1.3393,
+      "step": 532
+    },
+    {
+      "epoch": 0.12958910770726964,
+      "grad_norm": 36.5,
+      "learning_rate": 2.482526195222838e-06,
+      "loss": 1.5646,
+      "step": 533
+    },
+    {
+      "epoch": 0.12983223924142961,
+      "grad_norm": 20.5,
+      "learning_rate": 2.4824455812692495e-06,
+      "loss": 0.9242,
+      "step": 534
+    },
+    {
+      "epoch": 0.1300753707755896,
+      "grad_norm": 16.125,
+      "learning_rate": 2.4823647831053844e-06,
+      "loss": 0.8967,
+      "step": 535
+    },
+    {
+      "epoch": 0.13031850230974956,
+      "grad_norm": 18.25,
+      "learning_rate": 2.482283800743318e-06,
+      "loss": 1.033,
+      "step": 536
+    },
+    {
+      "epoch": 0.13056163384390956,
+      "grad_norm": 30.375,
+      "learning_rate": 2.482202634195156e-06,
+      "loss": 1.1283,
+      "step": 537
+    },
+    {
+      "epoch": 0.13080476537806954,
+      "grad_norm": 21.5,
+      "learning_rate": 2.482121283473029e-06,
+      "loss": 1.1388,
+      "step": 538
+    },
+    {
+      "epoch": 0.1310478969122295,
+      "grad_norm": 23.5,
+      "learning_rate": 2.482039748589097e-06,
+      "loss": 1.4592,
+      "step": 539
+    },
+    {
+      "epoch": 0.13129102844638948,
+      "grad_norm": 22.375,
+      "learning_rate": 2.4819580295555467e-06,
+      "loss": 1.0908,
+      "step": 540
+    },
+    {
+      "epoch": 0.1315341599805495,
+      "grad_norm": 20.875,
+      "learning_rate": 2.481876126384592e-06,
+      "loss": 1.2061,
+      "step": 541
+    },
+    {
+      "epoch": 0.13177729151470946,
+      "grad_norm": 19.625,
+      "learning_rate": 2.481794039088475e-06,
+      "loss": 1.0404,
+      "step": 542
+    },
+    {
+      "epoch": 0.13202042304886943,
+      "grad_norm": 22.5,
+      "learning_rate": 2.4817117676794647e-06,
+      "loss": 1.2118,
+      "step": 543
+    },
+    {
+      "epoch": 0.1322635545830294,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.4816293121698586e-06,
+      "loss": 0.967,
+      "step": 544
+    },
+    {
+      "epoch": 0.1325066861171894,
+      "grad_norm": 21.375,
+      "learning_rate": 2.481546672571981e-06,
+      "loss": 1.1325,
+      "step": 545
+    },
+    {
+      "epoch": 0.13274981765134938,
+      "grad_norm": 16.625,
+      "learning_rate": 2.481463848898183e-06,
+      "loss": 0.7889,
+      "step": 546
+    },
+    {
+      "epoch": 0.13299294918550936,
+      "grad_norm": 22.125,
+      "learning_rate": 2.481380841160845e-06,
+      "loss": 0.7796,
+      "step": 547
+    },
+    {
+      "epoch": 0.13323608071966933,
+      "grad_norm": 19.875,
+      "learning_rate": 2.481297649372374e-06,
+      "loss": 0.9637,
+      "step": 548
+    },
+    {
+      "epoch": 0.13347921225382933,
+      "grad_norm": 17.875,
+      "learning_rate": 2.481214273545204e-06,
+      "loss": 0.7144,
+      "step": 549
+    },
+    {
+      "epoch": 0.1337223437879893,
+      "grad_norm": 20.375,
+      "learning_rate": 2.4811307136917966e-06,
+      "loss": 1.0322,
+      "step": 550
+    },
+    {
+      "epoch": 0.13396547532214928,
+      "grad_norm": 29.75,
+      "learning_rate": 2.481046969824642e-06,
+      "loss": 1.5122,
+      "step": 551
+    },
+    {
+      "epoch": 0.13420860685630925,
+      "grad_norm": 18.625,
+      "learning_rate": 2.4809630419562567e-06,
+      "loss": 1.0251,
+      "step": 552
+    },
+    {
+      "epoch": 0.13445173839046926,
+      "grad_norm": 20.75,
+      "learning_rate": 2.4808789300991853e-06,
+      "loss": 1.6416,
+      "step": 553
+    },
+    {
+      "epoch": 0.13469486992462923,
+      "grad_norm": 23.5,
+      "learning_rate": 2.4807946342659995e-06,
+      "loss": 1.1206,
+      "step": 554
+    },
+    {
+      "epoch": 0.1349380014587892,
+      "grad_norm": 18.125,
+      "learning_rate": 2.4807101544692995e-06,
+      "loss": 0.9281,
+      "step": 555
+    },
+    {
+      "epoch": 0.13518113299294918,
+      "grad_norm": 17.75,
+      "learning_rate": 2.480625490721712e-06,
+      "loss": 1.0279,
+      "step": 556
+    },
+    {
+      "epoch": 0.13542426452710918,
+      "grad_norm": 27.375,
+      "learning_rate": 2.480540643035891e-06,
+      "loss": 1.4237,
+      "step": 557
+    },
+    {
+      "epoch": 0.13566739606126915,
+      "grad_norm": 24.5,
+      "learning_rate": 2.4804556114245183e-06,
+      "loss": 1.2811,
+      "step": 558
+    },
+    {
+      "epoch": 0.13591052759542913,
+      "grad_norm": 17.375,
+      "learning_rate": 2.4803703959003044e-06,
+      "loss": 0.62,
+      "step": 559
+    },
+    {
+      "epoch": 0.1361536591295891,
+      "grad_norm": 17.625,
+      "learning_rate": 2.480284996475985e-06,
+      "loss": 0.8716,
+      "step": 560
+    },
+    {
+      "epoch": 0.1363967906637491,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4801994131643255e-06,
+      "loss": 0.6587,
+      "step": 561
+    },
+    {
+      "epoch": 0.13663992219790908,
+      "grad_norm": 31.0,
+      "learning_rate": 2.4801136459781177e-06,
+      "loss": 1.0227,
+      "step": 562
+    },
+    {
+      "epoch": 0.13688305373206905,
+      "grad_norm": 19.5,
+      "learning_rate": 2.48002769493018e-06,
+      "loss": 1.1458,
+      "step": 563
+    },
+    {
+      "epoch": 0.13712618526622902,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4799415600333606e-06,
+      "loss": 0.9259,
+      "step": 564
+    },
+    {
+      "epoch": 0.13736931680038902,
+      "grad_norm": 18.125,
+      "learning_rate": 2.4798552413005327e-06,
+      "loss": 0.5576,
+      "step": 565
+    },
+    {
+      "epoch": 0.137612448334549,
+      "grad_norm": 25.125,
+      "learning_rate": 2.479768738744599e-06,
+      "loss": 1.3333,
+      "step": 566
+    },
+    {
+      "epoch": 0.13785557986870897,
+      "grad_norm": 17.625,
+      "learning_rate": 2.479682052378489e-06,
+      "loss": 1.0037,
+      "step": 567
+    },
+    {
+      "epoch": 0.13809871140286895,
+      "grad_norm": 21.5,
+      "learning_rate": 2.479595182215158e-06,
+      "loss": 1.0878,
+      "step": 568
+    },
+    {
+      "epoch": 0.13834184293702892,
+      "grad_norm": 20.25,
+      "learning_rate": 2.4795081282675917e-06,
+      "loss": 0.7082,
+      "step": 569
+    },
+    {
+      "epoch": 0.13858497447118892,
+      "grad_norm": 20.75,
+      "learning_rate": 2.479420890548801e-06,
+      "loss": 0.8939,
+      "step": 570
+    },
+    {
+      "epoch": 0.1388281060053489,
+      "grad_norm": 22.125,
+      "learning_rate": 2.4793334690718253e-06,
+      "loss": 0.9109,
+      "step": 571
+    },
+    {
+      "epoch": 0.13907123753950887,
+      "grad_norm": 21.0,
+      "learning_rate": 2.479245863849731e-06,
+      "loss": 0.8302,
+      "step": 572
+    },
+    {
+      "epoch": 0.13931436907366884,
+      "grad_norm": 22.0,
+      "learning_rate": 2.4791580748956133e-06,
+      "loss": 0.8713,
+      "step": 573
+    },
+    {
+      "epoch": 0.13955750060782884,
+      "grad_norm": 22.25,
+      "learning_rate": 2.479070102222593e-06,
+      "loss": 0.9753,
+      "step": 574
+    },
+    {
+      "epoch": 0.13980063214198882,
+      "grad_norm": 17.75,
+      "learning_rate": 2.478981945843819e-06,
+      "loss": 0.6024,
+      "step": 575
+    },
+    {
+      "epoch": 0.1400437636761488,
+      "grad_norm": 20.0,
+      "learning_rate": 2.478893605772468e-06,
+      "loss": 1.0587,
+      "step": 576
+    },
+    {
+      "epoch": 0.14028689521030877,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4788050820217437e-06,
+      "loss": 0.9547,
+      "step": 577
+    },
+    {
+      "epoch": 0.14053002674446877,
+      "grad_norm": 27.375,
+      "learning_rate": 2.4787163746048776e-06,
+      "loss": 1.2551,
+      "step": 578
+    },
+    {
+      "epoch": 0.14077315827862874,
+      "grad_norm": 19.125,
+      "learning_rate": 2.478627483535129e-06,
+      "loss": 0.9121,
+      "step": 579
+    },
+    {
+      "epoch": 0.14101628981278871,
+      "grad_norm": 21.75,
+      "learning_rate": 2.4785384088257835e-06,
+      "loss": 0.6421,
+      "step": 580
+    },
+    {
+      "epoch": 0.1412594213469487,
+      "grad_norm": 16.375,
+      "learning_rate": 2.478449150490155e-06,
+      "loss": 0.7209,
+      "step": 581
+    },
+    {
+      "epoch": 0.1415025528811087,
+      "grad_norm": 23.375,
+      "learning_rate": 2.4783597085415855e-06,
+      "loss": 1.075,
+      "step": 582
+    },
+    {
+      "epoch": 0.14174568441526866,
+      "grad_norm": 19.625,
+      "learning_rate": 2.4782700829934423e-06,
+      "loss": 1.0413,
+      "step": 583
+    },
+    {
+      "epoch": 0.14198881594942864,
+      "grad_norm": 17.75,
+      "learning_rate": 2.4781802738591232e-06,
+      "loss": 0.7188,
+      "step": 584
+    },
+    {
+      "epoch": 0.1422319474835886,
+      "grad_norm": 18.25,
+      "learning_rate": 2.4780902811520503e-06,
+      "loss": 0.8088,
+      "step": 585
+    },
+    {
+      "epoch": 0.1424750790177486,
+      "grad_norm": 20.625,
+      "learning_rate": 2.478000104885675e-06,
+      "loss": 1.2832,
+      "step": 586
+    },
+    {
+      "epoch": 0.1427182105519086,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.4779097450734756e-06,
+      "loss": 0.6821,
+      "step": 587
+    },
+    {
+      "epoch": 0.14296134208606856,
+      "grad_norm": 21.875,
+      "learning_rate": 2.477819201728958e-06,
+      "loss": 0.8267,
+      "step": 588
+    },
+    {
+      "epoch": 0.14320447362022853,
+      "grad_norm": 25.125,
+      "learning_rate": 2.477728474865656e-06,
+      "loss": 1.3358,
+      "step": 589
+    },
+    {
+      "epoch": 0.14344760515438854,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.4776375644971297e-06,
+      "loss": 0.6596,
+      "step": 590
+    },
+    {
+      "epoch": 0.1436907366885485,
+      "grad_norm": 19.25,
+      "learning_rate": 2.477546470636967e-06,
+      "loss": 0.9614,
+      "step": 591
+    },
+    {
+      "epoch": 0.14393386822270848,
+      "grad_norm": 16.75,
+      "learning_rate": 2.477455193298784e-06,
+      "loss": 0.8964,
+      "step": 592
+    },
+    {
+      "epoch": 0.14417699975686846,
+      "grad_norm": 15.375,
+      "learning_rate": 2.4773637324962236e-06,
+      "loss": 0.8571,
+      "step": 593
+    },
+    {
+      "epoch": 0.14442013129102846,
+      "grad_norm": 17.0,
+      "learning_rate": 2.4772720882429557e-06,
+      "loss": 0.7895,
+      "step": 594
+    },
+    {
+      "epoch": 0.14466326282518843,
+      "grad_norm": 18.375,
+      "learning_rate": 2.477180260552679e-06,
+      "loss": 0.8672,
+      "step": 595
+    },
+    {
+      "epoch": 0.1449063943593484,
+      "grad_norm": 21.5,
+      "learning_rate": 2.477088249439118e-06,
+      "loss": 1.0772,
+      "step": 596
+    },
+    {
+      "epoch": 0.14514952589350838,
+      "grad_norm": 18.625,
+      "learning_rate": 2.4769960549160255e-06,
+      "loss": 0.9394,
+      "step": 597
+    },
+    {
+      "epoch": 0.14539265742766838,
+      "grad_norm": 32.0,
+      "learning_rate": 2.4769036769971816e-06,
+      "loss": 1.3212,
+      "step": 598
+    },
+    {
+      "epoch": 0.14563578896182836,
+      "grad_norm": 18.125,
+      "learning_rate": 2.4768111156963944e-06,
+      "loss": 1.0702,
+      "step": 599
+    },
+    {
+      "epoch": 0.14587892049598833,
+      "grad_norm": 21.5,
+      "learning_rate": 2.4767183710274974e-06,
+      "loss": 1.245,
+      "step": 600
+    },
+    {
+      "epoch": 0.1461220520301483,
+      "grad_norm": 27.0,
+      "learning_rate": 2.476625443004354e-06,
+      "loss": 1.0605,
+      "step": 601
+    },
+    {
+      "epoch": 0.1463651835643083,
+      "grad_norm": 21.5,
+      "learning_rate": 2.4765323316408537e-06,
+      "loss": 0.4655,
+      "step": 602
+    },
+    {
+      "epoch": 0.14660831509846828,
+      "grad_norm": 21.625,
+      "learning_rate": 2.4764390369509133e-06,
+      "loss": 1.1133,
+      "step": 603
+    },
+    {
+      "epoch": 0.14685144663262825,
+      "grad_norm": 24.0,
+      "learning_rate": 2.4763455589484776e-06,
+      "loss": 1.1572,
+      "step": 604
+    },
+    {
+      "epoch": 0.14709457816678823,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4762518976475184e-06,
+      "loss": 0.7207,
+      "step": 605
+    },
+    {
+      "epoch": 0.1473377097009482,
+      "grad_norm": 22.75,
+      "learning_rate": 2.476158053062035e-06,
+      "loss": 0.9064,
+      "step": 606
+    },
+    {
+      "epoch": 0.1475808412351082,
+      "grad_norm": 26.875,
+      "learning_rate": 2.476064025206054e-06,
+      "loss": 1.184,
+      "step": 607
+    },
+    {
+      "epoch": 0.14782397276926817,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4759698140936294e-06,
+      "loss": 0.8613,
+      "step": 608
+    },
+    {
+      "epoch": 0.14806710430342815,
+      "grad_norm": 18.25,
+      "learning_rate": 2.4758754197388433e-06,
+      "loss": 0.6474,
+      "step": 609
+    },
+    {
+      "epoch": 0.14831023583758812,
+      "grad_norm": 14.75,
+      "learning_rate": 2.475780842155804e-06,
+      "loss": 0.6431,
+      "step": 610
+    },
+    {
+      "epoch": 0.14855336737174812,
+      "grad_norm": 17.5,
+      "learning_rate": 2.4756860813586474e-06,
+      "loss": 0.8249,
+      "step": 611
+    },
+    {
+      "epoch": 0.1487964989059081,
+      "grad_norm": 22.5,
+      "learning_rate": 2.4755911373615382e-06,
+      "loss": 1.7489,
+      "step": 612
+    },
+    {
+      "epoch": 0.14903963044006807,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4754960101786663e-06,
+      "loss": 0.9744,
+      "step": 613
+    },
+    {
+      "epoch": 0.14928276197422805,
+      "grad_norm": 16.875,
+      "learning_rate": 2.4754006998242513e-06,
+      "loss": 0.7147,
+      "step": 614
+    },
+    {
+      "epoch": 0.14952589350838805,
+      "grad_norm": 18.75,
+      "learning_rate": 2.4753052063125377e-06,
+      "loss": 0.8841,
+      "step": 615
+    },
+    {
+      "epoch": 0.14976902504254802,
+      "grad_norm": 17.0,
+      "learning_rate": 2.4752095296577996e-06,
+      "loss": 0.8324,
+      "step": 616
+    },
+    {
+      "epoch": 0.150012156576708,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4751136698743372e-06,
+      "loss": 0.9666,
+      "step": 617
+    },
+    {
+      "epoch": 0.15025528811086797,
+      "grad_norm": 19.875,
+      "learning_rate": 2.475017626976478e-06,
+      "loss": 0.9407,
+      "step": 618
+    },
+    {
+      "epoch": 0.15049841964502797,
+      "grad_norm": 21.875,
+      "learning_rate": 2.4749214009785784e-06,
+      "loss": 0.9423,
+      "step": 619
+    },
+    {
+      "epoch": 0.15074155117918794,
+      "grad_norm": 22.625,
+      "learning_rate": 2.4748249918950196e-06,
+      "loss": 1.0306,
+      "step": 620
+    },
+    {
+      "epoch": 0.15098468271334792,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4747283997402128e-06,
+      "loss": 1.0252,
+      "step": 621
+    },
+    {
+      "epoch": 0.1512278142475079,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4746316245285947e-06,
+      "loss": 0.955,
+      "step": 622
+    },
+    {
+      "epoch": 0.1514709457816679,
+      "grad_norm": 22.625,
+      "learning_rate": 2.47453466627463e-06,
+      "loss": 0.7389,
+      "step": 623
+    },
+    {
+      "epoch": 0.15171407731582787,
+      "grad_norm": 21.125,
+      "learning_rate": 2.474437524992811e-06,
+      "loss": 1.063,
+      "step": 624
+    },
+    {
+      "epoch": 0.15195720884998784,
+      "grad_norm": 18.125,
+      "learning_rate": 2.4743402006976573e-06,
+      "loss": 0.7157,
+      "step": 625
+    },
+    {
+      "epoch": 0.15220034038414781,
+      "grad_norm": 16.75,
+      "learning_rate": 2.4742426934037155e-06,
+      "loss": 0.8126,
+      "step": 626
+    },
+    {
+      "epoch": 0.15244347191830782,
+      "grad_norm": 22.5,
+      "learning_rate": 2.4741450031255595e-06,
+      "loss": 0.977,
+      "step": 627
+    },
+    {
+      "epoch": 0.1526866034524678,
+      "grad_norm": 24.5,
+      "learning_rate": 2.4740471298777914e-06,
+      "loss": 1.0679,
+      "step": 628
+    },
+    {
+      "epoch": 0.15292973498662776,
+      "grad_norm": 20.375,
+      "learning_rate": 2.4739490736750393e-06,
+      "loss": 0.6969,
+      "step": 629
+    },
+    {
+      "epoch": 0.15317286652078774,
+      "grad_norm": 23.625,
+      "learning_rate": 2.47385083453196e-06,
+      "loss": 0.8285,
+      "step": 630
+    },
+    {
+      "epoch": 0.15341599805494774,
+      "grad_norm": 21.0,
+      "learning_rate": 2.4737524124632373e-06,
+      "loss": 0.9942,
+      "step": 631
+    },
+    {
+      "epoch": 0.1536591295891077,
+      "grad_norm": 20.875,
+      "learning_rate": 2.4736538074835812e-06,
+      "loss": 0.9077,
+      "step": 632
+    },
+    {
+      "epoch": 0.1539022611232677,
+      "grad_norm": 15.125,
+      "learning_rate": 2.4735550196077304e-06,
+      "loss": 0.6105,
+      "step": 633
+    },
+    {
+      "epoch": 0.15414539265742766,
+      "grad_norm": 20.625,
+      "learning_rate": 2.4734560488504507e-06,
+      "loss": 1.1664,
+      "step": 634
+    },
+    {
+      "epoch": 0.15438852419158766,
+      "grad_norm": 18.75,
+      "learning_rate": 2.4733568952265342e-06,
+      "loss": 1.0338,
+      "step": 635
+    },
+    {
+      "epoch": 0.15463165572574764,
+      "grad_norm": 27.375,
+      "learning_rate": 2.4732575587508016e-06,
+      "loss": 0.6824,
+      "step": 636
+    },
+    {
+      "epoch": 0.1548747872599076,
+      "grad_norm": 28.5,
+      "learning_rate": 2.4731580394381005e-06,
+      "loss": 1.1426,
+      "step": 637
+    },
+    {
+      "epoch": 0.15511791879406758,
+      "grad_norm": 19.125,
+      "learning_rate": 2.473058337303306e-06,
+      "loss": 0.7891,
+      "step": 638
+    },
+    {
+      "epoch": 0.15536105032822758,
+      "grad_norm": 23.0,
+      "learning_rate": 2.4729584523613196e-06,
+      "loss": 0.9048,
+      "step": 639
+    },
+    {
+      "epoch": 0.15560418186238756,
+      "grad_norm": 17.125,
+      "learning_rate": 2.472858384627072e-06,
+      "loss": 1.0675,
+      "step": 640
+    },
+    {
+      "epoch": 0.15584731339654753,
+      "grad_norm": 18.75,
+      "learning_rate": 2.4727581341155186e-06,
+      "loss": 0.7432,
+      "step": 641
+    },
+    {
+      "epoch": 0.1560904449307075,
+      "grad_norm": 16.125,
+      "learning_rate": 2.472657700841645e-06,
+      "loss": 0.8413,
+      "step": 642
+    },
+    {
+      "epoch": 0.15633357646486748,
+      "grad_norm": 17.375,
+      "learning_rate": 2.4725570848204615e-06,
+      "loss": 0.5183,
+      "step": 643
+    },
+    {
+      "epoch": 0.15657670799902748,
+      "grad_norm": 20.625,
+      "learning_rate": 2.472456286067007e-06,
+      "loss": 0.778,
+      "step": 644
+    },
+    {
+      "epoch": 0.15681983953318746,
+      "grad_norm": 14.375,
+      "learning_rate": 2.4723553045963488e-06,
+      "loss": 0.6765,
+      "step": 645
+    },
+    {
+      "epoch": 0.15706297106734743,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4722541404235793e-06,
+      "loss": 0.9651,
+      "step": 646
+    },
+    {
+      "epoch": 0.1573061026015074,
+      "grad_norm": 20.5,
+      "learning_rate": 2.4721527935638194e-06,
+      "loss": 1.0763,
+      "step": 647
+    },
+    {
+      "epoch": 0.1575492341356674,
+      "grad_norm": 20.0,
+      "learning_rate": 2.472051264032217e-06,
+      "loss": 1.0098,
+      "step": 648
+    },
+    {
+      "epoch": 0.15779236566982738,
+      "grad_norm": 21.625,
+      "learning_rate": 2.471949551843948e-06,
+      "loss": 1.3047,
+      "step": 649
+    },
+    {
+      "epoch": 0.15803549720398735,
+      "grad_norm": 20.25,
+      "learning_rate": 2.4718476570142142e-06,
+      "loss": 1.3861,
+      "step": 650
+    },
+    {
+      "epoch": 0.15827862873814733,
+      "grad_norm": 19.25,
+      "learning_rate": 2.4717455795582462e-06,
+      "loss": 0.919,
+      "step": 651
+    },
+    {
+      "epoch": 0.15852176027230733,
+      "grad_norm": 16.125,
+      "learning_rate": 2.471643319491301e-06,
+      "loss": 0.5179,
+      "step": 652
+    },
+    {
+      "epoch": 0.1587648918064673,
+      "grad_norm": 22.375,
+      "learning_rate": 2.4715408768286638e-06,
+      "loss": 1.1422,
+      "step": 653
+    },
+    {
+      "epoch": 0.15900802334062727,
+      "grad_norm": 15.875,
+      "learning_rate": 2.471438251585645e-06,
+      "loss": 0.6059,
+      "step": 654
+    },
+    {
+      "epoch": 0.15925115487478725,
+      "grad_norm": 24.0,
+      "learning_rate": 2.471335443777585e-06,
+      "loss": 0.9181,
+      "step": 655
+    },
+    {
+      "epoch": 0.15949428640894725,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4712324534198497e-06,
+      "loss": 0.952,
+      "step": 656
+    },
+    {
+      "epoch": 0.15973741794310722,
+      "grad_norm": 15.625,
+      "learning_rate": 2.4711292805278327e-06,
+      "loss": 0.9161,
+      "step": 657
+    },
+    {
+      "epoch": 0.1599805494772672,
+      "grad_norm": 25.375,
+      "learning_rate": 2.471025925116955e-06,
+      "loss": 0.9753,
+      "step": 658
+    },
+    {
+      "epoch": 0.16022368101142717,
+      "grad_norm": 20.125,
+      "learning_rate": 2.470922387202665e-06,
+      "loss": 0.4339,
+      "step": 659
+    },
+    {
+      "epoch": 0.16046681254558717,
+      "grad_norm": 20.625,
+      "learning_rate": 2.470818666800438e-06,
+      "loss": 1.052,
+      "step": 660
+    },
+    {
+      "epoch": 0.16070994407974715,
+      "grad_norm": 23.375,
+      "learning_rate": 2.470714763925777e-06,
+      "loss": 0.8525,
+      "step": 661
+    },
+    {
+      "epoch": 0.16095307561390712,
+      "grad_norm": 15.625,
+      "learning_rate": 2.4706106785942123e-06,
+      "loss": 0.7993,
+      "step": 662
+    },
+    {
+      "epoch": 0.1611962071480671,
+      "grad_norm": 22.125,
+      "learning_rate": 2.470506410821301e-06,
+      "loss": 0.9482,
+      "step": 663
+    },
+    {
+      "epoch": 0.1614393386822271,
+      "grad_norm": 21.5,
+      "learning_rate": 2.470401960622628e-06,
+      "loss": 1.0482,
+      "step": 664
+    },
+    {
+      "epoch": 0.16168247021638707,
+      "grad_norm": 29.0,
+      "learning_rate": 2.4702973280138044e-06,
+      "loss": 1.6028,
+      "step": 665
+    },
+    {
+      "epoch": 0.16192560175054704,
+      "grad_norm": 17.375,
+      "learning_rate": 2.4701925130104705e-06,
+      "loss": 1.0979,
+      "step": 666
+    },
+    {
+      "epoch": 0.16216873328470702,
+      "grad_norm": 19.875,
+      "learning_rate": 2.4700875156282918e-06,
+      "loss": 0.8368,
+      "step": 667
+    },
+    {
+      "epoch": 0.16241186481886702,
+      "grad_norm": 18.125,
+      "learning_rate": 2.4699823358829616e-06,
+      "loss": 1.1269,
+      "step": 668
+    },
+    {
+      "epoch": 0.162654996353027,
+      "grad_norm": 21.5,
+      "learning_rate": 2.469876973790202e-06,
+      "loss": 1.0932,
+      "step": 669
+    },
+    {
+      "epoch": 0.16289812788718697,
+      "grad_norm": 15.25,
+      "learning_rate": 2.4697714293657608e-06,
+      "loss": 0.5482,
+      "step": 670
+    },
+    {
+      "epoch": 0.16314125942134694,
+      "grad_norm": 14.875,
+      "learning_rate": 2.4696657026254133e-06,
+      "loss": 0.4568,
+      "step": 671
+    },
+    {
+      "epoch": 0.16338439095550694,
+      "grad_norm": 21.0,
+      "learning_rate": 2.469559793584962e-06,
+      "loss": 1.2617,
+      "step": 672
+    },
+    {
+      "epoch": 0.16362752248966692,
+      "grad_norm": 16.375,
+      "learning_rate": 2.4694537022602367e-06,
+      "loss": 0.6635,
+      "step": 673
+    },
+    {
+      "epoch": 0.1638706540238269,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4693474286670955e-06,
+      "loss": 0.8394,
+      "step": 674
+    },
+    {
+      "epoch": 0.16411378555798686,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4692409728214216e-06,
+      "loss": 1.0343,
+      "step": 675
+    },
+    {
+      "epoch": 0.16435691709214686,
+      "grad_norm": 22.375,
+      "learning_rate": 2.469134334739128e-06,
+      "loss": 0.9508,
+      "step": 676
+    },
+    {
+      "epoch": 0.16460004862630684,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.4690275144361526e-06,
+      "loss": 0.6007,
+      "step": 677
+    },
+    {
+      "epoch": 0.1648431801604668,
+      "grad_norm": 19.125,
+      "learning_rate": 2.4689205119284618e-06,
+      "loss": 0.629,
+      "step": 678
+    },
+    {
+      "epoch": 0.16508631169462679,
+      "grad_norm": 34.75,
+      "learning_rate": 2.468813327232049e-06,
+      "loss": 1.4319,
+      "step": 679
+    },
+    {
+      "epoch": 0.16532944322878676,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4687059603629348e-06,
+      "loss": 0.8659,
+      "step": 680
+    },
+    {
+      "epoch": 0.16557257476294676,
+      "grad_norm": 16.5,
+      "learning_rate": 2.4685984113371668e-06,
+      "loss": 0.6219,
+      "step": 681
+    },
+    {
+      "epoch": 0.16581570629710674,
+      "grad_norm": 15.375,
+      "learning_rate": 2.468490680170821e-06,
+      "loss": 0.7307,
+      "step": 682
+    },
+    {
+      "epoch": 0.1660588378312667,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4683827668799985e-06,
+      "loss": 0.7113,
+      "step": 683
+    },
+    {
+      "epoch": 0.16630196936542668,
+      "grad_norm": 23.25,
+      "learning_rate": 2.468274671480829e-06,
+      "loss": 1.0579,
+      "step": 684
+    },
+    {
+      "epoch": 0.16654510089958668,
+      "grad_norm": 18.25,
+      "learning_rate": 2.4681663939894703e-06,
+      "loss": 1.0621,
+      "step": 685
+    },
+    {
+      "epoch": 0.16678823243374666,
+      "grad_norm": 14.25,
+      "learning_rate": 2.468057934422105e-06,
+      "loss": 0.5793,
+      "step": 686
+    },
+    {
+      "epoch": 0.16703136396790663,
+      "grad_norm": 27.625,
+      "learning_rate": 2.467949292794945e-06,
+      "loss": 1.1904,
+      "step": 687
+    },
+    {
+      "epoch": 0.1672744955020666,
+      "grad_norm": 15.125,
+      "learning_rate": 2.4678404691242285e-06,
+      "loss": 0.5465,
+      "step": 688
+    },
+    {
+      "epoch": 0.1675176270362266,
+      "grad_norm": 27.5,
+      "learning_rate": 2.4677314634262206e-06,
+      "loss": 1.4834,
+      "step": 689
+    },
+    {
+      "epoch": 0.16776075857038658,
+      "grad_norm": 17.875,
+      "learning_rate": 2.467622275717215e-06,
+      "loss": 0.9697,
+      "step": 690
+    },
+    {
+      "epoch": 0.16800389010454655,
+      "grad_norm": 20.125,
+      "learning_rate": 2.467512906013531e-06,
+      "loss": 1.0066,
+      "step": 691
+    },
+    {
+      "epoch": 0.16824702163870653,
+      "grad_norm": 22.375,
+      "learning_rate": 2.4674033543315164e-06,
+      "loss": 1.0141,
+      "step": 692
+    },
+    {
+      "epoch": 0.16849015317286653,
+      "grad_norm": 23.125,
+      "learning_rate": 2.467293620687545e-06,
+      "loss": 1.1554,
+      "step": 693
+    },
+    {
+      "epoch": 0.1687332847070265,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4671837050980186e-06,
+      "loss": 0.8953,
+      "step": 694
+    },
+    {
+      "epoch": 0.16897641624118648,
+      "grad_norm": 14.875,
+      "learning_rate": 2.467073607579366e-06,
+      "loss": 0.7851,
+      "step": 695
+    },
+    {
+      "epoch": 0.16921954777534645,
+      "grad_norm": 12.5,
+      "learning_rate": 2.466963328148043e-06,
+      "loss": 0.3995,
+      "step": 696
+    },
+    {
+      "epoch": 0.16946267930950645,
+      "grad_norm": 18.5,
+      "learning_rate": 2.466852866820533e-06,
+      "loss": 1.0252,
+      "step": 697
+    },
+    {
+      "epoch": 0.16970581084366643,
+      "grad_norm": 22.25,
+      "learning_rate": 2.4667422236133463e-06,
+      "loss": 0.9539,
+      "step": 698
+    },
+    {
+      "epoch": 0.1699489423778264,
+      "grad_norm": 23.75,
+      "learning_rate": 2.4666313985430205e-06,
+      "loss": 1.281,
+      "step": 699
+    },
+    {
+      "epoch": 0.17019207391198637,
+      "grad_norm": 17.875,
+      "learning_rate": 2.46652039162612e-06,
+      "loss": 0.8046,
+      "step": 700
+    },
+    {
+      "epoch": 0.17043520544614638,
+      "grad_norm": 14.0,
+      "learning_rate": 2.466409202879237e-06,
+      "loss": 0.4778,
+      "step": 701
+    },
+    {
+      "epoch": 0.17067833698030635,
+      "grad_norm": 19.875,
+      "learning_rate": 2.4662978323189907e-06,
+      "loss": 1.0883,
+      "step": 702
+    },
+    {
+      "epoch": 0.17092146851446632,
+      "grad_norm": 21.625,
+      "learning_rate": 2.4661862799620275e-06,
+      "loss": 0.9669,
+      "step": 703
+    },
+    {
+      "epoch": 0.1711646000486263,
+      "grad_norm": 19.375,
+      "learning_rate": 2.4660745458250197e-06,
+      "loss": 0.9314,
+      "step": 704
+    },
+    {
+      "epoch": 0.1714077315827863,
+      "grad_norm": 23.75,
+      "learning_rate": 2.465962629924669e-06,
+      "loss": 1.1983,
+      "step": 705
+    },
+    {
+      "epoch": 0.17165086311694627,
+      "grad_norm": 21.875,
+      "learning_rate": 2.4658505322777032e-06,
+      "loss": 0.9296,
+      "step": 706
+    },
+    {
+      "epoch": 0.17189399465110625,
+      "grad_norm": 20.25,
+      "learning_rate": 2.4657382529008765e-06,
+      "loss": 1.1447,
+      "step": 707
+    },
+    {
+      "epoch": 0.17213712618526622,
+      "grad_norm": 19.375,
+      "learning_rate": 2.4656257918109716e-06,
+      "loss": 1.5269,
+      "step": 708
+    },
+    {
+      "epoch": 0.17238025771942622,
+      "grad_norm": 19.375,
+      "learning_rate": 2.4655131490247974e-06,
+      "loss": 0.5523,
+      "step": 709
+    },
+    {
+      "epoch": 0.1726233892535862,
+      "grad_norm": 38.5,
+      "learning_rate": 2.4654003245591905e-06,
+      "loss": 1.1902,
+      "step": 710
+    },
+    {
+      "epoch": 0.17286652078774617,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4652873184310143e-06,
+      "loss": 0.7753,
+      "step": 711
+    },
+    {
+      "epoch": 0.17310965232190614,
+      "grad_norm": 16.25,
+      "learning_rate": 2.4651741306571596e-06,
+      "loss": 0.6987,
+      "step": 712
+    },
+    {
+      "epoch": 0.17335278385606614,
+      "grad_norm": 20.875,
+      "learning_rate": 2.465060761254544e-06,
+      "loss": 0.8424,
+      "step": 713
+    },
+    {
+      "epoch": 0.17359591539022612,
+      "grad_norm": 23.0,
+      "learning_rate": 2.4649472102401134e-06,
+      "loss": 0.976,
+      "step": 714
+    },
+    {
+      "epoch": 0.1738390469243861,
+      "grad_norm": 19.625,
+      "learning_rate": 2.4648334776308395e-06,
+      "loss": 1.0214,
+      "step": 715
+    },
+    {
+      "epoch": 0.17408217845854607,
+      "grad_norm": 18.0,
+      "learning_rate": 2.464719563443721e-06,
+      "loss": 0.5665,
+      "step": 716
+    },
+    {
+      "epoch": 0.17432530999270604,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4646054676957847e-06,
+      "loss": 0.7222,
+      "step": 717
+    },
+    {
+      "epoch": 0.17456844152686604,
+      "grad_norm": 23.5,
+      "learning_rate": 2.4644911904040846e-06,
+      "loss": 1.3754,
+      "step": 718
+    },
+    {
+      "epoch": 0.17481157306102602,
+      "grad_norm": 17.0,
+      "learning_rate": 2.4643767315857013e-06,
+      "loss": 0.8122,
+      "step": 719
+    },
+    {
+      "epoch": 0.175054704595186,
+      "grad_norm": 17.625,
+      "learning_rate": 2.464262091257742e-06,
+      "loss": 0.5982,
+      "step": 720
+    },
+    {
+      "epoch": 0.17529783612934596,
+      "grad_norm": 25.125,
+      "learning_rate": 2.4641472694373427e-06,
+      "loss": 1.4758,
+      "step": 721
+    },
+    {
+      "epoch": 0.17554096766350596,
+      "grad_norm": 23.125,
+      "learning_rate": 2.4640322661416645e-06,
+      "loss": 1.1522,
+      "step": 722
+    },
+    {
+      "epoch": 0.17578409919766594,
+      "grad_norm": 16.875,
+      "learning_rate": 2.463917081387897e-06,
+      "loss": 0.8721,
+      "step": 723
+    },
+    {
+      "epoch": 0.1760272307318259,
+      "grad_norm": 16.25,
+      "learning_rate": 2.4638017151932565e-06,
+      "loss": 0.7402,
+      "step": 724
+    },
+    {
+      "epoch": 0.17627036226598589,
+      "grad_norm": 17.375,
+      "learning_rate": 2.463686167574987e-06,
+      "loss": 0.9078,
+      "step": 725
+    },
+    {
+      "epoch": 0.1765134938001459,
+      "grad_norm": 23.875,
+      "learning_rate": 2.4635704385503585e-06,
+      "loss": 1.1702,
+      "step": 726
+    },
+    {
+      "epoch": 0.17675662533430586,
+      "grad_norm": 16.125,
+      "learning_rate": 2.4634545281366688e-06,
+      "loss": 0.4668,
+      "step": 727
+    },
+    {
+      "epoch": 0.17699975686846584,
+      "grad_norm": 44.25,
+      "learning_rate": 2.4633384363512424e-06,
+      "loss": 0.9433,
+      "step": 728
+    },
+    {
+      "epoch": 0.1772428884026258,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.463222163211432e-06,
+      "loss": 0.6542,
+      "step": 729
+    },
+    {
+      "epoch": 0.1774860199367858,
+      "grad_norm": 23.875,
+      "learning_rate": 2.4631057087346166e-06,
+      "loss": 1.027,
+      "step": 730
+    },
+    {
+      "epoch": 0.17772915147094578,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.4629890729382018e-06,
+      "loss": 0.5723,
+      "step": 731
+    },
+    {
+      "epoch": 0.17797228300510576,
+      "grad_norm": 16.5,
+      "learning_rate": 2.4628722558396206e-06,
+      "loss": 0.8273,
+      "step": 732
+    },
+    {
+      "epoch": 0.17821541453926573,
+      "grad_norm": 21.0,
+      "learning_rate": 2.462755257456334e-06,
+      "loss": 1.07,
+      "step": 733
+    },
+    {
+      "epoch": 0.17845854607342573,
+      "grad_norm": 28.875,
+      "learning_rate": 2.4626380778058293e-06,
+      "loss": 0.8394,
+      "step": 734
+    },
+    {
+      "epoch": 0.1787016776075857,
+      "grad_norm": 17.125,
+      "learning_rate": 2.4625207169056204e-06,
+      "loss": 1.3333,
+      "step": 735
+    },
+    {
+      "epoch": 0.17894480914174568,
+      "grad_norm": 25.875,
+      "learning_rate": 2.46240317477325e-06,
+      "loss": 1.1257,
+      "step": 736
+    },
+    {
+      "epoch": 0.17918794067590565,
+      "grad_norm": 19.5,
+      "learning_rate": 2.462285451426286e-06,
+      "loss": 0.9058,
+      "step": 737
+    },
+    {
+      "epoch": 0.17943107221006566,
+      "grad_norm": 22.5,
+      "learning_rate": 2.4621675468823243e-06,
+      "loss": 1.0879,
+      "step": 738
+    },
+    {
+      "epoch": 0.17967420374422563,
+      "grad_norm": 36.5,
+      "learning_rate": 2.4620494611589877e-06,
+      "loss": 0.7984,
+      "step": 739
+    },
+    {
+      "epoch": 0.1799173352783856,
+      "grad_norm": 21.0,
+      "learning_rate": 2.4619311942739266e-06,
+      "loss": 1.0539,
+      "step": 740
+    },
+    {
+      "epoch": 0.18016046681254558,
+      "grad_norm": 14.75,
+      "learning_rate": 2.4618127462448177e-06,
+      "loss": 0.6653,
+      "step": 741
+    },
+    {
+      "epoch": 0.18040359834670558,
+      "grad_norm": 23.75,
+      "learning_rate": 2.4616941170893647e-06,
+      "loss": 1.0389,
+      "step": 742
+    },
+    {
+      "epoch": 0.18064672988086555,
+      "grad_norm": 34.25,
+      "learning_rate": 2.4615753068253e-06,
+      "loss": 0.9567,
+      "step": 743
+    },
+    {
+      "epoch": 0.18088986141502553,
+      "grad_norm": 21.5,
+      "learning_rate": 2.4614563154703808e-06,
+      "loss": 0.8491,
+      "step": 744
+    },
+    {
+      "epoch": 0.1811329929491855,
+      "grad_norm": 18.0,
+      "learning_rate": 2.4613371430423925e-06,
+      "loss": 0.7903,
+      "step": 745
+    },
+    {
+      "epoch": 0.1813761244833455,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.4612177895591475e-06,
+      "loss": 0.7727,
+      "step": 746
+    },
+    {
+      "epoch": 0.18161925601750548,
+      "grad_norm": 19.5,
+      "learning_rate": 2.4610982550384855e-06,
+      "loss": 0.908,
+      "step": 747
+    },
+    {
+      "epoch": 0.18186238755166545,
+      "grad_norm": 24.0,
+      "learning_rate": 2.460978539498273e-06,
+      "loss": 1.0352,
+      "step": 748
+    },
+    {
+      "epoch": 0.18210551908582542,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4608586429564037e-06,
+      "loss": 0.5442,
+      "step": 749
+    },
+    {
+      "epoch": 0.18234865061998543,
+      "grad_norm": 25.125,
+      "learning_rate": 2.4607385654307976e-06,
+      "loss": 1.1106,
+      "step": 750
+    },
+    {
+      "epoch": 0.1825917821541454,
+      "grad_norm": 50.5,
+      "learning_rate": 2.460618306939403e-06,
+      "loss": 1.0396,
+      "step": 751
+    },
+    {
+      "epoch": 0.18283491368830537,
+      "grad_norm": 17.875,
+      "learning_rate": 2.460497867500194e-06,
+      "loss": 0.9884,
+      "step": 752
+    },
+    {
+      "epoch": 0.18307804522246535,
+      "grad_norm": 24.625,
+      "learning_rate": 2.4603772471311727e-06,
+      "loss": 1.0661,
+      "step": 753
+    },
+    {
+      "epoch": 0.18332117675662535,
+      "grad_norm": 18.375,
+      "learning_rate": 2.460256445850368e-06,
+      "loss": 0.7287,
+      "step": 754
+    },
+    {
+      "epoch": 0.18356430829078532,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4601354636758357e-06,
+      "loss": 1.3232,
+      "step": 755
+    },
+    {
+      "epoch": 0.1838074398249453,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4600143006256587e-06,
+      "loss": 0.7209,
+      "step": 756
+    },
+    {
+      "epoch": 0.18405057135910527,
+      "grad_norm": 18.125,
+      "learning_rate": 2.459892956717946e-06,
+      "loss": 0.7975,
+      "step": 757
+    },
+    {
+      "epoch": 0.18429370289326524,
+      "grad_norm": 36.5,
+      "learning_rate": 2.4597714319708365e-06,
+      "loss": 0.6597,
+      "step": 758
+    },
+    {
+      "epoch": 0.18453683442742524,
+      "grad_norm": 18.125,
+      "learning_rate": 2.4596497264024926e-06,
+      "loss": 0.9782,
+      "step": 759
+    },
+    {
+      "epoch": 0.18477996596158522,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4595278400311053e-06,
+      "loss": 1.129,
+      "step": 760
+    },
+    {
+      "epoch": 0.1850230974957452,
+      "grad_norm": 15.25,
+      "learning_rate": 2.4594057728748934e-06,
+      "loss": 0.5534,
+      "step": 761
+    },
+    {
+      "epoch": 0.18526622902990517,
+      "grad_norm": 17.0,
+      "learning_rate": 2.4592835249521013e-06,
+      "loss": 0.6545,
+      "step": 762
+    },
+    {
+      "epoch": 0.18550936056406517,
+      "grad_norm": 17.125,
+      "learning_rate": 2.4591610962810015e-06,
+      "loss": 0.7302,
+      "step": 763
+    },
+    {
+      "epoch": 0.18575249209822514,
+      "grad_norm": 19.875,
+      "learning_rate": 2.4590384868798933e-06,
+      "loss": 0.8268,
+      "step": 764
+    },
+    {
+      "epoch": 0.18599562363238512,
+      "grad_norm": 15.125,
+      "learning_rate": 2.458915696767102e-06,
+      "loss": 0.4773,
+      "step": 765
+    },
+    {
+      "epoch": 0.1862387551665451,
+      "grad_norm": 21.75,
+      "learning_rate": 2.458792725960981e-06,
+      "loss": 0.7745,
+      "step": 766
+    },
+    {
+      "epoch": 0.1864818867007051,
+      "grad_norm": 19.875,
+      "learning_rate": 2.458669574479911e-06,
+      "loss": 1.0215,
+      "step": 767
+    },
+    {
+      "epoch": 0.18672501823486506,
+      "grad_norm": 17.375,
+      "learning_rate": 2.4585462423422984e-06,
+      "loss": 0.7035,
+      "step": 768
+    },
+    {
+      "epoch": 0.18696814976902504,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4584227295665776e-06,
+      "loss": 0.9188,
+      "step": 769
+    },
+    {
+      "epoch": 0.187211281303185,
+      "grad_norm": 22.25,
+      "learning_rate": 2.4582990361712096e-06,
+      "loss": 0.9235,
+      "step": 770
+    },
+    {
+      "epoch": 0.187454412837345,
+      "grad_norm": 33.75,
+      "learning_rate": 2.4581751621746827e-06,
+      "loss": 1.1085,
+      "step": 771
+    },
+    {
+      "epoch": 0.187697544371505,
+      "grad_norm": 31.125,
+      "learning_rate": 2.458051107595512e-06,
+      "loss": 1.2226,
+      "step": 772
+    },
+    {
+      "epoch": 0.18794067590566496,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4579268724522392e-06,
+      "loss": 0.9582,
+      "step": 773
+    },
+    {
+      "epoch": 0.18818380743982493,
+      "grad_norm": 21.125,
+      "learning_rate": 2.457802456763434e-06,
+      "loss": 1.123,
+      "step": 774
+    },
+    {
+      "epoch": 0.18842693897398494,
+      "grad_norm": 17.0,
+      "learning_rate": 2.457677860547692e-06,
+      "loss": 0.7536,
+      "step": 775
+    },
+    {
+      "epoch": 0.1886700705081449,
+      "grad_norm": 20.25,
+      "learning_rate": 2.4575530838236364e-06,
+      "loss": 1.163,
+      "step": 776
+    },
+    {
+      "epoch": 0.18891320204230488,
+      "grad_norm": 23.875,
+      "learning_rate": 2.4574281266099172e-06,
+      "loss": 0.929,
+      "step": 777
+    },
+    {
+      "epoch": 0.18915633357646486,
+      "grad_norm": 22.875,
+      "learning_rate": 2.4573029889252115e-06,
+      "loss": 1.1094,
+      "step": 778
+    },
+    {
+      "epoch": 0.18939946511062486,
+      "grad_norm": 18.75,
+      "learning_rate": 2.4571776707882235e-06,
+      "loss": 1.2334,
+      "step": 779
+    },
+    {
+      "epoch": 0.18964259664478483,
+      "grad_norm": 23.75,
+      "learning_rate": 2.457052172217684e-06,
+      "loss": 0.958,
+      "step": 780
+    },
+    {
+      "epoch": 0.1898857281789448,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4569264932323505e-06,
+      "loss": 1.0465,
+      "step": 781
+    },
+    {
+      "epoch": 0.19012885971310478,
+      "grad_norm": 20.875,
+      "learning_rate": 2.456800633851008e-06,
+      "loss": 1.2705,
+      "step": 782
+    },
+    {
+      "epoch": 0.19037199124726478,
+      "grad_norm": 21.75,
+      "learning_rate": 2.456674594092469e-06,
+      "loss": 0.8074,
+      "step": 783
+    },
+    {
+      "epoch": 0.19061512278142476,
+      "grad_norm": 18.75,
+      "learning_rate": 2.456548373975572e-06,
+      "loss": 1.0174,
+      "step": 784
+    },
+    {
+      "epoch": 0.19085825431558473,
+      "grad_norm": 23.375,
+      "learning_rate": 2.4564219735191824e-06,
+      "loss": 1.3718,
+      "step": 785
+    },
+    {
+      "epoch": 0.1911013858497447,
+      "grad_norm": 20.375,
+      "learning_rate": 2.4562953927421935e-06,
+      "loss": 1.0486,
+      "step": 786
+    },
+    {
+      "epoch": 0.1913445173839047,
+      "grad_norm": 19.625,
+      "learning_rate": 2.4561686316635246e-06,
+      "loss": 0.9703,
+      "step": 787
+    },
+    {
+      "epoch": 0.19158764891806468,
+      "grad_norm": 20.25,
+      "learning_rate": 2.4560416903021224e-06,
+      "loss": 0.9202,
+      "step": 788
+    },
+    {
+      "epoch": 0.19183078045222465,
+      "grad_norm": 17.25,
+      "learning_rate": 2.455914568676961e-06,
+      "loss": 0.6329,
+      "step": 789
+    },
+    {
+      "epoch": 0.19207391198638463,
+      "grad_norm": 18.5,
+      "learning_rate": 2.45578726680704e-06,
+      "loss": 0.811,
+      "step": 790
+    },
+    {
+      "epoch": 0.19231704352054463,
+      "grad_norm": 20.75,
+      "learning_rate": 2.4556597847113873e-06,
+      "loss": 1.3838,
+      "step": 791
+    },
+    {
+      "epoch": 0.1925601750547046,
+      "grad_norm": 31.875,
+      "learning_rate": 2.455532122409057e-06,
+      "loss": 1.3928,
+      "step": 792
+    },
+    {
+      "epoch": 0.19280330658886458,
+      "grad_norm": 20.375,
+      "learning_rate": 2.4554042799191313e-06,
+      "loss": 0.9493,
+      "step": 793
+    },
+    {
+      "epoch": 0.19304643812302455,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4552762572607174e-06,
+      "loss": 0.7572,
+      "step": 794
+    },
+    {
+      "epoch": 0.19328956965718452,
+      "grad_norm": 10.875,
+      "learning_rate": 2.4551480544529518e-06,
+      "loss": 0.3469,
+      "step": 795
+    },
+    {
+      "epoch": 0.19353270119134452,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.4550196715149953e-06,
+      "loss": 0.4945,
+      "step": 796
+    },
+    {
+      "epoch": 0.1937758327255045,
+      "grad_norm": 18.75,
+      "learning_rate": 2.4548911084660375e-06,
+      "loss": 0.8778,
+      "step": 797
+    },
+    {
+      "epoch": 0.19401896425966447,
+      "grad_norm": 23.625,
+      "learning_rate": 2.4547623653252945e-06,
+      "loss": 1.0358,
+      "step": 798
+    },
+    {
+      "epoch": 0.19426209579382445,
+      "grad_norm": 20.875,
+      "learning_rate": 2.454633442112009e-06,
+      "loss": 0.9956,
+      "step": 799
+    },
+    {
+      "epoch": 0.19450522732798445,
+      "grad_norm": 21.875,
+      "learning_rate": 2.4545043388454505e-06,
+      "loss": 0.6224,
+      "step": 800
+    },
+    {
+      "epoch": 0.19474835886214442,
+      "grad_norm": 31.25,
+      "learning_rate": 2.454375055544916e-06,
+      "loss": 1.0289,
+      "step": 801
+    },
+    {
+      "epoch": 0.1949914903963044,
+      "grad_norm": 23.25,
+      "learning_rate": 2.4542455922297297e-06,
+      "loss": 1.1676,
+      "step": 802
+    },
+    {
+      "epoch": 0.19523462193046437,
+      "grad_norm": 21.125,
+      "learning_rate": 2.4541159489192414e-06,
+      "loss": 0.8504,
+      "step": 803
+    },
+    {
+      "epoch": 0.19547775346462437,
+      "grad_norm": 19.5,
+      "learning_rate": 2.4539861256328286e-06,
+      "loss": 0.9696,
+      "step": 804
+    },
+    {
+      "epoch": 0.19572088499878434,
+      "grad_norm": 19.125,
+      "learning_rate": 2.453856122389896e-06,
+      "loss": 1.1432,
+      "step": 805
+    },
+    {
+      "epoch": 0.19596401653294432,
+      "grad_norm": 16.5,
+      "learning_rate": 2.4537259392098745e-06,
+      "loss": 0.5648,
+      "step": 806
+    },
+    {
+      "epoch": 0.1962071480671043,
+      "grad_norm": 26.875,
+      "learning_rate": 2.4535955761122223e-06,
+      "loss": 1.1229,
+      "step": 807
+    },
+    {
+      "epoch": 0.1964502796012643,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.4534650331164247e-06,
+      "loss": 1.0883,
+      "step": 808
+    },
+    {
+      "epoch": 0.19669341113542427,
+      "grad_norm": 19.875,
+      "learning_rate": 2.4533343102419927e-06,
+      "loss": 0.8562,
+      "step": 809
+    },
+    {
+      "epoch": 0.19693654266958424,
+      "grad_norm": 19.125,
+      "learning_rate": 2.453203407508466e-06,
+      "loss": 1.0313,
+      "step": 810
+    },
+    {
+      "epoch": 0.19717967420374422,
+      "grad_norm": 23.0,
+      "learning_rate": 2.4530723249354105e-06,
+      "loss": 0.605,
+      "step": 811
+    },
+    {
+      "epoch": 0.19742280573790422,
+      "grad_norm": 21.625,
+      "learning_rate": 2.452941062542418e-06,
+      "loss": 1.0167,
+      "step": 812
+    },
+    {
+      "epoch": 0.1976659372720642,
+      "grad_norm": 24.5,
+      "learning_rate": 2.4528096203491074e-06,
+      "loss": 1.066,
+      "step": 813
+    },
+    {
+      "epoch": 0.19790906880622416,
+      "grad_norm": 22.875,
+      "learning_rate": 2.4526779983751266e-06,
+      "loss": 0.9114,
+      "step": 814
+    },
+    {
+      "epoch": 0.19815220034038414,
+      "grad_norm": 20.75,
+      "learning_rate": 2.4525461966401482e-06,
+      "loss": 0.8261,
+      "step": 815
+    },
+    {
+      "epoch": 0.19839533187454414,
+      "grad_norm": 21.625,
+      "learning_rate": 2.4524142151638712e-06,
+      "loss": 0.6334,
+      "step": 816
+    },
+    {
+      "epoch": 0.1986384634087041,
+      "grad_norm": 24.0,
+      "learning_rate": 2.452282053966024e-06,
+      "loss": 1.1767,
+      "step": 817
+    },
+    {
+      "epoch": 0.1988815949428641,
+      "grad_norm": 17.5,
+      "learning_rate": 2.4521497130663595e-06,
+      "loss": 0.7888,
+      "step": 818
+    },
+    {
+      "epoch": 0.19912472647702406,
+      "grad_norm": 23.25,
+      "learning_rate": 2.452017192484659e-06,
+      "loss": 1.0081,
+      "step": 819
+    },
+    {
+      "epoch": 0.19936785801118406,
+      "grad_norm": 20.75,
+      "learning_rate": 2.4518844922407287e-06,
+      "loss": 1.1283,
+      "step": 820
+    },
+    {
+      "epoch": 0.19961098954534404,
+      "grad_norm": 23.125,
+      "learning_rate": 2.451751612354404e-06,
+      "loss": 0.7548,
+      "step": 821
+    },
+    {
+      "epoch": 0.199854121079504,
+      "grad_norm": 18.375,
+      "learning_rate": 2.451618552845546e-06,
+      "loss": 1.0112,
+      "step": 822
+    },
+    {
+      "epoch": 0.20009725261366398,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.4514853137340427e-06,
+      "loss": 0.5844,
+      "step": 823
+    },
+    {
+      "epoch": 0.20034038414782399,
+      "grad_norm": 22.875,
+      "learning_rate": 2.4513518950398085e-06,
+      "loss": 1.0063,
+      "step": 824
+    },
+    {
+      "epoch": 0.20058351568198396,
+      "grad_norm": 21.375,
+      "learning_rate": 2.451218296782786e-06,
+      "loss": 0.7965,
+      "step": 825
+    },
+    {
+      "epoch": 0.20082664721614393,
+      "grad_norm": 24.125,
+      "learning_rate": 2.451084518982943e-06,
+      "loss": 1.2595,
+      "step": 826
+    },
+    {
+      "epoch": 0.2010697787503039,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4509505616602753e-06,
+      "loss": 0.7854,
+      "step": 827
+    },
+    {
+      "epoch": 0.2013129102844639,
+      "grad_norm": 19.875,
+      "learning_rate": 2.450816424834805e-06,
+      "loss": 0.7807,
+      "step": 828
+    },
+    {
+      "epoch": 0.20155604181862388,
+      "grad_norm": 22.375,
+      "learning_rate": 2.4506821085265813e-06,
+      "loss": 1.1725,
+      "step": 829
+    },
+    {
+      "epoch": 0.20179917335278386,
+      "grad_norm": 16.625,
+      "learning_rate": 2.45054761275568e-06,
+      "loss": 0.8272,
+      "step": 830
+    },
+    {
+      "epoch": 0.20204230488694383,
+      "grad_norm": 16.0,
+      "learning_rate": 2.4504129375422037e-06,
+      "loss": 0.7311,
+      "step": 831
+    },
+    {
+      "epoch": 0.2022854364211038,
+      "grad_norm": 17.875,
+      "learning_rate": 2.450278082906282e-06,
+      "loss": 0.9647,
+      "step": 832
+    },
+    {
+      "epoch": 0.2025285679552638,
+      "grad_norm": 21.75,
+      "learning_rate": 2.450143048868071e-06,
+      "loss": 0.802,
+      "step": 833
+    },
+    {
+      "epoch": 0.20277169948942378,
+      "grad_norm": 18.625,
+      "learning_rate": 2.4500078354477547e-06,
+      "loss": 0.8647,
+      "step": 834
+    },
+    {
+      "epoch": 0.20301483102358375,
+      "grad_norm": 18.75,
+      "learning_rate": 2.4498724426655424e-06,
+      "loss": 0.7957,
+      "step": 835
+    },
+    {
+      "epoch": 0.20325796255774373,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.449736870541671e-06,
+      "loss": 0.4855,
+      "step": 836
+    },
+    {
+      "epoch": 0.20350109409190373,
+      "grad_norm": 13.6875,
+      "learning_rate": 2.4496011190964044e-06,
+      "loss": 0.4245,
+      "step": 837
+    },
+    {
+      "epoch": 0.2037442256260637,
+      "grad_norm": 24.375,
+      "learning_rate": 2.449465188350032e-06,
+      "loss": 1.2063,
+      "step": 838
+    },
+    {
+      "epoch": 0.20398735716022368,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.4493290783228723e-06,
+      "loss": 0.5422,
+      "step": 839
+    },
+    {
+      "epoch": 0.20423048869438365,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4491927890352685e-06,
+      "loss": 0.9839,
+      "step": 840
+    },
+    {
+      "epoch": 0.20447362022854365,
+      "grad_norm": 17.5,
+      "learning_rate": 2.4490563205075916e-06,
+      "loss": 0.7722,
+      "step": 841
+    },
+    {
+      "epoch": 0.20471675176270362,
+      "grad_norm": 25.25,
+      "learning_rate": 2.448919672760239e-06,
+      "loss": 0.6736,
+      "step": 842
+    },
+    {
+      "epoch": 0.2049598832968636,
+      "grad_norm": 17.5,
+      "learning_rate": 2.4487828458136354e-06,
+      "loss": 0.6458,
+      "step": 843
+    },
+    {
+      "epoch": 0.20520301483102357,
+      "grad_norm": 20.25,
+      "learning_rate": 2.4486458396882317e-06,
+      "loss": 0.89,
+      "step": 844
+    },
+    {
+      "epoch": 0.20544614636518357,
+      "grad_norm": 25.25,
+      "learning_rate": 2.4485086544045063e-06,
+      "loss": 1.1589,
+      "step": 845
+    },
+    {
+      "epoch": 0.20568927789934355,
+      "grad_norm": 18.375,
+      "learning_rate": 2.4483712899829636e-06,
+      "loss": 0.9451,
+      "step": 846
+    },
+    {
+      "epoch": 0.20593240943350352,
+      "grad_norm": 19.125,
+      "learning_rate": 2.448233746444135e-06,
+      "loss": 1.145,
+      "step": 847
+    },
+    {
+      "epoch": 0.2061755409676635,
+      "grad_norm": 22.125,
+      "learning_rate": 2.448096023808578e-06,
+      "loss": 0.7977,
+      "step": 848
+    },
+    {
+      "epoch": 0.2064186725018235,
+      "grad_norm": 19.375,
+      "learning_rate": 2.447958122096879e-06,
+      "loss": 0.8558,
+      "step": 849
+    },
+    {
+      "epoch": 0.20666180403598347,
+      "grad_norm": 20.875,
+      "learning_rate": 2.4478200413296494e-06,
+      "loss": 1.0523,
+      "step": 850
+    },
+    {
+      "epoch": 0.20690493557014344,
+      "grad_norm": 22.0,
+      "learning_rate": 2.447681781527527e-06,
+      "loss": 0.955,
+      "step": 851
+    },
+    {
+      "epoch": 0.20714806710430342,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.447543342711178e-06,
+      "loss": 0.5719,
+      "step": 852
+    },
+    {
+      "epoch": 0.20739119863846342,
+      "grad_norm": 23.0,
+      "learning_rate": 2.447404724901294e-06,
+      "loss": 0.7843,
+      "step": 853
+    },
+    {
+      "epoch": 0.2076343301726234,
+      "grad_norm": 21.625,
+      "learning_rate": 2.447265928118594e-06,
+      "loss": 1.3464,
+      "step": 854
+    },
+    {
+      "epoch": 0.20787746170678337,
+      "grad_norm": 18.5,
+      "learning_rate": 2.447126952383824e-06,
+      "loss": 0.7989,
+      "step": 855
+    },
+    {
+      "epoch": 0.20812059324094334,
+      "grad_norm": 28.875,
+      "learning_rate": 2.446987797717755e-06,
+      "loss": 0.5581,
+      "step": 856
+    },
+    {
+      "epoch": 0.20836372477510334,
+      "grad_norm": 15.1875,
+      "learning_rate": 2.4468484641411877e-06,
+      "loss": 1.1626,
+      "step": 857
+    },
+    {
+      "epoch": 0.20860685630926332,
+      "grad_norm": 17.75,
+      "learning_rate": 2.446708951674947e-06,
+      "loss": 1.2429,
+      "step": 858
+    },
+    {
+      "epoch": 0.2088499878434233,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4465692603398854e-06,
+      "loss": 1.0368,
+      "step": 859
+    },
+    {
+      "epoch": 0.20909311937758326,
+      "grad_norm": 23.25,
+      "learning_rate": 2.4464293901568824e-06,
+      "loss": 1.3265,
+      "step": 860
+    },
+    {
+      "epoch": 0.20933625091174327,
+      "grad_norm": 22.25,
+      "learning_rate": 2.446289341146844e-06,
+      "loss": 0.842,
+      "step": 861
+    },
+    {
+      "epoch": 0.20957938244590324,
+      "grad_norm": 22.75,
+      "learning_rate": 2.446149113330703e-06,
+      "loss": 1.1281,
+      "step": 862
+    },
+    {
+      "epoch": 0.2098225139800632,
+      "grad_norm": 20.875,
+      "learning_rate": 2.4460087067294186e-06,
+      "loss": 1.2158,
+      "step": 863
+    },
+    {
+      "epoch": 0.2100656455142232,
+      "grad_norm": 15.875,
+      "learning_rate": 2.4458681213639773e-06,
+      "loss": 0.6732,
+      "step": 864
+    },
+    {
+      "epoch": 0.2103087770483832,
+      "grad_norm": 24.25,
+      "learning_rate": 2.445727357255392e-06,
+      "loss": 1.3589,
+      "step": 865
+    },
+    {
+      "epoch": 0.21055190858254316,
+      "grad_norm": 18.625,
+      "learning_rate": 2.4455864144247023e-06,
+      "loss": 1.2187,
+      "step": 866
+    },
+    {
+      "epoch": 0.21079504011670314,
+      "grad_norm": 26.625,
+      "learning_rate": 2.4454452928929746e-06,
+      "loss": 1.1716,
+      "step": 867
+    },
+    {
+      "epoch": 0.2110381716508631,
+      "grad_norm": 22.125,
+      "learning_rate": 2.4453039926813014e-06,
+      "loss": 0.9565,
+      "step": 868
+    },
+    {
+      "epoch": 0.21128130318502308,
+      "grad_norm": 22.0,
+      "learning_rate": 2.445162513810803e-06,
+      "loss": 0.9566,
+      "step": 869
+    },
+    {
+      "epoch": 0.21152443471918309,
+      "grad_norm": 22.375,
+      "learning_rate": 2.445020856302626e-06,
+      "loss": 1.2194,
+      "step": 870
+    },
+    {
+      "epoch": 0.21176756625334306,
+      "grad_norm": 22.625,
+      "learning_rate": 2.4448790201779428e-06,
+      "loss": 1.1004,
+      "step": 871
+    },
+    {
+      "epoch": 0.21201069778750303,
+      "grad_norm": 32.5,
+      "learning_rate": 2.4447370054579542e-06,
+      "loss": 1.3613,
+      "step": 872
+    },
+    {
+      "epoch": 0.212253829321663,
+      "grad_norm": 16.875,
+      "learning_rate": 2.444594812163886e-06,
+      "loss": 0.9028,
+      "step": 873
+    },
+    {
+      "epoch": 0.212496960855823,
+      "grad_norm": 19.5,
+      "learning_rate": 2.4444524403169922e-06,
+      "loss": 1.2903,
+      "step": 874
+    },
+    {
+      "epoch": 0.21274009238998298,
+      "grad_norm": 18.875,
+      "learning_rate": 2.444309889938552e-06,
+      "loss": 1.2591,
+      "step": 875
+    },
+    {
+      "epoch": 0.21298322392414296,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.4441671610498725e-06,
+      "loss": 0.7273,
+      "step": 876
+    },
+    {
+      "epoch": 0.21322635545830293,
+      "grad_norm": 22.75,
+      "learning_rate": 2.4440242536722863e-06,
+      "loss": 1.3297,
+      "step": 877
+    },
+    {
+      "epoch": 0.21346948699246293,
+      "grad_norm": 21.75,
+      "learning_rate": 2.4438811678271543e-06,
+      "loss": 0.8035,
+      "step": 878
+    },
+    {
+      "epoch": 0.2137126185266229,
+      "grad_norm": 34.0,
+      "learning_rate": 2.4437379035358626e-06,
+      "loss": 1.1964,
+      "step": 879
+    },
+    {
+      "epoch": 0.21395575006078288,
+      "grad_norm": 29.75,
+      "learning_rate": 2.4435944608198246e-06,
+      "loss": 0.9425,
+      "step": 880
+    },
+    {
+      "epoch": 0.21419888159494285,
+      "grad_norm": 21.0,
+      "learning_rate": 2.4434508397004806e-06,
+      "loss": 1.0503,
+      "step": 881
+    },
+    {
+      "epoch": 0.21444201312910285,
+      "grad_norm": 26.75,
+      "learning_rate": 2.4433070401992968e-06,
+      "loss": 1.116,
+      "step": 882
+    },
+    {
+      "epoch": 0.21468514466326283,
+      "grad_norm": 23.625,
+      "learning_rate": 2.4431630623377665e-06,
+      "loss": 1.0084,
+      "step": 883
+    },
+    {
+      "epoch": 0.2149282761974228,
+      "grad_norm": 18.0,
+      "learning_rate": 2.4430189061374103e-06,
+      "loss": 1.1068,
+      "step": 884
+    },
+    {
+      "epoch": 0.21517140773158278,
+      "grad_norm": 20.25,
+      "learning_rate": 2.4428745716197746e-06,
+      "loss": 0.8787,
+      "step": 885
+    },
+    {
+      "epoch": 0.21541453926574278,
+      "grad_norm": 19.875,
+      "learning_rate": 2.4427300588064316e-06,
+      "loss": 0.9147,
+      "step": 886
+    },
+    {
+      "epoch": 0.21565767079990275,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.4425853677189833e-06,
+      "loss": 0.8199,
+      "step": 887
+    },
+    {
+      "epoch": 0.21590080233406272,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4424404983790547e-06,
+      "loss": 0.9097,
+      "step": 888
+    },
+    {
+      "epoch": 0.2161439338682227,
+      "grad_norm": 22.375,
+      "learning_rate": 2.442295450808299e-06,
+      "loss": 0.9711,
+      "step": 889
+    },
+    {
+      "epoch": 0.2163870654023827,
+      "grad_norm": 19.375,
+      "learning_rate": 2.4421502250283966e-06,
+      "loss": 0.7945,
+      "step": 890
+    },
+    {
+      "epoch": 0.21663019693654267,
+      "grad_norm": 20.75,
+      "learning_rate": 2.4420048210610542e-06,
+      "loss": 0.9914,
+      "step": 891
+    },
+    {
+      "epoch": 0.21687332847070265,
+      "grad_norm": 19.0,
+      "learning_rate": 2.441859238928005e-06,
+      "loss": 0.9432,
+      "step": 892
+    },
+    {
+      "epoch": 0.21711646000486262,
+      "grad_norm": 21.0,
+      "learning_rate": 2.4417134786510077e-06,
+      "loss": 0.8256,
+      "step": 893
+    },
+    {
+      "epoch": 0.21735959153902262,
+      "grad_norm": 19.75,
+      "learning_rate": 2.44156754025185e-06,
+      "loss": 0.6091,
+      "step": 894
+    },
+    {
+      "epoch": 0.2176027230731826,
+      "grad_norm": 27.375,
+      "learning_rate": 2.441421423752344e-06,
+      "loss": 1.0419,
+      "step": 895
+    },
+    {
+      "epoch": 0.21784585460734257,
+      "grad_norm": 20.875,
+      "learning_rate": 2.4412751291743297e-06,
+      "loss": 1.296,
+      "step": 896
+    },
+    {
+      "epoch": 0.21808898614150254,
+      "grad_norm": 23.625,
+      "learning_rate": 2.4411286565396735e-06,
+      "loss": 0.8414,
+      "step": 897
+    },
+    {
+      "epoch": 0.21833211767566255,
+      "grad_norm": 17.375,
+      "learning_rate": 2.4409820058702678e-06,
+      "loss": 0.9116,
+      "step": 898
+    },
+    {
+      "epoch": 0.21857524920982252,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4408351771880324e-06,
+      "loss": 0.8075,
+      "step": 899
+    },
+    {
+      "epoch": 0.2188183807439825,
+      "grad_norm": 16.75,
+      "learning_rate": 2.4406881705149133e-06,
+      "loss": 0.9483,
+      "step": 900
+    },
+    {
+      "epoch": 0.21906151227814247,
+      "grad_norm": 28.75,
+      "learning_rate": 2.4405409858728836e-06,
+      "loss": 1.284,
+      "step": 901
+    },
+    {
+      "epoch": 0.21930464381230247,
+      "grad_norm": 22.875,
+      "learning_rate": 2.4403936232839418e-06,
+      "loss": 0.869,
+      "step": 902
+    },
+    {
+      "epoch": 0.21954777534646244,
+      "grad_norm": 18.375,
+      "learning_rate": 2.440246082770114e-06,
+      "loss": 1.0692,
+      "step": 903
+    },
+    {
+      "epoch": 0.21979090688062242,
+      "grad_norm": 25.125,
+      "learning_rate": 2.440098364353454e-06,
+      "loss": 0.92,
+      "step": 904
+    },
+    {
+      "epoch": 0.2200340384147824,
+      "grad_norm": 19.125,
+      "learning_rate": 2.4399504680560387e-06,
+      "loss": 1.0571,
+      "step": 905
+    },
+    {
+      "epoch": 0.22027716994894236,
+      "grad_norm": 23.875,
+      "learning_rate": 2.439802393899975e-06,
+      "loss": 1.2083,
+      "step": 906
+    },
+    {
+      "epoch": 0.22052030148310237,
+      "grad_norm": 23.5,
+      "learning_rate": 2.4396541419073947e-06,
+      "loss": 1.2092,
+      "step": 907
+    },
+    {
+      "epoch": 0.22076343301726234,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4395057121004573e-06,
+      "loss": 1.0874,
+      "step": 908
+    },
+    {
+      "epoch": 0.2210065645514223,
+      "grad_norm": 18.125,
+      "learning_rate": 2.4393571045013475e-06,
+      "loss": 0.6744,
+      "step": 909
+    },
+    {
+      "epoch": 0.2212496960855823,
+      "grad_norm": 22.375,
+      "learning_rate": 2.4392083191322774e-06,
+      "loss": 0.9102,
+      "step": 910
+    },
+    {
+      "epoch": 0.2214928276197423,
+      "grad_norm": 16.875,
+      "learning_rate": 2.439059356015486e-06,
+      "loss": 0.7533,
+      "step": 911
+    },
+    {
+      "epoch": 0.22173595915390226,
+      "grad_norm": 17.5,
+      "learning_rate": 2.438910215173238e-06,
+      "loss": 0.7811,
+      "step": 912
+    },
+    {
+      "epoch": 0.22197909068806224,
+      "grad_norm": 19.5,
+      "learning_rate": 2.438760896627825e-06,
+      "loss": 1.1445,
+      "step": 913
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4386114004015653e-06,
+      "loss": 1.0945,
+      "step": 914
+    },
+    {
+      "epoch": 0.2224653537563822,
+      "grad_norm": 23.625,
+      "learning_rate": 2.4384617265168043e-06,
+      "loss": 1.0269,
+      "step": 915
+    },
+    {
+      "epoch": 0.22270848529054219,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4383118749959122e-06,
+      "loss": 0.8623,
+      "step": 916
+    },
+    {
+      "epoch": 0.22295161682470216,
+      "grad_norm": 20.25,
+      "learning_rate": 2.438161845861288e-06,
+      "loss": 1.0068,
+      "step": 917
+    },
+    {
+      "epoch": 0.22319474835886213,
+      "grad_norm": 23.375,
+      "learning_rate": 2.438011639135355e-06,
+      "loss": 1.1542,
+      "step": 918
+    },
+    {
+      "epoch": 0.22343787989302213,
+      "grad_norm": 21.125,
+      "learning_rate": 2.4378612548405657e-06,
+      "loss": 0.9252,
+      "step": 919
+    },
+    {
+      "epoch": 0.2236810114271821,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4377106929993966e-06,
+      "loss": 0.5471,
+      "step": 920
+    },
+    {
+      "epoch": 0.22392414296134208,
+      "grad_norm": 16.625,
+      "learning_rate": 2.4375599536343515e-06,
+      "loss": 1.4131,
+      "step": 921
+    },
+    {
+      "epoch": 0.22416727449550206,
+      "grad_norm": 19.625,
+      "learning_rate": 2.437409036767962e-06,
+      "loss": 0.686,
+      "step": 922
+    },
+    {
+      "epoch": 0.22441040602966206,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4372579424227843e-06,
+      "loss": 1.0637,
+      "step": 923
+    },
+    {
+      "epoch": 0.22465353756382203,
+      "grad_norm": 24.375,
+      "learning_rate": 2.4371066706214026e-06,
+      "loss": 1.1009,
+      "step": 924
+    },
+    {
+      "epoch": 0.224896669097982,
+      "grad_norm": 23.25,
+      "learning_rate": 2.436955221386427e-06,
+      "loss": 1.106,
+      "step": 925
+    },
+    {
+      "epoch": 0.22513980063214198,
+      "grad_norm": 20.375,
+      "learning_rate": 2.436803594740494e-06,
+      "loss": 1.0571,
+      "step": 926
+    },
+    {
+      "epoch": 0.22538293216630198,
+      "grad_norm": 35.0,
+      "learning_rate": 2.436651790706267e-06,
+      "loss": 0.4582,
+      "step": 927
+    },
+    {
+      "epoch": 0.22562606370046195,
+      "grad_norm": 32.5,
+      "learning_rate": 2.4364998093064357e-06,
+      "loss": 0.9672,
+      "step": 928
+    },
+    {
+      "epoch": 0.22586919523462193,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.4363476505637162e-06,
+      "loss": 0.7489,
+      "step": 929
+    },
+    {
+      "epoch": 0.2261123267687819,
+      "grad_norm": 20.625,
+      "learning_rate": 2.4361953145008517e-06,
+      "loss": 0.8131,
+      "step": 930
+    },
+    {
+      "epoch": 0.2263554583029419,
+      "grad_norm": 15.4375,
+      "learning_rate": 2.436042801140611e-06,
+      "loss": 0.4668,
+      "step": 931
+    },
+    {
+      "epoch": 0.22659858983710188,
+      "grad_norm": 22.625,
+      "learning_rate": 2.4358901105057902e-06,
+      "loss": 0.7394,
+      "step": 932
+    },
+    {
+      "epoch": 0.22684172137126185,
+      "grad_norm": 26.75,
+      "learning_rate": 2.435737242619211e-06,
+      "loss": 0.9973,
+      "step": 933
+    },
+    {
+      "epoch": 0.22708485290542182,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4355841975037226e-06,
+      "loss": 0.9461,
+      "step": 934
+    },
+    {
+      "epoch": 0.22732798443958183,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4354309751822004e-06,
+      "loss": 0.6809,
+      "step": 935
+    },
+    {
+      "epoch": 0.2275711159737418,
+      "grad_norm": 20.0,
+      "learning_rate": 2.4352775756775453e-06,
+      "loss": 0.8722,
+      "step": 936
+    },
+    {
+      "epoch": 0.22781424750790177,
+      "grad_norm": 26.375,
+      "learning_rate": 2.435123999012687e-06,
+      "loss": 0.9036,
+      "step": 937
+    },
+    {
+      "epoch": 0.22805737904206175,
+      "grad_norm": 17.5,
+      "learning_rate": 2.4349702452105783e-06,
+      "loss": 0.7738,
+      "step": 938
+    },
+    {
+      "epoch": 0.22830051057622175,
+      "grad_norm": 18.25,
+      "learning_rate": 2.4348163142942017e-06,
+      "loss": 1.0418,
+      "step": 939
+    },
+    {
+      "epoch": 0.22854364211038172,
+      "grad_norm": 18.125,
+      "learning_rate": 2.4346622062865645e-06,
+      "loss": 0.8831,
+      "step": 940
+    },
+    {
+      "epoch": 0.2287867736445417,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4345079212107003e-06,
+      "loss": 0.9789,
+      "step": 941
+    },
+    {
+      "epoch": 0.22902990517870167,
+      "grad_norm": 18.25,
+      "learning_rate": 2.4343534590896705e-06,
+      "loss": 0.9599,
+      "step": 942
+    },
+    {
+      "epoch": 0.22927303671286167,
+      "grad_norm": 21.25,
+      "learning_rate": 2.434198819946562e-06,
+      "loss": 1.1084,
+      "step": 943
+    },
+    {
+      "epoch": 0.22951616824702165,
+      "grad_norm": 18.625,
+      "learning_rate": 2.4340440038044877e-06,
+      "loss": 0.8981,
+      "step": 944
+    },
+    {
+      "epoch": 0.22975929978118162,
+      "grad_norm": 19.875,
+      "learning_rate": 2.433889010686588e-06,
+      "loss": 0.842,
+      "step": 945
+    },
+    {
+      "epoch": 0.2300024313153416,
+      "grad_norm": 22.625,
+      "learning_rate": 2.433733840616029e-06,
+      "loss": 0.9982,
+      "step": 946
+    },
+    {
+      "epoch": 0.23024556284950157,
+      "grad_norm": 21.25,
+      "learning_rate": 2.433578493616004e-06,
+      "loss": 0.9833,
+      "step": 947
+    },
+    {
+      "epoch": 0.23048869438366157,
+      "grad_norm": 26.25,
+      "learning_rate": 2.4334229697097315e-06,
+      "loss": 1.354,
+      "step": 948
+    },
+    {
+      "epoch": 0.23073182591782154,
+      "grad_norm": 18.625,
+      "learning_rate": 2.4332672689204583e-06,
+      "loss": 0.9825,
+      "step": 949
+    },
+    {
+      "epoch": 0.23097495745198152,
+      "grad_norm": 19.875,
+      "learning_rate": 2.433111391271456e-06,
+      "loss": 0.9362,
+      "step": 950
+    },
+    {
+      "epoch": 0.2312180889861415,
+      "grad_norm": 16.875,
+      "learning_rate": 2.432955336786023e-06,
+      "loss": 0.7623,
+      "step": 951
+    },
+    {
+      "epoch": 0.2314612205203015,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4327991054874843e-06,
+      "loss": 0.8751,
+      "step": 952
+    },
+    {
+      "epoch": 0.23170435205446147,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.4326426973991922e-06,
+      "loss": 0.4809,
+      "step": 953
+    },
+    {
+      "epoch": 0.23194748358862144,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.4324861125445236e-06,
+      "loss": 0.7759,
+      "step": 954
+    },
+    {
+      "epoch": 0.2321906151227814,
+      "grad_norm": 19.625,
+      "learning_rate": 2.4323293509468837e-06,
+      "loss": 0.7713,
+      "step": 955
+    },
+    {
+      "epoch": 0.23243374665694141,
+      "grad_norm": 18.375,
+      "learning_rate": 2.4321724126297026e-06,
+      "loss": 1.0325,
+      "step": 956
+    },
+    {
+      "epoch": 0.2326768781911014,
+      "grad_norm": 20.5,
+      "learning_rate": 2.432015297616437e-06,
+      "loss": 0.8978,
+      "step": 957
+    },
+    {
+      "epoch": 0.23292000972526136,
+      "grad_norm": 23.75,
+      "learning_rate": 2.431858005930572e-06,
+      "loss": 1.1904,
+      "step": 958
+    },
+    {
+      "epoch": 0.23316314125942134,
+      "grad_norm": 23.75,
+      "learning_rate": 2.4317005375956163e-06,
+      "loss": 1.2113,
+      "step": 959
+    },
+    {
+      "epoch": 0.23340627279358134,
+      "grad_norm": 20.25,
+      "learning_rate": 2.4315428926351067e-06,
+      "loss": 0.8656,
+      "step": 960
+    },
+    {
+      "epoch": 0.2336494043277413,
+      "grad_norm": 19.5,
+      "learning_rate": 2.4313850710726054e-06,
+      "loss": 1.0967,
+      "step": 961
+    },
+    {
+      "epoch": 0.23389253586190129,
+      "grad_norm": 22.5,
+      "learning_rate": 2.4312270729317024e-06,
+      "loss": 0.7397,
+      "step": 962
+    },
+    {
+      "epoch": 0.23413566739606126,
+      "grad_norm": 29.25,
+      "learning_rate": 2.4310688982360125e-06,
+      "loss": 1.0042,
+      "step": 963
+    },
+    {
+      "epoch": 0.23437879893022126,
+      "grad_norm": 24.875,
+      "learning_rate": 2.430910547009178e-06,
+      "loss": 1.1196,
+      "step": 964
+    },
+    {
+      "epoch": 0.23462193046438123,
+      "grad_norm": 22.125,
+      "learning_rate": 2.4307520192748675e-06,
+      "loss": 0.9531,
+      "step": 965
+    },
+    {
+      "epoch": 0.2348650619985412,
+      "grad_norm": 20.125,
+      "learning_rate": 2.430593315056776e-06,
+      "loss": 0.7087,
+      "step": 966
+    },
+    {
+      "epoch": 0.23510819353270118,
+      "grad_norm": 17.0,
+      "learning_rate": 2.430434434378623e-06,
+      "loss": 0.7385,
+      "step": 967
+    },
+    {
+      "epoch": 0.23535132506686118,
+      "grad_norm": 23.5,
+      "learning_rate": 2.430275377264157e-06,
+      "loss": 1.1282,
+      "step": 968
+    },
+    {
+      "epoch": 0.23559445660102116,
+      "grad_norm": 17.0,
+      "learning_rate": 2.4301161437371525e-06,
+      "loss": 0.7602,
+      "step": 969
+    },
+    {
+      "epoch": 0.23583758813518113,
+      "grad_norm": 22.75,
+      "learning_rate": 2.4299567338214086e-06,
+      "loss": 1.0104,
+      "step": 970
+    },
+    {
+      "epoch": 0.2360807196693411,
+      "grad_norm": 17.375,
+      "learning_rate": 2.429797147540752e-06,
+      "loss": 0.7066,
+      "step": 971
+    },
+    {
+      "epoch": 0.2363238512035011,
+      "grad_norm": 35.25,
+      "learning_rate": 2.4296373849190363e-06,
+      "loss": 0.8827,
+      "step": 972
+    },
+    {
+      "epoch": 0.23656698273766108,
+      "grad_norm": 21.5,
+      "learning_rate": 2.42947744598014e-06,
+      "loss": 0.9952,
+      "step": 973
+    },
+    {
+      "epoch": 0.23681011427182105,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4293173307479696e-06,
+      "loss": 0.9081,
+      "step": 974
+    },
+    {
+      "epoch": 0.23705324580598103,
+      "grad_norm": 22.125,
+      "learning_rate": 2.4291570392464566e-06,
+      "loss": 1.469,
+      "step": 975
+    },
+    {
+      "epoch": 0.23729637734014103,
+      "grad_norm": 23.875,
+      "learning_rate": 2.4289965714995588e-06,
+      "loss": 1.0484,
+      "step": 976
+    },
+    {
+      "epoch": 0.237539508874301,
+      "grad_norm": 19.0,
+      "learning_rate": 2.428835927531262e-06,
+      "loss": 1.0291,
+      "step": 977
+    },
+    {
+      "epoch": 0.23778264040846098,
+      "grad_norm": 19.25,
+      "learning_rate": 2.428675107365576e-06,
+      "loss": 1.0113,
+      "step": 978
+    },
+    {
+      "epoch": 0.23802577194262095,
+      "grad_norm": 23.0,
+      "learning_rate": 2.42851411102654e-06,
+      "loss": 1.4051,
+      "step": 979
+    },
+    {
+      "epoch": 0.23826890347678095,
+      "grad_norm": 21.125,
+      "learning_rate": 2.4283529385382154e-06,
+      "loss": 1.0861,
+      "step": 980
+    },
+    {
+      "epoch": 0.23851203501094093,
+      "grad_norm": 12.4375,
+      "learning_rate": 2.4281915899246934e-06,
+      "loss": 0.6562,
+      "step": 981
+    },
+    {
+      "epoch": 0.2387551665451009,
+      "grad_norm": 20.5,
+      "learning_rate": 2.4280300652100904e-06,
+      "loss": 1.0857,
+      "step": 982
+    },
+    {
+      "epoch": 0.23899829807926087,
+      "grad_norm": 19.25,
+      "learning_rate": 2.4278683644185487e-06,
+      "loss": 0.8665,
+      "step": 983
+    },
+    {
+      "epoch": 0.23924142961342085,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4277064875742375e-06,
+      "loss": 0.7677,
+      "step": 984
+    },
+    {
+      "epoch": 0.23948456114758085,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4275444347013523e-06,
+      "loss": 0.9632,
+      "step": 985
+    },
+    {
+      "epoch": 0.23972769268174082,
+      "grad_norm": 21.0,
+      "learning_rate": 2.427382205824114e-06,
+      "loss": 0.9758,
+      "step": 986
+    },
+    {
+      "epoch": 0.2399708242159008,
+      "grad_norm": 20.25,
+      "learning_rate": 2.427219800966771e-06,
+      "loss": 0.8743,
+      "step": 987
+    },
+    {
+      "epoch": 0.24021395575006077,
+      "grad_norm": 23.0,
+      "learning_rate": 2.427057220153598e-06,
+      "loss": 1.2395,
+      "step": 988
+    },
+    {
+      "epoch": 0.24045708728422077,
+      "grad_norm": 21.125,
+      "learning_rate": 2.426894463408894e-06,
+      "loss": 1.0091,
+      "step": 989
+    },
+    {
+      "epoch": 0.24070021881838075,
+      "grad_norm": 17.125,
+      "learning_rate": 2.4267315307569876e-06,
+      "loss": 0.8063,
+      "step": 990
+    },
+    {
+      "epoch": 0.24094335035254072,
+      "grad_norm": 14.0,
+      "learning_rate": 2.4265684222222307e-06,
+      "loss": 0.4877,
+      "step": 991
+    },
+    {
+      "epoch": 0.2411864818867007,
+      "grad_norm": 34.75,
+      "learning_rate": 2.426405137829003e-06,
+      "loss": 1.0394,
+      "step": 992
+    },
+    {
+      "epoch": 0.2414296134208607,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4262416776017107e-06,
+      "loss": 1.3841,
+      "step": 993
+    },
+    {
+      "epoch": 0.24167274495502067,
+      "grad_norm": 17.25,
+      "learning_rate": 2.426078041564785e-06,
+      "loss": 0.8635,
+      "step": 994
+    },
+    {
+      "epoch": 0.24191587648918064,
+      "grad_norm": 19.25,
+      "learning_rate": 2.4259142297426846e-06,
+      "loss": 0.9947,
+      "step": 995
+    },
+    {
+      "epoch": 0.24215900802334062,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4257502421598934e-06,
+      "loss": 0.9596,
+      "step": 996
+    },
+    {
+      "epoch": 0.24240213955750062,
+      "grad_norm": 23.0,
+      "learning_rate": 2.425586078840923e-06,
+      "loss": 1.1512,
+      "step": 997
+    },
+    {
+      "epoch": 0.2426452710916606,
+      "grad_norm": 17.875,
+      "learning_rate": 2.42542173981031e-06,
+      "loss": 0.9097,
+      "step": 998
+    },
+    {
+      "epoch": 0.24288840262582057,
+      "grad_norm": 16.0,
+      "learning_rate": 2.4252572250926176e-06,
+      "loss": 0.7223,
+      "step": 999
+    },
+    {
+      "epoch": 0.24313153415998054,
+      "grad_norm": 14.375,
+      "learning_rate": 2.4250925347124353e-06,
+      "loss": 0.5323,
+      "step": 1000
+    },
+    {
+      "epoch": 0.24337466569414054,
+      "grad_norm": 24.875,
+      "learning_rate": 2.4249276686943797e-06,
+      "loss": 1.383,
+      "step": 1001
+    },
+    {
+      "epoch": 0.24361779722830051,
+      "grad_norm": 31.25,
+      "learning_rate": 2.424762627063092e-06,
+      "loss": 0.877,
+      "step": 1002
+    },
+    {
+      "epoch": 0.2438609287624605,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4245974098432406e-06,
+      "loss": 1.054,
+      "step": 1003
+    },
+    {
+      "epoch": 0.24410406029662046,
+      "grad_norm": 24.375,
+      "learning_rate": 2.4244320170595206e-06,
+      "loss": 1.2724,
+      "step": 1004
+    },
+    {
+      "epoch": 0.24434719183078046,
+      "grad_norm": 18.75,
+      "learning_rate": 2.4242664487366523e-06,
+      "loss": 1.0731,
+      "step": 1005
+    },
+    {
+      "epoch": 0.24459032336494044,
+      "grad_norm": 20.25,
+      "learning_rate": 2.424100704899383e-06,
+      "loss": 1.1288,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2448334548991004,
+      "grad_norm": 19.25,
+      "learning_rate": 2.4239347855724863e-06,
+      "loss": 1.3863,
+      "step": 1007
+    },
+    {
+      "epoch": 0.24507658643326038,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4237686907807612e-06,
+      "loss": 0.8535,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2453197179674204,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4236024205490335e-06,
+      "loss": 1.1126,
+      "step": 1009
+    },
+    {
+      "epoch": 0.24556284950158036,
+      "grad_norm": 16.75,
+      "learning_rate": 2.423435974902155e-06,
+      "loss": 0.9586,
+      "step": 1010
+    },
+    {
+      "epoch": 0.24580598103574033,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4232693538650043e-06,
+      "loss": 0.7974,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2460491125699003,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4231025574624855e-06,
+      "loss": 1.0039,
+      "step": 1012
+    },
+    {
+      "epoch": 0.2462922441040603,
+      "grad_norm": 14.625,
+      "learning_rate": 2.42293558571953e-06,
+      "loss": 0.5908,
+      "step": 1013
+    },
+    {
+      "epoch": 0.24653537563822028,
+      "grad_norm": 25.75,
+      "learning_rate": 2.4227684386610927e-06,
+      "loss": 1.2829,
+      "step": 1014
+    },
+    {
+      "epoch": 0.24677850717238026,
+      "grad_norm": 14.875,
+      "learning_rate": 2.422601116312159e-06,
+      "loss": 0.5478,
+      "step": 1015
+    },
+    {
+      "epoch": 0.24702163870654023,
+      "grad_norm": 22.0,
+      "learning_rate": 2.422433618697736e-06,
+      "loss": 1.4427,
+      "step": 1016
+    },
+    {
+      "epoch": 0.24726477024070023,
+      "grad_norm": 21.125,
+      "learning_rate": 2.4222659458428606e-06,
+      "loss": 0.9482,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2475079017748602,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.4220980977725934e-06,
+      "loss": 0.6992,
+      "step": 1018
+    },
+    {
+      "epoch": 0.24775103330902018,
+      "grad_norm": 26.25,
+      "learning_rate": 2.421930074512023e-06,
+      "loss": 1.6438,
+      "step": 1019
+    },
+    {
+      "epoch": 0.24799416484318015,
+      "grad_norm": 23.5,
+      "learning_rate": 2.421761876086263e-06,
+      "loss": 1.2188,
+      "step": 1020
+    },
+    {
+      "epoch": 0.24823729637734013,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.4215935025204536e-06,
+      "loss": 0.4709,
+      "step": 1021
+    },
+    {
+      "epoch": 0.24848042791150013,
+      "grad_norm": 19.125,
+      "learning_rate": 2.421424953839761e-06,
+      "loss": 1.0291,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2487235594456601,
+      "grad_norm": 18.25,
+      "learning_rate": 2.421256230069378e-06,
+      "loss": 1.1222,
+      "step": 1023
+    },
+    {
+      "epoch": 0.24896669097982008,
+      "grad_norm": 18.375,
+      "learning_rate": 2.421087331234523e-06,
+      "loss": 1.1241,
+      "step": 1024
+    },
+    {
+      "epoch": 0.24920982251398005,
+      "grad_norm": 20.625,
+      "learning_rate": 2.4209182573604414e-06,
+      "loss": 1.2151,
+      "step": 1025
+    },
+    {
+      "epoch": 0.24945295404814005,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4207490084724033e-06,
+      "loss": 1.3846,
+      "step": 1026
+    },
+    {
+      "epoch": 0.24969608558230003,
+      "grad_norm": 14.125,
+      "learning_rate": 2.4205795845957062e-06,
+      "loss": 0.4152,
+      "step": 1027
+    },
+    {
+      "epoch": 0.24993921711646,
+      "grad_norm": 16.25,
+      "learning_rate": 2.420409985755674e-06,
+      "loss": 0.7009,
+      "step": 1028
+    },
+    {
+      "epoch": 0.25018234865062,
+      "grad_norm": 31.0,
+      "learning_rate": 2.4202402119776556e-06,
+      "loss": 0.9549,
+      "step": 1029
+    },
+    {
+      "epoch": 0.25042548018477995,
+      "grad_norm": 23.375,
+      "learning_rate": 2.420070263287027e-06,
+      "loss": 0.9725,
+      "step": 1030
+    },
+    {
+      "epoch": 0.25066861171893995,
+      "grad_norm": 23.625,
+      "learning_rate": 2.4199001397091894e-06,
+      "loss": 1.3791,
+      "step": 1031
+    },
+    {
+      "epoch": 0.25091174325309995,
+      "grad_norm": 19.25,
+      "learning_rate": 2.4197298412695712e-06,
+      "loss": 0.9608,
+      "step": 1032
+    },
+    {
+      "epoch": 0.2511548747872599,
+      "grad_norm": 18.875,
+      "learning_rate": 2.419559367993626e-06,
+      "loss": 0.7395,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2513980063214199,
+      "grad_norm": 22.5,
+      "learning_rate": 2.4193887199068342e-06,
+      "loss": 1.0357,
+      "step": 1034
+    },
+    {
+      "epoch": 0.25164113785557984,
+      "grad_norm": 18.875,
+      "learning_rate": 2.419217897034703e-06,
+      "loss": 0.6159,
+      "step": 1035
+    },
+    {
+      "epoch": 0.25188426938973985,
+      "grad_norm": 19.375,
+      "learning_rate": 2.4190468994027633e-06,
+      "loss": 0.8969,
+      "step": 1036
+    },
+    {
+      "epoch": 0.25212740092389985,
+      "grad_norm": 13.8125,
+      "learning_rate": 2.4188757270365744e-06,
+      "loss": 0.5303,
+      "step": 1037
+    },
+    {
+      "epoch": 0.2523705324580598,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.418704379961721e-06,
+      "loss": 0.464,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2526136639922198,
+      "grad_norm": 22.625,
+      "learning_rate": 2.418532858203814e-06,
+      "loss": 1.5141,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2528567955263798,
+      "grad_norm": 17.25,
+      "learning_rate": 2.41836116178849e-06,
+      "loss": 0.8802,
+      "step": 1040
+    },
+    {
+      "epoch": 0.25309992706053974,
+      "grad_norm": 26.0,
+      "learning_rate": 2.4181892907414116e-06,
+      "loss": 1.1084,
+      "step": 1041
+    },
+    {
+      "epoch": 0.25334305859469974,
+      "grad_norm": 23.375,
+      "learning_rate": 2.418017245088269e-06,
+      "loss": 1.1972,
+      "step": 1042
+    },
+    {
+      "epoch": 0.2535861901288597,
+      "grad_norm": 30.875,
+      "learning_rate": 2.4178450248547763e-06,
+      "loss": 0.9698,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2538293216630197,
+      "grad_norm": 33.5,
+      "learning_rate": 2.4176726300666757e-06,
+      "loss": 0.8843,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2540724531971797,
+      "grad_norm": 18.375,
+      "learning_rate": 2.417500060749734e-06,
+      "loss": 0.6135,
+      "step": 1045
+    },
+    {
+      "epoch": 0.25431558473133964,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4173273169297446e-06,
+      "loss": 0.8931,
+      "step": 1046
+    },
+    {
+      "epoch": 0.25455871626549964,
+      "grad_norm": 16.0,
+      "learning_rate": 2.4171543986325272e-06,
+      "loss": 0.7161,
+      "step": 1047
+    },
+    {
+      "epoch": 0.25480184779965964,
+      "grad_norm": 16.375,
+      "learning_rate": 2.4169813058839277e-06,
+      "loss": 0.7568,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2550449793338196,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4168080387098175e-06,
+      "loss": 1.3455,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2552881108679796,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4166345971360944e-06,
+      "loss": 0.7672,
+      "step": 1050
+    },
+    {
+      "epoch": 0.25553124240213954,
+      "grad_norm": 24.25,
+      "learning_rate": 2.416460981188682e-06,
+      "loss": 1.2802,
+      "step": 1051
+    },
+    {
+      "epoch": 0.25577437393629954,
+      "grad_norm": 27.0,
+      "learning_rate": 2.4162871908935308e-06,
+      "loss": 1.5866,
+      "step": 1052
+    },
+    {
+      "epoch": 0.25601750547045954,
+      "grad_norm": 21.875,
+      "learning_rate": 2.4161132262766163e-06,
+      "loss": 0.988,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2562606370046195,
+      "grad_norm": 59.75,
+      "learning_rate": 2.415939087363941e-06,
+      "loss": 1.196,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2565037685387795,
+      "grad_norm": 17.5,
+      "learning_rate": 2.415764774181533e-06,
+      "loss": 0.967,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2567469000729395,
+      "grad_norm": 17.75,
+      "learning_rate": 2.415590286755445e-06,
+      "loss": 0.9417,
+      "step": 1056
+    },
+    {
+      "epoch": 0.25699003160709943,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4154156251117584e-06,
+      "loss": 0.8826,
+      "step": 1057
+    },
+    {
+      "epoch": 0.25723316314125944,
+      "grad_norm": 20.375,
+      "learning_rate": 2.4152407892765798e-06,
+      "loss": 1.0421,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2574762946754194,
+      "grad_norm": 17.375,
+      "learning_rate": 2.4150657792760404e-06,
+      "loss": 0.8625,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2577194262095794,
+      "grad_norm": 15.625,
+      "learning_rate": 2.414890595136299e-06,
+      "loss": 0.755,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2579625577437394,
+      "grad_norm": 18.875,
+      "learning_rate": 2.41471523688354e-06,
+      "loss": 0.953,
+      "step": 1061
+    },
+    {
+      "epoch": 0.25820568927789933,
+      "grad_norm": 16.5,
+      "learning_rate": 2.4145397045439734e-06,
+      "loss": 0.8457,
+      "step": 1062
+    },
+    {
+      "epoch": 0.25844882081205933,
+      "grad_norm": 16.25,
+      "learning_rate": 2.4143639981438357e-06,
+      "loss": 0.7202,
+      "step": 1063
+    },
+    {
+      "epoch": 0.2586919523462193,
+      "grad_norm": 16.0,
+      "learning_rate": 2.414188117709389e-06,
+      "loss": 0.9083,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2589350838803793,
+      "grad_norm": 25.375,
+      "learning_rate": 2.4140120632669216e-06,
+      "loss": 0.9105,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2591782154145393,
+      "grad_norm": 20.5,
+      "learning_rate": 2.413835834842749e-06,
+      "loss": 0.9964,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2594213469486992,
+      "grad_norm": 17.75,
+      "learning_rate": 2.4136594324632102e-06,
+      "loss": 0.9518,
+      "step": 1067
+    },
+    {
+      "epoch": 0.25966447848285923,
+      "grad_norm": 18.25,
+      "learning_rate": 2.413482856154672e-06,
+      "loss": 0.8343,
+      "step": 1068
+    },
+    {
+      "epoch": 0.25990761001701923,
+      "grad_norm": 17.375,
+      "learning_rate": 2.413306105943527e-06,
+      "loss": 0.5186,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2601507415511792,
+      "grad_norm": 16.25,
+      "learning_rate": 2.4131291818561937e-06,
+      "loss": 1.4428,
+      "step": 1070
+    },
+    {
+      "epoch": 0.2603938730853392,
+      "grad_norm": 16.75,
+      "learning_rate": 2.4129520839191162e-06,
+      "loss": 0.5751,
+      "step": 1071
+    },
+    {
+      "epoch": 0.2606370046194991,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4127748121587646e-06,
+      "loss": 0.971,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2608801361536591,
+      "grad_norm": 29.375,
+      "learning_rate": 2.412597366601636e-06,
+      "loss": 1.0751,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2611232676878191,
+      "grad_norm": 30.375,
+      "learning_rate": 2.4124197472742516e-06,
+      "loss": 1.3877,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2613663992219791,
+      "grad_norm": 22.875,
+      "learning_rate": 2.4122419542031607e-06,
+      "loss": 1.0804,
+      "step": 1075
+    },
+    {
+      "epoch": 0.2616095307561391,
+      "grad_norm": 27.125,
+      "learning_rate": 2.412063987414937e-06,
+      "loss": 0.8204,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2618526622902991,
+      "grad_norm": 19.375,
+      "learning_rate": 2.4118858469361813e-06,
+      "loss": 1.0674,
+      "step": 1077
+    },
+    {
+      "epoch": 0.262095793824459,
+      "grad_norm": 19.75,
+      "learning_rate": 2.4117075327935186e-06,
+      "loss": 0.8349,
+      "step": 1078
+    },
+    {
+      "epoch": 0.262338925358619,
+      "grad_norm": 18.375,
+      "learning_rate": 2.411529045013602e-06,
+      "loss": 0.6545,
+      "step": 1079
+    },
+    {
+      "epoch": 0.26258205689277897,
+      "grad_norm": 17.125,
+      "learning_rate": 2.4113503836231096e-06,
+      "loss": 0.9066,
+      "step": 1080
+    },
+    {
+      "epoch": 0.26282518842693897,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4111715486487447e-06,
+      "loss": 0.7835,
+      "step": 1081
+    },
+    {
+      "epoch": 0.263068319961099,
+      "grad_norm": 16.0,
+      "learning_rate": 2.4109925401172377e-06,
+      "loss": 0.6715,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2633114514952589,
+      "grad_norm": 17.375,
+      "learning_rate": 2.410813358055345e-06,
+      "loss": 0.8965,
+      "step": 1083
+    },
+    {
+      "epoch": 0.2635545830294189,
+      "grad_norm": 38.75,
+      "learning_rate": 2.4106340024898478e-06,
+      "loss": 1.6617,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2637977145635789,
+      "grad_norm": 27.375,
+      "learning_rate": 2.4104544734475544e-06,
+      "loss": 1.2628,
+      "step": 1085
+    },
+    {
+      "epoch": 0.26404084609773887,
+      "grad_norm": 17.5,
+      "learning_rate": 2.4102747709552975e-06,
+      "loss": 0.632,
+      "step": 1086
+    },
+    {
+      "epoch": 0.26428397763189887,
+      "grad_norm": 26.5,
+      "learning_rate": 2.410094895039938e-06,
+      "loss": 1.0735,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2645271091660588,
+      "grad_norm": 23.75,
+      "learning_rate": 2.4099148457283606e-06,
+      "loss": 1.2874,
+      "step": 1088
+    },
+    {
+      "epoch": 0.2647702407002188,
+      "grad_norm": 17.0,
+      "learning_rate": 2.4097346230474774e-06,
+      "loss": 0.5789,
+      "step": 1089
+    },
+    {
+      "epoch": 0.2650133722343788,
+      "grad_norm": 16.125,
+      "learning_rate": 2.409554227024225e-06,
+      "loss": 0.9835,
+      "step": 1090
+    },
+    {
+      "epoch": 0.26525650376853876,
+      "grad_norm": 20.625,
+      "learning_rate": 2.4093736576855675e-06,
+      "loss": 0.8613,
+      "step": 1091
+    },
+    {
+      "epoch": 0.26549963530269877,
+      "grad_norm": 15.625,
+      "learning_rate": 2.4091929150584935e-06,
+      "loss": 0.7219,
+      "step": 1092
+    },
+    {
+      "epoch": 0.26574276683685877,
+      "grad_norm": 22.25,
+      "learning_rate": 2.4090119991700187e-06,
+      "loss": 1.1011,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2659858983710187,
+      "grad_norm": 19.625,
+      "learning_rate": 2.408830910047184e-06,
+      "loss": 0.727,
+      "step": 1094
+    },
+    {
+      "epoch": 0.2662290299051787,
+      "grad_norm": 18.5,
+      "learning_rate": 2.4086496477170556e-06,
+      "loss": 0.9894,
+      "step": 1095
+    },
+    {
+      "epoch": 0.26647216143933866,
+      "grad_norm": 17.125,
+      "learning_rate": 2.408468212206727e-06,
+      "loss": 0.593,
+      "step": 1096
+    },
+    {
+      "epoch": 0.26671529297349866,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.4082866035433167e-06,
+      "loss": 0.6468,
+      "step": 1097
+    },
+    {
+      "epoch": 0.26695842450765866,
+      "grad_norm": 20.875,
+      "learning_rate": 2.4081048217539693e-06,
+      "loss": 0.8832,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2672015560418186,
+      "grad_norm": 18.5,
+      "learning_rate": 2.407922866865855e-06,
+      "loss": 1.0391,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2674446875759786,
+      "grad_norm": 18.875,
+      "learning_rate": 2.4077407389061703e-06,
+      "loss": 0.9531,
+      "step": 1100
+    },
+    {
+      "epoch": 0.26768781911013856,
+      "grad_norm": 19.625,
+      "learning_rate": 2.407558437902137e-06,
+      "loss": 1.1267,
+      "step": 1101
+    },
+    {
+      "epoch": 0.26793095064429856,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.4073759638810034e-06,
+      "loss": 0.4987,
+      "step": 1102
+    },
+    {
+      "epoch": 0.26817408217845856,
+      "grad_norm": 18.375,
+      "learning_rate": 2.407193316870044e-06,
+      "loss": 0.6768,
+      "step": 1103
+    },
+    {
+      "epoch": 0.2684172137126185,
+      "grad_norm": 21.875,
+      "learning_rate": 2.4070104968965572e-06,
+      "loss": 0.9036,
+      "step": 1104
+    },
+    {
+      "epoch": 0.2686603452467785,
+      "grad_norm": 20.625,
+      "learning_rate": 2.40682750398787e-06,
+      "loss": 1.1075,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2689034767809385,
+      "grad_norm": 19.5,
+      "learning_rate": 2.4066443381713332e-06,
+      "loss": 0.6668,
+      "step": 1106
+    },
+    {
+      "epoch": 0.26914660831509846,
+      "grad_norm": 19.625,
+      "learning_rate": 2.406460999474324e-06,
+      "loss": 1.027,
+      "step": 1107
+    },
+    {
+      "epoch": 0.26938973984925846,
+      "grad_norm": 28.125,
+      "learning_rate": 2.4062774879242454e-06,
+      "loss": 1.557,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2696328713834184,
+      "grad_norm": 25.25,
+      "learning_rate": 2.406093803548527e-06,
+      "loss": 1.5739,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2698760029175784,
+      "grad_norm": 15.0,
+      "learning_rate": 2.4059099463746228e-06,
+      "loss": 0.4134,
+      "step": 1110
+    },
+    {
+      "epoch": 0.2701191344517384,
+      "grad_norm": 16.75,
+      "learning_rate": 2.405725916430014e-06,
+      "loss": 0.5757,
+      "step": 1111
+    },
+    {
+      "epoch": 0.27036226598589835,
+      "grad_norm": 22.125,
+      "learning_rate": 2.4055417137422072e-06,
+      "loss": 1.015,
+      "step": 1112
+    },
+    {
+      "epoch": 0.27060539752005836,
+      "grad_norm": 22.625,
+      "learning_rate": 2.405357338338734e-06,
+      "loss": 0.7689,
+      "step": 1113
+    },
+    {
+      "epoch": 0.27084852905421836,
+      "grad_norm": 19.375,
+      "learning_rate": 2.4051727902471532e-06,
+      "loss": 1.0271,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2710916605883783,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4049880694950485e-06,
+      "loss": 0.9973,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2713347921225383,
+      "grad_norm": 17.75,
+      "learning_rate": 2.4048031761100286e-06,
+      "loss": 0.9478,
+      "step": 1116
+    },
+    {
+      "epoch": 0.27157792365669825,
+      "grad_norm": 17.0,
+      "learning_rate": 2.4046181101197307e-06,
+      "loss": 0.7035,
+      "step": 1117
+    },
+    {
+      "epoch": 0.27182105519085825,
+      "grad_norm": 17.625,
+      "learning_rate": 2.4044328715518154e-06,
+      "loss": 0.999,
+      "step": 1118
+    },
+    {
+      "epoch": 0.27206418672501825,
+      "grad_norm": 20.75,
+      "learning_rate": 2.4042474604339693e-06,
+      "loss": 0.8228,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2723073182591782,
+      "grad_norm": 17.875,
+      "learning_rate": 2.404061876793906e-06,
+      "loss": 0.8087,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2725504497933382,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.4038761206593636e-06,
+      "loss": 0.4776,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2727935813274982,
+      "grad_norm": 12.5625,
+      "learning_rate": 2.403690192058107e-06,
+      "loss": 0.4574,
+      "step": 1122
+    },
+    {
+      "epoch": 0.27303671286165815,
+      "grad_norm": 24.375,
+      "learning_rate": 2.4035040910179262e-06,
+      "loss": 1.0786,
+      "step": 1123
+    },
+    {
+      "epoch": 0.27327984439581815,
+      "grad_norm": 32.75,
+      "learning_rate": 2.403317817566637e-06,
+      "loss": 1.121,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2735229759299781,
+      "grad_norm": 19.0,
+      "learning_rate": 2.403131371732082e-06,
+      "loss": 1.0104,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2737661074641381,
+      "grad_norm": 17.0,
+      "learning_rate": 2.402944753542128e-06,
+      "loss": 0.5847,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2740092389982981,
+      "grad_norm": 17.125,
+      "learning_rate": 2.4027579630246683e-06,
+      "loss": 0.819,
+      "step": 1127
+    },
+    {
+      "epoch": 0.27425237053245805,
+      "grad_norm": 21.875,
+      "learning_rate": 2.4025710002076225e-06,
+      "loss": 1.0926,
+      "step": 1128
+    },
+    {
+      "epoch": 0.27449550206661805,
+      "grad_norm": 19.0,
+      "learning_rate": 2.402383865118935e-06,
+      "loss": 1.0038,
+      "step": 1129
+    },
+    {
+      "epoch": 0.27473863360077805,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.402196557786577e-06,
+      "loss": 1.035,
+      "step": 1130
+    },
+    {
+      "epoch": 0.274981765134938,
+      "grad_norm": 17.125,
+      "learning_rate": 2.4020090782385437e-06,
+      "loss": 0.8398,
+      "step": 1131
+    },
+    {
+      "epoch": 0.275224896669098,
+      "grad_norm": 16.75,
+      "learning_rate": 2.4018214265028577e-06,
+      "loss": 0.6444,
+      "step": 1132
+    },
+    {
+      "epoch": 0.27546802820325794,
+      "grad_norm": 18.75,
+      "learning_rate": 2.401633602607567e-06,
+      "loss": 0.7797,
+      "step": 1133
+    },
+    {
+      "epoch": 0.27571115973741794,
+      "grad_norm": 22.125,
+      "learning_rate": 2.4014456065807457e-06,
+      "loss": 1.0577,
+      "step": 1134
+    },
+    {
+      "epoch": 0.27595429127157795,
+      "grad_norm": 27.0,
+      "learning_rate": 2.4012574384504917e-06,
+      "loss": 1.0948,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2761974228057379,
+      "grad_norm": 30.875,
+      "learning_rate": 2.4010690982449307e-06,
+      "loss": 0.9635,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2764405543398979,
+      "grad_norm": 20.375,
+      "learning_rate": 2.400880585992213e-06,
+      "loss": 1.0681,
+      "step": 1137
+    },
+    {
+      "epoch": 0.27668368587405784,
+      "grad_norm": 20.5,
+      "learning_rate": 2.4006919017205158e-06,
+      "loss": 0.85,
+      "step": 1138
+    },
+    {
+      "epoch": 0.27692681740821784,
+      "grad_norm": 23.5,
+      "learning_rate": 2.4005030454580403e-06,
+      "loss": 1.1541,
+      "step": 1139
+    },
+    {
+      "epoch": 0.27716994894237784,
+      "grad_norm": 26.625,
+      "learning_rate": 2.4003140172330154e-06,
+      "loss": 0.8538,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2774130804765378,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4001248170736934e-06,
+      "loss": 1.057,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2776562120106978,
+      "grad_norm": 19.0,
+      "learning_rate": 2.3999354450083545e-06,
+      "loss": 1.1297,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2778993435448578,
+      "grad_norm": 19.0,
+      "learning_rate": 2.3997459010653033e-06,
+      "loss": 1.4319,
+      "step": 1143
+    },
+    {
+      "epoch": 0.27814247507901774,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.39955618527287e-06,
+      "loss": 0.5,
+      "step": 1144
+    },
+    {
+      "epoch": 0.27838560661317774,
+      "grad_norm": 20.75,
+      "learning_rate": 2.3993662976594116e-06,
+      "loss": 0.9077,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2786287381473377,
+      "grad_norm": 20.0,
+      "learning_rate": 2.3991762382533097e-06,
+      "loss": 1.0673,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2788718696814977,
+      "grad_norm": 23.0,
+      "learning_rate": 2.3989860070829724e-06,
+      "loss": 1.1778,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2791150012156577,
+      "grad_norm": 26.25,
+      "learning_rate": 2.3987956041768325e-06,
+      "loss": 0.9862,
+      "step": 1148
+    },
+    {
+      "epoch": 0.27935813274981763,
+      "grad_norm": 20.875,
+      "learning_rate": 2.3986050295633486e-06,
+      "loss": 0.8852,
+      "step": 1149
+    },
+    {
+      "epoch": 0.27960126428397764,
+      "grad_norm": 14.375,
+      "learning_rate": 2.3984142832710065e-06,
+      "loss": 0.3197,
+      "step": 1150
+    },
+    {
+      "epoch": 0.27984439581813764,
+      "grad_norm": 18.0,
+      "learning_rate": 2.3982233653283156e-06,
+      "loss": 1.1707,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2800875273522976,
+      "grad_norm": 18.125,
+      "learning_rate": 2.3980322757638124e-06,
+      "loss": 1.2708,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2803306588864576,
+      "grad_norm": 16.125,
+      "learning_rate": 2.397841014606059e-06,
+      "loss": 1.1025,
+      "step": 1153
+    },
+    {
+      "epoch": 0.28057379042061753,
+      "grad_norm": 17.875,
+      "learning_rate": 2.3976495818836408e-06,
+      "loss": 1.0117,
+      "step": 1154
+    },
+    {
+      "epoch": 0.28081692195477753,
+      "grad_norm": 15.625,
+      "learning_rate": 2.397457977625173e-06,
+      "loss": 0.7081,
+      "step": 1155
+    },
+    {
+      "epoch": 0.28106005348893753,
+      "grad_norm": 18.75,
+      "learning_rate": 2.397266201859293e-06,
+      "loss": 0.9149,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2813031850230975,
+      "grad_norm": 15.25,
+      "learning_rate": 2.3970742546146646e-06,
+      "loss": 0.5956,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2815463165572575,
+      "grad_norm": 22.75,
+      "learning_rate": 2.396882135919979e-06,
+      "loss": 1.4236,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2817894480914175,
+      "grad_norm": 19.875,
+      "learning_rate": 2.3966898458039505e-06,
+      "loss": 0.8948,
+      "step": 1159
+    },
+    {
+      "epoch": 0.28203257962557743,
+      "grad_norm": 20.375,
+      "learning_rate": 2.3964973842953202e-06,
+      "loss": 1.0794,
+      "step": 1160
+    },
+    {
+      "epoch": 0.28227571115973743,
+      "grad_norm": 23.5,
+      "learning_rate": 2.3963047514228556e-06,
+      "loss": 1.3466,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2825188426938974,
+      "grad_norm": 18.375,
+      "learning_rate": 2.3961119472153484e-06,
+      "loss": 0.7326,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2827619742280574,
+      "grad_norm": 17.125,
+      "learning_rate": 2.395918971701616e-06,
+      "loss": 0.9474,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2830051057622174,
+      "grad_norm": 20.375,
+      "learning_rate": 2.3957258249105035e-06,
+      "loss": 0.859,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2832482372963773,
+      "grad_norm": 17.625,
+      "learning_rate": 2.3955325068708788e-06,
+      "loss": 0.7619,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2834913688305373,
+      "grad_norm": 38.25,
+      "learning_rate": 2.395339017611637e-06,
+      "loss": 0.9685,
+      "step": 1166
+    },
+    {
+      "epoch": 0.28373450036469733,
+      "grad_norm": 21.0,
+      "learning_rate": 2.395145357161698e-06,
+      "loss": 0.9813,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2839776318988573,
+      "grad_norm": 24.75,
+      "learning_rate": 2.3949515255500083e-06,
+      "loss": 1.2902,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2842207634330173,
+      "grad_norm": 18.0,
+      "learning_rate": 2.394757522805539e-06,
+      "loss": 1.0547,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2844638949671772,
+      "grad_norm": 20.125,
+      "learning_rate": 2.3945633489572874e-06,
+      "loss": 1.0345,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2847070265013372,
+      "grad_norm": 19.0,
+      "learning_rate": 2.394369004034276e-06,
+      "loss": 1.0358,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2849501580354972,
+      "grad_norm": 24.375,
+      "learning_rate": 2.394174488065553e-06,
+      "loss": 1.0853,
+      "step": 1172
+    },
+    {
+      "epoch": 0.28519328956965717,
+      "grad_norm": 16.25,
+      "learning_rate": 2.3939798010801918e-06,
+      "loss": 0.5782,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2854364211038172,
+      "grad_norm": 20.25,
+      "learning_rate": 2.3937849431072924e-06,
+      "loss": 1.1041,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2856795526379771,
+      "grad_norm": 20.0,
+      "learning_rate": 2.3935899141759794e-06,
+      "loss": 0.7543,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2859226841721371,
+      "grad_norm": 17.125,
+      "learning_rate": 2.3933947143154036e-06,
+      "loss": 0.7341,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2861658157062971,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.39319934355474e-06,
+      "loss": 0.7032,
+      "step": 1177
+    },
+    {
+      "epoch": 0.28640894724045707,
+      "grad_norm": 28.75,
+      "learning_rate": 2.393003801923191e-06,
+      "loss": 1.172,
+      "step": 1178
+    },
+    {
+      "epoch": 0.28665207877461707,
+      "grad_norm": 18.25,
+      "learning_rate": 2.3928080894499835e-06,
+      "loss": 1.2408,
+      "step": 1179
+    },
+    {
+      "epoch": 0.28689521030877707,
+      "grad_norm": 23.0,
+      "learning_rate": 2.3926122061643703e-06,
+      "loss": 0.8347,
+      "step": 1180
+    },
+    {
+      "epoch": 0.287138341842937,
+      "grad_norm": 24.25,
+      "learning_rate": 2.392416152095629e-06,
+      "loss": 1.0152,
+      "step": 1181
+    },
+    {
+      "epoch": 0.287381473377097,
+      "grad_norm": 24.0,
+      "learning_rate": 2.3922199272730632e-06,
+      "loss": 1.8937,
+      "step": 1182
+    },
+    {
+      "epoch": 0.28762460491125696,
+      "grad_norm": 15.5,
+      "learning_rate": 2.392023531726003e-06,
+      "loss": 0.7869,
+      "step": 1183
+    },
+    {
+      "epoch": 0.28786773644541697,
+      "grad_norm": 16.125,
+      "learning_rate": 2.3918269654838028e-06,
+      "loss": 0.7737,
+      "step": 1184
+    },
+    {
+      "epoch": 0.28811086797957697,
+      "grad_norm": 20.75,
+      "learning_rate": 2.391630228575842e-06,
+      "loss": 0.7496,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2883539995137369,
+      "grad_norm": 20.125,
+      "learning_rate": 2.391433321031527e-06,
+      "loss": 0.8947,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2885971310478969,
+      "grad_norm": 16.75,
+      "learning_rate": 2.391236242880289e-06,
+      "loss": 0.7558,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2888402625820569,
+      "grad_norm": 17.875,
+      "learning_rate": 2.391038994151585e-06,
+      "loss": 0.773,
+      "step": 1188
+    },
+    {
+      "epoch": 0.28908339411621686,
+      "grad_norm": 27.875,
+      "learning_rate": 2.3908415748748964e-06,
+      "loss": 1.1688,
+      "step": 1189
+    },
+    {
+      "epoch": 0.28932652565037686,
+      "grad_norm": 19.375,
+      "learning_rate": 2.390643985079732e-06,
+      "loss": 0.9692,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2895696571845368,
+      "grad_norm": 19.25,
+      "learning_rate": 2.390446224795624e-06,
+      "loss": 0.8757,
+      "step": 1191
+    },
+    {
+      "epoch": 0.2898127887186968,
+      "grad_norm": 24.25,
+      "learning_rate": 2.3902482940521316e-06,
+      "loss": 0.9955,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2900559202528568,
+      "grad_norm": 22.625,
+      "learning_rate": 2.3900501928788386e-06,
+      "loss": 0.9887,
+      "step": 1193
+    },
+    {
+      "epoch": 0.29029905178701676,
+      "grad_norm": 20.0,
+      "learning_rate": 2.389851921305355e-06,
+      "loss": 0.9916,
+      "step": 1194
+    },
+    {
+      "epoch": 0.29054218332117676,
+      "grad_norm": 22.5,
+      "learning_rate": 2.3896534793613164e-06,
+      "loss": 1.0105,
+      "step": 1195
+    },
+    {
+      "epoch": 0.29078531485533676,
+      "grad_norm": 20.625,
+      "learning_rate": 2.3894548670763825e-06,
+      "loss": 0.8992,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2910284463894967,
+      "grad_norm": 20.25,
+      "learning_rate": 2.3892560844802394e-06,
+      "loss": 1.1296,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2912715779236567,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.389057131602599e-06,
+      "loss": 0.8563,
+      "step": 1198
+    },
+    {
+      "epoch": 0.29151470945781666,
+      "grad_norm": 22.125,
+      "learning_rate": 2.388858008473198e-06,
+      "loss": 0.8952,
+      "step": 1199
+    },
+    {
+      "epoch": 0.29175784099197666,
+      "grad_norm": 21.0,
+      "learning_rate": 2.3886587151217986e-06,
+      "loss": 0.8936,
+      "step": 1200
+    },
+    {
+      "epoch": 0.29200097252613666,
+      "grad_norm": 16.125,
+      "learning_rate": 2.3884592515781895e-06,
+      "loss": 0.6503,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2922441040602966,
+      "grad_norm": 21.5,
+      "learning_rate": 2.3882596178721835e-06,
+      "loss": 1.1833,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2924872355944566,
+      "grad_norm": 17.625,
+      "learning_rate": 2.3880598140336185e-06,
+      "loss": 0.9162,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2927303671286166,
+      "grad_norm": 17.0,
+      "learning_rate": 2.3878598400923597e-06,
+      "loss": 0.8729,
+      "step": 1204
+    },
+    {
+      "epoch": 0.29297349866277655,
+      "grad_norm": 21.125,
+      "learning_rate": 2.3876596960782967e-06,
+      "loss": 1.2932,
+      "step": 1205
+    },
+    {
+      "epoch": 0.29321663019693656,
+      "grad_norm": 23.375,
+      "learning_rate": 2.3874593820213434e-06,
+      "loss": 1.1353,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2934597617310965,
+      "grad_norm": 20.75,
+      "learning_rate": 2.387258897951441e-06,
+      "loss": 0.6816,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2937028932652565,
+      "grad_norm": 16.875,
+      "learning_rate": 2.3870582438985552e-06,
+      "loss": 0.662,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2939460247994165,
+      "grad_norm": 16.25,
+      "learning_rate": 2.386857419892677e-06,
+      "loss": 0.8001,
+      "step": 1209
+    },
+    {
+      "epoch": 0.29418915633357645,
+      "grad_norm": 23.5,
+      "learning_rate": 2.3866564259638237e-06,
+      "loss": 1.026,
+      "step": 1210
+    },
+    {
+      "epoch": 0.29443228786773645,
+      "grad_norm": 20.625,
+      "learning_rate": 2.3864552621420365e-06,
+      "loss": 0.8811,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2946754194018964,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.386253928457383e-06,
+      "loss": 0.7637,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2949185509360564,
+      "grad_norm": 18.125,
+      "learning_rate": 2.3860524249399564e-06,
+      "loss": 0.7777,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2951616824702164,
+      "grad_norm": 26.375,
+      "learning_rate": 2.385850751619874e-06,
+      "loss": 1.4308,
+      "step": 1214
+    },
+    {
+      "epoch": 0.29540481400437635,
+      "grad_norm": 16.0,
+      "learning_rate": 2.3856489085272806e-06,
+      "loss": 0.67,
+      "step": 1215
+    },
+    {
+      "epoch": 0.29564794553853635,
+      "grad_norm": 16.875,
+      "learning_rate": 2.3854468956923444e-06,
+      "loss": 0.6661,
+      "step": 1216
+    },
+    {
+      "epoch": 0.29589107707269635,
+      "grad_norm": 19.375,
+      "learning_rate": 2.3852447131452593e-06,
+      "loss": 0.992,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2961342086068563,
+      "grad_norm": 15.5,
+      "learning_rate": 2.385042360916246e-06,
+      "loss": 0.6511,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2963773401410163,
+      "grad_norm": 23.125,
+      "learning_rate": 2.384839839035549e-06,
+      "loss": 0.8863,
+      "step": 1219
+    },
+    {
+      "epoch": 0.29662047167517624,
+      "grad_norm": 26.25,
+      "learning_rate": 2.3846371475334382e-06,
+      "loss": 1.0645,
+      "step": 1220
+    },
+    {
+      "epoch": 0.29686360320933625,
+      "grad_norm": 21.0,
+      "learning_rate": 2.3844342864402103e-06,
+      "loss": 1.3082,
+      "step": 1221
+    },
+    {
+      "epoch": 0.29710673474349625,
+      "grad_norm": 17.125,
+      "learning_rate": 2.3842312557861854e-06,
+      "loss": 1.1422,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2973498662776562,
+      "grad_norm": 16.5,
+      "learning_rate": 2.384028055601711e-06,
+      "loss": 0.6125,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2975929978118162,
+      "grad_norm": 23.125,
+      "learning_rate": 2.3838246859171584e-06,
+      "loss": 1.024,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2978361293459762,
+      "grad_norm": 17.625,
+      "learning_rate": 2.383621146762924e-06,
+      "loss": 0.8368,
+      "step": 1225
+    },
+    {
+      "epoch": 0.29807926088013614,
+      "grad_norm": 17.375,
+      "learning_rate": 2.3834174381694314e-06,
+      "loss": 0.8363,
+      "step": 1226
+    },
+    {
+      "epoch": 0.29832239241429614,
+      "grad_norm": 20.25,
+      "learning_rate": 2.383213560167128e-06,
+      "loss": 1.2673,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2985655239484561,
+      "grad_norm": 29.25,
+      "learning_rate": 2.3830095127864867e-06,
+      "loss": 1.0421,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2988086554826161,
+      "grad_norm": 24.125,
+      "learning_rate": 2.3828052960580057e-06,
+      "loss": 0.8843,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2990517870167761,
+      "grad_norm": 23.625,
+      "learning_rate": 2.3826009100122087e-06,
+      "loss": 0.9703,
+      "step": 1230
+    },
+    {
+      "epoch": 0.29929491855093604,
+      "grad_norm": 21.0,
+      "learning_rate": 2.3823963546796456e-06,
+      "loss": 1.1192,
+      "step": 1231
+    },
+    {
+      "epoch": 0.29953805008509604,
+      "grad_norm": 22.125,
+      "learning_rate": 2.38219163009089e-06,
+      "loss": 0.6999,
+      "step": 1232
+    },
+    {
+      "epoch": 0.29978118161925604,
+      "grad_norm": 25.25,
+      "learning_rate": 2.381986736276542e-06,
+      "loss": 1.0703,
+      "step": 1233
+    },
+    {
+      "epoch": 0.300024313153416,
+      "grad_norm": 22.0,
+      "learning_rate": 2.3817816732672255e-06,
+      "loss": 0.9507,
+      "step": 1234
+    },
+    {
+      "epoch": 0.300267444687576,
+      "grad_norm": 18.5,
+      "learning_rate": 2.3815764410935914e-06,
+      "loss": 0.681,
+      "step": 1235
+    },
+    {
+      "epoch": 0.30051057622173594,
+      "grad_norm": 18.25,
+      "learning_rate": 2.3813710397863158e-06,
+      "loss": 0.7719,
+      "step": 1236
+    },
+    {
+      "epoch": 0.30075370775589594,
+      "grad_norm": 21.75,
+      "learning_rate": 2.3811654693760985e-06,
+      "loss": 0.9225,
+      "step": 1237
+    },
+    {
+      "epoch": 0.30099683929005594,
+      "grad_norm": 19.25,
+      "learning_rate": 2.3809597298936656e-06,
+      "loss": 0.8943,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3012399708242159,
+      "grad_norm": 16.375,
+      "learning_rate": 2.380753821369769e-06,
+      "loss": 0.624,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3014831023583759,
+      "grad_norm": 16.375,
+      "learning_rate": 2.380547743835185e-06,
+      "loss": 0.4559,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3017262338925359,
+      "grad_norm": 16.125,
+      "learning_rate": 2.3803414973207154e-06,
+      "loss": 0.5842,
+      "step": 1241
+    },
+    {
+      "epoch": 0.30196936542669583,
+      "grad_norm": 17.375,
+      "learning_rate": 2.3801350818571876e-06,
+      "loss": 0.9134,
+      "step": 1242
+    },
+    {
+      "epoch": 0.30221249696085584,
+      "grad_norm": 20.0,
+      "learning_rate": 2.3799284974754534e-06,
+      "loss": 1.0343,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3024556284950158,
+      "grad_norm": 20.5,
+      "learning_rate": 2.379721744206391e-06,
+      "loss": 0.8462,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3026987600291758,
+      "grad_norm": 16.625,
+      "learning_rate": 2.3795148220809027e-06,
+      "loss": 0.6042,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3029418915633358,
+      "grad_norm": 16.625,
+      "learning_rate": 2.379307731129917e-06,
+      "loss": 0.683,
+      "step": 1246
+    },
+    {
+      "epoch": 0.30318502309749573,
+      "grad_norm": 17.625,
+      "learning_rate": 2.379100471384387e-06,
+      "loss": 0.5934,
+      "step": 1247
+    },
+    {
+      "epoch": 0.30342815463165573,
+      "grad_norm": 17.25,
+      "learning_rate": 2.3788930428752914e-06,
+      "loss": 0.9201,
+      "step": 1248
+    },
+    {
+      "epoch": 0.3036712861658157,
+      "grad_norm": 23.625,
+      "learning_rate": 2.378685445633634e-06,
+      "loss": 1.0846,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3039144176999757,
+      "grad_norm": 20.875,
+      "learning_rate": 2.378477679690443e-06,
+      "loss": 0.904,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3041575492341357,
+      "grad_norm": 22.875,
+      "learning_rate": 2.378269745076774e-06,
+      "loss": 1.0509,
+      "step": 1251
+    },
+    {
+      "epoch": 0.30440068076829563,
+      "grad_norm": 21.625,
+      "learning_rate": 2.378061641823705e-06,
+      "loss": 0.9499,
+      "step": 1252
+    },
+    {
+      "epoch": 0.30464381230245563,
+      "grad_norm": 19.125,
+      "learning_rate": 2.377853369962342e-06,
+      "loss": 1.0333,
+      "step": 1253
+    },
+    {
+      "epoch": 0.30488694383661563,
+      "grad_norm": 23.5,
+      "learning_rate": 2.3776449295238142e-06,
+      "loss": 0.8748,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3051300753707756,
+      "grad_norm": 26.125,
+      "learning_rate": 2.377436320539276e-06,
+      "loss": 0.8633,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3053732069049356,
+      "grad_norm": 25.75,
+      "learning_rate": 2.3772275430399087e-06,
+      "loss": 1.3877,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3056163384390955,
+      "grad_norm": 26.0,
+      "learning_rate": 2.377018597056917e-06,
+      "loss": 0.8743,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3058594699732555,
+      "grad_norm": 25.5,
+      "learning_rate": 2.3768094826215317e-06,
+      "loss": 0.9658,
+      "step": 1258
+    },
+    {
+      "epoch": 0.30610260150741553,
+      "grad_norm": 30.875,
+      "learning_rate": 2.3766001997650086e-06,
+      "loss": 0.8682,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3063457330415755,
+      "grad_norm": 22.25,
+      "learning_rate": 2.3763907485186287e-06,
+      "loss": 0.8051,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3065888645757355,
+      "grad_norm": 24.0,
+      "learning_rate": 2.3761811289136978e-06,
+      "loss": 1.1045,
+      "step": 1261
+    },
+    {
+      "epoch": 0.3068319961098955,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.3759713409815473e-06,
+      "loss": 0.7073,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3070751276440554,
+      "grad_norm": 21.125,
+      "learning_rate": 2.375761384753534e-06,
+      "loss": 1.251,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3073182591782154,
+      "grad_norm": 17.625,
+      "learning_rate": 2.3755512602610386e-06,
+      "loss": 0.9009,
+      "step": 1264
+    },
+    {
+      "epoch": 0.30756139071237537,
+      "grad_norm": 25.5,
+      "learning_rate": 2.375340967535469e-06,
+      "loss": 0.9195,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3078045222465354,
+      "grad_norm": 18.25,
+      "learning_rate": 2.3751305066082563e-06,
+      "loss": 1.0106,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3080476537806954,
+      "grad_norm": 22.25,
+      "learning_rate": 2.3749198775108578e-06,
+      "loss": 1.1506,
+      "step": 1267
+    },
+    {
+      "epoch": 0.3082907853148553,
+      "grad_norm": 17.375,
+      "learning_rate": 2.3747090802747553e-06,
+      "loss": 1.0442,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3085339168490153,
+      "grad_norm": 17.0,
+      "learning_rate": 2.3744981149314567e-06,
+      "loss": 0.6376,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3087770483831753,
+      "grad_norm": 17.625,
+      "learning_rate": 2.3742869815124934e-06,
+      "loss": 1.0133,
+      "step": 1270
+    },
+    {
+      "epoch": 0.30902017991733527,
+      "grad_norm": 20.5,
+      "learning_rate": 2.374075680049424e-06,
+      "loss": 1.0628,
+      "step": 1271
+    },
+    {
+      "epoch": 0.30926331145149527,
+      "grad_norm": 17.125,
+      "learning_rate": 2.373864210573831e-06,
+      "loss": 0.9332,
+      "step": 1272
+    },
+    {
+      "epoch": 0.3095064429856552,
+      "grad_norm": 19.25,
+      "learning_rate": 2.3736525731173217e-06,
+      "loss": 0.9674,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3097495745198152,
+      "grad_norm": 17.75,
+      "learning_rate": 2.3734407677115295e-06,
+      "loss": 0.8424,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3099927060539752,
+      "grad_norm": 19.625,
+      "learning_rate": 2.3732287943881114e-06,
+      "loss": 0.695,
+      "step": 1275
+    },
+    {
+      "epoch": 0.31023583758813517,
+      "grad_norm": 19.0,
+      "learning_rate": 2.373016653178752e-06,
+      "loss": 1.032,
+      "step": 1276
+    },
+    {
+      "epoch": 0.31047896912229517,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.3728043441151584e-06,
+      "loss": 0.6565,
+      "step": 1277
+    },
+    {
+      "epoch": 0.31072210065645517,
+      "grad_norm": 18.5,
+      "learning_rate": 2.3725918672290637e-06,
+      "loss": 0.8705,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3109652321906151,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.3723792225522267e-06,
+      "loss": 0.4942,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3112083637247751,
+      "grad_norm": 25.125,
+      "learning_rate": 2.372166410116431e-06,
+      "loss": 1.1596,
+      "step": 1280
+    },
+    {
+      "epoch": 0.31145149525893506,
+      "grad_norm": 24.625,
+      "learning_rate": 2.3719534299534845e-06,
+      "loss": 0.9487,
+      "step": 1281
+    },
+    {
+      "epoch": 0.31169462679309506,
+      "grad_norm": 19.125,
+      "learning_rate": 2.3717402820952212e-06,
+      "loss": 1.0082,
+      "step": 1282
+    },
+    {
+      "epoch": 0.31193775832725507,
+      "grad_norm": 18.875,
+      "learning_rate": 2.3715269665734996e-06,
+      "loss": 1.0386,
+      "step": 1283
+    },
+    {
+      "epoch": 0.312180889861415,
+      "grad_norm": 18.375,
+      "learning_rate": 2.3713134834202033e-06,
+      "loss": 1.0792,
+      "step": 1284
+    },
+    {
+      "epoch": 0.312424021395575,
+      "grad_norm": 21.0,
+      "learning_rate": 2.371099832667241e-06,
+      "loss": 0.809,
+      "step": 1285
+    },
+    {
+      "epoch": 0.31266715292973496,
+      "grad_norm": 20.875,
+      "learning_rate": 2.3708860143465473e-06,
+      "loss": 0.7824,
+      "step": 1286
+    },
+    {
+      "epoch": 0.31291028446389496,
+      "grad_norm": 27.625,
+      "learning_rate": 2.37067202849008e-06,
+      "loss": 1.1836,
+      "step": 1287
+    },
+    {
+      "epoch": 0.31315341599805496,
+      "grad_norm": 18.0,
+      "learning_rate": 2.3704578751298237e-06,
+      "loss": 0.7504,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3133965475322149,
+      "grad_norm": 17.875,
+      "learning_rate": 2.3702435542977863e-06,
+      "loss": 0.7071,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3136396790663749,
+      "grad_norm": 18.75,
+      "learning_rate": 2.3700290660260026e-06,
+      "loss": 0.8358,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3138828106005349,
+      "grad_norm": 20.875,
+      "learning_rate": 2.369814410346532e-06,
+      "loss": 1.0616,
+      "step": 1291
+    },
+    {
+      "epoch": 0.31412594213469486,
+      "grad_norm": 21.0,
+      "learning_rate": 2.3695995872914573e-06,
+      "loss": 1.0078,
+      "step": 1292
+    },
+    {
+      "epoch": 0.31436907366885486,
+      "grad_norm": 22.375,
+      "learning_rate": 2.3693845968928885e-06,
+      "loss": 1.2884,
+      "step": 1293
+    },
+    {
+      "epoch": 0.3146122052030148,
+      "grad_norm": 25.5,
+      "learning_rate": 2.369169439182959e-06,
+      "loss": 1.3012,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3148553367371748,
+      "grad_norm": 20.25,
+      "learning_rate": 2.368954114193828e-06,
+      "loss": 1.1192,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3150984682713348,
+      "grad_norm": 19.25,
+      "learning_rate": 2.36873862195768e-06,
+      "loss": 1.0548,
+      "step": 1296
+    },
+    {
+      "epoch": 0.31534159980549475,
+      "grad_norm": 21.25,
+      "learning_rate": 2.3685229625067234e-06,
+      "loss": 1.0982,
+      "step": 1297
+    },
+    {
+      "epoch": 0.31558473133965476,
+      "grad_norm": 19.75,
+      "learning_rate": 2.3683071358731923e-06,
+      "loss": 0.912,
+      "step": 1298
+    },
+    {
+      "epoch": 0.31582786287381476,
+      "grad_norm": 22.375,
+      "learning_rate": 2.3680911420893464e-06,
+      "loss": 0.9697,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3160709944079747,
+      "grad_norm": 17.625,
+      "learning_rate": 2.367874981187469e-06,
+      "loss": 0.654,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3163141259421347,
+      "grad_norm": 17.5,
+      "learning_rate": 2.367658653199869e-06,
+      "loss": 0.854,
+      "step": 1301
+    },
+    {
+      "epoch": 0.31655725747629465,
+      "grad_norm": 15.5,
+      "learning_rate": 2.367442158158881e-06,
+      "loss": 0.5738,
+      "step": 1302
+    },
+    {
+      "epoch": 0.31680038901045465,
+      "grad_norm": 17.0,
+      "learning_rate": 2.367225496096864e-06,
+      "loss": 0.6767,
+      "step": 1303
+    },
+    {
+      "epoch": 0.31704352054461465,
+      "grad_norm": 20.0,
+      "learning_rate": 2.3670086670462007e-06,
+      "loss": 0.8495,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3172866520787746,
+      "grad_norm": 19.75,
+      "learning_rate": 2.366791671039301e-06,
+      "loss": 0.852,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3175297836129346,
+      "grad_norm": 15.625,
+      "learning_rate": 2.3665745081085983e-06,
+      "loss": 0.5346,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3177729151470946,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.3663571782865515e-06,
+      "loss": 0.5753,
+      "step": 1307
+    },
+    {
+      "epoch": 0.31801604668125455,
+      "grad_norm": 16.5,
+      "learning_rate": 2.366139681605644e-06,
+      "loss": 0.652,
+      "step": 1308
+    },
+    {
+      "epoch": 0.31825917821541455,
+      "grad_norm": 18.375,
+      "learning_rate": 2.365922018098385e-06,
+      "loss": 0.7502,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3185023097495745,
+      "grad_norm": 19.25,
+      "learning_rate": 2.365704187797308e-06,
+      "loss": 0.9262,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3187454412837345,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.3654861907349706e-06,
+      "loss": 0.6902,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3189885728178945,
+      "grad_norm": 17.125,
+      "learning_rate": 2.3652680269439575e-06,
+      "loss": 0.9122,
+      "step": 1312
+    },
+    {
+      "epoch": 0.31923170435205445,
+      "grad_norm": 18.375,
+      "learning_rate": 2.3650496964568765e-06,
+      "loss": 0.8371,
+      "step": 1313
+    },
+    {
+      "epoch": 0.31947483588621445,
+      "grad_norm": 22.125,
+      "learning_rate": 2.36483119930636e-06,
+      "loss": 1.1271,
+      "step": 1314
+    },
+    {
+      "epoch": 0.31971796742037445,
+      "grad_norm": 18.625,
+      "learning_rate": 2.3646125355250677e-06,
+      "loss": 0.8004,
+      "step": 1315
+    },
+    {
+      "epoch": 0.3199610989545344,
+      "grad_norm": 14.0,
+      "learning_rate": 2.3643937051456817e-06,
+      "loss": 0.5843,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3202042304886944,
+      "grad_norm": 21.875,
+      "learning_rate": 2.36417470820091e-06,
+      "loss": 0.9962,
+      "step": 1317
+    },
+    {
+      "epoch": 0.32044736202285434,
+      "grad_norm": 26.5,
+      "learning_rate": 2.363955544723486e-06,
+      "loss": 0.9226,
+      "step": 1318
+    },
+    {
+      "epoch": 0.32069049355701434,
+      "grad_norm": 17.875,
+      "learning_rate": 2.363736214746167e-06,
+      "loss": 0.9539,
+      "step": 1319
+    },
+    {
+      "epoch": 0.32093362509117435,
+      "grad_norm": 17.125,
+      "learning_rate": 2.363516718301736e-06,
+      "loss": 0.6266,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3211767566253343,
+      "grad_norm": 20.125,
+      "learning_rate": 2.363297055423e-06,
+      "loss": 0.8119,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3214198881594943,
+      "grad_norm": 15.375,
+      "learning_rate": 2.363077226142792e-06,
+      "loss": 0.5649,
+      "step": 1322
+    },
+    {
+      "epoch": 0.32166301969365424,
+      "grad_norm": 27.5,
+      "learning_rate": 2.362857230493969e-06,
+      "loss": 0.9691,
+      "step": 1323
+    },
+    {
+      "epoch": 0.32190615122781424,
+      "grad_norm": 21.0,
+      "learning_rate": 2.362637068509413e-06,
+      "loss": 0.8346,
+      "step": 1324
+    },
+    {
+      "epoch": 0.32214928276197424,
+      "grad_norm": 19.625,
+      "learning_rate": 2.3624167402220317e-06,
+      "loss": 1.1271,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3223924142961342,
+      "grad_norm": 22.875,
+      "learning_rate": 2.3621962456647564e-06,
+      "loss": 1.166,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3226355458302942,
+      "grad_norm": 20.25,
+      "learning_rate": 2.361975584870543e-06,
+      "loss": 0.7447,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3228786773644542,
+      "grad_norm": 19.625,
+      "learning_rate": 2.361754757872375e-06,
+      "loss": 0.7225,
+      "step": 1328
+    },
+    {
+      "epoch": 0.32312180889861414,
+      "grad_norm": 19.0,
+      "learning_rate": 2.361533764703258e-06,
+      "loss": 0.9487,
+      "step": 1329
+    },
+    {
+      "epoch": 0.32336494043277414,
+      "grad_norm": 17.625,
+      "learning_rate": 2.361312605396222e-06,
+      "loss": 1.0187,
+      "step": 1330
+    },
+    {
+      "epoch": 0.3236080719669341,
+      "grad_norm": 16.0,
+      "learning_rate": 2.3610912799843242e-06,
+      "loss": 0.5292,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3238512035010941,
+      "grad_norm": 20.375,
+      "learning_rate": 2.360869788500646e-06,
+      "loss": 1.4659,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3240943350352541,
+      "grad_norm": 21.0,
+      "learning_rate": 2.360648130978292e-06,
+      "loss": 0.8654,
+      "step": 1333
+    },
+    {
+      "epoch": 0.32433746656941403,
+      "grad_norm": 27.5,
+      "learning_rate": 2.3604263074503934e-06,
+      "loss": 0.697,
+      "step": 1334
+    },
+    {
+      "epoch": 0.32458059810357404,
+      "grad_norm": 22.375,
+      "learning_rate": 2.3602043179501056e-06,
+      "loss": 0.8634,
+      "step": 1335
+    },
+    {
+      "epoch": 0.32482372963773404,
+      "grad_norm": 18.5,
+      "learning_rate": 2.3599821625106086e-06,
+      "loss": 0.909,
+      "step": 1336
+    },
+    {
+      "epoch": 0.325066861171894,
+      "grad_norm": 16.625,
+      "learning_rate": 2.3597598411651072e-06,
+      "loss": 0.781,
+      "step": 1337
+    },
+    {
+      "epoch": 0.325309992706054,
+      "grad_norm": 17.0,
+      "learning_rate": 2.3595373539468315e-06,
+      "loss": 0.6706,
+      "step": 1338
+    },
+    {
+      "epoch": 0.32555312424021393,
+      "grad_norm": 22.25,
+      "learning_rate": 2.3593147008890356e-06,
+      "loss": 1.2848,
+      "step": 1339
+    },
+    {
+      "epoch": 0.32579625577437393,
+      "grad_norm": 20.375,
+      "learning_rate": 2.3590918820249993e-06,
+      "loss": 1.0801,
+      "step": 1340
+    },
+    {
+      "epoch": 0.32603938730853393,
+      "grad_norm": 18.375,
+      "learning_rate": 2.3588688973880268e-06,
+      "loss": 1.015,
+      "step": 1341
+    },
+    {
+      "epoch": 0.3262825188426939,
+      "grad_norm": 17.375,
+      "learning_rate": 2.3586457470114466e-06,
+      "loss": 0.5862,
+      "step": 1342
+    },
+    {
+      "epoch": 0.3265256503768539,
+      "grad_norm": 22.375,
+      "learning_rate": 2.3584224309286124e-06,
+      "loss": 1.3698,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3267687819110139,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.3581989491729028e-06,
+      "loss": 0.4445,
+      "step": 1344
+    },
+    {
+      "epoch": 0.32701191344517383,
+      "grad_norm": 16.25,
+      "learning_rate": 2.357975301777721e-06,
+      "loss": 0.8973,
+      "step": 1345
+    },
+    {
+      "epoch": 0.32725504497933383,
+      "grad_norm": 25.5,
+      "learning_rate": 2.357751488776495e-06,
+      "loss": 0.9841,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3274981765134938,
+      "grad_norm": 17.25,
+      "learning_rate": 2.3575275102026775e-06,
+      "loss": 0.7709,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3277413080476538,
+      "grad_norm": 17.25,
+      "learning_rate": 2.3573033660897457e-06,
+      "loss": 0.9584,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3279844395818138,
+      "grad_norm": 19.5,
+      "learning_rate": 2.357079056471202e-06,
+      "loss": 0.7564,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3282275711159737,
+      "grad_norm": 20.625,
+      "learning_rate": 2.3568545813805737e-06,
+      "loss": 0.839,
+      "step": 1350
+    },
+    {
+      "epoch": 0.32847070265013373,
+      "grad_norm": 18.375,
+      "learning_rate": 2.356629940851412e-06,
+      "loss": 0.8169,
+      "step": 1351
+    },
+    {
+      "epoch": 0.32871383418429373,
+      "grad_norm": 17.625,
+      "learning_rate": 2.3564051349172937e-06,
+      "loss": 0.9295,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3289569657184537,
+      "grad_norm": 19.0,
+      "learning_rate": 2.3561801636118197e-06,
+      "loss": 0.7949,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3292000972526137,
+      "grad_norm": 16.625,
+      "learning_rate": 2.3559550269686153e-06,
+      "loss": 0.8806,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3294432287867736,
+      "grad_norm": 13.5,
+      "learning_rate": 2.3557297250213324e-06,
+      "loss": 0.6512,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3296863603209336,
+      "grad_norm": 17.375,
+      "learning_rate": 2.355504257803645e-06,
+      "loss": 0.9964,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3299294918550936,
+      "grad_norm": 16.0,
+      "learning_rate": 2.3552786253492537e-06,
+      "loss": 0.5643,
+      "step": 1357
+    },
+    {
+      "epoch": 0.33017262338925357,
+      "grad_norm": 21.625,
+      "learning_rate": 2.3550528276918834e-06,
+      "loss": 1.1216,
+      "step": 1358
+    },
+    {
+      "epoch": 0.3304157549234136,
+      "grad_norm": 18.25,
+      "learning_rate": 2.354826864865282e-06,
+      "loss": 1.0007,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3306588864575735,
+      "grad_norm": 19.0,
+      "learning_rate": 2.3546007369032255e-06,
+      "loss": 0.8458,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3309020179917335,
+      "grad_norm": 20.625,
+      "learning_rate": 2.3543744438395117e-06,
+      "loss": 0.8335,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3311451495258935,
+      "grad_norm": 25.125,
+      "learning_rate": 2.3541479857079636e-06,
+      "loss": 1.193,
+      "step": 1362
+    },
+    {
+      "epoch": 0.33138828106005347,
+      "grad_norm": 20.25,
+      "learning_rate": 2.3539213625424304e-06,
+      "loss": 1.2014,
+      "step": 1363
+    },
+    {
+      "epoch": 0.33163141259421347,
+      "grad_norm": 19.75,
+      "learning_rate": 2.3536945743767836e-06,
+      "loss": 1.1478,
+      "step": 1364
+    },
+    {
+      "epoch": 0.33187454412837347,
+      "grad_norm": 21.5,
+      "learning_rate": 2.3534676212449214e-06,
+      "loss": 0.9309,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3321176756625334,
+      "grad_norm": 28.5,
+      "learning_rate": 2.353240503180766e-06,
+      "loss": 1.0866,
+      "step": 1366
+    },
+    {
+      "epoch": 0.3323608071966934,
+      "grad_norm": 24.125,
+      "learning_rate": 2.3530132202182633e-06,
+      "loss": 0.7259,
+      "step": 1367
+    },
+    {
+      "epoch": 0.33260393873085337,
+      "grad_norm": 18.0,
+      "learning_rate": 2.3527857723913853e-06,
+      "loss": 0.9255,
+      "step": 1368
+    },
+    {
+      "epoch": 0.33284707026501337,
+      "grad_norm": 18.625,
+      "learning_rate": 2.3525581597341278e-06,
+      "loss": 0.8591,
+      "step": 1369
+    },
+    {
+      "epoch": 0.33309020179917337,
+      "grad_norm": 21.625,
+      "learning_rate": 2.3523303822805117e-06,
+      "loss": 1.11,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 15.5,
+      "learning_rate": 2.352102440064582e-06,
+      "loss": 0.5432,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3335764648674933,
+      "grad_norm": 16.25,
+      "learning_rate": 2.3518743331204085e-06,
+      "loss": 0.8395,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3338195964016533,
+      "grad_norm": 18.125,
+      "learning_rate": 2.351646061482086e-06,
+      "loss": 0.794,
+      "step": 1373
+    },
+    {
+      "epoch": 0.33406272793581326,
+      "grad_norm": 66.5,
+      "learning_rate": 2.3514176251837332e-06,
+      "loss": 1.0758,
+      "step": 1374
+    },
+    {
+      "epoch": 0.33430585946997327,
+      "grad_norm": 23.75,
+      "learning_rate": 2.351189024259495e-06,
+      "loss": 0.9747,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3345489910041332,
+      "grad_norm": 16.375,
+      "learning_rate": 2.3509602587435383e-06,
+      "loss": 0.8169,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3347921225382932,
+      "grad_norm": 22.875,
+      "learning_rate": 2.3507313286700568e-06,
+      "loss": 0.8971,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3350352540724532,
+      "grad_norm": 21.375,
+      "learning_rate": 2.350502234073268e-06,
+      "loss": 0.9816,
+      "step": 1378
+    },
+    {
+      "epoch": 0.33527838560661316,
+      "grad_norm": 24.75,
+      "learning_rate": 2.3502729749874136e-06,
+      "loss": 1.3641,
+      "step": 1379
+    },
+    {
+      "epoch": 0.33552151714077316,
+      "grad_norm": 18.0,
+      "learning_rate": 2.3500435514467612e-06,
+      "loss": 0.8322,
+      "step": 1380
+    },
+    {
+      "epoch": 0.33576464867493316,
+      "grad_norm": 15.1875,
+      "learning_rate": 2.3498139634856017e-06,
+      "loss": 0.4193,
+      "step": 1381
+    },
+    {
+      "epoch": 0.3360077802090931,
+      "grad_norm": 21.625,
+      "learning_rate": 2.3495842111382505e-06,
+      "loss": 1.4416,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3362509117432531,
+      "grad_norm": 18.625,
+      "learning_rate": 2.3493542944390484e-06,
+      "loss": 1.1323,
+      "step": 1383
+    },
+    {
+      "epoch": 0.33649404327741306,
+      "grad_norm": 17.875,
+      "learning_rate": 2.349124213422361e-06,
+      "loss": 0.854,
+      "step": 1384
+    },
+    {
+      "epoch": 0.33673717481157306,
+      "grad_norm": 32.75,
+      "learning_rate": 2.348893968122577e-06,
+      "loss": 0.9347,
+      "step": 1385
+    },
+    {
+      "epoch": 0.33698030634573306,
+      "grad_norm": 21.875,
+      "learning_rate": 2.348663558574111e-06,
+      "loss": 1.0152,
+      "step": 1386
+    },
+    {
+      "epoch": 0.337223437879893,
+      "grad_norm": 24.125,
+      "learning_rate": 2.3484329848114017e-06,
+      "loss": 0.9265,
+      "step": 1387
+    },
+    {
+      "epoch": 0.337466569414053,
+      "grad_norm": 20.875,
+      "learning_rate": 2.348202246868912e-06,
+      "loss": 0.9097,
+      "step": 1388
+    },
+    {
+      "epoch": 0.337709700948213,
+      "grad_norm": 25.25,
+      "learning_rate": 2.34797134478113e-06,
+      "loss": 1.3648,
+      "step": 1389
+    },
+    {
+      "epoch": 0.33795283248237296,
+      "grad_norm": 17.75,
+      "learning_rate": 2.3477402785825683e-06,
+      "loss": 0.6084,
+      "step": 1390
+    },
+    {
+      "epoch": 0.33819596401653296,
+      "grad_norm": 16.875,
+      "learning_rate": 2.347509048307763e-06,
+      "loss": 0.6965,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3384390955506929,
+      "grad_norm": 15.375,
+      "learning_rate": 2.3472776539912756e-06,
+      "loss": 0.5375,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3386822270848529,
+      "grad_norm": 20.625,
+      "learning_rate": 2.347046095667692e-06,
+      "loss": 0.8974,
+      "step": 1393
+    },
+    {
+      "epoch": 0.3389253586190129,
+      "grad_norm": 28.625,
+      "learning_rate": 2.346814373371623e-06,
+      "loss": 1.0636,
+      "step": 1394
+    },
+    {
+      "epoch": 0.33916849015317285,
+      "grad_norm": 34.0,
+      "learning_rate": 2.3465824871377036e-06,
+      "loss": 1.1303,
+      "step": 1395
+    },
+    {
+      "epoch": 0.33941162168733285,
+      "grad_norm": 20.625,
+      "learning_rate": 2.3463504370005927e-06,
+      "loss": 1.0416,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3396547532214928,
+      "grad_norm": 19.375,
+      "learning_rate": 2.346118222994974e-06,
+      "loss": 0.81,
+      "step": 1397
+    },
+    {
+      "epoch": 0.3398978847556528,
+      "grad_norm": 17.5,
+      "learning_rate": 2.345885845155557e-06,
+      "loss": 0.9071,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3401410162898128,
+      "grad_norm": 18.375,
+      "learning_rate": 2.345653303517073e-06,
+      "loss": 0.7898,
+      "step": 1399
+    },
+    {
+      "epoch": 0.34038414782397275,
+      "grad_norm": 18.25,
+      "learning_rate": 2.345420598114281e-06,
+      "loss": 0.8184,
+      "step": 1400
+    },
+    {
+      "epoch": 0.34062727935813275,
+      "grad_norm": 21.25,
+      "learning_rate": 2.3451877289819614e-06,
+      "loss": 1.1307,
+      "step": 1401
+    },
+    {
+      "epoch": 0.34087041089229275,
+      "grad_norm": 21.25,
+      "learning_rate": 2.3449546961549215e-06,
+      "loss": 1.1001,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3411135424264527,
+      "grad_norm": 19.125,
+      "learning_rate": 2.3447214996679913e-06,
+      "loss": 1.2553,
+      "step": 1403
+    },
+    {
+      "epoch": 0.3413566739606127,
+      "grad_norm": 21.25,
+      "learning_rate": 2.344488139556027e-06,
+      "loss": 1.0843,
+      "step": 1404
+    },
+    {
+      "epoch": 0.34159980549477265,
+      "grad_norm": 22.375,
+      "learning_rate": 2.3442546158539074e-06,
+      "loss": 0.8769,
+      "step": 1405
+    },
+    {
+      "epoch": 0.34184293702893265,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.3440209285965374e-06,
+      "loss": 0.6061,
+      "step": 1406
+    },
+    {
+      "epoch": 0.34208606856309265,
+      "grad_norm": 24.875,
+      "learning_rate": 2.343787077818845e-06,
+      "loss": 0.6801,
+      "step": 1407
+    },
+    {
+      "epoch": 0.3423292000972526,
+      "grad_norm": 33.5,
+      "learning_rate": 2.3435530635557836e-06,
+      "loss": 1.4359,
+      "step": 1408
+    },
+    {
+      "epoch": 0.3425723316314126,
+      "grad_norm": 22.5,
+      "learning_rate": 2.34331888584233e-06,
+      "loss": 1.2318,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3428154631655726,
+      "grad_norm": 19.25,
+      "learning_rate": 2.343084544713487e-06,
+      "loss": 0.9751,
+      "step": 1410
+    },
+    {
+      "epoch": 0.34305859469973254,
+      "grad_norm": 24.875,
+      "learning_rate": 2.3428500402042807e-06,
+      "loss": 0.8773,
+      "step": 1411
+    },
+    {
+      "epoch": 0.34330172623389255,
+      "grad_norm": 25.125,
+      "learning_rate": 2.342615372349762e-06,
+      "loss": 0.8469,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3435448577680525,
+      "grad_norm": 36.0,
+      "learning_rate": 2.3423805411850054e-06,
+      "loss": 1.3903,
+      "step": 1413
+    },
+    {
+      "epoch": 0.3437879893022125,
+      "grad_norm": 21.125,
+      "learning_rate": 2.342145546745111e-06,
+      "loss": 0.9616,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3440311208363725,
+      "grad_norm": 16.75,
+      "learning_rate": 2.3419103890652023e-06,
+      "loss": 0.4577,
+      "step": 1415
+    },
+    {
+      "epoch": 0.34427425237053244,
+      "grad_norm": 19.75,
+      "learning_rate": 2.3416750681804283e-06,
+      "loss": 0.7604,
+      "step": 1416
+    },
+    {
+      "epoch": 0.34451738390469244,
+      "grad_norm": 15.25,
+      "learning_rate": 2.341439584125961e-06,
+      "loss": 0.9324,
+      "step": 1417
+    },
+    {
+      "epoch": 0.34476051543885244,
+      "grad_norm": 17.5,
+      "learning_rate": 2.341203936936999e-06,
+      "loss": 0.7031,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3450036469730124,
+      "grad_norm": 17.375,
+      "learning_rate": 2.3409681266487623e-06,
+      "loss": 1.0507,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3452467785071724,
+      "grad_norm": 14.875,
+      "learning_rate": 2.340732153296497e-06,
+      "loss": 0.4448,
+      "step": 1420
+    },
+    {
+      "epoch": 0.34548991004133234,
+      "grad_norm": 21.625,
+      "learning_rate": 2.3404960169154743e-06,
+      "loss": 0.8389,
+      "step": 1421
+    },
+    {
+      "epoch": 0.34573304157549234,
+      "grad_norm": 27.625,
+      "learning_rate": 2.3402597175409887e-06,
+      "loss": 0.8419,
+      "step": 1422
+    },
+    {
+      "epoch": 0.34597617310965234,
+      "grad_norm": 20.625,
+      "learning_rate": 2.3400232552083585e-06,
+      "loss": 0.9551,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3462193046438123,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.339786629952927e-06,
+      "loss": 0.4091,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3464624361779723,
+      "grad_norm": 25.375,
+      "learning_rate": 2.339549841810063e-06,
+      "loss": 1.3345,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3467055677121323,
+      "grad_norm": 19.75,
+      "learning_rate": 2.3393128908151576e-06,
+      "loss": 0.9239,
+      "step": 1426
+    },
+    {
+      "epoch": 0.34694869924629224,
+      "grad_norm": 16.25,
+      "learning_rate": 2.3390757770036277e-06,
+      "loss": 0.7221,
+      "step": 1427
+    },
+    {
+      "epoch": 0.34719183078045224,
+      "grad_norm": 20.0,
+      "learning_rate": 2.338838500410914e-06,
+      "loss": 0.7285,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3474349623146122,
+      "grad_norm": 16.625,
+      "learning_rate": 2.3386010610724817e-06,
+      "loss": 0.6575,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3476780938487722,
+      "grad_norm": 21.875,
+      "learning_rate": 2.33836345902382e-06,
+      "loss": 0.9959,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3479212253829322,
+      "grad_norm": 25.125,
+      "learning_rate": 2.3381256943004425e-06,
+      "loss": 1.1109,
+      "step": 1431
+    },
+    {
+      "epoch": 0.34816435691709213,
+      "grad_norm": 19.375,
+      "learning_rate": 2.3378877669378874e-06,
+      "loss": 1.3305,
+      "step": 1432
+    },
+    {
+      "epoch": 0.34840748845125213,
+      "grad_norm": 21.25,
+      "learning_rate": 2.337649676971717e-06,
+      "loss": 1.4021,
+      "step": 1433
+    },
+    {
+      "epoch": 0.3486506199854121,
+      "grad_norm": 17.5,
+      "learning_rate": 2.3374114244375177e-06,
+      "loss": 0.8123,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3488937515195721,
+      "grad_norm": 16.875,
+      "learning_rate": 2.337173009370902e-06,
+      "loss": 0.6366,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3491368830537321,
+      "grad_norm": 18.625,
+      "learning_rate": 2.336934431807503e-06,
+      "loss": 0.964,
+      "step": 1436
+    },
+    {
+      "epoch": 0.34938001458789203,
+      "grad_norm": 22.5,
+      "learning_rate": 2.336695691782981e-06,
+      "loss": 0.9122,
+      "step": 1437
+    },
+    {
+      "epoch": 0.34962314612205203,
+      "grad_norm": 17.75,
+      "learning_rate": 2.33645678933302e-06,
+      "loss": 0.8408,
+      "step": 1438
+    },
+    {
+      "epoch": 0.34986627765621203,
+      "grad_norm": 21.25,
+      "learning_rate": 2.336217724493328e-06,
+      "loss": 1.0306,
+      "step": 1439
+    },
+    {
+      "epoch": 0.350109409190372,
+      "grad_norm": 13.5625,
+      "learning_rate": 2.335978497299638e-06,
+      "loss": 0.8109,
+      "step": 1440
+    },
+    {
+      "epoch": 0.350352540724532,
+      "grad_norm": 22.375,
+      "learning_rate": 2.335739107787706e-06,
+      "loss": 0.939,
+      "step": 1441
+    },
+    {
+      "epoch": 0.3505956722586919,
+      "grad_norm": 19.5,
+      "learning_rate": 2.3354995559933127e-06,
+      "loss": 1.1162,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3508388037928519,
+      "grad_norm": 18.75,
+      "learning_rate": 2.335259841952264e-06,
+      "loss": 0.7332,
+      "step": 1443
+    },
+    {
+      "epoch": 0.35108193532701193,
+      "grad_norm": 22.5,
+      "learning_rate": 2.3350199657003882e-06,
+      "loss": 1.2186,
+      "step": 1444
+    },
+    {
+      "epoch": 0.3513250668611719,
+      "grad_norm": 21.875,
+      "learning_rate": 2.3347799272735398e-06,
+      "loss": 0.7035,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3515681983953319,
+      "grad_norm": 24.0,
+      "learning_rate": 2.3345397267075962e-06,
+      "loss": 0.9537,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3518113299294919,
+      "grad_norm": 18.5,
+      "learning_rate": 2.3342993640384604e-06,
+      "loss": 0.9893,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3520544614636518,
+      "grad_norm": 20.375,
+      "learning_rate": 2.334058839302058e-06,
+      "loss": 0.7534,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3522975929978118,
+      "grad_norm": 22.125,
+      "learning_rate": 2.3338181525343395e-06,
+      "loss": 1.2002,
+      "step": 1449
+    },
+    {
+      "epoch": 0.35254072453197177,
+      "grad_norm": 21.5,
+      "learning_rate": 2.33357730377128e-06,
+      "loss": 1.2812,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3527838560661318,
+      "grad_norm": 15.875,
+      "learning_rate": 2.3333362930488785e-06,
+      "loss": 0.7726,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3530269876002918,
+      "grad_norm": 21.375,
+      "learning_rate": 2.333095120403158e-06,
+      "loss": 0.9819,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3532701191344517,
+      "grad_norm": 21.125,
+      "learning_rate": 2.332853785870166e-06,
+      "loss": 0.8736,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3535132506686117,
+      "grad_norm": 23.25,
+      "learning_rate": 2.3326122894859745e-06,
+      "loss": 1.0426,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3537563822027717,
+      "grad_norm": 22.75,
+      "learning_rate": 2.3323706312866785e-06,
+      "loss": 1.1238,
+      "step": 1455
+    },
+    {
+      "epoch": 0.35399951373693167,
+      "grad_norm": 18.75,
+      "learning_rate": 2.332128811308399e-06,
+      "loss": 1.0771,
+      "step": 1456
+    },
+    {
+      "epoch": 0.35424264527109167,
+      "grad_norm": 20.875,
+      "learning_rate": 2.3318868295872793e-06,
+      "loss": 1.0684,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3544857768052516,
+      "grad_norm": 15.625,
+      "learning_rate": 2.3316446861594878e-06,
+      "loss": 0.5876,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3547289083394116,
+      "grad_norm": 18.125,
+      "learning_rate": 2.331402381061218e-06,
+      "loss": 0.859,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3549720398735716,
+      "grad_norm": 22.875,
+      "learning_rate": 2.3311599143286855e-06,
+      "loss": 0.8267,
+      "step": 1460
+    },
+    {
+      "epoch": 0.35521517140773157,
+      "grad_norm": 21.375,
+      "learning_rate": 2.3309172859981317e-06,
+      "loss": 1.0423,
+      "step": 1461
+    },
+    {
+      "epoch": 0.35545830294189157,
+      "grad_norm": 20.875,
+      "learning_rate": 2.330674496105821e-06,
+      "loss": 1.1045,
+      "step": 1462
+    },
+    {
+      "epoch": 0.35570143447605157,
+      "grad_norm": 22.125,
+      "learning_rate": 2.3304315446880434e-06,
+      "loss": 0.9494,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3559445660102115,
+      "grad_norm": 19.5,
+      "learning_rate": 2.330188431781111e-06,
+      "loss": 0.8851,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3561876975443715,
+      "grad_norm": 20.625,
+      "learning_rate": 2.329945157421363e-06,
+      "loss": 1.2262,
+      "step": 1465
+    },
+    {
+      "epoch": 0.35643082907853146,
+      "grad_norm": 20.75,
+      "learning_rate": 2.3297017216451597e-06,
+      "loss": 0.6847,
+      "step": 1466
+    },
+    {
+      "epoch": 0.35667396061269147,
+      "grad_norm": 27.25,
+      "learning_rate": 2.3294581244888867e-06,
+      "loss": 0.8174,
+      "step": 1467
+    },
+    {
+      "epoch": 0.35691709214685147,
+      "grad_norm": 19.5,
+      "learning_rate": 2.329214365988954e-06,
+      "loss": 1.0658,
+      "step": 1468
+    },
+    {
+      "epoch": 0.3571602236810114,
+      "grad_norm": 19.625,
+      "learning_rate": 2.328970446181796e-06,
+      "loss": 0.9411,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3574033552151714,
+      "grad_norm": 17.625,
+      "learning_rate": 2.32872636510387e-06,
+      "loss": 0.9679,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3576464867493314,
+      "grad_norm": 21.0,
+      "learning_rate": 2.3284821227916586e-06,
+      "loss": 0.9815,
+      "step": 1471
+    },
+    {
+      "epoch": 0.35788961828349136,
+      "grad_norm": 17.5,
+      "learning_rate": 2.3282377192816682e-06,
+      "loss": 1.068,
+      "step": 1472
+    },
+    {
+      "epoch": 0.35813274981765136,
+      "grad_norm": 24.625,
+      "learning_rate": 2.3279931546104286e-06,
+      "loss": 1.2518,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3583758813518113,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.3277484288144947e-06,
+      "loss": 0.5412,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3586190128859713,
+      "grad_norm": 21.375,
+      "learning_rate": 2.3275035419304443e-06,
+      "loss": 0.9545,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3588621444201313,
+      "grad_norm": 19.875,
+      "learning_rate": 2.3272584939948807e-06,
+      "loss": 0.9992,
+      "step": 1476
+    },
+    {
+      "epoch": 0.35910527595429126,
+      "grad_norm": 17.875,
+      "learning_rate": 2.3270132850444304e-06,
+      "loss": 0.4695,
+      "step": 1477
+    },
+    {
+      "epoch": 0.35934840748845126,
+      "grad_norm": 25.875,
+      "learning_rate": 2.3267679151157437e-06,
+      "loss": 0.7442,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3595915390226112,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.326522384245496e-06,
+      "loss": 0.5161,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3598346705567712,
+      "grad_norm": 20.75,
+      "learning_rate": 2.3262766924703856e-06,
+      "loss": 0.6954,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3600778020909312,
+      "grad_norm": 19.625,
+      "learning_rate": 2.3260308398271353e-06,
+      "loss": 1.154,
+      "step": 1481
+    },
+    {
+      "epoch": 0.36032093362509116,
+      "grad_norm": 25.875,
+      "learning_rate": 2.325784826352493e-06,
+      "loss": 1.123,
+      "step": 1482
+    },
+    {
+      "epoch": 0.36056406515925116,
+      "grad_norm": 12.9375,
+      "learning_rate": 2.3255386520832282e-06,
+      "loss": 0.4059,
+      "step": 1483
+    },
+    {
+      "epoch": 0.36080719669341116,
+      "grad_norm": 18.125,
+      "learning_rate": 2.325292317056137e-06,
+      "loss": 0.8285,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3610503282275711,
+      "grad_norm": 26.0,
+      "learning_rate": 2.3250458213080378e-06,
+      "loss": 1.1255,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3612934597617311,
+      "grad_norm": 24.75,
+      "learning_rate": 2.324799164875774e-06,
+      "loss": 0.9591,
+      "step": 1486
+    },
+    {
+      "epoch": 0.36153659129589105,
+      "grad_norm": 21.5,
+      "learning_rate": 2.3245523477962133e-06,
+      "loss": 1.1673,
+      "step": 1487
+    },
+    {
+      "epoch": 0.36177972283005105,
+      "grad_norm": 23.5,
+      "learning_rate": 2.324305370106245e-06,
+      "loss": 1.0086,
+      "step": 1488
+    },
+    {
+      "epoch": 0.36202285436421106,
+      "grad_norm": 24.125,
+      "learning_rate": 2.324058231842786e-06,
+      "loss": 0.9124,
+      "step": 1489
+    },
+    {
+      "epoch": 0.362265985898371,
+      "grad_norm": 24.125,
+      "learning_rate": 2.3238109330427746e-06,
+      "loss": 0.9154,
+      "step": 1490
+    },
+    {
+      "epoch": 0.362509117432531,
+      "grad_norm": 19.125,
+      "learning_rate": 2.323563473743173e-06,
+      "loss": 1.0681,
+      "step": 1491
+    },
+    {
+      "epoch": 0.362752248966691,
+      "grad_norm": 20.75,
+      "learning_rate": 2.32331585398097e-06,
+      "loss": 1.0651,
+      "step": 1492
+    },
+    {
+      "epoch": 0.36299538050085095,
+      "grad_norm": 16.875,
+      "learning_rate": 2.323068073793176e-06,
+      "loss": 0.9262,
+      "step": 1493
+    },
+    {
+      "epoch": 0.36323851203501095,
+      "grad_norm": 22.5,
+      "learning_rate": 2.3228201332168253e-06,
+      "loss": 0.8907,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3634816435691709,
+      "grad_norm": 23.375,
+      "learning_rate": 2.3225720322889778e-06,
+      "loss": 1.1956,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3637247751033309,
+      "grad_norm": 16.375,
+      "learning_rate": 2.3223237710467157e-06,
+      "loss": 0.6987,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3639679066374909,
+      "grad_norm": 25.25,
+      "learning_rate": 2.322075349527147e-06,
+      "loss": 0.9879,
+      "step": 1497
+    },
+    {
+      "epoch": 0.36421103817165085,
+      "grad_norm": 29.25,
+      "learning_rate": 2.321826767767401e-06,
+      "loss": 1.1628,
+      "step": 1498
+    },
+    {
+      "epoch": 0.36445416970581085,
+      "grad_norm": 34.25,
+      "learning_rate": 2.3215780258046344e-06,
+      "loss": 1.0867,
+      "step": 1499
+    },
+    {
+      "epoch": 0.36469730123997085,
+      "grad_norm": 20.5,
+      "learning_rate": 2.321329123676024e-06,
+      "loss": 1.0927,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3649404327741308,
+      "grad_norm": 16.875,
+      "learning_rate": 2.321080061418775e-06,
+      "loss": 0.7475,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3651835643082908,
+      "grad_norm": 16.625,
+      "learning_rate": 2.320830839070112e-06,
+      "loss": 0.5888,
+      "step": 1502
+    },
+    {
+      "epoch": 0.36542669584245074,
+      "grad_norm": 18.25,
+      "learning_rate": 2.3205814566672857e-06,
+      "loss": 0.7968,
+      "step": 1503
+    },
+    {
+      "epoch": 0.36566982737661075,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.320331914247571e-06,
+      "loss": 0.7453,
+      "step": 1504
+    },
+    {
+      "epoch": 0.36591295891077075,
+      "grad_norm": 17.5,
+      "learning_rate": 2.3200822118482675e-06,
+      "loss": 0.7019,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3661560904449307,
+      "grad_norm": 19.5,
+      "learning_rate": 2.3198323495066957e-06,
+      "loss": 0.916,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3663992219790907,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.319582327260203e-06,
+      "loss": 0.7411,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3666423535132507,
+      "grad_norm": 18.0,
+      "learning_rate": 2.319332145146159e-06,
+      "loss": 1.0345,
+      "step": 1508
+    },
+    {
+      "epoch": 0.36688548504741064,
+      "grad_norm": 24.5,
+      "learning_rate": 2.3190818032019578e-06,
+      "loss": 0.7271,
+      "step": 1509
+    },
+    {
+      "epoch": 0.36712861658157064,
+      "grad_norm": 27.0,
+      "learning_rate": 2.3188313014650178e-06,
+      "loss": 0.8754,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3673717481157306,
+      "grad_norm": 18.0,
+      "learning_rate": 2.31858063997278e-06,
+      "loss": 0.7795,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3676148796498906,
+      "grad_norm": 27.5,
+      "learning_rate": 2.3183298187627107e-06,
+      "loss": 0.9894,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3678580111840506,
+      "grad_norm": 18.0,
+      "learning_rate": 2.318078837872299e-06,
+      "loss": 1.0625,
+      "step": 1513
+    },
+    {
+      "epoch": 0.36810114271821054,
+      "grad_norm": 17.375,
+      "learning_rate": 2.317827697339059e-06,
+      "loss": 0.7488,
+      "step": 1514
+    },
+    {
+      "epoch": 0.36834427425237054,
+      "grad_norm": 29.75,
+      "learning_rate": 2.3175763972005277e-06,
+      "loss": 1.2496,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3685874057865305,
+      "grad_norm": 13.625,
+      "learning_rate": 2.3173249374942657e-06,
+      "loss": 0.8552,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3688305373206905,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.3170733182578586e-06,
+      "loss": 0.4992,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3690736688548505,
+      "grad_norm": 20.25,
+      "learning_rate": 2.3168215395289156e-06,
+      "loss": 1.0485,
+      "step": 1518
+    },
+    {
+      "epoch": 0.36931680038901044,
+      "grad_norm": 18.25,
+      "learning_rate": 2.3165696013450682e-06,
+      "loss": 0.7444,
+      "step": 1519
+    },
+    {
+      "epoch": 0.36955993192317044,
+      "grad_norm": 24.125,
+      "learning_rate": 2.316317503743974e-06,
+      "loss": 1.4502,
+      "step": 1520
+    },
+    {
+      "epoch": 0.36980306345733044,
+      "grad_norm": 21.125,
+      "learning_rate": 2.3160652467633127e-06,
+      "loss": 0.9617,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3700461949914904,
+      "grad_norm": 15.75,
+      "learning_rate": 2.315812830440789e-06,
+      "loss": 0.7754,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3702893265256504,
+      "grad_norm": 16.75,
+      "learning_rate": 2.3155602548141303e-06,
+      "loss": 0.7547,
+      "step": 1523
+    },
+    {
+      "epoch": 0.37053245805981033,
+      "grad_norm": 18.25,
+      "learning_rate": 2.3153075199210886e-06,
+      "loss": 0.8524,
+      "step": 1524
+    },
+    {
+      "epoch": 0.37077558959397033,
+      "grad_norm": 20.125,
+      "learning_rate": 2.3150546257994396e-06,
+      "loss": 0.9676,
+      "step": 1525
+    },
+    {
+      "epoch": 0.37101872112813034,
+      "grad_norm": 16.5,
+      "learning_rate": 2.314801572486983e-06,
+      "loss": 0.8736,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3712618526622903,
+      "grad_norm": 12.6875,
+      "learning_rate": 2.3145483600215414e-06,
+      "loss": 0.4003,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3715049841964503,
+      "grad_norm": 16.75,
+      "learning_rate": 2.3142949884409616e-06,
+      "loss": 0.6586,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3717481157306103,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.314041457783115e-06,
+      "loss": 0.4269,
+      "step": 1529
+    },
+    {
+      "epoch": 0.37199124726477023,
+      "grad_norm": 22.0,
+      "learning_rate": 2.313787768085896e-06,
+      "loss": 1.127,
+      "step": 1530
+    },
+    {
+      "epoch": 0.37223437879893023,
+      "grad_norm": 20.75,
+      "learning_rate": 2.313533919387223e-06,
+      "loss": 0.9467,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3724775103330902,
+      "grad_norm": 22.375,
+      "learning_rate": 2.3132799117250378e-06,
+      "loss": 0.9898,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3727206418672502,
+      "grad_norm": 23.875,
+      "learning_rate": 2.313025745137306e-06,
+      "loss": 1.0176,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3729637734014102,
+      "grad_norm": 17.75,
+      "learning_rate": 2.312771419662018e-06,
+      "loss": 0.4831,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3732069049355701,
+      "grad_norm": 24.75,
+      "learning_rate": 2.312516935337186e-06,
+      "loss": 0.9142,
+      "step": 1535
+    },
+    {
+      "epoch": 0.37345003646973013,
+      "grad_norm": 25.75,
+      "learning_rate": 2.312262292200848e-06,
+      "loss": 1.4025,
+      "step": 1536
+    },
+    {
+      "epoch": 0.37369316800389013,
+      "grad_norm": 22.25,
+      "learning_rate": 2.312007490291065e-06,
+      "loss": 1.1622,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3739362995380501,
+      "grad_norm": 24.25,
+      "learning_rate": 2.3117525296459203e-06,
+      "loss": 1.3053,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3741794310722101,
+      "grad_norm": 18.125,
+      "learning_rate": 2.3114974103035236e-06,
+      "loss": 0.9604,
+      "step": 1539
+    },
+    {
+      "epoch": 0.37442256260637,
+      "grad_norm": 16.125,
+      "learning_rate": 2.311242132302006e-06,
+      "loss": 0.8601,
+      "step": 1540
+    },
+    {
+      "epoch": 0.37466569414053,
+      "grad_norm": 18.5,
+      "learning_rate": 2.3109866956795234e-06,
+      "loss": 1.0024,
+      "step": 1541
+    },
+    {
+      "epoch": 0.37490882567469,
+      "grad_norm": 23.25,
+      "learning_rate": 2.310731100474255e-06,
+      "loss": 1.064,
+      "step": 1542
+    },
+    {
+      "epoch": 0.37515195720885,
+      "grad_norm": 22.25,
+      "learning_rate": 2.3104753467244045e-06,
+      "loss": 0.8009,
+      "step": 1543
+    },
+    {
+      "epoch": 0.37539508874301,
+      "grad_norm": 21.375,
+      "learning_rate": 2.310219434468198e-06,
+      "loss": 0.6426,
+      "step": 1544
+    },
+    {
+      "epoch": 0.37563822027717,
+      "grad_norm": 19.0,
+      "learning_rate": 2.309963363743887e-06,
+      "loss": 1.0863,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3758813518113299,
+      "grad_norm": 21.125,
+      "learning_rate": 2.309707134589745e-06,
+      "loss": 1.032,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3761244833454899,
+      "grad_norm": 19.125,
+      "learning_rate": 2.3094507470440697e-06,
+      "loss": 1.0391,
+      "step": 1547
+    },
+    {
+      "epoch": 0.37636761487964987,
+      "grad_norm": 21.75,
+      "learning_rate": 2.309194201145183e-06,
+      "loss": 0.7259,
+      "step": 1548
+    },
+    {
+      "epoch": 0.37661074641380987,
+      "grad_norm": 15.5,
+      "learning_rate": 2.3089374969314297e-06,
+      "loss": 0.5964,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3768538779479699,
+      "grad_norm": 21.5,
+      "learning_rate": 2.3086806344411795e-06,
+      "loss": 0.8881,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3770970094821298,
+      "grad_norm": 18.875,
+      "learning_rate": 2.308423613712824e-06,
+      "loss": 0.8602,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3773401410162898,
+      "grad_norm": 13.4375,
+      "learning_rate": 2.30816643478478e-06,
+      "loss": 0.6663,
+      "step": 1552
+    },
+    {
+      "epoch": 0.37758327255044977,
+      "grad_norm": 19.5,
+      "learning_rate": 2.307909097695487e-06,
+      "loss": 1.0368,
+      "step": 1553
+    },
+    {
+      "epoch": 0.37782640408460977,
+      "grad_norm": 17.375,
+      "learning_rate": 2.307651602483409e-06,
+      "loss": 0.955,
+      "step": 1554
+    },
+    {
+      "epoch": 0.37806953561876977,
+      "grad_norm": 23.125,
+      "learning_rate": 2.3073939491870326e-06,
+      "loss": 1.3472,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3783126671529297,
+      "grad_norm": 16.375,
+      "learning_rate": 2.307136137844869e-06,
+      "loss": 0.5445,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3785557986870897,
+      "grad_norm": 18.625,
+      "learning_rate": 2.3068781684954515e-06,
+      "loss": 1.0573,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3787989302212497,
+      "grad_norm": 18.375,
+      "learning_rate": 2.306620041177339e-06,
+      "loss": 1.0862,
+      "step": 1558
+    },
+    {
+      "epoch": 0.37904206175540966,
+      "grad_norm": 17.25,
+      "learning_rate": 2.306361755929113e-06,
+      "loss": 0.8455,
+      "step": 1559
+    },
+    {
+      "epoch": 0.37928519328956967,
+      "grad_norm": 16.625,
+      "learning_rate": 2.3061033127893788e-06,
+      "loss": 0.8344,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3795283248237296,
+      "grad_norm": 26.75,
+      "learning_rate": 2.3058447117967646e-06,
+      "loss": 1.2789,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3797714563578896,
+      "grad_norm": 22.375,
+      "learning_rate": 2.3055859529899235e-06,
+      "loss": 1.0004,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3800145878920496,
+      "grad_norm": 16.25,
+      "learning_rate": 2.305327036407531e-06,
+      "loss": 0.7607,
+      "step": 1563
+    },
+    {
+      "epoch": 0.38025771942620956,
+      "grad_norm": 22.625,
+      "learning_rate": 2.3050679620882865e-06,
+      "loss": 0.8534,
+      "step": 1564
+    },
+    {
+      "epoch": 0.38050085096036956,
+      "grad_norm": 132.0,
+      "learning_rate": 2.3048087300709137e-06,
+      "loss": 0.9035,
+      "step": 1565
+    },
+    {
+      "epoch": 0.38074398249452956,
+      "grad_norm": 19.625,
+      "learning_rate": 2.304549340394159e-06,
+      "loss": 0.7826,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3809871140286895,
+      "grad_norm": 15.0,
+      "learning_rate": 2.304289793096793e-06,
+      "loss": 0.5729,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3812302455628495,
+      "grad_norm": 18.25,
+      "learning_rate": 2.3040300882176084e-06,
+      "loss": 0.5804,
+      "step": 1568
+    },
+    {
+      "epoch": 0.38147337709700946,
+      "grad_norm": 27.0,
+      "learning_rate": 2.303770225795424e-06,
+      "loss": 1.1457,
+      "step": 1569
+    },
+    {
+      "epoch": 0.38171650863116946,
+      "grad_norm": 35.5,
+      "learning_rate": 2.30351020586908e-06,
+      "loss": 1.2839,
+      "step": 1570
+    },
+    {
+      "epoch": 0.38195964016532946,
+      "grad_norm": 21.5,
+      "learning_rate": 2.3032500284774407e-06,
+      "loss": 0.8459,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3822027716994894,
+      "grad_norm": 18.625,
+      "learning_rate": 2.302989693659395e-06,
+      "loss": 0.959,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3824459032336494,
+      "grad_norm": 17.375,
+      "learning_rate": 2.3027292014538533e-06,
+      "loss": 0.767,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3826890347678094,
+      "grad_norm": 20.5,
+      "learning_rate": 2.3024685518997514e-06,
+      "loss": 1.0778,
+      "step": 1574
+    },
+    {
+      "epoch": 0.38293216630196936,
+      "grad_norm": 28.125,
+      "learning_rate": 2.3022077450360474e-06,
+      "loss": 1.1952,
+      "step": 1575
+    },
+    {
+      "epoch": 0.38317529783612936,
+      "grad_norm": 17.125,
+      "learning_rate": 2.3019467809017235e-06,
+      "loss": 1.0121,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3834184293702893,
+      "grad_norm": 18.125,
+      "learning_rate": 2.301685659535786e-06,
+      "loss": 0.7615,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3836615609044493,
+      "grad_norm": 27.125,
+      "learning_rate": 2.301424380977263e-06,
+      "loss": 1.1592,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3839046924386093,
+      "grad_norm": 18.5,
+      "learning_rate": 2.301162945265208e-06,
+      "loss": 0.8433,
+      "step": 1579
+    },
+    {
+      "epoch": 0.38414782397276925,
+      "grad_norm": 21.125,
+      "learning_rate": 2.3009013524386963e-06,
+      "loss": 1.1105,
+      "step": 1580
+    },
+    {
+      "epoch": 0.38439095550692925,
+      "grad_norm": 16.75,
+      "learning_rate": 2.300639602536828e-06,
+      "loss": 0.7393,
+      "step": 1581
+    },
+    {
+      "epoch": 0.38463408704108926,
+      "grad_norm": 18.25,
+      "learning_rate": 2.3003776955987258e-06,
+      "loss": 0.6562,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3848772185752492,
+      "grad_norm": 18.0,
+      "learning_rate": 2.3001156316635362e-06,
+      "loss": 1.1729,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3851203501094092,
+      "grad_norm": 17.5,
+      "learning_rate": 2.2998534107704294e-06,
+      "loss": 0.7324,
+      "step": 1584
+    },
+    {
+      "epoch": 0.38536348164356915,
+      "grad_norm": 18.75,
+      "learning_rate": 2.2995910329585987e-06,
+      "loss": 0.9908,
+      "step": 1585
+    },
+    {
+      "epoch": 0.38560661317772915,
+      "grad_norm": 20.625,
+      "learning_rate": 2.2993284982672613e-06,
+      "loss": 0.6683,
+      "step": 1586
+    },
+    {
+      "epoch": 0.38584974471188915,
+      "grad_norm": 20.875,
+      "learning_rate": 2.2990658067356574e-06,
+      "loss": 0.9222,
+      "step": 1587
+    },
+    {
+      "epoch": 0.3860928762460491,
+      "grad_norm": 22.375,
+      "learning_rate": 2.2988029584030503e-06,
+      "loss": 0.7977,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3863360077802091,
+      "grad_norm": 17.375,
+      "learning_rate": 2.2985399533087275e-06,
+      "loss": 0.919,
+      "step": 1589
+    },
+    {
+      "epoch": 0.38657913931436905,
+      "grad_norm": 16.75,
+      "learning_rate": 2.2982767914920002e-06,
+      "loss": 0.9644,
+      "step": 1590
+    },
+    {
+      "epoch": 0.38682227084852905,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.2980134729922017e-06,
+      "loss": 0.9162,
+      "step": 1591
+    },
+    {
+      "epoch": 0.38706540238268905,
+      "grad_norm": 16.875,
+      "learning_rate": 2.29774999784869e-06,
+      "loss": 0.6784,
+      "step": 1592
+    },
+    {
+      "epoch": 0.387308533916849,
+      "grad_norm": 16.625,
+      "learning_rate": 2.2974863661008464e-06,
+      "loss": 0.9269,
+      "step": 1593
+    },
+    {
+      "epoch": 0.387551665451009,
+      "grad_norm": 19.375,
+      "learning_rate": 2.297222577788074e-06,
+      "loss": 1.1508,
+      "step": 1594
+    },
+    {
+      "epoch": 0.387794796985169,
+      "grad_norm": 16.5,
+      "learning_rate": 2.296958632949801e-06,
+      "loss": 0.7253,
+      "step": 1595
+    },
+    {
+      "epoch": 0.38803792851932895,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.296694531625479e-06,
+      "loss": 0.7891,
+      "step": 1596
+    },
+    {
+      "epoch": 0.38828106005348895,
+      "grad_norm": 17.625,
+      "learning_rate": 2.2964302738545823e-06,
+      "loss": 0.6991,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3885241915876489,
+      "grad_norm": 15.25,
+      "learning_rate": 2.2961658596766087e-06,
+      "loss": 0.9986,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3887673231218089,
+      "grad_norm": 19.625,
+      "learning_rate": 2.2959012891310794e-06,
+      "loss": 0.7878,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3890104546559689,
+      "grad_norm": 17.875,
+      "learning_rate": 2.2956365622575395e-06,
+      "loss": 0.9268,
+      "step": 1600
+    },
+    {
+      "epoch": 0.38925358619012884,
+      "grad_norm": 15.1875,
+      "learning_rate": 2.295371679095556e-06,
+      "loss": 0.7907,
+      "step": 1601
+    },
+    {
+      "epoch": 0.38949671772428884,
+      "grad_norm": 19.125,
+      "learning_rate": 2.295106639684721e-06,
+      "loss": 1.0365,
+      "step": 1602
+    },
+    {
+      "epoch": 0.38973984925844884,
+      "grad_norm": 20.375,
+      "learning_rate": 2.294841444064649e-06,
+      "loss": 0.909,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3899829807926088,
+      "grad_norm": 20.0,
+      "learning_rate": 2.2945760922749783e-06,
+      "loss": 0.8828,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3902261123267688,
+      "grad_norm": 18.25,
+      "learning_rate": 2.29431058435537e-06,
+      "loss": 0.7795,
+      "step": 1605
+    },
+    {
+      "epoch": 0.39046924386092874,
+      "grad_norm": 16.625,
+      "learning_rate": 2.2940449203455097e-06,
+      "loss": 0.5941,
+      "step": 1606
+    },
+    {
+      "epoch": 0.39071237539508874,
+      "grad_norm": 16.625,
+      "learning_rate": 2.293779100285104e-06,
+      "loss": 0.6087,
+      "step": 1607
+    },
+    {
+      "epoch": 0.39095550692924874,
+      "grad_norm": 23.5,
+      "learning_rate": 2.2935131242138855e-06,
+      "loss": 0.9364,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3911986384634087,
+      "grad_norm": 19.75,
+      "learning_rate": 2.293246992171608e-06,
+      "loss": 1.2017,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3914417699975687,
+      "grad_norm": 22.75,
+      "learning_rate": 2.2929807041980505e-06,
+      "loss": 0.9784,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3916849015317287,
+      "grad_norm": 17.75,
+      "learning_rate": 2.2927142603330137e-06,
+      "loss": 0.5636,
+      "step": 1611
+    },
+    {
+      "epoch": 0.39192803306588864,
+      "grad_norm": 19.75,
+      "learning_rate": 2.2924476606163223e-06,
+      "loss": 1.1359,
+      "step": 1612
+    },
+    {
+      "epoch": 0.39217116460004864,
+      "grad_norm": 12.4375,
+      "learning_rate": 2.2921809050878245e-06,
+      "loss": 0.3779,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3924142961342086,
+      "grad_norm": 16.25,
+      "learning_rate": 2.2919139937873915e-06,
+      "loss": 0.8724,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3926574276683686,
+      "grad_norm": 17.875,
+      "learning_rate": 2.291646926754917e-06,
+      "loss": 0.8597,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3929005592025286,
+      "grad_norm": 19.0,
+      "learning_rate": 2.29137970403032e-06,
+      "loss": 0.8451,
+      "step": 1616
+    },
+    {
+      "epoch": 0.39314369073668853,
+      "grad_norm": 19.5,
+      "learning_rate": 2.2911123256535407e-06,
+      "loss": 0.895,
+      "step": 1617
+    },
+    {
+      "epoch": 0.39338682227084854,
+      "grad_norm": 22.25,
+      "learning_rate": 2.2908447916645436e-06,
+      "loss": 1.1503,
+      "step": 1618
+    },
+    {
+      "epoch": 0.39362995380500854,
+      "grad_norm": 20.75,
+      "learning_rate": 2.2905771021033167e-06,
+      "loss": 0.9719,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3938730853391685,
+      "grad_norm": 26.25,
+      "learning_rate": 2.290309257009871e-06,
+      "loss": 0.9519,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3941162168733285,
+      "grad_norm": 19.0,
+      "learning_rate": 2.290041256424239e-06,
+      "loss": 1.1053,
+      "step": 1621
+    },
+    {
+      "epoch": 0.39435934840748843,
+      "grad_norm": 19.75,
+      "learning_rate": 2.2897731003864794e-06,
+      "loss": 0.975,
+      "step": 1622
+    },
+    {
+      "epoch": 0.39460247994164843,
+      "grad_norm": 18.0,
+      "learning_rate": 2.289504788936673e-06,
+      "loss": 0.7324,
+      "step": 1623
+    },
+    {
+      "epoch": 0.39484561147580843,
+      "grad_norm": 48.25,
+      "learning_rate": 2.2892363221149223e-06,
+      "loss": 1.228,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3950887430099684,
+      "grad_norm": 19.5,
+      "learning_rate": 2.288967699961355e-06,
+      "loss": 1.047,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3953318745441284,
+      "grad_norm": 22.0,
+      "learning_rate": 2.288698922516122e-06,
+      "loss": 1.057,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3955750060782883,
+      "grad_norm": 23.75,
+      "learning_rate": 2.2884299898193958e-06,
+      "loss": 0.9644,
+      "step": 1627
+    },
+    {
+      "epoch": 0.39581813761244833,
+      "grad_norm": 21.75,
+      "learning_rate": 2.2881609019113735e-06,
+      "loss": 1.0534,
+      "step": 1628
+    },
+    {
+      "epoch": 0.39606126914660833,
+      "grad_norm": 18.625,
+      "learning_rate": 2.2878916588322744e-06,
+      "loss": 0.875,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3963044006807683,
+      "grad_norm": 18.625,
+      "learning_rate": 2.287622260622342e-06,
+      "loss": 0.9888,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3965475322149283,
+      "grad_norm": 20.625,
+      "learning_rate": 2.2873527073218424e-06,
+      "loss": 0.6882,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3967906637490883,
+      "grad_norm": 20.875,
+      "learning_rate": 2.2870829989710653e-06,
+      "loss": 1.2406,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3970337952832482,
+      "grad_norm": 19.25,
+      "learning_rate": 2.2868131356103226e-06,
+      "loss": 0.9361,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3972769268174082,
+      "grad_norm": 17.875,
+      "learning_rate": 2.2865431172799504e-06,
+      "loss": 0.8386,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3975200583515682,
+      "grad_norm": 21.5,
+      "learning_rate": 2.2862729440203078e-06,
+      "loss": 1.2811,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3977631898857282,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.286002615871776e-06,
+      "loss": 0.7588,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3980063214198882,
+      "grad_norm": 20.25,
+      "learning_rate": 2.2857321328747615e-06,
+      "loss": 0.9633,
+      "step": 1637
+    },
+    {
+      "epoch": 0.3982494529540481,
+      "grad_norm": 18.875,
+      "learning_rate": 2.2854614950696923e-06,
+      "loss": 0.5652,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3984925844882081,
+      "grad_norm": 19.375,
+      "learning_rate": 2.2851907024970196e-06,
+      "loss": 0.9574,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3987357160223681,
+      "grad_norm": 24.25,
+      "learning_rate": 2.2849197551972173e-06,
+      "loss": 1.315,
+      "step": 1640
+    },
+    {
+      "epoch": 0.39897884755652807,
+      "grad_norm": 16.625,
+      "learning_rate": 2.284648653210784e-06,
+      "loss": 1.0215,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3992219790906881,
+      "grad_norm": 20.0,
+      "learning_rate": 2.2843773965782413e-06,
+      "loss": 0.922,
+      "step": 1642
+    },
+    {
+      "epoch": 0.399465110624848,
+      "grad_norm": 23.75,
+      "learning_rate": 2.2841059853401315e-06,
+      "loss": 0.8942,
+      "step": 1643
+    },
+    {
+      "epoch": 0.399708242159008,
+      "grad_norm": 20.625,
+      "learning_rate": 2.283834419537023e-06,
+      "loss": 1.0954,
+      "step": 1644
+    },
+    {
+      "epoch": 0.399951373693168,
+      "grad_norm": 20.125,
+      "learning_rate": 2.2835626992095055e-06,
+      "loss": 0.8404,
+      "step": 1645
+    },
+    {
+      "epoch": 0.40019450522732797,
+      "grad_norm": 20.75,
+      "learning_rate": 2.2832908243981923e-06,
+      "loss": 0.9693,
+      "step": 1646
+    },
+    {
+      "epoch": 0.40043763676148797,
+      "grad_norm": 22.625,
+      "learning_rate": 2.28301879514372e-06,
+      "loss": 0.9124,
+      "step": 1647
+    },
+    {
+      "epoch": 0.40068076829564797,
+      "grad_norm": 19.5,
+      "learning_rate": 2.282746611486748e-06,
+      "loss": 0.8976,
+      "step": 1648
+    },
+    {
+      "epoch": 0.4009238998298079,
+      "grad_norm": 21.0,
+      "learning_rate": 2.2824742734679585e-06,
+      "loss": 1.1095,
+      "step": 1649
+    },
+    {
+      "epoch": 0.4011670313639679,
+      "grad_norm": 28.625,
+      "learning_rate": 2.2822017811280573e-06,
+      "loss": 1.1894,
+      "step": 1650
+    },
+    {
+      "epoch": 0.40141016289812786,
+      "grad_norm": 17.0,
+      "learning_rate": 2.281929134507773e-06,
+      "loss": 0.6871,
+      "step": 1651
+    },
+    {
+      "epoch": 0.40165329443228787,
+      "grad_norm": 17.5,
+      "learning_rate": 2.2816563336478582e-06,
+      "loss": 0.61,
+      "step": 1652
+    },
+    {
+      "epoch": 0.40189642596644787,
+      "grad_norm": 33.25,
+      "learning_rate": 2.2813833785890864e-06,
+      "loss": 1.0288,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4021395575006078,
+      "grad_norm": 43.5,
+      "learning_rate": 2.2811102693722565e-06,
+      "loss": 1.2935,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4023826890347678,
+      "grad_norm": 22.25,
+      "learning_rate": 2.280837006038189e-06,
+      "loss": 1.3191,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4026258205689278,
+      "grad_norm": 12.9375,
+      "learning_rate": 2.280563588627727e-06,
+      "loss": 0.4425,
+      "step": 1656
+    },
+    {
+      "epoch": 0.40286895210308776,
+      "grad_norm": 16.25,
+      "learning_rate": 2.280290017181739e-06,
+      "loss": 0.7694,
+      "step": 1657
+    },
+    {
+      "epoch": 0.40311208363724776,
+      "grad_norm": 16.375,
+      "learning_rate": 2.280016291741114e-06,
+      "loss": 0.953,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4033552151714077,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.2797424123467656e-06,
+      "loss": 0.6801,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4035983467055677,
+      "grad_norm": 21.5,
+      "learning_rate": 2.279468379039629e-06,
+      "loss": 0.6453,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4038414782397277,
+      "grad_norm": 17.25,
+      "learning_rate": 2.279194191860663e-06,
+      "loss": 0.9078,
+      "step": 1661
+    },
+    {
+      "epoch": 0.40408460977388766,
+      "grad_norm": 18.125,
+      "learning_rate": 2.278919850850851e-06,
+      "loss": 0.8203,
+      "step": 1662
+    },
+    {
+      "epoch": 0.40432774130804766,
+      "grad_norm": 16.5,
+      "learning_rate": 2.2786453560511975e-06,
+      "loss": 0.537,
+      "step": 1663
+    },
+    {
+      "epoch": 0.4045708728422076,
+      "grad_norm": 18.625,
+      "learning_rate": 2.2783707075027295e-06,
+      "loss": 0.7241,
+      "step": 1664
+    },
+    {
+      "epoch": 0.4048140043763676,
+      "grad_norm": 18.25,
+      "learning_rate": 2.278095905246499e-06,
+      "loss": 0.9083,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4050571359105276,
+      "grad_norm": 20.0,
+      "learning_rate": 2.2778209493235794e-06,
+      "loss": 0.7556,
+      "step": 1666
+    },
+    {
+      "epoch": 0.40530026744468756,
+      "grad_norm": 18.375,
+      "learning_rate": 2.277545839775068e-06,
+      "loss": 0.8547,
+      "step": 1667
+    },
+    {
+      "epoch": 0.40554339897884756,
+      "grad_norm": 19.125,
+      "learning_rate": 2.277270576642084e-06,
+      "loss": 0.9679,
+      "step": 1668
+    },
+    {
+      "epoch": 0.40578653051300756,
+      "grad_norm": 16.625,
+      "learning_rate": 2.276995159965772e-06,
+      "loss": 0.6753,
+      "step": 1669
+    },
+    {
+      "epoch": 0.4060296620471675,
+      "grad_norm": 26.5,
+      "learning_rate": 2.2767195897872955e-06,
+      "loss": 0.552,
+      "step": 1670
+    },
+    {
+      "epoch": 0.4062727935813275,
+      "grad_norm": 35.25,
+      "learning_rate": 2.276443866147845e-06,
+      "loss": 1.304,
+      "step": 1671
+    },
+    {
+      "epoch": 0.40651592511548745,
+      "grad_norm": 17.25,
+      "learning_rate": 2.2761679890886307e-06,
+      "loss": 0.7139,
+      "step": 1672
+    },
+    {
+      "epoch": 0.40675905664964745,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.275891958650888e-06,
+      "loss": 0.6474,
+      "step": 1673
+    },
+    {
+      "epoch": 0.40700218818380746,
+      "grad_norm": 17.25,
+      "learning_rate": 2.2756157748758745e-06,
+      "loss": 1.0807,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4072453197179674,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.2753394378048705e-06,
+      "loss": 0.5286,
+      "step": 1675
+    },
+    {
+      "epoch": 0.4074884512521274,
+      "grad_norm": 19.625,
+      "learning_rate": 2.2750629474791792e-06,
+      "loss": 0.8319,
+      "step": 1676
+    },
+    {
+      "epoch": 0.4077315827862874,
+      "grad_norm": 25.5,
+      "learning_rate": 2.2747863039401267e-06,
+      "loss": 0.8369,
+      "step": 1677
+    },
+    {
+      "epoch": 0.40797471432044735,
+      "grad_norm": 19.75,
+      "learning_rate": 2.274509507229063e-06,
+      "loss": 0.6813,
+      "step": 1678
+    },
+    {
+      "epoch": 0.40821784585460735,
+      "grad_norm": 20.0,
+      "learning_rate": 2.274232557387359e-06,
+      "loss": 1.2366,
+      "step": 1679
+    },
+    {
+      "epoch": 0.4084609773887673,
+      "grad_norm": 19.125,
+      "learning_rate": 2.2739554544564107e-06,
+      "loss": 0.8795,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4087041089229273,
+      "grad_norm": 16.0,
+      "learning_rate": 2.2736781984776354e-06,
+      "loss": 0.7821,
+      "step": 1681
+    },
+    {
+      "epoch": 0.4089472404570873,
+      "grad_norm": 18.0,
+      "learning_rate": 2.273400789492473e-06,
+      "loss": 0.8354,
+      "step": 1682
+    },
+    {
+      "epoch": 0.40919037199124725,
+      "grad_norm": 17.375,
+      "learning_rate": 2.2731232275423886e-06,
+      "loss": 0.8526,
+      "step": 1683
+    },
+    {
+      "epoch": 0.40943350352540725,
+      "grad_norm": 27.75,
+      "learning_rate": 2.272845512668868e-06,
+      "loss": 1.19,
+      "step": 1684
+    },
+    {
+      "epoch": 0.40967663505956725,
+      "grad_norm": 18.75,
+      "learning_rate": 2.27256764491342e-06,
+      "loss": 1.0011,
+      "step": 1685
+    },
+    {
+      "epoch": 0.4099197665937272,
+      "grad_norm": 20.875,
+      "learning_rate": 2.2722896243175767e-06,
+      "loss": 0.9363,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4101628981278872,
+      "grad_norm": 21.875,
+      "learning_rate": 2.272011450922894e-06,
+      "loss": 0.6616,
+      "step": 1687
+    },
+    {
+      "epoch": 0.41040602966204714,
+      "grad_norm": 24.625,
+      "learning_rate": 2.2717331247709496e-06,
+      "loss": 0.9938,
+      "step": 1688
+    },
+    {
+      "epoch": 0.41064916119620715,
+      "grad_norm": 19.375,
+      "learning_rate": 2.271454645903343e-06,
+      "loss": 0.7729,
+      "step": 1689
+    },
+    {
+      "epoch": 0.41089229273036715,
+      "grad_norm": 18.375,
+      "learning_rate": 2.271176014361699e-06,
+      "loss": 1.2154,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4111354242645271,
+      "grad_norm": 20.625,
+      "learning_rate": 2.270897230187663e-06,
+      "loss": 0.8767,
+      "step": 1691
+    },
+    {
+      "epoch": 0.4113785557986871,
+      "grad_norm": 17.625,
+      "learning_rate": 2.270618293422905e-06,
+      "loss": 0.8918,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4116216873328471,
+      "grad_norm": 17.25,
+      "learning_rate": 2.2703392041091156e-06,
+      "loss": 0.7898,
+      "step": 1693
+    },
+    {
+      "epoch": 0.41186481886700704,
+      "grad_norm": 21.0,
+      "learning_rate": 2.2700599622880106e-06,
+      "loss": 1.0549,
+      "step": 1694
+    },
+    {
+      "epoch": 0.41210795040116704,
+      "grad_norm": 22.625,
+      "learning_rate": 2.2697805680013274e-06,
+      "loss": 0.7342,
+      "step": 1695
+    },
+    {
+      "epoch": 0.412351081935327,
+      "grad_norm": 20.375,
+      "learning_rate": 2.2695010212908256e-06,
+      "loss": 1.0939,
+      "step": 1696
+    },
+    {
+      "epoch": 0.412594213469487,
+      "grad_norm": 22.5,
+      "learning_rate": 2.269221322198289e-06,
+      "loss": 0.7617,
+      "step": 1697
+    },
+    {
+      "epoch": 0.412837345003647,
+      "grad_norm": 20.125,
+      "learning_rate": 2.2689414707655233e-06,
+      "loss": 1.1387,
+      "step": 1698
+    },
+    {
+      "epoch": 0.41308047653780694,
+      "grad_norm": 18.125,
+      "learning_rate": 2.268661467034357e-06,
+      "loss": 0.7587,
+      "step": 1699
+    },
+    {
+      "epoch": 0.41332360807196694,
+      "grad_norm": 16.375,
+      "learning_rate": 2.2683813110466417e-06,
+      "loss": 0.7771,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4135667396061269,
+      "grad_norm": 20.875,
+      "learning_rate": 2.2681010028442517e-06,
+      "loss": 0.9262,
+      "step": 1701
+    },
+    {
+      "epoch": 0.4138098711402869,
+      "grad_norm": 20.75,
+      "learning_rate": 2.267820542469083e-06,
+      "loss": 1.1631,
+      "step": 1702
+    },
+    {
+      "epoch": 0.4140530026744469,
+      "grad_norm": 16.625,
+      "learning_rate": 2.2675399299630563e-06,
+      "loss": 0.45,
+      "step": 1703
+    },
+    {
+      "epoch": 0.41429613420860684,
+      "grad_norm": 21.125,
+      "learning_rate": 2.267259165368113e-06,
+      "loss": 0.96,
+      "step": 1704
+    },
+    {
+      "epoch": 0.41453926574276684,
+      "grad_norm": 20.375,
+      "learning_rate": 2.2669782487262193e-06,
+      "loss": 0.9909,
+      "step": 1705
+    },
+    {
+      "epoch": 0.41478239727692684,
+      "grad_norm": 17.125,
+      "learning_rate": 2.2666971800793625e-06,
+      "loss": 0.535,
+      "step": 1706
+    },
+    {
+      "epoch": 0.4150255288110868,
+      "grad_norm": 17.375,
+      "learning_rate": 2.2664159594695527e-06,
+      "loss": 0.8767,
+      "step": 1707
+    },
+    {
+      "epoch": 0.4152686603452468,
+      "grad_norm": 18.125,
+      "learning_rate": 2.266134586938824e-06,
+      "loss": 0.6714,
+      "step": 1708
+    },
+    {
+      "epoch": 0.41551179187940673,
+      "grad_norm": 21.625,
+      "learning_rate": 2.265853062529232e-06,
+      "loss": 0.9772,
+      "step": 1709
+    },
+    {
+      "epoch": 0.41575492341356673,
+      "grad_norm": 18.25,
+      "learning_rate": 2.2655713862828554e-06,
+      "loss": 0.8306,
+      "step": 1710
+    },
+    {
+      "epoch": 0.41599805494772674,
+      "grad_norm": 15.375,
+      "learning_rate": 2.2652895582417955e-06,
+      "loss": 0.6248,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4162411864818867,
+      "grad_norm": 20.375,
+      "learning_rate": 2.2650075784481767e-06,
+      "loss": 1.3364,
+      "step": 1712
+    },
+    {
+      "epoch": 0.4164843180160467,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.2647254469441456e-06,
+      "loss": 0.7473,
+      "step": 1713
+    },
+    {
+      "epoch": 0.4167274495502067,
+      "grad_norm": 15.625,
+      "learning_rate": 2.2644431637718713e-06,
+      "loss": 0.5194,
+      "step": 1714
+    },
+    {
+      "epoch": 0.41697058108436663,
+      "grad_norm": 17.25,
+      "learning_rate": 2.2641607289735455e-06,
+      "loss": 1.151,
+      "step": 1715
+    },
+    {
+      "epoch": 0.41721371261852663,
+      "grad_norm": 21.125,
+      "learning_rate": 2.2638781425913846e-06,
+      "loss": 0.9437,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4174568441526866,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.2635954046676247e-06,
+      "loss": 0.7822,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4176999756868466,
+      "grad_norm": 23.0,
+      "learning_rate": 2.263312515244526e-06,
+      "loss": 1.0253,
+      "step": 1718
+    },
+    {
+      "epoch": 0.4179431072210066,
+      "grad_norm": 22.25,
+      "learning_rate": 2.2630294743643717e-06,
+      "loss": 1.1386,
+      "step": 1719
+    },
+    {
+      "epoch": 0.41818623875516653,
+      "grad_norm": 19.5,
+      "learning_rate": 2.2627462820694664e-06,
+      "loss": 1.0932,
+      "step": 1720
+    },
+    {
+      "epoch": 0.41842937028932653,
+      "grad_norm": 17.75,
+      "learning_rate": 2.2624629384021386e-06,
+      "loss": 0.7368,
+      "step": 1721
+    },
+    {
+      "epoch": 0.41867250182348653,
+      "grad_norm": 18.5,
+      "learning_rate": 2.262179443404739e-06,
+      "loss": 0.9348,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4189156333576465,
+      "grad_norm": 18.125,
+      "learning_rate": 2.2618957971196402e-06,
+      "loss": 0.8927,
+      "step": 1723
+    },
+    {
+      "epoch": 0.4191587648918065,
+      "grad_norm": 16.625,
+      "learning_rate": 2.2616119995892394e-06,
+      "loss": 0.7519,
+      "step": 1724
+    },
+    {
+      "epoch": 0.4194018964259664,
+      "grad_norm": 25.375,
+      "learning_rate": 2.2613280508559536e-06,
+      "loss": 1.2584,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4196450279601264,
+      "grad_norm": 22.5,
+      "learning_rate": 2.261043950962224e-06,
+      "loss": 0.8007,
+      "step": 1726
+    },
+    {
+      "epoch": 0.41988815949428643,
+      "grad_norm": 18.625,
+      "learning_rate": 2.260759699950515e-06,
+      "loss": 0.7818,
+      "step": 1727
+    },
+    {
+      "epoch": 0.4201312910284464,
+      "grad_norm": 35.25,
+      "learning_rate": 2.2604752978633124e-06,
+      "loss": 0.8461,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4203744225626064,
+      "grad_norm": 24.0,
+      "learning_rate": 2.2601907447431247e-06,
+      "loss": 1.2568,
+      "step": 1729
+    },
+    {
+      "epoch": 0.4206175540967664,
+      "grad_norm": 68.0,
+      "learning_rate": 2.2599060406324842e-06,
+      "loss": 1.2069,
+      "step": 1730
+    },
+    {
+      "epoch": 0.4208606856309263,
+      "grad_norm": 21.0,
+      "learning_rate": 2.259621185573944e-06,
+      "loss": 0.9982,
+      "step": 1731
+    },
+    {
+      "epoch": 0.4211038171650863,
+      "grad_norm": 29.75,
+      "learning_rate": 2.2593361796100803e-06,
+      "loss": 1.2981,
+      "step": 1732
+    },
+    {
+      "epoch": 0.42134694869924627,
+      "grad_norm": 22.125,
+      "learning_rate": 2.2590510227834937e-06,
+      "loss": 1.6931,
+      "step": 1733
+    },
+    {
+      "epoch": 0.4215900802334063,
+      "grad_norm": 21.375,
+      "learning_rate": 2.2587657151368044e-06,
+      "loss": 0.766,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4218332117675663,
+      "grad_norm": 27.375,
+      "learning_rate": 2.2584802567126567e-06,
+      "loss": 0.9785,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4220763433017262,
+      "grad_norm": 22.5,
+      "learning_rate": 2.2581946475537177e-06,
+      "loss": 1.2253,
+      "step": 1736
+    },
+    {
+      "epoch": 0.4223194748358862,
+      "grad_norm": 21.625,
+      "learning_rate": 2.2579088877026767e-06,
+      "loss": 0.7935,
+      "step": 1737
+    },
+    {
+      "epoch": 0.42256260637004617,
+      "grad_norm": 20.25,
+      "learning_rate": 2.2576229772022452e-06,
+      "loss": 0.8855,
+      "step": 1738
+    },
+    {
+      "epoch": 0.42280573790420617,
+      "grad_norm": 18.125,
+      "learning_rate": 2.2573369160951574e-06,
+      "loss": 0.9769,
+      "step": 1739
+    },
+    {
+      "epoch": 0.42304886943836617,
+      "grad_norm": 22.75,
+      "learning_rate": 2.2570507044241702e-06,
+      "loss": 0.9223,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4232920009725261,
+      "grad_norm": 17.0,
+      "learning_rate": 2.256764342232063e-06,
+      "loss": 0.6006,
+      "step": 1741
+    },
+    {
+      "epoch": 0.4235351325066861,
+      "grad_norm": 25.5,
+      "learning_rate": 2.2564778295616373e-06,
+      "loss": 1.0001,
+      "step": 1742
+    },
+    {
+      "epoch": 0.4237782640408461,
+      "grad_norm": 20.375,
+      "learning_rate": 2.2561911664557173e-06,
+      "loss": 0.7565,
+      "step": 1743
+    },
+    {
+      "epoch": 0.42402139557500607,
+      "grad_norm": 18.625,
+      "learning_rate": 2.25590435295715e-06,
+      "loss": 0.4233,
+      "step": 1744
+    },
+    {
+      "epoch": 0.42426452710916607,
+      "grad_norm": 21.75,
+      "learning_rate": 2.2556173891088047e-06,
+      "loss": 0.7461,
+      "step": 1745
+    },
+    {
+      "epoch": 0.424507658643326,
+      "grad_norm": 21.5,
+      "learning_rate": 2.2553302749535733e-06,
+      "loss": 0.7831,
+      "step": 1746
+    },
+    {
+      "epoch": 0.424750790177486,
+      "grad_norm": 40.25,
+      "learning_rate": 2.255043010534369e-06,
+      "loss": 0.9118,
+      "step": 1747
+    },
+    {
+      "epoch": 0.424993921711646,
+      "grad_norm": 23.5,
+      "learning_rate": 2.2547555958941296e-06,
+      "loss": 0.9295,
+      "step": 1748
+    },
+    {
+      "epoch": 0.42523705324580596,
+      "grad_norm": 20.625,
+      "learning_rate": 2.2544680310758136e-06,
+      "loss": 0.8796,
+      "step": 1749
+    },
+    {
+      "epoch": 0.42548018477996596,
+      "grad_norm": 16.125,
+      "learning_rate": 2.2541803161224023e-06,
+      "loss": 0.7127,
+      "step": 1750
+    },
+    {
+      "epoch": 0.42572331631412597,
+      "grad_norm": 18.5,
+      "learning_rate": 2.2538924510769004e-06,
+      "loss": 0.6023,
+      "step": 1751
+    },
+    {
+      "epoch": 0.4259664478482859,
+      "grad_norm": 17.5,
+      "learning_rate": 2.2536044359823338e-06,
+      "loss": 0.9475,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4262095793824459,
+      "grad_norm": 25.25,
+      "learning_rate": 2.253316270881751e-06,
+      "loss": 0.909,
+      "step": 1753
+    },
+    {
+      "epoch": 0.42645271091660586,
+      "grad_norm": 24.375,
+      "learning_rate": 2.253027955818224e-06,
+      "loss": 1.1146,
+      "step": 1754
+    },
+    {
+      "epoch": 0.42669584245076586,
+      "grad_norm": 18.0,
+      "learning_rate": 2.252739490834846e-06,
+      "loss": 0.6651,
+      "step": 1755
+    },
+    {
+      "epoch": 0.42693897398492586,
+      "grad_norm": 20.5,
+      "learning_rate": 2.252450875974733e-06,
+      "loss": 1.4401,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4271821055190858,
+      "grad_norm": 21.625,
+      "learning_rate": 2.2521621112810236e-06,
+      "loss": 0.8544,
+      "step": 1757
+    },
+    {
+      "epoch": 0.4274252370532458,
+      "grad_norm": 16.75,
+      "learning_rate": 2.2518731967968794e-06,
+      "loss": 0.8524,
+      "step": 1758
+    },
+    {
+      "epoch": 0.4276683685874058,
+      "grad_norm": 19.5,
+      "learning_rate": 2.2515841325654824e-06,
+      "loss": 0.7561,
+      "step": 1759
+    },
+    {
+      "epoch": 0.42791150012156576,
+      "grad_norm": 20.625,
+      "learning_rate": 2.251294918630039e-06,
+      "loss": 0.7684,
+      "step": 1760
+    },
+    {
+      "epoch": 0.42815463165572576,
+      "grad_norm": 35.25,
+      "learning_rate": 2.251005555033777e-06,
+      "loss": 1.1054,
+      "step": 1761
+    },
+    {
+      "epoch": 0.4283977631898857,
+      "grad_norm": 20.0,
+      "learning_rate": 2.250716041819947e-06,
+      "loss": 1.2366,
+      "step": 1762
+    },
+    {
+      "epoch": 0.4286408947240457,
+      "grad_norm": 15.0,
+      "learning_rate": 2.2504263790318215e-06,
+      "loss": 0.6197,
+      "step": 1763
+    },
+    {
+      "epoch": 0.4288840262582057,
+      "grad_norm": 20.5,
+      "learning_rate": 2.2501365667126954e-06,
+      "loss": 1.1374,
+      "step": 1764
+    },
+    {
+      "epoch": 0.42912715779236565,
+      "grad_norm": 24.75,
+      "learning_rate": 2.2498466049058866e-06,
+      "loss": 0.9917,
+      "step": 1765
+    },
+    {
+      "epoch": 0.42937028932652566,
+      "grad_norm": 22.125,
+      "learning_rate": 2.249556493654735e-06,
+      "loss": 1.0051,
+      "step": 1766
+    },
+    {
+      "epoch": 0.42961342086068566,
+      "grad_norm": 27.125,
+      "learning_rate": 2.249266233002602e-06,
+      "loss": 0.9072,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4298565523948456,
+      "grad_norm": 16.625,
+      "learning_rate": 2.248975822992873e-06,
+      "loss": 0.8493,
+      "step": 1768
+    },
+    {
+      "epoch": 0.4300996839290056,
+      "grad_norm": 16.5,
+      "learning_rate": 2.248685263668954e-06,
+      "loss": 0.7018,
+      "step": 1769
+    },
+    {
+      "epoch": 0.43034281546316555,
+      "grad_norm": 44.75,
+      "learning_rate": 2.248394555074275e-06,
+      "loss": 0.9015,
+      "step": 1770
+    },
+    {
+      "epoch": 0.43058594699732555,
+      "grad_norm": 17.25,
+      "learning_rate": 2.248103697252287e-06,
+      "loss": 0.9825,
+      "step": 1771
+    },
+    {
+      "epoch": 0.43082907853148555,
+      "grad_norm": 17.25,
+      "learning_rate": 2.247812690246463e-06,
+      "loss": 0.7986,
+      "step": 1772
+    },
+    {
+      "epoch": 0.4310722100656455,
+      "grad_norm": 21.125,
+      "learning_rate": 2.2475215341002998e-06,
+      "loss": 0.8086,
+      "step": 1773
+    },
+    {
+      "epoch": 0.4313153415998055,
+      "grad_norm": 15.5,
+      "learning_rate": 2.2472302288573153e-06,
+      "loss": 0.6696,
+      "step": 1774
+    },
+    {
+      "epoch": 0.43155847313396545,
+      "grad_norm": 21.75,
+      "learning_rate": 2.2469387745610504e-06,
+      "loss": 1.0677,
+      "step": 1775
+    },
+    {
+      "epoch": 0.43180160466812545,
+      "grad_norm": 22.375,
+      "learning_rate": 2.2466471712550682e-06,
+      "loss": 0.8187,
+      "step": 1776
+    },
+    {
+      "epoch": 0.43204473620228545,
+      "grad_norm": 18.0,
+      "learning_rate": 2.2463554189829534e-06,
+      "loss": 0.632,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4322878677364454,
+      "grad_norm": 24.125,
+      "learning_rate": 2.2460635177883137e-06,
+      "loss": 0.8745,
+      "step": 1778
+    },
+    {
+      "epoch": 0.4325309992706054,
+      "grad_norm": 20.375,
+      "learning_rate": 2.2457714677147786e-06,
+      "loss": 0.9126,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4327741308047654,
+      "grad_norm": 16.125,
+      "learning_rate": 2.2454792688060002e-06,
+      "loss": 0.6082,
+      "step": 1780
+    },
+    {
+      "epoch": 0.43301726233892535,
+      "grad_norm": 17.375,
+      "learning_rate": 2.245186921105652e-06,
+      "loss": 0.6234,
+      "step": 1781
+    },
+    {
+      "epoch": 0.43326039387308535,
+      "grad_norm": 32.25,
+      "learning_rate": 2.2448944246574314e-06,
+      "loss": 1.3383,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4335035254072453,
+      "grad_norm": 22.0,
+      "learning_rate": 2.2446017795050564e-06,
+      "loss": 0.7792,
+      "step": 1783
+    },
+    {
+      "epoch": 0.4337466569414053,
+      "grad_norm": 23.0,
+      "learning_rate": 2.2443089856922683e-06,
+      "loss": 1.1139,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4339897884755653,
+      "grad_norm": 28.0,
+      "learning_rate": 2.24401604326283e-06,
+      "loss": 1.2003,
+      "step": 1785
+    },
+    {
+      "epoch": 0.43423292000972524,
+      "grad_norm": 20.25,
+      "learning_rate": 2.243722952260527e-06,
+      "loss": 1.0905,
+      "step": 1786
+    },
+    {
+      "epoch": 0.43447605154388524,
+      "grad_norm": 16.5,
+      "learning_rate": 2.243429712729166e-06,
+      "loss": 0.6541,
+      "step": 1787
+    },
+    {
+      "epoch": 0.43471918307804525,
+      "grad_norm": 24.0,
+      "learning_rate": 2.2431363247125777e-06,
+      "loss": 0.5291,
+      "step": 1788
+    },
+    {
+      "epoch": 0.4349623146122052,
+      "grad_norm": 30.375,
+      "learning_rate": 2.2428427882546136e-06,
+      "loss": 1.2832,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4352054461463652,
+      "grad_norm": 16.5,
+      "learning_rate": 2.2425491033991474e-06,
+      "loss": 0.8466,
+      "step": 1790
+    },
+    {
+      "epoch": 0.43544857768052514,
+      "grad_norm": 31.375,
+      "learning_rate": 2.242255270190076e-06,
+      "loss": 0.9529,
+      "step": 1791
+    },
+    {
+      "epoch": 0.43569170921468514,
+      "grad_norm": 19.75,
+      "learning_rate": 2.241961288671318e-06,
+      "loss": 0.752,
+      "step": 1792
+    },
+    {
+      "epoch": 0.43593484074884514,
+      "grad_norm": 22.0,
+      "learning_rate": 2.2416671588868136e-06,
+      "loss": 0.766,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4361779722830051,
+      "grad_norm": 16.875,
+      "learning_rate": 2.2413728808805256e-06,
+      "loss": 1.022,
+      "step": 1794
+    },
+    {
+      "epoch": 0.4364211038171651,
+      "grad_norm": 24.125,
+      "learning_rate": 2.2410784546964385e-06,
+      "loss": 0.9557,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4366642353513251,
+      "grad_norm": 16.875,
+      "learning_rate": 2.2407838803785604e-06,
+      "loss": 0.8656,
+      "step": 1796
+    },
+    {
+      "epoch": 0.43690736688548504,
+      "grad_norm": 27.875,
+      "learning_rate": 2.240489157970919e-06,
+      "loss": 0.9394,
+      "step": 1797
+    },
+    {
+      "epoch": 0.43715049841964504,
+      "grad_norm": 23.0,
+      "learning_rate": 2.2401942875175675e-06,
+      "loss": 1.0365,
+      "step": 1798
+    },
+    {
+      "epoch": 0.437393629953805,
+      "grad_norm": 34.25,
+      "learning_rate": 2.2398992690625785e-06,
+      "loss": 1.6354,
+      "step": 1799
+    },
+    {
+      "epoch": 0.437636761487965,
+      "grad_norm": 20.75,
+      "learning_rate": 2.239604102650047e-06,
+      "loss": 1.0085,
+      "step": 1800
+    },
+    {
+      "epoch": 0.437879893022125,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.2393087883240917e-06,
+      "loss": 0.7051,
+      "step": 1801
+    },
+    {
+      "epoch": 0.43812302455628493,
+      "grad_norm": 21.625,
+      "learning_rate": 2.2390133261288523e-06,
+      "loss": 0.8884,
+      "step": 1802
+    },
+    {
+      "epoch": 0.43836615609044494,
+      "grad_norm": 17.5,
+      "learning_rate": 2.23871771610849e-06,
+      "loss": 0.5416,
+      "step": 1803
+    },
+    {
+      "epoch": 0.43860928762460494,
+      "grad_norm": 33.5,
+      "learning_rate": 2.238421958307189e-06,
+      "loss": 1.0586,
+      "step": 1804
+    },
+    {
+      "epoch": 0.4388524191587649,
+      "grad_norm": 26.5,
+      "learning_rate": 2.238126052769156e-06,
+      "loss": 0.9796,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4390955506929249,
+      "grad_norm": 22.0,
+      "learning_rate": 2.2378299995386194e-06,
+      "loss": 0.6153,
+      "step": 1806
+    },
+    {
+      "epoch": 0.43933868222708483,
+      "grad_norm": 20.75,
+      "learning_rate": 2.2375337986598282e-06,
+      "loss": 0.9212,
+      "step": 1807
+    },
+    {
+      "epoch": 0.43958181376124483,
+      "grad_norm": 23.25,
+      "learning_rate": 2.237237450177056e-06,
+      "loss": 1.2888,
+      "step": 1808
+    },
+    {
+      "epoch": 0.43982494529540483,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.2369409541345967e-06,
+      "loss": 0.6865,
+      "step": 1809
+    },
+    {
+      "epoch": 0.4400680768295648,
+      "grad_norm": 16.75,
+      "learning_rate": 2.2366443105767667e-06,
+      "loss": 0.8341,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4403112083637248,
+      "grad_norm": 26.125,
+      "learning_rate": 2.236347519547904e-06,
+      "loss": 1.2071,
+      "step": 1811
+    },
+    {
+      "epoch": 0.44055433989788473,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.23605058109237e-06,
+      "loss": 0.6755,
+      "step": 1812
+    },
+    {
+      "epoch": 0.44079747143204473,
+      "grad_norm": 18.25,
+      "learning_rate": 2.235753495254547e-06,
+      "loss": 0.6688,
+      "step": 1813
+    },
+    {
+      "epoch": 0.44104060296620473,
+      "grad_norm": 17.25,
+      "learning_rate": 2.23545626207884e-06,
+      "loss": 0.7241,
+      "step": 1814
+    },
+    {
+      "epoch": 0.4412837345003647,
+      "grad_norm": 18.875,
+      "learning_rate": 2.235158881609675e-06,
+      "loss": 0.7078,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4415268660345247,
+      "grad_norm": 17.625,
+      "learning_rate": 2.2348613538915004e-06,
+      "loss": 0.8524,
+      "step": 1816
+    },
+    {
+      "epoch": 0.4417699975686847,
+      "grad_norm": 17.625,
+      "learning_rate": 2.234563678968788e-06,
+      "loss": 1.007,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4420131291028446,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.2342658568860292e-06,
+      "loss": 0.6466,
+      "step": 1818
+    },
+    {
+      "epoch": 0.44225626063700463,
+      "grad_norm": 21.875,
+      "learning_rate": 2.2339678876877393e-06,
+      "loss": 1.3698,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4424993921711646,
+      "grad_norm": 21.375,
+      "learning_rate": 2.233669771418455e-06,
+      "loss": 0.5664,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4427425237053246,
+      "grad_norm": 22.875,
+      "learning_rate": 2.2333715081227347e-06,
+      "loss": 0.7115,
+      "step": 1821
+    },
+    {
+      "epoch": 0.4429856552394846,
+      "grad_norm": 18.625,
+      "learning_rate": 2.2330730978451593e-06,
+      "loss": 1.0623,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4432287867736445,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.2327745406303314e-06,
+      "loss": 0.7299,
+      "step": 1823
+    },
+    {
+      "epoch": 0.4434719183078045,
+      "grad_norm": 19.375,
+      "learning_rate": 2.2324758365228745e-06,
+      "loss": 0.7466,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4437150498419645,
+      "grad_norm": 17.125,
+      "learning_rate": 2.2321769855674365e-06,
+      "loss": 0.6095,
+      "step": 1825
+    },
+    {
+      "epoch": 0.44395818137612447,
+      "grad_norm": 21.5,
+      "learning_rate": 2.2318779878086853e-06,
+      "loss": 0.847,
+      "step": 1826
+    },
+    {
+      "epoch": 0.4442013129102845,
+      "grad_norm": 18.5,
+      "learning_rate": 2.231578843291311e-06,
+      "loss": 1.123,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 15.375,
+      "learning_rate": 2.231279552060026e-06,
+      "loss": 0.3888,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4446875759786044,
+      "grad_norm": 16.875,
+      "learning_rate": 2.230980114159565e-06,
+      "loss": 0.6976,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4449307075127644,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.2306805296346836e-06,
+      "loss": 0.5196,
+      "step": 1830
+    },
+    {
+      "epoch": 0.44517383904692437,
+      "grad_norm": 39.25,
+      "learning_rate": 2.23038079853016e-06,
+      "loss": 0.6276,
+      "step": 1831
+    },
+    {
+      "epoch": 0.44541697058108437,
+      "grad_norm": 18.5,
+      "learning_rate": 2.2300809208907943e-06,
+      "loss": 0.743,
+      "step": 1832
+    },
+    {
+      "epoch": 0.44566010211524437,
+      "grad_norm": 17.375,
+      "learning_rate": 2.2297808967614085e-06,
+      "loss": 0.9677,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4459032336494043,
+      "grad_norm": 22.625,
+      "learning_rate": 2.2294807261868463e-06,
+      "loss": 0.7887,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4461463651835643,
+      "grad_norm": 17.5,
+      "learning_rate": 2.2291804092119736e-06,
+      "loss": 0.8214,
+      "step": 1835
+    },
+    {
+      "epoch": 0.44638949671772427,
+      "grad_norm": 18.875,
+      "learning_rate": 2.228879945881677e-06,
+      "loss": 0.6916,
+      "step": 1836
+    },
+    {
+      "epoch": 0.44663262825188427,
+      "grad_norm": 16.375,
+      "learning_rate": 2.228579336240867e-06,
+      "loss": 0.7197,
+      "step": 1837
+    },
+    {
+      "epoch": 0.44687575978604427,
+      "grad_norm": 19.5,
+      "learning_rate": 2.228278580334475e-06,
+      "loss": 1.1411,
+      "step": 1838
+    },
+    {
+      "epoch": 0.4471188913202042,
+      "grad_norm": 17.375,
+      "learning_rate": 2.227977678207453e-06,
+      "loss": 1.0894,
+      "step": 1839
+    },
+    {
+      "epoch": 0.4473620228543642,
+      "grad_norm": 20.375,
+      "learning_rate": 2.227676629904777e-06,
+      "loss": 0.9801,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4476051543885242,
+      "grad_norm": 22.25,
+      "learning_rate": 2.2273754354714437e-06,
+      "loss": 1.063,
+      "step": 1841
+    },
+    {
+      "epoch": 0.44784828592268416,
+      "grad_norm": 19.0,
+      "learning_rate": 2.2270740949524717e-06,
+      "loss": 0.7653,
+      "step": 1842
+    },
+    {
+      "epoch": 0.44809141745684417,
+      "grad_norm": 22.375,
+      "learning_rate": 2.2267726083929015e-06,
+      "loss": 0.9432,
+      "step": 1843
+    },
+    {
+      "epoch": 0.4483345489910041,
+      "grad_norm": 16.5,
+      "learning_rate": 2.2264709758377957e-06,
+      "loss": 0.6867,
+      "step": 1844
+    },
+    {
+      "epoch": 0.4485776805251641,
+      "grad_norm": 15.75,
+      "learning_rate": 2.226169197332238e-06,
+      "loss": 0.4329,
+      "step": 1845
+    },
+    {
+      "epoch": 0.4488208120593241,
+      "grad_norm": 18.625,
+      "learning_rate": 2.225867272921335e-06,
+      "loss": 0.8247,
+      "step": 1846
+    },
+    {
+      "epoch": 0.44906394359348406,
+      "grad_norm": 17.375,
+      "learning_rate": 2.2255652026502144e-06,
+      "loss": 0.8816,
+      "step": 1847
+    },
+    {
+      "epoch": 0.44930707512764406,
+      "grad_norm": 16.625,
+      "learning_rate": 2.225262986564025e-06,
+      "loss": 0.6855,
+      "step": 1848
+    },
+    {
+      "epoch": 0.449550206661804,
+      "grad_norm": 21.375,
+      "learning_rate": 2.2249606247079397e-06,
+      "loss": 0.9725,
+      "step": 1849
+    },
+    {
+      "epoch": 0.449793338195964,
+      "grad_norm": 19.75,
+      "learning_rate": 2.2246581171271503e-06,
+      "loss": 0.8016,
+      "step": 1850
+    },
+    {
+      "epoch": 0.450036469730124,
+      "grad_norm": 23.0,
+      "learning_rate": 2.2243554638668727e-06,
+      "loss": 0.8972,
+      "step": 1851
+    },
+    {
+      "epoch": 0.45027960126428396,
+      "grad_norm": 22.5,
+      "learning_rate": 2.2240526649723433e-06,
+      "loss": 0.6963,
+      "step": 1852
+    },
+    {
+      "epoch": 0.45052273279844396,
+      "grad_norm": 15.375,
+      "learning_rate": 2.2237497204888205e-06,
+      "loss": 0.4336,
+      "step": 1853
+    },
+    {
+      "epoch": 0.45076586433260396,
+      "grad_norm": 22.25,
+      "learning_rate": 2.223446630461585e-06,
+      "loss": 0.962,
+      "step": 1854
+    },
+    {
+      "epoch": 0.4510089958667639,
+      "grad_norm": 17.875,
+      "learning_rate": 2.2231433949359384e-06,
+      "loss": 0.7468,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4512521274009239,
+      "grad_norm": 19.125,
+      "learning_rate": 2.2228400139572043e-06,
+      "loss": 1.029,
+      "step": 1856
+    },
+    {
+      "epoch": 0.45149525893508385,
+      "grad_norm": 21.25,
+      "learning_rate": 2.222536487570729e-06,
+      "loss": 0.9867,
+      "step": 1857
+    },
+    {
+      "epoch": 0.45173839046924386,
+      "grad_norm": 16.5,
+      "learning_rate": 2.2222328158218793e-06,
+      "loss": 0.7912,
+      "step": 1858
+    },
+    {
+      "epoch": 0.45198152200340386,
+      "grad_norm": 21.375,
+      "learning_rate": 2.221928998756044e-06,
+      "loss": 1.0065,
+      "step": 1859
+    },
+    {
+      "epoch": 0.4522246535375638,
+      "grad_norm": 18.625,
+      "learning_rate": 2.2216250364186344e-06,
+      "loss": 0.8939,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4524677850717238,
+      "grad_norm": 16.25,
+      "learning_rate": 2.2213209288550826e-06,
+      "loss": 0.7344,
+      "step": 1861
+    },
+    {
+      "epoch": 0.4527109166058838,
+      "grad_norm": 18.625,
+      "learning_rate": 2.2210166761108422e-06,
+      "loss": 0.9777,
+      "step": 1862
+    },
+    {
+      "epoch": 0.45295404814004375,
+      "grad_norm": 16.375,
+      "learning_rate": 2.2207122782313895e-06,
+      "loss": 0.6445,
+      "step": 1863
+    },
+    {
+      "epoch": 0.45319717967420375,
+      "grad_norm": 16.625,
+      "learning_rate": 2.220407735262223e-06,
+      "loss": 0.6399,
+      "step": 1864
+    },
+    {
+      "epoch": 0.4534403112083637,
+      "grad_norm": 22.25,
+      "learning_rate": 2.22010304724886e-06,
+      "loss": 0.6662,
+      "step": 1865
+    },
+    {
+      "epoch": 0.4536834427425237,
+      "grad_norm": 17.125,
+      "learning_rate": 2.2197982142368423e-06,
+      "loss": 0.8791,
+      "step": 1866
+    },
+    {
+      "epoch": 0.4539265742766837,
+      "grad_norm": 17.5,
+      "learning_rate": 2.219493236271733e-06,
+      "loss": 0.7613,
+      "step": 1867
+    },
+    {
+      "epoch": 0.45416970581084365,
+      "grad_norm": 22.875,
+      "learning_rate": 2.2191881133991154e-06,
+      "loss": 0.8419,
+      "step": 1868
+    },
+    {
+      "epoch": 0.45441283734500365,
+      "grad_norm": 16.125,
+      "learning_rate": 2.218882845664596e-06,
+      "loss": 0.6044,
+      "step": 1869
+    },
+    {
+      "epoch": 0.45465596887916365,
+      "grad_norm": 18.0,
+      "learning_rate": 2.2185774331138023e-06,
+      "loss": 1.0509,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4548991004133236,
+      "grad_norm": 25.75,
+      "learning_rate": 2.2182718757923834e-06,
+      "loss": 1.0906,
+      "step": 1871
+    },
+    {
+      "epoch": 0.4551422319474836,
+      "grad_norm": 18.875,
+      "learning_rate": 2.2179661737460096e-06,
+      "loss": 0.9405,
+      "step": 1872
+    },
+    {
+      "epoch": 0.45538536348164355,
+      "grad_norm": 20.625,
+      "learning_rate": 2.217660327020374e-06,
+      "loss": 1.2509,
+      "step": 1873
+    },
+    {
+      "epoch": 0.45562849501580355,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.2173543356611903e-06,
+      "loss": 0.6884,
+      "step": 1874
+    },
+    {
+      "epoch": 0.45587162654996355,
+      "grad_norm": 18.5,
+      "learning_rate": 2.217048199714194e-06,
+      "loss": 1.1921,
+      "step": 1875
+    },
+    {
+      "epoch": 0.4561147580841235,
+      "grad_norm": 16.75,
+      "learning_rate": 2.2167419192251435e-06,
+      "loss": 0.66,
+      "step": 1876
+    },
+    {
+      "epoch": 0.4563578896182835,
+      "grad_norm": 19.625,
+      "learning_rate": 2.216435494239817e-06,
+      "loss": 0.6735,
+      "step": 1877
+    },
+    {
+      "epoch": 0.4566010211524435,
+      "grad_norm": 19.75,
+      "learning_rate": 2.2161289248040144e-06,
+      "loss": 0.8849,
+      "step": 1878
+    },
+    {
+      "epoch": 0.45684415268660344,
+      "grad_norm": 20.75,
+      "learning_rate": 2.2158222109635583e-06,
+      "loss": 0.8143,
+      "step": 1879
+    },
+    {
+      "epoch": 0.45708728422076345,
+      "grad_norm": 19.0,
+      "learning_rate": 2.215515352764293e-06,
+      "loss": 0.8738,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4573304157549234,
+      "grad_norm": 33.5,
+      "learning_rate": 2.215208350252083e-06,
+      "loss": 1.2343,
+      "step": 1881
+    },
+    {
+      "epoch": 0.4575735472890834,
+      "grad_norm": 22.5,
+      "learning_rate": 2.214901203472815e-06,
+      "loss": 1.3959,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4578166788232434,
+      "grad_norm": 20.75,
+      "learning_rate": 2.214593912472398e-06,
+      "loss": 0.6936,
+      "step": 1883
+    },
+    {
+      "epoch": 0.45805981035740334,
+      "grad_norm": 20.75,
+      "learning_rate": 2.214286477296762e-06,
+      "loss": 0.7576,
+      "step": 1884
+    },
+    {
+      "epoch": 0.45830294189156334,
+      "grad_norm": 23.5,
+      "learning_rate": 2.2139788979918577e-06,
+      "loss": 1.4372,
+      "step": 1885
+    },
+    {
+      "epoch": 0.45854607342572334,
+      "grad_norm": 14.25,
+      "learning_rate": 2.2136711746036587e-06,
+      "loss": 0.5822,
+      "step": 1886
+    },
+    {
+      "epoch": 0.4587892049598833,
+      "grad_norm": 18.125,
+      "learning_rate": 2.2133633071781597e-06,
+      "loss": 0.8024,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4590323364940433,
+      "grad_norm": 13.3125,
+      "learning_rate": 2.213055295761376e-06,
+      "loss": 0.4761,
+      "step": 1888
+    },
+    {
+      "epoch": 0.45927546802820324,
+      "grad_norm": 31.0,
+      "learning_rate": 2.2127471403993463e-06,
+      "loss": 0.9431,
+      "step": 1889
+    },
+    {
+      "epoch": 0.45951859956236324,
+      "grad_norm": 18.75,
+      "learning_rate": 2.2124388411381294e-06,
+      "loss": 1.0756,
+      "step": 1890
+    },
+    {
+      "epoch": 0.45976173109652324,
+      "grad_norm": 24.375,
+      "learning_rate": 2.2121303980238053e-06,
+      "loss": 0.5698,
+      "step": 1891
+    },
+    {
+      "epoch": 0.4600048626306832,
+      "grad_norm": 16.625,
+      "learning_rate": 2.2118218111024768e-06,
+      "loss": 0.8241,
+      "step": 1892
+    },
+    {
+      "epoch": 0.4602479941648432,
+      "grad_norm": 18.875,
+      "learning_rate": 2.2115130804202676e-06,
+      "loss": 0.6232,
+      "step": 1893
+    },
+    {
+      "epoch": 0.46049112569900313,
+      "grad_norm": 20.5,
+      "learning_rate": 2.2112042060233225e-06,
+      "loss": 0.9004,
+      "step": 1894
+    },
+    {
+      "epoch": 0.46073425723316314,
+      "grad_norm": 15.25,
+      "learning_rate": 2.2108951879578082e-06,
+      "loss": 0.6967,
+      "step": 1895
+    },
+    {
+      "epoch": 0.46097738876732314,
+      "grad_norm": 21.125,
+      "learning_rate": 2.210586026269913e-06,
+      "loss": 1.2644,
+      "step": 1896
+    },
+    {
+      "epoch": 0.4612205203014831,
+      "grad_norm": 26.625,
+      "learning_rate": 2.210276721005846e-06,
+      "loss": 1.1362,
+      "step": 1897
+    },
+    {
+      "epoch": 0.4614636518356431,
+      "grad_norm": 18.0,
+      "learning_rate": 2.2099672722118386e-06,
+      "loss": 1.0583,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4617067833698031,
+      "grad_norm": 19.375,
+      "learning_rate": 2.2096576799341436e-06,
+      "loss": 0.6459,
+      "step": 1899
+    },
+    {
+      "epoch": 0.46194991490396303,
+      "grad_norm": 25.0,
+      "learning_rate": 2.209347944219034e-06,
+      "loss": 1.2253,
+      "step": 1900
+    },
+    {
+      "epoch": 0.46219304643812303,
+      "grad_norm": 22.625,
+      "learning_rate": 2.2090380651128056e-06,
+      "loss": 0.904,
+      "step": 1901
+    },
+    {
+      "epoch": 0.462436177972283,
+      "grad_norm": 22.0,
+      "learning_rate": 2.2087280426617754e-06,
+      "loss": 0.6962,
+      "step": 1902
+    },
+    {
+      "epoch": 0.462679309506443,
+      "grad_norm": 17.875,
+      "learning_rate": 2.208417876912281e-06,
+      "loss": 0.4962,
+      "step": 1903
+    },
+    {
+      "epoch": 0.462922441040603,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.208107567910683e-06,
+      "loss": 0.5747,
+      "step": 1904
+    },
+    {
+      "epoch": 0.46316557257476293,
+      "grad_norm": 21.5,
+      "learning_rate": 2.207797115703362e-06,
+      "loss": 0.6997,
+      "step": 1905
+    },
+    {
+      "epoch": 0.46340870410892293,
+      "grad_norm": 21.375,
+      "learning_rate": 2.2074865203367196e-06,
+      "loss": 1.1637,
+      "step": 1906
+    },
+    {
+      "epoch": 0.46365183564308293,
+      "grad_norm": 16.375,
+      "learning_rate": 2.2071757818571808e-06,
+      "loss": 1.3612,
+      "step": 1907
+    },
+    {
+      "epoch": 0.4638949671772429,
+      "grad_norm": 18.875,
+      "learning_rate": 2.2068649003111903e-06,
+      "loss": 0.786,
+      "step": 1908
+    },
+    {
+      "epoch": 0.4641380987114029,
+      "grad_norm": 14.0,
+      "learning_rate": 2.2065538757452148e-06,
+      "loss": 0.5215,
+      "step": 1909
+    },
+    {
+      "epoch": 0.4643812302455628,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.2062427082057427e-06,
+      "loss": 0.5998,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4646243617797228,
+      "grad_norm": 25.75,
+      "learning_rate": 2.2059313977392825e-06,
+      "loss": 0.954,
+      "step": 1911
+    },
+    {
+      "epoch": 0.46486749331388283,
+      "grad_norm": 17.375,
+      "learning_rate": 2.2056199443923656e-06,
+      "loss": 1.0733,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4651106248480428,
+      "grad_norm": 19.5,
+      "learning_rate": 2.205308348211544e-06,
+      "loss": 0.8767,
+      "step": 1913
+    },
+    {
+      "epoch": 0.4653537563822028,
+      "grad_norm": 19.0,
+      "learning_rate": 2.2049966092433906e-06,
+      "loss": 0.9691,
+      "step": 1914
+    },
+    {
+      "epoch": 0.4655968879163628,
+      "grad_norm": 24.75,
+      "learning_rate": 2.2046847275345007e-06,
+      "loss": 1.2806,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4658400194505227,
+      "grad_norm": 16.75,
+      "learning_rate": 2.2043727031314906e-06,
+      "loss": 0.9385,
+      "step": 1916
+    },
+    {
+      "epoch": 0.4660831509846827,
+      "grad_norm": 19.0,
+      "learning_rate": 2.2040605360809973e-06,
+      "loss": 0.5772,
+      "step": 1917
+    },
+    {
+      "epoch": 0.46632628251884267,
+      "grad_norm": 15.4375,
+      "learning_rate": 2.20374822642968e-06,
+      "loss": 0.5349,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4665694140530027,
+      "grad_norm": 18.625,
+      "learning_rate": 2.203435774224218e-06,
+      "loss": 0.8595,
+      "step": 1919
+    },
+    {
+      "epoch": 0.4668125455871627,
+      "grad_norm": 16.0,
+      "learning_rate": 2.203123179511313e-06,
+      "loss": 0.6659,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4670556771213226,
+      "grad_norm": 16.25,
+      "learning_rate": 2.202810442337688e-06,
+      "loss": 0.606,
+      "step": 1921
+    },
+    {
+      "epoch": 0.4672988086554826,
+      "grad_norm": 21.5,
+      "learning_rate": 2.202497562750087e-06,
+      "loss": 0.6258,
+      "step": 1922
+    },
+    {
+      "epoch": 0.4675419401896426,
+      "grad_norm": 23.125,
+      "learning_rate": 2.202184540795275e-06,
+      "loss": 1.0312,
+      "step": 1923
+    },
+    {
+      "epoch": 0.46778507172380257,
+      "grad_norm": 21.875,
+      "learning_rate": 2.2018713765200384e-06,
+      "loss": 1.1694,
+      "step": 1924
+    },
+    {
+      "epoch": 0.46802820325796257,
+      "grad_norm": 21.0,
+      "learning_rate": 2.201558069971185e-06,
+      "loss": 1.1825,
+      "step": 1925
+    },
+    {
+      "epoch": 0.4682713347921225,
+      "grad_norm": 17.5,
+      "learning_rate": 2.2012446211955445e-06,
+      "loss": 1.0307,
+      "step": 1926
+    },
+    {
+      "epoch": 0.4685144663262825,
+      "grad_norm": 26.375,
+      "learning_rate": 2.2009310302399666e-06,
+      "loss": 0.9426,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4687575978604425,
+      "grad_norm": 17.0,
+      "learning_rate": 2.2006172971513234e-06,
+      "loss": 0.9969,
+      "step": 1928
+    },
+    {
+      "epoch": 0.46900072939460247,
+      "grad_norm": 20.0,
+      "learning_rate": 2.200303421976507e-06,
+      "loss": 0.9373,
+      "step": 1929
+    },
+    {
+      "epoch": 0.46924386092876247,
+      "grad_norm": 21.0,
+      "learning_rate": 2.199989404762432e-06,
+      "loss": 0.7108,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4694869924629224,
+      "grad_norm": 23.5,
+      "learning_rate": 2.1996752455560337e-06,
+      "loss": 0.8034,
+      "step": 1931
+    },
+    {
+      "epoch": 0.4697301239970824,
+      "grad_norm": 13.0,
+      "learning_rate": 2.1993609444042683e-06,
+      "loss": 1.0756,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4699732555312424,
+      "grad_norm": 13.9375,
+      "learning_rate": 2.199046501354114e-06,
+      "loss": 0.4195,
+      "step": 1933
+    },
+    {
+      "epoch": 0.47021638706540236,
+      "grad_norm": 18.625,
+      "learning_rate": 2.1987319164525692e-06,
+      "loss": 0.9116,
+      "step": 1934
+    },
+    {
+      "epoch": 0.47045951859956237,
+      "grad_norm": 22.5,
+      "learning_rate": 2.1984171897466544e-06,
+      "loss": 1.0722,
+      "step": 1935
+    },
+    {
+      "epoch": 0.47070265013372237,
+      "grad_norm": 18.625,
+      "learning_rate": 2.198102321283411e-06,
+      "loss": 0.9446,
+      "step": 1936
+    },
+    {
+      "epoch": 0.4709457816678823,
+      "grad_norm": 17.375,
+      "learning_rate": 2.1977873111099014e-06,
+      "loss": 0.8527,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4711889132020423,
+      "grad_norm": 17.5,
+      "learning_rate": 2.197472159273209e-06,
+      "loss": 0.7372,
+      "step": 1938
+    },
+    {
+      "epoch": 0.47143204473620226,
+      "grad_norm": 17.625,
+      "learning_rate": 2.197156865820439e-06,
+      "loss": 0.7782,
+      "step": 1939
+    },
+    {
+      "epoch": 0.47167517627036226,
+      "grad_norm": 51.75,
+      "learning_rate": 2.1968414307987178e-06,
+      "loss": 1.2376,
+      "step": 1940
+    },
+    {
+      "epoch": 0.47191830780452226,
+      "grad_norm": 19.75,
+      "learning_rate": 2.196525854255192e-06,
+      "loss": 0.9464,
+      "step": 1941
+    },
+    {
+      "epoch": 0.4721614393386822,
+      "grad_norm": 21.75,
+      "learning_rate": 2.1962101362370305e-06,
+      "loss": 0.6516,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4724045708728422,
+      "grad_norm": 21.25,
+      "learning_rate": 2.1958942767914223e-06,
+      "loss": 1.1686,
+      "step": 1943
+    },
+    {
+      "epoch": 0.4726477024070022,
+      "grad_norm": 26.0,
+      "learning_rate": 2.195578275965578e-06,
+      "loss": 1.0714,
+      "step": 1944
+    },
+    {
+      "epoch": 0.47289083394116216,
+      "grad_norm": 17.75,
+      "learning_rate": 2.19526213380673e-06,
+      "loss": 0.7905,
+      "step": 1945
+    },
+    {
+      "epoch": 0.47313396547532216,
+      "grad_norm": 17.625,
+      "learning_rate": 2.1949458503621308e-06,
+      "loss": 0.5541,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4733770970094821,
+      "grad_norm": 30.0,
+      "learning_rate": 2.194629425679054e-06,
+      "loss": 1.2114,
+      "step": 1947
+    },
+    {
+      "epoch": 0.4736202285436421,
+      "grad_norm": 18.875,
+      "learning_rate": 2.1943128598047957e-06,
+      "loss": 0.4835,
+      "step": 1948
+    },
+    {
+      "epoch": 0.4738633600778021,
+      "grad_norm": 19.5,
+      "learning_rate": 2.1939961527866718e-06,
+      "loss": 1.0849,
+      "step": 1949
+    },
+    {
+      "epoch": 0.47410649161196206,
+      "grad_norm": 19.5,
+      "learning_rate": 2.1936793046720196e-06,
+      "loss": 0.9007,
+      "step": 1950
+    },
+    {
+      "epoch": 0.47434962314612206,
+      "grad_norm": 25.5,
+      "learning_rate": 2.1933623155081967e-06,
+      "loss": 1.0149,
+      "step": 1951
+    },
+    {
+      "epoch": 0.47459275468028206,
+      "grad_norm": 19.375,
+      "learning_rate": 2.1930451853425837e-06,
+      "loss": 0.912,
+      "step": 1952
+    },
+    {
+      "epoch": 0.474835886214442,
+      "grad_norm": 34.5,
+      "learning_rate": 2.1927279142225812e-06,
+      "loss": 1.1245,
+      "step": 1953
+    },
+    {
+      "epoch": 0.475079017748602,
+      "grad_norm": 22.625,
+      "learning_rate": 2.1924105021956097e-06,
+      "loss": 1.1844,
+      "step": 1954
+    },
+    {
+      "epoch": 0.47532214928276195,
+      "grad_norm": 23.875,
+      "learning_rate": 2.192092949309113e-06,
+      "loss": 1.0106,
+      "step": 1955
+    },
+    {
+      "epoch": 0.47556528081692195,
+      "grad_norm": 24.75,
+      "learning_rate": 2.191775255610555e-06,
+      "loss": 1.0633,
+      "step": 1956
+    },
+    {
+      "epoch": 0.47580841235108196,
+      "grad_norm": 18.125,
+      "learning_rate": 2.1914574211474194e-06,
+      "loss": 0.5309,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4760515438852419,
+      "grad_norm": 20.25,
+      "learning_rate": 2.191139445967213e-06,
+      "loss": 0.7897,
+      "step": 1958
+    },
+    {
+      "epoch": 0.4762946754194019,
+      "grad_norm": 19.25,
+      "learning_rate": 2.190821330117462e-06,
+      "loss": 0.8475,
+      "step": 1959
+    },
+    {
+      "epoch": 0.4765378069535619,
+      "grad_norm": 16.75,
+      "learning_rate": 2.190503073645715e-06,
+      "loss": 0.7451,
+      "step": 1960
+    },
+    {
+      "epoch": 0.47678093848772185,
+      "grad_norm": 20.5,
+      "learning_rate": 2.1901846765995406e-06,
+      "loss": 0.9894,
+      "step": 1961
+    },
+    {
+      "epoch": 0.47702407002188185,
+      "grad_norm": 18.0,
+      "learning_rate": 2.1898661390265287e-06,
+      "loss": 0.9663,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4772672015560418,
+      "grad_norm": 17.375,
+      "learning_rate": 2.1895474609742897e-06,
+      "loss": 0.9645,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4775103330902018,
+      "grad_norm": 21.25,
+      "learning_rate": 2.1892286424904567e-06,
+      "loss": 1.2093,
+      "step": 1964
+    },
+    {
+      "epoch": 0.4777534646243618,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.188909683622682e-06,
+      "loss": 0.4981,
+      "step": 1965
+    },
+    {
+      "epoch": 0.47799659615852175,
+      "grad_norm": 21.375,
+      "learning_rate": 2.1885905844186395e-06,
+      "loss": 1.0424,
+      "step": 1966
+    },
+    {
+      "epoch": 0.47823972769268175,
+      "grad_norm": 21.0,
+      "learning_rate": 2.188271344926024e-06,
+      "loss": 0.9301,
+      "step": 1967
+    },
+    {
+      "epoch": 0.4784828592268417,
+      "grad_norm": 18.375,
+      "learning_rate": 2.187951965192552e-06,
+      "loss": 0.8298,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4787259907610017,
+      "grad_norm": 18.75,
+      "learning_rate": 2.1876324452659593e-06,
+      "loss": 1.1689,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4789691222951617,
+      "grad_norm": 25.375,
+      "learning_rate": 2.187312785194004e-06,
+      "loss": 0.9826,
+      "step": 1970
+    },
+    {
+      "epoch": 0.47921225382932164,
+      "grad_norm": 20.625,
+      "learning_rate": 2.1869929850244655e-06,
+      "loss": 0.732,
+      "step": 1971
+    },
+    {
+      "epoch": 0.47945538536348165,
+      "grad_norm": 15.125,
+      "learning_rate": 2.1866730448051427e-06,
+      "loss": 0.7193,
+      "step": 1972
+    },
+    {
+      "epoch": 0.47969851689764165,
+      "grad_norm": 18.625,
+      "learning_rate": 2.1863529645838564e-06,
+      "loss": 0.9485,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4799416484318016,
+      "grad_norm": 18.0,
+      "learning_rate": 2.1860327444084483e-06,
+      "loss": 0.718,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4801847799659616,
+      "grad_norm": 16.875,
+      "learning_rate": 2.1857123843267808e-06,
+      "loss": 0.7812,
+      "step": 1975
+    },
+    {
+      "epoch": 0.48042791150012154,
+      "grad_norm": 18.25,
+      "learning_rate": 2.185391884386737e-06,
+      "loss": 0.8458,
+      "step": 1976
+    },
+    {
+      "epoch": 0.48067104303428154,
+      "grad_norm": 21.0,
+      "learning_rate": 2.185071244636221e-06,
+      "loss": 0.9139,
+      "step": 1977
+    },
+    {
+      "epoch": 0.48091417456844154,
+      "grad_norm": 15.4375,
+      "learning_rate": 2.1847504651231586e-06,
+      "loss": 0.4537,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4811573061026015,
+      "grad_norm": 24.25,
+      "learning_rate": 2.1844295458954956e-06,
+      "loss": 0.6585,
+      "step": 1979
+    },
+    {
+      "epoch": 0.4814004376367615,
+      "grad_norm": 22.125,
+      "learning_rate": 2.184108487001199e-06,
+      "loss": 0.8459,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4816435691709215,
+      "grad_norm": 19.75,
+      "learning_rate": 2.183787288488256e-06,
+      "loss": 0.7802,
+      "step": 1981
+    },
+    {
+      "epoch": 0.48188670070508144,
+      "grad_norm": 17.875,
+      "learning_rate": 2.183465950404676e-06,
+      "loss": 0.8244,
+      "step": 1982
+    },
+    {
+      "epoch": 0.48212983223924144,
+      "grad_norm": 16.375,
+      "learning_rate": 2.1831444727984877e-06,
+      "loss": 0.8622,
+      "step": 1983
+    },
+    {
+      "epoch": 0.4823729637734014,
+      "grad_norm": 18.125,
+      "learning_rate": 2.1828228557177424e-06,
+      "loss": 0.912,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4826160953075614,
+      "grad_norm": 21.75,
+      "learning_rate": 2.1825010992105107e-06,
+      "loss": 1.3709,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4828592268417214,
+      "grad_norm": 19.25,
+      "learning_rate": 2.1821792033248847e-06,
+      "loss": 0.7946,
+      "step": 1986
+    },
+    {
+      "epoch": 0.48310235837588134,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.181857168108978e-06,
+      "loss": 0.7402,
+      "step": 1987
+    },
+    {
+      "epoch": 0.48334548991004134,
+      "grad_norm": 17.25,
+      "learning_rate": 2.1815349936109233e-06,
+      "loss": 0.713,
+      "step": 1988
+    },
+    {
+      "epoch": 0.48358862144420134,
+      "grad_norm": 39.5,
+      "learning_rate": 2.1812126798788758e-06,
+      "loss": 0.987,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4838317529783613,
+      "grad_norm": 16.5,
+      "learning_rate": 2.1808902269610106e-06,
+      "loss": 0.8076,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4840748845125213,
+      "grad_norm": 18.125,
+      "learning_rate": 2.1805676349055244e-06,
+      "loss": 1.0758,
+      "step": 1991
+    },
+    {
+      "epoch": 0.48431801604668123,
+      "grad_norm": 20.5,
+      "learning_rate": 2.1802449037606333e-06,
+      "loss": 1.0461,
+      "step": 1992
+    },
+    {
+      "epoch": 0.48456114758084123,
+      "grad_norm": 15.25,
+      "learning_rate": 2.1799220335745753e-06,
+      "loss": 0.8085,
+      "step": 1993
+    },
+    {
+      "epoch": 0.48480427911500124,
+      "grad_norm": 15.25,
+      "learning_rate": 2.1795990243956094e-06,
+      "loss": 0.8672,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4850474106491612,
+      "grad_norm": 16.375,
+      "learning_rate": 2.1792758762720147e-06,
+      "loss": 0.7869,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4852905421833212,
+      "grad_norm": 19.0,
+      "learning_rate": 2.1789525892520906e-06,
+      "loss": 1.2812,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4855336737174812,
+      "grad_norm": 28.375,
+      "learning_rate": 2.178629163384159e-06,
+      "loss": 0.9107,
+      "step": 1997
+    },
+    {
+      "epoch": 0.48577680525164113,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.1783055987165604e-06,
+      "loss": 0.7029,
+      "step": 1998
+    },
+    {
+      "epoch": 0.48601993678580113,
+      "grad_norm": 20.5,
+      "learning_rate": 2.177981895297658e-06,
+      "loss": 0.8151,
+      "step": 1999
+    },
+    {
+      "epoch": 0.4862630683199611,
+      "grad_norm": 23.625,
+      "learning_rate": 2.1776580531758344e-06,
+      "loss": 1.1322,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4865061998541211,
+      "grad_norm": 16.75,
+      "learning_rate": 2.177334072399494e-06,
+      "loss": 0.6582,
+      "step": 2001
+    },
+    {
+      "epoch": 0.4867493313882811,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.1770099530170606e-06,
+      "loss": 0.571,
+      "step": 2002
+    },
+    {
+      "epoch": 0.486992462922441,
+      "grad_norm": 16.25,
+      "learning_rate": 2.1766856950769798e-06,
+      "loss": 0.8654,
+      "step": 2003
+    },
+    {
+      "epoch": 0.48723559445660103,
+      "grad_norm": 19.5,
+      "learning_rate": 2.176361298627717e-06,
+      "loss": 0.8847,
+      "step": 2004
+    },
+    {
+      "epoch": 0.487478725990761,
+      "grad_norm": 14.875,
+      "learning_rate": 2.1760367637177597e-06,
+      "loss": 0.6756,
+      "step": 2005
+    },
+    {
+      "epoch": 0.487721857524921,
+      "grad_norm": 23.125,
+      "learning_rate": 2.1757120903956146e-06,
+      "loss": 0.7871,
+      "step": 2006
+    },
+    {
+      "epoch": 0.487964989059081,
+      "grad_norm": 17.375,
+      "learning_rate": 2.17538727870981e-06,
+      "loss": 0.6673,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4882081205932409,
+      "grad_norm": 23.75,
+      "learning_rate": 2.1750623287088953e-06,
+      "loss": 0.9512,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4884512521274009,
+      "grad_norm": 17.5,
+      "learning_rate": 2.1747372404414385e-06,
+      "loss": 1.0196,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4886943836615609,
+      "grad_norm": 18.125,
+      "learning_rate": 2.1744120139560306e-06,
+      "loss": 0.8133,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4889375151957209,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.174086649301282e-06,
+      "loss": 0.4063,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4891806467298809,
+      "grad_norm": 14.625,
+      "learning_rate": 2.1737611465258242e-06,
+      "loss": 0.6986,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4894237782640408,
+      "grad_norm": 25.875,
+      "learning_rate": 2.1734355056783092e-06,
+      "loss": 0.8912,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4896669097982008,
+      "grad_norm": 21.375,
+      "learning_rate": 2.173109726807409e-06,
+      "loss": 1.2023,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4899100413323608,
+      "grad_norm": 19.0,
+      "learning_rate": 2.172783809961818e-06,
+      "loss": 0.6458,
+      "step": 2015
+    },
+    {
+      "epoch": 0.49015317286652077,
+      "grad_norm": 16.0,
+      "learning_rate": 2.1724577551902497e-06,
+      "loss": 0.6342,
+      "step": 2016
+    },
+    {
+      "epoch": 0.49039630440068077,
+      "grad_norm": 23.625,
+      "learning_rate": 2.172131562541438e-06,
+      "loss": 1.0564,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4906394359348408,
+      "grad_norm": 17.75,
+      "learning_rate": 2.171805232064139e-06,
+      "loss": 0.6947,
+      "step": 2018
+    },
+    {
+      "epoch": 0.4908825674690007,
+      "grad_norm": 13.75,
+      "learning_rate": 2.1714787638071276e-06,
+      "loss": 0.8488,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4911256990031607,
+      "grad_norm": 21.625,
+      "learning_rate": 2.1711521578192008e-06,
+      "loss": 1.1226,
+      "step": 2020
+    },
+    {
+      "epoch": 0.49136883053732067,
+      "grad_norm": 27.125,
+      "learning_rate": 2.170825414149175e-06,
+      "loss": 1.0455,
+      "step": 2021
+    },
+    {
+      "epoch": 0.49161196207148067,
+      "grad_norm": 17.125,
+      "learning_rate": 2.1704985328458877e-06,
+      "loss": 0.5323,
+      "step": 2022
+    },
+    {
+      "epoch": 0.49185509360564067,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.1701715139581974e-06,
+      "loss": 0.505,
+      "step": 2023
+    },
+    {
+      "epoch": 0.4920982251398006,
+      "grad_norm": 19.5,
+      "learning_rate": 2.1698443575349824e-06,
+      "loss": 0.7535,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4923413566739606,
+      "grad_norm": 23.125,
+      "learning_rate": 2.1695170636251416e-06,
+      "loss": 1.2536,
+      "step": 2025
+    },
+    {
+      "epoch": 0.4925844882081206,
+      "grad_norm": 18.0,
+      "learning_rate": 2.169189632277595e-06,
+      "loss": 0.8221,
+      "step": 2026
+    },
+    {
+      "epoch": 0.49282761974228056,
+      "grad_norm": 21.375,
+      "learning_rate": 2.168862063541283e-06,
+      "loss": 1.145,
+      "step": 2027
+    },
+    {
+      "epoch": 0.49307075127644057,
+      "grad_norm": 30.5,
+      "learning_rate": 2.168534357465167e-06,
+      "loss": 1.123,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4933138828106005,
+      "grad_norm": 17.5,
+      "learning_rate": 2.1682065140982266e-06,
+      "loss": 0.9509,
+      "step": 2029
+    },
+    {
+      "epoch": 0.4935570143447605,
+      "grad_norm": 20.625,
+      "learning_rate": 2.167878533489465e-06,
+      "loss": 0.9344,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4938001458789205,
+      "grad_norm": 16.5,
+      "learning_rate": 2.1675504156879047e-06,
+      "loss": 0.8253,
+      "step": 2031
+    },
+    {
+      "epoch": 0.49404327741308046,
+      "grad_norm": 18.375,
+      "learning_rate": 2.167222160742588e-06,
+      "loss": 0.8082,
+      "step": 2032
+    },
+    {
+      "epoch": 0.49428640894724046,
+      "grad_norm": 20.875,
+      "learning_rate": 2.166893768702578e-06,
+      "loss": 1.0062,
+      "step": 2033
+    },
+    {
+      "epoch": 0.49452954048140046,
+      "grad_norm": 28.0,
+      "learning_rate": 2.1665652396169593e-06,
+      "loss": 1.1932,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4947726720155604,
+      "grad_norm": 24.0,
+      "learning_rate": 2.1662365735348358e-06,
+      "loss": 1.0181,
+      "step": 2035
+    },
+    {
+      "epoch": 0.4950158035497204,
+      "grad_norm": 24.375,
+      "learning_rate": 2.165907770505332e-06,
+      "loss": 1.3114,
+      "step": 2036
+    },
+    {
+      "epoch": 0.49525893508388036,
+      "grad_norm": 23.875,
+      "learning_rate": 2.1655788305775945e-06,
+      "loss": 1.3005,
+      "step": 2037
+    },
+    {
+      "epoch": 0.49550206661804036,
+      "grad_norm": 18.375,
+      "learning_rate": 2.165249753800788e-06,
+      "loss": 1.0847,
+      "step": 2038
+    },
+    {
+      "epoch": 0.49574519815220036,
+      "grad_norm": 22.875,
+      "learning_rate": 2.1649205402240984e-06,
+      "loss": 1.1625,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4959883296863603,
+      "grad_norm": 21.0,
+      "learning_rate": 2.164591189896733e-06,
+      "loss": 0.849,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4962314612205203,
+      "grad_norm": 19.125,
+      "learning_rate": 2.164261702867919e-06,
+      "loss": 0.6409,
+      "step": 2041
+    },
+    {
+      "epoch": 0.49647459275468026,
+      "grad_norm": 24.125,
+      "learning_rate": 2.1639320791869035e-06,
+      "loss": 1.1276,
+      "step": 2042
+    },
+    {
+      "epoch": 0.49671772428884026,
+      "grad_norm": 18.625,
+      "learning_rate": 2.163602318902954e-06,
+      "loss": 1.0885,
+      "step": 2043
+    },
+    {
+      "epoch": 0.49696085582300026,
+      "grad_norm": 18.875,
+      "learning_rate": 2.1632724220653607e-06,
+      "loss": 0.8804,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4972039873571602,
+      "grad_norm": 20.625,
+      "learning_rate": 2.16294238872343e-06,
+      "loss": 1.1937,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4974471188913202,
+      "grad_norm": 18.375,
+      "learning_rate": 2.162612218926493e-06,
+      "loss": 0.9079,
+      "step": 2046
+    },
+    {
+      "epoch": 0.4976902504254802,
+      "grad_norm": 20.375,
+      "learning_rate": 2.162281912723898e-06,
+      "loss": 1.02,
+      "step": 2047
+    },
+    {
+      "epoch": 0.49793338195964015,
+      "grad_norm": 21.75,
+      "learning_rate": 2.161951470165016e-06,
+      "loss": 0.8726,
+      "step": 2048
+    },
+    {
+      "epoch": 0.49817651349380015,
+      "grad_norm": 20.375,
+      "learning_rate": 2.1616208912992363e-06,
+      "loss": 1.0263,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4984196450279601,
+      "grad_norm": 19.75,
+      "learning_rate": 2.161290176175971e-06,
+      "loss": 0.8888,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4986627765621201,
+      "grad_norm": 20.25,
+      "learning_rate": 2.16095932484465e-06,
+      "loss": 0.9729,
+      "step": 2051
+    },
+    {
+      "epoch": 0.4989059080962801,
+      "grad_norm": 19.5,
+      "learning_rate": 2.1606283373547246e-06,
+      "loss": 0.9825,
+      "step": 2052
+    },
+    {
+      "epoch": 0.49914903963044005,
+      "grad_norm": 19.625,
+      "learning_rate": 2.160297213755667e-06,
+      "loss": 0.6419,
+      "step": 2053
+    },
+    {
+      "epoch": 0.49939217116460005,
+      "grad_norm": 21.5,
+      "learning_rate": 2.1599659540969705e-06,
+      "loss": 1.0555,
+      "step": 2054
+    },
+    {
+      "epoch": 0.49963530269876005,
+      "grad_norm": 21.75,
+      "learning_rate": 2.1596345584281453e-06,
+      "loss": 1.0693,
+      "step": 2055
+    },
+    {
+      "epoch": 0.49987843423292,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.1593030267987262e-06,
+      "loss": 0.8538,
+      "step": 2056
+    },
+    {
+      "epoch": 0.50012156576708,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.158971359258265e-06,
+      "loss": 0.4969,
+      "step": 2057
+    },
+    {
+      "epoch": 0.50036469730124,
+      "grad_norm": 21.25,
+      "learning_rate": 2.1586395558563363e-06,
+      "loss": 1.3187,
+      "step": 2058
+    },
+    {
+      "epoch": 0.5006078288354,
+      "grad_norm": 18.5,
+      "learning_rate": 2.1583076166425328e-06,
+      "loss": 1.2774,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5008509603695599,
+      "grad_norm": 22.125,
+      "learning_rate": 2.157975541666469e-06,
+      "loss": 1.0634,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5010940919037199,
+      "grad_norm": 18.0,
+      "learning_rate": 2.1576433309777794e-06,
+      "loss": 0.8712,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5013372234378799,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.157310984626118e-06,
+      "loss": 0.6967,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5015803549720399,
+      "grad_norm": 23.375,
+      "learning_rate": 2.1569785026611605e-06,
+      "loss": 1.2955,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5018234865061999,
+      "grad_norm": 21.25,
+      "learning_rate": 2.1566458851326015e-06,
+      "loss": 0.8218,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5020666180403598,
+      "grad_norm": 15.125,
+      "learning_rate": 2.156313132090157e-06,
+      "loss": 0.6786,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5023097495745198,
+      "grad_norm": 18.875,
+      "learning_rate": 2.1559802435835623e-06,
+      "loss": 0.669,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5025528811086798,
+      "grad_norm": 16.375,
+      "learning_rate": 2.1556472196625733e-06,
+      "loss": 0.71,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5027960126428398,
+      "grad_norm": 16.25,
+      "learning_rate": 2.155314060376966e-06,
+      "loss": 0.6031,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5030391441769998,
+      "grad_norm": 19.375,
+      "learning_rate": 2.1549807657765375e-06,
+      "loss": 1.1718,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5032822757111597,
+      "grad_norm": 26.125,
+      "learning_rate": 2.1546473359111037e-06,
+      "loss": 0.8826,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5035254072453197,
+      "grad_norm": 19.625,
+      "learning_rate": 2.154313770830502e-06,
+      "loss": 0.8892,
+      "step": 2071
+    },
+    {
+      "epoch": 0.5037685387794797,
+      "grad_norm": 21.0,
+      "learning_rate": 2.1539800705845886e-06,
+      "loss": 1.0622,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5040116703136397,
+      "grad_norm": 17.375,
+      "learning_rate": 2.1536462352232416e-06,
+      "loss": 0.7742,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5042548018477997,
+      "grad_norm": 16.25,
+      "learning_rate": 2.153312264796359e-06,
+      "loss": 1.0432,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5044979333819597,
+      "grad_norm": 20.75,
+      "learning_rate": 2.152978159353857e-06,
+      "loss": 0.6966,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5047410649161196,
+      "grad_norm": 20.75,
+      "learning_rate": 2.152643918945674e-06,
+      "loss": 0.9008,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5049841964502796,
+      "grad_norm": 15.375,
+      "learning_rate": 2.1523095436217685e-06,
+      "loss": 0.6613,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5052273279844396,
+      "grad_norm": 20.75,
+      "learning_rate": 2.151975033432118e-06,
+      "loss": 1.3671,
+      "step": 2078
+    },
+    {
+      "epoch": 0.5054704595185996,
+      "grad_norm": 19.25,
+      "learning_rate": 2.151640388426721e-06,
+      "loss": 1.0413,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5057135910527596,
+      "grad_norm": 19.125,
+      "learning_rate": 2.151305608655597e-06,
+      "loss": 0.8604,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5059567225869195,
+      "grad_norm": 19.0,
+      "learning_rate": 2.1509706941687824e-06,
+      "loss": 0.794,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5061998541210795,
+      "grad_norm": 17.5,
+      "learning_rate": 2.150635645016338e-06,
+      "loss": 1.0629,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5064429856552395,
+      "grad_norm": 16.625,
+      "learning_rate": 2.150300461248342e-06,
+      "loss": 1.1524,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5066861171893995,
+      "grad_norm": 22.5,
+      "learning_rate": 2.149965142914893e-06,
+      "loss": 1.0185,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5069292487235595,
+      "grad_norm": 18.875,
+      "learning_rate": 2.1496296900661106e-06,
+      "loss": 1.0102,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5071723802577194,
+      "grad_norm": 16.75,
+      "learning_rate": 2.149294102752134e-06,
+      "loss": 0.8522,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5074155117918794,
+      "grad_norm": 17.75,
+      "learning_rate": 2.1489583810231217e-06,
+      "loss": 0.6616,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5076586433260394,
+      "grad_norm": 22.25,
+      "learning_rate": 2.148622524929255e-06,
+      "loss": 1.2406,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5079017748601994,
+      "grad_norm": 25.75,
+      "learning_rate": 2.148286534520731e-06,
+      "loss": 0.9268,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5081449063943594,
+      "grad_norm": 21.0,
+      "learning_rate": 2.147950409847771e-06,
+      "loss": 1.0492,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5083880379285193,
+      "grad_norm": 20.125,
+      "learning_rate": 2.1476141509606144e-06,
+      "loss": 1.084,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5086311694626793,
+      "grad_norm": 18.875,
+      "learning_rate": 2.1472777579095205e-06,
+      "loss": 0.9973,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5088743009968393,
+      "grad_norm": 16.125,
+      "learning_rate": 2.146941230744769e-06,
+      "loss": 0.3655,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5091174325309993,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.14660456951666e-06,
+      "loss": 0.7654,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5093605640651593,
+      "grad_norm": 18.625,
+      "learning_rate": 2.146267774275513e-06,
+      "loss": 0.9237,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5096036955993193,
+      "grad_norm": 17.25,
+      "learning_rate": 2.145930845071668e-06,
+      "loss": 0.7004,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5098468271334792,
+      "grad_norm": 26.5,
+      "learning_rate": 2.145593781955485e-06,
+      "loss": 0.9624,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5100899586676392,
+      "grad_norm": 22.75,
+      "learning_rate": 2.145256584977344e-06,
+      "loss": 0.8197,
+      "step": 2098
+    },
+    {
+      "epoch": 0.5103330902017992,
+      "grad_norm": 22.625,
+      "learning_rate": 2.1449192541876447e-06,
+      "loss": 1.3957,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5105762217359592,
+      "grad_norm": 25.5,
+      "learning_rate": 2.144581789636807e-06,
+      "loss": 0.9053,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5108193532701192,
+      "grad_norm": 17.125,
+      "learning_rate": 2.144244191375271e-06,
+      "loss": 0.8517,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5110624848042791,
+      "grad_norm": 17.125,
+      "learning_rate": 2.143906459453496e-06,
+      "loss": 0.8091,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5113056163384391,
+      "grad_norm": 23.125,
+      "learning_rate": 2.143568593921963e-06,
+      "loss": 0.8385,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5115487478725991,
+      "grad_norm": 20.75,
+      "learning_rate": 2.143230594831171e-06,
+      "loss": 0.9638,
+      "step": 2104
+    },
+    {
+      "epoch": 0.5117918794067591,
+      "grad_norm": 19.25,
+      "learning_rate": 2.1428924622316396e-06,
+      "loss": 0.9676,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5120350109409191,
+      "grad_norm": 19.125,
+      "learning_rate": 2.1425541961739093e-06,
+      "loss": 0.8893,
+      "step": 2106
+    },
+    {
+      "epoch": 0.512278142475079,
+      "grad_norm": 17.375,
+      "learning_rate": 2.1422157967085394e-06,
+      "loss": 0.8004,
+      "step": 2107
+    },
+    {
+      "epoch": 0.512521274009239,
+      "grad_norm": 25.625,
+      "learning_rate": 2.1418772638861095e-06,
+      "loss": 0.8912,
+      "step": 2108
+    },
+    {
+      "epoch": 0.512764405543399,
+      "grad_norm": 18.5,
+      "learning_rate": 2.141538597757219e-06,
+      "loss": 0.6878,
+      "step": 2109
+    },
+    {
+      "epoch": 0.513007537077559,
+      "grad_norm": 20.0,
+      "learning_rate": 2.141199798372488e-06,
+      "loss": 0.8896,
+      "step": 2110
+    },
+    {
+      "epoch": 0.513250668611719,
+      "grad_norm": 25.25,
+      "learning_rate": 2.140860865782556e-06,
+      "loss": 1.0707,
+      "step": 2111
+    },
+    {
+      "epoch": 0.513493800145879,
+      "grad_norm": 31.25,
+      "learning_rate": 2.1405218000380813e-06,
+      "loss": 0.9574,
+      "step": 2112
+    },
+    {
+      "epoch": 0.5137369316800389,
+      "grad_norm": 21.0,
+      "learning_rate": 2.1401826011897436e-06,
+      "loss": 0.8259,
+      "step": 2113
+    },
+    {
+      "epoch": 0.5139800632141989,
+      "grad_norm": 19.5,
+      "learning_rate": 2.1398432692882423e-06,
+      "loss": 1.1377,
+      "step": 2114
+    },
+    {
+      "epoch": 0.5142231947483589,
+      "grad_norm": 17.125,
+      "learning_rate": 2.1395038043842966e-06,
+      "loss": 0.5954,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5144663262825189,
+      "grad_norm": 21.75,
+      "learning_rate": 2.139164206528645e-06,
+      "loss": 0.9733,
+      "step": 2116
+    },
+    {
+      "epoch": 0.5147094578166789,
+      "grad_norm": 16.25,
+      "learning_rate": 2.138824475772046e-06,
+      "loss": 0.6823,
+      "step": 2117
+    },
+    {
+      "epoch": 0.5149525893508388,
+      "grad_norm": 15.375,
+      "learning_rate": 2.138484612165279e-06,
+      "loss": 1.3318,
+      "step": 2118
+    },
+    {
+      "epoch": 0.5151957208849988,
+      "grad_norm": 24.625,
+      "learning_rate": 2.138144615759142e-06,
+      "loss": 1.221,
+      "step": 2119
+    },
+    {
+      "epoch": 0.5154388524191588,
+      "grad_norm": 17.125,
+      "learning_rate": 2.137804486604453e-06,
+      "loss": 0.7901,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5156819839533188,
+      "grad_norm": 16.125,
+      "learning_rate": 2.1374642247520506e-06,
+      "loss": 0.819,
+      "step": 2121
+    },
+    {
+      "epoch": 0.5159251154874788,
+      "grad_norm": 17.375,
+      "learning_rate": 2.137123830252793e-06,
+      "loss": 0.6808,
+      "step": 2122
+    },
+    {
+      "epoch": 0.5161682470216387,
+      "grad_norm": 20.0,
+      "learning_rate": 2.1367833031575576e-06,
+      "loss": 0.6986,
+      "step": 2123
+    },
+    {
+      "epoch": 0.5164113785557987,
+      "grad_norm": 28.625,
+      "learning_rate": 2.1364426435172426e-06,
+      "loss": 0.9064,
+      "step": 2124
+    },
+    {
+      "epoch": 0.5166545100899587,
+      "grad_norm": 26.875,
+      "learning_rate": 2.136101851382765e-06,
+      "loss": 0.8917,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5168976416241187,
+      "grad_norm": 17.375,
+      "learning_rate": 2.1357609268050623e-06,
+      "loss": 0.7116,
+      "step": 2126
+    },
+    {
+      "epoch": 0.5171407731582787,
+      "grad_norm": 23.125,
+      "learning_rate": 2.135419869835091e-06,
+      "loss": 1.0916,
+      "step": 2127
+    },
+    {
+      "epoch": 0.5173839046924386,
+      "grad_norm": 26.25,
+      "learning_rate": 2.1350786805238287e-06,
+      "loss": 1.1954,
+      "step": 2128
+    },
+    {
+      "epoch": 0.5176270362265986,
+      "grad_norm": 28.0,
+      "learning_rate": 2.1347373589222718e-06,
+      "loss": 1.0237,
+      "step": 2129
+    },
+    {
+      "epoch": 0.5178701677607586,
+      "grad_norm": 18.75,
+      "learning_rate": 2.134395905081437e-06,
+      "loss": 0.6715,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5181132992949186,
+      "grad_norm": 18.25,
+      "learning_rate": 2.13405431905236e-06,
+      "loss": 1.0831,
+      "step": 2131
+    },
+    {
+      "epoch": 0.5183564308290786,
+      "grad_norm": 17.25,
+      "learning_rate": 2.1337126008860964e-06,
+      "loss": 0.7293,
+      "step": 2132
+    },
+    {
+      "epoch": 0.5185995623632386,
+      "grad_norm": 20.125,
+      "learning_rate": 2.1333707506337227e-06,
+      "loss": 0.7882,
+      "step": 2133
+    },
+    {
+      "epoch": 0.5188426938973985,
+      "grad_norm": 19.375,
+      "learning_rate": 2.133028768346334e-06,
+      "loss": 0.761,
+      "step": 2134
+    },
+    {
+      "epoch": 0.5190858254315585,
+      "grad_norm": 16.5,
+      "learning_rate": 2.132686654075045e-06,
+      "loss": 0.5887,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5193289569657185,
+      "grad_norm": 22.0,
+      "learning_rate": 2.132344407870992e-06,
+      "loss": 0.9873,
+      "step": 2136
+    },
+    {
+      "epoch": 0.5195720884998785,
+      "grad_norm": 28.25,
+      "learning_rate": 2.1320020297853276e-06,
+      "loss": 1.2165,
+      "step": 2137
+    },
+    {
+      "epoch": 0.5198152200340385,
+      "grad_norm": 18.75,
+      "learning_rate": 2.1316595198692274e-06,
+      "loss": 1.0632,
+      "step": 2138
+    },
+    {
+      "epoch": 0.5200583515681984,
+      "grad_norm": 25.625,
+      "learning_rate": 2.1313168781738855e-06,
+      "loss": 1.2686,
+      "step": 2139
+    },
+    {
+      "epoch": 0.5203014831023584,
+      "grad_norm": 17.125,
+      "learning_rate": 2.130974104750515e-06,
+      "loss": 0.75,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5205446146365184,
+      "grad_norm": 22.0,
+      "learning_rate": 2.13063119965035e-06,
+      "loss": 1.0972,
+      "step": 2141
+    },
+    {
+      "epoch": 0.5207877461706784,
+      "grad_norm": 16.125,
+      "learning_rate": 2.1302881629246426e-06,
+      "loss": 0.5599,
+      "step": 2142
+    },
+    {
+      "epoch": 0.5210308777048384,
+      "grad_norm": 19.875,
+      "learning_rate": 2.1299449946246666e-06,
+      "loss": 1.2064,
+      "step": 2143
+    },
+    {
+      "epoch": 0.5212740092389982,
+      "grad_norm": 21.75,
+      "learning_rate": 2.129601694801714e-06,
+      "loss": 1.2053,
+      "step": 2144
+    },
+    {
+      "epoch": 0.5215171407731582,
+      "grad_norm": 20.375,
+      "learning_rate": 2.1292582635070966e-06,
+      "loss": 0.8525,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5217602723073183,
+      "grad_norm": 14.3125,
+      "learning_rate": 2.128914700792146e-06,
+      "loss": 0.859,
+      "step": 2146
+    },
+    {
+      "epoch": 0.5220034038414783,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.1285710067082147e-06,
+      "loss": 0.4744,
+      "step": 2147
+    },
+    {
+      "epoch": 0.5222465353756383,
+      "grad_norm": 16.125,
+      "learning_rate": 2.128227181306673e-06,
+      "loss": 0.8446,
+      "step": 2148
+    },
+    {
+      "epoch": 0.5224896669097983,
+      "grad_norm": 34.5,
+      "learning_rate": 2.1278832246389116e-06,
+      "loss": 0.8184,
+      "step": 2149
+    },
+    {
+      "epoch": 0.5227327984439581,
+      "grad_norm": 24.375,
+      "learning_rate": 2.1275391367563403e-06,
+      "loss": 1.5662,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5229759299781181,
+      "grad_norm": 17.75,
+      "learning_rate": 2.1271949177103894e-06,
+      "loss": 0.6558,
+      "step": 2151
+    },
+    {
+      "epoch": 0.5232190615122781,
+      "grad_norm": 23.375,
+      "learning_rate": 2.1268505675525084e-06,
+      "loss": 0.8729,
+      "step": 2152
+    },
+    {
+      "epoch": 0.5234621930464382,
+      "grad_norm": 21.0,
+      "learning_rate": 2.1265060863341665e-06,
+      "loss": 1.039,
+      "step": 2153
+    },
+    {
+      "epoch": 0.5237053245805982,
+      "grad_norm": 17.5,
+      "learning_rate": 2.1261614741068522e-06,
+      "loss": 0.7471,
+      "step": 2154
+    },
+    {
+      "epoch": 0.523948456114758,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.1258167309220737e-06,
+      "loss": 0.5395,
+      "step": 2155
+    },
+    {
+      "epoch": 0.524191587648918,
+      "grad_norm": 17.125,
+      "learning_rate": 2.125471856831359e-06,
+      "loss": 0.7584,
+      "step": 2156
+    },
+    {
+      "epoch": 0.524434719183078,
+      "grad_norm": 41.25,
+      "learning_rate": 2.125126851886255e-06,
+      "loss": 0.7686,
+      "step": 2157
+    },
+    {
+      "epoch": 0.524677850717238,
+      "grad_norm": 15.5,
+      "learning_rate": 2.1247817161383295e-06,
+      "loss": 0.4855,
+      "step": 2158
+    },
+    {
+      "epoch": 0.524920982251398,
+      "grad_norm": 16.875,
+      "learning_rate": 2.1244364496391684e-06,
+      "loss": 0.7316,
+      "step": 2159
+    },
+    {
+      "epoch": 0.5251641137855579,
+      "grad_norm": 20.375,
+      "learning_rate": 2.124091052440378e-06,
+      "loss": 1.142,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5254072453197179,
+      "grad_norm": 22.75,
+      "learning_rate": 2.123745524593583e-06,
+      "loss": 0.729,
+      "step": 2161
+    },
+    {
+      "epoch": 0.5256503768538779,
+      "grad_norm": 21.0,
+      "learning_rate": 2.1233998661504297e-06,
+      "loss": 0.9577,
+      "step": 2162
+    },
+    {
+      "epoch": 0.525893508388038,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.1230540771625823e-06,
+      "loss": 0.657,
+      "step": 2163
+    },
+    {
+      "epoch": 0.526136639922198,
+      "grad_norm": 16.75,
+      "learning_rate": 2.122708157681725e-06,
+      "loss": 1.1673,
+      "step": 2164
+    },
+    {
+      "epoch": 0.5263797714563578,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.122362107759561e-06,
+      "loss": 0.7474,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5266229029905178,
+      "grad_norm": 19.625,
+      "learning_rate": 2.1220159274478145e-06,
+      "loss": 0.9294,
+      "step": 2166
+    },
+    {
+      "epoch": 0.5268660345246778,
+      "grad_norm": 17.25,
+      "learning_rate": 2.121669616798227e-06,
+      "loss": 0.7146,
+      "step": 2167
+    },
+    {
+      "epoch": 0.5271091660588378,
+      "grad_norm": 23.25,
+      "learning_rate": 2.1213231758625606e-06,
+      "loss": 0.9958,
+      "step": 2168
+    },
+    {
+      "epoch": 0.5273522975929978,
+      "grad_norm": 18.5,
+      "learning_rate": 2.1209766046925976e-06,
+      "loss": 0.7999,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5275954291271578,
+      "grad_norm": 15.0625,
+      "learning_rate": 2.120629903340139e-06,
+      "loss": 0.6183,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5278385606613177,
+      "grad_norm": 20.75,
+      "learning_rate": 2.120283071857005e-06,
+      "loss": 0.8308,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5280816921954777,
+      "grad_norm": 18.625,
+      "learning_rate": 2.1199361102950357e-06,
+      "loss": 0.6661,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5283248237296377,
+      "grad_norm": 22.375,
+      "learning_rate": 2.11958901870609e-06,
+      "loss": 0.8004,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5285679552637977,
+      "grad_norm": 23.5,
+      "learning_rate": 2.119241797142047e-06,
+      "loss": 1.2852,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5288110867979577,
+      "grad_norm": 27.0,
+      "learning_rate": 2.1188944456548054e-06,
+      "loss": 0.9778,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5290542183321176,
+      "grad_norm": 23.0,
+      "learning_rate": 2.1185469642962826e-06,
+      "loss": 0.8277,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5292973498662776,
+      "grad_norm": 17.75,
+      "learning_rate": 2.1181993531184156e-06,
+      "loss": 0.6125,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5295404814004376,
+      "grad_norm": 20.375,
+      "learning_rate": 2.117851612173161e-06,
+      "loss": 1.027,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5297836129345976,
+      "grad_norm": 17.875,
+      "learning_rate": 2.1175037415124947e-06,
+      "loss": 0.8801,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5300267444687576,
+      "grad_norm": 21.0,
+      "learning_rate": 2.1171557411884116e-06,
+      "loss": 1.0453,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5302698760029175,
+      "grad_norm": 25.875,
+      "learning_rate": 2.116807611252927e-06,
+      "loss": 1.467,
+      "step": 2181
+    },
+    {
+      "epoch": 0.5305130075370775,
+      "grad_norm": 16.875,
+      "learning_rate": 2.1164593517580746e-06,
+      "loss": 0.9979,
+      "step": 2182
+    },
+    {
+      "epoch": 0.5307561390712375,
+      "grad_norm": 15.375,
+      "learning_rate": 2.116110962755908e-06,
+      "loss": 0.6979,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5309992706053975,
+      "grad_norm": 31.0,
+      "learning_rate": 2.1157624442984993e-06,
+      "loss": 1.4906,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5312424021395575,
+      "grad_norm": 19.625,
+      "learning_rate": 2.115413796437941e-06,
+      "loss": 0.9826,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5314855336737175,
+      "grad_norm": 20.625,
+      "learning_rate": 2.115065019226345e-06,
+      "loss": 0.7923,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5317286652078774,
+      "grad_norm": 20.625,
+      "learning_rate": 2.114716112715842e-06,
+      "loss": 1.0459,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5319717967420374,
+      "grad_norm": 17.25,
+      "learning_rate": 2.114367076958581e-06,
+      "loss": 1.155,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5322149282761974,
+      "grad_norm": 18.625,
+      "learning_rate": 2.1140179120067324e-06,
+      "loss": 1.0049,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5324580598103574,
+      "grad_norm": 19.75,
+      "learning_rate": 2.113668617912485e-06,
+      "loss": 0.8863,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5327011913445174,
+      "grad_norm": 15.375,
+      "learning_rate": 2.1133191947280465e-06,
+      "loss": 0.7787,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5329443228786773,
+      "grad_norm": 21.125,
+      "learning_rate": 2.112969642505644e-06,
+      "loss": 0.8467,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5331874544128373,
+      "grad_norm": 22.25,
+      "learning_rate": 2.112619961297525e-06,
+      "loss": 0.7615,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5334305859469973,
+      "grad_norm": 23.625,
+      "learning_rate": 2.1122701511559548e-06,
+      "loss": 0.7331,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5336737174811573,
+      "grad_norm": 17.875,
+      "learning_rate": 2.1119202121332185e-06,
+      "loss": 0.751,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5339168490153173,
+      "grad_norm": 24.625,
+      "learning_rate": 2.11157014428162e-06,
+      "loss": 1.0139,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5341599805494772,
+      "grad_norm": 17.125,
+      "learning_rate": 2.111219947653484e-06,
+      "loss": 0.7356,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5344031120836372,
+      "grad_norm": 23.625,
+      "learning_rate": 2.1108696223011534e-06,
+      "loss": 1.0466,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5346462436177972,
+      "grad_norm": 19.0,
+      "learning_rate": 2.1105191682769895e-06,
+      "loss": 0.7842,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5348893751519572,
+      "grad_norm": 25.125,
+      "learning_rate": 2.1101685856333744e-06,
+      "loss": 1.2724,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5351325066861172,
+      "grad_norm": 20.5,
+      "learning_rate": 2.1098178744227088e-06,
+      "loss": 0.8252,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5353756382202771,
+      "grad_norm": 24.5,
+      "learning_rate": 2.109467034697412e-06,
+      "loss": 0.5705,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5356187697544371,
+      "grad_norm": 22.25,
+      "learning_rate": 2.1091160665099235e-06,
+      "loss": 0.9459,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5358619012885971,
+      "grad_norm": 19.875,
+      "learning_rate": 2.108764969912701e-06,
+      "loss": 1.189,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5361050328227571,
+      "grad_norm": 23.625,
+      "learning_rate": 2.108413744958223e-06,
+      "loss": 1.2873,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5363481643569171,
+      "grad_norm": 17.125,
+      "learning_rate": 2.108062391698985e-06,
+      "loss": 0.7731,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5365912958910771,
+      "grad_norm": 18.625,
+      "learning_rate": 2.1077109101875036e-06,
+      "loss": 0.8861,
+      "step": 2207
+    },
+    {
+      "epoch": 0.536834427425237,
+      "grad_norm": 18.5,
+      "learning_rate": 2.1073593004763134e-06,
+      "loss": 1.0385,
+      "step": 2208
+    },
+    {
+      "epoch": 0.537077558959397,
+      "grad_norm": 24.375,
+      "learning_rate": 2.1070075626179686e-06,
+      "loss": 0.8896,
+      "step": 2209
+    },
+    {
+      "epoch": 0.537320690493557,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.1066556966650427e-06,
+      "loss": 0.6847,
+      "step": 2210
+    },
+    {
+      "epoch": 0.537563822027717,
+      "grad_norm": 18.25,
+      "learning_rate": 2.1063037026701277e-06,
+      "loss": 0.8065,
+      "step": 2211
+    },
+    {
+      "epoch": 0.537806953561877,
+      "grad_norm": 24.75,
+      "learning_rate": 2.1059515806858357e-06,
+      "loss": 0.9747,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5380500850960369,
+      "grad_norm": 22.75,
+      "learning_rate": 2.105599330764797e-06,
+      "loss": 1.2146,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5382932166301969,
+      "grad_norm": 21.0,
+      "learning_rate": 2.105246952959662e-06,
+      "loss": 0.8581,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5385363481643569,
+      "grad_norm": 24.5,
+      "learning_rate": 2.104894447323099e-06,
+      "loss": 1.2318,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5387794796985169,
+      "grad_norm": 22.875,
+      "learning_rate": 2.104541813907796e-06,
+      "loss": 1.3717,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5390226112326769,
+      "grad_norm": 16.375,
+      "learning_rate": 2.104189052766461e-06,
+      "loss": 0.9492,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5392657427668368,
+      "grad_norm": 26.75,
+      "learning_rate": 2.1038361639518195e-06,
+      "loss": 0.9797,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5395088743009968,
+      "grad_norm": 35.5,
+      "learning_rate": 2.1034831475166166e-06,
+      "loss": 0.8712,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5397520058351568,
+      "grad_norm": 25.125,
+      "learning_rate": 2.103130003513618e-06,
+      "loss": 1.3308,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5399951373693168,
+      "grad_norm": 13.1875,
+      "learning_rate": 2.1027767319956055e-06,
+      "loss": 0.3165,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5402382689034768,
+      "grad_norm": 20.375,
+      "learning_rate": 2.1024233330153828e-06,
+      "loss": 0.7538,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5404814004376368,
+      "grad_norm": 20.25,
+      "learning_rate": 2.1020698066257707e-06,
+      "loss": 0.9751,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5407245319717967,
+      "grad_norm": 17.875,
+      "learning_rate": 2.101716152879611e-06,
+      "loss": 0.8774,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5409676635059567,
+      "grad_norm": 17.75,
+      "learning_rate": 2.1013623718297623e-06,
+      "loss": 0.8974,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5412107950401167,
+      "grad_norm": 18.875,
+      "learning_rate": 2.1010084635291036e-06,
+      "loss": 0.8043,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5414539265742767,
+      "grad_norm": 21.75,
+      "learning_rate": 2.1006544280305325e-06,
+      "loss": 0.8488,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5416970581084367,
+      "grad_norm": 22.0,
+      "learning_rate": 2.1003002653869658e-06,
+      "loss": 0.9424,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5419401896425966,
+      "grad_norm": 20.75,
+      "learning_rate": 2.099945975651339e-06,
+      "loss": 0.9605,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5421833211767566,
+      "grad_norm": 24.375,
+      "learning_rate": 2.0995915588766074e-06,
+      "loss": 1.0632,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5424264527109166,
+      "grad_norm": 25.0,
+      "learning_rate": 2.0992370151157444e-06,
+      "loss": 1.0454,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5426695842450766,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0988823444217426e-06,
+      "loss": 0.8084,
+      "step": 2232
+    },
+    {
+      "epoch": 0.5429127157792366,
+      "grad_norm": 17.75,
+      "learning_rate": 2.0985275468476137e-06,
+      "loss": 0.9918,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5431558473133965,
+      "grad_norm": 19.5,
+      "learning_rate": 2.098172622446388e-06,
+      "loss": 0.9706,
+      "step": 2234
+    },
+    {
+      "epoch": 0.5433989788475565,
+      "grad_norm": 23.625,
+      "learning_rate": 2.097817571271116e-06,
+      "loss": 0.8768,
+      "step": 2235
+    },
+    {
+      "epoch": 0.5436421103817165,
+      "grad_norm": 18.75,
+      "learning_rate": 2.0974623933748655e-06,
+      "loss": 0.9007,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5438852419158765,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.097107088810724e-06,
+      "loss": 0.5056,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5441283734500365,
+      "grad_norm": 17.25,
+      "learning_rate": 2.096751657631798e-06,
+      "loss": 1.0783,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5443715049841964,
+      "grad_norm": 22.875,
+      "learning_rate": 2.0963960998912132e-06,
+      "loss": 0.9539,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5446146365183564,
+      "grad_norm": 20.375,
+      "learning_rate": 2.0960404156421133e-06,
+      "loss": 0.9082,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5448577680525164,
+      "grad_norm": 23.375,
+      "learning_rate": 2.095684604937662e-06,
+      "loss": 1.0703,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5451008995866764,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.09532866783104e-06,
+      "loss": 0.6463,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5453440311208364,
+      "grad_norm": 18.5,
+      "learning_rate": 2.09497260437545e-06,
+      "loss": 0.8202,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5455871626549964,
+      "grad_norm": 21.0,
+      "learning_rate": 2.094616414624111e-06,
+      "loss": 0.8521,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5458302941891563,
+      "grad_norm": 33.0,
+      "learning_rate": 2.094260098630262e-06,
+      "loss": 1.0215,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5460734257233163,
+      "grad_norm": 22.375,
+      "learning_rate": 2.09390365644716e-06,
+      "loss": 1.2234,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5463165572574763,
+      "grad_norm": 20.5,
+      "learning_rate": 2.093547088128082e-06,
+      "loss": 0.7184,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5465596887916363,
+      "grad_norm": 34.25,
+      "learning_rate": 2.093190393726323e-06,
+      "loss": 1.2451,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5468028203257963,
+      "grad_norm": 14.375,
+      "learning_rate": 2.0928335732951976e-06,
+      "loss": 0.5257,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5470459518599562,
+      "grad_norm": 21.125,
+      "learning_rate": 2.0924766268880382e-06,
+      "loss": 0.7474,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5472890833941162,
+      "grad_norm": 34.0,
+      "learning_rate": 2.0921195545581967e-06,
+      "loss": 0.8276,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5475322149282762,
+      "grad_norm": 31.625,
+      "learning_rate": 2.091762356359044e-06,
+      "loss": 0.9416,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5477753464624362,
+      "grad_norm": 21.75,
+      "learning_rate": 2.0914050323439695e-06,
+      "loss": 1.4075,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5480184779965962,
+      "grad_norm": 24.5,
+      "learning_rate": 2.0910475825663813e-06,
+      "loss": 1.0821,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5482616095307561,
+      "grad_norm": 15.0,
+      "learning_rate": 2.0906900070797067e-06,
+      "loss": 0.6176,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5485047410649161,
+      "grad_norm": 12.8125,
+      "learning_rate": 2.090332305937391e-06,
+      "loss": 0.4484,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5487478725990761,
+      "grad_norm": 24.375,
+      "learning_rate": 2.089974479192899e-06,
+      "loss": 1.1816,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5489910041332361,
+      "grad_norm": 15.4375,
+      "learning_rate": 2.0896165268997145e-06,
+      "loss": 1.0037,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5492341356673961,
+      "grad_norm": 16.25,
+      "learning_rate": 2.089258449111339e-06,
+      "loss": 0.5961,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5494772672015561,
+      "grad_norm": 27.75,
+      "learning_rate": 2.088900245881294e-06,
+      "loss": 0.9547,
+      "step": 2260
+    },
+    {
+      "epoch": 0.549720398735716,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0885419172631192e-06,
+      "loss": 1.0561,
+      "step": 2261
+    },
+    {
+      "epoch": 0.549963530269876,
+      "grad_norm": 20.0,
+      "learning_rate": 2.088183463310373e-06,
+      "loss": 0.9799,
+      "step": 2262
+    },
+    {
+      "epoch": 0.550206661804036,
+      "grad_norm": 15.625,
+      "learning_rate": 2.0878248840766317e-06,
+      "loss": 0.5904,
+      "step": 2263
+    },
+    {
+      "epoch": 0.550449793338196,
+      "grad_norm": 24.75,
+      "learning_rate": 2.0874661796154916e-06,
+      "loss": 0.7311,
+      "step": 2264
+    },
+    {
+      "epoch": 0.550692924872356,
+      "grad_norm": 18.5,
+      "learning_rate": 2.087107349980568e-06,
+      "loss": 0.7602,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5509360564065159,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.086748395225493e-06,
+      "loss": 0.6092,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5511791879406759,
+      "grad_norm": 18.625,
+      "learning_rate": 2.08638931540392e-06,
+      "loss": 0.7429,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5514223194748359,
+      "grad_norm": 32.5,
+      "learning_rate": 2.086030110569518e-06,
+      "loss": 0.9813,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5516654510089959,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.0856707807759776e-06,
+      "loss": 0.6213,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5519085825431559,
+      "grad_norm": 21.5,
+      "learning_rate": 2.0853113260770063e-06,
+      "loss": 1.0692,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5521517140773158,
+      "grad_norm": 23.0,
+      "learning_rate": 2.0849517465263315e-06,
+      "loss": 1.0135,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5523948456114758,
+      "grad_norm": 18.0,
+      "learning_rate": 2.084592042177698e-06,
+      "loss": 1.0545,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5526379771456358,
+      "grad_norm": 12.3125,
+      "learning_rate": 2.084232213084869e-06,
+      "loss": 0.5815,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5528811086797958,
+      "grad_norm": 21.375,
+      "learning_rate": 2.0838722593016288e-06,
+      "loss": 0.8707,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5531242402139558,
+      "grad_norm": 25.625,
+      "learning_rate": 2.083512180881778e-06,
+      "loss": 0.76,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5533673717481157,
+      "grad_norm": 14.25,
+      "learning_rate": 2.0831519778791363e-06,
+      "loss": 0.4806,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5536105032822757,
+      "grad_norm": 13.75,
+      "learning_rate": 2.0827916503475425e-06,
+      "loss": 0.5682,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5538536348164357,
+      "grad_norm": 26.875,
+      "learning_rate": 2.082431198340854e-06,
+      "loss": 0.9897,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5540967663505957,
+      "grad_norm": 18.75,
+      "learning_rate": 2.082070621912946e-06,
+      "loss": 1.3372,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5543398978847557,
+      "grad_norm": 24.0,
+      "learning_rate": 2.0817099211177137e-06,
+      "loss": 1.2054,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5545830294189157,
+      "grad_norm": 26.5,
+      "learning_rate": 2.081349096009069e-06,
+      "loss": 0.9381,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5548261609530756,
+      "grad_norm": 22.375,
+      "learning_rate": 2.0809881466409444e-06,
+      "loss": 0.9875,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5550692924872356,
+      "grad_norm": 23.375,
+      "learning_rate": 2.080627073067289e-06,
+      "loss": 0.9427,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5553124240213956,
+      "grad_norm": 18.75,
+      "learning_rate": 2.0802658753420728e-06,
+      "loss": 0.8702,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 17.5,
+      "learning_rate": 2.0799045535192817e-06,
+      "loss": 0.5937,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5557986870897156,
+      "grad_norm": 21.75,
+      "learning_rate": 2.0795431076529226e-06,
+      "loss": 1.0784,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5560418186238755,
+      "grad_norm": 21.25,
+      "learning_rate": 2.0791815377970197e-06,
+      "loss": 1.1514,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5562849501580355,
+      "grad_norm": 16.25,
+      "learning_rate": 2.078819844005615e-06,
+      "loss": 0.6522,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5565280816921955,
+      "grad_norm": 17.5,
+      "learning_rate": 2.07845802633277e-06,
+      "loss": 0.9068,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5567712132263555,
+      "grad_norm": 21.0,
+      "learning_rate": 2.078096084832566e-06,
+      "loss": 0.8342,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5570143447605155,
+      "grad_norm": 21.5,
+      "learning_rate": 2.0777340195590996e-06,
+      "loss": 0.6393,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5572574762946754,
+      "grad_norm": 21.375,
+      "learning_rate": 2.0773718305664887e-06,
+      "loss": 0.6652,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5575006078288354,
+      "grad_norm": 22.75,
+      "learning_rate": 2.0770095179088688e-06,
+      "loss": 0.7814,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5577437393629954,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0766470816403935e-06,
+      "loss": 0.9092,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5579868708971554,
+      "grad_norm": 21.5,
+      "learning_rate": 2.076284521815235e-06,
+      "loss": 0.9198,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5582300024313154,
+      "grad_norm": 21.125,
+      "learning_rate": 2.075921838487584e-06,
+      "loss": 0.7984,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5584731339654754,
+      "grad_norm": 17.75,
+      "learning_rate": 2.07555903171165e-06,
+      "loss": 0.8968,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5587162654996353,
+      "grad_norm": 22.0,
+      "learning_rate": 2.0751961015416617e-06,
+      "loss": 1.0454,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5589593970337953,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.0748330480318637e-06,
+      "loss": 0.7281,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5592025285679553,
+      "grad_norm": 15.1875,
+      "learning_rate": 2.0744698712365215e-06,
+      "loss": 0.7593,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5594456601021153,
+      "grad_norm": 22.25,
+      "learning_rate": 2.074106571209918e-06,
+      "loss": 0.8927,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5596887916362753,
+      "grad_norm": 16.75,
+      "learning_rate": 2.073743148006354e-06,
+      "loss": 0.9008,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5599319231704352,
+      "grad_norm": 30.25,
+      "learning_rate": 2.073379601680151e-06,
+      "loss": 0.936,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5601750547045952,
+      "grad_norm": 19.25,
+      "learning_rate": 2.0730159322856454e-06,
+      "loss": 0.7952,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5604181862387552,
+      "grad_norm": 17.5,
+      "learning_rate": 2.0726521398771956e-06,
+      "loss": 0.6307,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5606613177729152,
+      "grad_norm": 19.125,
+      "learning_rate": 2.0722882245091753e-06,
+      "loss": 0.8667,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5609044493070752,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.0719241862359786e-06,
+      "loss": 0.623,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5611475808412351,
+      "grad_norm": 21.125,
+      "learning_rate": 2.0715600251120167e-06,
+      "loss": 1.173,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5613907123753951,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0711957411917207e-06,
+      "loss": 0.8189,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5616338439095551,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0708313345295384e-06,
+      "loss": 0.6672,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5618769754437151,
+      "grad_norm": 19.125,
+      "learning_rate": 2.070466805179937e-06,
+      "loss": 0.7677,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5621201069778751,
+      "grad_norm": 18.625,
+      "learning_rate": 2.0701021531974014e-06,
+      "loss": 0.9986,
+      "step": 2312
+    },
+    {
+      "epoch": 0.562363238512035,
+      "grad_norm": 26.0,
+      "learning_rate": 2.0697373786364357e-06,
+      "loss": 0.9489,
+      "step": 2313
+    },
+    {
+      "epoch": 0.562606370046195,
+      "grad_norm": 42.75,
+      "learning_rate": 2.0693724815515612e-06,
+      "loss": 1.7043,
+      "step": 2314
+    },
+    {
+      "epoch": 0.562849501580355,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0690074619973185e-06,
+      "loss": 0.7949,
+      "step": 2315
+    },
+    {
+      "epoch": 0.563092633114515,
+      "grad_norm": 19.875,
+      "learning_rate": 2.0686423200282652e-06,
+      "loss": 1.1615,
+      "step": 2316
+    },
+    {
+      "epoch": 0.563335764648675,
+      "grad_norm": 16.875,
+      "learning_rate": 2.0682770556989797e-06,
+      "loss": 0.8947,
+      "step": 2317
+    },
+    {
+      "epoch": 0.563578896182835,
+      "grad_norm": 15.375,
+      "learning_rate": 2.0679116690640556e-06,
+      "loss": 0.5107,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5638220277169949,
+      "grad_norm": 17.75,
+      "learning_rate": 2.0675461601781067e-06,
+      "loss": 0.8202,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5640651592511549,
+      "grad_norm": 20.75,
+      "learning_rate": 2.0671805290957646e-06,
+      "loss": 1.0881,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5643082907853149,
+      "grad_norm": 19.125,
+      "learning_rate": 2.0668147758716792e-06,
+      "loss": 0.9051,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5645514223194749,
+      "grad_norm": 28.375,
+      "learning_rate": 2.0664489005605187e-06,
+      "loss": 0.9758,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5647945538536349,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0660829032169695e-06,
+      "loss": 0.8816,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5650376853877948,
+      "grad_norm": 17.875,
+      "learning_rate": 2.0657167838957365e-06,
+      "loss": 0.7318,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5652808169219548,
+      "grad_norm": 17.5,
+      "learning_rate": 2.065350542651542e-06,
+      "loss": 0.6686,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5655239484561148,
+      "grad_norm": 12.125,
+      "learning_rate": 2.064984179539127e-06,
+      "loss": 0.3821,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5657670799902748,
+      "grad_norm": 22.375,
+      "learning_rate": 2.064617694613251e-06,
+      "loss": 1.1379,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5660102115244348,
+      "grad_norm": 17.75,
+      "learning_rate": 2.0642510879286924e-06,
+      "loss": 0.8428,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5662533430585946,
+      "grad_norm": 20.0,
+      "learning_rate": 2.0638843595402456e-06,
+      "loss": 1.07,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5664964745927547,
+      "grad_norm": 27.625,
+      "learning_rate": 2.063517509502725e-06,
+      "loss": 1.1483,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5667396061269147,
+      "grad_norm": 20.25,
+      "learning_rate": 2.063150537870963e-06,
+      "loss": 0.9096,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5669827376610747,
+      "grad_norm": 19.25,
+      "learning_rate": 2.062783444699809e-06,
+      "loss": 0.633,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5672258691952347,
+      "grad_norm": 30.875,
+      "learning_rate": 2.0624162300441327e-06,
+      "loss": 1.0003,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5674690007293947,
+      "grad_norm": 30.375,
+      "learning_rate": 2.062048893958819e-06,
+      "loss": 1.4803,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5677121322635545,
+      "grad_norm": 30.75,
+      "learning_rate": 2.0616814364987738e-06,
+      "loss": 1.3023,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5679552637977145,
+      "grad_norm": 23.375,
+      "learning_rate": 2.0613138577189203e-06,
+      "loss": 0.8766,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5681983953318746,
+      "grad_norm": 21.0,
+      "learning_rate": 2.060946157674198e-06,
+      "loss": 0.7785,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5684415268660346,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0605783364195676e-06,
+      "loss": 0.9711,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5686846584001946,
+      "grad_norm": 17.625,
+      "learning_rate": 2.060210394010005e-06,
+      "loss": 0.9325,
+      "step": 2339
+    },
+    {
+      "epoch": 0.5689277899343544,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0598423305005065e-06,
+      "loss": 0.8241,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5691709214685144,
+      "grad_norm": 17.875,
+      "learning_rate": 2.059474145946086e-06,
+      "loss": 0.7612,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5694140530026744,
+      "grad_norm": 16.0,
+      "learning_rate": 2.0591058404017735e-06,
+      "loss": 0.7962,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5696571845368344,
+      "grad_norm": 18.75,
+      "learning_rate": 2.058737413922619e-06,
+      "loss": 1.0329,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5699003160709945,
+      "grad_norm": 30.375,
+      "learning_rate": 2.0583688665636915e-06,
+      "loss": 1.0044,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5701434476051543,
+      "grad_norm": 21.75,
+      "learning_rate": 2.0580001983800752e-06,
+      "loss": 1.1053,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5703865791393143,
+      "grad_norm": 18.5,
+      "learning_rate": 2.0576314094268753e-06,
+      "loss": 1.1189,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5706297106734743,
+      "grad_norm": 16.375,
+      "learning_rate": 2.0572624997592124e-06,
+      "loss": 0.6924,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5708728422076343,
+      "grad_norm": 26.125,
+      "learning_rate": 2.0568934694322274e-06,
+      "loss": 1.156,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5711159737417943,
+      "grad_norm": 16.625,
+      "learning_rate": 2.0565243185010776e-06,
+      "loss": 0.633,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5713591052759542,
+      "grad_norm": 26.875,
+      "learning_rate": 2.056155047020939e-06,
+      "loss": 0.922,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5716022368101142,
+      "grad_norm": 19.875,
+      "learning_rate": 2.055785655047006e-06,
+      "loss": 0.7393,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5718453683442742,
+      "grad_norm": 16.5,
+      "learning_rate": 2.055416142634491e-06,
+      "loss": 0.7964,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5720884998784342,
+      "grad_norm": 20.5,
+      "learning_rate": 2.055046509838623e-06,
+      "loss": 0.9522,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5723316314125942,
+      "grad_norm": 14.625,
+      "learning_rate": 2.05467675671465e-06,
+      "loss": 0.7326,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5725747629467542,
+      "grad_norm": 20.0,
+      "learning_rate": 2.0543068833178394e-06,
+      "loss": 0.659,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5728178944809141,
+      "grad_norm": 20.125,
+      "learning_rate": 2.053936889703474e-06,
+      "loss": 0.7541,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5730610260150741,
+      "grad_norm": 16.625,
+      "learning_rate": 2.0535667759268556e-06,
+      "loss": 0.9131,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5733041575492341,
+      "grad_norm": 29.125,
+      "learning_rate": 2.0531965420433046e-06,
+      "loss": 1.2867,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5735472890833941,
+      "grad_norm": 21.25,
+      "learning_rate": 2.0528261881081587e-06,
+      "loss": 1.1126,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5737904206175541,
+      "grad_norm": 17.375,
+      "learning_rate": 2.052455714176774e-06,
+      "loss": 0.9647,
+      "step": 2360
+    },
+    {
+      "epoch": 0.574033552151714,
+      "grad_norm": 25.375,
+      "learning_rate": 2.0520851203045236e-06,
+      "loss": 0.9846,
+      "step": 2361
+    },
+    {
+      "epoch": 0.574276683685874,
+      "grad_norm": 20.5,
+      "learning_rate": 2.0517144065467993e-06,
+      "loss": 0.9121,
+      "step": 2362
+    },
+    {
+      "epoch": 0.574519815220034,
+      "grad_norm": 19.25,
+      "learning_rate": 2.0513435729590114e-06,
+      "loss": 0.7062,
+      "step": 2363
+    },
+    {
+      "epoch": 0.574762946754194,
+      "grad_norm": 20.5,
+      "learning_rate": 2.0509726195965863e-06,
+      "loss": 0.8853,
+      "step": 2364
+    },
+    {
+      "epoch": 0.575006078288354,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0506015465149704e-06,
+      "loss": 0.7734,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5752492098225139,
+      "grad_norm": 17.5,
+      "learning_rate": 2.050230353769626e-06,
+      "loss": 0.7197,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5754923413566739,
+      "grad_norm": 20.125,
+      "learning_rate": 2.049859041416035e-06,
+      "loss": 0.8364,
+      "step": 2367
+    },
+    {
+      "epoch": 0.5757354728908339,
+      "grad_norm": 14.5,
+      "learning_rate": 2.0494876095096964e-06,
+      "loss": 0.7339,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5759786044249939,
+      "grad_norm": 14.9375,
+      "learning_rate": 2.049116058106127e-06,
+      "loss": 0.757,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5762217359591539,
+      "grad_norm": 21.75,
+      "learning_rate": 2.0487443872608613e-06,
+      "loss": 1.1455,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5764648674933139,
+      "grad_norm": 21.5,
+      "learning_rate": 2.048372597029452e-06,
+      "loss": 0.7505,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5767079990274738,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.04800068746747e-06,
+      "loss": 0.6198,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5769511305616338,
+      "grad_norm": 18.0,
+      "learning_rate": 2.047628658630503e-06,
+      "loss": 0.9446,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5771942620957938,
+      "grad_norm": 15.5,
+      "learning_rate": 2.0472565105741578e-06,
+      "loss": 1.1734,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5774373936299538,
+      "grad_norm": 19.625,
+      "learning_rate": 2.0468842433540576e-06,
+      "loss": 1.1002,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5776805251641138,
+      "grad_norm": 16.625,
+      "learning_rate": 2.046511857025845e-06,
+      "loss": 1.0567,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5779236566982737,
+      "grad_norm": 17.875,
+      "learning_rate": 2.0461393516451785e-06,
+      "loss": 0.7922,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5781667882324337,
+      "grad_norm": 24.375,
+      "learning_rate": 2.0457667272677366e-06,
+      "loss": 1.0617,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5784099197665937,
+      "grad_norm": 24.375,
+      "learning_rate": 2.0453939839492137e-06,
+      "loss": 0.9944,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5786530513007537,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0450211217453235e-06,
+      "loss": 0.962,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5788961828349137,
+      "grad_norm": 15.5,
+      "learning_rate": 2.0446481407117953e-06,
+      "loss": 0.5756,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5791393143690736,
+      "grad_norm": 17.75,
+      "learning_rate": 2.044275040904379e-06,
+      "loss": 1.2556,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5793824459032336,
+      "grad_norm": 22.875,
+      "learning_rate": 2.0439018223788402e-06,
+      "loss": 0.9041,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5796255774373936,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.043528485190963e-06,
+      "loss": 0.414,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5798687089715536,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0431550293965486e-06,
+      "loss": 0.8813,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5801118405057136,
+      "grad_norm": 21.75,
+      "learning_rate": 2.042781455051417e-06,
+      "loss": 0.7461,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5803549720398735,
+      "grad_norm": 19.75,
+      "learning_rate": 2.042407762211405e-06,
+      "loss": 1.0001,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5805981035740335,
+      "grad_norm": 19.125,
+      "learning_rate": 2.042033950932368e-06,
+      "loss": 1.0402,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5808412351081935,
+      "grad_norm": 20.125,
+      "learning_rate": 2.0416600212701777e-06,
+      "loss": 1.1356,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5810843666423535,
+      "grad_norm": 21.125,
+      "learning_rate": 2.041285973280725e-06,
+      "loss": 0.9779,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5813274981765135,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.0409118070199175e-06,
+      "loss": 0.5775,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5815706297106735,
+      "grad_norm": 19.375,
+      "learning_rate": 2.040537522543681e-06,
+      "loss": 0.9431,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5818137612448334,
+      "grad_norm": 18.625,
+      "learning_rate": 2.0401631199079584e-06,
+      "loss": 0.9297,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5820568927789934,
+      "grad_norm": 25.0,
+      "learning_rate": 2.039788599168711e-06,
+      "loss": 1.3637,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5823000243131534,
+      "grad_norm": 23.75,
+      "learning_rate": 2.0394139603819176e-06,
+      "loss": 1.4877,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5825431558473134,
+      "grad_norm": 24.75,
+      "learning_rate": 2.039039203603574e-06,
+      "loss": 1.1025,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5827862873814734,
+      "grad_norm": 23.625,
+      "learning_rate": 2.0386643288896944e-06,
+      "loss": 1.1648,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5830294189156333,
+      "grad_norm": 20.375,
+      "learning_rate": 2.0382893362963102e-06,
+      "loss": 0.6469,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5832725504497933,
+      "grad_norm": 20.375,
+      "learning_rate": 2.0379142258794703e-06,
+      "loss": 0.8905,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5835156819839533,
+      "grad_norm": 14.4375,
+      "learning_rate": 2.0375389976952416e-06,
+      "loss": 0.4437,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5837588135181133,
+      "grad_norm": 19.625,
+      "learning_rate": 2.0371636517997085e-06,
+      "loss": 1.5071,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5840019450522733,
+      "grad_norm": 18.5,
+      "learning_rate": 2.0367881882489727e-06,
+      "loss": 0.8795,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5842450765864332,
+      "grad_norm": 21.5,
+      "learning_rate": 2.0364126070991543e-06,
+      "loss": 0.9467,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5844882081205932,
+      "grad_norm": 19.75,
+      "learning_rate": 2.036036908406389e-06,
+      "loss": 1.3867,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5847313396547532,
+      "grad_norm": 20.125,
+      "learning_rate": 2.0356610922268335e-06,
+      "loss": 0.7784,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5849744711889132,
+      "grad_norm": 23.125,
+      "learning_rate": 2.035285158616658e-06,
+      "loss": 0.912,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5852176027230732,
+      "grad_norm": 26.125,
+      "learning_rate": 2.034909107632054e-06,
+      "loss": 1.0614,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5854607342572332,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0345329393292272e-06,
+      "loss": 0.643,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5857038657913931,
+      "grad_norm": 19.625,
+      "learning_rate": 2.034156653764404e-06,
+      "loss": 0.7394,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5859469973255531,
+      "grad_norm": 18.625,
+      "learning_rate": 2.033780250993826e-06,
+      "loss": 0.7713,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5861901288597131,
+      "grad_norm": 21.5,
+      "learning_rate": 2.033403731073753e-06,
+      "loss": 1.119,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5864332603938731,
+      "grad_norm": 19.75,
+      "learning_rate": 2.033027094060462e-06,
+      "loss": 1.2968,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5866763919280331,
+      "grad_norm": 15.5,
+      "learning_rate": 2.0326503400102494e-06,
+      "loss": 0.6065,
+      "step": 2413
+    },
+    {
+      "epoch": 0.586919523462193,
+      "grad_norm": 18.0,
+      "learning_rate": 2.0322734689794262e-06,
+      "loss": 0.7435,
+      "step": 2414
+    },
+    {
+      "epoch": 0.587162654996353,
+      "grad_norm": 14.75,
+      "learning_rate": 2.0318964810243224e-06,
+      "loss": 0.4709,
+      "step": 2415
+    },
+    {
+      "epoch": 0.587405786530513,
+      "grad_norm": 27.125,
+      "learning_rate": 2.031519376201286e-06,
+      "loss": 0.9531,
+      "step": 2416
+    },
+    {
+      "epoch": 0.587648918064673,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0311421545666817e-06,
+      "loss": 0.7211,
+      "step": 2417
+    },
+    {
+      "epoch": 0.587892049598833,
+      "grad_norm": 17.625,
+      "learning_rate": 2.0307648161768914e-06,
+      "loss": 0.83,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5881351811329929,
+      "grad_norm": 17.25,
+      "learning_rate": 2.030387361088315e-06,
+      "loss": 0.8202,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5883783126671529,
+      "grad_norm": 13.625,
+      "learning_rate": 2.0300097893573694e-06,
+      "loss": 0.3608,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5886214442013129,
+      "grad_norm": 20.375,
+      "learning_rate": 2.02963210104049e-06,
+      "loss": 0.7299,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5888645757354729,
+      "grad_norm": 17.125,
+      "learning_rate": 2.0292542961941285e-06,
+      "loss": 0.8545,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5891077072696329,
+      "grad_norm": 22.0,
+      "learning_rate": 2.028876374874754e-06,
+      "loss": 1.1379,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5893508388037928,
+      "grad_norm": 18.75,
+      "learning_rate": 2.028498337138853e-06,
+      "loss": 1.0685,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5895939703379528,
+      "grad_norm": 15.3125,
+      "learning_rate": 2.0281201830429316e-06,
+      "loss": 0.9339,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5898371018721128,
+      "grad_norm": 17.75,
+      "learning_rate": 2.02774191264351e-06,
+      "loss": 1.0232,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5900802334062728,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0273635259971268e-06,
+      "loss": 1.0813,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5903233649404328,
+      "grad_norm": 23.625,
+      "learning_rate": 2.0269850231603393e-06,
+      "loss": 1.1019,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5905664964745928,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0266064041897216e-06,
+      "loss": 0.6596,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5908096280087527,
+      "grad_norm": 22.25,
+      "learning_rate": 2.026227669141864e-06,
+      "loss": 1.3176,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5910527595429127,
+      "grad_norm": 19.875,
+      "learning_rate": 2.0258488180733755e-06,
+      "loss": 0.7592,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5912958910770727,
+      "grad_norm": 17.875,
+      "learning_rate": 2.0254698510408815e-06,
+      "loss": 1.0825,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5915390226112327,
+      "grad_norm": 17.0,
+      "learning_rate": 2.0250907681010255e-06,
+      "loss": 1.1475,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5917821541453927,
+      "grad_norm": 17.125,
+      "learning_rate": 2.024711569310468e-06,
+      "loss": 0.8014,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5920252856795526,
+      "grad_norm": 20.0,
+      "learning_rate": 2.0243322547258866e-06,
+      "loss": 1.1512,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5922684172137126,
+      "grad_norm": 23.0,
+      "learning_rate": 2.0239528244039767e-06,
+      "loss": 0.7642,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5925115487478726,
+      "grad_norm": 23.75,
+      "learning_rate": 2.0235732784014507e-06,
+      "loss": 1.2959,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5927546802820326,
+      "grad_norm": 16.875,
+      "learning_rate": 2.0231936167750378e-06,
+      "loss": 0.6246,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5929978118161926,
+      "grad_norm": 24.125,
+      "learning_rate": 2.0228138395814854e-06,
+      "loss": 1.1494,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5932409433503525,
+      "grad_norm": 13.25,
+      "learning_rate": 2.022433946877558e-06,
+      "loss": 0.3006,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5934840748845125,
+      "grad_norm": 37.5,
+      "learning_rate": 2.0220539387200365e-06,
+      "loss": 1.3829,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5937272064186725,
+      "grad_norm": 15.75,
+      "learning_rate": 2.0216738151657208e-06,
+      "loss": 0.7871,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5939703379528325,
+      "grad_norm": 15.25,
+      "learning_rate": 2.0212935762714254e-06,
+      "loss": 0.817,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5942134694869925,
+      "grad_norm": 15.125,
+      "learning_rate": 2.0209132220939845e-06,
+      "loss": 0.5611,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5944566010211525,
+      "grad_norm": 17.375,
+      "learning_rate": 2.0205327526902486e-06,
+      "loss": 0.873,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5946997325553124,
+      "grad_norm": 17.125,
+      "learning_rate": 2.020152168117085e-06,
+      "loss": 1.2222,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5949428640894724,
+      "grad_norm": 21.875,
+      "learning_rate": 2.0197714684313786e-06,
+      "loss": 0.689,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5951859956236324,
+      "grad_norm": 22.5,
+      "learning_rate": 2.019390653690033e-06,
+      "loss": 1.0182,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5954291271577924,
+      "grad_norm": 20.0,
+      "learning_rate": 2.019009723949965e-06,
+      "loss": 1.0079,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5956722586919524,
+      "grad_norm": 23.5,
+      "learning_rate": 2.018628679268113e-06,
+      "loss": 0.9264,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5959153902261123,
+      "grad_norm": 21.625,
+      "learning_rate": 2.0182475197014306e-06,
+      "loss": 0.9782,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5961585217602723,
+      "grad_norm": 16.25,
+      "learning_rate": 2.0178662453068877e-06,
+      "loss": 0.6295,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5964016532944323,
+      "grad_norm": 19.375,
+      "learning_rate": 2.0174848561414734e-06,
+      "loss": 0.6381,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5966447848285923,
+      "grad_norm": 17.625,
+      "learning_rate": 2.017103352262192e-06,
+      "loss": 0.589,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5968879163627523,
+      "grad_norm": 16.875,
+      "learning_rate": 2.0167217337260665e-06,
+      "loss": 0.4894,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5971310478969122,
+      "grad_norm": 17.875,
+      "learning_rate": 2.0163400005901362e-06,
+      "loss": 0.8663,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5973741794310722,
+      "grad_norm": 13.875,
+      "learning_rate": 2.015958152911458e-06,
+      "loss": 0.6678,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5976173109652322,
+      "grad_norm": 16.625,
+      "learning_rate": 2.0155761907471043e-06,
+      "loss": 0.7652,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5978604424993922,
+      "grad_norm": 23.375,
+      "learning_rate": 2.015194114154168e-06,
+      "loss": 0.9417,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5981035740335522,
+      "grad_norm": 18.125,
+      "learning_rate": 2.0148119231897556e-06,
+      "loss": 0.912,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5983467055677121,
+      "grad_norm": 24.0,
+      "learning_rate": 2.0144296179109923e-06,
+      "loss": 0.8892,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5985898371018721,
+      "grad_norm": 26.25,
+      "learning_rate": 2.0140471983750205e-06,
+      "loss": 0.9228,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5988329686360321,
+      "grad_norm": 18.75,
+      "learning_rate": 2.0136646646389996e-06,
+      "loss": 1.0494,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5990761001701921,
+      "grad_norm": 15.5625,
+      "learning_rate": 2.013282016760105e-06,
+      "loss": 0.5649,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5993192317043521,
+      "grad_norm": 39.75,
+      "learning_rate": 2.0128992547955315e-06,
+      "loss": 1.1498,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5995623632385121,
+      "grad_norm": 22.125,
+      "learning_rate": 2.012516378802488e-06,
+      "loss": 1.0061,
+      "step": 2466
+    },
+    {
+      "epoch": 0.599805494772672,
+      "grad_norm": 23.75,
+      "learning_rate": 2.0121333888382032e-06,
+      "loss": 1.245,
+      "step": 2467
+    },
+    {
+      "epoch": 0.600048626306832,
+      "grad_norm": 16.5,
+      "learning_rate": 2.0117502849599204e-06,
+      "loss": 0.6547,
+      "step": 2468
+    },
+    {
+      "epoch": 0.600291757840992,
+      "grad_norm": 19.0,
+      "learning_rate": 2.011367067224902e-06,
+      "loss": 0.6211,
+      "step": 2469
+    },
+    {
+      "epoch": 0.600534889375152,
+      "grad_norm": 19.5,
+      "learning_rate": 2.0109837356904257e-06,
+      "loss": 0.8737,
+      "step": 2470
+    },
+    {
+      "epoch": 0.600778020909312,
+      "grad_norm": 23.375,
+      "learning_rate": 2.0106002904137877e-06,
+      "loss": 0.9821,
+      "step": 2471
+    },
+    {
+      "epoch": 0.6010211524434719,
+      "grad_norm": 18.125,
+      "learning_rate": 2.0102167314523004e-06,
+      "loss": 0.9827,
+      "step": 2472
+    },
+    {
+      "epoch": 0.6012642839776319,
+      "grad_norm": 16.75,
+      "learning_rate": 2.009833058863293e-06,
+      "loss": 0.8096,
+      "step": 2473
+    },
+    {
+      "epoch": 0.6015074155117919,
+      "grad_norm": 24.25,
+      "learning_rate": 2.0094492727041124e-06,
+      "loss": 0.8818,
+      "step": 2474
+    },
+    {
+      "epoch": 0.6017505470459519,
+      "grad_norm": 14.375,
+      "learning_rate": 2.009065373032122e-06,
+      "loss": 0.4699,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6019936785801119,
+      "grad_norm": 14.125,
+      "learning_rate": 2.0086813599047012e-06,
+      "loss": 0.6093,
+      "step": 2476
+    },
+    {
+      "epoch": 0.6022368101142718,
+      "grad_norm": 17.25,
+      "learning_rate": 2.0082972333792496e-06,
+      "loss": 0.6792,
+      "step": 2477
+    },
+    {
+      "epoch": 0.6024799416484318,
+      "grad_norm": 13.5,
+      "learning_rate": 2.007912993513179e-06,
+      "loss": 0.4687,
+      "step": 2478
+    },
+    {
+      "epoch": 0.6027230731825918,
+      "grad_norm": 23.875,
+      "learning_rate": 2.0075286403639226e-06,
+      "loss": 0.792,
+      "step": 2479
+    },
+    {
+      "epoch": 0.6029662047167518,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.0071441739889278e-06,
+      "loss": 0.7047,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6032093362509118,
+      "grad_norm": 25.0,
+      "learning_rate": 2.0067595944456598e-06,
+      "loss": 1.1387,
+      "step": 2481
+    },
+    {
+      "epoch": 0.6034524677850718,
+      "grad_norm": 37.25,
+      "learning_rate": 2.006374901791601e-06,
+      "loss": 1.6589,
+      "step": 2482
+    },
+    {
+      "epoch": 0.6036955993192317,
+      "grad_norm": 16.75,
+      "learning_rate": 2.0059900960842493e-06,
+      "loss": 0.6608,
+      "step": 2483
+    },
+    {
+      "epoch": 0.6039387308533917,
+      "grad_norm": 20.625,
+      "learning_rate": 2.005605177381122e-06,
+      "loss": 0.7116,
+      "step": 2484
+    },
+    {
+      "epoch": 0.6041818623875517,
+      "grad_norm": 32.5,
+      "learning_rate": 2.0052201457397507e-06,
+      "loss": 1.7751,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6044249939217117,
+      "grad_norm": 25.125,
+      "learning_rate": 2.004835001217686e-06,
+      "loss": 1.0185,
+      "step": 2486
+    },
+    {
+      "epoch": 0.6046681254558717,
+      "grad_norm": 15.75,
+      "learning_rate": 2.004449743872494e-06,
+      "loss": 0.5704,
+      "step": 2487
+    },
+    {
+      "epoch": 0.6049112569900316,
+      "grad_norm": 19.375,
+      "learning_rate": 2.0040643737617577e-06,
+      "loss": 0.7855,
+      "step": 2488
+    },
+    {
+      "epoch": 0.6051543885241916,
+      "grad_norm": 14.5625,
+      "learning_rate": 2.0036788909430774e-06,
+      "loss": 0.4763,
+      "step": 2489
+    },
+    {
+      "epoch": 0.6053975200583516,
+      "grad_norm": 20.75,
+      "learning_rate": 2.0032932954740707e-06,
+      "loss": 0.9844,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6056406515925116,
+      "grad_norm": 17.375,
+      "learning_rate": 2.002907587412371e-06,
+      "loss": 0.7559,
+      "step": 2491
+    },
+    {
+      "epoch": 0.6058837831266716,
+      "grad_norm": 20.25,
+      "learning_rate": 2.0025217668156295e-06,
+      "loss": 0.9571,
+      "step": 2492
+    },
+    {
+      "epoch": 0.6061269146608315,
+      "grad_norm": 16.5,
+      "learning_rate": 2.002135833741513e-06,
+      "loss": 0.667,
+      "step": 2493
+    },
+    {
+      "epoch": 0.6063700461949915,
+      "grad_norm": 23.125,
+      "learning_rate": 2.0017497882477068e-06,
+      "loss": 1.0873,
+      "step": 2494
+    },
+    {
+      "epoch": 0.6066131777291515,
+      "grad_norm": 21.375,
+      "learning_rate": 2.001363630391911e-06,
+      "loss": 0.8989,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6068563092633115,
+      "grad_norm": 28.5,
+      "learning_rate": 2.0009773602318444e-06,
+      "loss": 0.8248,
+      "step": 2496
+    },
+    {
+      "epoch": 0.6070994407974715,
+      "grad_norm": 24.125,
+      "learning_rate": 2.0005909778252415e-06,
+      "loss": 0.9701,
+      "step": 2497
+    },
+    {
+      "epoch": 0.6073425723316314,
+      "grad_norm": 15.5,
+      "learning_rate": 2.000204483229854e-06,
+      "loss": 0.7341,
+      "step": 2498
+    },
+    {
+      "epoch": 0.6075857038657914,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9998178765034496e-06,
+      "loss": 0.5449,
+      "step": 2499
+    },
+    {
+      "epoch": 0.6078288353999514,
+      "grad_norm": 19.625,
+      "learning_rate": 1.9994311577038146e-06,
+      "loss": 0.8267,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6080719669341114,
+      "grad_norm": 22.25,
+      "learning_rate": 1.999044326888749e-06,
+      "loss": 0.9458,
+      "step": 2501
+    },
+    {
+      "epoch": 0.6083150984682714,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.9986573841160728e-06,
+      "loss": 0.5654,
+      "step": 2502
+    },
+    {
+      "epoch": 0.6085582300024314,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.9982703294436206e-06,
+      "loss": 0.6877,
+      "step": 2503
+    },
+    {
+      "epoch": 0.6088013615365913,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9978831629292444e-06,
+      "loss": 0.7262,
+      "step": 2504
+    },
+    {
+      "epoch": 0.6090444930707513,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9974958846308136e-06,
+      "loss": 0.7039,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6092876246049113,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9971084946062126e-06,
+      "loss": 0.7864,
+      "step": 2506
+    },
+    {
+      "epoch": 0.6095307561390713,
+      "grad_norm": 20.375,
+      "learning_rate": 1.996720992913345e-06,
+      "loss": 0.7615,
+      "step": 2507
+    },
+    {
+      "epoch": 0.6097738876732313,
+      "grad_norm": 19.625,
+      "learning_rate": 1.9963333796101275e-06,
+      "loss": 0.6926,
+      "step": 2508
+    },
+    {
+      "epoch": 0.6100170192073912,
+      "grad_norm": 18.375,
+      "learning_rate": 1.995945654754497e-06,
+      "loss": 0.777,
+      "step": 2509
+    },
+    {
+      "epoch": 0.6102601507415512,
+      "grad_norm": 27.875,
+      "learning_rate": 1.9955578184044062e-06,
+      "loss": 1.2121,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6105032822757112,
+      "grad_norm": 18.375,
+      "learning_rate": 1.995169870617823e-06,
+      "loss": 0.7759,
+      "step": 2511
+    },
+    {
+      "epoch": 0.6107464138098712,
+      "grad_norm": 19.75,
+      "learning_rate": 1.994781811452733e-06,
+      "loss": 0.672,
+      "step": 2512
+    },
+    {
+      "epoch": 0.6109895453440312,
+      "grad_norm": 23.25,
+      "learning_rate": 1.994393640967138e-06,
+      "loss": 1.1134,
+      "step": 2513
+    },
+    {
+      "epoch": 0.611232676878191,
+      "grad_norm": 26.5,
+      "learning_rate": 1.994005359219058e-06,
+      "loss": 0.9399,
+      "step": 2514
+    },
+    {
+      "epoch": 0.611475808412351,
+      "grad_norm": 23.125,
+      "learning_rate": 1.993616966266527e-06,
+      "loss": 1.2102,
+      "step": 2515
+    },
+    {
+      "epoch": 0.611718939946511,
+      "grad_norm": 16.75,
+      "learning_rate": 1.993228462167598e-06,
+      "loss": 0.6831,
+      "step": 2516
+    },
+    {
+      "epoch": 0.611962071480671,
+      "grad_norm": 16.875,
+      "learning_rate": 1.992839846980339e-06,
+      "loss": 0.5729,
+      "step": 2517
+    },
+    {
+      "epoch": 0.6122052030148311,
+      "grad_norm": 16.625,
+      "learning_rate": 1.992451120762836e-06,
+      "loss": 0.8248,
+      "step": 2518
+    },
+    {
+      "epoch": 0.6124483345489911,
+      "grad_norm": 22.0,
+      "learning_rate": 1.99206228357319e-06,
+      "loss": 1.1109,
+      "step": 2519
+    },
+    {
+      "epoch": 0.612691466083151,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9916733354695204e-06,
+      "loss": 0.9915,
+      "step": 2520
+    },
+    {
+      "epoch": 0.612934597617311,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9912842765099617e-06,
+      "loss": 0.7568,
+      "step": 2521
+    },
+    {
+      "epoch": 0.613177729151471,
+      "grad_norm": 25.875,
+      "learning_rate": 1.990895106752665e-06,
+      "loss": 1.5433,
+      "step": 2522
+    },
+    {
+      "epoch": 0.613420860685631,
+      "grad_norm": 25.125,
+      "learning_rate": 1.9905058262557993e-06,
+      "loss": 0.8675,
+      "step": 2523
+    },
+    {
+      "epoch": 0.613663992219791,
+      "grad_norm": 18.75,
+      "learning_rate": 1.9901164350775482e-06,
+      "loss": 0.8637,
+      "step": 2524
+    },
+    {
+      "epoch": 0.6139071237539508,
+      "grad_norm": 22.625,
+      "learning_rate": 1.9897269332761145e-06,
+      "loss": 0.7131,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6141502552881108,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9893373209097142e-06,
+      "loss": 0.7901,
+      "step": 2526
+    },
+    {
+      "epoch": 0.6143933868222708,
+      "grad_norm": 17.5,
+      "learning_rate": 1.988947598036583e-06,
+      "loss": 0.7823,
+      "step": 2527
+    },
+    {
+      "epoch": 0.6146365183564308,
+      "grad_norm": 17.875,
+      "learning_rate": 1.988557764714971e-06,
+      "loss": 0.8806,
+      "step": 2528
+    },
+    {
+      "epoch": 0.6148796498905909,
+      "grad_norm": 20.5,
+      "learning_rate": 1.9881678210031462e-06,
+      "loss": 0.7563,
+      "step": 2529
+    },
+    {
+      "epoch": 0.6151227814247507,
+      "grad_norm": 21.0,
+      "learning_rate": 1.9877777669593917e-06,
+      "loss": 0.9966,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6153659129589107,
+      "grad_norm": 19.625,
+      "learning_rate": 1.987387602642008e-06,
+      "loss": 0.6233,
+      "step": 2531
+    },
+    {
+      "epoch": 0.6156090444930707,
+      "grad_norm": 18.25,
+      "learning_rate": 1.986997328109312e-06,
+      "loss": 1.0035,
+      "step": 2532
+    },
+    {
+      "epoch": 0.6158521760272307,
+      "grad_norm": 14.0,
+      "learning_rate": 1.9866069434196367e-06,
+      "loss": 0.8368,
+      "step": 2533
+    },
+    {
+      "epoch": 0.6160953075613907,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9862164486313323e-06,
+      "loss": 0.6484,
+      "step": 2534
+    },
+    {
+      "epoch": 0.6163384390955506,
+      "grad_norm": 17.0,
+      "learning_rate": 1.985825843802765e-06,
+      "loss": 0.6463,
+      "step": 2535
+    },
+    {
+      "epoch": 0.6165815706297106,
+      "grad_norm": 19.875,
+      "learning_rate": 1.985435128992317e-06,
+      "loss": 0.7574,
+      "step": 2536
+    },
+    {
+      "epoch": 0.6168247021638706,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9850443042583872e-06,
+      "loss": 0.7423,
+      "step": 2537
+    },
+    {
+      "epoch": 0.6170678336980306,
+      "grad_norm": 15.125,
+      "learning_rate": 1.984653369659392e-06,
+      "loss": 0.6347,
+      "step": 2538
+    },
+    {
+      "epoch": 0.6173109652321906,
+      "grad_norm": 27.125,
+      "learning_rate": 1.9842623252537624e-06,
+      "loss": 1.0904,
+      "step": 2539
+    },
+    {
+      "epoch": 0.6175540967663506,
+      "grad_norm": 21.375,
+      "learning_rate": 1.983871171099947e-06,
+      "loss": 0.8306,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6177972283005105,
+      "grad_norm": 18.75,
+      "learning_rate": 1.983479907256411e-06,
+      "loss": 1.1409,
+      "step": 2541
+    },
+    {
+      "epoch": 0.6180403598346705,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.983088533781635e-06,
+      "loss": 0.5206,
+      "step": 2542
+    },
+    {
+      "epoch": 0.6182834913688305,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9826970507341173e-06,
+      "loss": 0.9233,
+      "step": 2543
+    },
+    {
+      "epoch": 0.6185266229029905,
+      "grad_norm": 27.125,
+      "learning_rate": 1.982305458172371e-06,
+      "loss": 0.7664,
+      "step": 2544
+    },
+    {
+      "epoch": 0.6187697544371505,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9819137561549265e-06,
+      "loss": 1.3842,
+      "step": 2545
+    },
+    {
+      "epoch": 0.6190128859713104,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9815219447403305e-06,
+      "loss": 0.8527,
+      "step": 2546
+    },
+    {
+      "epoch": 0.6192560175054704,
+      "grad_norm": 19.0,
+      "learning_rate": 1.9811300239871463e-06,
+      "loss": 1.0483,
+      "step": 2547
+    },
+    {
+      "epoch": 0.6194991490396304,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9807379939539527e-06,
+      "loss": 0.7207,
+      "step": 2548
+    },
+    {
+      "epoch": 0.6197422805737904,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9803458546993456e-06,
+      "loss": 0.5849,
+      "step": 2549
+    },
+    {
+      "epoch": 0.6199854121079504,
+      "grad_norm": 19.625,
+      "learning_rate": 1.9799536062819376e-06,
+      "loss": 0.6721,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6202285436421103,
+      "grad_norm": 21.25,
+      "learning_rate": 1.9795612487603553e-06,
+      "loss": 0.7343,
+      "step": 2551
+    },
+    {
+      "epoch": 0.6204716751762703,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9791687821932456e-06,
+      "loss": 0.6469,
+      "step": 2552
+    },
+    {
+      "epoch": 0.6207148067104303,
+      "grad_norm": 13.375,
+      "learning_rate": 1.9787762066392675e-06,
+      "loss": 0.4419,
+      "step": 2553
+    },
+    {
+      "epoch": 0.6209579382445903,
+      "grad_norm": 17.25,
+      "learning_rate": 1.978383522157099e-06,
+      "loss": 0.6656,
+      "step": 2554
+    },
+    {
+      "epoch": 0.6212010697787503,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9779907288054332e-06,
+      "loss": 0.5274,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6214442013129103,
+      "grad_norm": 24.625,
+      "learning_rate": 1.977597826642981e-06,
+      "loss": 1.051,
+      "step": 2556
+    },
+    {
+      "epoch": 0.6216873328470702,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9772048157284666e-06,
+      "loss": 0.9632,
+      "step": 2557
+    },
+    {
+      "epoch": 0.6219304643812302,
+      "grad_norm": 19.0,
+      "learning_rate": 1.976811696120634e-06,
+      "loss": 1.0646,
+      "step": 2558
+    },
+    {
+      "epoch": 0.6221735959153902,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9764184678782406e-06,
+      "loss": 0.711,
+      "step": 2559
+    },
+    {
+      "epoch": 0.6224167274495502,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9760251310600614e-06,
+      "loss": 1.0041,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6226598589837102,
+      "grad_norm": 17.75,
+      "learning_rate": 1.9756316857248877e-06,
+      "loss": 0.7743,
+      "step": 2561
+    },
+    {
+      "epoch": 0.6229029905178701,
+      "grad_norm": 18.625,
+      "learning_rate": 1.9752381319315267e-06,
+      "loss": 0.9294,
+      "step": 2562
+    },
+    {
+      "epoch": 0.6231461220520301,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9748444697388008e-06,
+      "loss": 0.631,
+      "step": 2563
+    },
+    {
+      "epoch": 0.6233892535861901,
+      "grad_norm": 17.875,
+      "learning_rate": 1.974450699205551e-06,
+      "loss": 0.8134,
+      "step": 2564
+    },
+    {
+      "epoch": 0.6236323851203501,
+      "grad_norm": 16.375,
+      "learning_rate": 1.9740568203906325e-06,
+      "loss": 0.4797,
+      "step": 2565
+    },
+    {
+      "epoch": 0.6238755166545101,
+      "grad_norm": 26.0,
+      "learning_rate": 1.973662833352917e-06,
+      "loss": 1.2025,
+      "step": 2566
+    },
+    {
+      "epoch": 0.62411864818867,
+      "grad_norm": 17.25,
+      "learning_rate": 1.9732687381512933e-06,
+      "loss": 0.8941,
+      "step": 2567
+    },
+    {
+      "epoch": 0.62436177972283,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9728745348446654e-06,
+      "loss": 1.0143,
+      "step": 2568
+    },
+    {
+      "epoch": 0.62460491125699,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9724802234919535e-06,
+      "loss": 0.9828,
+      "step": 2569
+    },
+    {
+      "epoch": 0.62484804279115,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9720858041520944e-06,
+      "loss": 0.7606,
+      "step": 2570
+    },
+    {
+      "epoch": 0.62509117432531,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9716912768840417e-06,
+      "loss": 0.7916,
+      "step": 2571
+    },
+    {
+      "epoch": 0.6253343058594699,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9712966417467634e-06,
+      "loss": 0.9992,
+      "step": 2572
+    },
+    {
+      "epoch": 0.6255774373936299,
+      "grad_norm": 13.75,
+      "learning_rate": 1.970901898799244e-06,
+      "loss": 0.3656,
+      "step": 2573
+    },
+    {
+      "epoch": 0.6258205689277899,
+      "grad_norm": 19.0,
+      "learning_rate": 1.9705070481004862e-06,
+      "loss": 0.8741,
+      "step": 2574
+    },
+    {
+      "epoch": 0.6260637004619499,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9701120897095063e-06,
+      "loss": 0.6176,
+      "step": 2575
+    },
+    {
+      "epoch": 0.6263068319961099,
+      "grad_norm": 16.25,
+      "learning_rate": 1.969717023685338e-06,
+      "loss": 0.7102,
+      "step": 2576
+    },
+    {
+      "epoch": 0.6265499635302699,
+      "grad_norm": 25.625,
+      "learning_rate": 1.9693218500870303e-06,
+      "loss": 1.0332,
+      "step": 2577
+    },
+    {
+      "epoch": 0.6267930950644298,
+      "grad_norm": 17.375,
+      "learning_rate": 1.968926568973649e-06,
+      "loss": 0.8218,
+      "step": 2578
+    },
+    {
+      "epoch": 0.6270362265985898,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9685311804042756e-06,
+      "loss": 0.8964,
+      "step": 2579
+    },
+    {
+      "epoch": 0.6272793581327498,
+      "grad_norm": 17.75,
+      "learning_rate": 1.968135684438008e-06,
+      "loss": 1.0285,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6275224896669098,
+      "grad_norm": 25.375,
+      "learning_rate": 1.96774008113396e-06,
+      "loss": 1.0416,
+      "step": 2581
+    },
+    {
+      "epoch": 0.6277656212010698,
+      "grad_norm": 19.75,
+      "learning_rate": 1.9673443705512605e-06,
+      "loss": 0.8864,
+      "step": 2582
+    },
+    {
+      "epoch": 0.6280087527352297,
+      "grad_norm": 22.125,
+      "learning_rate": 1.9669485527490563e-06,
+      "loss": 1.0568,
+      "step": 2583
+    },
+    {
+      "epoch": 0.6282518842693897,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9665526277865084e-06,
+      "loss": 0.6917,
+      "step": 2584
+    },
+    {
+      "epoch": 0.6284950158035497,
+      "grad_norm": 21.5,
+      "learning_rate": 1.9661565957227954e-06,
+      "loss": 1.0235,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6287381473377097,
+      "grad_norm": 20.375,
+      "learning_rate": 1.96576045661711e-06,
+      "loss": 0.9047,
+      "step": 2586
+    },
+    {
+      "epoch": 0.6289812788718697,
+      "grad_norm": 21.375,
+      "learning_rate": 1.9653642105286636e-06,
+      "loss": 1.0859,
+      "step": 2587
+    },
+    {
+      "epoch": 0.6292244104060296,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9649678575166808e-06,
+      "loss": 0.7796,
+      "step": 2588
+    },
+    {
+      "epoch": 0.6294675419401896,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9645713976404036e-06,
+      "loss": 0.9683,
+      "step": 2589
+    },
+    {
+      "epoch": 0.6297106734743496,
+      "grad_norm": 17.0,
+      "learning_rate": 1.96417483095909e-06,
+      "loss": 0.3621,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6299538050085096,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.9637781575320138e-06,
+      "loss": 0.7793,
+      "step": 2591
+    },
+    {
+      "epoch": 0.6301969365426696,
+      "grad_norm": 20.25,
+      "learning_rate": 1.9633813774184646e-06,
+      "loss": 0.94,
+      "step": 2592
+    },
+    {
+      "epoch": 0.6304400680768296,
+      "grad_norm": 19.875,
+      "learning_rate": 1.9629844906777483e-06,
+      "loss": 0.7717,
+      "step": 2593
+    },
+    {
+      "epoch": 0.6306831996109895,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9625874973691856e-06,
+      "loss": 0.6891,
+      "step": 2594
+    },
+    {
+      "epoch": 0.6309263311451495,
+      "grad_norm": 19.375,
+      "learning_rate": 1.962190397552115e-06,
+      "loss": 0.9662,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6311694626793095,
+      "grad_norm": 21.75,
+      "learning_rate": 1.9617931912858897e-06,
+      "loss": 0.7517,
+      "step": 2596
+    },
+    {
+      "epoch": 0.6314125942134695,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.9613958786298783e-06,
+      "loss": 0.551,
+      "step": 2597
+    },
+    {
+      "epoch": 0.6316557257476295,
+      "grad_norm": 25.25,
+      "learning_rate": 1.960998459643467e-06,
+      "loss": 0.9711,
+      "step": 2598
+    },
+    {
+      "epoch": 0.6318988572817894,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9606009343860566e-06,
+      "loss": 1.0415,
+      "step": 2599
+    },
+    {
+      "epoch": 0.6321419888159494,
+      "grad_norm": 18.125,
+      "learning_rate": 1.9602033029170637e-06,
+      "loss": 1.009,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6323851203501094,
+      "grad_norm": 19.75,
+      "learning_rate": 1.959805565295922e-06,
+      "loss": 0.7581,
+      "step": 2601
+    },
+    {
+      "epoch": 0.6326282518842694,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9594077215820795e-06,
+      "loss": 0.6777,
+      "step": 2602
+    },
+    {
+      "epoch": 0.6328713834184294,
+      "grad_norm": 20.75,
+      "learning_rate": 1.959009771835001e-06,
+      "loss": 0.7276,
+      "step": 2603
+    },
+    {
+      "epoch": 0.6331145149525893,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9586117161141672e-06,
+      "loss": 1.1352,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6333576464867493,
+      "grad_norm": 45.0,
+      "learning_rate": 1.958213554479074e-06,
+      "loss": 0.9656,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6336007780209093,
+      "grad_norm": 16.75,
+      "learning_rate": 1.957815286989235e-06,
+      "loss": 0.7441,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6338439095550693,
+      "grad_norm": 15.5,
+      "learning_rate": 1.957416913704176e-06,
+      "loss": 0.5765,
+      "step": 2607
+    },
+    {
+      "epoch": 0.6340870410892293,
+      "grad_norm": 24.75,
+      "learning_rate": 1.9570184346834415e-06,
+      "loss": 1.3351,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6343301726233892,
+      "grad_norm": 24.0,
+      "learning_rate": 1.9566198499865917e-06,
+      "loss": 0.9138,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6345733041575492,
+      "grad_norm": 16.375,
+      "learning_rate": 1.9562211596732012e-06,
+      "loss": 0.5859,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6348164356917092,
+      "grad_norm": 17.5,
+      "learning_rate": 1.955822363802862e-06,
+      "loss": 0.6955,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6350595672258692,
+      "grad_norm": 22.375,
+      "learning_rate": 1.9554234624351807e-06,
+      "loss": 0.9278,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6353026987600292,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9550244556297794e-06,
+      "loss": 0.5837,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6355458302941892,
+      "grad_norm": 17.875,
+      "learning_rate": 1.954625343446297e-06,
+      "loss": 0.9824,
+      "step": 2614
+    },
+    {
+      "epoch": 0.6357889618283491,
+      "grad_norm": 23.625,
+      "learning_rate": 1.954226125944388e-06,
+      "loss": 1.0017,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6360320933625091,
+      "grad_norm": 21.125,
+      "learning_rate": 1.953826803183722e-06,
+      "loss": 1.2419,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6362752248966691,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9534273752239844e-06,
+      "loss": 0.44,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6365183564308291,
+      "grad_norm": 18.75,
+      "learning_rate": 1.953027842124878e-06,
+      "loss": 0.7213,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6367614879649891,
+      "grad_norm": 17.25,
+      "learning_rate": 1.9526282039461177e-06,
+      "loss": 0.7985,
+      "step": 2619
+    },
+    {
+      "epoch": 0.637004619499149,
+      "grad_norm": 20.0,
+      "learning_rate": 1.952228460747438e-06,
+      "loss": 0.8595,
+      "step": 2620
+    },
+    {
+      "epoch": 0.637247751033309,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9518286125885872e-06,
+      "loss": 0.7828,
+      "step": 2621
+    },
+    {
+      "epoch": 0.637490882567469,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9514286595293286e-06,
+      "loss": 0.9267,
+      "step": 2622
+    },
+    {
+      "epoch": 0.637734014101629,
+      "grad_norm": 12.125,
+      "learning_rate": 1.9510286016294432e-06,
+      "loss": 0.544,
+      "step": 2623
+    },
+    {
+      "epoch": 0.637977145635789,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9506284389487256e-06,
+      "loss": 0.8828,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6382202771699489,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9502281715469883e-06,
+      "loss": 0.9487,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6384634087041089,
+      "grad_norm": 21.75,
+      "learning_rate": 1.949827799484057e-06,
+      "loss": 0.8824,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6387065402382689,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9494273228197747e-06,
+      "loss": 1.0268,
+      "step": 2627
+    },
+    {
+      "epoch": 0.6389496717724289,
+      "grad_norm": 20.125,
+      "learning_rate": 1.949026741613999e-06,
+      "loss": 0.6952,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6391928033065889,
+      "grad_norm": 14.875,
+      "learning_rate": 1.948626055926605e-06,
+      "loss": 0.6886,
+      "step": 2629
+    },
+    {
+      "epoch": 0.6394359348407489,
+      "grad_norm": 19.875,
+      "learning_rate": 1.948225265817481e-06,
+      "loss": 0.7787,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6396790663749088,
+      "grad_norm": 24.625,
+      "learning_rate": 1.947824371346532e-06,
+      "loss": 1.0134,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6399221979090688,
+      "grad_norm": 16.0,
+      "learning_rate": 1.9474233725736787e-06,
+      "loss": 0.8336,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6401653294432288,
+      "grad_norm": 18.125,
+      "learning_rate": 1.947022269558858e-06,
+      "loss": 0.9251,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6404084609773888,
+      "grad_norm": 21.0,
+      "learning_rate": 1.9466210623620207e-06,
+      "loss": 1.2374,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6406515925115488,
+      "grad_norm": 16.75,
+      "learning_rate": 1.9462197510431346e-06,
+      "loss": 0.5718,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6408947240457087,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9458183356621826e-06,
+      "loss": 0.7142,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6411378555798687,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9454168162791635e-06,
+      "loss": 0.8723,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6413809871140287,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9450151929540908e-06,
+      "loss": 0.6637,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6416241186481887,
+      "grad_norm": 16.375,
+      "learning_rate": 1.944613465746994e-06,
+      "loss": 0.9261,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6418672501823487,
+      "grad_norm": 26.25,
+      "learning_rate": 1.944211634717918e-06,
+      "loss": 0.985,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6421103817165086,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9438096999269243e-06,
+      "loss": 0.9207,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6423535132506686,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9434076614340883e-06,
+      "loss": 0.7383,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6425966447848286,
+      "grad_norm": 15.875,
+      "learning_rate": 1.9430055192995016e-06,
+      "loss": 0.8852,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6428397763189886,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9426032735832717e-06,
+      "loss": 0.6596,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6430829078531486,
+      "grad_norm": 23.25,
+      "learning_rate": 1.94220092434552e-06,
+      "loss": 1.1477,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6433260393873085,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9417984716463868e-06,
+      "loss": 0.555,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6435691709214685,
+      "grad_norm": 19.0,
+      "learning_rate": 1.941395915546024e-06,
+      "loss": 0.7148,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6438123024556285,
+      "grad_norm": 26.125,
+      "learning_rate": 1.9409932561045995e-06,
+      "loss": 1.0506,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6440554339897885,
+      "grad_norm": 16.125,
+      "learning_rate": 1.9405904933823e-06,
+      "loss": 0.7608,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6442985655239485,
+      "grad_norm": 21.25,
+      "learning_rate": 1.940187627439325e-06,
+      "loss": 1.1644,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6445416970581085,
+      "grad_norm": 21.5,
+      "learning_rate": 1.939784658335888e-06,
+      "loss": 0.9409,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6447848285922684,
+      "grad_norm": 17.375,
+      "learning_rate": 1.939381586132221e-06,
+      "loss": 0.9638,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6450279601264284,
+      "grad_norm": 19.375,
+      "learning_rate": 1.93897841088857e-06,
+      "loss": 1.0011,
+      "step": 2653
+    },
+    {
+      "epoch": 0.6452710916605884,
+      "grad_norm": 19.375,
+      "learning_rate": 1.938575132665197e-06,
+      "loss": 0.9863,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6455142231947484,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9381717515223775e-06,
+      "loss": 0.7573,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6457573547289084,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9377682675204053e-06,
+      "loss": 0.6723,
+      "step": 2656
+    },
+    {
+      "epoch": 0.6460004862630683,
+      "grad_norm": 20.25,
+      "learning_rate": 1.9373646807195867e-06,
+      "loss": 1.2054,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6462436177972283,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9369609911802455e-06,
+      "loss": 1.0758,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6464867493313883,
+      "grad_norm": 17.375,
+      "learning_rate": 1.93655719896272e-06,
+      "loss": 0.9359,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6467298808655483,
+      "grad_norm": 17.125,
+      "learning_rate": 1.9361533041273643e-06,
+      "loss": 0.8533,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6469730123997083,
+      "grad_norm": 17.875,
+      "learning_rate": 1.935749306734547e-06,
+      "loss": 0.9048,
+      "step": 2661
+    },
+    {
+      "epoch": 0.6472161439338682,
+      "grad_norm": 16.875,
+      "learning_rate": 1.935345206844652e-06,
+      "loss": 0.7516,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6474592754680282,
+      "grad_norm": 15.125,
+      "learning_rate": 1.9349410045180796e-06,
+      "loss": 0.6485,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6477024070021882,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.9345366998152448e-06,
+      "loss": 0.7834,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6479455385363482,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9341322927965782e-06,
+      "loss": 1.2497,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6481886700705082,
+      "grad_norm": 17.25,
+      "learning_rate": 1.9337277835225248e-06,
+      "loss": 0.7493,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6484318016046682,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9333231720535456e-06,
+      "loss": 1.232,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6486749331388281,
+      "grad_norm": 21.625,
+      "learning_rate": 1.932918458450117e-06,
+      "loss": 1.0351,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6489180646729881,
+      "grad_norm": 22.5,
+      "learning_rate": 1.9325136427727302e-06,
+      "loss": 0.9951,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6491611962071481,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9321087250818927e-06,
+      "loss": 0.8068,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6494043277413081,
+      "grad_norm": 18.75,
+      "learning_rate": 1.9317037054381255e-06,
+      "loss": 0.5792,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6496474592754681,
+      "grad_norm": 13.75,
+      "learning_rate": 1.931298583901966e-06,
+      "loss": 0.6272,
+      "step": 2672
+    },
+    {
+      "epoch": 0.649890590809628,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9308933605339667e-06,
+      "loss": 0.647,
+      "step": 2673
+    },
+    {
+      "epoch": 0.650133722343788,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9304880353946952e-06,
+      "loss": 0.6139,
+      "step": 2674
+    },
+    {
+      "epoch": 0.650376853877948,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.9300826085447345e-06,
+      "loss": 0.671,
+      "step": 2675
+    },
+    {
+      "epoch": 0.650619985412108,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9296770800446825e-06,
+      "loss": 1.1881,
+      "step": 2676
+    },
+    {
+      "epoch": 0.650863116946268,
+      "grad_norm": 29.75,
+      "learning_rate": 1.9292714499551524e-06,
+      "loss": 0.8084,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6511062484804279,
+      "grad_norm": 22.0,
+      "learning_rate": 1.9288657183367725e-06,
+      "loss": 0.8615,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6513493800145879,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9284598852501867e-06,
+      "loss": 1.3256,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6515925115487479,
+      "grad_norm": 16.375,
+      "learning_rate": 1.928053950756054e-06,
+      "loss": 0.7895,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6518356430829079,
+      "grad_norm": 19.375,
+      "learning_rate": 1.9276479149150475e-06,
+      "loss": 0.5394,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6520787746170679,
+      "grad_norm": 24.25,
+      "learning_rate": 1.9272417777878573e-06,
+      "loss": 0.9726,
+      "step": 2682
+    },
+    {
+      "epoch": 0.6523219061512278,
+      "grad_norm": 25.25,
+      "learning_rate": 1.9268355394351862e-06,
+      "loss": 1.1387,
+      "step": 2683
+    },
+    {
+      "epoch": 0.6525650376853878,
+      "grad_norm": 22.875,
+      "learning_rate": 1.9264291999177547e-06,
+      "loss": 1.2903,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6528081692195478,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9260227592962976e-06,
+      "loss": 0.8315,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6530513007537078,
+      "grad_norm": 20.75,
+      "learning_rate": 1.925616217631563e-06,
+      "loss": 1.2539,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6532944322878678,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9252095749843162e-06,
+      "loss": 0.6728,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6535375638220278,
+      "grad_norm": 21.5,
+      "learning_rate": 1.9248028314153383e-06,
+      "loss": 0.796,
+      "step": 2688
+    },
+    {
+      "epoch": 0.6537806953561877,
+      "grad_norm": 15.625,
+      "learning_rate": 1.9243959869854222e-06,
+      "loss": 0.6722,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6540238268903477,
+      "grad_norm": 15.75,
+      "learning_rate": 1.9239890417553786e-06,
+      "loss": 0.8586,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6542669584245077,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9235819957860323e-06,
+      "loss": 0.9895,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6545100899586677,
+      "grad_norm": 20.875,
+      "learning_rate": 1.923174849138224e-06,
+      "loss": 0.9195,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6547532214928277,
+      "grad_norm": 18.0,
+      "learning_rate": 1.9227676018728087e-06,
+      "loss": 1.1034,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6549963530269876,
+      "grad_norm": 20.75,
+      "learning_rate": 1.922360254050655e-06,
+      "loss": 0.9611,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6552394845611476,
+      "grad_norm": 18.25,
+      "learning_rate": 1.9219528057326507e-06,
+      "loss": 0.5477,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6554826160953076,
+      "grad_norm": 17.625,
+      "learning_rate": 1.921545256979694e-06,
+      "loss": 0.7053,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6557257476294676,
+      "grad_norm": 26.75,
+      "learning_rate": 1.9211376078527003e-06,
+      "loss": 0.9475,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6559688791636276,
+      "grad_norm": 23.75,
+      "learning_rate": 1.9207298584126005e-06,
+      "loss": 0.6847,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6562120106977875,
+      "grad_norm": 16.875,
+      "learning_rate": 1.920322008720339e-06,
+      "loss": 0.5618,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6564551422319475,
+      "grad_norm": 16.625,
+      "learning_rate": 1.919914058836877e-06,
+      "loss": 0.8941,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6566982737661075,
+      "grad_norm": 22.25,
+      "learning_rate": 1.919506008823189e-06,
+      "loss": 0.6823,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6569414053002675,
+      "grad_norm": 19.375,
+      "learning_rate": 1.919097858740265e-06,
+      "loss": 0.9736,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6571845368344275,
+      "grad_norm": 16.75,
+      "learning_rate": 1.91868960864911e-06,
+      "loss": 0.8121,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6574276683685875,
+      "grad_norm": 16.5,
+      "learning_rate": 1.9182812586107454e-06,
+      "loss": 0.6385,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6576707999027473,
+      "grad_norm": 18.625,
+      "learning_rate": 1.917872808686204e-06,
+      "loss": 1.0717,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6579139314369074,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9174642589365372e-06,
+      "loss": 0.6511,
+      "step": 2706
+    },
+    {
+      "epoch": 0.6581570629710674,
+      "grad_norm": 25.0,
+      "learning_rate": 1.9170556094228092e-06,
+      "loss": 0.8808,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6584001945052274,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.9166468602061e-06,
+      "loss": 0.5074,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6586433260393874,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9162380113475045e-06,
+      "loss": 0.6399,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6588864575735472,
+      "grad_norm": 21.875,
+      "learning_rate": 1.9158290629081317e-06,
+      "loss": 1.036,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6591295891077072,
+      "grad_norm": 46.75,
+      "learning_rate": 1.915420014949106e-06,
+      "loss": 1.1031,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6593727206418672,
+      "grad_norm": 18.5,
+      "learning_rate": 1.915010867531567e-06,
+      "loss": 0.5675,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6596158521760273,
+      "grad_norm": 26.25,
+      "learning_rate": 1.9146016207166684e-06,
+      "loss": 0.987,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6598589837101873,
+      "grad_norm": 17.5,
+      "learning_rate": 1.91419227456558e-06,
+      "loss": 0.6809,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6601021152443471,
+      "grad_norm": 19.125,
+      "learning_rate": 1.913782829139485e-06,
+      "loss": 0.8617,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6603452467785071,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.9133732844995824e-06,
+      "loss": 0.5928,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6605883783126671,
+      "grad_norm": 20.0,
+      "learning_rate": 1.912963640707085e-06,
+      "loss": 1.1929,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6608315098468271,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.912553897823222e-06,
+      "loss": 0.5935,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6610746413809871,
+      "grad_norm": 23.25,
+      "learning_rate": 1.912144055909237e-06,
+      "loss": 1.0319,
+      "step": 2719
+    },
+    {
+      "epoch": 0.661317772915147,
+      "grad_norm": 25.875,
+      "learning_rate": 1.9117341150263864e-06,
+      "loss": 1.0592,
+      "step": 2720
+    },
+    {
+      "epoch": 0.661560904449307,
+      "grad_norm": 19.875,
+      "learning_rate": 1.911324075235944e-06,
+      "loss": 0.9669,
+      "step": 2721
+    },
+    {
+      "epoch": 0.661804035983467,
+      "grad_norm": 23.375,
+      "learning_rate": 1.910913936599197e-06,
+      "loss": 0.8921,
+      "step": 2722
+    },
+    {
+      "epoch": 0.662047167517627,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9105036991774476e-06,
+      "loss": 0.8377,
+      "step": 2723
+    },
+    {
+      "epoch": 0.662290299051787,
+      "grad_norm": 25.5,
+      "learning_rate": 1.9100933630320135e-06,
+      "loss": 0.8749,
+      "step": 2724
+    },
+    {
+      "epoch": 0.662533430585947,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9096829282242257e-06,
+      "loss": 0.7983,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6627765621201069,
+      "grad_norm": 18.75,
+      "learning_rate": 1.909272394815432e-06,
+      "loss": 0.935,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6630196936542669,
+      "grad_norm": 17.5,
+      "learning_rate": 1.908861762866992e-06,
+      "loss": 0.8205,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6632628251884269,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.908451032440283e-06,
+      "loss": 0.5174,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6635059567225869,
+      "grad_norm": 17.125,
+      "learning_rate": 1.908040203596695e-06,
+      "loss": 0.6204,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6637490882567469,
+      "grad_norm": 22.75,
+      "learning_rate": 1.9076292763976338e-06,
+      "loss": 0.6751,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6639922197909068,
+      "grad_norm": 23.875,
+      "learning_rate": 1.90721825090452e-06,
+      "loss": 0.939,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6642353513250668,
+      "grad_norm": 42.0,
+      "learning_rate": 1.906807127178788e-06,
+      "loss": 1.2166,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6644784828592268,
+      "grad_norm": 17.875,
+      "learning_rate": 1.906395905281887e-06,
+      "loss": 0.433,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6647216143933868,
+      "grad_norm": 30.5,
+      "learning_rate": 1.905984585275282e-06,
+      "loss": 1.1259,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6649647459275468,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9055731672204513e-06,
+      "loss": 1.1893,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6652078774617067,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.9051616511788886e-06,
+      "loss": 0.7084,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6654510089958667,
+      "grad_norm": 16.875,
+      "learning_rate": 1.9047500372121022e-06,
+      "loss": 0.8062,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6656941405300267,
+      "grad_norm": 19.625,
+      "learning_rate": 1.904338325381615e-06,
+      "loss": 0.9368,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6659372720641867,
+      "grad_norm": 21.375,
+      "learning_rate": 1.903926515748964e-06,
+      "loss": 0.8167,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6661804035983467,
+      "grad_norm": 23.75,
+      "learning_rate": 1.9035146083757012e-06,
+      "loss": 1.1495,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6664235351325067,
+      "grad_norm": 20.875,
+      "learning_rate": 1.903102603323394e-06,
+      "loss": 0.652,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 16.875,
+      "learning_rate": 1.9026905006536234e-06,
+      "loss": 0.693,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6669097982008266,
+      "grad_norm": 17.0,
+      "learning_rate": 1.9022783004279852e-06,
+      "loss": 0.8309,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6671529297349866,
+      "grad_norm": 16.75,
+      "learning_rate": 1.9018660027080893e-06,
+      "loss": 0.5793,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6673960612691466,
+      "grad_norm": 16.875,
+      "learning_rate": 1.9014536075555612e-06,
+      "loss": 0.7686,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6676391928033066,
+      "grad_norm": 14.75,
+      "learning_rate": 1.9010411150320408e-06,
+      "loss": 0.7207,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6678823243374665,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9006285251991818e-06,
+      "loss": 0.6781,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6681254558716265,
+      "grad_norm": 17.5,
+      "learning_rate": 1.9002158381186527e-06,
+      "loss": 0.5717,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6683685874057865,
+      "grad_norm": 18.875,
+      "learning_rate": 1.8998030538521373e-06,
+      "loss": 0.7242,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6686117189399465,
+      "grad_norm": 22.125,
+      "learning_rate": 1.8993901724613328e-06,
+      "loss": 0.6692,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6688548504741065,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8989771940079517e-06,
+      "loss": 0.7458,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6690979820082664,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8985641185537207e-06,
+      "loss": 1.3859,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6693411135424264,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8981509461603815e-06,
+      "loss": 0.6103,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6695842450765864,
+      "grad_norm": 17.5,
+      "learning_rate": 1.8977376768896888e-06,
+      "loss": 0.8004,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6698273766107464,
+      "grad_norm": 17.375,
+      "learning_rate": 1.897324310803414e-06,
+      "loss": 0.7593,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6700705081449064,
+      "grad_norm": 20.5,
+      "learning_rate": 1.8969108479633408e-06,
+      "loss": 0.9617,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6703136396790663,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8964972884312694e-06,
+      "loss": 0.6147,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6705567712132263,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8960836322690124e-06,
+      "loss": 1.0357,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6707999027473863,
+      "grad_norm": 24.875,
+      "learning_rate": 1.8956698795383985e-06,
+      "loss": 1.1743,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6710430342815463,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8952560303012702e-06,
+      "loss": 0.5745,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6712861658157063,
+      "grad_norm": 26.375,
+      "learning_rate": 1.8948420846194837e-06,
+      "loss": 1.2181,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6715292973498663,
+      "grad_norm": 17.875,
+      "learning_rate": 1.894428042554911e-06,
+      "loss": 1.0507,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6717724288840262,
+      "grad_norm": 17.0,
+      "learning_rate": 1.8940139041694377e-06,
+      "loss": 0.9185,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6720155604181862,
+      "grad_norm": 25.25,
+      "learning_rate": 1.8935996695249643e-06,
+      "loss": 0.8981,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6722586919523462,
+      "grad_norm": 21.0,
+      "learning_rate": 1.8931853386834047e-06,
+      "loss": 0.8092,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6725018234865062,
+      "grad_norm": 21.375,
+      "learning_rate": 1.8927709117066878e-06,
+      "loss": 0.9859,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6727449550206662,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8923563886567574e-06,
+      "loss": 1.413,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6729880865548261,
+      "grad_norm": 15.625,
+      "learning_rate": 1.8919417695955705e-06,
+      "loss": 0.6552,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6732312180889861,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.8915270545850998e-06,
+      "loss": 0.9794,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6734743496231461,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8911122436873313e-06,
+      "loss": 0.726,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6737174811573061,
+      "grad_norm": 13.5,
+      "learning_rate": 1.890697336964265e-06,
+      "loss": 0.4577,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6739606126914661,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.890282334477917e-06,
+      "loss": 0.637,
+      "step": 2772
+    },
+    {
+      "epoch": 0.674203744225626,
+      "grad_norm": 17.5,
+      "learning_rate": 1.889867236290316e-06,
+      "loss": 0.8888,
+      "step": 2773
+    },
+    {
+      "epoch": 0.674446875759786,
+      "grad_norm": 22.875,
+      "learning_rate": 1.8894520424635055e-06,
+      "loss": 0.9943,
+      "step": 2774
+    },
+    {
+      "epoch": 0.674690007293946,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8890367530595435e-06,
+      "loss": 1.2011,
+      "step": 2775
+    },
+    {
+      "epoch": 0.674933138828106,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8886213681405022e-06,
+      "loss": 0.8127,
+      "step": 2776
+    },
+    {
+      "epoch": 0.675176270362266,
+      "grad_norm": 17.0,
+      "learning_rate": 1.8882058877684684e-06,
+      "loss": 0.8076,
+      "step": 2777
+    },
+    {
+      "epoch": 0.675419401896426,
+      "grad_norm": 18.5,
+      "learning_rate": 1.887790312005542e-06,
+      "loss": 1.0948,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6756625334305859,
+      "grad_norm": 24.75,
+      "learning_rate": 1.887374640913839e-06,
+      "loss": 0.8811,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6759056649647459,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8869588745554874e-06,
+      "loss": 0.9598,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6761487964989059,
+      "grad_norm": 18.5,
+      "learning_rate": 1.8865430129926316e-06,
+      "loss": 0.8612,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6763919280330659,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.8861270562874295e-06,
+      "loss": 0.6872,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6766350595672259,
+      "grad_norm": 17.875,
+      "learning_rate": 1.8857110045020518e-06,
+      "loss": 0.836,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6768781911013858,
+      "grad_norm": 22.125,
+      "learning_rate": 1.885294857698686e-06,
+      "loss": 0.9759,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6771213226355458,
+      "grad_norm": 20.375,
+      "learning_rate": 1.8848786159395317e-06,
+      "loss": 1.1223,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6773644541697058,
+      "grad_norm": 20.375,
+      "learning_rate": 1.884462279286803e-06,
+      "loss": 0.7113,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6776075857038658,
+      "grad_norm": 18.625,
+      "learning_rate": 1.8840458478027296e-06,
+      "loss": 0.818,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6778507172380258,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8836293215495535e-06,
+      "loss": 1.1104,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6780938487721857,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8832127005895325e-06,
+      "loss": 0.6935,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6783369803063457,
+      "grad_norm": 20.0,
+      "learning_rate": 1.882795984984937e-06,
+      "loss": 0.8294,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6785801118405057,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8823791747980535e-06,
+      "loss": 0.7348,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6788232433746657,
+      "grad_norm": 17.875,
+      "learning_rate": 1.8819622700911804e-06,
+      "loss": 0.95,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6790663749088257,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8815452709266314e-06,
+      "loss": 0.8172,
+      "step": 2793
+    },
+    {
+      "epoch": 0.6793095064429856,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8811281773667347e-06,
+      "loss": 0.8765,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6795526379771456,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8807109894738317e-06,
+      "loss": 0.7575,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6797957695113056,
+      "grad_norm": 24.125,
+      "learning_rate": 1.8802937073102796e-06,
+      "loss": 0.8822,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6800389010454656,
+      "grad_norm": 14.875,
+      "learning_rate": 1.8798763309384463e-06,
+      "loss": 0.6037,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6802820325796256,
+      "grad_norm": 22.875,
+      "learning_rate": 1.8794588604207173e-06,
+      "loss": 0.8779,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6805251641137856,
+      "grad_norm": 24.125,
+      "learning_rate": 1.8790412958194903e-06,
+      "loss": 0.6963,
+      "step": 2799
+    },
+    {
+      "epoch": 0.6807682956479455,
+      "grad_norm": 17.375,
+      "learning_rate": 1.878623637197178e-06,
+      "loss": 0.6414,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6810114271821055,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.8782058846162065e-06,
+      "loss": 0.7041,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6812545587162655,
+      "grad_norm": 16.375,
+      "learning_rate": 1.8777880381390157e-06,
+      "loss": 0.4329,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6814976902504255,
+      "grad_norm": 22.75,
+      "learning_rate": 1.8773700978280607e-06,
+      "loss": 1.1243,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6817408217845855,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8769520637458094e-06,
+      "loss": 0.8327,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6819839533187454,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8765339359547441e-06,
+      "loss": 0.6192,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6822270848529054,
+      "grad_norm": 20.875,
+      "learning_rate": 1.8761157145173613e-06,
+      "loss": 1.2173,
+      "step": 2806
+    },
+    {
+      "epoch": 0.6824702163870654,
+      "grad_norm": 18.125,
+      "learning_rate": 1.875697399496172e-06,
+      "loss": 0.9912,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6827133479212254,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8752789909537005e-06,
+      "loss": 0.7923,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6829564794553854,
+      "grad_norm": 31.75,
+      "learning_rate": 1.8748604889524844e-06,
+      "loss": 1.0251,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6831996109895453,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8744418935550764e-06,
+      "loss": 0.625,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6834427425237053,
+      "grad_norm": 17.125,
+      "learning_rate": 1.874023204824043e-06,
+      "loss": 1.0231,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6836858740578653,
+      "grad_norm": 18.0,
+      "learning_rate": 1.8736044228219647e-06,
+      "loss": 0.8958,
+      "step": 2812
+    },
+    {
+      "epoch": 0.6839290055920253,
+      "grad_norm": 22.5,
+      "learning_rate": 1.8731855476114353e-06,
+      "loss": 1.2198,
+      "step": 2813
+    },
+    {
+      "epoch": 0.6841721371261853,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.8727665792550625e-06,
+      "loss": 0.4517,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6844152686603453,
+      "grad_norm": 20.875,
+      "learning_rate": 1.8723475178154693e-06,
+      "loss": 0.9555,
+      "step": 2815
+    },
+    {
+      "epoch": 0.6846584001945052,
+      "grad_norm": 17.875,
+      "learning_rate": 1.8719283633552913e-06,
+      "loss": 0.8075,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6849015317286652,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8715091159371781e-06,
+      "loss": 0.7464,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6851446632628252,
+      "grad_norm": 19.5,
+      "learning_rate": 1.8710897756237939e-06,
+      "loss": 0.9057,
+      "step": 2818
+    },
+    {
+      "epoch": 0.6853877947969852,
+      "grad_norm": 19.0,
+      "learning_rate": 1.8706703424778159e-06,
+      "loss": 0.8518,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6856309263311452,
+      "grad_norm": 18.375,
+      "learning_rate": 1.8702508165619363e-06,
+      "loss": 0.9205,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6858740578653051,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8698311979388594e-06,
+      "loss": 0.6844,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6861171893994651,
+      "grad_norm": 23.5,
+      "learning_rate": 1.8694114866713056e-06,
+      "loss": 1.3278,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6863603209336251,
+      "grad_norm": 30.875,
+      "learning_rate": 1.8689916828220075e-06,
+      "loss": 1.3844,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6866034524677851,
+      "grad_norm": 17.875,
+      "learning_rate": 1.8685717864537116e-06,
+      "loss": 0.4919,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6868465840019451,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8681517976291796e-06,
+      "loss": 0.8494,
+      "step": 2825
+    },
+    {
+      "epoch": 0.687089715536105,
+      "grad_norm": 20.5,
+      "learning_rate": 1.8677317164111856e-06,
+      "loss": 1.1265,
+      "step": 2826
+    },
+    {
+      "epoch": 0.687332847070265,
+      "grad_norm": 18.75,
+      "learning_rate": 1.867311542862518e-06,
+      "loss": 0.5187,
+      "step": 2827
+    },
+    {
+      "epoch": 0.687575978604425,
+      "grad_norm": 17.375,
+      "learning_rate": 1.8668912770459787e-06,
+      "loss": 0.8619,
+      "step": 2828
+    },
+    {
+      "epoch": 0.687819110138585,
+      "grad_norm": 18.625,
+      "learning_rate": 1.866470919024384e-06,
+      "loss": 0.9234,
+      "step": 2829
+    },
+    {
+      "epoch": 0.688062241672745,
+      "grad_norm": 20.0,
+      "learning_rate": 1.8660504688605638e-06,
+      "loss": 0.9266,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6883053732069049,
+      "grad_norm": 18.625,
+      "learning_rate": 1.8656299266173613e-06,
+      "loss": 0.9105,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6885485047410649,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8652092923576342e-06,
+      "loss": 0.8332,
+      "step": 2832
+    },
+    {
+      "epoch": 0.6887916362752249,
+      "grad_norm": 19.875,
+      "learning_rate": 1.864788566144253e-06,
+      "loss": 1.1016,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6890347678093849,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8643677480401032e-06,
+      "loss": 0.8181,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6892778993435449,
+      "grad_norm": 17.875,
+      "learning_rate": 1.8639468381080828e-06,
+      "loss": 0.6619,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6895210308777049,
+      "grad_norm": 15.75,
+      "learning_rate": 1.8635258364111042e-06,
+      "loss": 0.7536,
+      "step": 2836
+    },
+    {
+      "epoch": 0.6897641624118648,
+      "grad_norm": 19.5,
+      "learning_rate": 1.863104743012093e-06,
+      "loss": 0.9637,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6900072939460248,
+      "grad_norm": 20.75,
+      "learning_rate": 1.86268355797399e-06,
+      "loss": 1.1947,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6902504254801848,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8622622813597474e-06,
+      "loss": 0.8352,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6904935570143448,
+      "grad_norm": 22.125,
+      "learning_rate": 1.8618409132323329e-06,
+      "loss": 1.2988,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6907366885485048,
+      "grad_norm": 21.375,
+      "learning_rate": 1.861419453654727e-06,
+      "loss": 0.8299,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6909798200826647,
+      "grad_norm": 20.5,
+      "learning_rate": 1.8609979026899239e-06,
+      "loss": 0.5336,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6912229516168247,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8605762604009323e-06,
+      "loss": 0.6185,
+      "step": 2843
+    },
+    {
+      "epoch": 0.6914660831509847,
+      "grad_norm": 18.5,
+      "learning_rate": 1.8601545268507734e-06,
+      "loss": 0.6208,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6917092146851447,
+      "grad_norm": 23.625,
+      "learning_rate": 1.8597327021024825e-06,
+      "loss": 0.8914,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6919523462193047,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.8593107862191095e-06,
+      "loss": 0.6565,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6921954777534646,
+      "grad_norm": 20.0,
+      "learning_rate": 1.8588887792637158e-06,
+      "loss": 0.9686,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6924386092876246,
+      "grad_norm": 18.75,
+      "learning_rate": 1.858466681299378e-06,
+      "loss": 1.0165,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6926817408217846,
+      "grad_norm": 15.25,
+      "learning_rate": 1.8580444923891865e-06,
+      "loss": 0.6777,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6929248723559446,
+      "grad_norm": 20.75,
+      "learning_rate": 1.8576222125962442e-06,
+      "loss": 1.0999,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6931680038901046,
+      "grad_norm": 28.75,
+      "learning_rate": 1.8571998419836684e-06,
+      "loss": 1.1889,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6934111354242646,
+      "grad_norm": 15.875,
+      "learning_rate": 1.8567773806145892e-06,
+      "loss": 0.5628,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6936542669584245,
+      "grad_norm": 21.375,
+      "learning_rate": 1.8563548285521515e-06,
+      "loss": 1.0427,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6938973984925845,
+      "grad_norm": 22.875,
+      "learning_rate": 1.8559321858595121e-06,
+      "loss": 1.1127,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6941405300267445,
+      "grad_norm": 18.625,
+      "learning_rate": 1.855509452599843e-06,
+      "loss": 1.1072,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6943836615609045,
+      "grad_norm": 14.5,
+      "learning_rate": 1.8550866288363284e-06,
+      "loss": 0.471,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6946267930950645,
+      "grad_norm": 21.875,
+      "learning_rate": 1.8546637146321672e-06,
+      "loss": 0.9184,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6948699246292244,
+      "grad_norm": 122.5,
+      "learning_rate": 1.854240710050571e-06,
+      "loss": 0.9554,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6951130561633844,
+      "grad_norm": 21.125,
+      "learning_rate": 1.853817615154765e-06,
+      "loss": 0.8874,
+      "step": 2859
+    },
+    {
+      "epoch": 0.6953561876975444,
+      "grad_norm": 22.375,
+      "learning_rate": 1.8533944300079876e-06,
+      "loss": 0.8626,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6955993192317044,
+      "grad_norm": 15.25,
+      "learning_rate": 1.8529711546734925e-06,
+      "loss": 0.4943,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6958424507658644,
+      "grad_norm": 18.0,
+      "learning_rate": 1.852547789214544e-06,
+      "loss": 1.1291,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6960855823000243,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8521243336944227e-06,
+      "loss": 0.6409,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6963287138341843,
+      "grad_norm": 20.625,
+      "learning_rate": 1.85170078817642e-06,
+      "loss": 0.7132,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6965718453683443,
+      "grad_norm": 26.75,
+      "learning_rate": 1.8512771527238433e-06,
+      "loss": 0.9868,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6968149769025043,
+      "grad_norm": 12.625,
+      "learning_rate": 1.8508534274000114e-06,
+      "loss": 0.4037,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6970581084366643,
+      "grad_norm": 18.5,
+      "learning_rate": 1.8504296122682578e-06,
+      "loss": 0.9511,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6973012399708242,
+      "grad_norm": 18.75,
+      "learning_rate": 1.8500057073919286e-06,
+      "loss": 1.1812,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6975443715049842,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8495817128343844e-06,
+      "loss": 0.7531,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6977875030391442,
+      "grad_norm": 13.625,
+      "learning_rate": 1.849157628658998e-06,
+      "loss": 0.4609,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6980306345733042,
+      "grad_norm": 16.375,
+      "learning_rate": 1.8487334549291562e-06,
+      "loss": 0.7991,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6982737661074642,
+      "grad_norm": 14.125,
+      "learning_rate": 1.8483091917082586e-06,
+      "loss": 0.2674,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6985168976416242,
+      "grad_norm": 25.875,
+      "learning_rate": 1.8478848390597195e-06,
+      "loss": 1.0291,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6987600291757841,
+      "grad_norm": 19.5,
+      "learning_rate": 1.8474603970469653e-06,
+      "loss": 0.8569,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6990031607099441,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8470358657334363e-06,
+      "loss": 1.0268,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6992462922441041,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.846611245182586e-06,
+      "loss": 0.2849,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6994894237782641,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8461865354578814e-06,
+      "loss": 0.7604,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6997325553124241,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8457617366228027e-06,
+      "loss": 0.8744,
+      "step": 2878
+    },
+    {
+      "epoch": 0.699975686846584,
+      "grad_norm": 24.125,
+      "learning_rate": 1.8453368487408427e-06,
+      "loss": 1.1708,
+      "step": 2879
+    },
+    {
+      "epoch": 0.700218818380744,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8449118718755094e-06,
+      "loss": 0.6322,
+      "step": 2880
+    },
+    {
+      "epoch": 0.700461949914904,
+      "grad_norm": 20.125,
+      "learning_rate": 1.844486806090322e-06,
+      "loss": 0.7541,
+      "step": 2881
+    },
+    {
+      "epoch": 0.700705081449064,
+      "grad_norm": 19.0,
+      "learning_rate": 1.8440616514488146e-06,
+      "loss": 0.7688,
+      "step": 2882
+    },
+    {
+      "epoch": 0.700948212983224,
+      "grad_norm": 28.875,
+      "learning_rate": 1.8436364080145333e-06,
+      "loss": 1.0153,
+      "step": 2883
+    },
+    {
+      "epoch": 0.7011913445173839,
+      "grad_norm": 32.5,
+      "learning_rate": 1.8432110758510386e-06,
+      "loss": 1.027,
+      "step": 2884
+    },
+    {
+      "epoch": 0.7014344760515439,
+      "grad_norm": 24.25,
+      "learning_rate": 1.8427856550219038e-06,
+      "loss": 0.7096,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7016776075857039,
+      "grad_norm": 21.375,
+      "learning_rate": 1.8423601455907145e-06,
+      "loss": 0.7179,
+      "step": 2886
+    },
+    {
+      "epoch": 0.7019207391198639,
+      "grad_norm": 17.625,
+      "learning_rate": 1.8419345476210712e-06,
+      "loss": 0.9237,
+      "step": 2887
+    },
+    {
+      "epoch": 0.7021638706540239,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8415088611765866e-06,
+      "loss": 0.7091,
+      "step": 2888
+    },
+    {
+      "epoch": 0.7024070021881839,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.8410830863208873e-06,
+      "loss": 0.4741,
+      "step": 2889
+    },
+    {
+      "epoch": 0.7026501337223438,
+      "grad_norm": 19.5,
+      "learning_rate": 1.8406572231176124e-06,
+      "loss": 0.789,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7028932652565038,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8402312716304138e-06,
+      "loss": 0.7747,
+      "step": 2891
+    },
+    {
+      "epoch": 0.7031363967906638,
+      "grad_norm": 29.125,
+      "learning_rate": 1.8398052319229586e-06,
+      "loss": 0.9976,
+      "step": 2892
+    },
+    {
+      "epoch": 0.7033795283248238,
+      "grad_norm": 23.375,
+      "learning_rate": 1.8393791040589255e-06,
+      "loss": 1.1398,
+      "step": 2893
+    },
+    {
+      "epoch": 0.7036226598589838,
+      "grad_norm": 19.125,
+      "learning_rate": 1.8389528881020061e-06,
+      "loss": 0.8569,
+      "step": 2894
+    },
+    {
+      "epoch": 0.7038657913931436,
+      "grad_norm": 18.0,
+      "learning_rate": 1.8385265841159056e-06,
+      "loss": 0.8613,
+      "step": 2895
+    },
+    {
+      "epoch": 0.7041089229273036,
+      "grad_norm": 20.375,
+      "learning_rate": 1.8381001921643431e-06,
+      "loss": 0.7865,
+      "step": 2896
+    },
+    {
+      "epoch": 0.7043520544614637,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8376737123110503e-06,
+      "loss": 0.6729,
+      "step": 2897
+    },
+    {
+      "epoch": 0.7045951859956237,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8372471446197716e-06,
+      "loss": 0.7436,
+      "step": 2898
+    },
+    {
+      "epoch": 0.7048383175297837,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8368204891542648e-06,
+      "loss": 0.6284,
+      "step": 2899
+    },
+    {
+      "epoch": 0.7050814490639435,
+      "grad_norm": 23.625,
+      "learning_rate": 1.8363937459783016e-06,
+      "loss": 1.0442,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7053245805981035,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8359669151556652e-06,
+      "loss": 0.8138,
+      "step": 2901
+    },
+    {
+      "epoch": 0.7055677121322635,
+      "grad_norm": 19.0,
+      "learning_rate": 1.8355399967501538e-06,
+      "loss": 1.0141,
+      "step": 2902
+    },
+    {
+      "epoch": 0.7058108436664235,
+      "grad_norm": 20.75,
+      "learning_rate": 1.8351129908255767e-06,
+      "loss": 1.1416,
+      "step": 2903
+    },
+    {
+      "epoch": 0.7060539752005836,
+      "grad_norm": 18.0,
+      "learning_rate": 1.8346858974457585e-06,
+      "loss": 0.8712,
+      "step": 2904
+    },
+    {
+      "epoch": 0.7062971067347436,
+      "grad_norm": 20.0,
+      "learning_rate": 1.8342587166745346e-06,
+      "loss": 0.9932,
+      "step": 2905
+    },
+    {
+      "epoch": 0.7065402382689034,
+      "grad_norm": 22.375,
+      "learning_rate": 1.8338314485757553e-06,
+      "loss": 1.1834,
+      "step": 2906
+    },
+    {
+      "epoch": 0.7067833698030634,
+      "grad_norm": 22.5,
+      "learning_rate": 1.8334040932132825e-06,
+      "loss": 1.0299,
+      "step": 2907
+    },
+    {
+      "epoch": 0.7070265013372234,
+      "grad_norm": 34.0,
+      "learning_rate": 1.8329766506509925e-06,
+      "loss": 1.161,
+      "step": 2908
+    },
+    {
+      "epoch": 0.7072696328713834,
+      "grad_norm": 21.25,
+      "learning_rate": 1.8325491209527737e-06,
+      "loss": 1.3779,
+      "step": 2909
+    },
+    {
+      "epoch": 0.7075127644055434,
+      "grad_norm": 15.875,
+      "learning_rate": 1.8321215041825276e-06,
+      "loss": 0.6067,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7077558959397033,
+      "grad_norm": 26.0,
+      "learning_rate": 1.8316938004041695e-06,
+      "loss": 1.2639,
+      "step": 2911
+    },
+    {
+      "epoch": 0.7079990274738633,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.8312660096816265e-06,
+      "loss": 0.8692,
+      "step": 2912
+    },
+    {
+      "epoch": 0.7082421590080233,
+      "grad_norm": 24.5,
+      "learning_rate": 1.8308381320788397e-06,
+      "loss": 0.9224,
+      "step": 2913
+    },
+    {
+      "epoch": 0.7084852905421833,
+      "grad_norm": 29.5,
+      "learning_rate": 1.8304101676597624e-06,
+      "loss": 1.2138,
+      "step": 2914
+    },
+    {
+      "epoch": 0.7087284220763433,
+      "grad_norm": 28.5,
+      "learning_rate": 1.8299821164883613e-06,
+      "loss": 0.9979,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7089715536105032,
+      "grad_norm": 21.25,
+      "learning_rate": 1.829553978628617e-06,
+      "loss": 1.2982,
+      "step": 2916
+    },
+    {
+      "epoch": 0.7092146851446632,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8291257541445206e-06,
+      "loss": 0.807,
+      "step": 2917
+    },
+    {
+      "epoch": 0.7094578166788232,
+      "grad_norm": 23.75,
+      "learning_rate": 1.828697443100079e-06,
+      "loss": 1.2644,
+      "step": 2918
+    },
+    {
+      "epoch": 0.7097009482129832,
+      "grad_norm": 19.875,
+      "learning_rate": 1.8282690455593096e-06,
+      "loss": 0.8658,
+      "step": 2919
+    },
+    {
+      "epoch": 0.7099440797471432,
+      "grad_norm": 18.125,
+      "learning_rate": 1.8278405615862444e-06,
+      "loss": 0.876,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7101872112813031,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8274119912449279e-06,
+      "loss": 1.1041,
+      "step": 2921
+    },
+    {
+      "epoch": 0.7104303428154631,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8269833345994168e-06,
+      "loss": 0.5185,
+      "step": 2922
+    },
+    {
+      "epoch": 0.7106734743496231,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8265545917137817e-06,
+      "loss": 0.878,
+      "step": 2923
+    },
+    {
+      "epoch": 0.7109166058837831,
+      "grad_norm": 17.375,
+      "learning_rate": 1.826125762652105e-06,
+      "loss": 0.8941,
+      "step": 2924
+    },
+    {
+      "epoch": 0.7111597374179431,
+      "grad_norm": 19.5,
+      "learning_rate": 1.8256968474784835e-06,
+      "loss": 0.6803,
+      "step": 2925
+    },
+    {
+      "epoch": 0.7114028689521031,
+      "grad_norm": 18.875,
+      "learning_rate": 1.8252678462570253e-06,
+      "loss": 1.1147,
+      "step": 2926
+    },
+    {
+      "epoch": 0.711646000486263,
+      "grad_norm": 18.375,
+      "learning_rate": 1.8248387590518522e-06,
+      "loss": 1.1708,
+      "step": 2927
+    },
+    {
+      "epoch": 0.711889132020423,
+      "grad_norm": 19.875,
+      "learning_rate": 1.8244095859270992e-06,
+      "loss": 0.8755,
+      "step": 2928
+    },
+    {
+      "epoch": 0.712132263554583,
+      "grad_norm": 22.0,
+      "learning_rate": 1.8239803269469126e-06,
+      "loss": 0.9856,
+      "step": 2929
+    },
+    {
+      "epoch": 0.712375395088743,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8235509821754532e-06,
+      "loss": 0.7377,
+      "step": 2930
+    },
+    {
+      "epoch": 0.712618526622903,
+      "grad_norm": 20.625,
+      "learning_rate": 1.823121551676894e-06,
+      "loss": 0.9506,
+      "step": 2931
+    },
+    {
+      "epoch": 0.7128616581570629,
+      "grad_norm": 16.375,
+      "learning_rate": 1.822692035515421e-06,
+      "loss": 0.7385,
+      "step": 2932
+    },
+    {
+      "epoch": 0.7131047896912229,
+      "grad_norm": 20.25,
+      "learning_rate": 1.8222624337552325e-06,
+      "loss": 0.9211,
+      "step": 2933
+    },
+    {
+      "epoch": 0.7133479212253829,
+      "grad_norm": 23.375,
+      "learning_rate": 1.8218327464605397e-06,
+      "loss": 1.2839,
+      "step": 2934
+    },
+    {
+      "epoch": 0.7135910527595429,
+      "grad_norm": 22.375,
+      "learning_rate": 1.8214029736955675e-06,
+      "loss": 1.0768,
+      "step": 2935
+    },
+    {
+      "epoch": 0.7138341842937029,
+      "grad_norm": 22.75,
+      "learning_rate": 1.8209731155245523e-06,
+      "loss": 1.1243,
+      "step": 2936
+    },
+    {
+      "epoch": 0.7140773158278628,
+      "grad_norm": 23.375,
+      "learning_rate": 1.8205431720117436e-06,
+      "loss": 0.8342,
+      "step": 2937
+    },
+    {
+      "epoch": 0.7143204473620228,
+      "grad_norm": 15.875,
+      "learning_rate": 1.8201131432214045e-06,
+      "loss": 0.6138,
+      "step": 2938
+    },
+    {
+      "epoch": 0.7145635788961828,
+      "grad_norm": 17.625,
+      "learning_rate": 1.8196830292178097e-06,
+      "loss": 0.805,
+      "step": 2939
+    },
+    {
+      "epoch": 0.7148067104303428,
+      "grad_norm": 16.375,
+      "learning_rate": 1.8192528300652479e-06,
+      "loss": 0.7749,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7150498419645028,
+      "grad_norm": 18.625,
+      "learning_rate": 1.8188225458280187e-06,
+      "loss": 0.7135,
+      "step": 2941
+    },
+    {
+      "epoch": 0.7152929734986628,
+      "grad_norm": 16.25,
+      "learning_rate": 1.8183921765704365e-06,
+      "loss": 0.5857,
+      "step": 2942
+    },
+    {
+      "epoch": 0.7155361050328227,
+      "grad_norm": 19.125,
+      "learning_rate": 1.8179617223568269e-06,
+      "loss": 0.7907,
+      "step": 2943
+    },
+    {
+      "epoch": 0.7157792365669827,
+      "grad_norm": 15.25,
+      "learning_rate": 1.8175311832515289e-06,
+      "loss": 0.4061,
+      "step": 2944
+    },
+    {
+      "epoch": 0.7160223681011427,
+      "grad_norm": 17.0,
+      "learning_rate": 1.8171005593188939e-06,
+      "loss": 0.865,
+      "step": 2945
+    },
+    {
+      "epoch": 0.7162654996353027,
+      "grad_norm": 13.75,
+      "learning_rate": 1.816669850623286e-06,
+      "loss": 0.5594,
+      "step": 2946
+    },
+    {
+      "epoch": 0.7165086311694627,
+      "grad_norm": 29.75,
+      "learning_rate": 1.8162390572290828e-06,
+      "loss": 1.004,
+      "step": 2947
+    },
+    {
+      "epoch": 0.7167517627036226,
+      "grad_norm": 28.75,
+      "learning_rate": 1.8158081792006727e-06,
+      "loss": 1.5714,
+      "step": 2948
+    },
+    {
+      "epoch": 0.7169948942377826,
+      "grad_norm": 16.125,
+      "learning_rate": 1.8153772166024585e-06,
+      "loss": 0.6644,
+      "step": 2949
+    },
+    {
+      "epoch": 0.7172380257719426,
+      "grad_norm": 14.0,
+      "learning_rate": 1.8149461694988548e-06,
+      "loss": 0.4888,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7174811573061026,
+      "grad_norm": 18.625,
+      "learning_rate": 1.814515037954289e-06,
+      "loss": 0.5384,
+      "step": 2951
+    },
+    {
+      "epoch": 0.7177242888402626,
+      "grad_norm": 15.0,
+      "learning_rate": 1.8140838220332019e-06,
+      "loss": 0.6208,
+      "step": 2952
+    },
+    {
+      "epoch": 0.7179674203744225,
+      "grad_norm": 17.75,
+      "learning_rate": 1.8136525218000448e-06,
+      "loss": 0.6364,
+      "step": 2953
+    },
+    {
+      "epoch": 0.7182105519085825,
+      "grad_norm": 17.5,
+      "learning_rate": 1.8132211373192844e-06,
+      "loss": 0.8892,
+      "step": 2954
+    },
+    {
+      "epoch": 0.7184536834427425,
+      "grad_norm": 21.125,
+      "learning_rate": 1.8127896686553973e-06,
+      "loss": 0.8518,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7186968149769025,
+      "grad_norm": 19.25,
+      "learning_rate": 1.8123581158728744e-06,
+      "loss": 0.7045,
+      "step": 2956
+    },
+    {
+      "epoch": 0.7189399465110625,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.811926479036219e-06,
+      "loss": 0.5171,
+      "step": 2957
+    },
+    {
+      "epoch": 0.7191830780452224,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8114947582099466e-06,
+      "loss": 0.8784,
+      "step": 2958
+    },
+    {
+      "epoch": 0.7194262095793824,
+      "grad_norm": 18.625,
+      "learning_rate": 1.8110629534585854e-06,
+      "loss": 0.797,
+      "step": 2959
+    },
+    {
+      "epoch": 0.7196693411135424,
+      "grad_norm": 16.375,
+      "learning_rate": 1.8106310648466754e-06,
+      "loss": 0.7181,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7199124726477024,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8101990924387708e-06,
+      "loss": 1.164,
+      "step": 2961
+    },
+    {
+      "epoch": 0.7201556041818624,
+      "grad_norm": 17.25,
+      "learning_rate": 1.8097670362994368e-06,
+      "loss": 1.1201,
+      "step": 2962
+    },
+    {
+      "epoch": 0.7203987357160224,
+      "grad_norm": 20.875,
+      "learning_rate": 1.8093348964932516e-06,
+      "loss": 1.043,
+      "step": 2963
+    },
+    {
+      "epoch": 0.7206418672501823,
+      "grad_norm": 29.875,
+      "learning_rate": 1.808902673084806e-06,
+      "loss": 1.2297,
+      "step": 2964
+    },
+    {
+      "epoch": 0.7208849987843423,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8084703661387035e-06,
+      "loss": 0.9002,
+      "step": 2965
+    },
+    {
+      "epoch": 0.7211281303185023,
+      "grad_norm": 21.5,
+      "learning_rate": 1.8080379757195597e-06,
+      "loss": 0.912,
+      "step": 2966
+    },
+    {
+      "epoch": 0.7213712618526623,
+      "grad_norm": 15.375,
+      "learning_rate": 1.8076055018920024e-06,
+      "loss": 0.65,
+      "step": 2967
+    },
+    {
+      "epoch": 0.7216143933868223,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8071729447206731e-06,
+      "loss": 0.8992,
+      "step": 2968
+    },
+    {
+      "epoch": 0.7218575249209822,
+      "grad_norm": 19.875,
+      "learning_rate": 1.8067403042702241e-06,
+      "loss": 1.1088,
+      "step": 2969
+    },
+    {
+      "epoch": 0.7221006564551422,
+      "grad_norm": 18.875,
+      "learning_rate": 1.8063075806053219e-06,
+      "loss": 0.909,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7223437879893022,
+      "grad_norm": 16.875,
+      "learning_rate": 1.8058747737906436e-06,
+      "loss": 1.0959,
+      "step": 2971
+    },
+    {
+      "epoch": 0.7225869195234622,
+      "grad_norm": 17.875,
+      "learning_rate": 1.80544188389088e-06,
+      "loss": 1.126,
+      "step": 2972
+    },
+    {
+      "epoch": 0.7228300510576222,
+      "grad_norm": 14.875,
+      "learning_rate": 1.8050089109707345e-06,
+      "loss": 0.4312,
+      "step": 2973
+    },
+    {
+      "epoch": 0.7230731825917821,
+      "grad_norm": 19.375,
+      "learning_rate": 1.8045758550949217e-06,
+      "loss": 0.9033,
+      "step": 2974
+    },
+    {
+      "epoch": 0.7233163141259421,
+      "grad_norm": 19.125,
+      "learning_rate": 1.8041427163281693e-06,
+      "loss": 0.9799,
+      "step": 2975
+    },
+    {
+      "epoch": 0.7235594456601021,
+      "grad_norm": 20.375,
+      "learning_rate": 1.8037094947352177e-06,
+      "loss": 0.8835,
+      "step": 2976
+    },
+    {
+      "epoch": 0.7238025771942621,
+      "grad_norm": 21.625,
+      "learning_rate": 1.8032761903808194e-06,
+      "loss": 1.0635,
+      "step": 2977
+    },
+    {
+      "epoch": 0.7240457087284221,
+      "grad_norm": 17.375,
+      "learning_rate": 1.802842803329739e-06,
+      "loss": 0.768,
+      "step": 2978
+    },
+    {
+      "epoch": 0.7242888402625821,
+      "grad_norm": 20.5,
+      "learning_rate": 1.8024093336467535e-06,
+      "loss": 1.0363,
+      "step": 2979
+    },
+    {
+      "epoch": 0.724531971796742,
+      "grad_norm": 23.875,
+      "learning_rate": 1.8019757813966526e-06,
+      "loss": 0.9208,
+      "step": 2980
+    },
+    {
+      "epoch": 0.724775103330902,
+      "grad_norm": 19.75,
+      "learning_rate": 1.8015421466442385e-06,
+      "loss": 0.7719,
+      "step": 2981
+    },
+    {
+      "epoch": 0.725018234865062,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8011084294543245e-06,
+      "loss": 0.6558,
+      "step": 2982
+    },
+    {
+      "epoch": 0.725261366399222,
+      "grad_norm": 20.0,
+      "learning_rate": 1.8006746298917389e-06,
+      "loss": 0.8556,
+      "step": 2983
+    },
+    {
+      "epoch": 0.725504497933382,
+      "grad_norm": 26.125,
+      "learning_rate": 1.8002407480213183e-06,
+      "loss": 1.2889,
+      "step": 2984
+    },
+    {
+      "epoch": 0.7257476294675419,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7998067839079154e-06,
+      "loss": 0.9437,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7259907610017019,
+      "grad_norm": 23.875,
+      "learning_rate": 1.799372737616393e-06,
+      "loss": 1.1154,
+      "step": 2986
+    },
+    {
+      "epoch": 0.7262338925358619,
+      "grad_norm": 22.5,
+      "learning_rate": 1.798938609211627e-06,
+      "loss": 0.8806,
+      "step": 2987
+    },
+    {
+      "epoch": 0.7264770240700219,
+      "grad_norm": 19.875,
+      "learning_rate": 1.7985043987585054e-06,
+      "loss": 1.0027,
+      "step": 2988
+    },
+    {
+      "epoch": 0.7267201556041819,
+      "grad_norm": 22.875,
+      "learning_rate": 1.7980701063219286e-06,
+      "loss": 1.3771,
+      "step": 2989
+    },
+    {
+      "epoch": 0.7269632871383418,
+      "grad_norm": 24.0,
+      "learning_rate": 1.7976357319668086e-06,
+      "loss": 1.0942,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7272064186725018,
+      "grad_norm": 18.75,
+      "learning_rate": 1.7972012757580703e-06,
+      "loss": 0.6214,
+      "step": 2991
+    },
+    {
+      "epoch": 0.7274495502066618,
+      "grad_norm": 21.125,
+      "learning_rate": 1.7967667377606515e-06,
+      "loss": 1.0108,
+      "step": 2992
+    },
+    {
+      "epoch": 0.7276926817408218,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7963321180395004e-06,
+      "loss": 0.8376,
+      "step": 2993
+    },
+    {
+      "epoch": 0.7279358132749818,
+      "grad_norm": 34.0,
+      "learning_rate": 1.7958974166595788e-06,
+      "loss": 0.9713,
+      "step": 2994
+    },
+    {
+      "epoch": 0.7281789448091417,
+      "grad_norm": 21.5,
+      "learning_rate": 1.7954626336858602e-06,
+      "loss": 1.6373,
+      "step": 2995
+    },
+    {
+      "epoch": 0.7284220763433017,
+      "grad_norm": 12.0625,
+      "learning_rate": 1.7950277691833308e-06,
+      "loss": 0.3456,
+      "step": 2996
+    },
+    {
+      "epoch": 0.7286652078774617,
+      "grad_norm": 22.625,
+      "learning_rate": 1.7945928232169879e-06,
+      "loss": 0.9632,
+      "step": 2997
+    },
+    {
+      "epoch": 0.7289083394116217,
+      "grad_norm": 17.75,
+      "learning_rate": 1.7941577958518424e-06,
+      "loss": 0.7156,
+      "step": 2998
+    },
+    {
+      "epoch": 0.7291514709457817,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.7937226871529162e-06,
+      "loss": 0.9448,
+      "step": 2999
+    },
+    {
+      "epoch": 0.7293946024799417,
+      "grad_norm": 22.625,
+      "learning_rate": 1.7932874971852443e-06,
+      "loss": 1.0137,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7296377340141016,
+      "grad_norm": 21.125,
+      "learning_rate": 1.7928522260138729e-06,
+      "loss": 0.891,
+      "step": 3001
+    },
+    {
+      "epoch": 0.7298808655482616,
+      "grad_norm": 17.25,
+      "learning_rate": 1.7924168737038612e-06,
+      "loss": 0.8415,
+      "step": 3002
+    },
+    {
+      "epoch": 0.7301239970824216,
+      "grad_norm": 21.75,
+      "learning_rate": 1.791981440320279e-06,
+      "loss": 0.641,
+      "step": 3003
+    },
+    {
+      "epoch": 0.7303671286165816,
+      "grad_norm": 15.125,
+      "learning_rate": 1.791545925928211e-06,
+      "loss": 0.6934,
+      "step": 3004
+    },
+    {
+      "epoch": 0.7306102601507416,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7911103305927512e-06,
+      "loss": 0.8781,
+      "step": 3005
+    },
+    {
+      "epoch": 0.7308533916849015,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.7906746543790075e-06,
+      "loss": 0.711,
+      "step": 3006
+    },
+    {
+      "epoch": 0.7310965232190615,
+      "grad_norm": 17.0,
+      "learning_rate": 1.7902388973520987e-06,
+      "loss": 0.7602,
+      "step": 3007
+    },
+    {
+      "epoch": 0.7313396547532215,
+      "grad_norm": 20.5,
+      "learning_rate": 1.7898030595771566e-06,
+      "loss": 0.6901,
+      "step": 3008
+    },
+    {
+      "epoch": 0.7315827862873815,
+      "grad_norm": 19.125,
+      "learning_rate": 1.7893671411193244e-06,
+      "loss": 0.6929,
+      "step": 3009
+    },
+    {
+      "epoch": 0.7318259178215415,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7889311420437578e-06,
+      "loss": 1.0812,
+      "step": 3010
+    },
+    {
+      "epoch": 0.7320690493557014,
+      "grad_norm": 28.25,
+      "learning_rate": 1.7884950624156242e-06,
+      "loss": 1.1491,
+      "step": 3011
+    },
+    {
+      "epoch": 0.7323121808898614,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7880589023001036e-06,
+      "loss": 0.7281,
+      "step": 3012
+    },
+    {
+      "epoch": 0.7325553124240214,
+      "grad_norm": 27.25,
+      "learning_rate": 1.7876226617623874e-06,
+      "loss": 0.9335,
+      "step": 3013
+    },
+    {
+      "epoch": 0.7327984439581814,
+      "grad_norm": 36.25,
+      "learning_rate": 1.7871863408676796e-06,
+      "loss": 1.1777,
+      "step": 3014
+    },
+    {
+      "epoch": 0.7330415754923414,
+      "grad_norm": 23.0,
+      "learning_rate": 1.7867499396811949e-06,
+      "loss": 1.0634,
+      "step": 3015
+    },
+    {
+      "epoch": 0.7332847070265014,
+      "grad_norm": 20.5,
+      "learning_rate": 1.786313458268162e-06,
+      "loss": 0.9238,
+      "step": 3016
+    },
+    {
+      "epoch": 0.7335278385606613,
+      "grad_norm": 16.75,
+      "learning_rate": 1.785876896693821e-06,
+      "loss": 0.9873,
+      "step": 3017
+    },
+    {
+      "epoch": 0.7337709700948213,
+      "grad_norm": 17.125,
+      "learning_rate": 1.7854402550234218e-06,
+      "loss": 0.6296,
+      "step": 3018
+    },
+    {
+      "epoch": 0.7340141016289813,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7850035333222298e-06,
+      "loss": 0.6889,
+      "step": 3019
+    },
+    {
+      "epoch": 0.7342572331631413,
+      "grad_norm": 18.0,
+      "learning_rate": 1.7845667316555198e-06,
+      "loss": 0.4127,
+      "step": 3020
+    },
+    {
+      "epoch": 0.7345003646973013,
+      "grad_norm": 20.0,
+      "learning_rate": 1.7841298500885798e-06,
+      "loss": 0.682,
+      "step": 3021
+    },
+    {
+      "epoch": 0.7347434962314612,
+      "grad_norm": 23.0,
+      "learning_rate": 1.7836928886867082e-06,
+      "loss": 1.0004,
+      "step": 3022
+    },
+    {
+      "epoch": 0.7349866277656212,
+      "grad_norm": 20.75,
+      "learning_rate": 1.783255847515218e-06,
+      "loss": 0.78,
+      "step": 3023
+    },
+    {
+      "epoch": 0.7352297592997812,
+      "grad_norm": 19.625,
+      "learning_rate": 1.7828187266394312e-06,
+      "loss": 0.886,
+      "step": 3024
+    },
+    {
+      "epoch": 0.7354728908339412,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.7823815261246839e-06,
+      "loss": 0.6049,
+      "step": 3025
+    },
+    {
+      "epoch": 0.7357160223681012,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.7819442460363225e-06,
+      "loss": 0.9734,
+      "step": 3026
+    },
+    {
+      "epoch": 0.7359591539022611,
+      "grad_norm": 20.5,
+      "learning_rate": 1.781506886439707e-06,
+      "loss": 0.9641,
+      "step": 3027
+    },
+    {
+      "epoch": 0.7362022854364211,
+      "grad_norm": 37.75,
+      "learning_rate": 1.7810694474002076e-06,
+      "loss": 1.4406,
+      "step": 3028
+    },
+    {
+      "epoch": 0.7364454169705811,
+      "grad_norm": 22.125,
+      "learning_rate": 1.7806319289832078e-06,
+      "loss": 1.0294,
+      "step": 3029
+    },
+    {
+      "epoch": 0.7366885485047411,
+      "grad_norm": 24.0,
+      "learning_rate": 1.7801943312541014e-06,
+      "loss": 0.6694,
+      "step": 3030
+    },
+    {
+      "epoch": 0.7369316800389011,
+      "grad_norm": 18.625,
+      "learning_rate": 1.7797566542782956e-06,
+      "loss": 0.6523,
+      "step": 3031
+    },
+    {
+      "epoch": 0.737174811573061,
+      "grad_norm": 18.75,
+      "learning_rate": 1.779318898121209e-06,
+      "loss": 0.8146,
+      "step": 3032
+    },
+    {
+      "epoch": 0.737417943107221,
+      "grad_norm": 31.875,
+      "learning_rate": 1.7788810628482708e-06,
+      "loss": 0.816,
+      "step": 3033
+    },
+    {
+      "epoch": 0.737661074641381,
+      "grad_norm": 28.125,
+      "learning_rate": 1.778443148524924e-06,
+      "loss": 1.3549,
+      "step": 3034
+    },
+    {
+      "epoch": 0.737904206175541,
+      "grad_norm": 24.0,
+      "learning_rate": 1.778005155216622e-06,
+      "loss": 0.9372,
+      "step": 3035
+    },
+    {
+      "epoch": 0.738147337709701,
+      "grad_norm": 23.125,
+      "learning_rate": 1.7775670829888309e-06,
+      "loss": 0.8605,
+      "step": 3036
+    },
+    {
+      "epoch": 0.738390469243861,
+      "grad_norm": 20.0,
+      "learning_rate": 1.7771289319070276e-06,
+      "loss": 0.9511,
+      "step": 3037
+    },
+    {
+      "epoch": 0.7386336007780209,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7766907020367013e-06,
+      "loss": 0.708,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7388767323121809,
+      "grad_norm": 21.0,
+      "learning_rate": 1.7762523934433538e-06,
+      "loss": 0.8422,
+      "step": 3039
+    },
+    {
+      "epoch": 0.7391198638463409,
+      "grad_norm": 16.875,
+      "learning_rate": 1.7758140061924971e-06,
+      "loss": 0.686,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7393629953805009,
+      "grad_norm": 17.25,
+      "learning_rate": 1.7753755403496564e-06,
+      "loss": 0.73,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7396061269146609,
+      "grad_norm": 17.0,
+      "learning_rate": 1.774936995980367e-06,
+      "loss": 0.6003,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7398492584488208,
+      "grad_norm": 19.25,
+      "learning_rate": 1.7744983731501783e-06,
+      "loss": 0.9744,
+      "step": 3043
+    },
+    {
+      "epoch": 0.7400923899829808,
+      "grad_norm": 20.625,
+      "learning_rate": 1.774059671924649e-06,
+      "loss": 1.174,
+      "step": 3044
+    },
+    {
+      "epoch": 0.7403355215171408,
+      "grad_norm": 17.625,
+      "learning_rate": 1.773620892369351e-06,
+      "loss": 0.8853,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7405786530513008,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.7731820345498672e-06,
+      "loss": 0.5966,
+      "step": 3046
+    },
+    {
+      "epoch": 0.7408217845854608,
+      "grad_norm": 18.625,
+      "learning_rate": 1.7727430985317927e-06,
+      "loss": 1.2801,
+      "step": 3047
+    },
+    {
+      "epoch": 0.7410649161196207,
+      "grad_norm": 17.125,
+      "learning_rate": 1.7723040843807343e-06,
+      "loss": 0.8067,
+      "step": 3048
+    },
+    {
+      "epoch": 0.7413080476537807,
+      "grad_norm": 16.5,
+      "learning_rate": 1.7718649921623097e-06,
+      "loss": 0.635,
+      "step": 3049
+    },
+    {
+      "epoch": 0.7415511791879407,
+      "grad_norm": 16.125,
+      "learning_rate": 1.7714258219421493e-06,
+      "loss": 0.6125,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7417943107221007,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7709865737858945e-06,
+      "loss": 1.0174,
+      "step": 3051
+    },
+    {
+      "epoch": 0.7420374422562607,
+      "grad_norm": 16.625,
+      "learning_rate": 1.7705472477591982e-06,
+      "loss": 0.7565,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7422805737904207,
+      "grad_norm": 23.5,
+      "learning_rate": 1.7701078439277255e-06,
+      "loss": 0.4331,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7425237053245806,
+      "grad_norm": 22.5,
+      "learning_rate": 1.7696683623571533e-06,
+      "loss": 0.9624,
+      "step": 3054
+    },
+    {
+      "epoch": 0.7427668368587406,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.7692288031131694e-06,
+      "loss": 0.4313,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7430099683929006,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7687891662614733e-06,
+      "loss": 0.7108,
+      "step": 3056
+    },
+    {
+      "epoch": 0.7432530999270606,
+      "grad_norm": 21.875,
+      "learning_rate": 1.7683494518677766e-06,
+      "loss": 0.8518,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7434962314612206,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.7679096599978019e-06,
+      "loss": 0.62,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7437393629953805,
+      "grad_norm": 21.75,
+      "learning_rate": 1.7674697907172841e-06,
+      "loss": 1.3389,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7439824945295405,
+      "grad_norm": 22.5,
+      "learning_rate": 1.7670298440919692e-06,
+      "loss": 1.0756,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7442256260637005,
+      "grad_norm": 19.625,
+      "learning_rate": 1.766589820187614e-06,
+      "loss": 0.735,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7444687575978605,
+      "grad_norm": 24.625,
+      "learning_rate": 1.7661497190699894e-06,
+      "loss": 0.9854,
+      "step": 3062
+    },
+    {
+      "epoch": 0.7447118891320205,
+      "grad_norm": 17.75,
+      "learning_rate": 1.7657095408048744e-06,
+      "loss": 0.9337,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7449550206661804,
+      "grad_norm": 16.25,
+      "learning_rate": 1.7652692854580622e-06,
+      "loss": 0.5433,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7451981522003404,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7648289530953561e-06,
+      "loss": 0.8421,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7454412837345004,
+      "grad_norm": 31.875,
+      "learning_rate": 1.7643885437825715e-06,
+      "loss": 1.266,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7456844152686604,
+      "grad_norm": 18.625,
+      "learning_rate": 1.7639480575855356e-06,
+      "loss": 0.7353,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7459275468028204,
+      "grad_norm": 14.0,
+      "learning_rate": 1.7635074945700858e-06,
+      "loss": 0.684,
+      "step": 3068
+    },
+    {
+      "epoch": 0.7461706783369803,
+      "grad_norm": 20.75,
+      "learning_rate": 1.7630668548020726e-06,
+      "loss": 0.6465,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7464138098711403,
+      "grad_norm": 36.5,
+      "learning_rate": 1.762626138347357e-06,
+      "loss": 1.2077,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7466569414053003,
+      "grad_norm": 20.875,
+      "learning_rate": 1.7621853452718115e-06,
+      "loss": 1.0533,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7469000729394603,
+      "grad_norm": 13.875,
+      "learning_rate": 1.7617444756413205e-06,
+      "loss": 0.5045,
+      "step": 3072
+    },
+    {
+      "epoch": 0.7471432044736203,
+      "grad_norm": 16.5,
+      "learning_rate": 1.7613035295217795e-06,
+      "loss": 0.6456,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7473863360077803,
+      "grad_norm": 17.125,
+      "learning_rate": 1.7608625069790959e-06,
+      "loss": 0.8867,
+      "step": 3074
+    },
+    {
+      "epoch": 0.7476294675419402,
+      "grad_norm": 24.0,
+      "learning_rate": 1.760421408079187e-06,
+      "loss": 0.805,
+      "step": 3075
+    },
+    {
+      "epoch": 0.7478725990761002,
+      "grad_norm": 21.5,
+      "learning_rate": 1.759980232887984e-06,
+      "loss": 0.5942,
+      "step": 3076
+    },
+    {
+      "epoch": 0.7481157306102602,
+      "grad_norm": 12.375,
+      "learning_rate": 1.759538981471427e-06,
+      "loss": 0.3936,
+      "step": 3077
+    },
+    {
+      "epoch": 0.7483588621444202,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7590976538954696e-06,
+      "loss": 1.0391,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7486019936785802,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.7586562502260753e-06,
+      "loss": 0.4015,
+      "step": 3079
+    },
+    {
+      "epoch": 0.74884512521274,
+      "grad_norm": 22.0,
+      "learning_rate": 1.7582147705292192e-06,
+      "loss": 0.9547,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7490882567469,
+      "grad_norm": 18.5,
+      "learning_rate": 1.757773214870889e-06,
+      "loss": 0.93,
+      "step": 3081
+    },
+    {
+      "epoch": 0.74933138828106,
+      "grad_norm": 17.25,
+      "learning_rate": 1.7573315833170821e-06,
+      "loss": 0.8952,
+      "step": 3082
+    },
+    {
+      "epoch": 0.74957451981522,
+      "grad_norm": 19.125,
+      "learning_rate": 1.7568898759338082e-06,
+      "loss": 0.8293,
+      "step": 3083
+    },
+    {
+      "epoch": 0.74981765134938,
+      "grad_norm": 18.5,
+      "learning_rate": 1.756448092787088e-06,
+      "loss": 0.8411,
+      "step": 3084
+    },
+    {
+      "epoch": 0.75006078288354,
+      "grad_norm": 13.375,
+      "learning_rate": 1.7560062339429533e-06,
+      "loss": 0.4048,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7503039144177,
+      "grad_norm": 17.75,
+      "learning_rate": 1.7555642994674489e-06,
+      "loss": 1.0634,
+      "step": 3086
+    },
+    {
+      "epoch": 0.75054704595186,
+      "grad_norm": 18.0,
+      "learning_rate": 1.7551222894266278e-06,
+      "loss": 0.7873,
+      "step": 3087
+    },
+    {
+      "epoch": 0.75079017748602,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7546802038865568e-06,
+      "loss": 0.7158,
+      "step": 3088
+    },
+    {
+      "epoch": 0.75103330902018,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.7542380429133133e-06,
+      "loss": 0.3718,
+      "step": 3089
+    },
+    {
+      "epoch": 0.75127644055434,
+      "grad_norm": 12.375,
+      "learning_rate": 1.7537958065729857e-06,
+      "loss": 0.3316,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7515195720884998,
+      "grad_norm": 26.0,
+      "learning_rate": 1.7533534949316745e-06,
+      "loss": 1.5041,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7517627036226598,
+      "grad_norm": 23.25,
+      "learning_rate": 1.7529111080554894e-06,
+      "loss": 0.8524,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7520058351568198,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.7524686460105542e-06,
+      "loss": 0.6641,
+      "step": 3093
+    },
+    {
+      "epoch": 0.7522489666909798,
+      "grad_norm": 22.875,
+      "learning_rate": 1.7520261088630016e-06,
+      "loss": 0.5891,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7524920982251398,
+      "grad_norm": 17.875,
+      "learning_rate": 1.751583496678977e-06,
+      "loss": 0.9094,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7527352297592997,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.751140809524636e-06,
+      "loss": 0.4343,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7529783612934597,
+      "grad_norm": 22.25,
+      "learning_rate": 1.7506980474661462e-06,
+      "loss": 0.7665,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7532214928276197,
+      "grad_norm": 17.0,
+      "learning_rate": 1.750255210569686e-06,
+      "loss": 0.8135,
+      "step": 3098
+    },
+    {
+      "epoch": 0.7534646243617797,
+      "grad_norm": 20.75,
+      "learning_rate": 1.7498122989014443e-06,
+      "loss": 0.927,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7537077558959397,
+      "grad_norm": 18.5,
+      "learning_rate": 1.749369312527623e-06,
+      "loss": 0.7866,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7539508874300996,
+      "grad_norm": 18.5,
+      "learning_rate": 1.7489262515144333e-06,
+      "loss": 0.7273,
+      "step": 3101
+    },
+    {
+      "epoch": 0.7541940189642596,
+      "grad_norm": 18.5,
+      "learning_rate": 1.7484831159280986e-06,
+      "loss": 1.1789,
+      "step": 3102
+    },
+    {
+      "epoch": 0.7544371504984196,
+      "grad_norm": 16.25,
+      "learning_rate": 1.7480399058348529e-06,
+      "loss": 0.8168,
+      "step": 3103
+    },
+    {
+      "epoch": 0.7546802820325796,
+      "grad_norm": 22.625,
+      "learning_rate": 1.747596621300942e-06,
+      "loss": 0.8012,
+      "step": 3104
+    },
+    {
+      "epoch": 0.7549234135667396,
+      "grad_norm": 26.25,
+      "learning_rate": 1.7471532623926227e-06,
+      "loss": 1.0752,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7551665451008995,
+      "grad_norm": 36.25,
+      "learning_rate": 1.7467098291761616e-06,
+      "loss": 0.8673,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7554096766350595,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7462663217178382e-06,
+      "loss": 0.9313,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7556528081692195,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7458227400839422e-06,
+      "loss": 0.8523,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7558959397033795,
+      "grad_norm": 16.75,
+      "learning_rate": 1.7453790843407747e-06,
+      "loss": 0.7026,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7561390712375395,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.7449353545546477e-06,
+      "loss": 0.5233,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7563822027716995,
+      "grad_norm": 14.125,
+      "learning_rate": 1.7444915507918835e-06,
+      "loss": 0.653,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7566253343058594,
+      "grad_norm": 21.75,
+      "learning_rate": 1.7440476731188175e-06,
+      "loss": 1.1768,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7568684658400194,
+      "grad_norm": 17.75,
+      "learning_rate": 1.743603721601794e-06,
+      "loss": 0.5922,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7571115973741794,
+      "grad_norm": 16.375,
+      "learning_rate": 1.7431596963071695e-06,
+      "loss": 0.7568,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7573547289083394,
+      "grad_norm": 15.5,
+      "learning_rate": 1.742715597301311e-06,
+      "loss": 0.6965,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7575978604424994,
+      "grad_norm": 16.0,
+      "learning_rate": 1.7422714246505972e-06,
+      "loss": 0.5604,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7578409919766593,
+      "grad_norm": 24.375,
+      "learning_rate": 1.7418271784214174e-06,
+      "loss": 1.4837,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7580841235108193,
+      "grad_norm": 18.75,
+      "learning_rate": 1.7413828586801713e-06,
+      "loss": 0.9794,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7583272550449793,
+      "grad_norm": 20.125,
+      "learning_rate": 1.7409384654932707e-06,
+      "loss": 1.03,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7585703865791393,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7404939989271374e-06,
+      "loss": 0.923,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7588135181132993,
+      "grad_norm": 19.5,
+      "learning_rate": 1.7400494590482049e-06,
+      "loss": 1.0926,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7590566496474592,
+      "grad_norm": 21.875,
+      "learning_rate": 1.7396048459229175e-06,
+      "loss": 0.6412,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7592997811816192,
+      "grad_norm": 21.375,
+      "learning_rate": 1.73916015961773e-06,
+      "loss": 1.078,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7595429127157792,
+      "grad_norm": 17.125,
+      "learning_rate": 1.7387154001991086e-06,
+      "loss": 0.6388,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7597860442499392,
+      "grad_norm": 17.125,
+      "learning_rate": 1.73827056773353e-06,
+      "loss": 0.6687,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7600291757840992,
+      "grad_norm": 17.5,
+      "learning_rate": 1.7378256622874826e-06,
+      "loss": 0.8569,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7602723073182592,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7373806839274647e-06,
+      "loss": 1.1778,
+      "step": 3127
+    },
+    {
+      "epoch": 0.7605154388524191,
+      "grad_norm": 19.75,
+      "learning_rate": 1.7369356327199862e-06,
+      "loss": 1.0933,
+      "step": 3128
+    },
+    {
+      "epoch": 0.7607585703865791,
+      "grad_norm": 24.375,
+      "learning_rate": 1.736490508731568e-06,
+      "loss": 0.9143,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7610017019207391,
+      "grad_norm": 19.25,
+      "learning_rate": 1.736045312028741e-06,
+      "loss": 0.7533,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7612448334548991,
+      "grad_norm": 18.375,
+      "learning_rate": 1.735600042678048e-06,
+      "loss": 0.9688,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7614879649890591,
+      "grad_norm": 16.875,
+      "learning_rate": 1.735154700746042e-06,
+      "loss": 0.3887,
+      "step": 3132
+    },
+    {
+      "epoch": 0.761731096523219,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7347092862992871e-06,
+      "loss": 0.8986,
+      "step": 3133
+    },
+    {
+      "epoch": 0.761974228057379,
+      "grad_norm": 22.75,
+      "learning_rate": 1.7342637994043582e-06,
+      "loss": 1.1174,
+      "step": 3134
+    },
+    {
+      "epoch": 0.762217359591539,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.733818240127841e-06,
+      "loss": 0.7159,
+      "step": 3135
+    },
+    {
+      "epoch": 0.762460491125699,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.7333726085363317e-06,
+      "loss": 0.5699,
+      "step": 3136
+    },
+    {
+      "epoch": 0.762703622659859,
+      "grad_norm": 19.0,
+      "learning_rate": 1.732926904696438e-06,
+      "loss": 0.7077,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7629467541940189,
+      "grad_norm": 25.25,
+      "learning_rate": 1.7324811286747779e-06,
+      "loss": 0.9547,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7631898857281789,
+      "grad_norm": 16.375,
+      "learning_rate": 1.7320352805379807e-06,
+      "loss": 0.7508,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7634330172623389,
+      "grad_norm": 18.5,
+      "learning_rate": 1.7315893603526857e-06,
+      "loss": 1.1658,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7636761487964989,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7311433681855432e-06,
+      "loss": 1.1308,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7639192803306589,
+      "grad_norm": 17.75,
+      "learning_rate": 1.7306973041032145e-06,
+      "loss": 1.1094,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7641624118648188,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7302511681723721e-06,
+      "loss": 1.0631,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7644055433989788,
+      "grad_norm": 19.75,
+      "learning_rate": 1.729804960459699e-06,
+      "loss": 0.7407,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7646486749331388,
+      "grad_norm": 20.625,
+      "learning_rate": 1.7293586810318872e-06,
+      "loss": 1.0228,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7648918064672988,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.7289123299556419e-06,
+      "loss": 0.5856,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7651349380014588,
+      "grad_norm": 14.0,
+      "learning_rate": 1.7284659072976778e-06,
+      "loss": 0.7226,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7653780695356188,
+      "grad_norm": 17.875,
+      "learning_rate": 1.7280194131247208e-06,
+      "loss": 0.8066,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7656212010697787,
+      "grad_norm": 17.25,
+      "learning_rate": 1.7275728475035063e-06,
+      "loss": 0.6307,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7658643326039387,
+      "grad_norm": 19.125,
+      "learning_rate": 1.727126210500782e-06,
+      "loss": 0.8575,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7661074641380987,
+      "grad_norm": 23.625,
+      "learning_rate": 1.7266795021833052e-06,
+      "loss": 0.8573,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7663505956722587,
+      "grad_norm": 20.875,
+      "learning_rate": 1.7262327226178445e-06,
+      "loss": 1.252,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7665937272064187,
+      "grad_norm": 17.5,
+      "learning_rate": 1.7257858718711784e-06,
+      "loss": 0.9626,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7668368587405786,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7253389500100965e-06,
+      "loss": 0.6853,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7670799902747386,
+      "grad_norm": 16.125,
+      "learning_rate": 1.724891957101399e-06,
+      "loss": 0.7116,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7673231218088986,
+      "grad_norm": 16.75,
+      "learning_rate": 1.7244448932118976e-06,
+      "loss": 0.7574,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7675662533430586,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7239977584084122e-06,
+      "loss": 0.9469,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7678093848772186,
+      "grad_norm": 21.0,
+      "learning_rate": 1.723550552757776e-06,
+      "loss": 1.0499,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7680525164113785,
+      "grad_norm": 16.75,
+      "learning_rate": 1.7231032763268314e-06,
+      "loss": 0.4783,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7682956479455385,
+      "grad_norm": 18.75,
+      "learning_rate": 1.722655929182431e-06,
+      "loss": 1.0957,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7685387794796985,
+      "grad_norm": 17.375,
+      "learning_rate": 1.7222085113914388e-06,
+      "loss": 0.6827,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7687819110138585,
+      "grad_norm": 16.0,
+      "learning_rate": 1.7217610230207294e-06,
+      "loss": 0.682,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7690250425480185,
+      "grad_norm": 42.5,
+      "learning_rate": 1.7213134641371876e-06,
+      "loss": 0.893,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7692681740821785,
+      "grad_norm": 18.75,
+      "learning_rate": 1.7208658348077087e-06,
+      "loss": 0.5515,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7695113056163384,
+      "grad_norm": 19.25,
+      "learning_rate": 1.7204181350991987e-06,
+      "loss": 0.762,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7697544371504984,
+      "grad_norm": 17.5,
+      "learning_rate": 1.7199703650785738e-06,
+      "loss": 0.7527,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7699975686846584,
+      "grad_norm": 14.625,
+      "learning_rate": 1.7195225248127611e-06,
+      "loss": 0.764,
+      "step": 3167
+    },
+    {
+      "epoch": 0.7702407002188184,
+      "grad_norm": 14.875,
+      "learning_rate": 1.7190746143686986e-06,
+      "loss": 0.6568,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7704838317529784,
+      "grad_norm": 21.75,
+      "learning_rate": 1.7186266338133334e-06,
+      "loss": 0.9405,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7707269632871383,
+      "grad_norm": 16.875,
+      "learning_rate": 1.7181785832136245e-06,
+      "loss": 0.9862,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7709700948212983,
+      "grad_norm": 26.0,
+      "learning_rate": 1.7177304626365404e-06,
+      "loss": 1.051,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7712132263554583,
+      "grad_norm": 23.25,
+      "learning_rate": 1.717282272149061e-06,
+      "loss": 1.0199,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7714563578896183,
+      "grad_norm": 25.25,
+      "learning_rate": 1.7168340118181754e-06,
+      "loss": 1.2688,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7716994894237783,
+      "grad_norm": 18.875,
+      "learning_rate": 1.7163856817108845e-06,
+      "loss": 0.8713,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7719426209579382,
+      "grad_norm": 14.125,
+      "learning_rate": 1.715937281894199e-06,
+      "loss": 0.4008,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7721857524920982,
+      "grad_norm": 16.25,
+      "learning_rate": 1.715488812435139e-06,
+      "loss": 0.7364,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7724288840262582,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7150402734007372e-06,
+      "loss": 1.1006,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7726720155604182,
+      "grad_norm": 15.0,
+      "learning_rate": 1.7145916648580345e-06,
+      "loss": 0.5837,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7729151470945782,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.7141429868740843e-06,
+      "loss": 0.6657,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7731582786287381,
+      "grad_norm": 15.25,
+      "learning_rate": 1.7136942395159487e-06,
+      "loss": 0.4304,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7734014101628981,
+      "grad_norm": 16.125,
+      "learning_rate": 1.7132454228507002e-06,
+      "loss": 0.6216,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7736445416970581,
+      "grad_norm": 19.875,
+      "learning_rate": 1.7127965369454233e-06,
+      "loss": 1.0955,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7738876732312181,
+      "grad_norm": 18.25,
+      "learning_rate": 1.7123475818672108e-06,
+      "loss": 0.8218,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7741308047653781,
+      "grad_norm": 22.375,
+      "learning_rate": 1.7118985576831673e-06,
+      "loss": 1.0814,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7743739362995381,
+      "grad_norm": 20.75,
+      "learning_rate": 1.7114494644604072e-06,
+      "loss": 1.1863,
+      "step": 3185
+    },
+    {
+      "epoch": 0.774617067833698,
+      "grad_norm": 18.125,
+      "learning_rate": 1.7110003022660548e-06,
+      "loss": 0.825,
+      "step": 3186
+    },
+    {
+      "epoch": 0.774860199367858,
+      "grad_norm": 25.375,
+      "learning_rate": 1.7105510711672456e-06,
+      "loss": 0.7863,
+      "step": 3187
+    },
+    {
+      "epoch": 0.775103330902018,
+      "grad_norm": 23.5,
+      "learning_rate": 1.710101771231125e-06,
+      "loss": 0.903,
+      "step": 3188
+    },
+    {
+      "epoch": 0.775346462436178,
+      "grad_norm": 20.75,
+      "learning_rate": 1.7096524025248483e-06,
+      "loss": 1.0174,
+      "step": 3189
+    },
+    {
+      "epoch": 0.775589593970338,
+      "grad_norm": 27.0,
+      "learning_rate": 1.7092029651155816e-06,
+      "loss": 1.0477,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7758327255044979,
+      "grad_norm": 19.375,
+      "learning_rate": 1.7087534590705012e-06,
+      "loss": 0.965,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7760758570386579,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7083038844567931e-06,
+      "loss": 1.0624,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7763189885728179,
+      "grad_norm": 20.25,
+      "learning_rate": 1.7078542413416547e-06,
+      "loss": 1.0174,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7765621201069779,
+      "grad_norm": 12.125,
+      "learning_rate": 1.7074045297922924e-06,
+      "loss": 0.6654,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7768052516411379,
+      "grad_norm": 17.375,
+      "learning_rate": 1.7069547498759231e-06,
+      "loss": 0.8047,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7770483831752978,
+      "grad_norm": 18.125,
+      "learning_rate": 1.706504901659775e-06,
+      "loss": 0.7833,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7772915147094578,
+      "grad_norm": 16.625,
+      "learning_rate": 1.706054985211085e-06,
+      "loss": 0.676,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7775346462436178,
+      "grad_norm": 27.125,
+      "learning_rate": 1.7056050005971008e-06,
+      "loss": 1.1959,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 17.875,
+      "learning_rate": 1.7051549478850816e-06,
+      "loss": 0.9621,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7780209093119378,
+      "grad_norm": 19.25,
+      "learning_rate": 1.7047048271422937e-06,
+      "loss": 1.1471,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7782640408460978,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7042546384360171e-06,
+      "loss": 0.7744,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7785071723802577,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.7038043818335389e-06,
+      "loss": 0.691,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7787503039144177,
+      "grad_norm": 19.25,
+      "learning_rate": 1.7033540574021587e-06,
+      "loss": 0.6704,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7789934354485777,
+      "grad_norm": 19.0,
+      "learning_rate": 1.7029036652091846e-06,
+      "loss": 1.4227,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7792365669827377,
+      "grad_norm": 16.625,
+      "learning_rate": 1.7024532053219362e-06,
+      "loss": 0.6215,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7794796985168977,
+      "grad_norm": 16.25,
+      "learning_rate": 1.7020026778077423e-06,
+      "loss": 0.7057,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7797228300510576,
+      "grad_norm": 24.75,
+      "learning_rate": 1.7015520827339415e-06,
+      "loss": 0.8641,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7799659615852176,
+      "grad_norm": 16.125,
+      "learning_rate": 1.7011014201678832e-06,
+      "loss": 1.3396,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7802090931193776,
+      "grad_norm": 17.0,
+      "learning_rate": 1.7006506901769271e-06,
+      "loss": 0.9634,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7804522246535376,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.7001998928284423e-06,
+      "loss": 0.627,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7806953561876976,
+      "grad_norm": 22.5,
+      "learning_rate": 1.6997490281898082e-06,
+      "loss": 0.8008,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7809384877218575,
+      "grad_norm": 19.625,
+      "learning_rate": 1.6992980963284144e-06,
+      "loss": 0.8307,
+      "step": 3212
+    },
+    {
+      "epoch": 0.7811816192560175,
+      "grad_norm": 16.375,
+      "learning_rate": 1.6988470973116611e-06,
+      "loss": 0.7118,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7814247507901775,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6983960312069566e-06,
+      "loss": 0.92,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7816678823243375,
+      "grad_norm": 22.0,
+      "learning_rate": 1.6979448980817212e-06,
+      "loss": 1.061,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7819110138584975,
+      "grad_norm": 18.75,
+      "learning_rate": 1.6974936980033846e-06,
+      "loss": 0.8675,
+      "step": 3216
+    },
+    {
+      "epoch": 0.7821541453926574,
+      "grad_norm": 19.625,
+      "learning_rate": 1.6970424310393863e-06,
+      "loss": 0.705,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7823972769268174,
+      "grad_norm": 18.875,
+      "learning_rate": 1.6965910972571763e-06,
+      "loss": 0.833,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7826404084609774,
+      "grad_norm": 23.5,
+      "learning_rate": 1.6961396967242133e-06,
+      "loss": 1.1624,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7828835399951374,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6956882295079683e-06,
+      "loss": 1.2196,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7831266715292974,
+      "grad_norm": 19.625,
+      "learning_rate": 1.6952366956759195e-06,
+      "loss": 1.1915,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7833698030634574,
+      "grad_norm": 16.875,
+      "learning_rate": 1.6947850952955572e-06,
+      "loss": 0.7576,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7836129345976173,
+      "grad_norm": 18.75,
+      "learning_rate": 1.6943334284343807e-06,
+      "loss": 0.648,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7838560661317773,
+      "grad_norm": 23.875,
+      "learning_rate": 1.6938816951598993e-06,
+      "loss": 0.8941,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7840991976659373,
+      "grad_norm": 29.125,
+      "learning_rate": 1.6934298955396331e-06,
+      "loss": 1.1015,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7843423292000973,
+      "grad_norm": 21.125,
+      "learning_rate": 1.6929780296411099e-06,
+      "loss": 0.9174,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7845854607342573,
+      "grad_norm": 23.25,
+      "learning_rate": 1.69252609753187e-06,
+      "loss": 0.6624,
+      "step": 3227
+    },
+    {
+      "epoch": 0.7848285922684172,
+      "grad_norm": 17.625,
+      "learning_rate": 1.692074099279462e-06,
+      "loss": 0.7137,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7850717238025772,
+      "grad_norm": 21.375,
+      "learning_rate": 1.6916220349514451e-06,
+      "loss": 1.0116,
+      "step": 3229
+    },
+    {
+      "epoch": 0.7853148553367372,
+      "grad_norm": 21.5,
+      "learning_rate": 1.6911699046153884e-06,
+      "loss": 0.8142,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7855579868708972,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6907177083388693e-06,
+      "loss": 0.8175,
+      "step": 3231
+    },
+    {
+      "epoch": 0.7858011184050572,
+      "grad_norm": 21.125,
+      "learning_rate": 1.690265446189478e-06,
+      "loss": 0.7351,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7860442499392171,
+      "grad_norm": 26.875,
+      "learning_rate": 1.6898131182348118e-06,
+      "loss": 1.2597,
+      "step": 3233
+    },
+    {
+      "epoch": 0.7862873814733771,
+      "grad_norm": 22.5,
+      "learning_rate": 1.6893607245424792e-06,
+      "loss": 0.9505,
+      "step": 3234
+    },
+    {
+      "epoch": 0.7865305130075371,
+      "grad_norm": 19.375,
+      "learning_rate": 1.6889082651800983e-06,
+      "loss": 0.994,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7867736445416971,
+      "grad_norm": 41.0,
+      "learning_rate": 1.6884557402152971e-06,
+      "loss": 0.8545,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7870167760758571,
+      "grad_norm": 29.25,
+      "learning_rate": 1.6880031497157133e-06,
+      "loss": 0.4975,
+      "step": 3237
+    },
+    {
+      "epoch": 0.7872599076100171,
+      "grad_norm": 22.75,
+      "learning_rate": 1.687550493748994e-06,
+      "loss": 1.159,
+      "step": 3238
+    },
+    {
+      "epoch": 0.787503039144177,
+      "grad_norm": 28.625,
+      "learning_rate": 1.6870977723827963e-06,
+      "loss": 1.4826,
+      "step": 3239
+    },
+    {
+      "epoch": 0.787746170678337,
+      "grad_norm": 28.125,
+      "learning_rate": 1.686644985684788e-06,
+      "loss": 1.3861,
+      "step": 3240
+    },
+    {
+      "epoch": 0.787989302212497,
+      "grad_norm": 21.25,
+      "learning_rate": 1.6861921337226453e-06,
+      "loss": 1.2006,
+      "step": 3241
+    },
+    {
+      "epoch": 0.788232433746657,
+      "grad_norm": 25.5,
+      "learning_rate": 1.685739216564055e-06,
+      "loss": 1.059,
+      "step": 3242
+    },
+    {
+      "epoch": 0.788475565280817,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.6852862342767132e-06,
+      "loss": 0.6043,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7887186968149769,
+      "grad_norm": 16.75,
+      "learning_rate": 1.6848331869283263e-06,
+      "loss": 0.6907,
+      "step": 3244
+    },
+    {
+      "epoch": 0.7889618283491369,
+      "grad_norm": 17.375,
+      "learning_rate": 1.6843800745866095e-06,
+      "loss": 0.6402,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7892049598832969,
+      "grad_norm": 17.125,
+      "learning_rate": 1.6839268973192888e-06,
+      "loss": 0.814,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7894480914174569,
+      "grad_norm": 18.375,
+      "learning_rate": 1.683473655194099e-06,
+      "loss": 0.4058,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7896912229516169,
+      "grad_norm": 30.75,
+      "learning_rate": 1.683020348278785e-06,
+      "loss": 1.2858,
+      "step": 3248
+    },
+    {
+      "epoch": 0.7899343544857768,
+      "grad_norm": 19.75,
+      "learning_rate": 1.6825669766411015e-06,
+      "loss": 1.006,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7901774860199368,
+      "grad_norm": 22.25,
+      "learning_rate": 1.6821135403488126e-06,
+      "loss": 0.8786,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7904206175540968,
+      "grad_norm": 26.125,
+      "learning_rate": 1.6816600394696924e-06,
+      "loss": 1.0818,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7906637490882568,
+      "grad_norm": 27.75,
+      "learning_rate": 1.681206474071524e-06,
+      "loss": 1.0326,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7909068806224168,
+      "grad_norm": 21.75,
+      "learning_rate": 1.680752844222101e-06,
+      "loss": 1.0456,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7911500121565767,
+      "grad_norm": 23.625,
+      "learning_rate": 1.680299149989226e-06,
+      "loss": 0.6997,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7913931436907367,
+      "grad_norm": 29.25,
+      "learning_rate": 1.6798453914407115e-06,
+      "loss": 0.9938,
+      "step": 3255
+    },
+    {
+      "epoch": 0.7916362752248967,
+      "grad_norm": 22.5,
+      "learning_rate": 1.6793915686443798e-06,
+      "loss": 0.9355,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7918794067590567,
+      "grad_norm": 17.5,
+      "learning_rate": 1.6789376816680622e-06,
+      "loss": 0.5805,
+      "step": 3257
+    },
+    {
+      "epoch": 0.7921225382932167,
+      "grad_norm": 24.25,
+      "learning_rate": 1.6784837305796e-06,
+      "loss": 1.3705,
+      "step": 3258
+    },
+    {
+      "epoch": 0.7923656698273767,
+      "grad_norm": 19.5,
+      "learning_rate": 1.6780297154468442e-06,
+      "loss": 1.0128,
+      "step": 3259
+    },
+    {
+      "epoch": 0.7926088013615366,
+      "grad_norm": 13.4375,
+      "learning_rate": 1.677575636337655e-06,
+      "loss": 0.3846,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7928519328956966,
+      "grad_norm": 25.0,
+      "learning_rate": 1.6771214933199027e-06,
+      "loss": 1.0039,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7930950644298566,
+      "grad_norm": 21.0,
+      "learning_rate": 1.6766672864614658e-06,
+      "loss": 1.1159,
+      "step": 3262
+    },
+    {
+      "epoch": 0.7933381959640166,
+      "grad_norm": 20.5,
+      "learning_rate": 1.676213015830235e-06,
+      "loss": 1.0886,
+      "step": 3263
+    },
+    {
+      "epoch": 0.7935813274981766,
+      "grad_norm": 18.75,
+      "learning_rate": 1.6757586814941078e-06,
+      "loss": 1.0525,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7938244590323364,
+      "grad_norm": 19.0,
+      "learning_rate": 1.6753042835209924e-06,
+      "loss": 1.0886,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7940675905664965,
+      "grad_norm": 16.625,
+      "learning_rate": 1.6748498219788067e-06,
+      "loss": 0.6404,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7943107221006565,
+      "grad_norm": 18.875,
+      "learning_rate": 1.6743952969354777e-06,
+      "loss": 0.972,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7945538536348165,
+      "grad_norm": 16.25,
+      "learning_rate": 1.673940708458942e-06,
+      "loss": 0.5333,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7947969851689765,
+      "grad_norm": 22.25,
+      "learning_rate": 1.6734860566171454e-06,
+      "loss": 0.8507,
+      "step": 3269
+    },
+    {
+      "epoch": 0.7950401167031363,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.6730313414780442e-06,
+      "loss": 0.6042,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7952832482372963,
+      "grad_norm": 16.0,
+      "learning_rate": 1.672576563109603e-06,
+      "loss": 0.6339,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7955263797714563,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.6721217215797961e-06,
+      "loss": 0.7245,
+      "step": 3272
+    },
+    {
+      "epoch": 0.7957695113056164,
+      "grad_norm": 18.125,
+      "learning_rate": 1.6716668169566074e-06,
+      "loss": 0.7123,
+      "step": 3273
+    },
+    {
+      "epoch": 0.7960126428397764,
+      "grad_norm": 19.25,
+      "learning_rate": 1.671211849308031e-06,
+      "loss": 0.8693,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7962557743739364,
+      "grad_norm": 23.25,
+      "learning_rate": 1.670756818702069e-06,
+      "loss": 0.9725,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7964989059080962,
+      "grad_norm": 19.25,
+      "learning_rate": 1.6703017252067336e-06,
+      "loss": 1.1398,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7967420374422562,
+      "grad_norm": 23.625,
+      "learning_rate": 1.6698465688900462e-06,
+      "loss": 0.9758,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7969851689764162,
+      "grad_norm": 21.75,
+      "learning_rate": 1.6693913498200383e-06,
+      "loss": 0.9211,
+      "step": 3278
+    },
+    {
+      "epoch": 0.7972283005105762,
+      "grad_norm": 16.5,
+      "learning_rate": 1.6689360680647504e-06,
+      "loss": 0.589,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7974714320447363,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6684807236922318e-06,
+      "loss": 0.9694,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7977145635788961,
+      "grad_norm": 14.0,
+      "learning_rate": 1.6680253167705409e-06,
+      "loss": 0.4407,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7979576951130561,
+      "grad_norm": 16.0,
+      "learning_rate": 1.6675698473677473e-06,
+      "loss": 0.7154,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7982008266472161,
+      "grad_norm": 14.125,
+      "learning_rate": 1.6671143155519286e-06,
+      "loss": 0.6744,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7984439581813761,
+      "grad_norm": 14.125,
+      "learning_rate": 1.6666587213911714e-06,
+      "loss": 0.4512,
+      "step": 3284
+    },
+    {
+      "epoch": 0.7986870897155361,
+      "grad_norm": 21.25,
+      "learning_rate": 1.6662030649535725e-06,
+      "loss": 1.0703,
+      "step": 3285
+    },
+    {
+      "epoch": 0.798930221249696,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.665747346307237e-06,
+      "loss": 0.6652,
+      "step": 3286
+    },
+    {
+      "epoch": 0.799173352783856,
+      "grad_norm": 18.5,
+      "learning_rate": 1.665291565520281e-06,
+      "loss": 0.7123,
+      "step": 3287
+    },
+    {
+      "epoch": 0.799416484318016,
+      "grad_norm": 18.0,
+      "learning_rate": 1.6648357226608281e-06,
+      "loss": 0.6371,
+      "step": 3288
+    },
+    {
+      "epoch": 0.799659615852176,
+      "grad_norm": 20.125,
+      "learning_rate": 1.6643798177970113e-06,
+      "loss": 0.9583,
+      "step": 3289
+    },
+    {
+      "epoch": 0.799902747386336,
+      "grad_norm": 18.5,
+      "learning_rate": 1.663923850996975e-06,
+      "loss": 1.0215,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8001458789204959,
+      "grad_norm": 19.125,
+      "learning_rate": 1.66346782232887e-06,
+      "loss": 0.9595,
+      "step": 3291
+    },
+    {
+      "epoch": 0.8003890104546559,
+      "grad_norm": 17.875,
+      "learning_rate": 1.6630117318608585e-06,
+      "loss": 0.9341,
+      "step": 3292
+    },
+    {
+      "epoch": 0.8006321419888159,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.6625555796611102e-06,
+      "loss": 0.2916,
+      "step": 3293
+    },
+    {
+      "epoch": 0.8008752735229759,
+      "grad_norm": 19.5,
+      "learning_rate": 1.6620993657978054e-06,
+      "loss": 0.9743,
+      "step": 3294
+    },
+    {
+      "epoch": 0.8011184050571359,
+      "grad_norm": 23.375,
+      "learning_rate": 1.6616430903391337e-06,
+      "loss": 1.0148,
+      "step": 3295
+    },
+    {
+      "epoch": 0.8013615365912959,
+      "grad_norm": 22.875,
+      "learning_rate": 1.6611867533532921e-06,
+      "loss": 1.2657,
+      "step": 3296
+    },
+    {
+      "epoch": 0.8016046681254558,
+      "grad_norm": 20.0,
+      "learning_rate": 1.6607303549084897e-06,
+      "loss": 1.0109,
+      "step": 3297
+    },
+    {
+      "epoch": 0.8018477996596158,
+      "grad_norm": 29.625,
+      "learning_rate": 1.6602738950729417e-06,
+      "loss": 1.0593,
+      "step": 3298
+    },
+    {
+      "epoch": 0.8020909311937758,
+      "grad_norm": 17.625,
+      "learning_rate": 1.6598173739148744e-06,
+      "loss": 0.6808,
+      "step": 3299
+    },
+    {
+      "epoch": 0.8023340627279358,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.6593607915025227e-06,
+      "loss": 0.5933,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8025771942620958,
+      "grad_norm": 14.625,
+      "learning_rate": 1.658904147904131e-06,
+      "loss": 0.5634,
+      "step": 3301
+    },
+    {
+      "epoch": 0.8028203257962557,
+      "grad_norm": 14.75,
+      "learning_rate": 1.6584474431879527e-06,
+      "loss": 0.6136,
+      "step": 3302
+    },
+    {
+      "epoch": 0.8030634573304157,
+      "grad_norm": 18.875,
+      "learning_rate": 1.6579906774222493e-06,
+      "loss": 0.7261,
+      "step": 3303
+    },
+    {
+      "epoch": 0.8033065888645757,
+      "grad_norm": 16.875,
+      "learning_rate": 1.657533850675293e-06,
+      "loss": 0.5733,
+      "step": 3304
+    },
+    {
+      "epoch": 0.8035497203987357,
+      "grad_norm": 23.375,
+      "learning_rate": 1.6570769630153643e-06,
+      "loss": 0.7423,
+      "step": 3305
+    },
+    {
+      "epoch": 0.8037928519328957,
+      "grad_norm": 17.375,
+      "learning_rate": 1.6566200145107534e-06,
+      "loss": 0.5861,
+      "step": 3306
+    },
+    {
+      "epoch": 0.8040359834670556,
+      "grad_norm": 20.625,
+      "learning_rate": 1.6561630052297586e-06,
+      "loss": 0.9351,
+      "step": 3307
+    },
+    {
+      "epoch": 0.8042791150012156,
+      "grad_norm": 23.375,
+      "learning_rate": 1.655705935240688e-06,
+      "loss": 1.3245,
+      "step": 3308
+    },
+    {
+      "epoch": 0.8045222465353756,
+      "grad_norm": 16.75,
+      "learning_rate": 1.6552488046118588e-06,
+      "loss": 0.7572,
+      "step": 3309
+    },
+    {
+      "epoch": 0.8047653780695356,
+      "grad_norm": 15.5,
+      "learning_rate": 1.6547916134115964e-06,
+      "loss": 0.4335,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8050085096036956,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.6543343617082364e-06,
+      "loss": 0.6441,
+      "step": 3311
+    },
+    {
+      "epoch": 0.8052516411378556,
+      "grad_norm": 16.75,
+      "learning_rate": 1.653877049570123e-06,
+      "loss": 0.798,
+      "step": 3312
+    },
+    {
+      "epoch": 0.8054947726720155,
+      "grad_norm": 18.375,
+      "learning_rate": 1.6534196770656097e-06,
+      "loss": 0.6102,
+      "step": 3313
+    },
+    {
+      "epoch": 0.8057379042061755,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6529622442630583e-06,
+      "loss": 1.1213,
+      "step": 3314
+    },
+    {
+      "epoch": 0.8059810357403355,
+      "grad_norm": 20.625,
+      "learning_rate": 1.6525047512308398e-06,
+      "loss": 0.8487,
+      "step": 3315
+    },
+    {
+      "epoch": 0.8062241672744955,
+      "grad_norm": 61.5,
+      "learning_rate": 1.6520471980373348e-06,
+      "loss": 0.8042,
+      "step": 3316
+    },
+    {
+      "epoch": 0.8064672988086555,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6515895847509325e-06,
+      "loss": 0.9871,
+      "step": 3317
+    },
+    {
+      "epoch": 0.8067104303428154,
+      "grad_norm": 17.5,
+      "learning_rate": 1.6511319114400308e-06,
+      "loss": 0.6412,
+      "step": 3318
+    },
+    {
+      "epoch": 0.8069535618769754,
+      "grad_norm": 20.125,
+      "learning_rate": 1.6506741781730379e-06,
+      "loss": 0.6683,
+      "step": 3319
+    },
+    {
+      "epoch": 0.8071966934111354,
+      "grad_norm": 18.0,
+      "learning_rate": 1.6502163850183683e-06,
+      "loss": 0.94,
+      "step": 3320
+    },
+    {
+      "epoch": 0.8074398249452954,
+      "grad_norm": 38.5,
+      "learning_rate": 1.6497585320444487e-06,
+      "loss": 1.5201,
+      "step": 3321
+    },
+    {
+      "epoch": 0.8076829564794554,
+      "grad_norm": 21.625,
+      "learning_rate": 1.649300619319712e-06,
+      "loss": 0.8041,
+      "step": 3322
+    },
+    {
+      "epoch": 0.8079260880136153,
+      "grad_norm": 27.875,
+      "learning_rate": 1.6488426469126017e-06,
+      "loss": 0.9246,
+      "step": 3323
+    },
+    {
+      "epoch": 0.8081692195477753,
+      "grad_norm": 17.125,
+      "learning_rate": 1.6483846148915698e-06,
+      "loss": 0.7319,
+      "step": 3324
+    },
+    {
+      "epoch": 0.8084123510819353,
+      "grad_norm": 19.25,
+      "learning_rate": 1.6479265233250763e-06,
+      "loss": 1.2679,
+      "step": 3325
+    },
+    {
+      "epoch": 0.8086554826160953,
+      "grad_norm": 22.125,
+      "learning_rate": 1.647468372281592e-06,
+      "loss": 0.7942,
+      "step": 3326
+    },
+    {
+      "epoch": 0.8088986141502553,
+      "grad_norm": 25.125,
+      "learning_rate": 1.6470101618295946e-06,
+      "loss": 1.0763,
+      "step": 3327
+    },
+    {
+      "epoch": 0.8091417456844152,
+      "grad_norm": 24.125,
+      "learning_rate": 1.6465518920375723e-06,
+      "loss": 1.0571,
+      "step": 3328
+    },
+    {
+      "epoch": 0.8093848772185752,
+      "grad_norm": 17.625,
+      "learning_rate": 1.6460935629740207e-06,
+      "loss": 0.8949,
+      "step": 3329
+    },
+    {
+      "epoch": 0.8096280087527352,
+      "grad_norm": 20.25,
+      "learning_rate": 1.6456351747074454e-06,
+      "loss": 1.1852,
+      "step": 3330
+    },
+    {
+      "epoch": 0.8098711402868952,
+      "grad_norm": 18.625,
+      "learning_rate": 1.6451767273063605e-06,
+      "loss": 0.7703,
+      "step": 3331
+    },
+    {
+      "epoch": 0.8101142718210552,
+      "grad_norm": 19.0,
+      "learning_rate": 1.6447182208392887e-06,
+      "loss": 1.0443,
+      "step": 3332
+    },
+    {
+      "epoch": 0.8103574033552152,
+      "grad_norm": 16.5,
+      "learning_rate": 1.644259655374762e-06,
+      "loss": 0.7272,
+      "step": 3333
+    },
+    {
+      "epoch": 0.8106005348893751,
+      "grad_norm": 25.125,
+      "learning_rate": 1.6438010309813202e-06,
+      "loss": 1.2095,
+      "step": 3334
+    },
+    {
+      "epoch": 0.8108436664235351,
+      "grad_norm": 16.0,
+      "learning_rate": 1.6433423477275134e-06,
+      "loss": 0.9695,
+      "step": 3335
+    },
+    {
+      "epoch": 0.8110867979576951,
+      "grad_norm": 25.375,
+      "learning_rate": 1.6428836056818995e-06,
+      "loss": 1.247,
+      "step": 3336
+    },
+    {
+      "epoch": 0.8113299294918551,
+      "grad_norm": 18.625,
+      "learning_rate": 1.6424248049130453e-06,
+      "loss": 1.0522,
+      "step": 3337
+    },
+    {
+      "epoch": 0.8115730610260151,
+      "grad_norm": 20.0,
+      "learning_rate": 1.641965945489527e-06,
+      "loss": 0.7952,
+      "step": 3338
+    },
+    {
+      "epoch": 0.811816192560175,
+      "grad_norm": 20.625,
+      "learning_rate": 1.641507027479928e-06,
+      "loss": 0.8989,
+      "step": 3339
+    },
+    {
+      "epoch": 0.812059324094335,
+      "grad_norm": 16.5,
+      "learning_rate": 1.6410480509528427e-06,
+      "loss": 0.5721,
+      "step": 3340
+    },
+    {
+      "epoch": 0.812302455628495,
+      "grad_norm": 21.5,
+      "learning_rate": 1.6405890159768722e-06,
+      "loss": 1.0475,
+      "step": 3341
+    },
+    {
+      "epoch": 0.812545587162655,
+      "grad_norm": 19.125,
+      "learning_rate": 1.640129922620628e-06,
+      "loss": 1.0032,
+      "step": 3342
+    },
+    {
+      "epoch": 0.812788718696815,
+      "grad_norm": 12.5,
+      "learning_rate": 1.6396707709527287e-06,
+      "loss": 0.3883,
+      "step": 3343
+    },
+    {
+      "epoch": 0.8130318502309749,
+      "grad_norm": 20.75,
+      "learning_rate": 1.639211561041803e-06,
+      "loss": 0.6252,
+      "step": 3344
+    },
+    {
+      "epoch": 0.8132749817651349,
+      "grad_norm": 17.25,
+      "learning_rate": 1.6387522929564874e-06,
+      "loss": 0.5771,
+      "step": 3345
+    },
+    {
+      "epoch": 0.8135181132992949,
+      "grad_norm": 40.5,
+      "learning_rate": 1.6382929667654278e-06,
+      "loss": 1.5006,
+      "step": 3346
+    },
+    {
+      "epoch": 0.8137612448334549,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6378335825372786e-06,
+      "loss": 0.7383,
+      "step": 3347
+    },
+    {
+      "epoch": 0.8140043763676149,
+      "grad_norm": 20.125,
+      "learning_rate": 1.6373741403407018e-06,
+      "loss": 1.3033,
+      "step": 3348
+    },
+    {
+      "epoch": 0.8142475079017749,
+      "grad_norm": 16.625,
+      "learning_rate": 1.6369146402443698e-06,
+      "loss": 0.6044,
+      "step": 3349
+    },
+    {
+      "epoch": 0.8144906394359348,
+      "grad_norm": 17.25,
+      "learning_rate": 1.6364550823169625e-06,
+      "loss": 0.704,
+      "step": 3350
+    },
+    {
+      "epoch": 0.8147337709700948,
+      "grad_norm": 19.125,
+      "learning_rate": 1.6359954666271688e-06,
+      "loss": 1.004,
+      "step": 3351
+    },
+    {
+      "epoch": 0.8149769025042548,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.6355357932436863e-06,
+      "loss": 0.4525,
+      "step": 3352
+    },
+    {
+      "epoch": 0.8152200340384148,
+      "grad_norm": 22.5,
+      "learning_rate": 1.635076062235221e-06,
+      "loss": 1.1442,
+      "step": 3353
+    },
+    {
+      "epoch": 0.8154631655725748,
+      "grad_norm": 25.625,
+      "learning_rate": 1.6346162736704878e-06,
+      "loss": 1.3611,
+      "step": 3354
+    },
+    {
+      "epoch": 0.8157062971067347,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.6341564276182097e-06,
+      "loss": 0.8538,
+      "step": 3355
+    },
+    {
+      "epoch": 0.8159494286408947,
+      "grad_norm": 18.125,
+      "learning_rate": 1.6336965241471193e-06,
+      "loss": 0.6753,
+      "step": 3356
+    },
+    {
+      "epoch": 0.8161925601750547,
+      "grad_norm": 14.25,
+      "learning_rate": 1.6332365633259568e-06,
+      "loss": 0.7703,
+      "step": 3357
+    },
+    {
+      "epoch": 0.8164356917092147,
+      "grad_norm": 15.0,
+      "learning_rate": 1.6327765452234706e-06,
+      "loss": 0.4715,
+      "step": 3358
+    },
+    {
+      "epoch": 0.8166788232433747,
+      "grad_norm": 16.5,
+      "learning_rate": 1.6323164699084193e-06,
+      "loss": 0.6636,
+      "step": 3359
+    },
+    {
+      "epoch": 0.8169219547775346,
+      "grad_norm": 25.25,
+      "learning_rate": 1.6318563374495686e-06,
+      "loss": 1.0865,
+      "step": 3360
+    },
+    {
+      "epoch": 0.8171650863116946,
+      "grad_norm": 17.125,
+      "learning_rate": 1.6313961479156935e-06,
+      "loss": 0.5912,
+      "step": 3361
+    },
+    {
+      "epoch": 0.8174082178458546,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.6309359013755772e-06,
+      "loss": 0.2972,
+      "step": 3362
+    },
+    {
+      "epoch": 0.8176513493800146,
+      "grad_norm": 18.875,
+      "learning_rate": 1.630475597898011e-06,
+      "loss": 0.9035,
+      "step": 3363
+    },
+    {
+      "epoch": 0.8178944809141746,
+      "grad_norm": 19.75,
+      "learning_rate": 1.6300152375517964e-06,
+      "loss": 0.9052,
+      "step": 3364
+    },
+    {
+      "epoch": 0.8181376124483345,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.629554820405741e-06,
+      "loss": 0.411,
+      "step": 3365
+    },
+    {
+      "epoch": 0.8183807439824945,
+      "grad_norm": 24.25,
+      "learning_rate": 1.6290943465286623e-06,
+      "loss": 0.6659,
+      "step": 3366
+    },
+    {
+      "epoch": 0.8186238755166545,
+      "grad_norm": 22.5,
+      "learning_rate": 1.6286338159893867e-06,
+      "loss": 0.7502,
+      "step": 3367
+    },
+    {
+      "epoch": 0.8188670070508145,
+      "grad_norm": 22.875,
+      "learning_rate": 1.6281732288567482e-06,
+      "loss": 0.9068,
+      "step": 3368
+    },
+    {
+      "epoch": 0.8191101385849745,
+      "grad_norm": 22.125,
+      "learning_rate": 1.6277125851995892e-06,
+      "loss": 1.0115,
+      "step": 3369
+    },
+    {
+      "epoch": 0.8193532701191345,
+      "grad_norm": 23.0,
+      "learning_rate": 1.6272518850867609e-06,
+      "loss": 0.9395,
+      "step": 3370
+    },
+    {
+      "epoch": 0.8195964016532944,
+      "grad_norm": 23.25,
+      "learning_rate": 1.6267911285871233e-06,
+      "loss": 0.9708,
+      "step": 3371
+    },
+    {
+      "epoch": 0.8198395331874544,
+      "grad_norm": 22.0,
+      "learning_rate": 1.6263303157695438e-06,
+      "loss": 1.1156,
+      "step": 3372
+    },
+    {
+      "epoch": 0.8200826647216144,
+      "grad_norm": 18.625,
+      "learning_rate": 1.625869446702899e-06,
+      "loss": 0.7308,
+      "step": 3373
+    },
+    {
+      "epoch": 0.8203257962557744,
+      "grad_norm": 18.375,
+      "learning_rate": 1.6254085214560743e-06,
+      "loss": 0.7755,
+      "step": 3374
+    },
+    {
+      "epoch": 0.8205689277899344,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6249475400979625e-06,
+      "loss": 0.8015,
+      "step": 3375
+    },
+    {
+      "epoch": 0.8208120593240943,
+      "grad_norm": 18.625,
+      "learning_rate": 1.6244865026974654e-06,
+      "loss": 0.7887,
+      "step": 3376
+    },
+    {
+      "epoch": 0.8210551908582543,
+      "grad_norm": 16.125,
+      "learning_rate": 1.6240254093234925e-06,
+      "loss": 0.7635,
+      "step": 3377
+    },
+    {
+      "epoch": 0.8212983223924143,
+      "grad_norm": 14.75,
+      "learning_rate": 1.6235642600449628e-06,
+      "loss": 0.5193,
+      "step": 3378
+    },
+    {
+      "epoch": 0.8215414539265743,
+      "grad_norm": 21.125,
+      "learning_rate": 1.6231030549308024e-06,
+      "loss": 0.6491,
+      "step": 3379
+    },
+    {
+      "epoch": 0.8217845854607343,
+      "grad_norm": 21.0,
+      "learning_rate": 1.622641794049947e-06,
+      "loss": 0.8781,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8220277169948942,
+      "grad_norm": 17.625,
+      "learning_rate": 1.6221804774713397e-06,
+      "loss": 1.0416,
+      "step": 3381
+    },
+    {
+      "epoch": 0.8222708485290542,
+      "grad_norm": 17.625,
+      "learning_rate": 1.6217191052639323e-06,
+      "loss": 0.6293,
+      "step": 3382
+    },
+    {
+      "epoch": 0.8225139800632142,
+      "grad_norm": 16.875,
+      "learning_rate": 1.6212576774966848e-06,
+      "loss": 0.6833,
+      "step": 3383
+    },
+    {
+      "epoch": 0.8227571115973742,
+      "grad_norm": 22.75,
+      "learning_rate": 1.6207961942385655e-06,
+      "loss": 1.1186,
+      "step": 3384
+    },
+    {
+      "epoch": 0.8230002431315342,
+      "grad_norm": 16.25,
+      "learning_rate": 1.620334655558551e-06,
+      "loss": 0.8758,
+      "step": 3385
+    },
+    {
+      "epoch": 0.8232433746656942,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6198730615256267e-06,
+      "loss": 0.9285,
+      "step": 3386
+    },
+    {
+      "epoch": 0.8234865061998541,
+      "grad_norm": 17.75,
+      "learning_rate": 1.6194114122087852e-06,
+      "loss": 0.6273,
+      "step": 3387
+    },
+    {
+      "epoch": 0.8237296377340141,
+      "grad_norm": 20.25,
+      "learning_rate": 1.6189497076770282e-06,
+      "loss": 0.8995,
+      "step": 3388
+    },
+    {
+      "epoch": 0.8239727692681741,
+      "grad_norm": 19.0,
+      "learning_rate": 1.618487947999365e-06,
+      "loss": 0.6926,
+      "step": 3389
+    },
+    {
+      "epoch": 0.8242159008023341,
+      "grad_norm": 16.875,
+      "learning_rate": 1.6180261332448146e-06,
+      "loss": 0.6152,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8244590323364941,
+      "grad_norm": 22.25,
+      "learning_rate": 1.6175642634824025e-06,
+      "loss": 1.3543,
+      "step": 3391
+    },
+    {
+      "epoch": 0.824702163870654,
+      "grad_norm": 21.25,
+      "learning_rate": 1.6171023387811627e-06,
+      "loss": 1.1115,
+      "step": 3392
+    },
+    {
+      "epoch": 0.824945295404814,
+      "grad_norm": 18.75,
+      "learning_rate": 1.6166403592101384e-06,
+      "loss": 0.8612,
+      "step": 3393
+    },
+    {
+      "epoch": 0.825188426938974,
+      "grad_norm": 16.875,
+      "learning_rate": 1.6161783248383805e-06,
+      "loss": 0.5647,
+      "step": 3394
+    },
+    {
+      "epoch": 0.825431558473134,
+      "grad_norm": 18.5,
+      "learning_rate": 1.6157162357349482e-06,
+      "loss": 0.7841,
+      "step": 3395
+    },
+    {
+      "epoch": 0.825674690007294,
+      "grad_norm": 16.625,
+      "learning_rate": 1.6152540919689077e-06,
+      "loss": 0.5308,
+      "step": 3396
+    },
+    {
+      "epoch": 0.8259178215414539,
+      "grad_norm": 30.5,
+      "learning_rate": 1.6147918936093355e-06,
+      "loss": 1.1225,
+      "step": 3397
+    },
+    {
+      "epoch": 0.8261609530756139,
+      "grad_norm": 21.75,
+      "learning_rate": 1.6143296407253142e-06,
+      "loss": 0.9388,
+      "step": 3398
+    },
+    {
+      "epoch": 0.8264040846097739,
+      "grad_norm": 27.5,
+      "learning_rate": 1.613867333385936e-06,
+      "loss": 1.1364,
+      "step": 3399
+    },
+    {
+      "epoch": 0.8266472161439339,
+      "grad_norm": 18.5,
+      "learning_rate": 1.613404971660301e-06,
+      "loss": 0.8135,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8268903476780939,
+      "grad_norm": 14.625,
+      "learning_rate": 1.612942555617516e-06,
+      "loss": 0.5177,
+      "step": 3401
+    },
+    {
+      "epoch": 0.8271334792122538,
+      "grad_norm": 21.75,
+      "learning_rate": 1.6124800853266986e-06,
+      "loss": 1.094,
+      "step": 3402
+    },
+    {
+      "epoch": 0.8273766107464138,
+      "grad_norm": 24.625,
+      "learning_rate": 1.6120175608569718e-06,
+      "loss": 0.786,
+      "step": 3403
+    },
+    {
+      "epoch": 0.8276197422805738,
+      "grad_norm": 20.125,
+      "learning_rate": 1.6115549822774684e-06,
+      "loss": 0.6718,
+      "step": 3404
+    },
+    {
+      "epoch": 0.8278628738147338,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6110923496573283e-06,
+      "loss": 0.68,
+      "step": 3405
+    },
+    {
+      "epoch": 0.8281060053488938,
+      "grad_norm": 21.0,
+      "learning_rate": 1.6106296630657005e-06,
+      "loss": 1.1493,
+      "step": 3406
+    },
+    {
+      "epoch": 0.8283491368830538,
+      "grad_norm": 20.75,
+      "learning_rate": 1.6101669225717417e-06,
+      "loss": 0.554,
+      "step": 3407
+    },
+    {
+      "epoch": 0.8285922684172137,
+      "grad_norm": 23.25,
+      "learning_rate": 1.6097041282446152e-06,
+      "loss": 1.284,
+      "step": 3408
+    },
+    {
+      "epoch": 0.8288353999513737,
+      "grad_norm": 19.875,
+      "learning_rate": 1.6092412801534949e-06,
+      "loss": 0.8646,
+      "step": 3409
+    },
+    {
+      "epoch": 0.8290785314855337,
+      "grad_norm": 16.75,
+      "learning_rate": 1.6087783783675611e-06,
+      "loss": 0.7641,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8293216630196937,
+      "grad_norm": 24.5,
+      "learning_rate": 1.6083154229560022e-06,
+      "loss": 1.0882,
+      "step": 3411
+    },
+    {
+      "epoch": 0.8295647945538537,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.607852413988015e-06,
+      "loss": 0.4787,
+      "step": 3412
+    },
+    {
+      "epoch": 0.8298079260880136,
+      "grad_norm": 17.625,
+      "learning_rate": 1.607389351532804e-06,
+      "loss": 0.78,
+      "step": 3413
+    },
+    {
+      "epoch": 0.8300510576221736,
+      "grad_norm": 24.5,
+      "learning_rate": 1.6069262356595827e-06,
+      "loss": 0.9327,
+      "step": 3414
+    },
+    {
+      "epoch": 0.8302941891563336,
+      "grad_norm": 23.625,
+      "learning_rate": 1.6064630664375705e-06,
+      "loss": 1.0338,
+      "step": 3415
+    },
+    {
+      "epoch": 0.8305373206904936,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6059998439359967e-06,
+      "loss": 0.9083,
+      "step": 3416
+    },
+    {
+      "epoch": 0.8307804522246536,
+      "grad_norm": 20.375,
+      "learning_rate": 1.6055365682240985e-06,
+      "loss": 1.0087,
+      "step": 3417
+    },
+    {
+      "epoch": 0.8310235837588135,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6050732393711193e-06,
+      "loss": 0.9165,
+      "step": 3418
+    },
+    {
+      "epoch": 0.8312667152929735,
+      "grad_norm": 28.0,
+      "learning_rate": 1.6046098574463126e-06,
+      "loss": 1.3364,
+      "step": 3419
+    },
+    {
+      "epoch": 0.8315098468271335,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.6041464225189376e-06,
+      "loss": 0.5851,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8317529783612935,
+      "grad_norm": 16.25,
+      "learning_rate": 1.603682934658264e-06,
+      "loss": 0.3832,
+      "step": 3421
+    },
+    {
+      "epoch": 0.8319961098954535,
+      "grad_norm": 21.5,
+      "learning_rate": 1.6032193939335676e-06,
+      "loss": 0.7287,
+      "step": 3422
+    },
+    {
+      "epoch": 0.8322392414296135,
+      "grad_norm": 21.875,
+      "learning_rate": 1.6027558004141323e-06,
+      "loss": 0.9834,
+      "step": 3423
+    },
+    {
+      "epoch": 0.8324823729637734,
+      "grad_norm": 19.375,
+      "learning_rate": 1.6022921541692501e-06,
+      "loss": 1.0242,
+      "step": 3424
+    },
+    {
+      "epoch": 0.8327255044979334,
+      "grad_norm": 19.375,
+      "learning_rate": 1.6018284552682215e-06,
+      "loss": 0.9596,
+      "step": 3425
+    },
+    {
+      "epoch": 0.8329686360320934,
+      "grad_norm": 21.125,
+      "learning_rate": 1.6013647037803539e-06,
+      "loss": 0.8035,
+      "step": 3426
+    },
+    {
+      "epoch": 0.8332117675662534,
+      "grad_norm": 20.75,
+      "learning_rate": 1.6009008997749631e-06,
+      "loss": 1.0939,
+      "step": 3427
+    },
+    {
+      "epoch": 0.8334548991004134,
+      "grad_norm": 16.625,
+      "learning_rate": 1.600437043321372e-06,
+      "loss": 0.4163,
+      "step": 3428
+    },
+    {
+      "epoch": 0.8336980306345733,
+      "grad_norm": 23.625,
+      "learning_rate": 1.5999731344889132e-06,
+      "loss": 1.03,
+      "step": 3429
+    },
+    {
+      "epoch": 0.8339411621687333,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.599509173346925e-06,
+      "loss": 0.7969,
+      "step": 3430
+    },
+    {
+      "epoch": 0.8341842937028933,
+      "grad_norm": 17.875,
+      "learning_rate": 1.599045159964755e-06,
+      "loss": 1.0562,
+      "step": 3431
+    },
+    {
+      "epoch": 0.8344274252370533,
+      "grad_norm": 32.5,
+      "learning_rate": 1.598581094411757e-06,
+      "loss": 0.8758,
+      "step": 3432
+    },
+    {
+      "epoch": 0.8346705567712133,
+      "grad_norm": 24.0,
+      "learning_rate": 1.598116976757294e-06,
+      "loss": 1.3562,
+      "step": 3433
+    },
+    {
+      "epoch": 0.8349136883053732,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5976528070707376e-06,
+      "loss": 0.7141,
+      "step": 3434
+    },
+    {
+      "epoch": 0.8351568198395332,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5971885854214642e-06,
+      "loss": 0.9906,
+      "step": 3435
+    },
+    {
+      "epoch": 0.8353999513736932,
+      "grad_norm": 18.75,
+      "learning_rate": 1.596724311878861e-06,
+      "loss": 0.7147,
+      "step": 3436
+    },
+    {
+      "epoch": 0.8356430829078532,
+      "grad_norm": 20.25,
+      "learning_rate": 1.596259986512321e-06,
+      "loss": 0.7191,
+      "step": 3437
+    },
+    {
+      "epoch": 0.8358862144420132,
+      "grad_norm": 22.875,
+      "learning_rate": 1.5957956093912459e-06,
+      "loss": 1.1478,
+      "step": 3438
+    },
+    {
+      "epoch": 0.836129345976173,
+      "grad_norm": 31.5,
+      "learning_rate": 1.5953311805850448e-06,
+      "loss": 0.8481,
+      "step": 3439
+    },
+    {
+      "epoch": 0.8363724775103331,
+      "grad_norm": 16.375,
+      "learning_rate": 1.5948667001631352e-06,
+      "loss": 0.921,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8366156090444931,
+      "grad_norm": 22.375,
+      "learning_rate": 1.594402168194941e-06,
+      "loss": 0.8647,
+      "step": 3441
+    },
+    {
+      "epoch": 0.8368587405786531,
+      "grad_norm": 21.5,
+      "learning_rate": 1.5939375847498944e-06,
+      "loss": 1.1712,
+      "step": 3442
+    },
+    {
+      "epoch": 0.8371018721128131,
+      "grad_norm": 21.625,
+      "learning_rate": 1.5934729498974362e-06,
+      "loss": 0.8327,
+      "step": 3443
+    },
+    {
+      "epoch": 0.8373450036469731,
+      "grad_norm": 23.375,
+      "learning_rate": 1.5930082637070132e-06,
+      "loss": 1.1919,
+      "step": 3444
+    },
+    {
+      "epoch": 0.837588135181133,
+      "grad_norm": 20.875,
+      "learning_rate": 1.5925435262480815e-06,
+      "loss": 0.7159,
+      "step": 3445
+    },
+    {
+      "epoch": 0.837831266715293,
+      "grad_norm": 17.0,
+      "learning_rate": 1.592078737590104e-06,
+      "loss": 0.8424,
+      "step": 3446
+    },
+    {
+      "epoch": 0.838074398249453,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5916138978025509e-06,
+      "loss": 0.6072,
+      "step": 3447
+    },
+    {
+      "epoch": 0.838317529783613,
+      "grad_norm": 19.625,
+      "learning_rate": 1.591149006954901e-06,
+      "loss": 0.6674,
+      "step": 3448
+    },
+    {
+      "epoch": 0.838560661317773,
+      "grad_norm": 31.375,
+      "learning_rate": 1.5906840651166402e-06,
+      "loss": 1.2384,
+      "step": 3449
+    },
+    {
+      "epoch": 0.8388037928519329,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.5902190723572622e-06,
+      "loss": 0.7587,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8390469243860929,
+      "grad_norm": 22.75,
+      "learning_rate": 1.589754028746268e-06,
+      "loss": 1.1475,
+      "step": 3451
+    },
+    {
+      "epoch": 0.8392900559202529,
+      "grad_norm": 16.75,
+      "learning_rate": 1.5892889343531662e-06,
+      "loss": 0.665,
+      "step": 3452
+    },
+    {
+      "epoch": 0.8395331874544129,
+      "grad_norm": 29.5,
+      "learning_rate": 1.588823789247474e-06,
+      "loss": 1.2101,
+      "step": 3453
+    },
+    {
+      "epoch": 0.8397763189885729,
+      "grad_norm": 50.75,
+      "learning_rate": 1.588358593498714e-06,
+      "loss": 0.9499,
+      "step": 3454
+    },
+    {
+      "epoch": 0.8400194505227327,
+      "grad_norm": 15.375,
+      "learning_rate": 1.5878933471764192e-06,
+      "loss": 0.7118,
+      "step": 3455
+    },
+    {
+      "epoch": 0.8402625820568927,
+      "grad_norm": 17.5,
+      "learning_rate": 1.5874280503501278e-06,
+      "loss": 0.7053,
+      "step": 3456
+    },
+    {
+      "epoch": 0.8405057135910527,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5869627030893867e-06,
+      "loss": 1.09,
+      "step": 3457
+    },
+    {
+      "epoch": 0.8407488451252128,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5864973054637504e-06,
+      "loss": 1.0151,
+      "step": 3458
+    },
+    {
+      "epoch": 0.8409919766593728,
+      "grad_norm": 20.625,
+      "learning_rate": 1.5860318575427793e-06,
+      "loss": 0.9082,
+      "step": 3459
+    },
+    {
+      "epoch": 0.8412351081935328,
+      "grad_norm": 15.25,
+      "learning_rate": 1.5855663593960446e-06,
+      "loss": 0.7381,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8414782397276926,
+      "grad_norm": 25.5,
+      "learning_rate": 1.585100811093122e-06,
+      "loss": 1.1652,
+      "step": 3461
+    },
+    {
+      "epoch": 0.8417213712618526,
+      "grad_norm": 21.75,
+      "learning_rate": 1.5846352127035952e-06,
+      "loss": 1.0499,
+      "step": 3462
+    },
+    {
+      "epoch": 0.8419645027960126,
+      "grad_norm": 22.625,
+      "learning_rate": 1.584169564297057e-06,
+      "loss": 0.8066,
+      "step": 3463
+    },
+    {
+      "epoch": 0.8422076343301726,
+      "grad_norm": 17.875,
+      "learning_rate": 1.5837038659431059e-06,
+      "loss": 0.79,
+      "step": 3464
+    },
+    {
+      "epoch": 0.8424507658643327,
+      "grad_norm": 17.375,
+      "learning_rate": 1.583238117711349e-06,
+      "loss": 0.822,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8426938973984925,
+      "grad_norm": 17.375,
+      "learning_rate": 1.5827723196713998e-06,
+      "loss": 0.814,
+      "step": 3466
+    },
+    {
+      "epoch": 0.8429370289326525,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5823064718928807e-06,
+      "loss": 0.6785,
+      "step": 3467
+    },
+    {
+      "epoch": 0.8431801604668125,
+      "grad_norm": 18.75,
+      "learning_rate": 1.58184057444542e-06,
+      "loss": 0.6204,
+      "step": 3468
+    },
+    {
+      "epoch": 0.8434232920009725,
+      "grad_norm": 19.75,
+      "learning_rate": 1.5813746273986541e-06,
+      "loss": 1.026,
+      "step": 3469
+    },
+    {
+      "epoch": 0.8436664235351325,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5809086308222273e-06,
+      "loss": 0.8074,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8439095550692924,
+      "grad_norm": 20.75,
+      "learning_rate": 1.5804425847857908e-06,
+      "loss": 0.9195,
+      "step": 3471
+    },
+    {
+      "epoch": 0.8441526866034524,
+      "grad_norm": 21.25,
+      "learning_rate": 1.5799764893590033e-06,
+      "loss": 1.2829,
+      "step": 3472
+    },
+    {
+      "epoch": 0.8443958181376124,
+      "grad_norm": 20.5,
+      "learning_rate": 1.5795103446115302e-06,
+      "loss": 0.7941,
+      "step": 3473
+    },
+    {
+      "epoch": 0.8446389496717724,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.5790441506130453e-06,
+      "loss": 0.7508,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8448820812059324,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5785779074332292e-06,
+      "loss": 1.1244,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8451252127400923,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5781116151417703e-06,
+      "loss": 0.7926,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8453683442742523,
+      "grad_norm": 24.625,
+      "learning_rate": 1.5776452738083637e-06,
+      "loss": 0.9534,
+      "step": 3477
+    },
+    {
+      "epoch": 0.8456114758084123,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5771788835027122e-06,
+      "loss": 0.9546,
+      "step": 3478
+    },
+    {
+      "epoch": 0.8458546073425723,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5767124442945264e-06,
+      "loss": 0.7319,
+      "step": 3479
+    },
+    {
+      "epoch": 0.8460977388767323,
+      "grad_norm": 15.125,
+      "learning_rate": 1.576245956253523e-06,
+      "loss": 0.7444,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8463408704108923,
+      "grad_norm": 20.0,
+      "learning_rate": 1.575779419449427e-06,
+      "loss": 0.9531,
+      "step": 3481
+    },
+    {
+      "epoch": 0.8465840019450522,
+      "grad_norm": 20.875,
+      "learning_rate": 1.5753128339519702e-06,
+      "loss": 0.9928,
+      "step": 3482
+    },
+    {
+      "epoch": 0.8468271334792122,
+      "grad_norm": 20.5,
+      "learning_rate": 1.574846199830892e-06,
+      "loss": 0.7358,
+      "step": 3483
+    },
+    {
+      "epoch": 0.8470702650133722,
+      "grad_norm": 18.0,
+      "learning_rate": 1.5743795171559392e-06,
+      "loss": 0.78,
+      "step": 3484
+    },
+    {
+      "epoch": 0.8473133965475322,
+      "grad_norm": 17.0,
+      "learning_rate": 1.5739127859968652e-06,
+      "loss": 0.5908,
+      "step": 3485
+    },
+    {
+      "epoch": 0.8475565280816922,
+      "grad_norm": 20.875,
+      "learning_rate": 1.5734460064234314e-06,
+      "loss": 0.9047,
+      "step": 3486
+    },
+    {
+      "epoch": 0.8477996596158521,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.5729791785054056e-06,
+      "loss": 0.5851,
+      "step": 3487
+    },
+    {
+      "epoch": 0.8480427911500121,
+      "grad_norm": 14.75,
+      "learning_rate": 1.5725123023125633e-06,
+      "loss": 0.9931,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8482859226841721,
+      "grad_norm": 17.0,
+      "learning_rate": 1.572045377914688e-06,
+      "loss": 0.7318,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8485290542183321,
+      "grad_norm": 17.875,
+      "learning_rate": 1.5715784053815687e-06,
+      "loss": 0.7543,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8487721857524921,
+      "grad_norm": 17.25,
+      "learning_rate": 1.5711113847830029e-06,
+      "loss": 0.6945,
+      "step": 3491
+    },
+    {
+      "epoch": 0.849015317286652,
+      "grad_norm": 21.5,
+      "learning_rate": 1.5706443161887948e-06,
+      "loss": 0.6763,
+      "step": 3492
+    },
+    {
+      "epoch": 0.849258448820812,
+      "grad_norm": 22.625,
+      "learning_rate": 1.570177199668756e-06,
+      "loss": 1.0924,
+      "step": 3493
+    },
+    {
+      "epoch": 0.849501580354972,
+      "grad_norm": 23.125,
+      "learning_rate": 1.569710035292705e-06,
+      "loss": 0.9922,
+      "step": 3494
+    },
+    {
+      "epoch": 0.849744711889132,
+      "grad_norm": 24.125,
+      "learning_rate": 1.5692428231304676e-06,
+      "loss": 0.8303,
+      "step": 3495
+    },
+    {
+      "epoch": 0.849987843423292,
+      "grad_norm": 25.75,
+      "learning_rate": 1.5687755632518772e-06,
+      "loss": 0.9958,
+      "step": 3496
+    },
+    {
+      "epoch": 0.850230974957452,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.5683082557267728e-06,
+      "loss": 0.3681,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8504741064916119,
+      "grad_norm": 25.25,
+      "learning_rate": 1.567840900625003e-06,
+      "loss": 0.8429,
+      "step": 3498
+    },
+    {
+      "epoch": 0.8507172380257719,
+      "grad_norm": 23.75,
+      "learning_rate": 1.5673734980164204e-06,
+      "loss": 0.8081,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8509603695599319,
+      "grad_norm": 23.0,
+      "learning_rate": 1.5669060479708878e-06,
+      "loss": 0.9996,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8512035010940919,
+      "grad_norm": 19.125,
+      "learning_rate": 1.566438550558273e-06,
+      "loss": 0.906,
+      "step": 3501
+    },
+    {
+      "epoch": 0.8514466326282519,
+      "grad_norm": 16.375,
+      "learning_rate": 1.5659710058484518e-06,
+      "loss": 0.3945,
+      "step": 3502
+    },
+    {
+      "epoch": 0.8516897641624118,
+      "grad_norm": 21.5,
+      "learning_rate": 1.5655034139113072e-06,
+      "loss": 0.812,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8519328956965718,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.5650357748167278e-06,
+      "loss": 0.8199,
+      "step": 3504
+    },
+    {
+      "epoch": 0.8521760272307318,
+      "grad_norm": 23.0,
+      "learning_rate": 1.5645680886346112e-06,
+      "loss": 1.5766,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8524191587648918,
+      "grad_norm": 18.0,
+      "learning_rate": 1.564100355434861e-06,
+      "loss": 0.7244,
+      "step": 3506
+    },
+    {
+      "epoch": 0.8526622902990518,
+      "grad_norm": 15.25,
+      "learning_rate": 1.563632575287388e-06,
+      "loss": 0.6229,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8529054218332117,
+      "grad_norm": 19.75,
+      "learning_rate": 1.56316474826211e-06,
+      "loss": 0.9417,
+      "step": 3508
+    },
+    {
+      "epoch": 0.8531485533673717,
+      "grad_norm": 20.75,
+      "learning_rate": 1.5626968744289516e-06,
+      "loss": 0.8819,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8533916849015317,
+      "grad_norm": 20.0,
+      "learning_rate": 1.5622289538578453e-06,
+      "loss": 0.9293,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8536348164356917,
+      "grad_norm": 18.875,
+      "learning_rate": 1.5617609866187291e-06,
+      "loss": 0.8597,
+      "step": 3511
+    },
+    {
+      "epoch": 0.8538779479698517,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5612929727815494e-06,
+      "loss": 0.6089,
+      "step": 3512
+    },
+    {
+      "epoch": 0.8541210795040116,
+      "grad_norm": 17.5,
+      "learning_rate": 1.5608249124162586e-06,
+      "loss": 0.7872,
+      "step": 3513
+    },
+    {
+      "epoch": 0.8543642110381716,
+      "grad_norm": 19.875,
+      "learning_rate": 1.5603568055928164e-06,
+      "loss": 0.9658,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8546073425723316,
+      "grad_norm": 20.375,
+      "learning_rate": 1.5598886523811898e-06,
+      "loss": 0.8721,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8548504741064916,
+      "grad_norm": 18.625,
+      "learning_rate": 1.559420452851352e-06,
+      "loss": 0.7256,
+      "step": 3516
+    },
+    {
+      "epoch": 0.8550936056406516,
+      "grad_norm": 19.75,
+      "learning_rate": 1.5589522070732838e-06,
+      "loss": 0.8953,
+      "step": 3517
+    },
+    {
+      "epoch": 0.8553367371748116,
+      "grad_norm": 16.875,
+      "learning_rate": 1.558483915116972e-06,
+      "loss": 0.6124,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8555798687089715,
+      "grad_norm": 14.5,
+      "learning_rate": 1.5580155770524119e-06,
+      "loss": 0.4481,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8558230002431315,
+      "grad_norm": 17.25,
+      "learning_rate": 1.557547192949604e-06,
+      "loss": 0.9589,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8560661317772915,
+      "grad_norm": 21.375,
+      "learning_rate": 1.5570787628785563e-06,
+      "loss": 0.9213,
+      "step": 3521
+    },
+    {
+      "epoch": 0.8563092633114515,
+      "grad_norm": 20.5,
+      "learning_rate": 1.5566102869092847e-06,
+      "loss": 0.661,
+      "step": 3522
+    },
+    {
+      "epoch": 0.8565523948456115,
+      "grad_norm": 23.0,
+      "learning_rate": 1.5561417651118098e-06,
+      "loss": 0.9228,
+      "step": 3523
+    },
+    {
+      "epoch": 0.8567955263797714,
+      "grad_norm": 16.625,
+      "learning_rate": 1.5556731975561613e-06,
+      "loss": 0.3625,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8570386579139314,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5552045843123737e-06,
+      "loss": 1.1483,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8572817894480914,
+      "grad_norm": 19.625,
+      "learning_rate": 1.5547359254504903e-06,
+      "loss": 0.6586,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8575249209822514,
+      "grad_norm": 23.125,
+      "learning_rate": 1.5542672210405603e-06,
+      "loss": 0.9238,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8577680525164114,
+      "grad_norm": 21.875,
+      "learning_rate": 1.5537984711526382e-06,
+      "loss": 0.9914,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8580111840505713,
+      "grad_norm": 20.5,
+      "learning_rate": 1.5533296758567884e-06,
+      "loss": 0.9804,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8582543155847313,
+      "grad_norm": 24.875,
+      "learning_rate": 1.5528608352230798e-06,
+      "loss": 1.1042,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8584974471188913,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5523919493215888e-06,
+      "loss": 0.88,
+      "step": 3531
+    },
+    {
+      "epoch": 0.8587405786530513,
+      "grad_norm": 21.75,
+      "learning_rate": 1.5519230182223984e-06,
+      "loss": 0.8651,
+      "step": 3532
+    },
+    {
+      "epoch": 0.8589837101872113,
+      "grad_norm": 19.875,
+      "learning_rate": 1.5514540419955986e-06,
+      "loss": 0.7751,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8592268417213713,
+      "grad_norm": 17.5,
+      "learning_rate": 1.550985020711286e-06,
+      "loss": 0.6555,
+      "step": 3534
+    },
+    {
+      "epoch": 0.8594699732555312,
+      "grad_norm": 18.375,
+      "learning_rate": 1.550515954439564e-06,
+      "loss": 0.829,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8597131047896912,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5500468432505422e-06,
+      "loss": 0.7958,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8599562363238512,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5495776872143379e-06,
+      "loss": 0.8569,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8601993678580112,
+      "grad_norm": 18.25,
+      "learning_rate": 1.5491084864010741e-06,
+      "loss": 0.6053,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8604424993921712,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5486392408808818e-06,
+      "loss": 0.7797,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8606856309263311,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5481699507238965e-06,
+      "loss": 0.7783,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8609287624604911,
+      "grad_norm": 18.25,
+      "learning_rate": 1.5477006160002631e-06,
+      "loss": 0.6464,
+      "step": 3541
+    },
+    {
+      "epoch": 0.8611718939946511,
+      "grad_norm": 20.875,
+      "learning_rate": 1.547231236780131e-06,
+      "loss": 0.7141,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8614150255288111,
+      "grad_norm": 21.875,
+      "learning_rate": 1.546761813133657e-06,
+      "loss": 0.877,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8616581570629711,
+      "grad_norm": 24.125,
+      "learning_rate": 1.5462923451310049e-06,
+      "loss": 1.0813,
+      "step": 3544
+    },
+    {
+      "epoch": 0.861901288597131,
+      "grad_norm": 17.75,
+      "learning_rate": 1.5458228328423447e-06,
+      "loss": 0.6834,
+      "step": 3545
+    },
+    {
+      "epoch": 0.862144420131291,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.545353276337853e-06,
+      "loss": 0.4675,
+      "step": 3546
+    },
+    {
+      "epoch": 0.862387551665451,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5448836756877135e-06,
+      "loss": 0.8901,
+      "step": 3547
+    },
+    {
+      "epoch": 0.862630683199611,
+      "grad_norm": 21.625,
+      "learning_rate": 1.5444140309621153e-06,
+      "loss": 1.0051,
+      "step": 3548
+    },
+    {
+      "epoch": 0.862873814733771,
+      "grad_norm": 19.5,
+      "learning_rate": 1.5439443422312562e-06,
+      "loss": 0.8683,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8631169462679309,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.543474609565338e-06,
+      "loss": 0.7164,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8633600778020909,
+      "grad_norm": 17.75,
+      "learning_rate": 1.5430048330345712e-06,
+      "loss": 0.5809,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8636032093362509,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5425350127091716e-06,
+      "loss": 0.6418,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8638463408704109,
+      "grad_norm": 22.0,
+      "learning_rate": 1.5420651486593624e-06,
+      "loss": 0.9831,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8640894724045709,
+      "grad_norm": 25.25,
+      "learning_rate": 1.5415952409553721e-06,
+      "loss": 0.9321,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8643326039387309,
+      "grad_norm": 25.125,
+      "learning_rate": 1.5411252896674369e-06,
+      "loss": 0.6299,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8645757354728908,
+      "grad_norm": 24.25,
+      "learning_rate": 1.5406552948658e-06,
+      "loss": 1.4371,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8648188670070508,
+      "grad_norm": 23.0,
+      "learning_rate": 1.540185256620709e-06,
+      "loss": 1.1024,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8650619985412108,
+      "grad_norm": 19.125,
+      "learning_rate": 1.53971517500242e-06,
+      "loss": 0.7464,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8653051300753708,
+      "grad_norm": 20.625,
+      "learning_rate": 1.539245050081194e-06,
+      "loss": 0.9117,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8655482616095308,
+      "grad_norm": 22.125,
+      "learning_rate": 1.5387748819273001e-06,
+      "loss": 0.8153,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8657913931436907,
+      "grad_norm": 21.25,
+      "learning_rate": 1.5383046706110133e-06,
+      "loss": 1.0534,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8660345246778507,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.5378344162026137e-06,
+      "loss": 0.3717,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8662776562120107,
+      "grad_norm": 16.375,
+      "learning_rate": 1.5373641187723898e-06,
+      "loss": 0.666,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8665207877461707,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5368937783906352e-06,
+      "loss": 0.7635,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8667639192803307,
+      "grad_norm": 20.75,
+      "learning_rate": 1.5364233951276505e-06,
+      "loss": 0.6799,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8670070508144906,
+      "grad_norm": 82.0,
+      "learning_rate": 1.5359529690537431e-06,
+      "loss": 1.3364,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8672501823486506,
+      "grad_norm": 23.875,
+      "learning_rate": 1.5354825002392254e-06,
+      "loss": 1.0862,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8674933138828106,
+      "grad_norm": 21.5,
+      "learning_rate": 1.535011988754418e-06,
+      "loss": 1.0549,
+      "step": 3568
+    },
+    {
+      "epoch": 0.8677364454169706,
+      "grad_norm": 19.25,
+      "learning_rate": 1.5345414346696463e-06,
+      "loss": 0.6873,
+      "step": 3569
+    },
+    {
+      "epoch": 0.8679795769511306,
+      "grad_norm": 17.125,
+      "learning_rate": 1.5340708380552436e-06,
+      "loss": 0.7122,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8682227084852906,
+      "grad_norm": 17.0,
+      "learning_rate": 1.5336001989815472e-06,
+      "loss": 0.655,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8684658400194505,
+      "grad_norm": 19.75,
+      "learning_rate": 1.5331295175189034e-06,
+      "loss": 0.7531,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8687089715536105,
+      "grad_norm": 17.375,
+      "learning_rate": 1.5326587937376635e-06,
+      "loss": 0.7692,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8689521030877705,
+      "grad_norm": 22.5,
+      "learning_rate": 1.5321880277081852e-06,
+      "loss": 0.8734,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8691952346219305,
+      "grad_norm": 18.875,
+      "learning_rate": 1.5317172195008326e-06,
+      "loss": 0.583,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8694383661560905,
+      "grad_norm": 17.25,
+      "learning_rate": 1.531246369185976e-06,
+      "loss": 0.7171,
+      "step": 3576
+    },
+    {
+      "epoch": 0.8696814976902504,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5307754768339922e-06,
+      "loss": 0.9413,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8699246292244104,
+      "grad_norm": 20.625,
+      "learning_rate": 1.5303045425152643e-06,
+      "loss": 0.9347,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8701677607585704,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.5298335663001814e-06,
+      "loss": 0.7473,
+      "step": 3579
+    },
+    {
+      "epoch": 0.8704108922927304,
+      "grad_norm": 23.25,
+      "learning_rate": 1.5293625482591396e-06,
+      "loss": 1.1312,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8706540238268904,
+      "grad_norm": 18.625,
+      "learning_rate": 1.52889148846254e-06,
+      "loss": 0.5924,
+      "step": 3581
+    },
+    {
+      "epoch": 0.8708971553610503,
+      "grad_norm": 21.75,
+      "learning_rate": 1.5284203869807906e-06,
+      "loss": 0.9332,
+      "step": 3582
+    },
+    {
+      "epoch": 0.8711402868952103,
+      "grad_norm": 20.875,
+      "learning_rate": 1.5279492438843058e-06,
+      "loss": 0.7122,
+      "step": 3583
+    },
+    {
+      "epoch": 0.8713834184293703,
+      "grad_norm": 20.5,
+      "learning_rate": 1.5274780592435064e-06,
+      "loss": 1.2327,
+      "step": 3584
+    },
+    {
+      "epoch": 0.8716265499635303,
+      "grad_norm": 40.75,
+      "learning_rate": 1.527006833128819e-06,
+      "loss": 0.9345,
+      "step": 3585
+    },
+    {
+      "epoch": 0.8718696814976903,
+      "grad_norm": 33.75,
+      "learning_rate": 1.5265355656106757e-06,
+      "loss": 1.377,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8721128130318502,
+      "grad_norm": 18.125,
+      "learning_rate": 1.526064256759517e-06,
+      "loss": 1.1876,
+      "step": 3587
+    },
+    {
+      "epoch": 0.8723559445660102,
+      "grad_norm": 16.75,
+      "learning_rate": 1.5255929066457868e-06,
+      "loss": 1.1715,
+      "step": 3588
+    },
+    {
+      "epoch": 0.8725990761001702,
+      "grad_norm": 17.25,
+      "learning_rate": 1.525121515339937e-06,
+      "loss": 0.5743,
+      "step": 3589
+    },
+    {
+      "epoch": 0.8728422076343302,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5246500829124253e-06,
+      "loss": 0.9771,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8730853391684902,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.5241786094337151e-06,
+      "loss": 0.6867,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8733284707026502,
+      "grad_norm": 16.125,
+      "learning_rate": 1.5237070949742772e-06,
+      "loss": 0.5928,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8735716022368101,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.5232355396045864e-06,
+      "loss": 0.447,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8738147337709701,
+      "grad_norm": 28.375,
+      "learning_rate": 1.5227639433951252e-06,
+      "loss": 0.8632,
+      "step": 3594
+    },
+    {
+      "epoch": 0.8740578653051301,
+      "grad_norm": 20.25,
+      "learning_rate": 1.5222923064163822e-06,
+      "loss": 0.8428,
+      "step": 3595
+    },
+    {
+      "epoch": 0.8743009968392901,
+      "grad_norm": 23.0,
+      "learning_rate": 1.521820628738851e-06,
+      "loss": 1.3935,
+      "step": 3596
+    },
+    {
+      "epoch": 0.8745441283734501,
+      "grad_norm": 19.375,
+      "learning_rate": 1.5213489104330328e-06,
+      "loss": 0.6194,
+      "step": 3597
+    },
+    {
+      "epoch": 0.87478725990761,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5208771515694329e-06,
+      "loss": 0.7012,
+      "step": 3598
+    },
+    {
+      "epoch": 0.87503039144177,
+      "grad_norm": 16.75,
+      "learning_rate": 1.520405352218565e-06,
+      "loss": 0.681,
+      "step": 3599
+    },
+    {
+      "epoch": 0.87527352297593,
+      "grad_norm": 19.5,
+      "learning_rate": 1.519933512450947e-06,
+      "loss": 0.7969,
+      "step": 3600
+    },
+    {
+      "epoch": 0.87551665451009,
+      "grad_norm": 16.5,
+      "learning_rate": 1.5194616323371036e-06,
+      "loss": 0.6389,
+      "step": 3601
+    },
+    {
+      "epoch": 0.87575978604425,
+      "grad_norm": 19.75,
+      "learning_rate": 1.5189897119475654e-06,
+      "loss": 1.2287,
+      "step": 3602
+    },
+    {
+      "epoch": 0.8760029175784099,
+      "grad_norm": 23.875,
+      "learning_rate": 1.5185177513528693e-06,
+      "loss": 1.3549,
+      "step": 3603
+    },
+    {
+      "epoch": 0.8762460491125699,
+      "grad_norm": 18.125,
+      "learning_rate": 1.518045750623558e-06,
+      "loss": 1.167,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8764891806467299,
+      "grad_norm": 24.875,
+      "learning_rate": 1.5175737098301792e-06,
+      "loss": 1.3455,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8767323121808899,
+      "grad_norm": 22.625,
+      "learning_rate": 1.517101629043289e-06,
+      "loss": 0.9083,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8769754437150499,
+      "grad_norm": 17.5,
+      "learning_rate": 1.5166295083334473e-06,
+      "loss": 0.8156,
+      "step": 3607
+    },
+    {
+      "epoch": 0.8772185752492099,
+      "grad_norm": 16.375,
+      "learning_rate": 1.5161573477712205e-06,
+      "loss": 0.5163,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8774617067833698,
+      "grad_norm": 22.375,
+      "learning_rate": 1.5156851474271815e-06,
+      "loss": 1.2113,
+      "step": 3609
+    },
+    {
+      "epoch": 0.8777048383175298,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5152129073719085e-06,
+      "loss": 0.9579,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8779479698516898,
+      "grad_norm": 16.875,
+      "learning_rate": 1.5147406276759865e-06,
+      "loss": 0.4668,
+      "step": 3611
+    },
+    {
+      "epoch": 0.8781911013858498,
+      "grad_norm": 19.125,
+      "learning_rate": 1.514268308410005e-06,
+      "loss": 0.5061,
+      "step": 3612
+    },
+    {
+      "epoch": 0.8784342329200098,
+      "grad_norm": 17.125,
+      "learning_rate": 1.5137959496445612e-06,
+      "loss": 1.0777,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8786773644541697,
+      "grad_norm": 17.625,
+      "learning_rate": 1.5133235514502564e-06,
+      "loss": 0.7726,
+      "step": 3614
+    },
+    {
+      "epoch": 0.8789204959883297,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.5128511138976992e-06,
+      "loss": 0.7334,
+      "step": 3615
+    },
+    {
+      "epoch": 0.8791636275224897,
+      "grad_norm": 16.75,
+      "learning_rate": 1.5123786370575038e-06,
+      "loss": 0.5958,
+      "step": 3616
+    },
+    {
+      "epoch": 0.8794067590566497,
+      "grad_norm": 18.25,
+      "learning_rate": 1.5119061210002892e-06,
+      "loss": 0.8657,
+      "step": 3617
+    },
+    {
+      "epoch": 0.8796498905908097,
+      "grad_norm": 19.0,
+      "learning_rate": 1.5114335657966816e-06,
+      "loss": 1.0357,
+      "step": 3618
+    },
+    {
+      "epoch": 0.8798930221249696,
+      "grad_norm": 20.375,
+      "learning_rate": 1.5109609715173127e-06,
+      "loss": 0.795,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8801361536591296,
+      "grad_norm": 22.875,
+      "learning_rate": 1.5104883382328195e-06,
+      "loss": 0.9602,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8803792851932896,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5100156660138454e-06,
+      "loss": 0.9176,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8806224167274496,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5095429549310392e-06,
+      "loss": 0.5905,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8808655482616096,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5090702050550562e-06,
+      "loss": 0.6135,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8811086797957695,
+      "grad_norm": 23.0,
+      "learning_rate": 1.5085974164565567e-06,
+      "loss": 1.1195,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8813518113299295,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.5081245892062072e-06,
+      "loss": 0.6659,
+      "step": 3625
+    },
+    {
+      "epoch": 0.8815949428640895,
+      "grad_norm": 17.25,
+      "learning_rate": 1.5076517233746796e-06,
+      "loss": 0.8656,
+      "step": 3626
+    },
+    {
+      "epoch": 0.8818380743982495,
+      "grad_norm": 20.625,
+      "learning_rate": 1.5071788190326521e-06,
+      "loss": 0.7428,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8820812059324095,
+      "grad_norm": 27.0,
+      "learning_rate": 1.506705876250809e-06,
+      "loss": 1.2332,
+      "step": 3628
+    },
+    {
+      "epoch": 0.8823243374665695,
+      "grad_norm": 32.75,
+      "learning_rate": 1.5062328950998386e-06,
+      "loss": 0.8189,
+      "step": 3629
+    },
+    {
+      "epoch": 0.8825674690007294,
+      "grad_norm": 30.375,
+      "learning_rate": 1.5057598756504373e-06,
+      "loss": 1.1142,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8828106005348894,
+      "grad_norm": 33.5,
+      "learning_rate": 1.5052868179733054e-06,
+      "loss": 0.918,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8830537320690494,
+      "grad_norm": 26.125,
+      "learning_rate": 1.5048137221391493e-06,
+      "loss": 1.3675,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8832968636032094,
+      "grad_norm": 21.625,
+      "learning_rate": 1.5043405882186819e-06,
+      "loss": 0.8361,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8835399951373694,
+      "grad_norm": 18.875,
+      "learning_rate": 1.5038674162826205e-06,
+      "loss": 0.5911,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8837831266715293,
+      "grad_norm": 18.75,
+      "learning_rate": 1.50339420640169e-06,
+      "loss": 0.7616,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8840262582056893,
+      "grad_norm": 22.875,
+      "learning_rate": 1.5029209586466184e-06,
+      "loss": 1.0386,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8842693897398493,
+      "grad_norm": 23.0,
+      "learning_rate": 1.502447673088142e-06,
+      "loss": 0.999,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8845125212740093,
+      "grad_norm": 25.875,
+      "learning_rate": 1.5019743497970008e-06,
+      "loss": 0.8509,
+      "step": 3638
+    },
+    {
+      "epoch": 0.8847556528081693,
+      "grad_norm": 18.25,
+      "learning_rate": 1.5015009888439408e-06,
+      "loss": 0.5531,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8849987843423291,
+      "grad_norm": 23.375,
+      "learning_rate": 1.5010275902997148e-06,
+      "loss": 0.7159,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8852419158764891,
+      "grad_norm": 12.75,
+      "learning_rate": 1.5005541542350802e-06,
+      "loss": 0.3923,
+      "step": 3641
+    },
+    {
+      "epoch": 0.8854850474106492,
+      "grad_norm": 20.125,
+      "learning_rate": 1.5000806807207999e-06,
+      "loss": 0.7973,
+      "step": 3642
+    },
+    {
+      "epoch": 0.8857281789448092,
+      "grad_norm": 17.0,
+      "learning_rate": 1.499607169827643e-06,
+      "loss": 1.0972,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8859713104789692,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4991336216263833e-06,
+      "loss": 0.7585,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8862144420131292,
+      "grad_norm": 19.625,
+      "learning_rate": 1.4986600361878012e-06,
+      "loss": 1.1922,
+      "step": 3645
+    },
+    {
+      "epoch": 0.886457573547289,
+      "grad_norm": 26.375,
+      "learning_rate": 1.4981864135826823e-06,
+      "loss": 1.1931,
+      "step": 3646
+    },
+    {
+      "epoch": 0.886700705081449,
+      "grad_norm": 19.5,
+      "learning_rate": 1.497712753881818e-06,
+      "loss": 0.8558,
+      "step": 3647
+    },
+    {
+      "epoch": 0.886943836615609,
+      "grad_norm": 16.875,
+      "learning_rate": 1.4972390571560035e-06,
+      "loss": 0.6759,
+      "step": 3648
+    },
+    {
+      "epoch": 0.887186968149769,
+      "grad_norm": 20.0,
+      "learning_rate": 1.496765323476043e-06,
+      "loss": 1.4714,
+      "step": 3649
+    },
+    {
+      "epoch": 0.887430099683929,
+      "grad_norm": 18.25,
+      "learning_rate": 1.4962915529127426e-06,
+      "loss": 0.779,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8876732312180889,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.495817745536916e-06,
+      "loss": 0.7506,
+      "step": 3651
+    },
+    {
+      "epoch": 0.8879163627522489,
+      "grad_norm": 17.75,
+      "learning_rate": 1.495343901419382e-06,
+      "loss": 0.7888,
+      "step": 3652
+    },
+    {
+      "epoch": 0.888159494286409,
+      "grad_norm": 19.75,
+      "learning_rate": 1.4948700206309645e-06,
+      "loss": 0.9544,
+      "step": 3653
+    },
+    {
+      "epoch": 0.888402625820569,
+      "grad_norm": 17.375,
+      "learning_rate": 1.4943961032424939e-06,
+      "loss": 0.877,
+      "step": 3654
+    },
+    {
+      "epoch": 0.888645757354729,
+      "grad_norm": 17.5,
+      "learning_rate": 1.4939221493248043e-06,
+      "loss": 0.6839,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 18.0,
+      "learning_rate": 1.493448158948737e-06,
+      "loss": 0.6713,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8891320204230488,
+      "grad_norm": 18.875,
+      "learning_rate": 1.4929741321851376e-06,
+      "loss": 1.162,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8893751519572088,
+      "grad_norm": 20.75,
+      "learning_rate": 1.492500069104858e-06,
+      "loss": 0.7053,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8896182834913688,
+      "grad_norm": 13.5,
+      "learning_rate": 1.492025969778755e-06,
+      "loss": 0.5691,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8898614150255288,
+      "grad_norm": 20.25,
+      "learning_rate": 1.4915518342776909e-06,
+      "loss": 1.1944,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8901045465596887,
+      "grad_norm": 24.875,
+      "learning_rate": 1.4910776626725336e-06,
+      "loss": 1.1712,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8903476780938487,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.4906034550341559e-06,
+      "loss": 0.7042,
+      "step": 3662
+    },
+    {
+      "epoch": 0.8905908096280087,
+      "grad_norm": 18.875,
+      "learning_rate": 1.4901292114334362e-06,
+      "loss": 0.9218,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8908339411621687,
+      "grad_norm": 23.5,
+      "learning_rate": 1.489654931941259e-06,
+      "loss": 0.8784,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8910770726963287,
+      "grad_norm": 28.625,
+      "learning_rate": 1.4891806166285131e-06,
+      "loss": 1.4625,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8913202042304887,
+      "grad_norm": 17.5,
+      "learning_rate": 1.4887062655660933e-06,
+      "loss": 1.1436,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8915633357646486,
+      "grad_norm": 22.25,
+      "learning_rate": 1.4882318788248996e-06,
+      "loss": 0.7973,
+      "step": 3667
+    },
+    {
+      "epoch": 0.8918064672988086,
+      "grad_norm": 20.125,
+      "learning_rate": 1.487757456475837e-06,
+      "loss": 0.7844,
+      "step": 3668
+    },
+    {
+      "epoch": 0.8920495988329686,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4872829985898167e-06,
+      "loss": 0.9402,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8922927303671286,
+      "grad_norm": 21.0,
+      "learning_rate": 1.4868085052377538e-06,
+      "loss": 0.7525,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8925358619012886,
+      "grad_norm": 25.625,
+      "learning_rate": 1.4863339764905698e-06,
+      "loss": 1.0096,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8927789934354485,
+      "grad_norm": 19.625,
+      "learning_rate": 1.485859412419192e-06,
+      "loss": 0.5815,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8930221249696085,
+      "grad_norm": 23.875,
+      "learning_rate": 1.4853848130945514e-06,
+      "loss": 0.9197,
+      "step": 3673
+    },
+    {
+      "epoch": 0.8932652565037685,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4849101785875853e-06,
+      "loss": 0.8862,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8935083880379285,
+      "grad_norm": 16.625,
+      "learning_rate": 1.4844355089692352e-06,
+      "loss": 0.6091,
+      "step": 3675
+    },
+    {
+      "epoch": 0.8937515195720885,
+      "grad_norm": 20.125,
+      "learning_rate": 1.4839608043104503e-06,
+      "loss": 1.1555,
+      "step": 3676
+    },
+    {
+      "epoch": 0.8939946511062484,
+      "grad_norm": 16.75,
+      "learning_rate": 1.4834860646821825e-06,
+      "loss": 0.6236,
+      "step": 3677
+    },
+    {
+      "epoch": 0.8942377826404084,
+      "grad_norm": 24.25,
+      "learning_rate": 1.4830112901553898e-06,
+      "loss": 1.389,
+      "step": 3678
+    },
+    {
+      "epoch": 0.8944809141745684,
+      "grad_norm": 17.75,
+      "learning_rate": 1.482536480801035e-06,
+      "loss": 0.6193,
+      "step": 3679
+    },
+    {
+      "epoch": 0.8947240457087284,
+      "grad_norm": 21.375,
+      "learning_rate": 1.4820616366900878e-06,
+      "loss": 0.6483,
+      "step": 3680
+    },
+    {
+      "epoch": 0.8949671772428884,
+      "grad_norm": 19.875,
+      "learning_rate": 1.481586757893521e-06,
+      "loss": 0.8801,
+      "step": 3681
+    },
+    {
+      "epoch": 0.8952103087770484,
+      "grad_norm": 21.5,
+      "learning_rate": 1.4811118444823133e-06,
+      "loss": 0.6479,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8954534403112083,
+      "grad_norm": 22.0,
+      "learning_rate": 1.4806368965274492e-06,
+      "loss": 1.0339,
+      "step": 3683
+    },
+    {
+      "epoch": 0.8956965718453683,
+      "grad_norm": 18.25,
+      "learning_rate": 1.4801619140999176e-06,
+      "loss": 0.9632,
+      "step": 3684
+    },
+    {
+      "epoch": 0.8959397033795283,
+      "grad_norm": 15.625,
+      "learning_rate": 1.4796868972707132e-06,
+      "loss": 0.695,
+      "step": 3685
+    },
+    {
+      "epoch": 0.8961828349136883,
+      "grad_norm": 26.5,
+      "learning_rate": 1.479211846110835e-06,
+      "loss": 0.8159,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8964259664478483,
+      "grad_norm": 22.125,
+      "learning_rate": 1.4787367606912872e-06,
+      "loss": 1.1078,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8966690979820082,
+      "grad_norm": 18.375,
+      "learning_rate": 1.478261641083081e-06,
+      "loss": 0.6567,
+      "step": 3688
+    },
+    {
+      "epoch": 0.8969122295161682,
+      "grad_norm": 20.25,
+      "learning_rate": 1.47778648735723e-06,
+      "loss": 0.7557,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8971553610503282,
+      "grad_norm": 16.0,
+      "learning_rate": 1.4773112995847543e-06,
+      "loss": 0.7756,
+      "step": 3690
+    },
+    {
+      "epoch": 0.8973984925844882,
+      "grad_norm": 17.5,
+      "learning_rate": 1.4768360778366791e-06,
+      "loss": 0.5595,
+      "step": 3691
+    },
+    {
+      "epoch": 0.8976416241186482,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4763608221840346e-06,
+      "loss": 1.1363,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8978847556528081,
+      "grad_norm": 18.0,
+      "learning_rate": 1.475885532697856e-06,
+      "loss": 0.909,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8981278871869681,
+      "grad_norm": 21.625,
+      "learning_rate": 1.4754102094491826e-06,
+      "loss": 1.0297,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8983710187211281,
+      "grad_norm": 34.25,
+      "learning_rate": 1.4749348525090611e-06,
+      "loss": 0.9239,
+      "step": 3695
+    },
+    {
+      "epoch": 0.8986141502552881,
+      "grad_norm": 22.0,
+      "learning_rate": 1.4744594619485412e-06,
+      "loss": 1.225,
+      "step": 3696
+    },
+    {
+      "epoch": 0.8988572817894481,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.4739840378386782e-06,
+      "loss": 0.3779,
+      "step": 3697
+    },
+    {
+      "epoch": 0.899100413323608,
+      "grad_norm": 19.875,
+      "learning_rate": 1.473508580250532e-06,
+      "loss": 0.8047,
+      "step": 3698
+    },
+    {
+      "epoch": 0.899343544857768,
+      "grad_norm": 28.625,
+      "learning_rate": 1.4730330892551684e-06,
+      "loss": 0.9387,
+      "step": 3699
+    },
+    {
+      "epoch": 0.899586676391928,
+      "grad_norm": 20.375,
+      "learning_rate": 1.4725575649236578e-06,
+      "loss": 0.5998,
+      "step": 3700
+    },
+    {
+      "epoch": 0.899829807926088,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4720820073270755e-06,
+      "loss": 0.6035,
+      "step": 3701
+    },
+    {
+      "epoch": 0.900072939460248,
+      "grad_norm": 28.0,
+      "learning_rate": 1.4716064165365018e-06,
+      "loss": 0.7062,
+      "step": 3702
+    },
+    {
+      "epoch": 0.900316070994408,
+      "grad_norm": 15.75,
+      "learning_rate": 1.4711307926230216e-06,
+      "loss": 0.7638,
+      "step": 3703
+    },
+    {
+      "epoch": 0.9005592025285679,
+      "grad_norm": 17.625,
+      "learning_rate": 1.4706551356577253e-06,
+      "loss": 0.7957,
+      "step": 3704
+    },
+    {
+      "epoch": 0.9008023340627279,
+      "grad_norm": 17.25,
+      "learning_rate": 1.4701794457117088e-06,
+      "loss": 0.6963,
+      "step": 3705
+    },
+    {
+      "epoch": 0.9010454655968879,
+      "grad_norm": 20.875,
+      "learning_rate": 1.4697037228560706e-06,
+      "loss": 0.5563,
+      "step": 3706
+    },
+    {
+      "epoch": 0.9012885971310479,
+      "grad_norm": 24.375,
+      "learning_rate": 1.4692279671619173e-06,
+      "loss": 1.057,
+      "step": 3707
+    },
+    {
+      "epoch": 0.9015317286652079,
+      "grad_norm": 15.75,
+      "learning_rate": 1.4687521787003577e-06,
+      "loss": 0.8086,
+      "step": 3708
+    },
+    {
+      "epoch": 0.9017748601993678,
+      "grad_norm": 20.625,
+      "learning_rate": 1.4682763575425078e-06,
+      "loss": 0.876,
+      "step": 3709
+    },
+    {
+      "epoch": 0.9020179917335278,
+      "grad_norm": 21.375,
+      "learning_rate": 1.4678005037594853e-06,
+      "loss": 0.7953,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9022611232676878,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4673246174224166e-06,
+      "loss": 0.7017,
+      "step": 3711
+    },
+    {
+      "epoch": 0.9025042548018478,
+      "grad_norm": 17.125,
+      "learning_rate": 1.4668486986024305e-06,
+      "loss": 0.7379,
+      "step": 3712
+    },
+    {
+      "epoch": 0.9027473863360078,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4663727473706612e-06,
+      "loss": 0.6913,
+      "step": 3713
+    },
+    {
+      "epoch": 0.9029905178701677,
+      "grad_norm": 22.0,
+      "learning_rate": 1.465896763798248e-06,
+      "loss": 0.8068,
+      "step": 3714
+    },
+    {
+      "epoch": 0.9032336494043277,
+      "grad_norm": 23.375,
+      "learning_rate": 1.4654207479563349e-06,
+      "loss": 1.0214,
+      "step": 3715
+    },
+    {
+      "epoch": 0.9034767809384877,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4649446999160701e-06,
+      "loss": 1.1374,
+      "step": 3716
+    },
+    {
+      "epoch": 0.9037199124726477,
+      "grad_norm": 17.25,
+      "learning_rate": 1.464468619748608e-06,
+      "loss": 0.6149,
+      "step": 3717
+    },
+    {
+      "epoch": 0.9039630440068077,
+      "grad_norm": 20.625,
+      "learning_rate": 1.4639925075251065e-06,
+      "loss": 0.5648,
+      "step": 3718
+    },
+    {
+      "epoch": 0.9042061755409677,
+      "grad_norm": 16.375,
+      "learning_rate": 1.4635163633167296e-06,
+      "loss": 0.8324,
+      "step": 3719
+    },
+    {
+      "epoch": 0.9044493070751276,
+      "grad_norm": 21.5,
+      "learning_rate": 1.463040187194644e-06,
+      "loss": 0.8693,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9046924386092876,
+      "grad_norm": 23.875,
+      "learning_rate": 1.4625639792300233e-06,
+      "loss": 1.0723,
+      "step": 3721
+    },
+    {
+      "epoch": 0.9049355701434476,
+      "grad_norm": 16.25,
+      "learning_rate": 1.4620877394940447e-06,
+      "loss": 0.7044,
+      "step": 3722
+    },
+    {
+      "epoch": 0.9051787016776076,
+      "grad_norm": 23.75,
+      "learning_rate": 1.4616114680578905e-06,
+      "loss": 0.8572,
+      "step": 3723
+    },
+    {
+      "epoch": 0.9054218332117676,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4611351649927482e-06,
+      "loss": 0.5652,
+      "step": 3724
+    },
+    {
+      "epoch": 0.9056649647459275,
+      "grad_norm": 15.625,
+      "learning_rate": 1.4606588303698082e-06,
+      "loss": 0.3811,
+      "step": 3725
+    },
+    {
+      "epoch": 0.9059080962800875,
+      "grad_norm": 19.625,
+      "learning_rate": 1.4601824642602688e-06,
+      "loss": 1.3879,
+      "step": 3726
+    },
+    {
+      "epoch": 0.9061512278142475,
+      "grad_norm": 21.375,
+      "learning_rate": 1.4597060667353296e-06,
+      "loss": 0.9602,
+      "step": 3727
+    },
+    {
+      "epoch": 0.9063943593484075,
+      "grad_norm": 19.5,
+      "learning_rate": 1.4592296378661968e-06,
+      "loss": 1.0221,
+      "step": 3728
+    },
+    {
+      "epoch": 0.9066374908825675,
+      "grad_norm": 25.125,
+      "learning_rate": 1.4587531777240814e-06,
+      "loss": 1.2204,
+      "step": 3729
+    },
+    {
+      "epoch": 0.9068806224167274,
+      "grad_norm": 18.875,
+      "learning_rate": 1.4582766863801984e-06,
+      "loss": 0.971,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9071237539508874,
+      "grad_norm": 15.75,
+      "learning_rate": 1.4578001639057676e-06,
+      "loss": 0.602,
+      "step": 3731
+    },
+    {
+      "epoch": 0.9073668854850474,
+      "grad_norm": 19.5,
+      "learning_rate": 1.4573236103720132e-06,
+      "loss": 0.8516,
+      "step": 3732
+    },
+    {
+      "epoch": 0.9076100170192074,
+      "grad_norm": 16.5,
+      "learning_rate": 1.4568470258501646e-06,
+      "loss": 0.5341,
+      "step": 3733
+    },
+    {
+      "epoch": 0.9078531485533674,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4563704104114557e-06,
+      "loss": 0.7416,
+      "step": 3734
+    },
+    {
+      "epoch": 0.9080962800875274,
+      "grad_norm": 18.125,
+      "learning_rate": 1.4558937641271248e-06,
+      "loss": 0.6536,
+      "step": 3735
+    },
+    {
+      "epoch": 0.9083394116216873,
+      "grad_norm": 22.75,
+      "learning_rate": 1.455417087068415e-06,
+      "loss": 0.8411,
+      "step": 3736
+    },
+    {
+      "epoch": 0.9085825431558473,
+      "grad_norm": 18.0,
+      "learning_rate": 1.4549403793065737e-06,
+      "loss": 0.8491,
+      "step": 3737
+    },
+    {
+      "epoch": 0.9088256746900073,
+      "grad_norm": 17.875,
+      "learning_rate": 1.4544636409128538e-06,
+      "loss": 0.8147,
+      "step": 3738
+    },
+    {
+      "epoch": 0.9090688062241673,
+      "grad_norm": 18.625,
+      "learning_rate": 1.453986871958511e-06,
+      "loss": 0.6442,
+      "step": 3739
+    },
+    {
+      "epoch": 0.9093119377583273,
+      "grad_norm": 24.625,
+      "learning_rate": 1.4535100725148072e-06,
+      "loss": 0.7537,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9095550692924872,
+      "grad_norm": 19.0,
+      "learning_rate": 1.4530332426530086e-06,
+      "loss": 0.6678,
+      "step": 3741
+    },
+    {
+      "epoch": 0.9097982008266472,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.452556382444385e-06,
+      "loss": 1.1911,
+      "step": 3742
+    },
+    {
+      "epoch": 0.9100413323608072,
+      "grad_norm": 19.75,
+      "learning_rate": 1.4520794919602125e-06,
+      "loss": 1.2636,
+      "step": 3743
+    },
+    {
+      "epoch": 0.9102844638949672,
+      "grad_norm": 24.125,
+      "learning_rate": 1.4516025712717692e-06,
+      "loss": 0.9072,
+      "step": 3744
+    },
+    {
+      "epoch": 0.9105275954291272,
+      "grad_norm": 24.25,
+      "learning_rate": 1.4511256204503403e-06,
+      "loss": 0.6314,
+      "step": 3745
+    },
+    {
+      "epoch": 0.9107707269632871,
+      "grad_norm": 20.625,
+      "learning_rate": 1.4506486395672134e-06,
+      "loss": 0.6908,
+      "step": 3746
+    },
+    {
+      "epoch": 0.9110138584974471,
+      "grad_norm": 22.375,
+      "learning_rate": 1.4501716286936824e-06,
+      "loss": 1.0111,
+      "step": 3747
+    },
+    {
+      "epoch": 0.9112569900316071,
+      "grad_norm": 19.125,
+      "learning_rate": 1.449694587901044e-06,
+      "loss": 0.7819,
+      "step": 3748
+    },
+    {
+      "epoch": 0.9115001215657671,
+      "grad_norm": 21.0,
+      "learning_rate": 1.4492175172606006e-06,
+      "loss": 0.5429,
+      "step": 3749
+    },
+    {
+      "epoch": 0.9117432530999271,
+      "grad_norm": 21.75,
+      "learning_rate": 1.4487404168436593e-06,
+      "loss": 1.2505,
+      "step": 3750
+    },
+    {
+      "epoch": 0.911986384634087,
+      "grad_norm": 22.875,
+      "learning_rate": 1.44826328672153e-06,
+      "loss": 0.6566,
+      "step": 3751
+    },
+    {
+      "epoch": 0.912229516168247,
+      "grad_norm": 20.125,
+      "learning_rate": 1.447786126965528e-06,
+      "loss": 1.0915,
+      "step": 3752
+    },
+    {
+      "epoch": 0.912472647702407,
+      "grad_norm": 21.75,
+      "learning_rate": 1.4473089376469737e-06,
+      "loss": 1.2058,
+      "step": 3753
+    },
+    {
+      "epoch": 0.912715779236567,
+      "grad_norm": 17.375,
+      "learning_rate": 1.446831718837191e-06,
+      "loss": 0.8974,
+      "step": 3754
+    },
+    {
+      "epoch": 0.912958910770727,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.4463544706075088e-06,
+      "loss": 0.5681,
+      "step": 3755
+    },
+    {
+      "epoch": 0.913202042304887,
+      "grad_norm": 23.375,
+      "learning_rate": 1.4458771930292592e-06,
+      "loss": 0.84,
+      "step": 3756
+    },
+    {
+      "epoch": 0.9134451738390469,
+      "grad_norm": 22.875,
+      "learning_rate": 1.4453998861737808e-06,
+      "loss": 1.123,
+      "step": 3757
+    },
+    {
+      "epoch": 0.9136883053732069,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4449225501124146e-06,
+      "loss": 1.0275,
+      "step": 3758
+    },
+    {
+      "epoch": 0.9139314369073669,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4444451849165067e-06,
+      "loss": 0.8495,
+      "step": 3759
+    },
+    {
+      "epoch": 0.9141745684415269,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4439677906574076e-06,
+      "loss": 0.754,
+      "step": 3760
+    },
+    {
+      "epoch": 0.9144176999756869,
+      "grad_norm": 22.375,
+      "learning_rate": 1.4434903674064726e-06,
+      "loss": 0.7724,
+      "step": 3761
+    },
+    {
+      "epoch": 0.9146608315098468,
+      "grad_norm": 17.25,
+      "learning_rate": 1.4430129152350605e-06,
+      "loss": 0.496,
+      "step": 3762
+    },
+    {
+      "epoch": 0.9149039630440068,
+      "grad_norm": 24.0,
+      "learning_rate": 1.4425354342145346e-06,
+      "loss": 0.9878,
+      "step": 3763
+    },
+    {
+      "epoch": 0.9151470945781668,
+      "grad_norm": 15.75,
+      "learning_rate": 1.4420579244162627e-06,
+      "loss": 0.8022,
+      "step": 3764
+    },
+    {
+      "epoch": 0.9153902261123268,
+      "grad_norm": 20.75,
+      "learning_rate": 1.4415803859116173e-06,
+      "loss": 0.7064,
+      "step": 3765
+    },
+    {
+      "epoch": 0.9156333576464868,
+      "grad_norm": 18.5,
+      "learning_rate": 1.441102818771974e-06,
+      "loss": 0.8631,
+      "step": 3766
+    },
+    {
+      "epoch": 0.9158764891806467,
+      "grad_norm": 17.375,
+      "learning_rate": 1.4406252230687148e-06,
+      "loss": 1.3142,
+      "step": 3767
+    },
+    {
+      "epoch": 0.9161196207148067,
+      "grad_norm": 24.125,
+      "learning_rate": 1.440147598873223e-06,
+      "loss": 0.9144,
+      "step": 3768
+    },
+    {
+      "epoch": 0.9163627522489667,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4396699462568894e-06,
+      "loss": 0.5567,
+      "step": 3769
+    },
+    {
+      "epoch": 0.9166058837831267,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.439192265291106e-06,
+      "loss": 0.43,
+      "step": 3770
+    },
+    {
+      "epoch": 0.9168490153172867,
+      "grad_norm": 16.5,
+      "learning_rate": 1.4387145560472712e-06,
+      "loss": 1.0951,
+      "step": 3771
+    },
+    {
+      "epoch": 0.9170921468514467,
+      "grad_norm": 20.75,
+      "learning_rate": 1.4382368185967868e-06,
+      "loss": 0.6864,
+      "step": 3772
+    },
+    {
+      "epoch": 0.9173352783856066,
+      "grad_norm": 21.25,
+      "learning_rate": 1.4377590530110591e-06,
+      "loss": 1.1724,
+      "step": 3773
+    },
+    {
+      "epoch": 0.9175784099197666,
+      "grad_norm": 19.75,
+      "learning_rate": 1.4372812593614983e-06,
+      "loss": 1.2391,
+      "step": 3774
+    },
+    {
+      "epoch": 0.9178215414539266,
+      "grad_norm": 18.625,
+      "learning_rate": 1.4368034377195183e-06,
+      "loss": 0.8745,
+      "step": 3775
+    },
+    {
+      "epoch": 0.9180646729880866,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4363255881565389e-06,
+      "loss": 0.8104,
+      "step": 3776
+    },
+    {
+      "epoch": 0.9183078045222466,
+      "grad_norm": 20.625,
+      "learning_rate": 1.435847710743982e-06,
+      "loss": 0.8274,
+      "step": 3777
+    },
+    {
+      "epoch": 0.9185509360564065,
+      "grad_norm": 14.75,
+      "learning_rate": 1.435369805553275e-06,
+      "loss": 0.7391,
+      "step": 3778
+    },
+    {
+      "epoch": 0.9187940675905665,
+      "grad_norm": 19.625,
+      "learning_rate": 1.4348918726558495e-06,
+      "loss": 1.0393,
+      "step": 3779
+    },
+    {
+      "epoch": 0.9190371991247265,
+      "grad_norm": 21.0,
+      "learning_rate": 1.4344139121231402e-06,
+      "loss": 0.9398,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9192803306588865,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.433935924026587e-06,
+      "loss": 0.7216,
+      "step": 3781
+    },
+    {
+      "epoch": 0.9195234621930465,
+      "grad_norm": 18.0,
+      "learning_rate": 1.433457908437633e-06,
+      "loss": 0.9834,
+      "step": 3782
+    },
+    {
+      "epoch": 0.9197665937272064,
+      "grad_norm": 22.25,
+      "learning_rate": 1.432979865427726e-06,
+      "loss": 1.0621,
+      "step": 3783
+    },
+    {
+      "epoch": 0.9200097252613664,
+      "grad_norm": 18.625,
+      "learning_rate": 1.4325017950683182e-06,
+      "loss": 0.7211,
+      "step": 3784
+    },
+    {
+      "epoch": 0.9202528567955264,
+      "grad_norm": 25.875,
+      "learning_rate": 1.4320236974308652e-06,
+      "loss": 1.1432,
+      "step": 3785
+    },
+    {
+      "epoch": 0.9204959883296864,
+      "grad_norm": 21.125,
+      "learning_rate": 1.431545572586827e-06,
+      "loss": 0.7333,
+      "step": 3786
+    },
+    {
+      "epoch": 0.9207391198638464,
+      "grad_norm": 23.625,
+      "learning_rate": 1.4310674206076675e-06,
+      "loss": 1.1414,
+      "step": 3787
+    },
+    {
+      "epoch": 0.9209822513980063,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4305892415648549e-06,
+      "loss": 0.8259,
+      "step": 3788
+    },
+    {
+      "epoch": 0.9212253829321663,
+      "grad_norm": 19.5,
+      "learning_rate": 1.4301110355298612e-06,
+      "loss": 1.0309,
+      "step": 3789
+    },
+    {
+      "epoch": 0.9214685144663263,
+      "grad_norm": 17.125,
+      "learning_rate": 1.4296328025741626e-06,
+      "loss": 0.8037,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9217116460004863,
+      "grad_norm": 16.5,
+      "learning_rate": 1.4291545427692394e-06,
+      "loss": 1.0646,
+      "step": 3791
+    },
+    {
+      "epoch": 0.9219547775346463,
+      "grad_norm": 20.0,
+      "learning_rate": 1.4286762561865756e-06,
+      "loss": 0.8686,
+      "step": 3792
+    },
+    {
+      "epoch": 0.9221979090688063,
+      "grad_norm": 28.125,
+      "learning_rate": 1.4281979428976594e-06,
+      "loss": 1.1802,
+      "step": 3793
+    },
+    {
+      "epoch": 0.9224410406029662,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.4277196029739831e-06,
+      "loss": 0.7507,
+      "step": 3794
+    },
+    {
+      "epoch": 0.9226841721371262,
+      "grad_norm": 18.5,
+      "learning_rate": 1.427241236487043e-06,
+      "loss": 1.3671,
+      "step": 3795
+    },
+    {
+      "epoch": 0.9229273036712862,
+      "grad_norm": 21.25,
+      "learning_rate": 1.4267628435083388e-06,
+      "loss": 0.7674,
+      "step": 3796
+    },
+    {
+      "epoch": 0.9231704352054462,
+      "grad_norm": 22.375,
+      "learning_rate": 1.4262844241093749e-06,
+      "loss": 1.1878,
+      "step": 3797
+    },
+    {
+      "epoch": 0.9234135667396062,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.4258059783616596e-06,
+      "loss": 0.7436,
+      "step": 3798
+    },
+    {
+      "epoch": 0.9236566982737661,
+      "grad_norm": 20.875,
+      "learning_rate": 1.4253275063367038e-06,
+      "loss": 0.6861,
+      "step": 3799
+    },
+    {
+      "epoch": 0.9238998298079261,
+      "grad_norm": 18.625,
+      "learning_rate": 1.4248490081060248e-06,
+      "loss": 1.0555,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9241429613420861,
+      "grad_norm": 15.25,
+      "learning_rate": 1.4243704837411418e-06,
+      "loss": 0.5213,
+      "step": 3801
+    },
+    {
+      "epoch": 0.9243860928762461,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4238919333135778e-06,
+      "loss": 0.7359,
+      "step": 3802
+    },
+    {
+      "epoch": 0.9246292244104061,
+      "grad_norm": 19.625,
+      "learning_rate": 1.423413356894862e-06,
+      "loss": 0.8722,
+      "step": 3803
+    },
+    {
+      "epoch": 0.924872355944566,
+      "grad_norm": 20.5,
+      "learning_rate": 1.4229347545565248e-06,
+      "loss": 0.9464,
+      "step": 3804
+    },
+    {
+      "epoch": 0.925115487478726,
+      "grad_norm": 18.125,
+      "learning_rate": 1.422456126370102e-06,
+      "loss": 0.777,
+      "step": 3805
+    },
+    {
+      "epoch": 0.925358619012886,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4219774724071322e-06,
+      "loss": 0.8119,
+      "step": 3806
+    },
+    {
+      "epoch": 0.925601750547046,
+      "grad_norm": 16.25,
+      "learning_rate": 1.4214987927391594e-06,
+      "loss": 0.8855,
+      "step": 3807
+    },
+    {
+      "epoch": 0.925844882081206,
+      "grad_norm": 18.625,
+      "learning_rate": 1.42102008743773e-06,
+      "loss": 0.804,
+      "step": 3808
+    },
+    {
+      "epoch": 0.926088013615366,
+      "grad_norm": 16.625,
+      "learning_rate": 1.420541356574395e-06,
+      "loss": 0.4214,
+      "step": 3809
+    },
+    {
+      "epoch": 0.9263311451495259,
+      "grad_norm": 22.25,
+      "learning_rate": 1.4200626002207089e-06,
+      "loss": 0.8928,
+      "step": 3810
+    },
+    {
+      "epoch": 0.9265742766836859,
+      "grad_norm": 36.5,
+      "learning_rate": 1.41958381844823e-06,
+      "loss": 1.06,
+      "step": 3811
+    },
+    {
+      "epoch": 0.9268174082178459,
+      "grad_norm": 23.0,
+      "learning_rate": 1.419105011328521e-06,
+      "loss": 1.194,
+      "step": 3812
+    },
+    {
+      "epoch": 0.9270605397520059,
+      "grad_norm": 19.625,
+      "learning_rate": 1.4186261789331471e-06,
+      "loss": 0.8113,
+      "step": 3813
+    },
+    {
+      "epoch": 0.9273036712861659,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.4181473213336783e-06,
+      "loss": 0.7907,
+      "step": 3814
+    },
+    {
+      "epoch": 0.9275468028203258,
+      "grad_norm": 23.125,
+      "learning_rate": 1.4176684386016886e-06,
+      "loss": 0.9357,
+      "step": 3815
+    },
+    {
+      "epoch": 0.9277899343544858,
+      "grad_norm": 24.25,
+      "learning_rate": 1.417189530808755e-06,
+      "loss": 0.8202,
+      "step": 3816
+    },
+    {
+      "epoch": 0.9280330658886458,
+      "grad_norm": 22.375,
+      "learning_rate": 1.416710598026459e-06,
+      "loss": 0.9124,
+      "step": 3817
+    },
+    {
+      "epoch": 0.9282761974228058,
+      "grad_norm": 19.125,
+      "learning_rate": 1.416231640326384e-06,
+      "loss": 0.8411,
+      "step": 3818
+    },
+    {
+      "epoch": 0.9285193289569658,
+      "grad_norm": 16.5,
+      "learning_rate": 1.41575265778012e-06,
+      "loss": 0.5253,
+      "step": 3819
+    },
+    {
+      "epoch": 0.9287624604911257,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4152736504592587e-06,
+      "loss": 0.9743,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9290055920252857,
+      "grad_norm": 18.875,
+      "learning_rate": 1.4147946184353958e-06,
+      "loss": 0.4579,
+      "step": 3821
+    },
+    {
+      "epoch": 0.9292487235594457,
+      "grad_norm": 23.25,
+      "learning_rate": 1.414315561780131e-06,
+      "loss": 0.8227,
+      "step": 3822
+    },
+    {
+      "epoch": 0.9294918550936057,
+      "grad_norm": 18.875,
+      "learning_rate": 1.4138364805650679e-06,
+      "loss": 0.8655,
+      "step": 3823
+    },
+    {
+      "epoch": 0.9297349866277657,
+      "grad_norm": 22.375,
+      "learning_rate": 1.4133573748618135e-06,
+      "loss": 0.9211,
+      "step": 3824
+    },
+    {
+      "epoch": 0.9299781181619255,
+      "grad_norm": 19.0,
+      "learning_rate": 1.4128782447419775e-06,
+      "loss": 0.7798,
+      "step": 3825
+    },
+    {
+      "epoch": 0.9302212496960856,
+      "grad_norm": 19.625,
+      "learning_rate": 1.4123990902771747e-06,
+      "loss": 1.0926,
+      "step": 3826
+    },
+    {
+      "epoch": 0.9304643812302456,
+      "grad_norm": 13.0625,
+      "learning_rate": 1.411919911539024e-06,
+      "loss": 0.392,
+      "step": 3827
+    },
+    {
+      "epoch": 0.9307075127644056,
+      "grad_norm": 36.0,
+      "learning_rate": 1.4114407085991456e-06,
+      "loss": 0.8685,
+      "step": 3828
+    },
+    {
+      "epoch": 0.9309506442985656,
+      "grad_norm": 24.0,
+      "learning_rate": 1.4109614815291648e-06,
+      "loss": 0.9956,
+      "step": 3829
+    },
+    {
+      "epoch": 0.9311937758327256,
+      "grad_norm": 22.75,
+      "learning_rate": 1.410482230400711e-06,
+      "loss": 1.1699,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9314369073668854,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.410002955285416e-06,
+      "loss": 0.5616,
+      "step": 3831
+    },
+    {
+      "epoch": 0.9316800389010454,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4095236562549167e-06,
+      "loss": 0.9048,
+      "step": 3832
+    },
+    {
+      "epoch": 0.9319231704352055,
+      "grad_norm": 25.0,
+      "learning_rate": 1.409044333380851e-06,
+      "loss": 0.8924,
+      "step": 3833
+    },
+    {
+      "epoch": 0.9321663019693655,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4085649867348635e-06,
+      "loss": 0.9992,
+      "step": 3834
+    },
+    {
+      "epoch": 0.9324094335035255,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4080856163886001e-06,
+      "loss": 0.7677,
+      "step": 3835
+    },
+    {
+      "epoch": 0.9326525650376853,
+      "grad_norm": 21.5,
+      "learning_rate": 1.407606222413711e-06,
+      "loss": 0.9993,
+      "step": 3836
+    },
+    {
+      "epoch": 0.9328956965718453,
+      "grad_norm": 18.125,
+      "learning_rate": 1.4071268048818499e-06,
+      "loss": 0.9306,
+      "step": 3837
+    },
+    {
+      "epoch": 0.9331388281060053,
+      "grad_norm": 22.625,
+      "learning_rate": 1.406647363864674e-06,
+      "loss": 1.0934,
+      "step": 3838
+    },
+    {
+      "epoch": 0.9333819596401653,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4061678994338449e-06,
+      "loss": 0.744,
+      "step": 3839
+    },
+    {
+      "epoch": 0.9336250911743253,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4056884116610255e-06,
+      "loss": 1.2328,
+      "step": 3840
+    },
+    {
+      "epoch": 0.9338682227084852,
+      "grad_norm": 21.0,
+      "learning_rate": 1.4052089006178843e-06,
+      "loss": 0.9657,
+      "step": 3841
+    },
+    {
+      "epoch": 0.9341113542426452,
+      "grad_norm": 20.0,
+      "learning_rate": 1.4047293663760922e-06,
+      "loss": 0.7684,
+      "step": 3842
+    },
+    {
+      "epoch": 0.9343544857768052,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.4042498090073243e-06,
+      "loss": 0.5175,
+      "step": 3843
+    },
+    {
+      "epoch": 0.9345976173109652,
+      "grad_norm": 17.375,
+      "learning_rate": 1.4037702285832586e-06,
+      "loss": 0.9267,
+      "step": 3844
+    },
+    {
+      "epoch": 0.9348407488451252,
+      "grad_norm": 13.1875,
+      "learning_rate": 1.403290625175576e-06,
+      "loss": 0.3068,
+      "step": 3845
+    },
+    {
+      "epoch": 0.9350838803792852,
+      "grad_norm": 21.25,
+      "learning_rate": 1.4028109988559624e-06,
+      "loss": 1.2203,
+      "step": 3846
+    },
+    {
+      "epoch": 0.9353270119134451,
+      "grad_norm": 23.75,
+      "learning_rate": 1.4023313496961059e-06,
+      "loss": 1.0322,
+      "step": 3847
+    },
+    {
+      "epoch": 0.9355701434476051,
+      "grad_norm": 18.875,
+      "learning_rate": 1.401851677767698e-06,
+      "loss": 0.6709,
+      "step": 3848
+    },
+    {
+      "epoch": 0.9358132749817651,
+      "grad_norm": 22.875,
+      "learning_rate": 1.4013719831424347e-06,
+      "loss": 0.9684,
+      "step": 3849
+    },
+    {
+      "epoch": 0.9360564065159251,
+      "grad_norm": 20.625,
+      "learning_rate": 1.4008922658920138e-06,
+      "loss": 0.8002,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9362995380500851,
+      "grad_norm": 20.125,
+      "learning_rate": 1.4004125260881383e-06,
+      "loss": 1.0745,
+      "step": 3851
+    },
+    {
+      "epoch": 0.936542669584245,
+      "grad_norm": 12.8125,
+      "learning_rate": 1.3999327638025128e-06,
+      "loss": 0.4899,
+      "step": 3852
+    },
+    {
+      "epoch": 0.936785801118405,
+      "grad_norm": 17.375,
+      "learning_rate": 1.3994529791068461e-06,
+      "loss": 0.5899,
+      "step": 3853
+    },
+    {
+      "epoch": 0.937028932652565,
+      "grad_norm": 19.125,
+      "learning_rate": 1.398973172072851e-06,
+      "loss": 0.694,
+      "step": 3854
+    },
+    {
+      "epoch": 0.937272064186725,
+      "grad_norm": 24.5,
+      "learning_rate": 1.3984933427722419e-06,
+      "loss": 0.9549,
+      "step": 3855
+    },
+    {
+      "epoch": 0.937515195720885,
+      "grad_norm": 19.0,
+      "learning_rate": 1.3980134912767385e-06,
+      "loss": 0.899,
+      "step": 3856
+    },
+    {
+      "epoch": 0.9377583272550449,
+      "grad_norm": 16.5,
+      "learning_rate": 1.3975336176580625e-06,
+      "loss": 0.9902,
+      "step": 3857
+    },
+    {
+      "epoch": 0.9380014587892049,
+      "grad_norm": 21.25,
+      "learning_rate": 1.3970537219879395e-06,
+      "loss": 0.5704,
+      "step": 3858
+    },
+    {
+      "epoch": 0.9382445903233649,
+      "grad_norm": 17.25,
+      "learning_rate": 1.396573804338098e-06,
+      "loss": 0.6479,
+      "step": 3859
+    },
+    {
+      "epoch": 0.9384877218575249,
+      "grad_norm": 16.625,
+      "learning_rate": 1.3960938647802699e-06,
+      "loss": 1.0088,
+      "step": 3860
+    },
+    {
+      "epoch": 0.9387308533916849,
+      "grad_norm": 16.875,
+      "learning_rate": 1.3956139033861904e-06,
+      "loss": 0.8491,
+      "step": 3861
+    },
+    {
+      "epoch": 0.9389739849258448,
+      "grad_norm": 17.0,
+      "learning_rate": 1.3951339202275984e-06,
+      "loss": 0.7275,
+      "step": 3862
+    },
+    {
+      "epoch": 0.9392171164600048,
+      "grad_norm": 16.625,
+      "learning_rate": 1.394653915376236e-06,
+      "loss": 0.8649,
+      "step": 3863
+    },
+    {
+      "epoch": 0.9394602479941648,
+      "grad_norm": 19.125,
+      "learning_rate": 1.3941738889038467e-06,
+      "loss": 0.7899,
+      "step": 3864
+    },
+    {
+      "epoch": 0.9397033795283248,
+      "grad_norm": 24.875,
+      "learning_rate": 1.3936938408821804e-06,
+      "loss": 1.1052,
+      "step": 3865
+    },
+    {
+      "epoch": 0.9399465110624848,
+      "grad_norm": 16.125,
+      "learning_rate": 1.3932137713829878e-06,
+      "loss": 0.974,
+      "step": 3866
+    },
+    {
+      "epoch": 0.9401896425966448,
+      "grad_norm": 16.25,
+      "learning_rate": 1.3927336804780235e-06,
+      "loss": 0.5647,
+      "step": 3867
+    },
+    {
+      "epoch": 0.9404327741308047,
+      "grad_norm": 15.125,
+      "learning_rate": 1.3922535682390453e-06,
+      "loss": 0.678,
+      "step": 3868
+    },
+    {
+      "epoch": 0.9406759056649647,
+      "grad_norm": 20.625,
+      "learning_rate": 1.391773434737815e-06,
+      "loss": 0.7141,
+      "step": 3869
+    },
+    {
+      "epoch": 0.9409190371991247,
+      "grad_norm": 21.5,
+      "learning_rate": 1.3912932800460965e-06,
+      "loss": 0.9183,
+      "step": 3870
+    },
+    {
+      "epoch": 0.9411621687332847,
+      "grad_norm": 15.625,
+      "learning_rate": 1.3908131042356568e-06,
+      "loss": 0.6303,
+      "step": 3871
+    },
+    {
+      "epoch": 0.9414053002674447,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3903329073782668e-06,
+      "loss": 0.7484,
+      "step": 3872
+    },
+    {
+      "epoch": 0.9416484318016046,
+      "grad_norm": 19.625,
+      "learning_rate": 1.3898526895457e-06,
+      "loss": 0.9081,
+      "step": 3873
+    },
+    {
+      "epoch": 0.9418915633357646,
+      "grad_norm": 20.5,
+      "learning_rate": 1.3893724508097334e-06,
+      "loss": 1.2226,
+      "step": 3874
+    },
+    {
+      "epoch": 0.9421346948699246,
+      "grad_norm": 17.75,
+      "learning_rate": 1.3888921912421473e-06,
+      "loss": 0.7871,
+      "step": 3875
+    },
+    {
+      "epoch": 0.9423778264040846,
+      "grad_norm": 16.875,
+      "learning_rate": 1.388411910914724e-06,
+      "loss": 0.8181,
+      "step": 3876
+    },
+    {
+      "epoch": 0.9426209579382446,
+      "grad_norm": 22.625,
+      "learning_rate": 1.3879316098992507e-06,
+      "loss": 0.9265,
+      "step": 3877
+    },
+    {
+      "epoch": 0.9428640894724045,
+      "grad_norm": 15.875,
+      "learning_rate": 1.3874512882675156e-06,
+      "loss": 0.4322,
+      "step": 3878
+    },
+    {
+      "epoch": 0.9431072210065645,
+      "grad_norm": 17.75,
+      "learning_rate": 1.386970946091312e-06,
+      "loss": 0.6555,
+      "step": 3879
+    },
+    {
+      "epoch": 0.9433503525407245,
+      "grad_norm": 20.125,
+      "learning_rate": 1.3864905834424348e-06,
+      "loss": 0.8817,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9435934840748845,
+      "grad_norm": 19.0,
+      "learning_rate": 1.3860102003926827e-06,
+      "loss": 0.8455,
+      "step": 3881
+    },
+    {
+      "epoch": 0.9438366156090445,
+      "grad_norm": 17.125,
+      "learning_rate": 1.3855297970138571e-06,
+      "loss": 0.6743,
+      "step": 3882
+    },
+    {
+      "epoch": 0.9440797471432045,
+      "grad_norm": 21.5,
+      "learning_rate": 1.3850493733777622e-06,
+      "loss": 1.0024,
+      "step": 3883
+    },
+    {
+      "epoch": 0.9443228786773644,
+      "grad_norm": 20.875,
+      "learning_rate": 1.384568929556207e-06,
+      "loss": 0.8406,
+      "step": 3884
+    },
+    {
+      "epoch": 0.9445660102115244,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3840884656210007e-06,
+      "loss": 1.0866,
+      "step": 3885
+    },
+    {
+      "epoch": 0.9448091417456844,
+      "grad_norm": 17.875,
+      "learning_rate": 1.3836079816439575e-06,
+      "loss": 0.8666,
+      "step": 3886
+    },
+    {
+      "epoch": 0.9450522732798444,
+      "grad_norm": 19.625,
+      "learning_rate": 1.3831274776968936e-06,
+      "loss": 0.9899,
+      "step": 3887
+    },
+    {
+      "epoch": 0.9452954048140044,
+      "grad_norm": 21.125,
+      "learning_rate": 1.3826469538516292e-06,
+      "loss": 1.2078,
+      "step": 3888
+    },
+    {
+      "epoch": 0.9455385363481643,
+      "grad_norm": 17.375,
+      "learning_rate": 1.382166410179987e-06,
+      "loss": 0.8049,
+      "step": 3889
+    },
+    {
+      "epoch": 0.9457816678823243,
+      "grad_norm": 24.125,
+      "learning_rate": 1.381685846753792e-06,
+      "loss": 0.9659,
+      "step": 3890
+    },
+    {
+      "epoch": 0.9460247994164843,
+      "grad_norm": 16.25,
+      "learning_rate": 1.3812052636448728e-06,
+      "loss": 0.4699,
+      "step": 3891
+    },
+    {
+      "epoch": 0.9462679309506443,
+      "grad_norm": 17.625,
+      "learning_rate": 1.380724660925061e-06,
+      "loss": 0.905,
+      "step": 3892
+    },
+    {
+      "epoch": 0.9465110624848043,
+      "grad_norm": 19.0,
+      "learning_rate": 1.3802440386661908e-06,
+      "loss": 0.5628,
+      "step": 3893
+    },
+    {
+      "epoch": 0.9467541940189642,
+      "grad_norm": 16.5,
+      "learning_rate": 1.3797633969401e-06,
+      "loss": 0.7007,
+      "step": 3894
+    },
+    {
+      "epoch": 0.9469973255531242,
+      "grad_norm": 19.375,
+      "learning_rate": 1.3792827358186277e-06,
+      "loss": 1.1727,
+      "step": 3895
+    },
+    {
+      "epoch": 0.9472404570872842,
+      "grad_norm": 16.125,
+      "learning_rate": 1.3788020553736186e-06,
+      "loss": 0.7185,
+      "step": 3896
+    },
+    {
+      "epoch": 0.9474835886214442,
+      "grad_norm": 14.4375,
+      "learning_rate": 1.3783213556769177e-06,
+      "loss": 0.4945,
+      "step": 3897
+    },
+    {
+      "epoch": 0.9477267201556042,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3778406368003735e-06,
+      "loss": 0.5671,
+      "step": 3898
+    },
+    {
+      "epoch": 0.9479698516897641,
+      "grad_norm": 25.25,
+      "learning_rate": 1.3773598988158386e-06,
+      "loss": 1.0576,
+      "step": 3899
+    },
+    {
+      "epoch": 0.9482129832239241,
+      "grad_norm": 22.5,
+      "learning_rate": 1.3768791417951671e-06,
+      "loss": 0.6711,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9484561147580841,
+      "grad_norm": 20.0,
+      "learning_rate": 1.3763983658102168e-06,
+      "loss": 0.7188,
+      "step": 3901
+    },
+    {
+      "epoch": 0.9486992462922441,
+      "grad_norm": 23.375,
+      "learning_rate": 1.3759175709328476e-06,
+      "loss": 1.1597,
+      "step": 3902
+    },
+    {
+      "epoch": 0.9489423778264041,
+      "grad_norm": 18.625,
+      "learning_rate": 1.3754367572349225e-06,
+      "loss": 0.7496,
+      "step": 3903
+    },
+    {
+      "epoch": 0.9491855093605641,
+      "grad_norm": 20.875,
+      "learning_rate": 1.374955924788308e-06,
+      "loss": 0.664,
+      "step": 3904
+    },
+    {
+      "epoch": 0.949428640894724,
+      "grad_norm": 17.25,
+      "learning_rate": 1.3744750736648724e-06,
+      "loss": 0.7426,
+      "step": 3905
+    },
+    {
+      "epoch": 0.949671772428884,
+      "grad_norm": 15.875,
+      "learning_rate": 1.3739942039364876e-06,
+      "loss": 0.5051,
+      "step": 3906
+    },
+    {
+      "epoch": 0.949914903963044,
+      "grad_norm": 22.625,
+      "learning_rate": 1.3735133156750268e-06,
+      "loss": 1.3829,
+      "step": 3907
+    },
+    {
+      "epoch": 0.950158035497204,
+      "grad_norm": 23.125,
+      "learning_rate": 1.3730324089523683e-06,
+      "loss": 0.8434,
+      "step": 3908
+    },
+    {
+      "epoch": 0.950401167031364,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3725514838403914e-06,
+      "loss": 0.5742,
+      "step": 3909
+    },
+    {
+      "epoch": 0.9506442985655239,
+      "grad_norm": 23.0,
+      "learning_rate": 1.3720705404109787e-06,
+      "loss": 1.0885,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9508874300996839,
+      "grad_norm": 25.375,
+      "learning_rate": 1.3715895787360155e-06,
+      "loss": 1.2471,
+      "step": 3911
+    },
+    {
+      "epoch": 0.9511305616338439,
+      "grad_norm": 19.25,
+      "learning_rate": 1.3711085988873898e-06,
+      "loss": 0.6364,
+      "step": 3912
+    },
+    {
+      "epoch": 0.9513736931680039,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3706276009369925e-06,
+      "loss": 0.5563,
+      "step": 3913
+    },
+    {
+      "epoch": 0.9516168247021639,
+      "grad_norm": 17.75,
+      "learning_rate": 1.3701465849567167e-06,
+      "loss": 0.7277,
+      "step": 3914
+    },
+    {
+      "epoch": 0.9518599562363238,
+      "grad_norm": 30.875,
+      "learning_rate": 1.3696655510184592e-06,
+      "loss": 1.1632,
+      "step": 3915
+    },
+    {
+      "epoch": 0.9521030877704838,
+      "grad_norm": 16.375,
+      "learning_rate": 1.369184499194118e-06,
+      "loss": 0.7922,
+      "step": 3916
+    },
+    {
+      "epoch": 0.9523462193046438,
+      "grad_norm": 22.125,
+      "learning_rate": 1.3687034295555951e-06,
+      "loss": 1.055,
+      "step": 3917
+    },
+    {
+      "epoch": 0.9525893508388038,
+      "grad_norm": 21.375,
+      "learning_rate": 1.3682223421747948e-06,
+      "loss": 0.8957,
+      "step": 3918
+    },
+    {
+      "epoch": 0.9528324823729638,
+      "grad_norm": 27.625,
+      "learning_rate": 1.3677412371236232e-06,
+      "loss": 0.9606,
+      "step": 3919
+    },
+    {
+      "epoch": 0.9530756139071238,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.367260114473991e-06,
+      "loss": 0.4122,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9533187454412837,
+      "grad_norm": 17.0,
+      "learning_rate": 1.3667789742978089e-06,
+      "loss": 0.7189,
+      "step": 3921
+    },
+    {
+      "epoch": 0.9535618769754437,
+      "grad_norm": 16.75,
+      "learning_rate": 1.3662978166669924e-06,
+      "loss": 0.6071,
+      "step": 3922
+    },
+    {
+      "epoch": 0.9538050085096037,
+      "grad_norm": 22.0,
+      "learning_rate": 1.3658166416534588e-06,
+      "loss": 0.8416,
+      "step": 3923
+    },
+    {
+      "epoch": 0.9540481400437637,
+      "grad_norm": 20.25,
+      "learning_rate": 1.3653354493291276e-06,
+      "loss": 0.9223,
+      "step": 3924
+    },
+    {
+      "epoch": 0.9542912715779237,
+      "grad_norm": 14.75,
+      "learning_rate": 1.364854239765922e-06,
+      "loss": 0.494,
+      "step": 3925
+    },
+    {
+      "epoch": 0.9545344031120836,
+      "grad_norm": 21.875,
+      "learning_rate": 1.3643730130357662e-06,
+      "loss": 0.9353,
+      "step": 3926
+    },
+    {
+      "epoch": 0.9547775346462436,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3638917692105888e-06,
+      "loss": 0.7108,
+      "step": 3927
+    },
+    {
+      "epoch": 0.9550206661804036,
+      "grad_norm": 22.375,
+      "learning_rate": 1.3634105083623191e-06,
+      "loss": 0.9413,
+      "step": 3928
+    },
+    {
+      "epoch": 0.9552637977145636,
+      "grad_norm": 34.0,
+      "learning_rate": 1.3629292305628905e-06,
+      "loss": 0.3229,
+      "step": 3929
+    },
+    {
+      "epoch": 0.9555069292487236,
+      "grad_norm": 18.5,
+      "learning_rate": 1.362447935884238e-06,
+      "loss": 1.3861,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9557500607828835,
+      "grad_norm": 19.375,
+      "learning_rate": 1.3619666243982993e-06,
+      "loss": 0.6782,
+      "step": 3931
+    },
+    {
+      "epoch": 0.9559931923170435,
+      "grad_norm": 16.875,
+      "learning_rate": 1.361485296177015e-06,
+      "loss": 1.1356,
+      "step": 3932
+    },
+    {
+      "epoch": 0.9562363238512035,
+      "grad_norm": 18.75,
+      "learning_rate": 1.3610039512923278e-06,
+      "loss": 0.9059,
+      "step": 3933
+    },
+    {
+      "epoch": 0.9564794553853635,
+      "grad_norm": 18.25,
+      "learning_rate": 1.3605225898161828e-06,
+      "loss": 0.7756,
+      "step": 3934
+    },
+    {
+      "epoch": 0.9567225869195235,
+      "grad_norm": 17.125,
+      "learning_rate": 1.360041211820528e-06,
+      "loss": 0.6717,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9569657184536834,
+      "grad_norm": 16.0,
+      "learning_rate": 1.3595598173773137e-06,
+      "loss": 0.621,
+      "step": 3936
+    },
+    {
+      "epoch": 0.9572088499878434,
+      "grad_norm": 36.25,
+      "learning_rate": 1.3590784065584927e-06,
+      "loss": 1.2804,
+      "step": 3937
+    },
+    {
+      "epoch": 0.9574519815220034,
+      "grad_norm": 21.5,
+      "learning_rate": 1.3585969794360197e-06,
+      "loss": 0.6779,
+      "step": 3938
+    },
+    {
+      "epoch": 0.9576951130561634,
+      "grad_norm": 19.375,
+      "learning_rate": 1.3581155360818526e-06,
+      "loss": 1.0433,
+      "step": 3939
+    },
+    {
+      "epoch": 0.9579382445903234,
+      "grad_norm": 20.5,
+      "learning_rate": 1.3576340765679516e-06,
+      "loss": 1.1591,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9581813761244834,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.3571526009662784e-06,
+      "loss": 0.6295,
+      "step": 3941
+    },
+    {
+      "epoch": 0.9584245076586433,
+      "grad_norm": 15.625,
+      "learning_rate": 1.356671109348799e-06,
+      "loss": 0.7017,
+      "step": 3942
+    },
+    {
+      "epoch": 0.9586676391928033,
+      "grad_norm": 23.875,
+      "learning_rate": 1.3561896017874799e-06,
+      "loss": 0.8728,
+      "step": 3943
+    },
+    {
+      "epoch": 0.9589107707269633,
+      "grad_norm": 41.75,
+      "learning_rate": 1.355708078354291e-06,
+      "loss": 0.951,
+      "step": 3944
+    },
+    {
+      "epoch": 0.9591539022611233,
+      "grad_norm": 23.125,
+      "learning_rate": 1.3552265391212038e-06,
+      "loss": 1.0707,
+      "step": 3945
+    },
+    {
+      "epoch": 0.9593970337952833,
+      "grad_norm": 19.625,
+      "learning_rate": 1.3547449841601935e-06,
+      "loss": 1.3283,
+      "step": 3946
+    },
+    {
+      "epoch": 0.9596401653294432,
+      "grad_norm": 21.0,
+      "learning_rate": 1.354263413543236e-06,
+      "loss": 0.6235,
+      "step": 3947
+    },
+    {
+      "epoch": 0.9598832968636032,
+      "grad_norm": 21.125,
+      "learning_rate": 1.3537818273423103e-06,
+      "loss": 0.6245,
+      "step": 3948
+    },
+    {
+      "epoch": 0.9601264283977632,
+      "grad_norm": 25.5,
+      "learning_rate": 1.3533002256293987e-06,
+      "loss": 0.8943,
+      "step": 3949
+    },
+    {
+      "epoch": 0.9603695599319232,
+      "grad_norm": 19.75,
+      "learning_rate": 1.352818608476484e-06,
+      "loss": 0.9066,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9606126914660832,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3523369759555526e-06,
+      "loss": 0.6345,
+      "step": 3951
+    },
+    {
+      "epoch": 0.9608558230002431,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3518553281385929e-06,
+      "loss": 1.0054,
+      "step": 3952
+    },
+    {
+      "epoch": 0.9610989545344031,
+      "grad_norm": 18.625,
+      "learning_rate": 1.3513736650975947e-06,
+      "loss": 0.3696,
+      "step": 3953
+    },
+    {
+      "epoch": 0.9613420860685631,
+      "grad_norm": 21.25,
+      "learning_rate": 1.3508919869045522e-06,
+      "loss": 0.9926,
+      "step": 3954
+    },
+    {
+      "epoch": 0.9615852176027231,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3504102936314594e-06,
+      "loss": 1.0307,
+      "step": 3955
+    },
+    {
+      "epoch": 0.9618283491368831,
+      "grad_norm": 25.75,
+      "learning_rate": 1.3499285853503146e-06,
+      "loss": 1.0777,
+      "step": 3956
+    },
+    {
+      "epoch": 0.9620714806710431,
+      "grad_norm": 17.25,
+      "learning_rate": 1.349446862133116e-06,
+      "loss": 0.9713,
+      "step": 3957
+    },
+    {
+      "epoch": 0.962314612205203,
+      "grad_norm": 19.75,
+      "learning_rate": 1.348965124051867e-06,
+      "loss": 0.9016,
+      "step": 3958
+    },
+    {
+      "epoch": 0.962557743739363,
+      "grad_norm": 19.375,
+      "learning_rate": 1.348483371178571e-06,
+      "loss": 0.6962,
+      "step": 3959
+    },
+    {
+      "epoch": 0.962800875273523,
+      "grad_norm": 22.375,
+      "learning_rate": 1.3480016035852342e-06,
+      "loss": 1.2241,
+      "step": 3960
+    },
+    {
+      "epoch": 0.963044006807683,
+      "grad_norm": 18.125,
+      "learning_rate": 1.3475198213438651e-06,
+      "loss": 0.7728,
+      "step": 3961
+    },
+    {
+      "epoch": 0.963287138341843,
+      "grad_norm": 21.375,
+      "learning_rate": 1.3470380245264744e-06,
+      "loss": 0.9239,
+      "step": 3962
+    },
+    {
+      "epoch": 0.9635302698760029,
+      "grad_norm": 26.75,
+      "learning_rate": 1.3465562132050752e-06,
+      "loss": 0.9932,
+      "step": 3963
+    },
+    {
+      "epoch": 0.9637734014101629,
+      "grad_norm": 21.125,
+      "learning_rate": 1.3460743874516823e-06,
+      "loss": 0.8634,
+      "step": 3964
+    },
+    {
+      "epoch": 0.9640165329443229,
+      "grad_norm": 22.375,
+      "learning_rate": 1.3455925473383128e-06,
+      "loss": 1.2086,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9642596644784829,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3451106929369864e-06,
+      "loss": 0.6385,
+      "step": 3966
+    },
+    {
+      "epoch": 0.9645027960126429,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.3446288243197242e-06,
+      "loss": 0.7051,
+      "step": 3967
+    },
+    {
+      "epoch": 0.9647459275468028,
+      "grad_norm": 24.875,
+      "learning_rate": 1.3441469415585501e-06,
+      "loss": 1.1945,
+      "step": 3968
+    },
+    {
+      "epoch": 0.9649890590809628,
+      "grad_norm": 19.5,
+      "learning_rate": 1.3436650447254892e-06,
+      "loss": 0.7287,
+      "step": 3969
+    },
+    {
+      "epoch": 0.9652321906151228,
+      "grad_norm": 16.875,
+      "learning_rate": 1.3431831338925699e-06,
+      "loss": 0.6114,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9654753221492828,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.3427012091318224e-06,
+      "loss": 0.8383,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9657184536834428,
+      "grad_norm": 17.875,
+      "learning_rate": 1.3422192705152773e-06,
+      "loss": 0.697,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9659615852176027,
+      "grad_norm": 26.0,
+      "learning_rate": 1.3417373181149704e-06,
+      "loss": 0.8018,
+      "step": 3973
+    },
+    {
+      "epoch": 0.9662047167517627,
+      "grad_norm": 26.875,
+      "learning_rate": 1.3412553520029365e-06,
+      "loss": 1.0745,
+      "step": 3974
+    },
+    {
+      "epoch": 0.9664478482859227,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3407733722512144e-06,
+      "loss": 1.014,
+      "step": 3975
+    },
+    {
+      "epoch": 0.9666909798200827,
+      "grad_norm": 29.125,
+      "learning_rate": 1.3402913789318436e-06,
+      "loss": 1.9196,
+      "step": 3976
+    },
+    {
+      "epoch": 0.9669341113542427,
+      "grad_norm": 25.875,
+      "learning_rate": 1.3398093721168672e-06,
+      "loss": 0.949,
+      "step": 3977
+    },
+    {
+      "epoch": 0.9671772428884027,
+      "grad_norm": 23.5,
+      "learning_rate": 1.3393273518783292e-06,
+      "loss": 1.3465,
+      "step": 3978
+    },
+    {
+      "epoch": 0.9674203744225626,
+      "grad_norm": 27.5,
+      "learning_rate": 1.3388453182882757e-06,
+      "loss": 0.9278,
+      "step": 3979
+    },
+    {
+      "epoch": 0.9676635059567226,
+      "grad_norm": 40.5,
+      "learning_rate": 1.3383632714187547e-06,
+      "loss": 1.2588,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9679066374908826,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3378812113418168e-06,
+      "loss": 0.7095,
+      "step": 3981
+    },
+    {
+      "epoch": 0.9681497690250426,
+      "grad_norm": 20.125,
+      "learning_rate": 1.3373991381295142e-06,
+      "loss": 1.1938,
+      "step": 3982
+    },
+    {
+      "epoch": 0.9683929005592026,
+      "grad_norm": 19.875,
+      "learning_rate": 1.3369170518539013e-06,
+      "loss": 0.9127,
+      "step": 3983
+    },
+    {
+      "epoch": 0.9686360320933625,
+      "grad_norm": 15.5,
+      "learning_rate": 1.3364349525870332e-06,
+      "loss": 0.596,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9688791636275225,
+      "grad_norm": 18.125,
+      "learning_rate": 1.3359528404009691e-06,
+      "loss": 0.7775,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9691222951616825,
+      "grad_norm": 24.5,
+      "learning_rate": 1.3354707153677685e-06,
+      "loss": 0.8741,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9693654266958425,
+      "grad_norm": 18.25,
+      "learning_rate": 1.334988577559493e-06,
+      "loss": 0.9123,
+      "step": 3987
+    },
+    {
+      "epoch": 0.9696085582300025,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.3345064270482072e-06,
+      "loss": 0.5815,
+      "step": 3988
+    },
+    {
+      "epoch": 0.9698516897641624,
+      "grad_norm": 16.625,
+      "learning_rate": 1.3340242639059764e-06,
+      "loss": 0.8815,
+      "step": 3989
+    },
+    {
+      "epoch": 0.9700948212983224,
+      "grad_norm": 14.625,
+      "learning_rate": 1.3335420882048683e-06,
+      "loss": 0.9273,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9703379528324824,
+      "grad_norm": 17.125,
+      "learning_rate": 1.3330599000169519e-06,
+      "loss": 0.6381,
+      "step": 3991
+    },
+    {
+      "epoch": 0.9705810843666424,
+      "grad_norm": 36.5,
+      "learning_rate": 1.3325776994142991e-06,
+      "loss": 1.6005,
+      "step": 3992
+    },
+    {
+      "epoch": 0.9708242159008024,
+      "grad_norm": 19.75,
+      "learning_rate": 1.3320954864689831e-06,
+      "loss": 0.9846,
+      "step": 3993
+    },
+    {
+      "epoch": 0.9710673474349624,
+      "grad_norm": 19.625,
+      "learning_rate": 1.3316132612530786e-06,
+      "loss": 0.9429,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9713104789691223,
+      "grad_norm": 22.625,
+      "learning_rate": 1.3311310238386626e-06,
+      "loss": 0.8868,
+      "step": 3995
+    },
+    {
+      "epoch": 0.9715536105032823,
+      "grad_norm": 18.875,
+      "learning_rate": 1.3306487742978142e-06,
+      "loss": 0.6734,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9717967420374423,
+      "grad_norm": 21.75,
+      "learning_rate": 1.3301665127026137e-06,
+      "loss": 1.0677,
+      "step": 3997
+    },
+    {
+      "epoch": 0.9720398735716023,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.329684239125143e-06,
+      "loss": 0.7513,
+      "step": 3998
+    },
+    {
+      "epoch": 0.9722830051057623,
+      "grad_norm": 19.0,
+      "learning_rate": 1.3292019536374866e-06,
+      "loss": 1.0038,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9725261366399222,
+      "grad_norm": 16.875,
+      "learning_rate": 1.3287196563117308e-06,
+      "loss": 0.9442,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9727692681740822,
+      "grad_norm": 27.625,
+      "learning_rate": 1.3282373472199623e-06,
+      "loss": 0.8517,
+      "step": 4001
+    },
+    {
+      "epoch": 0.9730123997082422,
+      "grad_norm": 19.75,
+      "learning_rate": 1.3277550264342714e-06,
+      "loss": 1.0273,
+      "step": 4002
+    },
+    {
+      "epoch": 0.9732555312424022,
+      "grad_norm": 18.125,
+      "learning_rate": 1.3272726940267485e-06,
+      "loss": 0.5411,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9734986627765622,
+      "grad_norm": 19.875,
+      "learning_rate": 1.3267903500694875e-06,
+      "loss": 0.998,
+      "step": 4004
+    },
+    {
+      "epoch": 0.973741794310722,
+      "grad_norm": 24.125,
+      "learning_rate": 1.3263079946345822e-06,
+      "loss": 0.9674,
+      "step": 4005
+    },
+    {
+      "epoch": 0.973984925844882,
+      "grad_norm": 21.875,
+      "learning_rate": 1.3258256277941291e-06,
+      "loss": 1.068,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9742280573790421,
+      "grad_norm": 15.875,
+      "learning_rate": 1.3253432496202267e-06,
+      "loss": 0.667,
+      "step": 4007
+    },
+    {
+      "epoch": 0.9744711889132021,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.3248608601849741e-06,
+      "loss": 0.3742,
+      "step": 4008
+    },
+    {
+      "epoch": 0.9747143204473621,
+      "grad_norm": 20.875,
+      "learning_rate": 1.3243784595604733e-06,
+      "loss": 0.6048,
+      "step": 4009
+    },
+    {
+      "epoch": 0.974957451981522,
+      "grad_norm": 22.75,
+      "learning_rate": 1.323896047818827e-06,
+      "loss": 1.2635,
+      "step": 4010
+    },
+    {
+      "epoch": 0.975200583515682,
+      "grad_norm": 17.5,
+      "learning_rate": 1.3234136250321403e-06,
+      "loss": 0.6713,
+      "step": 4011
+    },
+    {
+      "epoch": 0.975443715049842,
+      "grad_norm": 19.375,
+      "learning_rate": 1.3229311912725193e-06,
+      "loss": 0.5918,
+      "step": 4012
+    },
+    {
+      "epoch": 0.975686846584002,
+      "grad_norm": 16.25,
+      "learning_rate": 1.322448746612072e-06,
+      "loss": 0.7757,
+      "step": 4013
+    },
+    {
+      "epoch": 0.975929978118162,
+      "grad_norm": 20.0,
+      "learning_rate": 1.321966291122909e-06,
+      "loss": 0.7835,
+      "step": 4014
+    },
+    {
+      "epoch": 0.976173109652322,
+      "grad_norm": 17.5,
+      "learning_rate": 1.3214838248771396e-06,
+      "loss": 1.0936,
+      "step": 4015
+    },
+    {
+      "epoch": 0.9764162411864818,
+      "grad_norm": 21.0,
+      "learning_rate": 1.3210013479468791e-06,
+      "loss": 0.9025,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9766593727206418,
+      "grad_norm": 23.5,
+      "learning_rate": 1.3205188604042407e-06,
+      "loss": 1.0179,
+      "step": 4017
+    },
+    {
+      "epoch": 0.9769025042548019,
+      "grad_norm": 19.125,
+      "learning_rate": 1.3200363623213406e-06,
+      "loss": 0.9663,
+      "step": 4018
+    },
+    {
+      "epoch": 0.9771456357889619,
+      "grad_norm": 21.5,
+      "learning_rate": 1.3195538537702965e-06,
+      "loss": 0.9284,
+      "step": 4019
+    },
+    {
+      "epoch": 0.9773887673231219,
+      "grad_norm": 16.5,
+      "learning_rate": 1.319071334823228e-06,
+      "loss": 0.912,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9776318988572817,
+      "grad_norm": 19.625,
+      "learning_rate": 1.3185888055522556e-06,
+      "loss": 0.7986,
+      "step": 4021
+    },
+    {
+      "epoch": 0.9778750303914417,
+      "grad_norm": 24.25,
+      "learning_rate": 1.3181062660295013e-06,
+      "loss": 1.0828,
+      "step": 4022
+    },
+    {
+      "epoch": 0.9781181619256017,
+      "grad_norm": 25.0,
+      "learning_rate": 1.3176237163270893e-06,
+      "loss": 0.8634,
+      "step": 4023
+    },
+    {
+      "epoch": 0.9783612934597617,
+      "grad_norm": 14.75,
+      "learning_rate": 1.3171411565171452e-06,
+      "loss": 0.7959,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9786044249939218,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3166585866717953e-06,
+      "loss": 0.7384,
+      "step": 4025
+    },
+    {
+      "epoch": 0.9788475565280816,
+      "grad_norm": 17.125,
+      "learning_rate": 1.3161760068631691e-06,
+      "loss": 0.7498,
+      "step": 4026
+    },
+    {
+      "epoch": 0.9790906880622416,
+      "grad_norm": 22.0,
+      "learning_rate": 1.315693417163395e-06,
+      "loss": 0.7234,
+      "step": 4027
+    },
+    {
+      "epoch": 0.9793338195964016,
+      "grad_norm": 17.0,
+      "learning_rate": 1.315210817644606e-06,
+      "loss": 0.9886,
+      "step": 4028
+    },
+    {
+      "epoch": 0.9795769511305616,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3147282083789337e-06,
+      "loss": 0.9663,
+      "step": 4029
+    },
+    {
+      "epoch": 0.9798200826647216,
+      "grad_norm": 15.5,
+      "learning_rate": 1.3142455894385125e-06,
+      "loss": 0.4486,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9800632141988816,
+      "grad_norm": 15.75,
+      "learning_rate": 1.3137629608954785e-06,
+      "loss": 0.8873,
+      "step": 4031
+    },
+    {
+      "epoch": 0.9803063457330415,
+      "grad_norm": 22.375,
+      "learning_rate": 1.3132803228219688e-06,
+      "loss": 0.794,
+      "step": 4032
+    },
+    {
+      "epoch": 0.9805494772672015,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.3127976752901222e-06,
+      "loss": 0.6404,
+      "step": 4033
+    },
+    {
+      "epoch": 0.9807926088013615,
+      "grad_norm": 17.125,
+      "learning_rate": 1.312315018372078e-06,
+      "loss": 0.9471,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9810357403355215,
+      "grad_norm": 21.0,
+      "learning_rate": 1.3118323521399787e-06,
+      "loss": 1.1323,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9812788718696815,
+      "grad_norm": 12.6875,
+      "learning_rate": 1.3113496766659661e-06,
+      "loss": 0.3732,
+      "step": 4036
+    },
+    {
+      "epoch": 0.9815220034038414,
+      "grad_norm": 21.125,
+      "learning_rate": 1.3108669920221848e-06,
+      "loss": 0.6477,
+      "step": 4037
+    },
+    {
+      "epoch": 0.9817651349380014,
+      "grad_norm": 18.0,
+      "learning_rate": 1.3103842982807802e-06,
+      "loss": 0.7884,
+      "step": 4038
+    },
+    {
+      "epoch": 0.9820082664721614,
+      "grad_norm": 22.25,
+      "learning_rate": 1.3099015955138997e-06,
+      "loss": 0.7812,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9822513980063214,
+      "grad_norm": 17.25,
+      "learning_rate": 1.3094188837936912e-06,
+      "loss": 0.9805,
+      "step": 4040
+    },
+    {
+      "epoch": 0.9824945295404814,
+      "grad_norm": 18.75,
+      "learning_rate": 1.3089361631923043e-06,
+      "loss": 0.8195,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9827376610746413,
+      "grad_norm": 21.75,
+      "learning_rate": 1.3084534337818896e-06,
+      "loss": 1.1097,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9829807926088013,
+      "grad_norm": 28.125,
+      "learning_rate": 1.3079706956345997e-06,
+      "loss": 1.3946,
+      "step": 4043
+    },
+    {
+      "epoch": 0.9832239241429613,
+      "grad_norm": 23.125,
+      "learning_rate": 1.3074879488225883e-06,
+      "loss": 0.8835,
+      "step": 4044
+    },
+    {
+      "epoch": 0.9834670556771213,
+      "grad_norm": 23.125,
+      "learning_rate": 1.3070051934180106e-06,
+      "loss": 0.9037,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9837101872112813,
+      "grad_norm": 17.0,
+      "learning_rate": 1.3065224294930213e-06,
+      "loss": 0.6628,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9839533187454412,
+      "grad_norm": 18.75,
+      "learning_rate": 1.3060396571197794e-06,
+      "loss": 0.8511,
+      "step": 4047
+    },
+    {
+      "epoch": 0.9841964502796012,
+      "grad_norm": 17.75,
+      "learning_rate": 1.3055568763704425e-06,
+      "loss": 0.9845,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9844395818137612,
+      "grad_norm": 17.5,
+      "learning_rate": 1.3050740873171714e-06,
+      "loss": 0.873,
+      "step": 4049
+    },
+    {
+      "epoch": 0.9846827133479212,
+      "grad_norm": 23.375,
+      "learning_rate": 1.3045912900321264e-06,
+      "loss": 1.1688,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9849258448820812,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3041084845874705e-06,
+      "loss": 0.71,
+      "step": 4051
+    },
+    {
+      "epoch": 0.9851689764162412,
+      "grad_norm": 19.5,
+      "learning_rate": 1.303625671055367e-06,
+      "loss": 0.6832,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9854121079504011,
+      "grad_norm": 18.75,
+      "learning_rate": 1.3031428495079807e-06,
+      "loss": 0.6501,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9856552394845611,
+      "grad_norm": 31.5,
+      "learning_rate": 1.302660020017478e-06,
+      "loss": 0.9338,
+      "step": 4054
+    },
+    {
+      "epoch": 0.9858983710187211,
+      "grad_norm": 24.875,
+      "learning_rate": 1.3021771826560256e-06,
+      "loss": 0.9302,
+      "step": 4055
+    },
+    {
+      "epoch": 0.9861415025528811,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.3016943374957922e-06,
+      "loss": 1.0072,
+      "step": 4056
+    },
+    {
+      "epoch": 0.9863846340870411,
+      "grad_norm": 20.875,
+      "learning_rate": 1.301211484608947e-06,
+      "loss": 0.7759,
+      "step": 4057
+    },
+    {
+      "epoch": 0.986627765621201,
+      "grad_norm": 20.5,
+      "learning_rate": 1.3007286240676614e-06,
+      "loss": 0.9501,
+      "step": 4058
+    },
+    {
+      "epoch": 0.986870897155361,
+      "grad_norm": 12.625,
+      "learning_rate": 1.300245755944107e-06,
+      "loss": 0.356,
+      "step": 4059
+    },
+    {
+      "epoch": 0.987114028689521,
+      "grad_norm": 19.75,
+      "learning_rate": 1.2997628803104563e-06,
+      "loss": 0.8057,
+      "step": 4060
+    },
+    {
+      "epoch": 0.987357160223681,
+      "grad_norm": 29.75,
+      "learning_rate": 1.2992799972388836e-06,
+      "loss": 0.9597,
+      "step": 4061
+    },
+    {
+      "epoch": 0.987600291757841,
+      "grad_norm": 18.75,
+      "learning_rate": 1.2987971068015643e-06,
+      "loss": 0.9781,
+      "step": 4062
+    },
+    {
+      "epoch": 0.9878434232920009,
+      "grad_norm": 14.625,
+      "learning_rate": 1.2983142090706744e-06,
+      "loss": 0.5611,
+      "step": 4063
+    },
+    {
+      "epoch": 0.9880865548261609,
+      "grad_norm": 22.125,
+      "learning_rate": 1.297831304118392e-06,
+      "loss": 1.1952,
+      "step": 4064
+    },
+    {
+      "epoch": 0.9883296863603209,
+      "grad_norm": 25.75,
+      "learning_rate": 1.2973483920168948e-06,
+      "loss": 1.1019,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9885728178944809,
+      "grad_norm": 18.375,
+      "learning_rate": 1.2968654728383629e-06,
+      "loss": 0.9886,
+      "step": 4066
+    },
+    {
+      "epoch": 0.9888159494286409,
+      "grad_norm": 16.75,
+      "learning_rate": 1.2963825466549765e-06,
+      "loss": 0.7126,
+      "step": 4067
+    },
+    {
+      "epoch": 0.9890590809628009,
+      "grad_norm": 18.875,
+      "learning_rate": 1.2958996135389174e-06,
+      "loss": 0.6975,
+      "step": 4068
+    },
+    {
+      "epoch": 0.9893022124969608,
+      "grad_norm": 28.125,
+      "learning_rate": 1.2954166735623682e-06,
+      "loss": 1.2519,
+      "step": 4069
+    },
+    {
+      "epoch": 0.9895453440311208,
+      "grad_norm": 16.625,
+      "learning_rate": 1.294933726797513e-06,
+      "loss": 0.7895,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9897884755652808,
+      "grad_norm": 14.5,
+      "learning_rate": 1.2944507733165367e-06,
+      "loss": 0.8475,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9900316070994408,
+      "grad_norm": 63.5,
+      "learning_rate": 1.293967813191624e-06,
+      "loss": 1.0635,
+      "step": 4072
+    },
+    {
+      "epoch": 0.9902747386336008,
+      "grad_norm": 20.875,
+      "learning_rate": 1.2934848464949625e-06,
+      "loss": 0.9305,
+      "step": 4073
+    },
+    {
+      "epoch": 0.9905178701677607,
+      "grad_norm": 20.25,
+      "learning_rate": 1.29300187329874e-06,
+      "loss": 0.7422,
+      "step": 4074
+    },
+    {
+      "epoch": 0.9907610017019207,
+      "grad_norm": 17.5,
+      "learning_rate": 1.2925188936751443e-06,
+      "loss": 1.0064,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9910041332360807,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2920359076963663e-06,
+      "loss": 0.786,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9912472647702407,
+      "grad_norm": 25.125,
+      "learning_rate": 1.291552915434595e-06,
+      "loss": 0.7397,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9914903963044007,
+      "grad_norm": 17.375,
+      "learning_rate": 1.2910699169620235e-06,
+      "loss": 0.6564,
+      "step": 4078
+    },
+    {
+      "epoch": 0.9917335278385606,
+      "grad_norm": 20.375,
+      "learning_rate": 1.2905869123508435e-06,
+      "loss": 0.6919,
+      "step": 4079
+    },
+    {
+      "epoch": 0.9919766593727206,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.290103901673248e-06,
+      "loss": 0.3727,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9922197909068806,
+      "grad_norm": 16.75,
+      "learning_rate": 1.2896208850014325e-06,
+      "loss": 0.8732,
+      "step": 4081
+    },
+    {
+      "epoch": 0.9924629224410406,
+      "grad_norm": 19.625,
+      "learning_rate": 1.2891378624075912e-06,
+      "loss": 0.8371,
+      "step": 4082
+    },
+    {
+      "epoch": 0.9927060539752006,
+      "grad_norm": 17.0,
+      "learning_rate": 1.2886548339639205e-06,
+      "loss": 0.646,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9929491855093605,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.288171799742617e-06,
+      "loss": 0.6926,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9931923170435205,
+      "grad_norm": 18.375,
+      "learning_rate": 1.287688759815879e-06,
+      "loss": 0.4883,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9934354485776805,
+      "grad_norm": 22.5,
+      "learning_rate": 1.2872057142559049e-06,
+      "loss": 0.7492,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9936785801118405,
+      "grad_norm": 18.5,
+      "learning_rate": 1.2867226631348943e-06,
+      "loss": 0.9716,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9939217116460005,
+      "grad_norm": 20.0,
+      "learning_rate": 1.2862396065250473e-06,
+      "loss": 1.0586,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9941648431801605,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.285756544498565e-06,
+      "loss": 0.5721,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9944079747143204,
+      "grad_norm": 22.375,
+      "learning_rate": 1.2852734771276504e-06,
+      "loss": 0.9043,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9946511062484804,
+      "grad_norm": 15.4375,
+      "learning_rate": 1.284790404484505e-06,
+      "loss": 0.4759,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9948942377826404,
+      "grad_norm": 21.0,
+      "learning_rate": 1.2843073266413323e-06,
+      "loss": 1.0459,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9951373693168004,
+      "grad_norm": 20.875,
+      "learning_rate": 1.2838242436703377e-06,
+      "loss": 1.2038,
+      "step": 4093
+    },
+    {
+      "epoch": 0.9953805008509604,
+      "grad_norm": 23.375,
+      "learning_rate": 1.2833411556437255e-06,
+      "loss": 0.9746,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9956236323851203,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.2828580626337024e-06,
+      "loss": 0.5975,
+      "step": 4095
+    },
+    {
+      "epoch": 0.9958667639192803,
+      "grad_norm": 19.875,
+      "learning_rate": 1.2823749647124733e-06,
+      "loss": 0.7218,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9961098954534403,
+      "grad_norm": 16.875,
+      "learning_rate": 1.2818918619522471e-06,
+      "loss": 0.5233,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9963530269876003,
+      "grad_norm": 21.5,
+      "learning_rate": 1.2814087544252316e-06,
+      "loss": 0.8869,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9965961585217603,
+      "grad_norm": 27.875,
+      "learning_rate": 1.2809256422036351e-06,
+      "loss": 0.9395,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9968392900559202,
+      "grad_norm": 20.375,
+      "learning_rate": 1.2804425253596672e-06,
+      "loss": 0.9283,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9970824215900802,
+      "grad_norm": 18.0,
+      "learning_rate": 1.279959403965538e-06,
+      "loss": 0.5885,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9973255531242402,
+      "grad_norm": 21.125,
+      "learning_rate": 1.2794762780934588e-06,
+      "loss": 0.929,
+      "step": 4102
+    },
+    {
+      "epoch": 0.9975686846584002,
+      "grad_norm": 16.75,
+      "learning_rate": 1.2789931478156406e-06,
+      "loss": 0.8388,
+      "step": 4103
+    },
+    {
+      "epoch": 0.9978118161925602,
+      "grad_norm": 16.75,
+      "learning_rate": 1.2785100132042954e-06,
+      "loss": 0.6561,
+      "step": 4104
+    },
+    {
+      "epoch": 0.9980549477267202,
+      "grad_norm": 18.125,
+      "learning_rate": 1.2780268743316369e-06,
+      "loss": 0.4373,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9982980792608801,
+      "grad_norm": 25.625,
+      "learning_rate": 1.2775437312698776e-06,
+      "loss": 1.1276,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9985412107950401,
+      "grad_norm": 23.25,
+      "learning_rate": 1.277060584091232e-06,
+      "loss": 1.1631,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9987843423292001,
+      "grad_norm": 13.0,
+      "learning_rate": 1.2765774328679147e-06,
+      "loss": 0.3814,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9990274738633601,
+      "grad_norm": 21.875,
+      "learning_rate": 1.2760942776721414e-06,
+      "loss": 1.1216,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9992706053975201,
+      "grad_norm": 15.75,
+      "learning_rate": 1.2756111185761277e-06,
+      "loss": 0.5756,
+      "step": 4110
+    },
+    {
+      "epoch": 0.99951373693168,
+      "grad_norm": 17.75,
+      "learning_rate": 1.2751279556520893e-06,
+      "loss": 0.7563,
+      "step": 4111
+    },
+    {
+      "epoch": 0.99975686846584,
+      "grad_norm": 16.375,
+      "learning_rate": 1.2746447889722446e-06,
+      "loss": 0.8834,
+      "step": 4112
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2741616186088103e-06,
+      "loss": 0.9585,
+      "step": 4113
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.1157740354537964,
+      "eval_runtime": 98.4299,
+      "eval_samples_per_second": 5.689,
+      "eval_steps_per_second": 5.689,
+      "step": 4113
+    },
+    {
+      "epoch": 1.00024313153416,
+      "grad_norm": 16.375,
+      "learning_rate": 1.273678444634005e-06,
+      "loss": 0.6687,
+      "step": 4114
+    },
+    {
+      "epoch": 1.00048626306832,
+      "grad_norm": 20.625,
+      "learning_rate": 1.273195267120047e-06,
+      "loss": 0.6081,
+      "step": 4115
+    },
+    {
+      "epoch": 1.00072939460248,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.272712086139156e-06,
+      "loss": 0.9411,
+      "step": 4116
+    },
+    {
+      "epoch": 1.00097252613664,
+      "grad_norm": 16.875,
+      "learning_rate": 1.2722289017635515e-06,
+      "loss": 0.4011,
+      "step": 4117
+    },
+    {
+      "epoch": 1.0012156576708,
+      "grad_norm": 21.5,
+      "learning_rate": 1.2717457140654533e-06,
+      "loss": 1.1048,
+      "step": 4118
+    },
+    {
+      "epoch": 1.0014587892049598,
+      "grad_norm": 21.625,
+      "learning_rate": 1.2712625231170827e-06,
+      "loss": 0.9995,
+      "step": 4119
+    },
+    {
+      "epoch": 1.0017019207391198,
+      "grad_norm": 20.0,
+      "learning_rate": 1.2707793289906609e-06,
+      "loss": 0.7247,
+      "step": 4120
+    },
+    {
+      "epoch": 1.0019450522732798,
+      "grad_norm": 16.375,
+      "learning_rate": 1.2702961317584096e-06,
+      "loss": 0.7889,
+      "step": 4121
+    },
+    {
+      "epoch": 1.0021881838074398,
+      "grad_norm": 20.0,
+      "learning_rate": 1.2698129314925508e-06,
+      "loss": 0.7779,
+      "step": 4122
+    },
+    {
+      "epoch": 1.0024313153415998,
+      "grad_norm": 17.125,
+      "learning_rate": 1.269329728265307e-06,
+      "loss": 0.5287,
+      "step": 4123
+    },
+    {
+      "epoch": 1.0026744468757598,
+      "grad_norm": 24.625,
+      "learning_rate": 1.2688465221489018e-06,
+      "loss": 1.009,
+      "step": 4124
+    },
+    {
+      "epoch": 1.0029175784099198,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.2683633132155582e-06,
+      "loss": 0.6406,
+      "step": 4125
+    },
+    {
+      "epoch": 1.0031607099440798,
+      "grad_norm": 17.125,
+      "learning_rate": 1.2678801015375003e-06,
+      "loss": 0.7235,
+      "step": 4126
+    },
+    {
+      "epoch": 1.0034038414782398,
+      "grad_norm": 15.75,
+      "learning_rate": 1.2673968871869524e-06,
+      "loss": 0.5363,
+      "step": 4127
+    },
+    {
+      "epoch": 1.0036469730123998,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.2669136702361396e-06,
+      "loss": 0.5582,
+      "step": 4128
+    },
+    {
+      "epoch": 1.0038901045465596,
+      "grad_norm": 21.0,
+      "learning_rate": 1.2664304507572864e-06,
+      "loss": 0.722,
+      "step": 4129
+    },
+    {
+      "epoch": 1.0041332360807196,
+      "grad_norm": 17.25,
+      "learning_rate": 1.2659472288226185e-06,
+      "loss": 0.5628,
+      "step": 4130
+    },
+    {
+      "epoch": 1.0043763676148796,
+      "grad_norm": 15.125,
+      "learning_rate": 1.265464004504362e-06,
+      "loss": 0.4223,
+      "step": 4131
+    },
+    {
+      "epoch": 1.0046194991490396,
+      "grad_norm": 20.25,
+      "learning_rate": 1.2649807778747428e-06,
+      "loss": 0.5436,
+      "step": 4132
+    },
+    {
+      "epoch": 1.0048626306831996,
+      "grad_norm": 22.75,
+      "learning_rate": 1.2644975490059875e-06,
+      "loss": 0.5505,
+      "step": 4133
+    },
+    {
+      "epoch": 1.0051057622173596,
+      "grad_norm": 19.25,
+      "learning_rate": 1.2640143179703235e-06,
+      "loss": 0.8603,
+      "step": 4134
+    },
+    {
+      "epoch": 1.0053488937515196,
+      "grad_norm": 20.75,
+      "learning_rate": 1.2635310848399773e-06,
+      "loss": 0.714,
+      "step": 4135
+    },
+    {
+      "epoch": 1.0055920252856796,
+      "grad_norm": 21.0,
+      "learning_rate": 1.2630478496871771e-06,
+      "loss": 0.6515,
+      "step": 4136
+    },
+    {
+      "epoch": 1.0058351568198396,
+      "grad_norm": 50.0,
+      "learning_rate": 1.26256461258415e-06,
+      "loss": 1.0408,
+      "step": 4137
+    },
+    {
+      "epoch": 1.0060782883539996,
+      "grad_norm": 17.0,
+      "learning_rate": 1.2620813736031248e-06,
+      "loss": 0.7504,
+      "step": 4138
+    },
+    {
+      "epoch": 1.0063214198881596,
+      "grad_norm": 21.625,
+      "learning_rate": 1.2615981328163296e-06,
+      "loss": 0.8906,
+      "step": 4139
+    },
+    {
+      "epoch": 1.0065645514223194,
+      "grad_norm": 18.25,
+      "learning_rate": 1.2611148902959932e-06,
+      "loss": 0.8087,
+      "step": 4140
+    },
+    {
+      "epoch": 1.0068076829564794,
+      "grad_norm": 16.125,
+      "learning_rate": 1.2606316461143447e-06,
+      "loss": 0.5794,
+      "step": 4141
+    },
+    {
+      "epoch": 1.0070508144906394,
+      "grad_norm": 29.0,
+      "learning_rate": 1.2601484003436127e-06,
+      "loss": 1.214,
+      "step": 4142
+    },
+    {
+      "epoch": 1.0072939460247994,
+      "grad_norm": 18.75,
+      "learning_rate": 1.2596651530560273e-06,
+      "loss": 0.5319,
+      "step": 4143
+    },
+    {
+      "epoch": 1.0075370775589594,
+      "grad_norm": 16.5,
+      "learning_rate": 1.2591819043238177e-06,
+      "loss": 0.7777,
+      "step": 4144
+    },
+    {
+      "epoch": 1.0077802090931194,
+      "grad_norm": 19.75,
+      "learning_rate": 1.258698654219214e-06,
+      "loss": 0.7306,
+      "step": 4145
+    },
+    {
+      "epoch": 1.0080233406272794,
+      "grad_norm": 21.375,
+      "learning_rate": 1.2582154028144457e-06,
+      "loss": 0.7118,
+      "step": 4146
+    },
+    {
+      "epoch": 1.0082664721614394,
+      "grad_norm": 19.5,
+      "learning_rate": 1.257732150181744e-06,
+      "loss": 0.9558,
+      "step": 4147
+    },
+    {
+      "epoch": 1.0085096036955994,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.2572488963933394e-06,
+      "loss": 0.5373,
+      "step": 4148
+    },
+    {
+      "epoch": 1.0087527352297594,
+      "grad_norm": 21.125,
+      "learning_rate": 1.2567656415214614e-06,
+      "loss": 0.6095,
+      "step": 4149
+    },
+    {
+      "epoch": 1.0089958667639194,
+      "grad_norm": 18.75,
+      "learning_rate": 1.2562823856383415e-06,
+      "loss": 0.7064,
+      "step": 4150
+    },
+    {
+      "epoch": 1.0092389982980792,
+      "grad_norm": 23.5,
+      "learning_rate": 1.2557991288162106e-06,
+      "loss": 0.8255,
+      "step": 4151
+    },
+    {
+      "epoch": 1.0094821298322392,
+      "grad_norm": 30.625,
+      "learning_rate": 1.2553158711272997e-06,
+      "loss": 1.0326,
+      "step": 4152
+    },
+    {
+      "epoch": 1.0097252613663992,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2548326126438403e-06,
+      "loss": 0.6889,
+      "step": 4153
+    },
+    {
+      "epoch": 1.0099683929005592,
+      "grad_norm": 23.25,
+      "learning_rate": 1.2543493534380632e-06,
+      "loss": 0.9978,
+      "step": 4154
+    },
+    {
+      "epoch": 1.0102115244347192,
+      "grad_norm": 19.0,
+      "learning_rate": 1.2538660935822004e-06,
+      "loss": 0.6695,
+      "step": 4155
+    },
+    {
+      "epoch": 1.0104546559688792,
+      "grad_norm": 16.5,
+      "learning_rate": 1.253382833148483e-06,
+      "loss": 0.4427,
+      "step": 4156
+    },
+    {
+      "epoch": 1.0106977875030392,
+      "grad_norm": 13.75,
+      "learning_rate": 1.2528995722091424e-06,
+      "loss": 0.296,
+      "step": 4157
+    },
+    {
+      "epoch": 1.0109409190371992,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.2524163108364113e-06,
+      "loss": 0.385,
+      "step": 4158
+    },
+    {
+      "epoch": 1.0111840505713592,
+      "grad_norm": 19.5,
+      "learning_rate": 1.2519330491025204e-06,
+      "loss": 0.7604,
+      "step": 4159
+    },
+    {
+      "epoch": 1.0114271821055192,
+      "grad_norm": 15.375,
+      "learning_rate": 1.251449787079702e-06,
+      "loss": 0.4404,
+      "step": 4160
+    },
+    {
+      "epoch": 1.011670313639679,
+      "grad_norm": 22.375,
+      "learning_rate": 1.2509665248401882e-06,
+      "loss": 0.5392,
+      "step": 4161
+    },
+    {
+      "epoch": 1.011913445173839,
+      "grad_norm": 22.5,
+      "learning_rate": 1.25048326245621e-06,
+      "loss": 0.8855,
+      "step": 4162
+    },
+    {
+      "epoch": 1.012156576707999,
+      "grad_norm": 19.625,
+      "learning_rate": 1.25e-06,
+      "loss": 0.6698,
+      "step": 4163
+    },
+    {
+      "epoch": 1.012399708242159,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2495167375437902e-06,
+      "loss": 0.6317,
+      "step": 4164
+    },
+    {
+      "epoch": 1.012642839776319,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2490334751598127e-06,
+      "loss": 0.6481,
+      "step": 4165
+    },
+    {
+      "epoch": 1.012885971310479,
+      "grad_norm": 20.125,
+      "learning_rate": 1.2485502129202986e-06,
+      "loss": 1.0355,
+      "step": 4166
+    },
+    {
+      "epoch": 1.013129102844639,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.2480669508974798e-06,
+      "loss": 0.3629,
+      "step": 4167
+    },
+    {
+      "epoch": 1.013372234378799,
+      "grad_norm": 22.875,
+      "learning_rate": 1.247583689163589e-06,
+      "loss": 1.0129,
+      "step": 4168
+    },
+    {
+      "epoch": 1.013615365912959,
+      "grad_norm": 21.625,
+      "learning_rate": 1.2471004277908578e-06,
+      "loss": 0.9057,
+      "step": 4169
+    },
+    {
+      "epoch": 1.013858497447119,
+      "grad_norm": 16.5,
+      "learning_rate": 1.2466171668515174e-06,
+      "loss": 0.5149,
+      "step": 4170
+    },
+    {
+      "epoch": 1.014101628981279,
+      "grad_norm": 18.75,
+      "learning_rate": 1.2461339064178e-06,
+      "loss": 0.6032,
+      "step": 4171
+    },
+    {
+      "epoch": 1.0143447605154388,
+      "grad_norm": 29.0,
+      "learning_rate": 1.2456506465619372e-06,
+      "loss": 0.7118,
+      "step": 4172
+    },
+    {
+      "epoch": 1.0145878920495988,
+      "grad_norm": 18.875,
+      "learning_rate": 1.2451673873561603e-06,
+      "loss": 0.8595,
+      "step": 4173
+    },
+    {
+      "epoch": 1.0148310235837588,
+      "grad_norm": 20.5,
+      "learning_rate": 1.2446841288727005e-06,
+      "loss": 0.7378,
+      "step": 4174
+    },
+    {
+      "epoch": 1.0150741551179188,
+      "grad_norm": 17.0,
+      "learning_rate": 1.2442008711837894e-06,
+      "loss": 0.4911,
+      "step": 4175
+    },
+    {
+      "epoch": 1.0153172866520788,
+      "grad_norm": 23.25,
+      "learning_rate": 1.243717614361659e-06,
+      "loss": 0.7517,
+      "step": 4176
+    },
+    {
+      "epoch": 1.0155604181862388,
+      "grad_norm": 19.5,
+      "learning_rate": 1.2432343584785388e-06,
+      "loss": 1.1732,
+      "step": 4177
+    },
+    {
+      "epoch": 1.0158035497203988,
+      "grad_norm": 17.625,
+      "learning_rate": 1.242751103606661e-06,
+      "loss": 0.7368,
+      "step": 4178
+    },
+    {
+      "epoch": 1.0160466812545588,
+      "grad_norm": 16.375,
+      "learning_rate": 1.2422678498182562e-06,
+      "loss": 0.5969,
+      "step": 4179
+    },
+    {
+      "epoch": 1.0162898127887188,
+      "grad_norm": 18.25,
+      "learning_rate": 1.2417845971855545e-06,
+      "loss": 0.5698,
+      "step": 4180
+    },
+    {
+      "epoch": 1.0165329443228788,
+      "grad_norm": 23.25,
+      "learning_rate": 1.2413013457807865e-06,
+      "loss": 1.1906,
+      "step": 4181
+    },
+    {
+      "epoch": 1.0167760758570386,
+      "grad_norm": 17.75,
+      "learning_rate": 1.240818095676183e-06,
+      "loss": 0.4573,
+      "step": 4182
+    },
+    {
+      "epoch": 1.0170192073911986,
+      "grad_norm": 20.25,
+      "learning_rate": 1.2403348469439731e-06,
+      "loss": 0.7516,
+      "step": 4183
+    },
+    {
+      "epoch": 1.0172623389253586,
+      "grad_norm": 26.0,
+      "learning_rate": 1.2398515996563875e-06,
+      "loss": 0.4984,
+      "step": 4184
+    },
+    {
+      "epoch": 1.0175054704595186,
+      "grad_norm": 24.875,
+      "learning_rate": 1.2393683538856556e-06,
+      "loss": 0.7951,
+      "step": 4185
+    },
+    {
+      "epoch": 1.0177486019936786,
+      "grad_norm": 19.375,
+      "learning_rate": 1.238885109704007e-06,
+      "loss": 0.9981,
+      "step": 4186
+    },
+    {
+      "epoch": 1.0179917335278386,
+      "grad_norm": 17.625,
+      "learning_rate": 1.2384018671836706e-06,
+      "loss": 0.523,
+      "step": 4187
+    },
+    {
+      "epoch": 1.0182348650619986,
+      "grad_norm": 20.625,
+      "learning_rate": 1.2379186263968754e-06,
+      "loss": 0.646,
+      "step": 4188
+    },
+    {
+      "epoch": 1.0184779965961586,
+      "grad_norm": 17.0,
+      "learning_rate": 1.2374353874158506e-06,
+      "loss": 0.4136,
+      "step": 4189
+    },
+    {
+      "epoch": 1.0187211281303186,
+      "grad_norm": 21.375,
+      "learning_rate": 1.2369521503128235e-06,
+      "loss": 1.1995,
+      "step": 4190
+    },
+    {
+      "epoch": 1.0189642596644786,
+      "grad_norm": 17.75,
+      "learning_rate": 1.2364689151600229e-06,
+      "loss": 0.4988,
+      "step": 4191
+    },
+    {
+      "epoch": 1.0192073911986386,
+      "grad_norm": 16.375,
+      "learning_rate": 1.2359856820296767e-06,
+      "loss": 0.3786,
+      "step": 4192
+    },
+    {
+      "epoch": 1.0194505227327983,
+      "grad_norm": 22.125,
+      "learning_rate": 1.2355024509940127e-06,
+      "loss": 0.737,
+      "step": 4193
+    },
+    {
+      "epoch": 1.0196936542669583,
+      "grad_norm": 12.5,
+      "learning_rate": 1.2350192221252576e-06,
+      "loss": 0.2558,
+      "step": 4194
+    },
+    {
+      "epoch": 1.0199367858011184,
+      "grad_norm": 20.25,
+      "learning_rate": 1.2345359954956384e-06,
+      "loss": 0.7269,
+      "step": 4195
+    },
+    {
+      "epoch": 1.0201799173352784,
+      "grad_norm": 27.625,
+      "learning_rate": 1.234052771177382e-06,
+      "loss": 0.7338,
+      "step": 4196
+    },
+    {
+      "epoch": 1.0204230488694384,
+      "grad_norm": 21.5,
+      "learning_rate": 1.2335695492427142e-06,
+      "loss": 0.9474,
+      "step": 4197
+    },
+    {
+      "epoch": 1.0206661804035984,
+      "grad_norm": 24.375,
+      "learning_rate": 1.2330863297638606e-06,
+      "loss": 0.7627,
+      "step": 4198
+    },
+    {
+      "epoch": 1.0209093119377584,
+      "grad_norm": 15.5,
+      "learning_rate": 1.2326031128130476e-06,
+      "loss": 0.6164,
+      "step": 4199
+    },
+    {
+      "epoch": 1.0211524434719184,
+      "grad_norm": 21.875,
+      "learning_rate": 1.2321198984624999e-06,
+      "loss": 0.335,
+      "step": 4200
+    },
+    {
+      "epoch": 1.0213955750060784,
+      "grad_norm": 20.875,
+      "learning_rate": 1.231636686784442e-06,
+      "loss": 0.5493,
+      "step": 4201
+    },
+    {
+      "epoch": 1.0216387065402384,
+      "grad_norm": 18.25,
+      "learning_rate": 1.2311534778510986e-06,
+      "loss": 0.525,
+      "step": 4202
+    },
+    {
+      "epoch": 1.0218818380743981,
+      "grad_norm": 19.25,
+      "learning_rate": 1.2306702717346935e-06,
+      "loss": 0.8582,
+      "step": 4203
+    },
+    {
+      "epoch": 1.0221249696085581,
+      "grad_norm": 22.75,
+      "learning_rate": 1.2301870685074498e-06,
+      "loss": 0.7255,
+      "step": 4204
+    },
+    {
+      "epoch": 1.0223681011427181,
+      "grad_norm": 15.375,
+      "learning_rate": 1.2297038682415909e-06,
+      "loss": 0.4337,
+      "step": 4205
+    },
+    {
+      "epoch": 1.0226112326768781,
+      "grad_norm": 33.0,
+      "learning_rate": 1.2292206710093391e-06,
+      "loss": 0.6941,
+      "step": 4206
+    },
+    {
+      "epoch": 1.0228543642110381,
+      "grad_norm": 17.375,
+      "learning_rate": 1.2287374768829173e-06,
+      "loss": 0.4654,
+      "step": 4207
+    },
+    {
+      "epoch": 1.0230974957451981,
+      "grad_norm": 16.25,
+      "learning_rate": 1.228254285934547e-06,
+      "loss": 0.3755,
+      "step": 4208
+    },
+    {
+      "epoch": 1.0233406272793582,
+      "grad_norm": 21.75,
+      "learning_rate": 1.227771098236449e-06,
+      "loss": 0.7989,
+      "step": 4209
+    },
+    {
+      "epoch": 1.0235837588135182,
+      "grad_norm": 23.125,
+      "learning_rate": 1.2272879138608446e-06,
+      "loss": 0.7926,
+      "step": 4210
+    },
+    {
+      "epoch": 1.0238268903476782,
+      "grad_norm": 21.125,
+      "learning_rate": 1.2268047328799534e-06,
+      "loss": 0.5802,
+      "step": 4211
+    },
+    {
+      "epoch": 1.0240700218818382,
+      "grad_norm": 22.0,
+      "learning_rate": 1.2263215553659953e-06,
+      "loss": 1.2618,
+      "step": 4212
+    },
+    {
+      "epoch": 1.0243131534159982,
+      "grad_norm": 15.0,
+      "learning_rate": 1.2258383813911903e-06,
+      "loss": 0.7276,
+      "step": 4213
+    },
+    {
+      "epoch": 1.024556284950158,
+      "grad_norm": 19.125,
+      "learning_rate": 1.2253552110277554e-06,
+      "loss": 0.487,
+      "step": 4214
+    },
+    {
+      "epoch": 1.024799416484318,
+      "grad_norm": 18.125,
+      "learning_rate": 1.2248720443479107e-06,
+      "loss": 0.8522,
+      "step": 4215
+    },
+    {
+      "epoch": 1.025042548018478,
+      "grad_norm": 17.5,
+      "learning_rate": 1.2243888814238727e-06,
+      "loss": 0.5451,
+      "step": 4216
+    },
+    {
+      "epoch": 1.025285679552638,
+      "grad_norm": 18.75,
+      "learning_rate": 1.223905722327859e-06,
+      "loss": 0.7036,
+      "step": 4217
+    },
+    {
+      "epoch": 1.025528811086798,
+      "grad_norm": 22.0,
+      "learning_rate": 1.2234225671320855e-06,
+      "loss": 0.7887,
+      "step": 4218
+    },
+    {
+      "epoch": 1.025771942620958,
+      "grad_norm": 18.75,
+      "learning_rate": 1.2229394159087682e-06,
+      "loss": 0.5522,
+      "step": 4219
+    },
+    {
+      "epoch": 1.026015074155118,
+      "grad_norm": 21.625,
+      "learning_rate": 1.2224562687301228e-06,
+      "loss": 0.8437,
+      "step": 4220
+    },
+    {
+      "epoch": 1.026258205689278,
+      "grad_norm": 23.5,
+      "learning_rate": 1.2219731256683633e-06,
+      "loss": 0.7484,
+      "step": 4221
+    },
+    {
+      "epoch": 1.026501337223438,
+      "grad_norm": 19.875,
+      "learning_rate": 1.2214899867957048e-06,
+      "loss": 0.6022,
+      "step": 4222
+    },
+    {
+      "epoch": 1.026744468757598,
+      "grad_norm": 26.5,
+      "learning_rate": 1.2210068521843598e-06,
+      "loss": 0.931,
+      "step": 4223
+    },
+    {
+      "epoch": 1.0269876002917577,
+      "grad_norm": 22.125,
+      "learning_rate": 1.2205237219065414e-06,
+      "loss": 0.7175,
+      "step": 4224
+    },
+    {
+      "epoch": 1.0272307318259177,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2200405960344622e-06,
+      "loss": 0.5514,
+      "step": 4225
+    },
+    {
+      "epoch": 1.0274738633600777,
+      "grad_norm": 16.125,
+      "learning_rate": 1.2195574746403332e-06,
+      "loss": 0.7143,
+      "step": 4226
+    },
+    {
+      "epoch": 1.0277169948942377,
+      "grad_norm": 17.125,
+      "learning_rate": 1.2190743577963655e-06,
+      "loss": 0.5525,
+      "step": 4227
+    },
+    {
+      "epoch": 1.0279601264283977,
+      "grad_norm": 22.5,
+      "learning_rate": 1.218591245574769e-06,
+      "loss": 0.9265,
+      "step": 4228
+    },
+    {
+      "epoch": 1.0282032579625577,
+      "grad_norm": 24.0,
+      "learning_rate": 1.218108138047753e-06,
+      "loss": 0.7215,
+      "step": 4229
+    },
+    {
+      "epoch": 1.0284463894967177,
+      "grad_norm": 21.875,
+      "learning_rate": 1.2176250352875267e-06,
+      "loss": 1.129,
+      "step": 4230
+    },
+    {
+      "epoch": 1.0286895210308777,
+      "grad_norm": 24.25,
+      "learning_rate": 1.217141937366298e-06,
+      "loss": 0.6935,
+      "step": 4231
+    },
+    {
+      "epoch": 1.0289326525650377,
+      "grad_norm": 23.625,
+      "learning_rate": 1.2166588443562747e-06,
+      "loss": 0.9392,
+      "step": 4232
+    },
+    {
+      "epoch": 1.0291757840991977,
+      "grad_norm": 41.5,
+      "learning_rate": 1.2161757563296625e-06,
+      "loss": 0.799,
+      "step": 4233
+    },
+    {
+      "epoch": 1.0294189156333577,
+      "grad_norm": 35.5,
+      "learning_rate": 1.2156926733586681e-06,
+      "loss": 1.0792,
+      "step": 4234
+    },
+    {
+      "epoch": 1.0296620471675175,
+      "grad_norm": 18.0,
+      "learning_rate": 1.2152095955154958e-06,
+      "loss": 0.5697,
+      "step": 4235
+    },
+    {
+      "epoch": 1.0299051787016775,
+      "grad_norm": 18.875,
+      "learning_rate": 1.2147265228723502e-06,
+      "loss": 0.4776,
+      "step": 4236
+    },
+    {
+      "epoch": 1.0301483102358375,
+      "grad_norm": 26.125,
+      "learning_rate": 1.2142434555014348e-06,
+      "loss": 0.8266,
+      "step": 4237
+    },
+    {
+      "epoch": 1.0303914417699975,
+      "grad_norm": 19.25,
+      "learning_rate": 1.213760393474953e-06,
+      "loss": 0.6032,
+      "step": 4238
+    },
+    {
+      "epoch": 1.0306345733041575,
+      "grad_norm": 32.25,
+      "learning_rate": 1.2132773368651061e-06,
+      "loss": 0.8117,
+      "step": 4239
+    },
+    {
+      "epoch": 1.0308777048383175,
+      "grad_norm": 17.375,
+      "learning_rate": 1.2127942857440953e-06,
+      "loss": 0.7527,
+      "step": 4240
+    },
+    {
+      "epoch": 1.0311208363724775,
+      "grad_norm": 19.875,
+      "learning_rate": 1.2123112401841212e-06,
+      "loss": 0.8044,
+      "step": 4241
+    },
+    {
+      "epoch": 1.0313639679066375,
+      "grad_norm": 21.375,
+      "learning_rate": 1.2118282002573835e-06,
+      "loss": 0.7041,
+      "step": 4242
+    },
+    {
+      "epoch": 1.0316070994407975,
+      "grad_norm": 18.625,
+      "learning_rate": 1.21134516603608e-06,
+      "loss": 0.5308,
+      "step": 4243
+    },
+    {
+      "epoch": 1.0318502309749575,
+      "grad_norm": 19.875,
+      "learning_rate": 1.2108621375924097e-06,
+      "loss": 0.7222,
+      "step": 4244
+    },
+    {
+      "epoch": 1.0320933625091175,
+      "grad_norm": 21.625,
+      "learning_rate": 1.2103791149985677e-06,
+      "loss": 0.6742,
+      "step": 4245
+    },
+    {
+      "epoch": 1.0323364940432773,
+      "grad_norm": 12.5625,
+      "learning_rate": 1.209896098326752e-06,
+      "loss": 0.2817,
+      "step": 4246
+    },
+    {
+      "epoch": 1.0325796255774373,
+      "grad_norm": 16.5,
+      "learning_rate": 1.209413087649157e-06,
+      "loss": 0.3466,
+      "step": 4247
+    },
+    {
+      "epoch": 1.0328227571115973,
+      "grad_norm": 25.375,
+      "learning_rate": 1.2089300830379767e-06,
+      "loss": 0.6938,
+      "step": 4248
+    },
+    {
+      "epoch": 1.0330658886457573,
+      "grad_norm": 26.625,
+      "learning_rate": 1.2084470845654054e-06,
+      "loss": 0.9129,
+      "step": 4249
+    },
+    {
+      "epoch": 1.0333090201799173,
+      "grad_norm": 17.625,
+      "learning_rate": 1.2079640923036345e-06,
+      "loss": 0.9526,
+      "step": 4250
+    },
+    {
+      "epoch": 1.0335521517140773,
+      "grad_norm": 23.125,
+      "learning_rate": 1.2074811063248564e-06,
+      "loss": 0.5601,
+      "step": 4251
+    },
+    {
+      "epoch": 1.0337952832482373,
+      "grad_norm": 19.25,
+      "learning_rate": 1.2069981267012603e-06,
+      "loss": 0.8824,
+      "step": 4252
+    },
+    {
+      "epoch": 1.0340384147823973,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.2065151535050377e-06,
+      "loss": 0.4185,
+      "step": 4253
+    },
+    {
+      "epoch": 1.0342815463165573,
+      "grad_norm": 18.625,
+      "learning_rate": 1.2060321868083761e-06,
+      "loss": 0.5302,
+      "step": 4254
+    },
+    {
+      "epoch": 1.0345246778507173,
+      "grad_norm": 20.0,
+      "learning_rate": 1.2055492266834637e-06,
+      "loss": 0.3381,
+      "step": 4255
+    },
+    {
+      "epoch": 1.0347678093848771,
+      "grad_norm": 22.375,
+      "learning_rate": 1.205066273202487e-06,
+      "loss": 0.5787,
+      "step": 4256
+    },
+    {
+      "epoch": 1.0350109409190371,
+      "grad_norm": 18.625,
+      "learning_rate": 1.204583326437632e-06,
+      "loss": 0.6395,
+      "step": 4257
+    },
+    {
+      "epoch": 1.0352540724531971,
+      "grad_norm": 14.125,
+      "learning_rate": 1.2041003864610832e-06,
+      "loss": 0.329,
+      "step": 4258
+    },
+    {
+      "epoch": 1.0354972039873571,
+      "grad_norm": 20.75,
+      "learning_rate": 1.2036174533450242e-06,
+      "loss": 0.575,
+      "step": 4259
+    },
+    {
+      "epoch": 1.0357403355215171,
+      "grad_norm": 21.0,
+      "learning_rate": 1.2031345271616376e-06,
+      "loss": 0.6864,
+      "step": 4260
+    },
+    {
+      "epoch": 1.0359834670556771,
+      "grad_norm": 20.0,
+      "learning_rate": 1.2026516079831054e-06,
+      "loss": 0.7953,
+      "step": 4261
+    },
+    {
+      "epoch": 1.0362265985898371,
+      "grad_norm": 20.125,
+      "learning_rate": 1.2021686958816084e-06,
+      "loss": 0.7184,
+      "step": 4262
+    },
+    {
+      "epoch": 1.0364697301239971,
+      "grad_norm": 23.0,
+      "learning_rate": 1.2016857909293258e-06,
+      "loss": 0.9226,
+      "step": 4263
+    },
+    {
+      "epoch": 1.0367128616581571,
+      "grad_norm": 22.5,
+      "learning_rate": 1.2012028931984362e-06,
+      "loss": 1.0413,
+      "step": 4264
+    },
+    {
+      "epoch": 1.0369559931923171,
+      "grad_norm": 20.0,
+      "learning_rate": 1.2007200027611168e-06,
+      "loss": 0.9578,
+      "step": 4265
+    },
+    {
+      "epoch": 1.0371991247264771,
+      "grad_norm": 23.125,
+      "learning_rate": 1.2002371196895444e-06,
+      "loss": 0.9385,
+      "step": 4266
+    },
+    {
+      "epoch": 1.037442256260637,
+      "grad_norm": 22.75,
+      "learning_rate": 1.1997542440558936e-06,
+      "loss": 0.761,
+      "step": 4267
+    },
+    {
+      "epoch": 1.037685387794797,
+      "grad_norm": 17.625,
+      "learning_rate": 1.1992713759323386e-06,
+      "loss": 0.3711,
+      "step": 4268
+    },
+    {
+      "epoch": 1.037928519328957,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.1987885153910527e-06,
+      "loss": 0.3699,
+      "step": 4269
+    },
+    {
+      "epoch": 1.038171650863117,
+      "grad_norm": 17.625,
+      "learning_rate": 1.1983056625042082e-06,
+      "loss": 0.5926,
+      "step": 4270
+    },
+    {
+      "epoch": 1.038414782397277,
+      "grad_norm": 21.5,
+      "learning_rate": 1.1978228173439746e-06,
+      "loss": 0.7479,
+      "step": 4271
+    },
+    {
+      "epoch": 1.038657913931437,
+      "grad_norm": 17.875,
+      "learning_rate": 1.1973399799825222e-06,
+      "loss": 0.8557,
+      "step": 4272
+    },
+    {
+      "epoch": 1.038901045465597,
+      "grad_norm": 26.375,
+      "learning_rate": 1.1968571504920198e-06,
+      "loss": 1.1538,
+      "step": 4273
+    },
+    {
+      "epoch": 1.039144176999757,
+      "grad_norm": 380.0,
+      "learning_rate": 1.1963743289446335e-06,
+      "loss": 0.7731,
+      "step": 4274
+    },
+    {
+      "epoch": 1.039387308533917,
+      "grad_norm": 24.5,
+      "learning_rate": 1.1958915154125303e-06,
+      "loss": 0.669,
+      "step": 4275
+    },
+    {
+      "epoch": 1.039630440068077,
+      "grad_norm": 15.375,
+      "learning_rate": 1.1954087099678736e-06,
+      "loss": 0.3717,
+      "step": 4276
+    },
+    {
+      "epoch": 1.0398735716022367,
+      "grad_norm": 21.75,
+      "learning_rate": 1.194925912682829e-06,
+      "loss": 0.5048,
+      "step": 4277
+    },
+    {
+      "epoch": 1.0401167031363967,
+      "grad_norm": 25.0,
+      "learning_rate": 1.1944431236295577e-06,
+      "loss": 0.9457,
+      "step": 4278
+    },
+    {
+      "epoch": 1.0403598346705567,
+      "grad_norm": 17.875,
+      "learning_rate": 1.1939603428802208e-06,
+      "loss": 0.564,
+      "step": 4279
+    },
+    {
+      "epoch": 1.0406029662047167,
+      "grad_norm": 24.25,
+      "learning_rate": 1.193477570506979e-06,
+      "loss": 0.8875,
+      "step": 4280
+    },
+    {
+      "epoch": 1.0408460977388767,
+      "grad_norm": 26.125,
+      "learning_rate": 1.1929948065819903e-06,
+      "loss": 0.9343,
+      "step": 4281
+    },
+    {
+      "epoch": 1.0410892292730367,
+      "grad_norm": 16.0,
+      "learning_rate": 1.192512051177412e-06,
+      "loss": 0.5083,
+      "step": 4282
+    },
+    {
+      "epoch": 1.0413323608071967,
+      "grad_norm": 13.875,
+      "learning_rate": 1.1920293043654002e-06,
+      "loss": 0.2945,
+      "step": 4283
+    },
+    {
+      "epoch": 1.0415754923413567,
+      "grad_norm": 15.625,
+      "learning_rate": 1.1915465662181109e-06,
+      "loss": 0.499,
+      "step": 4284
+    },
+    {
+      "epoch": 1.0418186238755167,
+      "grad_norm": 15.625,
+      "learning_rate": 1.1910638368076963e-06,
+      "loss": 0.3521,
+      "step": 4285
+    },
+    {
+      "epoch": 1.0420617554096767,
+      "grad_norm": 17.5,
+      "learning_rate": 1.1905811162063093e-06,
+      "loss": 0.6663,
+      "step": 4286
+    },
+    {
+      "epoch": 1.0423048869438367,
+      "grad_norm": 17.125,
+      "learning_rate": 1.1900984044861008e-06,
+      "loss": 0.8063,
+      "step": 4287
+    },
+    {
+      "epoch": 1.0425480184779965,
+      "grad_norm": 18.0,
+      "learning_rate": 1.18961570171922e-06,
+      "loss": 0.6233,
+      "step": 4288
+    },
+    {
+      "epoch": 1.0427911500121565,
+      "grad_norm": 17.375,
+      "learning_rate": 1.1891330079778156e-06,
+      "loss": 0.4607,
+      "step": 4289
+    },
+    {
+      "epoch": 1.0430342815463165,
+      "grad_norm": 18.375,
+      "learning_rate": 1.1886503233340347e-06,
+      "loss": 0.6102,
+      "step": 4290
+    },
+    {
+      "epoch": 1.0432774130804765,
+      "grad_norm": 14.375,
+      "learning_rate": 1.1881676478600217e-06,
+      "loss": 0.5538,
+      "step": 4291
+    },
+    {
+      "epoch": 1.0435205446146365,
+      "grad_norm": 22.0,
+      "learning_rate": 1.1876849816279222e-06,
+      "loss": 0.7654,
+      "step": 4292
+    },
+    {
+      "epoch": 1.0437636761487965,
+      "grad_norm": 18.875,
+      "learning_rate": 1.187202324709878e-06,
+      "loss": 0.8974,
+      "step": 4293
+    },
+    {
+      "epoch": 1.0440068076829565,
+      "grad_norm": 21.125,
+      "learning_rate": 1.1867196771780314e-06,
+      "loss": 0.5014,
+      "step": 4294
+    },
+    {
+      "epoch": 1.0442499392171165,
+      "grad_norm": 21.875,
+      "learning_rate": 1.1862370391045217e-06,
+      "loss": 0.7386,
+      "step": 4295
+    },
+    {
+      "epoch": 1.0444930707512765,
+      "grad_norm": 18.0,
+      "learning_rate": 1.185754410561488e-06,
+      "loss": 0.5077,
+      "step": 4296
+    },
+    {
+      "epoch": 1.0447362022854365,
+      "grad_norm": 21.375,
+      "learning_rate": 1.1852717916210672e-06,
+      "loss": 0.7701,
+      "step": 4297
+    },
+    {
+      "epoch": 1.0449793338195965,
+      "grad_norm": 17.125,
+      "learning_rate": 1.1847891823553947e-06,
+      "loss": 0.4472,
+      "step": 4298
+    },
+    {
+      "epoch": 1.0452224653537563,
+      "grad_norm": 27.125,
+      "learning_rate": 1.184306582836605e-06,
+      "loss": 0.6802,
+      "step": 4299
+    },
+    {
+      "epoch": 1.0454655968879163,
+      "grad_norm": 19.75,
+      "learning_rate": 1.183823993136831e-06,
+      "loss": 0.7162,
+      "step": 4300
+    },
+    {
+      "epoch": 1.0457087284220763,
+      "grad_norm": 16.25,
+      "learning_rate": 1.1833414133282049e-06,
+      "loss": 0.3975,
+      "step": 4301
+    },
+    {
+      "epoch": 1.0459518599562363,
+      "grad_norm": 18.625,
+      "learning_rate": 1.1828588434828552e-06,
+      "loss": 0.7045,
+      "step": 4302
+    },
+    {
+      "epoch": 1.0461949914903963,
+      "grad_norm": 16.125,
+      "learning_rate": 1.182376283672911e-06,
+      "loss": 0.3439,
+      "step": 4303
+    },
+    {
+      "epoch": 1.0464381230245563,
+      "grad_norm": 25.25,
+      "learning_rate": 1.1818937339704995e-06,
+      "loss": 0.6235,
+      "step": 4304
+    },
+    {
+      "epoch": 1.0466812545587163,
+      "grad_norm": 27.75,
+      "learning_rate": 1.1814111944477452e-06,
+      "loss": 0.7153,
+      "step": 4305
+    },
+    {
+      "epoch": 1.0469243860928763,
+      "grad_norm": 26.625,
+      "learning_rate": 1.1809286651767723e-06,
+      "loss": 1.2673,
+      "step": 4306
+    },
+    {
+      "epoch": 1.0471675176270363,
+      "grad_norm": 22.125,
+      "learning_rate": 1.1804461462297035e-06,
+      "loss": 0.7497,
+      "step": 4307
+    },
+    {
+      "epoch": 1.0474106491611963,
+      "grad_norm": 20.125,
+      "learning_rate": 1.1799636376786598e-06,
+      "loss": 0.648,
+      "step": 4308
+    },
+    {
+      "epoch": 1.047653780695356,
+      "grad_norm": 18.875,
+      "learning_rate": 1.1794811395957595e-06,
+      "loss": 0.7847,
+      "step": 4309
+    },
+    {
+      "epoch": 1.047896912229516,
+      "grad_norm": 17.0,
+      "learning_rate": 1.178998652053121e-06,
+      "loss": 0.5397,
+      "step": 4310
+    },
+    {
+      "epoch": 1.048140043763676,
+      "grad_norm": 17.0,
+      "learning_rate": 1.1785161751228606e-06,
+      "loss": 1.2018,
+      "step": 4311
+    },
+    {
+      "epoch": 1.048383175297836,
+      "grad_norm": 22.125,
+      "learning_rate": 1.178033708877092e-06,
+      "loss": 0.9839,
+      "step": 4312
+    },
+    {
+      "epoch": 1.048626306831996,
+      "grad_norm": 18.5,
+      "learning_rate": 1.1775512533879282e-06,
+      "loss": 1.0773,
+      "step": 4313
+    },
+    {
+      "epoch": 1.048869438366156,
+      "grad_norm": 24.75,
+      "learning_rate": 1.1770688087274809e-06,
+      "loss": 0.5905,
+      "step": 4314
+    },
+    {
+      "epoch": 1.049112569900316,
+      "grad_norm": 19.0,
+      "learning_rate": 1.17658637496786e-06,
+      "loss": 0.5783,
+      "step": 4315
+    },
+    {
+      "epoch": 1.049355701434476,
+      "grad_norm": 20.875,
+      "learning_rate": 1.1761039521811731e-06,
+      "loss": 0.5302,
+      "step": 4316
+    },
+    {
+      "epoch": 1.049598832968636,
+      "grad_norm": 19.75,
+      "learning_rate": 1.175621540439527e-06,
+      "loss": 0.7245,
+      "step": 4317
+    },
+    {
+      "epoch": 1.049841964502796,
+      "grad_norm": 24.375,
+      "learning_rate": 1.1751391398150263e-06,
+      "loss": 0.5827,
+      "step": 4318
+    },
+    {
+      "epoch": 1.050085096036956,
+      "grad_norm": 24.75,
+      "learning_rate": 1.1746567503797738e-06,
+      "loss": 0.7684,
+      "step": 4319
+    },
+    {
+      "epoch": 1.0503282275711159,
+      "grad_norm": 18.0,
+      "learning_rate": 1.174174372205871e-06,
+      "loss": 0.8336,
+      "step": 4320
+    },
+    {
+      "epoch": 1.0505713591052759,
+      "grad_norm": 19.0,
+      "learning_rate": 1.1736920053654183e-06,
+      "loss": 0.8152,
+      "step": 4321
+    },
+    {
+      "epoch": 1.0508144906394359,
+      "grad_norm": 26.625,
+      "learning_rate": 1.1732096499305127e-06,
+      "loss": 0.974,
+      "step": 4322
+    },
+    {
+      "epoch": 1.0510576221735959,
+      "grad_norm": 33.0,
+      "learning_rate": 1.1727273059732513e-06,
+      "loss": 0.7634,
+      "step": 4323
+    },
+    {
+      "epoch": 1.0513007537077559,
+      "grad_norm": 15.375,
+      "learning_rate": 1.1722449735657288e-06,
+      "loss": 1.243,
+      "step": 4324
+    },
+    {
+      "epoch": 1.0515438852419159,
+      "grad_norm": 17.75,
+      "learning_rate": 1.171762652780038e-06,
+      "loss": 0.6222,
+      "step": 4325
+    },
+    {
+      "epoch": 1.051787016776076,
+      "grad_norm": 17.375,
+      "learning_rate": 1.1712803436882696e-06,
+      "loss": 0.678,
+      "step": 4326
+    },
+    {
+      "epoch": 1.052030148310236,
+      "grad_norm": 23.625,
+      "learning_rate": 1.1707980463625136e-06,
+      "loss": 0.5647,
+      "step": 4327
+    },
+    {
+      "epoch": 1.052273279844396,
+      "grad_norm": 17.25,
+      "learning_rate": 1.1703157608748574e-06,
+      "loss": 0.7858,
+      "step": 4328
+    },
+    {
+      "epoch": 1.052516411378556,
+      "grad_norm": 22.875,
+      "learning_rate": 1.169833487297387e-06,
+      "loss": 1.1034,
+      "step": 4329
+    },
+    {
+      "epoch": 1.0527595429127157,
+      "grad_norm": 22.375,
+      "learning_rate": 1.1693512257021858e-06,
+      "loss": 0.4154,
+      "step": 4330
+    },
+    {
+      "epoch": 1.0530026744468757,
+      "grad_norm": 22.25,
+      "learning_rate": 1.1688689761613374e-06,
+      "loss": 0.6577,
+      "step": 4331
+    },
+    {
+      "epoch": 1.0532458059810357,
+      "grad_norm": 23.5,
+      "learning_rate": 1.168386738746922e-06,
+      "loss": 0.9481,
+      "step": 4332
+    },
+    {
+      "epoch": 1.0534889375151957,
+      "grad_norm": 23.0,
+      "learning_rate": 1.1679045135310175e-06,
+      "loss": 0.7352,
+      "step": 4333
+    },
+    {
+      "epoch": 1.0537320690493557,
+      "grad_norm": 23.5,
+      "learning_rate": 1.1674223005857013e-06,
+      "loss": 0.6326,
+      "step": 4334
+    },
+    {
+      "epoch": 1.0539752005835157,
+      "grad_norm": 21.125,
+      "learning_rate": 1.1669400999830487e-06,
+      "loss": 0.7645,
+      "step": 4335
+    },
+    {
+      "epoch": 1.0542183321176757,
+      "grad_norm": 16.25,
+      "learning_rate": 1.1664579117951326e-06,
+      "loss": 0.5578,
+      "step": 4336
+    },
+    {
+      "epoch": 1.0544614636518357,
+      "grad_norm": 14.6875,
+      "learning_rate": 1.1659757360940238e-06,
+      "loss": 0.3474,
+      "step": 4337
+    },
+    {
+      "epoch": 1.0547045951859957,
+      "grad_norm": 19.875,
+      "learning_rate": 1.1654935729517928e-06,
+      "loss": 0.9139,
+      "step": 4338
+    },
+    {
+      "epoch": 1.0549477267201557,
+      "grad_norm": 20.875,
+      "learning_rate": 1.165011422440507e-06,
+      "loss": 0.6661,
+      "step": 4339
+    },
+    {
+      "epoch": 1.0551908582543157,
+      "grad_norm": 21.375,
+      "learning_rate": 1.164529284632232e-06,
+      "loss": 0.9341,
+      "step": 4340
+    },
+    {
+      "epoch": 1.0554339897884755,
+      "grad_norm": 23.5,
+      "learning_rate": 1.164047159599031e-06,
+      "loss": 0.8347,
+      "step": 4341
+    },
+    {
+      "epoch": 1.0556771213226355,
+      "grad_norm": 14.125,
+      "learning_rate": 1.1635650474129672e-06,
+      "loss": 0.3434,
+      "step": 4342
+    },
+    {
+      "epoch": 1.0559202528567955,
+      "grad_norm": 18.125,
+      "learning_rate": 1.1630829481460993e-06,
+      "loss": 0.8507,
+      "step": 4343
+    },
+    {
+      "epoch": 1.0561633843909555,
+      "grad_norm": 21.5,
+      "learning_rate": 1.1626008618704862e-06,
+      "loss": 0.6312,
+      "step": 4344
+    },
+    {
+      "epoch": 1.0564065159251155,
+      "grad_norm": 21.5,
+      "learning_rate": 1.1621187886581832e-06,
+      "loss": 0.725,
+      "step": 4345
+    },
+    {
+      "epoch": 1.0566496474592755,
+      "grad_norm": 23.875,
+      "learning_rate": 1.1616367285812453e-06,
+      "loss": 0.8004,
+      "step": 4346
+    },
+    {
+      "epoch": 1.0568927789934355,
+      "grad_norm": 23.625,
+      "learning_rate": 1.1611546817117248e-06,
+      "loss": 0.7786,
+      "step": 4347
+    },
+    {
+      "epoch": 1.0571359105275955,
+      "grad_norm": 22.625,
+      "learning_rate": 1.160672648121671e-06,
+      "loss": 0.6954,
+      "step": 4348
+    },
+    {
+      "epoch": 1.0573790420617555,
+      "grad_norm": 22.875,
+      "learning_rate": 1.1601906278831332e-06,
+      "loss": 0.7504,
+      "step": 4349
+    },
+    {
+      "epoch": 1.0576221735959155,
+      "grad_norm": 26.125,
+      "learning_rate": 1.1597086210681566e-06,
+      "loss": 0.7485,
+      "step": 4350
+    },
+    {
+      "epoch": 1.0578653051300755,
+      "grad_norm": 20.5,
+      "learning_rate": 1.1592266277487862e-06,
+      "loss": 0.8543,
+      "step": 4351
+    },
+    {
+      "epoch": 1.0581084366642353,
+      "grad_norm": 18.375,
+      "learning_rate": 1.1587446479970642e-06,
+      "loss": 0.956,
+      "step": 4352
+    },
+    {
+      "epoch": 1.0583515681983953,
+      "grad_norm": 27.25,
+      "learning_rate": 1.1582626818850298e-06,
+      "loss": 0.7695,
+      "step": 4353
+    },
+    {
+      "epoch": 1.0585946997325553,
+      "grad_norm": 19.375,
+      "learning_rate": 1.1577807294847227e-06,
+      "loss": 0.6911,
+      "step": 4354
+    },
+    {
+      "epoch": 1.0588378312667153,
+      "grad_norm": 18.75,
+      "learning_rate": 1.157298790868178e-06,
+      "loss": 0.8286,
+      "step": 4355
+    },
+    {
+      "epoch": 1.0590809628008753,
+      "grad_norm": 26.25,
+      "learning_rate": 1.1568168661074303e-06,
+      "loss": 0.982,
+      "step": 4356
+    },
+    {
+      "epoch": 1.0593240943350353,
+      "grad_norm": 28.75,
+      "learning_rate": 1.156334955274511e-06,
+      "loss": 0.3835,
+      "step": 4357
+    },
+    {
+      "epoch": 1.0595672258691953,
+      "grad_norm": 27.5,
+      "learning_rate": 1.1558530584414505e-06,
+      "loss": 0.7446,
+      "step": 4358
+    },
+    {
+      "epoch": 1.0598103574033553,
+      "grad_norm": 16.875,
+      "learning_rate": 1.1553711756802762e-06,
+      "loss": 0.614,
+      "step": 4359
+    },
+    {
+      "epoch": 1.0600534889375153,
+      "grad_norm": 15.125,
+      "learning_rate": 1.1548893070630135e-06,
+      "loss": 0.3711,
+      "step": 4360
+    },
+    {
+      "epoch": 1.0602966204716753,
+      "grad_norm": 26.875,
+      "learning_rate": 1.1544074526616874e-06,
+      "loss": 0.9628,
+      "step": 4361
+    },
+    {
+      "epoch": 1.060539752005835,
+      "grad_norm": 19.0,
+      "learning_rate": 1.153925612548318e-06,
+      "loss": 0.4451,
+      "step": 4362
+    },
+    {
+      "epoch": 1.060782883539995,
+      "grad_norm": 15.25,
+      "learning_rate": 1.153443786794925e-06,
+      "loss": 0.492,
+      "step": 4363
+    },
+    {
+      "epoch": 1.061026015074155,
+      "grad_norm": 20.625,
+      "learning_rate": 1.152961975473526e-06,
+      "loss": 0.9892,
+      "step": 4364
+    },
+    {
+      "epoch": 1.061269146608315,
+      "grad_norm": 17.625,
+      "learning_rate": 1.1524801786561355e-06,
+      "loss": 0.6712,
+      "step": 4365
+    },
+    {
+      "epoch": 1.061512278142475,
+      "grad_norm": 20.125,
+      "learning_rate": 1.1519983964147667e-06,
+      "loss": 0.6189,
+      "step": 4366
+    },
+    {
+      "epoch": 1.061755409676635,
+      "grad_norm": 18.5,
+      "learning_rate": 1.1515166288214298e-06,
+      "loss": 0.7311,
+      "step": 4367
+    },
+    {
+      "epoch": 1.061998541210795,
+      "grad_norm": 18.875,
+      "learning_rate": 1.1510348759481332e-06,
+      "loss": 0.6576,
+      "step": 4368
+    },
+    {
+      "epoch": 1.062241672744955,
+      "grad_norm": 16.5,
+      "learning_rate": 1.150553137866884e-06,
+      "loss": 0.6377,
+      "step": 4369
+    },
+    {
+      "epoch": 1.062484804279115,
+      "grad_norm": 18.375,
+      "learning_rate": 1.1500714146496858e-06,
+      "loss": 0.4656,
+      "step": 4370
+    },
+    {
+      "epoch": 1.062727935813275,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.149589706368541e-06,
+      "loss": 0.5009,
+      "step": 4371
+    },
+    {
+      "epoch": 1.0629710673474349,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.149108013095448e-06,
+      "loss": 0.3528,
+      "step": 4372
+    },
+    {
+      "epoch": 1.0632141988815949,
+      "grad_norm": 23.625,
+      "learning_rate": 1.1486263349024055e-06,
+      "loss": 0.7346,
+      "step": 4373
+    },
+    {
+      "epoch": 1.0634573304157549,
+      "grad_norm": 25.625,
+      "learning_rate": 1.1481446718614077e-06,
+      "loss": 0.9986,
+      "step": 4374
+    },
+    {
+      "epoch": 1.0637004619499149,
+      "grad_norm": 21.625,
+      "learning_rate": 1.1476630240444478e-06,
+      "loss": 0.5542,
+      "step": 4375
+    },
+    {
+      "epoch": 1.0639435934840749,
+      "grad_norm": 16.375,
+      "learning_rate": 1.147181391523516e-06,
+      "loss": 0.788,
+      "step": 4376
+    },
+    {
+      "epoch": 1.0641867250182349,
+      "grad_norm": 26.375,
+      "learning_rate": 1.1466997743706015e-06,
+      "loss": 0.7264,
+      "step": 4377
+    },
+    {
+      "epoch": 1.0644298565523949,
+      "grad_norm": 17.5,
+      "learning_rate": 1.14621817265769e-06,
+      "loss": 0.8345,
+      "step": 4378
+    },
+    {
+      "epoch": 1.0646729880865549,
+      "grad_norm": 17.0,
+      "learning_rate": 1.1457365864567645e-06,
+      "loss": 0.9265,
+      "step": 4379
+    },
+    {
+      "epoch": 1.0649161196207149,
+      "grad_norm": 26.75,
+      "learning_rate": 1.145255015839807e-06,
+      "loss": 0.6017,
+      "step": 4380
+    },
+    {
+      "epoch": 1.0651592511548749,
+      "grad_norm": 20.625,
+      "learning_rate": 1.1447734608787966e-06,
+      "loss": 0.6863,
+      "step": 4381
+    },
+    {
+      "epoch": 1.0654023826890349,
+      "grad_norm": 17.5,
+      "learning_rate": 1.1442919216457095e-06,
+      "loss": 0.7533,
+      "step": 4382
+    },
+    {
+      "epoch": 1.0656455142231946,
+      "grad_norm": 17.125,
+      "learning_rate": 1.1438103982125206e-06,
+      "loss": 0.4611,
+      "step": 4383
+    },
+    {
+      "epoch": 1.0658886457573546,
+      "grad_norm": 23.0,
+      "learning_rate": 1.143328890651201e-06,
+      "loss": 0.5866,
+      "step": 4384
+    },
+    {
+      "epoch": 1.0661317772915146,
+      "grad_norm": 21.375,
+      "learning_rate": 1.1428473990337216e-06,
+      "loss": 1.1776,
+      "step": 4385
+    },
+    {
+      "epoch": 1.0663749088256747,
+      "grad_norm": 20.25,
+      "learning_rate": 1.1423659234320489e-06,
+      "loss": 0.8063,
+      "step": 4386
+    },
+    {
+      "epoch": 1.0666180403598347,
+      "grad_norm": 24.375,
+      "learning_rate": 1.1418844639181476e-06,
+      "loss": 0.9249,
+      "step": 4387
+    },
+    {
+      "epoch": 1.0668611718939947,
+      "grad_norm": 19.375,
+      "learning_rate": 1.141403020563981e-06,
+      "loss": 0.5387,
+      "step": 4388
+    },
+    {
+      "epoch": 1.0671043034281547,
+      "grad_norm": 18.25,
+      "learning_rate": 1.140921593441508e-06,
+      "loss": 0.5215,
+      "step": 4389
+    },
+    {
+      "epoch": 1.0673474349623147,
+      "grad_norm": 23.125,
+      "learning_rate": 1.140440182622687e-06,
+      "loss": 0.7367,
+      "step": 4390
+    },
+    {
+      "epoch": 1.0675905664964747,
+      "grad_norm": 17.875,
+      "learning_rate": 1.139958788179472e-06,
+      "loss": 0.7743,
+      "step": 4391
+    },
+    {
+      "epoch": 1.0678336980306347,
+      "grad_norm": 17.0,
+      "learning_rate": 1.1394774101838176e-06,
+      "loss": 0.5295,
+      "step": 4392
+    },
+    {
+      "epoch": 1.0680768295647947,
+      "grad_norm": 17.625,
+      "learning_rate": 1.1389960487076726e-06,
+      "loss": 0.4567,
+      "step": 4393
+    },
+    {
+      "epoch": 1.0683199610989544,
+      "grad_norm": 17.875,
+      "learning_rate": 1.1385147038229853e-06,
+      "loss": 0.7191,
+      "step": 4394
+    },
+    {
+      "epoch": 1.0685630926331144,
+      "grad_norm": 26.0,
+      "learning_rate": 1.1380333756017013e-06,
+      "loss": 0.684,
+      "step": 4395
+    },
+    {
+      "epoch": 1.0688062241672744,
+      "grad_norm": 17.25,
+      "learning_rate": 1.1375520641157626e-06,
+      "loss": 0.4503,
+      "step": 4396
+    },
+    {
+      "epoch": 1.0690493557014344,
+      "grad_norm": 27.25,
+      "learning_rate": 1.13707076943711e-06,
+      "loss": 0.7169,
+      "step": 4397
+    },
+    {
+      "epoch": 1.0692924872355944,
+      "grad_norm": 17.75,
+      "learning_rate": 1.1365894916376815e-06,
+      "loss": 0.6929,
+      "step": 4398
+    },
+    {
+      "epoch": 1.0695356187697544,
+      "grad_norm": 20.25,
+      "learning_rate": 1.1361082307894116e-06,
+      "loss": 0.7774,
+      "step": 4399
+    },
+    {
+      "epoch": 1.0697787503039144,
+      "grad_norm": 21.75,
+      "learning_rate": 1.135626986964234e-06,
+      "loss": 0.6137,
+      "step": 4400
+    },
+    {
+      "epoch": 1.0700218818380745,
+      "grad_norm": 21.875,
+      "learning_rate": 1.1351457602340783e-06,
+      "loss": 0.9395,
+      "step": 4401
+    },
+    {
+      "epoch": 1.0702650133722345,
+      "grad_norm": 34.25,
+      "learning_rate": 1.1346645506708726e-06,
+      "loss": 0.7565,
+      "step": 4402
+    },
+    {
+      "epoch": 1.0705081449063945,
+      "grad_norm": 49.0,
+      "learning_rate": 1.1341833583465418e-06,
+      "loss": 1.0853,
+      "step": 4403
+    },
+    {
+      "epoch": 1.0707512764405545,
+      "grad_norm": 20.0,
+      "learning_rate": 1.1337021833330079e-06,
+      "loss": 0.8362,
+      "step": 4404
+    },
+    {
+      "epoch": 1.0709944079747142,
+      "grad_norm": 26.875,
+      "learning_rate": 1.1332210257021917e-06,
+      "loss": 0.4942,
+      "step": 4405
+    },
+    {
+      "epoch": 1.0712375395088742,
+      "grad_norm": 19.375,
+      "learning_rate": 1.1327398855260097e-06,
+      "loss": 0.7441,
+      "step": 4406
+    },
+    {
+      "epoch": 1.0714806710430342,
+      "grad_norm": 20.25,
+      "learning_rate": 1.1322587628763768e-06,
+      "loss": 0.7362,
+      "step": 4407
+    },
+    {
+      "epoch": 1.0717238025771942,
+      "grad_norm": 29.0,
+      "learning_rate": 1.1317776578252054e-06,
+      "loss": 1.0423,
+      "step": 4408
+    },
+    {
+      "epoch": 1.0719669341113542,
+      "grad_norm": 30.5,
+      "learning_rate": 1.1312965704444053e-06,
+      "loss": 0.6931,
+      "step": 4409
+    },
+    {
+      "epoch": 1.0722100656455142,
+      "grad_norm": 16.75,
+      "learning_rate": 1.1308155008058823e-06,
+      "loss": 0.3339,
+      "step": 4410
+    },
+    {
+      "epoch": 1.0724531971796742,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.1303344489815412e-06,
+      "loss": 1.0337,
+      "step": 4411
+    },
+    {
+      "epoch": 1.0726963287138342,
+      "grad_norm": 21.25,
+      "learning_rate": 1.1298534150432835e-06,
+      "loss": 0.7422,
+      "step": 4412
+    },
+    {
+      "epoch": 1.0729394602479942,
+      "grad_norm": 29.5,
+      "learning_rate": 1.129372399063008e-06,
+      "loss": 1.074,
+      "step": 4413
+    },
+    {
+      "epoch": 1.0731825917821542,
+      "grad_norm": 22.125,
+      "learning_rate": 1.1288914011126104e-06,
+      "loss": 0.5714,
+      "step": 4414
+    },
+    {
+      "epoch": 1.073425723316314,
+      "grad_norm": 24.5,
+      "learning_rate": 1.1284104212639847e-06,
+      "loss": 0.6241,
+      "step": 4415
+    },
+    {
+      "epoch": 1.073668854850474,
+      "grad_norm": 20.5,
+      "learning_rate": 1.1279294595890215e-06,
+      "loss": 1.1062,
+      "step": 4416
+    },
+    {
+      "epoch": 1.073911986384634,
+      "grad_norm": 20.875,
+      "learning_rate": 1.1274485161596088e-06,
+      "loss": 0.7436,
+      "step": 4417
+    },
+    {
+      "epoch": 1.074155117918794,
+      "grad_norm": 22.125,
+      "learning_rate": 1.1269675910476319e-06,
+      "loss": 0.9513,
+      "step": 4418
+    },
+    {
+      "epoch": 1.074398249452954,
+      "grad_norm": 21.75,
+      "learning_rate": 1.1264866843249736e-06,
+      "loss": 1.0243,
+      "step": 4419
+    },
+    {
+      "epoch": 1.074641380987114,
+      "grad_norm": 24.5,
+      "learning_rate": 1.1260057960635132e-06,
+      "loss": 0.909,
+      "step": 4420
+    },
+    {
+      "epoch": 1.074884512521274,
+      "grad_norm": 35.5,
+      "learning_rate": 1.125524926335128e-06,
+      "loss": 0.6748,
+      "step": 4421
+    },
+    {
+      "epoch": 1.075127644055434,
+      "grad_norm": 32.25,
+      "learning_rate": 1.125044075211692e-06,
+      "loss": 0.9951,
+      "step": 4422
+    },
+    {
+      "epoch": 1.075370775589594,
+      "grad_norm": 26.625,
+      "learning_rate": 1.1245632427650777e-06,
+      "loss": 0.9535,
+      "step": 4423
+    },
+    {
+      "epoch": 1.075613907123754,
+      "grad_norm": 18.75,
+      "learning_rate": 1.1240824290671528e-06,
+      "loss": 0.5518,
+      "step": 4424
+    },
+    {
+      "epoch": 1.0758570386579138,
+      "grad_norm": 26.25,
+      "learning_rate": 1.1236016341897836e-06,
+      "loss": 0.6448,
+      "step": 4425
+    },
+    {
+      "epoch": 1.0761001701920738,
+      "grad_norm": 20.625,
+      "learning_rate": 1.1231208582048333e-06,
+      "loss": 0.5226,
+      "step": 4426
+    },
+    {
+      "epoch": 1.0763433017262338,
+      "grad_norm": 18.25,
+      "learning_rate": 1.1226401011841618e-06,
+      "loss": 0.6849,
+      "step": 4427
+    },
+    {
+      "epoch": 1.0765864332603938,
+      "grad_norm": 15.25,
+      "learning_rate": 1.122159363199627e-06,
+      "loss": 0.425,
+      "step": 4428
+    },
+    {
+      "epoch": 1.0768295647945538,
+      "grad_norm": 27.25,
+      "learning_rate": 1.1216786443230832e-06,
+      "loss": 0.8265,
+      "step": 4429
+    },
+    {
+      "epoch": 1.0770726963287138,
+      "grad_norm": 20.875,
+      "learning_rate": 1.1211979446263816e-06,
+      "loss": 0.9099,
+      "step": 4430
+    },
+    {
+      "epoch": 1.0773158278628738,
+      "grad_norm": 15.0,
+      "learning_rate": 1.1207172641813723e-06,
+      "loss": 0.369,
+      "step": 4431
+    },
+    {
+      "epoch": 1.0775589593970338,
+      "grad_norm": 28.5,
+      "learning_rate": 1.1202366030599003e-06,
+      "loss": 0.5714,
+      "step": 4432
+    },
+    {
+      "epoch": 1.0778020909311938,
+      "grad_norm": 28.75,
+      "learning_rate": 1.1197559613338094e-06,
+      "loss": 0.6227,
+      "step": 4433
+    },
+    {
+      "epoch": 1.0780452224653538,
+      "grad_norm": 20.875,
+      "learning_rate": 1.1192753390749394e-06,
+      "loss": 0.9075,
+      "step": 4434
+    },
+    {
+      "epoch": 1.0782883539995138,
+      "grad_norm": 23.0,
+      "learning_rate": 1.1187947363551276e-06,
+      "loss": 0.5589,
+      "step": 4435
+    },
+    {
+      "epoch": 1.0785314855336736,
+      "grad_norm": 19.25,
+      "learning_rate": 1.1183141532462085e-06,
+      "loss": 0.7107,
+      "step": 4436
+    },
+    {
+      "epoch": 1.0787746170678336,
+      "grad_norm": 23.75,
+      "learning_rate": 1.1178335898200135e-06,
+      "loss": 1.2436,
+      "step": 4437
+    },
+    {
+      "epoch": 1.0790177486019936,
+      "grad_norm": 19.875,
+      "learning_rate": 1.1173530461483708e-06,
+      "loss": 0.6621,
+      "step": 4438
+    },
+    {
+      "epoch": 1.0792608801361536,
+      "grad_norm": 20.875,
+      "learning_rate": 1.1168725223031062e-06,
+      "loss": 0.6918,
+      "step": 4439
+    },
+    {
+      "epoch": 1.0795040116703136,
+      "grad_norm": 21.125,
+      "learning_rate": 1.116392018356043e-06,
+      "loss": 0.8615,
+      "step": 4440
+    },
+    {
+      "epoch": 1.0797471432044736,
+      "grad_norm": 14.75,
+      "learning_rate": 1.1159115343789997e-06,
+      "loss": 0.629,
+      "step": 4441
+    },
+    {
+      "epoch": 1.0799902747386336,
+      "grad_norm": 18.875,
+      "learning_rate": 1.1154310704437934e-06,
+      "loss": 0.6937,
+      "step": 4442
+    },
+    {
+      "epoch": 1.0802334062727936,
+      "grad_norm": 19.875,
+      "learning_rate": 1.114950626622238e-06,
+      "loss": 0.4885,
+      "step": 4443
+    },
+    {
+      "epoch": 1.0804765378069536,
+      "grad_norm": 20.75,
+      "learning_rate": 1.1144702029861435e-06,
+      "loss": 0.6823,
+      "step": 4444
+    },
+    {
+      "epoch": 1.0807196693411136,
+      "grad_norm": 18.125,
+      "learning_rate": 1.1139897996073175e-06,
+      "loss": 0.7005,
+      "step": 4445
+    },
+    {
+      "epoch": 1.0809628008752736,
+      "grad_norm": 22.75,
+      "learning_rate": 1.1135094165575654e-06,
+      "loss": 0.9733,
+      "step": 4446
+    },
+    {
+      "epoch": 1.0812059324094334,
+      "grad_norm": 25.5,
+      "learning_rate": 1.1130290539086885e-06,
+      "loss": 0.743,
+      "step": 4447
+    },
+    {
+      "epoch": 1.0814490639435934,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.1125487117324846e-06,
+      "loss": 0.291,
+      "step": 4448
+    },
+    {
+      "epoch": 1.0816921954777534,
+      "grad_norm": 22.875,
+      "learning_rate": 1.1120683901007497e-06,
+      "loss": 1.1769,
+      "step": 4449
+    },
+    {
+      "epoch": 1.0819353270119134,
+      "grad_norm": 15.25,
+      "learning_rate": 1.1115880890852763e-06,
+      "loss": 0.3238,
+      "step": 4450
+    },
+    {
+      "epoch": 1.0821784585460734,
+      "grad_norm": 17.125,
+      "learning_rate": 1.1111078087578533e-06,
+      "loss": 0.5467,
+      "step": 4451
+    },
+    {
+      "epoch": 1.0824215900802334,
+      "grad_norm": 24.875,
+      "learning_rate": 1.110627549190267e-06,
+      "loss": 0.5565,
+      "step": 4452
+    },
+    {
+      "epoch": 1.0826647216143934,
+      "grad_norm": 21.875,
+      "learning_rate": 1.1101473104543003e-06,
+      "loss": 0.8823,
+      "step": 4453
+    },
+    {
+      "epoch": 1.0829078531485534,
+      "grad_norm": 17.125,
+      "learning_rate": 1.1096670926217338e-06,
+      "loss": 0.4465,
+      "step": 4454
+    },
+    {
+      "epoch": 1.0831509846827134,
+      "grad_norm": 21.375,
+      "learning_rate": 1.1091868957643436e-06,
+      "loss": 0.6273,
+      "step": 4455
+    },
+    {
+      "epoch": 1.0833941162168734,
+      "grad_norm": 18.75,
+      "learning_rate": 1.108706719953904e-06,
+      "loss": 0.5618,
+      "step": 4456
+    },
+    {
+      "epoch": 1.0836372477510332,
+      "grad_norm": 18.375,
+      "learning_rate": 1.1082265652621854e-06,
+      "loss": 0.5161,
+      "step": 4457
+    },
+    {
+      "epoch": 1.0838803792851932,
+      "grad_norm": 17.75,
+      "learning_rate": 1.1077464317609549e-06,
+      "loss": 0.4918,
+      "step": 4458
+    },
+    {
+      "epoch": 1.0841235108193532,
+      "grad_norm": 16.375,
+      "learning_rate": 1.107266319521977e-06,
+      "loss": 0.4809,
+      "step": 4459
+    },
+    {
+      "epoch": 1.0843666423535132,
+      "grad_norm": 19.75,
+      "learning_rate": 1.106786228617013e-06,
+      "loss": 0.7428,
+      "step": 4460
+    },
+    {
+      "epoch": 1.0846097738876732,
+      "grad_norm": 22.875,
+      "learning_rate": 1.1063061591178198e-06,
+      "loss": 0.6756,
+      "step": 4461
+    },
+    {
+      "epoch": 1.0848529054218332,
+      "grad_norm": 21.375,
+      "learning_rate": 1.1058261110961535e-06,
+      "loss": 0.9119,
+      "step": 4462
+    },
+    {
+      "epoch": 1.0850960369559932,
+      "grad_norm": 20.25,
+      "learning_rate": 1.1053460846237645e-06,
+      "loss": 0.4709,
+      "step": 4463
+    },
+    {
+      "epoch": 1.0853391684901532,
+      "grad_norm": 19.375,
+      "learning_rate": 1.1048660797724018e-06,
+      "loss": 0.6212,
+      "step": 4464
+    },
+    {
+      "epoch": 1.0855823000243132,
+      "grad_norm": 19.125,
+      "learning_rate": 1.1043860966138098e-06,
+      "loss": 0.7112,
+      "step": 4465
+    },
+    {
+      "epoch": 1.0858254315584732,
+      "grad_norm": 18.25,
+      "learning_rate": 1.1039061352197306e-06,
+      "loss": 0.5805,
+      "step": 4466
+    },
+    {
+      "epoch": 1.086068563092633,
+      "grad_norm": 25.875,
+      "learning_rate": 1.1034261956619028e-06,
+      "loss": 0.514,
+      "step": 4467
+    },
+    {
+      "epoch": 1.086311694626793,
+      "grad_norm": 21.5,
+      "learning_rate": 1.1029462780120611e-06,
+      "loss": 0.584,
+      "step": 4468
+    },
+    {
+      "epoch": 1.086554826160953,
+      "grad_norm": 16.125,
+      "learning_rate": 1.1024663823419377e-06,
+      "loss": 0.6386,
+      "step": 4469
+    },
+    {
+      "epoch": 1.086797957695113,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.1019865087232615e-06,
+      "loss": 0.5411,
+      "step": 4470
+    },
+    {
+      "epoch": 1.087041089229273,
+      "grad_norm": 15.75,
+      "learning_rate": 1.1015066572277583e-06,
+      "loss": 0.471,
+      "step": 4471
+    },
+    {
+      "epoch": 1.087284220763433,
+      "grad_norm": 16.125,
+      "learning_rate": 1.1010268279271495e-06,
+      "loss": 0.483,
+      "step": 4472
+    },
+    {
+      "epoch": 1.087527352297593,
+      "grad_norm": 16.0,
+      "learning_rate": 1.100547020893154e-06,
+      "loss": 0.7322,
+      "step": 4473
+    },
+    {
+      "epoch": 1.087770483831753,
+      "grad_norm": 18.25,
+      "learning_rate": 1.1000672361974879e-06,
+      "loss": 0.7431,
+      "step": 4474
+    },
+    {
+      "epoch": 1.088013615365913,
+      "grad_norm": 19.375,
+      "learning_rate": 1.0995874739118623e-06,
+      "loss": 0.7328,
+      "step": 4475
+    },
+    {
+      "epoch": 1.088256746900073,
+      "grad_norm": 18.5,
+      "learning_rate": 1.0991077341079862e-06,
+      "loss": 0.6849,
+      "step": 4476
+    },
+    {
+      "epoch": 1.088499878434233,
+      "grad_norm": 20.375,
+      "learning_rate": 1.0986280168575655e-06,
+      "loss": 0.9232,
+      "step": 4477
+    },
+    {
+      "epoch": 1.0887430099683928,
+      "grad_norm": 20.875,
+      "learning_rate": 1.0981483222323023e-06,
+      "loss": 1.1457,
+      "step": 4478
+    },
+    {
+      "epoch": 1.0889861415025528,
+      "grad_norm": 19.375,
+      "learning_rate": 1.0976686503038945e-06,
+      "loss": 0.677,
+      "step": 4479
+    },
+    {
+      "epoch": 1.0892292730367128,
+      "grad_norm": 24.25,
+      "learning_rate": 1.097189001144038e-06,
+      "loss": 1.3531,
+      "step": 4480
+    },
+    {
+      "epoch": 1.0894724045708728,
+      "grad_norm": 19.5,
+      "learning_rate": 1.0967093748244245e-06,
+      "loss": 0.733,
+      "step": 4481
+    },
+    {
+      "epoch": 1.0897155361050328,
+      "grad_norm": 20.125,
+      "learning_rate": 1.096229771416742e-06,
+      "loss": 0.6181,
+      "step": 4482
+    },
+    {
+      "epoch": 1.0899586676391928,
+      "grad_norm": 39.25,
+      "learning_rate": 1.0957501909926761e-06,
+      "loss": 0.8568,
+      "step": 4483
+    },
+    {
+      "epoch": 1.0902017991733528,
+      "grad_norm": 16.875,
+      "learning_rate": 1.0952706336239077e-06,
+      "loss": 0.5019,
+      "step": 4484
+    },
+    {
+      "epoch": 1.0904449307075128,
+      "grad_norm": 17.125,
+      "learning_rate": 1.0947910993821157e-06,
+      "loss": 0.4438,
+      "step": 4485
+    },
+    {
+      "epoch": 1.0906880622416728,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.0943115883389747e-06,
+      "loss": 0.3666,
+      "step": 4486
+    },
+    {
+      "epoch": 1.0909311937758328,
+      "grad_norm": 25.875,
+      "learning_rate": 1.0938321005661555e-06,
+      "loss": 0.9279,
+      "step": 4487
+    },
+    {
+      "epoch": 1.0911743253099928,
+      "grad_norm": 23.125,
+      "learning_rate": 1.0933526361353261e-06,
+      "loss": 0.8652,
+      "step": 4488
+    },
+    {
+      "epoch": 1.0914174568441526,
+      "grad_norm": 15.1875,
+      "learning_rate": 1.0928731951181505e-06,
+      "loss": 0.7262,
+      "step": 4489
+    },
+    {
+      "epoch": 1.0916605883783126,
+      "grad_norm": 19.75,
+      "learning_rate": 1.0923937775862894e-06,
+      "loss": 0.65,
+      "step": 4490
+    },
+    {
+      "epoch": 1.0919037199124726,
+      "grad_norm": 16.5,
+      "learning_rate": 1.0919143836114005e-06,
+      "loss": 0.8162,
+      "step": 4491
+    },
+    {
+      "epoch": 1.0921468514466326,
+      "grad_norm": 17.25,
+      "learning_rate": 1.0914350132651367e-06,
+      "loss": 0.6148,
+      "step": 4492
+    },
+    {
+      "epoch": 1.0923899829807926,
+      "grad_norm": 19.25,
+      "learning_rate": 1.090955666619149e-06,
+      "loss": 1.0146,
+      "step": 4493
+    },
+    {
+      "epoch": 1.0926331145149526,
+      "grad_norm": 13.6875,
+      "learning_rate": 1.0904763437450839e-06,
+      "loss": 0.2574,
+      "step": 4494
+    },
+    {
+      "epoch": 1.0928762460491126,
+      "grad_norm": 22.25,
+      "learning_rate": 1.0899970447145843e-06,
+      "loss": 0.8433,
+      "step": 4495
+    },
+    {
+      "epoch": 1.0931193775832726,
+      "grad_norm": 18.0,
+      "learning_rate": 1.0895177695992895e-06,
+      "loss": 0.5903,
+      "step": 4496
+    },
+    {
+      "epoch": 1.0933625091174326,
+      "grad_norm": 20.875,
+      "learning_rate": 1.0890385184708354e-06,
+      "loss": 0.488,
+      "step": 4497
+    },
+    {
+      "epoch": 1.0936056406515926,
+      "grad_norm": 27.0,
+      "learning_rate": 1.0885592914008552e-06,
+      "loss": 0.9404,
+      "step": 4498
+    },
+    {
+      "epoch": 1.0938487721857526,
+      "grad_norm": 16.875,
+      "learning_rate": 1.088080088460977e-06,
+      "loss": 0.5162,
+      "step": 4499
+    },
+    {
+      "epoch": 1.0940919037199124,
+      "grad_norm": 35.5,
+      "learning_rate": 1.0876009097228253e-06,
+      "loss": 1.0972,
+      "step": 4500
+    },
+    {
+      "epoch": 1.0943350352540724,
+      "grad_norm": 22.25,
+      "learning_rate": 1.0871217552580227e-06,
+      "loss": 0.7532,
+      "step": 4501
+    },
+    {
+      "epoch": 1.0945781667882324,
+      "grad_norm": 18.875,
+      "learning_rate": 1.0866426251381871e-06,
+      "loss": 0.8938,
+      "step": 4502
+    },
+    {
+      "epoch": 1.0948212983223924,
+      "grad_norm": 19.375,
+      "learning_rate": 1.0861635194349326e-06,
+      "loss": 0.4472,
+      "step": 4503
+    },
+    {
+      "epoch": 1.0950644298565524,
+      "grad_norm": 20.625,
+      "learning_rate": 1.0856844382198691e-06,
+      "loss": 0.505,
+      "step": 4504
+    },
+    {
+      "epoch": 1.0953075613907124,
+      "grad_norm": 19.625,
+      "learning_rate": 1.0852053815646048e-06,
+      "loss": 0.8325,
+      "step": 4505
+    },
+    {
+      "epoch": 1.0955506929248724,
+      "grad_norm": 20.75,
+      "learning_rate": 1.084726349540742e-06,
+      "loss": 0.7421,
+      "step": 4506
+    },
+    {
+      "epoch": 1.0957938244590324,
+      "grad_norm": 21.25,
+      "learning_rate": 1.0842473422198801e-06,
+      "loss": 1.3701,
+      "step": 4507
+    },
+    {
+      "epoch": 1.0960369559931924,
+      "grad_norm": 17.5,
+      "learning_rate": 1.083768359673616e-06,
+      "loss": 0.4848,
+      "step": 4508
+    },
+    {
+      "epoch": 1.0962800875273524,
+      "grad_norm": 32.0,
+      "learning_rate": 1.0832894019735416e-06,
+      "loss": 0.7744,
+      "step": 4509
+    },
+    {
+      "epoch": 1.0965232190615122,
+      "grad_norm": 22.75,
+      "learning_rate": 1.0828104691912452e-06,
+      "loss": 0.894,
+      "step": 4510
+    },
+    {
+      "epoch": 1.0967663505956722,
+      "grad_norm": 19.875,
+      "learning_rate": 1.0823315613983118e-06,
+      "loss": 0.913,
+      "step": 4511
+    },
+    {
+      "epoch": 1.0970094821298322,
+      "grad_norm": 18.75,
+      "learning_rate": 1.0818526786663221e-06,
+      "loss": 0.4371,
+      "step": 4512
+    },
+    {
+      "epoch": 1.0972526136639922,
+      "grad_norm": 23.875,
+      "learning_rate": 1.0813738210668537e-06,
+      "loss": 0.8429,
+      "step": 4513
+    },
+    {
+      "epoch": 1.0974957451981522,
+      "grad_norm": 20.0,
+      "learning_rate": 1.0808949886714798e-06,
+      "loss": 0.7703,
+      "step": 4514
+    },
+    {
+      "epoch": 1.0977388767323122,
+      "grad_norm": 25.5,
+      "learning_rate": 1.0804161815517702e-06,
+      "loss": 0.9715,
+      "step": 4515
+    },
+    {
+      "epoch": 1.0979820082664722,
+      "grad_norm": 18.625,
+      "learning_rate": 1.0799373997792913e-06,
+      "loss": 0.678,
+      "step": 4516
+    },
+    {
+      "epoch": 1.0982251398006322,
+      "grad_norm": 16.25,
+      "learning_rate": 1.0794586434256053e-06,
+      "loss": 0.6187,
+      "step": 4517
+    },
+    {
+      "epoch": 1.0984682713347922,
+      "grad_norm": 18.875,
+      "learning_rate": 1.0789799125622701e-06,
+      "loss": 0.5908,
+      "step": 4518
+    },
+    {
+      "epoch": 1.0987114028689522,
+      "grad_norm": 21.625,
+      "learning_rate": 1.0785012072608408e-06,
+      "loss": 0.631,
+      "step": 4519
+    },
+    {
+      "epoch": 1.098954534403112,
+      "grad_norm": 20.75,
+      "learning_rate": 1.0780225275928682e-06,
+      "loss": 0.8226,
+      "step": 4520
+    },
+    {
+      "epoch": 1.099197665937272,
+      "grad_norm": 24.875,
+      "learning_rate": 1.0775438736298987e-06,
+      "loss": 0.7697,
+      "step": 4521
+    },
+    {
+      "epoch": 1.099440797471432,
+      "grad_norm": 16.875,
+      "learning_rate": 1.0770652454434758e-06,
+      "loss": 0.4744,
+      "step": 4522
+    },
+    {
+      "epoch": 1.099683929005592,
+      "grad_norm": 20.25,
+      "learning_rate": 1.076586643105138e-06,
+      "loss": 0.4628,
+      "step": 4523
+    },
+    {
+      "epoch": 1.099927060539752,
+      "grad_norm": 24.75,
+      "learning_rate": 1.0761080666864222e-06,
+      "loss": 0.8995,
+      "step": 4524
+    },
+    {
+      "epoch": 1.100170192073912,
+      "grad_norm": 15.625,
+      "learning_rate": 1.0756295162588586e-06,
+      "loss": 0.5974,
+      "step": 4525
+    },
+    {
+      "epoch": 1.100413323608072,
+      "grad_norm": 24.75,
+      "learning_rate": 1.0751509918939754e-06,
+      "loss": 1.1693,
+      "step": 4526
+    },
+    {
+      "epoch": 1.100656455142232,
+      "grad_norm": 25.375,
+      "learning_rate": 1.0746724936632966e-06,
+      "loss": 0.9175,
+      "step": 4527
+    },
+    {
+      "epoch": 1.100899586676392,
+      "grad_norm": 18.625,
+      "learning_rate": 1.0741940216383412e-06,
+      "loss": 0.6916,
+      "step": 4528
+    },
+    {
+      "epoch": 1.101142718210552,
+      "grad_norm": 17.5,
+      "learning_rate": 1.0737155758906258e-06,
+      "loss": 0.5722,
+      "step": 4529
+    },
+    {
+      "epoch": 1.101385849744712,
+      "grad_norm": 21.25,
+      "learning_rate": 1.0732371564916614e-06,
+      "loss": 1.0089,
+      "step": 4530
+    },
+    {
+      "epoch": 1.1016289812788718,
+      "grad_norm": 18.125,
+      "learning_rate": 1.0727587635129574e-06,
+      "loss": 0.6362,
+      "step": 4531
+    },
+    {
+      "epoch": 1.1018721128130318,
+      "grad_norm": 25.5,
+      "learning_rate": 1.072280397026017e-06,
+      "loss": 0.8854,
+      "step": 4532
+    },
+    {
+      "epoch": 1.1021152443471918,
+      "grad_norm": 19.25,
+      "learning_rate": 1.0718020571023408e-06,
+      "loss": 1.0526,
+      "step": 4533
+    },
+    {
+      "epoch": 1.1023583758813518,
+      "grad_norm": 22.625,
+      "learning_rate": 1.0713237438134249e-06,
+      "loss": 0.7364,
+      "step": 4534
+    },
+    {
+      "epoch": 1.1026015074155118,
+      "grad_norm": 18.25,
+      "learning_rate": 1.070845457230761e-06,
+      "loss": 0.8941,
+      "step": 4535
+    },
+    {
+      "epoch": 1.1028446389496718,
+      "grad_norm": 19.375,
+      "learning_rate": 1.0703671974258378e-06,
+      "loss": 0.8003,
+      "step": 4536
+    },
+    {
+      "epoch": 1.1030877704838318,
+      "grad_norm": 24.0,
+      "learning_rate": 1.0698889644701394e-06,
+      "loss": 0.2289,
+      "step": 4537
+    },
+    {
+      "epoch": 1.1033309020179918,
+      "grad_norm": 19.0,
+      "learning_rate": 1.0694107584351453e-06,
+      "loss": 0.7821,
+      "step": 4538
+    },
+    {
+      "epoch": 1.1035740335521518,
+      "grad_norm": 23.375,
+      "learning_rate": 1.0689325793923327e-06,
+      "loss": 0.7209,
+      "step": 4539
+    },
+    {
+      "epoch": 1.1038171650863118,
+      "grad_norm": 20.625,
+      "learning_rate": 1.0684544274131731e-06,
+      "loss": 0.9191,
+      "step": 4540
+    },
+    {
+      "epoch": 1.1040602966204718,
+      "grad_norm": 18.5,
+      "learning_rate": 1.0679763025691352e-06,
+      "loss": 0.7183,
+      "step": 4541
+    },
+    {
+      "epoch": 1.1043034281546316,
+      "grad_norm": 20.25,
+      "learning_rate": 1.0674982049316822e-06,
+      "loss": 0.779,
+      "step": 4542
+    },
+    {
+      "epoch": 1.1045465596887916,
+      "grad_norm": 19.75,
+      "learning_rate": 1.0670201345722742e-06,
+      "loss": 0.7229,
+      "step": 4543
+    },
+    {
+      "epoch": 1.1047896912229516,
+      "grad_norm": 23.625,
+      "learning_rate": 1.0665420915623678e-06,
+      "loss": 0.8702,
+      "step": 4544
+    },
+    {
+      "epoch": 1.1050328227571116,
+      "grad_norm": 22.0,
+      "learning_rate": 1.0660640759734137e-06,
+      "loss": 0.7513,
+      "step": 4545
+    },
+    {
+      "epoch": 1.1052759542912716,
+      "grad_norm": 23.625,
+      "learning_rate": 1.06558608787686e-06,
+      "loss": 0.8785,
+      "step": 4546
+    },
+    {
+      "epoch": 1.1055190858254316,
+      "grad_norm": 22.625,
+      "learning_rate": 1.0651081273441507e-06,
+      "loss": 1.1261,
+      "step": 4547
+    },
+    {
+      "epoch": 1.1057622173595916,
+      "grad_norm": 25.875,
+      "learning_rate": 1.0646301944467252e-06,
+      "loss": 0.692,
+      "step": 4548
+    },
+    {
+      "epoch": 1.1060053488937516,
+      "grad_norm": 21.125,
+      "learning_rate": 1.0641522892560184e-06,
+      "loss": 1.2272,
+      "step": 4549
+    },
+    {
+      "epoch": 1.1062484804279116,
+      "grad_norm": 22.0,
+      "learning_rate": 1.0636744118434615e-06,
+      "loss": 0.6876,
+      "step": 4550
+    },
+    {
+      "epoch": 1.1064916119620716,
+      "grad_norm": 18.75,
+      "learning_rate": 1.0631965622804821e-06,
+      "loss": 0.7962,
+      "step": 4551
+    },
+    {
+      "epoch": 1.1067347434962316,
+      "grad_norm": 17.125,
+      "learning_rate": 1.0627187406385023e-06,
+      "loss": 0.4454,
+      "step": 4552
+    },
+    {
+      "epoch": 1.1069778750303914,
+      "grad_norm": 22.75,
+      "learning_rate": 1.0622409469889413e-06,
+      "loss": 0.8664,
+      "step": 4553
+    },
+    {
+      "epoch": 1.1072210065645514,
+      "grad_norm": 19.125,
+      "learning_rate": 1.0617631814032132e-06,
+      "loss": 0.8129,
+      "step": 4554
+    },
+    {
+      "epoch": 1.1074641380987114,
+      "grad_norm": 32.5,
+      "learning_rate": 1.061285443952729e-06,
+      "loss": 0.8578,
+      "step": 4555
+    },
+    {
+      "epoch": 1.1077072696328714,
+      "grad_norm": 20.875,
+      "learning_rate": 1.0608077347088943e-06,
+      "loss": 0.7495,
+      "step": 4556
+    },
+    {
+      "epoch": 1.1079504011670314,
+      "grad_norm": 22.375,
+      "learning_rate": 1.060330053743111e-06,
+      "loss": 0.8623,
+      "step": 4557
+    },
+    {
+      "epoch": 1.1081935327011914,
+      "grad_norm": 16.875,
+      "learning_rate": 1.0598524011267771e-06,
+      "loss": 0.4749,
+      "step": 4558
+    },
+    {
+      "epoch": 1.1084366642353514,
+      "grad_norm": 19.5,
+      "learning_rate": 1.0593747769312858e-06,
+      "loss": 1.0007,
+      "step": 4559
+    },
+    {
+      "epoch": 1.1086797957695114,
+      "grad_norm": 18.125,
+      "learning_rate": 1.0588971812280261e-06,
+      "loss": 0.6778,
+      "step": 4560
+    },
+    {
+      "epoch": 1.1089229273036714,
+      "grad_norm": 21.5,
+      "learning_rate": 1.058419614088383e-06,
+      "loss": 0.7019,
+      "step": 4561
+    },
+    {
+      "epoch": 1.1091660588378314,
+      "grad_norm": 15.75,
+      "learning_rate": 1.0579420755837375e-06,
+      "loss": 0.5354,
+      "step": 4562
+    },
+    {
+      "epoch": 1.1094091903719911,
+      "grad_norm": 16.75,
+      "learning_rate": 1.0574645657854659e-06,
+      "loss": 0.6889,
+      "step": 4563
+    },
+    {
+      "epoch": 1.1096523219061512,
+      "grad_norm": 22.25,
+      "learning_rate": 1.05698708476494e-06,
+      "loss": 1.0348,
+      "step": 4564
+    },
+    {
+      "epoch": 1.1098954534403112,
+      "grad_norm": 16.0,
+      "learning_rate": 1.0565096325935278e-06,
+      "loss": 0.6408,
+      "step": 4565
+    },
+    {
+      "epoch": 1.1101385849744712,
+      "grad_norm": 25.0,
+      "learning_rate": 1.0560322093425926e-06,
+      "loss": 0.8725,
+      "step": 4566
+    },
+    {
+      "epoch": 1.1103817165086312,
+      "grad_norm": 23.75,
+      "learning_rate": 1.0555548150834937e-06,
+      "loss": 0.6499,
+      "step": 4567
+    },
+    {
+      "epoch": 1.1106248480427912,
+      "grad_norm": 24.25,
+      "learning_rate": 1.055077449887586e-06,
+      "loss": 0.9151,
+      "step": 4568
+    },
+    {
+      "epoch": 1.1108679795769512,
+      "grad_norm": 26.25,
+      "learning_rate": 1.0546001138262196e-06,
+      "loss": 0.9657,
+      "step": 4569
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 24.75,
+      "learning_rate": 1.0541228069707408e-06,
+      "loss": 0.9306,
+      "step": 4570
+    },
+    {
+      "epoch": 1.1113542426452712,
+      "grad_norm": 21.375,
+      "learning_rate": 1.0536455293924914e-06,
+      "loss": 0.6871,
+      "step": 4571
+    },
+    {
+      "epoch": 1.1115973741794312,
+      "grad_norm": 17.375,
+      "learning_rate": 1.0531682811628092e-06,
+      "loss": 0.6849,
+      "step": 4572
+    },
+    {
+      "epoch": 1.111840505713591,
+      "grad_norm": 24.0,
+      "learning_rate": 1.0526910623530267e-06,
+      "loss": 0.7795,
+      "step": 4573
+    },
+    {
+      "epoch": 1.112083637247751,
+      "grad_norm": 19.75,
+      "learning_rate": 1.0522138730344722e-06,
+      "loss": 0.389,
+      "step": 4574
+    },
+    {
+      "epoch": 1.112326768781911,
+      "grad_norm": 20.625,
+      "learning_rate": 1.0517367132784707e-06,
+      "loss": 0.8928,
+      "step": 4575
+    },
+    {
+      "epoch": 1.112569900316071,
+      "grad_norm": 20.625,
+      "learning_rate": 1.0512595831563413e-06,
+      "loss": 0.7636,
+      "step": 4576
+    },
+    {
+      "epoch": 1.112813031850231,
+      "grad_norm": 18.0,
+      "learning_rate": 1.0507824827393994e-06,
+      "loss": 0.7822,
+      "step": 4577
+    },
+    {
+      "epoch": 1.113056163384391,
+      "grad_norm": 21.25,
+      "learning_rate": 1.0503054120989562e-06,
+      "loss": 0.6797,
+      "step": 4578
+    },
+    {
+      "epoch": 1.113299294918551,
+      "grad_norm": 16.875,
+      "learning_rate": 1.0498283713063182e-06,
+      "loss": 0.5428,
+      "step": 4579
+    },
+    {
+      "epoch": 1.113542426452711,
+      "grad_norm": 14.1875,
+      "learning_rate": 1.0493513604327868e-06,
+      "loss": 0.4768,
+      "step": 4580
+    },
+    {
+      "epoch": 1.113785557986871,
+      "grad_norm": 17.125,
+      "learning_rate": 1.0488743795496602e-06,
+      "loss": 0.5572,
+      "step": 4581
+    },
+    {
+      "epoch": 1.114028689521031,
+      "grad_norm": 24.625,
+      "learning_rate": 1.0483974287282313e-06,
+      "loss": 0.7668,
+      "step": 4582
+    },
+    {
+      "epoch": 1.114271821055191,
+      "grad_norm": 16.0,
+      "learning_rate": 1.0479205080397881e-06,
+      "loss": 0.7331,
+      "step": 4583
+    },
+    {
+      "epoch": 1.1145149525893507,
+      "grad_norm": 19.625,
+      "learning_rate": 1.047443617555615e-06,
+      "loss": 0.5156,
+      "step": 4584
+    },
+    {
+      "epoch": 1.1147580841235107,
+      "grad_norm": 17.125,
+      "learning_rate": 1.0469667573469916e-06,
+      "loss": 0.8256,
+      "step": 4585
+    },
+    {
+      "epoch": 1.1150012156576707,
+      "grad_norm": 23.625,
+      "learning_rate": 1.046489927485193e-06,
+      "loss": 0.973,
+      "step": 4586
+    },
+    {
+      "epoch": 1.1152443471918307,
+      "grad_norm": 17.75,
+      "learning_rate": 1.0460131280414896e-06,
+      "loss": 0.6681,
+      "step": 4587
+    },
+    {
+      "epoch": 1.1154874787259907,
+      "grad_norm": 20.625,
+      "learning_rate": 1.0455363590871468e-06,
+      "loss": 0.6001,
+      "step": 4588
+    },
+    {
+      "epoch": 1.1157306102601507,
+      "grad_norm": 21.5,
+      "learning_rate": 1.0450596206934267e-06,
+      "loss": 0.7672,
+      "step": 4589
+    },
+    {
+      "epoch": 1.1159737417943107,
+      "grad_norm": 34.5,
+      "learning_rate": 1.0445829129315854e-06,
+      "loss": 0.6228,
+      "step": 4590
+    },
+    {
+      "epoch": 1.1162168733284707,
+      "grad_norm": 18.875,
+      "learning_rate": 1.0441062358728757e-06,
+      "loss": 0.5794,
+      "step": 4591
+    },
+    {
+      "epoch": 1.1164600048626308,
+      "grad_norm": 19.0,
+      "learning_rate": 1.0436295895885445e-06,
+      "loss": 0.934,
+      "step": 4592
+    },
+    {
+      "epoch": 1.1167031363967908,
+      "grad_norm": 16.625,
+      "learning_rate": 1.0431529741498358e-06,
+      "loss": 0.4576,
+      "step": 4593
+    },
+    {
+      "epoch": 1.1169462679309508,
+      "grad_norm": 17.25,
+      "learning_rate": 1.0426763896279873e-06,
+      "loss": 0.5815,
+      "step": 4594
+    },
+    {
+      "epoch": 1.1171893994651105,
+      "grad_norm": 20.625,
+      "learning_rate": 1.0421998360942328e-06,
+      "loss": 0.6244,
+      "step": 4595
+    },
+    {
+      "epoch": 1.1174325309992705,
+      "grad_norm": 19.75,
+      "learning_rate": 1.041723313619802e-06,
+      "loss": 0.9553,
+      "step": 4596
+    },
+    {
+      "epoch": 1.1176756625334305,
+      "grad_norm": 19.125,
+      "learning_rate": 1.041246822275919e-06,
+      "loss": 0.4917,
+      "step": 4597
+    },
+    {
+      "epoch": 1.1179187940675905,
+      "grad_norm": 21.5,
+      "learning_rate": 1.0407703621338034e-06,
+      "loss": 0.5445,
+      "step": 4598
+    },
+    {
+      "epoch": 1.1181619256017505,
+      "grad_norm": 18.5,
+      "learning_rate": 1.040293933264671e-06,
+      "loss": 0.7912,
+      "step": 4599
+    },
+    {
+      "epoch": 1.1184050571359105,
+      "grad_norm": 17.5,
+      "learning_rate": 1.0398175357397314e-06,
+      "loss": 0.6897,
+      "step": 4600
+    },
+    {
+      "epoch": 1.1186481886700705,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.0393411696301918e-06,
+      "loss": 0.7391,
+      "step": 4601
+    },
+    {
+      "epoch": 1.1188913202042305,
+      "grad_norm": 19.125,
+      "learning_rate": 1.0388648350072522e-06,
+      "loss": 0.6874,
+      "step": 4602
+    },
+    {
+      "epoch": 1.1191344517383905,
+      "grad_norm": 23.75,
+      "learning_rate": 1.0383885319421097e-06,
+      "loss": 0.9047,
+      "step": 4603
+    },
+    {
+      "epoch": 1.1193775832725505,
+      "grad_norm": 18.875,
+      "learning_rate": 1.0379122605059557e-06,
+      "loss": 0.7662,
+      "step": 4604
+    },
+    {
+      "epoch": 1.1196207148067103,
+      "grad_norm": 16.75,
+      "learning_rate": 1.0374360207699771e-06,
+      "loss": 0.4598,
+      "step": 4605
+    },
+    {
+      "epoch": 1.1198638463408703,
+      "grad_norm": 22.75,
+      "learning_rate": 1.0369598128053565e-06,
+      "loss": 0.6787,
+      "step": 4606
+    },
+    {
+      "epoch": 1.1201069778750303,
+      "grad_norm": 18.125,
+      "learning_rate": 1.036483636683271e-06,
+      "loss": 0.619,
+      "step": 4607
+    },
+    {
+      "epoch": 1.1203501094091903,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.0360074924748934e-06,
+      "loss": 0.3686,
+      "step": 4608
+    },
+    {
+      "epoch": 1.1205932409433503,
+      "grad_norm": 18.5,
+      "learning_rate": 1.0355313802513922e-06,
+      "loss": 0.5171,
+      "step": 4609
+    },
+    {
+      "epoch": 1.1208363724775103,
+      "grad_norm": 19.125,
+      "learning_rate": 1.03505530008393e-06,
+      "loss": 0.8641,
+      "step": 4610
+    },
+    {
+      "epoch": 1.1210795040116703,
+      "grad_norm": 20.5,
+      "learning_rate": 1.0345792520436657e-06,
+      "loss": 0.6632,
+      "step": 4611
+    },
+    {
+      "epoch": 1.1213226355458303,
+      "grad_norm": 17.75,
+      "learning_rate": 1.0341032362017523e-06,
+      "loss": 0.8097,
+      "step": 4612
+    },
+    {
+      "epoch": 1.1215657670799903,
+      "grad_norm": 25.0,
+      "learning_rate": 1.0336272526293392e-06,
+      "loss": 0.5701,
+      "step": 4613
+    },
+    {
+      "epoch": 1.1218088986141503,
+      "grad_norm": 23.875,
+      "learning_rate": 1.03315130139757e-06,
+      "loss": 0.8333,
+      "step": 4614
+    },
+    {
+      "epoch": 1.1220520301483101,
+      "grad_norm": 20.5,
+      "learning_rate": 1.0326753825775837e-06,
+      "loss": 0.5231,
+      "step": 4615
+    },
+    {
+      "epoch": 1.1222951616824701,
+      "grad_norm": 17.25,
+      "learning_rate": 1.0321994962405147e-06,
+      "loss": 0.7017,
+      "step": 4616
+    },
+    {
+      "epoch": 1.1225382932166301,
+      "grad_norm": 19.0,
+      "learning_rate": 1.0317236424574929e-06,
+      "loss": 0.7746,
+      "step": 4617
+    },
+    {
+      "epoch": 1.1227814247507901,
+      "grad_norm": 22.375,
+      "learning_rate": 1.0312478212996425e-06,
+      "loss": 0.6846,
+      "step": 4618
+    },
+    {
+      "epoch": 1.1230245562849501,
+      "grad_norm": 17.125,
+      "learning_rate": 1.030772032838083e-06,
+      "loss": 0.5556,
+      "step": 4619
+    },
+    {
+      "epoch": 1.1232676878191101,
+      "grad_norm": 20.875,
+      "learning_rate": 1.0302962771439296e-06,
+      "loss": 0.8408,
+      "step": 4620
+    },
+    {
+      "epoch": 1.1235108193532701,
+      "grad_norm": 27.5,
+      "learning_rate": 1.029820554288292e-06,
+      "loss": 0.6969,
+      "step": 4621
+    },
+    {
+      "epoch": 1.1237539508874301,
+      "grad_norm": 16.375,
+      "learning_rate": 1.029344864342275e-06,
+      "loss": 0.5205,
+      "step": 4622
+    },
+    {
+      "epoch": 1.1239970824215901,
+      "grad_norm": 16.625,
+      "learning_rate": 1.0288692073769786e-06,
+      "loss": 0.4991,
+      "step": 4623
+    },
+    {
+      "epoch": 1.1242402139557501,
+      "grad_norm": 19.25,
+      "learning_rate": 1.0283935834634984e-06,
+      "loss": 0.6053,
+      "step": 4624
+    },
+    {
+      "epoch": 1.1244833454899101,
+      "grad_norm": 19.75,
+      "learning_rate": 1.0279179926729249e-06,
+      "loss": 0.8172,
+      "step": 4625
+    },
+    {
+      "epoch": 1.12472647702407,
+      "grad_norm": 13.3125,
+      "learning_rate": 1.0274424350763424e-06,
+      "loss": 0.2787,
+      "step": 4626
+    },
+    {
+      "epoch": 1.12496960855823,
+      "grad_norm": 18.875,
+      "learning_rate": 1.026966910744832e-06,
+      "loss": 0.6541,
+      "step": 4627
+    },
+    {
+      "epoch": 1.12521274009239,
+      "grad_norm": 17.125,
+      "learning_rate": 1.0264914197494685e-06,
+      "loss": 0.4244,
+      "step": 4628
+    },
+    {
+      "epoch": 1.12545587162655,
+      "grad_norm": 34.25,
+      "learning_rate": 1.0260159621613224e-06,
+      "loss": 0.6205,
+      "step": 4629
+    },
+    {
+      "epoch": 1.12569900316071,
+      "grad_norm": 19.625,
+      "learning_rate": 1.0255405380514594e-06,
+      "loss": 0.7377,
+      "step": 4630
+    },
+    {
+      "epoch": 1.12594213469487,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.0250651474909386e-06,
+      "loss": 0.4619,
+      "step": 4631
+    },
+    {
+      "epoch": 1.12618526622903,
+      "grad_norm": 26.0,
+      "learning_rate": 1.0245897905508174e-06,
+      "loss": 0.8511,
+      "step": 4632
+    },
+    {
+      "epoch": 1.12642839776319,
+      "grad_norm": 25.5,
+      "learning_rate": 1.0241144673021444e-06,
+      "loss": 0.5489,
+      "step": 4633
+    },
+    {
+      "epoch": 1.12667152929735,
+      "grad_norm": 21.5,
+      "learning_rate": 1.0236391778159658e-06,
+      "loss": 0.6719,
+      "step": 4634
+    },
+    {
+      "epoch": 1.12691466083151,
+      "grad_norm": 19.375,
+      "learning_rate": 1.0231639221633213e-06,
+      "loss": 0.459,
+      "step": 4635
+    },
+    {
+      "epoch": 1.12715779236567,
+      "grad_norm": 16.5,
+      "learning_rate": 1.022688700415246e-06,
+      "loss": 0.7492,
+      "step": 4636
+    },
+    {
+      "epoch": 1.1274009238998297,
+      "grad_norm": 19.875,
+      "learning_rate": 1.0222135126427708e-06,
+      "loss": 0.7292,
+      "step": 4637
+    },
+    {
+      "epoch": 1.1276440554339897,
+      "grad_norm": 22.0,
+      "learning_rate": 1.0217383589169196e-06,
+      "loss": 0.6016,
+      "step": 4638
+    },
+    {
+      "epoch": 1.1278871869681497,
+      "grad_norm": 21.375,
+      "learning_rate": 1.0212632393087126e-06,
+      "loss": 0.8727,
+      "step": 4639
+    },
+    {
+      "epoch": 1.1281303185023097,
+      "grad_norm": 18.75,
+      "learning_rate": 1.0207881538891654e-06,
+      "loss": 0.7021,
+      "step": 4640
+    },
+    {
+      "epoch": 1.1283734500364697,
+      "grad_norm": 18.25,
+      "learning_rate": 1.020313102729287e-06,
+      "loss": 0.7066,
+      "step": 4641
+    },
+    {
+      "epoch": 1.1286165815706297,
+      "grad_norm": 17.5,
+      "learning_rate": 1.0198380859000828e-06,
+      "loss": 0.6684,
+      "step": 4642
+    },
+    {
+      "epoch": 1.1288597131047897,
+      "grad_norm": 25.125,
+      "learning_rate": 1.019363103472551e-06,
+      "loss": 0.5839,
+      "step": 4643
+    },
+    {
+      "epoch": 1.1291028446389497,
+      "grad_norm": 21.0,
+      "learning_rate": 1.018888155517687e-06,
+      "loss": 0.7214,
+      "step": 4644
+    },
+    {
+      "epoch": 1.1293459761731097,
+      "grad_norm": 23.5,
+      "learning_rate": 1.0184132421064797e-06,
+      "loss": 0.6538,
+      "step": 4645
+    },
+    {
+      "epoch": 1.1295891077072697,
+      "grad_norm": 18.375,
+      "learning_rate": 1.0179383633099124e-06,
+      "loss": 0.5012,
+      "step": 4646
+    },
+    {
+      "epoch": 1.1298322392414297,
+      "grad_norm": 19.75,
+      "learning_rate": 1.017463519198965e-06,
+      "loss": 0.7287,
+      "step": 4647
+    },
+    {
+      "epoch": 1.1300753707755895,
+      "grad_norm": 20.5,
+      "learning_rate": 1.0169887098446106e-06,
+      "loss": 0.6575,
+      "step": 4648
+    },
+    {
+      "epoch": 1.1303185023097495,
+      "grad_norm": 19.875,
+      "learning_rate": 1.016513935317818e-06,
+      "loss": 0.6859,
+      "step": 4649
+    },
+    {
+      "epoch": 1.1305616338439095,
+      "grad_norm": 31.75,
+      "learning_rate": 1.01603919568955e-06,
+      "loss": 1.0938,
+      "step": 4650
+    },
+    {
+      "epoch": 1.1308047653780695,
+      "grad_norm": 19.875,
+      "learning_rate": 1.015564491030765e-06,
+      "loss": 0.7796,
+      "step": 4651
+    },
+    {
+      "epoch": 1.1310478969122295,
+      "grad_norm": 19.0,
+      "learning_rate": 1.0150898214124155e-06,
+      "loss": 0.6401,
+      "step": 4652
+    },
+    {
+      "epoch": 1.1312910284463895,
+      "grad_norm": 19.625,
+      "learning_rate": 1.0146151869054492e-06,
+      "loss": 0.9332,
+      "step": 4653
+    },
+    {
+      "epoch": 1.1315341599805495,
+      "grad_norm": 25.375,
+      "learning_rate": 1.0141405875808083e-06,
+      "loss": 0.8476,
+      "step": 4654
+    },
+    {
+      "epoch": 1.1317772915147095,
+      "grad_norm": 20.75,
+      "learning_rate": 1.01366602350943e-06,
+      "loss": 1.0623,
+      "step": 4655
+    },
+    {
+      "epoch": 1.1320204230488695,
+      "grad_norm": 15.25,
+      "learning_rate": 1.0131914947622466e-06,
+      "loss": 0.6053,
+      "step": 4656
+    },
+    {
+      "epoch": 1.1322635545830293,
+      "grad_norm": 17.5,
+      "learning_rate": 1.012717001410184e-06,
+      "loss": 0.6007,
+      "step": 4657
+    },
+    {
+      "epoch": 1.1325066861171895,
+      "grad_norm": 18.75,
+      "learning_rate": 1.0122425435241633e-06,
+      "loss": 0.7136,
+      "step": 4658
+    },
+    {
+      "epoch": 1.1327498176513493,
+      "grad_norm": 20.375,
+      "learning_rate": 1.011768121175101e-06,
+      "loss": 0.5298,
+      "step": 4659
+    },
+    {
+      "epoch": 1.1329929491855093,
+      "grad_norm": 25.25,
+      "learning_rate": 1.0112937344339071e-06,
+      "loss": 0.7498,
+      "step": 4660
+    },
+    {
+      "epoch": 1.1332360807196693,
+      "grad_norm": 18.375,
+      "learning_rate": 1.0108193833714875e-06,
+      "loss": 0.5374,
+      "step": 4661
+    },
+    {
+      "epoch": 1.1334792122538293,
+      "grad_norm": 23.625,
+      "learning_rate": 1.0103450680587412e-06,
+      "loss": 0.7976,
+      "step": 4662
+    },
+    {
+      "epoch": 1.1337223437879893,
+      "grad_norm": 24.75,
+      "learning_rate": 1.009870788566564e-06,
+      "loss": 0.7245,
+      "step": 4663
+    },
+    {
+      "epoch": 1.1339654753221493,
+      "grad_norm": 23.125,
+      "learning_rate": 1.0093965449658445e-06,
+      "loss": 0.8302,
+      "step": 4664
+    },
+    {
+      "epoch": 1.1342086068563093,
+      "grad_norm": 17.5,
+      "learning_rate": 1.0089223373274668e-06,
+      "loss": 0.5245,
+      "step": 4665
+    },
+    {
+      "epoch": 1.1344517383904693,
+      "grad_norm": 16.125,
+      "learning_rate": 1.0084481657223093e-06,
+      "loss": 0.2544,
+      "step": 4666
+    },
+    {
+      "epoch": 1.1346948699246293,
+      "grad_norm": 23.5,
+      "learning_rate": 1.0079740302212452e-06,
+      "loss": 0.9773,
+      "step": 4667
+    },
+    {
+      "epoch": 1.134938001458789,
+      "grad_norm": 17.125,
+      "learning_rate": 1.0074999308951426e-06,
+      "loss": 0.3679,
+      "step": 4668
+    },
+    {
+      "epoch": 1.135181132992949,
+      "grad_norm": 20.375,
+      "learning_rate": 1.0070258678148624e-06,
+      "loss": 0.7059,
+      "step": 4669
+    },
+    {
+      "epoch": 1.135424264527109,
+      "grad_norm": 22.5,
+      "learning_rate": 1.0065518410512634e-06,
+      "loss": 0.6372,
+      "step": 4670
+    },
+    {
+      "epoch": 1.135667396061269,
+      "grad_norm": 29.625,
+      "learning_rate": 1.006077850675196e-06,
+      "loss": 0.7484,
+      "step": 4671
+    },
+    {
+      "epoch": 1.135910527595429,
+      "grad_norm": 25.875,
+      "learning_rate": 1.0056038967575065e-06,
+      "loss": 0.9291,
+      "step": 4672
+    },
+    {
+      "epoch": 1.136153659129589,
+      "grad_norm": 17.875,
+      "learning_rate": 1.0051299793690359e-06,
+      "loss": 0.6821,
+      "step": 4673
+    },
+    {
+      "epoch": 1.136396790663749,
+      "grad_norm": 18.75,
+      "learning_rate": 1.0046560985806183e-06,
+      "loss": 0.679,
+      "step": 4674
+    },
+    {
+      "epoch": 1.136639922197909,
+      "grad_norm": 11.9375,
+      "learning_rate": 1.0041822544630844e-06,
+      "loss": 0.4055,
+      "step": 4675
+    },
+    {
+      "epoch": 1.136883053732069,
+      "grad_norm": 21.25,
+      "learning_rate": 1.003708447087258e-06,
+      "loss": 0.3208,
+      "step": 4676
+    },
+    {
+      "epoch": 1.137126185266229,
+      "grad_norm": 18.25,
+      "learning_rate": 1.0032346765239574e-06,
+      "loss": 0.7156,
+      "step": 4677
+    },
+    {
+      "epoch": 1.137369316800389,
+      "grad_norm": 17.125,
+      "learning_rate": 1.0027609428439963e-06,
+      "loss": 0.6838,
+      "step": 4678
+    },
+    {
+      "epoch": 1.1376124483345489,
+      "grad_norm": 18.25,
+      "learning_rate": 1.0022872461181823e-06,
+      "loss": 0.3967,
+      "step": 4679
+    },
+    {
+      "epoch": 1.1378555798687089,
+      "grad_norm": 17.0,
+      "learning_rate": 1.001813586417318e-06,
+      "loss": 0.6606,
+      "step": 4680
+    },
+    {
+      "epoch": 1.138098711402869,
+      "grad_norm": 20.875,
+      "learning_rate": 1.001339963812199e-06,
+      "loss": 1.1108,
+      "step": 4681
+    },
+    {
+      "epoch": 1.138341842937029,
+      "grad_norm": 27.125,
+      "learning_rate": 1.0008663783736172e-06,
+      "loss": 0.7837,
+      "step": 4682
+    },
+    {
+      "epoch": 1.138584974471189,
+      "grad_norm": 22.125,
+      "learning_rate": 1.0003928301723579e-06,
+      "loss": 0.4921,
+      "step": 4683
+    },
+    {
+      "epoch": 1.138828106005349,
+      "grad_norm": 23.375,
+      "learning_rate": 9.999193192792005e-07,
+      "loss": 0.837,
+      "step": 4684
+    },
+    {
+      "epoch": 1.139071237539509,
+      "grad_norm": 39.0,
+      "learning_rate": 9.9944584576492e-07,
+      "loss": 1.2307,
+      "step": 4685
+    },
+    {
+      "epoch": 1.139314369073669,
+      "grad_norm": 42.75,
+      "learning_rate": 9.989724097002852e-07,
+      "loss": 1.0931,
+      "step": 4686
+    },
+    {
+      "epoch": 1.139557500607829,
+      "grad_norm": 23.5,
+      "learning_rate": 9.984990111560594e-07,
+      "loss": 0.8322,
+      "step": 4687
+    },
+    {
+      "epoch": 1.139800632141989,
+      "grad_norm": 20.875,
+      "learning_rate": 9.980256502029998e-07,
+      "loss": 0.9468,
+      "step": 4688
+    },
+    {
+      "epoch": 1.140043763676149,
+      "grad_norm": 18.375,
+      "learning_rate": 9.975523269118583e-07,
+      "loss": 0.623,
+      "step": 4689
+    },
+    {
+      "epoch": 1.1402868952103087,
+      "grad_norm": 24.0,
+      "learning_rate": 9.97079041353382e-07,
+      "loss": 0.8993,
+      "step": 4690
+    },
+    {
+      "epoch": 1.1405300267444687,
+      "grad_norm": 21.25,
+      "learning_rate": 9.966057935983107e-07,
+      "loss": 0.8847,
+      "step": 4691
+    },
+    {
+      "epoch": 1.1407731582786287,
+      "grad_norm": 22.75,
+      "learning_rate": 9.961325837173797e-07,
+      "loss": 0.9124,
+      "step": 4692
+    },
+    {
+      "epoch": 1.1410162898127887,
+      "grad_norm": 28.5,
+      "learning_rate": 9.956594117813183e-07,
+      "loss": 0.5026,
+      "step": 4693
+    },
+    {
+      "epoch": 1.1412594213469487,
+      "grad_norm": 37.75,
+      "learning_rate": 9.951862778608511e-07,
+      "loss": 1.2797,
+      "step": 4694
+    },
+    {
+      "epoch": 1.1415025528811087,
+      "grad_norm": 15.9375,
+      "learning_rate": 9.94713182026695e-07,
+      "loss": 0.7834,
+      "step": 4695
+    },
+    {
+      "epoch": 1.1417456844152687,
+      "grad_norm": 20.25,
+      "learning_rate": 9.942401243495629e-07,
+      "loss": 0.5828,
+      "step": 4696
+    },
+    {
+      "epoch": 1.1419888159494287,
+      "grad_norm": 23.25,
+      "learning_rate": 9.937671049001616e-07,
+      "loss": 0.8871,
+      "step": 4697
+    },
+    {
+      "epoch": 1.1422319474835887,
+      "grad_norm": 17.375,
+      "learning_rate": 9.932941237491916e-07,
+      "loss": 0.617,
+      "step": 4698
+    },
+    {
+      "epoch": 1.1424750790177487,
+      "grad_norm": 24.375,
+      "learning_rate": 9.92821180967348e-07,
+      "loss": 0.4949,
+      "step": 4699
+    },
+    {
+      "epoch": 1.1427182105519087,
+      "grad_norm": 22.875,
+      "learning_rate": 9.923482766253204e-07,
+      "loss": 0.7303,
+      "step": 4700
+    },
+    {
+      "epoch": 1.1429613420860685,
+      "grad_norm": 16.625,
+      "learning_rate": 9.918754107937933e-07,
+      "loss": 0.5431,
+      "step": 4701
+    },
+    {
+      "epoch": 1.1432044736202285,
+      "grad_norm": 26.25,
+      "learning_rate": 9.914025835434435e-07,
+      "loss": 0.9557,
+      "step": 4702
+    },
+    {
+      "epoch": 1.1434476051543885,
+      "grad_norm": 11.6875,
+      "learning_rate": 9.90929794944944e-07,
+      "loss": 0.5551,
+      "step": 4703
+    },
+    {
+      "epoch": 1.1436907366885485,
+      "grad_norm": 24.375,
+      "learning_rate": 9.904570450689612e-07,
+      "loss": 0.9468,
+      "step": 4704
+    },
+    {
+      "epoch": 1.1439338682227085,
+      "grad_norm": 23.75,
+      "learning_rate": 9.89984333986155e-07,
+      "loss": 1.0116,
+      "step": 4705
+    },
+    {
+      "epoch": 1.1441769997568685,
+      "grad_norm": 17.375,
+      "learning_rate": 9.89511661767181e-07,
+      "loss": 0.6854,
+      "step": 4706
+    },
+    {
+      "epoch": 1.1444201312910285,
+      "grad_norm": 17.375,
+      "learning_rate": 9.89039028482688e-07,
+      "loss": 0.4921,
+      "step": 4707
+    },
+    {
+      "epoch": 1.1446632628251885,
+      "grad_norm": 12.5,
+      "learning_rate": 9.885664342033186e-07,
+      "loss": 0.4337,
+      "step": 4708
+    },
+    {
+      "epoch": 1.1449063943593485,
+      "grad_norm": 18.875,
+      "learning_rate": 9.88093878999711e-07,
+      "loss": 0.33,
+      "step": 4709
+    },
+    {
+      "epoch": 1.1451495258935083,
+      "grad_norm": 19.625,
+      "learning_rate": 9.876213629424966e-07,
+      "loss": 0.4763,
+      "step": 4710
+    },
+    {
+      "epoch": 1.1453926574276685,
+      "grad_norm": 24.0,
+      "learning_rate": 9.87148886102301e-07,
+      "loss": 1.1494,
+      "step": 4711
+    },
+    {
+      "epoch": 1.1456357889618283,
+      "grad_norm": 17.0,
+      "learning_rate": 9.866764485497438e-07,
+      "loss": 0.5958,
+      "step": 4712
+    },
+    {
+      "epoch": 1.1458789204959883,
+      "grad_norm": 17.0,
+      "learning_rate": 9.862040503554392e-07,
+      "loss": 0.8135,
+      "step": 4713
+    },
+    {
+      "epoch": 1.1461220520301483,
+      "grad_norm": 15.9375,
+      "learning_rate": 9.857316915899953e-07,
+      "loss": 0.4461,
+      "step": 4714
+    },
+    {
+      "epoch": 1.1463651835643083,
+      "grad_norm": 20.875,
+      "learning_rate": 9.852593723240142e-07,
+      "loss": 0.9367,
+      "step": 4715
+    },
+    {
+      "epoch": 1.1466083150984683,
+      "grad_norm": 19.875,
+      "learning_rate": 9.847870926280915e-07,
+      "loss": 0.7702,
+      "step": 4716
+    },
+    {
+      "epoch": 1.1468514466326283,
+      "grad_norm": 16.25,
+      "learning_rate": 9.843148525728187e-07,
+      "loss": 0.5595,
+      "step": 4717
+    },
+    {
+      "epoch": 1.1470945781667883,
+      "grad_norm": 21.125,
+      "learning_rate": 9.838426522287797e-07,
+      "loss": 0.7565,
+      "step": 4718
+    },
+    {
+      "epoch": 1.1473377097009483,
+      "grad_norm": 20.125,
+      "learning_rate": 9.83370491666553e-07,
+      "loss": 0.7135,
+      "step": 4719
+    },
+    {
+      "epoch": 1.1475808412351083,
+      "grad_norm": 18.0,
+      "learning_rate": 9.828983709567112e-07,
+      "loss": 1.0726,
+      "step": 4720
+    },
+    {
+      "epoch": 1.147823972769268,
+      "grad_norm": 13.5,
+      "learning_rate": 9.82426290169821e-07,
+      "loss": 0.3294,
+      "step": 4721
+    },
+    {
+      "epoch": 1.148067104303428,
+      "grad_norm": 22.375,
+      "learning_rate": 9.819542493764427e-07,
+      "loss": 0.8569,
+      "step": 4722
+    },
+    {
+      "epoch": 1.148310235837588,
+      "grad_norm": 18.25,
+      "learning_rate": 9.81482248647131e-07,
+      "loss": 0.5157,
+      "step": 4723
+    },
+    {
+      "epoch": 1.148553367371748,
+      "grad_norm": 25.25,
+      "learning_rate": 9.810102880524348e-07,
+      "loss": 0.9876,
+      "step": 4724
+    },
+    {
+      "epoch": 1.148796498905908,
+      "grad_norm": 18.75,
+      "learning_rate": 9.805383676628968e-07,
+      "loss": 0.8455,
+      "step": 4725
+    },
+    {
+      "epoch": 1.149039630440068,
+      "grad_norm": 19.5,
+      "learning_rate": 9.800664875490533e-07,
+      "loss": 0.5097,
+      "step": 4726
+    },
+    {
+      "epoch": 1.149282761974228,
+      "grad_norm": 22.875,
+      "learning_rate": 9.795946477814352e-07,
+      "loss": 0.702,
+      "step": 4727
+    },
+    {
+      "epoch": 1.149525893508388,
+      "grad_norm": 16.625,
+      "learning_rate": 9.791228484305675e-07,
+      "loss": 0.4389,
+      "step": 4728
+    },
+    {
+      "epoch": 1.149769025042548,
+      "grad_norm": 19.5,
+      "learning_rate": 9.786510895669678e-07,
+      "loss": 0.6537,
+      "step": 4729
+    },
+    {
+      "epoch": 1.150012156576708,
+      "grad_norm": 22.25,
+      "learning_rate": 9.781793712611492e-07,
+      "loss": 0.6354,
+      "step": 4730
+    },
+    {
+      "epoch": 1.150255288110868,
+      "grad_norm": 17.0,
+      "learning_rate": 9.77707693583618e-07,
+      "loss": 0.7206,
+      "step": 4731
+    },
+    {
+      "epoch": 1.1504984196450279,
+      "grad_norm": 17.875,
+      "learning_rate": 9.77236056604875e-07,
+      "loss": 0.717,
+      "step": 4732
+    },
+    {
+      "epoch": 1.1507415511791879,
+      "grad_norm": 15.5625,
+      "learning_rate": 9.767644603954138e-07,
+      "loss": 0.5995,
+      "step": 4733
+    },
+    {
+      "epoch": 1.1509846827133479,
+      "grad_norm": 20.25,
+      "learning_rate": 9.76292905025723e-07,
+      "loss": 0.617,
+      "step": 4734
+    },
+    {
+      "epoch": 1.1512278142475079,
+      "grad_norm": 12.625,
+      "learning_rate": 9.75821390566285e-07,
+      "loss": 0.4788,
+      "step": 4735
+    },
+    {
+      "epoch": 1.1514709457816679,
+      "grad_norm": 17.875,
+      "learning_rate": 9.75349917087575e-07,
+      "loss": 0.7572,
+      "step": 4736
+    },
+    {
+      "epoch": 1.1517140773158279,
+      "grad_norm": 21.875,
+      "learning_rate": 9.748784846600634e-07,
+      "loss": 1.1137,
+      "step": 4737
+    },
+    {
+      "epoch": 1.1519572088499879,
+      "grad_norm": 22.5,
+      "learning_rate": 9.744070933542139e-07,
+      "loss": 1.0709,
+      "step": 4738
+    },
+    {
+      "epoch": 1.1522003403841479,
+      "grad_norm": 19.125,
+      "learning_rate": 9.739357432404833e-07,
+      "loss": 0.8677,
+      "step": 4739
+    },
+    {
+      "epoch": 1.1524434719183079,
+      "grad_norm": 17.5,
+      "learning_rate": 9.734644343893243e-07,
+      "loss": 0.4708,
+      "step": 4740
+    },
+    {
+      "epoch": 1.1526866034524679,
+      "grad_norm": 21.375,
+      "learning_rate": 9.729931668711815e-07,
+      "loss": 0.461,
+      "step": 4741
+    },
+    {
+      "epoch": 1.1529297349866279,
+      "grad_norm": 16.25,
+      "learning_rate": 9.72521940756494e-07,
+      "loss": 0.5278,
+      "step": 4742
+    },
+    {
+      "epoch": 1.1531728665207877,
+      "grad_norm": 19.875,
+      "learning_rate": 9.720507561156944e-07,
+      "loss": 0.7804,
+      "step": 4743
+    },
+    {
+      "epoch": 1.1534159980549477,
+      "grad_norm": 18.625,
+      "learning_rate": 9.715796130192099e-07,
+      "loss": 0.7236,
+      "step": 4744
+    },
+    {
+      "epoch": 1.1536591295891077,
+      "grad_norm": 17.875,
+      "learning_rate": 9.711085115374608e-07,
+      "loss": 0.5271,
+      "step": 4745
+    },
+    {
+      "epoch": 1.1539022611232677,
+      "grad_norm": 22.625,
+      "learning_rate": 9.706374517408608e-07,
+      "loss": 0.5795,
+      "step": 4746
+    },
+    {
+      "epoch": 1.1541453926574277,
+      "grad_norm": 24.0,
+      "learning_rate": 9.701664336998183e-07,
+      "loss": 0.6542,
+      "step": 4747
+    },
+    {
+      "epoch": 1.1543885241915877,
+      "grad_norm": 21.125,
+      "learning_rate": 9.696954574847357e-07,
+      "loss": 1.0421,
+      "step": 4748
+    },
+    {
+      "epoch": 1.1546316557257477,
+      "grad_norm": 23.125,
+      "learning_rate": 9.69224523166008e-07,
+      "loss": 0.575,
+      "step": 4749
+    },
+    {
+      "epoch": 1.1548747872599077,
+      "grad_norm": 18.75,
+      "learning_rate": 9.687536308140244e-07,
+      "loss": 0.6088,
+      "step": 4750
+    },
+    {
+      "epoch": 1.1551179187940677,
+      "grad_norm": 24.5,
+      "learning_rate": 9.682827804991679e-07,
+      "loss": 0.759,
+      "step": 4751
+    },
+    {
+      "epoch": 1.1553610503282277,
+      "grad_norm": 17.25,
+      "learning_rate": 9.678119722918154e-07,
+      "loss": 0.4644,
+      "step": 4752
+    },
+    {
+      "epoch": 1.1556041818623877,
+      "grad_norm": 14.1875,
+      "learning_rate": 9.673412062623371e-07,
+      "loss": 0.3513,
+      "step": 4753
+    },
+    {
+      "epoch": 1.1558473133965474,
+      "grad_norm": 20.625,
+      "learning_rate": 9.668704824810968e-07,
+      "loss": 0.5674,
+      "step": 4754
+    },
+    {
+      "epoch": 1.1560904449307075,
+      "grad_norm": 20.125,
+      "learning_rate": 9.66399801018453e-07,
+      "loss": 0.5522,
+      "step": 4755
+    },
+    {
+      "epoch": 1.1563335764648675,
+      "grad_norm": 22.875,
+      "learning_rate": 9.65929161944757e-07,
+      "loss": 0.7269,
+      "step": 4756
+    },
+    {
+      "epoch": 1.1565767079990275,
+      "grad_norm": 21.375,
+      "learning_rate": 9.65458565330354e-07,
+      "loss": 0.7541,
+      "step": 4757
+    },
+    {
+      "epoch": 1.1568198395331875,
+      "grad_norm": 15.625,
+      "learning_rate": 9.649880112455823e-07,
+      "loss": 0.4317,
+      "step": 4758
+    },
+    {
+      "epoch": 1.1570629710673475,
+      "grad_norm": 19.0,
+      "learning_rate": 9.64517499760775e-07,
+      "loss": 0.426,
+      "step": 4759
+    },
+    {
+      "epoch": 1.1573061026015075,
+      "grad_norm": 19.875,
+      "learning_rate": 9.640470309462575e-07,
+      "loss": 0.7548,
+      "step": 4760
+    },
+    {
+      "epoch": 1.1575492341356675,
+      "grad_norm": 22.25,
+      "learning_rate": 9.635766048723497e-07,
+      "loss": 0.7269,
+      "step": 4761
+    },
+    {
+      "epoch": 1.1577923656698275,
+      "grad_norm": 18.875,
+      "learning_rate": 9.63106221609365e-07,
+      "loss": 0.8873,
+      "step": 4762
+    },
+    {
+      "epoch": 1.1580354972039872,
+      "grad_norm": 19.75,
+      "learning_rate": 9.626358812276104e-07,
+      "loss": 0.6826,
+      "step": 4763
+    },
+    {
+      "epoch": 1.1582786287381472,
+      "grad_norm": 26.875,
+      "learning_rate": 9.621655837973865e-07,
+      "loss": 0.7798,
+      "step": 4764
+    },
+    {
+      "epoch": 1.1585217602723072,
+      "grad_norm": 18.25,
+      "learning_rate": 9.616953293889871e-07,
+      "loss": 0.6742,
+      "step": 4765
+    },
+    {
+      "epoch": 1.1587648918064672,
+      "grad_norm": 26.625,
+      "learning_rate": 9.612251180727e-07,
+      "loss": 1.0155,
+      "step": 4766
+    },
+    {
+      "epoch": 1.1590080233406272,
+      "grad_norm": 17.75,
+      "learning_rate": 9.607549499188062e-07,
+      "loss": 0.3763,
+      "step": 4767
+    },
+    {
+      "epoch": 1.1592511548747872,
+      "grad_norm": 24.375,
+      "learning_rate": 9.602848249975805e-07,
+      "loss": 0.9412,
+      "step": 4768
+    },
+    {
+      "epoch": 1.1594942864089473,
+      "grad_norm": 30.25,
+      "learning_rate": 9.598147433792915e-07,
+      "loss": 0.855,
+      "step": 4769
+    },
+    {
+      "epoch": 1.1597374179431073,
+      "grad_norm": 14.75,
+      "learning_rate": 9.593447051342e-07,
+      "loss": 0.6047,
+      "step": 4770
+    },
+    {
+      "epoch": 1.1599805494772673,
+      "grad_norm": 16.75,
+      "learning_rate": 9.58874710332563e-07,
+      "loss": 0.4301,
+      "step": 4771
+    },
+    {
+      "epoch": 1.1602236810114273,
+      "grad_norm": 18.125,
+      "learning_rate": 9.58404759044628e-07,
+      "loss": 0.5674,
+      "step": 4772
+    },
+    {
+      "epoch": 1.1604668125455873,
+      "grad_norm": 21.625,
+      "learning_rate": 9.57934851340638e-07,
+      "loss": 0.7534,
+      "step": 4773
+    },
+    {
+      "epoch": 1.160709944079747,
+      "grad_norm": 22.125,
+      "learning_rate": 9.574649872908286e-07,
+      "loss": 0.9484,
+      "step": 4774
+    },
+    {
+      "epoch": 1.160953075613907,
+      "grad_norm": 21.625,
+      "learning_rate": 9.56995166965429e-07,
+      "loss": 0.9059,
+      "step": 4775
+    },
+    {
+      "epoch": 1.161196207148067,
+      "grad_norm": 31.0,
+      "learning_rate": 9.565253904346624e-07,
+      "loss": 0.9156,
+      "step": 4776
+    },
+    {
+      "epoch": 1.161439338682227,
+      "grad_norm": 23.875,
+      "learning_rate": 9.560556577687445e-07,
+      "loss": 0.563,
+      "step": 4777
+    },
+    {
+      "epoch": 1.161682470216387,
+      "grad_norm": 16.625,
+      "learning_rate": 9.555859690378846e-07,
+      "loss": 0.723,
+      "step": 4778
+    },
+    {
+      "epoch": 1.161925601750547,
+      "grad_norm": 16.375,
+      "learning_rate": 9.551163243122868e-07,
+      "loss": 0.7377,
+      "step": 4779
+    },
+    {
+      "epoch": 1.162168733284707,
+      "grad_norm": 15.9375,
+      "learning_rate": 9.546467236621472e-07,
+      "loss": 0.4267,
+      "step": 4780
+    },
+    {
+      "epoch": 1.162411864818867,
+      "grad_norm": 24.75,
+      "learning_rate": 9.541771671576557e-07,
+      "loss": 0.9024,
+      "step": 4781
+    },
+    {
+      "epoch": 1.162654996353027,
+      "grad_norm": 25.125,
+      "learning_rate": 9.537076548689953e-07,
+      "loss": 0.9069,
+      "step": 4782
+    },
+    {
+      "epoch": 1.162898127887187,
+      "grad_norm": 27.0,
+      "learning_rate": 9.532381868663436e-07,
+      "loss": 0.7099,
+      "step": 4783
+    },
+    {
+      "epoch": 1.163141259421347,
+      "grad_norm": 20.75,
+      "learning_rate": 9.527687632198697e-07,
+      "loss": 0.5455,
+      "step": 4784
+    },
+    {
+      "epoch": 1.1633843909555068,
+      "grad_norm": 25.625,
+      "learning_rate": 9.522993839997372e-07,
+      "loss": 0.5441,
+      "step": 4785
+    },
+    {
+      "epoch": 1.1636275224896668,
+      "grad_norm": 26.875,
+      "learning_rate": 9.518300492761035e-07,
+      "loss": 0.8369,
+      "step": 4786
+    },
+    {
+      "epoch": 1.1638706540238268,
+      "grad_norm": 20.25,
+      "learning_rate": 9.513607591191186e-07,
+      "loss": 0.6305,
+      "step": 4787
+    },
+    {
+      "epoch": 1.1641137855579868,
+      "grad_norm": 23.5,
+      "learning_rate": 9.508915135989261e-07,
+      "loss": 0.6392,
+      "step": 4788
+    },
+    {
+      "epoch": 1.1643569170921468,
+      "grad_norm": 16.625,
+      "learning_rate": 9.504223127856624e-07,
+      "loss": 0.5864,
+      "step": 4789
+    },
+    {
+      "epoch": 1.1646000486263068,
+      "grad_norm": 15.4375,
+      "learning_rate": 9.49953156749458e-07,
+      "loss": 0.4625,
+      "step": 4790
+    },
+    {
+      "epoch": 1.1648431801604668,
+      "grad_norm": 20.5,
+      "learning_rate": 9.494840455604366e-07,
+      "loss": 0.6627,
+      "step": 4791
+    },
+    {
+      "epoch": 1.1650863116946268,
+      "grad_norm": 19.625,
+      "learning_rate": 9.490149792887143e-07,
+      "loss": 0.7221,
+      "step": 4792
+    },
+    {
+      "epoch": 1.1653294432287868,
+      "grad_norm": 21.5,
+      "learning_rate": 9.485459580044014e-07,
+      "loss": 0.5299,
+      "step": 4793
+    },
+    {
+      "epoch": 1.1655725747629468,
+      "grad_norm": 18.5,
+      "learning_rate": 9.480769817776016e-07,
+      "loss": 0.4763,
+      "step": 4794
+    },
+    {
+      "epoch": 1.1658157062971068,
+      "grad_norm": 15.5625,
+      "learning_rate": 9.476080506784115e-07,
+      "loss": 0.3602,
+      "step": 4795
+    },
+    {
+      "epoch": 1.1660588378312666,
+      "grad_norm": 23.625,
+      "learning_rate": 9.471391647769204e-07,
+      "loss": 0.7227,
+      "step": 4796
+    },
+    {
+      "epoch": 1.1663019693654266,
+      "grad_norm": 18.375,
+      "learning_rate": 9.466703241432118e-07,
+      "loss": 0.6979,
+      "step": 4797
+    },
+    {
+      "epoch": 1.1665451008995866,
+      "grad_norm": 21.5,
+      "learning_rate": 9.462015288473622e-07,
+      "loss": 0.8033,
+      "step": 4798
+    },
+    {
+      "epoch": 1.1667882324337466,
+      "grad_norm": 24.75,
+      "learning_rate": 9.457327789594406e-07,
+      "loss": 0.5965,
+      "step": 4799
+    },
+    {
+      "epoch": 1.1670313639679066,
+      "grad_norm": 32.5,
+      "learning_rate": 9.452640745495104e-07,
+      "loss": 0.8128,
+      "step": 4800
+    },
+    {
+      "epoch": 1.1672744955020666,
+      "grad_norm": 24.875,
+      "learning_rate": 9.447954156876263e-07,
+      "loss": 0.8793,
+      "step": 4801
+    },
+    {
+      "epoch": 1.1675176270362266,
+      "grad_norm": 16.25,
+      "learning_rate": 9.443268024438393e-07,
+      "loss": 0.6922,
+      "step": 4802
+    },
+    {
+      "epoch": 1.1677607585703866,
+      "grad_norm": 21.5,
+      "learning_rate": 9.438582348881906e-07,
+      "loss": 0.9107,
+      "step": 4803
+    },
+    {
+      "epoch": 1.1680038901045466,
+      "grad_norm": 28.375,
+      "learning_rate": 9.433897130907157e-07,
+      "loss": 0.9011,
+      "step": 4804
+    },
+    {
+      "epoch": 1.1682470216387064,
+      "grad_norm": 23.125,
+      "learning_rate": 9.42921237121444e-07,
+      "loss": 0.5479,
+      "step": 4805
+    },
+    {
+      "epoch": 1.1684901531728666,
+      "grad_norm": 26.375,
+      "learning_rate": 9.424528070503967e-07,
+      "loss": 0.6548,
+      "step": 4806
+    },
+    {
+      "epoch": 1.1687332847070264,
+      "grad_norm": 21.875,
+      "learning_rate": 9.419844229475889e-07,
+      "loss": 0.7603,
+      "step": 4807
+    },
+    {
+      "epoch": 1.1689764162411864,
+      "grad_norm": 17.5,
+      "learning_rate": 9.415160848830279e-07,
+      "loss": 0.4958,
+      "step": 4808
+    },
+    {
+      "epoch": 1.1692195477753464,
+      "grad_norm": 18.625,
+      "learning_rate": 9.410477929267167e-07,
+      "loss": 0.5777,
+      "step": 4809
+    },
+    {
+      "epoch": 1.1694626793095064,
+      "grad_norm": 21.75,
+      "learning_rate": 9.405795471486483e-07,
+      "loss": 0.5548,
+      "step": 4810
+    },
+    {
+      "epoch": 1.1697058108436664,
+      "grad_norm": 23.875,
+      "learning_rate": 9.401113476188105e-07,
+      "loss": 0.6941,
+      "step": 4811
+    },
+    {
+      "epoch": 1.1699489423778264,
+      "grad_norm": 21.75,
+      "learning_rate": 9.396431944071839e-07,
+      "loss": 1.0553,
+      "step": 4812
+    },
+    {
+      "epoch": 1.1701920739119864,
+      "grad_norm": 18.0,
+      "learning_rate": 9.391750875837418e-07,
+      "loss": 0.504,
+      "step": 4813
+    },
+    {
+      "epoch": 1.1704352054461464,
+      "grad_norm": 22.875,
+      "learning_rate": 9.387070272184509e-07,
+      "loss": 0.923,
+      "step": 4814
+    },
+    {
+      "epoch": 1.1706783369803064,
+      "grad_norm": 24.375,
+      "learning_rate": 9.382390133812714e-07,
+      "loss": 0.708,
+      "step": 4815
+    },
+    {
+      "epoch": 1.1709214685144662,
+      "grad_norm": 19.125,
+      "learning_rate": 9.37771046142155e-07,
+      "loss": 0.443,
+      "step": 4816
+    },
+    {
+      "epoch": 1.1711646000486262,
+      "grad_norm": 22.25,
+      "learning_rate": 9.373031255710486e-07,
+      "loss": 0.9353,
+      "step": 4817
+    },
+    {
+      "epoch": 1.1714077315827862,
+      "grad_norm": 23.75,
+      "learning_rate": 9.368352517378903e-07,
+      "loss": 0.7583,
+      "step": 4818
+    },
+    {
+      "epoch": 1.1716508631169462,
+      "grad_norm": 14.5625,
+      "learning_rate": 9.363674247126126e-07,
+      "loss": 0.3643,
+      "step": 4819
+    },
+    {
+      "epoch": 1.1718939946511062,
+      "grad_norm": 22.875,
+      "learning_rate": 9.358996445651394e-07,
+      "loss": 0.537,
+      "step": 4820
+    },
+    {
+      "epoch": 1.1721371261852662,
+      "grad_norm": 25.25,
+      "learning_rate": 9.354319113653893e-07,
+      "loss": 1.0392,
+      "step": 4821
+    },
+    {
+      "epoch": 1.1723802577194262,
+      "grad_norm": 24.0,
+      "learning_rate": 9.349642251832729e-07,
+      "loss": 0.8923,
+      "step": 4822
+    },
+    {
+      "epoch": 1.1726233892535862,
+      "grad_norm": 18.875,
+      "learning_rate": 9.344965860886937e-07,
+      "loss": 0.7016,
+      "step": 4823
+    },
+    {
+      "epoch": 1.1728665207877462,
+      "grad_norm": 21.0,
+      "learning_rate": 9.340289941515483e-07,
+      "loss": 0.5655,
+      "step": 4824
+    },
+    {
+      "epoch": 1.1731096523219062,
+      "grad_norm": 20.125,
+      "learning_rate": 9.335614494417271e-07,
+      "loss": 0.6167,
+      "step": 4825
+    },
+    {
+      "epoch": 1.1733527838560662,
+      "grad_norm": 25.125,
+      "learning_rate": 9.330939520291127e-07,
+      "loss": 0.9493,
+      "step": 4826
+    },
+    {
+      "epoch": 1.173595915390226,
+      "grad_norm": 26.875,
+      "learning_rate": 9.326265019835798e-07,
+      "loss": 0.8987,
+      "step": 4827
+    },
+    {
+      "epoch": 1.173839046924386,
+      "grad_norm": 23.625,
+      "learning_rate": 9.321590993749977e-07,
+      "loss": 1.0924,
+      "step": 4828
+    },
+    {
+      "epoch": 1.174082178458546,
+      "grad_norm": 19.75,
+      "learning_rate": 9.316917442732277e-07,
+      "loss": 0.6558,
+      "step": 4829
+    },
+    {
+      "epoch": 1.174325309992706,
+      "grad_norm": 17.375,
+      "learning_rate": 9.312244367481234e-07,
+      "loss": 0.3978,
+      "step": 4830
+    },
+    {
+      "epoch": 1.174568441526866,
+      "grad_norm": 26.75,
+      "learning_rate": 9.307571768695327e-07,
+      "loss": 0.7809,
+      "step": 4831
+    },
+    {
+      "epoch": 1.174811573061026,
+      "grad_norm": 21.5,
+      "learning_rate": 9.302899647072951e-07,
+      "loss": 1.1568,
+      "step": 4832
+    },
+    {
+      "epoch": 1.175054704595186,
+      "grad_norm": 19.0,
+      "learning_rate": 9.298228003312443e-07,
+      "loss": 0.5374,
+      "step": 4833
+    },
+    {
+      "epoch": 1.175297836129346,
+      "grad_norm": 17.75,
+      "learning_rate": 9.293556838112056e-07,
+      "loss": 0.3968,
+      "step": 4834
+    },
+    {
+      "epoch": 1.175540967663506,
+      "grad_norm": 21.125,
+      "learning_rate": 9.288886152169974e-07,
+      "loss": 0.9098,
+      "step": 4835
+    },
+    {
+      "epoch": 1.175784099197666,
+      "grad_norm": 21.75,
+      "learning_rate": 9.284215946184319e-07,
+      "loss": 0.5139,
+      "step": 4836
+    },
+    {
+      "epoch": 1.176027230731826,
+      "grad_norm": 23.625,
+      "learning_rate": 9.279546220853125e-07,
+      "loss": 0.566,
+      "step": 4837
+    },
+    {
+      "epoch": 1.1762703622659858,
+      "grad_norm": 18.375,
+      "learning_rate": 9.27487697687437e-07,
+      "loss": 0.5967,
+      "step": 4838
+    },
+    {
+      "epoch": 1.1765134938001458,
+      "grad_norm": 20.5,
+      "learning_rate": 9.270208214945947e-07,
+      "loss": 0.5644,
+      "step": 4839
+    },
+    {
+      "epoch": 1.1767566253343058,
+      "grad_norm": 21.875,
+      "learning_rate": 9.265539935765691e-07,
+      "loss": 1.1081,
+      "step": 4840
+    },
+    {
+      "epoch": 1.1769997568684658,
+      "grad_norm": 21.625,
+      "learning_rate": 9.26087214003135e-07,
+      "loss": 0.605,
+      "step": 4841
+    },
+    {
+      "epoch": 1.1772428884026258,
+      "grad_norm": 22.625,
+      "learning_rate": 9.25620482844061e-07,
+      "loss": 0.8402,
+      "step": 4842
+    },
+    {
+      "epoch": 1.1774860199367858,
+      "grad_norm": 18.25,
+      "learning_rate": 9.251538001691084e-07,
+      "loss": 0.7289,
+      "step": 4843
+    },
+    {
+      "epoch": 1.1777291514709458,
+      "grad_norm": 22.375,
+      "learning_rate": 9.246871660480303e-07,
+      "loss": 0.5228,
+      "step": 4844
+    },
+    {
+      "epoch": 1.1779722830051058,
+      "grad_norm": 15.3125,
+      "learning_rate": 9.242205805505735e-07,
+      "loss": 0.49,
+      "step": 4845
+    },
+    {
+      "epoch": 1.1782154145392658,
+      "grad_norm": 19.25,
+      "learning_rate": 9.237540437464779e-07,
+      "loss": 0.759,
+      "step": 4846
+    },
+    {
+      "epoch": 1.1784585460734258,
+      "grad_norm": 17.875,
+      "learning_rate": 9.23287555705474e-07,
+      "loss": 0.6646,
+      "step": 4847
+    },
+    {
+      "epoch": 1.1787016776075858,
+      "grad_norm": 19.5,
+      "learning_rate": 9.228211164972879e-07,
+      "loss": 0.3533,
+      "step": 4848
+    },
+    {
+      "epoch": 1.1789448091417456,
+      "grad_norm": 21.625,
+      "learning_rate": 9.223547261916366e-07,
+      "loss": 0.5101,
+      "step": 4849
+    },
+    {
+      "epoch": 1.1791879406759056,
+      "grad_norm": 22.0,
+      "learning_rate": 9.218883848582302e-07,
+      "loss": 0.6311,
+      "step": 4850
+    },
+    {
+      "epoch": 1.1794310722100656,
+      "grad_norm": 20.875,
+      "learning_rate": 9.214220925667712e-07,
+      "loss": 1.1025,
+      "step": 4851
+    },
+    {
+      "epoch": 1.1796742037442256,
+      "grad_norm": 18.625,
+      "learning_rate": 9.209558493869551e-07,
+      "loss": 0.6219,
+      "step": 4852
+    },
+    {
+      "epoch": 1.1799173352783856,
+      "grad_norm": 18.375,
+      "learning_rate": 9.204896553884705e-07,
+      "loss": 0.5677,
+      "step": 4853
+    },
+    {
+      "epoch": 1.1801604668125456,
+      "grad_norm": 23.125,
+      "learning_rate": 9.200235106409974e-07,
+      "loss": 1.0367,
+      "step": 4854
+    },
+    {
+      "epoch": 1.1804035983467056,
+      "grad_norm": 16.375,
+      "learning_rate": 9.195574152142092e-07,
+      "loss": 0.4637,
+      "step": 4855
+    },
+    {
+      "epoch": 1.1806467298808656,
+      "grad_norm": 15.125,
+      "learning_rate": 9.190913691777726e-07,
+      "loss": 0.3284,
+      "step": 4856
+    },
+    {
+      "epoch": 1.1808898614150256,
+      "grad_norm": 23.25,
+      "learning_rate": 9.186253726013461e-07,
+      "loss": 0.8505,
+      "step": 4857
+    },
+    {
+      "epoch": 1.1811329929491854,
+      "grad_norm": 25.125,
+      "learning_rate": 9.181594255545805e-07,
+      "loss": 0.718,
+      "step": 4858
+    },
+    {
+      "epoch": 1.1813761244833456,
+      "grad_norm": 22.875,
+      "learning_rate": 9.176935281071198e-07,
+      "loss": 0.9073,
+      "step": 4859
+    },
+    {
+      "epoch": 1.1816192560175054,
+      "grad_norm": 18.875,
+      "learning_rate": 9.172276803286006e-07,
+      "loss": 0.8079,
+      "step": 4860
+    },
+    {
+      "epoch": 1.1818623875516654,
+      "grad_norm": 17.75,
+      "learning_rate": 9.167618822886516e-07,
+      "loss": 0.49,
+      "step": 4861
+    },
+    {
+      "epoch": 1.1821055190858254,
+      "grad_norm": 24.375,
+      "learning_rate": 9.162961340568944e-07,
+      "loss": 0.9699,
+      "step": 4862
+    },
+    {
+      "epoch": 1.1823486506199854,
+      "grad_norm": 23.875,
+      "learning_rate": 9.158304357029432e-07,
+      "loss": 0.5796,
+      "step": 4863
+    },
+    {
+      "epoch": 1.1825917821541454,
+      "grad_norm": 24.75,
+      "learning_rate": 9.15364787296405e-07,
+      "loss": 1.1586,
+      "step": 4864
+    },
+    {
+      "epoch": 1.1828349136883054,
+      "grad_norm": 18.375,
+      "learning_rate": 9.148991889068785e-07,
+      "loss": 0.5214,
+      "step": 4865
+    },
+    {
+      "epoch": 1.1830780452224654,
+      "grad_norm": 22.25,
+      "learning_rate": 9.144336406039556e-07,
+      "loss": 0.853,
+      "step": 4866
+    },
+    {
+      "epoch": 1.1833211767566254,
+      "grad_norm": 14.75,
+      "learning_rate": 9.139681424572208e-07,
+      "loss": 0.389,
+      "step": 4867
+    },
+    {
+      "epoch": 1.1835643082907854,
+      "grad_norm": 18.125,
+      "learning_rate": 9.135026945362505e-07,
+      "loss": 0.5399,
+      "step": 4868
+    },
+    {
+      "epoch": 1.1838074398249452,
+      "grad_norm": 21.25,
+      "learning_rate": 9.130372969106138e-07,
+      "loss": 0.6072,
+      "step": 4869
+    },
+    {
+      "epoch": 1.1840505713591052,
+      "grad_norm": 19.5,
+      "learning_rate": 9.125719496498723e-07,
+      "loss": 0.7865,
+      "step": 4870
+    },
+    {
+      "epoch": 1.1842937028932652,
+      "grad_norm": 24.5,
+      "learning_rate": 9.121066528235812e-07,
+      "loss": 0.8741,
+      "step": 4871
+    },
+    {
+      "epoch": 1.1845368344274252,
+      "grad_norm": 18.125,
+      "learning_rate": 9.116414065012861e-07,
+      "loss": 0.6396,
+      "step": 4872
+    },
+    {
+      "epoch": 1.1847799659615852,
+      "grad_norm": 19.875,
+      "learning_rate": 9.111762107525266e-07,
+      "loss": 0.6386,
+      "step": 4873
+    },
+    {
+      "epoch": 1.1850230974957452,
+      "grad_norm": 16.75,
+      "learning_rate": 9.107110656468342e-07,
+      "loss": 0.4619,
+      "step": 4874
+    },
+    {
+      "epoch": 1.1852662290299052,
+      "grad_norm": 17.625,
+      "learning_rate": 9.102459712537326e-07,
+      "loss": 0.4006,
+      "step": 4875
+    },
+    {
+      "epoch": 1.1855093605640652,
+      "grad_norm": 28.5,
+      "learning_rate": 9.097809276427382e-07,
+      "loss": 0.7335,
+      "step": 4876
+    },
+    {
+      "epoch": 1.1857524920982252,
+      "grad_norm": 18.75,
+      "learning_rate": 9.093159348833604e-07,
+      "loss": 0.6676,
+      "step": 4877
+    },
+    {
+      "epoch": 1.1859956236323852,
+      "grad_norm": 22.0,
+      "learning_rate": 9.08850993045099e-07,
+      "loss": 1.0245,
+      "step": 4878
+    },
+    {
+      "epoch": 1.1862387551665452,
+      "grad_norm": 19.25,
+      "learning_rate": 9.083861021974494e-07,
+      "loss": 0.5719,
+      "step": 4879
+    },
+    {
+      "epoch": 1.186481886700705,
+      "grad_norm": 15.375,
+      "learning_rate": 9.079212624098966e-07,
+      "loss": 0.3109,
+      "step": 4880
+    },
+    {
+      "epoch": 1.186725018234865,
+      "grad_norm": 27.25,
+      "learning_rate": 9.07456473751919e-07,
+      "loss": 0.7522,
+      "step": 4881
+    },
+    {
+      "epoch": 1.186968149769025,
+      "grad_norm": 13.6875,
+      "learning_rate": 9.069917362929873e-07,
+      "loss": 0.3014,
+      "step": 4882
+    },
+    {
+      "epoch": 1.187211281303185,
+      "grad_norm": 19.875,
+      "learning_rate": 9.065270501025645e-07,
+      "loss": 0.6737,
+      "step": 4883
+    },
+    {
+      "epoch": 1.187454412837345,
+      "grad_norm": 25.75,
+      "learning_rate": 9.060624152501062e-07,
+      "loss": 0.7117,
+      "step": 4884
+    },
+    {
+      "epoch": 1.187697544371505,
+      "grad_norm": 20.25,
+      "learning_rate": 9.055978318050597e-07,
+      "loss": 0.6932,
+      "step": 4885
+    },
+    {
+      "epoch": 1.187940675905665,
+      "grad_norm": 18.0,
+      "learning_rate": 9.051332998368651e-07,
+      "loss": 0.5434,
+      "step": 4886
+    },
+    {
+      "epoch": 1.188183807439825,
+      "grad_norm": 17.5,
+      "learning_rate": 9.046688194149552e-07,
+      "loss": 1.0233,
+      "step": 4887
+    },
+    {
+      "epoch": 1.188426938973985,
+      "grad_norm": 16.75,
+      "learning_rate": 9.042043906087544e-07,
+      "loss": 0.6163,
+      "step": 4888
+    },
+    {
+      "epoch": 1.188670070508145,
+      "grad_norm": 21.0,
+      "learning_rate": 9.037400134876793e-07,
+      "loss": 0.8709,
+      "step": 4889
+    },
+    {
+      "epoch": 1.188913202042305,
+      "grad_norm": 22.375,
+      "learning_rate": 9.032756881211394e-07,
+      "loss": 0.7121,
+      "step": 4890
+    },
+    {
+      "epoch": 1.1891563335764648,
+      "grad_norm": 19.5,
+      "learning_rate": 9.028114145785363e-07,
+      "loss": 0.7821,
+      "step": 4891
+    },
+    {
+      "epoch": 1.1893994651106248,
+      "grad_norm": 22.0,
+      "learning_rate": 9.023471929292632e-07,
+      "loss": 0.7796,
+      "step": 4892
+    },
+    {
+      "epoch": 1.1896425966447848,
+      "grad_norm": 20.125,
+      "learning_rate": 9.018830232427059e-07,
+      "loss": 0.9188,
+      "step": 4893
+    },
+    {
+      "epoch": 1.1898857281789448,
+      "grad_norm": 18.625,
+      "learning_rate": 9.014189055882433e-07,
+      "loss": 0.4755,
+      "step": 4894
+    },
+    {
+      "epoch": 1.1901288597131048,
+      "grad_norm": 17.25,
+      "learning_rate": 9.009548400352455e-07,
+      "loss": 0.6711,
+      "step": 4895
+    },
+    {
+      "epoch": 1.1903719912472648,
+      "grad_norm": 20.625,
+      "learning_rate": 9.004908266530754e-07,
+      "loss": 0.5938,
+      "step": 4896
+    },
+    {
+      "epoch": 1.1906151227814248,
+      "grad_norm": 18.625,
+      "learning_rate": 9.000268655110871e-07,
+      "loss": 0.5651,
+      "step": 4897
+    },
+    {
+      "epoch": 1.1908582543155848,
+      "grad_norm": 15.75,
+      "learning_rate": 8.995629566786282e-07,
+      "loss": 0.295,
+      "step": 4898
+    },
+    {
+      "epoch": 1.1911013858497448,
+      "grad_norm": 20.125,
+      "learning_rate": 8.990991002250376e-07,
+      "loss": 0.5402,
+      "step": 4899
+    },
+    {
+      "epoch": 1.1913445173839048,
+      "grad_norm": 22.75,
+      "learning_rate": 8.986352962196466e-07,
+      "loss": 1.0619,
+      "step": 4900
+    },
+    {
+      "epoch": 1.1915876489180648,
+      "grad_norm": 17.625,
+      "learning_rate": 8.981715447317788e-07,
+      "loss": 0.4252,
+      "step": 4901
+    },
+    {
+      "epoch": 1.1918307804522246,
+      "grad_norm": 21.375,
+      "learning_rate": 8.977078458307499e-07,
+      "loss": 0.8044,
+      "step": 4902
+    },
+    {
+      "epoch": 1.1920739119863846,
+      "grad_norm": 23.875,
+      "learning_rate": 8.972441995858681e-07,
+      "loss": 0.5866,
+      "step": 4903
+    },
+    {
+      "epoch": 1.1923170435205446,
+      "grad_norm": 24.0,
+      "learning_rate": 8.967806060664328e-07,
+      "loss": 1.1851,
+      "step": 4904
+    },
+    {
+      "epoch": 1.1925601750547046,
+      "grad_norm": 17.75,
+      "learning_rate": 8.963170653417364e-07,
+      "loss": 0.4021,
+      "step": 4905
+    },
+    {
+      "epoch": 1.1928033065888646,
+      "grad_norm": 22.75,
+      "learning_rate": 8.958535774810626e-07,
+      "loss": 0.8427,
+      "step": 4906
+    },
+    {
+      "epoch": 1.1930464381230246,
+      "grad_norm": 18.875,
+      "learning_rate": 8.953901425536881e-07,
+      "loss": 0.6998,
+      "step": 4907
+    },
+    {
+      "epoch": 1.1932895696571846,
+      "grad_norm": 15.75,
+      "learning_rate": 8.949267606288814e-07,
+      "loss": 0.318,
+      "step": 4908
+    },
+    {
+      "epoch": 1.1935327011913446,
+      "grad_norm": 20.5,
+      "learning_rate": 8.944634317759019e-07,
+      "loss": 0.5234,
+      "step": 4909
+    },
+    {
+      "epoch": 1.1937758327255046,
+      "grad_norm": 22.875,
+      "learning_rate": 8.940001560640034e-07,
+      "loss": 0.8072,
+      "step": 4910
+    },
+    {
+      "epoch": 1.1940189642596644,
+      "grad_norm": 17.625,
+      "learning_rate": 8.935369335624297e-07,
+      "loss": 0.4621,
+      "step": 4911
+    },
+    {
+      "epoch": 1.1942620957938244,
+      "grad_norm": 21.25,
+      "learning_rate": 8.930737643404178e-07,
+      "loss": 0.759,
+      "step": 4912
+    },
+    {
+      "epoch": 1.1945052273279844,
+      "grad_norm": 23.25,
+      "learning_rate": 8.926106484671962e-07,
+      "loss": 0.6255,
+      "step": 4913
+    },
+    {
+      "epoch": 1.1947483588621444,
+      "grad_norm": 19.5,
+      "learning_rate": 8.921475860119854e-07,
+      "loss": 0.8516,
+      "step": 4914
+    },
+    {
+      "epoch": 1.1949914903963044,
+      "grad_norm": 16.5,
+      "learning_rate": 8.916845770439984e-07,
+      "loss": 0.4676,
+      "step": 4915
+    },
+    {
+      "epoch": 1.1952346219304644,
+      "grad_norm": 20.25,
+      "learning_rate": 8.912216216324395e-07,
+      "loss": 0.5871,
+      "step": 4916
+    },
+    {
+      "epoch": 1.1954777534646244,
+      "grad_norm": 20.125,
+      "learning_rate": 8.907587198465051e-07,
+      "loss": 0.5728,
+      "step": 4917
+    },
+    {
+      "epoch": 1.1957208849987844,
+      "grad_norm": 23.75,
+      "learning_rate": 8.902958717553848e-07,
+      "loss": 0.6871,
+      "step": 4918
+    },
+    {
+      "epoch": 1.1959640165329444,
+      "grad_norm": 18.25,
+      "learning_rate": 8.898330774282588e-07,
+      "loss": 0.7488,
+      "step": 4919
+    },
+    {
+      "epoch": 1.1962071480671044,
+      "grad_norm": 23.875,
+      "learning_rate": 8.893703369342998e-07,
+      "loss": 0.8874,
+      "step": 4920
+    },
+    {
+      "epoch": 1.1964502796012644,
+      "grad_norm": 27.25,
+      "learning_rate": 8.889076503426719e-07,
+      "loss": 0.9963,
+      "step": 4921
+    },
+    {
+      "epoch": 1.1966934111354242,
+      "grad_norm": 17.875,
+      "learning_rate": 8.884450177225323e-07,
+      "loss": 0.7539,
+      "step": 4922
+    },
+    {
+      "epoch": 1.1969365426695842,
+      "grad_norm": 21.25,
+      "learning_rate": 8.879824391430289e-07,
+      "loss": 0.554,
+      "step": 4923
+    },
+    {
+      "epoch": 1.1971796742037442,
+      "grad_norm": 29.375,
+      "learning_rate": 8.875199146733018e-07,
+      "loss": 0.8313,
+      "step": 4924
+    },
+    {
+      "epoch": 1.1974228057379042,
+      "grad_norm": 25.0,
+      "learning_rate": 8.87057444382484e-07,
+      "loss": 0.8371,
+      "step": 4925
+    },
+    {
+      "epoch": 1.1976659372720642,
+      "grad_norm": 19.375,
+      "learning_rate": 8.865950283396995e-07,
+      "loss": 0.664,
+      "step": 4926
+    },
+    {
+      "epoch": 1.1979090688062242,
+      "grad_norm": 14.0625,
+      "learning_rate": 8.861326666140644e-07,
+      "loss": 0.4368,
+      "step": 4927
+    },
+    {
+      "epoch": 1.1981522003403842,
+      "grad_norm": 22.125,
+      "learning_rate": 8.856703592746862e-07,
+      "loss": 0.8901,
+      "step": 4928
+    },
+    {
+      "epoch": 1.1983953318745442,
+      "grad_norm": 17.25,
+      "learning_rate": 8.85208106390665e-07,
+      "loss": 0.519,
+      "step": 4929
+    },
+    {
+      "epoch": 1.1986384634087042,
+      "grad_norm": 17.75,
+      "learning_rate": 8.847459080310927e-07,
+      "loss": 0.7561,
+      "step": 4930
+    },
+    {
+      "epoch": 1.1988815949428642,
+      "grad_norm": 21.625,
+      "learning_rate": 8.842837642650526e-07,
+      "loss": 0.9093,
+      "step": 4931
+    },
+    {
+      "epoch": 1.1991247264770242,
+      "grad_norm": 23.875,
+      "learning_rate": 8.838216751616195e-07,
+      "loss": 0.6419,
+      "step": 4932
+    },
+    {
+      "epoch": 1.199367858011184,
+      "grad_norm": 16.25,
+      "learning_rate": 8.833596407898615e-07,
+      "loss": 0.4743,
+      "step": 4933
+    },
+    {
+      "epoch": 1.199610989545344,
+      "grad_norm": 17.5,
+      "learning_rate": 8.828976612188376e-07,
+      "loss": 0.6557,
+      "step": 4934
+    },
+    {
+      "epoch": 1.199854121079504,
+      "grad_norm": 16.0,
+      "learning_rate": 8.824357365175982e-07,
+      "loss": 0.538,
+      "step": 4935
+    },
+    {
+      "epoch": 1.200097252613664,
+      "grad_norm": 22.125,
+      "learning_rate": 8.819738667551857e-07,
+      "loss": 0.8751,
+      "step": 4936
+    },
+    {
+      "epoch": 1.200340384147824,
+      "grad_norm": 17.125,
+      "learning_rate": 8.815120520006352e-07,
+      "loss": 0.5195,
+      "step": 4937
+    },
+    {
+      "epoch": 1.200583515681984,
+      "grad_norm": 19.75,
+      "learning_rate": 8.810502923229724e-07,
+      "loss": 0.6029,
+      "step": 4938
+    },
+    {
+      "epoch": 1.200826647216144,
+      "grad_norm": 19.625,
+      "learning_rate": 8.805885877912156e-07,
+      "loss": 1.3229,
+      "step": 4939
+    },
+    {
+      "epoch": 1.201069778750304,
+      "grad_norm": 18.625,
+      "learning_rate": 8.801269384743735e-07,
+      "loss": 0.9567,
+      "step": 4940
+    },
+    {
+      "epoch": 1.201312910284464,
+      "grad_norm": 18.0,
+      "learning_rate": 8.796653444414491e-07,
+      "loss": 0.5203,
+      "step": 4941
+    },
+    {
+      "epoch": 1.201556041818624,
+      "grad_norm": 25.75,
+      "learning_rate": 8.792038057614348e-07,
+      "loss": 0.6189,
+      "step": 4942
+    },
+    {
+      "epoch": 1.201799173352784,
+      "grad_norm": 14.3125,
+      "learning_rate": 8.787423225033154e-07,
+      "loss": 0.6033,
+      "step": 4943
+    },
+    {
+      "epoch": 1.2020423048869437,
+      "grad_norm": 21.75,
+      "learning_rate": 8.782808947360682e-07,
+      "loss": 1.0152,
+      "step": 4944
+    },
+    {
+      "epoch": 1.2022854364211037,
+      "grad_norm": 24.125,
+      "learning_rate": 8.778195225286607e-07,
+      "loss": 0.8958,
+      "step": 4945
+    },
+    {
+      "epoch": 1.2025285679552637,
+      "grad_norm": 19.0,
+      "learning_rate": 8.773582059500534e-07,
+      "loss": 0.7713,
+      "step": 4946
+    },
+    {
+      "epoch": 1.2027716994894238,
+      "grad_norm": 25.5,
+      "learning_rate": 8.768969450691982e-07,
+      "loss": 0.5732,
+      "step": 4947
+    },
+    {
+      "epoch": 1.2030148310235838,
+      "grad_norm": 14.9375,
+      "learning_rate": 8.764357399550377e-07,
+      "loss": 0.3609,
+      "step": 4948
+    },
+    {
+      "epoch": 1.2032579625577438,
+      "grad_norm": 18.75,
+      "learning_rate": 8.759745906765079e-07,
+      "loss": 0.5953,
+      "step": 4949
+    },
+    {
+      "epoch": 1.2035010940919038,
+      "grad_norm": 31.0,
+      "learning_rate": 8.75513497302535e-07,
+      "loss": 1.0023,
+      "step": 4950
+    },
+    {
+      "epoch": 1.2037442256260638,
+      "grad_norm": 23.5,
+      "learning_rate": 8.75052459902038e-07,
+      "loss": 0.9095,
+      "step": 4951
+    },
+    {
+      "epoch": 1.2039873571602238,
+      "grad_norm": 21.125,
+      "learning_rate": 8.745914785439261e-07,
+      "loss": 0.9881,
+      "step": 4952
+    },
+    {
+      "epoch": 1.2042304886943835,
+      "grad_norm": 19.625,
+      "learning_rate": 8.741305532971011e-07,
+      "loss": 0.8702,
+      "step": 4953
+    },
+    {
+      "epoch": 1.2044736202285438,
+      "grad_norm": 17.5,
+      "learning_rate": 8.736696842304567e-07,
+      "loss": 0.671,
+      "step": 4954
+    },
+    {
+      "epoch": 1.2047167517627035,
+      "grad_norm": 17.125,
+      "learning_rate": 8.732088714128773e-07,
+      "loss": 0.6673,
+      "step": 4955
+    },
+    {
+      "epoch": 1.2049598832968635,
+      "grad_norm": 27.625,
+      "learning_rate": 8.727481149132394e-07,
+      "loss": 0.5197,
+      "step": 4956
+    },
+    {
+      "epoch": 1.2052030148310235,
+      "grad_norm": 20.75,
+      "learning_rate": 8.722874148004111e-07,
+      "loss": 1.0014,
+      "step": 4957
+    },
+    {
+      "epoch": 1.2054461463651835,
+      "grad_norm": 20.125,
+      "learning_rate": 8.718267711432524e-07,
+      "loss": 0.5627,
+      "step": 4958
+    },
+    {
+      "epoch": 1.2056892778993435,
+      "grad_norm": 18.375,
+      "learning_rate": 8.713661840106136e-07,
+      "loss": 0.8288,
+      "step": 4959
+    },
+    {
+      "epoch": 1.2059324094335035,
+      "grad_norm": 19.125,
+      "learning_rate": 8.709056534713378e-07,
+      "loss": 0.4407,
+      "step": 4960
+    },
+    {
+      "epoch": 1.2061755409676636,
+      "grad_norm": 29.625,
+      "learning_rate": 8.704451795942596e-07,
+      "loss": 1.067,
+      "step": 4961
+    },
+    {
+      "epoch": 1.2064186725018236,
+      "grad_norm": 19.75,
+      "learning_rate": 8.699847624482042e-07,
+      "loss": 0.7111,
+      "step": 4962
+    },
+    {
+      "epoch": 1.2066618040359836,
+      "grad_norm": 24.5,
+      "learning_rate": 8.69524402101989e-07,
+      "loss": 0.895,
+      "step": 4963
+    },
+    {
+      "epoch": 1.2069049355701433,
+      "grad_norm": 24.875,
+      "learning_rate": 8.690640986244231e-07,
+      "loss": 0.7428,
+      "step": 4964
+    },
+    {
+      "epoch": 1.2071480671043033,
+      "grad_norm": 22.75,
+      "learning_rate": 8.686038520843068e-07,
+      "loss": 0.8462,
+      "step": 4965
+    },
+    {
+      "epoch": 1.2073911986384633,
+      "grad_norm": 21.25,
+      "learning_rate": 8.681436625504316e-07,
+      "loss": 0.7423,
+      "step": 4966
+    },
+    {
+      "epoch": 1.2076343301726233,
+      "grad_norm": 20.875,
+      "learning_rate": 8.67683530091581e-07,
+      "loss": 0.567,
+      "step": 4967
+    },
+    {
+      "epoch": 1.2078774617067833,
+      "grad_norm": 15.625,
+      "learning_rate": 8.672234547765298e-07,
+      "loss": 0.5139,
+      "step": 4968
+    },
+    {
+      "epoch": 1.2081205932409433,
+      "grad_norm": 17.375,
+      "learning_rate": 8.667634366740439e-07,
+      "loss": 0.5036,
+      "step": 4969
+    },
+    {
+      "epoch": 1.2083637247751033,
+      "grad_norm": 18.875,
+      "learning_rate": 8.663034758528809e-07,
+      "loss": 0.4269,
+      "step": 4970
+    },
+    {
+      "epoch": 1.2086068563092633,
+      "grad_norm": 19.125,
+      "learning_rate": 8.658435723817902e-07,
+      "loss": 0.5858,
+      "step": 4971
+    },
+    {
+      "epoch": 1.2088499878434233,
+      "grad_norm": 18.625,
+      "learning_rate": 8.653837263295126e-07,
+      "loss": 0.5424,
+      "step": 4972
+    },
+    {
+      "epoch": 1.2090931193775833,
+      "grad_norm": 18.875,
+      "learning_rate": 8.649239377647791e-07,
+      "loss": 0.6346,
+      "step": 4973
+    },
+    {
+      "epoch": 1.2093362509117433,
+      "grad_norm": 21.0,
+      "learning_rate": 8.644642067563138e-07,
+      "loss": 0.7891,
+      "step": 4974
+    },
+    {
+      "epoch": 1.2095793824459031,
+      "grad_norm": 14.125,
+      "learning_rate": 8.640045333728316e-07,
+      "loss": 0.2975,
+      "step": 4975
+    },
+    {
+      "epoch": 1.2098225139800631,
+      "grad_norm": 20.125,
+      "learning_rate": 8.635449176830379e-07,
+      "loss": 0.7084,
+      "step": 4976
+    },
+    {
+      "epoch": 1.2100656455142231,
+      "grad_norm": 16.0,
+      "learning_rate": 8.630853597556308e-07,
+      "loss": 0.6825,
+      "step": 4977
+    },
+    {
+      "epoch": 1.2103087770483831,
+      "grad_norm": 32.75,
+      "learning_rate": 8.626258596592984e-07,
+      "loss": 0.7569,
+      "step": 4978
+    },
+    {
+      "epoch": 1.2105519085825431,
+      "grad_norm": 30.875,
+      "learning_rate": 8.62166417462722e-07,
+      "loss": 0.6905,
+      "step": 4979
+    },
+    {
+      "epoch": 1.2107950401167031,
+      "grad_norm": 20.5,
+      "learning_rate": 8.617070332345723e-07,
+      "loss": 0.6176,
+      "step": 4980
+    },
+    {
+      "epoch": 1.2110381716508631,
+      "grad_norm": 18.0,
+      "learning_rate": 8.612477070435127e-07,
+      "loss": 0.6297,
+      "step": 4981
+    },
+    {
+      "epoch": 1.2112813031850231,
+      "grad_norm": 20.5,
+      "learning_rate": 8.607884389581975e-07,
+      "loss": 0.8307,
+      "step": 4982
+    },
+    {
+      "epoch": 1.2115244347191831,
+      "grad_norm": 24.25,
+      "learning_rate": 8.603292290472717e-07,
+      "loss": 1.0392,
+      "step": 4983
+    },
+    {
+      "epoch": 1.2117675662533431,
+      "grad_norm": 16.125,
+      "learning_rate": 8.598700773793725e-07,
+      "loss": 0.7302,
+      "step": 4984
+    },
+    {
+      "epoch": 1.2120106977875031,
+      "grad_norm": 20.125,
+      "learning_rate": 8.594109840231282e-07,
+      "loss": 0.6587,
+      "step": 4985
+    },
+    {
+      "epoch": 1.212253829321663,
+      "grad_norm": 27.875,
+      "learning_rate": 8.589519490471576e-07,
+      "loss": 0.7282,
+      "step": 4986
+    },
+    {
+      "epoch": 1.212496960855823,
+      "grad_norm": 17.875,
+      "learning_rate": 8.58492972520072e-07,
+      "loss": 0.4925,
+      "step": 4987
+    },
+    {
+      "epoch": 1.212740092389983,
+      "grad_norm": 16.5,
+      "learning_rate": 8.580340545104735e-07,
+      "loss": 0.4443,
+      "step": 4988
+    },
+    {
+      "epoch": 1.212983223924143,
+      "grad_norm": 17.625,
+      "learning_rate": 8.57575195086955e-07,
+      "loss": 0.4446,
+      "step": 4989
+    },
+    {
+      "epoch": 1.213226355458303,
+      "grad_norm": 21.625,
+      "learning_rate": 8.571163943181008e-07,
+      "loss": 0.9306,
+      "step": 4990
+    },
+    {
+      "epoch": 1.213469486992463,
+      "grad_norm": 13.875,
+      "learning_rate": 8.566576522724869e-07,
+      "loss": 0.3169,
+      "step": 4991
+    },
+    {
+      "epoch": 1.213712618526623,
+      "grad_norm": 25.375,
+      "learning_rate": 8.561989690186803e-07,
+      "loss": 0.4153,
+      "step": 4992
+    },
+    {
+      "epoch": 1.213955750060783,
+      "grad_norm": 22.625,
+      "learning_rate": 8.557403446252388e-07,
+      "loss": 0.5255,
+      "step": 4993
+    },
+    {
+      "epoch": 1.214198881594943,
+      "grad_norm": 23.875,
+      "learning_rate": 8.552817791607115e-07,
+      "loss": 0.7175,
+      "step": 4994
+    },
+    {
+      "epoch": 1.214442013129103,
+      "grad_norm": 26.625,
+      "learning_rate": 8.548232726936396e-07,
+      "loss": 0.7659,
+      "step": 4995
+    },
+    {
+      "epoch": 1.214685144663263,
+      "grad_norm": 18.75,
+      "learning_rate": 8.543648252925549e-07,
+      "loss": 0.3982,
+      "step": 4996
+    },
+    {
+      "epoch": 1.2149282761974227,
+      "grad_norm": 26.625,
+      "learning_rate": 8.539064370259796e-07,
+      "loss": 1.0359,
+      "step": 4997
+    },
+    {
+      "epoch": 1.2151714077315827,
+      "grad_norm": 30.75,
+      "learning_rate": 8.53448107962428e-07,
+      "loss": 0.7103,
+      "step": 4998
+    },
+    {
+      "epoch": 1.2154145392657427,
+      "grad_norm": 19.375,
+      "learning_rate": 8.529898381704058e-07,
+      "loss": 0.4476,
+      "step": 4999
+    },
+    {
+      "epoch": 1.2156576707999027,
+      "grad_norm": 20.0,
+      "learning_rate": 8.525316277184084e-07,
+      "loss": 0.5951,
+      "step": 5000
+    },
+    {
+      "epoch": 1.2159008023340627,
+      "grad_norm": 21.75,
+      "learning_rate": 8.520734766749239e-07,
+      "loss": 0.7579,
+      "step": 5001
+    },
+    {
+      "epoch": 1.2161439338682227,
+      "grad_norm": 20.0,
+      "learning_rate": 8.516153851084305e-07,
+      "loss": 1.0184,
+      "step": 5002
+    },
+    {
+      "epoch": 1.2163870654023827,
+      "grad_norm": 20.5,
+      "learning_rate": 8.511573530873985e-07,
+      "loss": 0.88,
+      "step": 5003
+    },
+    {
+      "epoch": 1.2166301969365427,
+      "grad_norm": 23.375,
+      "learning_rate": 8.506993806802882e-07,
+      "loss": 0.6534,
+      "step": 5004
+    },
+    {
+      "epoch": 1.2168733284707027,
+      "grad_norm": 25.125,
+      "learning_rate": 8.502414679555515e-07,
+      "loss": 0.7731,
+      "step": 5005
+    },
+    {
+      "epoch": 1.2171164600048625,
+      "grad_norm": 25.25,
+      "learning_rate": 8.497836149816318e-07,
+      "loss": 0.6918,
+      "step": 5006
+    },
+    {
+      "epoch": 1.2173595915390227,
+      "grad_norm": 20.625,
+      "learning_rate": 8.493258218269627e-07,
+      "loss": 1.1542,
+      "step": 5007
+    },
+    {
+      "epoch": 1.2176027230731825,
+      "grad_norm": 20.0,
+      "learning_rate": 8.488680885599692e-07,
+      "loss": 0.7988,
+      "step": 5008
+    },
+    {
+      "epoch": 1.2178458546073425,
+      "grad_norm": 32.5,
+      "learning_rate": 8.484104152490677e-07,
+      "loss": 1.2164,
+      "step": 5009
+    },
+    {
+      "epoch": 1.2180889861415025,
+      "grad_norm": 15.125,
+      "learning_rate": 8.479528019626654e-07,
+      "loss": 0.4186,
+      "step": 5010
+    },
+    {
+      "epoch": 1.2183321176756625,
+      "grad_norm": 18.875,
+      "learning_rate": 8.474952487691607e-07,
+      "loss": 0.8304,
+      "step": 5011
+    },
+    {
+      "epoch": 1.2185752492098225,
+      "grad_norm": 18.875,
+      "learning_rate": 8.470377557369422e-07,
+      "loss": 0.536,
+      "step": 5012
+    },
+    {
+      "epoch": 1.2188183807439825,
+      "grad_norm": 21.0,
+      "learning_rate": 8.465803229343908e-07,
+      "loss": 0.6295,
+      "step": 5013
+    },
+    {
+      "epoch": 1.2190615122781425,
+      "grad_norm": 39.5,
+      "learning_rate": 8.461229504298772e-07,
+      "loss": 1.0606,
+      "step": 5014
+    },
+    {
+      "epoch": 1.2193046438123025,
+      "grad_norm": 22.375,
+      "learning_rate": 8.456656382917639e-07,
+      "loss": 0.6199,
+      "step": 5015
+    },
+    {
+      "epoch": 1.2195477753464625,
+      "grad_norm": 24.875,
+      "learning_rate": 8.452083865884044e-07,
+      "loss": 0.977,
+      "step": 5016
+    },
+    {
+      "epoch": 1.2197909068806223,
+      "grad_norm": 22.375,
+      "learning_rate": 8.447511953881416e-07,
+      "loss": 1.1204,
+      "step": 5017
+    },
+    {
+      "epoch": 1.2200340384147823,
+      "grad_norm": 17.625,
+      "learning_rate": 8.442940647593123e-07,
+      "loss": 0.5688,
+      "step": 5018
+    },
+    {
+      "epoch": 1.2202771699489423,
+      "grad_norm": 17.375,
+      "learning_rate": 8.438369947702416e-07,
+      "loss": 0.3245,
+      "step": 5019
+    },
+    {
+      "epoch": 1.2205203014831023,
+      "grad_norm": 18.625,
+      "learning_rate": 8.43379985489247e-07,
+      "loss": 0.5849,
+      "step": 5020
+    },
+    {
+      "epoch": 1.2207634330172623,
+      "grad_norm": 22.0,
+      "learning_rate": 8.429230369846358e-07,
+      "loss": 0.5065,
+      "step": 5021
+    },
+    {
+      "epoch": 1.2210065645514223,
+      "grad_norm": 27.375,
+      "learning_rate": 8.424661493247073e-07,
+      "loss": 0.7494,
+      "step": 5022
+    },
+    {
+      "epoch": 1.2212496960855823,
+      "grad_norm": 28.5,
+      "learning_rate": 8.420093225777513e-07,
+      "loss": 0.8999,
+      "step": 5023
+    },
+    {
+      "epoch": 1.2214928276197423,
+      "grad_norm": 18.125,
+      "learning_rate": 8.41552556812048e-07,
+      "loss": 0.5888,
+      "step": 5024
+    },
+    {
+      "epoch": 1.2217359591539023,
+      "grad_norm": 21.625,
+      "learning_rate": 8.410958520958691e-07,
+      "loss": 0.4691,
+      "step": 5025
+    },
+    {
+      "epoch": 1.2219790906880623,
+      "grad_norm": 21.25,
+      "learning_rate": 8.406392084974774e-07,
+      "loss": 0.5534,
+      "step": 5026
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 17.375,
+      "learning_rate": 8.40182626085126e-07,
+      "loss": 0.635,
+      "step": 5027
+    },
+    {
+      "epoch": 1.222465353756382,
+      "grad_norm": 17.5,
+      "learning_rate": 8.397261049270586e-07,
+      "loss": 0.4704,
+      "step": 5028
+    },
+    {
+      "epoch": 1.222708485290542,
+      "grad_norm": 18.0,
+      "learning_rate": 8.392696450915106e-07,
+      "loss": 0.5269,
+      "step": 5029
+    },
+    {
+      "epoch": 1.222951616824702,
+      "grad_norm": 22.0,
+      "learning_rate": 8.388132466467081e-07,
+      "loss": 0.6745,
+      "step": 5030
+    },
+    {
+      "epoch": 1.223194748358862,
+      "grad_norm": 24.25,
+      "learning_rate": 8.383569096608668e-07,
+      "loss": 0.8552,
+      "step": 5031
+    },
+    {
+      "epoch": 1.223437879893022,
+      "grad_norm": 19.0,
+      "learning_rate": 8.379006342021946e-07,
+      "loss": 0.6174,
+      "step": 5032
+    },
+    {
+      "epoch": 1.223681011427182,
+      "grad_norm": 20.75,
+      "learning_rate": 8.374444203388902e-07,
+      "loss": 0.8532,
+      "step": 5033
+    },
+    {
+      "epoch": 1.223924142961342,
+      "grad_norm": 26.125,
+      "learning_rate": 8.36988268139142e-07,
+      "loss": 0.665,
+      "step": 5034
+    },
+    {
+      "epoch": 1.2241672744955021,
+      "grad_norm": 25.375,
+      "learning_rate": 8.365321776711304e-07,
+      "loss": 0.738,
+      "step": 5035
+    },
+    {
+      "epoch": 1.2244104060296621,
+      "grad_norm": 16.875,
+      "learning_rate": 8.360761490030254e-07,
+      "loss": 0.5678,
+      "step": 5036
+    },
+    {
+      "epoch": 1.2246535375638221,
+      "grad_norm": 18.75,
+      "learning_rate": 8.356201822029889e-07,
+      "loss": 0.6996,
+      "step": 5037
+    },
+    {
+      "epoch": 1.2248966690979821,
+      "grad_norm": 29.125,
+      "learning_rate": 8.351642773391727e-07,
+      "loss": 0.9046,
+      "step": 5038
+    },
+    {
+      "epoch": 1.225139800632142,
+      "grad_norm": 21.875,
+      "learning_rate": 8.347084344797196e-07,
+      "loss": 0.9709,
+      "step": 5039
+    },
+    {
+      "epoch": 1.225382932166302,
+      "grad_norm": 15.8125,
+      "learning_rate": 8.342526536927629e-07,
+      "loss": 0.4506,
+      "step": 5040
+    },
+    {
+      "epoch": 1.225626063700462,
+      "grad_norm": 21.75,
+      "learning_rate": 8.337969350464278e-07,
+      "loss": 0.7138,
+      "step": 5041
+    },
+    {
+      "epoch": 1.225869195234622,
+      "grad_norm": 14.75,
+      "learning_rate": 8.333412786088289e-07,
+      "loss": 0.344,
+      "step": 5042
+    },
+    {
+      "epoch": 1.226112326768782,
+      "grad_norm": 18.5,
+      "learning_rate": 8.328856844480718e-07,
+      "loss": 0.5888,
+      "step": 5043
+    },
+    {
+      "epoch": 1.226355458302942,
+      "grad_norm": 21.375,
+      "learning_rate": 8.32430152632253e-07,
+      "loss": 0.9529,
+      "step": 5044
+    },
+    {
+      "epoch": 1.226598589837102,
+      "grad_norm": 19.75,
+      "learning_rate": 8.319746832294595e-07,
+      "loss": 0.6544,
+      "step": 5045
+    },
+    {
+      "epoch": 1.226841721371262,
+      "grad_norm": 16.75,
+      "learning_rate": 8.315192763077689e-07,
+      "loss": 0.4995,
+      "step": 5046
+    },
+    {
+      "epoch": 1.227084852905422,
+      "grad_norm": 22.75,
+      "learning_rate": 8.310639319352504e-07,
+      "loss": 0.9184,
+      "step": 5047
+    },
+    {
+      "epoch": 1.227327984439582,
+      "grad_norm": 35.25,
+      "learning_rate": 8.306086501799616e-07,
+      "loss": 0.6188,
+      "step": 5048
+    },
+    {
+      "epoch": 1.227571115973742,
+      "grad_norm": 20.25,
+      "learning_rate": 8.30153431109954e-07,
+      "loss": 0.519,
+      "step": 5049
+    },
+    {
+      "epoch": 1.2278142475079017,
+      "grad_norm": 18.375,
+      "learning_rate": 8.29698274793267e-07,
+      "loss": 0.5042,
+      "step": 5050
+    },
+    {
+      "epoch": 1.2280573790420617,
+      "grad_norm": 20.0,
+      "learning_rate": 8.292431812979315e-07,
+      "loss": 0.9208,
+      "step": 5051
+    },
+    {
+      "epoch": 1.2283005105762217,
+      "grad_norm": 20.75,
+      "learning_rate": 8.287881506919696e-07,
+      "loss": 0.5641,
+      "step": 5052
+    },
+    {
+      "epoch": 1.2285436421103817,
+      "grad_norm": 19.875,
+      "learning_rate": 8.283331830433928e-07,
+      "loss": 0.7488,
+      "step": 5053
+    },
+    {
+      "epoch": 1.2287867736445417,
+      "grad_norm": 16.5,
+      "learning_rate": 8.278782784202047e-07,
+      "loss": 0.5215,
+      "step": 5054
+    },
+    {
+      "epoch": 1.2290299051787017,
+      "grad_norm": 20.25,
+      "learning_rate": 8.274234368903978e-07,
+      "loss": 0.7046,
+      "step": 5055
+    },
+    {
+      "epoch": 1.2292730367128617,
+      "grad_norm": 19.625,
+      "learning_rate": 8.269686585219561e-07,
+      "loss": 0.3506,
+      "step": 5056
+    },
+    {
+      "epoch": 1.2295161682470217,
+      "grad_norm": 24.25,
+      "learning_rate": 8.265139433828548e-07,
+      "loss": 0.9019,
+      "step": 5057
+    },
+    {
+      "epoch": 1.2297592997811817,
+      "grad_norm": 27.125,
+      "learning_rate": 8.260592915410584e-07,
+      "loss": 0.6694,
+      "step": 5058
+    },
+    {
+      "epoch": 1.2300024313153415,
+      "grad_norm": 25.375,
+      "learning_rate": 8.256047030645228e-07,
+      "loss": 0.3019,
+      "step": 5059
+    },
+    {
+      "epoch": 1.2302455628495015,
+      "grad_norm": 34.0,
+      "learning_rate": 8.251501780211938e-07,
+      "loss": 0.8181,
+      "step": 5060
+    },
+    {
+      "epoch": 1.2304886943836615,
+      "grad_norm": 19.0,
+      "learning_rate": 8.246957164790082e-07,
+      "loss": 0.8579,
+      "step": 5061
+    },
+    {
+      "epoch": 1.2307318259178215,
+      "grad_norm": 21.625,
+      "learning_rate": 8.242413185058928e-07,
+      "loss": 0.8039,
+      "step": 5062
+    },
+    {
+      "epoch": 1.2309749574519815,
+      "grad_norm": 19.75,
+      "learning_rate": 8.237869841697652e-07,
+      "loss": 0.9045,
+      "step": 5063
+    },
+    {
+      "epoch": 1.2312180889861415,
+      "grad_norm": 19.25,
+      "learning_rate": 8.233327135385341e-07,
+      "loss": 0.5451,
+      "step": 5064
+    },
+    {
+      "epoch": 1.2314612205203015,
+      "grad_norm": 17.75,
+      "learning_rate": 8.228785066800977e-07,
+      "loss": 0.6289,
+      "step": 5065
+    },
+    {
+      "epoch": 1.2317043520544615,
+      "grad_norm": 24.375,
+      "learning_rate": 8.224243636623455e-07,
+      "loss": 0.8035,
+      "step": 5066
+    },
+    {
+      "epoch": 1.2319474835886215,
+      "grad_norm": 18.5,
+      "learning_rate": 8.219702845531563e-07,
+      "loss": 0.5471,
+      "step": 5067
+    },
+    {
+      "epoch": 1.2321906151227815,
+      "grad_norm": 19.875,
+      "learning_rate": 8.215162694204003e-07,
+      "loss": 0.5268,
+      "step": 5068
+    },
+    {
+      "epoch": 1.2324337466569415,
+      "grad_norm": 17.5,
+      "learning_rate": 8.210623183319383e-07,
+      "loss": 0.5326,
+      "step": 5069
+    },
+    {
+      "epoch": 1.2326768781911013,
+      "grad_norm": 18.875,
+      "learning_rate": 8.206084313556207e-07,
+      "loss": 0.4787,
+      "step": 5070
+    },
+    {
+      "epoch": 1.2329200097252613,
+      "grad_norm": 22.75,
+      "learning_rate": 8.201546085592884e-07,
+      "loss": 0.8955,
+      "step": 5071
+    },
+    {
+      "epoch": 1.2331631412594213,
+      "grad_norm": 14.9375,
+      "learning_rate": 8.19700850010774e-07,
+      "loss": 0.3327,
+      "step": 5072
+    },
+    {
+      "epoch": 1.2334062727935813,
+      "grad_norm": 16.5,
+      "learning_rate": 8.192471557778995e-07,
+      "loss": 0.7224,
+      "step": 5073
+    },
+    {
+      "epoch": 1.2336494043277413,
+      "grad_norm": 23.375,
+      "learning_rate": 8.187935259284762e-07,
+      "loss": 0.5969,
+      "step": 5074
+    },
+    {
+      "epoch": 1.2338925358619013,
+      "grad_norm": 16.75,
+      "learning_rate": 8.18339960530308e-07,
+      "loss": 0.5852,
+      "step": 5075
+    },
+    {
+      "epoch": 1.2341356673960613,
+      "grad_norm": 22.75,
+      "learning_rate": 8.178864596511879e-07,
+      "loss": 0.5865,
+      "step": 5076
+    },
+    {
+      "epoch": 1.2343787989302213,
+      "grad_norm": 20.625,
+      "learning_rate": 8.174330233588989e-07,
+      "loss": 0.7485,
+      "step": 5077
+    },
+    {
+      "epoch": 1.2346219304643813,
+      "grad_norm": 21.75,
+      "learning_rate": 8.169796517212157e-07,
+      "loss": 0.6504,
+      "step": 5078
+    },
+    {
+      "epoch": 1.2348650619985413,
+      "grad_norm": 23.125,
+      "learning_rate": 8.165263448059013e-07,
+      "loss": 0.7995,
+      "step": 5079
+    },
+    {
+      "epoch": 1.2351081935327013,
+      "grad_norm": 39.25,
+      "learning_rate": 8.160731026807116e-07,
+      "loss": 0.7201,
+      "step": 5080
+    },
+    {
+      "epoch": 1.235351325066861,
+      "grad_norm": 19.75,
+      "learning_rate": 8.156199254133907e-07,
+      "loss": 0.5835,
+      "step": 5081
+    },
+    {
+      "epoch": 1.235594456601021,
+      "grad_norm": 20.625,
+      "learning_rate": 8.151668130716739e-07,
+      "loss": 0.9944,
+      "step": 5082
+    },
+    {
+      "epoch": 1.235837588135181,
+      "grad_norm": 23.0,
+      "learning_rate": 8.14713765723287e-07,
+      "loss": 0.6267,
+      "step": 5083
+    },
+    {
+      "epoch": 1.236080719669341,
+      "grad_norm": 27.875,
+      "learning_rate": 8.142607834359453e-07,
+      "loss": 0.8266,
+      "step": 5084
+    },
+    {
+      "epoch": 1.236323851203501,
+      "grad_norm": 26.375,
+      "learning_rate": 8.138078662773549e-07,
+      "loss": 0.8328,
+      "step": 5085
+    },
+    {
+      "epoch": 1.236566982737661,
+      "grad_norm": 20.0,
+      "learning_rate": 8.133550143152126e-07,
+      "loss": 0.503,
+      "step": 5086
+    },
+    {
+      "epoch": 1.236810114271821,
+      "grad_norm": 15.4375,
+      "learning_rate": 8.129022276172038e-07,
+      "loss": 0.3991,
+      "step": 5087
+    },
+    {
+      "epoch": 1.237053245805981,
+      "grad_norm": 17.0,
+      "learning_rate": 8.124495062510065e-07,
+      "loss": 0.3644,
+      "step": 5088
+    },
+    {
+      "epoch": 1.237296377340141,
+      "grad_norm": 18.375,
+      "learning_rate": 8.11996850284287e-07,
+      "loss": 0.3558,
+      "step": 5089
+    },
+    {
+      "epoch": 1.237539508874301,
+      "grad_norm": 20.875,
+      "learning_rate": 8.115442597847033e-07,
+      "loss": 0.627,
+      "step": 5090
+    },
+    {
+      "epoch": 1.237782640408461,
+      "grad_norm": 19.125,
+      "learning_rate": 8.11091734819902e-07,
+      "loss": 0.6868,
+      "step": 5091
+    },
+    {
+      "epoch": 1.2380257719426209,
+      "grad_norm": 21.25,
+      "learning_rate": 8.106392754575211e-07,
+      "loss": 0.7741,
+      "step": 5092
+    },
+    {
+      "epoch": 1.2382689034767809,
+      "grad_norm": 17.5,
+      "learning_rate": 8.101868817651889e-07,
+      "loss": 0.4642,
+      "step": 5093
+    },
+    {
+      "epoch": 1.2385120350109409,
+      "grad_norm": 20.75,
+      "learning_rate": 8.097345538105223e-07,
+      "loss": 0.6459,
+      "step": 5094
+    },
+    {
+      "epoch": 1.2387551665451009,
+      "grad_norm": 22.5,
+      "learning_rate": 8.092822916611307e-07,
+      "loss": 0.9879,
+      "step": 5095
+    },
+    {
+      "epoch": 1.2389982980792609,
+      "grad_norm": 24.0,
+      "learning_rate": 8.08830095384612e-07,
+      "loss": 0.7213,
+      "step": 5096
+    },
+    {
+      "epoch": 1.2392414296134209,
+      "grad_norm": 21.625,
+      "learning_rate": 8.083779650485552e-07,
+      "loss": 0.8667,
+      "step": 5097
+    },
+    {
+      "epoch": 1.2394845611475809,
+      "grad_norm": 23.75,
+      "learning_rate": 8.079259007205381e-07,
+      "loss": 0.9636,
+      "step": 5098
+    },
+    {
+      "epoch": 1.2397276926817409,
+      "grad_norm": 19.75,
+      "learning_rate": 8.074739024681302e-07,
+      "loss": 0.6448,
+      "step": 5099
+    },
+    {
+      "epoch": 1.2399708242159009,
+      "grad_norm": 22.5,
+      "learning_rate": 8.070219703588905e-07,
+      "loss": 0.4778,
+      "step": 5100
+    },
+    {
+      "epoch": 1.2402139557500607,
+      "grad_norm": 20.5,
+      "learning_rate": 8.065701044603677e-07,
+      "loss": 0.9189,
+      "step": 5101
+    },
+    {
+      "epoch": 1.2404570872842209,
+      "grad_norm": 24.25,
+      "learning_rate": 8.061183048401005e-07,
+      "loss": 0.7258,
+      "step": 5102
+    },
+    {
+      "epoch": 1.2407002188183807,
+      "grad_norm": 20.875,
+      "learning_rate": 8.056665715656193e-07,
+      "loss": 0.7502,
+      "step": 5103
+    },
+    {
+      "epoch": 1.2409433503525407,
+      "grad_norm": 18.25,
+      "learning_rate": 8.05214904704443e-07,
+      "loss": 0.8119,
+      "step": 5104
+    },
+    {
+      "epoch": 1.2411864818867007,
+      "grad_norm": 20.25,
+      "learning_rate": 8.047633043240807e-07,
+      "loss": 0.6244,
+      "step": 5105
+    },
+    {
+      "epoch": 1.2414296134208607,
+      "grad_norm": 15.8125,
+      "learning_rate": 8.043117704920321e-07,
+      "loss": 0.6666,
+      "step": 5106
+    },
+    {
+      "epoch": 1.2416727449550207,
+      "grad_norm": 32.5,
+      "learning_rate": 8.038603032757869e-07,
+      "loss": 0.5042,
+      "step": 5107
+    },
+    {
+      "epoch": 1.2419158764891807,
+      "grad_norm": 24.625,
+      "learning_rate": 8.034089027428244e-07,
+      "loss": 0.8816,
+      "step": 5108
+    },
+    {
+      "epoch": 1.2421590080233407,
+      "grad_norm": 16.625,
+      "learning_rate": 8.029575689606139e-07,
+      "loss": 0.4124,
+      "step": 5109
+    },
+    {
+      "epoch": 1.2424021395575007,
+      "grad_norm": 16.375,
+      "learning_rate": 8.025063019966153e-07,
+      "loss": 0.4326,
+      "step": 5110
+    },
+    {
+      "epoch": 1.2426452710916607,
+      "grad_norm": 20.25,
+      "learning_rate": 8.02055101918279e-07,
+      "loss": 0.5714,
+      "step": 5111
+    },
+    {
+      "epoch": 1.2428884026258205,
+      "grad_norm": 20.25,
+      "learning_rate": 8.016039687930438e-07,
+      "loss": 0.7711,
+      "step": 5112
+    },
+    {
+      "epoch": 1.2431315341599805,
+      "grad_norm": 19.625,
+      "learning_rate": 8.011529026883394e-07,
+      "loss": 0.9322,
+      "step": 5113
+    },
+    {
+      "epoch": 1.2433746656941405,
+      "grad_norm": 17.375,
+      "learning_rate": 8.007019036715858e-07,
+      "loss": 0.6829,
+      "step": 5114
+    },
+    {
+      "epoch": 1.2436177972283005,
+      "grad_norm": 14.8125,
+      "learning_rate": 8.00250971810192e-07,
+      "loss": 0.5632,
+      "step": 5115
+    },
+    {
+      "epoch": 1.2438609287624605,
+      "grad_norm": 20.25,
+      "learning_rate": 7.998001071715581e-07,
+      "loss": 0.762,
+      "step": 5116
+    },
+    {
+      "epoch": 1.2441040602966205,
+      "grad_norm": 19.125,
+      "learning_rate": 7.993493098230729e-07,
+      "loss": 0.8899,
+      "step": 5117
+    },
+    {
+      "epoch": 1.2443471918307805,
+      "grad_norm": 15.0625,
+      "learning_rate": 7.988985798321169e-07,
+      "loss": 0.3197,
+      "step": 5118
+    },
+    {
+      "epoch": 1.2445903233649405,
+      "grad_norm": 23.75,
+      "learning_rate": 7.984479172660589e-07,
+      "loss": 0.7413,
+      "step": 5119
+    },
+    {
+      "epoch": 1.2448334548991005,
+      "grad_norm": 39.75,
+      "learning_rate": 7.979973221922581e-07,
+      "loss": 1.0104,
+      "step": 5120
+    },
+    {
+      "epoch": 1.2450765864332605,
+      "grad_norm": 31.375,
+      "learning_rate": 7.975467946780641e-07,
+      "loss": 0.7779,
+      "step": 5121
+    },
+    {
+      "epoch": 1.2453197179674205,
+      "grad_norm": 47.0,
+      "learning_rate": 7.970963347908156e-07,
+      "loss": 1.2212,
+      "step": 5122
+    },
+    {
+      "epoch": 1.2455628495015802,
+      "grad_norm": 21.375,
+      "learning_rate": 7.966459425978418e-07,
+      "loss": 0.8814,
+      "step": 5123
+    },
+    {
+      "epoch": 1.2458059810357403,
+      "grad_norm": 19.375,
+      "learning_rate": 7.961956181664616e-07,
+      "loss": 0.3904,
+      "step": 5124
+    },
+    {
+      "epoch": 1.2460491125699003,
+      "grad_norm": 22.375,
+      "learning_rate": 7.957453615639835e-07,
+      "loss": 0.6804,
+      "step": 5125
+    },
+    {
+      "epoch": 1.2462922441040603,
+      "grad_norm": 20.625,
+      "learning_rate": 7.952951728577064e-07,
+      "loss": 0.6559,
+      "step": 5126
+    },
+    {
+      "epoch": 1.2465353756382203,
+      "grad_norm": 18.0,
+      "learning_rate": 7.948450521149189e-07,
+      "loss": 0.4762,
+      "step": 5127
+    },
+    {
+      "epoch": 1.2467785071723803,
+      "grad_norm": 22.625,
+      "learning_rate": 7.943949994028993e-07,
+      "loss": 0.7315,
+      "step": 5128
+    },
+    {
+      "epoch": 1.2470216387065403,
+      "grad_norm": 17.0,
+      "learning_rate": 7.939450147889155e-07,
+      "loss": 0.7564,
+      "step": 5129
+    },
+    {
+      "epoch": 1.2472647702407003,
+      "grad_norm": 16.625,
+      "learning_rate": 7.934950983402254e-07,
+      "loss": 0.8133,
+      "step": 5130
+    },
+    {
+      "epoch": 1.2475079017748603,
+      "grad_norm": 14.9375,
+      "learning_rate": 7.930452501240773e-07,
+      "loss": 0.3438,
+      "step": 5131
+    },
+    {
+      "epoch": 1.2477510333090203,
+      "grad_norm": 19.875,
+      "learning_rate": 7.925954702077082e-07,
+      "loss": 0.7378,
+      "step": 5132
+    },
+    {
+      "epoch": 1.2479941648431803,
+      "grad_norm": 16.625,
+      "learning_rate": 7.921457586583456e-07,
+      "loss": 0.4752,
+      "step": 5133
+    },
+    {
+      "epoch": 1.24823729637734,
+      "grad_norm": 17.5,
+      "learning_rate": 7.916961155432069e-07,
+      "loss": 0.868,
+      "step": 5134
+    },
+    {
+      "epoch": 1.2484804279115,
+      "grad_norm": 18.375,
+      "learning_rate": 7.912465409294992e-07,
+      "loss": 0.5162,
+      "step": 5135
+    },
+    {
+      "epoch": 1.24872355944566,
+      "grad_norm": 22.0,
+      "learning_rate": 7.907970348844186e-07,
+      "loss": 0.5627,
+      "step": 5136
+    },
+    {
+      "epoch": 1.24896669097982,
+      "grad_norm": 21.0,
+      "learning_rate": 7.903475974751519e-07,
+      "loss": 0.5853,
+      "step": 5137
+    },
+    {
+      "epoch": 1.24920982251398,
+      "grad_norm": 21.0,
+      "learning_rate": 7.898982287688756e-07,
+      "loss": 0.9628,
+      "step": 5138
+    },
+    {
+      "epoch": 1.24945295404814,
+      "grad_norm": 17.75,
+      "learning_rate": 7.894489288327548e-07,
+      "loss": 0.3787,
+      "step": 5139
+    },
+    {
+      "epoch": 1.2496960855823,
+      "grad_norm": 22.875,
+      "learning_rate": 7.889996977339456e-07,
+      "loss": 0.7657,
+      "step": 5140
+    },
+    {
+      "epoch": 1.24993921711646,
+      "grad_norm": 26.5,
+      "learning_rate": 7.885505355395931e-07,
+      "loss": 0.6755,
+      "step": 5141
+    },
+    {
+      "epoch": 1.25018234865062,
+      "grad_norm": 24.0,
+      "learning_rate": 7.881014423168331e-07,
+      "loss": 1.288,
+      "step": 5142
+    },
+    {
+      "epoch": 1.2504254801847798,
+      "grad_norm": 17.0,
+      "learning_rate": 7.876524181327895e-07,
+      "loss": 0.4901,
+      "step": 5143
+    },
+    {
+      "epoch": 1.25066861171894,
+      "grad_norm": 24.625,
+      "learning_rate": 7.87203463054577e-07,
+      "loss": 0.9405,
+      "step": 5144
+    },
+    {
+      "epoch": 1.2509117432530998,
+      "grad_norm": 19.5,
+      "learning_rate": 7.867545771493002e-07,
+      "loss": 0.9092,
+      "step": 5145
+    },
+    {
+      "epoch": 1.2511548747872598,
+      "grad_norm": 19.625,
+      "learning_rate": 7.863057604840518e-07,
+      "loss": 0.4593,
+      "step": 5146
+    },
+    {
+      "epoch": 1.2513980063214198,
+      "grad_norm": 23.125,
+      "learning_rate": 7.858570131259161e-07,
+      "loss": 0.8733,
+      "step": 5147
+    },
+    {
+      "epoch": 1.2516411378555798,
+      "grad_norm": 25.375,
+      "learning_rate": 7.854083351419652e-07,
+      "loss": 0.9887,
+      "step": 5148
+    },
+    {
+      "epoch": 1.2518842693897398,
+      "grad_norm": 21.625,
+      "learning_rate": 7.849597265992628e-07,
+      "loss": 0.6092,
+      "step": 5149
+    },
+    {
+      "epoch": 1.2521274009238998,
+      "grad_norm": 20.875,
+      "learning_rate": 7.845111875648612e-07,
+      "loss": 0.793,
+      "step": 5150
+    },
+    {
+      "epoch": 1.2523705324580598,
+      "grad_norm": 17.75,
+      "learning_rate": 7.840627181058015e-07,
+      "loss": 0.606,
+      "step": 5151
+    },
+    {
+      "epoch": 1.2526136639922199,
+      "grad_norm": 20.875,
+      "learning_rate": 7.836143182891159e-07,
+      "loss": 0.689,
+      "step": 5152
+    },
+    {
+      "epoch": 1.2528567955263799,
+      "grad_norm": 30.125,
+      "learning_rate": 7.831659881818249e-07,
+      "loss": 0.9238,
+      "step": 5153
+    },
+    {
+      "epoch": 1.2530999270605396,
+      "grad_norm": 18.0,
+      "learning_rate": 7.827177278509394e-07,
+      "loss": 0.785,
+      "step": 5154
+    },
+    {
+      "epoch": 1.2533430585946999,
+      "grad_norm": 24.25,
+      "learning_rate": 7.822695373634601e-07,
+      "loss": 0.6184,
+      "step": 5155
+    },
+    {
+      "epoch": 1.2535861901288596,
+      "grad_norm": 24.625,
+      "learning_rate": 7.818214167863755e-07,
+      "loss": 0.6755,
+      "step": 5156
+    },
+    {
+      "epoch": 1.2538293216630196,
+      "grad_norm": 25.75,
+      "learning_rate": 7.813733661866668e-07,
+      "loss": 0.764,
+      "step": 5157
+    },
+    {
+      "epoch": 1.2540724531971796,
+      "grad_norm": 18.625,
+      "learning_rate": 7.809253856313018e-07,
+      "loss": 0.7369,
+      "step": 5158
+    },
+    {
+      "epoch": 1.2543155847313396,
+      "grad_norm": 23.25,
+      "learning_rate": 7.804774751872391e-07,
+      "loss": 0.7766,
+      "step": 5159
+    },
+    {
+      "epoch": 1.2545587162654996,
+      "grad_norm": 14.8125,
+      "learning_rate": 7.800296349214267e-07,
+      "loss": 0.36,
+      "step": 5160
+    },
+    {
+      "epoch": 1.2548018477996596,
+      "grad_norm": 17.0,
+      "learning_rate": 7.795818649008017e-07,
+      "loss": 0.5848,
+      "step": 5161
+    },
+    {
+      "epoch": 1.2550449793338196,
+      "grad_norm": 16.375,
+      "learning_rate": 7.79134165192292e-07,
+      "loss": 0.611,
+      "step": 5162
+    },
+    {
+      "epoch": 1.2552881108679796,
+      "grad_norm": 18.75,
+      "learning_rate": 7.786865358628129e-07,
+      "loss": 0.6683,
+      "step": 5163
+    },
+    {
+      "epoch": 1.2555312424021396,
+      "grad_norm": 16.75,
+      "learning_rate": 7.782389769792708e-07,
+      "loss": 0.5806,
+      "step": 5164
+    },
+    {
+      "epoch": 1.2557743739362994,
+      "grad_norm": 22.875,
+      "learning_rate": 7.777914886085613e-07,
+      "loss": 0.5107,
+      "step": 5165
+    },
+    {
+      "epoch": 1.2560175054704596,
+      "grad_norm": 25.625,
+      "learning_rate": 7.773440708175694e-07,
+      "loss": 0.9852,
+      "step": 5166
+    },
+    {
+      "epoch": 1.2562606370046194,
+      "grad_norm": 20.625,
+      "learning_rate": 7.768967236731692e-07,
+      "loss": 0.6061,
+      "step": 5167
+    },
+    {
+      "epoch": 1.2565037685387794,
+      "grad_norm": 16.25,
+      "learning_rate": 7.764494472422243e-07,
+      "loss": 0.54,
+      "step": 5168
+    },
+    {
+      "epoch": 1.2567469000729394,
+      "grad_norm": 18.75,
+      "learning_rate": 7.760022415915881e-07,
+      "loss": 0.4746,
+      "step": 5169
+    },
+    {
+      "epoch": 1.2569900316070994,
+      "grad_norm": 17.875,
+      "learning_rate": 7.75555106788103e-07,
+      "loss": 0.9075,
+      "step": 5170
+    },
+    {
+      "epoch": 1.2572331631412594,
+      "grad_norm": 27.375,
+      "learning_rate": 7.75108042898601e-07,
+      "loss": 0.7413,
+      "step": 5171
+    },
+    {
+      "epoch": 1.2574762946754194,
+      "grad_norm": 22.125,
+      "learning_rate": 7.746610499899036e-07,
+      "loss": 0.7065,
+      "step": 5172
+    },
+    {
+      "epoch": 1.2577194262095794,
+      "grad_norm": 20.375,
+      "learning_rate": 7.742141281288218e-07,
+      "loss": 0.4949,
+      "step": 5173
+    },
+    {
+      "epoch": 1.2579625577437394,
+      "grad_norm": 21.375,
+      "learning_rate": 7.73767277382156e-07,
+      "loss": 0.6434,
+      "step": 5174
+    },
+    {
+      "epoch": 1.2582056892778994,
+      "grad_norm": 25.0,
+      "learning_rate": 7.73320497816695e-07,
+      "loss": 0.7034,
+      "step": 5175
+    },
+    {
+      "epoch": 1.2584488208120592,
+      "grad_norm": 16.125,
+      "learning_rate": 7.728737894992186e-07,
+      "loss": 0.3293,
+      "step": 5176
+    },
+    {
+      "epoch": 1.2586919523462192,
+      "grad_norm": 18.125,
+      "learning_rate": 7.724271524964942e-07,
+      "loss": 0.6795,
+      "step": 5177
+    },
+    {
+      "epoch": 1.2589350838803792,
+      "grad_norm": 20.0,
+      "learning_rate": 7.7198058687528e-07,
+      "loss": 0.588,
+      "step": 5178
+    },
+    {
+      "epoch": 1.2591782154145392,
+      "grad_norm": 22.25,
+      "learning_rate": 7.715340927023224e-07,
+      "loss": 0.7101,
+      "step": 5179
+    },
+    {
+      "epoch": 1.2594213469486992,
+      "grad_norm": 15.125,
+      "learning_rate": 7.710876700443581e-07,
+      "loss": 0.363,
+      "step": 5180
+    },
+    {
+      "epoch": 1.2596644784828592,
+      "grad_norm": 22.875,
+      "learning_rate": 7.706413189681132e-07,
+      "loss": 0.5323,
+      "step": 5181
+    },
+    {
+      "epoch": 1.2599076100170192,
+      "grad_norm": 15.5,
+      "learning_rate": 7.701950395403015e-07,
+      "loss": 0.3437,
+      "step": 5182
+    },
+    {
+      "epoch": 1.2601507415511792,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.697488318276281e-07,
+      "loss": 0.36,
+      "step": 5183
+    },
+    {
+      "epoch": 1.2603938730853392,
+      "grad_norm": 18.125,
+      "learning_rate": 7.693026958967856e-07,
+      "loss": 0.4509,
+      "step": 5184
+    },
+    {
+      "epoch": 1.260637004619499,
+      "grad_norm": 16.75,
+      "learning_rate": 7.688566318144572e-07,
+      "loss": 0.5548,
+      "step": 5185
+    },
+    {
+      "epoch": 1.2608801361536592,
+      "grad_norm": 24.25,
+      "learning_rate": 7.684106396473151e-07,
+      "loss": 0.8648,
+      "step": 5186
+    },
+    {
+      "epoch": 1.261123267687819,
+      "grad_norm": 18.0,
+      "learning_rate": 7.679647194620194e-07,
+      "loss": 0.5511,
+      "step": 5187
+    },
+    {
+      "epoch": 1.261366399221979,
+      "grad_norm": 25.625,
+      "learning_rate": 7.675188713252222e-07,
+      "loss": 0.5117,
+      "step": 5188
+    },
+    {
+      "epoch": 1.261609530756139,
+      "grad_norm": 17.0,
+      "learning_rate": 7.670730953035623e-07,
+      "loss": 0.7025,
+      "step": 5189
+    },
+    {
+      "epoch": 1.261852662290299,
+      "grad_norm": 18.375,
+      "learning_rate": 7.666273914636685e-07,
+      "loss": 0.8594,
+      "step": 5190
+    },
+    {
+      "epoch": 1.262095793824459,
+      "grad_norm": 22.125,
+      "learning_rate": 7.661817598721596e-07,
+      "loss": 0.7871,
+      "step": 5191
+    },
+    {
+      "epoch": 1.262338925358619,
+      "grad_norm": 17.625,
+      "learning_rate": 7.657362005956423e-07,
+      "loss": 0.4572,
+      "step": 5192
+    },
+    {
+      "epoch": 1.262582056892779,
+      "grad_norm": 26.25,
+      "learning_rate": 7.652907137007135e-07,
+      "loss": 0.9232,
+      "step": 5193
+    },
+    {
+      "epoch": 1.262825188426939,
+      "grad_norm": 16.25,
+      "learning_rate": 7.648452992539584e-07,
+      "loss": 0.3624,
+      "step": 5194
+    },
+    {
+      "epoch": 1.263068319961099,
+      "grad_norm": 16.125,
+      "learning_rate": 7.643999573219521e-07,
+      "loss": 0.5707,
+      "step": 5195
+    },
+    {
+      "epoch": 1.2633114514952588,
+      "grad_norm": 17.125,
+      "learning_rate": 7.639546879712592e-07,
+      "loss": 0.4665,
+      "step": 5196
+    },
+    {
+      "epoch": 1.263554583029419,
+      "grad_norm": 15.875,
+      "learning_rate": 7.635094912684323e-07,
+      "loss": 0.5266,
+      "step": 5197
+    },
+    {
+      "epoch": 1.2637977145635788,
+      "grad_norm": 24.125,
+      "learning_rate": 7.630643672800142e-07,
+      "loss": 1.1576,
+      "step": 5198
+    },
+    {
+      "epoch": 1.2640408460977388,
+      "grad_norm": 13.3125,
+      "learning_rate": 7.626193160725357e-07,
+      "loss": 0.2176,
+      "step": 5199
+    },
+    {
+      "epoch": 1.2642839776318988,
+      "grad_norm": 15.25,
+      "learning_rate": 7.621743377125182e-07,
+      "loss": 0.2813,
+      "step": 5200
+    },
+    {
+      "epoch": 1.2645271091660588,
+      "grad_norm": 30.125,
+      "learning_rate": 7.617294322664706e-07,
+      "loss": 0.764,
+      "step": 5201
+    },
+    {
+      "epoch": 1.2647702407002188,
+      "grad_norm": 22.125,
+      "learning_rate": 7.612845998008919e-07,
+      "loss": 0.9649,
+      "step": 5202
+    },
+    {
+      "epoch": 1.2650133722343788,
+      "grad_norm": 24.5,
+      "learning_rate": 7.608398403822703e-07,
+      "loss": 0.8815,
+      "step": 5203
+    },
+    {
+      "epoch": 1.2652565037685388,
+      "grad_norm": 21.5,
+      "learning_rate": 7.603951540770828e-07,
+      "loss": 0.7787,
+      "step": 5204
+    },
+    {
+      "epoch": 1.2654996353026988,
+      "grad_norm": 20.25,
+      "learning_rate": 7.599505409517954e-07,
+      "loss": 0.5999,
+      "step": 5205
+    },
+    {
+      "epoch": 1.2657427668368588,
+      "grad_norm": 22.125,
+      "learning_rate": 7.595060010728629e-07,
+      "loss": 0.5027,
+      "step": 5206
+    },
+    {
+      "epoch": 1.2659858983710186,
+      "grad_norm": 19.375,
+      "learning_rate": 7.590615345067298e-07,
+      "loss": 0.6925,
+      "step": 5207
+    },
+    {
+      "epoch": 1.2662290299051788,
+      "grad_norm": 16.875,
+      "learning_rate": 7.586171413198293e-07,
+      "loss": 0.4931,
+      "step": 5208
+    },
+    {
+      "epoch": 1.2664721614393386,
+      "grad_norm": 20.125,
+      "learning_rate": 7.581728215785832e-07,
+      "loss": 1.0039,
+      "step": 5209
+    },
+    {
+      "epoch": 1.2667152929734986,
+      "grad_norm": 17.125,
+      "learning_rate": 7.57728575349403e-07,
+      "loss": 0.5779,
+      "step": 5210
+    },
+    {
+      "epoch": 1.2669584245076586,
+      "grad_norm": 22.625,
+      "learning_rate": 7.572844026986889e-07,
+      "loss": 0.9309,
+      "step": 5211
+    },
+    {
+      "epoch": 1.2672015560418186,
+      "grad_norm": 16.375,
+      "learning_rate": 7.56840303692831e-07,
+      "loss": 0.4784,
+      "step": 5212
+    },
+    {
+      "epoch": 1.2674446875759786,
+      "grad_norm": 20.5,
+      "learning_rate": 7.563962783982064e-07,
+      "loss": 0.8777,
+      "step": 5213
+    },
+    {
+      "epoch": 1.2676878191101386,
+      "grad_norm": 24.25,
+      "learning_rate": 7.559523268811828e-07,
+      "loss": 0.9291,
+      "step": 5214
+    },
+    {
+      "epoch": 1.2679309506442986,
+      "grad_norm": 21.75,
+      "learning_rate": 7.555084492081169e-07,
+      "loss": 0.6857,
+      "step": 5215
+    },
+    {
+      "epoch": 1.2681740821784586,
+      "grad_norm": 19.875,
+      "learning_rate": 7.550646454453529e-07,
+      "loss": 0.5013,
+      "step": 5216
+    },
+    {
+      "epoch": 1.2684172137126186,
+      "grad_norm": 20.0,
+      "learning_rate": 7.546209156592258e-07,
+      "loss": 0.4317,
+      "step": 5217
+    },
+    {
+      "epoch": 1.2686603452467784,
+      "grad_norm": 22.625,
+      "learning_rate": 7.541772599160577e-07,
+      "loss": 0.4605,
+      "step": 5218
+    },
+    {
+      "epoch": 1.2689034767809386,
+      "grad_norm": 17.0,
+      "learning_rate": 7.537336782821621e-07,
+      "loss": 0.7397,
+      "step": 5219
+    },
+    {
+      "epoch": 1.2691466083150984,
+      "grad_norm": 25.375,
+      "learning_rate": 7.532901708238387e-07,
+      "loss": 0.6636,
+      "step": 5220
+    },
+    {
+      "epoch": 1.2693897398492584,
+      "grad_norm": 18.75,
+      "learning_rate": 7.528467376073778e-07,
+      "loss": 0.717,
+      "step": 5221
+    },
+    {
+      "epoch": 1.2696328713834184,
+      "grad_norm": 18.75,
+      "learning_rate": 7.524033786990583e-07,
+      "loss": 0.6163,
+      "step": 5222
+    },
+    {
+      "epoch": 1.2698760029175784,
+      "grad_norm": 19.875,
+      "learning_rate": 7.519600941651472e-07,
+      "loss": 0.6147,
+      "step": 5223
+    },
+    {
+      "epoch": 1.2701191344517384,
+      "grad_norm": 21.0,
+      "learning_rate": 7.515168840719017e-07,
+      "loss": 0.608,
+      "step": 5224
+    },
+    {
+      "epoch": 1.2703622659858984,
+      "grad_norm": 19.5,
+      "learning_rate": 7.510737484855673e-07,
+      "loss": 0.9512,
+      "step": 5225
+    },
+    {
+      "epoch": 1.2706053975200584,
+      "grad_norm": 22.5,
+      "learning_rate": 7.506306874723774e-07,
+      "loss": 0.5054,
+      "step": 5226
+    },
+    {
+      "epoch": 1.2708485290542184,
+      "grad_norm": 33.0,
+      "learning_rate": 7.501877010985559e-07,
+      "loss": 0.671,
+      "step": 5227
+    },
+    {
+      "epoch": 1.2710916605883784,
+      "grad_norm": 21.375,
+      "learning_rate": 7.497447894303145e-07,
+      "loss": 0.6001,
+      "step": 5228
+    },
+    {
+      "epoch": 1.2713347921225382,
+      "grad_norm": 17.25,
+      "learning_rate": 7.493019525338542e-07,
+      "loss": 0.7246,
+      "step": 5229
+    },
+    {
+      "epoch": 1.2715779236566982,
+      "grad_norm": 21.875,
+      "learning_rate": 7.488591904753642e-07,
+      "loss": 0.5793,
+      "step": 5230
+    },
+    {
+      "epoch": 1.2718210551908582,
+      "grad_norm": 14.9375,
+      "learning_rate": 7.484165033210233e-07,
+      "loss": 0.3067,
+      "step": 5231
+    },
+    {
+      "epoch": 1.2720641867250182,
+      "grad_norm": 13.9375,
+      "learning_rate": 7.479738911369988e-07,
+      "loss": 0.3615,
+      "step": 5232
+    },
+    {
+      "epoch": 1.2723073182591782,
+      "grad_norm": 22.0,
+      "learning_rate": 7.47531353989446e-07,
+      "loss": 0.6091,
+      "step": 5233
+    },
+    {
+      "epoch": 1.2725504497933382,
+      "grad_norm": 20.125,
+      "learning_rate": 7.470888919445107e-07,
+      "loss": 0.7712,
+      "step": 5234
+    },
+    {
+      "epoch": 1.2727935813274982,
+      "grad_norm": 21.375,
+      "learning_rate": 7.466465050683261e-07,
+      "loss": 0.584,
+      "step": 5235
+    },
+    {
+      "epoch": 1.2730367128616582,
+      "grad_norm": 18.0,
+      "learning_rate": 7.462041934270145e-07,
+      "loss": 0.6518,
+      "step": 5236
+    },
+    {
+      "epoch": 1.2732798443958182,
+      "grad_norm": 31.375,
+      "learning_rate": 7.457619570866869e-07,
+      "loss": 0.9055,
+      "step": 5237
+    },
+    {
+      "epoch": 1.273522975929978,
+      "grad_norm": 18.125,
+      "learning_rate": 7.453197961134436e-07,
+      "loss": 0.5466,
+      "step": 5238
+    },
+    {
+      "epoch": 1.2737661074641382,
+      "grad_norm": 17.125,
+      "learning_rate": 7.448777105733729e-07,
+      "loss": 0.7802,
+      "step": 5239
+    },
+    {
+      "epoch": 1.274009238998298,
+      "grad_norm": 17.375,
+      "learning_rate": 7.444357005325519e-07,
+      "loss": 0.7875,
+      "step": 5240
+    },
+    {
+      "epoch": 1.274252370532458,
+      "grad_norm": 20.125,
+      "learning_rate": 7.439937660570464e-07,
+      "loss": 0.7861,
+      "step": 5241
+    },
+    {
+      "epoch": 1.274495502066618,
+      "grad_norm": 15.9375,
+      "learning_rate": 7.435519072129121e-07,
+      "loss": 0.3423,
+      "step": 5242
+    },
+    {
+      "epoch": 1.274738633600778,
+      "grad_norm": 18.0,
+      "learning_rate": 7.431101240661923e-07,
+      "loss": 0.5573,
+      "step": 5243
+    },
+    {
+      "epoch": 1.274981765134938,
+      "grad_norm": 19.125,
+      "learning_rate": 7.426684166829182e-07,
+      "loss": 0.6575,
+      "step": 5244
+    },
+    {
+      "epoch": 1.275224896669098,
+      "grad_norm": 25.125,
+      "learning_rate": 7.422267851291113e-07,
+      "loss": 1.0183,
+      "step": 5245
+    },
+    {
+      "epoch": 1.275468028203258,
+      "grad_norm": 22.375,
+      "learning_rate": 7.417852294707812e-07,
+      "loss": 0.4543,
+      "step": 5246
+    },
+    {
+      "epoch": 1.275711159737418,
+      "grad_norm": 18.25,
+      "learning_rate": 7.413437497739254e-07,
+      "loss": 0.5249,
+      "step": 5247
+    },
+    {
+      "epoch": 1.275954291271578,
+      "grad_norm": 20.375,
+      "learning_rate": 7.409023461045309e-07,
+      "loss": 0.6888,
+      "step": 5248
+    },
+    {
+      "epoch": 1.2761974228057378,
+      "grad_norm": 16.625,
+      "learning_rate": 7.404610185285729e-07,
+      "loss": 0.621,
+      "step": 5249
+    },
+    {
+      "epoch": 1.276440554339898,
+      "grad_norm": 14.625,
+      "learning_rate": 7.400197671120165e-07,
+      "loss": 0.5498,
+      "step": 5250
+    },
+    {
+      "epoch": 1.2766836858740578,
+      "grad_norm": 17.5,
+      "learning_rate": 7.395785919208131e-07,
+      "loss": 0.4632,
+      "step": 5251
+    },
+    {
+      "epoch": 1.2769268174082178,
+      "grad_norm": 17.25,
+      "learning_rate": 7.391374930209046e-07,
+      "loss": 0.6256,
+      "step": 5252
+    },
+    {
+      "epoch": 1.2771699489423778,
+      "grad_norm": 17.875,
+      "learning_rate": 7.386964704782208e-07,
+      "loss": 0.4981,
+      "step": 5253
+    },
+    {
+      "epoch": 1.2774130804765378,
+      "grad_norm": 19.875,
+      "learning_rate": 7.382555243586797e-07,
+      "loss": 0.8559,
+      "step": 5254
+    },
+    {
+      "epoch": 1.2776562120106978,
+      "grad_norm": 35.5,
+      "learning_rate": 7.378146547281888e-07,
+      "loss": 1.1875,
+      "step": 5255
+    },
+    {
+      "epoch": 1.2778993435448578,
+      "grad_norm": 18.75,
+      "learning_rate": 7.373738616526431e-07,
+      "loss": 0.7175,
+      "step": 5256
+    },
+    {
+      "epoch": 1.2781424750790178,
+      "grad_norm": 23.375,
+      "learning_rate": 7.369331451979278e-07,
+      "loss": 0.5727,
+      "step": 5257
+    },
+    {
+      "epoch": 1.2783856066131778,
+      "grad_norm": 18.0,
+      "learning_rate": 7.364925054299143e-07,
+      "loss": 0.654,
+      "step": 5258
+    },
+    {
+      "epoch": 1.2786287381473378,
+      "grad_norm": 17.375,
+      "learning_rate": 7.36051942414465e-07,
+      "loss": 0.3481,
+      "step": 5259
+    },
+    {
+      "epoch": 1.2788718696814976,
+      "grad_norm": 21.25,
+      "learning_rate": 7.356114562174287e-07,
+      "loss": 0.6114,
+      "step": 5260
+    },
+    {
+      "epoch": 1.2791150012156578,
+      "grad_norm": 32.5,
+      "learning_rate": 7.351710469046444e-07,
+      "loss": 0.7523,
+      "step": 5261
+    },
+    {
+      "epoch": 1.2793581327498176,
+      "grad_norm": 18.375,
+      "learning_rate": 7.347307145419383e-07,
+      "loss": 0.5799,
+      "step": 5262
+    },
+    {
+      "epoch": 1.2796012642839776,
+      "grad_norm": 22.75,
+      "learning_rate": 7.342904591951261e-07,
+      "loss": 0.8784,
+      "step": 5263
+    },
+    {
+      "epoch": 1.2798443958181376,
+      "grad_norm": 16.125,
+      "learning_rate": 7.338502809300111e-07,
+      "loss": 0.496,
+      "step": 5264
+    },
+    {
+      "epoch": 1.2800875273522976,
+      "grad_norm": 20.125,
+      "learning_rate": 7.334101798123858e-07,
+      "loss": 0.4859,
+      "step": 5265
+    },
+    {
+      "epoch": 1.2803306588864576,
+      "grad_norm": 19.125,
+      "learning_rate": 7.32970155908031e-07,
+      "loss": 0.6878,
+      "step": 5266
+    },
+    {
+      "epoch": 1.2805737904206176,
+      "grad_norm": 19.25,
+      "learning_rate": 7.325302092827162e-07,
+      "loss": 0.8037,
+      "step": 5267
+    },
+    {
+      "epoch": 1.2808169219547776,
+      "grad_norm": 15.6875,
+      "learning_rate": 7.320903400021983e-07,
+      "loss": 0.5192,
+      "step": 5268
+    },
+    {
+      "epoch": 1.2810600534889376,
+      "grad_norm": 15.9375,
+      "learning_rate": 7.316505481322237e-07,
+      "loss": 0.4567,
+      "step": 5269
+    },
+    {
+      "epoch": 1.2813031850230976,
+      "grad_norm": 18.125,
+      "learning_rate": 7.312108337385272e-07,
+      "loss": 0.4329,
+      "step": 5270
+    },
+    {
+      "epoch": 1.2815463165572574,
+      "grad_norm": 27.25,
+      "learning_rate": 7.307711968868311e-07,
+      "loss": 0.7742,
+      "step": 5271
+    },
+    {
+      "epoch": 1.2817894480914176,
+      "grad_norm": 21.375,
+      "learning_rate": 7.303316376428468e-07,
+      "loss": 0.5598,
+      "step": 5272
+    },
+    {
+      "epoch": 1.2820325796255774,
+      "grad_norm": 23.375,
+      "learning_rate": 7.298921560722744e-07,
+      "loss": 0.7967,
+      "step": 5273
+    },
+    {
+      "epoch": 1.2822757111597374,
+      "grad_norm": 17.625,
+      "learning_rate": 7.294527522408019e-07,
+      "loss": 0.394,
+      "step": 5274
+    },
+    {
+      "epoch": 1.2825188426938974,
+      "grad_norm": 29.875,
+      "learning_rate": 7.29013426214106e-07,
+      "loss": 0.9392,
+      "step": 5275
+    },
+    {
+      "epoch": 1.2827619742280574,
+      "grad_norm": 24.25,
+      "learning_rate": 7.285741780578513e-07,
+      "loss": 0.7566,
+      "step": 5276
+    },
+    {
+      "epoch": 1.2830051057622174,
+      "grad_norm": 15.625,
+      "learning_rate": 7.281350078376906e-07,
+      "loss": 0.3234,
+      "step": 5277
+    },
+    {
+      "epoch": 1.2832482372963774,
+      "grad_norm": 21.625,
+      "learning_rate": 7.276959156192664e-07,
+      "loss": 0.8778,
+      "step": 5278
+    },
+    {
+      "epoch": 1.2834913688305374,
+      "grad_norm": 20.25,
+      "learning_rate": 7.272569014682079e-07,
+      "loss": 0.8403,
+      "step": 5279
+    },
+    {
+      "epoch": 1.2837345003646974,
+      "grad_norm": 15.8125,
+      "learning_rate": 7.268179654501332e-07,
+      "loss": 0.4489,
+      "step": 5280
+    },
+    {
+      "epoch": 1.2839776318988574,
+      "grad_norm": 26.125,
+      "learning_rate": 7.263791076306492e-07,
+      "loss": 0.4786,
+      "step": 5281
+    },
+    {
+      "epoch": 1.2842207634330172,
+      "grad_norm": 30.625,
+      "learning_rate": 7.259403280753513e-07,
+      "loss": 1.072,
+      "step": 5282
+    },
+    {
+      "epoch": 1.2844638949671772,
+      "grad_norm": 19.125,
+      "learning_rate": 7.255016268498223e-07,
+      "loss": 0.6029,
+      "step": 5283
+    },
+    {
+      "epoch": 1.2847070265013372,
+      "grad_norm": 18.75,
+      "learning_rate": 7.25063004019633e-07,
+      "loss": 0.8734,
+      "step": 5284
+    },
+    {
+      "epoch": 1.2849501580354972,
+      "grad_norm": 16.5,
+      "learning_rate": 7.246244596503442e-07,
+      "loss": 0.6562,
+      "step": 5285
+    },
+    {
+      "epoch": 1.2851932895696572,
+      "grad_norm": 17.125,
+      "learning_rate": 7.241859938075035e-07,
+      "loss": 0.614,
+      "step": 5286
+    },
+    {
+      "epoch": 1.2854364211038172,
+      "grad_norm": 18.75,
+      "learning_rate": 7.237476065566466e-07,
+      "loss": 0.8221,
+      "step": 5287
+    },
+    {
+      "epoch": 1.2856795526379772,
+      "grad_norm": 22.625,
+      "learning_rate": 7.233092979632986e-07,
+      "loss": 0.7513,
+      "step": 5288
+    },
+    {
+      "epoch": 1.2859226841721372,
+      "grad_norm": 20.25,
+      "learning_rate": 7.22871068092973e-07,
+      "loss": 0.7276,
+      "step": 5289
+    },
+    {
+      "epoch": 1.2861658157062972,
+      "grad_norm": 19.5,
+      "learning_rate": 7.224329170111698e-07,
+      "loss": 0.4139,
+      "step": 5290
+    },
+    {
+      "epoch": 1.286408947240457,
+      "grad_norm": 22.25,
+      "learning_rate": 7.219948447833782e-07,
+      "loss": 0.5554,
+      "step": 5291
+    },
+    {
+      "epoch": 1.2866520787746172,
+      "grad_norm": 17.5,
+      "learning_rate": 7.215568514750765e-07,
+      "loss": 0.7128,
+      "step": 5292
+    },
+    {
+      "epoch": 1.286895210308777,
+      "grad_norm": 24.0,
+      "learning_rate": 7.211189371517297e-07,
+      "loss": 0.7536,
+      "step": 5293
+    },
+    {
+      "epoch": 1.287138341842937,
+      "grad_norm": 20.375,
+      "learning_rate": 7.206811018787915e-07,
+      "loss": 0.7304,
+      "step": 5294
+    },
+    {
+      "epoch": 1.287381473377097,
+      "grad_norm": 32.0,
+      "learning_rate": 7.202433457217043e-07,
+      "loss": 0.754,
+      "step": 5295
+    },
+    {
+      "epoch": 1.287624604911257,
+      "grad_norm": 16.75,
+      "learning_rate": 7.198056687458988e-07,
+      "loss": 0.5729,
+      "step": 5296
+    },
+    {
+      "epoch": 1.287867736445417,
+      "grad_norm": 17.25,
+      "learning_rate": 7.193680710167927e-07,
+      "loss": 0.3206,
+      "step": 5297
+    },
+    {
+      "epoch": 1.288110867979577,
+      "grad_norm": 17.25,
+      "learning_rate": 7.189305525997925e-07,
+      "loss": 0.734,
+      "step": 5298
+    },
+    {
+      "epoch": 1.288353999513737,
+      "grad_norm": 16.0,
+      "learning_rate": 7.184931135602933e-07,
+      "loss": 0.4035,
+      "step": 5299
+    },
+    {
+      "epoch": 1.288597131047897,
+      "grad_norm": 18.5,
+      "learning_rate": 7.180557539636779e-07,
+      "loss": 0.7211,
+      "step": 5300
+    },
+    {
+      "epoch": 1.288840262582057,
+      "grad_norm": 19.0,
+      "learning_rate": 7.176184738753166e-07,
+      "loss": 0.804,
+      "step": 5301
+    },
+    {
+      "epoch": 1.2890833941162168,
+      "grad_norm": 19.0,
+      "learning_rate": 7.171812733605694e-07,
+      "loss": 0.7541,
+      "step": 5302
+    },
+    {
+      "epoch": 1.289326525650377,
+      "grad_norm": 22.875,
+      "learning_rate": 7.167441524847824e-07,
+      "loss": 1.01,
+      "step": 5303
+    },
+    {
+      "epoch": 1.2895696571845368,
+      "grad_norm": 19.25,
+      "learning_rate": 7.163071113132919e-07,
+      "loss": 0.746,
+      "step": 5304
+    },
+    {
+      "epoch": 1.2898127887186968,
+      "grad_norm": 26.25,
+      "learning_rate": 7.158701499114205e-07,
+      "loss": 0.9137,
+      "step": 5305
+    },
+    {
+      "epoch": 1.2900559202528568,
+      "grad_norm": 19.375,
+      "learning_rate": 7.154332683444803e-07,
+      "loss": 0.8958,
+      "step": 5306
+    },
+    {
+      "epoch": 1.2902990517870168,
+      "grad_norm": 19.5,
+      "learning_rate": 7.149964666777706e-07,
+      "loss": 0.6978,
+      "step": 5307
+    },
+    {
+      "epoch": 1.2905421833211768,
+      "grad_norm": 20.25,
+      "learning_rate": 7.145597449765782e-07,
+      "loss": 0.4383,
+      "step": 5308
+    },
+    {
+      "epoch": 1.2907853148553368,
+      "grad_norm": 18.375,
+      "learning_rate": 7.141231033061797e-07,
+      "loss": 0.6132,
+      "step": 5309
+    },
+    {
+      "epoch": 1.2910284463894968,
+      "grad_norm": 37.0,
+      "learning_rate": 7.136865417318384e-07,
+      "loss": 0.7585,
+      "step": 5310
+    },
+    {
+      "epoch": 1.2912715779236568,
+      "grad_norm": 33.0,
+      "learning_rate": 7.132500603188054e-07,
+      "loss": 0.7202,
+      "step": 5311
+    },
+    {
+      "epoch": 1.2915147094578168,
+      "grad_norm": 22.75,
+      "learning_rate": 7.128136591323209e-07,
+      "loss": 1.0635,
+      "step": 5312
+    },
+    {
+      "epoch": 1.2917578409919765,
+      "grad_norm": 29.25,
+      "learning_rate": 7.12377338237613e-07,
+      "loss": 0.6653,
+      "step": 5313
+    },
+    {
+      "epoch": 1.2920009725261368,
+      "grad_norm": 20.375,
+      "learning_rate": 7.11941097699897e-07,
+      "loss": 0.7537,
+      "step": 5314
+    },
+    {
+      "epoch": 1.2922441040602966,
+      "grad_norm": 23.0,
+      "learning_rate": 7.115049375843761e-07,
+      "loss": 1.0126,
+      "step": 5315
+    },
+    {
+      "epoch": 1.2924872355944566,
+      "grad_norm": 21.875,
+      "learning_rate": 7.110688579562428e-07,
+      "loss": 0.7576,
+      "step": 5316
+    },
+    {
+      "epoch": 1.2927303671286166,
+      "grad_norm": 17.625,
+      "learning_rate": 7.106328588806762e-07,
+      "loss": 0.5833,
+      "step": 5317
+    },
+    {
+      "epoch": 1.2929734986627766,
+      "grad_norm": 18.25,
+      "learning_rate": 7.101969404228439e-07,
+      "loss": 0.7418,
+      "step": 5318
+    },
+    {
+      "epoch": 1.2932166301969366,
+      "grad_norm": 22.5,
+      "learning_rate": 7.097611026479015e-07,
+      "loss": 0.581,
+      "step": 5319
+    },
+    {
+      "epoch": 1.2934597617310966,
+      "grad_norm": 22.75,
+      "learning_rate": 7.093253456209928e-07,
+      "loss": 0.8539,
+      "step": 5320
+    },
+    {
+      "epoch": 1.2937028932652566,
+      "grad_norm": 23.875,
+      "learning_rate": 7.088896694072492e-07,
+      "loss": 0.7946,
+      "step": 5321
+    },
+    {
+      "epoch": 1.2939460247994166,
+      "grad_norm": 19.75,
+      "learning_rate": 7.084540740717892e-07,
+      "loss": 0.9326,
+      "step": 5322
+    },
+    {
+      "epoch": 1.2941891563335766,
+      "grad_norm": 32.25,
+      "learning_rate": 7.080185596797212e-07,
+      "loss": 0.7788,
+      "step": 5323
+    },
+    {
+      "epoch": 1.2944322878677363,
+      "grad_norm": 34.0,
+      "learning_rate": 7.075831262961399e-07,
+      "loss": 1.151,
+      "step": 5324
+    },
+    {
+      "epoch": 1.2946754194018963,
+      "grad_norm": 17.25,
+      "learning_rate": 7.071477739861276e-07,
+      "loss": 0.437,
+      "step": 5325
+    },
+    {
+      "epoch": 1.2949185509360563,
+      "grad_norm": 16.375,
+      "learning_rate": 7.067125028147559e-07,
+      "loss": 0.7148,
+      "step": 5326
+    },
+    {
+      "epoch": 1.2951616824702163,
+      "grad_norm": 20.5,
+      "learning_rate": 7.06277312847084e-07,
+      "loss": 0.7783,
+      "step": 5327
+    },
+    {
+      "epoch": 1.2954048140043763,
+      "grad_norm": 15.25,
+      "learning_rate": 7.058422041481581e-07,
+      "loss": 0.3319,
+      "step": 5328
+    },
+    {
+      "epoch": 1.2956479455385363,
+      "grad_norm": 25.5,
+      "learning_rate": 7.054071767830123e-07,
+      "loss": 0.7161,
+      "step": 5329
+    },
+    {
+      "epoch": 1.2958910770726964,
+      "grad_norm": 22.875,
+      "learning_rate": 7.049722308166699e-07,
+      "loss": 0.6475,
+      "step": 5330
+    },
+    {
+      "epoch": 1.2961342086068564,
+      "grad_norm": 20.375,
+      "learning_rate": 7.045373663141406e-07,
+      "loss": 0.9005,
+      "step": 5331
+    },
+    {
+      "epoch": 1.2963773401410164,
+      "grad_norm": 21.75,
+      "learning_rate": 7.041025833404216e-07,
+      "loss": 0.9873,
+      "step": 5332
+    },
+    {
+      "epoch": 1.2966204716751761,
+      "grad_norm": 17.25,
+      "learning_rate": 7.036678819605003e-07,
+      "loss": 0.6963,
+      "step": 5333
+    },
+    {
+      "epoch": 1.2968636032093364,
+      "grad_norm": 21.875,
+      "learning_rate": 7.032332622393488e-07,
+      "loss": 0.8343,
+      "step": 5334
+    },
+    {
+      "epoch": 1.2971067347434961,
+      "grad_norm": 26.125,
+      "learning_rate": 7.027987242419299e-07,
+      "loss": 1.0523,
+      "step": 5335
+    },
+    {
+      "epoch": 1.2973498662776561,
+      "grad_norm": 16.75,
+      "learning_rate": 7.023642680331915e-07,
+      "loss": 0.5916,
+      "step": 5336
+    },
+    {
+      "epoch": 1.2975929978118161,
+      "grad_norm": 16.875,
+      "learning_rate": 7.019298936780719e-07,
+      "loss": 0.6827,
+      "step": 5337
+    },
+    {
+      "epoch": 1.2978361293459761,
+      "grad_norm": 23.625,
+      "learning_rate": 7.01495601241495e-07,
+      "loss": 0.9354,
+      "step": 5338
+    },
+    {
+      "epoch": 1.2980792608801361,
+      "grad_norm": 24.75,
+      "learning_rate": 7.010613907883731e-07,
+      "loss": 0.7278,
+      "step": 5339
+    },
+    {
+      "epoch": 1.2983223924142961,
+      "grad_norm": 17.625,
+      "learning_rate": 7.006272623836075e-07,
+      "loss": 0.5342,
+      "step": 5340
+    },
+    {
+      "epoch": 1.2985655239484561,
+      "grad_norm": 16.875,
+      "learning_rate": 7.001932160920849e-07,
+      "loss": 0.8626,
+      "step": 5341
+    },
+    {
+      "epoch": 1.2988086554826161,
+      "grad_norm": 20.25,
+      "learning_rate": 6.99759251978682e-07,
+      "loss": 0.6627,
+      "step": 5342
+    },
+    {
+      "epoch": 1.2990517870167761,
+      "grad_norm": 24.5,
+      "learning_rate": 6.993253701082617e-07,
+      "loss": 0.8373,
+      "step": 5343
+    },
+    {
+      "epoch": 1.299294918550936,
+      "grad_norm": 24.0,
+      "learning_rate": 6.988915705456756e-07,
+      "loss": 0.9368,
+      "step": 5344
+    },
+    {
+      "epoch": 1.2995380500850962,
+      "grad_norm": 20.125,
+      "learning_rate": 6.984578533557624e-07,
+      "loss": 0.7427,
+      "step": 5345
+    },
+    {
+      "epoch": 1.299781181619256,
+      "grad_norm": 16.0,
+      "learning_rate": 6.980242186033477e-07,
+      "loss": 0.589,
+      "step": 5346
+    },
+    {
+      "epoch": 1.300024313153416,
+      "grad_norm": 11.6875,
+      "learning_rate": 6.975906663532473e-07,
+      "loss": 0.2786,
+      "step": 5347
+    },
+    {
+      "epoch": 1.300267444687576,
+      "grad_norm": 18.625,
+      "learning_rate": 6.97157196670262e-07,
+      "loss": 0.846,
+      "step": 5348
+    },
+    {
+      "epoch": 1.300510576221736,
+      "grad_norm": 13.875,
+      "learning_rate": 6.967238096191806e-07,
+      "loss": 0.249,
+      "step": 5349
+    },
+    {
+      "epoch": 1.300753707755896,
+      "grad_norm": 25.5,
+      "learning_rate": 6.962905052647824e-07,
+      "loss": 0.8746,
+      "step": 5350
+    },
+    {
+      "epoch": 1.300996839290056,
+      "grad_norm": 15.125,
+      "learning_rate": 6.958572836718311e-07,
+      "loss": 0.5124,
+      "step": 5351
+    },
+    {
+      "epoch": 1.301239970824216,
+      "grad_norm": 24.75,
+      "learning_rate": 6.95424144905079e-07,
+      "loss": 0.4673,
+      "step": 5352
+    },
+    {
+      "epoch": 1.301483102358376,
+      "grad_norm": 26.125,
+      "learning_rate": 6.949910890292657e-07,
+      "loss": 0.6952,
+      "step": 5353
+    },
+    {
+      "epoch": 1.301726233892536,
+      "grad_norm": 21.25,
+      "learning_rate": 6.945581161091203e-07,
+      "loss": 0.7587,
+      "step": 5354
+    },
+    {
+      "epoch": 1.3019693654266957,
+      "grad_norm": 17.625,
+      "learning_rate": 6.94125226209357e-07,
+      "loss": 0.4456,
+      "step": 5355
+    },
+    {
+      "epoch": 1.302212496960856,
+      "grad_norm": 21.25,
+      "learning_rate": 6.936924193946787e-07,
+      "loss": 0.7767,
+      "step": 5356
+    },
+    {
+      "epoch": 1.3024556284950157,
+      "grad_norm": 22.75,
+      "learning_rate": 6.932596957297757e-07,
+      "loss": 0.8849,
+      "step": 5357
+    },
+    {
+      "epoch": 1.3026987600291757,
+      "grad_norm": 22.0,
+      "learning_rate": 6.928270552793273e-07,
+      "loss": 0.7143,
+      "step": 5358
+    },
+    {
+      "epoch": 1.3029418915633357,
+      "grad_norm": 18.375,
+      "learning_rate": 6.923944981079981e-07,
+      "loss": 0.4373,
+      "step": 5359
+    },
+    {
+      "epoch": 1.3031850230974957,
+      "grad_norm": 23.375,
+      "learning_rate": 6.919620242804406e-07,
+      "loss": 0.7099,
+      "step": 5360
+    },
+    {
+      "epoch": 1.3034281546316557,
+      "grad_norm": 27.5,
+      "learning_rate": 6.91529633861297e-07,
+      "loss": 1.33,
+      "step": 5361
+    },
+    {
+      "epoch": 1.3036712861658157,
+      "grad_norm": 26.25,
+      "learning_rate": 6.910973269151946e-07,
+      "loss": 0.7442,
+      "step": 5362
+    },
+    {
+      "epoch": 1.3039144176999757,
+      "grad_norm": 19.0,
+      "learning_rate": 6.906651035067488e-07,
+      "loss": 0.8893,
+      "step": 5363
+    },
+    {
+      "epoch": 1.3041575492341357,
+      "grad_norm": 17.25,
+      "learning_rate": 6.902329637005639e-07,
+      "loss": 0.651,
+      "step": 5364
+    },
+    {
+      "epoch": 1.3044006807682957,
+      "grad_norm": 20.375,
+      "learning_rate": 6.898009075612294e-07,
+      "loss": 0.7097,
+      "step": 5365
+    },
+    {
+      "epoch": 1.3046438123024555,
+      "grad_norm": 19.25,
+      "learning_rate": 6.893689351533245e-07,
+      "loss": 0.688,
+      "step": 5366
+    },
+    {
+      "epoch": 1.3048869438366157,
+      "grad_norm": 18.625,
+      "learning_rate": 6.88937046541415e-07,
+      "loss": 0.9324,
+      "step": 5367
+    },
+    {
+      "epoch": 1.3051300753707755,
+      "grad_norm": 15.1875,
+      "learning_rate": 6.885052417900537e-07,
+      "loss": 0.502,
+      "step": 5368
+    },
+    {
+      "epoch": 1.3053732069049355,
+      "grad_norm": 18.625,
+      "learning_rate": 6.880735209637815e-07,
+      "loss": 0.7486,
+      "step": 5369
+    },
+    {
+      "epoch": 1.3056163384390955,
+      "grad_norm": 21.875,
+      "learning_rate": 6.87641884127126e-07,
+      "loss": 0.6823,
+      "step": 5370
+    },
+    {
+      "epoch": 1.3058594699732555,
+      "grad_norm": 14.6875,
+      "learning_rate": 6.872103313446036e-07,
+      "loss": 0.3315,
+      "step": 5371
+    },
+    {
+      "epoch": 1.3061026015074155,
+      "grad_norm": 17.5,
+      "learning_rate": 6.867788626807162e-07,
+      "loss": 0.7381,
+      "step": 5372
+    },
+    {
+      "epoch": 1.3063457330415755,
+      "grad_norm": 20.875,
+      "learning_rate": 6.863474781999552e-07,
+      "loss": 0.4472,
+      "step": 5373
+    },
+    {
+      "epoch": 1.3065888645757355,
+      "grad_norm": 24.875,
+      "learning_rate": 6.859161779667984e-07,
+      "loss": 1.0121,
+      "step": 5374
+    },
+    {
+      "epoch": 1.3068319961098955,
+      "grad_norm": 21.125,
+      "learning_rate": 6.85484962045711e-07,
+      "loss": 0.8792,
+      "step": 5375
+    },
+    {
+      "epoch": 1.3070751276440555,
+      "grad_norm": 21.75,
+      "learning_rate": 6.850538305011457e-07,
+      "loss": 0.596,
+      "step": 5376
+    },
+    {
+      "epoch": 1.3073182591782153,
+      "grad_norm": 24.875,
+      "learning_rate": 6.846227833975417e-07,
+      "loss": 0.6542,
+      "step": 5377
+    },
+    {
+      "epoch": 1.3075613907123753,
+      "grad_norm": 18.0,
+      "learning_rate": 6.841918207993279e-07,
+      "loss": 0.77,
+      "step": 5378
+    },
+    {
+      "epoch": 1.3078045222465353,
+      "grad_norm": 22.375,
+      "learning_rate": 6.837609427709179e-07,
+      "loss": 0.7198,
+      "step": 5379
+    },
+    {
+      "epoch": 1.3080476537806953,
+      "grad_norm": 24.75,
+      "learning_rate": 6.833301493767136e-07,
+      "loss": 0.5151,
+      "step": 5380
+    },
+    {
+      "epoch": 1.3082907853148553,
+      "grad_norm": 19.375,
+      "learning_rate": 6.828994406811062e-07,
+      "loss": 0.6333,
+      "step": 5381
+    },
+    {
+      "epoch": 1.3085339168490153,
+      "grad_norm": 19.375,
+      "learning_rate": 6.824688167484713e-07,
+      "loss": 0.8866,
+      "step": 5382
+    },
+    {
+      "epoch": 1.3087770483831753,
+      "grad_norm": 18.75,
+      "learning_rate": 6.820382776431732e-07,
+      "loss": 0.5126,
+      "step": 5383
+    },
+    {
+      "epoch": 1.3090201799173353,
+      "grad_norm": 18.875,
+      "learning_rate": 6.816078234295638e-07,
+      "loss": 0.6611,
+      "step": 5384
+    },
+    {
+      "epoch": 1.3092633114514953,
+      "grad_norm": 24.875,
+      "learning_rate": 6.811774541719818e-07,
+      "loss": 0.7533,
+      "step": 5385
+    },
+    {
+      "epoch": 1.309506442985655,
+      "grad_norm": 21.375,
+      "learning_rate": 6.807471699347531e-07,
+      "loss": 0.9855,
+      "step": 5386
+    },
+    {
+      "epoch": 1.3097495745198153,
+      "grad_norm": 17.375,
+      "learning_rate": 6.803169707821907e-07,
+      "loss": 0.6446,
+      "step": 5387
+    },
+    {
+      "epoch": 1.309992706053975,
+      "grad_norm": 17.625,
+      "learning_rate": 6.798868567785958e-07,
+      "loss": 0.4483,
+      "step": 5388
+    },
+    {
+      "epoch": 1.310235837588135,
+      "grad_norm": 21.125,
+      "learning_rate": 6.794568279882567e-07,
+      "loss": 0.4895,
+      "step": 5389
+    },
+    {
+      "epoch": 1.3104789691222951,
+      "grad_norm": 20.125,
+      "learning_rate": 6.79026884475448e-07,
+      "loss": 0.8769,
+      "step": 5390
+    },
+    {
+      "epoch": 1.3107221006564551,
+      "grad_norm": 20.0,
+      "learning_rate": 6.785970263044328e-07,
+      "loss": 0.6661,
+      "step": 5391
+    },
+    {
+      "epoch": 1.3109652321906151,
+      "grad_norm": 19.375,
+      "learning_rate": 6.781672535394605e-07,
+      "loss": 0.5786,
+      "step": 5392
+    },
+    {
+      "epoch": 1.3112083637247751,
+      "grad_norm": 21.25,
+      "learning_rate": 6.777375662447682e-07,
+      "loss": 0.5528,
+      "step": 5393
+    },
+    {
+      "epoch": 1.3114514952589351,
+      "grad_norm": 21.5,
+      "learning_rate": 6.773079644845792e-07,
+      "loss": 0.3921,
+      "step": 5394
+    },
+    {
+      "epoch": 1.3116946267930951,
+      "grad_norm": 22.25,
+      "learning_rate": 6.768784483231064e-07,
+      "loss": 0.9042,
+      "step": 5395
+    },
+    {
+      "epoch": 1.3119377583272551,
+      "grad_norm": 18.5,
+      "learning_rate": 6.76449017824547e-07,
+      "loss": 0.9287,
+      "step": 5396
+    },
+    {
+      "epoch": 1.312180889861415,
+      "grad_norm": 21.875,
+      "learning_rate": 6.760196730530874e-07,
+      "loss": 0.8828,
+      "step": 5397
+    },
+    {
+      "epoch": 1.3124240213955751,
+      "grad_norm": 17.25,
+      "learning_rate": 6.755904140729014e-07,
+      "loss": 0.4136,
+      "step": 5398
+    },
+    {
+      "epoch": 1.312667152929735,
+      "grad_norm": 19.875,
+      "learning_rate": 6.751612409481482e-07,
+      "loss": 0.7068,
+      "step": 5399
+    },
+    {
+      "epoch": 1.312910284463895,
+      "grad_norm": 25.125,
+      "learning_rate": 6.747321537429749e-07,
+      "loss": 0.9837,
+      "step": 5400
+    },
+    {
+      "epoch": 1.313153415998055,
+      "grad_norm": 21.25,
+      "learning_rate": 6.743031525215169e-07,
+      "loss": 0.4716,
+      "step": 5401
+    },
+    {
+      "epoch": 1.313396547532215,
+      "grad_norm": 20.5,
+      "learning_rate": 6.738742373478954e-07,
+      "loss": 0.6532,
+      "step": 5402
+    },
+    {
+      "epoch": 1.313639679066375,
+      "grad_norm": 29.0,
+      "learning_rate": 6.734454082862188e-07,
+      "loss": 1.0469,
+      "step": 5403
+    },
+    {
+      "epoch": 1.313882810600535,
+      "grad_norm": 34.75,
+      "learning_rate": 6.730166654005832e-07,
+      "loss": 0.7992,
+      "step": 5404
+    },
+    {
+      "epoch": 1.314125942134695,
+      "grad_norm": 34.75,
+      "learning_rate": 6.725880087550723e-07,
+      "loss": 1.3177,
+      "step": 5405
+    },
+    {
+      "epoch": 1.314369073668855,
+      "grad_norm": 17.75,
+      "learning_rate": 6.721594384137558e-07,
+      "loss": 0.6191,
+      "step": 5406
+    },
+    {
+      "epoch": 1.314612205203015,
+      "grad_norm": 17.875,
+      "learning_rate": 6.717309544406905e-07,
+      "loss": 0.8107,
+      "step": 5407
+    },
+    {
+      "epoch": 1.3148553367371747,
+      "grad_norm": 23.25,
+      "learning_rate": 6.713025568999216e-07,
+      "loss": 0.5281,
+      "step": 5408
+    },
+    {
+      "epoch": 1.315098468271335,
+      "grad_norm": 23.875,
+      "learning_rate": 6.708742458554798e-07,
+      "loss": 0.5955,
+      "step": 5409
+    },
+    {
+      "epoch": 1.3153415998054947,
+      "grad_norm": 19.5,
+      "learning_rate": 6.70446021371384e-07,
+      "loss": 0.5887,
+      "step": 5410
+    },
+    {
+      "epoch": 1.3155847313396547,
+      "grad_norm": 17.625,
+      "learning_rate": 6.700178835116383e-07,
+      "loss": 0.4819,
+      "step": 5411
+    },
+    {
+      "epoch": 1.3158278628738147,
+      "grad_norm": 20.0,
+      "learning_rate": 6.695898323402378e-07,
+      "loss": 0.6117,
+      "step": 5412
+    },
+    {
+      "epoch": 1.3160709944079747,
+      "grad_norm": 21.125,
+      "learning_rate": 6.691618679211609e-07,
+      "loss": 0.5834,
+      "step": 5413
+    },
+    {
+      "epoch": 1.3163141259421347,
+      "grad_norm": 16.75,
+      "learning_rate": 6.687339903183735e-07,
+      "loss": 0.4409,
+      "step": 5414
+    },
+    {
+      "epoch": 1.3165572574762947,
+      "grad_norm": 30.25,
+      "learning_rate": 6.683061995958308e-07,
+      "loss": 1.1829,
+      "step": 5415
+    },
+    {
+      "epoch": 1.3168003890104547,
+      "grad_norm": 32.5,
+      "learning_rate": 6.678784958174728e-07,
+      "loss": 0.8507,
+      "step": 5416
+    },
+    {
+      "epoch": 1.3170435205446147,
+      "grad_norm": 18.375,
+      "learning_rate": 6.674508790472264e-07,
+      "loss": 1.0468,
+      "step": 5417
+    },
+    {
+      "epoch": 1.3172866520787747,
+      "grad_norm": 21.375,
+      "learning_rate": 6.670233493490079e-07,
+      "loss": 0.6007,
+      "step": 5418
+    },
+    {
+      "epoch": 1.3175297836129345,
+      "grad_norm": 25.625,
+      "learning_rate": 6.665959067867175e-07,
+      "loss": 0.8114,
+      "step": 5419
+    },
+    {
+      "epoch": 1.3177729151470947,
+      "grad_norm": 22.125,
+      "learning_rate": 6.661685514242453e-07,
+      "loss": 0.5541,
+      "step": 5420
+    },
+    {
+      "epoch": 1.3180160466812545,
+      "grad_norm": 26.125,
+      "learning_rate": 6.657412833254655e-07,
+      "loss": 0.879,
+      "step": 5421
+    },
+    {
+      "epoch": 1.3182591782154145,
+      "grad_norm": 43.0,
+      "learning_rate": 6.65314102554242e-07,
+      "loss": 0.5993,
+      "step": 5422
+    },
+    {
+      "epoch": 1.3185023097495745,
+      "grad_norm": 27.0,
+      "learning_rate": 6.648870091744236e-07,
+      "loss": 0.5893,
+      "step": 5423
+    },
+    {
+      "epoch": 1.3187454412837345,
+      "grad_norm": 16.75,
+      "learning_rate": 6.644600032498466e-07,
+      "loss": 0.3996,
+      "step": 5424
+    },
+    {
+      "epoch": 1.3189885728178945,
+      "grad_norm": 18.25,
+      "learning_rate": 6.640330848443352e-07,
+      "loss": 0.6858,
+      "step": 5425
+    },
+    {
+      "epoch": 1.3192317043520545,
+      "grad_norm": 14.625,
+      "learning_rate": 6.636062540216987e-07,
+      "loss": 0.2918,
+      "step": 5426
+    },
+    {
+      "epoch": 1.3194748358862145,
+      "grad_norm": 25.375,
+      "learning_rate": 6.631795108457354e-07,
+      "loss": 0.9733,
+      "step": 5427
+    },
+    {
+      "epoch": 1.3197179674203745,
+      "grad_norm": 15.0,
+      "learning_rate": 6.627528553802285e-07,
+      "loss": 0.4875,
+      "step": 5428
+    },
+    {
+      "epoch": 1.3199610989545345,
+      "grad_norm": 16.75,
+      "learning_rate": 6.6232628768895e-07,
+      "loss": 0.429,
+      "step": 5429
+    },
+    {
+      "epoch": 1.3202042304886943,
+      "grad_norm": 16.5,
+      "learning_rate": 6.618998078356573e-07,
+      "loss": 0.5399,
+      "step": 5430
+    },
+    {
+      "epoch": 1.3204473620228543,
+      "grad_norm": 23.5,
+      "learning_rate": 6.614734158840946e-07,
+      "loss": 0.6783,
+      "step": 5431
+    },
+    {
+      "epoch": 1.3206904935570143,
+      "grad_norm": 19.0,
+      "learning_rate": 6.610471118979946e-07,
+      "loss": 0.7945,
+      "step": 5432
+    },
+    {
+      "epoch": 1.3209336250911743,
+      "grad_norm": 31.75,
+      "learning_rate": 6.606208959410754e-07,
+      "loss": 1.2546,
+      "step": 5433
+    },
+    {
+      "epoch": 1.3211767566253343,
+      "grad_norm": 24.125,
+      "learning_rate": 6.601947680770415e-07,
+      "loss": 0.8433,
+      "step": 5434
+    },
+    {
+      "epoch": 1.3214198881594943,
+      "grad_norm": 26.875,
+      "learning_rate": 6.59768728369586e-07,
+      "loss": 0.6929,
+      "step": 5435
+    },
+    {
+      "epoch": 1.3216630196936543,
+      "grad_norm": 17.25,
+      "learning_rate": 6.593427768823881e-07,
+      "loss": 0.4186,
+      "step": 5436
+    },
+    {
+      "epoch": 1.3219061512278143,
+      "grad_norm": 14.0625,
+      "learning_rate": 6.589169136791132e-07,
+      "loss": 0.5019,
+      "step": 5437
+    },
+    {
+      "epoch": 1.3221492827619743,
+      "grad_norm": 18.0,
+      "learning_rate": 6.584911388234135e-07,
+      "loss": 0.7286,
+      "step": 5438
+    },
+    {
+      "epoch": 1.322392414296134,
+      "grad_norm": 29.375,
+      "learning_rate": 6.580654523789292e-07,
+      "loss": 1.0972,
+      "step": 5439
+    },
+    {
+      "epoch": 1.3226355458302943,
+      "grad_norm": 20.25,
+      "learning_rate": 6.576398544092861e-07,
+      "loss": 0.8966,
+      "step": 5440
+    },
+    {
+      "epoch": 1.322878677364454,
+      "grad_norm": 29.125,
+      "learning_rate": 6.572143449780969e-07,
+      "loss": 0.7435,
+      "step": 5441
+    },
+    {
+      "epoch": 1.323121808898614,
+      "grad_norm": 20.0,
+      "learning_rate": 6.567889241489614e-07,
+      "loss": 0.6646,
+      "step": 5442
+    },
+    {
+      "epoch": 1.323364940432774,
+      "grad_norm": 22.0,
+      "learning_rate": 6.563635919854668e-07,
+      "loss": 0.6163,
+      "step": 5443
+    },
+    {
+      "epoch": 1.323608071966934,
+      "grad_norm": 20.875,
+      "learning_rate": 6.559383485511859e-07,
+      "loss": 0.5634,
+      "step": 5444
+    },
+    {
+      "epoch": 1.323851203501094,
+      "grad_norm": 17.5,
+      "learning_rate": 6.55513193909678e-07,
+      "loss": 0.5734,
+      "step": 5445
+    },
+    {
+      "epoch": 1.324094335035254,
+      "grad_norm": 17.875,
+      "learning_rate": 6.55088128124491e-07,
+      "loss": 0.7802,
+      "step": 5446
+    },
+    {
+      "epoch": 1.324337466569414,
+      "grad_norm": 16.125,
+      "learning_rate": 6.546631512591577e-07,
+      "loss": 0.4235,
+      "step": 5447
+    },
+    {
+      "epoch": 1.324580598103574,
+      "grad_norm": 24.0,
+      "learning_rate": 6.542382633771978e-07,
+      "loss": 0.943,
+      "step": 5448
+    },
+    {
+      "epoch": 1.324823729637734,
+      "grad_norm": 18.625,
+      "learning_rate": 6.53813464542119e-07,
+      "loss": 0.6957,
+      "step": 5449
+    },
+    {
+      "epoch": 1.3250668611718939,
+      "grad_norm": 17.625,
+      "learning_rate": 6.53388754817414e-07,
+      "loss": 0.7104,
+      "step": 5450
+    },
+    {
+      "epoch": 1.325309992706054,
+      "grad_norm": 25.75,
+      "learning_rate": 6.529641342665641e-07,
+      "loss": 0.9805,
+      "step": 5451
+    },
+    {
+      "epoch": 1.3255531242402139,
+      "grad_norm": 16.375,
+      "learning_rate": 6.525396029530349e-07,
+      "loss": 0.5378,
+      "step": 5452
+    },
+    {
+      "epoch": 1.3257962557743739,
+      "grad_norm": 16.5,
+      "learning_rate": 6.521151609402808e-07,
+      "loss": 0.3958,
+      "step": 5453
+    },
+    {
+      "epoch": 1.3260393873085339,
+      "grad_norm": 17.75,
+      "learning_rate": 6.51690808291742e-07,
+      "loss": 0.6653,
+      "step": 5454
+    },
+    {
+      "epoch": 1.3262825188426939,
+      "grad_norm": 18.125,
+      "learning_rate": 6.512665450708445e-07,
+      "loss": 0.4327,
+      "step": 5455
+    },
+    {
+      "epoch": 1.3265256503768539,
+      "grad_norm": 21.375,
+      "learning_rate": 6.508423713410027e-07,
+      "loss": 0.6051,
+      "step": 5456
+    },
+    {
+      "epoch": 1.3267687819110139,
+      "grad_norm": 20.5,
+      "learning_rate": 6.504182871656159e-07,
+      "loss": 0.6215,
+      "step": 5457
+    },
+    {
+      "epoch": 1.3270119134451739,
+      "grad_norm": 14.125,
+      "learning_rate": 6.499942926080716e-07,
+      "loss": 0.2769,
+      "step": 5458
+    },
+    {
+      "epoch": 1.3272550449793339,
+      "grad_norm": 20.625,
+      "learning_rate": 6.495703877317425e-07,
+      "loss": 0.872,
+      "step": 5459
+    },
+    {
+      "epoch": 1.3274981765134939,
+      "grad_norm": 19.25,
+      "learning_rate": 6.49146572599989e-07,
+      "loss": 0.7892,
+      "step": 5460
+    },
+    {
+      "epoch": 1.3277413080476537,
+      "grad_norm": 22.0,
+      "learning_rate": 6.487228472761572e-07,
+      "loss": 0.9695,
+      "step": 5461
+    },
+    {
+      "epoch": 1.327984439581814,
+      "grad_norm": 21.125,
+      "learning_rate": 6.482992118235801e-07,
+      "loss": 0.7597,
+      "step": 5462
+    },
+    {
+      "epoch": 1.3282275711159737,
+      "grad_norm": 20.375,
+      "learning_rate": 6.478756663055779e-07,
+      "loss": 0.805,
+      "step": 5463
+    },
+    {
+      "epoch": 1.3284707026501337,
+      "grad_norm": 20.25,
+      "learning_rate": 6.474522107854564e-07,
+      "loss": 1.0494,
+      "step": 5464
+    },
+    {
+      "epoch": 1.3287138341842937,
+      "grad_norm": 20.0,
+      "learning_rate": 6.470288453265079e-07,
+      "loss": 0.3861,
+      "step": 5465
+    },
+    {
+      "epoch": 1.3289569657184537,
+      "grad_norm": 33.25,
+      "learning_rate": 6.466055699920121e-07,
+      "loss": 0.9758,
+      "step": 5466
+    },
+    {
+      "epoch": 1.3292000972526137,
+      "grad_norm": 19.125,
+      "learning_rate": 6.461823848452353e-07,
+      "loss": 0.8472,
+      "step": 5467
+    },
+    {
+      "epoch": 1.3294432287867737,
+      "grad_norm": 19.625,
+      "learning_rate": 6.457592899494295e-07,
+      "loss": 0.7136,
+      "step": 5468
+    },
+    {
+      "epoch": 1.3296863603209337,
+      "grad_norm": 24.125,
+      "learning_rate": 6.45336285367833e-07,
+      "loss": 0.9375,
+      "step": 5469
+    },
+    {
+      "epoch": 1.3299294918550937,
+      "grad_norm": 14.75,
+      "learning_rate": 6.449133711636718e-07,
+      "loss": 0.2902,
+      "step": 5470
+    },
+    {
+      "epoch": 1.3301726233892537,
+      "grad_norm": 18.875,
+      "learning_rate": 6.444905474001576e-07,
+      "loss": 0.6764,
+      "step": 5471
+    },
+    {
+      "epoch": 1.3304157549234135,
+      "grad_norm": 25.375,
+      "learning_rate": 6.440678141404883e-07,
+      "loss": 0.4396,
+      "step": 5472
+    },
+    {
+      "epoch": 1.3306588864575735,
+      "grad_norm": 21.25,
+      "learning_rate": 6.436451714478487e-07,
+      "loss": 0.6173,
+      "step": 5473
+    },
+    {
+      "epoch": 1.3309020179917335,
+      "grad_norm": 21.75,
+      "learning_rate": 6.432226193854109e-07,
+      "loss": 0.6661,
+      "step": 5474
+    },
+    {
+      "epoch": 1.3311451495258935,
+      "grad_norm": 20.625,
+      "learning_rate": 6.42800158016332e-07,
+      "loss": 0.8773,
+      "step": 5475
+    },
+    {
+      "epoch": 1.3313882810600535,
+      "grad_norm": 16.5,
+      "learning_rate": 6.423777874037558e-07,
+      "loss": 0.3369,
+      "step": 5476
+    },
+    {
+      "epoch": 1.3316314125942135,
+      "grad_norm": 21.125,
+      "learning_rate": 6.419555076108137e-07,
+      "loss": 0.5003,
+      "step": 5477
+    },
+    {
+      "epoch": 1.3318745441283735,
+      "grad_norm": 25.75,
+      "learning_rate": 6.415333187006223e-07,
+      "loss": 0.7549,
+      "step": 5478
+    },
+    {
+      "epoch": 1.3321176756625335,
+      "grad_norm": 22.0,
+      "learning_rate": 6.411112207362846e-07,
+      "loss": 0.777,
+      "step": 5479
+    },
+    {
+      "epoch": 1.3323608071966935,
+      "grad_norm": 24.625,
+      "learning_rate": 6.406892137808907e-07,
+      "loss": 0.8066,
+      "step": 5480
+    },
+    {
+      "epoch": 1.3326039387308533,
+      "grad_norm": 21.5,
+      "learning_rate": 6.402672978975175e-07,
+      "loss": 0.8785,
+      "step": 5481
+    },
+    {
+      "epoch": 1.3328470702650135,
+      "grad_norm": 33.5,
+      "learning_rate": 6.398454731492271e-07,
+      "loss": 0.7769,
+      "step": 5482
+    },
+    {
+      "epoch": 1.3330902017991733,
+      "grad_norm": 21.125,
+      "learning_rate": 6.39423739599068e-07,
+      "loss": 0.7673,
+      "step": 5483
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 22.5,
+      "learning_rate": 6.390020973100763e-07,
+      "loss": 1.0259,
+      "step": 5484
+    },
+    {
+      "epoch": 1.3335764648674933,
+      "grad_norm": 29.125,
+      "learning_rate": 6.385805463452738e-07,
+      "loss": 0.5669,
+      "step": 5485
+    },
+    {
+      "epoch": 1.3338195964016533,
+      "grad_norm": 22.625,
+      "learning_rate": 6.381590867676677e-07,
+      "loss": 0.7433,
+      "step": 5486
+    },
+    {
+      "epoch": 1.3340627279358133,
+      "grad_norm": 16.125,
+      "learning_rate": 6.377377186402531e-07,
+      "loss": 0.5105,
+      "step": 5487
+    },
+    {
+      "epoch": 1.3343058594699733,
+      "grad_norm": 19.75,
+      "learning_rate": 6.373164420260103e-07,
+      "loss": 1.1968,
+      "step": 5488
+    },
+    {
+      "epoch": 1.3345489910041333,
+      "grad_norm": 22.875,
+      "learning_rate": 6.368952569879072e-07,
+      "loss": 0.9597,
+      "step": 5489
+    },
+    {
+      "epoch": 1.3347921225382933,
+      "grad_norm": 25.0,
+      "learning_rate": 6.36474163588896e-07,
+      "loss": 0.3777,
+      "step": 5490
+    },
+    {
+      "epoch": 1.3350352540724533,
+      "grad_norm": 33.0,
+      "learning_rate": 6.360531618919177e-07,
+      "loss": 0.9988,
+      "step": 5491
+    },
+    {
+      "epoch": 1.335278385606613,
+      "grad_norm": 25.125,
+      "learning_rate": 6.356322519598974e-07,
+      "loss": 0.7696,
+      "step": 5492
+    },
+    {
+      "epoch": 1.3355215171407733,
+      "grad_norm": 19.0,
+      "learning_rate": 6.352114338557472e-07,
+      "loss": 0.8448,
+      "step": 5493
+    },
+    {
+      "epoch": 1.335764648674933,
+      "grad_norm": 24.5,
+      "learning_rate": 6.347907076423662e-07,
+      "loss": 0.6717,
+      "step": 5494
+    },
+    {
+      "epoch": 1.336007780209093,
+      "grad_norm": 21.875,
+      "learning_rate": 6.343700733826393e-07,
+      "loss": 0.9157,
+      "step": 5495
+    },
+    {
+      "epoch": 1.336250911743253,
+      "grad_norm": 17.375,
+      "learning_rate": 6.339495311394365e-07,
+      "loss": 0.4617,
+      "step": 5496
+    },
+    {
+      "epoch": 1.336494043277413,
+      "grad_norm": 28.0,
+      "learning_rate": 6.33529080975616e-07,
+      "loss": 0.8627,
+      "step": 5497
+    },
+    {
+      "epoch": 1.336737174811573,
+      "grad_norm": 15.5,
+      "learning_rate": 6.331087229540217e-07,
+      "loss": 0.6037,
+      "step": 5498
+    },
+    {
+      "epoch": 1.336980306345733,
+      "grad_norm": 16.375,
+      "learning_rate": 6.326884571374826e-07,
+      "loss": 0.3355,
+      "step": 5499
+    },
+    {
+      "epoch": 1.337223437879893,
+      "grad_norm": 20.75,
+      "learning_rate": 6.322682835888145e-07,
+      "loss": 0.6281,
+      "step": 5500
+    },
+    {
+      "epoch": 1.337466569414053,
+      "grad_norm": 17.375,
+      "learning_rate": 6.318482023708205e-07,
+      "loss": 0.3225,
+      "step": 5501
+    },
+    {
+      "epoch": 1.337709700948213,
+      "grad_norm": 18.625,
+      "learning_rate": 6.314282135462888e-07,
+      "loss": 0.8917,
+      "step": 5502
+    },
+    {
+      "epoch": 1.3379528324823728,
+      "grad_norm": 26.375,
+      "learning_rate": 6.310083171779929e-07,
+      "loss": 0.7172,
+      "step": 5503
+    },
+    {
+      "epoch": 1.338195964016533,
+      "grad_norm": 20.125,
+      "learning_rate": 6.305885133286943e-07,
+      "loss": 0.5759,
+      "step": 5504
+    },
+    {
+      "epoch": 1.3384390955506928,
+      "grad_norm": 14.625,
+      "learning_rate": 6.301688020611408e-07,
+      "loss": 0.3775,
+      "step": 5505
+    },
+    {
+      "epoch": 1.3386822270848528,
+      "grad_norm": 17.0,
+      "learning_rate": 6.297491834380644e-07,
+      "loss": 0.7441,
+      "step": 5506
+    },
+    {
+      "epoch": 1.3389253586190129,
+      "grad_norm": 26.5,
+      "learning_rate": 6.293296575221842e-07,
+      "loss": 1.0262,
+      "step": 5507
+    },
+    {
+      "epoch": 1.3391684901531729,
+      "grad_norm": 16.125,
+      "learning_rate": 6.289102243762066e-07,
+      "loss": 0.4134,
+      "step": 5508
+    },
+    {
+      "epoch": 1.3394116216873329,
+      "grad_norm": 30.5,
+      "learning_rate": 6.284908840628225e-07,
+      "loss": 1.225,
+      "step": 5509
+    },
+    {
+      "epoch": 1.3396547532214929,
+      "grad_norm": 18.375,
+      "learning_rate": 6.280716366447092e-07,
+      "loss": 0.4572,
+      "step": 5510
+    },
+    {
+      "epoch": 1.3398978847556529,
+      "grad_norm": 18.5,
+      "learning_rate": 6.276524821845307e-07,
+      "loss": 0.6042,
+      "step": 5511
+    },
+    {
+      "epoch": 1.3401410162898129,
+      "grad_norm": 19.75,
+      "learning_rate": 6.272334207449376e-07,
+      "loss": 0.6372,
+      "step": 5512
+    },
+    {
+      "epoch": 1.3403841478239729,
+      "grad_norm": 19.125,
+      "learning_rate": 6.268144523885656e-07,
+      "loss": 0.8907,
+      "step": 5513
+    },
+    {
+      "epoch": 1.3406272793581326,
+      "grad_norm": 23.25,
+      "learning_rate": 6.263955771780355e-07,
+      "loss": 0.9603,
+      "step": 5514
+    },
+    {
+      "epoch": 1.3408704108922929,
+      "grad_norm": 17.0,
+      "learning_rate": 6.259767951759573e-07,
+      "loss": 0.6493,
+      "step": 5515
+    },
+    {
+      "epoch": 1.3411135424264526,
+      "grad_norm": 21.875,
+      "learning_rate": 6.255581064449242e-07,
+      "loss": 1.1151,
+      "step": 5516
+    },
+    {
+      "epoch": 1.3413566739606126,
+      "grad_norm": 16.5,
+      "learning_rate": 6.25139511047516e-07,
+      "loss": 0.5045,
+      "step": 5517
+    },
+    {
+      "epoch": 1.3415998054947726,
+      "grad_norm": 19.75,
+      "learning_rate": 6.247210090463002e-07,
+      "loss": 0.7112,
+      "step": 5518
+    },
+    {
+      "epoch": 1.3418429370289326,
+      "grad_norm": 30.375,
+      "learning_rate": 6.24302600503828e-07,
+      "loss": 0.6472,
+      "step": 5519
+    },
+    {
+      "epoch": 1.3420860685630926,
+      "grad_norm": 23.0,
+      "learning_rate": 6.238842854826388e-07,
+      "loss": 0.5853,
+      "step": 5520
+    },
+    {
+      "epoch": 1.3423292000972527,
+      "grad_norm": 20.625,
+      "learning_rate": 6.23466064045256e-07,
+      "loss": 0.8197,
+      "step": 5521
+    },
+    {
+      "epoch": 1.3425723316314127,
+      "grad_norm": 24.875,
+      "learning_rate": 6.23047936254191e-07,
+      "loss": 0.8313,
+      "step": 5522
+    },
+    {
+      "epoch": 1.3428154631655727,
+      "grad_norm": 18.75,
+      "learning_rate": 6.2262990217194e-07,
+      "loss": 0.4227,
+      "step": 5523
+    },
+    {
+      "epoch": 1.3430585946997327,
+      "grad_norm": 17.875,
+      "learning_rate": 6.222119618609844e-07,
+      "loss": 0.4389,
+      "step": 5524
+    },
+    {
+      "epoch": 1.3433017262338924,
+      "grad_norm": 18.5,
+      "learning_rate": 6.217941153837941e-07,
+      "loss": 0.6153,
+      "step": 5525
+    },
+    {
+      "epoch": 1.3435448577680524,
+      "grad_norm": 25.875,
+      "learning_rate": 6.213763628028226e-07,
+      "loss": 0.9794,
+      "step": 5526
+    },
+    {
+      "epoch": 1.3437879893022124,
+      "grad_norm": 22.125,
+      "learning_rate": 6.209587041805099e-07,
+      "loss": 0.8706,
+      "step": 5527
+    },
+    {
+      "epoch": 1.3440311208363724,
+      "grad_norm": 17.875,
+      "learning_rate": 6.205411395792827e-07,
+      "loss": 0.786,
+      "step": 5528
+    },
+    {
+      "epoch": 1.3442742523705324,
+      "grad_norm": 17.375,
+      "learning_rate": 6.201236690615541e-07,
+      "loss": 0.8461,
+      "step": 5529
+    },
+    {
+      "epoch": 1.3445173839046924,
+      "grad_norm": 20.0,
+      "learning_rate": 6.197062926897213e-07,
+      "loss": 0.9461,
+      "step": 5530
+    },
+    {
+      "epoch": 1.3447605154388524,
+      "grad_norm": 17.5,
+      "learning_rate": 6.192890105261682e-07,
+      "loss": 0.5896,
+      "step": 5531
+    },
+    {
+      "epoch": 1.3450036469730124,
+      "grad_norm": 15.6875,
+      "learning_rate": 6.188718226332656e-07,
+      "loss": 0.4941,
+      "step": 5532
+    },
+    {
+      "epoch": 1.3452467785071724,
+      "grad_norm": 31.0,
+      "learning_rate": 6.18454729073369e-07,
+      "loss": 0.824,
+      "step": 5533
+    },
+    {
+      "epoch": 1.3454899100413322,
+      "grad_norm": 20.25,
+      "learning_rate": 6.1803772990882e-07,
+      "loss": 0.7432,
+      "step": 5534
+    },
+    {
+      "epoch": 1.3457330415754925,
+      "grad_norm": 17.75,
+      "learning_rate": 6.176208252019467e-07,
+      "loss": 0.7477,
+      "step": 5535
+    },
+    {
+      "epoch": 1.3459761731096522,
+      "grad_norm": 21.5,
+      "learning_rate": 6.172040150150629e-07,
+      "loss": 0.7695,
+      "step": 5536
+    },
+    {
+      "epoch": 1.3462193046438122,
+      "grad_norm": 17.75,
+      "learning_rate": 6.167872994104679e-07,
+      "loss": 0.4621,
+      "step": 5537
+    },
+    {
+      "epoch": 1.3464624361779722,
+      "grad_norm": 17.875,
+      "learning_rate": 6.163706784504466e-07,
+      "loss": 1.0213,
+      "step": 5538
+    },
+    {
+      "epoch": 1.3467055677121322,
+      "grad_norm": 23.0,
+      "learning_rate": 6.159541521972709e-07,
+      "loss": 0.4595,
+      "step": 5539
+    },
+    {
+      "epoch": 1.3469486992462922,
+      "grad_norm": 18.25,
+      "learning_rate": 6.155377207131975e-07,
+      "loss": 0.3787,
+      "step": 5540
+    },
+    {
+      "epoch": 1.3471918307804522,
+      "grad_norm": 18.875,
+      "learning_rate": 6.151213840604688e-07,
+      "loss": 0.9116,
+      "step": 5541
+    },
+    {
+      "epoch": 1.3474349623146122,
+      "grad_norm": 17.75,
+      "learning_rate": 6.147051423013141e-07,
+      "loss": 0.363,
+      "step": 5542
+    },
+    {
+      "epoch": 1.3476780938487722,
+      "grad_norm": 20.375,
+      "learning_rate": 6.142889954979482e-07,
+      "loss": 0.9249,
+      "step": 5543
+    },
+    {
+      "epoch": 1.3479212253829322,
+      "grad_norm": 19.75,
+      "learning_rate": 6.138729437125711e-07,
+      "loss": 1.2321,
+      "step": 5544
+    },
+    {
+      "epoch": 1.348164356917092,
+      "grad_norm": 18.125,
+      "learning_rate": 6.134569870073684e-07,
+      "loss": 0.6476,
+      "step": 5545
+    },
+    {
+      "epoch": 1.3484074884512522,
+      "grad_norm": 18.125,
+      "learning_rate": 6.130411254445129e-07,
+      "loss": 0.4712,
+      "step": 5546
+    },
+    {
+      "epoch": 1.348650619985412,
+      "grad_norm": 18.375,
+      "learning_rate": 6.126253590861618e-07,
+      "loss": 0.5074,
+      "step": 5547
+    },
+    {
+      "epoch": 1.348893751519572,
+      "grad_norm": 18.0,
+      "learning_rate": 6.122096879944582e-07,
+      "loss": 0.741,
+      "step": 5548
+    },
+    {
+      "epoch": 1.349136883053732,
+      "grad_norm": 18.75,
+      "learning_rate": 6.117941122315323e-07,
+      "loss": 0.7042,
+      "step": 5549
+    },
+    {
+      "epoch": 1.349380014587892,
+      "grad_norm": 20.875,
+      "learning_rate": 6.113786318594979e-07,
+      "loss": 0.8076,
+      "step": 5550
+    },
+    {
+      "epoch": 1.349623146122052,
+      "grad_norm": 22.125,
+      "learning_rate": 6.109632469404568e-07,
+      "loss": 0.5434,
+      "step": 5551
+    },
+    {
+      "epoch": 1.349866277656212,
+      "grad_norm": 22.5,
+      "learning_rate": 6.105479575364946e-07,
+      "loss": 0.7464,
+      "step": 5552
+    },
+    {
+      "epoch": 1.350109409190372,
+      "grad_norm": 26.25,
+      "learning_rate": 6.101327637096844e-07,
+      "loss": 0.649,
+      "step": 5553
+    },
+    {
+      "epoch": 1.350352540724532,
+      "grad_norm": 22.0,
+      "learning_rate": 6.097176655220835e-07,
+      "loss": 0.6611,
+      "step": 5554
+    },
+    {
+      "epoch": 1.350595672258692,
+      "grad_norm": 18.125,
+      "learning_rate": 6.09302663035735e-07,
+      "loss": 0.4305,
+      "step": 5555
+    },
+    {
+      "epoch": 1.3508388037928518,
+      "grad_norm": 22.5,
+      "learning_rate": 6.088877563126694e-07,
+      "loss": 0.6376,
+      "step": 5556
+    },
+    {
+      "epoch": 1.351081935327012,
+      "grad_norm": 23.375,
+      "learning_rate": 6.084729454149008e-07,
+      "loss": 0.7713,
+      "step": 5557
+    },
+    {
+      "epoch": 1.3513250668611718,
+      "grad_norm": 21.125,
+      "learning_rate": 6.080582304044297e-07,
+      "loss": 0.7473,
+      "step": 5558
+    },
+    {
+      "epoch": 1.3515681983953318,
+      "grad_norm": 24.875,
+      "learning_rate": 6.076436113432427e-07,
+      "loss": 0.5891,
+      "step": 5559
+    },
+    {
+      "epoch": 1.3518113299294918,
+      "grad_norm": 20.875,
+      "learning_rate": 6.072290882933124e-07,
+      "loss": 0.6971,
+      "step": 5560
+    },
+    {
+      "epoch": 1.3520544614636518,
+      "grad_norm": 22.125,
+      "learning_rate": 6.068146613165959e-07,
+      "loss": 0.5151,
+      "step": 5561
+    },
+    {
+      "epoch": 1.3522975929978118,
+      "grad_norm": 19.875,
+      "learning_rate": 6.06400330475036e-07,
+      "loss": 0.3973,
+      "step": 5562
+    },
+    {
+      "epoch": 1.3525407245319718,
+      "grad_norm": 17.875,
+      "learning_rate": 6.059860958305625e-07,
+      "loss": 0.5082,
+      "step": 5563
+    },
+    {
+      "epoch": 1.3527838560661318,
+      "grad_norm": 19.5,
+      "learning_rate": 6.055719574450895e-07,
+      "loss": 0.4535,
+      "step": 5564
+    },
+    {
+      "epoch": 1.3530269876002918,
+      "grad_norm": 21.625,
+      "learning_rate": 6.051579153805166e-07,
+      "loss": 0.9679,
+      "step": 5565
+    },
+    {
+      "epoch": 1.3532701191344518,
+      "grad_norm": 21.25,
+      "learning_rate": 6.047439696987302e-07,
+      "loss": 0.7651,
+      "step": 5566
+    },
+    {
+      "epoch": 1.3535132506686116,
+      "grad_norm": 15.4375,
+      "learning_rate": 6.043301204616017e-07,
+      "loss": 0.2833,
+      "step": 5567
+    },
+    {
+      "epoch": 1.3537563822027718,
+      "grad_norm": 14.5,
+      "learning_rate": 6.039163677309881e-07,
+      "loss": 0.5003,
+      "step": 5568
+    },
+    {
+      "epoch": 1.3539995137369316,
+      "grad_norm": 17.125,
+      "learning_rate": 6.035027115687309e-07,
+      "loss": 0.595,
+      "step": 5569
+    },
+    {
+      "epoch": 1.3542426452710916,
+      "grad_norm": 21.375,
+      "learning_rate": 6.030891520366594e-07,
+      "loss": 0.8139,
+      "step": 5570
+    },
+    {
+      "epoch": 1.3544857768052516,
+      "grad_norm": 22.625,
+      "learning_rate": 6.026756891965867e-07,
+      "loss": 0.5346,
+      "step": 5571
+    },
+    {
+      "epoch": 1.3547289083394116,
+      "grad_norm": 17.875,
+      "learning_rate": 6.022623231103114e-07,
+      "loss": 0.8117,
+      "step": 5572
+    },
+    {
+      "epoch": 1.3549720398735716,
+      "grad_norm": 17.375,
+      "learning_rate": 6.018490538396188e-07,
+      "loss": 0.5895,
+      "step": 5573
+    },
+    {
+      "epoch": 1.3552151714077316,
+      "grad_norm": 19.875,
+      "learning_rate": 6.014358814462793e-07,
+      "loss": 0.4554,
+      "step": 5574
+    },
+    {
+      "epoch": 1.3554583029418916,
+      "grad_norm": 27.25,
+      "learning_rate": 6.010228059920487e-07,
+      "loss": 1.1678,
+      "step": 5575
+    },
+    {
+      "epoch": 1.3557014344760516,
+      "grad_norm": 24.375,
+      "learning_rate": 6.006098275386674e-07,
+      "loss": 0.8656,
+      "step": 5576
+    },
+    {
+      "epoch": 1.3559445660102116,
+      "grad_norm": 16.125,
+      "learning_rate": 6.001969461478631e-07,
+      "loss": 0.7357,
+      "step": 5577
+    },
+    {
+      "epoch": 1.3561876975443714,
+      "grad_norm": 14.125,
+      "learning_rate": 5.997841618813478e-07,
+      "loss": 0.3415,
+      "step": 5578
+    },
+    {
+      "epoch": 1.3564308290785314,
+      "grad_norm": 24.625,
+      "learning_rate": 5.993714748008186e-07,
+      "loss": 0.6959,
+      "step": 5579
+    },
+    {
+      "epoch": 1.3566739606126914,
+      "grad_norm": 17.125,
+      "learning_rate": 5.989588849679598e-07,
+      "loss": 0.3308,
+      "step": 5580
+    },
+    {
+      "epoch": 1.3569170921468514,
+      "grad_norm": 27.0,
+      "learning_rate": 5.985463924444388e-07,
+      "loss": 0.692,
+      "step": 5581
+    },
+    {
+      "epoch": 1.3571602236810114,
+      "grad_norm": 18.875,
+      "learning_rate": 5.981339972919112e-07,
+      "loss": 0.8544,
+      "step": 5582
+    },
+    {
+      "epoch": 1.3574033552151714,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.977216995720152e-07,
+      "loss": 0.4573,
+      "step": 5583
+    },
+    {
+      "epoch": 1.3576464867493314,
+      "grad_norm": 17.625,
+      "learning_rate": 5.97309499346377e-07,
+      "loss": 0.5268,
+      "step": 5584
+    },
+    {
+      "epoch": 1.3578896182834914,
+      "grad_norm": 19.875,
+      "learning_rate": 5.968973966766064e-07,
+      "loss": 0.7208,
+      "step": 5585
+    },
+    {
+      "epoch": 1.3581327498176514,
+      "grad_norm": 26.25,
+      "learning_rate": 5.964853916242988e-07,
+      "loss": 0.7436,
+      "step": 5586
+    },
+    {
+      "epoch": 1.3583758813518112,
+      "grad_norm": 21.125,
+      "learning_rate": 5.960734842510366e-07,
+      "loss": 0.4939,
+      "step": 5587
+    },
+    {
+      "epoch": 1.3586190128859714,
+      "grad_norm": 15.25,
+      "learning_rate": 5.956616746183859e-07,
+      "loss": 0.4749,
+      "step": 5588
+    },
+    {
+      "epoch": 1.3588621444201312,
+      "grad_norm": 21.25,
+      "learning_rate": 5.952499627878981e-07,
+      "loss": 0.5949,
+      "step": 5589
+    },
+    {
+      "epoch": 1.3591052759542912,
+      "grad_norm": 16.125,
+      "learning_rate": 5.948383488211114e-07,
+      "loss": 0.3801,
+      "step": 5590
+    },
+    {
+      "epoch": 1.3593484074884512,
+      "grad_norm": 15.3125,
+      "learning_rate": 5.94426832779549e-07,
+      "loss": 0.6721,
+      "step": 5591
+    },
+    {
+      "epoch": 1.3595915390226112,
+      "grad_norm": 19.625,
+      "learning_rate": 5.940154147247186e-07,
+      "loss": 0.7625,
+      "step": 5592
+    },
+    {
+      "epoch": 1.3598346705567712,
+      "grad_norm": 16.125,
+      "learning_rate": 5.936040947181132e-07,
+      "loss": 0.4876,
+      "step": 5593
+    },
+    {
+      "epoch": 1.3600778020909312,
+      "grad_norm": 16.125,
+      "learning_rate": 5.931928728212126e-07,
+      "loss": 0.2877,
+      "step": 5594
+    },
+    {
+      "epoch": 1.3603209336250912,
+      "grad_norm": 14.3125,
+      "learning_rate": 5.927817490954807e-07,
+      "loss": 0.2737,
+      "step": 5595
+    },
+    {
+      "epoch": 1.3605640651592512,
+      "grad_norm": 20.875,
+      "learning_rate": 5.923707236023663e-07,
+      "loss": 0.7164,
+      "step": 5596
+    },
+    {
+      "epoch": 1.3608071966934112,
+      "grad_norm": 17.75,
+      "learning_rate": 5.919597964033051e-07,
+      "loss": 0.8874,
+      "step": 5597
+    },
+    {
+      "epoch": 1.361050328227571,
+      "grad_norm": 21.75,
+      "learning_rate": 5.915489675597176e-07,
+      "loss": 0.5238,
+      "step": 5598
+    },
+    {
+      "epoch": 1.3612934597617312,
+      "grad_norm": 20.125,
+      "learning_rate": 5.911382371330087e-07,
+      "loss": 0.6454,
+      "step": 5599
+    },
+    {
+      "epoch": 1.361536591295891,
+      "grad_norm": 20.5,
+      "learning_rate": 5.907276051845686e-07,
+      "loss": 0.8754,
+      "step": 5600
+    },
+    {
+      "epoch": 1.361779722830051,
+      "grad_norm": 16.375,
+      "learning_rate": 5.903170717757745e-07,
+      "loss": 0.5905,
+      "step": 5601
+    },
+    {
+      "epoch": 1.362022854364211,
+      "grad_norm": 16.5,
+      "learning_rate": 5.89906636967987e-07,
+      "loss": 0.6097,
+      "step": 5602
+    },
+    {
+      "epoch": 1.362265985898371,
+      "grad_norm": 18.375,
+      "learning_rate": 5.894963008225525e-07,
+      "loss": 0.8333,
+      "step": 5603
+    },
+    {
+      "epoch": 1.362509117432531,
+      "grad_norm": 25.75,
+      "learning_rate": 5.89086063400803e-07,
+      "loss": 0.7765,
+      "step": 5604
+    },
+    {
+      "epoch": 1.362752248966691,
+      "grad_norm": 32.25,
+      "learning_rate": 5.886759247640564e-07,
+      "loss": 0.7475,
+      "step": 5605
+    },
+    {
+      "epoch": 1.362995380500851,
+      "grad_norm": 18.75,
+      "learning_rate": 5.88265884973614e-07,
+      "loss": 0.4395,
+      "step": 5606
+    },
+    {
+      "epoch": 1.363238512035011,
+      "grad_norm": 18.0,
+      "learning_rate": 5.878559440907635e-07,
+      "loss": 0.4385,
+      "step": 5607
+    },
+    {
+      "epoch": 1.363481643569171,
+      "grad_norm": 20.75,
+      "learning_rate": 5.87446102176778e-07,
+      "loss": 0.7954,
+      "step": 5608
+    },
+    {
+      "epoch": 1.3637247751033308,
+      "grad_norm": 19.875,
+      "learning_rate": 5.870363592929154e-07,
+      "loss": 0.4758,
+      "step": 5609
+    },
+    {
+      "epoch": 1.363967906637491,
+      "grad_norm": 22.375,
+      "learning_rate": 5.866267155004181e-07,
+      "loss": 0.9392,
+      "step": 5610
+    },
+    {
+      "epoch": 1.3642110381716508,
+      "grad_norm": 26.625,
+      "learning_rate": 5.862171708605155e-07,
+      "loss": 0.9368,
+      "step": 5611
+    },
+    {
+      "epoch": 1.3644541697058108,
+      "grad_norm": 29.375,
+      "learning_rate": 5.858077254344203e-07,
+      "loss": 0.625,
+      "step": 5612
+    },
+    {
+      "epoch": 1.3646973012399708,
+      "grad_norm": 22.25,
+      "learning_rate": 5.853983792833318e-07,
+      "loss": 0.6548,
+      "step": 5613
+    },
+    {
+      "epoch": 1.3649404327741308,
+      "grad_norm": 20.0,
+      "learning_rate": 5.849891324684331e-07,
+      "loss": 0.6221,
+      "step": 5614
+    },
+    {
+      "epoch": 1.3651835643082908,
+      "grad_norm": 23.125,
+      "learning_rate": 5.845799850508943e-07,
+      "loss": 0.9084,
+      "step": 5615
+    },
+    {
+      "epoch": 1.3654266958424508,
+      "grad_norm": 20.0,
+      "learning_rate": 5.84170937091869e-07,
+      "loss": 1.063,
+      "step": 5616
+    },
+    {
+      "epoch": 1.3656698273766108,
+      "grad_norm": 20.75,
+      "learning_rate": 5.837619886524957e-07,
+      "loss": 0.8154,
+      "step": 5617
+    },
+    {
+      "epoch": 1.3659129589107708,
+      "grad_norm": 20.25,
+      "learning_rate": 5.833531397939002e-07,
+      "loss": 0.5817,
+      "step": 5618
+    },
+    {
+      "epoch": 1.3661560904449308,
+      "grad_norm": 26.625,
+      "learning_rate": 5.829443905771913e-07,
+      "loss": 1.072,
+      "step": 5619
+    },
+    {
+      "epoch": 1.3663992219790906,
+      "grad_norm": 16.875,
+      "learning_rate": 5.825357410634628e-07,
+      "loss": 0.4841,
+      "step": 5620
+    },
+    {
+      "epoch": 1.3666423535132508,
+      "grad_norm": 16.625,
+      "learning_rate": 5.82127191313796e-07,
+      "loss": 0.6542,
+      "step": 5621
+    },
+    {
+      "epoch": 1.3668854850474106,
+      "grad_norm": 16.875,
+      "learning_rate": 5.817187413892552e-07,
+      "loss": 0.6381,
+      "step": 5622
+    },
+    {
+      "epoch": 1.3671286165815706,
+      "grad_norm": 22.5,
+      "learning_rate": 5.813103913508902e-07,
+      "loss": 0.8829,
+      "step": 5623
+    },
+    {
+      "epoch": 1.3673717481157306,
+      "grad_norm": 22.125,
+      "learning_rate": 5.809021412597353e-07,
+      "loss": 1.1265,
+      "step": 5624
+    },
+    {
+      "epoch": 1.3676148796498906,
+      "grad_norm": 27.0,
+      "learning_rate": 5.804939911768117e-07,
+      "loss": 0.8627,
+      "step": 5625
+    },
+    {
+      "epoch": 1.3678580111840506,
+      "grad_norm": 19.0,
+      "learning_rate": 5.800859411631236e-07,
+      "loss": 0.618,
+      "step": 5626
+    },
+    {
+      "epoch": 1.3681011427182106,
+      "grad_norm": 19.5,
+      "learning_rate": 5.796779912796608e-07,
+      "loss": 0.5365,
+      "step": 5627
+    },
+    {
+      "epoch": 1.3683442742523706,
+      "grad_norm": 21.5,
+      "learning_rate": 5.792701415873998e-07,
+      "loss": 0.3092,
+      "step": 5628
+    },
+    {
+      "epoch": 1.3685874057865304,
+      "grad_norm": 21.625,
+      "learning_rate": 5.788623921473001e-07,
+      "loss": 0.7268,
+      "step": 5629
+    },
+    {
+      "epoch": 1.3688305373206906,
+      "grad_norm": 19.625,
+      "learning_rate": 5.784547430203068e-07,
+      "loss": 0.6092,
+      "step": 5630
+    },
+    {
+      "epoch": 1.3690736688548504,
+      "grad_norm": 16.875,
+      "learning_rate": 5.780471942673498e-07,
+      "loss": 0.3973,
+      "step": 5631
+    },
+    {
+      "epoch": 1.3693168003890104,
+      "grad_norm": 21.875,
+      "learning_rate": 5.77639745949345e-07,
+      "loss": 0.5814,
+      "step": 5632
+    },
+    {
+      "epoch": 1.3695599319231704,
+      "grad_norm": 14.1875,
+      "learning_rate": 5.772323981271922e-07,
+      "loss": 0.3311,
+      "step": 5633
+    },
+    {
+      "epoch": 1.3698030634573304,
+      "grad_norm": 16.875,
+      "learning_rate": 5.768251508617762e-07,
+      "loss": 0.7044,
+      "step": 5634
+    },
+    {
+      "epoch": 1.3700461949914904,
+      "grad_norm": 16.625,
+      "learning_rate": 5.764180042139675e-07,
+      "loss": 0.5737,
+      "step": 5635
+    },
+    {
+      "epoch": 1.3702893265256504,
+      "grad_norm": 21.375,
+      "learning_rate": 5.760109582446218e-07,
+      "loss": 0.6558,
+      "step": 5636
+    },
+    {
+      "epoch": 1.3705324580598104,
+      "grad_norm": 30.375,
+      "learning_rate": 5.756040130145781e-07,
+      "loss": 0.6485,
+      "step": 5637
+    },
+    {
+      "epoch": 1.3707755895939704,
+      "grad_norm": 26.75,
+      "learning_rate": 5.751971685846622e-07,
+      "loss": 0.8077,
+      "step": 5638
+    },
+    {
+      "epoch": 1.3710187211281304,
+      "grad_norm": 16.875,
+      "learning_rate": 5.747904250156839e-07,
+      "loss": 0.6828,
+      "step": 5639
+    },
+    {
+      "epoch": 1.3712618526622902,
+      "grad_norm": 21.875,
+      "learning_rate": 5.743837823684377e-07,
+      "loss": 0.8346,
+      "step": 5640
+    },
+    {
+      "epoch": 1.3715049841964504,
+      "grad_norm": 15.0625,
+      "learning_rate": 5.739772407037031e-07,
+      "loss": 0.6601,
+      "step": 5641
+    },
+    {
+      "epoch": 1.3717481157306102,
+      "grad_norm": 18.625,
+      "learning_rate": 5.735708000822456e-07,
+      "loss": 0.3835,
+      "step": 5642
+    },
+    {
+      "epoch": 1.3719912472647702,
+      "grad_norm": 16.5,
+      "learning_rate": 5.73164460564814e-07,
+      "loss": 0.6508,
+      "step": 5643
+    },
+    {
+      "epoch": 1.3722343787989302,
+      "grad_norm": 15.25,
+      "learning_rate": 5.72758222212143e-07,
+      "loss": 0.4461,
+      "step": 5644
+    },
+    {
+      "epoch": 1.3724775103330902,
+      "grad_norm": 19.625,
+      "learning_rate": 5.723520850849526e-07,
+      "loss": 0.5593,
+      "step": 5645
+    },
+    {
+      "epoch": 1.3727206418672502,
+      "grad_norm": 21.5,
+      "learning_rate": 5.719460492439465e-07,
+      "loss": 0.7001,
+      "step": 5646
+    },
+    {
+      "epoch": 1.3729637734014102,
+      "grad_norm": 16.5,
+      "learning_rate": 5.715401147498137e-07,
+      "loss": 0.392,
+      "step": 5647
+    },
+    {
+      "epoch": 1.3732069049355702,
+      "grad_norm": 19.375,
+      "learning_rate": 5.711342816632277e-07,
+      "loss": 0.8772,
+      "step": 5648
+    },
+    {
+      "epoch": 1.3734500364697302,
+      "grad_norm": 16.875,
+      "learning_rate": 5.707285500448483e-07,
+      "loss": 0.2809,
+      "step": 5649
+    },
+    {
+      "epoch": 1.3736931680038902,
+      "grad_norm": 22.0,
+      "learning_rate": 5.703229199553178e-07,
+      "loss": 0.8582,
+      "step": 5650
+    },
+    {
+      "epoch": 1.37393629953805,
+      "grad_norm": 19.25,
+      "learning_rate": 5.699173914552655e-07,
+      "loss": 0.7408,
+      "step": 5651
+    },
+    {
+      "epoch": 1.3741794310722102,
+      "grad_norm": 17.5,
+      "learning_rate": 5.69511964605305e-07,
+      "loss": 0.4454,
+      "step": 5652
+    },
+    {
+      "epoch": 1.37442256260637,
+      "grad_norm": 14.9375,
+      "learning_rate": 5.691066394660337e-07,
+      "loss": 0.4825,
+      "step": 5653
+    },
+    {
+      "epoch": 1.37466569414053,
+      "grad_norm": 18.5,
+      "learning_rate": 5.687014160980346e-07,
+      "loss": 0.7598,
+      "step": 5654
+    },
+    {
+      "epoch": 1.37490882567469,
+      "grad_norm": 18.875,
+      "learning_rate": 5.682962945618749e-07,
+      "loss": 0.66,
+      "step": 5655
+    },
+    {
+      "epoch": 1.37515195720885,
+      "grad_norm": 28.25,
+      "learning_rate": 5.678912749181079e-07,
+      "loss": 1.0435,
+      "step": 5656
+    },
+    {
+      "epoch": 1.37539508874301,
+      "grad_norm": 18.25,
+      "learning_rate": 5.674863572272701e-07,
+      "loss": 0.938,
+      "step": 5657
+    },
+    {
+      "epoch": 1.37563822027717,
+      "grad_norm": 22.5,
+      "learning_rate": 5.670815415498829e-07,
+      "loss": 0.6949,
+      "step": 5658
+    },
+    {
+      "epoch": 1.37588135181133,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.666768279464546e-07,
+      "loss": 0.6275,
+      "step": 5659
+    },
+    {
+      "epoch": 1.37612448334549,
+      "grad_norm": 20.5,
+      "learning_rate": 5.662722164774757e-07,
+      "loss": 0.6347,
+      "step": 5660
+    },
+    {
+      "epoch": 1.37636761487965,
+      "grad_norm": 17.375,
+      "learning_rate": 5.65867707203422e-07,
+      "loss": 0.4161,
+      "step": 5661
+    },
+    {
+      "epoch": 1.3766107464138098,
+      "grad_norm": 16.75,
+      "learning_rate": 5.654633001847553e-07,
+      "loss": 0.643,
+      "step": 5662
+    },
+    {
+      "epoch": 1.37685387794797,
+      "grad_norm": 20.375,
+      "learning_rate": 5.650589954819208e-07,
+      "loss": 1.1652,
+      "step": 5663
+    },
+    {
+      "epoch": 1.3770970094821298,
+      "grad_norm": 17.875,
+      "learning_rate": 5.646547931553487e-07,
+      "loss": 0.3996,
+      "step": 5664
+    },
+    {
+      "epoch": 1.3773401410162898,
+      "grad_norm": 25.0,
+      "learning_rate": 5.642506932654536e-07,
+      "loss": 0.6739,
+      "step": 5665
+    },
+    {
+      "epoch": 1.3775832725504498,
+      "grad_norm": 21.375,
+      "learning_rate": 5.638466958726358e-07,
+      "loss": 0.8701,
+      "step": 5666
+    },
+    {
+      "epoch": 1.3778264040846098,
+      "grad_norm": 31.125,
+      "learning_rate": 5.6344280103728e-07,
+      "loss": 0.8251,
+      "step": 5667
+    },
+    {
+      "epoch": 1.3780695356187698,
+      "grad_norm": 17.125,
+      "learning_rate": 5.630390088197543e-07,
+      "loss": 0.6207,
+      "step": 5668
+    },
+    {
+      "epoch": 1.3783126671529298,
+      "grad_norm": 20.375,
+      "learning_rate": 5.626353192804135e-07,
+      "loss": 0.7564,
+      "step": 5669
+    },
+    {
+      "epoch": 1.3785557986870898,
+      "grad_norm": 28.25,
+      "learning_rate": 5.622317324795954e-07,
+      "loss": 0.7164,
+      "step": 5670
+    },
+    {
+      "epoch": 1.3787989302212498,
+      "grad_norm": 67.5,
+      "learning_rate": 5.61828248477623e-07,
+      "loss": 1.9273,
+      "step": 5671
+    },
+    {
+      "epoch": 1.3790420617554098,
+      "grad_norm": 17.875,
+      "learning_rate": 5.614248673348036e-07,
+      "loss": 0.5701,
+      "step": 5672
+    },
+    {
+      "epoch": 1.3792851932895696,
+      "grad_norm": 27.125,
+      "learning_rate": 5.610215891114304e-07,
+      "loss": 1.1853,
+      "step": 5673
+    },
+    {
+      "epoch": 1.3795283248237296,
+      "grad_norm": 20.0,
+      "learning_rate": 5.606184138677792e-07,
+      "loss": 0.7212,
+      "step": 5674
+    },
+    {
+      "epoch": 1.3797714563578896,
+      "grad_norm": 26.5,
+      "learning_rate": 5.602153416641121e-07,
+      "loss": 0.94,
+      "step": 5675
+    },
+    {
+      "epoch": 1.3800145878920496,
+      "grad_norm": 19.75,
+      "learning_rate": 5.598123725606757e-07,
+      "loss": 0.7837,
+      "step": 5676
+    },
+    {
+      "epoch": 1.3802577194262096,
+      "grad_norm": 19.875,
+      "learning_rate": 5.594095066177001e-07,
+      "loss": 0.6102,
+      "step": 5677
+    },
+    {
+      "epoch": 1.3805008509603696,
+      "grad_norm": 20.25,
+      "learning_rate": 5.590067438954003e-07,
+      "loss": 0.488,
+      "step": 5678
+    },
+    {
+      "epoch": 1.3807439824945296,
+      "grad_norm": 19.75,
+      "learning_rate": 5.586040844539769e-07,
+      "loss": 0.5539,
+      "step": 5679
+    },
+    {
+      "epoch": 1.3809871140286896,
+      "grad_norm": 18.125,
+      "learning_rate": 5.582015283536137e-07,
+      "loss": 0.6042,
+      "step": 5680
+    },
+    {
+      "epoch": 1.3812302455628496,
+      "grad_norm": 18.375,
+      "learning_rate": 5.577990756544797e-07,
+      "loss": 0.7593,
+      "step": 5681
+    },
+    {
+      "epoch": 1.3814733770970093,
+      "grad_norm": 17.5,
+      "learning_rate": 5.573967264167285e-07,
+      "loss": 0.7128,
+      "step": 5682
+    },
+    {
+      "epoch": 1.3817165086311696,
+      "grad_norm": 16.0,
+      "learning_rate": 5.569944807004986e-07,
+      "loss": 0.5725,
+      "step": 5683
+    },
+    {
+      "epoch": 1.3819596401653294,
+      "grad_norm": 16.375,
+      "learning_rate": 5.565923385659122e-07,
+      "loss": 0.3594,
+      "step": 5684
+    },
+    {
+      "epoch": 1.3822027716994894,
+      "grad_norm": 19.75,
+      "learning_rate": 5.561903000730759e-07,
+      "loss": 0.7702,
+      "step": 5685
+    },
+    {
+      "epoch": 1.3824459032336494,
+      "grad_norm": 21.0,
+      "learning_rate": 5.557883652820823e-07,
+      "loss": 0.7749,
+      "step": 5686
+    },
+    {
+      "epoch": 1.3826890347678094,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.553865342530068e-07,
+      "loss": 0.666,
+      "step": 5687
+    },
+    {
+      "epoch": 1.3829321663019694,
+      "grad_norm": 21.625,
+      "learning_rate": 5.549848070459102e-07,
+      "loss": 0.589,
+      "step": 5688
+    },
+    {
+      "epoch": 1.3831752978361294,
+      "grad_norm": 17.0,
+      "learning_rate": 5.545831837208366e-07,
+      "loss": 0.6232,
+      "step": 5689
+    },
+    {
+      "epoch": 1.3834184293702894,
+      "grad_norm": 19.25,
+      "learning_rate": 5.541816643378174e-07,
+      "loss": 0.5763,
+      "step": 5690
+    },
+    {
+      "epoch": 1.3836615609044494,
+      "grad_norm": 18.125,
+      "learning_rate": 5.537802489568657e-07,
+      "loss": 0.56,
+      "step": 5691
+    },
+    {
+      "epoch": 1.3839046924386094,
+      "grad_norm": 24.5,
+      "learning_rate": 5.533789376379795e-07,
+      "loss": 0.7102,
+      "step": 5692
+    },
+    {
+      "epoch": 1.3841478239727691,
+      "grad_norm": 27.875,
+      "learning_rate": 5.529777304411424e-07,
+      "loss": 1.0647,
+      "step": 5693
+    },
+    {
+      "epoch": 1.3843909555069294,
+      "grad_norm": 27.625,
+      "learning_rate": 5.525766274263217e-07,
+      "loss": 0.6592,
+      "step": 5694
+    },
+    {
+      "epoch": 1.3846340870410891,
+      "grad_norm": 14.0625,
+      "learning_rate": 5.521756286534682e-07,
+      "loss": 0.3931,
+      "step": 5695
+    },
+    {
+      "epoch": 1.3848772185752491,
+      "grad_norm": 18.5,
+      "learning_rate": 5.517747341825196e-07,
+      "loss": 0.6468,
+      "step": 5696
+    },
+    {
+      "epoch": 1.3851203501094091,
+      "grad_norm": 20.5,
+      "learning_rate": 5.513739440733953e-07,
+      "loss": 0.492,
+      "step": 5697
+    },
+    {
+      "epoch": 1.3853634816435692,
+      "grad_norm": 23.625,
+      "learning_rate": 5.50973258386001e-07,
+      "loss": 0.6066,
+      "step": 5698
+    },
+    {
+      "epoch": 1.3856066131777292,
+      "grad_norm": 22.5,
+      "learning_rate": 5.505726771802254e-07,
+      "loss": 0.4876,
+      "step": 5699
+    },
+    {
+      "epoch": 1.3858497447118892,
+      "grad_norm": 23.625,
+      "learning_rate": 5.501722005159432e-07,
+      "loss": 0.6459,
+      "step": 5700
+    },
+    {
+      "epoch": 1.3860928762460492,
+      "grad_norm": 21.5,
+      "learning_rate": 5.497718284530123e-07,
+      "loss": 0.8453,
+      "step": 5701
+    },
+    {
+      "epoch": 1.3863360077802092,
+      "grad_norm": 24.75,
+      "learning_rate": 5.493715610512742e-07,
+      "loss": 0.812,
+      "step": 5702
+    },
+    {
+      "epoch": 1.3865791393143692,
+      "grad_norm": 23.625,
+      "learning_rate": 5.489713983705572e-07,
+      "loss": 0.6894,
+      "step": 5703
+    },
+    {
+      "epoch": 1.386822270848529,
+      "grad_norm": 24.5,
+      "learning_rate": 5.485713404706719e-07,
+      "loss": 0.7908,
+      "step": 5704
+    },
+    {
+      "epoch": 1.3870654023826892,
+      "grad_norm": 18.5,
+      "learning_rate": 5.481713874114134e-07,
+      "loss": 0.4933,
+      "step": 5705
+    },
+    {
+      "epoch": 1.387308533916849,
+      "grad_norm": 26.375,
+      "learning_rate": 5.47771539252562e-07,
+      "loss": 0.6707,
+      "step": 5706
+    },
+    {
+      "epoch": 1.387551665451009,
+      "grad_norm": 25.875,
+      "learning_rate": 5.473717960538825e-07,
+      "loss": 0.8116,
+      "step": 5707
+    },
+    {
+      "epoch": 1.387794796985169,
+      "grad_norm": 16.625,
+      "learning_rate": 5.469721578751229e-07,
+      "loss": 0.4498,
+      "step": 5708
+    },
+    {
+      "epoch": 1.388037928519329,
+      "grad_norm": 16.125,
+      "learning_rate": 5.465726247760155e-07,
+      "loss": 0.7928,
+      "step": 5709
+    },
+    {
+      "epoch": 1.388281060053489,
+      "grad_norm": 23.25,
+      "learning_rate": 5.461731968162784e-07,
+      "loss": 0.736,
+      "step": 5710
+    },
+    {
+      "epoch": 1.388524191587649,
+      "grad_norm": 19.75,
+      "learning_rate": 5.457738740556125e-07,
+      "loss": 0.8674,
+      "step": 5711
+    },
+    {
+      "epoch": 1.388767323121809,
+      "grad_norm": 20.0,
+      "learning_rate": 5.453746565537031e-07,
+      "loss": 0.905,
+      "step": 5712
+    },
+    {
+      "epoch": 1.389010454655969,
+      "grad_norm": 24.5,
+      "learning_rate": 5.449755443702206e-07,
+      "loss": 0.7641,
+      "step": 5713
+    },
+    {
+      "epoch": 1.389253586190129,
+      "grad_norm": 22.5,
+      "learning_rate": 5.445765375648197e-07,
+      "loss": 0.8638,
+      "step": 5714
+    },
+    {
+      "epoch": 1.3894967177242887,
+      "grad_norm": 14.4375,
+      "learning_rate": 5.441776361971382e-07,
+      "loss": 0.2904,
+      "step": 5715
+    },
+    {
+      "epoch": 1.389739849258449,
+      "grad_norm": 24.75,
+      "learning_rate": 5.437788403267986e-07,
+      "loss": 0.8309,
+      "step": 5716
+    },
+    {
+      "epoch": 1.3899829807926087,
+      "grad_norm": 20.25,
+      "learning_rate": 5.433801500134087e-07,
+      "loss": 0.7246,
+      "step": 5717
+    },
+    {
+      "epoch": 1.3902261123267687,
+      "grad_norm": 20.375,
+      "learning_rate": 5.42981565316559e-07,
+      "loss": 0.8662,
+      "step": 5718
+    },
+    {
+      "epoch": 1.3904692438609287,
+      "grad_norm": 17.5,
+      "learning_rate": 5.425830862958245e-07,
+      "loss": 0.4607,
+      "step": 5719
+    },
+    {
+      "epoch": 1.3907123753950887,
+      "grad_norm": 14.375,
+      "learning_rate": 5.421847130107655e-07,
+      "loss": 0.2762,
+      "step": 5720
+    },
+    {
+      "epoch": 1.3909555069292487,
+      "grad_norm": 36.25,
+      "learning_rate": 5.417864455209257e-07,
+      "loss": 0.9315,
+      "step": 5721
+    },
+    {
+      "epoch": 1.3911986384634087,
+      "grad_norm": 22.125,
+      "learning_rate": 5.413882838858331e-07,
+      "loss": 0.8537,
+      "step": 5722
+    },
+    {
+      "epoch": 1.3914417699975687,
+      "grad_norm": 22.0,
+      "learning_rate": 5.409902281649991e-07,
+      "loss": 0.7202,
+      "step": 5723
+    },
+    {
+      "epoch": 1.3916849015317287,
+      "grad_norm": 21.0,
+      "learning_rate": 5.405922784179209e-07,
+      "loss": 0.603,
+      "step": 5724
+    },
+    {
+      "epoch": 1.3919280330658887,
+      "grad_norm": 19.625,
+      "learning_rate": 5.401944347040787e-07,
+      "loss": 0.5254,
+      "step": 5725
+    },
+    {
+      "epoch": 1.3921711646000485,
+      "grad_norm": 28.375,
+      "learning_rate": 5.397966970829365e-07,
+      "loss": 1.0057,
+      "step": 5726
+    },
+    {
+      "epoch": 1.3924142961342085,
+      "grad_norm": 22.0,
+      "learning_rate": 5.393990656139438e-07,
+      "loss": 0.6172,
+      "step": 5727
+    },
+    {
+      "epoch": 1.3926574276683685,
+      "grad_norm": 17.75,
+      "learning_rate": 5.390015403565331e-07,
+      "loss": 0.5187,
+      "step": 5728
+    },
+    {
+      "epoch": 1.3929005592025285,
+      "grad_norm": 19.75,
+      "learning_rate": 5.38604121370122e-07,
+      "loss": 0.5958,
+      "step": 5729
+    },
+    {
+      "epoch": 1.3931436907366885,
+      "grad_norm": 22.25,
+      "learning_rate": 5.382068087141105e-07,
+      "loss": 0.7699,
+      "step": 5730
+    },
+    {
+      "epoch": 1.3933868222708485,
+      "grad_norm": 20.875,
+      "learning_rate": 5.378096024478852e-07,
+      "loss": 0.8332,
+      "step": 5731
+    },
+    {
+      "epoch": 1.3936299538050085,
+      "grad_norm": 18.5,
+      "learning_rate": 5.374125026308148e-07,
+      "loss": 0.854,
+      "step": 5732
+    },
+    {
+      "epoch": 1.3938730853391685,
+      "grad_norm": 23.0,
+      "learning_rate": 5.37015509322252e-07,
+      "loss": 0.8707,
+      "step": 5733
+    },
+    {
+      "epoch": 1.3941162168733285,
+      "grad_norm": 23.5,
+      "learning_rate": 5.366186225815356e-07,
+      "loss": 0.7658,
+      "step": 5734
+    },
+    {
+      "epoch": 1.3943593484074883,
+      "grad_norm": 18.25,
+      "learning_rate": 5.362218424679862e-07,
+      "loss": 0.561,
+      "step": 5735
+    },
+    {
+      "epoch": 1.3946024799416485,
+      "grad_norm": 16.625,
+      "learning_rate": 5.358251690409103e-07,
+      "loss": 0.7389,
+      "step": 5736
+    },
+    {
+      "epoch": 1.3948456114758083,
+      "grad_norm": 17.875,
+      "learning_rate": 5.354286023595964e-07,
+      "loss": 0.4713,
+      "step": 5737
+    },
+    {
+      "epoch": 1.3950887430099683,
+      "grad_norm": 23.0,
+      "learning_rate": 5.350321424833196e-07,
+      "loss": 0.7443,
+      "step": 5738
+    },
+    {
+      "epoch": 1.3953318745441283,
+      "grad_norm": 17.0,
+      "learning_rate": 5.34635789471337e-07,
+      "loss": 0.5654,
+      "step": 5739
+    },
+    {
+      "epoch": 1.3955750060782883,
+      "grad_norm": 27.375,
+      "learning_rate": 5.342395433828899e-07,
+      "loss": 1.1186,
+      "step": 5740
+    },
+    {
+      "epoch": 1.3958181376124483,
+      "grad_norm": 29.875,
+      "learning_rate": 5.338434042772052e-07,
+      "loss": 1.1762,
+      "step": 5741
+    },
+    {
+      "epoch": 1.3960612691466083,
+      "grad_norm": 32.75,
+      "learning_rate": 5.334473722134923e-07,
+      "loss": 0.5788,
+      "step": 5742
+    },
+    {
+      "epoch": 1.3963044006807683,
+      "grad_norm": 22.625,
+      "learning_rate": 5.330514472509442e-07,
+      "loss": 0.9732,
+      "step": 5743
+    },
+    {
+      "epoch": 1.3965475322149283,
+      "grad_norm": 17.375,
+      "learning_rate": 5.326556294487396e-07,
+      "loss": 0.4992,
+      "step": 5744
+    },
+    {
+      "epoch": 1.3967906637490883,
+      "grad_norm": 18.125,
+      "learning_rate": 5.322599188660406e-07,
+      "loss": 0.3687,
+      "step": 5745
+    },
+    {
+      "epoch": 1.3970337952832481,
+      "grad_norm": 17.5,
+      "learning_rate": 5.318643155619924e-07,
+      "loss": 0.5838,
+      "step": 5746
+    },
+    {
+      "epoch": 1.3972769268174083,
+      "grad_norm": 17.875,
+      "learning_rate": 5.314688195957245e-07,
+      "loss": 0.6116,
+      "step": 5747
+    },
+    {
+      "epoch": 1.3975200583515681,
+      "grad_norm": 17.0,
+      "learning_rate": 5.310734310263515e-07,
+      "loss": 0.4648,
+      "step": 5748
+    },
+    {
+      "epoch": 1.3977631898857281,
+      "grad_norm": 23.375,
+      "learning_rate": 5.306781499129704e-07,
+      "loss": 0.5402,
+      "step": 5749
+    },
+    {
+      "epoch": 1.3980063214198881,
+      "grad_norm": 20.375,
+      "learning_rate": 5.302829763146625e-07,
+      "loss": 0.4471,
+      "step": 5750
+    },
+    {
+      "epoch": 1.3982494529540481,
+      "grad_norm": 19.375,
+      "learning_rate": 5.298879102904937e-07,
+      "loss": 0.6008,
+      "step": 5751
+    },
+    {
+      "epoch": 1.3984925844882081,
+      "grad_norm": 23.25,
+      "learning_rate": 5.29492951899514e-07,
+      "loss": 0.8676,
+      "step": 5752
+    },
+    {
+      "epoch": 1.3987357160223681,
+      "grad_norm": 22.5,
+      "learning_rate": 5.290981012007563e-07,
+      "loss": 0.5519,
+      "step": 5753
+    },
+    {
+      "epoch": 1.3989788475565281,
+      "grad_norm": 23.75,
+      "learning_rate": 5.28703358253237e-07,
+      "loss": 0.5874,
+      "step": 5754
+    },
+    {
+      "epoch": 1.3992219790906881,
+      "grad_norm": 21.125,
+      "learning_rate": 5.283087231159588e-07,
+      "loss": 0.533,
+      "step": 5755
+    },
+    {
+      "epoch": 1.3994651106248481,
+      "grad_norm": 24.625,
+      "learning_rate": 5.279141958479059e-07,
+      "loss": 1.2137,
+      "step": 5756
+    },
+    {
+      "epoch": 1.399708242159008,
+      "grad_norm": 12.9375,
+      "learning_rate": 5.275197765080469e-07,
+      "loss": 0.2967,
+      "step": 5757
+    },
+    {
+      "epoch": 1.3999513736931681,
+      "grad_norm": 17.625,
+      "learning_rate": 5.271254651553352e-07,
+      "loss": 0.4688,
+      "step": 5758
+    },
+    {
+      "epoch": 1.400194505227328,
+      "grad_norm": 19.375,
+      "learning_rate": 5.26731261848707e-07,
+      "loss": 0.7406,
+      "step": 5759
+    },
+    {
+      "epoch": 1.400437636761488,
+      "grad_norm": 20.875,
+      "learning_rate": 5.263371666470833e-07,
+      "loss": 0.5208,
+      "step": 5760
+    },
+    {
+      "epoch": 1.400680768295648,
+      "grad_norm": 19.375,
+      "learning_rate": 5.259431796093676e-07,
+      "loss": 0.594,
+      "step": 5761
+    },
+    {
+      "epoch": 1.400923899829808,
+      "grad_norm": 21.5,
+      "learning_rate": 5.255493007944492e-07,
+      "loss": 0.5439,
+      "step": 5762
+    },
+    {
+      "epoch": 1.401167031363968,
+      "grad_norm": 22.0,
+      "learning_rate": 5.251555302611995e-07,
+      "loss": 0.6506,
+      "step": 5763
+    },
+    {
+      "epoch": 1.401410162898128,
+      "grad_norm": 27.75,
+      "learning_rate": 5.247618680684738e-07,
+      "loss": 0.8935,
+      "step": 5764
+    },
+    {
+      "epoch": 1.401653294432288,
+      "grad_norm": 22.0,
+      "learning_rate": 5.243683142751127e-07,
+      "loss": 0.5009,
+      "step": 5765
+    },
+    {
+      "epoch": 1.401896425966448,
+      "grad_norm": 20.25,
+      "learning_rate": 5.239748689399387e-07,
+      "loss": 0.7143,
+      "step": 5766
+    },
+    {
+      "epoch": 1.402139557500608,
+      "grad_norm": 39.0,
+      "learning_rate": 5.235815321217598e-07,
+      "loss": 1.0894,
+      "step": 5767
+    },
+    {
+      "epoch": 1.4023826890347677,
+      "grad_norm": 25.0,
+      "learning_rate": 5.231883038793662e-07,
+      "loss": 0.9148,
+      "step": 5768
+    },
+    {
+      "epoch": 1.402625820568928,
+      "grad_norm": 21.625,
+      "learning_rate": 5.227951842715335e-07,
+      "loss": 0.6216,
+      "step": 5769
+    },
+    {
+      "epoch": 1.4028689521030877,
+      "grad_norm": 15.625,
+      "learning_rate": 5.224021733570197e-07,
+      "loss": 0.4572,
+      "step": 5770
+    },
+    {
+      "epoch": 1.4031120836372477,
+      "grad_norm": 21.5,
+      "learning_rate": 5.220092711945668e-07,
+      "loss": 0.9574,
+      "step": 5771
+    },
+    {
+      "epoch": 1.4033552151714077,
+      "grad_norm": 23.0,
+      "learning_rate": 5.216164778429016e-07,
+      "loss": 0.6536,
+      "step": 5772
+    },
+    {
+      "epoch": 1.4035983467055677,
+      "grad_norm": 15.5,
+      "learning_rate": 5.212237933607332e-07,
+      "loss": 0.5542,
+      "step": 5773
+    },
+    {
+      "epoch": 1.4038414782397277,
+      "grad_norm": 24.5,
+      "learning_rate": 5.208312178067551e-07,
+      "loss": 1.0369,
+      "step": 5774
+    },
+    {
+      "epoch": 1.4040846097738877,
+      "grad_norm": 26.5,
+      "learning_rate": 5.204387512396446e-07,
+      "loss": 0.9562,
+      "step": 5775
+    },
+    {
+      "epoch": 1.4043277413080477,
+      "grad_norm": 21.0,
+      "learning_rate": 5.20046393718063e-07,
+      "loss": 0.9378,
+      "step": 5776
+    },
+    {
+      "epoch": 1.4045708728422075,
+      "grad_norm": 17.0,
+      "learning_rate": 5.196541453006547e-07,
+      "loss": 0.6464,
+      "step": 5777
+    },
+    {
+      "epoch": 1.4048140043763677,
+      "grad_norm": 18.0,
+      "learning_rate": 5.192620060460475e-07,
+      "loss": 0.556,
+      "step": 5778
+    },
+    {
+      "epoch": 1.4050571359105275,
+      "grad_norm": 21.125,
+      "learning_rate": 5.188699760128542e-07,
+      "loss": 0.7421,
+      "step": 5779
+    },
+    {
+      "epoch": 1.4053002674446875,
+      "grad_norm": 19.375,
+      "learning_rate": 5.184780552596699e-07,
+      "loss": 0.7795,
+      "step": 5780
+    },
+    {
+      "epoch": 1.4055433989788475,
+      "grad_norm": 16.0,
+      "learning_rate": 5.180862438450739e-07,
+      "loss": 0.3672,
+      "step": 5781
+    },
+    {
+      "epoch": 1.4057865305130075,
+      "grad_norm": 15.3125,
+      "learning_rate": 5.176945418276292e-07,
+      "loss": 0.6121,
+      "step": 5782
+    },
+    {
+      "epoch": 1.4060296620471675,
+      "grad_norm": 17.25,
+      "learning_rate": 5.173029492658829e-07,
+      "loss": 0.5308,
+      "step": 5783
+    },
+    {
+      "epoch": 1.4062727935813275,
+      "grad_norm": 13.75,
+      "learning_rate": 5.16911466218365e-07,
+      "loss": 0.3052,
+      "step": 5784
+    },
+    {
+      "epoch": 1.4065159251154875,
+      "grad_norm": 25.0,
+      "learning_rate": 5.16520092743589e-07,
+      "loss": 0.9342,
+      "step": 5785
+    },
+    {
+      "epoch": 1.4067590566496475,
+      "grad_norm": 22.0,
+      "learning_rate": 5.161288289000532e-07,
+      "loss": 1.0008,
+      "step": 5786
+    },
+    {
+      "epoch": 1.4070021881838075,
+      "grad_norm": 23.875,
+      "learning_rate": 5.157376747462382e-07,
+      "loss": 0.8272,
+      "step": 5787
+    },
+    {
+      "epoch": 1.4072453197179673,
+      "grad_norm": 19.25,
+      "learning_rate": 5.153466303406085e-07,
+      "loss": 0.9004,
+      "step": 5788
+    },
+    {
+      "epoch": 1.4074884512521275,
+      "grad_norm": 23.125,
+      "learning_rate": 5.149556957416129e-07,
+      "loss": 0.9683,
+      "step": 5789
+    },
+    {
+      "epoch": 1.4077315827862873,
+      "grad_norm": 16.0,
+      "learning_rate": 5.145648710076835e-07,
+      "loss": 0.4709,
+      "step": 5790
+    },
+    {
+      "epoch": 1.4079747143204473,
+      "grad_norm": 18.625,
+      "learning_rate": 5.141741561972356e-07,
+      "loss": 0.7457,
+      "step": 5791
+    },
+    {
+      "epoch": 1.4082178458546073,
+      "grad_norm": 18.375,
+      "learning_rate": 5.137835513686677e-07,
+      "loss": 0.4687,
+      "step": 5792
+    },
+    {
+      "epoch": 1.4084609773887673,
+      "grad_norm": 29.875,
+      "learning_rate": 5.133930565803635e-07,
+      "loss": 0.7572,
+      "step": 5793
+    },
+    {
+      "epoch": 1.4087041089229273,
+      "grad_norm": 25.125,
+      "learning_rate": 5.130026718906886e-07,
+      "loss": 0.6181,
+      "step": 5794
+    },
+    {
+      "epoch": 1.4089472404570873,
+      "grad_norm": 114.0,
+      "learning_rate": 5.126123973579923e-07,
+      "loss": 0.6233,
+      "step": 5795
+    },
+    {
+      "epoch": 1.4091903719912473,
+      "grad_norm": 23.25,
+      "learning_rate": 5.122222330406089e-07,
+      "loss": 1.0922,
+      "step": 5796
+    },
+    {
+      "epoch": 1.4094335035254073,
+      "grad_norm": 29.125,
+      "learning_rate": 5.118321789968541e-07,
+      "loss": 1.1315,
+      "step": 5797
+    },
+    {
+      "epoch": 1.4096766350595673,
+      "grad_norm": 19.875,
+      "learning_rate": 5.114422352850291e-07,
+      "loss": 0.8951,
+      "step": 5798
+    },
+    {
+      "epoch": 1.409919766593727,
+      "grad_norm": 23.75,
+      "learning_rate": 5.110524019634171e-07,
+      "loss": 0.9874,
+      "step": 5799
+    },
+    {
+      "epoch": 1.4101628981278873,
+      "grad_norm": 22.375,
+      "learning_rate": 5.10662679090286e-07,
+      "loss": 0.515,
+      "step": 5800
+    },
+    {
+      "epoch": 1.410406029662047,
+      "grad_norm": 25.75,
+      "learning_rate": 5.102730667238862e-07,
+      "loss": 1.2052,
+      "step": 5801
+    },
+    {
+      "epoch": 1.410649161196207,
+      "grad_norm": 14.125,
+      "learning_rate": 5.098835649224519e-07,
+      "loss": 0.2532,
+      "step": 5802
+    },
+    {
+      "epoch": 1.410892292730367,
+      "grad_norm": 26.875,
+      "learning_rate": 5.094941737442014e-07,
+      "loss": 0.8128,
+      "step": 5803
+    },
+    {
+      "epoch": 1.411135424264527,
+      "grad_norm": 22.5,
+      "learning_rate": 5.091048932473356e-07,
+      "loss": 1.0882,
+      "step": 5804
+    },
+    {
+      "epoch": 1.411378555798687,
+      "grad_norm": 27.5,
+      "learning_rate": 5.087157234900389e-07,
+      "loss": 0.8706,
+      "step": 5805
+    },
+    {
+      "epoch": 1.411621687332847,
+      "grad_norm": 22.125,
+      "learning_rate": 5.083266645304796e-07,
+      "loss": 0.9997,
+      "step": 5806
+    },
+    {
+      "epoch": 1.411864818867007,
+      "grad_norm": 22.25,
+      "learning_rate": 5.0793771642681e-07,
+      "loss": 0.4459,
+      "step": 5807
+    },
+    {
+      "epoch": 1.412107950401167,
+      "grad_norm": 25.25,
+      "learning_rate": 5.075488792371644e-07,
+      "loss": 0.7581,
+      "step": 5808
+    },
+    {
+      "epoch": 1.412351081935327,
+      "grad_norm": 19.0,
+      "learning_rate": 5.07160153019661e-07,
+      "loss": 0.8137,
+      "step": 5809
+    },
+    {
+      "epoch": 1.4125942134694869,
+      "grad_norm": 17.25,
+      "learning_rate": 5.067715378324026e-07,
+      "loss": 0.6121,
+      "step": 5810
+    },
+    {
+      "epoch": 1.412837345003647,
+      "grad_norm": 19.5,
+      "learning_rate": 5.063830337334737e-07,
+      "loss": 0.8765,
+      "step": 5811
+    },
+    {
+      "epoch": 1.4130804765378069,
+      "grad_norm": 37.25,
+      "learning_rate": 5.059946407809427e-07,
+      "loss": 1.1682,
+      "step": 5812
+    },
+    {
+      "epoch": 1.4133236080719669,
+      "grad_norm": 26.5,
+      "learning_rate": 5.056063590328619e-07,
+      "loss": 0.6475,
+      "step": 5813
+    },
+    {
+      "epoch": 1.4135667396061269,
+      "grad_norm": 19.25,
+      "learning_rate": 5.052181885472676e-07,
+      "loss": 0.8764,
+      "step": 5814
+    },
+    {
+      "epoch": 1.413809871140287,
+      "grad_norm": 14.6875,
+      "learning_rate": 5.048301293821776e-07,
+      "loss": 0.4941,
+      "step": 5815
+    },
+    {
+      "epoch": 1.414053002674447,
+      "grad_norm": 21.125,
+      "learning_rate": 5.04442181595594e-07,
+      "loss": 0.9546,
+      "step": 5816
+    },
+    {
+      "epoch": 1.414296134208607,
+      "grad_norm": 22.5,
+      "learning_rate": 5.040543452455029e-07,
+      "loss": 0.7142,
+      "step": 5817
+    },
+    {
+      "epoch": 1.414539265742767,
+      "grad_norm": 18.25,
+      "learning_rate": 5.036666203898731e-07,
+      "loss": 0.5793,
+      "step": 5818
+    },
+    {
+      "epoch": 1.414782397276927,
+      "grad_norm": 24.875,
+      "learning_rate": 5.032790070866558e-07,
+      "loss": 0.6766,
+      "step": 5819
+    },
+    {
+      "epoch": 1.415025528811087,
+      "grad_norm": 24.0,
+      "learning_rate": 5.028915053937873e-07,
+      "loss": 0.8037,
+      "step": 5820
+    },
+    {
+      "epoch": 1.4152686603452467,
+      "grad_norm": 20.875,
+      "learning_rate": 5.025041153691868e-07,
+      "loss": 0.8649,
+      "step": 5821
+    },
+    {
+      "epoch": 1.4155117918794067,
+      "grad_norm": 16.375,
+      "learning_rate": 5.021168370707559e-07,
+      "loss": 0.3932,
+      "step": 5822
+    },
+    {
+      "epoch": 1.4157549234135667,
+      "grad_norm": 18.25,
+      "learning_rate": 5.017296705563797e-07,
+      "loss": 0.5906,
+      "step": 5823
+    },
+    {
+      "epoch": 1.4159980549477267,
+      "grad_norm": 24.5,
+      "learning_rate": 5.013426158839277e-07,
+      "loss": 0.7213,
+      "step": 5824
+    },
+    {
+      "epoch": 1.4162411864818867,
+      "grad_norm": 28.625,
+      "learning_rate": 5.009556731112515e-07,
+      "loss": 0.6516,
+      "step": 5825
+    },
+    {
+      "epoch": 1.4164843180160467,
+      "grad_norm": 21.125,
+      "learning_rate": 5.00568842296186e-07,
+      "loss": 0.7865,
+      "step": 5826
+    },
+    {
+      "epoch": 1.4167274495502067,
+      "grad_norm": 27.5,
+      "learning_rate": 5.001821234965507e-07,
+      "loss": 0.832,
+      "step": 5827
+    },
+    {
+      "epoch": 1.4169705810843667,
+      "grad_norm": 18.375,
+      "learning_rate": 4.997955167701463e-07,
+      "loss": 0.6087,
+      "step": 5828
+    },
+    {
+      "epoch": 1.4172137126185267,
+      "grad_norm": 19.75,
+      "learning_rate": 4.994090221747587e-07,
+      "loss": 0.6195,
+      "step": 5829
+    },
+    {
+      "epoch": 1.4174568441526865,
+      "grad_norm": 24.125,
+      "learning_rate": 4.990226397681556e-07,
+      "loss": 0.5386,
+      "step": 5830
+    },
+    {
+      "epoch": 1.4176999756868467,
+      "grad_norm": 19.875,
+      "learning_rate": 4.986363696080893e-07,
+      "loss": 0.8244,
+      "step": 5831
+    },
+    {
+      "epoch": 1.4179431072210065,
+      "grad_norm": 20.0,
+      "learning_rate": 4.98250211752294e-07,
+      "loss": 0.7723,
+      "step": 5832
+    },
+    {
+      "epoch": 1.4181862387551665,
+      "grad_norm": 23.75,
+      "learning_rate": 4.978641662584872e-07,
+      "loss": 0.6594,
+      "step": 5833
+    },
+    {
+      "epoch": 1.4184293702893265,
+      "grad_norm": 19.875,
+      "learning_rate": 4.974782331843711e-07,
+      "loss": 0.8725,
+      "step": 5834
+    },
+    {
+      "epoch": 1.4186725018234865,
+      "grad_norm": 18.75,
+      "learning_rate": 4.970924125876296e-07,
+      "loss": 1.0599,
+      "step": 5835
+    },
+    {
+      "epoch": 1.4189156333576465,
+      "grad_norm": 25.125,
+      "learning_rate": 4.967067045259297e-07,
+      "loss": 0.8882,
+      "step": 5836
+    },
+    {
+      "epoch": 1.4191587648918065,
+      "grad_norm": 23.625,
+      "learning_rate": 4.963211090569227e-07,
+      "loss": 0.6374,
+      "step": 5837
+    },
+    {
+      "epoch": 1.4194018964259665,
+      "grad_norm": 23.25,
+      "learning_rate": 4.959356262382428e-07,
+      "loss": 1.1284,
+      "step": 5838
+    },
+    {
+      "epoch": 1.4196450279601265,
+      "grad_norm": 16.375,
+      "learning_rate": 4.955502561275068e-07,
+      "loss": 0.6845,
+      "step": 5839
+    },
+    {
+      "epoch": 1.4198881594942865,
+      "grad_norm": 16.875,
+      "learning_rate": 4.951649987823141e-07,
+      "loss": 0.6355,
+      "step": 5840
+    },
+    {
+      "epoch": 1.4201312910284463,
+      "grad_norm": 21.25,
+      "learning_rate": 4.947798542602496e-07,
+      "loss": 0.389,
+      "step": 5841
+    },
+    {
+      "epoch": 1.4203744225626065,
+      "grad_norm": 81.5,
+      "learning_rate": 4.943948226188787e-07,
+      "loss": 0.7621,
+      "step": 5842
+    },
+    {
+      "epoch": 1.4206175540967663,
+      "grad_norm": 22.25,
+      "learning_rate": 4.94009903915751e-07,
+      "loss": 0.7873,
+      "step": 5843
+    },
+    {
+      "epoch": 1.4208606856309263,
+      "grad_norm": 20.125,
+      "learning_rate": 4.936250982083996e-07,
+      "loss": 0.8401,
+      "step": 5844
+    },
+    {
+      "epoch": 1.4211038171650863,
+      "grad_norm": 20.125,
+      "learning_rate": 4.932404055543406e-07,
+      "loss": 0.6991,
+      "step": 5845
+    },
+    {
+      "epoch": 1.4213469486992463,
+      "grad_norm": 24.25,
+      "learning_rate": 4.928558260110729e-07,
+      "loss": 0.8347,
+      "step": 5846
+    },
+    {
+      "epoch": 1.4215900802334063,
+      "grad_norm": 25.75,
+      "learning_rate": 4.924713596360778e-07,
+      "loss": 0.6477,
+      "step": 5847
+    },
+    {
+      "epoch": 1.4218332117675663,
+      "grad_norm": 17.25,
+      "learning_rate": 4.920870064868214e-07,
+      "loss": 0.9365,
+      "step": 5848
+    },
+    {
+      "epoch": 1.4220763433017263,
+      "grad_norm": 21.75,
+      "learning_rate": 4.917027666207514e-07,
+      "loss": 0.6038,
+      "step": 5849
+    },
+    {
+      "epoch": 1.4223194748358863,
+      "grad_norm": 40.0,
+      "learning_rate": 4.91318640095299e-07,
+      "loss": 0.9744,
+      "step": 5850
+    },
+    {
+      "epoch": 1.4225626063700463,
+      "grad_norm": 18.5,
+      "learning_rate": 4.909346269678785e-07,
+      "loss": 0.7813,
+      "step": 5851
+    },
+    {
+      "epoch": 1.422805737904206,
+      "grad_norm": 18.25,
+      "learning_rate": 4.905507272958879e-07,
+      "loss": 0.6914,
+      "step": 5852
+    },
+    {
+      "epoch": 1.4230488694383663,
+      "grad_norm": 23.875,
+      "learning_rate": 4.901669411367073e-07,
+      "loss": 0.435,
+      "step": 5853
+    },
+    {
+      "epoch": 1.423292000972526,
+      "grad_norm": 24.75,
+      "learning_rate": 4.897832685476997e-07,
+      "loss": 0.7319,
+      "step": 5854
+    },
+    {
+      "epoch": 1.423535132506686,
+      "grad_norm": 22.75,
+      "learning_rate": 4.893997095862126e-07,
+      "loss": 0.6193,
+      "step": 5855
+    },
+    {
+      "epoch": 1.423778264040846,
+      "grad_norm": 18.75,
+      "learning_rate": 4.890162643095747e-07,
+      "loss": 0.5112,
+      "step": 5856
+    },
+    {
+      "epoch": 1.424021395575006,
+      "grad_norm": 24.75,
+      "learning_rate": 4.886329327750984e-07,
+      "loss": 0.7389,
+      "step": 5857
+    },
+    {
+      "epoch": 1.424264527109166,
+      "grad_norm": 23.375,
+      "learning_rate": 4.8824971504008e-07,
+      "loss": 0.8028,
+      "step": 5858
+    },
+    {
+      "epoch": 1.424507658643326,
+      "grad_norm": 17.0,
+      "learning_rate": 4.878666111617972e-07,
+      "loss": 0.5618,
+      "step": 5859
+    },
+    {
+      "epoch": 1.424750790177486,
+      "grad_norm": 33.25,
+      "learning_rate": 4.874836211975122e-07,
+      "loss": 0.9848,
+      "step": 5860
+    },
+    {
+      "epoch": 1.424993921711646,
+      "grad_norm": 32.75,
+      "learning_rate": 4.871007452044686e-07,
+      "loss": 0.6187,
+      "step": 5861
+    },
+    {
+      "epoch": 1.425237053245806,
+      "grad_norm": 19.125,
+      "learning_rate": 4.867179832398949e-07,
+      "loss": 0.9497,
+      "step": 5862
+    },
+    {
+      "epoch": 1.4254801847799659,
+      "grad_norm": 20.75,
+      "learning_rate": 4.863353353610011e-07,
+      "loss": 0.821,
+      "step": 5863
+    },
+    {
+      "epoch": 1.425723316314126,
+      "grad_norm": 25.875,
+      "learning_rate": 4.859528016249796e-07,
+      "loss": 0.8392,
+      "step": 5864
+    },
+    {
+      "epoch": 1.4259664478482859,
+      "grad_norm": 23.75,
+      "learning_rate": 4.855703820890083e-07,
+      "loss": 0.7004,
+      "step": 5865
+    },
+    {
+      "epoch": 1.4262095793824459,
+      "grad_norm": 18.75,
+      "learning_rate": 4.851880768102453e-07,
+      "loss": 0.5366,
+      "step": 5866
+    },
+    {
+      "epoch": 1.4264527109166059,
+      "grad_norm": 19.0,
+      "learning_rate": 4.848058858458326e-07,
+      "loss": 0.6864,
+      "step": 5867
+    },
+    {
+      "epoch": 1.4266958424507659,
+      "grad_norm": 19.0,
+      "learning_rate": 4.844238092528955e-07,
+      "loss": 0.4801,
+      "step": 5868
+    },
+    {
+      "epoch": 1.4269389739849259,
+      "grad_norm": 26.375,
+      "learning_rate": 4.840418470885426e-07,
+      "loss": 0.6771,
+      "step": 5869
+    },
+    {
+      "epoch": 1.4271821055190859,
+      "grad_norm": 21.625,
+      "learning_rate": 4.836599994098643e-07,
+      "loss": 0.7433,
+      "step": 5870
+    },
+    {
+      "epoch": 1.4274252370532459,
+      "grad_norm": 16.25,
+      "learning_rate": 4.832782662739336e-07,
+      "loss": 0.5384,
+      "step": 5871
+    },
+    {
+      "epoch": 1.4276683685874059,
+      "grad_norm": 18.625,
+      "learning_rate": 4.828966477378083e-07,
+      "loss": 0.5939,
+      "step": 5872
+    },
+    {
+      "epoch": 1.4279115001215659,
+      "grad_norm": 18.125,
+      "learning_rate": 4.825151438585272e-07,
+      "loss": 0.6243,
+      "step": 5873
+    },
+    {
+      "epoch": 1.4281546316557256,
+      "grad_norm": 19.75,
+      "learning_rate": 4.821337546931125e-07,
+      "loss": 0.5652,
+      "step": 5874
+    },
+    {
+      "epoch": 1.4283977631898856,
+      "grad_norm": 22.125,
+      "learning_rate": 4.817524802985697e-07,
+      "loss": 0.7576,
+      "step": 5875
+    },
+    {
+      "epoch": 1.4286408947240457,
+      "grad_norm": 19.0,
+      "learning_rate": 4.813713207318871e-07,
+      "loss": 0.7209,
+      "step": 5876
+    },
+    {
+      "epoch": 1.4288840262582057,
+      "grad_norm": 15.8125,
+      "learning_rate": 4.809902760500354e-07,
+      "loss": 0.9383,
+      "step": 5877
+    },
+    {
+      "epoch": 1.4291271577923657,
+      "grad_norm": 36.75,
+      "learning_rate": 4.806093463099677e-07,
+      "loss": 0.7053,
+      "step": 5878
+    },
+    {
+      "epoch": 1.4293702893265257,
+      "grad_norm": 20.625,
+      "learning_rate": 4.802285315686216e-07,
+      "loss": 0.6179,
+      "step": 5879
+    },
+    {
+      "epoch": 1.4296134208606857,
+      "grad_norm": 28.125,
+      "learning_rate": 4.798478318829157e-07,
+      "loss": 0.986,
+      "step": 5880
+    },
+    {
+      "epoch": 1.4298565523948457,
+      "grad_norm": 22.875,
+      "learning_rate": 4.79467247309752e-07,
+      "loss": 0.7236,
+      "step": 5881
+    },
+    {
+      "epoch": 1.4300996839290057,
+      "grad_norm": 18.75,
+      "learning_rate": 4.790867779060155e-07,
+      "loss": 0.6398,
+      "step": 5882
+    },
+    {
+      "epoch": 1.4303428154631654,
+      "grad_norm": 17.625,
+      "learning_rate": 4.787064237285749e-07,
+      "loss": 0.5652,
+      "step": 5883
+    },
+    {
+      "epoch": 1.4305859469973257,
+      "grad_norm": 23.5,
+      "learning_rate": 4.783261848342799e-07,
+      "loss": 0.6914,
+      "step": 5884
+    },
+    {
+      "epoch": 1.4308290785314854,
+      "grad_norm": 22.125,
+      "learning_rate": 4.779460612799635e-07,
+      "loss": 0.6014,
+      "step": 5885
+    },
+    {
+      "epoch": 1.4310722100656454,
+      "grad_norm": 33.5,
+      "learning_rate": 4.775660531224423e-07,
+      "loss": 0.7086,
+      "step": 5886
+    },
+    {
+      "epoch": 1.4313153415998054,
+      "grad_norm": 24.125,
+      "learning_rate": 4.771861604185149e-07,
+      "loss": 0.7281,
+      "step": 5887
+    },
+    {
+      "epoch": 1.4315584731339654,
+      "grad_norm": 25.625,
+      "learning_rate": 4.7680638322496253e-07,
+      "loss": 1.0503,
+      "step": 5888
+    },
+    {
+      "epoch": 1.4318016046681254,
+      "grad_norm": 23.75,
+      "learning_rate": 4.764267215985499e-07,
+      "loss": 0.7713,
+      "step": 5889
+    },
+    {
+      "epoch": 1.4320447362022855,
+      "grad_norm": 28.25,
+      "learning_rate": 4.7604717559602347e-07,
+      "loss": 0.9048,
+      "step": 5890
+    },
+    {
+      "epoch": 1.4322878677364455,
+      "grad_norm": 19.5,
+      "learning_rate": 4.7566774527411373e-07,
+      "loss": 1.0033,
+      "step": 5891
+    },
+    {
+      "epoch": 1.4325309992706055,
+      "grad_norm": 15.75,
+      "learning_rate": 4.7528843068953224e-07,
+      "loss": 0.3963,
+      "step": 5892
+    },
+    {
+      "epoch": 1.4327741308047655,
+      "grad_norm": 21.875,
+      "learning_rate": 4.7490923189897494e-07,
+      "loss": 0.9483,
+      "step": 5893
+    },
+    {
+      "epoch": 1.4330172623389252,
+      "grad_norm": 16.75,
+      "learning_rate": 4.745301489591191e-07,
+      "loss": 0.7503,
+      "step": 5894
+    },
+    {
+      "epoch": 1.4332603938730855,
+      "grad_norm": 18.75,
+      "learning_rate": 4.741511819266249e-07,
+      "loss": 0.5521,
+      "step": 5895
+    },
+    {
+      "epoch": 1.4335035254072452,
+      "grad_norm": 15.6875,
+      "learning_rate": 4.7377233085813656e-07,
+      "loss": 0.3149,
+      "step": 5896
+    },
+    {
+      "epoch": 1.4337466569414052,
+      "grad_norm": 34.0,
+      "learning_rate": 4.7339359581027907e-07,
+      "loss": 0.53,
+      "step": 5897
+    },
+    {
+      "epoch": 1.4339897884755652,
+      "grad_norm": 28.5,
+      "learning_rate": 4.730149768396604e-07,
+      "loss": 0.7893,
+      "step": 5898
+    },
+    {
+      "epoch": 1.4342329200097252,
+      "grad_norm": 23.375,
+      "learning_rate": 4.726364740028733e-07,
+      "loss": 0.6011,
+      "step": 5899
+    },
+    {
+      "epoch": 1.4344760515438852,
+      "grad_norm": 24.25,
+      "learning_rate": 4.7225808735649067e-07,
+      "loss": 0.4862,
+      "step": 5900
+    },
+    {
+      "epoch": 1.4347191830780452,
+      "grad_norm": 28.625,
+      "learning_rate": 4.7187981695706886e-07,
+      "loss": 0.9825,
+      "step": 5901
+    },
+    {
+      "epoch": 1.4349623146122052,
+      "grad_norm": 19.5,
+      "learning_rate": 4.715016628611467e-07,
+      "loss": 0.601,
+      "step": 5902
+    },
+    {
+      "epoch": 1.4352054461463652,
+      "grad_norm": 19.0,
+      "learning_rate": 4.7112362512524655e-07,
+      "loss": 0.8823,
+      "step": 5903
+    },
+    {
+      "epoch": 1.4354485776805253,
+      "grad_norm": 17.75,
+      "learning_rate": 4.7074570380587226e-07,
+      "loss": 0.4759,
+      "step": 5904
+    },
+    {
+      "epoch": 1.435691709214685,
+      "grad_norm": 22.75,
+      "learning_rate": 4.703678989595099e-07,
+      "loss": 0.8828,
+      "step": 5905
+    },
+    {
+      "epoch": 1.4359348407488453,
+      "grad_norm": 19.5,
+      "learning_rate": 4.6999021064263057e-07,
+      "loss": 0.5674,
+      "step": 5906
+    },
+    {
+      "epoch": 1.436177972283005,
+      "grad_norm": 19.0,
+      "learning_rate": 4.696126389116856e-07,
+      "loss": 0.7699,
+      "step": 5907
+    },
+    {
+      "epoch": 1.436421103817165,
+      "grad_norm": 19.25,
+      "learning_rate": 4.6923518382310933e-07,
+      "loss": 0.6402,
+      "step": 5908
+    },
+    {
+      "epoch": 1.436664235351325,
+      "grad_norm": 21.25,
+      "learning_rate": 4.688578454333188e-07,
+      "loss": 0.7068,
+      "step": 5909
+    },
+    {
+      "epoch": 1.436907366885485,
+      "grad_norm": 17.5,
+      "learning_rate": 4.6848062379871454e-07,
+      "loss": 0.4421,
+      "step": 5910
+    },
+    {
+      "epoch": 1.437150498419645,
+      "grad_norm": 19.625,
+      "learning_rate": 4.681035189756783e-07,
+      "loss": 0.6486,
+      "step": 5911
+    },
+    {
+      "epoch": 1.437393629953805,
+      "grad_norm": 20.0,
+      "learning_rate": 4.6772653102057447e-07,
+      "loss": 0.5045,
+      "step": 5912
+    },
+    {
+      "epoch": 1.437636761487965,
+      "grad_norm": 19.625,
+      "learning_rate": 4.6734965998975106e-07,
+      "loss": 0.8552,
+      "step": 5913
+    },
+    {
+      "epoch": 1.437879893022125,
+      "grad_norm": 17.75,
+      "learning_rate": 4.669729059395381e-07,
+      "loss": 0.6541,
+      "step": 5914
+    },
+    {
+      "epoch": 1.438123024556285,
+      "grad_norm": 18.875,
+      "learning_rate": 4.6659626892624723e-07,
+      "loss": 0.6457,
+      "step": 5915
+    },
+    {
+      "epoch": 1.4383661560904448,
+      "grad_norm": 24.875,
+      "learning_rate": 4.6621974900617444e-07,
+      "loss": 0.9936,
+      "step": 5916
+    },
+    {
+      "epoch": 1.438609287624605,
+      "grad_norm": 18.5,
+      "learning_rate": 4.658433462355963e-07,
+      "loss": 0.6769,
+      "step": 5917
+    },
+    {
+      "epoch": 1.4388524191587648,
+      "grad_norm": 20.0,
+      "learning_rate": 4.65467060670773e-07,
+      "loss": 0.5245,
+      "step": 5918
+    },
+    {
+      "epoch": 1.4390955506929248,
+      "grad_norm": 21.375,
+      "learning_rate": 4.6509089236794645e-07,
+      "loss": 0.5577,
+      "step": 5919
+    },
+    {
+      "epoch": 1.4393386822270848,
+      "grad_norm": 18.375,
+      "learning_rate": 4.647148413833423e-07,
+      "loss": 0.4768,
+      "step": 5920
+    },
+    {
+      "epoch": 1.4395818137612448,
+      "grad_norm": 21.25,
+      "learning_rate": 4.643389077731669e-07,
+      "loss": 0.9441,
+      "step": 5921
+    },
+    {
+      "epoch": 1.4398249452954048,
+      "grad_norm": 22.875,
+      "learning_rate": 4.639630915936108e-07,
+      "loss": 0.7449,
+      "step": 5922
+    },
+    {
+      "epoch": 1.4400680768295648,
+      "grad_norm": 19.875,
+      "learning_rate": 4.635873929008462e-07,
+      "loss": 0.32,
+      "step": 5923
+    },
+    {
+      "epoch": 1.4403112083637248,
+      "grad_norm": 20.0,
+      "learning_rate": 4.6321181175102773e-07,
+      "loss": 0.458,
+      "step": 5924
+    },
+    {
+      "epoch": 1.4405543398978846,
+      "grad_norm": 16.125,
+      "learning_rate": 4.628363482002922e-07,
+      "loss": 0.5633,
+      "step": 5925
+    },
+    {
+      "epoch": 1.4407974714320448,
+      "grad_norm": 20.875,
+      "learning_rate": 4.6246100230475876e-07,
+      "loss": 0.8196,
+      "step": 5926
+    },
+    {
+      "epoch": 1.4410406029662046,
+      "grad_norm": 17.125,
+      "learning_rate": 4.620857741205302e-07,
+      "loss": 0.3961,
+      "step": 5927
+    },
+    {
+      "epoch": 1.4412837345003646,
+      "grad_norm": 21.375,
+      "learning_rate": 4.6171066370369056e-07,
+      "loss": 0.7585,
+      "step": 5928
+    },
+    {
+      "epoch": 1.4415268660345246,
+      "grad_norm": 22.125,
+      "learning_rate": 4.613356711103055e-07,
+      "loss": 0.8838,
+      "step": 5929
+    },
+    {
+      "epoch": 1.4417699975686846,
+      "grad_norm": 15.5625,
+      "learning_rate": 4.60960796396426e-07,
+      "loss": 0.3827,
+      "step": 5930
+    },
+    {
+      "epoch": 1.4420131291028446,
+      "grad_norm": 24.75,
+      "learning_rate": 4.605860396180828e-07,
+      "loss": 0.7036,
+      "step": 5931
+    },
+    {
+      "epoch": 1.4422562606370046,
+      "grad_norm": 17.25,
+      "learning_rate": 4.60211400831289e-07,
+      "loss": 0.6268,
+      "step": 5932
+    },
+    {
+      "epoch": 1.4424993921711646,
+      "grad_norm": 27.375,
+      "learning_rate": 4.598368800920419e-07,
+      "loss": 0.9218,
+      "step": 5933
+    },
+    {
+      "epoch": 1.4427425237053246,
+      "grad_norm": 24.125,
+      "learning_rate": 4.5946247745631973e-07,
+      "loss": 0.6123,
+      "step": 5934
+    },
+    {
+      "epoch": 1.4429856552394846,
+      "grad_norm": 19.0,
+      "learning_rate": 4.590881929800833e-07,
+      "loss": 0.5335,
+      "step": 5935
+    },
+    {
+      "epoch": 1.4432287867736444,
+      "grad_norm": 21.375,
+      "learning_rate": 4.5871402671927523e-07,
+      "loss": 0.727,
+      "step": 5936
+    },
+    {
+      "epoch": 1.4434719183078046,
+      "grad_norm": 25.25,
+      "learning_rate": 4.5833997872982263e-07,
+      "loss": 0.6745,
+      "step": 5937
+    },
+    {
+      "epoch": 1.4437150498419644,
+      "grad_norm": 19.125,
+      "learning_rate": 4.5796604906763254e-07,
+      "loss": 0.5413,
+      "step": 5938
+    },
+    {
+      "epoch": 1.4439581813761244,
+      "grad_norm": 22.375,
+      "learning_rate": 4.5759223778859495e-07,
+      "loss": 0.7198,
+      "step": 5939
+    },
+    {
+      "epoch": 1.4442013129102844,
+      "grad_norm": 16.0,
+      "learning_rate": 4.5721854494858333e-07,
+      "loss": 0.7373,
+      "step": 5940
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 26.125,
+      "learning_rate": 4.568449706034518e-07,
+      "loss": 0.792,
+      "step": 5941
+    },
+    {
+      "epoch": 1.4446875759786044,
+      "grad_norm": 20.25,
+      "learning_rate": 4.5647151480903775e-07,
+      "loss": 1.1198,
+      "step": 5942
+    },
+    {
+      "epoch": 1.4449307075127644,
+      "grad_norm": 25.75,
+      "learning_rate": 4.560981776211601e-07,
+      "loss": 0.4278,
+      "step": 5943
+    },
+    {
+      "epoch": 1.4451738390469244,
+      "grad_norm": 21.625,
+      "learning_rate": 4.55724959095621e-07,
+      "loss": 0.8805,
+      "step": 5944
+    },
+    {
+      "epoch": 1.4454169705810844,
+      "grad_norm": 21.875,
+      "learning_rate": 4.5535185928820474e-07,
+      "loss": 0.8968,
+      "step": 5945
+    },
+    {
+      "epoch": 1.4456601021152444,
+      "grad_norm": 21.75,
+      "learning_rate": 4.5497887825467684e-07,
+      "loss": 1.1248,
+      "step": 5946
+    },
+    {
+      "epoch": 1.4459032336494042,
+      "grad_norm": 21.0,
+      "learning_rate": 4.546060160507863e-07,
+      "loss": 0.5402,
+      "step": 5947
+    },
+    {
+      "epoch": 1.4461463651835644,
+      "grad_norm": 21.25,
+      "learning_rate": 4.5423327273226376e-07,
+      "loss": 0.6048,
+      "step": 5948
+    },
+    {
+      "epoch": 1.4463894967177242,
+      "grad_norm": 20.25,
+      "learning_rate": 4.538606483548215e-07,
+      "loss": 0.7775,
+      "step": 5949
+    },
+    {
+      "epoch": 1.4466326282518842,
+      "grad_norm": 17.875,
+      "learning_rate": 4.5348814297415555e-07,
+      "loss": 0.6131,
+      "step": 5950
+    },
+    {
+      "epoch": 1.4468757597860442,
+      "grad_norm": 20.875,
+      "learning_rate": 4.531157566459429e-07,
+      "loss": 0.4645,
+      "step": 5951
+    },
+    {
+      "epoch": 1.4471188913202042,
+      "grad_norm": 17.125,
+      "learning_rate": 4.527434894258427e-07,
+      "loss": 0.6405,
+      "step": 5952
+    },
+    {
+      "epoch": 1.4473620228543642,
+      "grad_norm": 19.375,
+      "learning_rate": 4.52371341369497e-07,
+      "loss": 0.5227,
+      "step": 5953
+    },
+    {
+      "epoch": 1.4476051543885242,
+      "grad_norm": 19.0,
+      "learning_rate": 4.519993125325303e-07,
+      "loss": 0.6499,
+      "step": 5954
+    },
+    {
+      "epoch": 1.4478482859226842,
+      "grad_norm": 19.75,
+      "learning_rate": 4.516274029705484e-07,
+      "loss": 0.7574,
+      "step": 5955
+    },
+    {
+      "epoch": 1.4480914174568442,
+      "grad_norm": 20.375,
+      "learning_rate": 4.512556127391389e-07,
+      "loss": 0.5497,
+      "step": 5956
+    },
+    {
+      "epoch": 1.4483345489910042,
+      "grad_norm": 17.5,
+      "learning_rate": 4.5088394189387347e-07,
+      "loss": 0.3428,
+      "step": 5957
+    },
+    {
+      "epoch": 1.448577680525164,
+      "grad_norm": 21.375,
+      "learning_rate": 4.5051239049030403e-07,
+      "loss": 0.6076,
+      "step": 5958
+    },
+    {
+      "epoch": 1.4488208120593242,
+      "grad_norm": 27.0,
+      "learning_rate": 4.50140958583965e-07,
+      "loss": 1.128,
+      "step": 5959
+    },
+    {
+      "epoch": 1.449063943593484,
+      "grad_norm": 20.875,
+      "learning_rate": 4.497696462303738e-07,
+      "loss": 0.6823,
+      "step": 5960
+    },
+    {
+      "epoch": 1.449307075127644,
+      "grad_norm": 20.25,
+      "learning_rate": 4.4939845348502985e-07,
+      "loss": 0.8267,
+      "step": 5961
+    },
+    {
+      "epoch": 1.449550206661804,
+      "grad_norm": 24.75,
+      "learning_rate": 4.4902738040341396e-07,
+      "loss": 0.9122,
+      "step": 5962
+    },
+    {
+      "epoch": 1.449793338195964,
+      "grad_norm": 24.25,
+      "learning_rate": 4.4865642704098893e-07,
+      "loss": 0.6708,
+      "step": 5963
+    },
+    {
+      "epoch": 1.450036469730124,
+      "grad_norm": 17.375,
+      "learning_rate": 4.4828559345320085e-07,
+      "loss": 0.4427,
+      "step": 5964
+    },
+    {
+      "epoch": 1.450279601264284,
+      "grad_norm": 21.625,
+      "learning_rate": 4.47914879695477e-07,
+      "loss": 0.8135,
+      "step": 5965
+    },
+    {
+      "epoch": 1.450522732798444,
+      "grad_norm": 19.125,
+      "learning_rate": 4.475442858232264e-07,
+      "loss": 0.5175,
+      "step": 5966
+    },
+    {
+      "epoch": 1.450765864332604,
+      "grad_norm": 14.9375,
+      "learning_rate": 4.4717381189184116e-07,
+      "loss": 0.3037,
+      "step": 5967
+    },
+    {
+      "epoch": 1.451008995866764,
+      "grad_norm": 18.0,
+      "learning_rate": 4.468034579566956e-07,
+      "loss": 0.7232,
+      "step": 5968
+    },
+    {
+      "epoch": 1.4512521274009238,
+      "grad_norm": 17.875,
+      "learning_rate": 4.464332240731448e-07,
+      "loss": 0.4817,
+      "step": 5969
+    },
+    {
+      "epoch": 1.4514952589350838,
+      "grad_norm": 25.5,
+      "learning_rate": 4.460631102965263e-07,
+      "loss": 0.6684,
+      "step": 5970
+    },
+    {
+      "epoch": 1.4517383904692438,
+      "grad_norm": 23.875,
+      "learning_rate": 4.4569311668216083e-07,
+      "loss": 0.5933,
+      "step": 5971
+    },
+    {
+      "epoch": 1.4519815220034038,
+      "grad_norm": 24.125,
+      "learning_rate": 4.453232432853501e-07,
+      "loss": 0.8794,
+      "step": 5972
+    },
+    {
+      "epoch": 1.4522246535375638,
+      "grad_norm": 23.375,
+      "learning_rate": 4.449534901613774e-07,
+      "loss": 0.9315,
+      "step": 5973
+    },
+    {
+      "epoch": 1.4524677850717238,
+      "grad_norm": 16.125,
+      "learning_rate": 4.4458385736550964e-07,
+      "loss": 0.3689,
+      "step": 5974
+    },
+    {
+      "epoch": 1.4527109166058838,
+      "grad_norm": 35.75,
+      "learning_rate": 4.44214344952994e-07,
+      "loss": 1.4904,
+      "step": 5975
+    },
+    {
+      "epoch": 1.4529540481400438,
+      "grad_norm": 23.75,
+      "learning_rate": 4.438449529790613e-07,
+      "loss": 0.9447,
+      "step": 5976
+    },
+    {
+      "epoch": 1.4531971796742038,
+      "grad_norm": 19.0,
+      "learning_rate": 4.4347568149892274e-07,
+      "loss": 0.6315,
+      "step": 5977
+    },
+    {
+      "epoch": 1.4534403112083636,
+      "grad_norm": 19.75,
+      "learning_rate": 4.4310653056777313e-07,
+      "loss": 0.5131,
+      "step": 5978
+    },
+    {
+      "epoch": 1.4536834427425238,
+      "grad_norm": 18.25,
+      "learning_rate": 4.427375002407881e-07,
+      "loss": 0.5167,
+      "step": 5979
+    },
+    {
+      "epoch": 1.4539265742766836,
+      "grad_norm": 19.625,
+      "learning_rate": 4.4236859057312523e-07,
+      "loss": 0.9916,
+      "step": 5980
+    },
+    {
+      "epoch": 1.4541697058108436,
+      "grad_norm": 20.375,
+      "learning_rate": 4.419998016199251e-07,
+      "loss": 0.9532,
+      "step": 5981
+    },
+    {
+      "epoch": 1.4544128373450036,
+      "grad_norm": 19.25,
+      "learning_rate": 4.416311334363092e-07,
+      "loss": 0.7733,
+      "step": 5982
+    },
+    {
+      "epoch": 1.4546559688791636,
+      "grad_norm": 23.125,
+      "learning_rate": 4.4126258607738115e-07,
+      "loss": 0.7843,
+      "step": 5983
+    },
+    {
+      "epoch": 1.4548991004133236,
+      "grad_norm": 30.5,
+      "learning_rate": 4.408941595982269e-07,
+      "loss": 0.7897,
+      "step": 5984
+    },
+    {
+      "epoch": 1.4551422319474836,
+      "grad_norm": 24.625,
+      "learning_rate": 4.4052585405391464e-07,
+      "loss": 0.6895,
+      "step": 5985
+    },
+    {
+      "epoch": 1.4553853634816436,
+      "grad_norm": 24.875,
+      "learning_rate": 4.401576694994937e-07,
+      "loss": 1.0368,
+      "step": 5986
+    },
+    {
+      "epoch": 1.4556284950158036,
+      "grad_norm": 20.0,
+      "learning_rate": 4.3978960598999505e-07,
+      "loss": 0.7658,
+      "step": 5987
+    },
+    {
+      "epoch": 1.4558716265499636,
+      "grad_norm": 23.25,
+      "learning_rate": 4.39421663580433e-07,
+      "loss": 0.469,
+      "step": 5988
+    },
+    {
+      "epoch": 1.4561147580841234,
+      "grad_norm": 27.25,
+      "learning_rate": 4.390538423258024e-07,
+      "loss": 1.2069,
+      "step": 5989
+    },
+    {
+      "epoch": 1.4563578896182836,
+      "grad_norm": 24.625,
+      "learning_rate": 4.386861422810802e-07,
+      "loss": 0.7784,
+      "step": 5990
+    },
+    {
+      "epoch": 1.4566010211524434,
+      "grad_norm": 22.125,
+      "learning_rate": 4.3831856350122603e-07,
+      "loss": 0.6391,
+      "step": 5991
+    },
+    {
+      "epoch": 1.4568441526866034,
+      "grad_norm": 21.625,
+      "learning_rate": 4.3795110604118117e-07,
+      "loss": 0.8035,
+      "step": 5992
+    },
+    {
+      "epoch": 1.4570872842207634,
+      "grad_norm": 19.875,
+      "learning_rate": 4.37583769955868e-07,
+      "loss": 0.75,
+      "step": 5993
+    },
+    {
+      "epoch": 1.4573304157549234,
+      "grad_norm": 21.75,
+      "learning_rate": 4.37216555300191e-07,
+      "loss": 0.8879,
+      "step": 5994
+    },
+    {
+      "epoch": 1.4575735472890834,
+      "grad_norm": 19.875,
+      "learning_rate": 4.3684946212903743e-07,
+      "loss": 0.6677,
+      "step": 5995
+    },
+    {
+      "epoch": 1.4578166788232434,
+      "grad_norm": 15.125,
+      "learning_rate": 4.364824904972753e-07,
+      "loss": 0.5898,
+      "step": 5996
+    },
+    {
+      "epoch": 1.4580598103574034,
+      "grad_norm": 23.25,
+      "learning_rate": 4.3611564045975453e-07,
+      "loss": 0.9381,
+      "step": 5997
+    },
+    {
+      "epoch": 1.4583029418915634,
+      "grad_norm": 18.5,
+      "learning_rate": 4.3574891207130767e-07,
+      "loss": 0.6625,
+      "step": 5998
+    },
+    {
+      "epoch": 1.4585460734257234,
+      "grad_norm": 17.0,
+      "learning_rate": 4.353823053867487e-07,
+      "loss": 0.4623,
+      "step": 5999
+    },
+    {
+      "epoch": 1.4587892049598832,
+      "grad_norm": 19.375,
+      "learning_rate": 4.350158204608733e-07,
+      "loss": 0.4713,
+      "step": 6000
+    },
+    {
+      "epoch": 1.4590323364940434,
+      "grad_norm": 19.375,
+      "learning_rate": 4.346494573484583e-07,
+      "loss": 0.7481,
+      "step": 6001
+    },
+    {
+      "epoch": 1.4592754680282032,
+      "grad_norm": 18.625,
+      "learning_rate": 4.34283216104264e-07,
+      "loss": 0.5707,
+      "step": 6002
+    },
+    {
+      "epoch": 1.4595185995623632,
+      "grad_norm": 26.125,
+      "learning_rate": 4.3391709678303083e-07,
+      "loss": 0.6832,
+      "step": 6003
+    },
+    {
+      "epoch": 1.4597617310965232,
+      "grad_norm": 28.5,
+      "learning_rate": 4.335510994394815e-07,
+      "loss": 0.9882,
+      "step": 6004
+    },
+    {
+      "epoch": 1.4600048626306832,
+      "grad_norm": 24.25,
+      "learning_rate": 4.3318522412832117e-07,
+      "loss": 1.2792,
+      "step": 6005
+    },
+    {
+      "epoch": 1.4602479941648432,
+      "grad_norm": 13.75,
+      "learning_rate": 4.3281947090423564e-07,
+      "loss": 0.2988,
+      "step": 6006
+    },
+    {
+      "epoch": 1.4604911256990032,
+      "grad_norm": 23.5,
+      "learning_rate": 4.324538398218939e-07,
+      "loss": 0.861,
+      "step": 6007
+    },
+    {
+      "epoch": 1.4607342572331632,
+      "grad_norm": 21.625,
+      "learning_rate": 4.320883309359447e-07,
+      "loss": 1.0645,
+      "step": 6008
+    },
+    {
+      "epoch": 1.4609773887673232,
+      "grad_norm": 12.3125,
+      "learning_rate": 4.3172294430102083e-07,
+      "loss": 0.2965,
+      "step": 6009
+    },
+    {
+      "epoch": 1.4612205203014832,
+      "grad_norm": 17.25,
+      "learning_rate": 4.3135767997173506e-07,
+      "loss": 0.5699,
+      "step": 6010
+    },
+    {
+      "epoch": 1.461463651835643,
+      "grad_norm": 15.6875,
+      "learning_rate": 4.3099253800268194e-07,
+      "loss": 0.3311,
+      "step": 6011
+    },
+    {
+      "epoch": 1.4617067833698032,
+      "grad_norm": 19.125,
+      "learning_rate": 4.306275184484393e-07,
+      "loss": 0.8128,
+      "step": 6012
+    },
+    {
+      "epoch": 1.461949914903963,
+      "grad_norm": 18.375,
+      "learning_rate": 4.302626213635645e-07,
+      "loss": 0.6445,
+      "step": 6013
+    },
+    {
+      "epoch": 1.462193046438123,
+      "grad_norm": 24.625,
+      "learning_rate": 4.2989784680259895e-07,
+      "loss": 0.9853,
+      "step": 6014
+    },
+    {
+      "epoch": 1.462436177972283,
+      "grad_norm": 20.875,
+      "learning_rate": 4.295331948200633e-07,
+      "loss": 0.5928,
+      "step": 6015
+    },
+    {
+      "epoch": 1.462679309506443,
+      "grad_norm": 22.875,
+      "learning_rate": 4.29168665470462e-07,
+      "loss": 0.9678,
+      "step": 6016
+    },
+    {
+      "epoch": 1.462922441040603,
+      "grad_norm": 20.375,
+      "learning_rate": 4.2880425880828e-07,
+      "loss": 0.7878,
+      "step": 6017
+    },
+    {
+      "epoch": 1.463165572574763,
+      "grad_norm": 24.5,
+      "learning_rate": 4.284399748879836e-07,
+      "loss": 0.9565,
+      "step": 6018
+    },
+    {
+      "epoch": 1.463408704108923,
+      "grad_norm": 24.875,
+      "learning_rate": 4.2807581376402214e-07,
+      "loss": 1.0191,
+      "step": 6019
+    },
+    {
+      "epoch": 1.463651835643083,
+      "grad_norm": 43.25,
+      "learning_rate": 4.2771177549082557e-07,
+      "loss": 0.7639,
+      "step": 6020
+    },
+    {
+      "epoch": 1.463894967177243,
+      "grad_norm": 21.25,
+      "learning_rate": 4.2734786012280495e-07,
+      "loss": 0.8329,
+      "step": 6021
+    },
+    {
+      "epoch": 1.4641380987114028,
+      "grad_norm": 15.75,
+      "learning_rate": 4.269840677143544e-07,
+      "loss": 0.5673,
+      "step": 6022
+    },
+    {
+      "epoch": 1.4643812302455628,
+      "grad_norm": 19.875,
+      "learning_rate": 4.2662039831984944e-07,
+      "loss": 0.7499,
+      "step": 6023
+    },
+    {
+      "epoch": 1.4646243617797228,
+      "grad_norm": 29.0,
+      "learning_rate": 4.2625685199364606e-07,
+      "loss": 0.7239,
+      "step": 6024
+    },
+    {
+      "epoch": 1.4648674933138828,
+      "grad_norm": 15.1875,
+      "learning_rate": 4.258934287900825e-07,
+      "loss": 0.5465,
+      "step": 6025
+    },
+    {
+      "epoch": 1.4651106248480428,
+      "grad_norm": 20.25,
+      "learning_rate": 4.2553012876347904e-07,
+      "loss": 0.7416,
+      "step": 6026
+    },
+    {
+      "epoch": 1.4653537563822028,
+      "grad_norm": 24.625,
+      "learning_rate": 4.251669519681369e-07,
+      "loss": 0.6116,
+      "step": 6027
+    },
+    {
+      "epoch": 1.4655968879163628,
+      "grad_norm": 23.25,
+      "learning_rate": 4.2480389845833884e-07,
+      "loss": 0.8547,
+      "step": 6028
+    },
+    {
+      "epoch": 1.4658400194505228,
+      "grad_norm": 20.25,
+      "learning_rate": 4.244409682883498e-07,
+      "loss": 0.8097,
+      "step": 6029
+    },
+    {
+      "epoch": 1.4660831509846828,
+      "grad_norm": 17.875,
+      "learning_rate": 4.240781615124162e-07,
+      "loss": 0.8004,
+      "step": 6030
+    },
+    {
+      "epoch": 1.4663262825188426,
+      "grad_norm": 17.125,
+      "learning_rate": 4.237154781847656e-07,
+      "loss": 0.6866,
+      "step": 6031
+    },
+    {
+      "epoch": 1.4665694140530028,
+      "grad_norm": 21.5,
+      "learning_rate": 4.233529183596069e-07,
+      "loss": 0.5433,
+      "step": 6032
+    },
+    {
+      "epoch": 1.4668125455871626,
+      "grad_norm": 25.0,
+      "learning_rate": 4.2299048209113154e-07,
+      "loss": 0.7478,
+      "step": 6033
+    },
+    {
+      "epoch": 1.4670556771213226,
+      "grad_norm": 21.125,
+      "learning_rate": 4.226281694335117e-07,
+      "loss": 0.5145,
+      "step": 6034
+    },
+    {
+      "epoch": 1.4672988086554826,
+      "grad_norm": 23.0,
+      "learning_rate": 4.2226598044090057e-07,
+      "loss": 0.8473,
+      "step": 6035
+    },
+    {
+      "epoch": 1.4675419401896426,
+      "grad_norm": 20.625,
+      "learning_rate": 4.2190391516743464e-07,
+      "loss": 1.1187,
+      "step": 6036
+    },
+    {
+      "epoch": 1.4677850717238026,
+      "grad_norm": 16.375,
+      "learning_rate": 4.2154197366722994e-07,
+      "loss": 0.6487,
+      "step": 6037
+    },
+    {
+      "epoch": 1.4680282032579626,
+      "grad_norm": 18.25,
+      "learning_rate": 4.2118015599438555e-07,
+      "loss": 0.7943,
+      "step": 6038
+    },
+    {
+      "epoch": 1.4682713347921226,
+      "grad_norm": 17.75,
+      "learning_rate": 4.2081846220298066e-07,
+      "loss": 0.6841,
+      "step": 6039
+    },
+    {
+      "epoch": 1.4685144663262826,
+      "grad_norm": 21.625,
+      "learning_rate": 4.204568923470775e-07,
+      "loss": 0.6428,
+      "step": 6040
+    },
+    {
+      "epoch": 1.4687575978604426,
+      "grad_norm": 19.625,
+      "learning_rate": 4.2009544648071847e-07,
+      "loss": 0.5299,
+      "step": 6041
+    },
+    {
+      "epoch": 1.4690007293946024,
+      "grad_norm": 17.5,
+      "learning_rate": 4.197341246579276e-07,
+      "loss": 0.2902,
+      "step": 6042
+    },
+    {
+      "epoch": 1.4692438609287626,
+      "grad_norm": 25.625,
+      "learning_rate": 4.1937292693271123e-07,
+      "loss": 0.7123,
+      "step": 6043
+    },
+    {
+      "epoch": 1.4694869924629224,
+      "grad_norm": 18.75,
+      "learning_rate": 4.1901185335905604e-07,
+      "loss": 0.5246,
+      "step": 6044
+    },
+    {
+      "epoch": 1.4697301239970824,
+      "grad_norm": 21.375,
+      "learning_rate": 4.1865090399093146e-07,
+      "loss": 0.6615,
+      "step": 6045
+    },
+    {
+      "epoch": 1.4699732555312424,
+      "grad_norm": 15.3125,
+      "learning_rate": 4.182900788822866e-07,
+      "loss": 0.5837,
+      "step": 6046
+    },
+    {
+      "epoch": 1.4702163870654024,
+      "grad_norm": 23.75,
+      "learning_rate": 4.1792937808705413e-07,
+      "loss": 0.948,
+      "step": 6047
+    },
+    {
+      "epoch": 1.4704595185995624,
+      "grad_norm": 20.25,
+      "learning_rate": 4.175688016591464e-07,
+      "loss": 0.9462,
+      "step": 6048
+    },
+    {
+      "epoch": 1.4707026501337224,
+      "grad_norm": 22.5,
+      "learning_rate": 4.1720834965245747e-07,
+      "loss": 0.6936,
+      "step": 6049
+    },
+    {
+      "epoch": 1.4709457816678824,
+      "grad_norm": 34.25,
+      "learning_rate": 4.16848022120864e-07,
+      "loss": 0.7489,
+      "step": 6050
+    },
+    {
+      "epoch": 1.4711889132020424,
+      "grad_norm": 22.25,
+      "learning_rate": 4.164878191182226e-07,
+      "loss": 0.8976,
+      "step": 6051
+    },
+    {
+      "epoch": 1.4714320447362024,
+      "grad_norm": 27.625,
+      "learning_rate": 4.161277406983713e-07,
+      "loss": 0.8001,
+      "step": 6052
+    },
+    {
+      "epoch": 1.4716751762703622,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.1576778691513084e-07,
+      "loss": 0.2567,
+      "step": 6053
+    },
+    {
+      "epoch": 1.4719183078045224,
+      "grad_norm": 18.625,
+      "learning_rate": 4.1540795782230264e-07,
+      "loss": 0.88,
+      "step": 6054
+    },
+    {
+      "epoch": 1.4721614393386822,
+      "grad_norm": 33.5,
+      "learning_rate": 4.150482534736691e-07,
+      "loss": 1.0864,
+      "step": 6055
+    },
+    {
+      "epoch": 1.4724045708728422,
+      "grad_norm": 24.0,
+      "learning_rate": 4.146886739229937e-07,
+      "loss": 0.9677,
+      "step": 6056
+    },
+    {
+      "epoch": 1.4726477024070022,
+      "grad_norm": 20.25,
+      "learning_rate": 4.143292192240228e-07,
+      "loss": 0.4961,
+      "step": 6057
+    },
+    {
+      "epoch": 1.4728908339411622,
+      "grad_norm": 14.5,
+      "learning_rate": 4.139698894304825e-07,
+      "loss": 0.3054,
+      "step": 6058
+    },
+    {
+      "epoch": 1.4731339654753222,
+      "grad_norm": 18.25,
+      "learning_rate": 4.1361068459608065e-07,
+      "loss": 0.788,
+      "step": 6059
+    },
+    {
+      "epoch": 1.4733770970094822,
+      "grad_norm": 25.75,
+      "learning_rate": 4.1325160477450694e-07,
+      "loss": 0.8181,
+      "step": 6060
+    },
+    {
+      "epoch": 1.4736202285436422,
+      "grad_norm": 15.8125,
+      "learning_rate": 4.1289265001943244e-07,
+      "loss": 0.6519,
+      "step": 6061
+    },
+    {
+      "epoch": 1.4738633600778022,
+      "grad_norm": 24.625,
+      "learning_rate": 4.1253382038450874e-07,
+      "loss": 0.6285,
+      "step": 6062
+    },
+    {
+      "epoch": 1.4741064916119622,
+      "grad_norm": 26.0,
+      "learning_rate": 4.121751159233686e-07,
+      "loss": 0.6115,
+      "step": 6063
+    },
+    {
+      "epoch": 1.474349623146122,
+      "grad_norm": 16.625,
+      "learning_rate": 4.118165366896278e-07,
+      "loss": 0.5134,
+      "step": 6064
+    },
+    {
+      "epoch": 1.4745927546802822,
+      "grad_norm": 19.75,
+      "learning_rate": 4.114580827368812e-07,
+      "loss": 0.8648,
+      "step": 6065
+    },
+    {
+      "epoch": 1.474835886214442,
+      "grad_norm": 24.75,
+      "learning_rate": 4.1109975411870603e-07,
+      "loss": 1.2879,
+      "step": 6066
+    },
+    {
+      "epoch": 1.475079017748602,
+      "grad_norm": 20.375,
+      "learning_rate": 4.107415508886613e-07,
+      "loss": 0.6188,
+      "step": 6067
+    },
+    {
+      "epoch": 1.475322149282762,
+      "grad_norm": 17.375,
+      "learning_rate": 4.1038347310028595e-07,
+      "loss": 0.5522,
+      "step": 6068
+    },
+    {
+      "epoch": 1.475565280816922,
+      "grad_norm": 22.25,
+      "learning_rate": 4.1002552080710146e-07,
+      "loss": 0.8245,
+      "step": 6069
+    },
+    {
+      "epoch": 1.475808412351082,
+      "grad_norm": 18.875,
+      "learning_rate": 4.0966769406260943e-07,
+      "loss": 0.6155,
+      "step": 6070
+    },
+    {
+      "epoch": 1.476051543885242,
+      "grad_norm": 27.25,
+      "learning_rate": 4.09309992920294e-07,
+      "loss": 0.6586,
+      "step": 6071
+    },
+    {
+      "epoch": 1.476294675419402,
+      "grad_norm": 21.125,
+      "learning_rate": 4.089524174336193e-07,
+      "loss": 0.6301,
+      "step": 6072
+    },
+    {
+      "epoch": 1.476537806953562,
+      "grad_norm": 14.875,
+      "learning_rate": 4.085949676560308e-07,
+      "loss": 0.5514,
+      "step": 6073
+    },
+    {
+      "epoch": 1.476780938487722,
+      "grad_norm": 33.5,
+      "learning_rate": 4.0823764364095643e-07,
+      "loss": 1.1112,
+      "step": 6074
+    },
+    {
+      "epoch": 1.4770240700218817,
+      "grad_norm": 26.375,
+      "learning_rate": 4.0788044544180354e-07,
+      "loss": 0.8851,
+      "step": 6075
+    },
+    {
+      "epoch": 1.4772672015560417,
+      "grad_norm": 19.375,
+      "learning_rate": 4.0752337311196224e-07,
+      "loss": 0.6695,
+      "step": 6076
+    },
+    {
+      "epoch": 1.4775103330902017,
+      "grad_norm": 20.25,
+      "learning_rate": 4.071664267048027e-07,
+      "loss": 1.2643,
+      "step": 6077
+    },
+    {
+      "epoch": 1.4777534646243617,
+      "grad_norm": 26.0,
+      "learning_rate": 4.0680960627367717e-07,
+      "loss": 0.9361,
+      "step": 6078
+    },
+    {
+      "epoch": 1.4779965961585217,
+      "grad_norm": 19.375,
+      "learning_rate": 4.0645291187191847e-07,
+      "loss": 0.6155,
+      "step": 6079
+    },
+    {
+      "epoch": 1.4782397276926817,
+      "grad_norm": 20.625,
+      "learning_rate": 4.060963435528402e-07,
+      "loss": 0.4762,
+      "step": 6080
+    },
+    {
+      "epoch": 1.4784828592268418,
+      "grad_norm": 20.25,
+      "learning_rate": 4.0573990136973855e-07,
+      "loss": 0.6792,
+      "step": 6081
+    },
+    {
+      "epoch": 1.4787259907610018,
+      "grad_norm": 18.5,
+      "learning_rate": 4.053835853758896e-07,
+      "loss": 0.6455,
+      "step": 6082
+    },
+    {
+      "epoch": 1.4789691222951618,
+      "grad_norm": 20.0,
+      "learning_rate": 4.0502739562455026e-07,
+      "loss": 0.764,
+      "step": 6083
+    },
+    {
+      "epoch": 1.4792122538293215,
+      "grad_norm": 16.375,
+      "learning_rate": 4.0467133216895994e-07,
+      "loss": 0.6791,
+      "step": 6084
+    },
+    {
+      "epoch": 1.4794553853634818,
+      "grad_norm": 16.75,
+      "learning_rate": 4.043153950623388e-07,
+      "loss": 0.5095,
+      "step": 6085
+    },
+    {
+      "epoch": 1.4796985168976415,
+      "grad_norm": 25.0,
+      "learning_rate": 4.0395958435788725e-07,
+      "loss": 0.9203,
+      "step": 6086
+    },
+    {
+      "epoch": 1.4799416484318015,
+      "grad_norm": 13.125,
+      "learning_rate": 4.0360390010878714e-07,
+      "loss": 0.3822,
+      "step": 6087
+    },
+    {
+      "epoch": 1.4801847799659615,
+      "grad_norm": 19.25,
+      "learning_rate": 4.0324834236820224e-07,
+      "loss": 0.7481,
+      "step": 6088
+    },
+    {
+      "epoch": 1.4804279115001215,
+      "grad_norm": 22.375,
+      "learning_rate": 4.0289291118927655e-07,
+      "loss": 0.4939,
+      "step": 6089
+    },
+    {
+      "epoch": 1.4806710430342815,
+      "grad_norm": 26.0,
+      "learning_rate": 4.025376066251348e-07,
+      "loss": 0.739,
+      "step": 6090
+    },
+    {
+      "epoch": 1.4809141745684415,
+      "grad_norm": 24.0,
+      "learning_rate": 4.0218242872888407e-07,
+      "loss": 0.6906,
+      "step": 6091
+    },
+    {
+      "epoch": 1.4811573061026015,
+      "grad_norm": 20.375,
+      "learning_rate": 4.018273775536119e-07,
+      "loss": 0.6155,
+      "step": 6092
+    },
+    {
+      "epoch": 1.4814004376367615,
+      "grad_norm": 18.25,
+      "learning_rate": 4.0147245315238685e-07,
+      "loss": 1.0447,
+      "step": 6093
+    },
+    {
+      "epoch": 1.4816435691709215,
+      "grad_norm": 16.875,
+      "learning_rate": 4.011176555782577e-07,
+      "loss": 0.2817,
+      "step": 6094
+    },
+    {
+      "epoch": 1.4818867007050813,
+      "grad_norm": 31.875,
+      "learning_rate": 4.0076298488425604e-07,
+      "loss": 0.7011,
+      "step": 6095
+    },
+    {
+      "epoch": 1.4821298322392416,
+      "grad_norm": 29.375,
+      "learning_rate": 4.0040844112339315e-07,
+      "loss": 0.723,
+      "step": 6096
+    },
+    {
+      "epoch": 1.4823729637734013,
+      "grad_norm": 19.375,
+      "learning_rate": 4.000540243486613e-07,
+      "loss": 0.729,
+      "step": 6097
+    },
+    {
+      "epoch": 1.4826160953075613,
+      "grad_norm": 19.875,
+      "learning_rate": 3.996997346130345e-07,
+      "loss": 0.5732,
+      "step": 6098
+    },
+    {
+      "epoch": 1.4828592268417213,
+      "grad_norm": 19.375,
+      "learning_rate": 3.99345571969468e-07,
+      "loss": 0.5748,
+      "step": 6099
+    },
+    {
+      "epoch": 1.4831023583758813,
+      "grad_norm": 17.875,
+      "learning_rate": 3.9899153647089713e-07,
+      "loss": 0.916,
+      "step": 6100
+    },
+    {
+      "epoch": 1.4833454899100413,
+      "grad_norm": 20.625,
+      "learning_rate": 3.986376281702381e-07,
+      "loss": 0.5299,
+      "step": 6101
+    },
+    {
+      "epoch": 1.4835886214442013,
+      "grad_norm": 20.125,
+      "learning_rate": 3.9828384712038956e-07,
+      "loss": 0.5411,
+      "step": 6102
+    },
+    {
+      "epoch": 1.4838317529783613,
+      "grad_norm": 18.75,
+      "learning_rate": 3.979301933742295e-07,
+      "loss": 0.9038,
+      "step": 6103
+    },
+    {
+      "epoch": 1.4840748845125213,
+      "grad_norm": 21.25,
+      "learning_rate": 3.975766669846176e-07,
+      "loss": 0.7698,
+      "step": 6104
+    },
+    {
+      "epoch": 1.4843180160466813,
+      "grad_norm": 32.5,
+      "learning_rate": 3.97223268004395e-07,
+      "loss": 0.8023,
+      "step": 6105
+    },
+    {
+      "epoch": 1.4845611475808411,
+      "grad_norm": 20.5,
+      "learning_rate": 3.968699964863826e-07,
+      "loss": 0.7003,
+      "step": 6106
+    },
+    {
+      "epoch": 1.4848042791150013,
+      "grad_norm": 18.75,
+      "learning_rate": 3.965168524833837e-07,
+      "loss": 0.8199,
+      "step": 6107
+    },
+    {
+      "epoch": 1.4850474106491611,
+      "grad_norm": 24.625,
+      "learning_rate": 3.9616383604818094e-07,
+      "loss": 1.1406,
+      "step": 6108
+    },
+    {
+      "epoch": 1.4852905421833211,
+      "grad_norm": 21.875,
+      "learning_rate": 3.958109472335396e-07,
+      "loss": 0.6098,
+      "step": 6109
+    },
+    {
+      "epoch": 1.4855336737174811,
+      "grad_norm": 23.25,
+      "learning_rate": 3.9545818609220436e-07,
+      "loss": 0.8669,
+      "step": 6110
+    },
+    {
+      "epoch": 1.4857768052516411,
+      "grad_norm": 21.125,
+      "learning_rate": 3.951055526769014e-07,
+      "loss": 0.525,
+      "step": 6111
+    },
+    {
+      "epoch": 1.4860199367858011,
+      "grad_norm": 23.625,
+      "learning_rate": 3.947530470403385e-07,
+      "loss": 0.6387,
+      "step": 6112
+    },
+    {
+      "epoch": 1.4862630683199611,
+      "grad_norm": 18.25,
+      "learning_rate": 3.9440066923520345e-07,
+      "loss": 0.693,
+      "step": 6113
+    },
+    {
+      "epoch": 1.4865061998541211,
+      "grad_norm": 21.875,
+      "learning_rate": 3.9404841931416457e-07,
+      "loss": 0.5969,
+      "step": 6114
+    },
+    {
+      "epoch": 1.4867493313882811,
+      "grad_norm": 24.25,
+      "learning_rate": 3.936962973298723e-07,
+      "loss": 0.9472,
+      "step": 6115
+    },
+    {
+      "epoch": 1.4869924629224411,
+      "grad_norm": 16.25,
+      "learning_rate": 3.9334430333495764e-07,
+      "loss": 0.8326,
+      "step": 6116
+    },
+    {
+      "epoch": 1.487235594456601,
+      "grad_norm": 19.375,
+      "learning_rate": 3.929924373820318e-07,
+      "loss": 0.7791,
+      "step": 6117
+    },
+    {
+      "epoch": 1.487478725990761,
+      "grad_norm": 24.0,
+      "learning_rate": 3.9264069952368674e-07,
+      "loss": 0.6789,
+      "step": 6118
+    },
+    {
+      "epoch": 1.487721857524921,
+      "grad_norm": 19.0,
+      "learning_rate": 3.9228908981249677e-07,
+      "loss": 0.4578,
+      "step": 6119
+    },
+    {
+      "epoch": 1.487964989059081,
+      "grad_norm": 17.125,
+      "learning_rate": 3.919376083010154e-07,
+      "loss": 0.4755,
+      "step": 6120
+    },
+    {
+      "epoch": 1.488208120593241,
+      "grad_norm": 19.75,
+      "learning_rate": 3.9158625504177746e-07,
+      "loss": 0.6604,
+      "step": 6121
+    },
+    {
+      "epoch": 1.488451252127401,
+      "grad_norm": 19.375,
+      "learning_rate": 3.9123503008729884e-07,
+      "loss": 0.5169,
+      "step": 6122
+    },
+    {
+      "epoch": 1.488694383661561,
+      "grad_norm": 22.75,
+      "learning_rate": 3.90883933490077e-07,
+      "loss": 0.7933,
+      "step": 6123
+    },
+    {
+      "epoch": 1.488937515195721,
+      "grad_norm": 22.5,
+      "learning_rate": 3.9053296530258854e-07,
+      "loss": 0.7724,
+      "step": 6124
+    },
+    {
+      "epoch": 1.489180646729881,
+      "grad_norm": 17.0,
+      "learning_rate": 3.9018212557729154e-07,
+      "loss": 0.574,
+      "step": 6125
+    },
+    {
+      "epoch": 1.4894237782640407,
+      "grad_norm": 23.125,
+      "learning_rate": 3.8983141436662597e-07,
+      "loss": 0.9739,
+      "step": 6126
+    },
+    {
+      "epoch": 1.489666909798201,
+      "grad_norm": 18.875,
+      "learning_rate": 3.8948083172301103e-07,
+      "loss": 0.6615,
+      "step": 6127
+    },
+    {
+      "epoch": 1.4899100413323607,
+      "grad_norm": 23.125,
+      "learning_rate": 3.891303776988471e-07,
+      "loss": 0.7575,
+      "step": 6128
+    },
+    {
+      "epoch": 1.4901531728665207,
+      "grad_norm": 30.0,
+      "learning_rate": 3.88780052346516e-07,
+      "loss": 0.7343,
+      "step": 6129
+    },
+    {
+      "epoch": 1.4903963044006807,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.8842985571838023e-07,
+      "loss": 0.3929,
+      "step": 6130
+    },
+    {
+      "epoch": 1.4906394359348407,
+      "grad_norm": 16.25,
+      "learning_rate": 3.880797878667823e-07,
+      "loss": 0.3027,
+      "step": 6131
+    },
+    {
+      "epoch": 1.4908825674690007,
+      "grad_norm": 22.25,
+      "learning_rate": 3.8772984884404564e-07,
+      "loss": 0.8387,
+      "step": 6132
+    },
+    {
+      "epoch": 1.4911256990031607,
+      "grad_norm": 18.25,
+      "learning_rate": 3.873800387024754e-07,
+      "loss": 0.5991,
+      "step": 6133
+    },
+    {
+      "epoch": 1.4913688305373207,
+      "grad_norm": 26.125,
+      "learning_rate": 3.8703035749435624e-07,
+      "loss": 0.4715,
+      "step": 6134
+    },
+    {
+      "epoch": 1.4916119620714807,
+      "grad_norm": 17.375,
+      "learning_rate": 3.866808052719538e-07,
+      "loss": 0.748,
+      "step": 6135
+    },
+    {
+      "epoch": 1.4918550936056407,
+      "grad_norm": 17.125,
+      "learning_rate": 3.863313820875154e-07,
+      "loss": 0.5773,
+      "step": 6136
+    },
+    {
+      "epoch": 1.4920982251398005,
+      "grad_norm": 18.25,
+      "learning_rate": 3.859820879932677e-07,
+      "loss": 0.6522,
+      "step": 6137
+    },
+    {
+      "epoch": 1.4923413566739607,
+      "grad_norm": 19.25,
+      "learning_rate": 3.856329230414195e-07,
+      "loss": 0.5437,
+      "step": 6138
+    },
+    {
+      "epoch": 1.4925844882081205,
+      "grad_norm": 20.125,
+      "learning_rate": 3.852838872841586e-07,
+      "loss": 0.6105,
+      "step": 6139
+    },
+    {
+      "epoch": 1.4928276197422805,
+      "grad_norm": 18.25,
+      "learning_rate": 3.8493498077365534e-07,
+      "loss": 0.5001,
+      "step": 6140
+    },
+    {
+      "epoch": 1.4930707512764405,
+      "grad_norm": 18.75,
+      "learning_rate": 3.8458620356205927e-07,
+      "loss": 0.8182,
+      "step": 6141
+    },
+    {
+      "epoch": 1.4933138828106005,
+      "grad_norm": 20.375,
+      "learning_rate": 3.84237555701501e-07,
+      "loss": 0.8294,
+      "step": 6142
+    },
+    {
+      "epoch": 1.4935570143447605,
+      "grad_norm": 22.125,
+      "learning_rate": 3.8388903724409275e-07,
+      "loss": 0.552,
+      "step": 6143
+    },
+    {
+      "epoch": 1.4938001458789205,
+      "grad_norm": 20.125,
+      "learning_rate": 3.83540648241926e-07,
+      "loss": 0.7927,
+      "step": 6144
+    },
+    {
+      "epoch": 1.4940432774130805,
+      "grad_norm": 37.0,
+      "learning_rate": 3.831923887470733e-07,
+      "loss": 1.1963,
+      "step": 6145
+    },
+    {
+      "epoch": 1.4942864089472405,
+      "grad_norm": 18.75,
+      "learning_rate": 3.828442588115884e-07,
+      "loss": 0.5773,
+      "step": 6146
+    },
+    {
+      "epoch": 1.4945295404814005,
+      "grad_norm": 21.625,
+      "learning_rate": 3.824962584875057e-07,
+      "loss": 0.9266,
+      "step": 6147
+    },
+    {
+      "epoch": 1.4947726720155603,
+      "grad_norm": 16.625,
+      "learning_rate": 3.821483878268395e-07,
+      "loss": 0.4933,
+      "step": 6148
+    },
+    {
+      "epoch": 1.4950158035497205,
+      "grad_norm": 24.625,
+      "learning_rate": 3.8180064688158464e-07,
+      "loss": 0.9639,
+      "step": 6149
+    },
+    {
+      "epoch": 1.4952589350838803,
+      "grad_norm": 22.125,
+      "learning_rate": 3.8145303570371774e-07,
+      "loss": 0.7591,
+      "step": 6150
+    },
+    {
+      "epoch": 1.4955020666180403,
+      "grad_norm": 25.875,
+      "learning_rate": 3.811055543451951e-07,
+      "loss": 0.6283,
+      "step": 6151
+    },
+    {
+      "epoch": 1.4957451981522003,
+      "grad_norm": 16.25,
+      "learning_rate": 3.807582028579532e-07,
+      "loss": 0.5153,
+      "step": 6152
+    },
+    {
+      "epoch": 1.4959883296863603,
+      "grad_norm": 32.5,
+      "learning_rate": 3.8041098129391023e-07,
+      "loss": 1.0004,
+      "step": 6153
+    },
+    {
+      "epoch": 1.4962314612205203,
+      "grad_norm": 34.0,
+      "learning_rate": 3.8006388970496493e-07,
+      "loss": 0.7645,
+      "step": 6154
+    },
+    {
+      "epoch": 1.4964745927546803,
+      "grad_norm": 23.625,
+      "learning_rate": 3.797169281429955e-07,
+      "loss": 0.9273,
+      "step": 6155
+    },
+    {
+      "epoch": 1.4967177242888403,
+      "grad_norm": 16.875,
+      "learning_rate": 3.793700966598611e-07,
+      "loss": 0.6612,
+      "step": 6156
+    },
+    {
+      "epoch": 1.4969608558230003,
+      "grad_norm": 25.75,
+      "learning_rate": 3.7902339530740255e-07,
+      "loss": 0.8358,
+      "step": 6157
+    },
+    {
+      "epoch": 1.4972039873571603,
+      "grad_norm": 17.25,
+      "learning_rate": 3.7867682413743974e-07,
+      "loss": 0.9172,
+      "step": 6158
+    },
+    {
+      "epoch": 1.49744711889132,
+      "grad_norm": 20.625,
+      "learning_rate": 3.7833038320177345e-07,
+      "loss": 0.3885,
+      "step": 6159
+    },
+    {
+      "epoch": 1.4976902504254803,
+      "grad_norm": 23.25,
+      "learning_rate": 3.7798407255218565e-07,
+      "loss": 0.8555,
+      "step": 6160
+    },
+    {
+      "epoch": 1.49793338195964,
+      "grad_norm": 20.375,
+      "learning_rate": 3.776378922404389e-07,
+      "loss": 0.6206,
+      "step": 6161
+    },
+    {
+      "epoch": 1.4981765134938,
+      "grad_norm": 19.375,
+      "learning_rate": 3.772918423182753e-07,
+      "loss": 1.0621,
+      "step": 6162
+    },
+    {
+      "epoch": 1.49841964502796,
+      "grad_norm": 18.125,
+      "learning_rate": 3.7694592283741767e-07,
+      "loss": 0.6415,
+      "step": 6163
+    },
+    {
+      "epoch": 1.49866277656212,
+      "grad_norm": 19.75,
+      "learning_rate": 3.766001338495705e-07,
+      "loss": 0.7096,
+      "step": 6164
+    },
+    {
+      "epoch": 1.49890590809628,
+      "grad_norm": 23.625,
+      "learning_rate": 3.762544754064175e-07,
+      "loss": 0.6475,
+      "step": 6165
+    },
+    {
+      "epoch": 1.49914903963044,
+      "grad_norm": 18.0,
+      "learning_rate": 3.759089475596227e-07,
+      "loss": 0.7357,
+      "step": 6166
+    },
+    {
+      "epoch": 1.4993921711646,
+      "grad_norm": 16.375,
+      "learning_rate": 3.7556355036083225e-07,
+      "loss": 0.4537,
+      "step": 6167
+    },
+    {
+      "epoch": 1.49963530269876,
+      "grad_norm": 16.625,
+      "learning_rate": 3.752182838616708e-07,
+      "loss": 0.508,
+      "step": 6168
+    },
+    {
+      "epoch": 1.49987843423292,
+      "grad_norm": 22.375,
+      "learning_rate": 3.7487314811374494e-07,
+      "loss": 0.9527,
+      "step": 6169
+    },
+    {
+      "epoch": 1.50012156576708,
+      "grad_norm": 19.375,
+      "learning_rate": 3.7452814316864134e-07,
+      "loss": 0.8152,
+      "step": 6170
+    },
+    {
+      "epoch": 1.5003646973012401,
+      "grad_norm": 17.625,
+      "learning_rate": 3.7418326907792664e-07,
+      "loss": 0.5876,
+      "step": 6171
+    },
+    {
+      "epoch": 1.5006078288354,
+      "grad_norm": 15.75,
+      "learning_rate": 3.738385258931483e-07,
+      "loss": 0.4343,
+      "step": 6172
+    },
+    {
+      "epoch": 1.50085096036956,
+      "grad_norm": 25.75,
+      "learning_rate": 3.7349391366583375e-07,
+      "loss": 0.6182,
+      "step": 6173
+    },
+    {
+      "epoch": 1.50109409190372,
+      "grad_norm": 15.25,
+      "learning_rate": 3.73149432447492e-07,
+      "loss": 0.2298,
+      "step": 6174
+    },
+    {
+      "epoch": 1.50133722343788,
+      "grad_norm": 17.0,
+      "learning_rate": 3.728050822896112e-07,
+      "loss": 0.3318,
+      "step": 6175
+    },
+    {
+      "epoch": 1.50158035497204,
+      "grad_norm": 16.875,
+      "learning_rate": 3.7246086324365977e-07,
+      "loss": 0.4708,
+      "step": 6176
+    },
+    {
+      "epoch": 1.5018234865062,
+      "grad_norm": 20.5,
+      "learning_rate": 3.721167753610888e-07,
+      "loss": 0.3161,
+      "step": 6177
+    },
+    {
+      "epoch": 1.50206661804036,
+      "grad_norm": 20.875,
+      "learning_rate": 3.717728186933273e-07,
+      "loss": 0.762,
+      "step": 6178
+    },
+    {
+      "epoch": 1.5023097495745197,
+      "grad_norm": 21.875,
+      "learning_rate": 3.714289932917856e-07,
+      "loss": 1.0256,
+      "step": 6179
+    },
+    {
+      "epoch": 1.50255288110868,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.710852992078538e-07,
+      "loss": 0.3279,
+      "step": 6180
+    },
+    {
+      "epoch": 1.5027960126428397,
+      "grad_norm": 21.75,
+      "learning_rate": 3.707417364929039e-07,
+      "loss": 0.7004,
+      "step": 6181
+    },
+    {
+      "epoch": 1.503039144177,
+      "grad_norm": 21.625,
+      "learning_rate": 3.703983051982868e-07,
+      "loss": 0.5921,
+      "step": 6182
+    },
+    {
+      "epoch": 1.5032822757111597,
+      "grad_norm": 19.875,
+      "learning_rate": 3.700550053753334e-07,
+      "loss": 0.429,
+      "step": 6183
+    },
+    {
+      "epoch": 1.5035254072453197,
+      "grad_norm": 23.625,
+      "learning_rate": 3.6971183707535733e-07,
+      "loss": 0.8703,
+      "step": 6184
+    },
+    {
+      "epoch": 1.5037685387794797,
+      "grad_norm": 19.125,
+      "learning_rate": 3.693688003496504e-07,
+      "loss": 0.7276,
+      "step": 6185
+    },
+    {
+      "epoch": 1.5040116703136397,
+      "grad_norm": 24.0,
+      "learning_rate": 3.690258952494849e-07,
+      "loss": 0.5193,
+      "step": 6186
+    },
+    {
+      "epoch": 1.5042548018477997,
+      "grad_norm": 20.75,
+      "learning_rate": 3.686831218261147e-07,
+      "loss": 0.562,
+      "step": 6187
+    },
+    {
+      "epoch": 1.5044979333819597,
+      "grad_norm": 16.5,
+      "learning_rate": 3.6834048013077275e-07,
+      "loss": 0.3605,
+      "step": 6188
+    },
+    {
+      "epoch": 1.5047410649161197,
+      "grad_norm": 19.5,
+      "learning_rate": 3.6799797021467283e-07,
+      "loss": 0.6409,
+      "step": 6189
+    },
+    {
+      "epoch": 1.5049841964502795,
+      "grad_norm": 21.125,
+      "learning_rate": 3.6765559212900867e-07,
+      "loss": 0.6381,
+      "step": 6190
+    },
+    {
+      "epoch": 1.5052273279844397,
+      "grad_norm": 20.875,
+      "learning_rate": 3.6731334592495477e-07,
+      "loss": 0.7849,
+      "step": 6191
+    },
+    {
+      "epoch": 1.5054704595185995,
+      "grad_norm": 18.875,
+      "learning_rate": 3.6697123165366637e-07,
+      "loss": 0.4924,
+      "step": 6192
+    },
+    {
+      "epoch": 1.5057135910527597,
+      "grad_norm": 19.75,
+      "learning_rate": 3.6662924936627735e-07,
+      "loss": 0.5789,
+      "step": 6193
+    },
+    {
+      "epoch": 1.5059567225869195,
+      "grad_norm": 18.0,
+      "learning_rate": 3.6628739911390377e-07,
+      "loss": 0.5383,
+      "step": 6194
+    },
+    {
+      "epoch": 1.5061998541210795,
+      "grad_norm": 20.375,
+      "learning_rate": 3.659456809476407e-07,
+      "loss": 0.6224,
+      "step": 6195
+    },
+    {
+      "epoch": 1.5064429856552395,
+      "grad_norm": 15.9375,
+      "learning_rate": 3.656040949185637e-07,
+      "loss": 0.4936,
+      "step": 6196
+    },
+    {
+      "epoch": 1.5066861171893995,
+      "grad_norm": 17.125,
+      "learning_rate": 3.652626410777282e-07,
+      "loss": 0.6152,
+      "step": 6197
+    },
+    {
+      "epoch": 1.5069292487235595,
+      "grad_norm": 24.0,
+      "learning_rate": 3.649213194761715e-07,
+      "loss": 0.9093,
+      "step": 6198
+    },
+    {
+      "epoch": 1.5071723802577193,
+      "grad_norm": 21.375,
+      "learning_rate": 3.6458013016490905e-07,
+      "loss": 0.5359,
+      "step": 6199
+    },
+    {
+      "epoch": 1.5074155117918795,
+      "grad_norm": 27.875,
+      "learning_rate": 3.6423907319493787e-07,
+      "loss": 0.7859,
+      "step": 6200
+    },
+    {
+      "epoch": 1.5076586433260393,
+      "grad_norm": 21.625,
+      "learning_rate": 3.638981486172352e-07,
+      "loss": 0.7023,
+      "step": 6201
+    },
+    {
+      "epoch": 1.5079017748601995,
+      "grad_norm": 24.25,
+      "learning_rate": 3.6355735648275785e-07,
+      "loss": 0.8522,
+      "step": 6202
+    },
+    {
+      "epoch": 1.5081449063943593,
+      "grad_norm": 14.5,
+      "learning_rate": 3.6321669684244274e-07,
+      "loss": 0.2743,
+      "step": 6203
+    },
+    {
+      "epoch": 1.5083880379285193,
+      "grad_norm": 17.0,
+      "learning_rate": 3.628761697472073e-07,
+      "loss": 0.4714,
+      "step": 6204
+    },
+    {
+      "epoch": 1.5086311694626793,
+      "grad_norm": 16.5,
+      "learning_rate": 3.625357752479498e-07,
+      "loss": 0.4457,
+      "step": 6205
+    },
+    {
+      "epoch": 1.5088743009968393,
+      "grad_norm": 18.5,
+      "learning_rate": 3.621955133955478e-07,
+      "loss": 0.8612,
+      "step": 6206
+    },
+    {
+      "epoch": 1.5091174325309993,
+      "grad_norm": 24.0,
+      "learning_rate": 3.618553842408584e-07,
+      "loss": 0.6889,
+      "step": 6207
+    },
+    {
+      "epoch": 1.5093605640651593,
+      "grad_norm": 19.375,
+      "learning_rate": 3.6151538783472134e-07,
+      "loss": 0.5208,
+      "step": 6208
+    },
+    {
+      "epoch": 1.5096036955993193,
+      "grad_norm": 16.125,
+      "learning_rate": 3.611755242279544e-07,
+      "loss": 0.6766,
+      "step": 6209
+    },
+    {
+      "epoch": 1.509846827133479,
+      "grad_norm": 19.0,
+      "learning_rate": 3.6083579347135534e-07,
+      "loss": 0.8017,
+      "step": 6210
+    },
+    {
+      "epoch": 1.5100899586676393,
+      "grad_norm": 20.875,
+      "learning_rate": 3.604961956157038e-07,
+      "loss": 0.7553,
+      "step": 6211
+    },
+    {
+      "epoch": 1.510333090201799,
+      "grad_norm": 24.5,
+      "learning_rate": 3.6015673071175804e-07,
+      "loss": 0.9282,
+      "step": 6212
+    },
+    {
+      "epoch": 1.5105762217359593,
+      "grad_norm": 19.75,
+      "learning_rate": 3.5981739881025715e-07,
+      "loss": 0.518,
+      "step": 6213
+    },
+    {
+      "epoch": 1.510819353270119,
+      "grad_norm": 19.25,
+      "learning_rate": 3.59478199961919e-07,
+      "loss": 0.3966,
+      "step": 6214
+    },
+    {
+      "epoch": 1.511062484804279,
+      "grad_norm": 14.625,
+      "learning_rate": 3.5913913421744453e-07,
+      "loss": 0.649,
+      "step": 6215
+    },
+    {
+      "epoch": 1.511305616338439,
+      "grad_norm": 18.0,
+      "learning_rate": 3.588002016275123e-07,
+      "loss": 0.5706,
+      "step": 6216
+    },
+    {
+      "epoch": 1.511548747872599,
+      "grad_norm": 17.25,
+      "learning_rate": 3.58461402242781e-07,
+      "loss": 0.505,
+      "step": 6217
+    },
+    {
+      "epoch": 1.511791879406759,
+      "grad_norm": 30.75,
+      "learning_rate": 3.581227361138909e-07,
+      "loss": 0.8332,
+      "step": 6218
+    },
+    {
+      "epoch": 1.512035010940919,
+      "grad_norm": 27.25,
+      "learning_rate": 3.577842032914612e-07,
+      "loss": 0.8551,
+      "step": 6219
+    },
+    {
+      "epoch": 1.512278142475079,
+      "grad_norm": 25.875,
+      "learning_rate": 3.574458038260914e-07,
+      "loss": 0.9446,
+      "step": 6220
+    },
+    {
+      "epoch": 1.5125212740092389,
+      "grad_norm": 23.25,
+      "learning_rate": 3.571075377683608e-07,
+      "loss": 1.0182,
+      "step": 6221
+    },
+    {
+      "epoch": 1.512764405543399,
+      "grad_norm": 17.125,
+      "learning_rate": 3.5676940516882933e-07,
+      "loss": 0.6912,
+      "step": 6222
+    },
+    {
+      "epoch": 1.5130075370775589,
+      "grad_norm": 17.25,
+      "learning_rate": 3.5643140607803745e-07,
+      "loss": 0.5258,
+      "step": 6223
+    },
+    {
+      "epoch": 1.513250668611719,
+      "grad_norm": 14.4375,
+      "learning_rate": 3.5609354054650395e-07,
+      "loss": 0.3123,
+      "step": 6224
+    },
+    {
+      "epoch": 1.5134938001458789,
+      "grad_norm": 25.875,
+      "learning_rate": 3.557558086247295e-07,
+      "loss": 1.3475,
+      "step": 6225
+    },
+    {
+      "epoch": 1.5137369316800389,
+      "grad_norm": 29.5,
+      "learning_rate": 3.5541821036319355e-07,
+      "loss": 0.3728,
+      "step": 6226
+    },
+    {
+      "epoch": 1.5139800632141989,
+      "grad_norm": 26.0,
+      "learning_rate": 3.550807458123556e-07,
+      "loss": 0.6712,
+      "step": 6227
+    },
+    {
+      "epoch": 1.5142231947483589,
+      "grad_norm": 23.0,
+      "learning_rate": 3.547434150226564e-07,
+      "loss": 0.7177,
+      "step": 6228
+    },
+    {
+      "epoch": 1.5144663262825189,
+      "grad_norm": 19.5,
+      "learning_rate": 3.5440621804451555e-07,
+      "loss": 0.8626,
+      "step": 6229
+    },
+    {
+      "epoch": 1.5147094578166789,
+      "grad_norm": 22.5,
+      "learning_rate": 3.5406915492833233e-07,
+      "loss": 0.832,
+      "step": 6230
+    },
+    {
+      "epoch": 1.5149525893508389,
+      "grad_norm": 20.125,
+      "learning_rate": 3.5373222572448724e-07,
+      "loss": 0.893,
+      "step": 6231
+    },
+    {
+      "epoch": 1.5151957208849987,
+      "grad_norm": 19.75,
+      "learning_rate": 3.5339543048334047e-07,
+      "loss": 0.42,
+      "step": 6232
+    },
+    {
+      "epoch": 1.5154388524191589,
+      "grad_norm": 32.75,
+      "learning_rate": 3.5305876925523146e-07,
+      "loss": 0.7095,
+      "step": 6233
+    },
+    {
+      "epoch": 1.5156819839533187,
+      "grad_norm": 21.125,
+      "learning_rate": 3.527222420904798e-07,
+      "loss": 0.696,
+      "step": 6234
+    },
+    {
+      "epoch": 1.5159251154874789,
+      "grad_norm": 23.625,
+      "learning_rate": 3.5238584903938584e-07,
+      "loss": 0.8977,
+      "step": 6235
+    },
+    {
+      "epoch": 1.5161682470216387,
+      "grad_norm": 24.0,
+      "learning_rate": 3.5204959015222916e-07,
+      "loss": 0.9621,
+      "step": 6236
+    },
+    {
+      "epoch": 1.5164113785557987,
+      "grad_norm": 18.875,
+      "learning_rate": 3.5171346547926907e-07,
+      "loss": 0.4414,
+      "step": 6237
+    },
+    {
+      "epoch": 1.5166545100899587,
+      "grad_norm": 19.375,
+      "learning_rate": 3.5137747507074537e-07,
+      "loss": 0.4407,
+      "step": 6238
+    },
+    {
+      "epoch": 1.5168976416241187,
+      "grad_norm": 21.625,
+      "learning_rate": 3.510416189768782e-07,
+      "loss": 0.8799,
+      "step": 6239
+    },
+    {
+      "epoch": 1.5171407731582787,
+      "grad_norm": 22.875,
+      "learning_rate": 3.5070589724786666e-07,
+      "loss": 1.233,
+      "step": 6240
+    },
+    {
+      "epoch": 1.5173839046924384,
+      "grad_norm": 32.5,
+      "learning_rate": 3.5037030993388965e-07,
+      "loss": 0.9792,
+      "step": 6241
+    },
+    {
+      "epoch": 1.5176270362265987,
+      "grad_norm": 20.5,
+      "learning_rate": 3.500348570851074e-07,
+      "loss": 0.6742,
+      "step": 6242
+    },
+    {
+      "epoch": 1.5178701677607584,
+      "grad_norm": 23.375,
+      "learning_rate": 3.496995387516587e-07,
+      "loss": 0.7058,
+      "step": 6243
+    },
+    {
+      "epoch": 1.5181132992949187,
+      "grad_norm": 17.75,
+      "learning_rate": 3.493643549836623e-07,
+      "loss": 0.5027,
+      "step": 6244
+    },
+    {
+      "epoch": 1.5183564308290785,
+      "grad_norm": 20.0,
+      "learning_rate": 3.490293058312175e-07,
+      "loss": 0.8238,
+      "step": 6245
+    },
+    {
+      "epoch": 1.5185995623632387,
+      "grad_norm": 19.875,
+      "learning_rate": 3.486943913444037e-07,
+      "loss": 0.4994,
+      "step": 6246
+    },
+    {
+      "epoch": 1.5188426938973985,
+      "grad_norm": 20.0,
+      "learning_rate": 3.4835961157327915e-07,
+      "loss": 0.8607,
+      "step": 6247
+    },
+    {
+      "epoch": 1.5190858254315585,
+      "grad_norm": 19.125,
+      "learning_rate": 3.480249665678821e-07,
+      "loss": 0.7661,
+      "step": 6248
+    },
+    {
+      "epoch": 1.5193289569657185,
+      "grad_norm": 25.375,
+      "learning_rate": 3.4769045637823184e-07,
+      "loss": 0.8371,
+      "step": 6249
+    },
+    {
+      "epoch": 1.5195720884998785,
+      "grad_norm": 16.25,
+      "learning_rate": 3.473560810543264e-07,
+      "loss": 0.5626,
+      "step": 6250
+    },
+    {
+      "epoch": 1.5198152200340385,
+      "grad_norm": 20.0,
+      "learning_rate": 3.470218406461434e-07,
+      "loss": 0.8862,
+      "step": 6251
+    },
+    {
+      "epoch": 1.5200583515681982,
+      "grad_norm": 22.75,
+      "learning_rate": 3.4668773520364173e-07,
+      "loss": 0.5772,
+      "step": 6252
+    },
+    {
+      "epoch": 1.5203014831023585,
+      "grad_norm": 20.5,
+      "learning_rate": 3.463537647767583e-07,
+      "loss": 0.9129,
+      "step": 6253
+    },
+    {
+      "epoch": 1.5205446146365182,
+      "grad_norm": 16.875,
+      "learning_rate": 3.4601992941541167e-07,
+      "loss": 0.4075,
+      "step": 6254
+    },
+    {
+      "epoch": 1.5207877461706785,
+      "grad_norm": 20.0,
+      "learning_rate": 3.4568622916949844e-07,
+      "loss": 1.0263,
+      "step": 6255
+    },
+    {
+      "epoch": 1.5210308777048382,
+      "grad_norm": 17.25,
+      "learning_rate": 3.453526640888967e-07,
+      "loss": 0.6099,
+      "step": 6256
+    },
+    {
+      "epoch": 1.5212740092389982,
+      "grad_norm": 16.25,
+      "learning_rate": 3.4501923422346304e-07,
+      "loss": 0.635,
+      "step": 6257
+    },
+    {
+      "epoch": 1.5215171407731582,
+      "grad_norm": 22.5,
+      "learning_rate": 3.446859396230341e-07,
+      "loss": 0.6274,
+      "step": 6258
+    },
+    {
+      "epoch": 1.5217602723073183,
+      "grad_norm": 16.375,
+      "learning_rate": 3.443527803374272e-07,
+      "loss": 0.5064,
+      "step": 6259
+    },
+    {
+      "epoch": 1.5220034038414783,
+      "grad_norm": 20.0,
+      "learning_rate": 3.4401975641643824e-07,
+      "loss": 0.8685,
+      "step": 6260
+    },
+    {
+      "epoch": 1.5222465353756383,
+      "grad_norm": 14.5625,
+      "learning_rate": 3.436868679098432e-07,
+      "loss": 0.3722,
+      "step": 6261
+    },
+    {
+      "epoch": 1.5224896669097983,
+      "grad_norm": 16.75,
+      "learning_rate": 3.433541148673983e-07,
+      "loss": 0.5688,
+      "step": 6262
+    },
+    {
+      "epoch": 1.522732798443958,
+      "grad_norm": 21.375,
+      "learning_rate": 3.430214973388396e-07,
+      "loss": 0.5267,
+      "step": 6263
+    },
+    {
+      "epoch": 1.5229759299781183,
+      "grad_norm": 20.5,
+      "learning_rate": 3.4268901537388223e-07,
+      "loss": 0.8173,
+      "step": 6264
+    },
+    {
+      "epoch": 1.523219061512278,
+      "grad_norm": 23.375,
+      "learning_rate": 3.4235666902222105e-07,
+      "loss": 0.5716,
+      "step": 6265
+    },
+    {
+      "epoch": 1.5234621930464383,
+      "grad_norm": 19.25,
+      "learning_rate": 3.4202445833353136e-07,
+      "loss": 0.6776,
+      "step": 6266
+    },
+    {
+      "epoch": 1.523705324580598,
+      "grad_norm": 23.375,
+      "learning_rate": 3.4169238335746786e-07,
+      "loss": 0.7741,
+      "step": 6267
+    },
+    {
+      "epoch": 1.523948456114758,
+      "grad_norm": 32.5,
+      "learning_rate": 3.4136044414366426e-07,
+      "loss": 1.1893,
+      "step": 6268
+    },
+    {
+      "epoch": 1.524191587648918,
+      "grad_norm": 23.875,
+      "learning_rate": 3.4102864074173486e-07,
+      "loss": 1.0365,
+      "step": 6269
+    },
+    {
+      "epoch": 1.524434719183078,
+      "grad_norm": 16.25,
+      "learning_rate": 3.406969732012741e-07,
+      "loss": 0.3805,
+      "step": 6270
+    },
+    {
+      "epoch": 1.524677850717238,
+      "grad_norm": 26.125,
+      "learning_rate": 3.403654415718549e-07,
+      "loss": 1.1491,
+      "step": 6271
+    },
+    {
+      "epoch": 1.524920982251398,
+      "grad_norm": 25.25,
+      "learning_rate": 3.4003404590303005e-07,
+      "loss": 1.0536,
+      "step": 6272
+    },
+    {
+      "epoch": 1.525164113785558,
+      "grad_norm": 17.125,
+      "learning_rate": 3.3970278624433306e-07,
+      "loss": 0.5627,
+      "step": 6273
+    },
+    {
+      "epoch": 1.5254072453197178,
+      "grad_norm": 21.0,
+      "learning_rate": 3.393716626452759e-07,
+      "loss": 0.4166,
+      "step": 6274
+    },
+    {
+      "epoch": 1.525650376853878,
+      "grad_norm": 20.25,
+      "learning_rate": 3.3904067515535056e-07,
+      "loss": 0.6059,
+      "step": 6275
+    },
+    {
+      "epoch": 1.5258935083880378,
+      "grad_norm": 21.875,
+      "learning_rate": 3.3870982382402927e-07,
+      "loss": 0.6766,
+      "step": 6276
+    },
+    {
+      "epoch": 1.526136639922198,
+      "grad_norm": 25.5,
+      "learning_rate": 3.383791087007636e-07,
+      "loss": 0.769,
+      "step": 6277
+    },
+    {
+      "epoch": 1.5263797714563578,
+      "grad_norm": 20.75,
+      "learning_rate": 3.380485298349843e-07,
+      "loss": 0.5913,
+      "step": 6278
+    },
+    {
+      "epoch": 1.5266229029905178,
+      "grad_norm": 24.5,
+      "learning_rate": 3.3771808727610186e-07,
+      "loss": 0.716,
+      "step": 6279
+    },
+    {
+      "epoch": 1.5268660345246778,
+      "grad_norm": 15.875,
+      "learning_rate": 3.3738778107350724e-07,
+      "loss": 0.5866,
+      "step": 6280
+    },
+    {
+      "epoch": 1.5271091660588378,
+      "grad_norm": 18.625,
+      "learning_rate": 3.370576112765703e-07,
+      "loss": 0.7001,
+      "step": 6281
+    },
+    {
+      "epoch": 1.5273522975929978,
+      "grad_norm": 29.5,
+      "learning_rate": 3.3672757793463974e-07,
+      "loss": 0.7475,
+      "step": 6282
+    },
+    {
+      "epoch": 1.5275954291271578,
+      "grad_norm": 24.0,
+      "learning_rate": 3.3639768109704607e-07,
+      "loss": 0.882,
+      "step": 6283
+    },
+    {
+      "epoch": 1.5278385606613178,
+      "grad_norm": 24.5,
+      "learning_rate": 3.3606792081309693e-07,
+      "loss": 0.617,
+      "step": 6284
+    },
+    {
+      "epoch": 1.5280816921954776,
+      "grad_norm": 18.25,
+      "learning_rate": 3.357382971320815e-07,
+      "loss": 0.6626,
+      "step": 6285
+    },
+    {
+      "epoch": 1.5283248237296378,
+      "grad_norm": 20.625,
+      "learning_rate": 3.354088101032671e-07,
+      "loss": 0.7945,
+      "step": 6286
+    },
+    {
+      "epoch": 1.5285679552637976,
+      "grad_norm": 23.625,
+      "learning_rate": 3.350794597759019e-07,
+      "loss": 0.9075,
+      "step": 6287
+    },
+    {
+      "epoch": 1.5288110867979579,
+      "grad_norm": 15.375,
+      "learning_rate": 3.347502461992126e-07,
+      "loss": 0.6195,
+      "step": 6288
+    },
+    {
+      "epoch": 1.5290542183321176,
+      "grad_norm": 19.75,
+      "learning_rate": 3.3442116942240575e-07,
+      "loss": 0.6145,
+      "step": 6289
+    },
+    {
+      "epoch": 1.5292973498662776,
+      "grad_norm": 16.625,
+      "learning_rate": 3.3409222949466786e-07,
+      "loss": 0.6245,
+      "step": 6290
+    },
+    {
+      "epoch": 1.5295404814004376,
+      "grad_norm": 24.875,
+      "learning_rate": 3.337634264651647e-07,
+      "loss": 0.5847,
+      "step": 6291
+    },
+    {
+      "epoch": 1.5297836129345976,
+      "grad_norm": 21.5,
+      "learning_rate": 3.3343476038304095e-07,
+      "loss": 0.8813,
+      "step": 6292
+    },
+    {
+      "epoch": 1.5300267444687576,
+      "grad_norm": 41.75,
+      "learning_rate": 3.3310623129742205e-07,
+      "loss": 0.7915,
+      "step": 6293
+    },
+    {
+      "epoch": 1.5302698760029174,
+      "grad_norm": 23.75,
+      "learning_rate": 3.327778392574124e-07,
+      "loss": 0.8187,
+      "step": 6294
+    },
+    {
+      "epoch": 1.5305130075370776,
+      "grad_norm": 30.625,
+      "learning_rate": 3.324495843120956e-07,
+      "loss": 0.7529,
+      "step": 6295
+    },
+    {
+      "epoch": 1.5307561390712374,
+      "grad_norm": 23.5,
+      "learning_rate": 3.321214665105349e-07,
+      "loss": 0.8982,
+      "step": 6296
+    },
+    {
+      "epoch": 1.5309992706053976,
+      "grad_norm": 17.5,
+      "learning_rate": 3.3179348590177353e-07,
+      "loss": 0.4632,
+      "step": 6297
+    },
+    {
+      "epoch": 1.5312424021395574,
+      "grad_norm": 24.875,
+      "learning_rate": 3.314656425348338e-07,
+      "loss": 0.6157,
+      "step": 6298
+    },
+    {
+      "epoch": 1.5314855336737176,
+      "grad_norm": 21.125,
+      "learning_rate": 3.3113793645871696e-07,
+      "loss": 0.847,
+      "step": 6299
+    },
+    {
+      "epoch": 1.5317286652078774,
+      "grad_norm": 24.0,
+      "learning_rate": 3.308103677224049e-07,
+      "loss": 0.7655,
+      "step": 6300
+    },
+    {
+      "epoch": 1.5319717967420374,
+      "grad_norm": 27.25,
+      "learning_rate": 3.3048293637485865e-07,
+      "loss": 0.9015,
+      "step": 6301
+    },
+    {
+      "epoch": 1.5322149282761974,
+      "grad_norm": 25.375,
+      "learning_rate": 3.301556424650182e-07,
+      "loss": 1.109,
+      "step": 6302
+    },
+    {
+      "epoch": 1.5324580598103574,
+      "grad_norm": 18.75,
+      "learning_rate": 3.298284860418027e-07,
+      "loss": 0.7065,
+      "step": 6303
+    },
+    {
+      "epoch": 1.5327011913445174,
+      "grad_norm": 24.125,
+      "learning_rate": 3.2950146715411246e-07,
+      "loss": 1.0324,
+      "step": 6304
+    },
+    {
+      "epoch": 1.5329443228786772,
+      "grad_norm": 27.0,
+      "learning_rate": 3.291745858508255e-07,
+      "loss": 1.2628,
+      "step": 6305
+    },
+    {
+      "epoch": 1.5331874544128374,
+      "grad_norm": 21.125,
+      "learning_rate": 3.2884784218079944e-07,
+      "loss": 0.8716,
+      "step": 6306
+    },
+    {
+      "epoch": 1.5334305859469972,
+      "grad_norm": 15.75,
+      "learning_rate": 3.2852123619287217e-07,
+      "loss": 0.7188,
+      "step": 6307
+    },
+    {
+      "epoch": 1.5336737174811574,
+      "grad_norm": 19.875,
+      "learning_rate": 3.2819476793586114e-07,
+      "loss": 0.6904,
+      "step": 6308
+    },
+    {
+      "epoch": 1.5339168490153172,
+      "grad_norm": 14.6875,
+      "learning_rate": 3.2786843745856207e-07,
+      "loss": 0.3485,
+      "step": 6309
+    },
+    {
+      "epoch": 1.5341599805494772,
+      "grad_norm": 33.5,
+      "learning_rate": 3.2754224480975055e-07,
+      "loss": 0.4056,
+      "step": 6310
+    },
+    {
+      "epoch": 1.5344031120836372,
+      "grad_norm": 22.875,
+      "learning_rate": 3.272161900381822e-07,
+      "loss": 0.8901,
+      "step": 6311
+    },
+    {
+      "epoch": 1.5346462436177972,
+      "grad_norm": 16.375,
+      "learning_rate": 3.2689027319259133e-07,
+      "loss": 0.5373,
+      "step": 6312
+    },
+    {
+      "epoch": 1.5348893751519572,
+      "grad_norm": 29.25,
+      "learning_rate": 3.265644943216913e-07,
+      "loss": 0.679,
+      "step": 6313
+    },
+    {
+      "epoch": 1.5351325066861172,
+      "grad_norm": 22.125,
+      "learning_rate": 3.262388534741763e-07,
+      "loss": 0.8547,
+      "step": 6314
+    },
+    {
+      "epoch": 1.5353756382202772,
+      "grad_norm": 21.875,
+      "learning_rate": 3.259133506987182e-07,
+      "loss": 0.7246,
+      "step": 6315
+    },
+    {
+      "epoch": 1.535618769754437,
+      "grad_norm": 24.625,
+      "learning_rate": 3.255879860439698e-07,
+      "loss": 1.0471,
+      "step": 6316
+    },
+    {
+      "epoch": 1.5358619012885972,
+      "grad_norm": 22.125,
+      "learning_rate": 3.252627595585615e-07,
+      "loss": 1.1357,
+      "step": 6317
+    },
+    {
+      "epoch": 1.536105032822757,
+      "grad_norm": 29.25,
+      "learning_rate": 3.2493767129110507e-07,
+      "loss": 0.9973,
+      "step": 6318
+    },
+    {
+      "epoch": 1.5363481643569172,
+      "grad_norm": 23.25,
+      "learning_rate": 3.2461272129019006e-07,
+      "loss": 0.6529,
+      "step": 6319
+    },
+    {
+      "epoch": 1.536591295891077,
+      "grad_norm": 26.125,
+      "learning_rate": 3.2428790960438536e-07,
+      "loss": 0.8078,
+      "step": 6320
+    },
+    {
+      "epoch": 1.536834427425237,
+      "grad_norm": 36.5,
+      "learning_rate": 3.239632362822406e-07,
+      "loss": 0.5222,
+      "step": 6321
+    },
+    {
+      "epoch": 1.537077558959397,
+      "grad_norm": 21.125,
+      "learning_rate": 3.2363870137228305e-07,
+      "loss": 0.9279,
+      "step": 6322
+    },
+    {
+      "epoch": 1.537320690493557,
+      "grad_norm": 19.75,
+      "learning_rate": 3.233143049230207e-07,
+      "loss": 0.7125,
+      "step": 6323
+    },
+    {
+      "epoch": 1.537563822027717,
+      "grad_norm": 18.75,
+      "learning_rate": 3.229900469829396e-07,
+      "loss": 0.5521,
+      "step": 6324
+    },
+    {
+      "epoch": 1.537806953561877,
+      "grad_norm": 18.0,
+      "learning_rate": 3.2266592760050635e-07,
+      "loss": 0.6292,
+      "step": 6325
+    },
+    {
+      "epoch": 1.538050085096037,
+      "grad_norm": 23.0,
+      "learning_rate": 3.223419468241658e-07,
+      "loss": 0.6819,
+      "step": 6326
+    },
+    {
+      "epoch": 1.5382932166301968,
+      "grad_norm": 20.875,
+      "learning_rate": 3.220181047023421e-07,
+      "loss": 1.0832,
+      "step": 6327
+    },
+    {
+      "epoch": 1.538536348164357,
+      "grad_norm": 27.5,
+      "learning_rate": 3.216944012834398e-07,
+      "loss": 0.6635,
+      "step": 6328
+    },
+    {
+      "epoch": 1.5387794796985168,
+      "grad_norm": 24.25,
+      "learning_rate": 3.2137083661584175e-07,
+      "loss": 0.9284,
+      "step": 6329
+    },
+    {
+      "epoch": 1.539022611232677,
+      "grad_norm": 22.25,
+      "learning_rate": 3.210474107479097e-07,
+      "loss": 0.8478,
+      "step": 6330
+    },
+    {
+      "epoch": 1.5392657427668368,
+      "grad_norm": 26.625,
+      "learning_rate": 3.2072412372798565e-07,
+      "loss": 0.6834,
+      "step": 6331
+    },
+    {
+      "epoch": 1.5395088743009968,
+      "grad_norm": 35.75,
+      "learning_rate": 3.204009756043909e-07,
+      "loss": 1.0705,
+      "step": 6332
+    },
+    {
+      "epoch": 1.5397520058351568,
+      "grad_norm": 20.125,
+      "learning_rate": 3.2007796642542513e-07,
+      "loss": 0.869,
+      "step": 6333
+    },
+    {
+      "epoch": 1.5399951373693168,
+      "grad_norm": 24.75,
+      "learning_rate": 3.197550962393671e-07,
+      "loss": 0.9519,
+      "step": 6334
+    },
+    {
+      "epoch": 1.5402382689034768,
+      "grad_norm": 19.75,
+      "learning_rate": 3.1943236509447625e-07,
+      "loss": 0.7127,
+      "step": 6335
+    },
+    {
+      "epoch": 1.5404814004376368,
+      "grad_norm": 19.5,
+      "learning_rate": 3.1910977303898985e-07,
+      "loss": 0.5472,
+      "step": 6336
+    },
+    {
+      "epoch": 1.5407245319717968,
+      "grad_norm": 22.25,
+      "learning_rate": 3.1878732012112447e-07,
+      "loss": 1.1632,
+      "step": 6337
+    },
+    {
+      "epoch": 1.5409676635059566,
+      "grad_norm": 22.875,
+      "learning_rate": 3.184650063890768e-07,
+      "loss": 0.9107,
+      "step": 6338
+    },
+    {
+      "epoch": 1.5412107950401168,
+      "grad_norm": 15.625,
+      "learning_rate": 3.181428318910225e-07,
+      "loss": 0.3992,
+      "step": 6339
+    },
+    {
+      "epoch": 1.5414539265742766,
+      "grad_norm": 19.5,
+      "learning_rate": 3.1782079667511554e-07,
+      "loss": 0.764,
+      "step": 6340
+    },
+    {
+      "epoch": 1.5416970581084368,
+      "grad_norm": 21.875,
+      "learning_rate": 3.174989007894896e-07,
+      "loss": 0.7726,
+      "step": 6341
+    },
+    {
+      "epoch": 1.5419401896425966,
+      "grad_norm": 17.5,
+      "learning_rate": 3.171771442822581e-07,
+      "loss": 0.3811,
+      "step": 6342
+    },
+    {
+      "epoch": 1.5421833211767566,
+      "grad_norm": 22.5,
+      "learning_rate": 3.168555272015128e-07,
+      "loss": 0.8022,
+      "step": 6343
+    },
+    {
+      "epoch": 1.5424264527109166,
+      "grad_norm": 23.75,
+      "learning_rate": 3.165340495953245e-07,
+      "loss": 1.0406,
+      "step": 6344
+    },
+    {
+      "epoch": 1.5426695842450766,
+      "grad_norm": 17.625,
+      "learning_rate": 3.162127115117444e-07,
+      "loss": 0.6932,
+      "step": 6345
+    },
+    {
+      "epoch": 1.5429127157792366,
+      "grad_norm": 21.25,
+      "learning_rate": 3.1589151299880133e-07,
+      "loss": 0.5317,
+      "step": 6346
+    },
+    {
+      "epoch": 1.5431558473133964,
+      "grad_norm": 13.25,
+      "learning_rate": 3.155704541045046e-07,
+      "loss": 0.2932,
+      "step": 6347
+    },
+    {
+      "epoch": 1.5433989788475566,
+      "grad_norm": 20.5,
+      "learning_rate": 3.152495348768413e-07,
+      "loss": 0.672,
+      "step": 6348
+    },
+    {
+      "epoch": 1.5436421103817164,
+      "grad_norm": 25.875,
+      "learning_rate": 3.1492875536377906e-07,
+      "loss": 0.8262,
+      "step": 6349
+    },
+    {
+      "epoch": 1.5438852419158766,
+      "grad_norm": 25.75,
+      "learning_rate": 3.1460811561326353e-07,
+      "loss": 0.6919,
+      "step": 6350
+    },
+    {
+      "epoch": 1.5441283734500364,
+      "grad_norm": 23.375,
+      "learning_rate": 3.1428761567321953e-07,
+      "loss": 0.6993,
+      "step": 6351
+    },
+    {
+      "epoch": 1.5443715049841964,
+      "grad_norm": 23.25,
+      "learning_rate": 3.1396725559155213e-07,
+      "loss": 0.7133,
+      "step": 6352
+    },
+    {
+      "epoch": 1.5446146365183564,
+      "grad_norm": 19.875,
+      "learning_rate": 3.1364703541614374e-07,
+      "loss": 0.7139,
+      "step": 6353
+    },
+    {
+      "epoch": 1.5448577680525164,
+      "grad_norm": 17.5,
+      "learning_rate": 3.133269551948577e-07,
+      "loss": 0.6425,
+      "step": 6354
+    },
+    {
+      "epoch": 1.5451008995866764,
+      "grad_norm": 23.75,
+      "learning_rate": 3.130070149755347e-07,
+      "loss": 0.8137,
+      "step": 6355
+    },
+    {
+      "epoch": 1.5453440311208364,
+      "grad_norm": 18.875,
+      "learning_rate": 3.126872148059962e-07,
+      "loss": 0.7958,
+      "step": 6356
+    },
+    {
+      "epoch": 1.5455871626549964,
+      "grad_norm": 20.25,
+      "learning_rate": 3.123675547340414e-07,
+      "loss": 0.6986,
+      "step": 6357
+    },
+    {
+      "epoch": 1.5458302941891562,
+      "grad_norm": 18.875,
+      "learning_rate": 3.1204803480744867e-07,
+      "loss": 0.7421,
+      "step": 6358
+    },
+    {
+      "epoch": 1.5460734257233164,
+      "grad_norm": 25.25,
+      "learning_rate": 3.1172865507397635e-07,
+      "loss": 0.8952,
+      "step": 6359
+    },
+    {
+      "epoch": 1.5463165572574762,
+      "grad_norm": 16.5,
+      "learning_rate": 3.1140941558136117e-07,
+      "loss": 0.4089,
+      "step": 6360
+    },
+    {
+      "epoch": 1.5465596887916364,
+      "grad_norm": 21.625,
+      "learning_rate": 3.110903163773184e-07,
+      "loss": 0.6667,
+      "step": 6361
+    },
+    {
+      "epoch": 1.5468028203257962,
+      "grad_norm": 17.5,
+      "learning_rate": 3.107713575095435e-07,
+      "loss": 0.379,
+      "step": 6362
+    },
+    {
+      "epoch": 1.5470459518599562,
+      "grad_norm": 17.375,
+      "learning_rate": 3.1045253902571044e-07,
+      "loss": 0.5482,
+      "step": 6363
+    },
+    {
+      "epoch": 1.5472890833941162,
+      "grad_norm": 20.375,
+      "learning_rate": 3.1013386097347205e-07,
+      "loss": 0.5824,
+      "step": 6364
+    },
+    {
+      "epoch": 1.5475322149282762,
+      "grad_norm": 19.25,
+      "learning_rate": 3.0981532340045985e-07,
+      "loss": 1.0763,
+      "step": 6365
+    },
+    {
+      "epoch": 1.5477753464624362,
+      "grad_norm": 18.375,
+      "learning_rate": 3.094969263542855e-07,
+      "loss": 0.537,
+      "step": 6366
+    },
+    {
+      "epoch": 1.5480184779965962,
+      "grad_norm": 41.75,
+      "learning_rate": 3.0917866988253857e-07,
+      "loss": 0.6596,
+      "step": 6367
+    },
+    {
+      "epoch": 1.5482616095307562,
+      "grad_norm": 21.25,
+      "learning_rate": 3.0886055403278756e-07,
+      "loss": 0.9174,
+      "step": 6368
+    },
+    {
+      "epoch": 1.548504741064916,
+      "grad_norm": 26.5,
+      "learning_rate": 3.085425788525807e-07,
+      "loss": 1.025,
+      "step": 6369
+    },
+    {
+      "epoch": 1.5487478725990762,
+      "grad_norm": 16.75,
+      "learning_rate": 3.082247443894455e-07,
+      "loss": 0.6325,
+      "step": 6370
+    },
+    {
+      "epoch": 1.548991004133236,
+      "grad_norm": 21.875,
+      "learning_rate": 3.0790705069088724e-07,
+      "loss": 0.6059,
+      "step": 6371
+    },
+    {
+      "epoch": 1.5492341356673962,
+      "grad_norm": 18.875,
+      "learning_rate": 3.075894978043903e-07,
+      "loss": 0.6588,
+      "step": 6372
+    },
+    {
+      "epoch": 1.549477267201556,
+      "grad_norm": 21.375,
+      "learning_rate": 3.0727208577741946e-07,
+      "loss": 0.6662,
+      "step": 6373
+    },
+    {
+      "epoch": 1.549720398735716,
+      "grad_norm": 25.125,
+      "learning_rate": 3.0695481465741665e-07,
+      "loss": 1.0003,
+      "step": 6374
+    },
+    {
+      "epoch": 1.549963530269876,
+      "grad_norm": 18.875,
+      "learning_rate": 3.0663768449180354e-07,
+      "loss": 0.6756,
+      "step": 6375
+    },
+    {
+      "epoch": 1.550206661804036,
+      "grad_norm": 17.25,
+      "learning_rate": 3.063206953279811e-07,
+      "loss": 0.3996,
+      "step": 6376
+    },
+    {
+      "epoch": 1.550449793338196,
+      "grad_norm": 17.75,
+      "learning_rate": 3.060038472133285e-07,
+      "loss": 0.4859,
+      "step": 6377
+    },
+    {
+      "epoch": 1.550692924872356,
+      "grad_norm": 18.875,
+      "learning_rate": 3.0568714019520455e-07,
+      "loss": 0.8684,
+      "step": 6378
+    },
+    {
+      "epoch": 1.550936056406516,
+      "grad_norm": 15.9375,
+      "learning_rate": 3.0537057432094603e-07,
+      "loss": 0.4608,
+      "step": 6379
+    },
+    {
+      "epoch": 1.5511791879406758,
+      "grad_norm": 20.625,
+      "learning_rate": 3.0505414963786977e-07,
+      "loss": 0.6788,
+      "step": 6380
+    },
+    {
+      "epoch": 1.551422319474836,
+      "grad_norm": 22.125,
+      "learning_rate": 3.047378661932707e-07,
+      "loss": 0.7022,
+      "step": 6381
+    },
+    {
+      "epoch": 1.5516654510089958,
+      "grad_norm": 37.25,
+      "learning_rate": 3.0442172403442244e-07,
+      "loss": 0.8,
+      "step": 6382
+    },
+    {
+      "epoch": 1.551908582543156,
+      "grad_norm": 13.375,
+      "learning_rate": 3.041057232085785e-07,
+      "loss": 0.4686,
+      "step": 6383
+    },
+    {
+      "epoch": 1.5521517140773158,
+      "grad_norm": 15.8125,
+      "learning_rate": 3.037898637629701e-07,
+      "loss": 0.4005,
+      "step": 6384
+    },
+    {
+      "epoch": 1.5523948456114758,
+      "grad_norm": 15.0,
+      "learning_rate": 3.034741457448084e-07,
+      "loss": 0.2845,
+      "step": 6385
+    },
+    {
+      "epoch": 1.5526379771456358,
+      "grad_norm": 23.125,
+      "learning_rate": 3.031585692012826e-07,
+      "loss": 0.9698,
+      "step": 6386
+    },
+    {
+      "epoch": 1.5528811086797958,
+      "grad_norm": 17.25,
+      "learning_rate": 3.0284313417956126e-07,
+      "loss": 0.6786,
+      "step": 6387
+    },
+    {
+      "epoch": 1.5531242402139558,
+      "grad_norm": 20.125,
+      "learning_rate": 3.025278407267915e-07,
+      "loss": 0.6165,
+      "step": 6388
+    },
+    {
+      "epoch": 1.5533673717481156,
+      "grad_norm": 21.375,
+      "learning_rate": 3.0221268889009903e-07,
+      "loss": 0.4938,
+      "step": 6389
+    },
+    {
+      "epoch": 1.5536105032822758,
+      "grad_norm": 24.0,
+      "learning_rate": 3.0189767871658947e-07,
+      "loss": 0.8756,
+      "step": 6390
+    },
+    {
+      "epoch": 1.5538536348164356,
+      "grad_norm": 22.75,
+      "learning_rate": 3.015828102533461e-07,
+      "loss": 0.4819,
+      "step": 6391
+    },
+    {
+      "epoch": 1.5540967663505958,
+      "grad_norm": 19.625,
+      "learning_rate": 3.012680835474312e-07,
+      "loss": 0.7592,
+      "step": 6392
+    },
+    {
+      "epoch": 1.5543398978847556,
+      "grad_norm": 28.5,
+      "learning_rate": 3.0095349864588617e-07,
+      "loss": 0.6524,
+      "step": 6393
+    },
+    {
+      "epoch": 1.5545830294189158,
+      "grad_norm": 21.875,
+      "learning_rate": 3.0063905559573186e-07,
+      "loss": 0.8198,
+      "step": 6394
+    },
+    {
+      "epoch": 1.5548261609530756,
+      "grad_norm": 15.1875,
+      "learning_rate": 3.003247544439666e-07,
+      "loss": 0.61,
+      "step": 6395
+    },
+    {
+      "epoch": 1.5550692924872356,
+      "grad_norm": 18.125,
+      "learning_rate": 3.0001059523756816e-07,
+      "loss": 0.7242,
+      "step": 6396
+    },
+    {
+      "epoch": 1.5553124240213956,
+      "grad_norm": 22.75,
+      "learning_rate": 2.9969657802349324e-07,
+      "loss": 0.8387,
+      "step": 6397
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 17.375,
+      "learning_rate": 2.993827028486772e-07,
+      "loss": 0.405,
+      "step": 6398
+    },
+    {
+      "epoch": 1.5557986870897156,
+      "grad_norm": 26.625,
+      "learning_rate": 2.990689697600335e-07,
+      "loss": 0.9947,
+      "step": 6399
+    },
+    {
+      "epoch": 1.5560418186238754,
+      "grad_norm": 15.125,
+      "learning_rate": 2.987553788044555e-07,
+      "loss": 0.5964,
+      "step": 6400
+    },
+    {
+      "epoch": 1.5562849501580356,
+      "grad_norm": 21.625,
+      "learning_rate": 2.9844193002881493e-07,
+      "loss": 0.8416,
+      "step": 6401
+    },
+    {
+      "epoch": 1.5565280816921954,
+      "grad_norm": 22.625,
+      "learning_rate": 2.9812862347996197e-07,
+      "loss": 0.8532,
+      "step": 6402
+    },
+    {
+      "epoch": 1.5567712132263556,
+      "grad_norm": 19.875,
+      "learning_rate": 2.9781545920472526e-07,
+      "loss": 0.7714,
+      "step": 6403
+    },
+    {
+      "epoch": 1.5570143447605154,
+      "grad_norm": 13.625,
+      "learning_rate": 2.9750243724991327e-07,
+      "loss": 0.453,
+      "step": 6404
+    },
+    {
+      "epoch": 1.5572574762946754,
+      "grad_norm": 25.375,
+      "learning_rate": 2.971895576623124e-07,
+      "loss": 0.8208,
+      "step": 6405
+    },
+    {
+      "epoch": 1.5575006078288354,
+      "grad_norm": 20.25,
+      "learning_rate": 2.9687682048868724e-07,
+      "loss": 0.9632,
+      "step": 6406
+    },
+    {
+      "epoch": 1.5577437393629954,
+      "grad_norm": 14.75,
+      "learning_rate": 2.9656422577578226e-07,
+      "loss": 0.3296,
+      "step": 6407
+    },
+    {
+      "epoch": 1.5579868708971554,
+      "grad_norm": 29.875,
+      "learning_rate": 2.9625177357032046e-07,
+      "loss": 1.1691,
+      "step": 6408
+    },
+    {
+      "epoch": 1.5582300024313154,
+      "grad_norm": 22.125,
+      "learning_rate": 2.95939463919003e-07,
+      "loss": 1.1166,
+      "step": 6409
+    },
+    {
+      "epoch": 1.5584731339654754,
+      "grad_norm": 19.125,
+      "learning_rate": 2.956272968685095e-07,
+      "loss": 0.8856,
+      "step": 6410
+    },
+    {
+      "epoch": 1.5587162654996352,
+      "grad_norm": 24.625,
+      "learning_rate": 2.9531527246549937e-07,
+      "loss": 1.0867,
+      "step": 6411
+    },
+    {
+      "epoch": 1.5589593970337954,
+      "grad_norm": 14.625,
+      "learning_rate": 2.950033907566098e-07,
+      "loss": 0.5691,
+      "step": 6412
+    },
+    {
+      "epoch": 1.5592025285679552,
+      "grad_norm": 21.5,
+      "learning_rate": 2.9469165178845655e-07,
+      "loss": 0.8177,
+      "step": 6413
+    },
+    {
+      "epoch": 1.5594456601021154,
+      "grad_norm": 20.5,
+      "learning_rate": 2.943800556076348e-07,
+      "loss": 0.5111,
+      "step": 6414
+    },
+    {
+      "epoch": 1.5596887916362752,
+      "grad_norm": 18.125,
+      "learning_rate": 2.940686022607178e-07,
+      "loss": 0.5951,
+      "step": 6415
+    },
+    {
+      "epoch": 1.5599319231704352,
+      "grad_norm": 16.5,
+      "learning_rate": 2.9375729179425784e-07,
+      "loss": 0.4576,
+      "step": 6416
+    },
+    {
+      "epoch": 1.5601750547045952,
+      "grad_norm": 24.0,
+      "learning_rate": 2.9344612425478523e-07,
+      "loss": 0.8859,
+      "step": 6417
+    },
+    {
+      "epoch": 1.5604181862387552,
+      "grad_norm": 30.5,
+      "learning_rate": 2.9313509968881005e-07,
+      "loss": 0.5793,
+      "step": 6418
+    },
+    {
+      "epoch": 1.5606613177729152,
+      "grad_norm": 17.125,
+      "learning_rate": 2.9282421814281965e-07,
+      "loss": 0.3839,
+      "step": 6419
+    },
+    {
+      "epoch": 1.5609044493070752,
+      "grad_norm": 20.375,
+      "learning_rate": 2.9251347966328073e-07,
+      "loss": 0.5984,
+      "step": 6420
+    },
+    {
+      "epoch": 1.5611475808412352,
+      "grad_norm": 18.5,
+      "learning_rate": 2.922028842966389e-07,
+      "loss": 0.4034,
+      "step": 6421
+    },
+    {
+      "epoch": 1.561390712375395,
+      "grad_norm": 20.125,
+      "learning_rate": 2.9189243208931765e-07,
+      "loss": 0.7853,
+      "step": 6422
+    },
+    {
+      "epoch": 1.5616338439095552,
+      "grad_norm": 18.625,
+      "learning_rate": 2.9158212308771915e-07,
+      "loss": 0.2526,
+      "step": 6423
+    },
+    {
+      "epoch": 1.561876975443715,
+      "grad_norm": 30.5,
+      "learning_rate": 2.9127195733822485e-07,
+      "loss": 0.7495,
+      "step": 6424
+    },
+    {
+      "epoch": 1.5621201069778752,
+      "grad_norm": 21.25,
+      "learning_rate": 2.909619348871948e-07,
+      "loss": 0.6682,
+      "step": 6425
+    },
+    {
+      "epoch": 1.562363238512035,
+      "grad_norm": 19.25,
+      "learning_rate": 2.9065205578096667e-07,
+      "loss": 0.514,
+      "step": 6426
+    },
+    {
+      "epoch": 1.562606370046195,
+      "grad_norm": 25.0,
+      "learning_rate": 2.9034232006585685e-07,
+      "loss": 1.1263,
+      "step": 6427
+    },
+    {
+      "epoch": 1.562849501580355,
+      "grad_norm": 23.625,
+      "learning_rate": 2.900327277881616e-07,
+      "loss": 0.7936,
+      "step": 6428
+    },
+    {
+      "epoch": 1.563092633114515,
+      "grad_norm": 25.625,
+      "learning_rate": 2.8972327899415437e-07,
+      "loss": 0.9782,
+      "step": 6429
+    },
+    {
+      "epoch": 1.563335764648675,
+      "grad_norm": 18.625,
+      "learning_rate": 2.8941397373008746e-07,
+      "loss": 0.8169,
+      "step": 6430
+    },
+    {
+      "epoch": 1.563578896182835,
+      "grad_norm": 21.5,
+      "learning_rate": 2.891048120421919e-07,
+      "loss": 0.859,
+      "step": 6431
+    },
+    {
+      "epoch": 1.563822027716995,
+      "grad_norm": 16.75,
+      "learning_rate": 2.887957939766778e-07,
+      "loss": 0.499,
+      "step": 6432
+    },
+    {
+      "epoch": 1.5640651592511547,
+      "grad_norm": 21.375,
+      "learning_rate": 2.884869195797328e-07,
+      "loss": 0.8267,
+      "step": 6433
+    },
+    {
+      "epoch": 1.564308290785315,
+      "grad_norm": 18.75,
+      "learning_rate": 2.881781888975232e-07,
+      "loss": 0.7452,
+      "step": 6434
+    },
+    {
+      "epoch": 1.5645514223194747,
+      "grad_norm": 23.375,
+      "learning_rate": 2.878696019761951e-07,
+      "loss": 0.7286,
+      "step": 6435
+    },
+    {
+      "epoch": 1.564794553853635,
+      "grad_norm": 18.0,
+      "learning_rate": 2.8756115886187125e-07,
+      "loss": 0.4586,
+      "step": 6436
+    },
+    {
+      "epoch": 1.5650376853877948,
+      "grad_norm": 17.0,
+      "learning_rate": 2.8725285960065396e-07,
+      "loss": 0.5311,
+      "step": 6437
+    },
+    {
+      "epoch": 1.5652808169219548,
+      "grad_norm": 23.375,
+      "learning_rate": 2.869447042386239e-07,
+      "loss": 1.0925,
+      "step": 6438
+    },
+    {
+      "epoch": 1.5655239484561148,
+      "grad_norm": 19.5,
+      "learning_rate": 2.8663669282184075e-07,
+      "loss": 0.5943,
+      "step": 6439
+    },
+    {
+      "epoch": 1.5657670799902748,
+      "grad_norm": 34.25,
+      "learning_rate": 2.863288253963417e-07,
+      "loss": 1.2476,
+      "step": 6440
+    },
+    {
+      "epoch": 1.5660102115244348,
+      "grad_norm": 34.25,
+      "learning_rate": 2.8602110200814257e-07,
+      "loss": 0.6716,
+      "step": 6441
+    },
+    {
+      "epoch": 1.5662533430585945,
+      "grad_norm": 16.625,
+      "learning_rate": 2.8571352270323857e-07,
+      "loss": 0.4485,
+      "step": 6442
+    },
+    {
+      "epoch": 1.5664964745927548,
+      "grad_norm": 17.75,
+      "learning_rate": 2.8540608752760227e-07,
+      "loss": 0.7211,
+      "step": 6443
+    },
+    {
+      "epoch": 1.5667396061269145,
+      "grad_norm": 18.0,
+      "learning_rate": 2.850987965271851e-07,
+      "loss": 0.6797,
+      "step": 6444
+    },
+    {
+      "epoch": 1.5669827376610748,
+      "grad_norm": 19.25,
+      "learning_rate": 2.847916497479175e-07,
+      "loss": 0.6479,
+      "step": 6445
+    },
+    {
+      "epoch": 1.5672258691952345,
+      "grad_norm": 22.0,
+      "learning_rate": 2.844846472357073e-07,
+      "loss": 0.9931,
+      "step": 6446
+    },
+    {
+      "epoch": 1.5674690007293948,
+      "grad_norm": 17.125,
+      "learning_rate": 2.8417778903644155e-07,
+      "loss": 0.5199,
+      "step": 6447
+    },
+    {
+      "epoch": 1.5677121322635545,
+      "grad_norm": 18.625,
+      "learning_rate": 2.8387107519598584e-07,
+      "loss": 0.3357,
+      "step": 6448
+    },
+    {
+      "epoch": 1.5679552637977145,
+      "grad_norm": 21.25,
+      "learning_rate": 2.8356450576018366e-07,
+      "loss": 0.6823,
+      "step": 6449
+    },
+    {
+      "epoch": 1.5681983953318746,
+      "grad_norm": 16.125,
+      "learning_rate": 2.8325808077485693e-07,
+      "loss": 0.5947,
+      "step": 6450
+    },
+    {
+      "epoch": 1.5684415268660346,
+      "grad_norm": 18.0,
+      "learning_rate": 2.829518002858059e-07,
+      "loss": 0.7884,
+      "step": 6451
+    },
+    {
+      "epoch": 1.5686846584001946,
+      "grad_norm": 18.5,
+      "learning_rate": 2.826456643388102e-07,
+      "loss": 0.5786,
+      "step": 6452
+    },
+    {
+      "epoch": 1.5689277899343543,
+      "grad_norm": 22.25,
+      "learning_rate": 2.823396729796267e-07,
+      "loss": 1.0034,
+      "step": 6453
+    },
+    {
+      "epoch": 1.5691709214685146,
+      "grad_norm": 21.25,
+      "learning_rate": 2.8203382625399056e-07,
+      "loss": 1.0719,
+      "step": 6454
+    },
+    {
+      "epoch": 1.5694140530026743,
+      "grad_norm": 19.75,
+      "learning_rate": 2.817281242076171e-07,
+      "loss": 0.8782,
+      "step": 6455
+    },
+    {
+      "epoch": 1.5696571845368346,
+      "grad_norm": 23.5,
+      "learning_rate": 2.814225668861981e-07,
+      "loss": 0.7174,
+      "step": 6456
+    },
+    {
+      "epoch": 1.5699003160709943,
+      "grad_norm": 18.375,
+      "learning_rate": 2.8111715433540437e-07,
+      "loss": 0.7646,
+      "step": 6457
+    },
+    {
+      "epoch": 1.5701434476051543,
+      "grad_norm": 31.875,
+      "learning_rate": 2.8081188660088464e-07,
+      "loss": 0.9543,
+      "step": 6458
+    },
+    {
+      "epoch": 1.5703865791393143,
+      "grad_norm": 20.625,
+      "learning_rate": 2.805067637282674e-07,
+      "loss": 0.6195,
+      "step": 6459
+    },
+    {
+      "epoch": 1.5706297106734743,
+      "grad_norm": 19.75,
+      "learning_rate": 2.802017857631581e-07,
+      "loss": 0.779,
+      "step": 6460
+    },
+    {
+      "epoch": 1.5708728422076343,
+      "grad_norm": 25.875,
+      "learning_rate": 2.798969527511401e-07,
+      "loss": 0.9677,
+      "step": 6461
+    },
+    {
+      "epoch": 1.5711159737417943,
+      "grad_norm": 25.625,
+      "learning_rate": 2.795922647377776e-07,
+      "loss": 0.9384,
+      "step": 6462
+    },
+    {
+      "epoch": 1.5713591052759543,
+      "grad_norm": 19.375,
+      "learning_rate": 2.792877217686106e-07,
+      "loss": 0.7889,
+      "step": 6463
+    },
+    {
+      "epoch": 1.5716022368101141,
+      "grad_norm": 19.75,
+      "learning_rate": 2.7898332388915787e-07,
+      "loss": 0.6775,
+      "step": 6464
+    },
+    {
+      "epoch": 1.5718453683442744,
+      "grad_norm": 16.5,
+      "learning_rate": 2.786790711449179e-07,
+      "loss": 0.3814,
+      "step": 6465
+    },
+    {
+      "epoch": 1.5720884998784341,
+      "grad_norm": 17.125,
+      "learning_rate": 2.7837496358136607e-07,
+      "loss": 0.7327,
+      "step": 6466
+    },
+    {
+      "epoch": 1.5723316314125944,
+      "grad_norm": 19.5,
+      "learning_rate": 2.780710012439565e-07,
+      "loss": 0.8297,
+      "step": 6467
+    },
+    {
+      "epoch": 1.5725747629467541,
+      "grad_norm": 16.875,
+      "learning_rate": 2.777671841781211e-07,
+      "loss": 0.5352,
+      "step": 6468
+    },
+    {
+      "epoch": 1.5728178944809141,
+      "grad_norm": 28.375,
+      "learning_rate": 2.7746351242927114e-07,
+      "loss": 0.8466,
+      "step": 6469
+    },
+    {
+      "epoch": 1.5730610260150741,
+      "grad_norm": 17.0,
+      "learning_rate": 2.7715998604279595e-07,
+      "loss": 0.4403,
+      "step": 6470
+    },
+    {
+      "epoch": 1.5733041575492341,
+      "grad_norm": 17.5,
+      "learning_rate": 2.76856605064062e-07,
+      "loss": 0.3576,
+      "step": 6471
+    },
+    {
+      "epoch": 1.5735472890833941,
+      "grad_norm": 20.125,
+      "learning_rate": 2.765533695384155e-07,
+      "loss": 0.7101,
+      "step": 6472
+    },
+    {
+      "epoch": 1.5737904206175541,
+      "grad_norm": 19.125,
+      "learning_rate": 2.762502795111799e-07,
+      "loss": 0.4901,
+      "step": 6473
+    },
+    {
+      "epoch": 1.5740335521517141,
+      "grad_norm": 25.625,
+      "learning_rate": 2.759473350276573e-07,
+      "loss": 1.4545,
+      "step": 6474
+    },
+    {
+      "epoch": 1.574276683685874,
+      "grad_norm": 30.375,
+      "learning_rate": 2.756445361331275e-07,
+      "loss": 1.1691,
+      "step": 6475
+    },
+    {
+      "epoch": 1.5745198152200341,
+      "grad_norm": 23.25,
+      "learning_rate": 2.7534188287285006e-07,
+      "loss": 0.9106,
+      "step": 6476
+    },
+    {
+      "epoch": 1.574762946754194,
+      "grad_norm": 24.125,
+      "learning_rate": 2.7503937529206063e-07,
+      "loss": 0.7725,
+      "step": 6477
+    },
+    {
+      "epoch": 1.5750060782883541,
+      "grad_norm": 21.5,
+      "learning_rate": 2.7473701343597474e-07,
+      "loss": 0.8067,
+      "step": 6478
+    },
+    {
+      "epoch": 1.575249209822514,
+      "grad_norm": 20.875,
+      "learning_rate": 2.74434797349786e-07,
+      "loss": 0.7489,
+      "step": 6479
+    },
+    {
+      "epoch": 1.575492341356674,
+      "grad_norm": 20.5,
+      "learning_rate": 2.741327270786652e-07,
+      "loss": 0.8283,
+      "step": 6480
+    },
+    {
+      "epoch": 1.575735472890834,
+      "grad_norm": 20.375,
+      "learning_rate": 2.7383080266776203e-07,
+      "loss": 0.6953,
+      "step": 6481
+    },
+    {
+      "epoch": 1.575978604424994,
+      "grad_norm": 20.5,
+      "learning_rate": 2.7352902416220463e-07,
+      "loss": 0.6859,
+      "step": 6482
+    },
+    {
+      "epoch": 1.576221735959154,
+      "grad_norm": 20.25,
+      "learning_rate": 2.7322739160709877e-07,
+      "loss": 0.8769,
+      "step": 6483
+    },
+    {
+      "epoch": 1.576464867493314,
+      "grad_norm": 18.5,
+      "learning_rate": 2.7292590504752884e-07,
+      "loss": 0.4509,
+      "step": 6484
+    },
+    {
+      "epoch": 1.576707999027474,
+      "grad_norm": 24.375,
+      "learning_rate": 2.726245645285562e-07,
+      "loss": 0.3359,
+      "step": 6485
+    },
+    {
+      "epoch": 1.5769511305616337,
+      "grad_norm": 26.5,
+      "learning_rate": 2.723233700952231e-07,
+      "loss": 1.3967,
+      "step": 6486
+    },
+    {
+      "epoch": 1.577194262095794,
+      "grad_norm": 20.375,
+      "learning_rate": 2.720223217925473e-07,
+      "loss": 0.9031,
+      "step": 6487
+    },
+    {
+      "epoch": 1.5774373936299537,
+      "grad_norm": 27.5,
+      "learning_rate": 2.7172141966552545e-07,
+      "loss": 0.9432,
+      "step": 6488
+    },
+    {
+      "epoch": 1.577680525164114,
+      "grad_norm": 19.625,
+      "learning_rate": 2.714206637591332e-07,
+      "loss": 0.7459,
+      "step": 6489
+    },
+    {
+      "epoch": 1.5779236566982737,
+      "grad_norm": 17.5,
+      "learning_rate": 2.7112005411832345e-07,
+      "loss": 0.6623,
+      "step": 6490
+    },
+    {
+      "epoch": 1.5781667882324337,
+      "grad_norm": 21.0,
+      "learning_rate": 2.7081959078802734e-07,
+      "loss": 0.8332,
+      "step": 6491
+    },
+    {
+      "epoch": 1.5784099197665937,
+      "grad_norm": 17.875,
+      "learning_rate": 2.705192738131539e-07,
+      "loss": 0.5069,
+      "step": 6492
+    },
+    {
+      "epoch": 1.5786530513007537,
+      "grad_norm": 20.125,
+      "learning_rate": 2.7021910323859173e-07,
+      "loss": 0.7234,
+      "step": 6493
+    },
+    {
+      "epoch": 1.5788961828349137,
+      "grad_norm": 18.0,
+      "learning_rate": 2.699190791092061e-07,
+      "loss": 0.5635,
+      "step": 6494
+    },
+    {
+      "epoch": 1.5791393143690735,
+      "grad_norm": 16.75,
+      "learning_rate": 2.6961920146984024e-07,
+      "loss": 0.4454,
+      "step": 6495
+    },
+    {
+      "epoch": 1.5793824459032337,
+      "grad_norm": 14.8125,
+      "learning_rate": 2.6931947036531696e-07,
+      "loss": 0.3257,
+      "step": 6496
+    },
+    {
+      "epoch": 1.5796255774373935,
+      "grad_norm": 18.0,
+      "learning_rate": 2.690198858404357e-07,
+      "loss": 0.7404,
+      "step": 6497
+    },
+    {
+      "epoch": 1.5798687089715537,
+      "grad_norm": 18.125,
+      "learning_rate": 2.6872044793997424e-07,
+      "loss": 0.7507,
+      "step": 6498
+    },
+    {
+      "epoch": 1.5801118405057135,
+      "grad_norm": 23.25,
+      "learning_rate": 2.6842115670868947e-07,
+      "loss": 1.0221,
+      "step": 6499
+    },
+    {
+      "epoch": 1.5803549720398735,
+      "grad_norm": 20.25,
+      "learning_rate": 2.6812201219131503e-07,
+      "loss": 1.0032,
+      "step": 6500
+    },
+    {
+      "epoch": 1.5805981035740335,
+      "grad_norm": 22.25,
+      "learning_rate": 2.6782301443256374e-07,
+      "loss": 0.9321,
+      "step": 6501
+    },
+    {
+      "epoch": 1.5808412351081935,
+      "grad_norm": 27.0,
+      "learning_rate": 2.675241634771253e-07,
+      "loss": 0.8659,
+      "step": 6502
+    },
+    {
+      "epoch": 1.5810843666423535,
+      "grad_norm": 21.625,
+      "learning_rate": 2.672254593696691e-07,
+      "loss": 0.9163,
+      "step": 6503
+    },
+    {
+      "epoch": 1.5813274981765135,
+      "grad_norm": 28.0,
+      "learning_rate": 2.6692690215484096e-07,
+      "loss": 0.6282,
+      "step": 6504
+    },
+    {
+      "epoch": 1.5815706297106735,
+      "grad_norm": 17.25,
+      "learning_rate": 2.6662849187726533e-07,
+      "loss": 0.4821,
+      "step": 6505
+    },
+    {
+      "epoch": 1.5818137612448333,
+      "grad_norm": 18.375,
+      "learning_rate": 2.663302285815453e-07,
+      "loss": 0.7284,
+      "step": 6506
+    },
+    {
+      "epoch": 1.5820568927789935,
+      "grad_norm": 22.375,
+      "learning_rate": 2.660321123122611e-07,
+      "loss": 1.2149,
+      "step": 6507
+    },
+    {
+      "epoch": 1.5823000243131533,
+      "grad_norm": 27.625,
+      "learning_rate": 2.6573414311397107e-07,
+      "loss": 0.9231,
+      "step": 6508
+    },
+    {
+      "epoch": 1.5825431558473135,
+      "grad_norm": 20.125,
+      "learning_rate": 2.6543632103121234e-07,
+      "loss": 0.6957,
+      "step": 6509
+    },
+    {
+      "epoch": 1.5827862873814733,
+      "grad_norm": 22.75,
+      "learning_rate": 2.651386461084997e-07,
+      "loss": 0.6521,
+      "step": 6510
+    },
+    {
+      "epoch": 1.5830294189156333,
+      "grad_norm": 18.625,
+      "learning_rate": 2.648411183903256e-07,
+      "loss": 0.7038,
+      "step": 6511
+    },
+    {
+      "epoch": 1.5832725504497933,
+      "grad_norm": 18.25,
+      "learning_rate": 2.645437379211603e-07,
+      "loss": 0.8506,
+      "step": 6512
+    },
+    {
+      "epoch": 1.5835156819839533,
+      "grad_norm": 20.0,
+      "learning_rate": 2.64246504745453e-07,
+      "loss": 0.6279,
+      "step": 6513
+    },
+    {
+      "epoch": 1.5837588135181133,
+      "grad_norm": 17.125,
+      "learning_rate": 2.6394941890763025e-07,
+      "loss": 0.6043,
+      "step": 6514
+    },
+    {
+      "epoch": 1.5840019450522733,
+      "grad_norm": 17.25,
+      "learning_rate": 2.636524804520961e-07,
+      "loss": 0.6408,
+      "step": 6515
+    },
+    {
+      "epoch": 1.5842450765864333,
+      "grad_norm": 21.25,
+      "learning_rate": 2.6335568942323365e-07,
+      "loss": 0.7061,
+      "step": 6516
+    },
+    {
+      "epoch": 1.584488208120593,
+      "grad_norm": 23.5,
+      "learning_rate": 2.6305904586540367e-07,
+      "loss": 0.6464,
+      "step": 6517
+    },
+    {
+      "epoch": 1.5847313396547533,
+      "grad_norm": 21.125,
+      "learning_rate": 2.6276254982294427e-07,
+      "loss": 0.7062,
+      "step": 6518
+    },
+    {
+      "epoch": 1.584974471188913,
+      "grad_norm": 18.125,
+      "learning_rate": 2.6246620134017174e-07,
+      "loss": 0.6233,
+      "step": 6519
+    },
+    {
+      "epoch": 1.5852176027230733,
+      "grad_norm": 21.75,
+      "learning_rate": 2.62170000461381e-07,
+      "loss": 0.8793,
+      "step": 6520
+    },
+    {
+      "epoch": 1.585460734257233,
+      "grad_norm": 18.75,
+      "learning_rate": 2.6187394723084403e-07,
+      "loss": 0.6398,
+      "step": 6521
+    },
+    {
+      "epoch": 1.585703865791393,
+      "grad_norm": 31.125,
+      "learning_rate": 2.61578041692811e-07,
+      "loss": 0.837,
+      "step": 6522
+    },
+    {
+      "epoch": 1.585946997325553,
+      "grad_norm": 22.375,
+      "learning_rate": 2.612822838915102e-07,
+      "loss": 0.7556,
+      "step": 6523
+    },
+    {
+      "epoch": 1.586190128859713,
+      "grad_norm": 21.5,
+      "learning_rate": 2.609866738711481e-07,
+      "loss": 0.561,
+      "step": 6524
+    },
+    {
+      "epoch": 1.5864332603938731,
+      "grad_norm": 23.625,
+      "learning_rate": 2.6069121167590846e-07,
+      "loss": 0.8384,
+      "step": 6525
+    },
+    {
+      "epoch": 1.5866763919280331,
+      "grad_norm": 18.75,
+      "learning_rate": 2.603958973499529e-07,
+      "loss": 0.5325,
+      "step": 6526
+    },
+    {
+      "epoch": 1.5869195234621931,
+      "grad_norm": 17.25,
+      "learning_rate": 2.601007309374219e-07,
+      "loss": 0.6433,
+      "step": 6527
+    },
+    {
+      "epoch": 1.587162654996353,
+      "grad_norm": 18.0,
+      "learning_rate": 2.598057124824328e-07,
+      "loss": 0.4489,
+      "step": 6528
+    },
+    {
+      "epoch": 1.5874057865305131,
+      "grad_norm": 18.625,
+      "learning_rate": 2.595108420290808e-07,
+      "loss": 0.6789,
+      "step": 6529
+    },
+    {
+      "epoch": 1.587648918064673,
+      "grad_norm": 21.625,
+      "learning_rate": 2.5921611962144014e-07,
+      "loss": 0.7601,
+      "step": 6530
+    },
+    {
+      "epoch": 1.5878920495988331,
+      "grad_norm": 20.125,
+      "learning_rate": 2.5892154530356166e-07,
+      "loss": 0.5507,
+      "step": 6531
+    },
+    {
+      "epoch": 1.588135181132993,
+      "grad_norm": 21.875,
+      "learning_rate": 2.586271191194749e-07,
+      "loss": 1.0553,
+      "step": 6532
+    },
+    {
+      "epoch": 1.588378312667153,
+      "grad_norm": 24.0,
+      "learning_rate": 2.5833284111318647e-07,
+      "loss": 0.8897,
+      "step": 6533
+    },
+    {
+      "epoch": 1.588621444201313,
+      "grad_norm": 21.0,
+      "learning_rate": 2.580387113286821e-07,
+      "loss": 0.9927,
+      "step": 6534
+    },
+    {
+      "epoch": 1.588864575735473,
+      "grad_norm": 22.75,
+      "learning_rate": 2.5774472980992394e-07,
+      "loss": 0.4649,
+      "step": 6535
+    },
+    {
+      "epoch": 1.589107707269633,
+      "grad_norm": 15.75,
+      "learning_rate": 2.574508966008525e-07,
+      "loss": 0.5479,
+      "step": 6536
+    },
+    {
+      "epoch": 1.5893508388037927,
+      "grad_norm": 19.625,
+      "learning_rate": 2.571572117453867e-07,
+      "loss": 0.7894,
+      "step": 6537
+    },
+    {
+      "epoch": 1.589593970337953,
+      "grad_norm": 17.625,
+      "learning_rate": 2.5686367528742277e-07,
+      "loss": 0.603,
+      "step": 6538
+    },
+    {
+      "epoch": 1.5898371018721127,
+      "grad_norm": 19.75,
+      "learning_rate": 2.5657028727083424e-07,
+      "loss": 1.1031,
+      "step": 6539
+    },
+    {
+      "epoch": 1.590080233406273,
+      "grad_norm": 27.125,
+      "learning_rate": 2.5627704773947326e-07,
+      "loss": 0.8461,
+      "step": 6540
+    },
+    {
+      "epoch": 1.5903233649404327,
+      "grad_norm": 19.125,
+      "learning_rate": 2.559839567371701e-07,
+      "loss": 0.416,
+      "step": 6541
+    },
+    {
+      "epoch": 1.590566496474593,
+      "grad_norm": 21.125,
+      "learning_rate": 2.5569101430773195e-07,
+      "loss": 0.8185,
+      "step": 6542
+    },
+    {
+      "epoch": 1.5908096280087527,
+      "grad_norm": 17.0,
+      "learning_rate": 2.553982204949436e-07,
+      "loss": 0.5952,
+      "step": 6543
+    },
+    {
+      "epoch": 1.5910527595429127,
+      "grad_norm": 22.5,
+      "learning_rate": 2.551055753425689e-07,
+      "loss": 0.6843,
+      "step": 6544
+    },
+    {
+      "epoch": 1.5912958910770727,
+      "grad_norm": 26.125,
+      "learning_rate": 2.548130788943484e-07,
+      "loss": 0.7056,
+      "step": 6545
+    },
+    {
+      "epoch": 1.5915390226112327,
+      "grad_norm": 24.875,
+      "learning_rate": 2.545207311940004e-07,
+      "loss": 0.9516,
+      "step": 6546
+    },
+    {
+      "epoch": 1.5917821541453927,
+      "grad_norm": 24.125,
+      "learning_rate": 2.5422853228522155e-07,
+      "loss": 0.7309,
+      "step": 6547
+    },
+    {
+      "epoch": 1.5920252856795525,
+      "grad_norm": 30.375,
+      "learning_rate": 2.539364822116866e-07,
+      "loss": 0.76,
+      "step": 6548
+    },
+    {
+      "epoch": 1.5922684172137127,
+      "grad_norm": 21.5,
+      "learning_rate": 2.5364458101704694e-07,
+      "loss": 0.7654,
+      "step": 6549
+    },
+    {
+      "epoch": 1.5925115487478725,
+      "grad_norm": 20.75,
+      "learning_rate": 2.5335282874493204e-07,
+      "loss": 0.5495,
+      "step": 6550
+    },
+    {
+      "epoch": 1.5927546802820327,
+      "grad_norm": 16.75,
+      "learning_rate": 2.5306122543894975e-07,
+      "loss": 0.6322,
+      "step": 6551
+    },
+    {
+      "epoch": 1.5929978118161925,
+      "grad_norm": 25.0,
+      "learning_rate": 2.527697711426852e-07,
+      "loss": 0.7557,
+      "step": 6552
+    },
+    {
+      "epoch": 1.5932409433503525,
+      "grad_norm": 19.5,
+      "learning_rate": 2.5247846589970065e-07,
+      "loss": 0.7574,
+      "step": 6553
+    },
+    {
+      "epoch": 1.5934840748845125,
+      "grad_norm": 19.0,
+      "learning_rate": 2.5218730975353725e-07,
+      "loss": 0.6171,
+      "step": 6554
+    },
+    {
+      "epoch": 1.5937272064186725,
+      "grad_norm": 21.25,
+      "learning_rate": 2.518963027477135e-07,
+      "loss": 0.871,
+      "step": 6555
+    },
+    {
+      "epoch": 1.5939703379528325,
+      "grad_norm": 23.625,
+      "learning_rate": 2.5160544492572517e-07,
+      "loss": 0.9971,
+      "step": 6556
+    },
+    {
+      "epoch": 1.5942134694869925,
+      "grad_norm": 16.25,
+      "learning_rate": 2.5131473633104575e-07,
+      "loss": 0.4762,
+      "step": 6557
+    },
+    {
+      "epoch": 1.5944566010211525,
+      "grad_norm": 19.625,
+      "learning_rate": 2.5102417700712714e-07,
+      "loss": 0.8368,
+      "step": 6558
+    },
+    {
+      "epoch": 1.5946997325553123,
+      "grad_norm": 22.375,
+      "learning_rate": 2.5073376699739805e-07,
+      "loss": 0.7926,
+      "step": 6559
+    },
+    {
+      "epoch": 1.5949428640894725,
+      "grad_norm": 20.25,
+      "learning_rate": 2.504435063452652e-07,
+      "loss": 0.5631,
+      "step": 6560
+    },
+    {
+      "epoch": 1.5951859956236323,
+      "grad_norm": 18.5,
+      "learning_rate": 2.501533950941136e-07,
+      "loss": 0.7315,
+      "step": 6561
+    },
+    {
+      "epoch": 1.5954291271577925,
+      "grad_norm": 19.625,
+      "learning_rate": 2.498634332873047e-07,
+      "loss": 0.7402,
+      "step": 6562
+    },
+    {
+      "epoch": 1.5956722586919523,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.49573620968179e-07,
+      "loss": 0.3719,
+      "step": 6563
+    },
+    {
+      "epoch": 1.5959153902261123,
+      "grad_norm": 21.875,
+      "learning_rate": 2.4928395818005324e-07,
+      "loss": 0.8335,
+      "step": 6564
+    },
+    {
+      "epoch": 1.5961585217602723,
+      "grad_norm": 21.875,
+      "learning_rate": 2.489944449662232e-07,
+      "loss": 1.0901,
+      "step": 6565
+    },
+    {
+      "epoch": 1.5964016532944323,
+      "grad_norm": 11.5,
+      "learning_rate": 2.487050813699614e-07,
+      "loss": 0.2278,
+      "step": 6566
+    },
+    {
+      "epoch": 1.5966447848285923,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.484158674345179e-07,
+      "loss": 0.4144,
+      "step": 6567
+    },
+    {
+      "epoch": 1.5968879163627523,
+      "grad_norm": 22.875,
+      "learning_rate": 2.481268032031212e-07,
+      "loss": 0.89,
+      "step": 6568
+    },
+    {
+      "epoch": 1.5971310478969123,
+      "grad_norm": 24.625,
+      "learning_rate": 2.4783788871897654e-07,
+      "loss": 0.6295,
+      "step": 6569
+    },
+    {
+      "epoch": 1.597374179431072,
+      "grad_norm": 22.375,
+      "learning_rate": 2.4754912402526727e-07,
+      "loss": 0.511,
+      "step": 6570
+    },
+    {
+      "epoch": 1.5976173109652323,
+      "grad_norm": 31.875,
+      "learning_rate": 2.472605091651542e-07,
+      "loss": 0.7927,
+      "step": 6571
+    },
+    {
+      "epoch": 1.597860442499392,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4697204418177634e-07,
+      "loss": 0.5545,
+      "step": 6572
+    },
+    {
+      "epoch": 1.5981035740335523,
+      "grad_norm": 22.875,
+      "learning_rate": 2.466837291182493e-07,
+      "loss": 1.0362,
+      "step": 6573
+    },
+    {
+      "epoch": 1.598346705567712,
+      "grad_norm": 16.625,
+      "learning_rate": 2.4639556401766655e-07,
+      "loss": 0.2959,
+      "step": 6574
+    },
+    {
+      "epoch": 1.598589837101872,
+      "grad_norm": 22.125,
+      "learning_rate": 2.461075489230999e-07,
+      "loss": 0.4179,
+      "step": 6575
+    },
+    {
+      "epoch": 1.598832968636032,
+      "grad_norm": 17.125,
+      "learning_rate": 2.4581968387759795e-07,
+      "loss": 0.7986,
+      "step": 6576
+    },
+    {
+      "epoch": 1.599076100170192,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.455319689241867e-07,
+      "loss": 0.3135,
+      "step": 6577
+    },
+    {
+      "epoch": 1.599319231704352,
+      "grad_norm": 24.25,
+      "learning_rate": 2.4524440410587047e-07,
+      "loss": 0.906,
+      "step": 6578
+    },
+    {
+      "epoch": 1.599562363238512,
+      "grad_norm": 22.0,
+      "learning_rate": 2.4495698946563103e-07,
+      "loss": 0.6496,
+      "step": 6579
+    },
+    {
+      "epoch": 1.599805494772672,
+      "grad_norm": 17.25,
+      "learning_rate": 2.4466972504642724e-07,
+      "loss": 0.8979,
+      "step": 6580
+    },
+    {
+      "epoch": 1.6000486263068319,
+      "grad_norm": 17.875,
+      "learning_rate": 2.4438261089119535e-07,
+      "loss": 0.7324,
+      "step": 6581
+    },
+    {
+      "epoch": 1.600291757840992,
+      "grad_norm": 15.5,
+      "learning_rate": 2.4409564704285013e-07,
+      "loss": 0.476,
+      "step": 6582
+    },
+    {
+      "epoch": 1.6005348893751519,
+      "grad_norm": 14.1875,
+      "learning_rate": 2.4380883354428316e-07,
+      "loss": 0.3257,
+      "step": 6583
+    },
+    {
+      "epoch": 1.600778020909312,
+      "grad_norm": 19.875,
+      "learning_rate": 2.435221704383632e-07,
+      "loss": 0.7453,
+      "step": 6584
+    },
+    {
+      "epoch": 1.6010211524434719,
+      "grad_norm": 15.625,
+      "learning_rate": 2.4323565776793735e-07,
+      "loss": 0.353,
+      "step": 6585
+    },
+    {
+      "epoch": 1.6012642839776319,
+      "grad_norm": 17.375,
+      "learning_rate": 2.429492955758302e-07,
+      "loss": 0.5536,
+      "step": 6586
+    },
+    {
+      "epoch": 1.6015074155117919,
+      "grad_norm": 24.0,
+      "learning_rate": 2.4266308390484315e-07,
+      "loss": 0.8279,
+      "step": 6587
+    },
+    {
+      "epoch": 1.6017505470459519,
+      "grad_norm": 23.375,
+      "learning_rate": 2.423770227977551e-07,
+      "loss": 0.6825,
+      "step": 6588
+    },
+    {
+      "epoch": 1.6019936785801119,
+      "grad_norm": 20.125,
+      "learning_rate": 2.4209111229732363e-07,
+      "loss": 0.6046,
+      "step": 6589
+    },
+    {
+      "epoch": 1.6022368101142717,
+      "grad_norm": 21.375,
+      "learning_rate": 2.4180535244628265e-07,
+      "loss": 0.7474,
+      "step": 6590
+    },
+    {
+      "epoch": 1.6024799416484319,
+      "grad_norm": 22.125,
+      "learning_rate": 2.415197432873437e-07,
+      "loss": 0.8632,
+      "step": 6591
+    },
+    {
+      "epoch": 1.6027230731825917,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.4123428486319627e-07,
+      "loss": 0.4441,
+      "step": 6592
+    },
+    {
+      "epoch": 1.6029662047167519,
+      "grad_norm": 17.5,
+      "learning_rate": 2.4094897721650675e-07,
+      "loss": 0.6094,
+      "step": 6593
+    },
+    {
+      "epoch": 1.6032093362509117,
+      "grad_norm": 23.125,
+      "learning_rate": 2.4066382038991973e-07,
+      "loss": 0.7961,
+      "step": 6594
+    },
+    {
+      "epoch": 1.6034524677850719,
+      "grad_norm": 19.0,
+      "learning_rate": 2.4037881442605633e-07,
+      "loss": 0.5949,
+      "step": 6595
+    },
+    {
+      "epoch": 1.6036955993192317,
+      "grad_norm": 18.625,
+      "learning_rate": 2.400939593675161e-07,
+      "loss": 0.7009,
+      "step": 6596
+    },
+    {
+      "epoch": 1.6039387308533917,
+      "grad_norm": 20.875,
+      "learning_rate": 2.398092552568755e-07,
+      "loss": 0.833,
+      "step": 6597
+    },
+    {
+      "epoch": 1.6041818623875517,
+      "grad_norm": 15.0,
+      "learning_rate": 2.3952470213668785e-07,
+      "loss": 0.3738,
+      "step": 6598
+    },
+    {
+      "epoch": 1.6044249939217117,
+      "grad_norm": 18.625,
+      "learning_rate": 2.392403000494853e-07,
+      "loss": 0.8037,
+      "step": 6599
+    },
+    {
+      "epoch": 1.6046681254558717,
+      "grad_norm": 19.875,
+      "learning_rate": 2.389560490377764e-07,
+      "loss": 0.6583,
+      "step": 6600
+    },
+    {
+      "epoch": 1.6049112569900315,
+      "grad_norm": 21.375,
+      "learning_rate": 2.38671949144047e-07,
+      "loss": 0.7524,
+      "step": 6601
+    },
+    {
+      "epoch": 1.6051543885241917,
+      "grad_norm": 18.375,
+      "learning_rate": 2.3838800041076096e-07,
+      "loss": 0.5015,
+      "step": 6602
+    },
+    {
+      "epoch": 1.6053975200583515,
+      "grad_norm": 17.25,
+      "learning_rate": 2.3810420288035971e-07,
+      "loss": 0.4567,
+      "step": 6603
+    },
+    {
+      "epoch": 1.6056406515925117,
+      "grad_norm": 20.5,
+      "learning_rate": 2.378205565952614e-07,
+      "loss": 0.6397,
+      "step": 6604
+    },
+    {
+      "epoch": 1.6058837831266715,
+      "grad_norm": 21.625,
+      "learning_rate": 2.3753706159786158e-07,
+      "loss": 0.7376,
+      "step": 6605
+    },
+    {
+      "epoch": 1.6061269146608315,
+      "grad_norm": 92.0,
+      "learning_rate": 2.372537179305341e-07,
+      "loss": 1.0405,
+      "step": 6606
+    },
+    {
+      "epoch": 1.6063700461949915,
+      "grad_norm": 20.625,
+      "learning_rate": 2.3697052563562918e-07,
+      "loss": 0.8989,
+      "step": 6607
+    },
+    {
+      "epoch": 1.6066131777291515,
+      "grad_norm": 21.375,
+      "learning_rate": 2.366874847554744e-07,
+      "loss": 0.7002,
+      "step": 6608
+    },
+    {
+      "epoch": 1.6068563092633115,
+      "grad_norm": 23.375,
+      "learning_rate": 2.3640459533237556e-07,
+      "loss": 0.8436,
+      "step": 6609
+    },
+    {
+      "epoch": 1.6070994407974715,
+      "grad_norm": 25.125,
+      "learning_rate": 2.3612185740861577e-07,
+      "loss": 0.911,
+      "step": 6610
+    },
+    {
+      "epoch": 1.6073425723316315,
+      "grad_norm": 32.25,
+      "learning_rate": 2.3583927102645467e-07,
+      "loss": 1.0438,
+      "step": 6611
+    },
+    {
+      "epoch": 1.6075857038657912,
+      "grad_norm": 22.625,
+      "learning_rate": 2.3555683622812922e-07,
+      "loss": 0.7154,
+      "step": 6612
+    },
+    {
+      "epoch": 1.6078288353999515,
+      "grad_norm": 16.125,
+      "learning_rate": 2.3527455305585506e-07,
+      "loss": 0.3308,
+      "step": 6613
+    },
+    {
+      "epoch": 1.6080719669341113,
+      "grad_norm": 19.875,
+      "learning_rate": 2.3499242155182387e-07,
+      "loss": 0.7239,
+      "step": 6614
+    },
+    {
+      "epoch": 1.6083150984682715,
+      "grad_norm": 24.625,
+      "learning_rate": 2.3471044175820468e-07,
+      "loss": 0.6991,
+      "step": 6615
+    },
+    {
+      "epoch": 1.6085582300024313,
+      "grad_norm": 14.875,
+      "learning_rate": 2.3442861371714477e-07,
+      "loss": 0.3567,
+      "step": 6616
+    },
+    {
+      "epoch": 1.6088013615365913,
+      "grad_norm": 35.5,
+      "learning_rate": 2.341469374707682e-07,
+      "loss": 0.6461,
+      "step": 6617
+    },
+    {
+      "epoch": 1.6090444930707513,
+      "grad_norm": 20.25,
+      "learning_rate": 2.338654130611763e-07,
+      "loss": 1.3156,
+      "step": 6618
+    },
+    {
+      "epoch": 1.6092876246049113,
+      "grad_norm": 21.0,
+      "learning_rate": 2.3358404053044736e-07,
+      "loss": 0.7001,
+      "step": 6619
+    },
+    {
+      "epoch": 1.6095307561390713,
+      "grad_norm": 13.0625,
+      "learning_rate": 2.3330281992063803e-07,
+      "loss": 0.381,
+      "step": 6620
+    },
+    {
+      "epoch": 1.6097738876732313,
+      "grad_norm": 22.625,
+      "learning_rate": 2.330217512737812e-07,
+      "loss": 1.0427,
+      "step": 6621
+    },
+    {
+      "epoch": 1.6100170192073913,
+      "grad_norm": 21.75,
+      "learning_rate": 2.3274083463188712e-07,
+      "loss": 0.6807,
+      "step": 6622
+    },
+    {
+      "epoch": 1.610260150741551,
+      "grad_norm": 21.75,
+      "learning_rate": 2.324600700369442e-07,
+      "loss": 0.5963,
+      "step": 6623
+    },
+    {
+      "epoch": 1.6105032822757113,
+      "grad_norm": 20.625,
+      "learning_rate": 2.321794575309172e-07,
+      "loss": 0.7102,
+      "step": 6624
+    },
+    {
+      "epoch": 1.610746413809871,
+      "grad_norm": 31.625,
+      "learning_rate": 2.3189899715574892e-07,
+      "loss": 0.6249,
+      "step": 6625
+    },
+    {
+      "epoch": 1.6109895453440313,
+      "grad_norm": 22.625,
+      "learning_rate": 2.3161868895335838e-07,
+      "loss": 0.6783,
+      "step": 6626
+    },
+    {
+      "epoch": 1.611232676878191,
+      "grad_norm": 24.125,
+      "learning_rate": 2.3133853296564313e-07,
+      "loss": 1.0236,
+      "step": 6627
+    },
+    {
+      "epoch": 1.611475808412351,
+      "grad_norm": 21.75,
+      "learning_rate": 2.3105852923447692e-07,
+      "loss": 1.2369,
+      "step": 6628
+    },
+    {
+      "epoch": 1.611718939946511,
+      "grad_norm": 23.875,
+      "learning_rate": 2.3077867780171112e-07,
+      "loss": 0.6579,
+      "step": 6629
+    },
+    {
+      "epoch": 1.611962071480671,
+      "grad_norm": 22.125,
+      "learning_rate": 2.3049897870917466e-07,
+      "loss": 0.6771,
+      "step": 6630
+    },
+    {
+      "epoch": 1.612205203014831,
+      "grad_norm": 21.25,
+      "learning_rate": 2.3021943199867295e-07,
+      "loss": 0.7821,
+      "step": 6631
+    },
+    {
+      "epoch": 1.612448334548991,
+      "grad_norm": 24.25,
+      "learning_rate": 2.2994003771198967e-07,
+      "loss": 0.9348,
+      "step": 6632
+    },
+    {
+      "epoch": 1.612691466083151,
+      "grad_norm": 25.375,
+      "learning_rate": 2.2966079589088446e-07,
+      "loss": 0.7367,
+      "step": 6633
+    },
+    {
+      "epoch": 1.6129345976173108,
+      "grad_norm": 12.375,
+      "learning_rate": 2.2938170657709562e-07,
+      "loss": 0.2971,
+      "step": 6634
+    },
+    {
+      "epoch": 1.613177729151471,
+      "grad_norm": 20.375,
+      "learning_rate": 2.2910276981233728e-07,
+      "loss": 1.1531,
+      "step": 6635
+    },
+    {
+      "epoch": 1.6134208606856308,
+      "grad_norm": 20.125,
+      "learning_rate": 2.2882398563830122e-07,
+      "loss": 0.6736,
+      "step": 6636
+    },
+    {
+      "epoch": 1.613663992219791,
+      "grad_norm": 21.625,
+      "learning_rate": 2.2854535409665715e-07,
+      "loss": 0.5975,
+      "step": 6637
+    },
+    {
+      "epoch": 1.6139071237539508,
+      "grad_norm": 18.625,
+      "learning_rate": 2.2826687522905096e-07,
+      "loss": 0.6806,
+      "step": 6638
+    },
+    {
+      "epoch": 1.6141502552881108,
+      "grad_norm": 24.125,
+      "learning_rate": 2.2798854907710596e-07,
+      "loss": 0.7261,
+      "step": 6639
+    },
+    {
+      "epoch": 1.6143933868222708,
+      "grad_norm": 19.5,
+      "learning_rate": 2.277103756824231e-07,
+      "loss": 0.9476,
+      "step": 6640
+    },
+    {
+      "epoch": 1.6146365183564308,
+      "grad_norm": 23.5,
+      "learning_rate": 2.274323550865805e-07,
+      "loss": 1.0675,
+      "step": 6641
+    },
+    {
+      "epoch": 1.6148796498905909,
+      "grad_norm": 18.5,
+      "learning_rate": 2.2715448733113264e-07,
+      "loss": 0.9203,
+      "step": 6642
+    },
+    {
+      "epoch": 1.6151227814247506,
+      "grad_norm": 22.375,
+      "learning_rate": 2.2687677245761153e-07,
+      "loss": 0.5243,
+      "step": 6643
+    },
+    {
+      "epoch": 1.6153659129589109,
+      "grad_norm": 21.5,
+      "learning_rate": 2.2659921050752708e-07,
+      "loss": 0.8116,
+      "step": 6644
+    },
+    {
+      "epoch": 1.6156090444930706,
+      "grad_norm": 20.25,
+      "learning_rate": 2.2632180152236532e-07,
+      "loss": 0.6487,
+      "step": 6645
+    },
+    {
+      "epoch": 1.6158521760272309,
+      "grad_norm": 19.375,
+      "learning_rate": 2.260445455435896e-07,
+      "loss": 0.6804,
+      "step": 6646
+    },
+    {
+      "epoch": 1.6160953075613906,
+      "grad_norm": 19.5,
+      "learning_rate": 2.2576744261264077e-07,
+      "loss": 0.5433,
+      "step": 6647
+    },
+    {
+      "epoch": 1.6163384390955506,
+      "grad_norm": 18.25,
+      "learning_rate": 2.2549049277093712e-07,
+      "loss": 0.8044,
+      "step": 6648
+    },
+    {
+      "epoch": 1.6165815706297106,
+      "grad_norm": 16.75,
+      "learning_rate": 2.2521369605987332e-07,
+      "loss": 0.2774,
+      "step": 6649
+    },
+    {
+      "epoch": 1.6168247021638706,
+      "grad_norm": 24.875,
+      "learning_rate": 2.2493705252082081e-07,
+      "loss": 0.7089,
+      "step": 6650
+    },
+    {
+      "epoch": 1.6170678336980306,
+      "grad_norm": 13.5,
+      "learning_rate": 2.2466056219512976e-07,
+      "loss": 0.2571,
+      "step": 6651
+    },
+    {
+      "epoch": 1.6173109652321906,
+      "grad_norm": 16.125,
+      "learning_rate": 2.2438422512412572e-07,
+      "loss": 0.6296,
+      "step": 6652
+    },
+    {
+      "epoch": 1.6175540967663506,
+      "grad_norm": 27.25,
+      "learning_rate": 2.2410804134911201e-07,
+      "loss": 0.9578,
+      "step": 6653
+    },
+    {
+      "epoch": 1.6177972283005104,
+      "grad_norm": 21.75,
+      "learning_rate": 2.2383201091136965e-07,
+      "loss": 0.8257,
+      "step": 6654
+    },
+    {
+      "epoch": 1.6180403598346706,
+      "grad_norm": 17.5,
+      "learning_rate": 2.2355613385215538e-07,
+      "loss": 0.6513,
+      "step": 6655
+    },
+    {
+      "epoch": 1.6182834913688304,
+      "grad_norm": 17.0,
+      "learning_rate": 2.2328041021270467e-07,
+      "loss": 0.739,
+      "step": 6656
+    },
+    {
+      "epoch": 1.6185266229029907,
+      "grad_norm": 30.875,
+      "learning_rate": 2.230048400342283e-07,
+      "loss": 1.4384,
+      "step": 6657
+    },
+    {
+      "epoch": 1.6187697544371504,
+      "grad_norm": 26.875,
+      "learning_rate": 2.2272942335791566e-07,
+      "loss": 0.8301,
+      "step": 6658
+    },
+    {
+      "epoch": 1.6190128859713104,
+      "grad_norm": 21.75,
+      "learning_rate": 2.2245416022493236e-07,
+      "loss": 0.6196,
+      "step": 6659
+    },
+    {
+      "epoch": 1.6192560175054704,
+      "grad_norm": 16.625,
+      "learning_rate": 2.2217905067642083e-07,
+      "loss": 0.9399,
+      "step": 6660
+    },
+    {
+      "epoch": 1.6194991490396304,
+      "grad_norm": 15.8125,
+      "learning_rate": 2.2190409475350158e-07,
+      "loss": 0.5162,
+      "step": 6661
+    },
+    {
+      "epoch": 1.6197422805737904,
+      "grad_norm": 18.25,
+      "learning_rate": 2.2162929249727087e-07,
+      "loss": 0.7829,
+      "step": 6662
+    },
+    {
+      "epoch": 1.6199854121079504,
+      "grad_norm": 20.125,
+      "learning_rate": 2.213546439488033e-07,
+      "loss": 0.7781,
+      "step": 6663
+    },
+    {
+      "epoch": 1.6202285436421104,
+      "grad_norm": 16.5,
+      "learning_rate": 2.210801491491492e-07,
+      "loss": 0.3432,
+      "step": 6664
+    },
+    {
+      "epoch": 1.6204716751762702,
+      "grad_norm": 18.75,
+      "learning_rate": 2.2080580813933717e-07,
+      "loss": 0.717,
+      "step": 6665
+    },
+    {
+      "epoch": 1.6207148067104304,
+      "grad_norm": 18.375,
+      "learning_rate": 2.205316209603718e-07,
+      "loss": 0.6753,
+      "step": 6666
+    },
+    {
+      "epoch": 1.6209579382445902,
+      "grad_norm": 23.25,
+      "learning_rate": 2.2025758765323506e-07,
+      "loss": 0.7017,
+      "step": 6667
+    },
+    {
+      "epoch": 1.6212010697787504,
+      "grad_norm": 20.125,
+      "learning_rate": 2.199837082588864e-07,
+      "loss": 0.6525,
+      "step": 6668
+    },
+    {
+      "epoch": 1.6214442013129102,
+      "grad_norm": 19.375,
+      "learning_rate": 2.197099828182614e-07,
+      "loss": 0.3315,
+      "step": 6669
+    },
+    {
+      "epoch": 1.6216873328470702,
+      "grad_norm": 141.0,
+      "learning_rate": 2.19436411372273e-07,
+      "loss": 0.8343,
+      "step": 6670
+    },
+    {
+      "epoch": 1.6219304643812302,
+      "grad_norm": 22.5,
+      "learning_rate": 2.1916299396181146e-07,
+      "loss": 1.3092,
+      "step": 6671
+    },
+    {
+      "epoch": 1.6221735959153902,
+      "grad_norm": 17.0,
+      "learning_rate": 2.1888973062774376e-07,
+      "loss": 0.4956,
+      "step": 6672
+    },
+    {
+      "epoch": 1.6224167274495502,
+      "grad_norm": 16.375,
+      "learning_rate": 2.186166214109138e-07,
+      "loss": 0.4634,
+      "step": 6673
+    },
+    {
+      "epoch": 1.6226598589837102,
+      "grad_norm": 19.875,
+      "learning_rate": 2.183436663521421e-07,
+      "loss": 0.6935,
+      "step": 6674
+    },
+    {
+      "epoch": 1.6229029905178702,
+      "grad_norm": 19.375,
+      "learning_rate": 2.18070865492227e-07,
+      "loss": 0.7186,
+      "step": 6675
+    },
+    {
+      "epoch": 1.62314612205203,
+      "grad_norm": 24.75,
+      "learning_rate": 2.1779821887194323e-07,
+      "loss": 1.0167,
+      "step": 6676
+    },
+    {
+      "epoch": 1.6233892535861902,
+      "grad_norm": 28.875,
+      "learning_rate": 2.1752572653204198e-07,
+      "loss": 1.1666,
+      "step": 6677
+    },
+    {
+      "epoch": 1.62363238512035,
+      "grad_norm": 20.875,
+      "learning_rate": 2.172533885132523e-07,
+      "loss": 0.7182,
+      "step": 6678
+    },
+    {
+      "epoch": 1.6238755166545102,
+      "grad_norm": 27.625,
+      "learning_rate": 2.1698120485628027e-07,
+      "loss": 1.1638,
+      "step": 6679
+    },
+    {
+      "epoch": 1.62411864818867,
+      "grad_norm": 31.875,
+      "learning_rate": 2.1670917560180805e-07,
+      "loss": 0.8286,
+      "step": 6680
+    },
+    {
+      "epoch": 1.62436177972283,
+      "grad_norm": 30.875,
+      "learning_rate": 2.1643730079049463e-07,
+      "loss": 1.0799,
+      "step": 6681
+    },
+    {
+      "epoch": 1.62460491125699,
+      "grad_norm": 23.0,
+      "learning_rate": 2.1616558046297724e-07,
+      "loss": 0.9894,
+      "step": 6682
+    },
+    {
+      "epoch": 1.62484804279115,
+      "grad_norm": 14.6875,
+      "learning_rate": 2.1589401465986883e-07,
+      "loss": 0.1656,
+      "step": 6683
+    },
+    {
+      "epoch": 1.62509117432531,
+      "grad_norm": 27.875,
+      "learning_rate": 2.1562260342175913e-07,
+      "loss": 0.8089,
+      "step": 6684
+    },
+    {
+      "epoch": 1.6253343058594698,
+      "grad_norm": 22.125,
+      "learning_rate": 2.1535134678921585e-07,
+      "loss": 0.9627,
+      "step": 6685
+    },
+    {
+      "epoch": 1.62557743739363,
+      "grad_norm": 21.5,
+      "learning_rate": 2.1508024480278292e-07,
+      "loss": 0.7536,
+      "step": 6686
+    },
+    {
+      "epoch": 1.6258205689277898,
+      "grad_norm": 21.375,
+      "learning_rate": 2.1480929750298126e-07,
+      "loss": 0.5848,
+      "step": 6687
+    },
+    {
+      "epoch": 1.62606370046195,
+      "grad_norm": 23.0,
+      "learning_rate": 2.1453850493030795e-07,
+      "loss": 1.0846,
+      "step": 6688
+    },
+    {
+      "epoch": 1.6263068319961098,
+      "grad_norm": 16.375,
+      "learning_rate": 2.142678671252385e-07,
+      "loss": 0.8118,
+      "step": 6689
+    },
+    {
+      "epoch": 1.62654996353027,
+      "grad_norm": 21.375,
+      "learning_rate": 2.1399738412822406e-07,
+      "loss": 0.5555,
+      "step": 6690
+    },
+    {
+      "epoch": 1.6267930950644298,
+      "grad_norm": 26.0,
+      "learning_rate": 2.1372705597969266e-07,
+      "loss": 1.2868,
+      "step": 6691
+    },
+    {
+      "epoch": 1.6270362265985898,
+      "grad_norm": 22.625,
+      "learning_rate": 2.1345688272005005e-07,
+      "loss": 0.6446,
+      "step": 6692
+    },
+    {
+      "epoch": 1.6272793581327498,
+      "grad_norm": 21.125,
+      "learning_rate": 2.131868643896777e-07,
+      "loss": 0.5819,
+      "step": 6693
+    },
+    {
+      "epoch": 1.6275224896669098,
+      "grad_norm": 22.125,
+      "learning_rate": 2.1291700102893526e-07,
+      "loss": 0.6422,
+      "step": 6694
+    },
+    {
+      "epoch": 1.6277656212010698,
+      "grad_norm": 20.625,
+      "learning_rate": 2.1264729267815767e-07,
+      "loss": 0.6746,
+      "step": 6695
+    },
+    {
+      "epoch": 1.6280087527352296,
+      "grad_norm": 18.0,
+      "learning_rate": 2.1237773937765827e-07,
+      "loss": 0.5748,
+      "step": 6696
+    },
+    {
+      "epoch": 1.6282518842693898,
+      "grad_norm": 18.375,
+      "learning_rate": 2.1210834116772605e-07,
+      "loss": 0.6676,
+      "step": 6697
+    },
+    {
+      "epoch": 1.6284950158035496,
+      "grad_norm": 19.375,
+      "learning_rate": 2.1183909808862686e-07,
+      "loss": 0.6756,
+      "step": 6698
+    },
+    {
+      "epoch": 1.6287381473377098,
+      "grad_norm": 30.875,
+      "learning_rate": 2.1157001018060452e-07,
+      "loss": 0.7719,
+      "step": 6699
+    },
+    {
+      "epoch": 1.6289812788718696,
+      "grad_norm": 15.75,
+      "learning_rate": 2.1130107748387849e-07,
+      "loss": 0.7149,
+      "step": 6700
+    },
+    {
+      "epoch": 1.6292244104060296,
+      "grad_norm": 18.25,
+      "learning_rate": 2.1103230003864467e-07,
+      "loss": 0.7606,
+      "step": 6701
+    },
+    {
+      "epoch": 1.6294675419401896,
+      "grad_norm": 28.875,
+      "learning_rate": 2.1076367788507787e-07,
+      "loss": 0.666,
+      "step": 6702
+    },
+    {
+      "epoch": 1.6297106734743496,
+      "grad_norm": 38.0,
+      "learning_rate": 2.1049521106332752e-07,
+      "loss": 0.7115,
+      "step": 6703
+    },
+    {
+      "epoch": 1.6299538050085096,
+      "grad_norm": 22.0,
+      "learning_rate": 2.1022689961352094e-07,
+      "loss": 1.1008,
+      "step": 6704
+    },
+    {
+      "epoch": 1.6301969365426696,
+      "grad_norm": 19.625,
+      "learning_rate": 2.0995874357576136e-07,
+      "loss": 0.5102,
+      "step": 6705
+    },
+    {
+      "epoch": 1.6304400680768296,
+      "grad_norm": 31.75,
+      "learning_rate": 2.0969074299012986e-07,
+      "loss": 0.5456,
+      "step": 6706
+    },
+    {
+      "epoch": 1.6306831996109894,
+      "grad_norm": 20.875,
+      "learning_rate": 2.0942289789668377e-07,
+      "loss": 0.8614,
+      "step": 6707
+    },
+    {
+      "epoch": 1.6309263311451496,
+      "grad_norm": 16.875,
+      "learning_rate": 2.0915520833545658e-07,
+      "loss": 0.6767,
+      "step": 6708
+    },
+    {
+      "epoch": 1.6311694626793094,
+      "grad_norm": 22.125,
+      "learning_rate": 2.088876743464595e-07,
+      "loss": 0.5842,
+      "step": 6709
+    },
+    {
+      "epoch": 1.6314125942134696,
+      "grad_norm": 15.375,
+      "learning_rate": 2.0862029596968036e-07,
+      "loss": 0.3612,
+      "step": 6710
+    },
+    {
+      "epoch": 1.6316557257476294,
+      "grad_norm": 27.0,
+      "learning_rate": 2.0835307324508336e-07,
+      "loss": 0.5747,
+      "step": 6711
+    },
+    {
+      "epoch": 1.6318988572817894,
+      "grad_norm": 23.625,
+      "learning_rate": 2.0808600621260916e-07,
+      "loss": 1.1054,
+      "step": 6712
+    },
+    {
+      "epoch": 1.6321419888159494,
+      "grad_norm": 27.5,
+      "learning_rate": 2.0781909491217596e-07,
+      "loss": 0.7226,
+      "step": 6713
+    },
+    {
+      "epoch": 1.6323851203501094,
+      "grad_norm": 18.375,
+      "learning_rate": 2.0755233938367815e-07,
+      "loss": 0.3595,
+      "step": 6714
+    },
+    {
+      "epoch": 1.6326282518842694,
+      "grad_norm": 20.75,
+      "learning_rate": 2.0728573966698667e-07,
+      "loss": 0.9376,
+      "step": 6715
+    },
+    {
+      "epoch": 1.6328713834184294,
+      "grad_norm": 19.375,
+      "learning_rate": 2.0701929580194972e-07,
+      "loss": 0.7214,
+      "step": 6716
+    },
+    {
+      "epoch": 1.6331145149525894,
+      "grad_norm": 23.75,
+      "learning_rate": 2.0675300782839218e-07,
+      "loss": 0.9348,
+      "step": 6717
+    },
+    {
+      "epoch": 1.6333576464867492,
+      "grad_norm": 15.4375,
+      "learning_rate": 2.064868757861148e-07,
+      "loss": 0.516,
+      "step": 6718
+    },
+    {
+      "epoch": 1.6336007780209094,
+      "grad_norm": 23.25,
+      "learning_rate": 2.0622089971489624e-07,
+      "loss": 0.9958,
+      "step": 6719
+    },
+    {
+      "epoch": 1.6338439095550692,
+      "grad_norm": 19.125,
+      "learning_rate": 2.0595507965449083e-07,
+      "loss": 0.6534,
+      "step": 6720
+    },
+    {
+      "epoch": 1.6340870410892294,
+      "grad_norm": 19.375,
+      "learning_rate": 2.0568941564463014e-07,
+      "loss": 0.637,
+      "step": 6721
+    },
+    {
+      "epoch": 1.6343301726233892,
+      "grad_norm": 20.75,
+      "learning_rate": 2.0542390772502183e-07,
+      "loss": 0.737,
+      "step": 6722
+    },
+    {
+      "epoch": 1.6345733041575492,
+      "grad_norm": 19.625,
+      "learning_rate": 2.0515855593535124e-07,
+      "loss": 0.5851,
+      "step": 6723
+    },
+    {
+      "epoch": 1.6348164356917092,
+      "grad_norm": 23.25,
+      "learning_rate": 2.048933603152793e-07,
+      "loss": 1.1741,
+      "step": 6724
+    },
+    {
+      "epoch": 1.6350595672258692,
+      "grad_norm": 19.0,
+      "learning_rate": 2.0462832090444417e-07,
+      "loss": 0.5789,
+      "step": 6725
+    },
+    {
+      "epoch": 1.6353026987600292,
+      "grad_norm": 20.625,
+      "learning_rate": 2.0436343774246094e-07,
+      "loss": 0.5312,
+      "step": 6726
+    },
+    {
+      "epoch": 1.6355458302941892,
+      "grad_norm": 18.875,
+      "learning_rate": 2.0409871086892087e-07,
+      "loss": 0.6926,
+      "step": 6727
+    },
+    {
+      "epoch": 1.6357889618283492,
+      "grad_norm": 19.875,
+      "learning_rate": 2.0383414032339162e-07,
+      "loss": 0.6181,
+      "step": 6728
+    },
+    {
+      "epoch": 1.636032093362509,
+      "grad_norm": 22.625,
+      "learning_rate": 2.0356972614541782e-07,
+      "loss": 0.7802,
+      "step": 6729
+    },
+    {
+      "epoch": 1.6362752248966692,
+      "grad_norm": 29.375,
+      "learning_rate": 2.033054683745213e-07,
+      "loss": 0.7724,
+      "step": 6730
+    },
+    {
+      "epoch": 1.636518356430829,
+      "grad_norm": 22.375,
+      "learning_rate": 2.0304136705019933e-07,
+      "loss": 0.9762,
+      "step": 6731
+    },
+    {
+      "epoch": 1.6367614879649892,
+      "grad_norm": 20.0,
+      "learning_rate": 2.0277742221192623e-07,
+      "loss": 0.9337,
+      "step": 6732
+    },
+    {
+      "epoch": 1.637004619499149,
+      "grad_norm": 18.25,
+      "learning_rate": 2.0251363389915414e-07,
+      "loss": 0.831,
+      "step": 6733
+    },
+    {
+      "epoch": 1.637247751033309,
+      "grad_norm": 17.5,
+      "learning_rate": 2.022500021513102e-07,
+      "loss": 0.8069,
+      "step": 6734
+    },
+    {
+      "epoch": 1.637490882567469,
+      "grad_norm": 27.125,
+      "learning_rate": 2.0198652700779823e-07,
+      "loss": 0.8059,
+      "step": 6735
+    },
+    {
+      "epoch": 1.637734014101629,
+      "grad_norm": 22.125,
+      "learning_rate": 2.017232085080001e-07,
+      "loss": 0.4712,
+      "step": 6736
+    },
+    {
+      "epoch": 1.637977145635789,
+      "grad_norm": 20.5,
+      "learning_rate": 2.0146004669127279e-07,
+      "loss": 0.8741,
+      "step": 6737
+    },
+    {
+      "epoch": 1.6382202771699488,
+      "grad_norm": 20.375,
+      "learning_rate": 2.0119704159695038e-07,
+      "loss": 0.4799,
+      "step": 6738
+    },
+    {
+      "epoch": 1.638463408704109,
+      "grad_norm": 20.0,
+      "learning_rate": 2.0093419326434326e-07,
+      "loss": 0.6843,
+      "step": 6739
+    },
+    {
+      "epoch": 1.6387065402382688,
+      "grad_norm": 15.6875,
+      "learning_rate": 2.0067150173273888e-07,
+      "loss": 0.3769,
+      "step": 6740
+    },
+    {
+      "epoch": 1.638949671772429,
+      "grad_norm": 19.5,
+      "learning_rate": 2.0040896704140153e-07,
+      "loss": 0.7515,
+      "step": 6741
+    },
+    {
+      "epoch": 1.6391928033065888,
+      "grad_norm": 21.875,
+      "learning_rate": 2.0014658922957077e-07,
+      "loss": 0.4303,
+      "step": 6742
+    },
+    {
+      "epoch": 1.639435934840749,
+      "grad_norm": 34.5,
+      "learning_rate": 1.9988436833646415e-07,
+      "loss": 1.0822,
+      "step": 6743
+    },
+    {
+      "epoch": 1.6396790663749088,
+      "grad_norm": 28.5,
+      "learning_rate": 1.9962230440127474e-07,
+      "loss": 1.1679,
+      "step": 6744
+    },
+    {
+      "epoch": 1.6399221979090688,
+      "grad_norm": 26.625,
+      "learning_rate": 1.9936039746317261e-07,
+      "loss": 1.0491,
+      "step": 6745
+    },
+    {
+      "epoch": 1.6401653294432288,
+      "grad_norm": 21.625,
+      "learning_rate": 1.9909864756130395e-07,
+      "loss": 0.8999,
+      "step": 6746
+    },
+    {
+      "epoch": 1.6404084609773888,
+      "grad_norm": 19.25,
+      "learning_rate": 1.9883705473479217e-07,
+      "loss": 0.5207,
+      "step": 6747
+    },
+    {
+      "epoch": 1.6406515925115488,
+      "grad_norm": 16.625,
+      "learning_rate": 1.985756190227371e-07,
+      "loss": 0.7254,
+      "step": 6748
+    },
+    {
+      "epoch": 1.6408947240457086,
+      "grad_norm": 16.625,
+      "learning_rate": 1.9831434046421405e-07,
+      "loss": 0.6691,
+      "step": 6749
+    },
+    {
+      "epoch": 1.6411378555798688,
+      "grad_norm": 18.375,
+      "learning_rate": 1.9805321909827645e-07,
+      "loss": 0.6687,
+      "step": 6750
+    },
+    {
+      "epoch": 1.6413809871140286,
+      "grad_norm": 19.875,
+      "learning_rate": 1.9779225496395298e-07,
+      "loss": 0.7149,
+      "step": 6751
+    },
+    {
+      "epoch": 1.6416241186481888,
+      "grad_norm": 20.0,
+      "learning_rate": 1.9753144810024937e-07,
+      "loss": 1.0038,
+      "step": 6752
+    },
+    {
+      "epoch": 1.6418672501823486,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9727079854614723e-07,
+      "loss": 0.7706,
+      "step": 6753
+    },
+    {
+      "epoch": 1.6421103817165086,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.9701030634060578e-07,
+      "loss": 0.4311,
+      "step": 6754
+    },
+    {
+      "epoch": 1.6423535132506686,
+      "grad_norm": 18.125,
+      "learning_rate": 1.9674997152255944e-07,
+      "loss": 0.4856,
+      "step": 6755
+    },
+    {
+      "epoch": 1.6425966447848286,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9648979413092017e-07,
+      "loss": 0.6617,
+      "step": 6756
+    },
+    {
+      "epoch": 1.6428397763189886,
+      "grad_norm": 18.625,
+      "learning_rate": 1.9622977420457628e-07,
+      "loss": 0.5935,
+      "step": 6757
+    },
+    {
+      "epoch": 1.6430829078531486,
+      "grad_norm": 25.0,
+      "learning_rate": 1.9596991178239183e-07,
+      "loss": 0.6934,
+      "step": 6758
+    },
+    {
+      "epoch": 1.6433260393873086,
+      "grad_norm": 22.125,
+      "learning_rate": 1.9571020690320754e-07,
+      "loss": 0.8906,
+      "step": 6759
+    },
+    {
+      "epoch": 1.6435691709214684,
+      "grad_norm": 20.375,
+      "learning_rate": 1.9545065960584127e-07,
+      "loss": 1.0914,
+      "step": 6760
+    },
+    {
+      "epoch": 1.6438123024556286,
+      "grad_norm": 19.125,
+      "learning_rate": 1.951912699290867e-07,
+      "loss": 0.7872,
+      "step": 6761
+    },
+    {
+      "epoch": 1.6440554339897884,
+      "grad_norm": 20.75,
+      "learning_rate": 1.9493203791171393e-07,
+      "loss": 0.5704,
+      "step": 6762
+    },
+    {
+      "epoch": 1.6442985655239486,
+      "grad_norm": 22.75,
+      "learning_rate": 1.946729635924692e-07,
+      "loss": 0.988,
+      "step": 6763
+    },
+    {
+      "epoch": 1.6445416970581084,
+      "grad_norm": 21.5,
+      "learning_rate": 1.944140470100768e-07,
+      "loss": 0.6622,
+      "step": 6764
+    },
+    {
+      "epoch": 1.6447848285922684,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9415528820323562e-07,
+      "loss": 0.744,
+      "step": 6765
+    },
+    {
+      "epoch": 1.6450279601264284,
+      "grad_norm": 24.25,
+      "learning_rate": 1.9389668721062142e-07,
+      "loss": 0.609,
+      "step": 6766
+    },
+    {
+      "epoch": 1.6452710916605884,
+      "grad_norm": 20.625,
+      "learning_rate": 1.9363824407088714e-07,
+      "loss": 0.7015,
+      "step": 6767
+    },
+    {
+      "epoch": 1.6455142231947484,
+      "grad_norm": 19.875,
+      "learning_rate": 1.9337995882266133e-07,
+      "loss": 0.5253,
+      "step": 6768
+    },
+    {
+      "epoch": 1.6457573547289084,
+      "grad_norm": 21.625,
+      "learning_rate": 1.931218315045491e-07,
+      "loss": 0.5222,
+      "step": 6769
+    },
+    {
+      "epoch": 1.6460004862630684,
+      "grad_norm": 20.125,
+      "learning_rate": 1.9286386215513139e-07,
+      "loss": 0.733,
+      "step": 6770
+    },
+    {
+      "epoch": 1.6462436177972282,
+      "grad_norm": 23.75,
+      "learning_rate": 1.9260605081296763e-07,
+      "loss": 0.6605,
+      "step": 6771
+    },
+    {
+      "epoch": 1.6464867493313884,
+      "grad_norm": 19.875,
+      "learning_rate": 1.923483975165913e-07,
+      "loss": 0.8826,
+      "step": 6772
+    },
+    {
+      "epoch": 1.6467298808655482,
+      "grad_norm": 18.875,
+      "learning_rate": 1.9209090230451283e-07,
+      "loss": 0.7168,
+      "step": 6773
+    },
+    {
+      "epoch": 1.6469730123997084,
+      "grad_norm": 19.25,
+      "learning_rate": 1.918335652152202e-07,
+      "loss": 0.823,
+      "step": 6774
+    },
+    {
+      "epoch": 1.6472161439338682,
+      "grad_norm": 20.75,
+      "learning_rate": 1.915763862871764e-07,
+      "loss": 0.7573,
+      "step": 6775
+    },
+    {
+      "epoch": 1.6474592754680282,
+      "grad_norm": 18.5,
+      "learning_rate": 1.9131936555882093e-07,
+      "loss": 0.4875,
+      "step": 6776
+    },
+    {
+      "epoch": 1.6477024070021882,
+      "grad_norm": 24.375,
+      "learning_rate": 1.910625030685706e-07,
+      "loss": 0.6212,
+      "step": 6777
+    },
+    {
+      "epoch": 1.6479455385363482,
+      "grad_norm": 19.125,
+      "learning_rate": 1.9080579885481745e-07,
+      "loss": 0.6861,
+      "step": 6778
+    },
+    {
+      "epoch": 1.6481886700705082,
+      "grad_norm": 23.5,
+      "learning_rate": 1.905492529559308e-07,
+      "loss": 1.1977,
+      "step": 6779
+    },
+    {
+      "epoch": 1.6484318016046682,
+      "grad_norm": 20.375,
+      "learning_rate": 1.902928654102554e-07,
+      "loss": 0.4884,
+      "step": 6780
+    },
+    {
+      "epoch": 1.6486749331388282,
+      "grad_norm": 20.0,
+      "learning_rate": 1.900366362561132e-07,
+      "loss": 0.6032,
+      "step": 6781
+    },
+    {
+      "epoch": 1.648918064672988,
+      "grad_norm": 19.125,
+      "learning_rate": 1.8978056553180205e-07,
+      "loss": 0.4183,
+      "step": 6782
+    },
+    {
+      "epoch": 1.6491611962071482,
+      "grad_norm": 15.25,
+      "learning_rate": 1.8952465327559565e-07,
+      "loss": 0.4614,
+      "step": 6783
+    },
+    {
+      "epoch": 1.649404327741308,
+      "grad_norm": 22.125,
+      "learning_rate": 1.8926889952574523e-07,
+      "loss": 0.8141,
+      "step": 6784
+    },
+    {
+      "epoch": 1.6496474592754682,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.8901330432047713e-07,
+      "loss": 0.3837,
+      "step": 6785
+    },
+    {
+      "epoch": 1.649890590809628,
+      "grad_norm": 20.125,
+      "learning_rate": 1.887578676979944e-07,
+      "loss": 0.6217,
+      "step": 6786
+    },
+    {
+      "epoch": 1.650133722343788,
+      "grad_norm": 18.625,
+      "learning_rate": 1.885025896964765e-07,
+      "loss": 0.6743,
+      "step": 6787
+    },
+    {
+      "epoch": 1.650376853877948,
+      "grad_norm": 24.0,
+      "learning_rate": 1.8824747035407972e-07,
+      "loss": 1.2319,
+      "step": 6788
+    },
+    {
+      "epoch": 1.650619985412108,
+      "grad_norm": 33.25,
+      "learning_rate": 1.879925097089355e-07,
+      "loss": 1.034,
+      "step": 6789
+    },
+    {
+      "epoch": 1.650863116946268,
+      "grad_norm": 23.875,
+      "learning_rate": 1.8773770779915185e-07,
+      "loss": 0.8589,
+      "step": 6790
+    },
+    {
+      "epoch": 1.6511062484804278,
+      "grad_norm": 24.625,
+      "learning_rate": 1.8748306466281411e-07,
+      "loss": 0.8367,
+      "step": 6791
+    },
+    {
+      "epoch": 1.651349380014588,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8722858033798253e-07,
+      "loss": 0.6895,
+      "step": 6792
+    },
+    {
+      "epoch": 1.6515925115487478,
+      "grad_norm": 20.5,
+      "learning_rate": 1.8697425486269404e-07,
+      "loss": 0.7425,
+      "step": 6793
+    },
+    {
+      "epoch": 1.651835643082908,
+      "grad_norm": 22.375,
+      "learning_rate": 1.867200882749623e-07,
+      "loss": 0.8686,
+      "step": 6794
+    },
+    {
+      "epoch": 1.6520787746170678,
+      "grad_norm": 22.625,
+      "learning_rate": 1.8646608061277717e-07,
+      "loss": 0.705,
+      "step": 6795
+    },
+    {
+      "epoch": 1.6523219061512278,
+      "grad_norm": 20.125,
+      "learning_rate": 1.862122319141041e-07,
+      "loss": 0.5989,
+      "step": 6796
+    },
+    {
+      "epoch": 1.6525650376853878,
+      "grad_norm": 21.125,
+      "learning_rate": 1.8595854221688497e-07,
+      "loss": 0.685,
+      "step": 6797
+    },
+    {
+      "epoch": 1.6528081692195478,
+      "grad_norm": 22.625,
+      "learning_rate": 1.857050115590385e-07,
+      "loss": 0.6797,
+      "step": 6798
+    },
+    {
+      "epoch": 1.6530513007537078,
+      "grad_norm": 20.625,
+      "learning_rate": 1.8545163997845921e-07,
+      "loss": 0.7317,
+      "step": 6799
+    },
+    {
+      "epoch": 1.6532944322878678,
+      "grad_norm": 21.125,
+      "learning_rate": 1.851984275130174e-07,
+      "loss": 0.5705,
+      "step": 6800
+    },
+    {
+      "epoch": 1.6535375638220278,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.8494537420056038e-07,
+      "loss": 0.6592,
+      "step": 6801
+    },
+    {
+      "epoch": 1.6537806953561875,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8469248007891154e-07,
+      "loss": 0.5066,
+      "step": 6802
+    },
+    {
+      "epoch": 1.6540238268903478,
+      "grad_norm": 20.125,
+      "learning_rate": 1.844397451858701e-07,
+      "loss": 0.8888,
+      "step": 6803
+    },
+    {
+      "epoch": 1.6542669584245075,
+      "grad_norm": 17.5,
+      "learning_rate": 1.8418716955921124e-07,
+      "loss": 0.7053,
+      "step": 6804
+    },
+    {
+      "epoch": 1.6545100899586678,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8393475323668739e-07,
+      "loss": 0.3783,
+      "step": 6805
+    },
+    {
+      "epoch": 1.6547532214928276,
+      "grad_norm": 21.875,
+      "learning_rate": 1.836824962560263e-07,
+      "loss": 1.0748,
+      "step": 6806
+    },
+    {
+      "epoch": 1.6549963530269876,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.8343039865493184e-07,
+      "loss": 0.6374,
+      "step": 6807
+    },
+    {
+      "epoch": 1.6552394845611476,
+      "grad_norm": 20.0,
+      "learning_rate": 1.8317846047108484e-07,
+      "loss": 0.546,
+      "step": 6808
+    },
+    {
+      "epoch": 1.6554826160953076,
+      "grad_norm": 19.625,
+      "learning_rate": 1.829266817421413e-07,
+      "loss": 0.7105,
+      "step": 6809
+    },
+    {
+      "epoch": 1.6557257476294676,
+      "grad_norm": 25.875,
+      "learning_rate": 1.8267506250573441e-07,
+      "loss": 0.9099,
+      "step": 6810
+    },
+    {
+      "epoch": 1.6559688791636276,
+      "grad_norm": 18.5,
+      "learning_rate": 1.824236027994726e-07,
+      "loss": 0.7669,
+      "step": 6811
+    },
+    {
+      "epoch": 1.6562120106977876,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8217230266094122e-07,
+      "loss": 0.8342,
+      "step": 6812
+    },
+    {
+      "epoch": 1.6564551422319473,
+      "grad_norm": 17.125,
+      "learning_rate": 1.8192116212770116e-07,
+      "loss": 0.4895,
+      "step": 6813
+    },
+    {
+      "epoch": 1.6566982737661076,
+      "grad_norm": 27.25,
+      "learning_rate": 1.8167018123728967e-07,
+      "loss": 0.4567,
+      "step": 6814
+    },
+    {
+      "epoch": 1.6569414053002673,
+      "grad_norm": 23.625,
+      "learning_rate": 1.8141936002722044e-07,
+      "loss": 0.9059,
+      "step": 6815
+    },
+    {
+      "epoch": 1.6571845368344276,
+      "grad_norm": 17.0,
+      "learning_rate": 1.8116869853498295e-07,
+      "loss": 0.5689,
+      "step": 6816
+    },
+    {
+      "epoch": 1.6574276683685873,
+      "grad_norm": 21.625,
+      "learning_rate": 1.8091819679804252e-07,
+      "loss": 0.9049,
+      "step": 6817
+    },
+    {
+      "epoch": 1.6576707999027473,
+      "grad_norm": 21.125,
+      "learning_rate": 1.8066785485384125e-07,
+      "loss": 0.7582,
+      "step": 6818
+    },
+    {
+      "epoch": 1.6579139314369074,
+      "grad_norm": 23.375,
+      "learning_rate": 1.804176727397973e-07,
+      "loss": 1.0467,
+      "step": 6819
+    },
+    {
+      "epoch": 1.6581570629710674,
+      "grad_norm": 17.5,
+      "learning_rate": 1.801676504933046e-07,
+      "loss": 0.659,
+      "step": 6820
+    },
+    {
+      "epoch": 1.6584001945052274,
+      "grad_norm": 17.75,
+      "learning_rate": 1.7991778815173278e-07,
+      "loss": 0.4853,
+      "step": 6821
+    },
+    {
+      "epoch": 1.6586433260393874,
+      "grad_norm": 19.75,
+      "learning_rate": 1.7966808575242883e-07,
+      "loss": 0.5262,
+      "step": 6822
+    },
+    {
+      "epoch": 1.6588864575735474,
+      "grad_norm": 18.25,
+      "learning_rate": 1.7941854333271474e-07,
+      "loss": 0.5657,
+      "step": 6823
+    },
+    {
+      "epoch": 1.6591295891077071,
+      "grad_norm": 17.0,
+      "learning_rate": 1.791691609298886e-07,
+      "loss": 0.4819,
+      "step": 6824
+    },
+    {
+      "epoch": 1.6593727206418674,
+      "grad_norm": 17.125,
+      "learning_rate": 1.7891993858122528e-07,
+      "loss": 0.6681,
+      "step": 6825
+    },
+    {
+      "epoch": 1.6596158521760271,
+      "grad_norm": 22.125,
+      "learning_rate": 1.786708763239757e-07,
+      "loss": 0.4846,
+      "step": 6826
+    },
+    {
+      "epoch": 1.6598589837101874,
+      "grad_norm": 22.375,
+      "learning_rate": 1.7842197419536605e-07,
+      "loss": 0.8214,
+      "step": 6827
+    },
+    {
+      "epoch": 1.6601021152443471,
+      "grad_norm": 23.125,
+      "learning_rate": 1.781732322325988e-07,
+      "loss": 0.7494,
+      "step": 6828
+    },
+    {
+      "epoch": 1.6603452467785071,
+      "grad_norm": 17.5,
+      "learning_rate": 1.7792465047285352e-07,
+      "loss": 0.724,
+      "step": 6829
+    },
+    {
+      "epoch": 1.6605883783126671,
+      "grad_norm": 23.0,
+      "learning_rate": 1.7767622895328448e-07,
+      "loss": 1.0386,
+      "step": 6830
+    },
+    {
+      "epoch": 1.6608315098468271,
+      "grad_norm": 18.0,
+      "learning_rate": 1.7742796771102257e-07,
+      "loss": 0.378,
+      "step": 6831
+    },
+    {
+      "epoch": 1.6610746413809871,
+      "grad_norm": 23.875,
+      "learning_rate": 1.7717986678317465e-07,
+      "loss": 0.6293,
+      "step": 6832
+    },
+    {
+      "epoch": 1.661317772915147,
+      "grad_norm": 18.25,
+      "learning_rate": 1.7693192620682433e-07,
+      "loss": 0.9571,
+      "step": 6833
+    },
+    {
+      "epoch": 1.6615609044493072,
+      "grad_norm": 22.75,
+      "learning_rate": 1.7668414601903007e-07,
+      "loss": 0.8777,
+      "step": 6834
+    },
+    {
+      "epoch": 1.661804035983467,
+      "grad_norm": 19.125,
+      "learning_rate": 1.764365262568267e-07,
+      "loss": 0.5456,
+      "step": 6835
+    },
+    {
+      "epoch": 1.6620471675176272,
+      "grad_norm": 20.125,
+      "learning_rate": 1.7618906695722597e-07,
+      "loss": 1.2321,
+      "step": 6836
+    },
+    {
+      "epoch": 1.662290299051787,
+      "grad_norm": 27.375,
+      "learning_rate": 1.759417681572144e-07,
+      "loss": 0.4281,
+      "step": 6837
+    },
+    {
+      "epoch": 1.6625334305859472,
+      "grad_norm": 31.125,
+      "learning_rate": 1.7569462989375493e-07,
+      "loss": 0.641,
+      "step": 6838
+    },
+    {
+      "epoch": 1.662776562120107,
+      "grad_norm": 29.25,
+      "learning_rate": 1.7544765220378734e-07,
+      "loss": 0.7262,
+      "step": 6839
+    },
+    {
+      "epoch": 1.663019693654267,
+      "grad_norm": 23.0,
+      "learning_rate": 1.7520083512422584e-07,
+      "loss": 1.0293,
+      "step": 6840
+    },
+    {
+      "epoch": 1.663262825188427,
+      "grad_norm": 17.875,
+      "learning_rate": 1.749541786919623e-07,
+      "loss": 0.4684,
+      "step": 6841
+    },
+    {
+      "epoch": 1.663505956722587,
+      "grad_norm": 23.5,
+      "learning_rate": 1.747076829438632e-07,
+      "loss": 0.5234,
+      "step": 6842
+    },
+    {
+      "epoch": 1.663749088256747,
+      "grad_norm": 24.5,
+      "learning_rate": 1.7446134791677214e-07,
+      "loss": 1.2282,
+      "step": 6843
+    },
+    {
+      "epoch": 1.6639922197909067,
+      "grad_norm": 18.375,
+      "learning_rate": 1.7421517364750771e-07,
+      "loss": 0.8967,
+      "step": 6844
+    },
+    {
+      "epoch": 1.664235351325067,
+      "grad_norm": 20.5,
+      "learning_rate": 1.7396916017286477e-07,
+      "loss": 0.7396,
+      "step": 6845
+    },
+    {
+      "epoch": 1.6644784828592267,
+      "grad_norm": 19.5,
+      "learning_rate": 1.7372330752961492e-07,
+      "loss": 0.7407,
+      "step": 6846
+    },
+    {
+      "epoch": 1.664721614393387,
+      "grad_norm": 16.0,
+      "learning_rate": 1.7347761575450457e-07,
+      "loss": 0.5837,
+      "step": 6847
+    },
+    {
+      "epoch": 1.6649647459275467,
+      "grad_norm": 16.625,
+      "learning_rate": 1.7323208488425656e-07,
+      "loss": 0.6942,
+      "step": 6848
+    },
+    {
+      "epoch": 1.6652078774617067,
+      "grad_norm": 19.5,
+      "learning_rate": 1.7298671495556974e-07,
+      "loss": 0.5391,
+      "step": 6849
+    },
+    {
+      "epoch": 1.6654510089958667,
+      "grad_norm": 22.125,
+      "learning_rate": 1.727415060051195e-07,
+      "loss": 0.77,
+      "step": 6850
+    },
+    {
+      "epoch": 1.6656941405300267,
+      "grad_norm": 37.0,
+      "learning_rate": 1.7249645806955594e-07,
+      "loss": 0.8063,
+      "step": 6851
+    },
+    {
+      "epoch": 1.6659372720641867,
+      "grad_norm": 14.75,
+      "learning_rate": 1.7225157118550565e-07,
+      "loss": 0.482,
+      "step": 6852
+    },
+    {
+      "epoch": 1.6661804035983467,
+      "grad_norm": 16.75,
+      "learning_rate": 1.7200684538957166e-07,
+      "loss": 0.7561,
+      "step": 6853
+    },
+    {
+      "epoch": 1.6664235351325067,
+      "grad_norm": 20.125,
+      "learning_rate": 1.7176228071833223e-07,
+      "loss": 0.6574,
+      "step": 6854
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 37.0,
+      "learning_rate": 1.7151787720834146e-07,
+      "loss": 0.814,
+      "step": 6855
+    },
+    {
+      "epoch": 1.6669097982008267,
+      "grad_norm": 20.375,
+      "learning_rate": 1.7127363489612997e-07,
+      "loss": 1.0402,
+      "step": 6856
+    },
+    {
+      "epoch": 1.6671529297349865,
+      "grad_norm": 24.875,
+      "learning_rate": 1.7102955381820427e-07,
+      "loss": 0.5734,
+      "step": 6857
+    },
+    {
+      "epoch": 1.6673960612691467,
+      "grad_norm": 22.875,
+      "learning_rate": 1.7078563401104613e-07,
+      "loss": 0.8925,
+      "step": 6858
+    },
+    {
+      "epoch": 1.6676391928033065,
+      "grad_norm": 36.5,
+      "learning_rate": 1.7054187551111351e-07,
+      "loss": 0.9493,
+      "step": 6859
+    },
+    {
+      "epoch": 1.6678823243374665,
+      "grad_norm": 19.625,
+      "learning_rate": 1.702982783548407e-07,
+      "loss": 0.7106,
+      "step": 6860
+    },
+    {
+      "epoch": 1.6681254558716265,
+      "grad_norm": 23.375,
+      "learning_rate": 1.7005484257863733e-07,
+      "loss": 0.6031,
+      "step": 6861
+    },
+    {
+      "epoch": 1.6683685874057865,
+      "grad_norm": 45.75,
+      "learning_rate": 1.6981156821888875e-07,
+      "loss": 0.7476,
+      "step": 6862
+    },
+    {
+      "epoch": 1.6686117189399465,
+      "grad_norm": 18.25,
+      "learning_rate": 1.6956845531195684e-07,
+      "loss": 0.6627,
+      "step": 6863
+    },
+    {
+      "epoch": 1.6688548504741065,
+      "grad_norm": 18.875,
+      "learning_rate": 1.693255038941792e-07,
+      "loss": 0.6377,
+      "step": 6864
+    },
+    {
+      "epoch": 1.6690979820082665,
+      "grad_norm": 28.375,
+      "learning_rate": 1.6908271400186887e-07,
+      "loss": 0.9894,
+      "step": 6865
+    },
+    {
+      "epoch": 1.6693411135424263,
+      "grad_norm": 21.375,
+      "learning_rate": 1.6884008567131473e-07,
+      "loss": 0.8733,
+      "step": 6866
+    },
+    {
+      "epoch": 1.6695842450765865,
+      "grad_norm": 15.9375,
+      "learning_rate": 1.6859761893878237e-07,
+      "loss": 0.4895,
+      "step": 6867
+    },
+    {
+      "epoch": 1.6698273766107463,
+      "grad_norm": 22.125,
+      "learning_rate": 1.6835531384051238e-07,
+      "loss": 0.7276,
+      "step": 6868
+    },
+    {
+      "epoch": 1.6700705081449065,
+      "grad_norm": 19.875,
+      "learning_rate": 1.6811317041272104e-07,
+      "loss": 0.8208,
+      "step": 6869
+    },
+    {
+      "epoch": 1.6703136396790663,
+      "grad_norm": 17.0,
+      "learning_rate": 1.678711886916015e-07,
+      "loss": 0.6899,
+      "step": 6870
+    },
+    {
+      "epoch": 1.6705567712132263,
+      "grad_norm": 21.375,
+      "learning_rate": 1.6762936871332166e-07,
+      "loss": 0.5199,
+      "step": 6871
+    },
+    {
+      "epoch": 1.6707999027473863,
+      "grad_norm": 29.875,
+      "learning_rate": 1.6738771051402604e-07,
+      "loss": 0.7683,
+      "step": 6872
+    },
+    {
+      "epoch": 1.6710430342815463,
+      "grad_norm": 20.75,
+      "learning_rate": 1.6714621412983413e-07,
+      "loss": 0.6856,
+      "step": 6873
+    },
+    {
+      "epoch": 1.6712861658157063,
+      "grad_norm": 18.75,
+      "learning_rate": 1.669048795968424e-07,
+      "loss": 0.7694,
+      "step": 6874
+    },
+    {
+      "epoch": 1.6715292973498663,
+      "grad_norm": 22.75,
+      "learning_rate": 1.6666370695112206e-07,
+      "loss": 1.1907,
+      "step": 6875
+    },
+    {
+      "epoch": 1.6717724288840263,
+      "grad_norm": 20.125,
+      "learning_rate": 1.6642269622872035e-07,
+      "loss": 0.7955,
+      "step": 6876
+    },
+    {
+      "epoch": 1.672015560418186,
+      "grad_norm": 20.875,
+      "learning_rate": 1.6618184746566097e-07,
+      "loss": 0.7652,
+      "step": 6877
+    },
+    {
+      "epoch": 1.6722586919523463,
+      "grad_norm": 23.5,
+      "learning_rate": 1.6594116069794262e-07,
+      "loss": 0.7562,
+      "step": 6878
+    },
+    {
+      "epoch": 1.6725018234865061,
+      "grad_norm": 18.0,
+      "learning_rate": 1.6570063596153998e-07,
+      "loss": 0.5837,
+      "step": 6879
+    },
+    {
+      "epoch": 1.6727449550206663,
+      "grad_norm": 22.375,
+      "learning_rate": 1.6546027329240365e-07,
+      "loss": 0.5496,
+      "step": 6880
+    },
+    {
+      "epoch": 1.6729880865548261,
+      "grad_norm": 18.75,
+      "learning_rate": 1.6522007272646052e-07,
+      "loss": 0.8426,
+      "step": 6881
+    },
+    {
+      "epoch": 1.6732312180889861,
+      "grad_norm": 20.875,
+      "learning_rate": 1.649800342996123e-07,
+      "loss": 0.7899,
+      "step": 6882
+    },
+    {
+      "epoch": 1.6734743496231461,
+      "grad_norm": 18.875,
+      "learning_rate": 1.647401580477366e-07,
+      "loss": 0.681,
+      "step": 6883
+    },
+    {
+      "epoch": 1.6737174811573061,
+      "grad_norm": 23.125,
+      "learning_rate": 1.645004440066876e-07,
+      "loss": 0.792,
+      "step": 6884
+    },
+    {
+      "epoch": 1.6739606126914661,
+      "grad_norm": 20.5,
+      "learning_rate": 1.6426089221229456e-07,
+      "loss": 0.7977,
+      "step": 6885
+    },
+    {
+      "epoch": 1.674203744225626,
+      "grad_norm": 19.75,
+      "learning_rate": 1.640215027003622e-07,
+      "loss": 0.7682,
+      "step": 6886
+    },
+    {
+      "epoch": 1.6744468757597861,
+      "grad_norm": 38.25,
+      "learning_rate": 1.6378227550667176e-07,
+      "loss": 0.5646,
+      "step": 6887
+    },
+    {
+      "epoch": 1.674690007293946,
+      "grad_norm": 19.875,
+      "learning_rate": 1.6354321066698015e-07,
+      "loss": 0.5912,
+      "step": 6888
+    },
+    {
+      "epoch": 1.6749331388281061,
+      "grad_norm": 20.875,
+      "learning_rate": 1.633043082170195e-07,
+      "loss": 0.7568,
+      "step": 6889
+    },
+    {
+      "epoch": 1.675176270362266,
+      "grad_norm": 20.75,
+      "learning_rate": 1.6306556819249767e-07,
+      "loss": 0.674,
+      "step": 6890
+    },
+    {
+      "epoch": 1.6754194018964261,
+      "grad_norm": 15.875,
+      "learning_rate": 1.6282699062909886e-07,
+      "loss": 0.4707,
+      "step": 6891
+    },
+    {
+      "epoch": 1.675662533430586,
+      "grad_norm": 24.125,
+      "learning_rate": 1.6258857556248244e-07,
+      "loss": 0.8059,
+      "step": 6892
+    },
+    {
+      "epoch": 1.675905664964746,
+      "grad_norm": 19.125,
+      "learning_rate": 1.623503230282833e-07,
+      "loss": 0.7181,
+      "step": 6893
+    },
+    {
+      "epoch": 1.676148796498906,
+      "grad_norm": 15.625,
+      "learning_rate": 1.6211223306211275e-07,
+      "loss": 0.3808,
+      "step": 6894
+    },
+    {
+      "epoch": 1.676391928033066,
+      "grad_norm": 14.5625,
+      "learning_rate": 1.6187430569955782e-07,
+      "loss": 0.4352,
+      "step": 6895
+    },
+    {
+      "epoch": 1.676635059567226,
+      "grad_norm": 26.125,
+      "learning_rate": 1.6163654097618054e-07,
+      "loss": 0.9999,
+      "step": 6896
+    },
+    {
+      "epoch": 1.6768781911013857,
+      "grad_norm": 23.125,
+      "learning_rate": 1.6139893892751848e-07,
+      "loss": 0.6284,
+      "step": 6897
+    },
+    {
+      "epoch": 1.677121322635546,
+      "grad_norm": 20.5,
+      "learning_rate": 1.6116149958908604e-07,
+      "loss": 0.605,
+      "step": 6898
+    },
+    {
+      "epoch": 1.6773644541697057,
+      "grad_norm": 34.0,
+      "learning_rate": 1.6092422299637243e-07,
+      "loss": 0.7257,
+      "step": 6899
+    },
+    {
+      "epoch": 1.677607585703866,
+      "grad_norm": 15.6875,
+      "learning_rate": 1.6068710918484248e-07,
+      "loss": 0.433,
+      "step": 6900
+    },
+    {
+      "epoch": 1.6778507172380257,
+      "grad_norm": 19.375,
+      "learning_rate": 1.6045015818993732e-07,
+      "loss": 0.7528,
+      "step": 6901
+    },
+    {
+      "epoch": 1.6780938487721857,
+      "grad_norm": 31.375,
+      "learning_rate": 1.60213370047073e-07,
+      "loss": 0.757,
+      "step": 6902
+    },
+    {
+      "epoch": 1.6783369803063457,
+      "grad_norm": 30.125,
+      "learning_rate": 1.59976744791642e-07,
+      "loss": 0.7798,
+      "step": 6903
+    },
+    {
+      "epoch": 1.6785801118405057,
+      "grad_norm": 20.25,
+      "learning_rate": 1.5974028245901162e-07,
+      "loss": 1.1059,
+      "step": 6904
+    },
+    {
+      "epoch": 1.6788232433746657,
+      "grad_norm": 17.25,
+      "learning_rate": 1.5950398308452575e-07,
+      "loss": 0.483,
+      "step": 6905
+    },
+    {
+      "epoch": 1.6790663749088257,
+      "grad_norm": 22.125,
+      "learning_rate": 1.5926784670350305e-07,
+      "loss": 0.6642,
+      "step": 6906
+    },
+    {
+      "epoch": 1.6793095064429857,
+      "grad_norm": 27.125,
+      "learning_rate": 1.5903187335123805e-07,
+      "loss": 0.5887,
+      "step": 6907
+    },
+    {
+      "epoch": 1.6795526379771455,
+      "grad_norm": 21.875,
+      "learning_rate": 1.5879606306300157e-07,
+      "loss": 0.6357,
+      "step": 6908
+    },
+    {
+      "epoch": 1.6797957695113057,
+      "grad_norm": 25.875,
+      "learning_rate": 1.5856041587403907e-07,
+      "loss": 0.8933,
+      "step": 6909
+    },
+    {
+      "epoch": 1.6800389010454655,
+      "grad_norm": 24.375,
+      "learning_rate": 1.5832493181957203e-07,
+      "loss": 1.0323,
+      "step": 6910
+    },
+    {
+      "epoch": 1.6802820325796257,
+      "grad_norm": 18.5,
+      "learning_rate": 1.5808961093479776e-07,
+      "loss": 0.6313,
+      "step": 6911
+    },
+    {
+      "epoch": 1.6805251641137855,
+      "grad_norm": 25.25,
+      "learning_rate": 1.578544532548895e-07,
+      "loss": 0.6426,
+      "step": 6912
+    },
+    {
+      "epoch": 1.6807682956479455,
+      "grad_norm": 15.875,
+      "learning_rate": 1.5761945881499514e-07,
+      "loss": 0.6022,
+      "step": 6913
+    },
+    {
+      "epoch": 1.6810114271821055,
+      "grad_norm": 21.25,
+      "learning_rate": 1.5738462765023831e-07,
+      "loss": 0.8832,
+      "step": 6914
+    },
+    {
+      "epoch": 1.6812545587162655,
+      "grad_norm": 22.625,
+      "learning_rate": 1.571499597957195e-07,
+      "loss": 0.5752,
+      "step": 6915
+    },
+    {
+      "epoch": 1.6814976902504255,
+      "grad_norm": 27.125,
+      "learning_rate": 1.569154552865132e-07,
+      "loss": 0.9953,
+      "step": 6916
+    },
+    {
+      "epoch": 1.6817408217845855,
+      "grad_norm": 28.625,
+      "learning_rate": 1.5668111415767004e-07,
+      "loss": 1.1177,
+      "step": 6917
+    },
+    {
+      "epoch": 1.6819839533187455,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.5644693644421672e-07,
+      "loss": 1.0306,
+      "step": 6918
+    },
+    {
+      "epoch": 1.6822270848529053,
+      "grad_norm": 20.625,
+      "learning_rate": 1.5621292218115537e-07,
+      "loss": 0.786,
+      "step": 6919
+    },
+    {
+      "epoch": 1.6824702163870655,
+      "grad_norm": 18.875,
+      "learning_rate": 1.5597907140346314e-07,
+      "loss": 0.44,
+      "step": 6920
+    },
+    {
+      "epoch": 1.6827133479212253,
+      "grad_norm": 22.5,
+      "learning_rate": 1.557453841460928e-07,
+      "loss": 0.7679,
+      "step": 6921
+    },
+    {
+      "epoch": 1.6829564794553855,
+      "grad_norm": 19.875,
+      "learning_rate": 1.5551186044397342e-07,
+      "loss": 0.8135,
+      "step": 6922
+    },
+    {
+      "epoch": 1.6831996109895453,
+      "grad_norm": 18.875,
+      "learning_rate": 1.5527850033200897e-07,
+      "loss": 1.1998,
+      "step": 6923
+    },
+    {
+      "epoch": 1.6834427425237053,
+      "grad_norm": 18.25,
+      "learning_rate": 1.55045303845079e-07,
+      "loss": 0.7589,
+      "step": 6924
+    },
+    {
+      "epoch": 1.6836858740578653,
+      "grad_norm": 17.875,
+      "learning_rate": 1.5481227101803886e-07,
+      "loss": 0.729,
+      "step": 6925
+    },
+    {
+      "epoch": 1.6839290055920253,
+      "grad_norm": 18.25,
+      "learning_rate": 1.5457940188571953e-07,
+      "loss": 0.629,
+      "step": 6926
+    },
+    {
+      "epoch": 1.6841721371261853,
+      "grad_norm": 25.625,
+      "learning_rate": 1.5434669648292724e-07,
+      "loss": 0.7621,
+      "step": 6927
+    },
+    {
+      "epoch": 1.6844152686603453,
+      "grad_norm": 26.25,
+      "learning_rate": 1.5411415484444344e-07,
+      "loss": 0.4665,
+      "step": 6928
+    },
+    {
+      "epoch": 1.6846584001945053,
+      "grad_norm": 25.625,
+      "learning_rate": 1.5388177700502604e-07,
+      "loss": 0.7456,
+      "step": 6929
+    },
+    {
+      "epoch": 1.684901531728665,
+      "grad_norm": 16.75,
+      "learning_rate": 1.536495629994078e-07,
+      "loss": 0.5649,
+      "step": 6930
+    },
+    {
+      "epoch": 1.6851446632628253,
+      "grad_norm": 17.125,
+      "learning_rate": 1.5341751286229667e-07,
+      "loss": 0.6208,
+      "step": 6931
+    },
+    {
+      "epoch": 1.685387794796985,
+      "grad_norm": 18.5,
+      "learning_rate": 1.531856266283771e-07,
+      "loss": 0.4659,
+      "step": 6932
+    },
+    {
+      "epoch": 1.6856309263311453,
+      "grad_norm": 18.5,
+      "learning_rate": 1.5295390433230805e-07,
+      "loss": 0.7431,
+      "step": 6933
+    },
+    {
+      "epoch": 1.685874057865305,
+      "grad_norm": 20.5,
+      "learning_rate": 1.5272234600872488e-07,
+      "loss": 0.7232,
+      "step": 6934
+    },
+    {
+      "epoch": 1.686117189399465,
+      "grad_norm": 16.375,
+      "learning_rate": 1.524909516922375e-07,
+      "loss": 0.4456,
+      "step": 6935
+    },
+    {
+      "epoch": 1.686360320933625,
+      "grad_norm": 18.75,
+      "learning_rate": 1.5225972141743235e-07,
+      "loss": 0.7808,
+      "step": 6936
+    },
+    {
+      "epoch": 1.686603452467785,
+      "grad_norm": 24.125,
+      "learning_rate": 1.520286552188703e-07,
+      "loss": 0.7493,
+      "step": 6937
+    },
+    {
+      "epoch": 1.686846584001945,
+      "grad_norm": 22.25,
+      "learning_rate": 1.5179775313108825e-07,
+      "loss": 0.9569,
+      "step": 6938
+    },
+    {
+      "epoch": 1.6870897155361049,
+      "grad_norm": 21.625,
+      "learning_rate": 1.5156701518859882e-07,
+      "loss": 0.7673,
+      "step": 6939
+    },
+    {
+      "epoch": 1.687332847070265,
+      "grad_norm": 22.0,
+      "learning_rate": 1.5133644142588916e-07,
+      "loss": 0.5399,
+      "step": 6940
+    },
+    {
+      "epoch": 1.6875759786044249,
+      "grad_norm": 28.75,
+      "learning_rate": 1.5110603187742336e-07,
+      "loss": 1.0731,
+      "step": 6941
+    },
+    {
+      "epoch": 1.687819110138585,
+      "grad_norm": 22.0,
+      "learning_rate": 1.5087578657763933e-07,
+      "loss": 1.0513,
+      "step": 6942
+    },
+    {
+      "epoch": 1.6880622416727449,
+      "grad_norm": 23.125,
+      "learning_rate": 1.506457055609517e-07,
+      "loss": 0.962,
+      "step": 6943
+    },
+    {
+      "epoch": 1.6883053732069049,
+      "grad_norm": 26.5,
+      "learning_rate": 1.5041578886174995e-07,
+      "loss": 0.7448,
+      "step": 6944
+    },
+    {
+      "epoch": 1.6885485047410649,
+      "grad_norm": 23.625,
+      "learning_rate": 1.501860365143988e-07,
+      "loss": 0.7738,
+      "step": 6945
+    },
+    {
+      "epoch": 1.6887916362752249,
+      "grad_norm": 22.625,
+      "learning_rate": 1.4995644855323926e-07,
+      "loss": 0.7364,
+      "step": 6946
+    },
+    {
+      "epoch": 1.6890347678093849,
+      "grad_norm": 19.0,
+      "learning_rate": 1.497270250125868e-07,
+      "loss": 0.674,
+      "step": 6947
+    },
+    {
+      "epoch": 1.6892778993435449,
+      "grad_norm": 24.875,
+      "learning_rate": 1.4949776592673258e-07,
+      "loss": 0.8241,
+      "step": 6948
+    },
+    {
+      "epoch": 1.6895210308777049,
+      "grad_norm": 15.5,
+      "learning_rate": 1.4926867132994348e-07,
+      "loss": 0.3412,
+      "step": 6949
+    },
+    {
+      "epoch": 1.6897641624118647,
+      "grad_norm": 21.625,
+      "learning_rate": 1.4903974125646205e-07,
+      "loss": 0.7731,
+      "step": 6950
+    },
+    {
+      "epoch": 1.690007293946025,
+      "grad_norm": 22.125,
+      "learning_rate": 1.4881097574050554e-07,
+      "loss": 0.6524,
+      "step": 6951
+    },
+    {
+      "epoch": 1.6902504254801847,
+      "grad_norm": 17.125,
+      "learning_rate": 1.4858237481626668e-07,
+      "loss": 0.6306,
+      "step": 6952
+    },
+    {
+      "epoch": 1.690493557014345,
+      "grad_norm": 21.625,
+      "learning_rate": 1.4835393851791432e-07,
+      "loss": 0.6385,
+      "step": 6953
+    },
+    {
+      "epoch": 1.6907366885485047,
+      "grad_norm": 20.125,
+      "learning_rate": 1.481256668795919e-07,
+      "loss": 0.7038,
+      "step": 6954
+    },
+    {
+      "epoch": 1.6909798200826647,
+      "grad_norm": 21.0,
+      "learning_rate": 1.4789755993541835e-07,
+      "loss": 0.609,
+      "step": 6955
+    },
+    {
+      "epoch": 1.6912229516168247,
+      "grad_norm": 18.5,
+      "learning_rate": 1.4766961771948835e-07,
+      "loss": 0.5567,
+      "step": 6956
+    },
+    {
+      "epoch": 1.6914660831509847,
+      "grad_norm": 17.625,
+      "learning_rate": 1.4744184026587227e-07,
+      "loss": 0.554,
+      "step": 6957
+    },
+    {
+      "epoch": 1.6917092146851447,
+      "grad_norm": 19.5,
+      "learning_rate": 1.4721422760861497e-07,
+      "loss": 0.6309,
+      "step": 6958
+    },
+    {
+      "epoch": 1.6919523462193047,
+      "grad_norm": 18.25,
+      "learning_rate": 1.4698677978173675e-07,
+      "loss": 0.5148,
+      "step": 6959
+    },
+    {
+      "epoch": 1.6921954777534647,
+      "grad_norm": 22.375,
+      "learning_rate": 1.4675949681923443e-07,
+      "loss": 0.8596,
+      "step": 6960
+    },
+    {
+      "epoch": 1.6924386092876245,
+      "grad_norm": 21.625,
+      "learning_rate": 1.4653237875507876e-07,
+      "loss": 0.6779,
+      "step": 6961
+    },
+    {
+      "epoch": 1.6926817408217847,
+      "grad_norm": 19.375,
+      "learning_rate": 1.4630542562321645e-07,
+      "loss": 0.9988,
+      "step": 6962
+    },
+    {
+      "epoch": 1.6929248723559445,
+      "grad_norm": 18.75,
+      "learning_rate": 1.4607863745756998e-07,
+      "loss": 0.4227,
+      "step": 6963
+    },
+    {
+      "epoch": 1.6931680038901047,
+      "grad_norm": 24.0,
+      "learning_rate": 1.458520142920364e-07,
+      "loss": 0.7502,
+      "step": 6964
+    },
+    {
+      "epoch": 1.6934111354242645,
+      "grad_norm": 19.875,
+      "learning_rate": 1.456255561604887e-07,
+      "loss": 0.7303,
+      "step": 6965
+    },
+    {
+      "epoch": 1.6936542669584245,
+      "grad_norm": 16.5,
+      "learning_rate": 1.453992630967746e-07,
+      "loss": 0.4584,
+      "step": 6966
+    },
+    {
+      "epoch": 1.6938973984925845,
+      "grad_norm": 18.125,
+      "learning_rate": 1.4517313513471798e-07,
+      "loss": 0.3222,
+      "step": 6967
+    },
+    {
+      "epoch": 1.6941405300267445,
+      "grad_norm": 25.125,
+      "learning_rate": 1.4494717230811736e-07,
+      "loss": 0.8891,
+      "step": 6968
+    },
+    {
+      "epoch": 1.6943836615609045,
+      "grad_norm": 23.25,
+      "learning_rate": 1.4472137465074656e-07,
+      "loss": 0.4947,
+      "step": 6969
+    },
+    {
+      "epoch": 1.6946267930950645,
+      "grad_norm": 19.75,
+      "learning_rate": 1.4449574219635525e-07,
+      "loss": 0.5912,
+      "step": 6970
+    },
+    {
+      "epoch": 1.6948699246292245,
+      "grad_norm": 18.125,
+      "learning_rate": 1.4427027497866784e-07,
+      "loss": 0.6814,
+      "step": 6971
+    },
+    {
+      "epoch": 1.6951130561633843,
+      "grad_norm": 20.0,
+      "learning_rate": 1.4404497303138477e-07,
+      "loss": 1.0506,
+      "step": 6972
+    },
+    {
+      "epoch": 1.6953561876975445,
+      "grad_norm": 19.0,
+      "learning_rate": 1.4381983638818055e-07,
+      "loss": 0.6342,
+      "step": 6973
+    },
+    {
+      "epoch": 1.6955993192317043,
+      "grad_norm": 25.75,
+      "learning_rate": 1.4359486508270654e-07,
+      "loss": 0.7947,
+      "step": 6974
+    },
+    {
+      "epoch": 1.6958424507658645,
+      "grad_norm": 23.375,
+      "learning_rate": 1.433700591485883e-07,
+      "loss": 0.7217,
+      "step": 6975
+    },
+    {
+      "epoch": 1.6960855823000243,
+      "grad_norm": 16.75,
+      "learning_rate": 1.4314541861942644e-07,
+      "loss": 0.6003,
+      "step": 6976
+    },
+    {
+      "epoch": 1.6963287138341843,
+      "grad_norm": 22.125,
+      "learning_rate": 1.4292094352879816e-07,
+      "loss": 0.762,
+      "step": 6977
+    },
+    {
+      "epoch": 1.6965718453683443,
+      "grad_norm": 17.375,
+      "learning_rate": 1.426966339102548e-07,
+      "loss": 0.9147,
+      "step": 6978
+    },
+    {
+      "epoch": 1.6968149769025043,
+      "grad_norm": 25.375,
+      "learning_rate": 1.4247248979732273e-07,
+      "loss": 0.6724,
+      "step": 6979
+    },
+    {
+      "epoch": 1.6970581084366643,
+      "grad_norm": 21.375,
+      "learning_rate": 1.422485112235053e-07,
+      "loss": 0.6231,
+      "step": 6980
+    },
+    {
+      "epoch": 1.697301239970824,
+      "grad_norm": 21.25,
+      "learning_rate": 1.4202469822227936e-07,
+      "loss": 0.8129,
+      "step": 6981
+    },
+    {
+      "epoch": 1.6975443715049843,
+      "grad_norm": 21.375,
+      "learning_rate": 1.418010508270977e-07,
+      "loss": 0.798,
+      "step": 6982
+    },
+    {
+      "epoch": 1.697787503039144,
+      "grad_norm": 23.125,
+      "learning_rate": 1.4157756907138804e-07,
+      "loss": 0.9029,
+      "step": 6983
+    },
+    {
+      "epoch": 1.6980306345733043,
+      "grad_norm": 21.875,
+      "learning_rate": 1.4135425298855394e-07,
+      "loss": 0.3993,
+      "step": 6984
+    },
+    {
+      "epoch": 1.698273766107464,
+      "grad_norm": 19.125,
+      "learning_rate": 1.4113110261197368e-07,
+      "loss": 0.6003,
+      "step": 6985
+    },
+    {
+      "epoch": 1.6985168976416243,
+      "grad_norm": 26.25,
+      "learning_rate": 1.409081179750009e-07,
+      "loss": 0.7194,
+      "step": 6986
+    },
+    {
+      "epoch": 1.698760029175784,
+      "grad_norm": 23.375,
+      "learning_rate": 1.4068529911096433e-07,
+      "loss": 0.673,
+      "step": 6987
+    },
+    {
+      "epoch": 1.699003160709944,
+      "grad_norm": 19.25,
+      "learning_rate": 1.404626460531687e-07,
+      "loss": 0.5228,
+      "step": 6988
+    },
+    {
+      "epoch": 1.699246292244104,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.4024015883489312e-07,
+      "loss": 0.3539,
+      "step": 6989
+    },
+    {
+      "epoch": 1.699489423778264,
+      "grad_norm": 21.0,
+      "learning_rate": 1.4001783748939152e-07,
+      "loss": 0.7357,
+      "step": 6990
+    },
+    {
+      "epoch": 1.699732555312424,
+      "grad_norm": 21.75,
+      "learning_rate": 1.3979568204989456e-07,
+      "loss": 0.6048,
+      "step": 6991
+    },
+    {
+      "epoch": 1.6999756868465838,
+      "grad_norm": 26.125,
+      "learning_rate": 1.3957369254960675e-07,
+      "loss": 0.7623,
+      "step": 6992
+    },
+    {
+      "epoch": 1.700218818380744,
+      "grad_norm": 21.25,
+      "learning_rate": 1.3935186902170815e-07,
+      "loss": 0.7856,
+      "step": 6993
+    },
+    {
+      "epoch": 1.7004619499149038,
+      "grad_norm": 18.625,
+      "learning_rate": 1.3913021149935425e-07,
+      "loss": 0.4841,
+      "step": 6994
+    },
+    {
+      "epoch": 1.700705081449064,
+      "grad_norm": 19.875,
+      "learning_rate": 1.3890872001567578e-07,
+      "loss": 0.8518,
+      "step": 6995
+    },
+    {
+      "epoch": 1.7009482129832239,
+      "grad_norm": 18.25,
+      "learning_rate": 1.3868739460377824e-07,
+      "loss": 0.7088,
+      "step": 6996
+    },
+    {
+      "epoch": 1.7011913445173839,
+      "grad_norm": 19.625,
+      "learning_rate": 1.384662352967428e-07,
+      "loss": 0.509,
+      "step": 6997
+    },
+    {
+      "epoch": 1.7014344760515439,
+      "grad_norm": 22.125,
+      "learning_rate": 1.3824524212762522e-07,
+      "loss": 0.5836,
+      "step": 6998
+    },
+    {
+      "epoch": 1.7016776075857039,
+      "grad_norm": 20.125,
+      "learning_rate": 1.3802441512945698e-07,
+      "loss": 0.4863,
+      "step": 6999
+    },
+    {
+      "epoch": 1.7019207391198639,
+      "grad_norm": 16.875,
+      "learning_rate": 1.378037543352441e-07,
+      "loss": 0.4595,
+      "step": 7000
+    },
+    {
+      "epoch": 1.7021638706540239,
+      "grad_norm": 32.5,
+      "learning_rate": 1.3758325977796875e-07,
+      "loss": 0.8221,
+      "step": 7001
+    },
+    {
+      "epoch": 1.7024070021881839,
+      "grad_norm": 23.625,
+      "learning_rate": 1.3736293149058695e-07,
+      "loss": 1.1364,
+      "step": 7002
+    },
+    {
+      "epoch": 1.7026501337223436,
+      "grad_norm": 22.25,
+      "learning_rate": 1.371427695060311e-07,
+      "loss": 0.5761,
+      "step": 7003
+    },
+    {
+      "epoch": 1.7028932652565039,
+      "grad_norm": 16.25,
+      "learning_rate": 1.369227738572082e-07,
+      "loss": 0.7817,
+      "step": 7004
+    },
+    {
+      "epoch": 1.7031363967906636,
+      "grad_norm": 19.875,
+      "learning_rate": 1.367029445770003e-07,
+      "loss": 0.5736,
+      "step": 7005
+    },
+    {
+      "epoch": 1.7033795283248239,
+      "grad_norm": 18.625,
+      "learning_rate": 1.364832816982646e-07,
+      "loss": 0.5242,
+      "step": 7006
+    },
+    {
+      "epoch": 1.7036226598589836,
+      "grad_norm": 19.375,
+      "learning_rate": 1.362637852538333e-07,
+      "loss": 0.7528,
+      "step": 7007
+    },
+    {
+      "epoch": 1.7038657913931436,
+      "grad_norm": 24.25,
+      "learning_rate": 1.3604445527651444e-07,
+      "loss": 0.8263,
+      "step": 7008
+    },
+    {
+      "epoch": 1.7041089229273036,
+      "grad_norm": 16.75,
+      "learning_rate": 1.3582529179909044e-07,
+      "loss": 0.4456,
+      "step": 7009
+    },
+    {
+      "epoch": 1.7043520544614637,
+      "grad_norm": 19.375,
+      "learning_rate": 1.3560629485431853e-07,
+      "loss": 0.7074,
+      "step": 7010
+    },
+    {
+      "epoch": 1.7045951859956237,
+      "grad_norm": 21.5,
+      "learning_rate": 1.3538746447493257e-07,
+      "loss": 1.0045,
+      "step": 7011
+    },
+    {
+      "epoch": 1.7048383175297837,
+      "grad_norm": 13.9375,
+      "learning_rate": 1.351688006936401e-07,
+      "loss": 0.3327,
+      "step": 7012
+    },
+    {
+      "epoch": 1.7050814490639437,
+      "grad_norm": 17.25,
+      "learning_rate": 1.3495030354312392e-07,
+      "loss": 0.9107,
+      "step": 7013
+    },
+    {
+      "epoch": 1.7053245805981034,
+      "grad_norm": 21.875,
+      "learning_rate": 1.3473197305604272e-07,
+      "loss": 0.705,
+      "step": 7014
+    },
+    {
+      "epoch": 1.7055677121322637,
+      "grad_norm": 24.25,
+      "learning_rate": 1.3451380926502944e-07,
+      "loss": 1.0641,
+      "step": 7015
+    },
+    {
+      "epoch": 1.7058108436664234,
+      "grad_norm": 24.0,
+      "learning_rate": 1.3429581220269245e-07,
+      "loss": 0.8427,
+      "step": 7016
+    },
+    {
+      "epoch": 1.7060539752005837,
+      "grad_norm": 21.875,
+      "learning_rate": 1.3407798190161512e-07,
+      "loss": 0.699,
+      "step": 7017
+    },
+    {
+      "epoch": 1.7062971067347434,
+      "grad_norm": 18.5,
+      "learning_rate": 1.3386031839435585e-07,
+      "loss": 0.3591,
+      "step": 7018
+    },
+    {
+      "epoch": 1.7065402382689034,
+      "grad_norm": 23.0,
+      "learning_rate": 1.3364282171344874e-07,
+      "loss": 0.7765,
+      "step": 7019
+    },
+    {
+      "epoch": 1.7067833698030634,
+      "grad_norm": 20.25,
+      "learning_rate": 1.3342549189140186e-07,
+      "loss": 0.5656,
+      "step": 7020
+    },
+    {
+      "epoch": 1.7070265013372234,
+      "grad_norm": 20.625,
+      "learning_rate": 1.3320832896069933e-07,
+      "loss": 1.0029,
+      "step": 7021
+    },
+    {
+      "epoch": 1.7072696328713834,
+      "grad_norm": 17.375,
+      "learning_rate": 1.3299133295379961e-07,
+      "loss": 0.3076,
+      "step": 7022
+    },
+    {
+      "epoch": 1.7075127644055434,
+      "grad_norm": 20.875,
+      "learning_rate": 1.3277450390313667e-07,
+      "loss": 0.8906,
+      "step": 7023
+    },
+    {
+      "epoch": 1.7077558959397034,
+      "grad_norm": 22.125,
+      "learning_rate": 1.32557841841119e-07,
+      "loss": 0.8653,
+      "step": 7024
+    },
+    {
+      "epoch": 1.7079990274738632,
+      "grad_norm": 20.125,
+      "learning_rate": 1.3234134680013085e-07,
+      "loss": 0.4807,
+      "step": 7025
+    },
+    {
+      "epoch": 1.7082421590080235,
+      "grad_norm": 22.875,
+      "learning_rate": 1.321250188125313e-07,
+      "loss": 0.7162,
+      "step": 7026
+    },
+    {
+      "epoch": 1.7084852905421832,
+      "grad_norm": 26.625,
+      "learning_rate": 1.3190885791065368e-07,
+      "loss": 0.847,
+      "step": 7027
+    },
+    {
+      "epoch": 1.7087284220763435,
+      "grad_norm": 20.5,
+      "learning_rate": 1.3169286412680768e-07,
+      "loss": 0.9272,
+      "step": 7028
+    },
+    {
+      "epoch": 1.7089715536105032,
+      "grad_norm": 27.875,
+      "learning_rate": 1.3147703749327695e-07,
+      "loss": 0.6254,
+      "step": 7029
+    },
+    {
+      "epoch": 1.7092146851446632,
+      "grad_norm": 30.0,
+      "learning_rate": 1.3126137804232025e-07,
+      "loss": 0.6389,
+      "step": 7030
+    },
+    {
+      "epoch": 1.7094578166788232,
+      "grad_norm": 26.25,
+      "learning_rate": 1.310458858061722e-07,
+      "loss": 0.9339,
+      "step": 7031
+    },
+    {
+      "epoch": 1.7097009482129832,
+      "grad_norm": 17.5,
+      "learning_rate": 1.308305608170414e-07,
+      "loss": 0.3279,
+      "step": 7032
+    },
+    {
+      "epoch": 1.7099440797471432,
+      "grad_norm": 21.75,
+      "learning_rate": 1.306154031071119e-07,
+      "loss": 1.0276,
+      "step": 7033
+    },
+    {
+      "epoch": 1.710187211281303,
+      "grad_norm": 19.5,
+      "learning_rate": 1.3040041270854286e-07,
+      "loss": 0.8698,
+      "step": 7034
+    },
+    {
+      "epoch": 1.7104303428154632,
+      "grad_norm": 19.125,
+      "learning_rate": 1.3018558965346844e-07,
+      "loss": 0.5881,
+      "step": 7035
+    },
+    {
+      "epoch": 1.710673474349623,
+      "grad_norm": 23.625,
+      "learning_rate": 1.2997093397399754e-07,
+      "loss": 0.5417,
+      "step": 7036
+    },
+    {
+      "epoch": 1.7109166058837832,
+      "grad_norm": 24.5,
+      "learning_rate": 1.2975644570221394e-07,
+      "loss": 1.1227,
+      "step": 7037
+    },
+    {
+      "epoch": 1.711159737417943,
+      "grad_norm": 17.625,
+      "learning_rate": 1.2954212487017697e-07,
+      "loss": 0.4961,
+      "step": 7038
+    },
+    {
+      "epoch": 1.7114028689521033,
+      "grad_norm": 19.375,
+      "learning_rate": 1.2932797150992048e-07,
+      "loss": 1.0014,
+      "step": 7039
+    },
+    {
+      "epoch": 1.711646000486263,
+      "grad_norm": 13.625,
+      "learning_rate": 1.2911398565345317e-07,
+      "loss": 0.3568,
+      "step": 7040
+    },
+    {
+      "epoch": 1.711889132020423,
+      "grad_norm": 18.5,
+      "learning_rate": 1.2890016733275877e-07,
+      "loss": 0.676,
+      "step": 7041
+    },
+    {
+      "epoch": 1.712132263554583,
+      "grad_norm": 19.5,
+      "learning_rate": 1.2868651657979682e-07,
+      "loss": 0.6297,
+      "step": 7042
+    },
+    {
+      "epoch": 1.712375395088743,
+      "grad_norm": 16.25,
+      "learning_rate": 1.284730334265008e-07,
+      "loss": 0.318,
+      "step": 7043
+    },
+    {
+      "epoch": 1.712618526622903,
+      "grad_norm": 19.75,
+      "learning_rate": 1.2825971790477912e-07,
+      "loss": 0.741,
+      "step": 7044
+    },
+    {
+      "epoch": 1.7128616581570628,
+      "grad_norm": 17.375,
+      "learning_rate": 1.2804657004651583e-07,
+      "loss": 0.5294,
+      "step": 7045
+    },
+    {
+      "epoch": 1.713104789691223,
+      "grad_norm": 17.375,
+      "learning_rate": 1.278335898835696e-07,
+      "loss": 0.6027,
+      "step": 7046
+    },
+    {
+      "epoch": 1.7133479212253828,
+      "grad_norm": 24.25,
+      "learning_rate": 1.2762077744777363e-07,
+      "loss": 0.5592,
+      "step": 7047
+    },
+    {
+      "epoch": 1.713591052759543,
+      "grad_norm": 17.75,
+      "learning_rate": 1.2740813277093673e-07,
+      "loss": 0.6728,
+      "step": 7048
+    },
+    {
+      "epoch": 1.7138341842937028,
+      "grad_norm": 22.0,
+      "learning_rate": 1.2719565588484208e-07,
+      "loss": 0.6397,
+      "step": 7049
+    },
+    {
+      "epoch": 1.7140773158278628,
+      "grad_norm": 22.25,
+      "learning_rate": 1.2698334682124834e-07,
+      "loss": 0.8683,
+      "step": 7050
+    },
+    {
+      "epoch": 1.7143204473620228,
+      "grad_norm": 17.75,
+      "learning_rate": 1.2677120561188834e-07,
+      "loss": 0.4273,
+      "step": 7051
+    },
+    {
+      "epoch": 1.7145635788961828,
+      "grad_norm": 20.75,
+      "learning_rate": 1.2655923228847082e-07,
+      "loss": 0.9747,
+      "step": 7052
+    },
+    {
+      "epoch": 1.7148067104303428,
+      "grad_norm": 21.5,
+      "learning_rate": 1.263474268826785e-07,
+      "loss": 0.9047,
+      "step": 7053
+    },
+    {
+      "epoch": 1.7150498419645028,
+      "grad_norm": 19.5,
+      "learning_rate": 1.2613578942616902e-07,
+      "loss": 0.6919,
+      "step": 7054
+    },
+    {
+      "epoch": 1.7152929734986628,
+      "grad_norm": 20.625,
+      "learning_rate": 1.2592431995057608e-07,
+      "loss": 0.4946,
+      "step": 7055
+    },
+    {
+      "epoch": 1.7155361050328226,
+      "grad_norm": 24.0,
+      "learning_rate": 1.257130184875066e-07,
+      "loss": 0.9144,
+      "step": 7056
+    },
+    {
+      "epoch": 1.7157792365669828,
+      "grad_norm": 18.0,
+      "learning_rate": 1.2550188506854383e-07,
+      "loss": 0.5277,
+      "step": 7057
+    },
+    {
+      "epoch": 1.7160223681011426,
+      "grad_norm": 21.5,
+      "learning_rate": 1.2529091972524486e-07,
+      "loss": 0.8893,
+      "step": 7058
+    },
+    {
+      "epoch": 1.7162654996353028,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.2508012248914265e-07,
+      "loss": 0.3351,
+      "step": 7059
+    },
+    {
+      "epoch": 1.7165086311694626,
+      "grad_norm": 22.125,
+      "learning_rate": 1.248694933917441e-07,
+      "loss": 0.5553,
+      "step": 7060
+    },
+    {
+      "epoch": 1.7167517627036226,
+      "grad_norm": 26.5,
+      "learning_rate": 1.2465903246453123e-07,
+      "loss": 1.0793,
+      "step": 7061
+    },
+    {
+      "epoch": 1.7169948942377826,
+      "grad_norm": 22.625,
+      "learning_rate": 1.2444873973896153e-07,
+      "loss": 0.6153,
+      "step": 7062
+    },
+    {
+      "epoch": 1.7172380257719426,
+      "grad_norm": 21.5,
+      "learning_rate": 1.242386152464667e-07,
+      "loss": 0.7403,
+      "step": 7063
+    },
+    {
+      "epoch": 1.7174811573061026,
+      "grad_norm": 17.875,
+      "learning_rate": 1.2402865901845294e-07,
+      "loss": 0.325,
+      "step": 7064
+    },
+    {
+      "epoch": 1.7177242888402626,
+      "grad_norm": 23.125,
+      "learning_rate": 1.238188710863024e-07,
+      "loss": 0.9473,
+      "step": 7065
+    },
+    {
+      "epoch": 1.7179674203744226,
+      "grad_norm": 36.75,
+      "learning_rate": 1.2360925148137165e-07,
+      "loss": 0.9482,
+      "step": 7066
+    },
+    {
+      "epoch": 1.7182105519085824,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.2339980023499176e-07,
+      "loss": 0.4364,
+      "step": 7067
+    },
+    {
+      "epoch": 1.7184536834427426,
+      "grad_norm": 30.0,
+      "learning_rate": 1.2319051737846838e-07,
+      "loss": 1.1723,
+      "step": 7068
+    },
+    {
+      "epoch": 1.7186968149769024,
+      "grad_norm": 17.125,
+      "learning_rate": 1.229814029430833e-07,
+      "loss": 0.6533,
+      "step": 7069
+    },
+    {
+      "epoch": 1.7189399465110626,
+      "grad_norm": 20.0,
+      "learning_rate": 1.2277245696009175e-07,
+      "loss": 0.5493,
+      "step": 7070
+    },
+    {
+      "epoch": 1.7191830780452224,
+      "grad_norm": 19.0,
+      "learning_rate": 1.2256367946072405e-07,
+      "loss": 0.8707,
+      "step": 7071
+    },
+    {
+      "epoch": 1.7194262095793824,
+      "grad_norm": 23.0,
+      "learning_rate": 1.2235507047618612e-07,
+      "loss": 0.6375,
+      "step": 7072
+    },
+    {
+      "epoch": 1.7196693411135424,
+      "grad_norm": 25.5,
+      "learning_rate": 1.2214663003765808e-07,
+      "loss": 1.1586,
+      "step": 7073
+    },
+    {
+      "epoch": 1.7199124726477024,
+      "grad_norm": 18.375,
+      "learning_rate": 1.2193835817629492e-07,
+      "loss": 0.6262,
+      "step": 7074
+    },
+    {
+      "epoch": 1.7201556041818624,
+      "grad_norm": 12.9375,
+      "learning_rate": 1.2173025492322633e-07,
+      "loss": 0.4056,
+      "step": 7075
+    },
+    {
+      "epoch": 1.7203987357160224,
+      "grad_norm": 18.75,
+      "learning_rate": 1.215223203095571e-07,
+      "loss": 0.6889,
+      "step": 7076
+    },
+    {
+      "epoch": 1.7206418672501824,
+      "grad_norm": 24.0,
+      "learning_rate": 1.213145543663667e-07,
+      "loss": 0.9116,
+      "step": 7077
+    },
+    {
+      "epoch": 1.7208849987843422,
+      "grad_norm": 25.25,
+      "learning_rate": 1.2110695712470893e-07,
+      "loss": 0.4975,
+      "step": 7078
+    },
+    {
+      "epoch": 1.7211281303185024,
+      "grad_norm": 17.0,
+      "learning_rate": 1.208995286156131e-07,
+      "loss": 0.6069,
+      "step": 7079
+    },
+    {
+      "epoch": 1.7213712618526622,
+      "grad_norm": 17.75,
+      "learning_rate": 1.206922688700833e-07,
+      "loss": 0.6969,
+      "step": 7080
+    },
+    {
+      "epoch": 1.7216143933868224,
+      "grad_norm": 14.75,
+      "learning_rate": 1.2048517791909755e-07,
+      "loss": 0.333,
+      "step": 7081
+    },
+    {
+      "epoch": 1.7218575249209822,
+      "grad_norm": 19.25,
+      "learning_rate": 1.2027825579360927e-07,
+      "loss": 0.7456,
+      "step": 7082
+    },
+    {
+      "epoch": 1.7221006564551422,
+      "grad_norm": 21.875,
+      "learning_rate": 1.2007150252454676e-07,
+      "loss": 0.8768,
+      "step": 7083
+    },
+    {
+      "epoch": 1.7223437879893022,
+      "grad_norm": 18.25,
+      "learning_rate": 1.198649181428127e-07,
+      "loss": 0.7055,
+      "step": 7084
+    },
+    {
+      "epoch": 1.7225869195234622,
+      "grad_norm": 17.125,
+      "learning_rate": 1.1965850267928458e-07,
+      "loss": 0.334,
+      "step": 7085
+    },
+    {
+      "epoch": 1.7228300510576222,
+      "grad_norm": 23.75,
+      "learning_rate": 1.1945225616481523e-07,
+      "loss": 0.5971,
+      "step": 7086
+    },
+    {
+      "epoch": 1.723073182591782,
+      "grad_norm": 19.0,
+      "learning_rate": 1.192461786302311e-07,
+      "loss": 0.7236,
+      "step": 7087
+    },
+    {
+      "epoch": 1.7233163141259422,
+      "grad_norm": 22.5,
+      "learning_rate": 1.190402701063345e-07,
+      "loss": 0.6064,
+      "step": 7088
+    },
+    {
+      "epoch": 1.723559445660102,
+      "grad_norm": 18.75,
+      "learning_rate": 1.1883453062390179e-07,
+      "loss": 0.9833,
+      "step": 7089
+    },
+    {
+      "epoch": 1.7238025771942622,
+      "grad_norm": 30.25,
+      "learning_rate": 1.1862896021368461e-07,
+      "loss": 1.0316,
+      "step": 7090
+    },
+    {
+      "epoch": 1.724045708728422,
+      "grad_norm": 23.125,
+      "learning_rate": 1.1842355890640866e-07,
+      "loss": 1.0296,
+      "step": 7091
+    },
+    {
+      "epoch": 1.7242888402625822,
+      "grad_norm": 18.25,
+      "learning_rate": 1.182183267327748e-07,
+      "loss": 0.5016,
+      "step": 7092
+    },
+    {
+      "epoch": 1.724531971796742,
+      "grad_norm": 17.375,
+      "learning_rate": 1.1801326372345865e-07,
+      "loss": 0.4573,
+      "step": 7093
+    },
+    {
+      "epoch": 1.724775103330902,
+      "grad_norm": 17.25,
+      "learning_rate": 1.1780836990911037e-07,
+      "loss": 0.6343,
+      "step": 7094
+    },
+    {
+      "epoch": 1.725018234865062,
+      "grad_norm": 33.75,
+      "learning_rate": 1.1760364532035454e-07,
+      "loss": 0.7275,
+      "step": 7095
+    },
+    {
+      "epoch": 1.725261366399222,
+      "grad_norm": 29.5,
+      "learning_rate": 1.1739908998779109e-07,
+      "loss": 0.6388,
+      "step": 7096
+    },
+    {
+      "epoch": 1.725504497933382,
+      "grad_norm": 20.625,
+      "learning_rate": 1.1719470394199461e-07,
+      "loss": 0.6406,
+      "step": 7097
+    },
+    {
+      "epoch": 1.7257476294675418,
+      "grad_norm": 20.5,
+      "learning_rate": 1.1699048721351386e-07,
+      "loss": 0.6727,
+      "step": 7098
+    },
+    {
+      "epoch": 1.725990761001702,
+      "grad_norm": 16.125,
+      "learning_rate": 1.1678643983287233e-07,
+      "loss": 0.5676,
+      "step": 7099
+    },
+    {
+      "epoch": 1.7262338925358618,
+      "grad_norm": 17.625,
+      "learning_rate": 1.1658256183056883e-07,
+      "loss": 0.7208,
+      "step": 7100
+    },
+    {
+      "epoch": 1.726477024070022,
+      "grad_norm": 16.125,
+      "learning_rate": 1.1637885323707618e-07,
+      "loss": 0.3214,
+      "step": 7101
+    },
+    {
+      "epoch": 1.7267201556041818,
+      "grad_norm": 23.25,
+      "learning_rate": 1.1617531408284213e-07,
+      "loss": 0.6523,
+      "step": 7102
+    },
+    {
+      "epoch": 1.7269632871383418,
+      "grad_norm": 20.0,
+      "learning_rate": 1.1597194439828916e-07,
+      "loss": 0.7844,
+      "step": 7103
+    },
+    {
+      "epoch": 1.7272064186725018,
+      "grad_norm": 19.75,
+      "learning_rate": 1.1576874421381473e-07,
+      "loss": 0.5816,
+      "step": 7104
+    },
+    {
+      "epoch": 1.7274495502066618,
+      "grad_norm": 22.5,
+      "learning_rate": 1.1556571355979027e-07,
+      "loss": 0.6491,
+      "step": 7105
+    },
+    {
+      "epoch": 1.7276926817408218,
+      "grad_norm": 28.25,
+      "learning_rate": 1.1536285246656203e-07,
+      "loss": 1.1528,
+      "step": 7106
+    },
+    {
+      "epoch": 1.7279358132749818,
+      "grad_norm": 20.75,
+      "learning_rate": 1.1516016096445162e-07,
+      "loss": 0.8362,
+      "step": 7107
+    },
+    {
+      "epoch": 1.7281789448091418,
+      "grad_norm": 32.0,
+      "learning_rate": 1.1495763908375452e-07,
+      "loss": 0.6713,
+      "step": 7108
+    },
+    {
+      "epoch": 1.7284220763433016,
+      "grad_norm": 17.5,
+      "learning_rate": 1.147552868547408e-07,
+      "loss": 0.7046,
+      "step": 7109
+    },
+    {
+      "epoch": 1.7286652078774618,
+      "grad_norm": 14.375,
+      "learning_rate": 1.1455310430765601e-07,
+      "loss": 0.4023,
+      "step": 7110
+    },
+    {
+      "epoch": 1.7289083394116216,
+      "grad_norm": 20.125,
+      "learning_rate": 1.1435109147271972e-07,
+      "loss": 0.9681,
+      "step": 7111
+    },
+    {
+      "epoch": 1.7291514709457818,
+      "grad_norm": 17.625,
+      "learning_rate": 1.1414924838012611e-07,
+      "loss": 0.3903,
+      "step": 7112
+    },
+    {
+      "epoch": 1.7293946024799416,
+      "grad_norm": 20.25,
+      "learning_rate": 1.1394757506004397e-07,
+      "loss": 0.7485,
+      "step": 7113
+    },
+    {
+      "epoch": 1.7296377340141016,
+      "grad_norm": 18.625,
+      "learning_rate": 1.1374607154261724e-07,
+      "loss": 1.008,
+      "step": 7114
+    },
+    {
+      "epoch": 1.7298808655482616,
+      "grad_norm": 16.125,
+      "learning_rate": 1.1354473785796405e-07,
+      "loss": 0.3276,
+      "step": 7115
+    },
+    {
+      "epoch": 1.7301239970824216,
+      "grad_norm": 24.5,
+      "learning_rate": 1.1334357403617671e-07,
+      "loss": 1.0734,
+      "step": 7116
+    },
+    {
+      "epoch": 1.7303671286165816,
+      "grad_norm": 16.375,
+      "learning_rate": 1.1314258010732312e-07,
+      "loss": 0.4096,
+      "step": 7117
+    },
+    {
+      "epoch": 1.7306102601507416,
+      "grad_norm": 18.5,
+      "learning_rate": 1.1294175610144495e-07,
+      "loss": 0.6694,
+      "step": 7118
+    },
+    {
+      "epoch": 1.7308533916849016,
+      "grad_norm": 20.875,
+      "learning_rate": 1.1274110204855929e-07,
+      "loss": 0.6163,
+      "step": 7119
+    },
+    {
+      "epoch": 1.7310965232190614,
+      "grad_norm": 23.125,
+      "learning_rate": 1.1254061797865687e-07,
+      "loss": 0.5054,
+      "step": 7120
+    },
+    {
+      "epoch": 1.7313396547532216,
+      "grad_norm": 26.25,
+      "learning_rate": 1.1234030392170386e-07,
+      "loss": 0.5857,
+      "step": 7121
+    },
+    {
+      "epoch": 1.7315827862873814,
+      "grad_norm": 18.0,
+      "learning_rate": 1.1214015990764049e-07,
+      "loss": 0.4037,
+      "step": 7122
+    },
+    {
+      "epoch": 1.7318259178215416,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.1194018596638156e-07,
+      "loss": 0.6404,
+      "step": 7123
+    },
+    {
+      "epoch": 1.7320690493557014,
+      "grad_norm": 20.75,
+      "learning_rate": 1.1174038212781693e-07,
+      "loss": 0.7418,
+      "step": 7124
+    },
+    {
+      "epoch": 1.7323121808898614,
+      "grad_norm": 23.0,
+      "learning_rate": 1.1154074842181075e-07,
+      "loss": 0.4778,
+      "step": 7125
+    },
+    {
+      "epoch": 1.7325553124240214,
+      "grad_norm": 22.875,
+      "learning_rate": 1.1134128487820126e-07,
+      "loss": 0.6934,
+      "step": 7126
+    },
+    {
+      "epoch": 1.7327984439581814,
+      "grad_norm": 16.5,
+      "learning_rate": 1.1114199152680208e-07,
+      "loss": 0.4853,
+      "step": 7127
+    },
+    {
+      "epoch": 1.7330415754923414,
+      "grad_norm": 24.25,
+      "learning_rate": 1.1094286839740134e-07,
+      "loss": 0.9262,
+      "step": 7128
+    },
+    {
+      "epoch": 1.7332847070265014,
+      "grad_norm": 21.25,
+      "learning_rate": 1.1074391551976108e-07,
+      "loss": 0.6301,
+      "step": 7129
+    },
+    {
+      "epoch": 1.7335278385606614,
+      "grad_norm": 28.125,
+      "learning_rate": 1.1054513292361804e-07,
+      "loss": 0.9771,
+      "step": 7130
+    },
+    {
+      "epoch": 1.7337709700948212,
+      "grad_norm": 23.375,
+      "learning_rate": 1.1034652063868417e-07,
+      "loss": 0.9385,
+      "step": 7131
+    },
+    {
+      "epoch": 1.7340141016289814,
+      "grad_norm": 21.375,
+      "learning_rate": 1.1014807869464516e-07,
+      "loss": 0.7921,
+      "step": 7132
+    },
+    {
+      "epoch": 1.7342572331631412,
+      "grad_norm": 18.0,
+      "learning_rate": 1.099498071211616e-07,
+      "loss": 0.5227,
+      "step": 7133
+    },
+    {
+      "epoch": 1.7345003646973014,
+      "grad_norm": 19.875,
+      "learning_rate": 1.097517059478688e-07,
+      "loss": 0.4977,
+      "step": 7134
+    },
+    {
+      "epoch": 1.7347434962314612,
+      "grad_norm": 20.875,
+      "learning_rate": 1.095537752043764e-07,
+      "loss": 0.7416,
+      "step": 7135
+    },
+    {
+      "epoch": 1.7349866277656212,
+      "grad_norm": 16.25,
+      "learning_rate": 1.0935601492026854e-07,
+      "loss": 0.5215,
+      "step": 7136
+    },
+    {
+      "epoch": 1.7352297592997812,
+      "grad_norm": 23.0,
+      "learning_rate": 1.0915842512510364e-07,
+      "loss": 0.4592,
+      "step": 7137
+    },
+    {
+      "epoch": 1.7354728908339412,
+      "grad_norm": 24.375,
+      "learning_rate": 1.0896100584841543e-07,
+      "loss": 1.4003,
+      "step": 7138
+    },
+    {
+      "epoch": 1.7357160223681012,
+      "grad_norm": 20.25,
+      "learning_rate": 1.0876375711971115e-07,
+      "loss": 0.6258,
+      "step": 7139
+    },
+    {
+      "epoch": 1.735959153902261,
+      "grad_norm": 29.125,
+      "learning_rate": 1.0856667896847306e-07,
+      "loss": 0.9952,
+      "step": 7140
+    },
+    {
+      "epoch": 1.7362022854364212,
+      "grad_norm": 20.375,
+      "learning_rate": 1.08369771424158e-07,
+      "loss": 0.6542,
+      "step": 7141
+    },
+    {
+      "epoch": 1.736445416970581,
+      "grad_norm": 31.125,
+      "learning_rate": 1.0817303451619756e-07,
+      "loss": 0.728,
+      "step": 7142
+    },
+    {
+      "epoch": 1.7366885485047412,
+      "grad_norm": 24.25,
+      "learning_rate": 1.0797646827399714e-07,
+      "loss": 1.1587,
+      "step": 7143
+    },
+    {
+      "epoch": 1.736931680038901,
+      "grad_norm": 22.625,
+      "learning_rate": 1.0778007272693666e-07,
+      "loss": 0.8443,
+      "step": 7144
+    },
+    {
+      "epoch": 1.737174811573061,
+      "grad_norm": 22.125,
+      "learning_rate": 1.0758384790437129e-07,
+      "loss": 0.5823,
+      "step": 7145
+    },
+    {
+      "epoch": 1.737417943107221,
+      "grad_norm": 23.375,
+      "learning_rate": 1.0738779383563019e-07,
+      "loss": 0.6295,
+      "step": 7146
+    },
+    {
+      "epoch": 1.737661074641381,
+      "grad_norm": 19.625,
+      "learning_rate": 1.071919105500166e-07,
+      "loss": 0.5701,
+      "step": 7147
+    },
+    {
+      "epoch": 1.737904206175541,
+      "grad_norm": 21.125,
+      "learning_rate": 1.0699619807680916e-07,
+      "loss": 0.7625,
+      "step": 7148
+    },
+    {
+      "epoch": 1.738147337709701,
+      "grad_norm": 18.875,
+      "learning_rate": 1.068006564452602e-07,
+      "loss": 0.5946,
+      "step": 7149
+    },
+    {
+      "epoch": 1.738390469243861,
+      "grad_norm": 14.5,
+      "learning_rate": 1.06605285684597e-07,
+      "loss": 0.3013,
+      "step": 7150
+    },
+    {
+      "epoch": 1.7386336007780208,
+      "grad_norm": 20.0,
+      "learning_rate": 1.0641008582402065e-07,
+      "loss": 0.725,
+      "step": 7151
+    },
+    {
+      "epoch": 1.738876732312181,
+      "grad_norm": 15.875,
+      "learning_rate": 1.0621505689270783e-07,
+      "loss": 0.6137,
+      "step": 7152
+    },
+    {
+      "epoch": 1.7391198638463408,
+      "grad_norm": 19.0,
+      "learning_rate": 1.0602019891980856e-07,
+      "loss": 0.7135,
+      "step": 7153
+    },
+    {
+      "epoch": 1.739362995380501,
+      "grad_norm": 22.375,
+      "learning_rate": 1.0582551193444743e-07,
+      "loss": 0.9001,
+      "step": 7154
+    },
+    {
+      "epoch": 1.7396061269146608,
+      "grad_norm": 17.25,
+      "learning_rate": 1.0563099596572452e-07,
+      "loss": 0.537,
+      "step": 7155
+    },
+    {
+      "epoch": 1.7398492584488208,
+      "grad_norm": 20.375,
+      "learning_rate": 1.0543665104271309e-07,
+      "loss": 0.4311,
+      "step": 7156
+    },
+    {
+      "epoch": 1.7400923899829808,
+      "grad_norm": 22.25,
+      "learning_rate": 1.0524247719446129e-07,
+      "loss": 0.5902,
+      "step": 7157
+    },
+    {
+      "epoch": 1.7403355215171408,
+      "grad_norm": 22.25,
+      "learning_rate": 1.0504847444999189e-07,
+      "loss": 0.4986,
+      "step": 7158
+    },
+    {
+      "epoch": 1.7405786530513008,
+      "grad_norm": 21.0,
+      "learning_rate": 1.0485464283830224e-07,
+      "loss": 0.5163,
+      "step": 7159
+    },
+    {
+      "epoch": 1.7408217845854608,
+      "grad_norm": 21.125,
+      "learning_rate": 1.0466098238836347e-07,
+      "loss": 0.482,
+      "step": 7160
+    },
+    {
+      "epoch": 1.7410649161196208,
+      "grad_norm": 18.75,
+      "learning_rate": 1.0446749312912147e-07,
+      "loss": 0.5752,
+      "step": 7161
+    },
+    {
+      "epoch": 1.7413080476537806,
+      "grad_norm": 20.25,
+      "learning_rate": 1.0427417508949669e-07,
+      "loss": 0.6879,
+      "step": 7162
+    },
+    {
+      "epoch": 1.7415511791879408,
+      "grad_norm": 19.0,
+      "learning_rate": 1.0408102829838395e-07,
+      "loss": 0.5716,
+      "step": 7163
+    },
+    {
+      "epoch": 1.7417943107221006,
+      "grad_norm": 12.25,
+      "learning_rate": 1.0388805278465211e-07,
+      "loss": 0.24,
+      "step": 7164
+    },
+    {
+      "epoch": 1.7420374422562608,
+      "grad_norm": 18.5,
+      "learning_rate": 1.036952485771446e-07,
+      "loss": 0.4406,
+      "step": 7165
+    },
+    {
+      "epoch": 1.7422805737904206,
+      "grad_norm": 19.875,
+      "learning_rate": 1.0350261570467993e-07,
+      "loss": 0.6768,
+      "step": 7166
+    },
+    {
+      "epoch": 1.7425237053245806,
+      "grad_norm": 20.875,
+      "learning_rate": 1.0331015419605003e-07,
+      "loss": 0.6819,
+      "step": 7167
+    },
+    {
+      "epoch": 1.7427668368587406,
+      "grad_norm": 23.125,
+      "learning_rate": 1.0311786408002137e-07,
+      "loss": 0.9034,
+      "step": 7168
+    },
+    {
+      "epoch": 1.7430099683929006,
+      "grad_norm": 23.75,
+      "learning_rate": 1.0292574538533537e-07,
+      "loss": 0.7285,
+      "step": 7169
+    },
+    {
+      "epoch": 1.7432530999270606,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.0273379814070756e-07,
+      "loss": 0.242,
+      "step": 7170
+    },
+    {
+      "epoch": 1.7434962314612206,
+      "grad_norm": 30.625,
+      "learning_rate": 1.0254202237482733e-07,
+      "loss": 0.838,
+      "step": 7171
+    },
+    {
+      "epoch": 1.7437393629953806,
+      "grad_norm": 19.75,
+      "learning_rate": 1.0235041811635899e-07,
+      "loss": 0.9775,
+      "step": 7172
+    },
+    {
+      "epoch": 1.7439824945295404,
+      "grad_norm": 21.5,
+      "learning_rate": 1.0215898539394156e-07,
+      "loss": 0.8749,
+      "step": 7173
+    },
+    {
+      "epoch": 1.7442256260637006,
+      "grad_norm": 23.5,
+      "learning_rate": 1.0196772423618772e-07,
+      "loss": 0.8868,
+      "step": 7174
+    },
+    {
+      "epoch": 1.7444687575978604,
+      "grad_norm": 18.25,
+      "learning_rate": 1.0177663467168447e-07,
+      "loss": 0.3057,
+      "step": 7175
+    },
+    {
+      "epoch": 1.7447118891320206,
+      "grad_norm": 20.5,
+      "learning_rate": 1.0158571672899381e-07,
+      "loss": 0.8536,
+      "step": 7176
+    },
+    {
+      "epoch": 1.7449550206661804,
+      "grad_norm": 23.25,
+      "learning_rate": 1.0139497043665166e-07,
+      "loss": 0.5509,
+      "step": 7177
+    },
+    {
+      "epoch": 1.7451981522003404,
+      "grad_norm": 20.125,
+      "learning_rate": 1.0120439582316802e-07,
+      "loss": 0.4302,
+      "step": 7178
+    },
+    {
+      "epoch": 1.7454412837345004,
+      "grad_norm": 24.25,
+      "learning_rate": 1.0101399291702813e-07,
+      "loss": 0.6958,
+      "step": 7179
+    },
+    {
+      "epoch": 1.7456844152686604,
+      "grad_norm": 31.75,
+      "learning_rate": 1.0082376174669034e-07,
+      "loss": 0.7287,
+      "step": 7180
+    },
+    {
+      "epoch": 1.7459275468028204,
+      "grad_norm": 21.125,
+      "learning_rate": 1.0063370234058859e-07,
+      "loss": 0.6602,
+      "step": 7181
+    },
+    {
+      "epoch": 1.7461706783369801,
+      "grad_norm": 15.5,
+      "learning_rate": 1.0044381472712999e-07,
+      "loss": 0.5224,
+      "step": 7182
+    },
+    {
+      "epoch": 1.7464138098711404,
+      "grad_norm": 17.125,
+      "learning_rate": 1.0025409893469701e-07,
+      "loss": 0.6278,
+      "step": 7183
+    },
+    {
+      "epoch": 1.7466569414053001,
+      "grad_norm": 24.75,
+      "learning_rate": 1.0006455499164582e-07,
+      "loss": 0.8102,
+      "step": 7184
+    },
+    {
+      "epoch": 1.7469000729394604,
+      "grad_norm": 18.75,
+      "learning_rate": 9.987518292630672e-08,
+      "loss": 0.6583,
+      "step": 7185
+    },
+    {
+      "epoch": 1.7471432044736201,
+      "grad_norm": 17.0,
+      "learning_rate": 9.968598276698508e-08,
+      "loss": 0.4623,
+      "step": 7186
+    },
+    {
+      "epoch": 1.7473863360077804,
+      "grad_norm": 13.9375,
+      "learning_rate": 9.949695454195996e-08,
+      "loss": 0.2664,
+      "step": 7187
+    },
+    {
+      "epoch": 1.7476294675419402,
+      "grad_norm": 20.0,
+      "learning_rate": 9.93080982794846e-08,
+      "loss": 0.7451,
+      "step": 7188
+    },
+    {
+      "epoch": 1.7478725990761002,
+      "grad_norm": 21.375,
+      "learning_rate": 9.91194140077871e-08,
+      "loss": 0.8819,
+      "step": 7189
+    },
+    {
+      "epoch": 1.7481157306102602,
+      "grad_norm": 18.5,
+      "learning_rate": 9.893090175506979e-08,
+      "loss": 0.7237,
+      "step": 7190
+    },
+    {
+      "epoch": 1.7483588621444202,
+      "grad_norm": 18.75,
+      "learning_rate": 9.874256154950885e-08,
+      "loss": 0.6904,
+      "step": 7191
+    },
+    {
+      "epoch": 1.7486019936785802,
+      "grad_norm": 18.75,
+      "learning_rate": 9.855439341925482e-08,
+      "loss": 0.6456,
+      "step": 7192
+    },
+    {
+      "epoch": 1.74884512521274,
+      "grad_norm": 18.125,
+      "learning_rate": 9.8366397392433e-08,
+      "loss": 0.5506,
+      "step": 7193
+    },
+    {
+      "epoch": 1.7490882567469002,
+      "grad_norm": 16.75,
+      "learning_rate": 9.817857349714244e-08,
+      "loss": 0.4657,
+      "step": 7194
+    },
+    {
+      "epoch": 1.74933138828106,
+      "grad_norm": 19.0,
+      "learning_rate": 9.799092176145664e-08,
+      "loss": 0.5334,
+      "step": 7195
+    },
+    {
+      "epoch": 1.7495745198152202,
+      "grad_norm": 22.625,
+      "learning_rate": 9.780344221342344e-08,
+      "loss": 0.8858,
+      "step": 7196
+    },
+    {
+      "epoch": 1.74981765134938,
+      "grad_norm": 18.125,
+      "learning_rate": 9.761613488106503e-08,
+      "loss": 0.4806,
+      "step": 7197
+    },
+    {
+      "epoch": 1.75006078288354,
+      "grad_norm": 24.0,
+      "learning_rate": 9.742899979237774e-08,
+      "loss": 1.0021,
+      "step": 7198
+    },
+    {
+      "epoch": 1.7503039144177,
+      "grad_norm": 19.5,
+      "learning_rate": 9.724203697533172e-08,
+      "loss": 0.6598,
+      "step": 7199
+    },
+    {
+      "epoch": 1.75054704595186,
+      "grad_norm": 19.625,
+      "learning_rate": 9.705524645787237e-08,
+      "loss": 0.6493,
+      "step": 7200
+    },
+    {
+      "epoch": 1.75079017748602,
+      "grad_norm": 20.5,
+      "learning_rate": 9.686862826791849e-08,
+      "loss": 0.7907,
+      "step": 7201
+    },
+    {
+      "epoch": 1.75103330902018,
+      "grad_norm": 49.0,
+      "learning_rate": 9.668218243336317e-08,
+      "loss": 0.7562,
+      "step": 7202
+    },
+    {
+      "epoch": 1.75127644055434,
+      "grad_norm": 21.75,
+      "learning_rate": 9.649590898207412e-08,
+      "loss": 0.8688,
+      "step": 7203
+    },
+    {
+      "epoch": 1.7515195720884997,
+      "grad_norm": 15.0,
+      "learning_rate": 9.630980794189338e-08,
+      "loss": 0.4958,
+      "step": 7204
+    },
+    {
+      "epoch": 1.75176270362266,
+      "grad_norm": 24.625,
+      "learning_rate": 9.612387934063674e-08,
+      "loss": 0.8671,
+      "step": 7205
+    },
+    {
+      "epoch": 1.7520058351568197,
+      "grad_norm": 23.125,
+      "learning_rate": 9.593812320609436e-08,
+      "loss": 0.7222,
+      "step": 7206
+    },
+    {
+      "epoch": 1.75224896669098,
+      "grad_norm": 21.625,
+      "learning_rate": 9.575253956603095e-08,
+      "loss": 0.8224,
+      "step": 7207
+    },
+    {
+      "epoch": 1.7524920982251397,
+      "grad_norm": 13.125,
+      "learning_rate": 9.556712844818502e-08,
+      "loss": 0.278,
+      "step": 7208
+    },
+    {
+      "epoch": 1.7527352297592997,
+      "grad_norm": 22.5,
+      "learning_rate": 9.538188988026928e-08,
+      "loss": 0.7725,
+      "step": 7209
+    },
+    {
+      "epoch": 1.7529783612934597,
+      "grad_norm": 22.625,
+      "learning_rate": 9.519682388997142e-08,
+      "loss": 0.9668,
+      "step": 7210
+    },
+    {
+      "epoch": 1.7532214928276197,
+      "grad_norm": 18.5,
+      "learning_rate": 9.501193050495197e-08,
+      "loss": 0.6622,
+      "step": 7211
+    },
+    {
+      "epoch": 1.7534646243617797,
+      "grad_norm": 22.625,
+      "learning_rate": 9.482720975284715e-08,
+      "loss": 0.6568,
+      "step": 7212
+    },
+    {
+      "epoch": 1.7537077558959397,
+      "grad_norm": 27.125,
+      "learning_rate": 9.464266166126613e-08,
+      "loss": 1.0969,
+      "step": 7213
+    },
+    {
+      "epoch": 1.7539508874300997,
+      "grad_norm": 20.875,
+      "learning_rate": 9.445828625779321e-08,
+      "loss": 0.7734,
+      "step": 7214
+    },
+    {
+      "epoch": 1.7541940189642595,
+      "grad_norm": 19.0,
+      "learning_rate": 9.427408356998624e-08,
+      "loss": 0.3951,
+      "step": 7215
+    },
+    {
+      "epoch": 1.7544371504984198,
+      "grad_norm": 17.875,
+      "learning_rate": 9.409005362537749e-08,
+      "loss": 0.6622,
+      "step": 7216
+    },
+    {
+      "epoch": 1.7546802820325795,
+      "grad_norm": 21.25,
+      "learning_rate": 9.390619645147355e-08,
+      "loss": 0.5264,
+      "step": 7217
+    },
+    {
+      "epoch": 1.7549234135667398,
+      "grad_norm": 22.25,
+      "learning_rate": 9.372251207575483e-08,
+      "loss": 0.5337,
+      "step": 7218
+    },
+    {
+      "epoch": 1.7551665451008995,
+      "grad_norm": 21.25,
+      "learning_rate": 9.353900052567658e-08,
+      "loss": 0.9108,
+      "step": 7219
+    },
+    {
+      "epoch": 1.7554096766350595,
+      "grad_norm": 16.375,
+      "learning_rate": 9.335566182866712e-08,
+      "loss": 0.6408,
+      "step": 7220
+    },
+    {
+      "epoch": 1.7556528081692195,
+      "grad_norm": 15.0,
+      "learning_rate": 9.317249601213025e-08,
+      "loss": 0.5524,
+      "step": 7221
+    },
+    {
+      "epoch": 1.7558959397033795,
+      "grad_norm": 19.125,
+      "learning_rate": 9.298950310344293e-08,
+      "loss": 0.7348,
+      "step": 7222
+    },
+    {
+      "epoch": 1.7561390712375395,
+      "grad_norm": 16.75,
+      "learning_rate": 9.28066831299565e-08,
+      "loss": 0.344,
+      "step": 7223
+    },
+    {
+      "epoch": 1.7563822027716995,
+      "grad_norm": 27.875,
+      "learning_rate": 9.262403611899673e-08,
+      "loss": 0.7012,
+      "step": 7224
+    },
+    {
+      "epoch": 1.7566253343058595,
+      "grad_norm": 19.0,
+      "learning_rate": 9.244156209786345e-08,
+      "loss": 0.5723,
+      "step": 7225
+    },
+    {
+      "epoch": 1.7568684658400193,
+      "grad_norm": 23.625,
+      "learning_rate": 9.225926109383026e-08,
+      "loss": 0.7488,
+      "step": 7226
+    },
+    {
+      "epoch": 1.7571115973741795,
+      "grad_norm": 23.25,
+      "learning_rate": 9.207713313414523e-08,
+      "loss": 0.7095,
+      "step": 7227
+    },
+    {
+      "epoch": 1.7573547289083393,
+      "grad_norm": 18.0,
+      "learning_rate": 9.189517824603103e-08,
+      "loss": 0.6345,
+      "step": 7228
+    },
+    {
+      "epoch": 1.7575978604424995,
+      "grad_norm": 21.75,
+      "learning_rate": 9.171339645668353e-08,
+      "loss": 0.7663,
+      "step": 7229
+    },
+    {
+      "epoch": 1.7578409919766593,
+      "grad_norm": 44.75,
+      "learning_rate": 9.15317877932731e-08,
+      "loss": 0.6386,
+      "step": 7230
+    },
+    {
+      "epoch": 1.7580841235108193,
+      "grad_norm": 15.625,
+      "learning_rate": 9.135035228294453e-08,
+      "loss": 0.4772,
+      "step": 7231
+    },
+    {
+      "epoch": 1.7583272550449793,
+      "grad_norm": 23.125,
+      "learning_rate": 9.116908995281642e-08,
+      "loss": 0.8511,
+      "step": 7232
+    },
+    {
+      "epoch": 1.7585703865791393,
+      "grad_norm": 22.125,
+      "learning_rate": 9.098800082998141e-08,
+      "loss": 0.7585,
+      "step": 7233
+    },
+    {
+      "epoch": 1.7588135181132993,
+      "grad_norm": 30.875,
+      "learning_rate": 9.08070849415063e-08,
+      "loss": 0.6268,
+      "step": 7234
+    },
+    {
+      "epoch": 1.7590566496474591,
+      "grad_norm": 24.25,
+      "learning_rate": 9.062634231443268e-08,
+      "loss": 0.9315,
+      "step": 7235
+    },
+    {
+      "epoch": 1.7592997811816193,
+      "grad_norm": 20.875,
+      "learning_rate": 9.044577297577517e-08,
+      "loss": 0.7464,
+      "step": 7236
+    },
+    {
+      "epoch": 1.7595429127157791,
+      "grad_norm": 21.75,
+      "learning_rate": 9.026537695252302e-08,
+      "loss": 0.4642,
+      "step": 7237
+    },
+    {
+      "epoch": 1.7597860442499393,
+      "grad_norm": 19.375,
+      "learning_rate": 9.008515427163966e-08,
+      "loss": 0.664,
+      "step": 7238
+    },
+    {
+      "epoch": 1.7600291757840991,
+      "grad_norm": 19.75,
+      "learning_rate": 8.990510496006244e-08,
+      "loss": 0.6459,
+      "step": 7239
+    },
+    {
+      "epoch": 1.7602723073182593,
+      "grad_norm": 21.0,
+      "learning_rate": 8.97252290447026e-08,
+      "loss": 0.9834,
+      "step": 7240
+    },
+    {
+      "epoch": 1.7605154388524191,
+      "grad_norm": 24.75,
+      "learning_rate": 8.954552655244627e-08,
+      "loss": 0.8953,
+      "step": 7241
+    },
+    {
+      "epoch": 1.7607585703865791,
+      "grad_norm": 20.375,
+      "learning_rate": 8.93659975101524e-08,
+      "loss": 0.8663,
+      "step": 7242
+    },
+    {
+      "epoch": 1.7610017019207391,
+      "grad_norm": 16.625,
+      "learning_rate": 8.91866419446552e-08,
+      "loss": 0.7115,
+      "step": 7243
+    },
+    {
+      "epoch": 1.7612448334548991,
+      "grad_norm": 16.375,
+      "learning_rate": 8.900745988276227e-08,
+      "loss": 0.7156,
+      "step": 7244
+    },
+    {
+      "epoch": 1.7614879649890591,
+      "grad_norm": 24.625,
+      "learning_rate": 8.882845135125551e-08,
+      "loss": 1.1081,
+      "step": 7245
+    },
+    {
+      "epoch": 1.761731096523219,
+      "grad_norm": 22.625,
+      "learning_rate": 8.864961637689102e-08,
+      "loss": 1.007,
+      "step": 7246
+    },
+    {
+      "epoch": 1.7619742280573791,
+      "grad_norm": 26.75,
+      "learning_rate": 8.847095498639823e-08,
+      "loss": 0.7365,
+      "step": 7247
+    },
+    {
+      "epoch": 1.762217359591539,
+      "grad_norm": 21.125,
+      "learning_rate": 8.829246720648165e-08,
+      "loss": 0.52,
+      "step": 7248
+    },
+    {
+      "epoch": 1.7624604911256991,
+      "grad_norm": 19.0,
+      "learning_rate": 8.811415306381924e-08,
+      "loss": 0.5325,
+      "step": 7249
+    },
+    {
+      "epoch": 1.762703622659859,
+      "grad_norm": 18.375,
+      "learning_rate": 8.793601258506299e-08,
+      "loss": 0.6283,
+      "step": 7250
+    },
+    {
+      "epoch": 1.762946754194019,
+      "grad_norm": 15.375,
+      "learning_rate": 8.775804579683939e-08,
+      "loss": 0.297,
+      "step": 7251
+    },
+    {
+      "epoch": 1.763189885728179,
+      "grad_norm": 19.75,
+      "learning_rate": 8.758025272574854e-08,
+      "loss": 0.8917,
+      "step": 7252
+    },
+    {
+      "epoch": 1.763433017262339,
+      "grad_norm": 15.1875,
+      "learning_rate": 8.740263339836449e-08,
+      "loss": 0.3076,
+      "step": 7253
+    },
+    {
+      "epoch": 1.763676148796499,
+      "grad_norm": 24.375,
+      "learning_rate": 8.722518784123557e-08,
+      "loss": 0.9268,
+      "step": 7254
+    },
+    {
+      "epoch": 1.763919280330659,
+      "grad_norm": 16.75,
+      "learning_rate": 8.704791608088417e-08,
+      "loss": 0.5323,
+      "step": 7255
+    },
+    {
+      "epoch": 1.764162411864819,
+      "grad_norm": 19.25,
+      "learning_rate": 8.687081814380674e-08,
+      "loss": 0.5292,
+      "step": 7256
+    },
+    {
+      "epoch": 1.7644055433989787,
+      "grad_norm": 16.875,
+      "learning_rate": 8.669389405647294e-08,
+      "loss": 0.3933,
+      "step": 7257
+    },
+    {
+      "epoch": 1.764648674933139,
+      "grad_norm": 19.5,
+      "learning_rate": 8.651714384532814e-08,
+      "loss": 1.0579,
+      "step": 7258
+    },
+    {
+      "epoch": 1.7648918064672987,
+      "grad_norm": 23.75,
+      "learning_rate": 8.634056753679024e-08,
+      "loss": 0.5955,
+      "step": 7259
+    },
+    {
+      "epoch": 1.765134938001459,
+      "grad_norm": 21.125,
+      "learning_rate": 8.616416515725174e-08,
+      "loss": 0.8159,
+      "step": 7260
+    },
+    {
+      "epoch": 1.7653780695356187,
+      "grad_norm": 17.75,
+      "learning_rate": 8.598793673307848e-08,
+      "loss": 0.5359,
+      "step": 7261
+    },
+    {
+      "epoch": 1.7656212010697787,
+      "grad_norm": 17.25,
+      "learning_rate": 8.581188229061163e-08,
+      "loss": 0.5799,
+      "step": 7262
+    },
+    {
+      "epoch": 1.7658643326039387,
+      "grad_norm": 19.75,
+      "learning_rate": 8.563600185616513e-08,
+      "loss": 0.6923,
+      "step": 7263
+    },
+    {
+      "epoch": 1.7661074641380987,
+      "grad_norm": 19.25,
+      "learning_rate": 8.546029545602713e-08,
+      "loss": 0.6216,
+      "step": 7264
+    },
+    {
+      "epoch": 1.7663505956722587,
+      "grad_norm": 17.875,
+      "learning_rate": 8.528476311646036e-08,
+      "loss": 0.8677,
+      "step": 7265
+    },
+    {
+      "epoch": 1.7665937272064187,
+      "grad_norm": 22.0,
+      "learning_rate": 8.51094048637012e-08,
+      "loss": 0.8781,
+      "step": 7266
+    },
+    {
+      "epoch": 1.7668368587405787,
+      "grad_norm": 18.125,
+      "learning_rate": 8.493422072395979e-08,
+      "loss": 0.5803,
+      "step": 7267
+    },
+    {
+      "epoch": 1.7670799902747385,
+      "grad_norm": 17.875,
+      "learning_rate": 8.475921072342047e-08,
+      "loss": 0.6115,
+      "step": 7268
+    },
+    {
+      "epoch": 1.7673231218088987,
+      "grad_norm": 24.125,
+      "learning_rate": 8.458437488824162e-08,
+      "loss": 0.549,
+      "step": 7269
+    },
+    {
+      "epoch": 1.7675662533430585,
+      "grad_norm": 19.625,
+      "learning_rate": 8.440971324455538e-08,
+      "loss": 0.8231,
+      "step": 7270
+    },
+    {
+      "epoch": 1.7678093848772187,
+      "grad_norm": 22.5,
+      "learning_rate": 8.423522581846783e-08,
+      "loss": 0.9291,
+      "step": 7271
+    },
+    {
+      "epoch": 1.7680525164113785,
+      "grad_norm": 17.75,
+      "learning_rate": 8.406091263605934e-08,
+      "loss": 1.018,
+      "step": 7272
+    },
+    {
+      "epoch": 1.7682956479455385,
+      "grad_norm": 17.375,
+      "learning_rate": 8.388677372338366e-08,
+      "loss": 0.6525,
+      "step": 7273
+    },
+    {
+      "epoch": 1.7685387794796985,
+      "grad_norm": 23.5,
+      "learning_rate": 8.371280910646914e-08,
+      "loss": 0.5221,
+      "step": 7274
+    },
+    {
+      "epoch": 1.7687819110138585,
+      "grad_norm": 19.5,
+      "learning_rate": 8.353901881131804e-08,
+      "loss": 0.894,
+      "step": 7275
+    },
+    {
+      "epoch": 1.7690250425480185,
+      "grad_norm": 15.125,
+      "learning_rate": 8.336540286390596e-08,
+      "loss": 0.9805,
+      "step": 7276
+    },
+    {
+      "epoch": 1.7692681740821785,
+      "grad_norm": 17.125,
+      "learning_rate": 8.319196129018298e-08,
+      "loss": 0.4495,
+      "step": 7277
+    },
+    {
+      "epoch": 1.7695113056163385,
+      "grad_norm": 223.0,
+      "learning_rate": 8.301869411607266e-08,
+      "loss": 0.6571,
+      "step": 7278
+    },
+    {
+      "epoch": 1.7697544371504983,
+      "grad_norm": 18.375,
+      "learning_rate": 8.284560136747318e-08,
+      "loss": 0.5611,
+      "step": 7279
+    },
+    {
+      "epoch": 1.7699975686846585,
+      "grad_norm": 21.625,
+      "learning_rate": 8.26726830702558e-08,
+      "loss": 0.9508,
+      "step": 7280
+    },
+    {
+      "epoch": 1.7702407002188183,
+      "grad_norm": 16.625,
+      "learning_rate": 8.249993925026636e-08,
+      "loss": 0.5974,
+      "step": 7281
+    },
+    {
+      "epoch": 1.7704838317529785,
+      "grad_norm": 18.625,
+      "learning_rate": 8.232736993332464e-08,
+      "loss": 0.5968,
+      "step": 7282
+    },
+    {
+      "epoch": 1.7707269632871383,
+      "grad_norm": 26.375,
+      "learning_rate": 8.215497514522386e-08,
+      "loss": 0.9187,
+      "step": 7283
+    },
+    {
+      "epoch": 1.7709700948212983,
+      "grad_norm": 23.0,
+      "learning_rate": 8.198275491173121e-08,
+      "loss": 0.8458,
+      "step": 7284
+    },
+    {
+      "epoch": 1.7712132263554583,
+      "grad_norm": 15.9375,
+      "learning_rate": 8.181070925858847e-08,
+      "loss": 0.2648,
+      "step": 7285
+    },
+    {
+      "epoch": 1.7714563578896183,
+      "grad_norm": 22.875,
+      "learning_rate": 8.163883821151047e-08,
+      "loss": 0.6686,
+      "step": 7286
+    },
+    {
+      "epoch": 1.7716994894237783,
+      "grad_norm": 18.625,
+      "learning_rate": 8.146714179618653e-08,
+      "loss": 0.854,
+      "step": 7287
+    },
+    {
+      "epoch": 1.771942620957938,
+      "grad_norm": 19.625,
+      "learning_rate": 8.129562003827903e-08,
+      "loss": 0.4788,
+      "step": 7288
+    },
+    {
+      "epoch": 1.7721857524920983,
+      "grad_norm": 26.125,
+      "learning_rate": 8.112427296342568e-08,
+      "loss": 1.1571,
+      "step": 7289
+    },
+    {
+      "epoch": 1.772428884026258,
+      "grad_norm": 18.5,
+      "learning_rate": 8.095310059723694e-08,
+      "loss": 0.8155,
+      "step": 7290
+    },
+    {
+      "epoch": 1.7726720155604183,
+      "grad_norm": 14.875,
+      "learning_rate": 8.078210296529734e-08,
+      "loss": 1.2172,
+      "step": 7291
+    },
+    {
+      "epoch": 1.772915147094578,
+      "grad_norm": 20.25,
+      "learning_rate": 8.061128009316577e-08,
+      "loss": 0.7776,
+      "step": 7292
+    },
+    {
+      "epoch": 1.773158278628738,
+      "grad_norm": 17.125,
+      "learning_rate": 8.044063200637428e-08,
+      "loss": 0.4722,
+      "step": 7293
+    },
+    {
+      "epoch": 1.773401410162898,
+      "grad_norm": 17.5,
+      "learning_rate": 8.027015873042942e-08,
+      "loss": 0.4354,
+      "step": 7294
+    },
+    {
+      "epoch": 1.773644541697058,
+      "grad_norm": 18.125,
+      "learning_rate": 8.009986029081097e-08,
+      "loss": 0.4938,
+      "step": 7295
+    },
+    {
+      "epoch": 1.773887673231218,
+      "grad_norm": 24.375,
+      "learning_rate": 7.992973671297338e-08,
+      "loss": 0.838,
+      "step": 7296
+    },
+    {
+      "epoch": 1.774130804765378,
+      "grad_norm": 23.625,
+      "learning_rate": 7.975978802234468e-08,
+      "loss": 0.9385,
+      "step": 7297
+    },
+    {
+      "epoch": 1.774373936299538,
+      "grad_norm": 17.5,
+      "learning_rate": 7.959001424432608e-08,
+      "loss": 0.7159,
+      "step": 7298
+    },
+    {
+      "epoch": 1.7746170678336979,
+      "grad_norm": 22.5,
+      "learning_rate": 7.942041540429379e-08,
+      "loss": 0.6502,
+      "step": 7299
+    },
+    {
+      "epoch": 1.774860199367858,
+      "grad_norm": 19.625,
+      "learning_rate": 7.925099152759714e-08,
+      "loss": 0.4799,
+      "step": 7300
+    },
+    {
+      "epoch": 1.7751033309020179,
+      "grad_norm": 20.875,
+      "learning_rate": 7.908174263955917e-08,
+      "loss": 0.6748,
+      "step": 7301
+    },
+    {
+      "epoch": 1.775346462436178,
+      "grad_norm": 19.375,
+      "learning_rate": 7.891266876547717e-08,
+      "loss": 0.9074,
+      "step": 7302
+    },
+    {
+      "epoch": 1.7755895939703379,
+      "grad_norm": 24.375,
+      "learning_rate": 7.874376993062205e-08,
+      "loss": 0.488,
+      "step": 7303
+    },
+    {
+      "epoch": 1.7758327255044979,
+      "grad_norm": 21.25,
+      "learning_rate": 7.857504616023915e-08,
+      "loss": 0.7793,
+      "step": 7304
+    },
+    {
+      "epoch": 1.776075857038658,
+      "grad_norm": 21.25,
+      "learning_rate": 7.840649747954648e-08,
+      "loss": 0.7053,
+      "step": 7305
+    },
+    {
+      "epoch": 1.776318988572818,
+      "grad_norm": 16.625,
+      "learning_rate": 7.823812391373711e-08,
+      "loss": 0.7816,
+      "step": 7306
+    },
+    {
+      "epoch": 1.776562120106978,
+      "grad_norm": 19.875,
+      "learning_rate": 7.806992548797729e-08,
+      "loss": 0.642,
+      "step": 7307
+    },
+    {
+      "epoch": 1.776805251641138,
+      "grad_norm": 19.375,
+      "learning_rate": 7.790190222740662e-08,
+      "loss": 0.8313,
+      "step": 7308
+    },
+    {
+      "epoch": 1.777048383175298,
+      "grad_norm": 19.25,
+      "learning_rate": 7.773405415713988e-08,
+      "loss": 0.5274,
+      "step": 7309
+    },
+    {
+      "epoch": 1.7772915147094577,
+      "grad_norm": 21.25,
+      "learning_rate": 7.756638130226438e-08,
+      "loss": 0.4838,
+      "step": 7310
+    },
+    {
+      "epoch": 1.777534646243618,
+      "grad_norm": 19.625,
+      "learning_rate": 7.739888368784171e-08,
+      "loss": 0.5142,
+      "step": 7311
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 35.25,
+      "learning_rate": 7.723156133890727e-08,
+      "loss": 0.8411,
+      "step": 7312
+    },
+    {
+      "epoch": 1.778020909311938,
+      "grad_norm": 14.625,
+      "learning_rate": 7.706441428047065e-08,
+      "loss": 0.3793,
+      "step": 7313
+    },
+    {
+      "epoch": 1.7782640408460977,
+      "grad_norm": 15.0,
+      "learning_rate": 7.689744253751463e-08,
+      "loss": 0.6047,
+      "step": 7314
+    },
+    {
+      "epoch": 1.7785071723802577,
+      "grad_norm": 14.875,
+      "learning_rate": 7.673064613499578e-08,
+      "loss": 0.6834,
+      "step": 7315
+    },
+    {
+      "epoch": 1.7787503039144177,
+      "grad_norm": 36.5,
+      "learning_rate": 7.656402509784527e-08,
+      "loss": 0.6551,
+      "step": 7316
+    },
+    {
+      "epoch": 1.7789934354485777,
+      "grad_norm": 24.625,
+      "learning_rate": 7.639757945096693e-08,
+      "loss": 0.5765,
+      "step": 7317
+    },
+    {
+      "epoch": 1.7792365669827377,
+      "grad_norm": 29.0,
+      "learning_rate": 7.623130921923932e-08,
+      "loss": 0.974,
+      "step": 7318
+    },
+    {
+      "epoch": 1.7794796985168977,
+      "grad_norm": 18.25,
+      "learning_rate": 7.606521442751383e-08,
+      "loss": 0.816,
+      "step": 7319
+    },
+    {
+      "epoch": 1.7797228300510577,
+      "grad_norm": 14.4375,
+      "learning_rate": 7.589929510061683e-08,
+      "loss": 0.3443,
+      "step": 7320
+    },
+    {
+      "epoch": 1.7799659615852175,
+      "grad_norm": 22.125,
+      "learning_rate": 7.57335512633478e-08,
+      "loss": 0.9432,
+      "step": 7321
+    },
+    {
+      "epoch": 1.7802090931193777,
+      "grad_norm": 18.375,
+      "learning_rate": 7.556798294047943e-08,
+      "loss": 0.5595,
+      "step": 7322
+    },
+    {
+      "epoch": 1.7804522246535375,
+      "grad_norm": 17.5,
+      "learning_rate": 7.540259015675955e-08,
+      "loss": 0.2119,
+      "step": 7323
+    },
+    {
+      "epoch": 1.7806953561876977,
+      "grad_norm": 23.125,
+      "learning_rate": 7.523737293690838e-08,
+      "loss": 0.6449,
+      "step": 7324
+    },
+    {
+      "epoch": 1.7809384877218575,
+      "grad_norm": 17.75,
+      "learning_rate": 7.507233130562064e-08,
+      "loss": 0.6951,
+      "step": 7325
+    },
+    {
+      "epoch": 1.7811816192560175,
+      "grad_norm": 18.875,
+      "learning_rate": 7.490746528756476e-08,
+      "loss": 0.8004,
+      "step": 7326
+    },
+    {
+      "epoch": 1.7814247507901775,
+      "grad_norm": 13.0,
+      "learning_rate": 7.474277490738257e-08,
+      "loss": 0.4268,
+      "step": 7327
+    },
+    {
+      "epoch": 1.7816678823243375,
+      "grad_norm": 20.875,
+      "learning_rate": 7.457826018969037e-08,
+      "loss": 1.2651,
+      "step": 7328
+    },
+    {
+      "epoch": 1.7819110138584975,
+      "grad_norm": 26.625,
+      "learning_rate": 7.441392115907706e-08,
+      "loss": 0.8745,
+      "step": 7329
+    },
+    {
+      "epoch": 1.7821541453926573,
+      "grad_norm": 19.375,
+      "learning_rate": 7.424975784010662e-08,
+      "loss": 0.4935,
+      "step": 7330
+    },
+    {
+      "epoch": 1.7823972769268175,
+      "grad_norm": 23.375,
+      "learning_rate": 7.40857702573157e-08,
+      "loss": 0.7011,
+      "step": 7331
+    },
+    {
+      "epoch": 1.7826404084609773,
+      "grad_norm": 15.0,
+      "learning_rate": 7.392195843521507e-08,
+      "loss": 0.2392,
+      "step": 7332
+    },
+    {
+      "epoch": 1.7828835399951375,
+      "grad_norm": 18.25,
+      "learning_rate": 7.375832239828948e-08,
+      "loss": 0.7848,
+      "step": 7333
+    },
+    {
+      "epoch": 1.7831266715292973,
+      "grad_norm": 15.9375,
+      "learning_rate": 7.35948621709967e-08,
+      "loss": 0.3441,
+      "step": 7334
+    },
+    {
+      "epoch": 1.7833698030634575,
+      "grad_norm": 14.6875,
+      "learning_rate": 7.34315777777693e-08,
+      "loss": 0.4485,
+      "step": 7335
+    },
+    {
+      "epoch": 1.7836129345976173,
+      "grad_norm": 20.25,
+      "learning_rate": 7.326846924301245e-08,
+      "loss": 0.6855,
+      "step": 7336
+    },
+    {
+      "epoch": 1.7838560661317773,
+      "grad_norm": 23.625,
+      "learning_rate": 7.310553659110584e-08,
+      "loss": 0.9441,
+      "step": 7337
+    },
+    {
+      "epoch": 1.7840991976659373,
+      "grad_norm": 17.375,
+      "learning_rate": 7.294277984640261e-08,
+      "loss": 0.3482,
+      "step": 7338
+    },
+    {
+      "epoch": 1.7843423292000973,
+      "grad_norm": 18.75,
+      "learning_rate": 7.2780199033229e-08,
+      "loss": 0.959,
+      "step": 7339
+    },
+    {
+      "epoch": 1.7845854607342573,
+      "grad_norm": 18.0,
+      "learning_rate": 7.261779417588628e-08,
+      "loss": 0.5354,
+      "step": 7340
+    },
+    {
+      "epoch": 1.784828592268417,
+      "grad_norm": 20.125,
+      "learning_rate": 7.245556529864834e-08,
+      "loss": 0.2532,
+      "step": 7341
+    },
+    {
+      "epoch": 1.7850717238025773,
+      "grad_norm": 25.25,
+      "learning_rate": 7.229351242576274e-08,
+      "loss": 0.5816,
+      "step": 7342
+    },
+    {
+      "epoch": 1.785314855336737,
+      "grad_norm": 19.0,
+      "learning_rate": 7.21316355814515e-08,
+      "loss": 0.9048,
+      "step": 7343
+    },
+    {
+      "epoch": 1.7855579868708973,
+      "grad_norm": 21.375,
+      "learning_rate": 7.196993478990999e-08,
+      "loss": 0.6518,
+      "step": 7344
+    },
+    {
+      "epoch": 1.785801118405057,
+      "grad_norm": 15.25,
+      "learning_rate": 7.180841007530693e-08,
+      "loss": 0.398,
+      "step": 7345
+    },
+    {
+      "epoch": 1.786044249939217,
+      "grad_norm": 18.875,
+      "learning_rate": 7.164706146178493e-08,
+      "loss": 0.6759,
+      "step": 7346
+    },
+    {
+      "epoch": 1.786287381473377,
+      "grad_norm": 20.5,
+      "learning_rate": 7.14858889734607e-08,
+      "loss": 1.0445,
+      "step": 7347
+    },
+    {
+      "epoch": 1.786530513007537,
+      "grad_norm": 17.125,
+      "learning_rate": 7.132489263442399e-08,
+      "loss": 0.9325,
+      "step": 7348
+    },
+    {
+      "epoch": 1.786773644541697,
+      "grad_norm": 17.25,
+      "learning_rate": 7.116407246873818e-08,
+      "loss": 0.4169,
+      "step": 7349
+    },
+    {
+      "epoch": 1.787016776075857,
+      "grad_norm": 21.0,
+      "learning_rate": 7.100342850044101e-08,
+      "loss": 0.8287,
+      "step": 7350
+    },
+    {
+      "epoch": 1.787259907610017,
+      "grad_norm": 16.0,
+      "learning_rate": 7.084296075354367e-08,
+      "loss": 0.3084,
+      "step": 7351
+    },
+    {
+      "epoch": 1.7875030391441769,
+      "grad_norm": 20.375,
+      "learning_rate": 7.068266925203057e-08,
+      "loss": 0.6098,
+      "step": 7352
+    },
+    {
+      "epoch": 1.787746170678337,
+      "grad_norm": 22.75,
+      "learning_rate": 7.052255401985994e-08,
+      "loss": 1.1271,
+      "step": 7353
+    },
+    {
+      "epoch": 1.7879893022124969,
+      "grad_norm": 14.0625,
+      "learning_rate": 7.036261508096387e-08,
+      "loss": 0.2789,
+      "step": 7354
+    },
+    {
+      "epoch": 1.788232433746657,
+      "grad_norm": 21.125,
+      "learning_rate": 7.020285245924824e-08,
+      "loss": 0.8638,
+      "step": 7355
+    },
+    {
+      "epoch": 1.7884755652808169,
+      "grad_norm": 14.5625,
+      "learning_rate": 7.004326617859187e-08,
+      "loss": 0.27,
+      "step": 7356
+    },
+    {
+      "epoch": 1.7887186968149769,
+      "grad_norm": 25.375,
+      "learning_rate": 6.988385626284802e-08,
+      "loss": 0.6371,
+      "step": 7357
+    },
+    {
+      "epoch": 1.7889618283491369,
+      "grad_norm": 17.625,
+      "learning_rate": 6.972462273584307e-08,
+      "loss": 0.6549,
+      "step": 7358
+    },
+    {
+      "epoch": 1.7892049598832969,
+      "grad_norm": 22.625,
+      "learning_rate": 6.956556562137743e-08,
+      "loss": 0.6225,
+      "step": 7359
+    },
+    {
+      "epoch": 1.7894480914174569,
+      "grad_norm": 19.25,
+      "learning_rate": 6.94066849432247e-08,
+      "loss": 0.7344,
+      "step": 7360
+    },
+    {
+      "epoch": 1.7896912229516169,
+      "grad_norm": 23.75,
+      "learning_rate": 6.924798072513256e-08,
+      "loss": 1.1601,
+      "step": 7361
+    },
+    {
+      "epoch": 1.7899343544857769,
+      "grad_norm": 21.125,
+      "learning_rate": 6.908945299082203e-08,
+      "loss": 0.8278,
+      "step": 7362
+    },
+    {
+      "epoch": 1.7901774860199366,
+      "grad_norm": 18.375,
+      "learning_rate": 6.893110176398765e-08,
+      "loss": 0.6753,
+      "step": 7363
+    },
+    {
+      "epoch": 1.7904206175540969,
+      "grad_norm": 17.375,
+      "learning_rate": 6.877292706829796e-08,
+      "loss": 0.5124,
+      "step": 7364
+    },
+    {
+      "epoch": 1.7906637490882567,
+      "grad_norm": 22.5,
+      "learning_rate": 6.861492892739477e-08,
+      "loss": 0.8909,
+      "step": 7365
+    },
+    {
+      "epoch": 1.7909068806224169,
+      "grad_norm": 26.625,
+      "learning_rate": 6.845710736489375e-08,
+      "loss": 0.6656,
+      "step": 7366
+    },
+    {
+      "epoch": 1.7911500121565767,
+      "grad_norm": 21.5,
+      "learning_rate": 6.829946240438382e-08,
+      "loss": 0.67,
+      "step": 7367
+    },
+    {
+      "epoch": 1.7913931436907367,
+      "grad_norm": 18.375,
+      "learning_rate": 6.81419940694282e-08,
+      "loss": 0.783,
+      "step": 7368
+    },
+    {
+      "epoch": 1.7916362752248967,
+      "grad_norm": 15.5625,
+      "learning_rate": 6.79847023835628e-08,
+      "loss": 0.3322,
+      "step": 7369
+    },
+    {
+      "epoch": 1.7918794067590567,
+      "grad_norm": 23.75,
+      "learning_rate": 6.782758737029771e-08,
+      "loss": 0.6024,
+      "step": 7370
+    },
+    {
+      "epoch": 1.7921225382932167,
+      "grad_norm": 18.5,
+      "learning_rate": 6.767064905311649e-08,
+      "loss": 0.672,
+      "step": 7371
+    },
+    {
+      "epoch": 1.7923656698273767,
+      "grad_norm": 15.875,
+      "learning_rate": 6.751388745547649e-08,
+      "loss": 0.4361,
+      "step": 7372
+    },
+    {
+      "epoch": 1.7926088013615367,
+      "grad_norm": 20.25,
+      "learning_rate": 6.7357302600808e-08,
+      "loss": 0.6361,
+      "step": 7373
+    },
+    {
+      "epoch": 1.7928519328956964,
+      "grad_norm": 18.875,
+      "learning_rate": 6.720089451251563e-08,
+      "loss": 0.7835,
+      "step": 7374
+    },
+    {
+      "epoch": 1.7930950644298567,
+      "grad_norm": 26.375,
+      "learning_rate": 6.704466321397734e-08,
+      "loss": 0.8165,
+      "step": 7375
+    },
+    {
+      "epoch": 1.7933381959640164,
+      "grad_norm": 25.25,
+      "learning_rate": 6.688860872854446e-08,
+      "loss": 1.1557,
+      "step": 7376
+    },
+    {
+      "epoch": 1.7935813274981767,
+      "grad_norm": 23.125,
+      "learning_rate": 6.673273107954195e-08,
+      "loss": 0.5129,
+      "step": 7377
+    },
+    {
+      "epoch": 1.7938244590323364,
+      "grad_norm": 21.375,
+      "learning_rate": 6.657703029026865e-08,
+      "loss": 0.832,
+      "step": 7378
+    },
+    {
+      "epoch": 1.7940675905664965,
+      "grad_norm": 19.375,
+      "learning_rate": 6.642150638399653e-08,
+      "loss": 0.4706,
+      "step": 7379
+    },
+    {
+      "epoch": 1.7943107221006565,
+      "grad_norm": 21.125,
+      "learning_rate": 6.626615938397127e-08,
+      "loss": 0.7821,
+      "step": 7380
+    },
+    {
+      "epoch": 1.7945538536348165,
+      "grad_norm": 22.375,
+      "learning_rate": 6.611098931341237e-08,
+      "loss": 0.4564,
+      "step": 7381
+    },
+    {
+      "epoch": 1.7947969851689765,
+      "grad_norm": 19.0,
+      "learning_rate": 6.595599619551266e-08,
+      "loss": 0.6673,
+      "step": 7382
+    },
+    {
+      "epoch": 1.7950401167031362,
+      "grad_norm": 24.125,
+      "learning_rate": 6.580118005343847e-08,
+      "loss": 0.8958,
+      "step": 7383
+    },
+    {
+      "epoch": 1.7952832482372965,
+      "grad_norm": 17.25,
+      "learning_rate": 6.564654091032949e-08,
+      "loss": 0.5238,
+      "step": 7384
+    },
+    {
+      "epoch": 1.7955263797714562,
+      "grad_norm": 23.75,
+      "learning_rate": 6.549207878929972e-08,
+      "loss": 0.7977,
+      "step": 7385
+    },
+    {
+      "epoch": 1.7957695113056165,
+      "grad_norm": 18.25,
+      "learning_rate": 6.533779371343599e-08,
+      "loss": 0.7698,
+      "step": 7386
+    },
+    {
+      "epoch": 1.7960126428397762,
+      "grad_norm": 18.625,
+      "learning_rate": 6.518368570579859e-08,
+      "loss": 0.7963,
+      "step": 7387
+    },
+    {
+      "epoch": 1.7962557743739365,
+      "grad_norm": 20.0,
+      "learning_rate": 6.502975478942186e-08,
+      "loss": 0.8037,
+      "step": 7388
+    },
+    {
+      "epoch": 1.7964989059080962,
+      "grad_norm": 15.75,
+      "learning_rate": 6.487600098731353e-08,
+      "loss": 0.6715,
+      "step": 7389
+    },
+    {
+      "epoch": 1.7967420374422562,
+      "grad_norm": 18.125,
+      "learning_rate": 6.47224243224548e-08,
+      "loss": 0.969,
+      "step": 7390
+    },
+    {
+      "epoch": 1.7969851689764162,
+      "grad_norm": 22.625,
+      "learning_rate": 6.456902481779992e-08,
+      "loss": 0.5991,
+      "step": 7391
+    },
+    {
+      "epoch": 1.7972283005105762,
+      "grad_norm": 18.625,
+      "learning_rate": 6.441580249627751e-08,
+      "loss": 0.4721,
+      "step": 7392
+    },
+    {
+      "epoch": 1.7974714320447363,
+      "grad_norm": 26.25,
+      "learning_rate": 6.426275738078928e-08,
+      "loss": 0.8928,
+      "step": 7393
+    },
+    {
+      "epoch": 1.797714563578896,
+      "grad_norm": 19.0,
+      "learning_rate": 6.410988949421007e-08,
+      "loss": 0.3274,
+      "step": 7394
+    },
+    {
+      "epoch": 1.7979576951130563,
+      "grad_norm": 21.125,
+      "learning_rate": 6.395719885938914e-08,
+      "loss": 0.6072,
+      "step": 7395
+    },
+    {
+      "epoch": 1.798200826647216,
+      "grad_norm": 25.875,
+      "learning_rate": 6.380468549914837e-08,
+      "loss": 0.8625,
+      "step": 7396
+    },
+    {
+      "epoch": 1.7984439581813763,
+      "grad_norm": 23.125,
+      "learning_rate": 6.365234943628382e-08,
+      "loss": 0.6045,
+      "step": 7397
+    },
+    {
+      "epoch": 1.798687089715536,
+      "grad_norm": 23.5,
+      "learning_rate": 6.350019069356436e-08,
+      "loss": 0.9674,
+      "step": 7398
+    },
+    {
+      "epoch": 1.798930221249696,
+      "grad_norm": 17.125,
+      "learning_rate": 6.33482092937332e-08,
+      "loss": 0.8304,
+      "step": 7399
+    },
+    {
+      "epoch": 1.799173352783856,
+      "grad_norm": 21.75,
+      "learning_rate": 6.319640525950632e-08,
+      "loss": 0.7151,
+      "step": 7400
+    },
+    {
+      "epoch": 1.799416484318016,
+      "grad_norm": 20.875,
+      "learning_rate": 6.304477861357322e-08,
+      "loss": 0.6211,
+      "step": 7401
+    },
+    {
+      "epoch": 1.799659615852176,
+      "grad_norm": 17.25,
+      "learning_rate": 6.289332937859757e-08,
+      "loss": 0.45,
+      "step": 7402
+    },
+    {
+      "epoch": 1.799902747386336,
+      "grad_norm": 20.0,
+      "learning_rate": 6.274205757721599e-08,
+      "loss": 0.4845,
+      "step": 7403
+    },
+    {
+      "epoch": 1.800145878920496,
+      "grad_norm": 18.875,
+      "learning_rate": 6.259096323203832e-08,
+      "loss": 0.5576,
+      "step": 7404
+    },
+    {
+      "epoch": 1.8003890104546558,
+      "grad_norm": 16.75,
+      "learning_rate": 6.244004636564855e-08,
+      "loss": 0.3467,
+      "step": 7405
+    },
+    {
+      "epoch": 1.800632141988816,
+      "grad_norm": 15.6875,
+      "learning_rate": 6.228930700060379e-08,
+      "loss": 0.3939,
+      "step": 7406
+    },
+    {
+      "epoch": 1.8008752735229758,
+      "grad_norm": 21.125,
+      "learning_rate": 6.21387451594345e-08,
+      "loss": 1.0078,
+      "step": 7407
+    },
+    {
+      "epoch": 1.801118405057136,
+      "grad_norm": 22.75,
+      "learning_rate": 6.198836086464474e-08,
+      "loss": 0.9298,
+      "step": 7408
+    },
+    {
+      "epoch": 1.8013615365912958,
+      "grad_norm": 21.5,
+      "learning_rate": 6.183815413871238e-08,
+      "loss": 0.651,
+      "step": 7409
+    },
+    {
+      "epoch": 1.8016046681254558,
+      "grad_norm": 17.0,
+      "learning_rate": 6.168812500408808e-08,
+      "loss": 0.8019,
+      "step": 7410
+    },
+    {
+      "epoch": 1.8018477996596158,
+      "grad_norm": 26.25,
+      "learning_rate": 6.15382734831961e-08,
+      "loss": 1.1185,
+      "step": 7411
+    },
+    {
+      "epoch": 1.8020909311937758,
+      "grad_norm": 19.0,
+      "learning_rate": 6.138859959843466e-08,
+      "loss": 0.5034,
+      "step": 7412
+    },
+    {
+      "epoch": 1.8023340627279358,
+      "grad_norm": 18.625,
+      "learning_rate": 6.123910337217528e-08,
+      "loss": 0.691,
+      "step": 7413
+    },
+    {
+      "epoch": 1.8025771942620958,
+      "grad_norm": 26.375,
+      "learning_rate": 6.108978482676233e-08,
+      "loss": 1.2395,
+      "step": 7414
+    },
+    {
+      "epoch": 1.8028203257962558,
+      "grad_norm": 16.625,
+      "learning_rate": 6.094064398451421e-08,
+      "loss": 0.4114,
+      "step": 7415
+    },
+    {
+      "epoch": 1.8030634573304156,
+      "grad_norm": 22.875,
+      "learning_rate": 6.07916808677228e-08,
+      "loss": 0.596,
+      "step": 7416
+    },
+    {
+      "epoch": 1.8033065888645758,
+      "grad_norm": 21.125,
+      "learning_rate": 6.064289549865293e-08,
+      "loss": 0.4784,
+      "step": 7417
+    },
+    {
+      "epoch": 1.8035497203987356,
+      "grad_norm": 18.0,
+      "learning_rate": 6.049428789954307e-08,
+      "loss": 0.7692,
+      "step": 7418
+    },
+    {
+      "epoch": 1.8037928519328958,
+      "grad_norm": 15.25,
+      "learning_rate": 6.034585809260543e-08,
+      "loss": 0.3401,
+      "step": 7419
+    },
+    {
+      "epoch": 1.8040359834670556,
+      "grad_norm": 19.0,
+      "learning_rate": 6.019760610002548e-08,
+      "loss": 0.4493,
+      "step": 7420
+    },
+    {
+      "epoch": 1.8042791150012156,
+      "grad_norm": 17.875,
+      "learning_rate": 6.004953194396187e-08,
+      "loss": 0.6573,
+      "step": 7421
+    },
+    {
+      "epoch": 1.8045222465353756,
+      "grad_norm": 15.4375,
+      "learning_rate": 5.990163564654663e-08,
+      "loss": 0.3954,
+      "step": 7422
+    },
+    {
+      "epoch": 1.8047653780695356,
+      "grad_norm": 16.875,
+      "learning_rate": 5.975391722988597e-08,
+      "loss": 0.4296,
+      "step": 7423
+    },
+    {
+      "epoch": 1.8050085096036956,
+      "grad_norm": 25.75,
+      "learning_rate": 5.96063767160586e-08,
+      "loss": 0.6389,
+      "step": 7424
+    },
+    {
+      "epoch": 1.8052516411378556,
+      "grad_norm": 18.625,
+      "learning_rate": 5.9459014127116743e-08,
+      "loss": 0.538,
+      "step": 7425
+    },
+    {
+      "epoch": 1.8054947726720156,
+      "grad_norm": 20.875,
+      "learning_rate": 5.931182948508696e-08,
+      "loss": 0.7424,
+      "step": 7426
+    },
+    {
+      "epoch": 1.8057379042061754,
+      "grad_norm": 23.25,
+      "learning_rate": 5.916482281196775e-08,
+      "loss": 0.7739,
+      "step": 7427
+    },
+    {
+      "epoch": 1.8059810357403356,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.901799412973252e-08,
+      "loss": 0.3448,
+      "step": 7428
+    },
+    {
+      "epoch": 1.8062241672744954,
+      "grad_norm": 21.0,
+      "learning_rate": 5.8871343460326916e-08,
+      "loss": 0.6944,
+      "step": 7429
+    },
+    {
+      "epoch": 1.8064672988086556,
+      "grad_norm": 16.375,
+      "learning_rate": 5.872487082567061e-08,
+      "loss": 0.5195,
+      "step": 7430
+    },
+    {
+      "epoch": 1.8067104303428154,
+      "grad_norm": 21.0,
+      "learning_rate": 5.857857624765637e-08,
+      "loss": 0.4985,
+      "step": 7431
+    },
+    {
+      "epoch": 1.8069535618769754,
+      "grad_norm": 18.75,
+      "learning_rate": 5.8432459748150315e-08,
+      "loss": 0.7131,
+      "step": 7432
+    },
+    {
+      "epoch": 1.8071966934111354,
+      "grad_norm": 21.625,
+      "learning_rate": 5.8286521348992484e-08,
+      "loss": 0.3244,
+      "step": 7433
+    },
+    {
+      "epoch": 1.8074398249452954,
+      "grad_norm": 17.875,
+      "learning_rate": 5.814076107199557e-08,
+      "loss": 1.1122,
+      "step": 7434
+    },
+    {
+      "epoch": 1.8076829564794554,
+      "grad_norm": 24.875,
+      "learning_rate": 5.799517893894588e-08,
+      "loss": 0.5813,
+      "step": 7435
+    },
+    {
+      "epoch": 1.8079260880136152,
+      "grad_norm": 22.375,
+      "learning_rate": 5.7849774971603376e-08,
+      "loss": 0.7805,
+      "step": 7436
+    },
+    {
+      "epoch": 1.8081692195477754,
+      "grad_norm": 20.75,
+      "learning_rate": 5.7704549191701236e-08,
+      "loss": 0.7114,
+      "step": 7437
+    },
+    {
+      "epoch": 1.8084123510819352,
+      "grad_norm": 19.375,
+      "learning_rate": 5.755950162094598e-08,
+      "loss": 0.837,
+      "step": 7438
+    },
+    {
+      "epoch": 1.8086554826160954,
+      "grad_norm": 19.0,
+      "learning_rate": 5.7414632281017206e-08,
+      "loss": 0.5499,
+      "step": 7439
+    },
+    {
+      "epoch": 1.8088986141502552,
+      "grad_norm": 19.375,
+      "learning_rate": 5.72699411935683e-08,
+      "loss": 0.4997,
+      "step": 7440
+    },
+    {
+      "epoch": 1.8091417456844152,
+      "grad_norm": 18.875,
+      "learning_rate": 5.712542838022597e-08,
+      "loss": 0.5257,
+      "step": 7441
+    },
+    {
+      "epoch": 1.8093848772185752,
+      "grad_norm": 20.125,
+      "learning_rate": 5.6981093862589904e-08,
+      "loss": 0.9424,
+      "step": 7442
+    },
+    {
+      "epoch": 1.8096280087527352,
+      "grad_norm": 26.125,
+      "learning_rate": 5.6836937662233385e-08,
+      "loss": 0.6369,
+      "step": 7443
+    },
+    {
+      "epoch": 1.8098711402868952,
+      "grad_norm": 22.75,
+      "learning_rate": 5.6692959800703356e-08,
+      "loss": 1.0135,
+      "step": 7444
+    },
+    {
+      "epoch": 1.8101142718210552,
+      "grad_norm": 16.125,
+      "learning_rate": 5.654916029951968e-08,
+      "loss": 0.7887,
+      "step": 7445
+    },
+    {
+      "epoch": 1.8103574033552152,
+      "grad_norm": 22.625,
+      "learning_rate": 5.640553918017544e-08,
+      "loss": 0.8612,
+      "step": 7446
+    },
+    {
+      "epoch": 1.810600534889375,
+      "grad_norm": 19.25,
+      "learning_rate": 5.6262096464137635e-08,
+      "loss": 0.6758,
+      "step": 7447
+    },
+    {
+      "epoch": 1.8108436664235352,
+      "grad_norm": 25.0,
+      "learning_rate": 5.6118832172845914e-08,
+      "loss": 0.7105,
+      "step": 7448
+    },
+    {
+      "epoch": 1.811086797957695,
+      "grad_norm": 19.625,
+      "learning_rate": 5.5975746327713854e-08,
+      "loss": 0.6348,
+      "step": 7449
+    },
+    {
+      "epoch": 1.8113299294918552,
+      "grad_norm": 21.25,
+      "learning_rate": 5.583283895012781e-08,
+      "loss": 1.0206,
+      "step": 7450
+    },
+    {
+      "epoch": 1.811573061026015,
+      "grad_norm": 20.125,
+      "learning_rate": 5.56901100614482e-08,
+      "loss": 0.6636,
+      "step": 7451
+    },
+    {
+      "epoch": 1.811816192560175,
+      "grad_norm": 17.875,
+      "learning_rate": 5.55475596830081e-08,
+      "loss": 0.9417,
+      "step": 7452
+    },
+    {
+      "epoch": 1.812059324094335,
+      "grad_norm": 20.75,
+      "learning_rate": 5.540518783611393e-08,
+      "loss": 0.7553,
+      "step": 7453
+    },
+    {
+      "epoch": 1.812302455628495,
+      "grad_norm": 30.625,
+      "learning_rate": 5.526299454204603e-08,
+      "loss": 1.0515,
+      "step": 7454
+    },
+    {
+      "epoch": 1.812545587162655,
+      "grad_norm": 19.625,
+      "learning_rate": 5.512097982205741e-08,
+      "loss": 0.5147,
+      "step": 7455
+    },
+    {
+      "epoch": 1.812788718696815,
+      "grad_norm": 23.5,
+      "learning_rate": 5.497914369737442e-08,
+      "loss": 0.7567,
+      "step": 7456
+    },
+    {
+      "epoch": 1.813031850230975,
+      "grad_norm": 25.0,
+      "learning_rate": 5.483748618919732e-08,
+      "loss": 0.9327,
+      "step": 7457
+    },
+    {
+      "epoch": 1.8132749817651348,
+      "grad_norm": 25.5,
+      "learning_rate": 5.4696007318698894e-08,
+      "loss": 1.2147,
+      "step": 7458
+    },
+    {
+      "epoch": 1.813518113299295,
+      "grad_norm": 26.875,
+      "learning_rate": 5.4554707107025846e-08,
+      "loss": 0.4982,
+      "step": 7459
+    },
+    {
+      "epoch": 1.8137612448334548,
+      "grad_norm": 20.75,
+      "learning_rate": 5.44135855752978e-08,
+      "loss": 0.5704,
+      "step": 7460
+    },
+    {
+      "epoch": 1.814004376367615,
+      "grad_norm": 13.875,
+      "learning_rate": 5.4272642744608166e-08,
+      "loss": 0.4078,
+      "step": 7461
+    },
+    {
+      "epoch": 1.8142475079017748,
+      "grad_norm": 19.5,
+      "learning_rate": 5.413187863602287e-08,
+      "loss": 0.8482,
+      "step": 7462
+    },
+    {
+      "epoch": 1.8144906394359348,
+      "grad_norm": 19.25,
+      "learning_rate": 5.399129327058147e-08,
+      "loss": 0.4841,
+      "step": 7463
+    },
+    {
+      "epoch": 1.8147337709700948,
+      "grad_norm": 22.875,
+      "learning_rate": 5.3850886669297304e-08,
+      "loss": 0.6655,
+      "step": 7464
+    },
+    {
+      "epoch": 1.8149769025042548,
+      "grad_norm": 18.375,
+      "learning_rate": 5.371065885315635e-08,
+      "loss": 0.5621,
+      "step": 7465
+    },
+    {
+      "epoch": 1.8152200340384148,
+      "grad_norm": 18.75,
+      "learning_rate": 5.357060984311796e-08,
+      "loss": 0.3608,
+      "step": 7466
+    },
+    {
+      "epoch": 1.8154631655725748,
+      "grad_norm": 18.0,
+      "learning_rate": 5.3430739660114835e-08,
+      "loss": 0.6578,
+      "step": 7467
+    },
+    {
+      "epoch": 1.8157062971067348,
+      "grad_norm": 18.125,
+      "learning_rate": 5.329104832505344e-08,
+      "loss": 0.725,
+      "step": 7468
+    },
+    {
+      "epoch": 1.8159494286408946,
+      "grad_norm": 23.25,
+      "learning_rate": 5.3151535858812634e-08,
+      "loss": 0.7315,
+      "step": 7469
+    },
+    {
+      "epoch": 1.8161925601750548,
+      "grad_norm": 17.75,
+      "learning_rate": 5.301220228224491e-08,
+      "loss": 0.5221,
+      "step": 7470
+    },
+    {
+      "epoch": 1.8164356917092146,
+      "grad_norm": 19.625,
+      "learning_rate": 5.287304761617651e-08,
+      "loss": 0.7665,
+      "step": 7471
+    },
+    {
+      "epoch": 1.8166788232433748,
+      "grad_norm": 18.5,
+      "learning_rate": 5.2734071881406226e-08,
+      "loss": 0.5429,
+      "step": 7472
+    },
+    {
+      "epoch": 1.8169219547775346,
+      "grad_norm": 20.625,
+      "learning_rate": 5.2595275098706186e-08,
+      "loss": 0.7075,
+      "step": 7473
+    },
+    {
+      "epoch": 1.8171650863116946,
+      "grad_norm": 19.5,
+      "learning_rate": 5.245665728882215e-08,
+      "loss": 0.6894,
+      "step": 7474
+    },
+    {
+      "epoch": 1.8174082178458546,
+      "grad_norm": 20.875,
+      "learning_rate": 5.231821847247326e-08,
+      "loss": 0.6145,
+      "step": 7475
+    },
+    {
+      "epoch": 1.8176513493800146,
+      "grad_norm": 23.5,
+      "learning_rate": 5.217995867035114e-08,
+      "loss": 1.0259,
+      "step": 7476
+    },
+    {
+      "epoch": 1.8178944809141746,
+      "grad_norm": 18.5,
+      "learning_rate": 5.204187790312121e-08,
+      "loss": 0.5431,
+      "step": 7477
+    },
+    {
+      "epoch": 1.8181376124483344,
+      "grad_norm": 23.0,
+      "learning_rate": 5.190397619142223e-08,
+      "loss": 0.6012,
+      "step": 7478
+    },
+    {
+      "epoch": 1.8183807439824946,
+      "grad_norm": 20.0,
+      "learning_rate": 5.176625355586579e-08,
+      "loss": 0.8422,
+      "step": 7479
+    },
+    {
+      "epoch": 1.8186238755166544,
+      "grad_norm": 23.25,
+      "learning_rate": 5.162871001703693e-08,
+      "loss": 0.9246,
+      "step": 7480
+    },
+    {
+      "epoch": 1.8188670070508146,
+      "grad_norm": 23.75,
+      "learning_rate": 5.149134559549379e-08,
+      "loss": 0.79,
+      "step": 7481
+    },
+    {
+      "epoch": 1.8191101385849744,
+      "grad_norm": 19.25,
+      "learning_rate": 5.135416031176829e-08,
+      "loss": 0.6807,
+      "step": 7482
+    },
+    {
+      "epoch": 1.8193532701191346,
+      "grad_norm": 21.125,
+      "learning_rate": 5.121715418636472e-08,
+      "loss": 0.7234,
+      "step": 7483
+    },
+    {
+      "epoch": 1.8195964016532944,
+      "grad_norm": 20.375,
+      "learning_rate": 5.108032723976114e-08,
+      "loss": 0.8167,
+      "step": 7484
+    },
+    {
+      "epoch": 1.8198395331874544,
+      "grad_norm": 17.75,
+      "learning_rate": 5.094367949240883e-08,
+      "loss": 0.586,
+      "step": 7485
+    },
+    {
+      "epoch": 1.8200826647216144,
+      "grad_norm": 32.5,
+      "learning_rate": 5.0807210964732014e-08,
+      "loss": 0.7678,
+      "step": 7486
+    },
+    {
+      "epoch": 1.8203257962557744,
+      "grad_norm": 28.625,
+      "learning_rate": 5.067092167712811e-08,
+      "loss": 1.0422,
+      "step": 7487
+    },
+    {
+      "epoch": 1.8205689277899344,
+      "grad_norm": 17.25,
+      "learning_rate": 5.053481164996835e-08,
+      "loss": 0.6595,
+      "step": 7488
+    },
+    {
+      "epoch": 1.8208120593240942,
+      "grad_norm": 24.5,
+      "learning_rate": 5.0398880903596294e-08,
+      "loss": 0.6553,
+      "step": 7489
+    },
+    {
+      "epoch": 1.8210551908582544,
+      "grad_norm": 21.75,
+      "learning_rate": 5.026312945832931e-08,
+      "loss": 0.805,
+      "step": 7490
+    },
+    {
+      "epoch": 1.8212983223924142,
+      "grad_norm": 19.75,
+      "learning_rate": 5.0127557334457846e-08,
+      "loss": 0.9541,
+      "step": 7491
+    },
+    {
+      "epoch": 1.8215414539265744,
+      "grad_norm": 21.875,
+      "learning_rate": 4.999216455224554e-08,
+      "loss": 1.0441,
+      "step": 7492
+    },
+    {
+      "epoch": 1.8217845854607342,
+      "grad_norm": 28.875,
+      "learning_rate": 4.9856951131928983e-08,
+      "loss": 0.909,
+      "step": 7493
+    },
+    {
+      "epoch": 1.8220277169948942,
+      "grad_norm": 21.0,
+      "learning_rate": 4.972191709371826e-08,
+      "loss": 0.8385,
+      "step": 7494
+    },
+    {
+      "epoch": 1.8222708485290542,
+      "grad_norm": 21.375,
+      "learning_rate": 4.9587062457796684e-08,
+      "loss": 0.9484,
+      "step": 7495
+    },
+    {
+      "epoch": 1.8225139800632142,
+      "grad_norm": 28.625,
+      "learning_rate": 4.945238724432047e-08,
+      "loss": 0.88,
+      "step": 7496
+    },
+    {
+      "epoch": 1.8227571115973742,
+      "grad_norm": 20.0,
+      "learning_rate": 4.931789147341895e-08,
+      "loss": 0.7199,
+      "step": 7497
+    },
+    {
+      "epoch": 1.8230002431315342,
+      "grad_norm": 18.625,
+      "learning_rate": 4.918357516519506e-08,
+      "loss": 0.9908,
+      "step": 7498
+    },
+    {
+      "epoch": 1.8232433746656942,
+      "grad_norm": 16.625,
+      "learning_rate": 4.9049438339724836e-08,
+      "loss": 0.6022,
+      "step": 7499
+    },
+    {
+      "epoch": 1.823486506199854,
+      "grad_norm": 40.5,
+      "learning_rate": 4.8915481017057235e-08,
+      "loss": 1.7718,
+      "step": 7500
+    },
+    {
+      "epoch": 1.8237296377340142,
+      "grad_norm": 20.75,
+      "learning_rate": 4.878170321721415e-08,
+      "loss": 0.9465,
+      "step": 7501
+    },
+    {
+      "epoch": 1.823972769268174,
+      "grad_norm": 20.125,
+      "learning_rate": 4.864810496019154e-08,
+      "loss": 0.6056,
+      "step": 7502
+    },
+    {
+      "epoch": 1.8242159008023342,
+      "grad_norm": 15.25,
+      "learning_rate": 4.851468626595773e-08,
+      "loss": 0.2311,
+      "step": 7503
+    },
+    {
+      "epoch": 1.824459032336494,
+      "grad_norm": 27.25,
+      "learning_rate": 4.8381447154454144e-08,
+      "loss": 0.5883,
+      "step": 7504
+    },
+    {
+      "epoch": 1.824702163870654,
+      "grad_norm": 30.125,
+      "learning_rate": 4.824838764559608e-08,
+      "loss": 0.7411,
+      "step": 7505
+    },
+    {
+      "epoch": 1.824945295404814,
+      "grad_norm": 22.0,
+      "learning_rate": 4.811550775927168e-08,
+      "loss": 0.6528,
+      "step": 7506
+    },
+    {
+      "epoch": 1.825188426938974,
+      "grad_norm": 21.625,
+      "learning_rate": 4.798280751534171e-08,
+      "loss": 0.8038,
+      "step": 7507
+    },
+    {
+      "epoch": 1.825431558473134,
+      "grad_norm": 28.0,
+      "learning_rate": 4.7850286933640716e-08,
+      "loss": 0.9102,
+      "step": 7508
+    },
+    {
+      "epoch": 1.825674690007294,
+      "grad_norm": 18.5,
+      "learning_rate": 4.77179460339762e-08,
+      "loss": 0.7418,
+      "step": 7509
+    },
+    {
+      "epoch": 1.825917821541454,
+      "grad_norm": 18.75,
+      "learning_rate": 4.758578483612886e-08,
+      "loss": 1.2349,
+      "step": 7510
+    },
+    {
+      "epoch": 1.8261609530756138,
+      "grad_norm": 18.375,
+      "learning_rate": 4.7453803359852196e-08,
+      "loss": 0.7239,
+      "step": 7511
+    },
+    {
+      "epoch": 1.826404084609774,
+      "grad_norm": 26.625,
+      "learning_rate": 4.732200162487335e-08,
+      "loss": 1.275,
+      "step": 7512
+    },
+    {
+      "epoch": 1.8266472161439338,
+      "grad_norm": 20.0,
+      "learning_rate": 4.719037965089254e-08,
+      "loss": 0.5573,
+      "step": 7513
+    },
+    {
+      "epoch": 1.826890347678094,
+      "grad_norm": 21.375,
+      "learning_rate": 4.705893745758264e-08,
+      "loss": 0.7901,
+      "step": 7514
+    },
+    {
+      "epoch": 1.8271334792122538,
+      "grad_norm": 16.625,
+      "learning_rate": 4.692767506458987e-08,
+      "loss": 0.3731,
+      "step": 7515
+    },
+    {
+      "epoch": 1.8273766107464138,
+      "grad_norm": 18.0,
+      "learning_rate": 4.6796592491534094e-08,
+      "loss": 0.5301,
+      "step": 7516
+    },
+    {
+      "epoch": 1.8276197422805738,
+      "grad_norm": 19.25,
+      "learning_rate": 4.666568975800756e-08,
+      "loss": 0.4829,
+      "step": 7517
+    },
+    {
+      "epoch": 1.8278628738147338,
+      "grad_norm": 27.625,
+      "learning_rate": 4.653496688357586e-08,
+      "loss": 1.1146,
+      "step": 7518
+    },
+    {
+      "epoch": 1.8281060053488938,
+      "grad_norm": 23.625,
+      "learning_rate": 4.640442388777797e-08,
+      "loss": 1.0188,
+      "step": 7519
+    },
+    {
+      "epoch": 1.8283491368830538,
+      "grad_norm": 15.125,
+      "learning_rate": 4.6274060790125754e-08,
+      "loss": 0.2492,
+      "step": 7520
+    },
+    {
+      "epoch": 1.8285922684172138,
+      "grad_norm": 20.875,
+      "learning_rate": 4.614387761010433e-08,
+      "loss": 0.5477,
+      "step": 7521
+    },
+    {
+      "epoch": 1.8288353999513736,
+      "grad_norm": 18.125,
+      "learning_rate": 4.6013874367171475e-08,
+      "loss": 0.4903,
+      "step": 7522
+    },
+    {
+      "epoch": 1.8290785314855338,
+      "grad_norm": 21.25,
+      "learning_rate": 4.5884051080758856e-08,
+      "loss": 0.7898,
+      "step": 7523
+    },
+    {
+      "epoch": 1.8293216630196936,
+      "grad_norm": 19.875,
+      "learning_rate": 4.575440777027054e-08,
+      "loss": 0.6642,
+      "step": 7524
+    },
+    {
+      "epoch": 1.8295647945538538,
+      "grad_norm": 18.625,
+      "learning_rate": 4.5624944455083944e-08,
+      "loss": 0.6796,
+      "step": 7525
+    },
+    {
+      "epoch": 1.8298079260880136,
+      "grad_norm": 16.5,
+      "learning_rate": 4.5495661154549844e-08,
+      "loss": 0.6323,
+      "step": 7526
+    },
+    {
+      "epoch": 1.8300510576221736,
+      "grad_norm": 14.5625,
+      "learning_rate": 4.5366557887991516e-08,
+      "loss": 0.283,
+      "step": 7527
+    },
+    {
+      "epoch": 1.8302941891563336,
+      "grad_norm": 16.25,
+      "learning_rate": 4.523763467470591e-08,
+      "loss": 0.6075,
+      "step": 7528
+    },
+    {
+      "epoch": 1.8305373206904936,
+      "grad_norm": 16.625,
+      "learning_rate": 4.510889153396286e-08,
+      "loss": 0.4247,
+      "step": 7529
+    },
+    {
+      "epoch": 1.8307804522246536,
+      "grad_norm": 37.75,
+      "learning_rate": 4.498032848500505e-08,
+      "loss": 1.2243,
+      "step": 7530
+    },
+    {
+      "epoch": 1.8310235837588134,
+      "grad_norm": 20.125,
+      "learning_rate": 4.4851945547048774e-08,
+      "loss": 0.8127,
+      "step": 7531
+    },
+    {
+      "epoch": 1.8312667152929736,
+      "grad_norm": 23.625,
+      "learning_rate": 4.472374273928257e-08,
+      "loss": 0.7337,
+      "step": 7532
+    },
+    {
+      "epoch": 1.8315098468271334,
+      "grad_norm": 20.75,
+      "learning_rate": 4.4595720080869016e-08,
+      "loss": 0.692,
+      "step": 7533
+    },
+    {
+      "epoch": 1.8317529783612936,
+      "grad_norm": 19.5,
+      "learning_rate": 4.446787759094323e-08,
+      "loss": 0.3627,
+      "step": 7534
+    },
+    {
+      "epoch": 1.8319961098954534,
+      "grad_norm": 36.25,
+      "learning_rate": 4.434021528861299e-08,
+      "loss": 0.9994,
+      "step": 7535
+    },
+    {
+      "epoch": 1.8322392414296136,
+      "grad_norm": 19.5,
+      "learning_rate": 4.4212733192960377e-08,
+      "loss": 0.7424,
+      "step": 7536
+    },
+    {
+      "epoch": 1.8324823729637734,
+      "grad_norm": 23.125,
+      "learning_rate": 4.408543132303947e-08,
+      "loss": 0.8957,
+      "step": 7537
+    },
+    {
+      "epoch": 1.8327255044979334,
+      "grad_norm": 19.125,
+      "learning_rate": 4.3958309697877815e-08,
+      "loss": 0.5535,
+      "step": 7538
+    },
+    {
+      "epoch": 1.8329686360320934,
+      "grad_norm": 16.125,
+      "learning_rate": 4.38313683364755e-08,
+      "loss": 0.4912,
+      "step": 7539
+    },
+    {
+      "epoch": 1.8332117675662534,
+      "grad_norm": 26.0,
+      "learning_rate": 4.3704607257806644e-08,
+      "loss": 0.7422,
+      "step": 7540
+    },
+    {
+      "epoch": 1.8334548991004134,
+      "grad_norm": 40.25,
+      "learning_rate": 4.357802648081777e-08,
+      "loss": 1.1415,
+      "step": 7541
+    },
+    {
+      "epoch": 1.8336980306345732,
+      "grad_norm": 25.125,
+      "learning_rate": 4.3451626024428315e-08,
+      "loss": 0.6099,
+      "step": 7542
+    },
+    {
+      "epoch": 1.8339411621687334,
+      "grad_norm": 19.25,
+      "learning_rate": 4.332540590753109e-08,
+      "loss": 0.6833,
+      "step": 7543
+    },
+    {
+      "epoch": 1.8341842937028932,
+      "grad_norm": 18.0,
+      "learning_rate": 4.3199366148992115e-08,
+      "loss": 0.6793,
+      "step": 7544
+    },
+    {
+      "epoch": 1.8344274252370534,
+      "grad_norm": 20.5,
+      "learning_rate": 4.3073506767649794e-08,
+      "loss": 0.7335,
+      "step": 7545
+    },
+    {
+      "epoch": 1.8346705567712132,
+      "grad_norm": 22.25,
+      "learning_rate": 4.294782778231657e-08,
+      "loss": 0.8019,
+      "step": 7546
+    },
+    {
+      "epoch": 1.8349136883053732,
+      "grad_norm": 17.125,
+      "learning_rate": 4.282232921177687e-08,
+      "loss": 0.5312,
+      "step": 7547
+    },
+    {
+      "epoch": 1.8351568198395332,
+      "grad_norm": 24.625,
+      "learning_rate": 4.269701107478874e-08,
+      "loss": 0.853,
+      "step": 7548
+    },
+    {
+      "epoch": 1.8353999513736932,
+      "grad_norm": 20.875,
+      "learning_rate": 4.257187339008303e-08,
+      "loss": 0.6265,
+      "step": 7549
+    },
+    {
+      "epoch": 1.8356430829078532,
+      "grad_norm": 15.4375,
+      "learning_rate": 4.2446916176363955e-08,
+      "loss": 0.3864,
+      "step": 7550
+    },
+    {
+      "epoch": 1.8358862144420132,
+      "grad_norm": 19.125,
+      "learning_rate": 4.232213945230837e-08,
+      "loss": 0.4786,
+      "step": 7551
+    },
+    {
+      "epoch": 1.8361293459761732,
+      "grad_norm": 16.875,
+      "learning_rate": 4.219754323656636e-08,
+      "loss": 0.8536,
+      "step": 7552
+    },
+    {
+      "epoch": 1.836372477510333,
+      "grad_norm": 24.125,
+      "learning_rate": 4.207312754776094e-08,
+      "loss": 0.554,
+      "step": 7553
+    },
+    {
+      "epoch": 1.8366156090444932,
+      "grad_norm": 16.625,
+      "learning_rate": 4.1948892404488487e-08,
+      "loss": 0.3813,
+      "step": 7554
+    },
+    {
+      "epoch": 1.836858740578653,
+      "grad_norm": 18.0,
+      "learning_rate": 4.1824837825317626e-08,
+      "loss": 0.6299,
+      "step": 7555
+    },
+    {
+      "epoch": 1.8371018721128132,
+      "grad_norm": 19.0,
+      "learning_rate": 4.1700963828790606e-08,
+      "loss": 0.6437,
+      "step": 7556
+    },
+    {
+      "epoch": 1.837345003646973,
+      "grad_norm": 18.125,
+      "learning_rate": 4.1577270433422745e-08,
+      "loss": 0.6631,
+      "step": 7557
+    },
+    {
+      "epoch": 1.837588135181133,
+      "grad_norm": 28.25,
+      "learning_rate": 4.145375765770177e-08,
+      "loss": 1.2545,
+      "step": 7558
+    },
+    {
+      "epoch": 1.837831266715293,
+      "grad_norm": 24.25,
+      "learning_rate": 4.133042552008915e-08,
+      "loss": 0.8158,
+      "step": 7559
+    },
+    {
+      "epoch": 1.838074398249453,
+      "grad_norm": 17.5,
+      "learning_rate": 4.120727403901889e-08,
+      "loss": 0.6079,
+      "step": 7560
+    },
+    {
+      "epoch": 1.838317529783613,
+      "grad_norm": 15.875,
+      "learning_rate": 4.108430323289822e-08,
+      "loss": 0.3736,
+      "step": 7561
+    },
+    {
+      "epoch": 1.838560661317773,
+      "grad_norm": 22.5,
+      "learning_rate": 4.0961513120107015e-08,
+      "loss": 0.5513,
+      "step": 7562
+    },
+    {
+      "epoch": 1.838803792851933,
+      "grad_norm": 15.875,
+      "learning_rate": 4.0838903718998516e-08,
+      "loss": 0.6372,
+      "step": 7563
+    },
+    {
+      "epoch": 1.8390469243860927,
+      "grad_norm": 23.875,
+      "learning_rate": 4.071647504789875e-08,
+      "loss": 0.8594,
+      "step": 7564
+    },
+    {
+      "epoch": 1.839290055920253,
+      "grad_norm": 17.875,
+      "learning_rate": 4.059422712510697e-08,
+      "loss": 0.8465,
+      "step": 7565
+    },
+    {
+      "epoch": 1.8395331874544127,
+      "grad_norm": 24.5,
+      "learning_rate": 4.047215996889481e-08,
+      "loss": 0.8671,
+      "step": 7566
+    },
+    {
+      "epoch": 1.839776318988573,
+      "grad_norm": 21.5,
+      "learning_rate": 4.035027359750782e-08,
+      "loss": 0.8169,
+      "step": 7567
+    },
+    {
+      "epoch": 1.8400194505227327,
+      "grad_norm": 16.75,
+      "learning_rate": 4.022856802916392e-08,
+      "loss": 0.4717,
+      "step": 7568
+    },
+    {
+      "epoch": 1.8402625820568927,
+      "grad_norm": 18.5,
+      "learning_rate": 4.01070432820537e-08,
+      "loss": 0.5276,
+      "step": 7569
+    },
+    {
+      "epoch": 1.8405057135910527,
+      "grad_norm": 19.0,
+      "learning_rate": 3.9985699374341656e-08,
+      "loss": 0.6881,
+      "step": 7570
+    },
+    {
+      "epoch": 1.8407488451252128,
+      "grad_norm": 23.625,
+      "learning_rate": 3.9864536324164536e-08,
+      "loss": 0.7364,
+      "step": 7571
+    },
+    {
+      "epoch": 1.8409919766593728,
+      "grad_norm": 18.5,
+      "learning_rate": 3.97435541496323e-08,
+      "loss": 0.5181,
+      "step": 7572
+    },
+    {
+      "epoch": 1.8412351081935328,
+      "grad_norm": 17.75,
+      "learning_rate": 3.962275286882741e-08,
+      "loss": 0.4092,
+      "step": 7573
+    },
+    {
+      "epoch": 1.8414782397276928,
+      "grad_norm": 19.875,
+      "learning_rate": 3.950213249980614e-08,
+      "loss": 0.8533,
+      "step": 7574
+    },
+    {
+      "epoch": 1.8417213712618525,
+      "grad_norm": 26.875,
+      "learning_rate": 3.9381693060597385e-08,
+      "loss": 0.9544,
+      "step": 7575
+    },
+    {
+      "epoch": 1.8419645027960128,
+      "grad_norm": 21.125,
+      "learning_rate": 3.926143456920259e-08,
+      "loss": 0.7246,
+      "step": 7576
+    },
+    {
+      "epoch": 1.8422076343301725,
+      "grad_norm": 17.375,
+      "learning_rate": 3.914135704359667e-08,
+      "loss": 0.5233,
+      "step": 7577
+    },
+    {
+      "epoch": 1.8424507658643328,
+      "grad_norm": 17.125,
+      "learning_rate": 3.9021460501727086e-08,
+      "loss": 0.5068,
+      "step": 7578
+    },
+    {
+      "epoch": 1.8426938973984925,
+      "grad_norm": 22.0,
+      "learning_rate": 3.89017449615145e-08,
+      "loss": 0.4583,
+      "step": 7579
+    },
+    {
+      "epoch": 1.8429370289326525,
+      "grad_norm": 18.625,
+      "learning_rate": 3.878221044085265e-08,
+      "loss": 0.4468,
+      "step": 7580
+    },
+    {
+      "epoch": 1.8431801604668125,
+      "grad_norm": 20.25,
+      "learning_rate": 3.866285695760794e-08,
+      "loss": 1.0066,
+      "step": 7581
+    },
+    {
+      "epoch": 1.8434232920009725,
+      "grad_norm": 20.125,
+      "learning_rate": 3.854368452961957e-08,
+      "loss": 0.2508,
+      "step": 7582
+    },
+    {
+      "epoch": 1.8436664235351325,
+      "grad_norm": 21.75,
+      "learning_rate": 3.842469317470024e-08,
+      "loss": 0.5249,
+      "step": 7583
+    },
+    {
+      "epoch": 1.8439095550692923,
+      "grad_norm": 25.25,
+      "learning_rate": 3.830588291063517e-08,
+      "loss": 0.7665,
+      "step": 7584
+    },
+    {
+      "epoch": 1.8441526866034526,
+      "grad_norm": 27.125,
+      "learning_rate": 3.818725375518265e-08,
+      "loss": 0.7356,
+      "step": 7585
+    },
+    {
+      "epoch": 1.8443958181376123,
+      "grad_norm": 17.875,
+      "learning_rate": 3.8068805726073634e-08,
+      "loss": 0.4984,
+      "step": 7586
+    },
+    {
+      "epoch": 1.8446389496717726,
+      "grad_norm": 15.4375,
+      "learning_rate": 3.7950538841012434e-08,
+      "loss": 0.5548,
+      "step": 7587
+    },
+    {
+      "epoch": 1.8448820812059323,
+      "grad_norm": 17.625,
+      "learning_rate": 3.7832453117676024e-08,
+      "loss": 0.4331,
+      "step": 7588
+    },
+    {
+      "epoch": 1.8451252127400923,
+      "grad_norm": 30.125,
+      "learning_rate": 3.7714548573714313e-08,
+      "loss": 0.8156,
+      "step": 7589
+    },
+    {
+      "epoch": 1.8453683442742523,
+      "grad_norm": 18.625,
+      "learning_rate": 3.759682522675015e-08,
+      "loss": 0.6269,
+      "step": 7590
+    },
+    {
+      "epoch": 1.8456114758084123,
+      "grad_norm": 19.125,
+      "learning_rate": 3.7479283094379596e-08,
+      "loss": 0.615,
+      "step": 7591
+    },
+    {
+      "epoch": 1.8458546073425723,
+      "grad_norm": 22.875,
+      "learning_rate": 3.736192219417109e-08,
+      "loss": 0.7458,
+      "step": 7592
+    },
+    {
+      "epoch": 1.8460977388767323,
+      "grad_norm": 17.0,
+      "learning_rate": 3.724474254366617e-08,
+      "loss": 0.7284,
+      "step": 7593
+    },
+    {
+      "epoch": 1.8463408704108923,
+      "grad_norm": 18.5,
+      "learning_rate": 3.7127744160379565e-08,
+      "loss": 0.544,
+      "step": 7594
+    },
+    {
+      "epoch": 1.8465840019450521,
+      "grad_norm": 17.0,
+      "learning_rate": 3.7010927061798676e-08,
+      "loss": 1.2383,
+      "step": 7595
+    },
+    {
+      "epoch": 1.8468271334792123,
+      "grad_norm": 22.625,
+      "learning_rate": 3.689429126538372e-08,
+      "loss": 0.5412,
+      "step": 7596
+    },
+    {
+      "epoch": 1.8470702650133721,
+      "grad_norm": 26.75,
+      "learning_rate": 3.67778367885678e-08,
+      "loss": 0.8867,
+      "step": 7597
+    },
+    {
+      "epoch": 1.8473133965475323,
+      "grad_norm": 21.375,
+      "learning_rate": 3.6661563648757586e-08,
+      "loss": 0.7603,
+      "step": 7598
+    },
+    {
+      "epoch": 1.8475565280816921,
+      "grad_norm": 19.0,
+      "learning_rate": 3.6545471863331656e-08,
+      "loss": 0.7846,
+      "step": 7599
+    },
+    {
+      "epoch": 1.8477996596158521,
+      "grad_norm": 21.375,
+      "learning_rate": 3.642956144964183e-08,
+      "loss": 0.5344,
+      "step": 7600
+    },
+    {
+      "epoch": 1.8480427911500121,
+      "grad_norm": 21.125,
+      "learning_rate": 3.631383242501341e-08,
+      "loss": 1.0846,
+      "step": 7601
+    },
+    {
+      "epoch": 1.8482859226841721,
+      "grad_norm": 22.5,
+      "learning_rate": 3.6198284806743814e-08,
+      "loss": 0.6092,
+      "step": 7602
+    },
+    {
+      "epoch": 1.8485290542183321,
+      "grad_norm": 20.875,
+      "learning_rate": 3.608291861210339e-08,
+      "loss": 0.7028,
+      "step": 7603
+    },
+    {
+      "epoch": 1.8487721857524921,
+      "grad_norm": 16.5,
+      "learning_rate": 3.596773385833613e-08,
+      "loss": 0.399,
+      "step": 7604
+    },
+    {
+      "epoch": 1.8490153172866521,
+      "grad_norm": 27.5,
+      "learning_rate": 3.585273056265784e-08,
+      "loss": 1.1686,
+      "step": 7605
+    },
+    {
+      "epoch": 1.849258448820812,
+      "grad_norm": 19.5,
+      "learning_rate": 3.573790874225824e-08,
+      "loss": 0.6415,
+      "step": 7606
+    },
+    {
+      "epoch": 1.8495015803549721,
+      "grad_norm": 19.625,
+      "learning_rate": 3.562326841429903e-08,
+      "loss": 0.7175,
+      "step": 7607
+    },
+    {
+      "epoch": 1.849744711889132,
+      "grad_norm": 23.875,
+      "learning_rate": 3.550880959591552e-08,
+      "loss": 0.8093,
+      "step": 7608
+    },
+    {
+      "epoch": 1.8499878434232921,
+      "grad_norm": 22.25,
+      "learning_rate": 3.5394532304215413e-08,
+      "loss": 0.6922,
+      "step": 7609
+    },
+    {
+      "epoch": 1.850230974957452,
+      "grad_norm": 13.125,
+      "learning_rate": 3.5280436556279216e-08,
+      "loss": 0.2974,
+      "step": 7610
+    },
+    {
+      "epoch": 1.850474106491612,
+      "grad_norm": 26.25,
+      "learning_rate": 3.516652236916093e-08,
+      "loss": 0.9919,
+      "step": 7611
+    },
+    {
+      "epoch": 1.850717238025772,
+      "grad_norm": 32.75,
+      "learning_rate": 3.5052789759886656e-08,
+      "loss": 0.7669,
+      "step": 7612
+    },
+    {
+      "epoch": 1.850960369559932,
+      "grad_norm": 20.375,
+      "learning_rate": 3.493923874545585e-08,
+      "loss": 0.7524,
+      "step": 7613
+    },
+    {
+      "epoch": 1.851203501094092,
+      "grad_norm": 23.25,
+      "learning_rate": 3.482586934284049e-08,
+      "loss": 0.8704,
+      "step": 7614
+    },
+    {
+      "epoch": 1.851446632628252,
+      "grad_norm": 15.75,
+      "learning_rate": 3.471268156898605e-08,
+      "loss": 0.7977,
+      "step": 7615
+    },
+    {
+      "epoch": 1.851689764162412,
+      "grad_norm": 16.5,
+      "learning_rate": 3.459967544080997e-08,
+      "loss": 0.3564,
+      "step": 7616
+    },
+    {
+      "epoch": 1.8519328956965717,
+      "grad_norm": 13.6875,
+      "learning_rate": 3.4486850975202904e-08,
+      "loss": 0.4405,
+      "step": 7617
+    },
+    {
+      "epoch": 1.852176027230732,
+      "grad_norm": 20.75,
+      "learning_rate": 3.437420818902873e-08,
+      "loss": 0.57,
+      "step": 7618
+    },
+    {
+      "epoch": 1.8524191587648917,
+      "grad_norm": 20.25,
+      "learning_rate": 3.4261747099123834e-08,
+      "loss": 0.7879,
+      "step": 7619
+    },
+    {
+      "epoch": 1.852662290299052,
+      "grad_norm": 17.125,
+      "learning_rate": 3.414946772229713e-08,
+      "loss": 0.6246,
+      "step": 7620
+    },
+    {
+      "epoch": 1.8529054218332117,
+      "grad_norm": 24.0,
+      "learning_rate": 3.4037370075331035e-08,
+      "loss": 0.9031,
+      "step": 7621
+    },
+    {
+      "epoch": 1.8531485533673717,
+      "grad_norm": 22.875,
+      "learning_rate": 3.392545417498047e-08,
+      "loss": 0.6658,
+      "step": 7622
+    },
+    {
+      "epoch": 1.8533916849015317,
+      "grad_norm": 19.75,
+      "learning_rate": 3.3813720037973034e-08,
+      "loss": 0.9077,
+      "step": 7623
+    },
+    {
+      "epoch": 1.8536348164356917,
+      "grad_norm": 18.875,
+      "learning_rate": 3.37021676810094e-08,
+      "loss": 1.0388,
+      "step": 7624
+    },
+    {
+      "epoch": 1.8538779479698517,
+      "grad_norm": 12.8125,
+      "learning_rate": 3.359079712076316e-08,
+      "loss": 0.4981,
+      "step": 7625
+    },
+    {
+      "epoch": 1.8541210795040115,
+      "grad_norm": 18.875,
+      "learning_rate": 3.347960837388031e-08,
+      "loss": 0.6078,
+      "step": 7626
+    },
+    {
+      "epoch": 1.8543642110381717,
+      "grad_norm": 19.875,
+      "learning_rate": 3.336860145697979e-08,
+      "loss": 0.6266,
+      "step": 7627
+    },
+    {
+      "epoch": 1.8546073425723315,
+      "grad_norm": 17.875,
+      "learning_rate": 3.3257776386653865e-08,
+      "loss": 0.6471,
+      "step": 7628
+    },
+    {
+      "epoch": 1.8548504741064917,
+      "grad_norm": 22.875,
+      "learning_rate": 3.3147133179467215e-08,
+      "loss": 0.5607,
+      "step": 7629
+    },
+    {
+      "epoch": 1.8550936056406515,
+      "grad_norm": 20.75,
+      "learning_rate": 3.30366718519573e-08,
+      "loss": 0.5692,
+      "step": 7630
+    },
+    {
+      "epoch": 1.8553367371748117,
+      "grad_norm": 14.125,
+      "learning_rate": 3.292639242063439e-08,
+      "loss": 0.5095,
+      "step": 7631
+    },
+    {
+      "epoch": 1.8555798687089715,
+      "grad_norm": 23.375,
+      "learning_rate": 3.281629490198182e-08,
+      "loss": 1.0919,
+      "step": 7632
+    },
+    {
+      "epoch": 1.8558230002431315,
+      "grad_norm": 22.0,
+      "learning_rate": 3.270637931245546e-08,
+      "loss": 0.77,
+      "step": 7633
+    },
+    {
+      "epoch": 1.8560661317772915,
+      "grad_norm": 25.625,
+      "learning_rate": 3.2596645668483974e-08,
+      "loss": 0.6726,
+      "step": 7634
+    },
+    {
+      "epoch": 1.8563092633114515,
+      "grad_norm": 19.125,
+      "learning_rate": 3.24870939864691e-08,
+      "loss": 0.6936,
+      "step": 7635
+    },
+    {
+      "epoch": 1.8565523948456115,
+      "grad_norm": 21.25,
+      "learning_rate": 3.237772428278524e-08,
+      "loss": 0.8342,
+      "step": 7636
+    },
+    {
+      "epoch": 1.8567955263797713,
+      "grad_norm": 25.5,
+      "learning_rate": 3.2268536573779596e-08,
+      "loss": 0.8295,
+      "step": 7637
+    },
+    {
+      "epoch": 1.8570386579139315,
+      "grad_norm": 22.25,
+      "learning_rate": 3.2159530875771896e-08,
+      "loss": 0.6691,
+      "step": 7638
+    },
+    {
+      "epoch": 1.8572817894480913,
+      "grad_norm": 19.125,
+      "learning_rate": 3.205070720505535e-08,
+      "loss": 0.726,
+      "step": 7639
+    },
+    {
+      "epoch": 1.8575249209822515,
+      "grad_norm": 15.5625,
+      "learning_rate": 3.194206557789542e-08,
+      "loss": 0.6589,
+      "step": 7640
+    },
+    {
+      "epoch": 1.8577680525164113,
+      "grad_norm": 18.75,
+      "learning_rate": 3.1833606010530094e-08,
+      "loss": 0.5514,
+      "step": 7641
+    },
+    {
+      "epoch": 1.8580111840505713,
+      "grad_norm": 22.125,
+      "learning_rate": 3.1725328519170984e-08,
+      "loss": 0.8581,
+      "step": 7642
+    },
+    {
+      "epoch": 1.8582543155847313,
+      "grad_norm": 21.75,
+      "learning_rate": 3.1617233120001814e-08,
+      "loss": 0.9477,
+      "step": 7643
+    },
+    {
+      "epoch": 1.8584974471188913,
+      "grad_norm": 21.25,
+      "learning_rate": 3.150931982917937e-08,
+      "loss": 1.2574,
+      "step": 7644
+    },
+    {
+      "epoch": 1.8587405786530513,
+      "grad_norm": 16.375,
+      "learning_rate": 3.140158866283313e-08,
+      "loss": 0.5894,
+      "step": 7645
+    },
+    {
+      "epoch": 1.8589837101872113,
+      "grad_norm": 22.625,
+      "learning_rate": 3.129403963706548e-08,
+      "loss": 0.5343,
+      "step": 7646
+    },
+    {
+      "epoch": 1.8592268417213713,
+      "grad_norm": 17.5,
+      "learning_rate": 3.1186672767951345e-08,
+      "loss": 0.5812,
+      "step": 7647
+    },
+    {
+      "epoch": 1.859469973255531,
+      "grad_norm": 18.5,
+      "learning_rate": 3.107948807153843e-08,
+      "loss": 0.6572,
+      "step": 7648
+    },
+    {
+      "epoch": 1.8597131047896913,
+      "grad_norm": 20.625,
+      "learning_rate": 3.0972485563847684e-08,
+      "loss": 0.7052,
+      "step": 7649
+    },
+    {
+      "epoch": 1.859956236323851,
+      "grad_norm": 23.0,
+      "learning_rate": 3.0865665260872287e-08,
+      "loss": 0.5064,
+      "step": 7650
+    },
+    {
+      "epoch": 1.8601993678580113,
+      "grad_norm": 21.0,
+      "learning_rate": 3.0759027178578345e-08,
+      "loss": 0.839,
+      "step": 7651
+    },
+    {
+      "epoch": 1.860442499392171,
+      "grad_norm": 19.875,
+      "learning_rate": 3.065257133290478e-08,
+      "loss": 0.63,
+      "step": 7652
+    },
+    {
+      "epoch": 1.860685630926331,
+      "grad_norm": 14.9375,
+      "learning_rate": 3.054629773976331e-08,
+      "loss": 0.3206,
+      "step": 7653
+    },
+    {
+      "epoch": 1.860928762460491,
+      "grad_norm": 30.625,
+      "learning_rate": 3.044020641503845e-08,
+      "loss": 0.7292,
+      "step": 7654
+    },
+    {
+      "epoch": 1.861171893994651,
+      "grad_norm": 18.25,
+      "learning_rate": 3.03342973745871e-08,
+      "loss": 0.7985,
+      "step": 7655
+    },
+    {
+      "epoch": 1.861415025528811,
+      "grad_norm": 20.5,
+      "learning_rate": 3.0228570634239505e-08,
+      "loss": 0.4857,
+      "step": 7656
+    },
+    {
+      "epoch": 1.861658157062971,
+      "grad_norm": 18.75,
+      "learning_rate": 3.0123026209798176e-08,
+      "loss": 0.6108,
+      "step": 7657
+    },
+    {
+      "epoch": 1.861901288597131,
+      "grad_norm": 20.75,
+      "learning_rate": 3.001766411703855e-08,
+      "loss": 0.523,
+      "step": 7658
+    },
+    {
+      "epoch": 1.862144420131291,
+      "grad_norm": 22.25,
+      "learning_rate": 2.991248437170871e-08,
+      "loss": 0.9962,
+      "step": 7659
+    },
+    {
+      "epoch": 1.8623875516654511,
+      "grad_norm": 19.75,
+      "learning_rate": 2.980748698952998e-08,
+      "loss": 0.4358,
+      "step": 7660
+    },
+    {
+      "epoch": 1.862630683199611,
+      "grad_norm": 17.625,
+      "learning_rate": 2.9702671986195642e-08,
+      "loss": 0.4995,
+      "step": 7661
+    },
+    {
+      "epoch": 1.8628738147337711,
+      "grad_norm": 18.875,
+      "learning_rate": 2.9598039377372184e-08,
+      "loss": 0.7364,
+      "step": 7662
+    },
+    {
+      "epoch": 1.863116946267931,
+      "grad_norm": 17.875,
+      "learning_rate": 2.9493589178698905e-08,
+      "loss": 0.4898,
+      "step": 7663
+    },
+    {
+      "epoch": 1.863360077802091,
+      "grad_norm": 20.5,
+      "learning_rate": 2.9389321405787623e-08,
+      "loss": 0.8234,
+      "step": 7664
+    },
+    {
+      "epoch": 1.863603209336251,
+      "grad_norm": 18.125,
+      "learning_rate": 2.9285236074222817e-08,
+      "loss": 0.7785,
+      "step": 7665
+    },
+    {
+      "epoch": 1.863846340870411,
+      "grad_norm": 21.875,
+      "learning_rate": 2.918133319956204e-08,
+      "loss": 0.8563,
+      "step": 7666
+    },
+    {
+      "epoch": 1.864089472404571,
+      "grad_norm": 17.5,
+      "learning_rate": 2.907761279733523e-08,
+      "loss": 0.654,
+      "step": 7667
+    },
+    {
+      "epoch": 1.864332603938731,
+      "grad_norm": 15.625,
+      "learning_rate": 2.8974074883045406e-08,
+      "loss": 0.3309,
+      "step": 7668
+    },
+    {
+      "epoch": 1.864575735472891,
+      "grad_norm": 20.75,
+      "learning_rate": 2.8870719472167684e-08,
+      "loss": 1.0329,
+      "step": 7669
+    },
+    {
+      "epoch": 1.8648188670070507,
+      "grad_norm": 23.125,
+      "learning_rate": 2.8767546580150823e-08,
+      "loss": 0.9319,
+      "step": 7670
+    },
+    {
+      "epoch": 1.865061998541211,
+      "grad_norm": 20.875,
+      "learning_rate": 2.8664556222415403e-08,
+      "loss": 0.6446,
+      "step": 7671
+    },
+    {
+      "epoch": 1.8653051300753707,
+      "grad_norm": 17.375,
+      "learning_rate": 2.8561748414355217e-08,
+      "loss": 0.6958,
+      "step": 7672
+    },
+    {
+      "epoch": 1.865548261609531,
+      "grad_norm": 25.75,
+      "learning_rate": 2.8459123171336868e-08,
+      "loss": 0.8497,
+      "step": 7673
+    },
+    {
+      "epoch": 1.8657913931436907,
+      "grad_norm": 15.875,
+      "learning_rate": 2.835668050869905e-08,
+      "loss": 0.6374,
+      "step": 7674
+    },
+    {
+      "epoch": 1.8660345246778507,
+      "grad_norm": 25.875,
+      "learning_rate": 2.825442044175397e-08,
+      "loss": 0.6558,
+      "step": 7675
+    },
+    {
+      "epoch": 1.8662776562120107,
+      "grad_norm": 23.375,
+      "learning_rate": 2.8152342985786062e-08,
+      "loss": 0.709,
+      "step": 7676
+    },
+    {
+      "epoch": 1.8665207877461707,
+      "grad_norm": 23.125,
+      "learning_rate": 2.8050448156052566e-08,
+      "loss": 1.2235,
+      "step": 7677
+    },
+    {
+      "epoch": 1.8667639192803307,
+      "grad_norm": 25.375,
+      "learning_rate": 2.7948735967783247e-08,
+      "loss": 1.0966,
+      "step": 7678
+    },
+    {
+      "epoch": 1.8670070508144905,
+      "grad_norm": 18.625,
+      "learning_rate": 2.784720643618094e-08,
+      "loss": 0.6256,
+      "step": 7679
+    },
+    {
+      "epoch": 1.8672501823486507,
+      "grad_norm": 21.875,
+      "learning_rate": 2.774585957642101e-08,
+      "loss": 0.8769,
+      "step": 7680
+    },
+    {
+      "epoch": 1.8674933138828105,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.7644695403651472e-08,
+      "loss": 0.5008,
+      "step": 7681
+    },
+    {
+      "epoch": 1.8677364454169707,
+      "grad_norm": 17.5,
+      "learning_rate": 2.7543713932992878e-08,
+      "loss": 0.4466,
+      "step": 7682
+    },
+    {
+      "epoch": 1.8679795769511305,
+      "grad_norm": 20.25,
+      "learning_rate": 2.7442915179538843e-08,
+      "loss": 0.5773,
+      "step": 7683
+    },
+    {
+      "epoch": 1.8682227084852907,
+      "grad_norm": 19.875,
+      "learning_rate": 2.7342299158355374e-08,
+      "loss": 0.4528,
+      "step": 7684
+    },
+    {
+      "epoch": 1.8684658400194505,
+      "grad_norm": 17.5,
+      "learning_rate": 2.7241865884481412e-08,
+      "loss": 0.4812,
+      "step": 7685
+    },
+    {
+      "epoch": 1.8687089715536105,
+      "grad_norm": 19.125,
+      "learning_rate": 2.714161537292828e-08,
+      "loss": 0.9347,
+      "step": 7686
+    },
+    {
+      "epoch": 1.8689521030877705,
+      "grad_norm": 18.0,
+      "learning_rate": 2.7041547638680388e-08,
+      "loss": 0.4699,
+      "step": 7687
+    },
+    {
+      "epoch": 1.8691952346219305,
+      "grad_norm": 18.5,
+      "learning_rate": 2.6941662696694236e-08,
+      "loss": 0.6178,
+      "step": 7688
+    },
+    {
+      "epoch": 1.8694383661560905,
+      "grad_norm": 14.375,
+      "learning_rate": 2.6841960561899555e-08,
+      "loss": 0.2273,
+      "step": 7689
+    },
+    {
+      "epoch": 1.8696814976902503,
+      "grad_norm": 17.125,
+      "learning_rate": 2.6742441249198586e-08,
+      "loss": 0.7081,
+      "step": 7690
+    },
+    {
+      "epoch": 1.8699246292244105,
+      "grad_norm": 16.125,
+      "learning_rate": 2.664310477346624e-08,
+      "loss": 0.5804,
+      "step": 7691
+    },
+    {
+      "epoch": 1.8701677607585703,
+      "grad_norm": 21.625,
+      "learning_rate": 2.654395114954994e-08,
+      "loss": 0.6758,
+      "step": 7692
+    },
+    {
+      "epoch": 1.8704108922927305,
+      "grad_norm": 42.5,
+      "learning_rate": 2.6444980392269782e-08,
+      "loss": 0.7383,
+      "step": 7693
+    },
+    {
+      "epoch": 1.8706540238268903,
+      "grad_norm": 15.9375,
+      "learning_rate": 2.6346192516419065e-08,
+      "loss": 0.7225,
+      "step": 7694
+    },
+    {
+      "epoch": 1.8708971553610503,
+      "grad_norm": 24.25,
+      "learning_rate": 2.624758753676307e-08,
+      "loss": 0.6042,
+      "step": 7695
+    },
+    {
+      "epoch": 1.8711402868952103,
+      "grad_norm": 19.5,
+      "learning_rate": 2.614916546804e-08,
+      "loss": 1.126,
+      "step": 7696
+    },
+    {
+      "epoch": 1.8713834184293703,
+      "grad_norm": 21.0,
+      "learning_rate": 2.6050926324960728e-08,
+      "loss": 0.733,
+      "step": 7697
+    },
+    {
+      "epoch": 1.8716265499635303,
+      "grad_norm": 19.0,
+      "learning_rate": 2.595287012220893e-08,
+      "loss": 0.7603,
+      "step": 7698
+    },
+    {
+      "epoch": 1.8718696814976903,
+      "grad_norm": 29.5,
+      "learning_rate": 2.585499687444079e-08,
+      "loss": 0.7504,
+      "step": 7699
+    },
+    {
+      "epoch": 1.8721128130318503,
+      "grad_norm": 24.25,
+      "learning_rate": 2.5757306596284892e-08,
+      "loss": 0.8763,
+      "step": 7700
+    },
+    {
+      "epoch": 1.87235594456601,
+      "grad_norm": 18.5,
+      "learning_rate": 2.5659799302343026e-08,
+      "loss": 0.9594,
+      "step": 7701
+    },
+    {
+      "epoch": 1.8725990761001703,
+      "grad_norm": 17.125,
+      "learning_rate": 2.5562475007189364e-08,
+      "loss": 0.4957,
+      "step": 7702
+    },
+    {
+      "epoch": 1.87284220763433,
+      "grad_norm": 20.375,
+      "learning_rate": 2.546533372537033e-08,
+      "loss": 0.9014,
+      "step": 7703
+    },
+    {
+      "epoch": 1.8730853391684903,
+      "grad_norm": 20.125,
+      "learning_rate": 2.536837547140583e-08,
+      "loss": 0.4721,
+      "step": 7704
+    },
+    {
+      "epoch": 1.87332847070265,
+      "grad_norm": 26.75,
+      "learning_rate": 2.5271600259787617e-08,
+      "loss": 0.875,
+      "step": 7705
+    },
+    {
+      "epoch": 1.87357160223681,
+      "grad_norm": 23.5,
+      "learning_rate": 2.5175008104980636e-08,
+      "loss": 0.8767,
+      "step": 7706
+    },
+    {
+      "epoch": 1.87381473377097,
+      "grad_norm": 17.5,
+      "learning_rate": 2.5078599021421957e-08,
+      "loss": 0.5381,
+      "step": 7707
+    },
+    {
+      "epoch": 1.87405786530513,
+      "grad_norm": 17.75,
+      "learning_rate": 2.4982373023521994e-08,
+      "loss": 0.4825,
+      "step": 7708
+    },
+    {
+      "epoch": 1.87430099683929,
+      "grad_norm": 33.5,
+      "learning_rate": 2.4886330125663133e-08,
+      "loss": 1.1793,
+      "step": 7709
+    },
+    {
+      "epoch": 1.87454412837345,
+      "grad_norm": 26.0,
+      "learning_rate": 2.4790470342200556e-08,
+      "loss": 0.8028,
+      "step": 7710
+    },
+    {
+      "epoch": 1.87478725990761,
+      "grad_norm": 19.625,
+      "learning_rate": 2.4694793687462386e-08,
+      "loss": 1.033,
+      "step": 7711
+    },
+    {
+      "epoch": 1.8750303914417699,
+      "grad_norm": 25.125,
+      "learning_rate": 2.459930017574913e-08,
+      "loss": 0.9768,
+      "step": 7712
+    },
+    {
+      "epoch": 1.87527352297593,
+      "grad_norm": 20.25,
+      "learning_rate": 2.4503989821333675e-08,
+      "loss": 0.5419,
+      "step": 7713
+    },
+    {
+      "epoch": 1.8755166545100899,
+      "grad_norm": 20.125,
+      "learning_rate": 2.440886263846212e-08,
+      "loss": 0.9091,
+      "step": 7714
+    },
+    {
+      "epoch": 1.87575978604425,
+      "grad_norm": 30.25,
+      "learning_rate": 2.4313918641352684e-08,
+      "loss": 0.5382,
+      "step": 7715
+    },
+    {
+      "epoch": 1.8760029175784099,
+      "grad_norm": 27.0,
+      "learning_rate": 2.4219157844196512e-08,
+      "loss": 0.6541,
+      "step": 7716
+    },
+    {
+      "epoch": 1.8762460491125699,
+      "grad_norm": 18.125,
+      "learning_rate": 2.4124580261156994e-08,
+      "loss": 0.3218,
+      "step": 7717
+    },
+    {
+      "epoch": 1.8764891806467299,
+      "grad_norm": 23.5,
+      "learning_rate": 2.403018590637074e-08,
+      "loss": 0.6784,
+      "step": 7718
+    },
+    {
+      "epoch": 1.8767323121808899,
+      "grad_norm": 19.25,
+      "learning_rate": 2.3935974793946466e-08,
+      "loss": 0.5894,
+      "step": 7719
+    },
+    {
+      "epoch": 1.8769754437150499,
+      "grad_norm": 16.125,
+      "learning_rate": 2.3841946937965404e-08,
+      "loss": 0.7901,
+      "step": 7720
+    },
+    {
+      "epoch": 1.8772185752492099,
+      "grad_norm": 17.75,
+      "learning_rate": 2.3748102352481873e-08,
+      "loss": 0.6835,
+      "step": 7721
+    },
+    {
+      "epoch": 1.8774617067833699,
+      "grad_norm": 41.25,
+      "learning_rate": 2.3654441051522704e-08,
+      "loss": 1.0942,
+      "step": 7722
+    },
+    {
+      "epoch": 1.8777048383175297,
+      "grad_norm": 22.125,
+      "learning_rate": 2.3560963049086975e-08,
+      "loss": 0.7283,
+      "step": 7723
+    },
+    {
+      "epoch": 1.8779479698516899,
+      "grad_norm": 27.5,
+      "learning_rate": 2.346766835914671e-08,
+      "loss": 0.7578,
+      "step": 7724
+    },
+    {
+      "epoch": 1.8781911013858497,
+      "grad_norm": 17.5,
+      "learning_rate": 2.3374556995646307e-08,
+      "loss": 0.3706,
+      "step": 7725
+    },
+    {
+      "epoch": 1.8784342329200099,
+      "grad_norm": 25.375,
+      "learning_rate": 2.328162897250297e-08,
+      "loss": 0.8971,
+      "step": 7726
+    },
+    {
+      "epoch": 1.8786773644541697,
+      "grad_norm": 24.5,
+      "learning_rate": 2.3188884303606135e-08,
+      "loss": 0.7438,
+      "step": 7727
+    },
+    {
+      "epoch": 1.8789204959883297,
+      "grad_norm": 18.5,
+      "learning_rate": 2.309632300281847e-08,
+      "loss": 0.6857,
+      "step": 7728
+    },
+    {
+      "epoch": 1.8791636275224897,
+      "grad_norm": 20.5,
+      "learning_rate": 2.3003945083974745e-08,
+      "loss": 0.7616,
+      "step": 7729
+    },
+    {
+      "epoch": 1.8794067590566497,
+      "grad_norm": 16.875,
+      "learning_rate": 2.291175056088224e-08,
+      "loss": 0.5579,
+      "step": 7730
+    },
+    {
+      "epoch": 1.8796498905908097,
+      "grad_norm": 28.125,
+      "learning_rate": 2.2819739447321188e-08,
+      "loss": 0.8129,
+      "step": 7731
+    },
+    {
+      "epoch": 1.8798930221249694,
+      "grad_norm": 18.875,
+      "learning_rate": 2.2727911757044334e-08,
+      "loss": 0.4981,
+      "step": 7732
+    },
+    {
+      "epoch": 1.8801361536591297,
+      "grad_norm": 21.5,
+      "learning_rate": 2.2636267503776817e-08,
+      "loss": 0.7,
+      "step": 7733
+    },
+    {
+      "epoch": 1.8803792851932895,
+      "grad_norm": 22.0,
+      "learning_rate": 2.2544806701216145e-08,
+      "loss": 0.6185,
+      "step": 7734
+    },
+    {
+      "epoch": 1.8806224167274497,
+      "grad_norm": 19.75,
+      "learning_rate": 2.2453529363033328e-08,
+      "loss": 0.8049,
+      "step": 7735
+    },
+    {
+      "epoch": 1.8808655482616095,
+      "grad_norm": 16.75,
+      "learning_rate": 2.236243550287079e-08,
+      "loss": 0.4919,
+      "step": 7736
+    },
+    {
+      "epoch": 1.8811086797957695,
+      "grad_norm": 19.625,
+      "learning_rate": 2.2271525134344302e-08,
+      "loss": 0.9488,
+      "step": 7737
+    },
+    {
+      "epoch": 1.8813518113299295,
+      "grad_norm": 19.125,
+      "learning_rate": 2.2180798271042027e-08,
+      "loss": 0.8047,
+      "step": 7738
+    },
+    {
+      "epoch": 1.8815949428640895,
+      "grad_norm": 25.75,
+      "learning_rate": 2.209025492652464e-08,
+      "loss": 0.8352,
+      "step": 7739
+    },
+    {
+      "epoch": 1.8818380743982495,
+      "grad_norm": 20.875,
+      "learning_rate": 2.1999895114325488e-08,
+      "loss": 0.8531,
+      "step": 7740
+    },
+    {
+      "epoch": 1.8820812059324095,
+      "grad_norm": 17.875,
+      "learning_rate": 2.190971884795015e-08,
+      "loss": 1.0134,
+      "step": 7741
+    },
+    {
+      "epoch": 1.8823243374665695,
+      "grad_norm": 18.75,
+      "learning_rate": 2.1819726140877294e-08,
+      "loss": 0.5415,
+      "step": 7742
+    },
+    {
+      "epoch": 1.8825674690007292,
+      "grad_norm": 16.625,
+      "learning_rate": 2.172991700655769e-08,
+      "loss": 0.298,
+      "step": 7743
+    },
+    {
+      "epoch": 1.8828106005348895,
+      "grad_norm": 22.25,
+      "learning_rate": 2.1640291458415036e-08,
+      "loss": 0.617,
+      "step": 7744
+    },
+    {
+      "epoch": 1.8830537320690492,
+      "grad_norm": 19.5,
+      "learning_rate": 2.1550849509845152e-08,
+      "loss": 0.7055,
+      "step": 7745
+    },
+    {
+      "epoch": 1.8832968636032095,
+      "grad_norm": 14.0625,
+      "learning_rate": 2.146159117421706e-08,
+      "loss": 0.3976,
+      "step": 7746
+    },
+    {
+      "epoch": 1.8835399951373692,
+      "grad_norm": 17.75,
+      "learning_rate": 2.1372516464871623e-08,
+      "loss": 0.4532,
+      "step": 7747
+    },
+    {
+      "epoch": 1.8837831266715293,
+      "grad_norm": 17.25,
+      "learning_rate": 2.128362539512277e-08,
+      "loss": 0.4402,
+      "step": 7748
+    },
+    {
+      "epoch": 1.8840262582056893,
+      "grad_norm": 16.5,
+      "learning_rate": 2.119491797825682e-08,
+      "loss": 0.3731,
+      "step": 7749
+    },
+    {
+      "epoch": 1.8842693897398493,
+      "grad_norm": 22.875,
+      "learning_rate": 2.1106394227532605e-08,
+      "loss": 0.8768,
+      "step": 7750
+    },
+    {
+      "epoch": 1.8845125212740093,
+      "grad_norm": 26.0,
+      "learning_rate": 2.1018054156181355e-08,
+      "loss": 0.6892,
+      "step": 7751
+    },
+    {
+      "epoch": 1.8847556528081693,
+      "grad_norm": 25.625,
+      "learning_rate": 2.0929897777407226e-08,
+      "loss": 0.8972,
+      "step": 7752
+    },
+    {
+      "epoch": 1.8849987843423293,
+      "grad_norm": 16.5,
+      "learning_rate": 2.0841925104386764e-08,
+      "loss": 0.5226,
+      "step": 7753
+    },
+    {
+      "epoch": 1.885241915876489,
+      "grad_norm": 18.5,
+      "learning_rate": 2.075413615026875e-08,
+      "loss": 0.5157,
+      "step": 7754
+    },
+    {
+      "epoch": 1.8854850474106493,
+      "grad_norm": 22.375,
+      "learning_rate": 2.0666530928174917e-08,
+      "loss": 0.6178,
+      "step": 7755
+    },
+    {
+      "epoch": 1.885728178944809,
+      "grad_norm": 21.75,
+      "learning_rate": 2.057910945119937e-08,
+      "loss": 0.7248,
+      "step": 7756
+    },
+    {
+      "epoch": 1.8859713104789693,
+      "grad_norm": 19.75,
+      "learning_rate": 2.049187173240888e-08,
+      "loss": 0.651,
+      "step": 7757
+    },
+    {
+      "epoch": 1.886214442013129,
+      "grad_norm": 19.75,
+      "learning_rate": 2.0404817784842323e-08,
+      "loss": 0.4942,
+      "step": 7758
+    },
+    {
+      "epoch": 1.886457573547289,
+      "grad_norm": 20.0,
+      "learning_rate": 2.0317947621511653e-08,
+      "loss": 1.0092,
+      "step": 7759
+    },
+    {
+      "epoch": 1.886700705081449,
+      "grad_norm": 17.5,
+      "learning_rate": 2.0231261255401065e-08,
+      "loss": 0.8084,
+      "step": 7760
+    },
+    {
+      "epoch": 1.886943836615609,
+      "grad_norm": 19.875,
+      "learning_rate": 2.0144758699467276e-08,
+      "loss": 0.7384,
+      "step": 7761
+    },
+    {
+      "epoch": 1.887186968149769,
+      "grad_norm": 38.5,
+      "learning_rate": 2.005843996663967e-08,
+      "loss": 1.0055,
+      "step": 7762
+    },
+    {
+      "epoch": 1.887430099683929,
+      "grad_norm": 21.125,
+      "learning_rate": 1.997230506982001e-08,
+      "loss": 0.736,
+      "step": 7763
+    },
+    {
+      "epoch": 1.887673231218089,
+      "grad_norm": 17.875,
+      "learning_rate": 1.9886354021882852e-08,
+      "loss": 0.4741,
+      "step": 7764
+    },
+    {
+      "epoch": 1.8879163627522488,
+      "grad_norm": 28.125,
+      "learning_rate": 1.9800586835674596e-08,
+      "loss": 0.7851,
+      "step": 7765
+    },
+    {
+      "epoch": 1.888159494286409,
+      "grad_norm": 19.5,
+      "learning_rate": 1.971500352401512e-08,
+      "loss": 0.9229,
+      "step": 7766
+    },
+    {
+      "epoch": 1.8884026258205688,
+      "grad_norm": 24.75,
+      "learning_rate": 1.9629604099695997e-08,
+      "loss": 0.946,
+      "step": 7767
+    },
+    {
+      "epoch": 1.888645757354729,
+      "grad_norm": 37.0,
+      "learning_rate": 1.954438857548188e-08,
+      "loss": 1.1072,
+      "step": 7768
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 20.875,
+      "learning_rate": 1.9459356964109527e-08,
+      "loss": 0.4563,
+      "step": 7769
+    },
+    {
+      "epoch": 1.8891320204230488,
+      "grad_norm": 23.75,
+      "learning_rate": 1.9374509278288488e-08,
+      "loss": 0.5726,
+      "step": 7770
+    },
+    {
+      "epoch": 1.8893751519572088,
+      "grad_norm": 18.125,
+      "learning_rate": 1.9289845530700707e-08,
+      "loss": 0.9522,
+      "step": 7771
+    },
+    {
+      "epoch": 1.8896182834913688,
+      "grad_norm": 27.5,
+      "learning_rate": 1.9205365734000503e-08,
+      "loss": 1.043,
+      "step": 7772
+    },
+    {
+      "epoch": 1.8898614150255288,
+      "grad_norm": 17.625,
+      "learning_rate": 1.9121069900814998e-08,
+      "loss": 0.7864,
+      "step": 7773
+    },
+    {
+      "epoch": 1.8901045465596886,
+      "grad_norm": 19.625,
+      "learning_rate": 1.9036958043743697e-08,
+      "loss": 0.5282,
+      "step": 7774
+    },
+    {
+      "epoch": 1.8903476780938488,
+      "grad_norm": 22.125,
+      "learning_rate": 1.895303017535835e-08,
+      "loss": 0.5427,
+      "step": 7775
+    },
+    {
+      "epoch": 1.8905908096280086,
+      "grad_norm": 22.0,
+      "learning_rate": 1.8869286308203506e-08,
+      "loss": 1.0772,
+      "step": 7776
+    },
+    {
+      "epoch": 1.8908339411621689,
+      "grad_norm": 14.9375,
+      "learning_rate": 1.878572645479637e-08,
+      "loss": 0.3036,
+      "step": 7777
+    },
+    {
+      "epoch": 1.8910770726963286,
+      "grad_norm": 19.875,
+      "learning_rate": 1.8702350627626125e-08,
+      "loss": 0.8388,
+      "step": 7778
+    },
+    {
+      "epoch": 1.8913202042304889,
+      "grad_norm": 22.125,
+      "learning_rate": 1.8619158839154883e-08,
+      "loss": 1.0803,
+      "step": 7779
+    },
+    {
+      "epoch": 1.8915633357646486,
+      "grad_norm": 20.125,
+      "learning_rate": 1.8536151101817003e-08,
+      "loss": 1.1311,
+      "step": 7780
+    },
+    {
+      "epoch": 1.8918064672988086,
+      "grad_norm": 17.25,
+      "learning_rate": 1.84533274280195e-08,
+      "loss": 0.5389,
+      "step": 7781
+    },
+    {
+      "epoch": 1.8920495988329686,
+      "grad_norm": 21.875,
+      "learning_rate": 1.8370687830141508e-08,
+      "loss": 1.2689,
+      "step": 7782
+    },
+    {
+      "epoch": 1.8922927303671286,
+      "grad_norm": 16.625,
+      "learning_rate": 1.8288232320535504e-08,
+      "loss": 0.6092,
+      "step": 7783
+    },
+    {
+      "epoch": 1.8925358619012886,
+      "grad_norm": 19.625,
+      "learning_rate": 1.8205960911525518e-08,
+      "loss": 0.9303,
+      "step": 7784
+    },
+    {
+      "epoch": 1.8927789934354484,
+      "grad_norm": 20.75,
+      "learning_rate": 1.8123873615408515e-08,
+      "loss": 0.6299,
+      "step": 7785
+    },
+    {
+      "epoch": 1.8930221249696086,
+      "grad_norm": 17.625,
+      "learning_rate": 1.8041970444453716e-08,
+      "loss": 0.7145,
+      "step": 7786
+    },
+    {
+      "epoch": 1.8932652565037684,
+      "grad_norm": 15.3125,
+      "learning_rate": 1.7960251410903128e-08,
+      "loss": 0.5215,
+      "step": 7787
+    },
+    {
+      "epoch": 1.8935083880379286,
+      "grad_norm": 24.875,
+      "learning_rate": 1.7878716526971152e-08,
+      "loss": 0.7656,
+      "step": 7788
+    },
+    {
+      "epoch": 1.8937515195720884,
+      "grad_norm": 15.8125,
+      "learning_rate": 1.7797365804844285e-08,
+      "loss": 0.6144,
+      "step": 7789
+    },
+    {
+      "epoch": 1.8939946511062484,
+      "grad_norm": 17.5,
+      "learning_rate": 1.771619925668197e-08,
+      "loss": 0.3488,
+      "step": 7790
+    },
+    {
+      "epoch": 1.8942377826404084,
+      "grad_norm": 26.375,
+      "learning_rate": 1.763521689461603e-08,
+      "loss": 0.6864,
+      "step": 7791
+    },
+    {
+      "epoch": 1.8944809141745684,
+      "grad_norm": 19.125,
+      "learning_rate": 1.7554418730750673e-08,
+      "loss": 0.8534,
+      "step": 7792
+    },
+    {
+      "epoch": 1.8947240457087284,
+      "grad_norm": 17.625,
+      "learning_rate": 1.7473804777162346e-08,
+      "loss": 0.7827,
+      "step": 7793
+    },
+    {
+      "epoch": 1.8949671772428884,
+      "grad_norm": 17.0,
+      "learning_rate": 1.739337504590044e-08,
+      "loss": 0.4682,
+      "step": 7794
+    },
+    {
+      "epoch": 1.8952103087770484,
+      "grad_norm": 19.125,
+      "learning_rate": 1.731312954898659e-08,
+      "loss": 0.7828,
+      "step": 7795
+    },
+    {
+      "epoch": 1.8954534403112082,
+      "grad_norm": 20.5,
+      "learning_rate": 1.7233068298414536e-08,
+      "loss": 0.7769,
+      "step": 7796
+    },
+    {
+      "epoch": 1.8956965718453684,
+      "grad_norm": 22.5,
+      "learning_rate": 1.715319130615123e-08,
+      "loss": 0.8071,
+      "step": 7797
+    },
+    {
+      "epoch": 1.8959397033795282,
+      "grad_norm": 18.75,
+      "learning_rate": 1.7073498584135457e-08,
+      "loss": 0.5649,
+      "step": 7798
+    },
+    {
+      "epoch": 1.8961828349136884,
+      "grad_norm": 22.375,
+      "learning_rate": 1.699399014427852e-08,
+      "loss": 0.5957,
+      "step": 7799
+    },
+    {
+      "epoch": 1.8964259664478482,
+      "grad_norm": 21.625,
+      "learning_rate": 1.6914665998464808e-08,
+      "loss": 0.7547,
+      "step": 7800
+    },
+    {
+      "epoch": 1.8966690979820082,
+      "grad_norm": 25.125,
+      "learning_rate": 1.6835526158550248e-08,
+      "loss": 0.9502,
+      "step": 7801
+    },
+    {
+      "epoch": 1.8969122295161682,
+      "grad_norm": 20.5,
+      "learning_rate": 1.675657063636385e-08,
+      "loss": 1.0027,
+      "step": 7802
+    },
+    {
+      "epoch": 1.8971553610503282,
+      "grad_norm": 21.5,
+      "learning_rate": 1.6677799443706593e-08,
+      "loss": 0.6465,
+      "step": 7803
+    },
+    {
+      "epoch": 1.8973984925844882,
+      "grad_norm": 27.125,
+      "learning_rate": 1.659921259235267e-08,
+      "loss": 0.71,
+      "step": 7804
+    },
+    {
+      "epoch": 1.8976416241186482,
+      "grad_norm": 28.625,
+      "learning_rate": 1.6520810094047963e-08,
+      "loss": 0.8231,
+      "step": 7805
+    },
+    {
+      "epoch": 1.8978847556528082,
+      "grad_norm": 20.125,
+      "learning_rate": 1.6442591960510873e-08,
+      "loss": 0.5697,
+      "step": 7806
+    },
+    {
+      "epoch": 1.898127887186968,
+      "grad_norm": 25.375,
+      "learning_rate": 1.6364558203432885e-08,
+      "loss": 0.7279,
+      "step": 7807
+    },
+    {
+      "epoch": 1.8983710187211282,
+      "grad_norm": 21.5,
+      "learning_rate": 1.6286708834477304e-08,
+      "loss": 0.6939,
+      "step": 7808
+    },
+    {
+      "epoch": 1.898614150255288,
+      "grad_norm": 19.125,
+      "learning_rate": 1.6209043865280105e-08,
+      "loss": 0.7817,
+      "step": 7809
+    },
+    {
+      "epoch": 1.8988572817894482,
+      "grad_norm": 22.125,
+      "learning_rate": 1.61315633074495e-08,
+      "loss": 0.4963,
+      "step": 7810
+    },
+    {
+      "epoch": 1.899100413323608,
+      "grad_norm": 23.0,
+      "learning_rate": 1.605426717256636e-08,
+      "loss": 1.0589,
+      "step": 7811
+    },
+    {
+      "epoch": 1.899343544857768,
+      "grad_norm": 19.625,
+      "learning_rate": 1.5977155472184093e-08,
+      "loss": 0.7394,
+      "step": 7812
+    },
+    {
+      "epoch": 1.899586676391928,
+      "grad_norm": 19.875,
+      "learning_rate": 1.5900228217828052e-08,
+      "loss": 0.5944,
+      "step": 7813
+    },
+    {
+      "epoch": 1.899829807926088,
+      "grad_norm": 21.375,
+      "learning_rate": 1.582348542099682e-08,
+      "loss": 0.7664,
+      "step": 7814
+    },
+    {
+      "epoch": 1.900072939460248,
+      "grad_norm": 15.625,
+      "learning_rate": 1.574692709316053e-08,
+      "loss": 0.3386,
+      "step": 7815
+    },
+    {
+      "epoch": 1.900316070994408,
+      "grad_norm": 20.5,
+      "learning_rate": 1.567055324576225e-08,
+      "loss": 0.7334,
+      "step": 7816
+    },
+    {
+      "epoch": 1.900559202528568,
+      "grad_norm": 17.75,
+      "learning_rate": 1.559436389021743e-08,
+      "loss": 0.473,
+      "step": 7817
+    },
+    {
+      "epoch": 1.9008023340627278,
+      "grad_norm": 24.5,
+      "learning_rate": 1.5518359037913905e-08,
+      "loss": 0.7397,
+      "step": 7818
+    },
+    {
+      "epoch": 1.901045465596888,
+      "grad_norm": 18.25,
+      "learning_rate": 1.5442538700211756e-08,
+      "loss": 0.6509,
+      "step": 7819
+    },
+    {
+      "epoch": 1.9012885971310478,
+      "grad_norm": 13.8125,
+      "learning_rate": 1.5366902888443723e-08,
+      "loss": 0.3035,
+      "step": 7820
+    },
+    {
+      "epoch": 1.901531728665208,
+      "grad_norm": 18.125,
+      "learning_rate": 1.529145161391493e-08,
+      "loss": 0.5966,
+      "step": 7821
+    },
+    {
+      "epoch": 1.9017748601993678,
+      "grad_norm": 19.75,
+      "learning_rate": 1.5216184887902884e-08,
+      "loss": 0.4218,
+      "step": 7822
+    },
+    {
+      "epoch": 1.9020179917335278,
+      "grad_norm": 18.375,
+      "learning_rate": 1.5141102721657475e-08,
+      "loss": 0.7222,
+      "step": 7823
+    },
+    {
+      "epoch": 1.9022611232676878,
+      "grad_norm": 18.125,
+      "learning_rate": 1.5066205126400973e-08,
+      "loss": 0.6308,
+      "step": 7824
+    },
+    {
+      "epoch": 1.9025042548018478,
+      "grad_norm": 23.5,
+      "learning_rate": 1.4991492113328177e-08,
+      "loss": 0.792,
+      "step": 7825
+    },
+    {
+      "epoch": 1.9027473863360078,
+      "grad_norm": 17.75,
+      "learning_rate": 1.491696369360626e-08,
+      "loss": 0.7579,
+      "step": 7826
+    },
+    {
+      "epoch": 1.9029905178701676,
+      "grad_norm": 23.625,
+      "learning_rate": 1.4842619878374654e-08,
+      "loss": 0.8057,
+      "step": 7827
+    },
+    {
+      "epoch": 1.9032336494043278,
+      "grad_norm": 16.0,
+      "learning_rate": 1.4768460678745573e-08,
+      "loss": 0.4927,
+      "step": 7828
+    },
+    {
+      "epoch": 1.9034767809384876,
+      "grad_norm": 19.0,
+      "learning_rate": 1.469448610580307e-08,
+      "loss": 0.6272,
+      "step": 7829
+    },
+    {
+      "epoch": 1.9037199124726478,
+      "grad_norm": 20.625,
+      "learning_rate": 1.4620696170604132e-08,
+      "loss": 0.4111,
+      "step": 7830
+    },
+    {
+      "epoch": 1.9039630440068076,
+      "grad_norm": 25.0,
+      "learning_rate": 1.4547090884177855e-08,
+      "loss": 0.9337,
+      "step": 7831
+    },
+    {
+      "epoch": 1.9042061755409678,
+      "grad_norm": 17.5,
+      "learning_rate": 1.4473670257525996e-08,
+      "loss": 0.5651,
+      "step": 7832
+    },
+    {
+      "epoch": 1.9044493070751276,
+      "grad_norm": 16.625,
+      "learning_rate": 1.4400434301622135e-08,
+      "loss": 0.4158,
+      "step": 7833
+    },
+    {
+      "epoch": 1.9046924386092876,
+      "grad_norm": 18.125,
+      "learning_rate": 1.432738302741321e-08,
+      "loss": 0.5588,
+      "step": 7834
+    },
+    {
+      "epoch": 1.9049355701434476,
+      "grad_norm": 18.0,
+      "learning_rate": 1.4254516445817573e-08,
+      "loss": 0.7193,
+      "step": 7835
+    },
+    {
+      "epoch": 1.9051787016776076,
+      "grad_norm": 23.875,
+      "learning_rate": 1.4181834567726372e-08,
+      "loss": 1.0141,
+      "step": 7836
+    },
+    {
+      "epoch": 1.9054218332117676,
+      "grad_norm": 17.375,
+      "learning_rate": 1.4109337404003414e-08,
+      "loss": 0.4704,
+      "step": 7837
+    },
+    {
+      "epoch": 1.9056649647459274,
+      "grad_norm": 20.0,
+      "learning_rate": 1.4037024965484618e-08,
+      "loss": 0.6783,
+      "step": 7838
+    },
+    {
+      "epoch": 1.9059080962800876,
+      "grad_norm": 16.0,
+      "learning_rate": 1.3964897262978283e-08,
+      "loss": 0.5158,
+      "step": 7839
+    },
+    {
+      "epoch": 1.9061512278142474,
+      "grad_norm": 21.75,
+      "learning_rate": 1.389295430726495e-08,
+      "loss": 0.6328,
+      "step": 7840
+    },
+    {
+      "epoch": 1.9063943593484076,
+      "grad_norm": 26.25,
+      "learning_rate": 1.38211961090981e-08,
+      "loss": 0.6498,
+      "step": 7841
+    },
+    {
+      "epoch": 1.9066374908825674,
+      "grad_norm": 17.875,
+      "learning_rate": 1.3749622679202907e-08,
+      "loss": 0.7024,
+      "step": 7842
+    },
+    {
+      "epoch": 1.9068806224167274,
+      "grad_norm": 20.375,
+      "learning_rate": 1.3678234028277476e-08,
+      "loss": 0.5721,
+      "step": 7843
+    },
+    {
+      "epoch": 1.9071237539508874,
+      "grad_norm": 15.5625,
+      "learning_rate": 1.3607030166991747e-08,
+      "loss": 0.5827,
+      "step": 7844
+    },
+    {
+      "epoch": 1.9073668854850474,
+      "grad_norm": 24.125,
+      "learning_rate": 1.3536011105988872e-08,
+      "loss": 1.0471,
+      "step": 7845
+    },
+    {
+      "epoch": 1.9076100170192074,
+      "grad_norm": 20.0,
+      "learning_rate": 1.3465176855883555e-08,
+      "loss": 0.9289,
+      "step": 7846
+    },
+    {
+      "epoch": 1.9078531485533674,
+      "grad_norm": 19.0,
+      "learning_rate": 1.339452742726316e-08,
+      "loss": 0.6694,
+      "step": 7847
+    },
+    {
+      "epoch": 1.9080962800875274,
+      "grad_norm": 19.5,
+      "learning_rate": 1.3324062830687712e-08,
+      "loss": 0.7547,
+      "step": 7848
+    },
+    {
+      "epoch": 1.9083394116216872,
+      "grad_norm": 17.125,
+      "learning_rate": 1.3253783076689064e-08,
+      "loss": 0.5016,
+      "step": 7849
+    },
+    {
+      "epoch": 1.9085825431558474,
+      "grad_norm": 18.375,
+      "learning_rate": 1.3183688175772008e-08,
+      "loss": 0.8364,
+      "step": 7850
+    },
+    {
+      "epoch": 1.9088256746900072,
+      "grad_norm": 17.0,
+      "learning_rate": 1.3113778138413164e-08,
+      "loss": 0.4185,
+      "step": 7851
+    },
+    {
+      "epoch": 1.9090688062241674,
+      "grad_norm": 18.0,
+      "learning_rate": 1.304405297506195e-08,
+      "loss": 0.4495,
+      "step": 7852
+    },
+    {
+      "epoch": 1.9093119377583272,
+      "grad_norm": 16.625,
+      "learning_rate": 1.2974512696140167e-08,
+      "loss": 0.583,
+      "step": 7853
+    },
+    {
+      "epoch": 1.9095550692924872,
+      "grad_norm": 20.875,
+      "learning_rate": 1.290515731204145e-08,
+      "loss": 0.6287,
+      "step": 7854
+    },
+    {
+      "epoch": 1.9097982008266472,
+      "grad_norm": 18.0,
+      "learning_rate": 1.2835986833132502e-08,
+      "loss": 0.5895,
+      "step": 7855
+    },
+    {
+      "epoch": 1.9100413323608072,
+      "grad_norm": 25.875,
+      "learning_rate": 1.2767001269751722e-08,
+      "loss": 0.9779,
+      "step": 7856
+    },
+    {
+      "epoch": 1.9102844638949672,
+      "grad_norm": 21.0,
+      "learning_rate": 1.2698200632210445e-08,
+      "loss": 0.6093,
+      "step": 7857
+    },
+    {
+      "epoch": 1.9105275954291272,
+      "grad_norm": 19.0,
+      "learning_rate": 1.2629584930791972e-08,
+      "loss": 0.6462,
+      "step": 7858
+    },
+    {
+      "epoch": 1.9107707269632872,
+      "grad_norm": 18.5,
+      "learning_rate": 1.2561154175752266e-08,
+      "loss": 0.9409,
+      "step": 7859
+    },
+    {
+      "epoch": 1.911013858497447,
+      "grad_norm": 26.875,
+      "learning_rate": 1.2492908377319257e-08,
+      "loss": 0.6514,
+      "step": 7860
+    },
+    {
+      "epoch": 1.9112569900316072,
+      "grad_norm": 20.375,
+      "learning_rate": 1.2424847545693536e-08,
+      "loss": 0.732,
+      "step": 7861
+    },
+    {
+      "epoch": 1.911500121565767,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.2356971691048075e-08,
+      "loss": 0.4285,
+      "step": 7862
+    },
+    {
+      "epoch": 1.9117432530999272,
+      "grad_norm": 27.125,
+      "learning_rate": 1.2289280823528094e-08,
+      "loss": 0.9337,
+      "step": 7863
+    },
+    {
+      "epoch": 1.911986384634087,
+      "grad_norm": 23.625,
+      "learning_rate": 1.2221774953251053e-08,
+      "loss": 0.8571,
+      "step": 7864
+    },
+    {
+      "epoch": 1.912229516168247,
+      "grad_norm": 19.5,
+      "learning_rate": 1.2154454090306939e-08,
+      "loss": 0.6406,
+      "step": 7865
+    },
+    {
+      "epoch": 1.912472647702407,
+      "grad_norm": 20.75,
+      "learning_rate": 1.2087318244757978e-08,
+      "loss": 0.5717,
+      "step": 7866
+    },
+    {
+      "epoch": 1.912715779236567,
+      "grad_norm": 21.75,
+      "learning_rate": 1.2020367426638785e-08,
+      "loss": 0.7081,
+      "step": 7867
+    },
+    {
+      "epoch": 1.912958910770727,
+      "grad_norm": 20.0,
+      "learning_rate": 1.1953601645956214e-08,
+      "loss": 0.8989,
+      "step": 7868
+    },
+    {
+      "epoch": 1.913202042304887,
+      "grad_norm": 20.625,
+      "learning_rate": 1.1887020912689784e-08,
+      "loss": 0.6977,
+      "step": 7869
+    },
+    {
+      "epoch": 1.913445173839047,
+      "grad_norm": 22.75,
+      "learning_rate": 1.1820625236791116e-08,
+      "loss": 0.6997,
+      "step": 7870
+    },
+    {
+      "epoch": 1.9136883053732068,
+      "grad_norm": 21.0,
+      "learning_rate": 1.1754414628183941e-08,
+      "loss": 0.6479,
+      "step": 7871
+    },
+    {
+      "epoch": 1.913931436907367,
+      "grad_norm": 20.125,
+      "learning_rate": 1.1688389096764923e-08,
+      "loss": 0.8286,
+      "step": 7872
+    },
+    {
+      "epoch": 1.9141745684415268,
+      "grad_norm": 19.875,
+      "learning_rate": 1.162254865240242e-08,
+      "loss": 0.8877,
+      "step": 7873
+    },
+    {
+      "epoch": 1.914417699975687,
+      "grad_norm": 19.25,
+      "learning_rate": 1.1556893304937588e-08,
+      "loss": 0.8329,
+      "step": 7874
+    },
+    {
+      "epoch": 1.9146608315098468,
+      "grad_norm": 18.875,
+      "learning_rate": 1.1491423064183549e-08,
+      "loss": 0.7368,
+      "step": 7875
+    },
+    {
+      "epoch": 1.9149039630440068,
+      "grad_norm": 25.375,
+      "learning_rate": 1.1426137939926362e-08,
+      "loss": 0.8761,
+      "step": 7876
+    },
+    {
+      "epoch": 1.9151470945781668,
+      "grad_norm": 24.25,
+      "learning_rate": 1.1361037941923641e-08,
+      "loss": 1.0683,
+      "step": 7877
+    },
+    {
+      "epoch": 1.9153902261123268,
+      "grad_norm": 28.5,
+      "learning_rate": 1.1296123079905796e-08,
+      "loss": 0.8153,
+      "step": 7878
+    },
+    {
+      "epoch": 1.9156333576464868,
+      "grad_norm": 23.875,
+      "learning_rate": 1.123139336357562e-08,
+      "loss": 0.8908,
+      "step": 7879
+    },
+    {
+      "epoch": 1.9158764891806466,
+      "grad_norm": 18.25,
+      "learning_rate": 1.1166848802607877e-08,
+      "loss": 0.6323,
+      "step": 7880
+    },
+    {
+      "epoch": 1.9161196207148068,
+      "grad_norm": 18.875,
+      "learning_rate": 1.1102489406649986e-08,
+      "loss": 0.6903,
+      "step": 7881
+    },
+    {
+      "epoch": 1.9163627522489666,
+      "grad_norm": 19.875,
+      "learning_rate": 1.1038315185321613e-08,
+      "loss": 0.7712,
+      "step": 7882
+    },
+    {
+      "epoch": 1.9166058837831268,
+      "grad_norm": 19.875,
+      "learning_rate": 1.097432614821467e-08,
+      "loss": 0.5251,
+      "step": 7883
+    },
+    {
+      "epoch": 1.9168490153172866,
+      "grad_norm": 24.0,
+      "learning_rate": 1.0910522304893312e-08,
+      "loss": 0.7394,
+      "step": 7884
+    },
+    {
+      "epoch": 1.9170921468514468,
+      "grad_norm": 22.875,
+      "learning_rate": 1.0846903664894215e-08,
+      "loss": 0.4537,
+      "step": 7885
+    },
+    {
+      "epoch": 1.9173352783856066,
+      "grad_norm": 20.125,
+      "learning_rate": 1.07834702377263e-08,
+      "loss": 0.5776,
+      "step": 7886
+    },
+    {
+      "epoch": 1.9175784099197666,
+      "grad_norm": 21.125,
+      "learning_rate": 1.0720222032870874e-08,
+      "loss": 0.857,
+      "step": 7887
+    },
+    {
+      "epoch": 1.9178215414539266,
+      "grad_norm": 26.75,
+      "learning_rate": 1.0657159059781208e-08,
+      "loss": 0.9864,
+      "step": 7888
+    },
+    {
+      "epoch": 1.9180646729880866,
+      "grad_norm": 18.0,
+      "learning_rate": 1.059428132788337e-08,
+      "loss": 0.6663,
+      "step": 7889
+    },
+    {
+      "epoch": 1.9183078045222466,
+      "grad_norm": 22.125,
+      "learning_rate": 1.0531588846575403e-08,
+      "loss": 0.6014,
+      "step": 7890
+    },
+    {
+      "epoch": 1.9185509360564064,
+      "grad_norm": 24.0,
+      "learning_rate": 1.0469081625228e-08,
+      "loss": 0.5684,
+      "step": 7891
+    },
+    {
+      "epoch": 1.9187940675905666,
+      "grad_norm": 17.375,
+      "learning_rate": 1.0406759673183554e-08,
+      "loss": 0.5461,
+      "step": 7892
+    },
+    {
+      "epoch": 1.9190371991247264,
+      "grad_norm": 21.125,
+      "learning_rate": 1.0344622999757525e-08,
+      "loss": 0.4815,
+      "step": 7893
+    },
+    {
+      "epoch": 1.9192803306588866,
+      "grad_norm": 14.8125,
+      "learning_rate": 1.028267161423721e-08,
+      "loss": 0.3031,
+      "step": 7894
+    },
+    {
+      "epoch": 1.9195234621930464,
+      "grad_norm": 16.375,
+      "learning_rate": 1.0220905525882285e-08,
+      "loss": 0.451,
+      "step": 7895
+    },
+    {
+      "epoch": 1.9197665937272064,
+      "grad_norm": 18.5,
+      "learning_rate": 1.0159324743924809e-08,
+      "loss": 0.3167,
+      "step": 7896
+    },
+    {
+      "epoch": 1.9200097252613664,
+      "grad_norm": 23.875,
+      "learning_rate": 1.0097929277568946e-08,
+      "loss": 0.7654,
+      "step": 7897
+    },
+    {
+      "epoch": 1.9202528567955264,
+      "grad_norm": 20.25,
+      "learning_rate": 1.0036719135991385e-08,
+      "loss": 0.577,
+      "step": 7898
+    },
+    {
+      "epoch": 1.9204959883296864,
+      "grad_norm": 21.625,
+      "learning_rate": 9.975694328341057e-09,
+      "loss": 0.8609,
+      "step": 7899
+    },
+    {
+      "epoch": 1.9207391198638464,
+      "grad_norm": 18.125,
+      "learning_rate": 9.914854863739138e-09,
+      "loss": 0.501,
+      "step": 7900
+    },
+    {
+      "epoch": 1.9209822513980064,
+      "grad_norm": 18.75,
+      "learning_rate": 9.854200751279325e-09,
+      "loss": 0.4953,
+      "step": 7901
+    },
+    {
+      "epoch": 1.9212253829321662,
+      "grad_norm": 15.0625,
+      "learning_rate": 9.793732000027145e-09,
+      "loss": 0.4904,
+      "step": 7902
+    },
+    {
+      "epoch": 1.9214685144663264,
+      "grad_norm": 15.0625,
+      "learning_rate": 9.733448619020786e-09,
+      "loss": 0.612,
+      "step": 7903
+    },
+    {
+      "epoch": 1.9217116460004862,
+      "grad_norm": 16.125,
+      "learning_rate": 9.673350617270816e-09,
+      "loss": 0.619,
+      "step": 7904
+    },
+    {
+      "epoch": 1.9219547775346464,
+      "grad_norm": 22.25,
+      "learning_rate": 9.613438003759634e-09,
+      "loss": 0.8661,
+      "step": 7905
+    },
+    {
+      "epoch": 1.9221979090688062,
+      "grad_norm": 18.75,
+      "learning_rate": 9.553710787442438e-09,
+      "loss": 0.3533,
+      "step": 7906
+    },
+    {
+      "epoch": 1.9224410406029662,
+      "grad_norm": 19.5,
+      "learning_rate": 9.494168977246394e-09,
+      "loss": 0.5105,
+      "step": 7907
+    },
+    {
+      "epoch": 1.9226841721371262,
+      "grad_norm": 13.5,
+      "learning_rate": 9.434812582071045e-09,
+      "loss": 0.3481,
+      "step": 7908
+    },
+    {
+      "epoch": 1.9229273036712862,
+      "grad_norm": 19.5,
+      "learning_rate": 9.375641610788327e-09,
+      "loss": 0.815,
+      "step": 7909
+    },
+    {
+      "epoch": 1.9231704352054462,
+      "grad_norm": 20.625,
+      "learning_rate": 9.316656072242275e-09,
+      "loss": 0.712,
+      "step": 7910
+    },
+    {
+      "epoch": 1.9234135667396062,
+      "grad_norm": 27.25,
+      "learning_rate": 9.257855975249308e-09,
+      "loss": 0.9866,
+      "step": 7911
+    },
+    {
+      "epoch": 1.9236566982737662,
+      "grad_norm": 17.75,
+      "learning_rate": 9.199241328598092e-09,
+      "loss": 1.1286,
+      "step": 7912
+    },
+    {
+      "epoch": 1.923899829807926,
+      "grad_norm": 20.125,
+      "learning_rate": 9.14081214104967e-09,
+      "loss": 0.7686,
+      "step": 7913
+    },
+    {
+      "epoch": 1.9241429613420862,
+      "grad_norm": 20.25,
+      "learning_rate": 9.0825684213372e-09,
+      "loss": 0.8863,
+      "step": 7914
+    },
+    {
+      "epoch": 1.924386092876246,
+      "grad_norm": 21.125,
+      "learning_rate": 9.024510178166357e-09,
+      "loss": 0.8258,
+      "step": 7915
+    },
+    {
+      "epoch": 1.9246292244104062,
+      "grad_norm": 25.875,
+      "learning_rate": 8.966637420214779e-09,
+      "loss": 0.8853,
+      "step": 7916
+    },
+    {
+      "epoch": 1.924872355944566,
+      "grad_norm": 19.5,
+      "learning_rate": 8.908950156132635e-09,
+      "loss": 0.7537,
+      "step": 7917
+    },
+    {
+      "epoch": 1.925115487478726,
+      "grad_norm": 21.0,
+      "learning_rate": 8.851448394542328e-09,
+      "loss": 0.7481,
+      "step": 7918
+    },
+    {
+      "epoch": 1.925358619012886,
+      "grad_norm": 20.75,
+      "learning_rate": 8.794132144038376e-09,
+      "loss": 0.7676,
+      "step": 7919
+    },
+    {
+      "epoch": 1.925601750547046,
+      "grad_norm": 22.875,
+      "learning_rate": 8.737001413187813e-09,
+      "loss": 0.6935,
+      "step": 7920
+    },
+    {
+      "epoch": 1.925844882081206,
+      "grad_norm": 16.625,
+      "learning_rate": 8.680056210529642e-09,
+      "loss": 0.3174,
+      "step": 7921
+    },
+    {
+      "epoch": 1.926088013615366,
+      "grad_norm": 16.75,
+      "learning_rate": 8.623296544575526e-09,
+      "loss": 0.5871,
+      "step": 7922
+    },
+    {
+      "epoch": 1.926331145149526,
+      "grad_norm": 30.5,
+      "learning_rate": 8.56672242380896e-09,
+      "loss": 0.7486,
+      "step": 7923
+    },
+    {
+      "epoch": 1.9265742766836857,
+      "grad_norm": 26.5,
+      "learning_rate": 8.510333856686092e-09,
+      "loss": 0.8505,
+      "step": 7924
+    },
+    {
+      "epoch": 1.926817408217846,
+      "grad_norm": 24.25,
+      "learning_rate": 8.454130851635045e-09,
+      "loss": 0.8154,
+      "step": 7925
+    },
+    {
+      "epoch": 1.9270605397520058,
+      "grad_norm": 23.375,
+      "learning_rate": 8.39811341705632e-09,
+      "loss": 0.8016,
+      "step": 7926
+    },
+    {
+      "epoch": 1.927303671286166,
+      "grad_norm": 22.75,
+      "learning_rate": 8.342281561322801e-09,
+      "loss": 0.7308,
+      "step": 7927
+    },
+    {
+      "epoch": 1.9275468028203258,
+      "grad_norm": 18.875,
+      "learning_rate": 8.286635292779621e-09,
+      "loss": 0.6389,
+      "step": 7928
+    },
+    {
+      "epoch": 1.9277899343544858,
+      "grad_norm": 30.125,
+      "learning_rate": 8.231174619743736e-09,
+      "loss": 0.9354,
+      "step": 7929
+    },
+    {
+      "epoch": 1.9280330658886458,
+      "grad_norm": 19.375,
+      "learning_rate": 8.175899550504906e-09,
+      "loss": 0.7407,
+      "step": 7930
+    },
+    {
+      "epoch": 1.9282761974228058,
+      "grad_norm": 26.0,
+      "learning_rate": 8.120810093324993e-09,
+      "loss": 0.6151,
+      "step": 7931
+    },
+    {
+      "epoch": 1.9285193289569658,
+      "grad_norm": 17.75,
+      "learning_rate": 8.065906256438105e-09,
+      "loss": 0.7461,
+      "step": 7932
+    },
+    {
+      "epoch": 1.9287624604911255,
+      "grad_norm": 24.25,
+      "learning_rate": 8.011188048050316e-09,
+      "loss": 0.7335,
+      "step": 7933
+    },
+    {
+      "epoch": 1.9290055920252858,
+      "grad_norm": 32.0,
+      "learning_rate": 7.956655476340503e-09,
+      "loss": 1.4917,
+      "step": 7934
+    },
+    {
+      "epoch": 1.9292487235594455,
+      "grad_norm": 20.5,
+      "learning_rate": 7.902308549459226e-09,
+      "loss": 0.8437,
+      "step": 7935
+    },
+    {
+      "epoch": 1.9294918550936058,
+      "grad_norm": 22.375,
+      "learning_rate": 7.84814727552971e-09,
+      "loss": 1.0596,
+      "step": 7936
+    },
+    {
+      "epoch": 1.9297349866277655,
+      "grad_norm": 22.75,
+      "learning_rate": 7.794171662647287e-09,
+      "loss": 0.7269,
+      "step": 7937
+    },
+    {
+      "epoch": 1.9299781181619255,
+      "grad_norm": 19.875,
+      "learning_rate": 7.740381718879669e-09,
+      "loss": 1.0005,
+      "step": 7938
+    },
+    {
+      "epoch": 1.9302212496960856,
+      "grad_norm": 28.375,
+      "learning_rate": 7.686777452266537e-09,
+      "loss": 0.7057,
+      "step": 7939
+    },
+    {
+      "epoch": 1.9304643812302456,
+      "grad_norm": 20.375,
+      "learning_rate": 7.633358870819812e-09,
+      "loss": 0.6135,
+      "step": 7940
+    },
+    {
+      "epoch": 1.9307075127644056,
+      "grad_norm": 17.375,
+      "learning_rate": 7.580125982524223e-09,
+      "loss": 0.5086,
+      "step": 7941
+    },
+    {
+      "epoch": 1.9309506442985656,
+      "grad_norm": 27.25,
+      "learning_rate": 7.527078795336179e-09,
+      "loss": 0.5235,
+      "step": 7942
+    },
+    {
+      "epoch": 1.9311937758327256,
+      "grad_norm": 24.75,
+      "learning_rate": 7.47421731718434e-09,
+      "loss": 1.2145,
+      "step": 7943
+    },
+    {
+      "epoch": 1.9314369073668853,
+      "grad_norm": 20.125,
+      "learning_rate": 7.421541555969885e-09,
+      "loss": 0.6006,
+      "step": 7944
+    },
+    {
+      "epoch": 1.9316800389010456,
+      "grad_norm": 15.75,
+      "learning_rate": 7.369051519566101e-09,
+      "loss": 0.3053,
+      "step": 7945
+    },
+    {
+      "epoch": 1.9319231704352053,
+      "grad_norm": 22.375,
+      "learning_rate": 7.316747215818654e-09,
+      "loss": 1.1693,
+      "step": 7946
+    },
+    {
+      "epoch": 1.9321663019693656,
+      "grad_norm": 16.75,
+      "learning_rate": 7.264628652545042e-09,
+      "loss": 0.4853,
+      "step": 7947
+    },
+    {
+      "epoch": 1.9324094335035253,
+      "grad_norm": 23.5,
+      "learning_rate": 7.212695837535561e-09,
+      "loss": 0.7163,
+      "step": 7948
+    },
+    {
+      "epoch": 1.9326525650376853,
+      "grad_norm": 24.0,
+      "learning_rate": 7.160948778552474e-09,
+      "loss": 0.9888,
+      "step": 7949
+    },
+    {
+      "epoch": 1.9328956965718453,
+      "grad_norm": 19.5,
+      "learning_rate": 7.10938748333001e-09,
+      "loss": 0.5207,
+      "step": 7950
+    },
+    {
+      "epoch": 1.9331388281060053,
+      "grad_norm": 19.75,
+      "learning_rate": 7.0580119595752e-09,
+      "loss": 0.3175,
+      "step": 7951
+    },
+    {
+      "epoch": 1.9333819596401653,
+      "grad_norm": 24.125,
+      "learning_rate": 7.0068222149666246e-09,
+      "loss": 0.851,
+      "step": 7952
+    },
+    {
+      "epoch": 1.9336250911743253,
+      "grad_norm": 19.5,
+      "learning_rate": 6.955818257155938e-09,
+      "loss": 0.6074,
+      "step": 7953
+    },
+    {
+      "epoch": 1.9338682227084854,
+      "grad_norm": 21.125,
+      "learning_rate": 6.905000093766212e-09,
+      "loss": 0.5639,
+      "step": 7954
+    },
+    {
+      "epoch": 1.9341113542426451,
+      "grad_norm": 17.5,
+      "learning_rate": 6.854367732393175e-09,
+      "loss": 0.4098,
+      "step": 7955
+    },
+    {
+      "epoch": 1.9343544857768054,
+      "grad_norm": 17.75,
+      "learning_rate": 6.8039211806048e-09,
+      "loss": 0.9482,
+      "step": 7956
+    },
+    {
+      "epoch": 1.9345976173109651,
+      "grad_norm": 17.125,
+      "learning_rate": 6.7536604459411684e-09,
+      "loss": 0.4689,
+      "step": 7957
+    },
+    {
+      "epoch": 1.9348407488451254,
+      "grad_norm": 20.875,
+      "learning_rate": 6.703585535914603e-09,
+      "loss": 0.8622,
+      "step": 7958
+    },
+    {
+      "epoch": 1.9350838803792851,
+      "grad_norm": 19.875,
+      "learning_rate": 6.653696458009673e-09,
+      "loss": 0.6347,
+      "step": 7959
+    },
+    {
+      "epoch": 1.9353270119134451,
+      "grad_norm": 23.375,
+      "learning_rate": 6.6039932196830524e-09,
+      "loss": 0.8108,
+      "step": 7960
+    },
+    {
+      "epoch": 1.9355701434476051,
+      "grad_norm": 25.0,
+      "learning_rate": 6.554475828363799e-09,
+      "loss": 1.2805,
+      "step": 7961
+    },
+    {
+      "epoch": 1.9358132749817651,
+      "grad_norm": 17.125,
+      "learning_rate": 6.505144291453352e-09,
+      "loss": 0.4351,
+      "step": 7962
+    },
+    {
+      "epoch": 1.9360564065159251,
+      "grad_norm": 14.5,
+      "learning_rate": 6.455998616324843e-09,
+      "loss": 0.3876,
+      "step": 7963
+    },
+    {
+      "epoch": 1.9362995380500851,
+      "grad_norm": 20.5,
+      "learning_rate": 6.407038810324062e-09,
+      "loss": 1.4158,
+      "step": 7964
+    },
+    {
+      "epoch": 1.9365426695842451,
+      "grad_norm": 17.25,
+      "learning_rate": 6.358264880769044e-09,
+      "loss": 0.627,
+      "step": 7965
+    },
+    {
+      "epoch": 1.936785801118405,
+      "grad_norm": 26.0,
+      "learning_rate": 6.309676834949791e-09,
+      "loss": 0.9844,
+      "step": 7966
+    },
+    {
+      "epoch": 1.9370289326525651,
+      "grad_norm": 28.5,
+      "learning_rate": 6.261274680128549e-09,
+      "loss": 0.914,
+      "step": 7967
+    },
+    {
+      "epoch": 1.937272064186725,
+      "grad_norm": 19.25,
+      "learning_rate": 6.213058423539809e-09,
+      "loss": 0.5061,
+      "step": 7968
+    },
+    {
+      "epoch": 1.9375151957208852,
+      "grad_norm": 28.0,
+      "learning_rate": 6.165028072390583e-09,
+      "loss": 1.0491,
+      "step": 7969
+    },
+    {
+      "epoch": 1.937758327255045,
+      "grad_norm": 26.375,
+      "learning_rate": 6.117183633859714e-09,
+      "loss": 0.7604,
+      "step": 7970
+    },
+    {
+      "epoch": 1.938001458789205,
+      "grad_norm": 16.75,
+      "learning_rate": 6.0695251150982844e-09,
+      "loss": 0.6431,
+      "step": 7971
+    },
+    {
+      "epoch": 1.938244590323365,
+      "grad_norm": 15.1875,
+      "learning_rate": 6.0220525232297655e-09,
+      "loss": 0.4362,
+      "step": 7972
+    },
+    {
+      "epoch": 1.938487721857525,
+      "grad_norm": 15.6875,
+      "learning_rate": 5.97476586534973e-09,
+      "loss": 0.545,
+      "step": 7973
+    },
+    {
+      "epoch": 1.938730853391685,
+      "grad_norm": 22.0,
+      "learning_rate": 5.927665148525858e-09,
+      "loss": 0.6529,
+      "step": 7974
+    },
+    {
+      "epoch": 1.9389739849258447,
+      "grad_norm": 19.125,
+      "learning_rate": 5.880750379798489e-09,
+      "loss": 0.6962,
+      "step": 7975
+    },
+    {
+      "epoch": 1.939217116460005,
+      "grad_norm": 16.625,
+      "learning_rate": 5.834021566179515e-09,
+      "loss": 0.7686,
+      "step": 7976
+    },
+    {
+      "epoch": 1.9394602479941647,
+      "grad_norm": 16.5,
+      "learning_rate": 5.787478714653627e-09,
+      "loss": 0.454,
+      "step": 7977
+    },
+    {
+      "epoch": 1.939703379528325,
+      "grad_norm": 24.5,
+      "learning_rate": 5.741121832177343e-09,
+      "loss": 0.7541,
+      "step": 7978
+    },
+    {
+      "epoch": 1.9399465110624847,
+      "grad_norm": 20.625,
+      "learning_rate": 5.694950925679427e-09,
+      "loss": 0.5118,
+      "step": 7979
+    },
+    {
+      "epoch": 1.940189642596645,
+      "grad_norm": 19.5,
+      "learning_rate": 5.648966002061024e-09,
+      "loss": 0.6334,
+      "step": 7980
+    },
+    {
+      "epoch": 1.9404327741308047,
+      "grad_norm": 20.375,
+      "learning_rate": 5.603167068195386e-09,
+      "loss": 0.7428,
+      "step": 7981
+    },
+    {
+      "epoch": 1.9406759056649647,
+      "grad_norm": 16.125,
+      "learning_rate": 5.557554130927872e-09,
+      "loss": 0.3143,
+      "step": 7982
+    },
+    {
+      "epoch": 1.9409190371991247,
+      "grad_norm": 18.125,
+      "learning_rate": 5.512127197076084e-09,
+      "loss": 0.5427,
+      "step": 7983
+    },
+    {
+      "epoch": 1.9411621687332847,
+      "grad_norm": 19.125,
+      "learning_rate": 5.466886273430005e-09,
+      "loss": 0.6041,
+      "step": 7984
+    },
+    {
+      "epoch": 1.9414053002674447,
+      "grad_norm": 21.375,
+      "learning_rate": 5.421831366751451e-09,
+      "loss": 0.8419,
+      "step": 7985
+    },
+    {
+      "epoch": 1.9416484318016045,
+      "grad_norm": 18.125,
+      "learning_rate": 5.376962483775033e-09,
+      "loss": 0.8757,
+      "step": 7986
+    },
+    {
+      "epoch": 1.9418915633357647,
+      "grad_norm": 18.375,
+      "learning_rate": 5.3322796312069156e-09,
+      "loss": 0.4442,
+      "step": 7987
+    },
+    {
+      "epoch": 1.9421346948699245,
+      "grad_norm": 24.0,
+      "learning_rate": 5.287782815725645e-09,
+      "loss": 1.2253,
+      "step": 7988
+    },
+    {
+      "epoch": 1.9423778264040847,
+      "grad_norm": 22.875,
+      "learning_rate": 5.243472043982289e-09,
+      "loss": 1.0615,
+      "step": 7989
+    },
+    {
+      "epoch": 1.9426209579382445,
+      "grad_norm": 22.875,
+      "learning_rate": 5.199347322599607e-09,
+      "loss": 0.5945,
+      "step": 7990
+    },
+    {
+      "epoch": 1.9428640894724045,
+      "grad_norm": 24.875,
+      "learning_rate": 5.155408658173017e-09,
+      "loss": 0.7143,
+      "step": 7991
+    },
+    {
+      "epoch": 1.9431072210065645,
+      "grad_norm": 17.5,
+      "learning_rate": 5.1116560572696264e-09,
+      "loss": 0.6492,
+      "step": 7992
+    },
+    {
+      "epoch": 1.9433503525407245,
+      "grad_norm": 19.5,
+      "learning_rate": 5.068089526429482e-09,
+      "loss": 0.5401,
+      "step": 7993
+    },
+    {
+      "epoch": 1.9435934840748845,
+      "grad_norm": 21.5,
+      "learning_rate": 5.024709072163903e-09,
+      "loss": 0.5217,
+      "step": 7994
+    },
+    {
+      "epoch": 1.9438366156090445,
+      "grad_norm": 21.75,
+      "learning_rate": 4.981514700957008e-09,
+      "loss": 0.7356,
+      "step": 7995
+    },
+    {
+      "epoch": 1.9440797471432045,
+      "grad_norm": 23.25,
+      "learning_rate": 4.938506419265021e-09,
+      "loss": 0.6327,
+      "step": 7996
+    },
+    {
+      "epoch": 1.9443228786773643,
+      "grad_norm": 15.5,
+      "learning_rate": 4.895684233516274e-09,
+      "loss": 0.7631,
+      "step": 7997
+    },
+    {
+      "epoch": 1.9445660102115245,
+      "grad_norm": 19.125,
+      "learning_rate": 4.8530481501110615e-09,
+      "loss": 0.596,
+      "step": 7998
+    },
+    {
+      "epoch": 1.9448091417456843,
+      "grad_norm": 20.25,
+      "learning_rate": 4.810598175422204e-09,
+      "loss": 0.7739,
+      "step": 7999
+    },
+    {
+      "epoch": 1.9450522732798445,
+      "grad_norm": 18.875,
+      "learning_rate": 4.768334315794765e-09,
+      "loss": 0.8091,
+      "step": 8000
+    },
+    {
+      "epoch": 1.9452954048140043,
+      "grad_norm": 19.125,
+      "learning_rate": 4.726256577545635e-09,
+      "loss": 0.6666,
+      "step": 8001
+    },
+    {
+      "epoch": 1.9455385363481643,
+      "grad_norm": 17.875,
+      "learning_rate": 4.68436496696395e-09,
+      "loss": 0.6151,
+      "step": 8002
+    },
+    {
+      "epoch": 1.9457816678823243,
+      "grad_norm": 19.875,
+      "learning_rate": 4.642659490311508e-09,
+      "loss": 0.7217,
+      "step": 8003
+    },
+    {
+      "epoch": 1.9460247994164843,
+      "grad_norm": 18.75,
+      "learning_rate": 4.601140153821515e-09,
+      "loss": 0.506,
+      "step": 8004
+    },
+    {
+      "epoch": 1.9462679309506443,
+      "grad_norm": 25.25,
+      "learning_rate": 4.559806963699842e-09,
+      "loss": 0.9511,
+      "step": 8005
+    },
+    {
+      "epoch": 1.9465110624848043,
+      "grad_norm": 18.25,
+      "learning_rate": 4.518659926124602e-09,
+      "loss": 0.584,
+      "step": 8006
+    },
+    {
+      "epoch": 1.9467541940189643,
+      "grad_norm": 31.125,
+      "learning_rate": 4.477699047245876e-09,
+      "loss": 1.1278,
+      "step": 8007
+    },
+    {
+      "epoch": 1.946997325553124,
+      "grad_norm": 18.5,
+      "learning_rate": 4.436924333185849e-09,
+      "loss": 0.4784,
+      "step": 8008
+    },
+    {
+      "epoch": 1.9472404570872843,
+      "grad_norm": 18.375,
+      "learning_rate": 4.3963357900390915e-09,
+      "loss": 0.2685,
+      "step": 8009
+    },
+    {
+      "epoch": 1.947483588621444,
+      "grad_norm": 36.25,
+      "learning_rate": 4.355933423872416e-09,
+      "loss": 0.7514,
+      "step": 8010
+    },
+    {
+      "epoch": 1.9477267201556043,
+      "grad_norm": 28.0,
+      "learning_rate": 4.315717240724465e-09,
+      "loss": 0.5849,
+      "step": 8011
+    },
+    {
+      "epoch": 1.947969851689764,
+      "grad_norm": 23.625,
+      "learning_rate": 4.275687246606125e-09,
+      "loss": 0.7061,
+      "step": 8012
+    },
+    {
+      "epoch": 1.948212983223924,
+      "grad_norm": 20.25,
+      "learning_rate": 4.235843447500942e-09,
+      "loss": 0.5486,
+      "step": 8013
+    },
+    {
+      "epoch": 1.9484561147580841,
+      "grad_norm": 19.0,
+      "learning_rate": 4.196185849363876e-09,
+      "loss": 0.6862,
+      "step": 8014
+    },
+    {
+      "epoch": 1.9486992462922441,
+      "grad_norm": 20.625,
+      "learning_rate": 4.156714458122685e-09,
+      "loss": 0.8226,
+      "step": 8015
+    },
+    {
+      "epoch": 1.9489423778264041,
+      "grad_norm": 20.875,
+      "learning_rate": 4.117429279676954e-09,
+      "loss": 0.6409,
+      "step": 8016
+    },
+    {
+      "epoch": 1.9491855093605641,
+      "grad_norm": 18.375,
+      "learning_rate": 4.078330319898655e-09,
+      "loss": 0.5071,
+      "step": 8017
+    },
+    {
+      "epoch": 1.9494286408947241,
+      "grad_norm": 21.5,
+      "learning_rate": 4.039417584631583e-09,
+      "loss": 0.6227,
+      "step": 8018
+    },
+    {
+      "epoch": 1.949671772428884,
+      "grad_norm": 25.375,
+      "learning_rate": 4.00069107969206e-09,
+      "loss": 1.1477,
+      "step": 8019
+    },
+    {
+      "epoch": 1.9499149039630441,
+      "grad_norm": 28.25,
+      "learning_rate": 3.962150810868509e-09,
+      "loss": 0.9367,
+      "step": 8020
+    },
+    {
+      "epoch": 1.950158035497204,
+      "grad_norm": 28.5,
+      "learning_rate": 3.923796783921185e-09,
+      "loss": 1.0795,
+      "step": 8021
+    },
+    {
+      "epoch": 1.9504011670313641,
+      "grad_norm": 20.0,
+      "learning_rate": 3.8856290045830025e-09,
+      "loss": 0.8159,
+      "step": 8022
+    },
+    {
+      "epoch": 1.950644298565524,
+      "grad_norm": 17.375,
+      "learning_rate": 3.847647478558564e-09,
+      "loss": 0.9348,
+      "step": 8023
+    },
+    {
+      "epoch": 1.950887430099684,
+      "grad_norm": 23.875,
+      "learning_rate": 3.809852211525134e-09,
+      "loss": 1.037,
+      "step": 8024
+    },
+    {
+      "epoch": 1.951130561633844,
+      "grad_norm": 17.25,
+      "learning_rate": 3.772243209131804e-09,
+      "loss": 0.5541,
+      "step": 8025
+    },
+    {
+      "epoch": 1.951373693168004,
+      "grad_norm": 14.875,
+      "learning_rate": 3.734820476999773e-09,
+      "loss": 0.3642,
+      "step": 8026
+    },
+    {
+      "epoch": 1.951616824702164,
+      "grad_norm": 18.875,
+      "learning_rate": 3.6975840207224832e-09,
+      "loss": 0.6307,
+      "step": 8027
+    },
+    {
+      "epoch": 1.9518599562363237,
+      "grad_norm": 21.75,
+      "learning_rate": 3.6605338458657604e-09,
+      "loss": 0.5342,
+      "step": 8028
+    },
+    {
+      "epoch": 1.952103087770484,
+      "grad_norm": 17.5,
+      "learning_rate": 3.6236699579672574e-09,
+      "loss": 0.8729,
+      "step": 8029
+    },
+    {
+      "epoch": 1.9523462193046437,
+      "grad_norm": 19.625,
+      "learning_rate": 3.586992362536873e-09,
+      "loss": 0.9134,
+      "step": 8030
+    },
+    {
+      "epoch": 1.952589350838804,
+      "grad_norm": 20.375,
+      "learning_rate": 3.5505010650567493e-09,
+      "loss": 0.5217,
+      "step": 8031
+    },
+    {
+      "epoch": 1.9528324823729637,
+      "grad_norm": 17.75,
+      "learning_rate": 3.514196070981274e-09,
+      "loss": 0.7421,
+      "step": 8032
+    },
+    {
+      "epoch": 1.953075613907124,
+      "grad_norm": 17.5,
+      "learning_rate": 3.4780773857366613e-09,
+      "loss": 0.468,
+      "step": 8033
+    },
+    {
+      "epoch": 1.9533187454412837,
+      "grad_norm": 20.125,
+      "learning_rate": 3.4421450147216483e-09,
+      "loss": 0.6938,
+      "step": 8034
+    },
+    {
+      "epoch": 1.9535618769754437,
+      "grad_norm": 20.125,
+      "learning_rate": 3.4063989633069395e-09,
+      "loss": 0.7182,
+      "step": 8035
+    },
+    {
+      "epoch": 1.9538050085096037,
+      "grad_norm": 21.875,
+      "learning_rate": 3.370839236835066e-09,
+      "loss": 0.669,
+      "step": 8036
+    },
+    {
+      "epoch": 1.9540481400437637,
+      "grad_norm": 21.5,
+      "learning_rate": 3.335465840621499e-09,
+      "loss": 0.6374,
+      "step": 8037
+    },
+    {
+      "epoch": 1.9542912715779237,
+      "grad_norm": 21.75,
+      "learning_rate": 3.30027877995312e-09,
+      "loss": 0.644,
+      "step": 8038
+    },
+    {
+      "epoch": 1.9545344031120835,
+      "grad_norm": 20.375,
+      "learning_rate": 3.2652780600894717e-09,
+      "loss": 0.7191,
+      "step": 8039
+    },
+    {
+      "epoch": 1.9547775346462437,
+      "grad_norm": 25.0,
+      "learning_rate": 3.230463686261648e-09,
+      "loss": 0.9739,
+      "step": 8040
+    },
+    {
+      "epoch": 1.9550206661804035,
+      "grad_norm": 14.5,
+      "learning_rate": 3.19583566367368e-09,
+      "loss": 0.2626,
+      "step": 8041
+    },
+    {
+      "epoch": 1.9552637977145637,
+      "grad_norm": 18.25,
+      "learning_rate": 3.161393997501011e-09,
+      "loss": 0.4982,
+      "step": 8042
+    },
+    {
+      "epoch": 1.9555069292487235,
+      "grad_norm": 19.5,
+      "learning_rate": 3.127138692891607e-09,
+      "loss": 0.6999,
+      "step": 8043
+    },
+    {
+      "epoch": 1.9557500607828835,
+      "grad_norm": 18.5,
+      "learning_rate": 3.0930697549656776e-09,
+      "loss": 0.7614,
+      "step": 8044
+    },
+    {
+      "epoch": 1.9559931923170435,
+      "grad_norm": 19.375,
+      "learning_rate": 3.0591871888152604e-09,
+      "loss": 0.6758,
+      "step": 8045
+    },
+    {
+      "epoch": 1.9562363238512035,
+      "grad_norm": 18.25,
+      "learning_rate": 3.0254909995047764e-09,
+      "loss": 0.8464,
+      "step": 8046
+    },
+    {
+      "epoch": 1.9564794553853635,
+      "grad_norm": 19.375,
+      "learning_rate": 2.991981192070614e-09,
+      "loss": 0.8545,
+      "step": 8047
+    },
+    {
+      "epoch": 1.9567225869195235,
+      "grad_norm": 19.5,
+      "learning_rate": 2.958657771521406e-09,
+      "loss": 0.4463,
+      "step": 8048
+    },
+    {
+      "epoch": 1.9569657184536835,
+      "grad_norm": 22.25,
+      "learning_rate": 2.925520742838028e-09,
+      "loss": 0.7963,
+      "step": 8049
+    },
+    {
+      "epoch": 1.9572088499878433,
+      "grad_norm": 29.25,
+      "learning_rate": 2.892570110973186e-09,
+      "loss": 0.7685,
+      "step": 8050
+    },
+    {
+      "epoch": 1.9574519815220035,
+      "grad_norm": 25.25,
+      "learning_rate": 2.8598058808521066e-09,
+      "loss": 0.9221,
+      "step": 8051
+    },
+    {
+      "epoch": 1.9576951130561633,
+      "grad_norm": 17.75,
+      "learning_rate": 2.8272280573718445e-09,
+      "loss": 0.7689,
+      "step": 8052
+    },
+    {
+      "epoch": 1.9579382445903235,
+      "grad_norm": 30.5,
+      "learning_rate": 2.7948366454016997e-09,
+      "loss": 0.7574,
+      "step": 8053
+    },
+    {
+      "epoch": 1.9581813761244833,
+      "grad_norm": 18.25,
+      "learning_rate": 2.7626316497833537e-09,
+      "loss": 0.5203,
+      "step": 8054
+    },
+    {
+      "epoch": 1.9584245076586433,
+      "grad_norm": 23.875,
+      "learning_rate": 2.7306130753301796e-09,
+      "loss": 0.3809,
+      "step": 8055
+    },
+    {
+      "epoch": 1.9586676391928033,
+      "grad_norm": 19.75,
+      "learning_rate": 2.6987809268280708e-09,
+      "loss": 0.4265,
+      "step": 8056
+    },
+    {
+      "epoch": 1.9589107707269633,
+      "grad_norm": 25.625,
+      "learning_rate": 2.6671352090346102e-09,
+      "loss": 0.9329,
+      "step": 8057
+    },
+    {
+      "epoch": 1.9591539022611233,
+      "grad_norm": 21.0,
+      "learning_rate": 2.6356759266800424e-09,
+      "loss": 0.6036,
+      "step": 8058
+    },
+    {
+      "epoch": 1.9593970337952833,
+      "grad_norm": 55.75,
+      "learning_rate": 2.6044030844664394e-09,
+      "loss": 0.9563,
+      "step": 8059
+    },
+    {
+      "epoch": 1.9596401653294433,
+      "grad_norm": 24.0,
+      "learning_rate": 2.573316687068117e-09,
+      "loss": 0.7326,
+      "step": 8060
+    },
+    {
+      "epoch": 1.959883296863603,
+      "grad_norm": 19.875,
+      "learning_rate": 2.5424167391313594e-09,
+      "loss": 0.887,
+      "step": 8061
+    },
+    {
+      "epoch": 1.9601264283977633,
+      "grad_norm": 15.125,
+      "learning_rate": 2.5117032452748326e-09,
+      "loss": 0.4569,
+      "step": 8062
+    },
+    {
+      "epoch": 1.960369559931923,
+      "grad_norm": 27.375,
+      "learning_rate": 2.4811762100890313e-09,
+      "loss": 0.7148,
+      "step": 8063
+    },
+    {
+      "epoch": 1.9606126914660833,
+      "grad_norm": 24.25,
+      "learning_rate": 2.4508356381368338e-09,
+      "loss": 0.5914,
+      "step": 8064
+    },
+    {
+      "epoch": 1.960855823000243,
+      "grad_norm": 27.125,
+      "learning_rate": 2.420681533953223e-09,
+      "loss": 0.6904,
+      "step": 8065
+    },
+    {
+      "epoch": 1.961098954534403,
+      "grad_norm": 18.25,
+      "learning_rate": 2.3907139020451498e-09,
+      "loss": 0.7931,
+      "step": 8066
+    },
+    {
+      "epoch": 1.961342086068563,
+      "grad_norm": 19.75,
+      "learning_rate": 2.3609327468919475e-09,
+      "loss": 0.5704,
+      "step": 8067
+    },
+    {
+      "epoch": 1.961585217602723,
+      "grad_norm": 12.625,
+      "learning_rate": 2.3313380729446387e-09,
+      "loss": 0.2747,
+      "step": 8068
+    },
+    {
+      "epoch": 1.961828349136883,
+      "grad_norm": 22.25,
+      "learning_rate": 2.3019298846270465e-09,
+      "loss": 0.6828,
+      "step": 8069
+    },
+    {
+      "epoch": 1.962071480671043,
+      "grad_norm": 23.5,
+      "learning_rate": 2.2727081863344046e-09,
+      "loss": 0.9959,
+      "step": 8070
+    },
+    {
+      "epoch": 1.962314612205203,
+      "grad_norm": 16.25,
+      "learning_rate": 2.243672982434608e-09,
+      "loss": 0.7359,
+      "step": 8071
+    },
+    {
+      "epoch": 1.9625577437393629,
+      "grad_norm": 16.25,
+      "learning_rate": 2.2148242772672414e-09,
+      "loss": 0.6741,
+      "step": 8072
+    },
+    {
+      "epoch": 1.962800875273523,
+      "grad_norm": 20.625,
+      "learning_rate": 2.186162075144549e-09,
+      "loss": 0.6812,
+      "step": 8073
+    },
+    {
+      "epoch": 1.9630440068076829,
+      "grad_norm": 18.25,
+      "learning_rate": 2.157686380350188e-09,
+      "loss": 0.5299,
+      "step": 8074
+    },
+    {
+      "epoch": 1.963287138341843,
+      "grad_norm": 18.75,
+      "learning_rate": 2.1293971971407535e-09,
+      "loss": 0.6591,
+      "step": 8075
+    },
+    {
+      "epoch": 1.9635302698760029,
+      "grad_norm": 25.25,
+      "learning_rate": 2.101294529744391e-09,
+      "loss": 0.5148,
+      "step": 8076
+    },
+    {
+      "epoch": 1.9637734014101629,
+      "grad_norm": 20.75,
+      "learning_rate": 2.0733783823614906e-09,
+      "loss": 0.6525,
+      "step": 8077
+    },
+    {
+      "epoch": 1.9640165329443229,
+      "grad_norm": 19.75,
+      "learning_rate": 2.045648759164548e-09,
+      "loss": 0.4643,
+      "step": 8078
+    },
+    {
+      "epoch": 1.9642596644784829,
+      "grad_norm": 19.5,
+      "learning_rate": 2.018105664298442e-09,
+      "loss": 0.6874,
+      "step": 8079
+    },
+    {
+      "epoch": 1.9645027960126429,
+      "grad_norm": 17.375,
+      "learning_rate": 1.9907491018797407e-09,
+      "loss": 0.2986,
+      "step": 8080
+    },
+    {
+      "epoch": 1.9647459275468027,
+      "grad_norm": 19.75,
+      "learning_rate": 1.963579075997535e-09,
+      "loss": 0.758,
+      "step": 8081
+    },
+    {
+      "epoch": 1.9649890590809629,
+      "grad_norm": 17.875,
+      "learning_rate": 1.936595590712742e-09,
+      "loss": 0.622,
+      "step": 8082
+    },
+    {
+      "epoch": 1.9652321906151227,
+      "grad_norm": 24.125,
+      "learning_rate": 1.9097986500585252e-09,
+      "loss": 0.7895,
+      "step": 8083
+    },
+    {
+      "epoch": 1.9654753221492829,
+      "grad_norm": 24.625,
+      "learning_rate": 1.883188258040153e-09,
+      "loss": 0.9597,
+      "step": 8084
+    },
+    {
+      "epoch": 1.9657184536834427,
+      "grad_norm": 24.0,
+      "learning_rate": 1.8567644186349987e-09,
+      "loss": 1.202,
+      "step": 8085
+    },
+    {
+      "epoch": 1.9659615852176027,
+      "grad_norm": 25.375,
+      "learning_rate": 1.8305271357925425e-09,
+      "loss": 1.1381,
+      "step": 8086
+    },
+    {
+      "epoch": 1.9662047167517627,
+      "grad_norm": 16.75,
+      "learning_rate": 1.8044764134346471e-09,
+      "loss": 1.1715,
+      "step": 8087
+    },
+    {
+      "epoch": 1.9664478482859227,
+      "grad_norm": 14.5,
+      "learning_rate": 1.778612255454587e-09,
+      "loss": 0.3874,
+      "step": 8088
+    },
+    {
+      "epoch": 1.9666909798200827,
+      "grad_norm": 22.625,
+      "learning_rate": 1.7529346657185754e-09,
+      "loss": 0.6248,
+      "step": 8089
+    },
+    {
+      "epoch": 1.9669341113542427,
+      "grad_norm": 20.75,
+      "learning_rate": 1.7274436480645141e-09,
+      "loss": 0.5769,
+      "step": 8090
+    },
+    {
+      "epoch": 1.9671772428884027,
+      "grad_norm": 16.375,
+      "learning_rate": 1.7021392063022724e-09,
+      "loss": 0.6754,
+      "step": 8091
+    },
+    {
+      "epoch": 1.9674203744225625,
+      "grad_norm": 26.875,
+      "learning_rate": 1.6770213442142412e-09,
+      "loss": 0.7949,
+      "step": 8092
+    },
+    {
+      "epoch": 1.9676635059567227,
+      "grad_norm": 19.875,
+      "learning_rate": 1.6520900655546401e-09,
+      "loss": 0.63,
+      "step": 8093
+    },
+    {
+      "epoch": 1.9679066374908825,
+      "grad_norm": 23.375,
+      "learning_rate": 1.6273453740499323e-09,
+      "loss": 0.5896,
+      "step": 8094
+    },
+    {
+      "epoch": 1.9681497690250427,
+      "grad_norm": 20.0,
+      "learning_rate": 1.6027872733985483e-09,
+      "loss": 0.4427,
+      "step": 8095
+    },
+    {
+      "epoch": 1.9683929005592025,
+      "grad_norm": 19.75,
+      "learning_rate": 1.578415767271163e-09,
+      "loss": 0.7467,
+      "step": 8096
+    },
+    {
+      "epoch": 1.9686360320933625,
+      "grad_norm": 18.0,
+      "learning_rate": 1.554230859310557e-09,
+      "loss": 0.7658,
+      "step": 8097
+    },
+    {
+      "epoch": 1.9688791636275225,
+      "grad_norm": 18.0,
+      "learning_rate": 1.5302325531316164e-09,
+      "loss": 0.5146,
+      "step": 8098
+    },
+    {
+      "epoch": 1.9691222951616825,
+      "grad_norm": 20.75,
+      "learning_rate": 1.5064208523211942e-09,
+      "loss": 0.6021,
+      "step": 8099
+    },
+    {
+      "epoch": 1.9693654266958425,
+      "grad_norm": 16.5,
+      "learning_rate": 1.482795760438388e-09,
+      "loss": 0.4044,
+      "step": 8100
+    },
+    {
+      "epoch": 1.9696085582300025,
+      "grad_norm": 20.125,
+      "learning_rate": 1.4593572810144008e-09,
+      "loss": 0.607,
+      "step": 8101
+    },
+    {
+      "epoch": 1.9698516897641625,
+      "grad_norm": 17.75,
+      "learning_rate": 1.4361054175524025e-09,
+      "loss": 0.5903,
+      "step": 8102
+    },
+    {
+      "epoch": 1.9700948212983223,
+      "grad_norm": 14.3125,
+      "learning_rate": 1.4130401735280852e-09,
+      "loss": 0.2701,
+      "step": 8103
+    },
+    {
+      "epoch": 1.9703379528324825,
+      "grad_norm": 30.75,
+      "learning_rate": 1.3901615523886914e-09,
+      "loss": 1.1624,
+      "step": 8104
+    },
+    {
+      "epoch": 1.9705810843666423,
+      "grad_norm": 27.0,
+      "learning_rate": 1.3674695575538465e-09,
+      "loss": 0.8752,
+      "step": 8105
+    },
+    {
+      "epoch": 1.9708242159008025,
+      "grad_norm": 16.375,
+      "learning_rate": 1.3449641924152823e-09,
+      "loss": 0.4863,
+      "step": 8106
+    },
+    {
+      "epoch": 1.9710673474349623,
+      "grad_norm": 19.75,
+      "learning_rate": 1.3226454603369744e-09,
+      "loss": 0.7416,
+      "step": 8107
+    },
+    {
+      "epoch": 1.9713104789691223,
+      "grad_norm": 16.0,
+      "learning_rate": 1.3005133646545875e-09,
+      "loss": 0.4618,
+      "step": 8108
+    },
+    {
+      "epoch": 1.9715536105032823,
+      "grad_norm": 28.0,
+      "learning_rate": 1.2785679086763092e-09,
+      "loss": 0.8066,
+      "step": 8109
+    },
+    {
+      "epoch": 1.9717967420374423,
+      "grad_norm": 18.25,
+      "learning_rate": 1.2568090956821543e-09,
+      "loss": 0.5981,
+      "step": 8110
+    },
+    {
+      "epoch": 1.9720398735716023,
+      "grad_norm": 17.375,
+      "learning_rate": 1.2352369289245214e-09,
+      "loss": 0.5006,
+      "step": 8111
+    },
+    {
+      "epoch": 1.9722830051057623,
+      "grad_norm": 15.875,
+      "learning_rate": 1.2138514116276369e-09,
+      "loss": 0.398,
+      "step": 8112
+    },
+    {
+      "epoch": 1.9725261366399223,
+      "grad_norm": 15.125,
+      "learning_rate": 1.1926525469878326e-09,
+      "loss": 0.5582,
+      "step": 8113
+    },
+    {
+      "epoch": 1.972769268174082,
+      "grad_norm": 23.625,
+      "learning_rate": 1.171640338173824e-09,
+      "loss": 0.8987,
+      "step": 8114
+    },
+    {
+      "epoch": 1.9730123997082423,
+      "grad_norm": 18.5,
+      "learning_rate": 1.1508147883261544e-09,
+      "loss": 0.5504,
+      "step": 8115
+    },
+    {
+      "epoch": 1.973255531242402,
+      "grad_norm": 17.0,
+      "learning_rate": 1.1301759005576118e-09,
+      "loss": 0.6745,
+      "step": 8116
+    },
+    {
+      "epoch": 1.9734986627765623,
+      "grad_norm": 18.5,
+      "learning_rate": 1.1097236779530895e-09,
+      "loss": 0.7148,
+      "step": 8117
+    },
+    {
+      "epoch": 1.973741794310722,
+      "grad_norm": 23.375,
+      "learning_rate": 1.0894581235693091e-09,
+      "loss": 1.4508,
+      "step": 8118
+    },
+    {
+      "epoch": 1.973984925844882,
+      "grad_norm": 23.25,
+      "learning_rate": 1.0693792404355141e-09,
+      "loss": 1.0882,
+      "step": 8119
+    },
+    {
+      "epoch": 1.974228057379042,
+      "grad_norm": 23.125,
+      "learning_rate": 1.049487031552776e-09,
+      "loss": 1.2678,
+      "step": 8120
+    },
+    {
+      "epoch": 1.974471188913202,
+      "grad_norm": 19.125,
+      "learning_rate": 1.0297814998942723e-09,
+      "loss": 0.6853,
+      "step": 8121
+    },
+    {
+      "epoch": 1.974714320447362,
+      "grad_norm": 16.375,
+      "learning_rate": 1.0102626484054245e-09,
+      "loss": 0.7197,
+      "step": 8122
+    },
+    {
+      "epoch": 1.9749574519815218,
+      "grad_norm": 14.6875,
+      "learning_rate": 9.909304800036213e-10,
+      "loss": 0.3409,
+      "step": 8123
+    },
+    {
+      "epoch": 1.975200583515682,
+      "grad_norm": 21.0,
+      "learning_rate": 9.717849975783567e-10,
+      "loss": 0.593,
+      "step": 8124
+    },
+    {
+      "epoch": 1.9754437150498418,
+      "grad_norm": 18.125,
+      "learning_rate": 9.528262039912306e-10,
+      "loss": 0.5253,
+      "step": 8125
+    },
+    {
+      "epoch": 1.975686846584002,
+      "grad_norm": 22.5,
+      "learning_rate": 9.340541020762262e-10,
+      "loss": 0.9469,
+      "step": 8126
+    },
+    {
+      "epoch": 1.9759299781181618,
+      "grad_norm": 25.875,
+      "learning_rate": 9.154686946387381e-10,
+      "loss": 0.716,
+      "step": 8127
+    },
+    {
+      "epoch": 1.976173109652322,
+      "grad_norm": 16.5,
+      "learning_rate": 8.970699844569608e-10,
+      "loss": 0.4917,
+      "step": 8128
+    },
+    {
+      "epoch": 1.9764162411864818,
+      "grad_norm": 16.25,
+      "learning_rate": 8.788579742809167e-10,
+      "loss": 0.4885,
+      "step": 8129
+    },
+    {
+      "epoch": 1.9766593727206418,
+      "grad_norm": 20.0,
+      "learning_rate": 8.608326668325951e-10,
+      "loss": 0.7131,
+      "step": 8130
+    },
+    {
+      "epoch": 1.9769025042548019,
+      "grad_norm": 18.125,
+      "learning_rate": 8.429940648060908e-10,
+      "loss": 0.4055,
+      "step": 8131
+    },
+    {
+      "epoch": 1.9771456357889619,
+      "grad_norm": 20.5,
+      "learning_rate": 8.253421708678822e-10,
+      "loss": 0.533,
+      "step": 8132
+    },
+    {
+      "epoch": 1.9773887673231219,
+      "grad_norm": 19.25,
+      "learning_rate": 8.078769876562753e-10,
+      "loss": 0.5524,
+      "step": 8133
+    },
+    {
+      "epoch": 1.9776318988572816,
+      "grad_norm": 17.875,
+      "learning_rate": 7.90598517781821e-10,
+      "loss": 0.6555,
+      "step": 8134
+    },
+    {
+      "epoch": 1.9778750303914419,
+      "grad_norm": 16.875,
+      "learning_rate": 7.735067638268978e-10,
+      "loss": 0.6905,
+      "step": 8135
+    },
+    {
+      "epoch": 1.9781181619256016,
+      "grad_norm": 19.375,
+      "learning_rate": 7.566017283462679e-10,
+      "loss": 0.9201,
+      "step": 8136
+    },
+    {
+      "epoch": 1.9783612934597619,
+      "grad_norm": 25.625,
+      "learning_rate": 7.398834138667987e-10,
+      "loss": 0.986,
+      "step": 8137
+    },
+    {
+      "epoch": 1.9786044249939216,
+      "grad_norm": 14.1875,
+      "learning_rate": 7.233518228871861e-10,
+      "loss": 0.2434,
+      "step": 8138
+    },
+    {
+      "epoch": 1.9788475565280816,
+      "grad_norm": 17.5,
+      "learning_rate": 7.070069578783701e-10,
+      "loss": 0.5198,
+      "step": 8139
+    },
+    {
+      "epoch": 1.9790906880622416,
+      "grad_norm": 20.5,
+      "learning_rate": 6.908488212833963e-10,
+      "loss": 0.708,
+      "step": 8140
+    },
+    {
+      "epoch": 1.9793338195964016,
+      "grad_norm": 17.75,
+      "learning_rate": 6.748774155174165e-10,
+      "loss": 0.4918,
+      "step": 8141
+    },
+    {
+      "epoch": 1.9795769511305616,
+      "grad_norm": 16.25,
+      "learning_rate": 6.590927429675487e-10,
+      "loss": 0.6461,
+      "step": 8142
+    },
+    {
+      "epoch": 1.9798200826647216,
+      "grad_norm": 17.875,
+      "learning_rate": 6.434948059931557e-10,
+      "loss": 0.5496,
+      "step": 8143
+    },
+    {
+      "epoch": 1.9800632141988816,
+      "grad_norm": 19.125,
+      "learning_rate": 6.280836069255669e-10,
+      "loss": 0.51,
+      "step": 8144
+    },
+    {
+      "epoch": 1.9803063457330414,
+      "grad_norm": 17.875,
+      "learning_rate": 6.128591480683566e-10,
+      "loss": 0.5675,
+      "step": 8145
+    },
+    {
+      "epoch": 1.9805494772672017,
+      "grad_norm": 23.625,
+      "learning_rate": 5.978214316969267e-10,
+      "loss": 0.6228,
+      "step": 8146
+    },
+    {
+      "epoch": 1.9807926088013614,
+      "grad_norm": 25.875,
+      "learning_rate": 5.829704600590624e-10,
+      "loss": 0.7185,
+      "step": 8147
+    },
+    {
+      "epoch": 1.9810357403355217,
+      "grad_norm": 20.25,
+      "learning_rate": 5.683062353745161e-10,
+      "loss": 1.0372,
+      "step": 8148
+    },
+    {
+      "epoch": 1.9812788718696814,
+      "grad_norm": 24.375,
+      "learning_rate": 5.53828759834868e-10,
+      "loss": 0.6557,
+      "step": 8149
+    },
+    {
+      "epoch": 1.9815220034038414,
+      "grad_norm": 20.25,
+      "learning_rate": 5.39538035604359e-10,
+      "loss": 0.9553,
+      "step": 8150
+    },
+    {
+      "epoch": 1.9817651349380014,
+      "grad_norm": 19.75,
+      "learning_rate": 5.254340648186418e-10,
+      "loss": 0.4918,
+      "step": 8151
+    },
+    {
+      "epoch": 1.9820082664721614,
+      "grad_norm": 17.625,
+      "learning_rate": 5.115168495861689e-10,
+      "loss": 0.5586,
+      "step": 8152
+    },
+    {
+      "epoch": 1.9822513980063214,
+      "grad_norm": 17.0,
+      "learning_rate": 4.977863919868042e-10,
+      "loss": 0.6619,
+      "step": 8153
+    },
+    {
+      "epoch": 1.9824945295404814,
+      "grad_norm": 19.75,
+      "learning_rate": 4.842426940729339e-10,
+      "loss": 0.5333,
+      "step": 8154
+    },
+    {
+      "epoch": 1.9827376610746414,
+      "grad_norm": 37.0,
+      "learning_rate": 4.70885757868772e-10,
+      "loss": 1.0077,
+      "step": 8155
+    },
+    {
+      "epoch": 1.9829807926088012,
+      "grad_norm": 28.5,
+      "learning_rate": 4.577155853709159e-10,
+      "loss": 0.7056,
+      "step": 8156
+    },
+    {
+      "epoch": 1.9832239241429614,
+      "grad_norm": 17.375,
+      "learning_rate": 4.4473217854792974e-10,
+      "loss": 0.8223,
+      "step": 8157
+    },
+    {
+      "epoch": 1.9834670556771212,
+      "grad_norm": 16.375,
+      "learning_rate": 4.3193553934020584e-10,
+      "loss": 0.3124,
+      "step": 8158
+    },
+    {
+      "epoch": 1.9837101872112815,
+      "grad_norm": 21.0,
+      "learning_rate": 4.193256696605197e-10,
+      "loss": 0.6194,
+      "step": 8159
+    },
+    {
+      "epoch": 1.9839533187454412,
+      "grad_norm": 22.75,
+      "learning_rate": 4.069025713934749e-10,
+      "loss": 0.6732,
+      "step": 8160
+    },
+    {
+      "epoch": 1.9841964502796012,
+      "grad_norm": 17.375,
+      "learning_rate": 3.94666246396197e-10,
+      "loss": 0.3517,
+      "step": 8161
+    },
+    {
+      "epoch": 1.9844395818137612,
+      "grad_norm": 19.75,
+      "learning_rate": 3.826166964975009e-10,
+      "loss": 0.8733,
+      "step": 8162
+    },
+    {
+      "epoch": 1.9846827133479212,
+      "grad_norm": 23.125,
+      "learning_rate": 3.707539234983071e-10,
+      "loss": 0.7119,
+      "step": 8163
+    },
+    {
+      "epoch": 1.9849258448820812,
+      "grad_norm": 20.5,
+      "learning_rate": 3.5907792917178055e-10,
+      "loss": 0.7217,
+      "step": 8164
+    },
+    {
+      "epoch": 1.9851689764162412,
+      "grad_norm": 16.0,
+      "learning_rate": 3.475887152631918e-10,
+      "loss": 0.6005,
+      "step": 8165
+    },
+    {
+      "epoch": 1.9854121079504012,
+      "grad_norm": 20.25,
+      "learning_rate": 3.362862834896397e-10,
+      "loss": 0.6347,
+      "step": 8166
+    },
+    {
+      "epoch": 1.985655239484561,
+      "grad_norm": 22.875,
+      "learning_rate": 3.251706355404671e-10,
+      "loss": 0.5622,
+      "step": 8167
+    },
+    {
+      "epoch": 1.9858983710187212,
+      "grad_norm": 17.375,
+      "learning_rate": 3.1424177307726177e-10,
+      "loss": 0.4076,
+      "step": 8168
+    },
+    {
+      "epoch": 1.986141502552881,
+      "grad_norm": 16.875,
+      "learning_rate": 3.0349969773343924e-10,
+      "loss": 0.4844,
+      "step": 8169
+    },
+    {
+      "epoch": 1.9863846340870412,
+      "grad_norm": 22.375,
+      "learning_rate": 2.9294441111465956e-10,
+      "loss": 0.5908,
+      "step": 8170
+    },
+    {
+      "epoch": 1.986627765621201,
+      "grad_norm": 20.5,
+      "learning_rate": 2.8257591479841083e-10,
+      "loss": 1.0264,
+      "step": 8171
+    },
+    {
+      "epoch": 1.986870897155361,
+      "grad_norm": 19.625,
+      "learning_rate": 2.7239421033470324e-10,
+      "loss": 0.6554,
+      "step": 8172
+    },
+    {
+      "epoch": 1.987114028689521,
+      "grad_norm": 24.25,
+      "learning_rate": 2.623992992450974e-10,
+      "loss": 0.9349,
+      "step": 8173
+    },
+    {
+      "epoch": 1.987357160223681,
+      "grad_norm": 25.25,
+      "learning_rate": 2.52591183023676e-10,
+      "loss": 0.8927,
+      "step": 8174
+    },
+    {
+      "epoch": 1.987600291757841,
+      "grad_norm": 22.0,
+      "learning_rate": 2.4296986313634973e-10,
+      "loss": 0.7329,
+      "step": 8175
+    },
+    {
+      "epoch": 1.9878434232920008,
+      "grad_norm": 19.0,
+      "learning_rate": 2.335353410212737e-10,
+      "loss": 0.4344,
+      "step": 8176
+    },
+    {
+      "epoch": 1.988086554826161,
+      "grad_norm": 24.5,
+      "learning_rate": 2.2428761808857002e-10,
+      "loss": 1.004,
+      "step": 8177
+    },
+    {
+      "epoch": 1.9883296863603208,
+      "grad_norm": 16.875,
+      "learning_rate": 2.1522669572032752e-10,
+      "loss": 0.6985,
+      "step": 8178
+    },
+    {
+      "epoch": 1.988572817894481,
+      "grad_norm": 18.625,
+      "learning_rate": 2.0635257527115705e-10,
+      "loss": 0.4614,
+      "step": 8179
+    },
+    {
+      "epoch": 1.9888159494286408,
+      "grad_norm": 23.125,
+      "learning_rate": 1.9766525806708126e-10,
+      "loss": 0.9136,
+      "step": 8180
+    },
+    {
+      "epoch": 1.989059080962801,
+      "grad_norm": 18.125,
+      "learning_rate": 1.8916474540692231e-10,
+      "loss": 1.0467,
+      "step": 8181
+    },
+    {
+      "epoch": 1.9893022124969608,
+      "grad_norm": 19.375,
+      "learning_rate": 1.808510385610529e-10,
+      "loss": 0.9221,
+      "step": 8182
+    },
+    {
+      "epoch": 1.9895453440311208,
+      "grad_norm": 17.875,
+      "learning_rate": 1.7272413877209015e-10,
+      "loss": 0.7192,
+      "step": 8183
+    },
+    {
+      "epoch": 1.9897884755652808,
+      "grad_norm": 26.75,
+      "learning_rate": 1.6478404725475683e-10,
+      "loss": 1.3566,
+      "step": 8184
+    },
+    {
+      "epoch": 1.9900316070994408,
+      "grad_norm": 25.125,
+      "learning_rate": 1.5703076519588135e-10,
+      "loss": 0.7287,
+      "step": 8185
+    },
+    {
+      "epoch": 1.9902747386336008,
+      "grad_norm": 23.625,
+      "learning_rate": 1.4946429375425896e-10,
+      "loss": 0.665,
+      "step": 8186
+    },
+    {
+      "epoch": 1.9905178701677606,
+      "grad_norm": 15.0625,
+      "learning_rate": 1.4208463406092941e-10,
+      "loss": 0.3656,
+      "step": 8187
+    },
+    {
+      "epoch": 1.9907610017019208,
+      "grad_norm": 25.125,
+      "learning_rate": 1.3489178721876052e-10,
+      "loss": 0.8067,
+      "step": 8188
+    },
+    {
+      "epoch": 1.9910041332360806,
+      "grad_norm": 21.125,
+      "learning_rate": 1.2788575430300321e-10,
+      "loss": 0.8212,
+      "step": 8189
+    },
+    {
+      "epoch": 1.9912472647702408,
+      "grad_norm": 20.75,
+      "learning_rate": 1.210665363607366e-10,
+      "loss": 0.6031,
+      "step": 8190
+    },
+    {
+      "epoch": 1.9914903963044006,
+      "grad_norm": 23.125,
+      "learning_rate": 1.1443413441114548e-10,
+      "loss": 0.8211,
+      "step": 8191
+    },
+    {
+      "epoch": 1.9917335278385606,
+      "grad_norm": 20.125,
+      "learning_rate": 1.0798854944579774e-10,
+      "loss": 0.855,
+      "step": 8192
+    },
+    {
+      "epoch": 1.9919766593727206,
+      "grad_norm": 20.25,
+      "learning_rate": 1.0172978242781184e-10,
+      "loss": 0.5624,
+      "step": 8193
+    },
+    {
+      "epoch": 1.9922197909068806,
+      "grad_norm": 24.0,
+      "learning_rate": 9.565783429282827e-11,
+      "loss": 1.0241,
+      "step": 8194
+    },
+    {
+      "epoch": 1.9924629224410406,
+      "grad_norm": 25.25,
+      "learning_rate": 8.977270594845433e-11,
+      "loss": 0.5422,
+      "step": 8195
+    },
+    {
+      "epoch": 1.9927060539752006,
+      "grad_norm": 28.5,
+      "learning_rate": 8.407439827412545e-11,
+      "loss": 0.5915,
+      "step": 8196
+    },
+    {
+      "epoch": 1.9929491855093606,
+      "grad_norm": 15.125,
+      "learning_rate": 7.856291212179901e-11,
+      "loss": 0.3425,
+      "step": 8197
+    },
+    {
+      "epoch": 1.9931923170435204,
+      "grad_norm": 21.25,
+      "learning_rate": 7.323824831512171e-11,
+      "loss": 0.506,
+      "step": 8198
+    },
+    {
+      "epoch": 1.9934354485776806,
+      "grad_norm": 64.0,
+      "learning_rate": 6.810040764984593e-11,
+      "loss": 0.915,
+      "step": 8199
+    },
+    {
+      "epoch": 1.9936785801118404,
+      "grad_norm": 19.625,
+      "learning_rate": 6.314939089410721e-11,
+      "loss": 0.445,
+      "step": 8200
+    },
+    {
+      "epoch": 1.9939217116460006,
+      "grad_norm": 18.25,
+      "learning_rate": 5.838519878786919e-11,
+      "loss": 0.4293,
+      "step": 8201
+    },
+    {
+      "epoch": 1.9941648431801604,
+      "grad_norm": 17.625,
+      "learning_rate": 5.380783204320117e-11,
+      "loss": 0.8994,
+      "step": 8202
+    },
+    {
+      "epoch": 1.9944079747143204,
+      "grad_norm": 24.75,
+      "learning_rate": 4.9417291344278086e-11,
+      "loss": 0.6165,
+      "step": 8203
+    },
+    {
+      "epoch": 1.9946511062484804,
+      "grad_norm": 22.125,
+      "learning_rate": 4.5213577347241744e-11,
+      "loss": 0.7128,
+      "step": 8204
+    },
+    {
+      "epoch": 1.9948942377826404,
+      "grad_norm": 16.625,
+      "learning_rate": 4.119669068061716e-11,
+      "loss": 0.3796,
+      "step": 8205
+    },
+    {
+      "epoch": 1.9951373693168004,
+      "grad_norm": 20.5,
+      "learning_rate": 3.7366631944618646e-11,
+      "loss": 0.8241,
+      "step": 8206
+    },
+    {
+      "epoch": 1.9953805008509604,
+      "grad_norm": 23.25,
+      "learning_rate": 3.372340171184374e-11,
+      "loss": 0.7091,
+      "step": 8207
+    },
+    {
+      "epoch": 1.9956236323851204,
+      "grad_norm": 24.125,
+      "learning_rate": 3.026700052671805e-11,
+      "loss": 1.0373,
+      "step": 8208
+    },
+    {
+      "epoch": 1.9958667639192802,
+      "grad_norm": 18.0,
+      "learning_rate": 2.6997428905911617e-11,
+      "loss": 0.6254,
+      "step": 8209
+    },
+    {
+      "epoch": 1.9961098954534404,
+      "grad_norm": 20.0,
+      "learning_rate": 2.391468733806135e-11,
+      "loss": 0.6479,
+      "step": 8210
+    },
+    {
+      "epoch": 1.9963530269876002,
+      "grad_norm": 18.75,
+      "learning_rate": 2.101877628418736e-11,
+      "loss": 0.8676,
+      "step": 8211
+    },
+    {
+      "epoch": 1.9965961585217604,
+      "grad_norm": 16.25,
+      "learning_rate": 1.8309696176721514e-11,
+      "loss": 0.4131,
+      "step": 8212
+    },
+    {
+      "epoch": 1.9968392900559202,
+      "grad_norm": 21.25,
+      "learning_rate": 1.578744742089522e-11,
+      "loss": 0.5246,
+      "step": 8213
+    },
+    {
+      "epoch": 1.9970824215900802,
+      "grad_norm": 21.0,
+      "learning_rate": 1.345203039362919e-11,
+      "loss": 0.7029,
+      "step": 8214
+    },
+    {
+      "epoch": 1.9973255531242402,
+      "grad_norm": 19.875,
+      "learning_rate": 1.130344544394979e-11,
+      "loss": 0.686,
+      "step": 8215
+    },
+    {
+      "epoch": 1.9975686846584002,
+      "grad_norm": 30.625,
+      "learning_rate": 9.341692892989029e-12,
+      "loss": 1.0333,
+      "step": 8216
+    },
+    {
+      "epoch": 1.9978118161925602,
+      "grad_norm": 17.375,
+      "learning_rate": 7.566773034123342e-12,
+      "loss": 0.9167,
+      "step": 8217
+    },
+    {
+      "epoch": 1.9980549477267202,
+      "grad_norm": 17.875,
+      "learning_rate": 5.9786861324184765e-12,
+      "loss": 0.67,
+      "step": 8218
+    },
+    {
+      "epoch": 1.9982980792608802,
+      "grad_norm": 18.75,
+      "learning_rate": 4.57743242532338e-12,
+      "loss": 0.4391,
+      "step": 8219
+    },
+    {
+      "epoch": 1.99854121079504,
+      "grad_norm": 16.375,
+      "learning_rate": 3.3630121223926502e-12,
+      "loss": 0.6415,
+      "step": 8220
+    },
+    {
+      "epoch": 1.9987843423292002,
+      "grad_norm": 22.375,
+      "learning_rate": 2.3354254050089733e-12,
+      "loss": 0.7354,
+      "step": 8221
+    },
+    {
+      "epoch": 1.99902747386336,
+      "grad_norm": 20.25,
+      "learning_rate": 1.4946724266606815e-12,
+      "loss": 0.518,
+      "step": 8222
+    },
+    {
+      "epoch": 1.9992706053975202,
+      "grad_norm": 16.25,
+      "learning_rate": 8.407533133580891e-13,
+      "loss": 0.4478,
+      "step": 8223
+    },
+    {
+      "epoch": 1.99951373693168,
+      "grad_norm": 34.75,
+      "learning_rate": 3.7366816252326633e-13,
+      "loss": 1.1643,
+      "step": 8224
+    },
+    {
+      "epoch": 1.99975686846584,
+      "grad_norm": 32.75,
+      "learning_rate": 9.341704410026354e-14,
+      "loss": 0.9797,
+      "step": 8225
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 22.875,
+      "learning_rate": 0.0,
+      "loss": 0.6045,
+      "step": 8226
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 8226,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 2057,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.629779584710083e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}