diff --git "a/checkpoint-1415/trainer_state.json" "b/checkpoint-1415/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1415/trainer_state.json"
@@ -0,0 +1,9938 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 4.99558693733451,
+  "eval_steps": 500,
+  "global_step": 1415,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00353045013239188,
+      "grad_norm": 20.178845075428434,
+      "learning_rate": 1.1764705882352942e-07,
+      "loss": 0.7173,
+      "step": 1
+    },
+    {
+      "epoch": 0.00706090026478376,
+      "grad_norm": 21.408372160532007,
+      "learning_rate": 2.3529411764705883e-07,
+      "loss": 0.8369,
+      "step": 2
+    },
+    {
+      "epoch": 0.01059135039717564,
+      "grad_norm": 22.339887896461786,
+      "learning_rate": 3.529411764705883e-07,
+      "loss": 0.7742,
+      "step": 3
+    },
+    {
+      "epoch": 0.01412180052956752,
+      "grad_norm": 20.114111492682788,
+      "learning_rate": 4.7058823529411767e-07,
+      "loss": 0.6685,
+      "step": 4
+    },
+    {
+      "epoch": 0.0176522506619594,
+      "grad_norm": 19.395831354668825,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.724,
+      "step": 5
+    },
+    {
+      "epoch": 0.02118270079435128,
+      "grad_norm": 20.441308282025314,
+      "learning_rate": 7.058823529411766e-07,
+      "loss": 0.8051,
+      "step": 6
+    },
+    {
+      "epoch": 0.02471315092674316,
+      "grad_norm": 21.439520175660693,
+      "learning_rate": 8.235294117647059e-07,
+      "loss": 0.7853,
+      "step": 7
+    },
+    {
+      "epoch": 0.02824360105913504,
+      "grad_norm": 16.361184069343224,
+      "learning_rate": 9.411764705882353e-07,
+      "loss": 0.6029,
+      "step": 8
+    },
+    {
+      "epoch": 0.03177405119152692,
+      "grad_norm": 16.50966475117204,
+      "learning_rate": 1.0588235294117648e-06,
+      "loss": 0.607,
+      "step": 9
+    },
+    {
+      "epoch": 0.0353045013239188,
+      "grad_norm": 10.733426578588741,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.5416,
+      "step": 10
+    },
+    {
+      "epoch": 0.038834951456310676,
+      "grad_norm": 10.395447257616585,
+      "learning_rate": 1.2941176470588237e-06,
+      "loss": 0.4926,
+      "step": 11
+    },
+    {
+      "epoch": 0.04236540158870256,
+      "grad_norm": 12.789389493634516,
+      "learning_rate": 1.4117647058823531e-06,
+      "loss": 0.5638,
+      "step": 12
+    },
+    {
+      "epoch": 0.04589585172109444,
+      "grad_norm": 8.94856882496586,
+      "learning_rate": 1.5294117647058826e-06,
+      "loss": 0.459,
+      "step": 13
+    },
+    {
+      "epoch": 0.04942630185348632,
+      "grad_norm": 4.421740571415892,
+      "learning_rate": 1.6470588235294118e-06,
+      "loss": 0.322,
+      "step": 14
+    },
+    {
+      "epoch": 0.0529567519858782,
+      "grad_norm": 4.035083034631525,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.3224,
+      "step": 15
+    },
+    {
+      "epoch": 0.05648720211827008,
+      "grad_norm": 4.319827920176901,
+      "learning_rate": 1.8823529411764707e-06,
+      "loss": 0.3079,
+      "step": 16
+    },
+    {
+      "epoch": 0.06001765225066196,
+      "grad_norm": 4.9583914837299,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.3102,
+      "step": 17
+    },
+    {
+      "epoch": 0.06354810238305383,
+      "grad_norm": 4.011276537282153,
+      "learning_rate": 2.1176470588235296e-06,
+      "loss": 0.2933,
+      "step": 18
+    },
+    {
+      "epoch": 0.06707855251544571,
+      "grad_norm": 3.1423743425131607,
+      "learning_rate": 2.2352941176470592e-06,
+      "loss": 0.2419,
+      "step": 19
+    },
+    {
+      "epoch": 0.0706090026478376,
+      "grad_norm": 3.526182340817241,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.2649,
+      "step": 20
+    },
+    {
+      "epoch": 0.07413945278022947,
+      "grad_norm": 2.1822830906532205,
+      "learning_rate": 2.470588235294118e-06,
+      "loss": 0.241,
+      "step": 21
+    },
+    {
+      "epoch": 0.07766990291262135,
+      "grad_norm": 3.01235390581002,
+      "learning_rate": 2.5882352941176473e-06,
+      "loss": 0.2352,
+      "step": 22
+    },
+    {
+      "epoch": 0.08120035304501325,
+      "grad_norm": 3.34743138297086,
+      "learning_rate": 2.7058823529411766e-06,
+      "loss": 0.2196,
+      "step": 23
+    },
+    {
+      "epoch": 0.08473080317740513,
+      "grad_norm": 2.453669230978404,
+      "learning_rate": 2.8235294117647062e-06,
+      "loss": 0.2577,
+      "step": 24
+    },
+    {
+      "epoch": 0.088261253309797,
+      "grad_norm": 2.0737233723937765,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.2099,
+      "step": 25
+    },
+    {
+      "epoch": 0.09179170344218888,
+      "grad_norm": 2.3481110465941653,
+      "learning_rate": 3.058823529411765e-06,
+      "loss": 0.2302,
+      "step": 26
+    },
+    {
+      "epoch": 0.09532215357458076,
+      "grad_norm": 2.4508501080365086,
+      "learning_rate": 3.1764705882352943e-06,
+      "loss": 0.2183,
+      "step": 27
+    },
+    {
+      "epoch": 0.09885260370697264,
+      "grad_norm": 1.9310659307898554,
+      "learning_rate": 3.2941176470588236e-06,
+      "loss": 0.22,
+      "step": 28
+    },
+    {
+      "epoch": 0.10238305383936452,
+      "grad_norm": 1.731861214941761,
+      "learning_rate": 3.4117647058823532e-06,
+      "loss": 0.1702,
+      "step": 29
+    },
+    {
+      "epoch": 0.1059135039717564,
+      "grad_norm": 1.7963688467168324,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.2141,
+      "step": 30
+    },
+    {
+      "epoch": 0.10944395410414828,
+      "grad_norm": 1.8118405787472915,
+      "learning_rate": 3.6470588235294117e-06,
+      "loss": 0.2155,
+      "step": 31
+    },
+    {
+      "epoch": 0.11297440423654016,
+      "grad_norm": 2.1536428426895764,
+      "learning_rate": 3.7647058823529414e-06,
+      "loss": 0.2348,
+      "step": 32
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 1.3361162643597526,
+      "learning_rate": 3.882352941176471e-06,
+      "loss": 0.211,
+      "step": 33
+    },
+    {
+      "epoch": 0.12003530450132392,
+      "grad_norm": 1.3679923730470394,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.2006,
+      "step": 34
+    },
+    {
+      "epoch": 0.1235657546337158,
+      "grad_norm": 1.6951013269663189,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.2173,
+      "step": 35
+    },
+    {
+      "epoch": 0.12709620476610767,
+      "grad_norm": 1.1243675718278554,
+      "learning_rate": 4.235294117647059e-06,
+      "loss": 0.21,
+      "step": 36
+    },
+    {
+      "epoch": 0.13062665489849956,
+      "grad_norm": 1.798056949309266,
+      "learning_rate": 4.352941176470588e-06,
+      "loss": 0.2151,
+      "step": 37
+    },
+    {
+      "epoch": 0.13415710503089143,
+      "grad_norm": 1.5535634835962755,
+      "learning_rate": 4.4705882352941184e-06,
+      "loss": 0.2136,
+      "step": 38
+    },
+    {
+      "epoch": 0.13768755516328332,
+      "grad_norm": 1.3322693781225434,
+      "learning_rate": 4.588235294117647e-06,
+      "loss": 0.2016,
+      "step": 39
+    },
+    {
+      "epoch": 0.1412180052956752,
+      "grad_norm": 1.2353366230952958,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.2158,
+      "step": 40
+    },
+    {
+      "epoch": 0.14474845542806708,
+      "grad_norm": 1.649116854499638,
+      "learning_rate": 4.823529411764706e-06,
+      "loss": 0.2085,
+      "step": 41
+    },
+    {
+      "epoch": 0.14827890556045895,
+      "grad_norm": 1.5327812092407431,
+      "learning_rate": 4.941176470588236e-06,
+      "loss": 0.223,
+      "step": 42
+    },
+    {
+      "epoch": 0.15180935569285084,
+      "grad_norm": 1.3654123529518498,
+      "learning_rate": 5.058823529411765e-06,
+      "loss": 0.1897,
+      "step": 43
+    },
+    {
+      "epoch": 0.1553398058252427,
+      "grad_norm": 1.2874875896720268,
+      "learning_rate": 5.176470588235295e-06,
+      "loss": 0.208,
+      "step": 44
+    },
+    {
+      "epoch": 0.1588702559576346,
+      "grad_norm": 1.5894455663917926,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.2114,
+      "step": 45
+    },
+    {
+      "epoch": 0.1624007060900265,
+      "grad_norm": 1.4199861690645073,
+      "learning_rate": 5.411764705882353e-06,
+      "loss": 0.2204,
+      "step": 46
+    },
+    {
+      "epoch": 0.16593115622241836,
+      "grad_norm": 1.6111716054466962,
+      "learning_rate": 5.529411764705883e-06,
+      "loss": 0.2296,
+      "step": 47
+    },
+    {
+      "epoch": 0.16946160635481025,
+      "grad_norm": 1.4732964348658326,
+      "learning_rate": 5.6470588235294125e-06,
+      "loss": 0.2236,
+      "step": 48
+    },
+    {
+      "epoch": 0.17299205648720212,
+      "grad_norm": 1.7125830950347147,
+      "learning_rate": 5.764705882352941e-06,
+      "loss": 0.1889,
+      "step": 49
+    },
+    {
+      "epoch": 0.176522506619594,
+      "grad_norm": 1.4473702644438498,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.2002,
+      "step": 50
+    },
+    {
+      "epoch": 0.18005295675198588,
+      "grad_norm": 1.5337680099050695,
+      "learning_rate": 6e-06,
+      "loss": 0.2162,
+      "step": 51
+    },
+    {
+      "epoch": 0.18358340688437777,
+      "grad_norm": 1.6166188689363534,
+      "learning_rate": 6.11764705882353e-06,
+      "loss": 0.1984,
+      "step": 52
+    },
+    {
+      "epoch": 0.18711385701676964,
+      "grad_norm": 1.193587355481407,
+      "learning_rate": 6.2352941176470595e-06,
+      "loss": 0.1952,
+      "step": 53
+    },
+    {
+      "epoch": 0.19064430714916153,
+      "grad_norm": 1.150057201671324,
+      "learning_rate": 6.352941176470589e-06,
+      "loss": 0.2181,
+      "step": 54
+    },
+    {
+      "epoch": 0.1941747572815534,
+      "grad_norm": 1.6229574548509673,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.2064,
+      "step": 55
+    },
+    {
+      "epoch": 0.1977052074139453,
+      "grad_norm": 1.3194440409324901,
+      "learning_rate": 6.588235294117647e-06,
+      "loss": 0.2024,
+      "step": 56
+    },
+    {
+      "epoch": 0.20123565754633715,
+      "grad_norm": 1.5072216772819462,
+      "learning_rate": 6.705882352941176e-06,
+      "loss": 0.2041,
+      "step": 57
+    },
+    {
+      "epoch": 0.20476610767872905,
+      "grad_norm": 1.2184767982361704,
+      "learning_rate": 6.8235294117647065e-06,
+      "loss": 0.2165,
+      "step": 58
+    },
+    {
+      "epoch": 0.2082965578111209,
+      "grad_norm": 1.4524431095544894,
+      "learning_rate": 6.941176470588236e-06,
+      "loss": 0.2108,
+      "step": 59
+    },
+    {
+      "epoch": 0.2118270079435128,
+      "grad_norm": 1.132320106750761,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 0.1904,
+      "step": 60
+    },
+    {
+      "epoch": 0.21535745807590467,
+      "grad_norm": 1.215224093914693,
+      "learning_rate": 7.176470588235295e-06,
+      "loss": 0.1994,
+      "step": 61
+    },
+    {
+      "epoch": 0.21888790820829657,
+      "grad_norm": 1.0022175847945094,
+      "learning_rate": 7.294117647058823e-06,
+      "loss": 0.186,
+      "step": 62
+    },
+    {
+      "epoch": 0.22241835834068843,
+      "grad_norm": 1.3628751075073207,
+      "learning_rate": 7.4117647058823535e-06,
+      "loss": 0.1911,
+      "step": 63
+    },
+    {
+      "epoch": 0.22594880847308033,
+      "grad_norm": 1.2505219061147377,
+      "learning_rate": 7.529411764705883e-06,
+      "loss": 0.2018,
+      "step": 64
+    },
+    {
+      "epoch": 0.2294792586054722,
+      "grad_norm": 1.3707206838948995,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 0.1997,
+      "step": 65
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 1.3412107722466498,
+      "learning_rate": 7.764705882352941e-06,
+      "loss": 0.2174,
+      "step": 66
+    },
+    {
+      "epoch": 0.23654015887025595,
+      "grad_norm": 1.3129029846369458,
+      "learning_rate": 7.882352941176471e-06,
+      "loss": 0.1947,
+      "step": 67
+    },
+    {
+      "epoch": 0.24007060900264784,
+      "grad_norm": 1.6510384082728948,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.1724,
+      "step": 68
+    },
+    {
+      "epoch": 0.2436010591350397,
+      "grad_norm": 1.2774008213531902,
+      "learning_rate": 8.11764705882353e-06,
+      "loss": 0.2002,
+      "step": 69
+    },
+    {
+      "epoch": 0.2471315092674316,
+      "grad_norm": 1.3287314560957624,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.1928,
+      "step": 70
+    },
+    {
+      "epoch": 0.2506619593998235,
+      "grad_norm": 1.0681415016582574,
+      "learning_rate": 8.35294117647059e-06,
+      "loss": 0.1721,
+      "step": 71
+    },
+    {
+      "epoch": 0.25419240953221534,
+      "grad_norm": 1.21518473330371,
+      "learning_rate": 8.470588235294118e-06,
+      "loss": 0.1985,
+      "step": 72
+    },
+    {
+      "epoch": 0.25772285966460723,
+      "grad_norm": 1.4756769784911512,
+      "learning_rate": 8.588235294117647e-06,
+      "loss": 0.2281,
+      "step": 73
+    },
+    {
+      "epoch": 0.2612533097969991,
+      "grad_norm": 0.8760360831121267,
+      "learning_rate": 8.705882352941177e-06,
+      "loss": 0.1737,
+      "step": 74
+    },
+    {
+      "epoch": 0.264783759929391,
+      "grad_norm": 1.336571680306999,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.1901,
+      "step": 75
+    },
+    {
+      "epoch": 0.26831421006178285,
+      "grad_norm": 1.4034937627691704,
+      "learning_rate": 8.941176470588237e-06,
+      "loss": 0.2238,
+      "step": 76
+    },
+    {
+      "epoch": 0.27184466019417475,
+      "grad_norm": 1.1091052692159387,
+      "learning_rate": 9.058823529411765e-06,
+      "loss": 0.2,
+      "step": 77
+    },
+    {
+      "epoch": 0.27537511032656664,
+      "grad_norm": 1.5465224562234414,
+      "learning_rate": 9.176470588235294e-06,
+      "loss": 0.192,
+      "step": 78
+    },
+    {
+      "epoch": 0.27890556045895853,
+      "grad_norm": 1.4355050182008238,
+      "learning_rate": 9.294117647058824e-06,
+      "loss": 0.1853,
+      "step": 79
+    },
+    {
+      "epoch": 0.2824360105913504,
+      "grad_norm": 1.6148908553292696,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.1944,
+      "step": 80
+    },
+    {
+      "epoch": 0.28596646072374227,
+      "grad_norm": 1.006979010091179,
+      "learning_rate": 9.529411764705882e-06,
+      "loss": 0.1842,
+      "step": 81
+    },
+    {
+      "epoch": 0.28949691085613416,
+      "grad_norm": 1.2631451183457176,
+      "learning_rate": 9.647058823529412e-06,
+      "loss": 0.1889,
+      "step": 82
+    },
+    {
+      "epoch": 0.29302736098852605,
+      "grad_norm": 1.0554142641488224,
+      "learning_rate": 9.764705882352942e-06,
+      "loss": 0.1954,
+      "step": 83
+    },
+    {
+      "epoch": 0.2965578111209179,
+      "grad_norm": 1.2871157434331197,
+      "learning_rate": 9.882352941176472e-06,
+      "loss": 0.1783,
+      "step": 84
+    },
+    {
+      "epoch": 0.3000882612533098,
+      "grad_norm": 1.8307918651683295,
+      "learning_rate": 1e-05,
+      "loss": 0.2095,
+      "step": 85
+    },
+    {
+      "epoch": 0.3036187113857017,
+      "grad_norm": 1.27120661506548,
+      "learning_rate": 9.999996725424495e-06,
+      "loss": 0.1818,
+      "step": 86
+    },
+    {
+      "epoch": 0.30714916151809357,
+      "grad_norm": 1.3010925765539272,
+      "learning_rate": 9.999986901702262e-06,
+      "loss": 0.1967,
+      "step": 87
+    },
+    {
+      "epoch": 0.3106796116504854,
+      "grad_norm": 1.012603025593678,
+      "learning_rate": 9.999970528846173e-06,
+      "loss": 0.2084,
+      "step": 88
+    },
+    {
+      "epoch": 0.3142100617828773,
+      "grad_norm": 1.042462350269043,
+      "learning_rate": 9.99994760687767e-06,
+      "loss": 0.2113,
+      "step": 89
+    },
+    {
+      "epoch": 0.3177405119152692,
+      "grad_norm": 1.479637631853533,
+      "learning_rate": 9.999918135826783e-06,
+      "loss": 0.1923,
+      "step": 90
+    },
+    {
+      "epoch": 0.3212709620476611,
+      "grad_norm": 1.6536152705406355,
+      "learning_rate": 9.999882115732109e-06,
+      "loss": 0.1757,
+      "step": 91
+    },
+    {
+      "epoch": 0.324801412180053,
+      "grad_norm": 1.0022195112884287,
+      "learning_rate": 9.999839546640827e-06,
+      "loss": 0.2014,
+      "step": 92
+    },
+    {
+      "epoch": 0.3283318623124448,
+      "grad_norm": 1.0752815706731398,
+      "learning_rate": 9.9997904286087e-06,
+      "loss": 0.1947,
+      "step": 93
+    },
+    {
+      "epoch": 0.3318623124448367,
+      "grad_norm": 0.9640451554876094,
+      "learning_rate": 9.999734761700061e-06,
+      "loss": 0.1947,
+      "step": 94
+    },
+    {
+      "epoch": 0.3353927625772286,
+      "grad_norm": 0.9472529792678961,
+      "learning_rate": 9.999672545987826e-06,
+      "loss": 0.1933,
+      "step": 95
+    },
+    {
+      "epoch": 0.3389232127096205,
+      "grad_norm": 0.8262321839814756,
+      "learning_rate": 9.999603781553487e-06,
+      "loss": 0.1824,
+      "step": 96
+    },
+    {
+      "epoch": 0.34245366284201234,
+      "grad_norm": 1.302452468053817,
+      "learning_rate": 9.999528468487113e-06,
+      "loss": 0.2068,
+      "step": 97
+    },
+    {
+      "epoch": 0.34598411297440423,
+      "grad_norm": 0.903902282892577,
+      "learning_rate": 9.999446606887349e-06,
+      "loss": 0.1978,
+      "step": 98
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 1.0828870366995145,
+      "learning_rate": 9.999358196861422e-06,
+      "loss": 0.218,
+      "step": 99
+    },
+    {
+      "epoch": 0.353045013239188,
+      "grad_norm": 1.1262059857251303,
+      "learning_rate": 9.999263238525135e-06,
+      "loss": 0.2054,
+      "step": 100
+    },
+    {
+      "epoch": 0.35657546337157986,
+      "grad_norm": 0.9433649602156967,
+      "learning_rate": 9.999161732002867e-06,
+      "loss": 0.1811,
+      "step": 101
+    },
+    {
+      "epoch": 0.36010591350397175,
+      "grad_norm": 1.2486258616646937,
+      "learning_rate": 9.999053677427573e-06,
+      "loss": 0.2028,
+      "step": 102
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 1.0344680877191332,
+      "learning_rate": 9.998939074940788e-06,
+      "loss": 0.2063,
+      "step": 103
+    },
+    {
+      "epoch": 0.36716681376875554,
+      "grad_norm": 1.3236038216383461,
+      "learning_rate": 9.99881792469262e-06,
+      "loss": 0.2014,
+      "step": 104
+    },
+    {
+      "epoch": 0.3706972639011474,
+      "grad_norm": 1.0486814111295473,
+      "learning_rate": 9.998690226841756e-06,
+      "loss": 0.2005,
+      "step": 105
+    },
+    {
+      "epoch": 0.37422771403353927,
+      "grad_norm": 0.9052783364228919,
+      "learning_rate": 9.998555981555459e-06,
+      "loss": 0.1806,
+      "step": 106
+    },
+    {
+      "epoch": 0.37775816416593117,
+      "grad_norm": 0.8762864385995432,
+      "learning_rate": 9.998415189009566e-06,
+      "loss": 0.1963,
+      "step": 107
+    },
+    {
+      "epoch": 0.38128861429832306,
+      "grad_norm": 1.0133330127905669,
+      "learning_rate": 9.998267849388494e-06,
+      "loss": 0.2063,
+      "step": 108
+    },
+    {
+      "epoch": 0.3848190644307149,
+      "grad_norm": 1.1032798889585722,
+      "learning_rate": 9.99811396288523e-06,
+      "loss": 0.2136,
+      "step": 109
+    },
+    {
+      "epoch": 0.3883495145631068,
+      "grad_norm": 0.9526777003854597,
+      "learning_rate": 9.99795352970134e-06,
+      "loss": 0.1997,
+      "step": 110
+    },
+    {
+      "epoch": 0.3918799646954987,
+      "grad_norm": 1.3675080989097932,
+      "learning_rate": 9.997786550046965e-06,
+      "loss": 0.1866,
+      "step": 111
+    },
+    {
+      "epoch": 0.3954104148278906,
+      "grad_norm": 1.0304325204839548,
+      "learning_rate": 9.99761302414082e-06,
+      "loss": 0.2001,
+      "step": 112
+    },
+    {
+      "epoch": 0.3989408649602824,
+      "grad_norm": 0.8029445134568592,
+      "learning_rate": 9.997432952210193e-06,
+      "loss": 0.1806,
+      "step": 113
+    },
+    {
+      "epoch": 0.4024713150926743,
+      "grad_norm": 0.9662432601060303,
+      "learning_rate": 9.997246334490952e-06,
+      "loss": 0.1983,
+      "step": 114
+    },
+    {
+      "epoch": 0.4060017652250662,
+      "grad_norm": 0.6650992554809413,
+      "learning_rate": 9.997053171227527e-06,
+      "loss": 0.1904,
+      "step": 115
+    },
+    {
+      "epoch": 0.4095322153574581,
+      "grad_norm": 1.2497143912397224,
+      "learning_rate": 9.996853462672935e-06,
+      "loss": 0.1921,
+      "step": 116
+    },
+    {
+      "epoch": 0.41306266548984993,
+      "grad_norm": 1.2105700846879488,
+      "learning_rate": 9.996647209088759e-06,
+      "loss": 0.1913,
+      "step": 117
+    },
+    {
+      "epoch": 0.4165931156222418,
+      "grad_norm": 0.8745391363159488,
+      "learning_rate": 9.996434410745158e-06,
+      "loss": 0.1877,
+      "step": 118
+    },
+    {
+      "epoch": 0.4201235657546337,
+      "grad_norm": 1.155984605743064,
+      "learning_rate": 9.996215067920855e-06,
+      "loss": 0.1931,
+      "step": 119
+    },
+    {
+      "epoch": 0.4236540158870256,
+      "grad_norm": 0.9959150019854762,
+      "learning_rate": 9.99598918090316e-06,
+      "loss": 0.1901,
+      "step": 120
+    },
+    {
+      "epoch": 0.42718446601941745,
+      "grad_norm": 1.051448796745704,
+      "learning_rate": 9.995756749987942e-06,
+      "loss": 0.1642,
+      "step": 121
+    },
+    {
+      "epoch": 0.43071491615180935,
+      "grad_norm": 0.935738059968614,
+      "learning_rate": 9.995517775479647e-06,
+      "loss": 0.1702,
+      "step": 122
+    },
+    {
+      "epoch": 0.43424536628420124,
+      "grad_norm": 1.085401876613817,
+      "learning_rate": 9.995272257691291e-06,
+      "loss": 0.1992,
+      "step": 123
+    },
+    {
+      "epoch": 0.43777581641659313,
+      "grad_norm": 0.9924411904534596,
+      "learning_rate": 9.99502019694446e-06,
+      "loss": 0.1922,
+      "step": 124
+    },
+    {
+      "epoch": 0.44130626654898497,
+      "grad_norm": 1.3346252020931497,
+      "learning_rate": 9.994761593569312e-06,
+      "loss": 0.1894,
+      "step": 125
+    },
+    {
+      "epoch": 0.44483671668137686,
+      "grad_norm": 0.996928968088298,
+      "learning_rate": 9.994496447904573e-06,
+      "loss": 0.18,
+      "step": 126
+    },
+    {
+      "epoch": 0.44836716681376876,
+      "grad_norm": 1.185034118977359,
+      "learning_rate": 9.99422476029754e-06,
+      "loss": 0.2015,
+      "step": 127
+    },
+    {
+      "epoch": 0.45189761694616065,
+      "grad_norm": 1.079465604221569,
+      "learning_rate": 9.993946531104075e-06,
+      "loss": 0.1795,
+      "step": 128
+    },
+    {
+      "epoch": 0.4554280670785525,
+      "grad_norm": 1.0665272432739623,
+      "learning_rate": 9.993661760688611e-06,
+      "loss": 0.199,
+      "step": 129
+    },
+    {
+      "epoch": 0.4589585172109444,
+      "grad_norm": 1.2364225145584533,
+      "learning_rate": 9.993370449424153e-06,
+      "loss": 0.2198,
+      "step": 130
+    },
+    {
+      "epoch": 0.4624889673433363,
+      "grad_norm": 0.755126320981263,
+      "learning_rate": 9.993072597692266e-06,
+      "loss": 0.2006,
+      "step": 131
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 1.3871006955467502,
+      "learning_rate": 9.992768205883085e-06,
+      "loss": 0.2037,
+      "step": 132
+    },
+    {
+      "epoch": 0.46954986760812,
+      "grad_norm": 0.7426689064084516,
+      "learning_rate": 9.992457274395312e-06,
+      "loss": 0.1774,
+      "step": 133
+    },
+    {
+      "epoch": 0.4730803177405119,
+      "grad_norm": 0.9196839690873588,
+      "learning_rate": 9.992139803636218e-06,
+      "loss": 0.1973,
+      "step": 134
+    },
+    {
+      "epoch": 0.4766107678729038,
+      "grad_norm": 0.7786869026927636,
+      "learning_rate": 9.99181579402163e-06,
+      "loss": 0.1842,
+      "step": 135
+    },
+    {
+      "epoch": 0.4801412180052957,
+      "grad_norm": 1.03244467425382,
+      "learning_rate": 9.991485245975949e-06,
+      "loss": 0.1663,
+      "step": 136
+    },
+    {
+      "epoch": 0.4836716681376876,
+      "grad_norm": 1.0713507973139096,
+      "learning_rate": 9.991148159932134e-06,
+      "loss": 0.1941,
+      "step": 137
+    },
+    {
+      "epoch": 0.4872021182700794,
+      "grad_norm": 0.9453234857914946,
+      "learning_rate": 9.990804536331715e-06,
+      "loss": 0.2007,
+      "step": 138
+    },
+    {
+      "epoch": 0.4907325684024713,
+      "grad_norm": 0.7607319035095186,
+      "learning_rate": 9.990454375624778e-06,
+      "loss": 0.18,
+      "step": 139
+    },
+    {
+      "epoch": 0.4942630185348632,
+      "grad_norm": 0.9915191845505621,
+      "learning_rate": 9.990097678269972e-06,
+      "loss": 0.1832,
+      "step": 140
+    },
+    {
+      "epoch": 0.4977934686672551,
+      "grad_norm": 0.6818554079855815,
+      "learning_rate": 9.989734444734513e-06,
+      "loss": 0.1779,
+      "step": 141
+    },
+    {
+      "epoch": 0.501323918799647,
+      "grad_norm": 0.8679707616739033,
+      "learning_rate": 9.989364675494175e-06,
+      "loss": 0.1585,
+      "step": 142
+    },
+    {
+      "epoch": 0.5048543689320388,
+      "grad_norm": 0.834867345483289,
+      "learning_rate": 9.988988371033292e-06,
+      "loss": 0.1684,
+      "step": 143
+    },
+    {
+      "epoch": 0.5083848190644307,
+      "grad_norm": 0.7262565816433312,
+      "learning_rate": 9.98860553184476e-06,
+      "loss": 0.1688,
+      "step": 144
+    },
+    {
+      "epoch": 0.5119152691968226,
+      "grad_norm": 0.9922761294466914,
+      "learning_rate": 9.988216158430033e-06,
+      "loss": 0.1846,
+      "step": 145
+    },
+    {
+      "epoch": 0.5154457193292145,
+      "grad_norm": 1.0610599926539903,
+      "learning_rate": 9.987820251299121e-06,
+      "loss": 0.1965,
+      "step": 146
+    },
+    {
+      "epoch": 0.5189761694616064,
+      "grad_norm": 0.8658381475355275,
+      "learning_rate": 9.987417810970601e-06,
+      "loss": 0.1923,
+      "step": 147
+    },
+    {
+      "epoch": 0.5225066195939982,
+      "grad_norm": 1.0687893341105537,
+      "learning_rate": 9.987008837971595e-06,
+      "loss": 0.2084,
+      "step": 148
+    },
+    {
+      "epoch": 0.5260370697263901,
+      "grad_norm": 1.0221589345087516,
+      "learning_rate": 9.986593332837795e-06,
+      "loss": 0.1825,
+      "step": 149
+    },
+    {
+      "epoch": 0.529567519858782,
+      "grad_norm": 0.9977245435165767,
+      "learning_rate": 9.986171296113437e-06,
+      "loss": 0.2014,
+      "step": 150
+    },
+    {
+      "epoch": 0.5330979699911739,
+      "grad_norm": 1.0012549684835264,
+      "learning_rate": 9.985742728351317e-06,
+      "loss": 0.1885,
+      "step": 151
+    },
+    {
+      "epoch": 0.5366284201235657,
+      "grad_norm": 1.2108459546620967,
+      "learning_rate": 9.98530763011279e-06,
+      "loss": 0.1888,
+      "step": 152
+    },
+    {
+      "epoch": 0.5401588702559577,
+      "grad_norm": 0.849586927833,
+      "learning_rate": 9.98486600196776e-06,
+      "loss": 0.1957,
+      "step": 153
+    },
+    {
+      "epoch": 0.5436893203883495,
+      "grad_norm": 0.929219836766702,
+      "learning_rate": 9.984417844494681e-06,
+      "loss": 0.2051,
+      "step": 154
+    },
+    {
+      "epoch": 0.5472197705207414,
+      "grad_norm": 0.7551838277587243,
+      "learning_rate": 9.983963158280568e-06,
+      "loss": 0.1687,
+      "step": 155
+    },
+    {
+      "epoch": 0.5507502206531333,
+      "grad_norm": 0.8965218043033845,
+      "learning_rate": 9.983501943920978e-06,
+      "loss": 0.2092,
+      "step": 156
+    },
+    {
+      "epoch": 0.5542806707855251,
+      "grad_norm": 0.9118641687914303,
+      "learning_rate": 9.98303420202003e-06,
+      "loss": 0.1857,
+      "step": 157
+    },
+    {
+      "epoch": 0.5578111209179171,
+      "grad_norm": 1.0557315457012684,
+      "learning_rate": 9.98255993319038e-06,
+      "loss": 0.1796,
+      "step": 158
+    },
+    {
+      "epoch": 0.5613415710503089,
+      "grad_norm": 0.9917786126115438,
+      "learning_rate": 9.982079138053243e-06,
+      "loss": 0.1797,
+      "step": 159
+    },
+    {
+      "epoch": 0.5648720211827007,
+      "grad_norm": 0.9264989403126399,
+      "learning_rate": 9.981591817238379e-06,
+      "loss": 0.1814,
+      "step": 160
+    },
+    {
+      "epoch": 0.5684024713150927,
+      "grad_norm": 1.1038265073069693,
+      "learning_rate": 9.981097971384094e-06,
+      "loss": 0.2039,
+      "step": 161
+    },
+    {
+      "epoch": 0.5719329214474845,
+      "grad_norm": 1.1450789849461764,
+      "learning_rate": 9.980597601137244e-06,
+      "loss": 0.2262,
+      "step": 162
+    },
+    {
+      "epoch": 0.5754633715798765,
+      "grad_norm": 1.0093808973803213,
+      "learning_rate": 9.980090707153226e-06,
+      "loss": 0.1767,
+      "step": 163
+    },
+    {
+      "epoch": 0.5789938217122683,
+      "grad_norm": 0.8305835473538516,
+      "learning_rate": 9.979577290095988e-06,
+      "loss": 0.1854,
+      "step": 164
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.8814645905158526,
+      "learning_rate": 9.97905735063802e-06,
+      "loss": 0.207,
+      "step": 165
+    },
+    {
+      "epoch": 0.5860547219770521,
+      "grad_norm": 1.1283393561929782,
+      "learning_rate": 9.978530889460351e-06,
+      "loss": 0.1482,
+      "step": 166
+    },
+    {
+      "epoch": 0.589585172109444,
+      "grad_norm": 0.8882147790176694,
+      "learning_rate": 9.977997907252557e-06,
+      "loss": 0.1778,
+      "step": 167
+    },
+    {
+      "epoch": 0.5931156222418358,
+      "grad_norm": 0.8063636359843489,
+      "learning_rate": 9.977458404712755e-06,
+      "loss": 0.1922,
+      "step": 168
+    },
+    {
+      "epoch": 0.5966460723742277,
+      "grad_norm": 0.8023695735241588,
+      "learning_rate": 9.976912382547603e-06,
+      "loss": 0.1979,
+      "step": 169
+    },
+    {
+      "epoch": 0.6001765225066196,
+      "grad_norm": 0.8128057874275411,
+      "learning_rate": 9.976359841472294e-06,
+      "loss": 0.196,
+      "step": 170
+    },
+    {
+      "epoch": 0.6037069726390115,
+      "grad_norm": 0.8202708710729476,
+      "learning_rate": 9.975800782210565e-06,
+      "loss": 0.1829,
+      "step": 171
+    },
+    {
+      "epoch": 0.6072374227714034,
+      "grad_norm": 0.9514477549047732,
+      "learning_rate": 9.975235205494689e-06,
+      "loss": 0.218,
+      "step": 172
+    },
+    {
+      "epoch": 0.6107678729037952,
+      "grad_norm": 1.0222112927480378,
+      "learning_rate": 9.974663112065474e-06,
+      "loss": 0.1783,
+      "step": 173
+    },
+    {
+      "epoch": 0.6142983230361871,
+      "grad_norm": 0.891305185008567,
+      "learning_rate": 9.974084502672266e-06,
+      "loss": 0.1588,
+      "step": 174
+    },
+    {
+      "epoch": 0.617828773168579,
+      "grad_norm": 0.8521407342632297,
+      "learning_rate": 9.973499378072947e-06,
+      "loss": 0.1997,
+      "step": 175
+    },
+    {
+      "epoch": 0.6213592233009708,
+      "grad_norm": 0.8261272762704142,
+      "learning_rate": 9.972907739033927e-06,
+      "loss": 0.2064,
+      "step": 176
+    },
+    {
+      "epoch": 0.6248896734333628,
+      "grad_norm": 0.9844072163334512,
+      "learning_rate": 9.972309586330156e-06,
+      "loss": 0.1934,
+      "step": 177
+    },
+    {
+      "epoch": 0.6284201235657546,
+      "grad_norm": 0.9760362259758513,
+      "learning_rate": 9.97170492074511e-06,
+      "loss": 0.1711,
+      "step": 178
+    },
+    {
+      "epoch": 0.6319505736981466,
+      "grad_norm": 0.8171800425370801,
+      "learning_rate": 9.9710937430708e-06,
+      "loss": 0.1814,
+      "step": 179
+    },
+    {
+      "epoch": 0.6354810238305384,
+      "grad_norm": 0.7848321856443494,
+      "learning_rate": 9.970476054107763e-06,
+      "loss": 0.1688,
+      "step": 180
+    },
+    {
+      "epoch": 0.6390114739629302,
+      "grad_norm": 0.7304874742153669,
+      "learning_rate": 9.969851854665072e-06,
+      "loss": 0.1975,
+      "step": 181
+    },
+    {
+      "epoch": 0.6425419240953222,
+      "grad_norm": 0.8122652782447508,
+      "learning_rate": 9.969221145560315e-06,
+      "loss": 0.1856,
+      "step": 182
+    },
+    {
+      "epoch": 0.646072374227714,
+      "grad_norm": 0.5663405788505017,
+      "learning_rate": 9.968583927619618e-06,
+      "loss": 0.1823,
+      "step": 183
+    },
+    {
+      "epoch": 0.649602824360106,
+      "grad_norm": 0.8631102591025225,
+      "learning_rate": 9.967940201677628e-06,
+      "loss": 0.191,
+      "step": 184
+    },
+    {
+      "epoch": 0.6531332744924978,
+      "grad_norm": 0.9219061678650854,
+      "learning_rate": 9.967289968577515e-06,
+      "loss": 0.1867,
+      "step": 185
+    },
+    {
+      "epoch": 0.6566637246248896,
+      "grad_norm": 0.8104447482059561,
+      "learning_rate": 9.966633229170975e-06,
+      "loss": 0.2085,
+      "step": 186
+    },
+    {
+      "epoch": 0.6601941747572816,
+      "grad_norm": 0.696325768189324,
+      "learning_rate": 9.965969984318227e-06,
+      "loss": 0.1955,
+      "step": 187
+    },
+    {
+      "epoch": 0.6637246248896734,
+      "grad_norm": 0.957747594144563,
+      "learning_rate": 9.965300234888007e-06,
+      "loss": 0.1973,
+      "step": 188
+    },
+    {
+      "epoch": 0.6672550750220653,
+      "grad_norm": 0.8838107190805117,
+      "learning_rate": 9.964623981757572e-06,
+      "loss": 0.1862,
+      "step": 189
+    },
+    {
+      "epoch": 0.6707855251544572,
+      "grad_norm": 1.0086692709079665,
+      "learning_rate": 9.9639412258127e-06,
+      "loss": 0.1876,
+      "step": 190
+    },
+    {
+      "epoch": 0.6743159752868491,
+      "grad_norm": 0.7851582403291492,
+      "learning_rate": 9.963251967947687e-06,
+      "loss": 0.1925,
+      "step": 191
+    },
+    {
+      "epoch": 0.677846425419241,
+      "grad_norm": 0.9006177750516297,
+      "learning_rate": 9.962556209065343e-06,
+      "loss": 0.1651,
+      "step": 192
+    },
+    {
+      "epoch": 0.6813768755516328,
+      "grad_norm": 0.8917643734840932,
+      "learning_rate": 9.961853950076992e-06,
+      "loss": 0.164,
+      "step": 193
+    },
+    {
+      "epoch": 0.6849073256840247,
+      "grad_norm": 0.8833175487061256,
+      "learning_rate": 9.961145191902475e-06,
+      "loss": 0.1858,
+      "step": 194
+    },
+    {
+      "epoch": 0.6884377758164166,
+      "grad_norm": 0.7703074827448632,
+      "learning_rate": 9.960429935470146e-06,
+      "loss": 0.1719,
+      "step": 195
+    },
+    {
+      "epoch": 0.6919682259488085,
+      "grad_norm": 0.978543268699946,
+      "learning_rate": 9.959708181716869e-06,
+      "loss": 0.1843,
+      "step": 196
+    },
+    {
+      "epoch": 0.6954986760812003,
+      "grad_norm": 1.0182659329814947,
+      "learning_rate": 9.958979931588018e-06,
+      "loss": 0.1698,
+      "step": 197
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 1.0948049910788071,
+      "learning_rate": 9.958245186037477e-06,
+      "loss": 0.1892,
+      "step": 198
+    },
+    {
+      "epoch": 0.7025595763459841,
+      "grad_norm": 0.6933804652704536,
+      "learning_rate": 9.957503946027638e-06,
+      "loss": 0.1672,
+      "step": 199
+    },
+    {
+      "epoch": 0.706090026478376,
+      "grad_norm": 0.8077103768832906,
+      "learning_rate": 9.956756212529402e-06,
+      "loss": 0.1585,
+      "step": 200
+    },
+    {
+      "epoch": 0.7096204766107679,
+      "grad_norm": 0.822023166461492,
+      "learning_rate": 9.95600198652217e-06,
+      "loss": 0.1801,
+      "step": 201
+    },
+    {
+      "epoch": 0.7131509267431597,
+      "grad_norm": 1.1206342687945703,
+      "learning_rate": 9.955241268993852e-06,
+      "loss": 0.201,
+      "step": 202
+    },
+    {
+      "epoch": 0.7166813768755517,
+      "grad_norm": 0.974499750770202,
+      "learning_rate": 9.954474060940856e-06,
+      "loss": 0.2085,
+      "step": 203
+    },
+    {
+      "epoch": 0.7202118270079435,
+      "grad_norm": 0.6598260944598164,
+      "learning_rate": 9.953700363368097e-06,
+      "loss": 0.1859,
+      "step": 204
+    },
+    {
+      "epoch": 0.7237422771403353,
+      "grad_norm": 0.8633415747390508,
+      "learning_rate": 9.952920177288985e-06,
+      "loss": 0.1963,
+      "step": 205
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.5898944244131702,
+      "learning_rate": 9.952133503725435e-06,
+      "loss": 0.2049,
+      "step": 206
+    },
+    {
+      "epoch": 0.7308031774051191,
+      "grad_norm": 1.2714984543033625,
+      "learning_rate": 9.951340343707852e-06,
+      "loss": 0.1837,
+      "step": 207
+    },
+    {
+      "epoch": 0.7343336275375111,
+      "grad_norm": 0.9097980008759828,
+      "learning_rate": 9.950540698275144e-06,
+      "loss": 0.1849,
+      "step": 208
+    },
+    {
+      "epoch": 0.7378640776699029,
+      "grad_norm": 0.6941779662092539,
+      "learning_rate": 9.949734568474708e-06,
+      "loss": 0.176,
+      "step": 209
+    },
+    {
+      "epoch": 0.7413945278022948,
+      "grad_norm": 1.0789428661342089,
+      "learning_rate": 9.948921955362438e-06,
+      "loss": 0.1845,
+      "step": 210
+    },
+    {
+      "epoch": 0.7449249779346867,
+      "grad_norm": 1.002969780293264,
+      "learning_rate": 9.94810286000272e-06,
+      "loss": 0.1932,
+      "step": 211
+    },
+    {
+      "epoch": 0.7484554280670785,
+      "grad_norm": 0.7346857479212374,
+      "learning_rate": 9.94727728346843e-06,
+      "loss": 0.1895,
+      "step": 212
+    },
+    {
+      "epoch": 0.7519858781994704,
+      "grad_norm": 1.1214956368968594,
+      "learning_rate": 9.946445226840932e-06,
+      "loss": 0.1872,
+      "step": 213
+    },
+    {
+      "epoch": 0.7555163283318623,
+      "grad_norm": 0.8098364912886733,
+      "learning_rate": 9.945606691210081e-06,
+      "loss": 0.182,
+      "step": 214
+    },
+    {
+      "epoch": 0.7590467784642542,
+      "grad_norm": 0.9351430688337309,
+      "learning_rate": 9.944761677674214e-06,
+      "loss": 0.194,
+      "step": 215
+    },
+    {
+      "epoch": 0.7625772285966461,
+      "grad_norm": 1.1657299434860284,
+      "learning_rate": 9.943910187340154e-06,
+      "loss": 0.1875,
+      "step": 216
+    },
+    {
+      "epoch": 0.766107678729038,
+      "grad_norm": 1.0056936253752848,
+      "learning_rate": 9.943052221323212e-06,
+      "loss": 0.1993,
+      "step": 217
+    },
+    {
+      "epoch": 0.7696381288614298,
+      "grad_norm": 1.018936012562021,
+      "learning_rate": 9.942187780747177e-06,
+      "loss": 0.205,
+      "step": 218
+    },
+    {
+      "epoch": 0.7731685789938217,
+      "grad_norm": 0.9704710377674022,
+      "learning_rate": 9.94131686674432e-06,
+      "loss": 0.1803,
+      "step": 219
+    },
+    {
+      "epoch": 0.7766990291262136,
+      "grad_norm": 0.844440847315935,
+      "learning_rate": 9.940439480455386e-06,
+      "loss": 0.1903,
+      "step": 220
+    },
+    {
+      "epoch": 0.7802294792586054,
+      "grad_norm": 0.6911229516194293,
+      "learning_rate": 9.939555623029608e-06,
+      "loss": 0.1648,
+      "step": 221
+    },
+    {
+      "epoch": 0.7837599293909974,
+      "grad_norm": 0.977598516917381,
+      "learning_rate": 9.938665295624686e-06,
+      "loss": 0.1775,
+      "step": 222
+    },
+    {
+      "epoch": 0.7872903795233892,
+      "grad_norm": 0.7894216874582392,
+      "learning_rate": 9.937768499406798e-06,
+      "loss": 0.1734,
+      "step": 223
+    },
+    {
+      "epoch": 0.7908208296557812,
+      "grad_norm": 0.8720889047054564,
+      "learning_rate": 9.936865235550595e-06,
+      "loss": 0.1818,
+      "step": 224
+    },
+    {
+      "epoch": 0.794351279788173,
+      "grad_norm": 0.7210753676381008,
+      "learning_rate": 9.9359555052392e-06,
+      "loss": 0.2001,
+      "step": 225
+    },
+    {
+      "epoch": 0.7978817299205648,
+      "grad_norm": 0.6358344164302707,
+      "learning_rate": 9.935039309664204e-06,
+      "loss": 0.1703,
+      "step": 226
+    },
+    {
+      "epoch": 0.8014121800529568,
+      "grad_norm": 0.8138407178306513,
+      "learning_rate": 9.93411665002567e-06,
+      "loss": 0.172,
+      "step": 227
+    },
+    {
+      "epoch": 0.8049426301853486,
+      "grad_norm": 0.974591399579874,
+      "learning_rate": 9.93318752753212e-06,
+      "loss": 0.2067,
+      "step": 228
+    },
+    {
+      "epoch": 0.8084730803177406,
+      "grad_norm": 0.8238739378175247,
+      "learning_rate": 9.932251943400554e-06,
+      "loss": 0.1607,
+      "step": 229
+    },
+    {
+      "epoch": 0.8120035304501324,
+      "grad_norm": 1.0422244856011833,
+      "learning_rate": 9.931309898856423e-06,
+      "loss": 0.1693,
+      "step": 230
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 0.812884439844158,
+      "learning_rate": 9.930361395133647e-06,
+      "loss": 0.1683,
+      "step": 231
+    },
+    {
+      "epoch": 0.8190644307149162,
+      "grad_norm": 0.6222138282668747,
+      "learning_rate": 9.929406433474606e-06,
+      "loss": 0.1636,
+      "step": 232
+    },
+    {
+      "epoch": 0.822594880847308,
+      "grad_norm": 0.8762443957945536,
+      "learning_rate": 9.928445015130136e-06,
+      "loss": 0.191,
+      "step": 233
+    },
+    {
+      "epoch": 0.8261253309796999,
+      "grad_norm": 0.9859315198495622,
+      "learning_rate": 9.927477141359533e-06,
+      "loss": 0.1895,
+      "step": 234
+    },
+    {
+      "epoch": 0.8296557811120918,
+      "grad_norm": 0.9589349463946001,
+      "learning_rate": 9.926502813430545e-06,
+      "loss": 0.197,
+      "step": 235
+    },
+    {
+      "epoch": 0.8331862312444837,
+      "grad_norm": 0.8718214863000249,
+      "learning_rate": 9.92552203261938e-06,
+      "loss": 0.1799,
+      "step": 236
+    },
+    {
+      "epoch": 0.8367166813768756,
+      "grad_norm": 1.1335307020440895,
+      "learning_rate": 9.92453480021069e-06,
+      "loss": 0.198,
+      "step": 237
+    },
+    {
+      "epoch": 0.8402471315092674,
+      "grad_norm": 0.8430408249858925,
+      "learning_rate": 9.923541117497586e-06,
+      "loss": 0.1478,
+      "step": 238
+    },
+    {
+      "epoch": 0.8437775816416593,
+      "grad_norm": 0.6769109520694074,
+      "learning_rate": 9.922540985781621e-06,
+      "loss": 0.1923,
+      "step": 239
+    },
+    {
+      "epoch": 0.8473080317740512,
+      "grad_norm": 0.6377271343571659,
+      "learning_rate": 9.921534406372797e-06,
+      "loss": 0.1879,
+      "step": 240
+    },
+    {
+      "epoch": 0.8508384819064431,
+      "grad_norm": 0.7271721148140101,
+      "learning_rate": 9.920521380589566e-06,
+      "loss": 0.1938,
+      "step": 241
+    },
+    {
+      "epoch": 0.8543689320388349,
+      "grad_norm": 0.9160321910693624,
+      "learning_rate": 9.919501909758815e-06,
+      "loss": 0.1689,
+      "step": 242
+    },
+    {
+      "epoch": 0.8578993821712269,
+      "grad_norm": 0.8451022682499405,
+      "learning_rate": 9.918475995215881e-06,
+      "loss": 0.1567,
+      "step": 243
+    },
+    {
+      "epoch": 0.8614298323036187,
+      "grad_norm": 0.7383618763118188,
+      "learning_rate": 9.917443638304536e-06,
+      "loss": 0.1998,
+      "step": 244
+    },
+    {
+      "epoch": 0.8649602824360106,
+      "grad_norm": 0.9977901978127347,
+      "learning_rate": 9.916404840376993e-06,
+      "loss": 0.2184,
+      "step": 245
+    },
+    {
+      "epoch": 0.8684907325684025,
+      "grad_norm": 0.905766448884252,
+      "learning_rate": 9.915359602793902e-06,
+      "loss": 0.1822,
+      "step": 246
+    },
+    {
+      "epoch": 0.8720211827007943,
+      "grad_norm": 0.668257498288602,
+      "learning_rate": 9.914307926924344e-06,
+      "loss": 0.1755,
+      "step": 247
+    },
+    {
+      "epoch": 0.8755516328331863,
+      "grad_norm": 1.2898306231343724,
+      "learning_rate": 9.913249814145839e-06,
+      "loss": 0.1943,
+      "step": 248
+    },
+    {
+      "epoch": 0.8790820829655781,
+      "grad_norm": 0.7801616088781465,
+      "learning_rate": 9.912185265844333e-06,
+      "loss": 0.1961,
+      "step": 249
+    },
+    {
+      "epoch": 0.8826125330979699,
+      "grad_norm": 0.8499032108492505,
+      "learning_rate": 9.911114283414204e-06,
+      "loss": 0.2022,
+      "step": 250
+    },
+    {
+      "epoch": 0.8861429832303619,
+      "grad_norm": 0.9986716197227599,
+      "learning_rate": 9.910036868258256e-06,
+      "loss": 0.1935,
+      "step": 251
+    },
+    {
+      "epoch": 0.8896734333627537,
+      "grad_norm": 0.8164649483937467,
+      "learning_rate": 9.908953021787725e-06,
+      "loss": 0.1772,
+      "step": 252
+    },
+    {
+      "epoch": 0.8932038834951457,
+      "grad_norm": 0.6585245832704895,
+      "learning_rate": 9.907862745422258e-06,
+      "loss": 0.1762,
+      "step": 253
+    },
+    {
+      "epoch": 0.8967343336275375,
+      "grad_norm": 0.7669263548180228,
+      "learning_rate": 9.906766040589937e-06,
+      "loss": 0.1673,
+      "step": 254
+    },
+    {
+      "epoch": 0.9002647837599294,
+      "grad_norm": 0.8311965531630117,
+      "learning_rate": 9.905662908727258e-06,
+      "loss": 0.1657,
+      "step": 255
+    },
+    {
+      "epoch": 0.9037952338923213,
+      "grad_norm": 0.6844670189568429,
+      "learning_rate": 9.904553351279139e-06,
+      "loss": 0.1594,
+      "step": 256
+    },
+    {
+      "epoch": 0.9073256840247131,
+      "grad_norm": 0.8500217928271374,
+      "learning_rate": 9.903437369698906e-06,
+      "loss": 0.2076,
+      "step": 257
+    },
+    {
+      "epoch": 0.910856134157105,
+      "grad_norm": 0.8553603507750409,
+      "learning_rate": 9.902314965448309e-06,
+      "loss": 0.1838,
+      "step": 258
+    },
+    {
+      "epoch": 0.9143865842894969,
+      "grad_norm": 0.8770431320611423,
+      "learning_rate": 9.901186139997507e-06,
+      "loss": 0.1935,
+      "step": 259
+    },
+    {
+      "epoch": 0.9179170344218888,
+      "grad_norm": 0.7281161731708585,
+      "learning_rate": 9.90005089482507e-06,
+      "loss": 0.1888,
+      "step": 260
+    },
+    {
+      "epoch": 0.9214474845542807,
+      "grad_norm": 0.6456235823671523,
+      "learning_rate": 9.898909231417976e-06,
+      "loss": 0.1764,
+      "step": 261
+    },
+    {
+      "epoch": 0.9249779346866726,
+      "grad_norm": 0.6768287107038422,
+      "learning_rate": 9.89776115127161e-06,
+      "loss": 0.163,
+      "step": 262
+    },
+    {
+      "epoch": 0.9285083848190644,
+      "grad_norm": 0.834905164466666,
+      "learning_rate": 9.89660665588976e-06,
+      "loss": 0.1842,
+      "step": 263
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 0.6460482257383012,
+      "learning_rate": 9.895445746784623e-06,
+      "loss": 0.165,
+      "step": 264
+    },
+    {
+      "epoch": 0.9355692850838482,
+      "grad_norm": 0.9772687205581816,
+      "learning_rate": 9.89427842547679e-06,
+      "loss": 0.2003,
+      "step": 265
+    },
+    {
+      "epoch": 0.93909973521624,
+      "grad_norm": 0.7686896072265816,
+      "learning_rate": 9.893104693495255e-06,
+      "loss": 0.1753,
+      "step": 266
+    },
+    {
+      "epoch": 0.942630185348632,
+      "grad_norm": 0.7667904026391998,
+      "learning_rate": 9.891924552377405e-06,
+      "loss": 0.1802,
+      "step": 267
+    },
+    {
+      "epoch": 0.9461606354810238,
+      "grad_norm": 0.6993230964449527,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.1701,
+      "step": 268
+    },
+    {
+      "epoch": 0.9496910856134158,
+      "grad_norm": 0.7632215639559887,
+      "learning_rate": 9.8895450489243e-06,
+      "loss": 0.1607,
+      "step": 269
+    },
+    {
+      "epoch": 0.9532215357458076,
+      "grad_norm": 0.8307898449698382,
+      "learning_rate": 9.88834568970579e-06,
+      "loss": 0.1631,
+      "step": 270
+    },
+    {
+      "epoch": 0.9567519858781994,
+      "grad_norm": 0.7176869974666727,
+      "learning_rate": 9.887139927584453e-06,
+      "loss": 0.1665,
+      "step": 271
+    },
+    {
+      "epoch": 0.9602824360105914,
+      "grad_norm": 0.924669948462229,
+      "learning_rate": 9.885927764139633e-06,
+      "loss": 0.2065,
+      "step": 272
+    },
+    {
+      "epoch": 0.9638128861429832,
+      "grad_norm": 0.7237412872405012,
+      "learning_rate": 9.88470920095906e-06,
+      "loss": 0.1802,
+      "step": 273
+    },
+    {
+      "epoch": 0.9673433362753752,
+      "grad_norm": 0.6998624857232224,
+      "learning_rate": 9.883484239638842e-06,
+      "loss": 0.1642,
+      "step": 274
+    },
+    {
+      "epoch": 0.970873786407767,
+      "grad_norm": 0.8569858798701728,
+      "learning_rate": 9.882252881783472e-06,
+      "loss": 0.1956,
+      "step": 275
+    },
+    {
+      "epoch": 0.9744042365401588,
+      "grad_norm": 0.7621961932465209,
+      "learning_rate": 9.881015129005822e-06,
+      "loss": 0.1841,
+      "step": 276
+    },
+    {
+      "epoch": 0.9779346866725508,
+      "grad_norm": 0.7592469083950414,
+      "learning_rate": 9.879770982927135e-06,
+      "loss": 0.1937,
+      "step": 277
+    },
+    {
+      "epoch": 0.9814651368049426,
+      "grad_norm": 0.6865266207305956,
+      "learning_rate": 9.87852044517703e-06,
+      "loss": 0.1957,
+      "step": 278
+    },
+    {
+      "epoch": 0.9849955869373345,
+      "grad_norm": 0.7228023197310592,
+      "learning_rate": 9.877263517393503e-06,
+      "loss": 0.1786,
+      "step": 279
+    },
+    {
+      "epoch": 0.9885260370697264,
+      "grad_norm": 0.9465468675591545,
+      "learning_rate": 9.876000201222912e-06,
+      "loss": 0.1661,
+      "step": 280
+    },
+    {
+      "epoch": 0.9920564872021183,
+      "grad_norm": 0.8521304202598817,
+      "learning_rate": 9.87473049831999e-06,
+      "loss": 0.154,
+      "step": 281
+    },
+    {
+      "epoch": 0.9955869373345102,
+      "grad_norm": 0.669879136764297,
+      "learning_rate": 9.87345441034783e-06,
+      "loss": 0.1624,
+      "step": 282
+    },
+    {
+      "epoch": 0.999117387466902,
+      "grad_norm": 0.9684619377618936,
+      "learning_rate": 9.872171938977895e-06,
+      "loss": 0.1881,
+      "step": 283
+    },
+    {
+      "epoch": 1.002647837599294,
+      "grad_norm": 0.85439046023442,
+      "learning_rate": 9.870883085889997e-06,
+      "loss": 0.1865,
+      "step": 284
+    },
+    {
+      "epoch": 1.0061782877316858,
+      "grad_norm": 0.8191806695595788,
+      "learning_rate": 9.86958785277232e-06,
+      "loss": 0.163,
+      "step": 285
+    },
+    {
+      "epoch": 1.0097087378640777,
+      "grad_norm": 0.9382251565209051,
+      "learning_rate": 9.8682862413214e-06,
+      "loss": 0.1758,
+      "step": 286
+    },
+    {
+      "epoch": 1.0132391879964695,
+      "grad_norm": 0.5193721161505495,
+      "learning_rate": 9.866978253242123e-06,
+      "loss": 0.1751,
+      "step": 287
+    },
+    {
+      "epoch": 1.0167696381288613,
+      "grad_norm": 0.8198070617026992,
+      "learning_rate": 9.865663890247735e-06,
+      "loss": 0.1535,
+      "step": 288
+    },
+    {
+      "epoch": 1.0203000882612534,
+      "grad_norm": 0.9683137087655609,
+      "learning_rate": 9.864343154059825e-06,
+      "loss": 0.193,
+      "step": 289
+    },
+    {
+      "epoch": 1.0238305383936452,
+      "grad_norm": 1.0396575123509755,
+      "learning_rate": 9.863016046408336e-06,
+      "loss": 0.1584,
+      "step": 290
+    },
+    {
+      "epoch": 1.027360988526037,
+      "grad_norm": 0.8171505980553876,
+      "learning_rate": 9.861682569031553e-06,
+      "loss": 0.1681,
+      "step": 291
+    },
+    {
+      "epoch": 1.030891438658429,
+      "grad_norm": 0.6927072039886698,
+      "learning_rate": 9.860342723676105e-06,
+      "loss": 0.1841,
+      "step": 292
+    },
+    {
+      "epoch": 1.0344218887908208,
+      "grad_norm": 0.9193349315868659,
+      "learning_rate": 9.85899651209696e-06,
+      "loss": 0.1786,
+      "step": 293
+    },
+    {
+      "epoch": 1.0379523389232128,
+      "grad_norm": 0.6395800368416922,
+      "learning_rate": 9.857643936057431e-06,
+      "loss": 0.1714,
+      "step": 294
+    },
+    {
+      "epoch": 1.0414827890556047,
+      "grad_norm": 0.7076027666270328,
+      "learning_rate": 9.856284997329158e-06,
+      "loss": 0.1628,
+      "step": 295
+    },
+    {
+      "epoch": 1.0450132391879965,
+      "grad_norm": 0.5571080103911318,
+      "learning_rate": 9.854919697692123e-06,
+      "loss": 0.1765,
+      "step": 296
+    },
+    {
+      "epoch": 1.0485436893203883,
+      "grad_norm": 0.8245809567555668,
+      "learning_rate": 9.853548038934635e-06,
+      "loss": 0.178,
+      "step": 297
+    },
+    {
+      "epoch": 1.0520741394527802,
+      "grad_norm": 1.3356935667806615,
+      "learning_rate": 9.852170022853335e-06,
+      "loss": 0.1629,
+      "step": 298
+    },
+    {
+      "epoch": 1.055604589585172,
+      "grad_norm": 0.5648163532404421,
+      "learning_rate": 9.85078565125319e-06,
+      "loss": 0.1638,
+      "step": 299
+    },
+    {
+      "epoch": 1.059135039717564,
+      "grad_norm": 0.8602300175351031,
+      "learning_rate": 9.849394925947493e-06,
+      "loss": 0.1801,
+      "step": 300
+    },
+    {
+      "epoch": 1.062665489849956,
+      "grad_norm": 0.8714458340169661,
+      "learning_rate": 9.847997848757855e-06,
+      "loss": 0.1494,
+      "step": 301
+    },
+    {
+      "epoch": 1.0661959399823477,
+      "grad_norm": 0.7793786634487098,
+      "learning_rate": 9.846594421514213e-06,
+      "loss": 0.1627,
+      "step": 302
+    },
+    {
+      "epoch": 1.0697263901147396,
+      "grad_norm": 0.7080125846941803,
+      "learning_rate": 9.845184646054817e-06,
+      "loss": 0.1776,
+      "step": 303
+    },
+    {
+      "epoch": 1.0732568402471314,
+      "grad_norm": 0.6880805275018903,
+      "learning_rate": 9.843768524226233e-06,
+      "loss": 0.1597,
+      "step": 304
+    },
+    {
+      "epoch": 1.0767872903795235,
+      "grad_norm": 0.7703863699473644,
+      "learning_rate": 9.842346057883341e-06,
+      "loss": 0.1603,
+      "step": 305
+    },
+    {
+      "epoch": 1.0803177405119153,
+      "grad_norm": 0.833743875181872,
+      "learning_rate": 9.84091724888933e-06,
+      "loss": 0.1613,
+      "step": 306
+    },
+    {
+      "epoch": 1.0838481906443072,
+      "grad_norm": 0.7125550254627817,
+      "learning_rate": 9.839482099115696e-06,
+      "loss": 0.1511,
+      "step": 307
+    },
+    {
+      "epoch": 1.087378640776699,
+      "grad_norm": 0.6773769437101177,
+      "learning_rate": 9.838040610442243e-06,
+      "loss": 0.1578,
+      "step": 308
+    },
+    {
+      "epoch": 1.0909090909090908,
+      "grad_norm": 0.7853675801397739,
+      "learning_rate": 9.836592784757077e-06,
+      "loss": 0.1548,
+      "step": 309
+    },
+    {
+      "epoch": 1.0944395410414829,
+      "grad_norm": 0.9629392348111224,
+      "learning_rate": 9.835138623956603e-06,
+      "loss": 0.1787,
+      "step": 310
+    },
+    {
+      "epoch": 1.0979699911738747,
+      "grad_norm": 1.0276282328498774,
+      "learning_rate": 9.833678129945524e-06,
+      "loss": 0.1798,
+      "step": 311
+    },
+    {
+      "epoch": 1.1015004413062666,
+      "grad_norm": 0.9134089726849168,
+      "learning_rate": 9.83221130463684e-06,
+      "loss": 0.1888,
+      "step": 312
+    },
+    {
+      "epoch": 1.1050308914386584,
+      "grad_norm": 0.9917300375319624,
+      "learning_rate": 9.830738149951843e-06,
+      "loss": 0.1884,
+      "step": 313
+    },
+    {
+      "epoch": 1.1085613415710502,
+      "grad_norm": 0.7089950679554146,
+      "learning_rate": 9.829258667820114e-06,
+      "loss": 0.1686,
+      "step": 314
+    },
+    {
+      "epoch": 1.1120917917034423,
+      "grad_norm": 0.578692441690528,
+      "learning_rate": 9.827772860179528e-06,
+      "loss": 0.1804,
+      "step": 315
+    },
+    {
+      "epoch": 1.1156222418358341,
+      "grad_norm": 0.739188688334005,
+      "learning_rate": 9.826280728976234e-06,
+      "loss": 0.1699,
+      "step": 316
+    },
+    {
+      "epoch": 1.119152691968226,
+      "grad_norm": 1.0203778463290891,
+      "learning_rate": 9.824782276164678e-06,
+      "loss": 0.1679,
+      "step": 317
+    },
+    {
+      "epoch": 1.1226831421006178,
+      "grad_norm": 0.9103291412358969,
+      "learning_rate": 9.823277503707571e-06,
+      "loss": 0.1885,
+      "step": 318
+    },
+    {
+      "epoch": 1.1262135922330097,
+      "grad_norm": 1.1240057634668683,
+      "learning_rate": 9.821766413575915e-06,
+      "loss": 0.1467,
+      "step": 319
+    },
+    {
+      "epoch": 1.1297440423654015,
+      "grad_norm": 0.7591033795680634,
+      "learning_rate": 9.820249007748978e-06,
+      "loss": 0.1708,
+      "step": 320
+    },
+    {
+      "epoch": 1.1332744924977936,
+      "grad_norm": 0.6461323484499221,
+      "learning_rate": 9.818725288214309e-06,
+      "loss": 0.1576,
+      "step": 321
+    },
+    {
+      "epoch": 1.1368049426301854,
+      "grad_norm": 0.8717085794666202,
+      "learning_rate": 9.817195256967715e-06,
+      "loss": 0.1846,
+      "step": 322
+    },
+    {
+      "epoch": 1.1403353927625772,
+      "grad_norm": 0.5766027007180119,
+      "learning_rate": 9.815658916013283e-06,
+      "loss": 0.1502,
+      "step": 323
+    },
+    {
+      "epoch": 1.143865842894969,
+      "grad_norm": 0.8409708523916547,
+      "learning_rate": 9.814116267363355e-06,
+      "loss": 0.1522,
+      "step": 324
+    },
+    {
+      "epoch": 1.147396293027361,
+      "grad_norm": 0.9388226084700922,
+      "learning_rate": 9.812567313038542e-06,
+      "loss": 0.1681,
+      "step": 325
+    },
+    {
+      "epoch": 1.150926743159753,
+      "grad_norm": 0.9369552573302047,
+      "learning_rate": 9.811012055067708e-06,
+      "loss": 0.1905,
+      "step": 326
+    },
+    {
+      "epoch": 1.1544571932921448,
+      "grad_norm": 1.0468936130746385,
+      "learning_rate": 9.809450495487977e-06,
+      "loss": 0.1838,
+      "step": 327
+    },
+    {
+      "epoch": 1.1579876434245366,
+      "grad_norm": 0.8466637747576687,
+      "learning_rate": 9.80788263634473e-06,
+      "loss": 0.1536,
+      "step": 328
+    },
+    {
+      "epoch": 1.1615180935569285,
+      "grad_norm": 0.8133260085099403,
+      "learning_rate": 9.806308479691595e-06,
+      "loss": 0.185,
+      "step": 329
+    },
+    {
+      "epoch": 1.1650485436893203,
+      "grad_norm": 0.634815173007273,
+      "learning_rate": 9.80472802759045e-06,
+      "loss": 0.1763,
+      "step": 330
+    },
+    {
+      "epoch": 1.1685789938217122,
+      "grad_norm": 0.7155087062007245,
+      "learning_rate": 9.803141282111414e-06,
+      "loss": 0.1689,
+      "step": 331
+    },
+    {
+      "epoch": 1.1721094439541042,
+      "grad_norm": 0.918184676277347,
+      "learning_rate": 9.801548245332863e-06,
+      "loss": 0.1771,
+      "step": 332
+    },
+    {
+      "epoch": 1.175639894086496,
+      "grad_norm": 0.847842933527483,
+      "learning_rate": 9.799948919341398e-06,
+      "loss": 0.1982,
+      "step": 333
+    },
+    {
+      "epoch": 1.179170344218888,
+      "grad_norm": 0.950371294714106,
+      "learning_rate": 9.798343306231869e-06,
+      "loss": 0.1715,
+      "step": 334
+    },
+    {
+      "epoch": 1.1827007943512797,
+      "grad_norm": 0.8452751079657315,
+      "learning_rate": 9.796731408107349e-06,
+      "loss": 0.1898,
+      "step": 335
+    },
+    {
+      "epoch": 1.1862312444836718,
+      "grad_norm": 1.0132828873057207,
+      "learning_rate": 9.79511322707916e-06,
+      "loss": 0.1832,
+      "step": 336
+    },
+    {
+      "epoch": 1.1897616946160636,
+      "grad_norm": 0.6796613513123204,
+      "learning_rate": 9.793488765266838e-06,
+      "loss": 0.1795,
+      "step": 337
+    },
+    {
+      "epoch": 1.1932921447484555,
+      "grad_norm": 0.7516255039006127,
+      "learning_rate": 9.791858024798157e-06,
+      "loss": 0.1682,
+      "step": 338
+    },
+    {
+      "epoch": 1.1968225948808473,
+      "grad_norm": 0.6882391031969943,
+      "learning_rate": 9.790221007809106e-06,
+      "loss": 0.158,
+      "step": 339
+    },
+    {
+      "epoch": 1.2003530450132391,
+      "grad_norm": 0.7235480772368882,
+      "learning_rate": 9.788577716443903e-06,
+      "loss": 0.1685,
+      "step": 340
+    },
+    {
+      "epoch": 1.203883495145631,
+      "grad_norm": 0.8054680221859308,
+      "learning_rate": 9.786928152854977e-06,
+      "loss": 0.1679,
+      "step": 341
+    },
+    {
+      "epoch": 1.207413945278023,
+      "grad_norm": 0.6537575157321099,
+      "learning_rate": 9.785272319202978e-06,
+      "loss": 0.1535,
+      "step": 342
+    },
+    {
+      "epoch": 1.2109443954104149,
+      "grad_norm": 0.8631226400216473,
+      "learning_rate": 9.783610217656766e-06,
+      "loss": 0.1686,
+      "step": 343
+    },
+    {
+      "epoch": 1.2144748455428067,
+      "grad_norm": 0.7074978790315094,
+      "learning_rate": 9.781941850393412e-06,
+      "loss": 0.1557,
+      "step": 344
+    },
+    {
+      "epoch": 1.2180052956751986,
+      "grad_norm": 0.7508923027783329,
+      "learning_rate": 9.780267219598196e-06,
+      "loss": 0.1734,
+      "step": 345
+    },
+    {
+      "epoch": 1.2215357458075904,
+      "grad_norm": 0.6041756898423745,
+      "learning_rate": 9.7785863274646e-06,
+      "loss": 0.1676,
+      "step": 346
+    },
+    {
+      "epoch": 1.2250661959399824,
+      "grad_norm": 0.8916407715059216,
+      "learning_rate": 9.776899176194302e-06,
+      "loss": 0.1612,
+      "step": 347
+    },
+    {
+      "epoch": 1.2285966460723743,
+      "grad_norm": 0.9080331160273555,
+      "learning_rate": 9.775205767997188e-06,
+      "loss": 0.1718,
+      "step": 348
+    },
+    {
+      "epoch": 1.2321270962047661,
+      "grad_norm": 0.844342085425393,
+      "learning_rate": 9.773506105091336e-06,
+      "loss": 0.1673,
+      "step": 349
+    },
+    {
+      "epoch": 1.235657546337158,
+      "grad_norm": 0.7564502855821623,
+      "learning_rate": 9.771800189703014e-06,
+      "loss": 0.1535,
+      "step": 350
+    },
+    {
+      "epoch": 1.2391879964695498,
+      "grad_norm": 0.6268844715263006,
+      "learning_rate": 9.770088024066681e-06,
+      "loss": 0.147,
+      "step": 351
+    },
+    {
+      "epoch": 1.2427184466019416,
+      "grad_norm": 0.7824396074378344,
+      "learning_rate": 9.768369610424984e-06,
+      "loss": 0.1865,
+      "step": 352
+    },
+    {
+      "epoch": 1.2462488967343337,
+      "grad_norm": 0.7832814173090706,
+      "learning_rate": 9.766644951028755e-06,
+      "loss": 0.1792,
+      "step": 353
+    },
+    {
+      "epoch": 1.2497793468667255,
+      "grad_norm": 0.9379854009237389,
+      "learning_rate": 9.764914048137001e-06,
+      "loss": 0.1657,
+      "step": 354
+    },
+    {
+      "epoch": 1.2533097969991174,
+      "grad_norm": 0.8058449874416329,
+      "learning_rate": 9.763176904016914e-06,
+      "loss": 0.1718,
+      "step": 355
+    },
+    {
+      "epoch": 1.2568402471315092,
+      "grad_norm": 0.8085192046750732,
+      "learning_rate": 9.761433520943856e-06,
+      "loss": 0.1776,
+      "step": 356
+    },
+    {
+      "epoch": 1.2603706972639013,
+      "grad_norm": 0.9591752385660486,
+      "learning_rate": 9.759683901201364e-06,
+      "loss": 0.1979,
+      "step": 357
+    },
+    {
+      "epoch": 1.2639011473962931,
+      "grad_norm": 1.0461750891605788,
+      "learning_rate": 9.757928047081144e-06,
+      "loss": 0.1926,
+      "step": 358
+    },
+    {
+      "epoch": 1.267431597528685,
+      "grad_norm": 0.7954582531988508,
+      "learning_rate": 9.756165960883063e-06,
+      "loss": 0.1648,
+      "step": 359
+    },
+    {
+      "epoch": 1.2709620476610768,
+      "grad_norm": 0.8863013446887811,
+      "learning_rate": 9.75439764491516e-06,
+      "loss": 0.1807,
+      "step": 360
+    },
+    {
+      "epoch": 1.2744924977934686,
+      "grad_norm": 0.8321220073097337,
+      "learning_rate": 9.752623101493624e-06,
+      "loss": 0.1771,
+      "step": 361
+    },
+    {
+      "epoch": 1.2780229479258605,
+      "grad_norm": 0.8204672244613037,
+      "learning_rate": 9.750842332942805e-06,
+      "loss": 0.1669,
+      "step": 362
+    },
+    {
+      "epoch": 1.2815533980582523,
+      "grad_norm": 0.8746356076847612,
+      "learning_rate": 9.749055341595213e-06,
+      "loss": 0.1511,
+      "step": 363
+    },
+    {
+      "epoch": 1.2850838481906444,
+      "grad_norm": 0.7387304926466468,
+      "learning_rate": 9.747262129791497e-06,
+      "loss": 0.1696,
+      "step": 364
+    },
+    {
+      "epoch": 1.2886142983230362,
+      "grad_norm": 0.6434577859365603,
+      "learning_rate": 9.745462699880464e-06,
+      "loss": 0.1617,
+      "step": 365
+    },
+    {
+      "epoch": 1.292144748455428,
+      "grad_norm": 0.8994503573558672,
+      "learning_rate": 9.74365705421906e-06,
+      "loss": 0.1771,
+      "step": 366
+    },
+    {
+      "epoch": 1.2956751985878199,
+      "grad_norm": 0.9104255097758082,
+      "learning_rate": 9.741845195172376e-06,
+      "loss": 0.1507,
+      "step": 367
+    },
+    {
+      "epoch": 1.299205648720212,
+      "grad_norm": 0.8237793644377707,
+      "learning_rate": 9.740027125113638e-06,
+      "loss": 0.1619,
+      "step": 368
+    },
+    {
+      "epoch": 1.3027360988526038,
+      "grad_norm": 0.7759651867155161,
+      "learning_rate": 9.73820284642421e-06,
+      "loss": 0.1671,
+      "step": 369
+    },
+    {
+      "epoch": 1.3062665489849956,
+      "grad_norm": 0.8529971231768904,
+      "learning_rate": 9.736372361493584e-06,
+      "loss": 0.1601,
+      "step": 370
+    },
+    {
+      "epoch": 1.3097969991173875,
+      "grad_norm": 0.6232528534187706,
+      "learning_rate": 9.734535672719391e-06,
+      "loss": 0.1662,
+      "step": 371
+    },
+    {
+      "epoch": 1.3133274492497793,
+      "grad_norm": 0.9946599681832587,
+      "learning_rate": 9.732692782507376e-06,
+      "loss": 0.1687,
+      "step": 372
+    },
+    {
+      "epoch": 1.3168578993821711,
+      "grad_norm": 0.7623188795800929,
+      "learning_rate": 9.730843693271413e-06,
+      "loss": 0.1683,
+      "step": 373
+    },
+    {
+      "epoch": 1.3203883495145632,
+      "grad_norm": 0.5696437347911626,
+      "learning_rate": 9.728988407433497e-06,
+      "loss": 0.1555,
+      "step": 374
+    },
+    {
+      "epoch": 1.323918799646955,
+      "grad_norm": 0.7695474373449546,
+      "learning_rate": 9.727126927423736e-06,
+      "loss": 0.1891,
+      "step": 375
+    },
+    {
+      "epoch": 1.3274492497793469,
+      "grad_norm": 0.7640349287167723,
+      "learning_rate": 9.725259255680352e-06,
+      "loss": 0.1822,
+      "step": 376
+    },
+    {
+      "epoch": 1.3309796999117387,
+      "grad_norm": 0.7909928993460938,
+      "learning_rate": 9.72338539464968e-06,
+      "loss": 0.1742,
+      "step": 377
+    },
+    {
+      "epoch": 1.3345101500441308,
+      "grad_norm": 1.1113868103845201,
+      "learning_rate": 9.72150534678616e-06,
+      "loss": 0.1647,
+      "step": 378
+    },
+    {
+      "epoch": 1.3380406001765226,
+      "grad_norm": 1.0428448743119225,
+      "learning_rate": 9.719619114552332e-06,
+      "loss": 0.1354,
+      "step": 379
+    },
+    {
+      "epoch": 1.3415710503089144,
+      "grad_norm": 0.8715636876334808,
+      "learning_rate": 9.717726700418842e-06,
+      "loss": 0.1772,
+      "step": 380
+    },
+    {
+      "epoch": 1.3451015004413063,
+      "grad_norm": 0.878842141186717,
+      "learning_rate": 9.715828106864434e-06,
+      "loss": 0.1657,
+      "step": 381
+    },
+    {
+      "epoch": 1.3486319505736981,
+      "grad_norm": 0.6697201435871655,
+      "learning_rate": 9.713923336375936e-06,
+      "loss": 0.1585,
+      "step": 382
+    },
+    {
+      "epoch": 1.35216240070609,
+      "grad_norm": 0.9696617071339634,
+      "learning_rate": 9.712012391448283e-06,
+      "loss": 0.1939,
+      "step": 383
+    },
+    {
+      "epoch": 1.3556928508384818,
+      "grad_norm": 0.8187281006125846,
+      "learning_rate": 9.710095274584482e-06,
+      "loss": 0.1653,
+      "step": 384
+    },
+    {
+      "epoch": 1.3592233009708738,
+      "grad_norm": 0.8974963672760499,
+      "learning_rate": 9.70817198829563e-06,
+      "loss": 0.1737,
+      "step": 385
+    },
+    {
+      "epoch": 1.3627537511032657,
+      "grad_norm": 0.7456778356448868,
+      "learning_rate": 9.70624253510091e-06,
+      "loss": 0.1635,
+      "step": 386
+    },
+    {
+      "epoch": 1.3662842012356575,
+      "grad_norm": 1.1629774820732057,
+      "learning_rate": 9.704306917527575e-06,
+      "loss": 0.2045,
+      "step": 387
+    },
+    {
+      "epoch": 1.3698146513680494,
+      "grad_norm": 0.8069603325247644,
+      "learning_rate": 9.702365138110958e-06,
+      "loss": 0.1603,
+      "step": 388
+    },
+    {
+      "epoch": 1.3733451015004414,
+      "grad_norm": 0.9350513325401503,
+      "learning_rate": 9.700417199394457e-06,
+      "loss": 0.1764,
+      "step": 389
+    },
+    {
+      "epoch": 1.3768755516328333,
+      "grad_norm": 0.5965930357229629,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.1688,
+      "step": 390
+    },
+    {
+      "epoch": 1.380406001765225,
+      "grad_norm": 0.9881802502016954,
+      "learning_rate": 9.69650285427575e-06,
+      "loss": 0.1996,
+      "step": 391
+    },
+    {
+      "epoch": 1.383936451897617,
+      "grad_norm": 0.6943128810137366,
+      "learning_rate": 9.694536453000668e-06,
+      "loss": 0.1685,
+      "step": 392
+    },
+    {
+      "epoch": 1.3874669020300088,
+      "grad_norm": 0.6982977323410529,
+      "learning_rate": 9.692563902679954e-06,
+      "loss": 0.1563,
+      "step": 393
+    },
+    {
+      "epoch": 1.3909973521624006,
+      "grad_norm": 0.7295520907046743,
+      "learning_rate": 9.69058520589731e-06,
+      "loss": 0.1692,
+      "step": 394
+    },
+    {
+      "epoch": 1.3945278022947925,
+      "grad_norm": 0.80781982226485,
+      "learning_rate": 9.688600365244495e-06,
+      "loss": 0.1762,
+      "step": 395
+    },
+    {
+      "epoch": 1.3980582524271845,
+      "grad_norm": 0.6182663519447118,
+      "learning_rate": 9.686609383321316e-06,
+      "loss": 0.1686,
+      "step": 396
+    },
+    {
+      "epoch": 1.4015887025595763,
+      "grad_norm": 0.5145110224464513,
+      "learning_rate": 9.684612262735614e-06,
+      "loss": 0.1626,
+      "step": 397
+    },
+    {
+      "epoch": 1.4051191526919682,
+      "grad_norm": 0.8760399209586848,
+      "learning_rate": 9.682609006103283e-06,
+      "loss": 0.1754,
+      "step": 398
+    },
+    {
+      "epoch": 1.40864960282436,
+      "grad_norm": 0.8161048235153104,
+      "learning_rate": 9.680599616048247e-06,
+      "loss": 0.164,
+      "step": 399
+    },
+    {
+      "epoch": 1.412180052956752,
+      "grad_norm": 0.7009270806175059,
+      "learning_rate": 9.678584095202468e-06,
+      "loss": 0.1613,
+      "step": 400
+    },
+    {
+      "epoch": 1.415710503089144,
+      "grad_norm": 0.6567037684930597,
+      "learning_rate": 9.676562446205936e-06,
+      "loss": 0.1498,
+      "step": 401
+    },
+    {
+      "epoch": 1.4192409532215358,
+      "grad_norm": 0.7885493080145554,
+      "learning_rate": 9.674534671706666e-06,
+      "loss": 0.1834,
+      "step": 402
+    },
+    {
+      "epoch": 1.4227714033539276,
+      "grad_norm": 0.561383249284057,
+      "learning_rate": 9.672500774360698e-06,
+      "loss": 0.1664,
+      "step": 403
+    },
+    {
+      "epoch": 1.4263018534863194,
+      "grad_norm": 0.8130649369686304,
+      "learning_rate": 9.670460756832094e-06,
+      "loss": 0.1375,
+      "step": 404
+    },
+    {
+      "epoch": 1.4298323036187113,
+      "grad_norm": 0.730228217232664,
+      "learning_rate": 9.668414621792928e-06,
+      "loss": 0.1602,
+      "step": 405
+    },
+    {
+      "epoch": 1.4333627537511033,
+      "grad_norm": 0.701767347722211,
+      "learning_rate": 9.666362371923294e-06,
+      "loss": 0.1619,
+      "step": 406
+    },
+    {
+      "epoch": 1.4368932038834952,
+      "grad_norm": 0.6142164548390184,
+      "learning_rate": 9.664304009911286e-06,
+      "loss": 0.1578,
+      "step": 407
+    },
+    {
+      "epoch": 1.440423654015887,
+      "grad_norm": 1.1635404922505252,
+      "learning_rate": 9.66223953845301e-06,
+      "loss": 0.1835,
+      "step": 408
+    },
+    {
+      "epoch": 1.4439541041482788,
+      "grad_norm": 0.8373651861820354,
+      "learning_rate": 9.660168960252575e-06,
+      "loss": 0.17,
+      "step": 409
+    },
+    {
+      "epoch": 1.447484554280671,
+      "grad_norm": 0.5626164848515763,
+      "learning_rate": 9.658092278022088e-06,
+      "loss": 0.1638,
+      "step": 410
+    },
+    {
+      "epoch": 1.4510150044130627,
+      "grad_norm": 0.9647932895601038,
+      "learning_rate": 9.656009494481646e-06,
+      "loss": 0.1737,
+      "step": 411
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 0.6769901378936639,
+      "learning_rate": 9.653920612359344e-06,
+      "loss": 0.1719,
+      "step": 412
+    },
+    {
+      "epoch": 1.4580759046778464,
+      "grad_norm": 0.5912616414391476,
+      "learning_rate": 9.651825634391262e-06,
+      "loss": 0.1547,
+      "step": 413
+    },
+    {
+      "epoch": 1.4616063548102383,
+      "grad_norm": 0.6130890022296159,
+      "learning_rate": 9.649724563321465e-06,
+      "loss": 0.1509,
+      "step": 414
+    },
+    {
+      "epoch": 1.46513680494263,
+      "grad_norm": 0.6981352025940074,
+      "learning_rate": 9.647617401902003e-06,
+      "loss": 0.1612,
+      "step": 415
+    },
+    {
+      "epoch": 1.468667255075022,
+      "grad_norm": 0.7832373213646434,
+      "learning_rate": 9.645504152892895e-06,
+      "loss": 0.1867,
+      "step": 416
+    },
+    {
+      "epoch": 1.472197705207414,
+      "grad_norm": 0.8544155478419707,
+      "learning_rate": 9.64338481906214e-06,
+      "loss": 0.1721,
+      "step": 417
+    },
+    {
+      "epoch": 1.4757281553398058,
+      "grad_norm": 0.9502151610480191,
+      "learning_rate": 9.641259403185706e-06,
+      "loss": 0.1743,
+      "step": 418
+    },
+    {
+      "epoch": 1.4792586054721977,
+      "grad_norm": 0.7733505393875248,
+      "learning_rate": 9.639127908047527e-06,
+      "loss": 0.1734,
+      "step": 419
+    },
+    {
+      "epoch": 1.4827890556045895,
+      "grad_norm": 0.8264946927060004,
+      "learning_rate": 9.6369903364395e-06,
+      "loss": 0.1911,
+      "step": 420
+    },
+    {
+      "epoch": 1.4863195057369816,
+      "grad_norm": 0.8237623458796101,
+      "learning_rate": 9.63484669116148e-06,
+      "loss": 0.169,
+      "step": 421
+    },
+    {
+      "epoch": 1.4898499558693734,
+      "grad_norm": 1.0384652928610334,
+      "learning_rate": 9.632696975021278e-06,
+      "loss": 0.1761,
+      "step": 422
+    },
+    {
+      "epoch": 1.4933804060017652,
+      "grad_norm": 0.9824278139544069,
+      "learning_rate": 9.630541190834659e-06,
+      "loss": 0.1854,
+      "step": 423
+    },
+    {
+      "epoch": 1.496910856134157,
+      "grad_norm": 0.7655842553137074,
+      "learning_rate": 9.628379341425332e-06,
+      "loss": 0.1762,
+      "step": 424
+    },
+    {
+      "epoch": 1.500441306266549,
+      "grad_norm": 0.7573467640480261,
+      "learning_rate": 9.626211429624955e-06,
+      "loss": 0.1719,
+      "step": 425
+    },
+    {
+      "epoch": 1.5039717563989408,
+      "grad_norm": 0.6342899695339446,
+      "learning_rate": 9.624037458273121e-06,
+      "loss": 0.1663,
+      "step": 426
+    },
+    {
+      "epoch": 1.5075022065313326,
+      "grad_norm": 0.7241886079251669,
+      "learning_rate": 9.621857430217366e-06,
+      "loss": 0.1612,
+      "step": 427
+    },
+    {
+      "epoch": 1.5110326566637247,
+      "grad_norm": 0.7713579132775925,
+      "learning_rate": 9.619671348313156e-06,
+      "loss": 0.1414,
+      "step": 428
+    },
+    {
+      "epoch": 1.5145631067961165,
+      "grad_norm": 0.6070241168065806,
+      "learning_rate": 9.617479215423887e-06,
+      "loss": 0.1752,
+      "step": 429
+    },
+    {
+      "epoch": 1.5180935569285083,
+      "grad_norm": 0.8462668684183862,
+      "learning_rate": 9.615281034420882e-06,
+      "loss": 0.1769,
+      "step": 430
+    },
+    {
+      "epoch": 1.5216240070609004,
+      "grad_norm": 0.7218104922456425,
+      "learning_rate": 9.613076808183382e-06,
+      "loss": 0.1531,
+      "step": 431
+    },
+    {
+      "epoch": 1.5251544571932922,
+      "grad_norm": 0.7261686662514826,
+      "learning_rate": 9.610866539598551e-06,
+      "loss": 0.1638,
+      "step": 432
+    },
+    {
+      "epoch": 1.528684907325684,
+      "grad_norm": 1.2087576640937627,
+      "learning_rate": 9.608650231561466e-06,
+      "loss": 0.1848,
+      "step": 433
+    },
+    {
+      "epoch": 1.532215357458076,
+      "grad_norm": 0.6381008529495783,
+      "learning_rate": 9.606427886975113e-06,
+      "loss": 0.1719,
+      "step": 434
+    },
+    {
+      "epoch": 1.5357458075904677,
+      "grad_norm": 0.9801974645943355,
+      "learning_rate": 9.604199508750386e-06,
+      "loss": 0.189,
+      "step": 435
+    },
+    {
+      "epoch": 1.5392762577228596,
+      "grad_norm": 0.7249244172345439,
+      "learning_rate": 9.601965099806085e-06,
+      "loss": 0.1729,
+      "step": 436
+    },
+    {
+      "epoch": 1.5428067078552514,
+      "grad_norm": 0.852227223208167,
+      "learning_rate": 9.599724663068904e-06,
+      "loss": 0.1522,
+      "step": 437
+    },
+    {
+      "epoch": 1.5463371579876433,
+      "grad_norm": 0.70519515152008,
+      "learning_rate": 9.597478201473433e-06,
+      "loss": 0.1682,
+      "step": 438
+    },
+    {
+      "epoch": 1.5498676081200353,
+      "grad_norm": 0.7777614025947788,
+      "learning_rate": 9.595225717962157e-06,
+      "loss": 0.16,
+      "step": 439
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 0.6144264510883923,
+      "learning_rate": 9.592967215485449e-06,
+      "loss": 0.1737,
+      "step": 440
+    },
+    {
+      "epoch": 1.5569285083848192,
+      "grad_norm": 0.6723553147570778,
+      "learning_rate": 9.590702697001559e-06,
+      "loss": 0.1864,
+      "step": 441
+    },
+    {
+      "epoch": 1.560458958517211,
+      "grad_norm": 0.786872255669317,
+      "learning_rate": 9.588432165476626e-06,
+      "loss": 0.1678,
+      "step": 442
+    },
+    {
+      "epoch": 1.563989408649603,
+      "grad_norm": 1.060953562194783,
+      "learning_rate": 9.58615562388466e-06,
+      "loss": 0.1743,
+      "step": 443
+    },
+    {
+      "epoch": 1.5675198587819947,
+      "grad_norm": 0.8590680802220704,
+      "learning_rate": 9.583873075207541e-06,
+      "loss": 0.1449,
+      "step": 444
+    },
+    {
+      "epoch": 1.5710503089143866,
+      "grad_norm": 0.8592072746191091,
+      "learning_rate": 9.581584522435025e-06,
+      "loss": 0.1689,
+      "step": 445
+    },
+    {
+      "epoch": 1.5745807590467784,
+      "grad_norm": 0.6431750629028619,
+      "learning_rate": 9.579289968564723e-06,
+      "loss": 0.1549,
+      "step": 446
+    },
+    {
+      "epoch": 1.5781112091791702,
+      "grad_norm": 0.7880523755608316,
+      "learning_rate": 9.576989416602113e-06,
+      "loss": 0.1569,
+      "step": 447
+    },
+    {
+      "epoch": 1.581641659311562,
+      "grad_norm": 0.7826920782979999,
+      "learning_rate": 9.574682869560528e-06,
+      "loss": 0.1655,
+      "step": 448
+    },
+    {
+      "epoch": 1.5851721094439541,
+      "grad_norm": 0.6507531502926956,
+      "learning_rate": 9.57237033046115e-06,
+      "loss": 0.1704,
+      "step": 449
+    },
+    {
+      "epoch": 1.588702559576346,
+      "grad_norm": 0.6131816635944837,
+      "learning_rate": 9.570051802333017e-06,
+      "loss": 0.1722,
+      "step": 450
+    },
+    {
+      "epoch": 1.5922330097087378,
+      "grad_norm": 0.8615528127088715,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.1671,
+      "step": 451
+    },
+    {
+      "epoch": 1.5957634598411299,
+      "grad_norm": 0.7175178771131497,
+      "learning_rate": 9.565396791145833e-06,
+      "loss": 0.1646,
+      "step": 452
+    },
+    {
+      "epoch": 1.5992939099735217,
+      "grad_norm": 0.6985637923301037,
+      "learning_rate": 9.563060314184055e-06,
+      "loss": 0.1919,
+      "step": 453
+    },
+    {
+      "epoch": 1.6028243601059136,
+      "grad_norm": 0.854067298117589,
+      "learning_rate": 9.560717860388061e-06,
+      "loss": 0.1819,
+      "step": 454
+    },
+    {
+      "epoch": 1.6063548102383054,
+      "grad_norm": 0.884963758284598,
+      "learning_rate": 9.558369432826067e-06,
+      "loss": 0.169,
+      "step": 455
+    },
+    {
+      "epoch": 1.6098852603706972,
+      "grad_norm": 0.7696374374267597,
+      "learning_rate": 9.556015034574114e-06,
+      "loss": 0.1538,
+      "step": 456
+    },
+    {
+      "epoch": 1.613415710503089,
+      "grad_norm": 0.6854959076334332,
+      "learning_rate": 9.553654668716065e-06,
+      "loss": 0.1743,
+      "step": 457
+    },
+    {
+      "epoch": 1.616946160635481,
+      "grad_norm": 0.7796697126004593,
+      "learning_rate": 9.551288338343597e-06,
+      "loss": 0.1841,
+      "step": 458
+    },
+    {
+      "epoch": 1.6204766107678727,
+      "grad_norm": 0.8659991614483475,
+      "learning_rate": 9.548916046556202e-06,
+      "loss": 0.1624,
+      "step": 459
+    },
+    {
+      "epoch": 1.6240070609002648,
+      "grad_norm": 0.9487399709116949,
+      "learning_rate": 9.54653779646118e-06,
+      "loss": 0.1855,
+      "step": 460
+    },
+    {
+      "epoch": 1.6275375110326566,
+      "grad_norm": 0.6655499212908461,
+      "learning_rate": 9.544153591173632e-06,
+      "loss": 0.1582,
+      "step": 461
+    },
+    {
+      "epoch": 1.6310679611650487,
+      "grad_norm": 0.6850321878653797,
+      "learning_rate": 9.541763433816467e-06,
+      "loss": 0.161,
+      "step": 462
+    },
+    {
+      "epoch": 1.6345984112974405,
+      "grad_norm": 0.6344259176675145,
+      "learning_rate": 9.539367327520382e-06,
+      "loss": 0.1899,
+      "step": 463
+    },
+    {
+      "epoch": 1.6381288614298324,
+      "grad_norm": 0.8057880340454765,
+      "learning_rate": 9.536965275423869e-06,
+      "loss": 0.1849,
+      "step": 464
+    },
+    {
+      "epoch": 1.6416593115622242,
+      "grad_norm": 0.6524278173237419,
+      "learning_rate": 9.53455728067321e-06,
+      "loss": 0.1616,
+      "step": 465
+    },
+    {
+      "epoch": 1.645189761694616,
+      "grad_norm": 0.721136853352441,
+      "learning_rate": 9.532143346422466e-06,
+      "loss": 0.1561,
+      "step": 466
+    },
+    {
+      "epoch": 1.648720211827008,
+      "grad_norm": 0.9490769657599735,
+      "learning_rate": 9.529723475833485e-06,
+      "loss": 0.1754,
+      "step": 467
+    },
+    {
+      "epoch": 1.6522506619593997,
+      "grad_norm": 0.5644282783067912,
+      "learning_rate": 9.527297672075886e-06,
+      "loss": 0.1627,
+      "step": 468
+    },
+    {
+      "epoch": 1.6557811120917916,
+      "grad_norm": 0.7598201706804275,
+      "learning_rate": 9.524865938327058e-06,
+      "loss": 0.1716,
+      "step": 469
+    },
+    {
+      "epoch": 1.6593115622241836,
+      "grad_norm": 0.7591422889029776,
+      "learning_rate": 9.522428277772162e-06,
+      "loss": 0.1451,
+      "step": 470
+    },
+    {
+      "epoch": 1.6628420123565755,
+      "grad_norm": 0.6982093938144709,
+      "learning_rate": 9.519984693604116e-06,
+      "loss": 0.1645,
+      "step": 471
+    },
+    {
+      "epoch": 1.6663724624889673,
+      "grad_norm": 0.5228701865370198,
+      "learning_rate": 9.517535189023602e-06,
+      "loss": 0.1563,
+      "step": 472
+    },
+    {
+      "epoch": 1.6699029126213594,
+      "grad_norm": 0.8204330233451095,
+      "learning_rate": 9.515079767239056e-06,
+      "loss": 0.1678,
+      "step": 473
+    },
+    {
+      "epoch": 1.6734333627537512,
+      "grad_norm": 0.9523128967867128,
+      "learning_rate": 9.512618431466662e-06,
+      "loss": 0.1664,
+      "step": 474
+    },
+    {
+      "epoch": 1.676963812886143,
+      "grad_norm": 0.7078287409806696,
+      "learning_rate": 9.510151184930354e-06,
+      "loss": 0.1658,
+      "step": 475
+    },
+    {
+      "epoch": 1.6804942630185349,
+      "grad_norm": 0.7520543206855101,
+      "learning_rate": 9.507678030861805e-06,
+      "loss": 0.1646,
+      "step": 476
+    },
+    {
+      "epoch": 1.6840247131509267,
+      "grad_norm": 0.9284654185473334,
+      "learning_rate": 9.505198972500426e-06,
+      "loss": 0.1684,
+      "step": 477
+    },
+    {
+      "epoch": 1.6875551632833186,
+      "grad_norm": 0.6224185496516447,
+      "learning_rate": 9.502714013093364e-06,
+      "loss": 0.1475,
+      "step": 478
+    },
+    {
+      "epoch": 1.6910856134157104,
+      "grad_norm": 0.7200200167776666,
+      "learning_rate": 9.500223155895491e-06,
+      "loss": 0.1739,
+      "step": 479
+    },
+    {
+      "epoch": 1.6946160635481022,
+      "grad_norm": 0.7893065612440167,
+      "learning_rate": 9.497726404169411e-06,
+      "loss": 0.1577,
+      "step": 480
+    },
+    {
+      "epoch": 1.6981465136804943,
+      "grad_norm": 0.806998234582965,
+      "learning_rate": 9.495223761185443e-06,
+      "loss": 0.1692,
+      "step": 481
+    },
+    {
+      "epoch": 1.7016769638128861,
+      "grad_norm": 0.8896910612567586,
+      "learning_rate": 9.492715230221623e-06,
+      "loss": 0.155,
+      "step": 482
+    },
+    {
+      "epoch": 1.705207413945278,
+      "grad_norm": 0.7929727685131104,
+      "learning_rate": 9.490200814563704e-06,
+      "loss": 0.1693,
+      "step": 483
+    },
+    {
+      "epoch": 1.70873786407767,
+      "grad_norm": 0.8281002779014005,
+      "learning_rate": 9.48768051750514e-06,
+      "loss": 0.162,
+      "step": 484
+    },
+    {
+      "epoch": 1.7122683142100619,
+      "grad_norm": 0.7574881698882824,
+      "learning_rate": 9.485154342347093e-06,
+      "loss": 0.1822,
+      "step": 485
+    },
+    {
+      "epoch": 1.7157987643424537,
+      "grad_norm": 0.7603690459222415,
+      "learning_rate": 9.482622292398424e-06,
+      "loss": 0.1705,
+      "step": 486
+    },
+    {
+      "epoch": 1.7193292144748455,
+      "grad_norm": 0.6922874253842599,
+      "learning_rate": 9.480084370975691e-06,
+      "loss": 0.1687,
+      "step": 487
+    },
+    {
+      "epoch": 1.7228596646072374,
+      "grad_norm": 0.80394336074322,
+      "learning_rate": 9.477540581403137e-06,
+      "loss": 0.1688,
+      "step": 488
+    },
+    {
+      "epoch": 1.7263901147396292,
+      "grad_norm": 0.8802309764956326,
+      "learning_rate": 9.474990927012694e-06,
+      "loss": 0.1304,
+      "step": 489
+    },
+    {
+      "epoch": 1.729920564872021,
+      "grad_norm": 0.6137981646644614,
+      "learning_rate": 9.472435411143979e-06,
+      "loss": 0.1664,
+      "step": 490
+    },
+    {
+      "epoch": 1.733451015004413,
+      "grad_norm": 0.9282185888184011,
+      "learning_rate": 9.469874037144281e-06,
+      "loss": 0.2012,
+      "step": 491
+    },
+    {
+      "epoch": 1.736981465136805,
+      "grad_norm": 0.6630691029948931,
+      "learning_rate": 9.467306808368568e-06,
+      "loss": 0.1502,
+      "step": 492
+    },
+    {
+      "epoch": 1.7405119152691968,
+      "grad_norm": 0.7577885818393465,
+      "learning_rate": 9.46473372817947e-06,
+      "loss": 0.1702,
+      "step": 493
+    },
+    {
+      "epoch": 1.7440423654015889,
+      "grad_norm": 0.5899998168235006,
+      "learning_rate": 9.46215479994729e-06,
+      "loss": 0.1602,
+      "step": 494
+    },
+    {
+      "epoch": 1.7475728155339807,
+      "grad_norm": 1.0306863382019156,
+      "learning_rate": 9.459570027049983e-06,
+      "loss": 0.1334,
+      "step": 495
+    },
+    {
+      "epoch": 1.7511032656663725,
+      "grad_norm": 0.865412362993785,
+      "learning_rate": 9.456979412873163e-06,
+      "loss": 0.1852,
+      "step": 496
+    },
+    {
+      "epoch": 1.7546337157987644,
+      "grad_norm": 0.8407963911360593,
+      "learning_rate": 9.454382960810093e-06,
+      "loss": 0.1691,
+      "step": 497
+    },
+    {
+      "epoch": 1.7581641659311562,
+      "grad_norm": 0.6988074166274713,
+      "learning_rate": 9.451780674261687e-06,
+      "loss": 0.16,
+      "step": 498
+    },
+    {
+      "epoch": 1.761694616063548,
+      "grad_norm": 0.6567772580817499,
+      "learning_rate": 9.4491725566365e-06,
+      "loss": 0.1678,
+      "step": 499
+    },
+    {
+      "epoch": 1.7652250661959399,
+      "grad_norm": 0.8826194122518776,
+      "learning_rate": 9.446558611350718e-06,
+      "loss": 0.177,
+      "step": 500
+    },
+    {
+      "epoch": 1.7687555163283317,
+      "grad_norm": 0.6900513296683881,
+      "learning_rate": 9.44393884182817e-06,
+      "loss": 0.1852,
+      "step": 501
+    },
+    {
+      "epoch": 1.7722859664607238,
+      "grad_norm": 0.6605181197941491,
+      "learning_rate": 9.441313251500307e-06,
+      "loss": 0.1557,
+      "step": 502
+    },
+    {
+      "epoch": 1.7758164165931156,
+      "grad_norm": 0.6146198835230601,
+      "learning_rate": 9.438681843806207e-06,
+      "loss": 0.1594,
+      "step": 503
+    },
+    {
+      "epoch": 1.7793468667255075,
+      "grad_norm": 0.5149059845127192,
+      "learning_rate": 9.436044622192569e-06,
+      "loss": 0.1665,
+      "step": 504
+    },
+    {
+      "epoch": 1.7828773168578995,
+      "grad_norm": 0.9135519362694595,
+      "learning_rate": 9.433401590113702e-06,
+      "loss": 0.1797,
+      "step": 505
+    },
+    {
+      "epoch": 1.7864077669902914,
+      "grad_norm": 0.8916844443544416,
+      "learning_rate": 9.43075275103153e-06,
+      "loss": 0.17,
+      "step": 506
+    },
+    {
+      "epoch": 1.7899382171226832,
+      "grad_norm": 0.8009121594911396,
+      "learning_rate": 9.428098108415587e-06,
+      "loss": 0.1599,
+      "step": 507
+    },
+    {
+      "epoch": 1.793468667255075,
+      "grad_norm": 0.8269496601871587,
+      "learning_rate": 9.425437665742998e-06,
+      "loss": 0.1633,
+      "step": 508
+    },
+    {
+      "epoch": 1.7969991173874669,
+      "grad_norm": 0.6798890374153103,
+      "learning_rate": 9.422771426498493e-06,
+      "loss": 0.1605,
+      "step": 509
+    },
+    {
+      "epoch": 1.8005295675198587,
+      "grad_norm": 0.6315062922758498,
+      "learning_rate": 9.420099394174397e-06,
+      "loss": 0.1795,
+      "step": 510
+    },
+    {
+      "epoch": 1.8040600176522505,
+      "grad_norm": 0.5746772810241558,
+      "learning_rate": 9.417421572270612e-06,
+      "loss": 0.1567,
+      "step": 511
+    },
+    {
+      "epoch": 1.8075904677846424,
+      "grad_norm": 0.8050650153052257,
+      "learning_rate": 9.414737964294636e-06,
+      "loss": 0.1729,
+      "step": 512
+    },
+    {
+      "epoch": 1.8111209179170344,
+      "grad_norm": 0.480452501514603,
+      "learning_rate": 9.412048573761535e-06,
+      "loss": 0.1766,
+      "step": 513
+    },
+    {
+      "epoch": 1.8146513680494263,
+      "grad_norm": 0.9104720825841831,
+      "learning_rate": 9.409353404193958e-06,
+      "loss": 0.1745,
+      "step": 514
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 0.8780636579991754,
+      "learning_rate": 9.406652459122115e-06,
+      "loss": 0.1762,
+      "step": 515
+    },
+    {
+      "epoch": 1.8217122683142102,
+      "grad_norm": 0.6984924204172243,
+      "learning_rate": 9.403945742083788e-06,
+      "loss": 0.1793,
+      "step": 516
+    },
+    {
+      "epoch": 1.825242718446602,
+      "grad_norm": 0.8238701991708997,
+      "learning_rate": 9.401233256624318e-06,
+      "loss": 0.1485,
+      "step": 517
+    },
+    {
+      "epoch": 1.8287731685789939,
+      "grad_norm": 0.8696582194623638,
+      "learning_rate": 9.398515006296598e-06,
+      "loss": 0.1891,
+      "step": 518
+    },
+    {
+      "epoch": 1.8323036187113857,
+      "grad_norm": 0.699771115532138,
+      "learning_rate": 9.395790994661075e-06,
+      "loss": 0.1553,
+      "step": 519
+    },
+    {
+      "epoch": 1.8358340688437775,
+      "grad_norm": 0.9125794295579036,
+      "learning_rate": 9.393061225285743e-06,
+      "loss": 0.163,
+      "step": 520
+    },
+    {
+      "epoch": 1.8393645189761694,
+      "grad_norm": 0.8950419906565266,
+      "learning_rate": 9.390325701746135e-06,
+      "loss": 0.167,
+      "step": 521
+    },
+    {
+      "epoch": 1.8428949691085612,
+      "grad_norm": 0.8710778282196566,
+      "learning_rate": 9.387584427625323e-06,
+      "loss": 0.1585,
+      "step": 522
+    },
+    {
+      "epoch": 1.8464254192409533,
+      "grad_norm": 0.8901232516687565,
+      "learning_rate": 9.38483740651391e-06,
+      "loss": 0.171,
+      "step": 523
+    },
+    {
+      "epoch": 1.849955869373345,
+      "grad_norm": 0.7244577661750393,
+      "learning_rate": 9.382084642010026e-06,
+      "loss": 0.1718,
+      "step": 524
+    },
+    {
+      "epoch": 1.853486319505737,
+      "grad_norm": 0.9044306478777916,
+      "learning_rate": 9.379326137719329e-06,
+      "loss": 0.1762,
+      "step": 525
+    },
+    {
+      "epoch": 1.857016769638129,
+      "grad_norm": 0.6730866435868901,
+      "learning_rate": 9.376561897254987e-06,
+      "loss": 0.1567,
+      "step": 526
+    },
+    {
+      "epoch": 1.8605472197705208,
+      "grad_norm": 0.816989704732525,
+      "learning_rate": 9.37379192423769e-06,
+      "loss": 0.1919,
+      "step": 527
+    },
+    {
+      "epoch": 1.8640776699029127,
+      "grad_norm": 0.7655388484499017,
+      "learning_rate": 9.371016222295628e-06,
+      "loss": 0.1736,
+      "step": 528
+    },
+    {
+      "epoch": 1.8676081200353045,
+      "grad_norm": 0.8699212931081676,
+      "learning_rate": 9.368234795064502e-06,
+      "loss": 0.1625,
+      "step": 529
+    },
+    {
+      "epoch": 1.8711385701676964,
+      "grad_norm": 1.0534597898924092,
+      "learning_rate": 9.365447646187509e-06,
+      "loss": 0.1381,
+      "step": 530
+    },
+    {
+      "epoch": 1.8746690203000882,
+      "grad_norm": 0.6614479535387474,
+      "learning_rate": 9.36265477931534e-06,
+      "loss": 0.1467,
+      "step": 531
+    },
+    {
+      "epoch": 1.87819947043248,
+      "grad_norm": 0.8393186062366995,
+      "learning_rate": 9.359856198106176e-06,
+      "loss": 0.1913,
+      "step": 532
+    },
+    {
+      "epoch": 1.8817299205648719,
+      "grad_norm": 0.5866280093193419,
+      "learning_rate": 9.357051906225685e-06,
+      "loss": 0.1566,
+      "step": 533
+    },
+    {
+      "epoch": 1.885260370697264,
+      "grad_norm": 0.9996266820247056,
+      "learning_rate": 9.35424190734701e-06,
+      "loss": 0.1919,
+      "step": 534
+    },
+    {
+      "epoch": 1.8887908208296558,
+      "grad_norm": 0.8127958941754003,
+      "learning_rate": 9.351426205150778e-06,
+      "loss": 0.1695,
+      "step": 535
+    },
+    {
+      "epoch": 1.8923212709620476,
+      "grad_norm": 0.6634743250735253,
+      "learning_rate": 9.348604803325075e-06,
+      "loss": 0.1555,
+      "step": 536
+    },
+    {
+      "epoch": 1.8958517210944397,
+      "grad_norm": 0.812340369225641,
+      "learning_rate": 9.34577770556546e-06,
+      "loss": 0.1645,
+      "step": 537
+    },
+    {
+      "epoch": 1.8993821712268315,
+      "grad_norm": 0.7497218737066156,
+      "learning_rate": 9.342944915574952e-06,
+      "loss": 0.1668,
+      "step": 538
+    },
+    {
+      "epoch": 1.9029126213592233,
+      "grad_norm": 0.5695613946823557,
+      "learning_rate": 9.340106437064026e-06,
+      "loss": 0.1625,
+      "step": 539
+    },
+    {
+      "epoch": 1.9064430714916152,
+      "grad_norm": 0.8336306538843815,
+      "learning_rate": 9.337262273750604e-06,
+      "loss": 0.1679,
+      "step": 540
+    },
+    {
+      "epoch": 1.909973521624007,
+      "grad_norm": 0.7985645323717766,
+      "learning_rate": 9.334412429360058e-06,
+      "loss": 0.1717,
+      "step": 541
+    },
+    {
+      "epoch": 1.9135039717563989,
+      "grad_norm": 0.803743435053368,
+      "learning_rate": 9.3315569076252e-06,
+      "loss": 0.1811,
+      "step": 542
+    },
+    {
+      "epoch": 1.9170344218887907,
+      "grad_norm": 0.7792918327227041,
+      "learning_rate": 9.328695712286283e-06,
+      "loss": 0.178,
+      "step": 543
+    },
+    {
+      "epoch": 1.9205648720211828,
+      "grad_norm": 1.0216399627321537,
+      "learning_rate": 9.32582884709098e-06,
+      "loss": 0.1702,
+      "step": 544
+    },
+    {
+      "epoch": 1.9240953221535746,
+      "grad_norm": 0.8524211551346756,
+      "learning_rate": 9.322956315794403e-06,
+      "loss": 0.1769,
+      "step": 545
+    },
+    {
+      "epoch": 1.9276257722859664,
+      "grad_norm": 1.0061876178961442,
+      "learning_rate": 9.320078122159077e-06,
+      "loss": 0.1725,
+      "step": 546
+    },
+    {
+      "epoch": 1.9311562224183585,
+      "grad_norm": 0.6449173961835437,
+      "learning_rate": 9.317194269954948e-06,
+      "loss": 0.1669,
+      "step": 547
+    },
+    {
+      "epoch": 1.9346866725507503,
+      "grad_norm": 0.6716320314286484,
+      "learning_rate": 9.314304762959373e-06,
+      "loss": 0.1837,
+      "step": 548
+    },
+    {
+      "epoch": 1.9382171226831422,
+      "grad_norm": 0.7847559077293133,
+      "learning_rate": 9.311409604957116e-06,
+      "loss": 0.1697,
+      "step": 549
+    },
+    {
+      "epoch": 1.941747572815534,
+      "grad_norm": 0.6479063264651344,
+      "learning_rate": 9.30850879974034e-06,
+      "loss": 0.1887,
+      "step": 550
+    },
+    {
+      "epoch": 1.9452780229479258,
+      "grad_norm": 0.5778944362755724,
+      "learning_rate": 9.305602351108613e-06,
+      "loss": 0.1732,
+      "step": 551
+    },
+    {
+      "epoch": 1.9488084730803177,
+      "grad_norm": 0.5993363081022134,
+      "learning_rate": 9.302690262868882e-06,
+      "loss": 0.1524,
+      "step": 552
+    },
+    {
+      "epoch": 1.9523389232127095,
+      "grad_norm": 0.7186259440033298,
+      "learning_rate": 9.299772538835492e-06,
+      "loss": 0.1562,
+      "step": 553
+    },
+    {
+      "epoch": 1.9558693733451014,
+      "grad_norm": 0.7547445542865213,
+      "learning_rate": 9.296849182830167e-06,
+      "loss": 0.159,
+      "step": 554
+    },
+    {
+      "epoch": 1.9593998234774934,
+      "grad_norm": 0.742454944410959,
+      "learning_rate": 9.293920198682004e-06,
+      "loss": 0.178,
+      "step": 555
+    },
+    {
+      "epoch": 1.9629302736098853,
+      "grad_norm": 0.731704919378445,
+      "learning_rate": 9.290985590227478e-06,
+      "loss": 0.178,
+      "step": 556
+    },
+    {
+      "epoch": 1.966460723742277,
+      "grad_norm": 0.6326777732721525,
+      "learning_rate": 9.288045361310423e-06,
+      "loss": 0.1572,
+      "step": 557
+    },
+    {
+      "epoch": 1.9699911738746692,
+      "grad_norm": 0.6096205323065964,
+      "learning_rate": 9.285099515782045e-06,
+      "loss": 0.1609,
+      "step": 558
+    },
+    {
+      "epoch": 1.973521624007061,
+      "grad_norm": 0.968981563236468,
+      "learning_rate": 9.2821480575009e-06,
+      "loss": 0.1554,
+      "step": 559
+    },
+    {
+      "epoch": 1.9770520741394528,
+      "grad_norm": 0.6731379335046332,
+      "learning_rate": 9.279190990332896e-06,
+      "loss": 0.1786,
+      "step": 560
+    },
+    {
+      "epoch": 1.9805825242718447,
+      "grad_norm": 0.5474632434518878,
+      "learning_rate": 9.276228318151288e-06,
+      "loss": 0.1825,
+      "step": 561
+    },
+    {
+      "epoch": 1.9841129744042365,
+      "grad_norm": 0.5300134793710058,
+      "learning_rate": 9.273260044836675e-06,
+      "loss": 0.1507,
+      "step": 562
+    },
+    {
+      "epoch": 1.9876434245366283,
+      "grad_norm": 0.6263089144886929,
+      "learning_rate": 9.270286174276992e-06,
+      "loss": 0.1753,
+      "step": 563
+    },
+    {
+      "epoch": 1.9911738746690202,
+      "grad_norm": 0.7895430535402537,
+      "learning_rate": 9.267306710367503e-06,
+      "loss": 0.1772,
+      "step": 564
+    },
+    {
+      "epoch": 1.994704324801412,
+      "grad_norm": 0.7740002132732854,
+      "learning_rate": 9.2643216570108e-06,
+      "loss": 0.1874,
+      "step": 565
+    },
+    {
+      "epoch": 1.998234774933804,
+      "grad_norm": 0.9874775418494401,
+      "learning_rate": 9.261331018116795e-06,
+      "loss": 0.173,
+      "step": 566
+    },
+    {
+      "epoch": 2.001765225066196,
+      "grad_norm": 0.6888142224041766,
+      "learning_rate": 9.25833479760272e-06,
+      "loss": 0.1811,
+      "step": 567
+    },
+    {
+      "epoch": 2.005295675198588,
+      "grad_norm": 0.6459025197429961,
+      "learning_rate": 9.255332999393114e-06,
+      "loss": 0.1507,
+      "step": 568
+    },
+    {
+      "epoch": 2.00882612533098,
+      "grad_norm": 0.6618092460843229,
+      "learning_rate": 9.252325627419821e-06,
+      "loss": 0.1302,
+      "step": 569
+    },
+    {
+      "epoch": 2.0123565754633717,
+      "grad_norm": 0.8386091353618218,
+      "learning_rate": 9.249312685621989e-06,
+      "loss": 0.1531,
+      "step": 570
+    },
+    {
+      "epoch": 2.0158870255957635,
+      "grad_norm": 0.7735686550828623,
+      "learning_rate": 9.246294177946062e-06,
+      "loss": 0.1554,
+      "step": 571
+    },
+    {
+      "epoch": 2.0194174757281553,
+      "grad_norm": 0.6350949023955103,
+      "learning_rate": 9.24327010834577e-06,
+      "loss": 0.1161,
+      "step": 572
+    },
+    {
+      "epoch": 2.022947925860547,
+      "grad_norm": 0.7901812966196844,
+      "learning_rate": 9.24024048078213e-06,
+      "loss": 0.1498,
+      "step": 573
+    },
+    {
+      "epoch": 2.026478375992939,
+      "grad_norm": 0.7745266020928578,
+      "learning_rate": 9.237205299223444e-06,
+      "loss": 0.1665,
+      "step": 574
+    },
+    {
+      "epoch": 2.030008826125331,
+      "grad_norm": 0.6972058102385509,
+      "learning_rate": 9.234164567645278e-06,
+      "loss": 0.1257,
+      "step": 575
+    },
+    {
+      "epoch": 2.0335392762577227,
+      "grad_norm": 0.6137168822739238,
+      "learning_rate": 9.231118290030481e-06,
+      "loss": 0.121,
+      "step": 576
+    },
+    {
+      "epoch": 2.0370697263901145,
+      "grad_norm": 0.7533855618361971,
+      "learning_rate": 9.228066470369155e-06,
+      "loss": 0.1352,
+      "step": 577
+    },
+    {
+      "epoch": 2.040600176522507,
+      "grad_norm": 0.792710448056873,
+      "learning_rate": 9.225009112658667e-06,
+      "loss": 0.144,
+      "step": 578
+    },
+    {
+      "epoch": 2.0441306266548986,
+      "grad_norm": 0.865387452735705,
+      "learning_rate": 9.221946220903634e-06,
+      "loss": 0.1521,
+      "step": 579
+    },
+    {
+      "epoch": 2.0476610767872905,
+      "grad_norm": 1.0235916124259214,
+      "learning_rate": 9.218877799115929e-06,
+      "loss": 0.1441,
+      "step": 580
+    },
+    {
+      "epoch": 2.0511915269196823,
+      "grad_norm": 1.0508276358903932,
+      "learning_rate": 9.21580385131466e-06,
+      "loss": 0.1685,
+      "step": 581
+    },
+    {
+      "epoch": 2.054721977052074,
+      "grad_norm": 0.8910905064383835,
+      "learning_rate": 9.212724381526177e-06,
+      "loss": 0.1736,
+      "step": 582
+    },
+    {
+      "epoch": 2.058252427184466,
+      "grad_norm": 1.0571601952147707,
+      "learning_rate": 9.209639393784064e-06,
+      "loss": 0.1565,
+      "step": 583
+    },
+    {
+      "epoch": 2.061782877316858,
+      "grad_norm": 0.8815436783906608,
+      "learning_rate": 9.206548892129129e-06,
+      "loss": 0.1371,
+      "step": 584
+    },
+    {
+      "epoch": 2.0653133274492497,
+      "grad_norm": 1.0517504956260713,
+      "learning_rate": 9.203452880609407e-06,
+      "loss": 0.1464,
+      "step": 585
+    },
+    {
+      "epoch": 2.0688437775816415,
+      "grad_norm": 0.8529237538392984,
+      "learning_rate": 9.200351363280144e-06,
+      "loss": 0.1266,
+      "step": 586
+    },
+    {
+      "epoch": 2.0723742277140333,
+      "grad_norm": 0.874098127040573,
+      "learning_rate": 9.197244344203804e-06,
+      "loss": 0.1349,
+      "step": 587
+    },
+    {
+      "epoch": 2.0759046778464256,
+      "grad_norm": 0.8109927532199982,
+      "learning_rate": 9.194131827450054e-06,
+      "loss": 0.1445,
+      "step": 588
+    },
+    {
+      "epoch": 2.0794351279788175,
+      "grad_norm": 0.6489560116761195,
+      "learning_rate": 9.191013817095762e-06,
+      "loss": 0.1294,
+      "step": 589
+    },
+    {
+      "epoch": 2.0829655781112093,
+      "grad_norm": 0.7508144606670216,
+      "learning_rate": 9.18789031722499e-06,
+      "loss": 0.1135,
+      "step": 590
+    },
+    {
+      "epoch": 2.086496028243601,
+      "grad_norm": 0.7295793522432309,
+      "learning_rate": 9.184761331928997e-06,
+      "loss": 0.1363,
+      "step": 591
+    },
+    {
+      "epoch": 2.090026478375993,
+      "grad_norm": 0.861095430673198,
+      "learning_rate": 9.18162686530622e-06,
+      "loss": 0.1279,
+      "step": 592
+    },
+    {
+      "epoch": 2.093556928508385,
+      "grad_norm": 0.6861216733576722,
+      "learning_rate": 9.178486921462276e-06,
+      "loss": 0.1224,
+      "step": 593
+    },
+    {
+      "epoch": 2.0970873786407767,
+      "grad_norm": 0.8019378401083357,
+      "learning_rate": 9.175341504509961e-06,
+      "loss": 0.1545,
+      "step": 594
+    },
+    {
+      "epoch": 2.1006178287731685,
+      "grad_norm": 1.526043856422149,
+      "learning_rate": 9.172190618569236e-06,
+      "loss": 0.1306,
+      "step": 595
+    },
+    {
+      "epoch": 2.1041482789055603,
+      "grad_norm": 0.8476895681346993,
+      "learning_rate": 9.16903426776723e-06,
+      "loss": 0.149,
+      "step": 596
+    },
+    {
+      "epoch": 2.107678729037952,
+      "grad_norm": 0.8161608102533534,
+      "learning_rate": 9.16587245623822e-06,
+      "loss": 0.1364,
+      "step": 597
+    },
+    {
+      "epoch": 2.111209179170344,
+      "grad_norm": 0.7223711233696612,
+      "learning_rate": 9.162705188123647e-06,
+      "loss": 0.142,
+      "step": 598
+    },
+    {
+      "epoch": 2.1147396293027363,
+      "grad_norm": 0.8507390490828752,
+      "learning_rate": 9.159532467572094e-06,
+      "loss": 0.1547,
+      "step": 599
+    },
+    {
+      "epoch": 2.118270079435128,
+      "grad_norm": 0.8316619115260057,
+      "learning_rate": 9.156354298739285e-06,
+      "loss": 0.1563,
+      "step": 600
+    },
+    {
+      "epoch": 2.12180052956752,
+      "grad_norm": 0.7544897788791859,
+      "learning_rate": 9.153170685788081e-06,
+      "loss": 0.1421,
+      "step": 601
+    },
+    {
+      "epoch": 2.125330979699912,
+      "grad_norm": 0.800819425552279,
+      "learning_rate": 9.149981632888476e-06,
+      "loss": 0.1579,
+      "step": 602
+    },
+    {
+      "epoch": 2.1288614298323036,
+      "grad_norm": 0.7838909562518437,
+      "learning_rate": 9.146787144217588e-06,
+      "loss": 0.1681,
+      "step": 603
+    },
+    {
+      "epoch": 2.1323918799646955,
+      "grad_norm": 0.9303054432448055,
+      "learning_rate": 9.143587223959653e-06,
+      "loss": 0.1397,
+      "step": 604
+    },
+    {
+      "epoch": 2.1359223300970873,
+      "grad_norm": 0.7625989738070533,
+      "learning_rate": 9.140381876306026e-06,
+      "loss": 0.1581,
+      "step": 605
+    },
+    {
+      "epoch": 2.139452780229479,
+      "grad_norm": 0.7857755303539646,
+      "learning_rate": 9.137171105455165e-06,
+      "loss": 0.1258,
+      "step": 606
+    },
+    {
+      "epoch": 2.142983230361871,
+      "grad_norm": 0.6441805120671155,
+      "learning_rate": 9.133954915612635e-06,
+      "loss": 0.1305,
+      "step": 607
+    },
+    {
+      "epoch": 2.146513680494263,
+      "grad_norm": 0.8467018145382312,
+      "learning_rate": 9.130733310991102e-06,
+      "loss": 0.1432,
+      "step": 608
+    },
+    {
+      "epoch": 2.1500441306266547,
+      "grad_norm": 0.7571331987070536,
+      "learning_rate": 9.127506295810316e-06,
+      "loss": 0.1536,
+      "step": 609
+    },
+    {
+      "epoch": 2.153574580759047,
+      "grad_norm": 1.0166648071075606,
+      "learning_rate": 9.124273874297123e-06,
+      "loss": 0.1641,
+      "step": 610
+    },
+    {
+      "epoch": 2.157105030891439,
+      "grad_norm": 1.0624683994204978,
+      "learning_rate": 9.121036050685443e-06,
+      "loss": 0.1568,
+      "step": 611
+    },
+    {
+      "epoch": 2.1606354810238306,
+      "grad_norm": 0.754254288642853,
+      "learning_rate": 9.11779282921628e-06,
+      "loss": 0.1456,
+      "step": 612
+    },
+    {
+      "epoch": 2.1641659311562225,
+      "grad_norm": 0.8711250730686613,
+      "learning_rate": 9.114544214137699e-06,
+      "loss": 0.1664,
+      "step": 613
+    },
+    {
+      "epoch": 2.1676963812886143,
+      "grad_norm": 0.8545203984797597,
+      "learning_rate": 9.111290209704835e-06,
+      "loss": 0.1494,
+      "step": 614
+    },
+    {
+      "epoch": 2.171226831421006,
+      "grad_norm": 0.8374462073082292,
+      "learning_rate": 9.108030820179882e-06,
+      "loss": 0.1541,
+      "step": 615
+    },
+    {
+      "epoch": 2.174757281553398,
+      "grad_norm": 0.7357997781637011,
+      "learning_rate": 9.104766049832088e-06,
+      "loss": 0.1491,
+      "step": 616
+    },
+    {
+      "epoch": 2.17828773168579,
+      "grad_norm": 0.7606936499199851,
+      "learning_rate": 9.101495902937744e-06,
+      "loss": 0.1577,
+      "step": 617
+    },
+    {
+      "epoch": 2.1818181818181817,
+      "grad_norm": 0.8031929036786638,
+      "learning_rate": 9.098220383780191e-06,
+      "loss": 0.1649,
+      "step": 618
+    },
+    {
+      "epoch": 2.1853486319505735,
+      "grad_norm": 1.064933842253509,
+      "learning_rate": 9.094939496649803e-06,
+      "loss": 0.1218,
+      "step": 619
+    },
+    {
+      "epoch": 2.1888790820829658,
+      "grad_norm": 0.7301032836107995,
+      "learning_rate": 9.091653245843984e-06,
+      "loss": 0.1376,
+      "step": 620
+    },
+    {
+      "epoch": 2.1924095322153576,
+      "grad_norm": 0.8431618205640518,
+      "learning_rate": 9.088361635667162e-06,
+      "loss": 0.148,
+      "step": 621
+    },
+    {
+      "epoch": 2.1959399823477495,
+      "grad_norm": 1.340479354749448,
+      "learning_rate": 9.08506467043079e-06,
+      "loss": 0.1616,
+      "step": 622
+    },
+    {
+      "epoch": 2.1994704324801413,
+      "grad_norm": 0.8064011593740745,
+      "learning_rate": 9.081762354453333e-06,
+      "loss": 0.1488,
+      "step": 623
+    },
+    {
+      "epoch": 2.203000882612533,
+      "grad_norm": 0.8004191241563091,
+      "learning_rate": 9.078454692060264e-06,
+      "loss": 0.1403,
+      "step": 624
+    },
+    {
+      "epoch": 2.206531332744925,
+      "grad_norm": 0.8885936871876676,
+      "learning_rate": 9.075141687584056e-06,
+      "loss": 0.1298,
+      "step": 625
+    },
+    {
+      "epoch": 2.210061782877317,
+      "grad_norm": 1.013124408250628,
+      "learning_rate": 9.071823345364187e-06,
+      "loss": 0.1461,
+      "step": 626
+    },
+    {
+      "epoch": 2.2135922330097086,
+      "grad_norm": 0.8812285753845427,
+      "learning_rate": 9.068499669747122e-06,
+      "loss": 0.1475,
+      "step": 627
+    },
+    {
+      "epoch": 2.2171226831421005,
+      "grad_norm": 0.8390832015738438,
+      "learning_rate": 9.065170665086307e-06,
+      "loss": 0.152,
+      "step": 628
+    },
+    {
+      "epoch": 2.2206531332744923,
+      "grad_norm": 0.6214987457344594,
+      "learning_rate": 9.061836335742175e-06,
+      "loss": 0.1418,
+      "step": 629
+    },
+    {
+      "epoch": 2.2241835834068846,
+      "grad_norm": 0.8210501747316103,
+      "learning_rate": 9.058496686082132e-06,
+      "loss": 0.1385,
+      "step": 630
+    },
+    {
+      "epoch": 2.2277140335392764,
+      "grad_norm": 0.8877640524225171,
+      "learning_rate": 9.055151720480553e-06,
+      "loss": 0.1636,
+      "step": 631
+    },
+    {
+      "epoch": 2.2312444836716683,
+      "grad_norm": 0.6993342223349599,
+      "learning_rate": 9.051801443318772e-06,
+      "loss": 0.1267,
+      "step": 632
+    },
+    {
+      "epoch": 2.23477493380406,
+      "grad_norm": 0.915383376214395,
+      "learning_rate": 9.048445858985085e-06,
+      "loss": 0.1346,
+      "step": 633
+    },
+    {
+      "epoch": 2.238305383936452,
+      "grad_norm": 0.8226201621022076,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.145,
+      "step": 634
+    },
+    {
+      "epoch": 2.241835834068844,
+      "grad_norm": 0.949771207964201,
+      "learning_rate": 9.041718786389921e-06,
+      "loss": 0.1729,
+      "step": 635
+    },
+    {
+      "epoch": 2.2453662842012356,
+      "grad_norm": 0.7669826168302643,
+      "learning_rate": 9.038347306939768e-06,
+      "loss": 0.1491,
+      "step": 636
+    },
+    {
+      "epoch": 2.2488967343336275,
+      "grad_norm": 0.8401151453900115,
+      "learning_rate": 9.034970537940342e-06,
+      "loss": 0.1587,
+      "step": 637
+    },
+    {
+      "epoch": 2.2524271844660193,
+      "grad_norm": 0.8362745634336628,
+      "learning_rate": 9.031588483814639e-06,
+      "loss": 0.152,
+      "step": 638
+    },
+    {
+      "epoch": 2.255957634598411,
+      "grad_norm": 0.828931692995688,
+      "learning_rate": 9.028201148992572e-06,
+      "loss": 0.127,
+      "step": 639
+    },
+    {
+      "epoch": 2.259488084730803,
+      "grad_norm": 0.8901277435562056,
+      "learning_rate": 9.024808537910981e-06,
+      "loss": 0.1472,
+      "step": 640
+    },
+    {
+      "epoch": 2.2630185348631953,
+      "grad_norm": 0.8458917100010721,
+      "learning_rate": 9.021410655013604e-06,
+      "loss": 0.1549,
+      "step": 641
+    },
+    {
+      "epoch": 2.266548984995587,
+      "grad_norm": 0.7511334646906924,
+      "learning_rate": 9.018007504751094e-06,
+      "loss": 0.1401,
+      "step": 642
+    },
+    {
+      "epoch": 2.270079435127979,
+      "grad_norm": 0.8998427764695459,
+      "learning_rate": 9.014599091581e-06,
+      "loss": 0.1665,
+      "step": 643
+    },
+    {
+      "epoch": 2.2736098852603708,
+      "grad_norm": 0.8241701592002284,
+      "learning_rate": 9.011185419967761e-06,
+      "loss": 0.136,
+      "step": 644
+    },
+    {
+      "epoch": 2.2771403353927626,
+      "grad_norm": 0.7830126296047867,
+      "learning_rate": 9.007766494382713e-06,
+      "loss": 0.151,
+      "step": 645
+    },
+    {
+      "epoch": 2.2806707855251545,
+      "grad_norm": 1.0742703925847628,
+      "learning_rate": 9.004342319304063e-06,
+      "loss": 0.1606,
+      "step": 646
+    },
+    {
+      "epoch": 2.2842012356575463,
+      "grad_norm": 0.9139889110017969,
+      "learning_rate": 9.000912899216902e-06,
+      "loss": 0.1556,
+      "step": 647
+    },
+    {
+      "epoch": 2.287731685789938,
+      "grad_norm": 0.9228198414722247,
+      "learning_rate": 8.997478238613186e-06,
+      "loss": 0.1635,
+      "step": 648
+    },
+    {
+      "epoch": 2.29126213592233,
+      "grad_norm": 0.6979642688439006,
+      "learning_rate": 8.994038341991739e-06,
+      "loss": 0.1443,
+      "step": 649
+    },
+    {
+      "epoch": 2.294792586054722,
+      "grad_norm": 0.7432603342893744,
+      "learning_rate": 8.990593213858239e-06,
+      "loss": 0.1497,
+      "step": 650
+    },
+    {
+      "epoch": 2.2983230361871136,
+      "grad_norm": 0.9788738069579437,
+      "learning_rate": 8.987142858725221e-06,
+      "loss": 0.1388,
+      "step": 651
+    },
+    {
+      "epoch": 2.301853486319506,
+      "grad_norm": 0.9451424967898011,
+      "learning_rate": 8.983687281112066e-06,
+      "loss": 0.1558,
+      "step": 652
+    },
+    {
+      "epoch": 2.3053839364518978,
+      "grad_norm": 0.796516553881763,
+      "learning_rate": 8.98022648554499e-06,
+      "loss": 0.1395,
+      "step": 653
+    },
+    {
+      "epoch": 2.3089143865842896,
+      "grad_norm": 0.6535029248700488,
+      "learning_rate": 8.976760476557049e-06,
+      "loss": 0.1466,
+      "step": 654
+    },
+    {
+      "epoch": 2.3124448367166814,
+      "grad_norm": 1.0145090677139883,
+      "learning_rate": 8.973289258688125e-06,
+      "loss": 0.1601,
+      "step": 655
+    },
+    {
+      "epoch": 2.3159752868490733,
+      "grad_norm": 0.9360528769068794,
+      "learning_rate": 8.969812836484929e-06,
+      "loss": 0.164,
+      "step": 656
+    },
+    {
+      "epoch": 2.319505736981465,
+      "grad_norm": 0.8047358252615086,
+      "learning_rate": 8.966331214500978e-06,
+      "loss": 0.1756,
+      "step": 657
+    },
+    {
+      "epoch": 2.323036187113857,
+      "grad_norm": 0.7975682754458384,
+      "learning_rate": 8.962844397296608e-06,
+      "loss": 0.1435,
+      "step": 658
+    },
+    {
+      "epoch": 2.326566637246249,
+      "grad_norm": 0.7277448447990951,
+      "learning_rate": 8.959352389438959e-06,
+      "loss": 0.1352,
+      "step": 659
+    },
+    {
+      "epoch": 2.3300970873786406,
+      "grad_norm": 0.8639595145485759,
+      "learning_rate": 8.955855195501966e-06,
+      "loss": 0.1497,
+      "step": 660
+    },
+    {
+      "epoch": 2.3336275375110325,
+      "grad_norm": 0.9741801125022028,
+      "learning_rate": 8.952352820066359e-06,
+      "loss": 0.1743,
+      "step": 661
+    },
+    {
+      "epoch": 2.3371579876434243,
+      "grad_norm": 0.7011036253403068,
+      "learning_rate": 8.948845267719658e-06,
+      "loss": 0.1483,
+      "step": 662
+    },
+    {
+      "epoch": 2.3406884377758166,
+      "grad_norm": 0.7915113319296646,
+      "learning_rate": 8.945332543056158e-06,
+      "loss": 0.1674,
+      "step": 663
+    },
+    {
+      "epoch": 2.3442188879082084,
+      "grad_norm": 0.7246375584288023,
+      "learning_rate": 8.941814650676934e-06,
+      "loss": 0.1429,
+      "step": 664
+    },
+    {
+      "epoch": 2.3477493380406003,
+      "grad_norm": 0.6572418012169463,
+      "learning_rate": 8.938291595189826e-06,
+      "loss": 0.1464,
+      "step": 665
+    },
+    {
+      "epoch": 2.351279788172992,
+      "grad_norm": 1.0088944178687222,
+      "learning_rate": 8.93476338120944e-06,
+      "loss": 0.1723,
+      "step": 666
+    },
+    {
+      "epoch": 2.354810238305384,
+      "grad_norm": 0.8873624823759165,
+      "learning_rate": 8.931230013357136e-06,
+      "loss": 0.1478,
+      "step": 667
+    },
+    {
+      "epoch": 2.358340688437776,
+      "grad_norm": 0.8232927235637216,
+      "learning_rate": 8.927691496261025e-06,
+      "loss": 0.1509,
+      "step": 668
+    },
+    {
+      "epoch": 2.3618711385701676,
+      "grad_norm": 0.7448944968086746,
+      "learning_rate": 8.924147834555966e-06,
+      "loss": 0.144,
+      "step": 669
+    },
+    {
+      "epoch": 2.3654015887025595,
+      "grad_norm": 0.6545431722863853,
+      "learning_rate": 8.920599032883553e-06,
+      "loss": 0.1442,
+      "step": 670
+    },
+    {
+      "epoch": 2.3689320388349513,
+      "grad_norm": 0.9857690033050072,
+      "learning_rate": 8.917045095892115e-06,
+      "loss": 0.1534,
+      "step": 671
+    },
+    {
+      "epoch": 2.3724624889673436,
+      "grad_norm": 0.8890672612580922,
+      "learning_rate": 8.913486028236703e-06,
+      "loss": 0.1414,
+      "step": 672
+    },
+    {
+      "epoch": 2.375992939099735,
+      "grad_norm": 0.7720214368079358,
+      "learning_rate": 8.909921834579093e-06,
+      "loss": 0.1379,
+      "step": 673
+    },
+    {
+      "epoch": 2.3795233892321273,
+      "grad_norm": 0.7968382535551254,
+      "learning_rate": 8.906352519587774e-06,
+      "loss": 0.1665,
+      "step": 674
+    },
+    {
+      "epoch": 2.383053839364519,
+      "grad_norm": 1.012361224051788,
+      "learning_rate": 8.902778087937943e-06,
+      "loss": 0.1506,
+      "step": 675
+    },
+    {
+      "epoch": 2.386584289496911,
+      "grad_norm": 1.6922659485205467,
+      "learning_rate": 8.899198544311496e-06,
+      "loss": 0.1621,
+      "step": 676
+    },
+    {
+      "epoch": 2.3901147396293028,
+      "grad_norm": 0.8810742656212996,
+      "learning_rate": 8.89561389339703e-06,
+      "loss": 0.1342,
+      "step": 677
+    },
+    {
+      "epoch": 2.3936451897616946,
+      "grad_norm": 0.8922813495137972,
+      "learning_rate": 8.892024139889827e-06,
+      "loss": 0.1348,
+      "step": 678
+    },
+    {
+      "epoch": 2.3971756398940864,
+      "grad_norm": 0.6428583850540706,
+      "learning_rate": 8.888429288491857e-06,
+      "loss": 0.1365,
+      "step": 679
+    },
+    {
+      "epoch": 2.4007060900264783,
+      "grad_norm": 0.959500169941628,
+      "learning_rate": 8.884829343911763e-06,
+      "loss": 0.1579,
+      "step": 680
+    },
+    {
+      "epoch": 2.40423654015887,
+      "grad_norm": 0.8928078544095518,
+      "learning_rate": 8.88122431086486e-06,
+      "loss": 0.158,
+      "step": 681
+    },
+    {
+      "epoch": 2.407766990291262,
+      "grad_norm": 0.7299606081515114,
+      "learning_rate": 8.877614194073134e-06,
+      "loss": 0.1332,
+      "step": 682
+    },
+    {
+      "epoch": 2.4112974404236542,
+      "grad_norm": 0.867643203493481,
+      "learning_rate": 8.87399899826522e-06,
+      "loss": 0.1567,
+      "step": 683
+    },
+    {
+      "epoch": 2.414827890556046,
+      "grad_norm": 0.8544098798566452,
+      "learning_rate": 8.870378728176412e-06,
+      "loss": 0.1479,
+      "step": 684
+    },
+    {
+      "epoch": 2.418358340688438,
+      "grad_norm": 1.0065430757516332,
+      "learning_rate": 8.86675338854865e-06,
+      "loss": 0.1826,
+      "step": 685
+    },
+    {
+      "epoch": 2.4218887908208298,
+      "grad_norm": 0.9392726338359665,
+      "learning_rate": 8.863122984130514e-06,
+      "loss": 0.1573,
+      "step": 686
+    },
+    {
+      "epoch": 2.4254192409532216,
+      "grad_norm": 0.815039972544626,
+      "learning_rate": 8.859487519677215e-06,
+      "loss": 0.1296,
+      "step": 687
+    },
+    {
+      "epoch": 2.4289496910856134,
+      "grad_norm": 0.8603703909617912,
+      "learning_rate": 8.855846999950595e-06,
+      "loss": 0.1547,
+      "step": 688
+    },
+    {
+      "epoch": 2.4324801412180053,
+      "grad_norm": 1.0058587648363357,
+      "learning_rate": 8.852201429719119e-06,
+      "loss": 0.1394,
+      "step": 689
+    },
+    {
+      "epoch": 2.436010591350397,
+      "grad_norm": 0.895199851785628,
+      "learning_rate": 8.848550813757861e-06,
+      "loss": 0.1379,
+      "step": 690
+    },
+    {
+      "epoch": 2.439541041482789,
+      "grad_norm": 0.7927306345545818,
+      "learning_rate": 8.844895156848511e-06,
+      "loss": 0.1347,
+      "step": 691
+    },
+    {
+      "epoch": 2.443071491615181,
+      "grad_norm": 0.7501822257778045,
+      "learning_rate": 8.841234463779359e-06,
+      "loss": 0.1472,
+      "step": 692
+    },
+    {
+      "epoch": 2.4466019417475726,
+      "grad_norm": 0.8155634914790267,
+      "learning_rate": 8.83756873934529e-06,
+      "loss": 0.1637,
+      "step": 693
+    },
+    {
+      "epoch": 2.450132391879965,
+      "grad_norm": 0.9416808234395326,
+      "learning_rate": 8.83389798834778e-06,
+      "loss": 0.1619,
+      "step": 694
+    },
+    {
+      "epoch": 2.4536628420123567,
+      "grad_norm": 0.7762931486236544,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.1607,
+      "step": 695
+    },
+    {
+      "epoch": 2.4571932921447486,
+      "grad_norm": 0.6647840629115083,
+      "learning_rate": 8.82654142590126e-06,
+      "loss": 0.1474,
+      "step": 696
+    },
+    {
+      "epoch": 2.4607237422771404,
+      "grad_norm": 0.809281030127039,
+      "learning_rate": 8.822855624088099e-06,
+      "loss": 0.1329,
+      "step": 697
+    },
+    {
+      "epoch": 2.4642541924095323,
+      "grad_norm": 0.7105112267949759,
+      "learning_rate": 8.819164814983179e-06,
+      "loss": 0.1513,
+      "step": 698
+    },
+    {
+      "epoch": 2.467784642541924,
+      "grad_norm": 0.6643065745395029,
+      "learning_rate": 8.815469003420835e-06,
+      "loss": 0.1355,
+      "step": 699
+    },
+    {
+      "epoch": 2.471315092674316,
+      "grad_norm": 0.6786894914054595,
+      "learning_rate": 8.811768194241951e-06,
+      "loss": 0.1491,
+      "step": 700
+    },
+    {
+      "epoch": 2.4748455428067078,
+      "grad_norm": 0.8985899448983452,
+      "learning_rate": 8.808062392293964e-06,
+      "loss": 0.1683,
+      "step": 701
+    },
+    {
+      "epoch": 2.4783759929390996,
+      "grad_norm": 0.8051939382089317,
+      "learning_rate": 8.804351602430838e-06,
+      "loss": 0.1316,
+      "step": 702
+    },
+    {
+      "epoch": 2.4819064430714914,
+      "grad_norm": 0.9657848784459468,
+      "learning_rate": 8.800635829513083e-06,
+      "loss": 0.1695,
+      "step": 703
+    },
+    {
+      "epoch": 2.4854368932038833,
+      "grad_norm": 0.8063728250819603,
+      "learning_rate": 8.796915078407729e-06,
+      "loss": 0.1524,
+      "step": 704
+    },
+    {
+      "epoch": 2.4889673433362756,
+      "grad_norm": 0.6461209785079439,
+      "learning_rate": 8.793189353988325e-06,
+      "loss": 0.1487,
+      "step": 705
+    },
+    {
+      "epoch": 2.4924977934686674,
+      "grad_norm": 0.6869278039173332,
+      "learning_rate": 8.789458661134943e-06,
+      "loss": 0.1442,
+      "step": 706
+    },
+    {
+      "epoch": 2.4960282436010592,
+      "grad_norm": 1.1034616460480864,
+      "learning_rate": 8.785723004734153e-06,
+      "loss": 0.1714,
+      "step": 707
+    },
+    {
+      "epoch": 2.499558693733451,
+      "grad_norm": 0.6104590542245811,
+      "learning_rate": 8.781982389679033e-06,
+      "loss": 0.1339,
+      "step": 708
+    },
+    {
+      "epoch": 2.503089143865843,
+      "grad_norm": 0.8481696722493157,
+      "learning_rate": 8.778236820869152e-06,
+      "loss": 0.1536,
+      "step": 709
+    },
+    {
+      "epoch": 2.5066195939982348,
+      "grad_norm": 0.8111930888716574,
+      "learning_rate": 8.77448630321057e-06,
+      "loss": 0.1422,
+      "step": 710
+    },
+    {
+      "epoch": 2.5101500441306266,
+      "grad_norm": 0.7036188732668607,
+      "learning_rate": 8.770730841615827e-06,
+      "loss": 0.1445,
+      "step": 711
+    },
+    {
+      "epoch": 2.5136804942630184,
+      "grad_norm": 0.9697606827791236,
+      "learning_rate": 8.766970441003942e-06,
+      "loss": 0.167,
+      "step": 712
+    },
+    {
+      "epoch": 2.5172109443954103,
+      "grad_norm": 0.8387174374399691,
+      "learning_rate": 8.7632051063004e-06,
+      "loss": 0.1585,
+      "step": 713
+    },
+    {
+      "epoch": 2.5207413945278025,
+      "grad_norm": 0.6631550381542344,
+      "learning_rate": 8.75943484243715e-06,
+      "loss": 0.1428,
+      "step": 714
+    },
+    {
+      "epoch": 2.524271844660194,
+      "grad_norm": 0.7695360425616815,
+      "learning_rate": 8.755659654352599e-06,
+      "loss": 0.1463,
+      "step": 715
+    },
+    {
+      "epoch": 2.5278022947925862,
+      "grad_norm": 0.7011902192003712,
+      "learning_rate": 8.751879546991601e-06,
+      "loss": 0.1406,
+      "step": 716
+    },
+    {
+      "epoch": 2.531332744924978,
+      "grad_norm": 0.7991566673347344,
+      "learning_rate": 8.748094525305457e-06,
+      "loss": 0.1545,
+      "step": 717
+    },
+    {
+      "epoch": 2.53486319505737,
+      "grad_norm": 0.8486318435641594,
+      "learning_rate": 8.744304594251898e-06,
+      "loss": 0.1565,
+      "step": 718
+    },
+    {
+      "epoch": 2.5383936451897617,
+      "grad_norm": 0.6882492324618286,
+      "learning_rate": 8.740509758795094e-06,
+      "loss": 0.141,
+      "step": 719
+    },
+    {
+      "epoch": 2.5419240953221536,
+      "grad_norm": 0.7803012919767831,
+      "learning_rate": 8.736710023905634e-06,
+      "loss": 0.1671,
+      "step": 720
+    },
+    {
+      "epoch": 2.5454545454545454,
+      "grad_norm": 1.0460659891563937,
+      "learning_rate": 8.732905394560526e-06,
+      "loss": 0.1662,
+      "step": 721
+    },
+    {
+      "epoch": 2.5489849955869373,
+      "grad_norm": 0.8104918877045203,
+      "learning_rate": 8.729095875743187e-06,
+      "loss": 0.1707,
+      "step": 722
+    },
+    {
+      "epoch": 2.552515445719329,
+      "grad_norm": 0.822076128070119,
+      "learning_rate": 8.725281472443442e-06,
+      "loss": 0.1434,
+      "step": 723
+    },
+    {
+      "epoch": 2.556045895851721,
+      "grad_norm": 0.8428018638045196,
+      "learning_rate": 8.72146218965751e-06,
+      "loss": 0.1452,
+      "step": 724
+    },
+    {
+      "epoch": 2.559576345984113,
+      "grad_norm": 0.6860938138056261,
+      "learning_rate": 8.717638032388002e-06,
+      "loss": 0.1452,
+      "step": 725
+    },
+    {
+      "epoch": 2.5631067961165046,
+      "grad_norm": 0.6679034911744651,
+      "learning_rate": 8.713809005643919e-06,
+      "loss": 0.1415,
+      "step": 726
+    },
+    {
+      "epoch": 2.566637246248897,
+      "grad_norm": 0.8156715771369578,
+      "learning_rate": 8.709975114440632e-06,
+      "loss": 0.1664,
+      "step": 727
+    },
+    {
+      "epoch": 2.5701676963812887,
+      "grad_norm": 0.950713499966215,
+      "learning_rate": 8.706136363799888e-06,
+      "loss": 0.1579,
+      "step": 728
+    },
+    {
+      "epoch": 2.5736981465136806,
+      "grad_norm": 0.6736037466099428,
+      "learning_rate": 8.7022927587498e-06,
+      "loss": 0.1376,
+      "step": 729
+    },
+    {
+      "epoch": 2.5772285966460724,
+      "grad_norm": 0.8007987125298717,
+      "learning_rate": 8.698444304324837e-06,
+      "loss": 0.1604,
+      "step": 730
+    },
+    {
+      "epoch": 2.5807590467784642,
+      "grad_norm": 0.7909053805653098,
+      "learning_rate": 8.694591005565818e-06,
+      "loss": 0.1525,
+      "step": 731
+    },
+    {
+      "epoch": 2.584289496910856,
+      "grad_norm": 0.7510519443516507,
+      "learning_rate": 8.690732867519915e-06,
+      "loss": 0.152,
+      "step": 732
+    },
+    {
+      "epoch": 2.587819947043248,
+      "grad_norm": 0.7070326571398531,
+      "learning_rate": 8.686869895240631e-06,
+      "loss": 0.1478,
+      "step": 733
+    },
+    {
+      "epoch": 2.5913503971756398,
+      "grad_norm": 0.7843881178771545,
+      "learning_rate": 8.683002093787805e-06,
+      "loss": 0.155,
+      "step": 734
+    },
+    {
+      "epoch": 2.5948808473080316,
+      "grad_norm": 0.74442690372609,
+      "learning_rate": 8.679129468227601e-06,
+      "loss": 0.1742,
+      "step": 735
+    },
+    {
+      "epoch": 2.598411297440424,
+      "grad_norm": 0.8873243162061539,
+      "learning_rate": 8.675252023632497e-06,
+      "loss": 0.1433,
+      "step": 736
+    },
+    {
+      "epoch": 2.6019417475728153,
+      "grad_norm": 1.0558814263186358,
+      "learning_rate": 8.671369765081291e-06,
+      "loss": 0.1383,
+      "step": 737
+    },
+    {
+      "epoch": 2.6054721977052075,
+      "grad_norm": 0.9661052979159337,
+      "learning_rate": 8.66748269765908e-06,
+      "loss": 0.1519,
+      "step": 738
+    },
+    {
+      "epoch": 2.6090026478375994,
+      "grad_norm": 0.7940230362582856,
+      "learning_rate": 8.663590826457264e-06,
+      "loss": 0.1449,
+      "step": 739
+    },
+    {
+      "epoch": 2.6125330979699912,
+      "grad_norm": 0.7712249737573764,
+      "learning_rate": 8.659694156573533e-06,
+      "loss": 0.1567,
+      "step": 740
+    },
+    {
+      "epoch": 2.616063548102383,
+      "grad_norm": 0.7354585280144065,
+      "learning_rate": 8.655792693111864e-06,
+      "loss": 0.147,
+      "step": 741
+    },
+    {
+      "epoch": 2.619593998234775,
+      "grad_norm": 0.7686478452231044,
+      "learning_rate": 8.651886441182509e-06,
+      "loss": 0.1462,
+      "step": 742
+    },
+    {
+      "epoch": 2.6231244483671667,
+      "grad_norm": 0.9615765356144627,
+      "learning_rate": 8.647975405901996e-06,
+      "loss": 0.1598,
+      "step": 743
+    },
+    {
+      "epoch": 2.6266548984995586,
+      "grad_norm": 0.7899472619604829,
+      "learning_rate": 8.644059592393117e-06,
+      "loss": 0.1473,
+      "step": 744
+    },
+    {
+      "epoch": 2.6301853486319504,
+      "grad_norm": 0.8716879941644811,
+      "learning_rate": 8.640139005784924e-06,
+      "loss": 0.1512,
+      "step": 745
+    },
+    {
+      "epoch": 2.6337157987643423,
+      "grad_norm": 0.9497477417594384,
+      "learning_rate": 8.636213651212719e-06,
+      "loss": 0.1501,
+      "step": 746
+    },
+    {
+      "epoch": 2.6372462488967345,
+      "grad_norm": 0.8255927742841409,
+      "learning_rate": 8.632283533818048e-06,
+      "loss": 0.147,
+      "step": 747
+    },
+    {
+      "epoch": 2.6407766990291264,
+      "grad_norm": 0.6589014957846149,
+      "learning_rate": 8.628348658748702e-06,
+      "loss": 0.146,
+      "step": 748
+    },
+    {
+      "epoch": 2.644307149161518,
+      "grad_norm": 0.7383869029734058,
+      "learning_rate": 8.624409031158696e-06,
+      "loss": 0.1444,
+      "step": 749
+    },
+    {
+      "epoch": 2.64783759929391,
+      "grad_norm": 0.8754969443335403,
+      "learning_rate": 8.620464656208272e-06,
+      "loss": 0.1533,
+      "step": 750
+    },
+    {
+      "epoch": 2.651368049426302,
+      "grad_norm": 0.8093544382322876,
+      "learning_rate": 8.616515539063894e-06,
+      "loss": 0.1466,
+      "step": 751
+    },
+    {
+      "epoch": 2.6548984995586937,
+      "grad_norm": 0.8094232555314647,
+      "learning_rate": 8.612561684898235e-06,
+      "loss": 0.1414,
+      "step": 752
+    },
+    {
+      "epoch": 2.6584289496910856,
+      "grad_norm": 0.9276446354153174,
+      "learning_rate": 8.60860309889017e-06,
+      "loss": 0.1608,
+      "step": 753
+    },
+    {
+      "epoch": 2.6619593998234774,
+      "grad_norm": 0.8670503330866877,
+      "learning_rate": 8.604639786224778e-06,
+      "loss": 0.148,
+      "step": 754
+    },
+    {
+      "epoch": 2.6654898499558692,
+      "grad_norm": 0.9236407101629923,
+      "learning_rate": 8.600671752093323e-06,
+      "loss": 0.1465,
+      "step": 755
+    },
+    {
+      "epoch": 2.6690203000882615,
+      "grad_norm": 0.7947533360687092,
+      "learning_rate": 8.596699001693257e-06,
+      "loss": 0.1382,
+      "step": 756
+    },
+    {
+      "epoch": 2.672550750220653,
+      "grad_norm": 0.9401531663981636,
+      "learning_rate": 8.592721540228207e-06,
+      "loss": 0.1545,
+      "step": 757
+    },
+    {
+      "epoch": 2.676081200353045,
+      "grad_norm": 0.6849210881232853,
+      "learning_rate": 8.588739372907974e-06,
+      "loss": 0.1422,
+      "step": 758
+    },
+    {
+      "epoch": 2.679611650485437,
+      "grad_norm": 0.9020222283025484,
+      "learning_rate": 8.584752504948522e-06,
+      "loss": 0.148,
+      "step": 759
+    },
+    {
+      "epoch": 2.683142100617829,
+      "grad_norm": 0.763357957155723,
+      "learning_rate": 8.580760941571968e-06,
+      "loss": 0.1387,
+      "step": 760
+    },
+    {
+      "epoch": 2.6866725507502207,
+      "grad_norm": 0.8677207545705664,
+      "learning_rate": 8.576764688006583e-06,
+      "loss": 0.1694,
+      "step": 761
+    },
+    {
+      "epoch": 2.6902030008826125,
+      "grad_norm": 0.7387471803373706,
+      "learning_rate": 8.572763749486784e-06,
+      "loss": 0.1507,
+      "step": 762
+    },
+    {
+      "epoch": 2.6937334510150044,
+      "grad_norm": 0.6986240692420573,
+      "learning_rate": 8.568758131253115e-06,
+      "loss": 0.1632,
+      "step": 763
+    },
+    {
+      "epoch": 2.6972639011473962,
+      "grad_norm": 0.9057220759646399,
+      "learning_rate": 8.564747838552262e-06,
+      "loss": 0.1607,
+      "step": 764
+    },
+    {
+      "epoch": 2.700794351279788,
+      "grad_norm": 0.8635819293383579,
+      "learning_rate": 8.560732876637023e-06,
+      "loss": 0.1446,
+      "step": 765
+    },
+    {
+      "epoch": 2.70432480141218,
+      "grad_norm": 0.773759396974051,
+      "learning_rate": 8.55671325076632e-06,
+      "loss": 0.152,
+      "step": 766
+    },
+    {
+      "epoch": 2.707855251544572,
+      "grad_norm": 0.7726341748367627,
+      "learning_rate": 8.552688966205176e-06,
+      "loss": 0.1507,
+      "step": 767
+    },
+    {
+      "epoch": 2.7113857016769636,
+      "grad_norm": 0.8172707078015327,
+      "learning_rate": 8.548660028224724e-06,
+      "loss": 0.1624,
+      "step": 768
+    },
+    {
+      "epoch": 2.714916151809356,
+      "grad_norm": 0.7096834505741385,
+      "learning_rate": 8.544626442102188e-06,
+      "loss": 0.1451,
+      "step": 769
+    },
+    {
+      "epoch": 2.7184466019417477,
+      "grad_norm": 0.7394117544451063,
+      "learning_rate": 8.54058821312088e-06,
+      "loss": 0.163,
+      "step": 770
+    },
+    {
+      "epoch": 2.7219770520741395,
+      "grad_norm": 0.9101784701617734,
+      "learning_rate": 8.536545346570195e-06,
+      "loss": 0.1597,
+      "step": 771
+    },
+    {
+      "epoch": 2.7255075022065314,
+      "grad_norm": 0.8528637952954311,
+      "learning_rate": 8.532497847745602e-06,
+      "loss": 0.1455,
+      "step": 772
+    },
+    {
+      "epoch": 2.729037952338923,
+      "grad_norm": 0.7128839210498725,
+      "learning_rate": 8.528445721948637e-06,
+      "loss": 0.1562,
+      "step": 773
+    },
+    {
+      "epoch": 2.732568402471315,
+      "grad_norm": 0.7514251001247261,
+      "learning_rate": 8.524388974486896e-06,
+      "loss": 0.1787,
+      "step": 774
+    },
+    {
+      "epoch": 2.736098852603707,
+      "grad_norm": 0.8717097957173104,
+      "learning_rate": 8.520327610674029e-06,
+      "loss": 0.1686,
+      "step": 775
+    },
+    {
+      "epoch": 2.7396293027360987,
+      "grad_norm": 0.8134061335902161,
+      "learning_rate": 8.516261635829734e-06,
+      "loss": 0.1444,
+      "step": 776
+    },
+    {
+      "epoch": 2.7431597528684906,
+      "grad_norm": 0.8130838438482634,
+      "learning_rate": 8.512191055279749e-06,
+      "loss": 0.1621,
+      "step": 777
+    },
+    {
+      "epoch": 2.746690203000883,
+      "grad_norm": 0.7609974982707016,
+      "learning_rate": 8.50811587435584e-06,
+      "loss": 0.1639,
+      "step": 778
+    },
+    {
+      "epoch": 2.7502206531332742,
+      "grad_norm": 0.8656038452478626,
+      "learning_rate": 8.504036098395805e-06,
+      "loss": 0.1584,
+      "step": 779
+    },
+    {
+      "epoch": 2.7537511032656665,
+      "grad_norm": 0.7488569534666748,
+      "learning_rate": 8.499951732743457e-06,
+      "loss": 0.148,
+      "step": 780
+    },
+    {
+      "epoch": 2.7572815533980584,
+      "grad_norm": 0.77448853187229,
+      "learning_rate": 8.495862782748619e-06,
+      "loss": 0.1406,
+      "step": 781
+    },
+    {
+      "epoch": 2.76081200353045,
+      "grad_norm": 0.6757102359665434,
+      "learning_rate": 8.491769253767126e-06,
+      "loss": 0.1301,
+      "step": 782
+    },
+    {
+      "epoch": 2.764342453662842,
+      "grad_norm": 0.7775212534985085,
+      "learning_rate": 8.487671151160801e-06,
+      "loss": 0.1595,
+      "step": 783
+    },
+    {
+      "epoch": 2.767872903795234,
+      "grad_norm": 0.7117920982524355,
+      "learning_rate": 8.483568480297467e-06,
+      "loss": 0.1378,
+      "step": 784
+    },
+    {
+      "epoch": 2.7714033539276257,
+      "grad_norm": 0.7937162920192259,
+      "learning_rate": 8.479461246550922e-06,
+      "loss": 0.152,
+      "step": 785
+    },
+    {
+      "epoch": 2.7749338040600176,
+      "grad_norm": 0.7164290770402717,
+      "learning_rate": 8.475349455300947e-06,
+      "loss": 0.1249,
+      "step": 786
+    },
+    {
+      "epoch": 2.7784642541924094,
+      "grad_norm": 0.7347153157580293,
+      "learning_rate": 8.471233111933291e-06,
+      "loss": 0.169,
+      "step": 787
+    },
+    {
+      "epoch": 2.7819947043248012,
+      "grad_norm": 0.6345954182380398,
+      "learning_rate": 8.467112221839664e-06,
+      "loss": 0.1476,
+      "step": 788
+    },
+    {
+      "epoch": 2.7855251544571935,
+      "grad_norm": 0.7710274326642901,
+      "learning_rate": 8.462986790417733e-06,
+      "loss": 0.1266,
+      "step": 789
+    },
+    {
+      "epoch": 2.789055604589585,
+      "grad_norm": 0.7730009077854115,
+      "learning_rate": 8.458856823071111e-06,
+      "loss": 0.1347,
+      "step": 790
+    },
+    {
+      "epoch": 2.792586054721977,
+      "grad_norm": 0.9320951665857266,
+      "learning_rate": 8.454722325209354e-06,
+      "loss": 0.1544,
+      "step": 791
+    },
+    {
+      "epoch": 2.796116504854369,
+      "grad_norm": 0.854024287144252,
+      "learning_rate": 8.450583302247954e-06,
+      "loss": 0.1776,
+      "step": 792
+    },
+    {
+      "epoch": 2.799646954986761,
+      "grad_norm": 0.8267045018857162,
+      "learning_rate": 8.44643975960833e-06,
+      "loss": 0.1805,
+      "step": 793
+    },
+    {
+      "epoch": 2.8031774051191527,
+      "grad_norm": 0.8689865555638648,
+      "learning_rate": 8.442291702717814e-06,
+      "loss": 0.1643,
+      "step": 794
+    },
+    {
+      "epoch": 2.8067078552515445,
+      "grad_norm": 0.78855645463223,
+      "learning_rate": 8.43813913700966e-06,
+      "loss": 0.1663,
+      "step": 795
+    },
+    {
+      "epoch": 2.8102383053839364,
+      "grad_norm": 0.8423324894257369,
+      "learning_rate": 8.433982067923021e-06,
+      "loss": 0.1539,
+      "step": 796
+    },
+    {
+      "epoch": 2.813768755516328,
+      "grad_norm": 0.6841465341626565,
+      "learning_rate": 8.429820500902957e-06,
+      "loss": 0.1707,
+      "step": 797
+    },
+    {
+      "epoch": 2.81729920564872,
+      "grad_norm": 0.637410884505517,
+      "learning_rate": 8.425654441400408e-06,
+      "loss": 0.1631,
+      "step": 798
+    },
+    {
+      "epoch": 2.820829655781112,
+      "grad_norm": 0.7435035501128142,
+      "learning_rate": 8.421483894872208e-06,
+      "loss": 0.1518,
+      "step": 799
+    },
+    {
+      "epoch": 2.824360105913504,
+      "grad_norm": 0.7390502021511453,
+      "learning_rate": 8.417308866781065e-06,
+      "loss": 0.1513,
+      "step": 800
+    },
+    {
+      "epoch": 2.827890556045896,
+      "grad_norm": 0.7074112942115228,
+      "learning_rate": 8.413129362595555e-06,
+      "loss": 0.1626,
+      "step": 801
+    },
+    {
+      "epoch": 2.831421006178288,
+      "grad_norm": 0.8297265070014697,
+      "learning_rate": 8.408945387790121e-06,
+      "loss": 0.1447,
+      "step": 802
+    },
+    {
+      "epoch": 2.8349514563106797,
+      "grad_norm": 0.6691649963937963,
+      "learning_rate": 8.404756947845056e-06,
+      "loss": 0.1453,
+      "step": 803
+    },
+    {
+      "epoch": 2.8384819064430715,
+      "grad_norm": 0.7275626975495146,
+      "learning_rate": 8.40056404824651e-06,
+      "loss": 0.162,
+      "step": 804
+    },
+    {
+      "epoch": 2.8420123565754634,
+      "grad_norm": 0.9804621444439445,
+      "learning_rate": 8.396366694486466e-06,
+      "loss": 0.1711,
+      "step": 805
+    },
+    {
+      "epoch": 2.845542806707855,
+      "grad_norm": 0.7841205320098567,
+      "learning_rate": 8.392164892062746e-06,
+      "loss": 0.1682,
+      "step": 806
+    },
+    {
+      "epoch": 2.849073256840247,
+      "grad_norm": 0.7259305687219134,
+      "learning_rate": 8.387958646478998e-06,
+      "loss": 0.1349,
+      "step": 807
+    },
+    {
+      "epoch": 2.852603706972639,
+      "grad_norm": 0.6285666615213146,
+      "learning_rate": 8.383747963244688e-06,
+      "loss": 0.1622,
+      "step": 808
+    },
+    {
+      "epoch": 2.856134157105031,
+      "grad_norm": 0.8191314538526839,
+      "learning_rate": 8.379532847875099e-06,
+      "loss": 0.1765,
+      "step": 809
+    },
+    {
+      "epoch": 2.8596646072374226,
+      "grad_norm": 0.7451704716031862,
+      "learning_rate": 8.375313305891312e-06,
+      "loss": 0.1451,
+      "step": 810
+    },
+    {
+      "epoch": 2.863195057369815,
+      "grad_norm": 0.6744145077220661,
+      "learning_rate": 8.371089342820216e-06,
+      "loss": 0.1643,
+      "step": 811
+    },
+    {
+      "epoch": 2.8667255075022067,
+      "grad_norm": 0.6887811045404645,
+      "learning_rate": 8.366860964194479e-06,
+      "loss": 0.1522,
+      "step": 812
+    },
+    {
+      "epoch": 2.8702559576345985,
+      "grad_norm": 0.63573464613807,
+      "learning_rate": 8.362628175552566e-06,
+      "loss": 0.1511,
+      "step": 813
+    },
+    {
+      "epoch": 2.8737864077669903,
+      "grad_norm": 0.6897759046360326,
+      "learning_rate": 8.358390982438706e-06,
+      "loss": 0.1434,
+      "step": 814
+    },
+    {
+      "epoch": 2.877316857899382,
+      "grad_norm": 0.7279048382426626,
+      "learning_rate": 8.354149390402904e-06,
+      "loss": 0.1351,
+      "step": 815
+    },
+    {
+      "epoch": 2.880847308031774,
+      "grad_norm": 0.7458563780785276,
+      "learning_rate": 8.349903405000926e-06,
+      "loss": 0.1517,
+      "step": 816
+    },
+    {
+      "epoch": 2.884377758164166,
+      "grad_norm": 0.7101890658117982,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.1462,
+      "step": 817
+    },
+    {
+      "epoch": 2.8879082082965577,
+      "grad_norm": 0.7742297467061107,
+      "learning_rate": 8.341398276350268e-06,
+      "loss": 0.1555,
+      "step": 818
+    },
+    {
+      "epoch": 2.8914386584289495,
+      "grad_norm": 0.7438278623504806,
+      "learning_rate": 8.337139144241864e-06,
+      "loss": 0.1396,
+      "step": 819
+    },
+    {
+      "epoch": 2.894969108561342,
+      "grad_norm": 0.724140366387057,
+      "learning_rate": 8.332875641047817e-06,
+      "loss": 0.1361,
+      "step": 820
+    },
+    {
+      "epoch": 2.898499558693733,
+      "grad_norm": 0.7091482359305543,
+      "learning_rate": 8.328607772352593e-06,
+      "loss": 0.1477,
+      "step": 821
+    },
+    {
+      "epoch": 2.9020300088261255,
+      "grad_norm": 0.714577592926291,
+      "learning_rate": 8.324335543746377e-06,
+      "loss": 0.155,
+      "step": 822
+    },
+    {
+      "epoch": 2.9055604589585173,
+      "grad_norm": 0.6821542598622823,
+      "learning_rate": 8.32005896082506e-06,
+      "loss": 0.1449,
+      "step": 823
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 0.7674294440664721,
+      "learning_rate": 8.31577802919024e-06,
+      "loss": 0.162,
+      "step": 824
+    },
+    {
+      "epoch": 2.912621359223301,
+      "grad_norm": 0.6635918469316766,
+      "learning_rate": 8.311492754449213e-06,
+      "loss": 0.1445,
+      "step": 825
+    },
+    {
+      "epoch": 2.916151809355693,
+      "grad_norm": 0.7039005991376192,
+      "learning_rate": 8.307203142214959e-06,
+      "loss": 0.1473,
+      "step": 826
+    },
+    {
+      "epoch": 2.9196822594880847,
+      "grad_norm": 0.7210854795121673,
+      "learning_rate": 8.302909198106144e-06,
+      "loss": 0.1431,
+      "step": 827
+    },
+    {
+      "epoch": 2.9232127096204765,
+      "grad_norm": 0.8804727621267001,
+      "learning_rate": 8.298610927747104e-06,
+      "loss": 0.1514,
+      "step": 828
+    },
+    {
+      "epoch": 2.9267431597528684,
+      "grad_norm": 0.6567640224299616,
+      "learning_rate": 8.294308336767843e-06,
+      "loss": 0.1356,
+      "step": 829
+    },
+    {
+      "epoch": 2.93027360988526,
+      "grad_norm": 0.7266522819100054,
+      "learning_rate": 8.290001430804026e-06,
+      "loss": 0.1515,
+      "step": 830
+    },
+    {
+      "epoch": 2.9338040600176525,
+      "grad_norm": 0.7006333424868092,
+      "learning_rate": 8.285690215496966e-06,
+      "loss": 0.1518,
+      "step": 831
+    },
+    {
+      "epoch": 2.937334510150044,
+      "grad_norm": 1.0457753137928258,
+      "learning_rate": 8.281374696493628e-06,
+      "loss": 0.1623,
+      "step": 832
+    },
+    {
+      "epoch": 2.940864960282436,
+      "grad_norm": 0.7835519475042352,
+      "learning_rate": 8.277054879446603e-06,
+      "loss": 0.1512,
+      "step": 833
+    },
+    {
+      "epoch": 2.944395410414828,
+      "grad_norm": 0.694852963337788,
+      "learning_rate": 8.27273077001412e-06,
+      "loss": 0.1408,
+      "step": 834
+    },
+    {
+      "epoch": 2.94792586054722,
+      "grad_norm": 0.6914754193933184,
+      "learning_rate": 8.26840237386003e-06,
+      "loss": 0.1553,
+      "step": 835
+    },
+    {
+      "epoch": 2.9514563106796117,
+      "grad_norm": 0.7640107634127185,
+      "learning_rate": 8.264069696653795e-06,
+      "loss": 0.1563,
+      "step": 836
+    },
+    {
+      "epoch": 2.9549867608120035,
+      "grad_norm": 0.7031909948366248,
+      "learning_rate": 8.25973274407049e-06,
+      "loss": 0.1513,
+      "step": 837
+    },
+    {
+      "epoch": 2.9585172109443953,
+      "grad_norm": 0.7212903191170139,
+      "learning_rate": 8.255391521790782e-06,
+      "loss": 0.1452,
+      "step": 838
+    },
+    {
+      "epoch": 2.962047661076787,
+      "grad_norm": 0.9288624940870822,
+      "learning_rate": 8.251046035500937e-06,
+      "loss": 0.1611,
+      "step": 839
+    },
+    {
+      "epoch": 2.965578111209179,
+      "grad_norm": 0.8307453353673667,
+      "learning_rate": 8.246696290892804e-06,
+      "loss": 0.1511,
+      "step": 840
+    },
+    {
+      "epoch": 2.969108561341571,
+      "grad_norm": 0.7346059651517513,
+      "learning_rate": 8.24234229366381e-06,
+      "loss": 0.1645,
+      "step": 841
+    },
+    {
+      "epoch": 2.972639011473963,
+      "grad_norm": 0.8211385940924918,
+      "learning_rate": 8.237984049516952e-06,
+      "loss": 0.1576,
+      "step": 842
+    },
+    {
+      "epoch": 2.9761694616063545,
+      "grad_norm": 0.6365132434025104,
+      "learning_rate": 8.23362156416079e-06,
+      "loss": 0.1228,
+      "step": 843
+    },
+    {
+      "epoch": 2.979699911738747,
+      "grad_norm": 0.8824809696079167,
+      "learning_rate": 8.229254843309438e-06,
+      "loss": 0.1509,
+      "step": 844
+    },
+    {
+      "epoch": 2.9832303618711387,
+      "grad_norm": 0.8601855800826128,
+      "learning_rate": 8.22488389268256e-06,
+      "loss": 0.1615,
+      "step": 845
+    },
+    {
+      "epoch": 2.9867608120035305,
+      "grad_norm": 0.731123229093688,
+      "learning_rate": 8.220508718005359e-06,
+      "loss": 0.1378,
+      "step": 846
+    },
+    {
+      "epoch": 2.9902912621359223,
+      "grad_norm": 0.6227120430460796,
+      "learning_rate": 8.21612932500857e-06,
+      "loss": 0.1399,
+      "step": 847
+    },
+    {
+      "epoch": 2.993821712268314,
+      "grad_norm": 0.7545771750137871,
+      "learning_rate": 8.211745719428456e-06,
+      "loss": 0.1701,
+      "step": 848
+    },
+    {
+      "epoch": 2.997352162400706,
+      "grad_norm": 0.8046466791224096,
+      "learning_rate": 8.207357907006794e-06,
+      "loss": 0.1549,
+      "step": 849
+    },
+    {
+      "epoch": 3.000882612533098,
+      "grad_norm": 0.7642023797697144,
+      "learning_rate": 8.202965893490877e-06,
+      "loss": 0.1368,
+      "step": 850
+    },
+    {
+      "epoch": 3.0044130626654897,
+      "grad_norm": 0.6716017754172908,
+      "learning_rate": 8.198569684633492e-06,
+      "loss": 0.1162,
+      "step": 851
+    },
+    {
+      "epoch": 3.0079435127978815,
+      "grad_norm": 0.6431034665161904,
+      "learning_rate": 8.194169286192928e-06,
+      "loss": 0.1195,
+      "step": 852
+    },
+    {
+      "epoch": 3.011473962930274,
+      "grad_norm": 0.7143235899902817,
+      "learning_rate": 8.18976470393296e-06,
+      "loss": 0.112,
+      "step": 853
+    },
+    {
+      "epoch": 3.0150044130626656,
+      "grad_norm": 0.698231889557549,
+      "learning_rate": 8.185355943622845e-06,
+      "loss": 0.1234,
+      "step": 854
+    },
+    {
+      "epoch": 3.0185348631950575,
+      "grad_norm": 0.9000098025363069,
+      "learning_rate": 8.180943011037306e-06,
+      "loss": 0.1249,
+      "step": 855
+    },
+    {
+      "epoch": 3.0220653133274493,
+      "grad_norm": 0.8015546761444414,
+      "learning_rate": 8.176525911956539e-06,
+      "loss": 0.1245,
+      "step": 856
+    },
+    {
+      "epoch": 3.025595763459841,
+      "grad_norm": 0.8094451013988107,
+      "learning_rate": 8.172104652166194e-06,
+      "loss": 0.1018,
+      "step": 857
+    },
+    {
+      "epoch": 3.029126213592233,
+      "grad_norm": 0.720799250577229,
+      "learning_rate": 8.167679237457368e-06,
+      "loss": 0.0992,
+      "step": 858
+    },
+    {
+      "epoch": 3.032656663724625,
+      "grad_norm": 0.8000553496291071,
+      "learning_rate": 8.163249673626603e-06,
+      "loss": 0.1194,
+      "step": 859
+    },
+    {
+      "epoch": 3.0361871138570167,
+      "grad_norm": 0.8669635173424244,
+      "learning_rate": 8.158815966475877e-06,
+      "loss": 0.0816,
+      "step": 860
+    },
+    {
+      "epoch": 3.0397175639894085,
+      "grad_norm": 0.9025742229651821,
+      "learning_rate": 8.154378121812593e-06,
+      "loss": 0.1018,
+      "step": 861
+    },
+    {
+      "epoch": 3.0432480141218003,
+      "grad_norm": 0.9845870522804518,
+      "learning_rate": 8.149936145449575e-06,
+      "loss": 0.1251,
+      "step": 862
+    },
+    {
+      "epoch": 3.046778464254192,
+      "grad_norm": 1.0256068011823871,
+      "learning_rate": 8.145490043205056e-06,
+      "loss": 0.1239,
+      "step": 863
+    },
+    {
+      "epoch": 3.0503089143865845,
+      "grad_norm": 1.148156571965961,
+      "learning_rate": 8.141039820902676e-06,
+      "loss": 0.1104,
+      "step": 864
+    },
+    {
+      "epoch": 3.0538393645189763,
+      "grad_norm": 1.2193696303353918,
+      "learning_rate": 8.136585484371471e-06,
+      "loss": 0.1152,
+      "step": 865
+    },
+    {
+      "epoch": 3.057369814651368,
+      "grad_norm": 1.2915877393106403,
+      "learning_rate": 8.132127039445864e-06,
+      "loss": 0.118,
+      "step": 866
+    },
+    {
+      "epoch": 3.06090026478376,
+      "grad_norm": 1.2149067405554672,
+      "learning_rate": 8.127664491965662e-06,
+      "loss": 0.1142,
+      "step": 867
+    },
+    {
+      "epoch": 3.064430714916152,
+      "grad_norm": 0.990566681159559,
+      "learning_rate": 8.123197847776043e-06,
+      "loss": 0.1078,
+      "step": 868
+    },
+    {
+      "epoch": 3.0679611650485437,
+      "grad_norm": 0.9565804354564945,
+      "learning_rate": 8.118727112727556e-06,
+      "loss": 0.1135,
+      "step": 869
+    },
+    {
+      "epoch": 3.0714916151809355,
+      "grad_norm": 1.0200728098865612,
+      "learning_rate": 8.1142522926761e-06,
+      "loss": 0.1195,
+      "step": 870
+    },
+    {
+      "epoch": 3.0750220653133273,
+      "grad_norm": 1.0243439506618488,
+      "learning_rate": 8.109773393482934e-06,
+      "loss": 0.1024,
+      "step": 871
+    },
+    {
+      "epoch": 3.078552515445719,
+      "grad_norm": 1.0049145977279004,
+      "learning_rate": 8.105290421014653e-06,
+      "loss": 0.1076,
+      "step": 872
+    },
+    {
+      "epoch": 3.082082965578111,
+      "grad_norm": 0.8112529372591591,
+      "learning_rate": 8.10080338114319e-06,
+      "loss": 0.0982,
+      "step": 873
+    },
+    {
+      "epoch": 3.0856134157105033,
+      "grad_norm": 1.0002477688608336,
+      "learning_rate": 8.096312279745804e-06,
+      "loss": 0.13,
+      "step": 874
+    },
+    {
+      "epoch": 3.089143865842895,
+      "grad_norm": 1.0623544903602675,
+      "learning_rate": 8.09181712270508e-06,
+      "loss": 0.1025,
+      "step": 875
+    },
+    {
+      "epoch": 3.092674315975287,
+      "grad_norm": 1.0410125624967994,
+      "learning_rate": 8.087317915908904e-06,
+      "loss": 0.1063,
+      "step": 876
+    },
+    {
+      "epoch": 3.096204766107679,
+      "grad_norm": 1.0118590449093119,
+      "learning_rate": 8.082814665250476e-06,
+      "loss": 0.0954,
+      "step": 877
+    },
+    {
+      "epoch": 3.0997352162400706,
+      "grad_norm": 0.9546611695399589,
+      "learning_rate": 8.078307376628292e-06,
+      "loss": 0.1245,
+      "step": 878
+    },
+    {
+      "epoch": 3.1032656663724625,
+      "grad_norm": 0.9269391315446083,
+      "learning_rate": 8.073796055946133e-06,
+      "loss": 0.1174,
+      "step": 879
+    },
+    {
+      "epoch": 3.1067961165048543,
+      "grad_norm": 1.0305117514176123,
+      "learning_rate": 8.06928070911306e-06,
+      "loss": 0.1026,
+      "step": 880
+    },
+    {
+      "epoch": 3.110326566637246,
+      "grad_norm": 1.208839666085655,
+      "learning_rate": 8.064761342043415e-06,
+      "loss": 0.1248,
+      "step": 881
+    },
+    {
+      "epoch": 3.113857016769638,
+      "grad_norm": 0.952513604276628,
+      "learning_rate": 8.060237960656799e-06,
+      "loss": 0.1085,
+      "step": 882
+    },
+    {
+      "epoch": 3.11738746690203,
+      "grad_norm": 0.8730540650291831,
+      "learning_rate": 8.055710570878076e-06,
+      "loss": 0.1134,
+      "step": 883
+    },
+    {
+      "epoch": 3.120917917034422,
+      "grad_norm": 1.006482604780559,
+      "learning_rate": 8.051179178637356e-06,
+      "loss": 0.1049,
+      "step": 884
+    },
+    {
+      "epoch": 3.124448367166814,
+      "grad_norm": 0.9112291303575337,
+      "learning_rate": 8.046643789869991e-06,
+      "loss": 0.108,
+      "step": 885
+    },
+    {
+      "epoch": 3.127978817299206,
+      "grad_norm": 1.0693444544873563,
+      "learning_rate": 8.042104410516576e-06,
+      "loss": 0.1313,
+      "step": 886
+    },
+    {
+      "epoch": 3.1315092674315976,
+      "grad_norm": 1.0053001607105967,
+      "learning_rate": 8.037561046522923e-06,
+      "loss": 0.1142,
+      "step": 887
+    },
+    {
+      "epoch": 3.1350397175639895,
+      "grad_norm": 0.9933955602184302,
+      "learning_rate": 8.033013703840067e-06,
+      "loss": 0.1278,
+      "step": 888
+    },
+    {
+      "epoch": 3.1385701676963813,
+      "grad_norm": 1.0821161416075462,
+      "learning_rate": 8.028462388424258e-06,
+      "loss": 0.1213,
+      "step": 889
+    },
+    {
+      "epoch": 3.142100617828773,
+      "grad_norm": 1.0448661581558534,
+      "learning_rate": 8.023907106236944e-06,
+      "loss": 0.1016,
+      "step": 890
+    },
+    {
+      "epoch": 3.145631067961165,
+      "grad_norm": 1.1354463844063831,
+      "learning_rate": 8.019347863244773e-06,
+      "loss": 0.1198,
+      "step": 891
+    },
+    {
+      "epoch": 3.149161518093557,
+      "grad_norm": 1.008070141194459,
+      "learning_rate": 8.014784665419577e-06,
+      "loss": 0.1252,
+      "step": 892
+    },
+    {
+      "epoch": 3.1526919682259487,
+      "grad_norm": 1.06009230334142,
+      "learning_rate": 8.01021751873837e-06,
+      "loss": 0.1286,
+      "step": 893
+    },
+    {
+      "epoch": 3.1562224183583405,
+      "grad_norm": 1.138468713844925,
+      "learning_rate": 8.005646429183342e-06,
+      "loss": 0.1129,
+      "step": 894
+    },
+    {
+      "epoch": 3.159752868490733,
+      "grad_norm": 1.2328861278363776,
+      "learning_rate": 8.001071402741843e-06,
+      "loss": 0.1235,
+      "step": 895
+    },
+    {
+      "epoch": 3.1632833186231246,
+      "grad_norm": 1.0853132184834922,
+      "learning_rate": 7.996492445406377e-06,
+      "loss": 0.1282,
+      "step": 896
+    },
+    {
+      "epoch": 3.1668137687555165,
+      "grad_norm": 0.933433925231955,
+      "learning_rate": 7.991909563174605e-06,
+      "loss": 0.1325,
+      "step": 897
+    },
+    {
+      "epoch": 3.1703442188879083,
+      "grad_norm": 0.8319438313846836,
+      "learning_rate": 7.987322762049323e-06,
+      "loss": 0.0987,
+      "step": 898
+    },
+    {
+      "epoch": 3.1738746690203,
+      "grad_norm": 0.9872815349839481,
+      "learning_rate": 7.982732048038462e-06,
+      "loss": 0.1195,
+      "step": 899
+    },
+    {
+      "epoch": 3.177405119152692,
+      "grad_norm": 0.9612402010886701,
+      "learning_rate": 7.978137427155078e-06,
+      "loss": 0.1258,
+      "step": 900
+    },
+    {
+      "epoch": 3.180935569285084,
+      "grad_norm": 0.8927263036280142,
+      "learning_rate": 7.973538905417344e-06,
+      "loss": 0.1072,
+      "step": 901
+    },
+    {
+      "epoch": 3.1844660194174756,
+      "grad_norm": 0.8841286752626785,
+      "learning_rate": 7.968936488848541e-06,
+      "loss": 0.1047,
+      "step": 902
+    },
+    {
+      "epoch": 3.1879964695498675,
+      "grad_norm": 0.912445757407591,
+      "learning_rate": 7.964330183477057e-06,
+      "loss": 0.1276,
+      "step": 903
+    },
+    {
+      "epoch": 3.1915269196822593,
+      "grad_norm": 1.059803669034267,
+      "learning_rate": 7.959719995336364e-06,
+      "loss": 0.1032,
+      "step": 904
+    },
+    {
+      "epoch": 3.195057369814651,
+      "grad_norm": 0.9513868377440023,
+      "learning_rate": 7.955105930465032e-06,
+      "loss": 0.0911,
+      "step": 905
+    },
+    {
+      "epoch": 3.1985878199470434,
+      "grad_norm": 0.8797745669418451,
+      "learning_rate": 7.9504879949067e-06,
+      "loss": 0.1143,
+      "step": 906
+    },
+    {
+      "epoch": 3.2021182700794353,
+      "grad_norm": 1.0029598114248315,
+      "learning_rate": 7.945866194710079e-06,
+      "loss": 0.1161,
+      "step": 907
+    },
+    {
+      "epoch": 3.205648720211827,
+      "grad_norm": 1.1319418617063524,
+      "learning_rate": 7.941240535928942e-06,
+      "loss": 0.1268,
+      "step": 908
+    },
+    {
+      "epoch": 3.209179170344219,
+      "grad_norm": 0.9958247294998045,
+      "learning_rate": 7.936611024622117e-06,
+      "loss": 0.1074,
+      "step": 909
+    },
+    {
+      "epoch": 3.212709620476611,
+      "grad_norm": 1.0922819941075517,
+      "learning_rate": 7.93197766685348e-06,
+      "loss": 0.1179,
+      "step": 910
+    },
+    {
+      "epoch": 3.2162400706090026,
+      "grad_norm": 1.136941227912527,
+      "learning_rate": 7.92734046869194e-06,
+      "loss": 0.1331,
+      "step": 911
+    },
+    {
+      "epoch": 3.2197705207413945,
+      "grad_norm": 1.0047937727286098,
+      "learning_rate": 7.922699436211438e-06,
+      "loss": 0.1104,
+      "step": 912
+    },
+    {
+      "epoch": 3.2233009708737863,
+      "grad_norm": 0.9964788630911344,
+      "learning_rate": 7.918054575490943e-06,
+      "loss": 0.1252,
+      "step": 913
+    },
+    {
+      "epoch": 3.226831421006178,
+      "grad_norm": 1.0926246384683764,
+      "learning_rate": 7.913405892614433e-06,
+      "loss": 0.1044,
+      "step": 914
+    },
+    {
+      "epoch": 3.23036187113857,
+      "grad_norm": 0.9688331888143029,
+      "learning_rate": 7.908753393670891e-06,
+      "loss": 0.0945,
+      "step": 915
+    },
+    {
+      "epoch": 3.233892321270962,
+      "grad_norm": 1.111353673933071,
+      "learning_rate": 7.904097084754301e-06,
+      "loss": 0.1148,
+      "step": 916
+    },
+    {
+      "epoch": 3.237422771403354,
+      "grad_norm": 1.092326321596579,
+      "learning_rate": 7.899436971963637e-06,
+      "loss": 0.1276,
+      "step": 917
+    },
+    {
+      "epoch": 3.240953221535746,
+      "grad_norm": 0.7874168063997805,
+      "learning_rate": 7.894773061402855e-06,
+      "loss": 0.0932,
+      "step": 918
+    },
+    {
+      "epoch": 3.244483671668138,
+      "grad_norm": 0.9597684028837579,
+      "learning_rate": 7.89010535918089e-06,
+      "loss": 0.1158,
+      "step": 919
+    },
+    {
+      "epoch": 3.2480141218005296,
+      "grad_norm": 1.0816606744001607,
+      "learning_rate": 7.885433871411634e-06,
+      "loss": 0.1204,
+      "step": 920
+    },
+    {
+      "epoch": 3.2515445719329215,
+      "grad_norm": 1.035099273475842,
+      "learning_rate": 7.880758604213945e-06,
+      "loss": 0.125,
+      "step": 921
+    },
+    {
+      "epoch": 3.2550750220653133,
+      "grad_norm": 0.8982660262483314,
+      "learning_rate": 7.876079563711631e-06,
+      "loss": 0.104,
+      "step": 922
+    },
+    {
+      "epoch": 3.258605472197705,
+      "grad_norm": 1.0276050394216014,
+      "learning_rate": 7.871396756033438e-06,
+      "loss": 0.119,
+      "step": 923
+    },
+    {
+      "epoch": 3.262135922330097,
+      "grad_norm": 1.0128946832960732,
+      "learning_rate": 7.866710187313048e-06,
+      "loss": 0.118,
+      "step": 924
+    },
+    {
+      "epoch": 3.265666372462489,
+      "grad_norm": 0.8704316103506846,
+      "learning_rate": 7.862019863689075e-06,
+      "loss": 0.1139,
+      "step": 925
+    },
+    {
+      "epoch": 3.269196822594881,
+      "grad_norm": 0.9671125371629974,
+      "learning_rate": 7.857325791305042e-06,
+      "loss": 0.137,
+      "step": 926
+    },
+    {
+      "epoch": 3.2727272727272725,
+      "grad_norm": 0.868940825788469,
+      "learning_rate": 7.852627976309388e-06,
+      "loss": 0.1124,
+      "step": 927
+    },
+    {
+      "epoch": 3.2762577228596648,
+      "grad_norm": 0.9630053987036548,
+      "learning_rate": 7.847926424855454e-06,
+      "loss": 0.1253,
+      "step": 928
+    },
+    {
+      "epoch": 3.2797881729920566,
+      "grad_norm": 1.0029586796082006,
+      "learning_rate": 7.843221143101474e-06,
+      "loss": 0.1137,
+      "step": 929
+    },
+    {
+      "epoch": 3.2833186231244484,
+      "grad_norm": 0.8820085770728214,
+      "learning_rate": 7.838512137210565e-06,
+      "loss": 0.1255,
+      "step": 930
+    },
+    {
+      "epoch": 3.2868490732568403,
+      "grad_norm": 0.9828634895838252,
+      "learning_rate": 7.833799413350732e-06,
+      "loss": 0.1035,
+      "step": 931
+    },
+    {
+      "epoch": 3.290379523389232,
+      "grad_norm": 1.319511988574606,
+      "learning_rate": 7.829082977694835e-06,
+      "loss": 0.1325,
+      "step": 932
+    },
+    {
+      "epoch": 3.293909973521624,
+      "grad_norm": 1.2144177327586667,
+      "learning_rate": 7.824362836420609e-06,
+      "loss": 0.1403,
+      "step": 933
+    },
+    {
+      "epoch": 3.297440423654016,
+      "grad_norm": 1.0719443808897844,
+      "learning_rate": 7.819638995710636e-06,
+      "loss": 0.1185,
+      "step": 934
+    },
+    {
+      "epoch": 3.3009708737864076,
+      "grad_norm": 1.0518286795857918,
+      "learning_rate": 7.814911461752346e-06,
+      "loss": 0.1065,
+      "step": 935
+    },
+    {
+      "epoch": 3.3045013239187995,
+      "grad_norm": 1.024493911176814,
+      "learning_rate": 7.810180240738004e-06,
+      "loss": 0.1151,
+      "step": 936
+    },
+    {
+      "epoch": 3.3080317740511918,
+      "grad_norm": 1.127575182515073,
+      "learning_rate": 7.805445338864709e-06,
+      "loss": 0.1308,
+      "step": 937
+    },
+    {
+      "epoch": 3.3115622241835836,
+      "grad_norm": 0.9456157508829596,
+      "learning_rate": 7.800706762334375e-06,
+      "loss": 0.107,
+      "step": 938
+    },
+    {
+      "epoch": 3.3150926743159754,
+      "grad_norm": 0.9074545977721913,
+      "learning_rate": 7.795964517353734e-06,
+      "loss": 0.1138,
+      "step": 939
+    },
+    {
+      "epoch": 3.3186231244483673,
+      "grad_norm": 1.094922914297947,
+      "learning_rate": 7.791218610134324e-06,
+      "loss": 0.1349,
+      "step": 940
+    },
+    {
+      "epoch": 3.322153574580759,
+      "grad_norm": 1.133601456805787,
+      "learning_rate": 7.786469046892475e-06,
+      "loss": 0.1069,
+      "step": 941
+    },
+    {
+      "epoch": 3.325684024713151,
+      "grad_norm": 1.0182500924518352,
+      "learning_rate": 7.781715833849311e-06,
+      "loss": 0.1154,
+      "step": 942
+    },
+    {
+      "epoch": 3.329214474845543,
+      "grad_norm": 1.046201515291234,
+      "learning_rate": 7.776958977230731e-06,
+      "loss": 0.1251,
+      "step": 943
+    },
+    {
+      "epoch": 3.3327449249779346,
+      "grad_norm": 0.957386976461524,
+      "learning_rate": 7.77219848326741e-06,
+      "loss": 0.1101,
+      "step": 944
+    },
+    {
+      "epoch": 3.3362753751103265,
+      "grad_norm": 1.1740948292667477,
+      "learning_rate": 7.76743435819479e-06,
+      "loss": 0.1443,
+      "step": 945
+    },
+    {
+      "epoch": 3.3398058252427183,
+      "grad_norm": 1.850891607235183,
+      "learning_rate": 7.76266660825306e-06,
+      "loss": 0.1018,
+      "step": 946
+    },
+    {
+      "epoch": 3.34333627537511,
+      "grad_norm": 0.9097207671724479,
+      "learning_rate": 7.757895239687168e-06,
+      "loss": 0.1041,
+      "step": 947
+    },
+    {
+      "epoch": 3.3468667255075024,
+      "grad_norm": 0.9016475882306597,
+      "learning_rate": 7.753120258746793e-06,
+      "loss": 0.1151,
+      "step": 948
+    },
+    {
+      "epoch": 3.3503971756398943,
+      "grad_norm": 1.399326889021019,
+      "learning_rate": 7.748341671686355e-06,
+      "loss": 0.1361,
+      "step": 949
+    },
+    {
+      "epoch": 3.353927625772286,
+      "grad_norm": 0.8958929274360947,
+      "learning_rate": 7.743559484764987e-06,
+      "loss": 0.109,
+      "step": 950
+    },
+    {
+      "epoch": 3.357458075904678,
+      "grad_norm": 0.9661628430088748,
+      "learning_rate": 7.73877370424654e-06,
+      "loss": 0.1184,
+      "step": 951
+    },
+    {
+      "epoch": 3.3609885260370698,
+      "grad_norm": 1.0856106807644708,
+      "learning_rate": 7.733984336399578e-06,
+      "loss": 0.1258,
+      "step": 952
+    },
+    {
+      "epoch": 3.3645189761694616,
+      "grad_norm": 1.1339299007520192,
+      "learning_rate": 7.729191387497357e-06,
+      "loss": 0.1313,
+      "step": 953
+    },
+    {
+      "epoch": 3.3680494263018534,
+      "grad_norm": 0.9462346995224554,
+      "learning_rate": 7.72439486381783e-06,
+      "loss": 0.1299,
+      "step": 954
+    },
+    {
+      "epoch": 3.3715798764342453,
+      "grad_norm": 0.9773900040503245,
+      "learning_rate": 7.719594771643623e-06,
+      "loss": 0.115,
+      "step": 955
+    },
+    {
+      "epoch": 3.375110326566637,
+      "grad_norm": 0.8812615869761742,
+      "learning_rate": 7.714791117262047e-06,
+      "loss": 0.1178,
+      "step": 956
+    },
+    {
+      "epoch": 3.378640776699029,
+      "grad_norm": 0.872321741116879,
+      "learning_rate": 7.709983906965071e-06,
+      "loss": 0.1177,
+      "step": 957
+    },
+    {
+      "epoch": 3.382171226831421,
+      "grad_norm": 0.9590632048655139,
+      "learning_rate": 7.705173147049326e-06,
+      "loss": 0.1209,
+      "step": 958
+    },
+    {
+      "epoch": 3.385701676963813,
+      "grad_norm": 0.9774010811015693,
+      "learning_rate": 7.700358843816087e-06,
+      "loss": 0.0988,
+      "step": 959
+    },
+    {
+      "epoch": 3.389232127096205,
+      "grad_norm": 1.05084415330861,
+      "learning_rate": 7.695541003571275e-06,
+      "loss": 0.1281,
+      "step": 960
+    },
+    {
+      "epoch": 3.3927625772285968,
+      "grad_norm": 0.9815810438992207,
+      "learning_rate": 7.690719632625447e-06,
+      "loss": 0.1127,
+      "step": 961
+    },
+    {
+      "epoch": 3.3962930273609886,
+      "grad_norm": 0.9979507342808466,
+      "learning_rate": 7.685894737293773e-06,
+      "loss": 0.1224,
+      "step": 962
+    },
+    {
+      "epoch": 3.3998234774933804,
+      "grad_norm": 1.0098708954426872,
+      "learning_rate": 7.681066323896053e-06,
+      "loss": 0.1418,
+      "step": 963
+    },
+    {
+      "epoch": 3.4033539276257723,
+      "grad_norm": 0.8683560303898923,
+      "learning_rate": 7.676234398756688e-06,
+      "loss": 0.11,
+      "step": 964
+    },
+    {
+      "epoch": 3.406884377758164,
+      "grad_norm": 0.9354248879782321,
+      "learning_rate": 7.671398968204675e-06,
+      "loss": 0.1125,
+      "step": 965
+    },
+    {
+      "epoch": 3.410414827890556,
+      "grad_norm": 0.9688499014186751,
+      "learning_rate": 7.66656003857361e-06,
+      "loss": 0.1233,
+      "step": 966
+    },
+    {
+      "epoch": 3.413945278022948,
+      "grad_norm": 1.0624016948474533,
+      "learning_rate": 7.66171761620167e-06,
+      "loss": 0.1051,
+      "step": 967
+    },
+    {
+      "epoch": 3.4174757281553396,
+      "grad_norm": 0.9463189432354194,
+      "learning_rate": 7.656871707431604e-06,
+      "loss": 0.1171,
+      "step": 968
+    },
+    {
+      "epoch": 3.4210061782877315,
+      "grad_norm": 0.9894483338890546,
+      "learning_rate": 7.65202231861073e-06,
+      "loss": 0.1108,
+      "step": 969
+    },
+    {
+      "epoch": 3.4245366284201237,
+      "grad_norm": 0.9687362431205614,
+      "learning_rate": 7.647169456090925e-06,
+      "loss": 0.1116,
+      "step": 970
+    },
+    {
+      "epoch": 3.4280670785525156,
+      "grad_norm": 0.9557897471038936,
+      "learning_rate": 7.642313126228618e-06,
+      "loss": 0.1153,
+      "step": 971
+    },
+    {
+      "epoch": 3.4315975286849074,
+      "grad_norm": 1.0869572132117586,
+      "learning_rate": 7.63745333538477e-06,
+      "loss": 0.1236,
+      "step": 972
+    },
+    {
+      "epoch": 3.4351279788172993,
+      "grad_norm": 0.9605589326914042,
+      "learning_rate": 7.632590089924885e-06,
+      "loss": 0.1065,
+      "step": 973
+    },
+    {
+      "epoch": 3.438658428949691,
+      "grad_norm": 1.0323514330728734,
+      "learning_rate": 7.627723396218988e-06,
+      "loss": 0.1175,
+      "step": 974
+    },
+    {
+      "epoch": 3.442188879082083,
+      "grad_norm": 1.0118187975842183,
+      "learning_rate": 7.622853260641623e-06,
+      "loss": 0.1186,
+      "step": 975
+    },
+    {
+      "epoch": 3.4457193292144748,
+      "grad_norm": 1.024656799596295,
+      "learning_rate": 7.61797968957184e-06,
+      "loss": 0.1271,
+      "step": 976
+    },
+    {
+      "epoch": 3.4492497793468666,
+      "grad_norm": 0.8727047804672294,
+      "learning_rate": 7.613102689393188e-06,
+      "loss": 0.1144,
+      "step": 977
+    },
+    {
+      "epoch": 3.4527802294792584,
+      "grad_norm": 0.8394841490325925,
+      "learning_rate": 7.608222266493712e-06,
+      "loss": 0.1123,
+      "step": 978
+    },
+    {
+      "epoch": 3.4563106796116507,
+      "grad_norm": 1.1351511857664562,
+      "learning_rate": 7.603338427265935e-06,
+      "loss": 0.1012,
+      "step": 979
+    },
+    {
+      "epoch": 3.459841129744042,
+      "grad_norm": 1.128335444374776,
+      "learning_rate": 7.598451178106857e-06,
+      "loss": 0.1206,
+      "step": 980
+    },
+    {
+      "epoch": 3.4633715798764344,
+      "grad_norm": 0.8576208704238955,
+      "learning_rate": 7.593560525417947e-06,
+      "loss": 0.103,
+      "step": 981
+    },
+    {
+      "epoch": 3.4669020300088262,
+      "grad_norm": 1.1074290695297295,
+      "learning_rate": 7.588666475605127e-06,
+      "loss": 0.1254,
+      "step": 982
+    },
+    {
+      "epoch": 3.470432480141218,
+      "grad_norm": 1.1040068892498724,
+      "learning_rate": 7.583769035078772e-06,
+      "loss": 0.1143,
+      "step": 983
+    },
+    {
+      "epoch": 3.47396293027361,
+      "grad_norm": 1.0664311386837333,
+      "learning_rate": 7.578868210253698e-06,
+      "loss": 0.1353,
+      "step": 984
+    },
+    {
+      "epoch": 3.4774933804060018,
+      "grad_norm": 1.059615015944644,
+      "learning_rate": 7.5739640075491546e-06,
+      "loss": 0.1386,
+      "step": 985
+    },
+    {
+      "epoch": 3.4810238305383936,
+      "grad_norm": 0.9899439186497315,
+      "learning_rate": 7.569056433388813e-06,
+      "loss": 0.1258,
+      "step": 986
+    },
+    {
+      "epoch": 3.4845542806707854,
+      "grad_norm": 0.9224864093833172,
+      "learning_rate": 7.5641454942007606e-06,
+      "loss": 0.1225,
+      "step": 987
+    },
+    {
+      "epoch": 3.4880847308031773,
+      "grad_norm": 0.887546649860455,
+      "learning_rate": 7.559231196417495e-06,
+      "loss": 0.1249,
+      "step": 988
+    },
+    {
+      "epoch": 3.491615180935569,
+      "grad_norm": 0.9660667346820182,
+      "learning_rate": 7.554313546475915e-06,
+      "loss": 0.1163,
+      "step": 989
+    },
+    {
+      "epoch": 3.4951456310679614,
+      "grad_norm": 0.9707962317309367,
+      "learning_rate": 7.5493925508173025e-06,
+      "loss": 0.1321,
+      "step": 990
+    },
+    {
+      "epoch": 3.4986760812003532,
+      "grad_norm": 1.1180387890566459,
+      "learning_rate": 7.5444682158873305e-06,
+      "loss": 0.1221,
+      "step": 991
+    },
+    {
+      "epoch": 3.502206531332745,
+      "grad_norm": 0.8799016210637428,
+      "learning_rate": 7.539540548136037e-06,
+      "loss": 0.1371,
+      "step": 992
+    },
+    {
+      "epoch": 3.505736981465137,
+      "grad_norm": 1.0482875409664072,
+      "learning_rate": 7.534609554017833e-06,
+      "loss": 0.1175,
+      "step": 993
+    },
+    {
+      "epoch": 3.5092674315975287,
+      "grad_norm": 0.8991365777316297,
+      "learning_rate": 7.529675239991483e-06,
+      "loss": 0.1293,
+      "step": 994
+    },
+    {
+      "epoch": 3.5127978817299206,
+      "grad_norm": 1.0386653386530387,
+      "learning_rate": 7.524737612520101e-06,
+      "loss": 0.1229,
+      "step": 995
+    },
+    {
+      "epoch": 3.5163283318623124,
+      "grad_norm": 1.0711541879063016,
+      "learning_rate": 7.5197966780711385e-06,
+      "loss": 0.1201,
+      "step": 996
+    },
+    {
+      "epoch": 3.5198587819947043,
+      "grad_norm": 1.0318173451390964,
+      "learning_rate": 7.5148524431163825e-06,
+      "loss": 0.1316,
+      "step": 997
+    },
+    {
+      "epoch": 3.523389232127096,
+      "grad_norm": 0.984403787758931,
+      "learning_rate": 7.509904914131941e-06,
+      "loss": 0.1197,
+      "step": 998
+    },
+    {
+      "epoch": 3.526919682259488,
+      "grad_norm": 0.8522121035569924,
+      "learning_rate": 7.504954097598238e-06,
+      "loss": 0.1061,
+      "step": 999
+    },
+    {
+      "epoch": 3.5304501323918798,
+      "grad_norm": 0.849756117125948,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.1067,
+      "step": 1000
+    },
+    {
+      "epoch": 3.533980582524272,
+      "grad_norm": 0.7751208269103894,
+      "learning_rate": 7.495042627826257e-06,
+      "loss": 0.1091,
+      "step": 1001
+    },
+    {
+      "epoch": 3.537511032656664,
+      "grad_norm": 0.9495031542830646,
+      "learning_rate": 7.490081987570321e-06,
+      "loss": 0.121,
+      "step": 1002
+    },
+    {
+      "epoch": 3.5410414827890557,
+      "grad_norm": 1.0312096390159498,
+      "learning_rate": 7.48511808572979e-06,
+      "loss": 0.1304,
+      "step": 1003
+    },
+    {
+      "epoch": 3.5445719329214476,
+      "grad_norm": 1.054899906830368,
+      "learning_rate": 7.480150928806534e-06,
+      "loss": 0.1361,
+      "step": 1004
+    },
+    {
+      "epoch": 3.5481023830538394,
+      "grad_norm": 0.979527615798883,
+      "learning_rate": 7.475180523306684e-06,
+      "loss": 0.1277,
+      "step": 1005
+    },
+    {
+      "epoch": 3.5516328331862312,
+      "grad_norm": 1.0015458182337016,
+      "learning_rate": 7.470206875740627e-06,
+      "loss": 0.1149,
+      "step": 1006
+    },
+    {
+      "epoch": 3.555163283318623,
+      "grad_norm": 1.0755807694863864,
+      "learning_rate": 7.465229992622996e-06,
+      "loss": 0.1178,
+      "step": 1007
+    },
+    {
+      "epoch": 3.558693733451015,
+      "grad_norm": 1.0077743388921196,
+      "learning_rate": 7.4602498804726646e-06,
+      "loss": 0.1267,
+      "step": 1008
+    },
+    {
+      "epoch": 3.5622241835834068,
+      "grad_norm": 0.9760852883342952,
+      "learning_rate": 7.455266545812733e-06,
+      "loss": 0.1144,
+      "step": 1009
+    },
+    {
+      "epoch": 3.565754633715799,
+      "grad_norm": 0.9599179886281429,
+      "learning_rate": 7.450279995170524e-06,
+      "loss": 0.1233,
+      "step": 1010
+    },
+    {
+      "epoch": 3.5692850838481904,
+      "grad_norm": 1.0506465130972065,
+      "learning_rate": 7.445290235077571e-06,
+      "loss": 0.1223,
+      "step": 1011
+    },
+    {
+      "epoch": 3.5728155339805827,
+      "grad_norm": 0.9904879926238966,
+      "learning_rate": 7.440297272069615e-06,
+      "loss": 0.1103,
+      "step": 1012
+    },
+    {
+      "epoch": 3.5763459841129746,
+      "grad_norm": 0.9373307300830205,
+      "learning_rate": 7.435301112686588e-06,
+      "loss": 0.1089,
+      "step": 1013
+    },
+    {
+      "epoch": 3.5798764342453664,
+      "grad_norm": 0.9642807996392302,
+      "learning_rate": 7.430301763472612e-06,
+      "loss": 0.1397,
+      "step": 1014
+    },
+    {
+      "epoch": 3.5834068843777582,
+      "grad_norm": 1.0798478829987526,
+      "learning_rate": 7.425299230975982e-06,
+      "loss": 0.1271,
+      "step": 1015
+    },
+    {
+      "epoch": 3.58693733451015,
+      "grad_norm": 0.9757898668217716,
+      "learning_rate": 7.420293521749169e-06,
+      "loss": 0.1254,
+      "step": 1016
+    },
+    {
+      "epoch": 3.590467784642542,
+      "grad_norm": 0.7957246885082836,
+      "learning_rate": 7.415284642348801e-06,
+      "loss": 0.1024,
+      "step": 1017
+    },
+    {
+      "epoch": 3.5939982347749337,
+      "grad_norm": 1.0499954601099537,
+      "learning_rate": 7.410272599335662e-06,
+      "loss": 0.1104,
+      "step": 1018
+    },
+    {
+      "epoch": 3.5975286849073256,
+      "grad_norm": 1.017468888424547,
+      "learning_rate": 7.4052573992746735e-06,
+      "loss": 0.1329,
+      "step": 1019
+    },
+    {
+      "epoch": 3.6010591350397174,
+      "grad_norm": 0.9050306794358908,
+      "learning_rate": 7.400239048734899e-06,
+      "loss": 0.1001,
+      "step": 1020
+    },
+    {
+      "epoch": 3.6045895851721097,
+      "grad_norm": 0.9892172267357415,
+      "learning_rate": 7.395217554289524e-06,
+      "loss": 0.1116,
+      "step": 1021
+    },
+    {
+      "epoch": 3.608120035304501,
+      "grad_norm": 0.825558862940724,
+      "learning_rate": 7.390192922515855e-06,
+      "loss": 0.1178,
+      "step": 1022
+    },
+    {
+      "epoch": 3.6116504854368934,
+      "grad_norm": 1.0870666189939848,
+      "learning_rate": 7.385165159995306e-06,
+      "loss": 0.1188,
+      "step": 1023
+    },
+    {
+      "epoch": 3.615180935569285,
+      "grad_norm": 0.9461150823847166,
+      "learning_rate": 7.38013427331339e-06,
+      "loss": 0.1387,
+      "step": 1024
+    },
+    {
+      "epoch": 3.618711385701677,
+      "grad_norm": 0.8931705317958296,
+      "learning_rate": 7.3751002690597175e-06,
+      "loss": 0.1145,
+      "step": 1025
+    },
+    {
+      "epoch": 3.622241835834069,
+      "grad_norm": 1.0044315340520118,
+      "learning_rate": 7.370063153827978e-06,
+      "loss": 0.1311,
+      "step": 1026
+    },
+    {
+      "epoch": 3.6257722859664607,
+      "grad_norm": 0.9366772554201618,
+      "learning_rate": 7.365022934215935e-06,
+      "loss": 0.1347,
+      "step": 1027
+    },
+    {
+      "epoch": 3.6293027360988526,
+      "grad_norm": 1.0287287282798234,
+      "learning_rate": 7.3599796168254256e-06,
+      "loss": 0.12,
+      "step": 1028
+    },
+    {
+      "epoch": 3.6328331862312444,
+      "grad_norm": 1.2655633592932094,
+      "learning_rate": 7.3549332082623336e-06,
+      "loss": 0.1225,
+      "step": 1029
+    },
+    {
+      "epoch": 3.6363636363636362,
+      "grad_norm": 0.962932765361898,
+      "learning_rate": 7.349883715136601e-06,
+      "loss": 0.1163,
+      "step": 1030
+    },
+    {
+      "epoch": 3.639894086496028,
+      "grad_norm": 1.1898739705783477,
+      "learning_rate": 7.344831144062204e-06,
+      "loss": 0.1342,
+      "step": 1031
+    },
+    {
+      "epoch": 3.6434245366284204,
+      "grad_norm": 0.9775960634954581,
+      "learning_rate": 7.339775501657156e-06,
+      "loss": 0.1159,
+      "step": 1032
+    },
+    {
+      "epoch": 3.6469549867608118,
+      "grad_norm": 0.9985532411205827,
+      "learning_rate": 7.334716794543487e-06,
+      "loss": 0.1215,
+      "step": 1033
+    },
+    {
+      "epoch": 3.650485436893204,
+      "grad_norm": 0.8371099521051029,
+      "learning_rate": 7.329655029347246e-06,
+      "loss": 0.133,
+      "step": 1034
+    },
+    {
+      "epoch": 3.654015887025596,
+      "grad_norm": 0.8282699034669408,
+      "learning_rate": 7.324590212698486e-06,
+      "loss": 0.1128,
+      "step": 1035
+    },
+    {
+      "epoch": 3.6575463371579877,
+      "grad_norm": 0.8892127258122363,
+      "learning_rate": 7.319522351231256e-06,
+      "loss": 0.1193,
+      "step": 1036
+    },
+    {
+      "epoch": 3.6610767872903796,
+      "grad_norm": 1.0611463535886227,
+      "learning_rate": 7.314451451583593e-06,
+      "loss": 0.1371,
+      "step": 1037
+    },
+    {
+      "epoch": 3.6646072374227714,
+      "grad_norm": 0.834905111221197,
+      "learning_rate": 7.309377520397517e-06,
+      "loss": 0.1198,
+      "step": 1038
+    },
+    {
+      "epoch": 3.6681376875551632,
+      "grad_norm": 0.8795916947723678,
+      "learning_rate": 7.304300564319013e-06,
+      "loss": 0.1154,
+      "step": 1039
+    },
+    {
+      "epoch": 3.671668137687555,
+      "grad_norm": 0.932134034033176,
+      "learning_rate": 7.299220589998037e-06,
+      "loss": 0.1172,
+      "step": 1040
+    },
+    {
+      "epoch": 3.675198587819947,
+      "grad_norm": 1.060320989726038,
+      "learning_rate": 7.294137604088487e-06,
+      "loss": 0.1272,
+      "step": 1041
+    },
+    {
+      "epoch": 3.6787290379523387,
+      "grad_norm": 1.0865767224751157,
+      "learning_rate": 7.2890516132482145e-06,
+      "loss": 0.1174,
+      "step": 1042
+    },
+    {
+      "epoch": 3.682259488084731,
+      "grad_norm": 0.9524420011955063,
+      "learning_rate": 7.283962624139004e-06,
+      "loss": 0.1025,
+      "step": 1043
+    },
+    {
+      "epoch": 3.6857899382171224,
+      "grad_norm": 1.0723266113080743,
+      "learning_rate": 7.278870643426565e-06,
+      "loss": 0.1188,
+      "step": 1044
+    },
+    {
+      "epoch": 3.6893203883495147,
+      "grad_norm": 0.9889202356506911,
+      "learning_rate": 7.27377567778053e-06,
+      "loss": 0.1171,
+      "step": 1045
+    },
+    {
+      "epoch": 3.6928508384819065,
+      "grad_norm": 1.1387504413356822,
+      "learning_rate": 7.268677733874438e-06,
+      "loss": 0.1347,
+      "step": 1046
+    },
+    {
+      "epoch": 3.6963812886142984,
+      "grad_norm": 0.9054547213610565,
+      "learning_rate": 7.263576818385728e-06,
+      "loss": 0.1192,
+      "step": 1047
+    },
+    {
+      "epoch": 3.69991173874669,
+      "grad_norm": 0.9389240088652642,
+      "learning_rate": 7.258472937995736e-06,
+      "loss": 0.105,
+      "step": 1048
+    },
+    {
+      "epoch": 3.703442188879082,
+      "grad_norm": 1.0097808039069447,
+      "learning_rate": 7.253366099389678e-06,
+      "loss": 0.1283,
+      "step": 1049
+    },
+    {
+      "epoch": 3.706972639011474,
+      "grad_norm": 0.9312349763358698,
+      "learning_rate": 7.2482563092566446e-06,
+      "loss": 0.1177,
+      "step": 1050
+    },
+    {
+      "epoch": 3.7105030891438657,
+      "grad_norm": 0.8462385685589844,
+      "learning_rate": 7.243143574289595e-06,
+      "loss": 0.1176,
+      "step": 1051
+    },
+    {
+      "epoch": 3.7140335392762576,
+      "grad_norm": 0.9999611373105524,
+      "learning_rate": 7.23802790118534e-06,
+      "loss": 0.12,
+      "step": 1052
+    },
+    {
+      "epoch": 3.7175639894086494,
+      "grad_norm": 1.029486778996518,
+      "learning_rate": 7.232909296644548e-06,
+      "loss": 0.1313,
+      "step": 1053
+    },
+    {
+      "epoch": 3.7210944395410417,
+      "grad_norm": 0.9598825118610063,
+      "learning_rate": 7.227787767371717e-06,
+      "loss": 0.1057,
+      "step": 1054
+    },
+    {
+      "epoch": 3.7246248896734335,
+      "grad_norm": 0.8914389737346524,
+      "learning_rate": 7.222663320075185e-06,
+      "loss": 0.1196,
+      "step": 1055
+    },
+    {
+      "epoch": 3.7281553398058254,
+      "grad_norm": 0.9459232821476742,
+      "learning_rate": 7.217535961467105e-06,
+      "loss": 0.1254,
+      "step": 1056
+    },
+    {
+      "epoch": 3.731685789938217,
+      "grad_norm": 1.121524174912343,
+      "learning_rate": 7.212405698263446e-06,
+      "loss": 0.1296,
+      "step": 1057
+    },
+    {
+      "epoch": 3.735216240070609,
+      "grad_norm": 1.0305189372866497,
+      "learning_rate": 7.2072725371839826e-06,
+      "loss": 0.1123,
+      "step": 1058
+    },
+    {
+      "epoch": 3.738746690203001,
+      "grad_norm": 1.0095800993562205,
+      "learning_rate": 7.202136484952286e-06,
+      "loss": 0.1109,
+      "step": 1059
+    },
+    {
+      "epoch": 3.7422771403353927,
+      "grad_norm": 1.091036374867166,
+      "learning_rate": 7.1969975482957075e-06,
+      "loss": 0.125,
+      "step": 1060
+    },
+    {
+      "epoch": 3.7458075904677846,
+      "grad_norm": 1.1262861615878523,
+      "learning_rate": 7.191855733945388e-06,
+      "loss": 0.1304,
+      "step": 1061
+    },
+    {
+      "epoch": 3.7493380406001764,
+      "grad_norm": 1.0644115251365422,
+      "learning_rate": 7.1867110486362255e-06,
+      "loss": 0.1221,
+      "step": 1062
+    },
+    {
+      "epoch": 3.7528684907325687,
+      "grad_norm": 1.0506019899313102,
+      "learning_rate": 7.1815634991068895e-06,
+      "loss": 0.1205,
+      "step": 1063
+    },
+    {
+      "epoch": 3.75639894086496,
+      "grad_norm": 1.0044867932848505,
+      "learning_rate": 7.176413092099791e-06,
+      "loss": 0.1234,
+      "step": 1064
+    },
+    {
+      "epoch": 3.7599293909973523,
+      "grad_norm": 0.9095279203965683,
+      "learning_rate": 7.171259834361092e-06,
+      "loss": 0.1205,
+      "step": 1065
+    },
+    {
+      "epoch": 3.763459841129744,
+      "grad_norm": 1.1445775552595612,
+      "learning_rate": 7.1661037326406825e-06,
+      "loss": 0.1275,
+      "step": 1066
+    },
+    {
+      "epoch": 3.766990291262136,
+      "grad_norm": 0.946883327804174,
+      "learning_rate": 7.1609447936921814e-06,
+      "loss": 0.1398,
+      "step": 1067
+    },
+    {
+      "epoch": 3.770520741394528,
+      "grad_norm": 0.9041398561076188,
+      "learning_rate": 7.155783024272925e-06,
+      "loss": 0.13,
+      "step": 1068
+    },
+    {
+      "epoch": 3.7740511915269197,
+      "grad_norm": 0.8457469332005921,
+      "learning_rate": 7.150618431143951e-06,
+      "loss": 0.1143,
+      "step": 1069
+    },
+    {
+      "epoch": 3.7775816416593115,
+      "grad_norm": 0.8578352898530813,
+      "learning_rate": 7.145451021070003e-06,
+      "loss": 0.1247,
+      "step": 1070
+    },
+    {
+      "epoch": 3.7811120917917034,
+      "grad_norm": 0.9352514012935873,
+      "learning_rate": 7.140280800819507e-06,
+      "loss": 0.1344,
+      "step": 1071
+    },
+    {
+      "epoch": 3.784642541924095,
+      "grad_norm": 0.9799389793882611,
+      "learning_rate": 7.135107777164576e-06,
+      "loss": 0.1223,
+      "step": 1072
+    },
+    {
+      "epoch": 3.788172992056487,
+      "grad_norm": 1.065041516403633,
+      "learning_rate": 7.129931956880992e-06,
+      "loss": 0.1375,
+      "step": 1073
+    },
+    {
+      "epoch": 3.7917034421888793,
+      "grad_norm": 1.0449759566953258,
+      "learning_rate": 7.124753346748201e-06,
+      "loss": 0.1171,
+      "step": 1074
+    },
+    {
+      "epoch": 3.7952338923212707,
+      "grad_norm": 1.1549382686805798,
+      "learning_rate": 7.119571953549305e-06,
+      "loss": 0.1281,
+      "step": 1075
+    },
+    {
+      "epoch": 3.798764342453663,
+      "grad_norm": 1.007228257856292,
+      "learning_rate": 7.114387784071044e-06,
+      "loss": 0.1269,
+      "step": 1076
+    },
+    {
+      "epoch": 3.802294792586055,
+      "grad_norm": 1.0225753082093523,
+      "learning_rate": 7.109200845103806e-06,
+      "loss": 0.1199,
+      "step": 1077
+    },
+    {
+      "epoch": 3.8058252427184467,
+      "grad_norm": 0.9355931698169853,
+      "learning_rate": 7.104011143441596e-06,
+      "loss": 0.1169,
+      "step": 1078
+    },
+    {
+      "epoch": 3.8093556928508385,
+      "grad_norm": 0.963698358188344,
+      "learning_rate": 7.098818685882044e-06,
+      "loss": 0.1179,
+      "step": 1079
+    },
+    {
+      "epoch": 3.8128861429832304,
+      "grad_norm": 0.7777902932062336,
+      "learning_rate": 7.093623479226386e-06,
+      "loss": 0.1186,
+      "step": 1080
+    },
+    {
+      "epoch": 3.816416593115622,
+      "grad_norm": 0.9518180229607743,
+      "learning_rate": 7.088425530279461e-06,
+      "loss": 0.1062,
+      "step": 1081
+    },
+    {
+      "epoch": 3.819947043248014,
+      "grad_norm": 1.034322676608895,
+      "learning_rate": 7.083224845849701e-06,
+      "loss": 0.1144,
+      "step": 1082
+    },
+    {
+      "epoch": 3.823477493380406,
+      "grad_norm": 1.0557347999282005,
+      "learning_rate": 7.078021432749121e-06,
+      "loss": 0.1243,
+      "step": 1083
+    },
+    {
+      "epoch": 3.8270079435127977,
+      "grad_norm": 0.8967864020526236,
+      "learning_rate": 7.072815297793303e-06,
+      "loss": 0.1189,
+      "step": 1084
+    },
+    {
+      "epoch": 3.83053839364519,
+      "grad_norm": 0.9394124246894394,
+      "learning_rate": 7.067606447801405e-06,
+      "loss": 0.1116,
+      "step": 1085
+    },
+    {
+      "epoch": 3.8340688437775814,
+      "grad_norm": 0.9390953780920334,
+      "learning_rate": 7.062394889596133e-06,
+      "loss": 0.1258,
+      "step": 1086
+    },
+    {
+      "epoch": 3.8375992939099737,
+      "grad_norm": 0.9777220254987733,
+      "learning_rate": 7.057180630003746e-06,
+      "loss": 0.1241,
+      "step": 1087
+    },
+    {
+      "epoch": 3.8411297440423655,
+      "grad_norm": 0.8779726966694773,
+      "learning_rate": 7.051963675854035e-06,
+      "loss": 0.1267,
+      "step": 1088
+    },
+    {
+      "epoch": 3.8446601941747574,
+      "grad_norm": 1.0359165893132334,
+      "learning_rate": 7.046744033980329e-06,
+      "loss": 0.1168,
+      "step": 1089
+    },
+    {
+      "epoch": 3.848190644307149,
+      "grad_norm": 0.9976911554856348,
+      "learning_rate": 7.041521711219468e-06,
+      "loss": 0.1317,
+      "step": 1090
+    },
+    {
+      "epoch": 3.851721094439541,
+      "grad_norm": 1.0169743441045915,
+      "learning_rate": 7.03629671441181e-06,
+      "loss": 0.1398,
+      "step": 1091
+    },
+    {
+      "epoch": 3.855251544571933,
+      "grad_norm": 0.9568634693122333,
+      "learning_rate": 7.031069050401215e-06,
+      "loss": 0.0919,
+      "step": 1092
+    },
+    {
+      "epoch": 3.8587819947043247,
+      "grad_norm": 0.9083188838175256,
+      "learning_rate": 7.025838726035032e-06,
+      "loss": 0.1098,
+      "step": 1093
+    },
+    {
+      "epoch": 3.8623124448367165,
+      "grad_norm": 0.9312177288214185,
+      "learning_rate": 7.020605748164101e-06,
+      "loss": 0.1275,
+      "step": 1094
+    },
+    {
+      "epoch": 3.8658428949691084,
+      "grad_norm": 0.8178614978701555,
+      "learning_rate": 7.015370123642733e-06,
+      "loss": 0.11,
+      "step": 1095
+    },
+    {
+      "epoch": 3.8693733451015007,
+      "grad_norm": 0.9943777665520181,
+      "learning_rate": 7.010131859328706e-06,
+      "loss": 0.1141,
+      "step": 1096
+    },
+    {
+      "epoch": 3.872903795233892,
+      "grad_norm": 0.7533041541573118,
+      "learning_rate": 7.00489096208326e-06,
+      "loss": 0.0988,
+      "step": 1097
+    },
+    {
+      "epoch": 3.8764342453662843,
+      "grad_norm": 0.9539898985521694,
+      "learning_rate": 6.999647438771076e-06,
+      "loss": 0.1158,
+      "step": 1098
+    },
+    {
+      "epoch": 3.879964695498676,
+      "grad_norm": 0.9694635032226555,
+      "learning_rate": 6.994401296260285e-06,
+      "loss": 0.1476,
+      "step": 1099
+    },
+    {
+      "epoch": 3.883495145631068,
+      "grad_norm": 0.9466515792078597,
+      "learning_rate": 6.989152541422438e-06,
+      "loss": 0.1207,
+      "step": 1100
+    },
+    {
+      "epoch": 3.88702559576346,
+      "grad_norm": 0.9232735267824136,
+      "learning_rate": 6.983901181132514e-06,
+      "loss": 0.1219,
+      "step": 1101
+    },
+    {
+      "epoch": 3.8905560458958517,
+      "grad_norm": 0.9035726923781363,
+      "learning_rate": 6.978647222268904e-06,
+      "loss": 0.1139,
+      "step": 1102
+    },
+    {
+      "epoch": 3.8940864960282435,
+      "grad_norm": 0.9447988957374035,
+      "learning_rate": 6.973390671713401e-06,
+      "loss": 0.1106,
+      "step": 1103
+    },
+    {
+      "epoch": 3.8976169461606354,
+      "grad_norm": 0.9993165292258056,
+      "learning_rate": 6.9681315363511946e-06,
+      "loss": 0.1126,
+      "step": 1104
+    },
+    {
+      "epoch": 3.901147396293027,
+      "grad_norm": 0.9567127281227304,
+      "learning_rate": 6.9628698230708604e-06,
+      "loss": 0.126,
+      "step": 1105
+    },
+    {
+      "epoch": 3.904677846425419,
+      "grad_norm": 0.9704513493237618,
+      "learning_rate": 6.957605538764347e-06,
+      "loss": 0.1135,
+      "step": 1106
+    },
+    {
+      "epoch": 3.9082082965578113,
+      "grad_norm": 0.9049083128378328,
+      "learning_rate": 6.9523386903269715e-06,
+      "loss": 0.118,
+      "step": 1107
+    },
+    {
+      "epoch": 3.911738746690203,
+      "grad_norm": 1.1410548052057965,
+      "learning_rate": 6.947069284657416e-06,
+      "loss": 0.1232,
+      "step": 1108
+    },
+    {
+      "epoch": 3.915269196822595,
+      "grad_norm": 0.9628185789378251,
+      "learning_rate": 6.941797328657702e-06,
+      "loss": 0.1012,
+      "step": 1109
+    },
+    {
+      "epoch": 3.918799646954987,
+      "grad_norm": 0.986500497408887,
+      "learning_rate": 6.936522829233202e-06,
+      "loss": 0.1267,
+      "step": 1110
+    },
+    {
+      "epoch": 3.9223300970873787,
+      "grad_norm": 0.9404574261439161,
+      "learning_rate": 6.93124579329261e-06,
+      "loss": 0.1177,
+      "step": 1111
+    },
+    {
+      "epoch": 3.9258605472197705,
+      "grad_norm": 0.9652812442302577,
+      "learning_rate": 6.92596622774795e-06,
+      "loss": 0.1099,
+      "step": 1112
+    },
+    {
+      "epoch": 3.9293909973521624,
+      "grad_norm": 0.9255614343816538,
+      "learning_rate": 6.920684139514555e-06,
+      "loss": 0.1157,
+      "step": 1113
+    },
+    {
+      "epoch": 3.932921447484554,
+      "grad_norm": 1.1133925624513497,
+      "learning_rate": 6.9153995355110645e-06,
+      "loss": 0.13,
+      "step": 1114
+    },
+    {
+      "epoch": 3.936451897616946,
+      "grad_norm": 1.0496747828703927,
+      "learning_rate": 6.9101124226594106e-06,
+      "loss": 0.1123,
+      "step": 1115
+    },
+    {
+      "epoch": 3.9399823477493383,
+      "grad_norm": 1.0246048820560874,
+      "learning_rate": 6.904822807884816e-06,
+      "loss": 0.1158,
+      "step": 1116
+    },
+    {
+      "epoch": 3.9435127978817297,
+      "grad_norm": 1.0333168676182993,
+      "learning_rate": 6.899530698115776e-06,
+      "loss": 0.1315,
+      "step": 1117
+    },
+    {
+      "epoch": 3.947043248014122,
+      "grad_norm": 1.0064120983759648,
+      "learning_rate": 6.894236100284058e-06,
+      "loss": 0.1282,
+      "step": 1118
+    },
+    {
+      "epoch": 3.950573698146514,
+      "grad_norm": 0.963684336758527,
+      "learning_rate": 6.888939021324681e-06,
+      "loss": 0.1217,
+      "step": 1119
+    },
+    {
+      "epoch": 3.9541041482789057,
+      "grad_norm": 0.9892019560831151,
+      "learning_rate": 6.883639468175926e-06,
+      "loss": 0.1239,
+      "step": 1120
+    },
+    {
+      "epoch": 3.9576345984112975,
+      "grad_norm": 1.0120690523201414,
+      "learning_rate": 6.878337447779304e-06,
+      "loss": 0.1335,
+      "step": 1121
+    },
+    {
+      "epoch": 3.9611650485436893,
+      "grad_norm": 0.8218662981254,
+      "learning_rate": 6.873032967079562e-06,
+      "loss": 0.1185,
+      "step": 1122
+    },
+    {
+      "epoch": 3.964695498676081,
+      "grad_norm": 0.9915669809027783,
+      "learning_rate": 6.867726033024667e-06,
+      "loss": 0.1181,
+      "step": 1123
+    },
+    {
+      "epoch": 3.968225948808473,
+      "grad_norm": 0.7674002943828144,
+      "learning_rate": 6.862416652565804e-06,
+      "loss": 0.11,
+      "step": 1124
+    },
+    {
+      "epoch": 3.971756398940865,
+      "grad_norm": 0.9775474573427773,
+      "learning_rate": 6.85710483265736e-06,
+      "loss": 0.1357,
+      "step": 1125
+    },
+    {
+      "epoch": 3.9752868490732567,
+      "grad_norm": 0.9046890209780649,
+      "learning_rate": 6.8517905802569164e-06,
+      "loss": 0.1145,
+      "step": 1126
+    },
+    {
+      "epoch": 3.978817299205649,
+      "grad_norm": 0.921557821840316,
+      "learning_rate": 6.846473902325243e-06,
+      "loss": 0.1392,
+      "step": 1127
+    },
+    {
+      "epoch": 3.9823477493380404,
+      "grad_norm": 0.9794967931322031,
+      "learning_rate": 6.8411548058262824e-06,
+      "loss": 0.1088,
+      "step": 1128
+    },
+    {
+      "epoch": 3.9858781994704326,
+      "grad_norm": 0.8740159825381886,
+      "learning_rate": 6.835833297727148e-06,
+      "loss": 0.1139,
+      "step": 1129
+    },
+    {
+      "epoch": 3.9894086496028245,
+      "grad_norm": 0.8628586385174893,
+      "learning_rate": 6.830509384998114e-06,
+      "loss": 0.1082,
+      "step": 1130
+    },
+    {
+      "epoch": 3.9929390997352163,
+      "grad_norm": 0.9488799231960956,
+      "learning_rate": 6.8251830746126e-06,
+      "loss": 0.1316,
+      "step": 1131
+    },
+    {
+      "epoch": 3.996469549867608,
+      "grad_norm": 1.0703456877526685,
+      "learning_rate": 6.81985437354717e-06,
+      "loss": 0.1279,
+      "step": 1132
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.967853650509713,
+      "learning_rate": 6.814523288781516e-06,
+      "loss": 0.1305,
+      "step": 1133
+    },
+    {
+      "epoch": 4.003530450132392,
+      "grad_norm": 0.7981320722327793,
+      "learning_rate": 6.809189827298457e-06,
+      "loss": 0.0885,
+      "step": 1134
+    },
+    {
+      "epoch": 4.007060900264784,
+      "grad_norm": 0.5854902750468941,
+      "learning_rate": 6.803853996083918e-06,
+      "loss": 0.0793,
+      "step": 1135
+    },
+    {
+      "epoch": 4.010591350397176,
+      "grad_norm": 0.7049273298585778,
+      "learning_rate": 6.7985158021269335e-06,
+      "loss": 0.0758,
+      "step": 1136
+    },
+    {
+      "epoch": 4.014121800529567,
+      "grad_norm": 0.6869340400131053,
+      "learning_rate": 6.79317525241963e-06,
+      "loss": 0.0933,
+      "step": 1137
+    },
+    {
+      "epoch": 4.01765225066196,
+      "grad_norm": 0.5900057189092714,
+      "learning_rate": 6.787832353957225e-06,
+      "loss": 0.0706,
+      "step": 1138
+    },
+    {
+      "epoch": 4.021182700794351,
+      "grad_norm": 0.7601438078339049,
+      "learning_rate": 6.782487113738002e-06,
+      "loss": 0.0769,
+      "step": 1139
+    },
+    {
+      "epoch": 4.024713150926743,
+      "grad_norm": 0.851783451355363,
+      "learning_rate": 6.777139538763324e-06,
+      "loss": 0.0711,
+      "step": 1140
+    },
+    {
+      "epoch": 4.028243601059135,
+      "grad_norm": 0.82880112354727,
+      "learning_rate": 6.7717896360376025e-06,
+      "loss": 0.0717,
+      "step": 1141
+    },
+    {
+      "epoch": 4.031774051191527,
+      "grad_norm": 1.1143799167216393,
+      "learning_rate": 6.766437412568302e-06,
+      "loss": 0.0784,
+      "step": 1142
+    },
+    {
+      "epoch": 4.035304501323918,
+      "grad_norm": 0.8210313928830595,
+      "learning_rate": 6.761082875365929e-06,
+      "loss": 0.0626,
+      "step": 1143
+    },
+    {
+      "epoch": 4.038834951456311,
+      "grad_norm": 0.7779155627289819,
+      "learning_rate": 6.755726031444014e-06,
+      "loss": 0.0649,
+      "step": 1144
+    },
+    {
+      "epoch": 4.042365401588703,
+      "grad_norm": 1.1156761361527734,
+      "learning_rate": 6.750366887819116e-06,
+      "loss": 0.0787,
+      "step": 1145
+    },
+    {
+      "epoch": 4.045895851721094,
+      "grad_norm": 1.11782008947278,
+      "learning_rate": 6.745005451510804e-06,
+      "loss": 0.0742,
+      "step": 1146
+    },
+    {
+      "epoch": 4.049426301853487,
+      "grad_norm": 1.0344771965558572,
+      "learning_rate": 6.739641729541645e-06,
+      "loss": 0.0838,
+      "step": 1147
+    },
+    {
+      "epoch": 4.052956751985878,
+      "grad_norm": 2.0154736605675763,
+      "learning_rate": 6.734275728937208e-06,
+      "loss": 0.1034,
+      "step": 1148
+    },
+    {
+      "epoch": 4.05648720211827,
+      "grad_norm": 1.2294533756515769,
+      "learning_rate": 6.728907456726042e-06,
+      "loss": 0.0619,
+      "step": 1149
+    },
+    {
+      "epoch": 4.060017652250662,
+      "grad_norm": 1.4290104269014912,
+      "learning_rate": 6.723536919939669e-06,
+      "loss": 0.0686,
+      "step": 1150
+    },
+    {
+      "epoch": 4.063548102383054,
+      "grad_norm": 1.5020515871139606,
+      "learning_rate": 6.718164125612585e-06,
+      "loss": 0.0761,
+      "step": 1151
+    },
+    {
+      "epoch": 4.067078552515445,
+      "grad_norm": 1.4908120498463364,
+      "learning_rate": 6.712789080782235e-06,
+      "loss": 0.0828,
+      "step": 1152
+    },
+    {
+      "epoch": 4.070609002647838,
+      "grad_norm": 1.2267050202060312,
+      "learning_rate": 6.707411792489016e-06,
+      "loss": 0.0643,
+      "step": 1153
+    },
+    {
+      "epoch": 4.074139452780229,
+      "grad_norm": 1.3569581964127149,
+      "learning_rate": 6.7020322677762614e-06,
+      "loss": 0.082,
+      "step": 1154
+    },
+    {
+      "epoch": 4.077669902912621,
+      "grad_norm": 1.1848448933110132,
+      "learning_rate": 6.696650513690238e-06,
+      "loss": 0.0639,
+      "step": 1155
+    },
+    {
+      "epoch": 4.081200353045014,
+      "grad_norm": 1.181765635994709,
+      "learning_rate": 6.691266537280128e-06,
+      "loss": 0.057,
+      "step": 1156
+    },
+    {
+      "epoch": 4.084730803177405,
+      "grad_norm": 1.016848718981542,
+      "learning_rate": 6.685880345598027e-06,
+      "loss": 0.0537,
+      "step": 1157
+    },
+    {
+      "epoch": 4.088261253309797,
+      "grad_norm": 1.3224355302362136,
+      "learning_rate": 6.680491945698929e-06,
+      "loss": 0.064,
+      "step": 1158
+    },
+    {
+      "epoch": 4.091791703442189,
+      "grad_norm": 1.4746292032964976,
+      "learning_rate": 6.675101344640727e-06,
+      "loss": 0.0892,
+      "step": 1159
+    },
+    {
+      "epoch": 4.095322153574581,
+      "grad_norm": 1.5383513145198393,
+      "learning_rate": 6.66970854948419e-06,
+      "loss": 0.0766,
+      "step": 1160
+    },
+    {
+      "epoch": 4.098852603706972,
+      "grad_norm": 1.207790515633538,
+      "learning_rate": 6.664313567292966e-06,
+      "loss": 0.0693,
+      "step": 1161
+    },
+    {
+      "epoch": 4.102383053839365,
+      "grad_norm": 1.288949461541623,
+      "learning_rate": 6.658916405133562e-06,
+      "loss": 0.0826,
+      "step": 1162
+    },
+    {
+      "epoch": 4.105913503971756,
+      "grad_norm": 1.0936618806184264,
+      "learning_rate": 6.6535170700753506e-06,
+      "loss": 0.0789,
+      "step": 1163
+    },
+    {
+      "epoch": 4.109443954104148,
+      "grad_norm": 1.087774369178708,
+      "learning_rate": 6.6481155691905375e-06,
+      "loss": 0.0571,
+      "step": 1164
+    },
+    {
+      "epoch": 4.112974404236541,
+      "grad_norm": 1.3919202257673968,
+      "learning_rate": 6.6427119095541745e-06,
+      "loss": 0.0655,
+      "step": 1165
+    },
+    {
+      "epoch": 4.116504854368932,
+      "grad_norm": 1.0692129775554304,
+      "learning_rate": 6.637306098244138e-06,
+      "loss": 0.0758,
+      "step": 1166
+    },
+    {
+      "epoch": 4.120035304501324,
+      "grad_norm": 1.1487653554819723,
+      "learning_rate": 6.631898142341122e-06,
+      "loss": 0.0619,
+      "step": 1167
+    },
+    {
+      "epoch": 4.123565754633716,
+      "grad_norm": 1.036223404954323,
+      "learning_rate": 6.626488048928632e-06,
+      "loss": 0.0593,
+      "step": 1168
+    },
+    {
+      "epoch": 4.127096204766108,
+      "grad_norm": 1.0121114014928942,
+      "learning_rate": 6.621075825092972e-06,
+      "loss": 0.0667,
+      "step": 1169
+    },
+    {
+      "epoch": 4.130626654898499,
+      "grad_norm": 1.1972992188904066,
+      "learning_rate": 6.6156614779232345e-06,
+      "loss": 0.0733,
+      "step": 1170
+    },
+    {
+      "epoch": 4.134157105030892,
+      "grad_norm": 1.197590502375495,
+      "learning_rate": 6.610245014511296e-06,
+      "loss": 0.0611,
+      "step": 1171
+    },
+    {
+      "epoch": 4.137687555163283,
+      "grad_norm": 1.0881571550345615,
+      "learning_rate": 6.604826441951803e-06,
+      "loss": 0.061,
+      "step": 1172
+    },
+    {
+      "epoch": 4.141218005295675,
+      "grad_norm": 0.9686194494515865,
+      "learning_rate": 6.599405767342167e-06,
+      "loss": 0.0559,
+      "step": 1173
+    },
+    {
+      "epoch": 4.144748455428067,
+      "grad_norm": 1.3831994016631128,
+      "learning_rate": 6.593982997782549e-06,
+      "loss": 0.0642,
+      "step": 1174
+    },
+    {
+      "epoch": 4.148278905560459,
+      "grad_norm": 1.241798068520593,
+      "learning_rate": 6.588558140375858e-06,
+      "loss": 0.0707,
+      "step": 1175
+    },
+    {
+      "epoch": 4.151809355692851,
+      "grad_norm": 1.2592043421219463,
+      "learning_rate": 6.583131202227736e-06,
+      "loss": 0.0851,
+      "step": 1176
+    },
+    {
+      "epoch": 4.155339805825243,
+      "grad_norm": 1.1890876424522738,
+      "learning_rate": 6.577702190446552e-06,
+      "loss": 0.0764,
+      "step": 1177
+    },
+    {
+      "epoch": 4.158870255957635,
+      "grad_norm": 1.1892762885244956,
+      "learning_rate": 6.572271112143386e-06,
+      "loss": 0.0823,
+      "step": 1178
+    },
+    {
+      "epoch": 4.162400706090026,
+      "grad_norm": 1.098687094404278,
+      "learning_rate": 6.566837974432032e-06,
+      "loss": 0.0582,
+      "step": 1179
+    },
+    {
+      "epoch": 4.165931156222419,
+      "grad_norm": 1.1795811578347426,
+      "learning_rate": 6.561402784428974e-06,
+      "loss": 0.0787,
+      "step": 1180
+    },
+    {
+      "epoch": 4.16946160635481,
+      "grad_norm": 1.0850929765127277,
+      "learning_rate": 6.5559655492533915e-06,
+      "loss": 0.0576,
+      "step": 1181
+    },
+    {
+      "epoch": 4.172992056487202,
+      "grad_norm": 1.2039461760737675,
+      "learning_rate": 6.550526276027138e-06,
+      "loss": 0.075,
+      "step": 1182
+    },
+    {
+      "epoch": 4.176522506619594,
+      "grad_norm": 1.1924260467355354,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.0759,
+      "step": 1183
+    },
+    {
+      "epoch": 4.180052956751986,
+      "grad_norm": 1.349601668880594,
+      "learning_rate": 6.539641643923376e-06,
+      "loss": 0.0762,
+      "step": 1184
+    },
+    {
+      "epoch": 4.183583406884377,
+      "grad_norm": 1.5302048823422634,
+      "learning_rate": 6.534196299302887e-06,
+      "loss": 0.0721,
+      "step": 1185
+    },
+    {
+      "epoch": 4.18711385701677,
+      "grad_norm": 1.1161238197658445,
+      "learning_rate": 6.528748945145749e-06,
+      "loss": 0.0752,
+      "step": 1186
+    },
+    {
+      "epoch": 4.190644307149162,
+      "grad_norm": 1.0876160222524982,
+      "learning_rate": 6.523299588587069e-06,
+      "loss": 0.0672,
+      "step": 1187
+    },
+    {
+      "epoch": 4.194174757281553,
+      "grad_norm": 1.1607986364903538,
+      "learning_rate": 6.51784823676458e-06,
+      "loss": 0.0743,
+      "step": 1188
+    },
+    {
+      "epoch": 4.197705207413946,
+      "grad_norm": 1.0772412583188433,
+      "learning_rate": 6.5123948968186275e-06,
+      "loss": 0.0673,
+      "step": 1189
+    },
+    {
+      "epoch": 4.201235657546337,
+      "grad_norm": 1.2617922130605492,
+      "learning_rate": 6.506939575892161e-06,
+      "loss": 0.0619,
+      "step": 1190
+    },
+    {
+      "epoch": 4.204766107678729,
+      "grad_norm": 1.2483711396156587,
+      "learning_rate": 6.501482281130724e-06,
+      "loss": 0.0871,
+      "step": 1191
+    },
+    {
+      "epoch": 4.208296557811121,
+      "grad_norm": 1.1989951304725375,
+      "learning_rate": 6.496023019682447e-06,
+      "loss": 0.068,
+      "step": 1192
+    },
+    {
+      "epoch": 4.211827007943513,
+      "grad_norm": 1.038016185342428,
+      "learning_rate": 6.490561798698033e-06,
+      "loss": 0.0672,
+      "step": 1193
+    },
+    {
+      "epoch": 4.215357458075904,
+      "grad_norm": 1.2879497667669522,
+      "learning_rate": 6.4850986253307566e-06,
+      "loss": 0.0813,
+      "step": 1194
+    },
+    {
+      "epoch": 4.218887908208297,
+      "grad_norm": 1.138562370359682,
+      "learning_rate": 6.479633506736447e-06,
+      "loss": 0.0798,
+      "step": 1195
+    },
+    {
+      "epoch": 4.222418358340688,
+      "grad_norm": 1.2141342731122955,
+      "learning_rate": 6.474166450073482e-06,
+      "loss": 0.0753,
+      "step": 1196
+    },
+    {
+      "epoch": 4.22594880847308,
+      "grad_norm": 1.1524180729165578,
+      "learning_rate": 6.468697462502776e-06,
+      "loss": 0.0669,
+      "step": 1197
+    },
+    {
+      "epoch": 4.229479258605473,
+      "grad_norm": 1.2503618917005421,
+      "learning_rate": 6.463226551187776e-06,
+      "loss": 0.0802,
+      "step": 1198
+    },
+    {
+      "epoch": 4.233009708737864,
+      "grad_norm": 1.0897556077719,
+      "learning_rate": 6.457753723294443e-06,
+      "loss": 0.0687,
+      "step": 1199
+    },
+    {
+      "epoch": 4.236540158870256,
+      "grad_norm": 1.2215195146222155,
+      "learning_rate": 6.452278985991257e-06,
+      "loss": 0.0859,
+      "step": 1200
+    },
+    {
+      "epoch": 4.240070609002648,
+      "grad_norm": 1.123982384909271,
+      "learning_rate": 6.4468023464491906e-06,
+      "loss": 0.0693,
+      "step": 1201
+    },
+    {
+      "epoch": 4.24360105913504,
+      "grad_norm": 1.161677655741415,
+      "learning_rate": 6.441323811841715e-06,
+      "loss": 0.0749,
+      "step": 1202
+    },
+    {
+      "epoch": 4.247131509267431,
+      "grad_norm": 1.1029803998203793,
+      "learning_rate": 6.435843389344778e-06,
+      "loss": 0.0666,
+      "step": 1203
+    },
+    {
+      "epoch": 4.250661959399824,
+      "grad_norm": 1.3744931507694842,
+      "learning_rate": 6.4303610861368026e-06,
+      "loss": 0.0844,
+      "step": 1204
+    },
+    {
+      "epoch": 4.254192409532215,
+      "grad_norm": 1.0659314315032296,
+      "learning_rate": 6.4248769093986765e-06,
+      "loss": 0.0723,
+      "step": 1205
+    },
+    {
+      "epoch": 4.257722859664607,
+      "grad_norm": 1.2659662434004557,
+      "learning_rate": 6.419390866313741e-06,
+      "loss": 0.0786,
+      "step": 1206
+    },
+    {
+      "epoch": 4.261253309796999,
+      "grad_norm": 1.2058156571448462,
+      "learning_rate": 6.413902964067776e-06,
+      "loss": 0.0708,
+      "step": 1207
+    },
+    {
+      "epoch": 4.264783759929391,
+      "grad_norm": 1.4552116581087131,
+      "learning_rate": 6.408413209849007e-06,
+      "loss": 0.0802,
+      "step": 1208
+    },
+    {
+      "epoch": 4.268314210061783,
+      "grad_norm": 1.3140338475558748,
+      "learning_rate": 6.402921610848076e-06,
+      "loss": 0.0701,
+      "step": 1209
+    },
+    {
+      "epoch": 4.271844660194175,
+      "grad_norm": 1.4948147235906752,
+      "learning_rate": 6.397428174258048e-06,
+      "loss": 0.0765,
+      "step": 1210
+    },
+    {
+      "epoch": 4.275375110326567,
+      "grad_norm": 1.4432674805143244,
+      "learning_rate": 6.391932907274389e-06,
+      "loss": 0.0736,
+      "step": 1211
+    },
+    {
+      "epoch": 4.278905560458958,
+      "grad_norm": 1.1269206127304876,
+      "learning_rate": 6.386435817094971e-06,
+      "loss": 0.0762,
+      "step": 1212
+    },
+    {
+      "epoch": 4.282436010591351,
+      "grad_norm": 1.4426602486276199,
+      "learning_rate": 6.380936910920046e-06,
+      "loss": 0.0834,
+      "step": 1213
+    },
+    {
+      "epoch": 4.285966460723742,
+      "grad_norm": 0.9454852560163308,
+      "learning_rate": 6.375436195952245e-06,
+      "loss": 0.0677,
+      "step": 1214
+    },
+    {
+      "epoch": 4.289496910856134,
+      "grad_norm": 1.1295627967839161,
+      "learning_rate": 6.369933679396571e-06,
+      "loss": 0.057,
+      "step": 1215
+    },
+    {
+      "epoch": 4.293027360988526,
+      "grad_norm": 1.19896192193708,
+      "learning_rate": 6.364429368460391e-06,
+      "loss": 0.0786,
+      "step": 1216
+    },
+    {
+      "epoch": 4.296557811120918,
+      "grad_norm": 1.2164688961559822,
+      "learning_rate": 6.358923270353412e-06,
+      "loss": 0.0743,
+      "step": 1217
+    },
+    {
+      "epoch": 4.300088261253309,
+      "grad_norm": 1.4316527320361974,
+      "learning_rate": 6.353415392287692e-06,
+      "loss": 0.0735,
+      "step": 1218
+    },
+    {
+      "epoch": 4.303618711385702,
+      "grad_norm": 1.1280198948959894,
+      "learning_rate": 6.347905741477613e-06,
+      "loss": 0.0638,
+      "step": 1219
+    },
+    {
+      "epoch": 4.307149161518094,
+      "grad_norm": 1.3537055389950547,
+      "learning_rate": 6.342394325139886e-06,
+      "loss": 0.0844,
+      "step": 1220
+    },
+    {
+      "epoch": 4.310679611650485,
+      "grad_norm": 1.175521881817613,
+      "learning_rate": 6.336881150493525e-06,
+      "loss": 0.0697,
+      "step": 1221
+    },
+    {
+      "epoch": 4.314210061782878,
+      "grad_norm": 1.3032526032205891,
+      "learning_rate": 6.3313662247598575e-06,
+      "loss": 0.0681,
+      "step": 1222
+    },
+    {
+      "epoch": 4.317740511915269,
+      "grad_norm": 1.2291592653948193,
+      "learning_rate": 6.325849555162496e-06,
+      "loss": 0.078,
+      "step": 1223
+    },
+    {
+      "epoch": 4.321270962047661,
+      "grad_norm": 1.2292408189938648,
+      "learning_rate": 6.320331148927344e-06,
+      "loss": 0.0605,
+      "step": 1224
+    },
+    {
+      "epoch": 4.324801412180053,
+      "grad_norm": 1.317695842582305,
+      "learning_rate": 6.314811013282574e-06,
+      "loss": 0.074,
+      "step": 1225
+    },
+    {
+      "epoch": 4.328331862312445,
+      "grad_norm": 1.3300537195408657,
+      "learning_rate": 6.309289155458629e-06,
+      "loss": 0.0781,
+      "step": 1226
+    },
+    {
+      "epoch": 4.331862312444836,
+      "grad_norm": 1.2795961802893723,
+      "learning_rate": 6.303765582688204e-06,
+      "loss": 0.0758,
+      "step": 1227
+    },
+    {
+      "epoch": 4.335392762577229,
+      "grad_norm": 1.089546205227726,
+      "learning_rate": 6.298240302206242e-06,
+      "loss": 0.0716,
+      "step": 1228
+    },
+    {
+      "epoch": 4.338923212709621,
+      "grad_norm": 1.4436985362559518,
+      "learning_rate": 6.29271332124992e-06,
+      "loss": 0.0733,
+      "step": 1229
+    },
+    {
+      "epoch": 4.342453662842012,
+      "grad_norm": 1.2123854120814914,
+      "learning_rate": 6.287184647058648e-06,
+      "loss": 0.063,
+      "step": 1230
+    },
+    {
+      "epoch": 4.345984112974405,
+      "grad_norm": 1.1599856785429075,
+      "learning_rate": 6.281654286874047e-06,
+      "loss": 0.0703,
+      "step": 1231
+    },
+    {
+      "epoch": 4.349514563106796,
+      "grad_norm": 1.2863269775769035,
+      "learning_rate": 6.276122247939953e-06,
+      "loss": 0.08,
+      "step": 1232
+    },
+    {
+      "epoch": 4.353045013239188,
+      "grad_norm": 0.9802224693174152,
+      "learning_rate": 6.2705885375023955e-06,
+      "loss": 0.0669,
+      "step": 1233
+    },
+    {
+      "epoch": 4.35657546337158,
+      "grad_norm": 1.3003686599476758,
+      "learning_rate": 6.265053162809597e-06,
+      "loss": 0.0758,
+      "step": 1234
+    },
+    {
+      "epoch": 4.360105913503972,
+      "grad_norm": 1.2476813765188703,
+      "learning_rate": 6.259516131111957e-06,
+      "loss": 0.0719,
+      "step": 1235
+    },
+    {
+      "epoch": 4.363636363636363,
+      "grad_norm": 1.0504270751123217,
+      "learning_rate": 6.253977449662047e-06,
+      "loss": 0.0614,
+      "step": 1236
+    },
+    {
+      "epoch": 4.367166813768756,
+      "grad_norm": 1.1435259908690352,
+      "learning_rate": 6.2484371257146e-06,
+      "loss": 0.0768,
+      "step": 1237
+    },
+    {
+      "epoch": 4.370697263901147,
+      "grad_norm": 1.3840251434761308,
+      "learning_rate": 6.242895166526502e-06,
+      "loss": 0.0742,
+      "step": 1238
+    },
+    {
+      "epoch": 4.374227714033539,
+      "grad_norm": 1.2422183469141699,
+      "learning_rate": 6.2373515793567715e-06,
+      "loss": 0.08,
+      "step": 1239
+    },
+    {
+      "epoch": 4.3777581641659316,
+      "grad_norm": 1.1315889355948883,
+      "learning_rate": 6.231806371466574e-06,
+      "loss": 0.0697,
+      "step": 1240
+    },
+    {
+      "epoch": 4.381288614298323,
+      "grad_norm": 1.081317701161799,
+      "learning_rate": 6.226259550119187e-06,
+      "loss": 0.0672,
+      "step": 1241
+    },
+    {
+      "epoch": 4.384819064430715,
+      "grad_norm": 1.139178949237585,
+      "learning_rate": 6.220711122580004e-06,
+      "loss": 0.0762,
+      "step": 1242
+    },
+    {
+      "epoch": 4.388349514563107,
+      "grad_norm": 1.2934745666925676,
+      "learning_rate": 6.215161096116524e-06,
+      "loss": 0.081,
+      "step": 1243
+    },
+    {
+      "epoch": 4.391879964695499,
+      "grad_norm": 1.4026385643923822,
+      "learning_rate": 6.209609477998339e-06,
+      "loss": 0.0759,
+      "step": 1244
+    },
+    {
+      "epoch": 4.39541041482789,
+      "grad_norm": 1.2849606159130602,
+      "learning_rate": 6.2040562754971265e-06,
+      "loss": 0.0713,
+      "step": 1245
+    },
+    {
+      "epoch": 4.398940864960283,
+      "grad_norm": 1.0919624470047933,
+      "learning_rate": 6.1985014958866386e-06,
+      "loss": 0.0731,
+      "step": 1246
+    },
+    {
+      "epoch": 4.402471315092674,
+      "grad_norm": 1.2552988710153292,
+      "learning_rate": 6.192945146442693e-06,
+      "loss": 0.0802,
+      "step": 1247
+    },
+    {
+      "epoch": 4.406001765225066,
+      "grad_norm": 1.2944459168853082,
+      "learning_rate": 6.187387234443164e-06,
+      "loss": 0.084,
+      "step": 1248
+    },
+    {
+      "epoch": 4.4095322153574585,
+      "grad_norm": 1.1228808317446295,
+      "learning_rate": 6.181827767167974e-06,
+      "loss": 0.0716,
+      "step": 1249
+    },
+    {
+      "epoch": 4.41306266548985,
+      "grad_norm": 1.1881280477511678,
+      "learning_rate": 6.176266751899078e-06,
+      "loss": 0.067,
+      "step": 1250
+    },
+    {
+      "epoch": 4.416593115622242,
+      "grad_norm": 1.1569504030576783,
+      "learning_rate": 6.170704195920465e-06,
+      "loss": 0.0702,
+      "step": 1251
+    },
+    {
+      "epoch": 4.420123565754634,
+      "grad_norm": 1.189002472009121,
+      "learning_rate": 6.165140106518137e-06,
+      "loss": 0.0694,
+      "step": 1252
+    },
+    {
+      "epoch": 4.423654015887026,
+      "grad_norm": 1.232505313641185,
+      "learning_rate": 6.1595744909801085e-06,
+      "loss": 0.0754,
+      "step": 1253
+    },
+    {
+      "epoch": 4.427184466019417,
+      "grad_norm": 1.2147460452464223,
+      "learning_rate": 6.154007356596387e-06,
+      "loss": 0.0652,
+      "step": 1254
+    },
+    {
+      "epoch": 4.43071491615181,
+      "grad_norm": 1.1480977785419901,
+      "learning_rate": 6.148438710658979e-06,
+      "loss": 0.0625,
+      "step": 1255
+    },
+    {
+      "epoch": 4.434245366284201,
+      "grad_norm": 1.218331723091377,
+      "learning_rate": 6.142868560461859e-06,
+      "loss": 0.0821,
+      "step": 1256
+    },
+    {
+      "epoch": 4.437775816416593,
+      "grad_norm": 1.1738427410555219,
+      "learning_rate": 6.137296913300982e-06,
+      "loss": 0.074,
+      "step": 1257
+    },
+    {
+      "epoch": 4.441306266548985,
+      "grad_norm": 1.19855168208864,
+      "learning_rate": 6.131723776474258e-06,
+      "loss": 0.0738,
+      "step": 1258
+    },
+    {
+      "epoch": 4.444836716681377,
+      "grad_norm": 1.3103544915169332,
+      "learning_rate": 6.12614915728155e-06,
+      "loss": 0.0781,
+      "step": 1259
+    },
+    {
+      "epoch": 4.448367166813769,
+      "grad_norm": 1.0115359906207875,
+      "learning_rate": 6.120573063024663e-06,
+      "loss": 0.0715,
+      "step": 1260
+    },
+    {
+      "epoch": 4.451897616946161,
+      "grad_norm": 0.9684920252271274,
+      "learning_rate": 6.114995501007334e-06,
+      "loss": 0.0626,
+      "step": 1261
+    },
+    {
+      "epoch": 4.455428067078553,
+      "grad_norm": 1.0836045218047463,
+      "learning_rate": 6.109416478535222e-06,
+      "loss": 0.0707,
+      "step": 1262
+    },
+    {
+      "epoch": 4.458958517210944,
+      "grad_norm": 1.1090014070135534,
+      "learning_rate": 6.103836002915898e-06,
+      "loss": 0.0638,
+      "step": 1263
+    },
+    {
+      "epoch": 4.4624889673433366,
+      "grad_norm": 1.300961015257033,
+      "learning_rate": 6.098254081458839e-06,
+      "loss": 0.0727,
+      "step": 1264
+    },
+    {
+      "epoch": 4.466019417475728,
+      "grad_norm": 1.2337280261124597,
+      "learning_rate": 6.092670721475414e-06,
+      "loss": 0.0748,
+      "step": 1265
+    },
+    {
+      "epoch": 4.46954986760812,
+      "grad_norm": 1.1278101885787395,
+      "learning_rate": 6.0870859302788756e-06,
+      "loss": 0.0703,
+      "step": 1266
+    },
+    {
+      "epoch": 4.473080317740512,
+      "grad_norm": 1.0631423710066719,
+      "learning_rate": 6.081499715184353e-06,
+      "loss": 0.0706,
+      "step": 1267
+    },
+    {
+      "epoch": 4.476610767872904,
+      "grad_norm": 1.143193764341101,
+      "learning_rate": 6.075912083508839e-06,
+      "loss": 0.0694,
+      "step": 1268
+    },
+    {
+      "epoch": 4.480141218005295,
+      "grad_norm": 1.1879992175907763,
+      "learning_rate": 6.070323042571184e-06,
+      "loss": 0.0683,
+      "step": 1269
+    },
+    {
+      "epoch": 4.483671668137688,
+      "grad_norm": 1.4154753236711644,
+      "learning_rate": 6.064732599692079e-06,
+      "loss": 0.0841,
+      "step": 1270
+    },
+    {
+      "epoch": 4.48720211827008,
+      "grad_norm": 1.0845224887804554,
+      "learning_rate": 6.059140762194059e-06,
+      "loss": 0.0753,
+      "step": 1271
+    },
+    {
+      "epoch": 4.490732568402471,
+      "grad_norm": 1.1903474371803968,
+      "learning_rate": 6.053547537401477e-06,
+      "loss": 0.0706,
+      "step": 1272
+    },
+    {
+      "epoch": 4.4942630185348635,
+      "grad_norm": 1.616858067887646,
+      "learning_rate": 6.047952932640513e-06,
+      "loss": 0.0912,
+      "step": 1273
+    },
+    {
+      "epoch": 4.497793468667255,
+      "grad_norm": 1.0913358558919948,
+      "learning_rate": 6.042356955239145e-06,
+      "loss": 0.0604,
+      "step": 1274
+    },
+    {
+      "epoch": 4.501323918799647,
+      "grad_norm": 1.1880905119942522,
+      "learning_rate": 6.036759612527157e-06,
+      "loss": 0.0759,
+      "step": 1275
+    },
+    {
+      "epoch": 4.504854368932039,
+      "grad_norm": 1.1747169454730277,
+      "learning_rate": 6.031160911836112e-06,
+      "loss": 0.0679,
+      "step": 1276
+    },
+    {
+      "epoch": 4.508384819064431,
+      "grad_norm": 1.211427366915763,
+      "learning_rate": 6.025560860499365e-06,
+      "loss": 0.0689,
+      "step": 1277
+    },
+    {
+      "epoch": 4.511915269196822,
+      "grad_norm": 1.417931942561815,
+      "learning_rate": 6.019959465852025e-06,
+      "loss": 0.0868,
+      "step": 1278
+    },
+    {
+      "epoch": 4.515445719329215,
+      "grad_norm": 1.2209993004826432,
+      "learning_rate": 6.014356735230973e-06,
+      "loss": 0.0646,
+      "step": 1279
+    },
+    {
+      "epoch": 4.518976169461606,
+      "grad_norm": 1.1649085744269059,
+      "learning_rate": 6.00875267597483e-06,
+      "loss": 0.0639,
+      "step": 1280
+    },
+    {
+      "epoch": 4.522506619593998,
+      "grad_norm": 1.011949914685396,
+      "learning_rate": 6.003147295423967e-06,
+      "loss": 0.0597,
+      "step": 1281
+    },
+    {
+      "epoch": 4.5260370697263905,
+      "grad_norm": 1.1309980576687069,
+      "learning_rate": 5.997540600920479e-06,
+      "loss": 0.0782,
+      "step": 1282
+    },
+    {
+      "epoch": 4.529567519858782,
+      "grad_norm": 1.0299510613826361,
+      "learning_rate": 5.9919325998081825e-06,
+      "loss": 0.066,
+      "step": 1283
+    },
+    {
+      "epoch": 4.533097969991174,
+      "grad_norm": 1.0789969131073938,
+      "learning_rate": 5.986323299432608e-06,
+      "loss": 0.0719,
+      "step": 1284
+    },
+    {
+      "epoch": 4.536628420123566,
+      "grad_norm": 1.0170912861541357,
+      "learning_rate": 5.980712707140985e-06,
+      "loss": 0.064,
+      "step": 1285
+    },
+    {
+      "epoch": 4.540158870255958,
+      "grad_norm": 1.1901299269101344,
+      "learning_rate": 5.97510083028224e-06,
+      "loss": 0.0707,
+      "step": 1286
+    },
+    {
+      "epoch": 4.543689320388349,
+      "grad_norm": 1.236036942535051,
+      "learning_rate": 5.969487676206975e-06,
+      "loss": 0.0695,
+      "step": 1287
+    },
+    {
+      "epoch": 4.5472197705207416,
+      "grad_norm": 1.0739729905219972,
+      "learning_rate": 5.9638732522674716e-06,
+      "loss": 0.0683,
+      "step": 1288
+    },
+    {
+      "epoch": 4.550750220653133,
+      "grad_norm": 1.1896504864727457,
+      "learning_rate": 5.95825756581767e-06,
+      "loss": 0.0692,
+      "step": 1289
+    },
+    {
+      "epoch": 4.554280670785525,
+      "grad_norm": 1.0099956859061028,
+      "learning_rate": 5.952640624213166e-06,
+      "loss": 0.0717,
+      "step": 1290
+    },
+    {
+      "epoch": 4.557811120917917,
+      "grad_norm": 1.1995738747246374,
+      "learning_rate": 5.947022434811202e-06,
+      "loss": 0.0611,
+      "step": 1291
+    },
+    {
+      "epoch": 4.561341571050309,
+      "grad_norm": 1.6264487607523304,
+      "learning_rate": 5.94140300497065e-06,
+      "loss": 0.0817,
+      "step": 1292
+    },
+    {
+      "epoch": 4.564872021182701,
+      "grad_norm": 1.2030247661685802,
+      "learning_rate": 5.935782342052007e-06,
+      "loss": 0.0605,
+      "step": 1293
+    },
+    {
+      "epoch": 4.568402471315093,
+      "grad_norm": 1.117896941640317,
+      "learning_rate": 5.93016045341739e-06,
+      "loss": 0.0521,
+      "step": 1294
+    },
+    {
+      "epoch": 4.571932921447485,
+      "grad_norm": 1.282540263365002,
+      "learning_rate": 5.924537346430517e-06,
+      "loss": 0.0719,
+      "step": 1295
+    },
+    {
+      "epoch": 4.575463371579876,
+      "grad_norm": 1.4173574483743017,
+      "learning_rate": 5.918913028456705e-06,
+      "loss": 0.0665,
+      "step": 1296
+    },
+    {
+      "epoch": 4.5789938217122685,
+      "grad_norm": 1.7378141839355061,
+      "learning_rate": 5.9132875068628555e-06,
+      "loss": 0.082,
+      "step": 1297
+    },
+    {
+      "epoch": 4.58252427184466,
+      "grad_norm": 1.5474239856959229,
+      "learning_rate": 5.907660789017446e-06,
+      "loss": 0.0747,
+      "step": 1298
+    },
+    {
+      "epoch": 4.586054721977052,
+      "grad_norm": 1.2885896943234003,
+      "learning_rate": 5.902032882290518e-06,
+      "loss": 0.0744,
+      "step": 1299
+    },
+    {
+      "epoch": 4.589585172109444,
+      "grad_norm": 1.3460860331275364,
+      "learning_rate": 5.896403794053679e-06,
+      "loss": 0.0767,
+      "step": 1300
+    },
+    {
+      "epoch": 4.593115622241836,
+      "grad_norm": 1.1096734280955396,
+      "learning_rate": 5.890773531680076e-06,
+      "loss": 0.0777,
+      "step": 1301
+    },
+    {
+      "epoch": 4.596646072374227,
+      "grad_norm": 1.1779234981799707,
+      "learning_rate": 5.885142102544398e-06,
+      "loss": 0.0636,
+      "step": 1302
+    },
+    {
+      "epoch": 4.60017652250662,
+      "grad_norm": 1.3624220771981437,
+      "learning_rate": 5.8795095140228595e-06,
+      "loss": 0.0894,
+      "step": 1303
+    },
+    {
+      "epoch": 4.603706972639012,
+      "grad_norm": 1.343188625919169,
+      "learning_rate": 5.8738757734931975e-06,
+      "loss": 0.0809,
+      "step": 1304
+    },
+    {
+      "epoch": 4.607237422771403,
+      "grad_norm": 1.4446432282791357,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.0766,
+      "step": 1305
+    },
+    {
+      "epoch": 4.6107678729037955,
+      "grad_norm": 1.3556894299906552,
+      "learning_rate": 5.86260486592797e-06,
+      "loss": 0.0797,
+      "step": 1306
+    },
+    {
+      "epoch": 4.614298323036187,
+      "grad_norm": 1.278747621381562,
+      "learning_rate": 5.856967713655379e-06,
+      "loss": 0.0833,
+      "step": 1307
+    },
+    {
+      "epoch": 4.617828773168579,
+      "grad_norm": 1.1918217353301246,
+      "learning_rate": 5.851329438900595e-06,
+      "loss": 0.0799,
+      "step": 1308
+    },
+    {
+      "epoch": 4.621359223300971,
+      "grad_norm": 1.1278725314966394,
+      "learning_rate": 5.845690049048799e-06,
+      "loss": 0.0643,
+      "step": 1309
+    },
+    {
+      "epoch": 4.624889673433363,
+      "grad_norm": 1.2557557372315369,
+      "learning_rate": 5.8400495514866355e-06,
+      "loss": 0.0879,
+      "step": 1310
+    },
+    {
+      "epoch": 4.628420123565754,
+      "grad_norm": 1.2461622162231867,
+      "learning_rate": 5.8344079536021955e-06,
+      "loss": 0.0809,
+      "step": 1311
+    },
+    {
+      "epoch": 4.631950573698147,
+      "grad_norm": 1.2846191093338162,
+      "learning_rate": 5.82876526278502e-06,
+      "loss": 0.0795,
+      "step": 1312
+    },
+    {
+      "epoch": 4.635481023830538,
+      "grad_norm": 1.1240893202153388,
+      "learning_rate": 5.823121486426069e-06,
+      "loss": 0.073,
+      "step": 1313
+    },
+    {
+      "epoch": 4.63901147396293,
+      "grad_norm": 1.2575691959822162,
+      "learning_rate": 5.817476631917736e-06,
+      "loss": 0.0843,
+      "step": 1314
+    },
+    {
+      "epoch": 4.6425419240953225,
+      "grad_norm": 1.017222394290193,
+      "learning_rate": 5.811830706653819e-06,
+      "loss": 0.0673,
+      "step": 1315
+    },
+    {
+      "epoch": 4.646072374227714,
+      "grad_norm": 1.0729116322930856,
+      "learning_rate": 5.806183718029524e-06,
+      "loss": 0.0732,
+      "step": 1316
+    },
+    {
+      "epoch": 4.649602824360106,
+      "grad_norm": 1.1579892771980353,
+      "learning_rate": 5.800535673441446e-06,
+      "loss": 0.0775,
+      "step": 1317
+    },
+    {
+      "epoch": 4.653133274492498,
+      "grad_norm": 1.0293530970686642,
+      "learning_rate": 5.794886580287565e-06,
+      "loss": 0.0636,
+      "step": 1318
+    },
+    {
+      "epoch": 4.65666372462489,
+      "grad_norm": 0.9376266948248516,
+      "learning_rate": 5.789236445967233e-06,
+      "loss": 0.0603,
+      "step": 1319
+    },
+    {
+      "epoch": 4.660194174757281,
+      "grad_norm": 0.9987883216154062,
+      "learning_rate": 5.783585277881167e-06,
+      "loss": 0.0762,
+      "step": 1320
+    },
+    {
+      "epoch": 4.6637246248896735,
+      "grad_norm": 1.2329836983396925,
+      "learning_rate": 5.777933083431437e-06,
+      "loss": 0.0831,
+      "step": 1321
+    },
+    {
+      "epoch": 4.667255075022065,
+      "grad_norm": 1.1400090314981657,
+      "learning_rate": 5.77227987002146e-06,
+      "loss": 0.0715,
+      "step": 1322
+    },
+    {
+      "epoch": 4.670785525154457,
+      "grad_norm": 1.108926279713076,
+      "learning_rate": 5.766625645055982e-06,
+      "loss": 0.0727,
+      "step": 1323
+    },
+    {
+      "epoch": 4.674315975286849,
+      "grad_norm": 1.036285195186968,
+      "learning_rate": 5.760970415941082e-06,
+      "loss": 0.0619,
+      "step": 1324
+    },
+    {
+      "epoch": 4.677846425419241,
+      "grad_norm": 1.2377457117763253,
+      "learning_rate": 5.755314190084146e-06,
+      "loss": 0.0679,
+      "step": 1325
+    },
+    {
+      "epoch": 4.681376875551633,
+      "grad_norm": 1.2816857024923805,
+      "learning_rate": 5.749656974893873e-06,
+      "loss": 0.0761,
+      "step": 1326
+    },
+    {
+      "epoch": 4.684907325684025,
+      "grad_norm": 1.4320175248552778,
+      "learning_rate": 5.743998777780252e-06,
+      "loss": 0.0735,
+      "step": 1327
+    },
+    {
+      "epoch": 4.688437775816417,
+      "grad_norm": 1.21805904446881,
+      "learning_rate": 5.73833960615456e-06,
+      "loss": 0.0654,
+      "step": 1328
+    },
+    {
+      "epoch": 4.691968225948808,
+      "grad_norm": 1.2720158896144405,
+      "learning_rate": 5.732679467429353e-06,
+      "loss": 0.0872,
+      "step": 1329
+    },
+    {
+      "epoch": 4.6954986760812005,
+      "grad_norm": 1.362220253742237,
+      "learning_rate": 5.72701836901845e-06,
+      "loss": 0.0763,
+      "step": 1330
+    },
+    {
+      "epoch": 4.699029126213592,
+      "grad_norm": 1.3376103897361957,
+      "learning_rate": 5.7213563183369295e-06,
+      "loss": 0.0718,
+      "step": 1331
+    },
+    {
+      "epoch": 4.702559576345984,
+      "grad_norm": 1.2102663618571865,
+      "learning_rate": 5.715693322801117e-06,
+      "loss": 0.0738,
+      "step": 1332
+    },
+    {
+      "epoch": 4.7060900264783765,
+      "grad_norm": 1.2277058279665478,
+      "learning_rate": 5.710029389828572e-06,
+      "loss": 0.0719,
+      "step": 1333
+    },
+    {
+      "epoch": 4.709620476610768,
+      "grad_norm": 1.2773556567896192,
+      "learning_rate": 5.70436452683809e-06,
+      "loss": 0.0902,
+      "step": 1334
+    },
+    {
+      "epoch": 4.713150926743159,
+      "grad_norm": 1.4332928684103645,
+      "learning_rate": 5.698698741249676e-06,
+      "loss": 0.0709,
+      "step": 1335
+    },
+    {
+      "epoch": 4.716681376875552,
+      "grad_norm": 1.1461862595511798,
+      "learning_rate": 5.6930320404845475e-06,
+      "loss": 0.0689,
+      "step": 1336
+    },
+    {
+      "epoch": 4.720211827007944,
+      "grad_norm": 1.141265802102587,
+      "learning_rate": 5.687364431965121e-06,
+      "loss": 0.0836,
+      "step": 1337
+    },
+    {
+      "epoch": 4.723742277140335,
+      "grad_norm": 1.503230303094657,
+      "learning_rate": 5.681695923115002e-06,
+      "loss": 0.0819,
+      "step": 1338
+    },
+    {
+      "epoch": 4.7272727272727275,
+      "grad_norm": 1.1626902290316583,
+      "learning_rate": 5.676026521358973e-06,
+      "loss": 0.0781,
+      "step": 1339
+    },
+    {
+      "epoch": 4.730803177405119,
+      "grad_norm": 1.1620016918977858,
+      "learning_rate": 5.6703562341229886e-06,
+      "loss": 0.0805,
+      "step": 1340
+    },
+    {
+      "epoch": 4.734333627537511,
+      "grad_norm": 0.980205333032981,
+      "learning_rate": 5.664685068834164e-06,
+      "loss": 0.06,
+      "step": 1341
+    },
+    {
+      "epoch": 4.737864077669903,
+      "grad_norm": 1.0919120901198711,
+      "learning_rate": 5.659013032920758e-06,
+      "loss": 0.0667,
+      "step": 1342
+    },
+    {
+      "epoch": 4.741394527802295,
+      "grad_norm": 1.058369377556783,
+      "learning_rate": 5.653340133812179e-06,
+      "loss": 0.0678,
+      "step": 1343
+    },
+    {
+      "epoch": 4.744924977934687,
+      "grad_norm": 1.2729504832126617,
+      "learning_rate": 5.647666378938959e-06,
+      "loss": 0.0746,
+      "step": 1344
+    },
+    {
+      "epoch": 4.7484554280670785,
+      "grad_norm": 1.2480081282387157,
+      "learning_rate": 5.641991775732756e-06,
+      "loss": 0.0711,
+      "step": 1345
+    },
+    {
+      "epoch": 4.75198587819947,
+      "grad_norm": 1.1576335449365858,
+      "learning_rate": 5.636316331626332e-06,
+      "loss": 0.074,
+      "step": 1346
+    },
+    {
+      "epoch": 4.755516328331862,
+      "grad_norm": 1.1637647202647143,
+      "learning_rate": 5.630640054053562e-06,
+      "loss": 0.0695,
+      "step": 1347
+    },
+    {
+      "epoch": 4.7590467784642545,
+      "grad_norm": 1.1265709555415027,
+      "learning_rate": 5.624962950449402e-06,
+      "loss": 0.0716,
+      "step": 1348
+    },
+    {
+      "epoch": 4.762577228596646,
+      "grad_norm": 1.2129102271686025,
+      "learning_rate": 5.6192850282498925e-06,
+      "loss": 0.0728,
+      "step": 1349
+    },
+    {
+      "epoch": 4.766107678729038,
+      "grad_norm": 1.4098311899227094,
+      "learning_rate": 5.6136062948921486e-06,
+      "loss": 0.0695,
+      "step": 1350
+    },
+    {
+      "epoch": 4.76963812886143,
+      "grad_norm": 1.0254747813352643,
+      "learning_rate": 5.607926757814347e-06,
+      "loss": 0.072,
+      "step": 1351
+    },
+    {
+      "epoch": 4.773168578993822,
+      "grad_norm": 1.3062783164855614,
+      "learning_rate": 5.6022464244557175e-06,
+      "loss": 0.0836,
+      "step": 1352
+    },
+    {
+      "epoch": 4.776699029126213,
+      "grad_norm": 1.1787618293744981,
+      "learning_rate": 5.596565302256531e-06,
+      "loss": 0.0644,
+      "step": 1353
+    },
+    {
+      "epoch": 4.7802294792586055,
+      "grad_norm": 1.4645406649621868,
+      "learning_rate": 5.590883398658095e-06,
+      "loss": 0.0786,
+      "step": 1354
+    },
+    {
+      "epoch": 4.783759929390998,
+      "grad_norm": 1.3381703719371458,
+      "learning_rate": 5.585200721102737e-06,
+      "loss": 0.0855,
+      "step": 1355
+    },
+    {
+      "epoch": 4.787290379523389,
+      "grad_norm": 1.5162385054662686,
+      "learning_rate": 5.579517277033799e-06,
+      "loss": 0.0727,
+      "step": 1356
+    },
+    {
+      "epoch": 4.7908208296557815,
+      "grad_norm": 1.333030859041725,
+      "learning_rate": 5.573833073895629e-06,
+      "loss": 0.0872,
+      "step": 1357
+    },
+    {
+      "epoch": 4.794351279788173,
+      "grad_norm": 1.2513683261705562,
+      "learning_rate": 5.568148119133566e-06,
+      "loss": 0.0661,
+      "step": 1358
+    },
+    {
+      "epoch": 4.797881729920565,
+      "grad_norm": 1.204520949412178,
+      "learning_rate": 5.562462420193939e-06,
+      "loss": 0.0769,
+      "step": 1359
+    },
+    {
+      "epoch": 4.801412180052957,
+      "grad_norm": 1.2922206835673506,
+      "learning_rate": 5.556775984524044e-06,
+      "loss": 0.0718,
+      "step": 1360
+    },
+    {
+      "epoch": 4.804942630185349,
+      "grad_norm": 1.4591640646940425,
+      "learning_rate": 5.55108881957215e-06,
+      "loss": 0.0648,
+      "step": 1361
+    },
+    {
+      "epoch": 4.80847308031774,
+      "grad_norm": 1.2720780355230428,
+      "learning_rate": 5.545400932787476e-06,
+      "loss": 0.0735,
+      "step": 1362
+    },
+    {
+      "epoch": 4.8120035304501325,
+      "grad_norm": 1.2599249791592424,
+      "learning_rate": 5.539712331620186e-06,
+      "loss": 0.062,
+      "step": 1363
+    },
+    {
+      "epoch": 4.815533980582524,
+      "grad_norm": 1.0986643518131567,
+      "learning_rate": 5.534023023521384e-06,
+      "loss": 0.0701,
+      "step": 1364
+    },
+    {
+      "epoch": 4.819064430714916,
+      "grad_norm": 1.194220635218748,
+      "learning_rate": 5.528333015943097e-06,
+      "loss": 0.0663,
+      "step": 1365
+    },
+    {
+      "epoch": 4.8225948808473085,
+      "grad_norm": 1.1823226864675451,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.059,
+      "step": 1366
+    },
+    {
+      "epoch": 4.8261253309797,
+      "grad_norm": 1.225355888651712,
+      "learning_rate": 5.516950932160748e-06,
+      "loss": 0.0755,
+      "step": 1367
+    },
+    {
+      "epoch": 4.829655781112092,
+      "grad_norm": 1.417142754172787,
+      "learning_rate": 5.511258870865283e-06,
+      "loss": 0.0907,
+      "step": 1368
+    },
+    {
+      "epoch": 4.8331862312444835,
+      "grad_norm": 1.315652572963653,
+      "learning_rate": 5.50556613990751e-06,
+      "loss": 0.0807,
+      "step": 1369
+    },
+    {
+      "epoch": 4.836716681376876,
+      "grad_norm": 1.4848341875180038,
+      "learning_rate": 5.499872746743935e-06,
+      "loss": 0.1056,
+      "step": 1370
+    },
+    {
+      "epoch": 4.840247131509267,
+      "grad_norm": 1.5291378895324677,
+      "learning_rate": 5.49417869883194e-06,
+      "loss": 0.0837,
+      "step": 1371
+    },
+    {
+      "epoch": 4.8437775816416595,
+      "grad_norm": 1.226839957471945,
+      "learning_rate": 5.488484003629759e-06,
+      "loss": 0.0779,
+      "step": 1372
+    },
+    {
+      "epoch": 4.847308031774051,
+      "grad_norm": 1.3463992269178928,
+      "learning_rate": 5.4827886685964774e-06,
+      "loss": 0.0976,
+      "step": 1373
+    },
+    {
+      "epoch": 4.850838481906443,
+      "grad_norm": 1.1494326044050664,
+      "learning_rate": 5.477092701192014e-06,
+      "loss": 0.0649,
+      "step": 1374
+    },
+    {
+      "epoch": 4.854368932038835,
+      "grad_norm": 1.2447420578267399,
+      "learning_rate": 5.471396108877123e-06,
+      "loss": 0.072,
+      "step": 1375
+    },
+    {
+      "epoch": 4.857899382171227,
+      "grad_norm": 1.3612258681021532,
+      "learning_rate": 5.465698899113371e-06,
+      "loss": 0.0807,
+      "step": 1376
+    },
+    {
+      "epoch": 4.861429832303619,
+      "grad_norm": 1.4193052420845504,
+      "learning_rate": 5.460001079363135e-06,
+      "loss": 0.0811,
+      "step": 1377
+    },
+    {
+      "epoch": 4.8649602824360105,
+      "grad_norm": 1.2743805042555498,
+      "learning_rate": 5.454302657089592e-06,
+      "loss": 0.08,
+      "step": 1378
+    },
+    {
+      "epoch": 4.868490732568403,
+      "grad_norm": 1.078809447494733,
+      "learning_rate": 5.448603639756708e-06,
+      "loss": 0.062,
+      "step": 1379
+    },
+    {
+      "epoch": 4.872021182700794,
+      "grad_norm": 1.0799797925482022,
+      "learning_rate": 5.442904034829226e-06,
+      "loss": 0.075,
+      "step": 1380
+    },
+    {
+      "epoch": 4.8755516328331865,
+      "grad_norm": 1.141767854824888,
+      "learning_rate": 5.437203849772664e-06,
+      "loss": 0.0636,
+      "step": 1381
+    },
+    {
+      "epoch": 4.879082082965578,
+      "grad_norm": 1.123958849288234,
+      "learning_rate": 5.431503092053294e-06,
+      "loss": 0.077,
+      "step": 1382
+    },
+    {
+      "epoch": 4.88261253309797,
+      "grad_norm": 0.9850627842600758,
+      "learning_rate": 5.4258017691381415e-06,
+      "loss": 0.0602,
+      "step": 1383
+    },
+    {
+      "epoch": 4.886142983230362,
+      "grad_norm": 1.1111371089369984,
+      "learning_rate": 5.420099888494972e-06,
+      "loss": 0.072,
+      "step": 1384
+    },
+    {
+      "epoch": 4.889673433362754,
+      "grad_norm": 1.116994328139007,
+      "learning_rate": 5.414397457592278e-06,
+      "loss": 0.0769,
+      "step": 1385
+    },
+    {
+      "epoch": 4.893203883495145,
+      "grad_norm": 1.4416102109718356,
+      "learning_rate": 5.408694483899281e-06,
+      "loss": 0.0813,
+      "step": 1386
+    },
+    {
+      "epoch": 4.8967343336275375,
+      "grad_norm": 1.4613051629608635,
+      "learning_rate": 5.402990974885904e-06,
+      "loss": 0.0821,
+      "step": 1387
+    },
+    {
+      "epoch": 4.90026478375993,
+      "grad_norm": 1.1868315059556709,
+      "learning_rate": 5.3972869380227765e-06,
+      "loss": 0.0804,
+      "step": 1388
+    },
+    {
+      "epoch": 4.903795233892321,
+      "grad_norm": 1.1666204562256506,
+      "learning_rate": 5.3915823807812196e-06,
+      "loss": 0.0785,
+      "step": 1389
+    },
+    {
+      "epoch": 4.9073256840247135,
+      "grad_norm": 1.1215733098359313,
+      "learning_rate": 5.385877310633233e-06,
+      "loss": 0.0745,
+      "step": 1390
+    },
+    {
+      "epoch": 4.910856134157105,
+      "grad_norm": 1.585401934451052,
+      "learning_rate": 5.38017173505149e-06,
+      "loss": 0.0797,
+      "step": 1391
+    },
+    {
+      "epoch": 4.914386584289497,
+      "grad_norm": 1.3919507079497462,
+      "learning_rate": 5.374465661509326e-06,
+      "loss": 0.0853,
+      "step": 1392
+    },
+    {
+      "epoch": 4.9179170344218885,
+      "grad_norm": 1.1020309883219317,
+      "learning_rate": 5.368759097480728e-06,
+      "loss": 0.0738,
+      "step": 1393
+    },
+    {
+      "epoch": 4.921447484554281,
+      "grad_norm": 1.2530954184669498,
+      "learning_rate": 5.363052050440327e-06,
+      "loss": 0.0649,
+      "step": 1394
+    },
+    {
+      "epoch": 4.924977934686672,
+      "grad_norm": 1.3563436418670745,
+      "learning_rate": 5.357344527863385e-06,
+      "loss": 0.0848,
+      "step": 1395
+    },
+    {
+      "epoch": 4.9285083848190645,
+      "grad_norm": 1.2161047750969436,
+      "learning_rate": 5.35163653722579e-06,
+      "loss": 0.0711,
+      "step": 1396
+    },
+    {
+      "epoch": 4.932038834951456,
+      "grad_norm": 1.1405877554596378,
+      "learning_rate": 5.345928086004034e-06,
+      "loss": 0.0759,
+      "step": 1397
+    },
+    {
+      "epoch": 4.935569285083848,
+      "grad_norm": 1.5340642291990807,
+      "learning_rate": 5.340219181675227e-06,
+      "loss": 0.0756,
+      "step": 1398
+    },
+    {
+      "epoch": 4.9390997352162405,
+      "grad_norm": 1.4009707021891677,
+      "learning_rate": 5.334509831717058e-06,
+      "loss": 0.0773,
+      "step": 1399
+    },
+    {
+      "epoch": 4.942630185348632,
+      "grad_norm": 1.2971256965481643,
+      "learning_rate": 5.328800043607809e-06,
+      "loss": 0.0891,
+      "step": 1400
+    },
+    {
+      "epoch": 4.946160635481024,
+      "grad_norm": 1.378471603765613,
+      "learning_rate": 5.323089824826332e-06,
+      "loss": 0.0812,
+      "step": 1401
+    },
+    {
+      "epoch": 4.9496910856134155,
+      "grad_norm": 0.9246558166305276,
+      "learning_rate": 5.317379182852044e-06,
+      "loss": 0.0652,
+      "step": 1402
+    },
+    {
+      "epoch": 4.953221535745808,
+      "grad_norm": 1.1283491586166101,
+      "learning_rate": 5.311668125164916e-06,
+      "loss": 0.0747,
+      "step": 1403
+    },
+    {
+      "epoch": 4.956751985878199,
+      "grad_norm": 1.1947879187764758,
+      "learning_rate": 5.305956659245466e-06,
+      "loss": 0.0871,
+      "step": 1404
+    },
+    {
+      "epoch": 4.9602824360105915,
+      "grad_norm": 1.1562516815747672,
+      "learning_rate": 5.300244792574743e-06,
+      "loss": 0.0818,
+      "step": 1405
+    },
+    {
+      "epoch": 4.963812886142983,
+      "grad_norm": 1.2658874903814799,
+      "learning_rate": 5.294532532634321e-06,
+      "loss": 0.0784,
+      "step": 1406
+    },
+    {
+      "epoch": 4.967343336275375,
+      "grad_norm": 1.0490558876913674,
+      "learning_rate": 5.288819886906293e-06,
+      "loss": 0.0618,
+      "step": 1407
+    },
+    {
+      "epoch": 4.970873786407767,
+      "grad_norm": 1.2147121704532486,
+      "learning_rate": 5.283106862873253e-06,
+      "loss": 0.0799,
+      "step": 1408
+    },
+    {
+      "epoch": 4.974404236540159,
+      "grad_norm": 1.1451504862662256,
+      "learning_rate": 5.2773934680182935e-06,
+      "loss": 0.0668,
+      "step": 1409
+    },
+    {
+      "epoch": 4.977934686672551,
+      "grad_norm": 1.1648655228082387,
+      "learning_rate": 5.271679709824993e-06,
+      "loss": 0.0742,
+      "step": 1410
+    },
+    {
+      "epoch": 4.9814651368049425,
+      "grad_norm": 1.1046312634759032,
+      "learning_rate": 5.265965595777402e-06,
+      "loss": 0.0814,
+      "step": 1411
+    },
+    {
+      "epoch": 4.984995586937335,
+      "grad_norm": 1.1273037101841474,
+      "learning_rate": 5.2602511333600414e-06,
+      "loss": 0.0785,
+      "step": 1412
+    },
+    {
+      "epoch": 4.988526037069726,
+      "grad_norm": 1.354622519190956,
+      "learning_rate": 5.254536330057884e-06,
+      "loss": 0.0785,
+      "step": 1413
+    },
+    {
+      "epoch": 4.9920564872021185,
+      "grad_norm": 1.3583127081670356,
+      "learning_rate": 5.248821193356356e-06,
+      "loss": 0.0789,
+      "step": 1414
+    },
+    {
+      "epoch": 4.99558693733451,
+      "grad_norm": 1.2297540102464188,
+      "learning_rate": 5.243105730741312e-06,
+      "loss": 0.0782,
+      "step": 1415
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2830,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 1415,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 110584559550464.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}